diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5467 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.983922829581994,
+  "eval_steps": 500,
+  "global_step": 775,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.006430868167202572,
+      "grad_norm": 7.0162146245461425,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 1.0664,
+      "step": 1
+    },
+    {
+      "epoch": 0.012861736334405145,
+      "grad_norm": 7.042137063997965,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.0667,
+      "step": 2
+    },
+    {
+      "epoch": 0.01929260450160772,
+      "grad_norm": 6.952690913098092,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 1.0695,
+      "step": 3
+    },
+    {
+      "epoch": 0.02572347266881029,
+      "grad_norm": 6.7538270332311425,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.0442,
+      "step": 4
+    },
+    {
+      "epoch": 0.03215434083601286,
+      "grad_norm": 6.376454462253263,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 1.051,
+      "step": 5
+    },
+    {
+      "epoch": 0.03858520900321544,
+      "grad_norm": 4.956917374969652,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.0002,
+      "step": 6
+    },
+    {
+      "epoch": 0.04501607717041801,
+      "grad_norm": 3.4841396218367806,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 0.9527,
+      "step": 7
+    },
+    {
+      "epoch": 0.05144694533762058,
+      "grad_norm": 3.0118863884782923,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 0.9526,
+      "step": 8
+    },
+    {
+      "epoch": 0.05787781350482315,
+      "grad_norm": 2.713290461312111,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.9463,
+      "step": 9
+    },
+    {
+      "epoch": 0.06430868167202572,
+      "grad_norm": 4.377833431621951,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 0.9295,
+      "step": 10
+    },
+    {
+      "epoch": 0.0707395498392283,
+      "grad_norm": 4.501301458537532,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 0.9365,
+      "step": 11
+    },
+    {
+      "epoch": 0.07717041800643087,
+      "grad_norm": 4.5293375751127805,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.9149,
+      "step": 12
+    },
+    {
+      "epoch": 0.08360128617363344,
+      "grad_norm": 3.4522004166791684,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.853,
+      "step": 13
+    },
+    {
+      "epoch": 0.09003215434083602,
+      "grad_norm": 3.290798888126205,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.8424,
+      "step": 14
+    },
+    {
+      "epoch": 0.09646302250803858,
+      "grad_norm": 2.6130066798433687,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.8309,
+      "step": 15
+    },
+    {
+      "epoch": 0.10289389067524116,
+      "grad_norm": 1.9609557414837218,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.8064,
+      "step": 16
+    },
+    {
+      "epoch": 0.10932475884244373,
+      "grad_norm": 1.7416084098064664,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 0.775,
+      "step": 17
+    },
+    {
+      "epoch": 0.1157556270096463,
+      "grad_norm": 1.9234596660609349,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 0.7639,
+      "step": 18
+    },
+    {
+      "epoch": 0.12218649517684887,
+      "grad_norm": 1.842944347391724,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 0.7516,
+      "step": 19
+    },
+    {
+      "epoch": 0.12861736334405144,
+      "grad_norm": 1.385542784297474,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 0.7362,
+      "step": 20
+    },
+    {
+      "epoch": 0.13504823151125403,
+      "grad_norm": 1.3434715178283498,
+      "learning_rate": 1.076923076923077e-05,
+      "loss": 0.7299,
+      "step": 21
+    },
+    {
+      "epoch": 0.1414790996784566,
+      "grad_norm": 1.2642455423940069,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 0.7297,
+      "step": 22
+    },
+    {
+      "epoch": 0.14790996784565916,
+      "grad_norm": 1.3810683927585838,
+      "learning_rate": 1.1794871794871796e-05,
+      "loss": 0.7151,
+      "step": 23
+    },
+    {
+      "epoch": 0.15434083601286175,
+      "grad_norm": 1.1763012918310838,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 0.7028,
+      "step": 24
+    },
+    {
+      "epoch": 0.1607717041800643,
+      "grad_norm": 0.8874899050987758,
+      "learning_rate": 1.2820512820512823e-05,
+      "loss": 0.7069,
+      "step": 25
+    },
+    {
+      "epoch": 0.16720257234726688,
+      "grad_norm": 1.1430717427922794,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.6947,
+      "step": 26
+    },
+    {
+      "epoch": 0.17363344051446947,
+      "grad_norm": 1.0152472518153854,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 0.6988,
+      "step": 27
+    },
+    {
+      "epoch": 0.18006430868167203,
+      "grad_norm": 0.8608429155517024,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 0.6953,
+      "step": 28
+    },
+    {
+      "epoch": 0.1864951768488746,
+      "grad_norm": 1.0509945599068666,
+      "learning_rate": 1.4871794871794874e-05,
+      "loss": 0.6853,
+      "step": 29
+    },
+    {
+      "epoch": 0.19292604501607716,
+      "grad_norm": 0.8044463029072009,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 0.6826,
+      "step": 30
+    },
+    {
+      "epoch": 0.19935691318327975,
+      "grad_norm": 1.2310559296987191,
+      "learning_rate": 1.5897435897435897e-05,
+      "loss": 0.6692,
+      "step": 31
+    },
+    {
+      "epoch": 0.2057877813504823,
+      "grad_norm": 0.8236025480165208,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 0.6762,
+      "step": 32
+    },
+    {
+      "epoch": 0.21221864951768488,
+      "grad_norm": 1.0067292632938298,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 0.6586,
+      "step": 33
+    },
+    {
+      "epoch": 0.21864951768488747,
+      "grad_norm": 0.9173211085785304,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 0.6572,
+      "step": 34
+    },
+    {
+      "epoch": 0.22508038585209003,
+      "grad_norm": 0.8516866412021882,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 0.6528,
+      "step": 35
+    },
+    {
+      "epoch": 0.2315112540192926,
+      "grad_norm": 0.7427049230572615,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 0.6412,
+      "step": 36
+    },
+    {
+      "epoch": 0.2379421221864952,
+      "grad_norm": 0.7669874118997514,
+      "learning_rate": 1.8974358974358975e-05,
+      "loss": 0.6677,
+      "step": 37
+    },
+    {
+      "epoch": 0.24437299035369775,
+      "grad_norm": 0.8036867534704667,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 0.6502,
+      "step": 38
+    },
+    {
+      "epoch": 0.2508038585209003,
+      "grad_norm": 0.8316635283700579,
+      "learning_rate": 2e-05,
+      "loss": 0.6567,
+      "step": 39
+    },
+    {
+      "epoch": 0.2572347266881029,
+      "grad_norm": 0.6486238726177489,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.6623,
+      "step": 40
+    },
+    {
+      "epoch": 0.26366559485530544,
+      "grad_norm": 0.7643455134890489,
+      "learning_rate": 2.102564102564103e-05,
+      "loss": 0.6574,
+      "step": 41
+    },
+    {
+      "epoch": 0.27009646302250806,
+      "grad_norm": 0.7635295839879329,
+      "learning_rate": 2.153846153846154e-05,
+      "loss": 0.6479,
+      "step": 42
+    },
+    {
+      "epoch": 0.2765273311897106,
+      "grad_norm": 0.9085553824119198,
+      "learning_rate": 2.2051282051282056e-05,
+      "loss": 0.6367,
+      "step": 43
+    },
+    {
+      "epoch": 0.2829581993569132,
+      "grad_norm": 0.7451374015372272,
+      "learning_rate": 2.2564102564102566e-05,
+      "loss": 0.6349,
+      "step": 44
+    },
+    {
+      "epoch": 0.28938906752411575,
+      "grad_norm": 0.7243157956772703,
+      "learning_rate": 2.3076923076923076e-05,
+      "loss": 0.6389,
+      "step": 45
+    },
+    {
+      "epoch": 0.2958199356913183,
+      "grad_norm": 0.6619765588408462,
+      "learning_rate": 2.3589743589743593e-05,
+      "loss": 0.6297,
+      "step": 46
+    },
+    {
+      "epoch": 0.3022508038585209,
+      "grad_norm": 0.6301343131970055,
+      "learning_rate": 2.4102564102564103e-05,
+      "loss": 0.6293,
+      "step": 47
+    },
+    {
+      "epoch": 0.3086816720257235,
+      "grad_norm": 0.56673780583801,
+      "learning_rate": 2.461538461538462e-05,
+      "loss": 0.6243,
+      "step": 48
+    },
+    {
+      "epoch": 0.31511254019292606,
+      "grad_norm": 0.5906702938651524,
+      "learning_rate": 2.512820512820513e-05,
+      "loss": 0.6402,
+      "step": 49
+    },
+    {
+      "epoch": 0.3215434083601286,
+      "grad_norm": 0.617600273963167,
+      "learning_rate": 2.5641025641025646e-05,
+      "loss": 0.6205,
+      "step": 50
+    },
+    {
+      "epoch": 0.3279742765273312,
+      "grad_norm": 0.6024732168010081,
+      "learning_rate": 2.6153846153846157e-05,
+      "loss": 0.6264,
+      "step": 51
+    },
+    {
+      "epoch": 0.33440514469453375,
+      "grad_norm": 0.7303388495485045,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.6151,
+      "step": 52
+    },
+    {
+      "epoch": 0.3408360128617363,
+      "grad_norm": 0.8010109628064097,
+      "learning_rate": 2.7179487179487183e-05,
+      "loss": 0.6295,
+      "step": 53
+    },
+    {
+      "epoch": 0.34726688102893893,
+      "grad_norm": 0.8545639494695195,
+      "learning_rate": 2.7692307692307694e-05,
+      "loss": 0.6304,
+      "step": 54
+    },
+    {
+      "epoch": 0.3536977491961415,
+      "grad_norm": 1.0203340239752252,
+      "learning_rate": 2.820512820512821e-05,
+      "loss": 0.6312,
+      "step": 55
+    },
+    {
+      "epoch": 0.36012861736334406,
+      "grad_norm": 0.9162410459030489,
+      "learning_rate": 2.871794871794872e-05,
+      "loss": 0.6206,
+      "step": 56
+    },
+    {
+      "epoch": 0.3665594855305466,
+      "grad_norm": 0.7963404691986334,
+      "learning_rate": 2.923076923076923e-05,
+      "loss": 0.6147,
+      "step": 57
+    },
+    {
+      "epoch": 0.3729903536977492,
+      "grad_norm": 0.8047583466593461,
+      "learning_rate": 2.9743589743589747e-05,
+      "loss": 0.6162,
+      "step": 58
+    },
+    {
+      "epoch": 0.37942122186495175,
+      "grad_norm": 1.1795766167401995,
+      "learning_rate": 3.0256410256410257e-05,
+      "loss": 0.6209,
+      "step": 59
+    },
+    {
+      "epoch": 0.3858520900321543,
+      "grad_norm": 1.0864121325683949,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 0.6212,
+      "step": 60
+    },
+    {
+      "epoch": 0.39228295819935693,
+      "grad_norm": 0.9521715572952941,
+      "learning_rate": 3.128205128205129e-05,
+      "loss": 0.6109,
+      "step": 61
+    },
+    {
+      "epoch": 0.3987138263665595,
+      "grad_norm": 1.1765099457452546,
+      "learning_rate": 3.1794871794871795e-05,
+      "loss": 0.6151,
+      "step": 62
+    },
+    {
+      "epoch": 0.40514469453376206,
+      "grad_norm": 0.9084885261943422,
+      "learning_rate": 3.230769230769231e-05,
+      "loss": 0.6093,
+      "step": 63
+    },
+    {
+      "epoch": 0.4115755627009646,
+      "grad_norm": 0.8547080139160044,
+      "learning_rate": 3.282051282051282e-05,
+      "loss": 0.6104,
+      "step": 64
+    },
+    {
+      "epoch": 0.4180064308681672,
+      "grad_norm": 1.0357274479216767,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.5979,
+      "step": 65
+    },
+    {
+      "epoch": 0.42443729903536975,
+      "grad_norm": 0.9819557484788577,
+      "learning_rate": 3.384615384615385e-05,
+      "loss": 0.6219,
+      "step": 66
+    },
+    {
+      "epoch": 0.43086816720257237,
+      "grad_norm": 0.9344724035411234,
+      "learning_rate": 3.435897435897436e-05,
+      "loss": 0.5993,
+      "step": 67
+    },
+    {
+      "epoch": 0.43729903536977494,
+      "grad_norm": 1.187832408129545,
+      "learning_rate": 3.4871794871794875e-05,
+      "loss": 0.6114,
+      "step": 68
+    },
+    {
+      "epoch": 0.4437299035369775,
+      "grad_norm": 1.0951664420643048,
+      "learning_rate": 3.538461538461539e-05,
+      "loss": 0.6179,
+      "step": 69
+    },
+    {
+      "epoch": 0.45016077170418006,
+      "grad_norm": 0.9808183249663346,
+      "learning_rate": 3.58974358974359e-05,
+      "loss": 0.6128,
+      "step": 70
+    },
+    {
+      "epoch": 0.4565916398713826,
+      "grad_norm": 1.2238304227998327,
+      "learning_rate": 3.6410256410256416e-05,
+      "loss": 0.6271,
+      "step": 71
+    },
+    {
+      "epoch": 0.4630225080385852,
+      "grad_norm": 0.8193662166037599,
+      "learning_rate": 3.692307692307693e-05,
+      "loss": 0.5979,
+      "step": 72
+    },
+    {
+      "epoch": 0.4694533762057878,
+      "grad_norm": 1.4077316753429578,
+      "learning_rate": 3.7435897435897436e-05,
+      "loss": 0.6097,
+      "step": 73
+    },
+    {
+      "epoch": 0.4758842443729904,
+      "grad_norm": 0.8471545008291038,
+      "learning_rate": 3.794871794871795e-05,
+      "loss": 0.6085,
+      "step": 74
+    },
+    {
+      "epoch": 0.48231511254019294,
+      "grad_norm": 1.4281620515346747,
+      "learning_rate": 3.846153846153846e-05,
+      "loss": 0.614,
+      "step": 75
+    },
+    {
+      "epoch": 0.4887459807073955,
+      "grad_norm": 0.9249111277715849,
+      "learning_rate": 3.8974358974358976e-05,
+      "loss": 0.6112,
+      "step": 76
+    },
+    {
+      "epoch": 0.49517684887459806,
+      "grad_norm": 1.2848864370241548,
+      "learning_rate": 3.948717948717949e-05,
+      "loss": 0.6087,
+      "step": 77
+    },
+    {
+      "epoch": 0.5016077170418006,
+      "grad_norm": 1.1679911595246018,
+      "learning_rate": 4e-05,
+      "loss": 0.6294,
+      "step": 78
+    },
+    {
+      "epoch": 0.5080385852090032,
+      "grad_norm": 0.9036375040978208,
+      "learning_rate": 3.999979684222212e-05,
+      "loss": 0.5991,
+      "step": 79
+    },
+    {
+      "epoch": 0.5144694533762058,
+      "grad_norm": 1.1455097611632525,
+      "learning_rate": 3.999918737301574e-05,
+      "loss": 0.6033,
+      "step": 80
+    },
+    {
+      "epoch": 0.5209003215434084,
+      "grad_norm": 1.0471182646218358,
+      "learning_rate": 3.9998171604762734e-05,
+      "loss": 0.6081,
+      "step": 81
+    },
+    {
+      "epoch": 0.5273311897106109,
+      "grad_norm": 0.9771344197266095,
+      "learning_rate": 3.999674955809922e-05,
+      "loss": 0.6128,
+      "step": 82
+    },
+    {
+      "epoch": 0.5337620578778135,
+      "grad_norm": 1.5328128828675538,
+      "learning_rate": 3.9994921261915174e-05,
+      "loss": 0.6239,
+      "step": 83
+    },
+    {
+      "epoch": 0.5401929260450161,
+      "grad_norm": 0.8557687933935886,
+      "learning_rate": 3.999268675335385e-05,
+      "loss": 0.6027,
+      "step": 84
+    },
+    {
+      "epoch": 0.5466237942122186,
+      "grad_norm": 1.7264034234419094,
+      "learning_rate": 3.999004607781104e-05,
+      "loss": 0.6178,
+      "step": 85
+    },
+    {
+      "epoch": 0.5530546623794212,
+      "grad_norm": 1.1232399205417798,
+      "learning_rate": 3.998699928893412e-05,
+      "loss": 0.6159,
+      "step": 86
+    },
+    {
+      "epoch": 0.5594855305466238,
+      "grad_norm": 1.6425765521799327,
+      "learning_rate": 3.998354644862097e-05,
+      "loss": 0.6224,
+      "step": 87
+    },
+    {
+      "epoch": 0.5659163987138264,
+      "grad_norm": 1.283945305348683,
+      "learning_rate": 3.997968762701873e-05,
+      "loss": 0.6168,
+      "step": 88
+    },
+    {
+      "epoch": 0.572347266881029,
+      "grad_norm": 1.2029505880341773,
+      "learning_rate": 3.997542290252236e-05,
+      "loss": 0.5944,
+      "step": 89
+    },
+    {
+      "epoch": 0.5787781350482315,
+      "grad_norm": 1.1813171489073961,
+      "learning_rate": 3.997075236177305e-05,
+      "loss": 0.6196,
+      "step": 90
+    },
+    {
+      "epoch": 0.5852090032154341,
+      "grad_norm": 1.0761647565610426,
+      "learning_rate": 3.9965676099656486e-05,
+      "loss": 0.6113,
+      "step": 91
+    },
+    {
+      "epoch": 0.5916398713826366,
+      "grad_norm": 1.048195128041139,
+      "learning_rate": 3.9960194219300856e-05,
+      "loss": 0.6005,
+      "step": 92
+    },
+    {
+      "epoch": 0.5980707395498392,
+      "grad_norm": 0.8282012654238775,
+      "learning_rate": 3.995430683207485e-05,
+      "loss": 0.5957,
+      "step": 93
+    },
+    {
+      "epoch": 0.6045016077170418,
+      "grad_norm": 0.9916485945453747,
+      "learning_rate": 3.9948014057585294e-05,
+      "loss": 0.6016,
+      "step": 94
+    },
+    {
+      "epoch": 0.6109324758842444,
+      "grad_norm": 0.8888456887367502,
+      "learning_rate": 3.994131602367481e-05,
+      "loss": 0.591,
+      "step": 95
+    },
+    {
+      "epoch": 0.617363344051447,
+      "grad_norm": 0.8995120605242533,
+      "learning_rate": 3.993421286641917e-05,
+      "loss": 0.5922,
+      "step": 96
+    },
+    {
+      "epoch": 0.6237942122186495,
+      "grad_norm": 0.829725405785306,
+      "learning_rate": 3.992670473012453e-05,
+      "loss": 0.611,
+      "step": 97
+    },
+    {
+      "epoch": 0.6302250803858521,
+      "grad_norm": 0.856820505259012,
+      "learning_rate": 3.9918791767324514e-05,
+      "loss": 0.6017,
+      "step": 98
+    },
+    {
+      "epoch": 0.6366559485530546,
+      "grad_norm": 0.69477434703996,
+      "learning_rate": 3.991047413877713e-05,
+      "loss": 0.5993,
+      "step": 99
+    },
+    {
+      "epoch": 0.6430868167202572,
+      "grad_norm": 0.8075695315569323,
+      "learning_rate": 3.990175201346146e-05,
+      "loss": 0.6007,
+      "step": 100
+    },
+    {
+      "epoch": 0.6495176848874598,
+      "grad_norm": 0.5822906086515444,
+      "learning_rate": 3.9892625568574274e-05,
+      "loss": 0.5923,
+      "step": 101
+    },
+    {
+      "epoch": 0.6559485530546624,
+      "grad_norm": 0.7068787530306003,
+      "learning_rate": 3.988309498952639e-05,
+      "loss": 0.5862,
+      "step": 102
+    },
+    {
+      "epoch": 0.662379421221865,
+      "grad_norm": 0.5952919980585671,
+      "learning_rate": 3.987316046993893e-05,
+      "loss": 0.5956,
+      "step": 103
+    },
+    {
+      "epoch": 0.6688102893890675,
+      "grad_norm": 0.6006957005647934,
+      "learning_rate": 3.98628222116394e-05,
+      "loss": 0.599,
+      "step": 104
+    },
+    {
+      "epoch": 0.6752411575562701,
+      "grad_norm": 0.824401600699459,
+      "learning_rate": 3.9852080424657545e-05,
+      "loss": 0.6034,
+      "step": 105
+    },
+    {
+      "epoch": 0.6816720257234726,
+      "grad_norm": 0.7284135925904331,
+      "learning_rate": 3.9840935327221136e-05,
+      "loss": 0.5941,
+      "step": 106
+    },
+    {
+      "epoch": 0.6881028938906752,
+      "grad_norm": 0.8065751381449505,
+      "learning_rate": 3.982938714575148e-05,
+      "loss": 0.5795,
+      "step": 107
+    },
+    {
+      "epoch": 0.6945337620578779,
+      "grad_norm": 0.7732795226244582,
+      "learning_rate": 3.981743611485888e-05,
+      "loss": 0.6033,
+      "step": 108
+    },
+    {
+      "epoch": 0.7009646302250804,
+      "grad_norm": 0.6959441052980924,
+      "learning_rate": 3.9805082477337815e-05,
+      "loss": 0.6061,
+      "step": 109
+    },
+    {
+      "epoch": 0.707395498392283,
+      "grad_norm": 0.648097019437012,
+      "learning_rate": 3.979232648416204e-05,
+      "loss": 0.5928,
+      "step": 110
+    },
+    {
+      "epoch": 0.7138263665594855,
+      "grad_norm": 0.6178376407880537,
+      "learning_rate": 3.9779168394479486e-05,
+      "loss": 0.6003,
+      "step": 111
+    },
+    {
+      "epoch": 0.7202572347266881,
+      "grad_norm": 0.6070407072437944,
+      "learning_rate": 3.976560847560697e-05,
+      "loss": 0.5986,
+      "step": 112
+    },
+    {
+      "epoch": 0.7266881028938906,
+      "grad_norm": 0.637403337085659,
+      "learning_rate": 3.97516470030248e-05,
+      "loss": 0.6021,
+      "step": 113
+    },
+    {
+      "epoch": 0.7331189710610932,
+      "grad_norm": 0.7480752175838351,
+      "learning_rate": 3.9737284260371144e-05,
+      "loss": 0.5953,
+      "step": 114
+    },
+    {
+      "epoch": 0.7395498392282959,
+      "grad_norm": 0.5965472370954971,
+      "learning_rate": 3.9722520539436286e-05,
+      "loss": 0.5966,
+      "step": 115
+    },
+    {
+      "epoch": 0.7459807073954984,
+      "grad_norm": 0.5824333806675311,
+      "learning_rate": 3.970735614015671e-05,
+      "loss": 0.5826,
+      "step": 116
+    },
+    {
+      "epoch": 0.752411575562701,
+      "grad_norm": 0.7790481671687746,
+      "learning_rate": 3.969179137060897e-05,
+      "loss": 0.5796,
+      "step": 117
+    },
+    {
+      "epoch": 0.7588424437299035,
+      "grad_norm": 0.5800091327175808,
+      "learning_rate": 3.967582654700348e-05,
+      "loss": 0.595,
+      "step": 118
+    },
+    {
+      "epoch": 0.7652733118971061,
+      "grad_norm": 0.5615113582167717,
+      "learning_rate": 3.965946199367804e-05,
+      "loss": 0.5922,
+      "step": 119
+    },
+    {
+      "epoch": 0.7717041800643086,
+      "grad_norm": 0.6526372250018604,
+      "learning_rate": 3.964269804309128e-05,
+      "loss": 0.5982,
+      "step": 120
+    },
+    {
+      "epoch": 0.7781350482315113,
+      "grad_norm": 0.4802696011361092,
+      "learning_rate": 3.962553503581589e-05,
+      "loss": 0.5984,
+      "step": 121
+    },
+    {
+      "epoch": 0.7845659163987139,
+      "grad_norm": 0.61327036480935,
+      "learning_rate": 3.9607973320531725e-05,
+      "loss": 0.5967,
+      "step": 122
+    },
+    {
+      "epoch": 0.7909967845659164,
+      "grad_norm": 0.5081736432497024,
+      "learning_rate": 3.959001325401868e-05,
+      "loss": 0.5941,
+      "step": 123
+    },
+    {
+      "epoch": 0.797427652733119,
+      "grad_norm": 0.5914030914116616,
+      "learning_rate": 3.957165520114948e-05,
+      "loss": 0.584,
+      "step": 124
+    },
+    {
+      "epoch": 0.8038585209003215,
+      "grad_norm": 0.7355488768513988,
+      "learning_rate": 3.955289953488224e-05,
+      "loss": 0.5793,
+      "step": 125
+    },
+    {
+      "epoch": 0.8102893890675241,
+      "grad_norm": 0.5649350788584953,
+      "learning_rate": 3.9533746636252915e-05,
+      "loss": 0.5772,
+      "step": 126
+    },
+    {
+      "epoch": 0.8167202572347267,
+      "grad_norm": 0.5980875860422536,
+      "learning_rate": 3.951419689436754e-05,
+      "loss": 0.5929,
+      "step": 127
+    },
+    {
+      "epoch": 0.8231511254019293,
+      "grad_norm": 0.7802462905110522,
+      "learning_rate": 3.949425070639432e-05,
+      "loss": 0.581,
+      "step": 128
+    },
+    {
+      "epoch": 0.8295819935691319,
+      "grad_norm": 0.7177307217999693,
+      "learning_rate": 3.947390847755559e-05,
+      "loss": 0.5992,
+      "step": 129
+    },
+    {
+      "epoch": 0.8360128617363344,
+      "grad_norm": 0.7520066479156595,
+      "learning_rate": 3.945317062111953e-05,
+      "loss": 0.5832,
+      "step": 130
+    },
+    {
+      "epoch": 0.842443729903537,
+      "grad_norm": 0.8529868006432115,
+      "learning_rate": 3.943203755839184e-05,
+      "loss": 0.5949,
+      "step": 131
+    },
+    {
+      "epoch": 0.8488745980707395,
+      "grad_norm": 0.687104044853933,
+      "learning_rate": 3.941050971870712e-05,
+      "loss": 0.5812,
+      "step": 132
+    },
+    {
+      "epoch": 0.8553054662379421,
+      "grad_norm": 0.7059912707956133,
+      "learning_rate": 3.938858753942018e-05,
+      "loss": 0.5821,
+      "step": 133
+    },
+    {
+      "epoch": 0.8617363344051447,
+      "grad_norm": 0.5652564763976348,
+      "learning_rate": 3.936627146589715e-05,
+      "loss": 0.588,
+      "step": 134
+    },
+    {
+      "epoch": 0.8681672025723473,
+      "grad_norm": 0.7350998795480612,
+      "learning_rate": 3.93435619515064e-05,
+      "loss": 0.5887,
+      "step": 135
+    },
+    {
+      "epoch": 0.8745980707395499,
+      "grad_norm": 0.7795178233691,
+      "learning_rate": 3.932045945760939e-05,
+      "loss": 0.5995,
+      "step": 136
+    },
+    {
+      "epoch": 0.8810289389067524,
+      "grad_norm": 0.6484496137663821,
+      "learning_rate": 3.929696445355126e-05,
+      "loss": 0.5758,
+      "step": 137
+    },
+    {
+      "epoch": 0.887459807073955,
+      "grad_norm": 0.8796073072875976,
+      "learning_rate": 3.9273077416651284e-05,
+      "loss": 0.5921,
+      "step": 138
+    },
+    {
+      "epoch": 0.8938906752411575,
+      "grad_norm": 0.8362168482003571,
+      "learning_rate": 3.92487988321932e-05,
+      "loss": 0.582,
+      "step": 139
+    },
+    {
+      "epoch": 0.9003215434083601,
+      "grad_norm": 0.6496668989131343,
+      "learning_rate": 3.922412919341533e-05,
+      "loss": 0.5885,
+      "step": 140
+    },
+    {
+      "epoch": 0.9067524115755627,
+      "grad_norm": 0.8822524561323366,
+      "learning_rate": 3.919906900150056e-05,
+      "loss": 0.5894,
+      "step": 141
+    },
+    {
+      "epoch": 0.9131832797427653,
+      "grad_norm": 0.5995833469337996,
+      "learning_rate": 3.917361876556621e-05,
+      "loss": 0.5776,
+      "step": 142
+    },
+    {
+      "epoch": 0.9196141479099679,
+      "grad_norm": 0.7665962043324072,
+      "learning_rate": 3.9147779002653606e-05,
+      "loss": 0.5827,
+      "step": 143
+    },
+    {
+      "epoch": 0.9260450160771704,
+      "grad_norm": 0.5851137089743218,
+      "learning_rate": 3.912155023771762e-05,
+      "loss": 0.5824,
+      "step": 144
+    },
+    {
+      "epoch": 0.932475884244373,
+      "grad_norm": 0.6254248895857109,
+      "learning_rate": 3.909493300361602e-05,
+      "loss": 0.5926,
+      "step": 145
+    },
+    {
+      "epoch": 0.9389067524115756,
+      "grad_norm": 0.6460601057624329,
+      "learning_rate": 3.9067927841098614e-05,
+      "loss": 0.5855,
+      "step": 146
+    },
+    {
+      "epoch": 0.9453376205787781,
+      "grad_norm": 0.6772434483573326,
+      "learning_rate": 3.904053529879629e-05,
+      "loss": 0.5752,
+      "step": 147
+    },
+    {
+      "epoch": 0.9517684887459807,
+      "grad_norm": 0.5477103785756737,
+      "learning_rate": 3.9012755933209846e-05,
+      "loss": 0.5895,
+      "step": 148
+    },
+    {
+      "epoch": 0.9581993569131833,
+      "grad_norm": 0.6151671038330779,
+      "learning_rate": 3.89845903086987e-05,
+      "loss": 0.586,
+      "step": 149
+    },
+    {
+      "epoch": 0.9646302250803859,
+      "grad_norm": 0.5940489395434554,
+      "learning_rate": 3.8956038997469436e-05,
+      "loss": 0.5814,
+      "step": 150
+    },
+    {
+      "epoch": 0.9710610932475884,
+      "grad_norm": 0.6777541959298194,
+      "learning_rate": 3.892710257956412e-05,
+      "loss": 0.5758,
+      "step": 151
+    },
+    {
+      "epoch": 0.977491961414791,
+      "grad_norm": 0.5474009034501883,
+      "learning_rate": 3.889778164284861e-05,
+      "loss": 0.5791,
+      "step": 152
+    },
+    {
+      "epoch": 0.9839228295819936,
+      "grad_norm": 0.7227382690094412,
+      "learning_rate": 3.886807678300053e-05,
+      "loss": 0.5916,
+      "step": 153
+    },
+    {
+      "epoch": 0.9903536977491961,
+      "grad_norm": 0.5613251512532227,
+      "learning_rate": 3.883798860349722e-05,
+      "loss": 0.5764,
+      "step": 154
+    },
+    {
+      "epoch": 0.9967845659163987,
+      "grad_norm": 0.746629668361697,
+      "learning_rate": 3.880751771560344e-05,
+      "loss": 0.5722,
+      "step": 155
+    },
+    {
+      "epoch": 1.0032154340836013,
+      "grad_norm": 0.5558760763654468,
+      "learning_rate": 3.877666473835899e-05,
+      "loss": 0.5427,
+      "step": 156
+    },
+    {
+      "epoch": 1.0096463022508038,
+      "grad_norm": 0.6546883920523154,
+      "learning_rate": 3.874543029856609e-05,
+      "loss": 0.5225,
+      "step": 157
+    },
+    {
+      "epoch": 1.0160771704180065,
+      "grad_norm": 0.7295882424551627,
+      "learning_rate": 3.871381503077668e-05,
+      "loss": 0.5165,
+      "step": 158
+    },
+    {
+      "epoch": 1.022508038585209,
+      "grad_norm": 0.5818356209448946,
+      "learning_rate": 3.8681819577279515e-05,
+      "loss": 0.498,
+      "step": 159
+    },
+    {
+      "epoch": 1.0289389067524115,
+      "grad_norm": 0.5163467947849115,
+      "learning_rate": 3.864944458808712e-05,
+      "loss": 0.5176,
+      "step": 160
+    },
+    {
+      "epoch": 1.0353697749196142,
+      "grad_norm": 0.6552225476813373,
+      "learning_rate": 3.861669072092258e-05,
+      "loss": 0.5121,
+      "step": 161
+    },
+    {
+      "epoch": 1.0418006430868167,
+      "grad_norm": 0.5568631360983052,
+      "learning_rate": 3.8583558641206183e-05,
+      "loss": 0.5092,
+      "step": 162
+    },
+    {
+      "epoch": 1.0482315112540193,
+      "grad_norm": 0.5841979058113783,
+      "learning_rate": 3.85500490220419e-05,
+      "loss": 0.5106,
+      "step": 163
+    },
+    {
+      "epoch": 1.0546623794212218,
+      "grad_norm": 0.5286925244184147,
+      "learning_rate": 3.85161625442037e-05,
+      "loss": 0.5009,
+      "step": 164
+    },
+    {
+      "epoch": 1.0610932475884245,
+      "grad_norm": 0.6140452699928527,
+      "learning_rate": 3.8481899896121744e-05,
+      "loss": 0.5056,
+      "step": 165
+    },
+    {
+      "epoch": 1.067524115755627,
+      "grad_norm": 0.6789030109707668,
+      "learning_rate": 3.8447261773868376e-05,
+      "loss": 0.5035,
+      "step": 166
+    },
+    {
+      "epoch": 1.0739549839228295,
+      "grad_norm": 0.4949970402702986,
+      "learning_rate": 3.8412248881144e-05,
+      "loss": 0.5102,
+      "step": 167
+    },
+    {
+      "epoch": 1.0803858520900322,
+      "grad_norm": 0.6787952900352953,
+      "learning_rate": 3.837686192926274e-05,
+      "loss": 0.5114,
+      "step": 168
+    },
+    {
+      "epoch": 1.0868167202572347,
+      "grad_norm": 0.6224501362637438,
+      "learning_rate": 3.834110163713806e-05,
+      "loss": 0.5141,
+      "step": 169
+    },
+    {
+      "epoch": 1.0932475884244373,
+      "grad_norm": 0.6575755149617953,
+      "learning_rate": 3.830496873126812e-05,
+      "loss": 0.5043,
+      "step": 170
+    },
+    {
+      "epoch": 1.09967845659164,
+      "grad_norm": 0.5366056783535434,
+      "learning_rate": 3.826846394572099e-05,
+      "loss": 0.5126,
+      "step": 171
+    },
+    {
+      "epoch": 1.1061093247588425,
+      "grad_norm": 0.6160754843937626,
+      "learning_rate": 3.823158802211979e-05,
+      "loss": 0.5042,
+      "step": 172
+    },
+    {
+      "epoch": 1.112540192926045,
+      "grad_norm": 0.603243808589432,
+      "learning_rate": 3.8194341709627595e-05,
+      "loss": 0.5199,
+      "step": 173
+    },
+    {
+      "epoch": 1.1189710610932475,
+      "grad_norm": 0.7456444308087263,
+      "learning_rate": 3.81567257649322e-05,
+      "loss": 0.5221,
+      "step": 174
+    },
+    {
+      "epoch": 1.1254019292604502,
+      "grad_norm": 0.570014715918431,
+      "learning_rate": 3.811874095223078e-05,
+      "loss": 0.5202,
+      "step": 175
+    },
+    {
+      "epoch": 1.1318327974276527,
+      "grad_norm": 0.7494743834398412,
+      "learning_rate": 3.808038804321436e-05,
+      "loss": 0.5207,
+      "step": 176
+    },
+    {
+      "epoch": 1.1382636655948553,
+      "grad_norm": 0.6385871647582012,
+      "learning_rate": 3.804166781705211e-05,
+      "loss": 0.5037,
+      "step": 177
+    },
+    {
+      "epoch": 1.144694533762058,
+      "grad_norm": 0.5964207675799855,
+      "learning_rate": 3.8002581060375535e-05,
+      "loss": 0.5036,
+      "step": 178
+    },
+    {
+      "epoch": 1.1511254019292605,
+      "grad_norm": 0.6978587450402056,
+      "learning_rate": 3.796312856726252e-05,
+      "loss": 0.5163,
+      "step": 179
+    },
+    {
+      "epoch": 1.157556270096463,
+      "grad_norm": 0.5608846001320263,
+      "learning_rate": 3.7923311139221114e-05,
+      "loss": 0.5172,
+      "step": 180
+    },
+    {
+      "epoch": 1.1639871382636655,
+      "grad_norm": 0.6413392060193942,
+      "learning_rate": 3.7883129585173365e-05,
+      "loss": 0.5053,
+      "step": 181
+    },
+    {
+      "epoch": 1.1704180064308682,
+      "grad_norm": 0.5629662380693857,
+      "learning_rate": 3.784258472143878e-05,
+      "loss": 0.5112,
+      "step": 182
+    },
+    {
+      "epoch": 1.1768488745980707,
+      "grad_norm": 0.5571340832411922,
+      "learning_rate": 3.7801677371717825e-05,
+      "loss": 0.5292,
+      "step": 183
+    },
+    {
+      "epoch": 1.1832797427652733,
+      "grad_norm": 0.5902956511857084,
+      "learning_rate": 3.77604083670751e-05,
+      "loss": 0.5095,
+      "step": 184
+    },
+    {
+      "epoch": 1.189710610932476,
+      "grad_norm": 0.5391565218719164,
+      "learning_rate": 3.771877854592254e-05,
+      "loss": 0.5263,
+      "step": 185
+    },
+    {
+      "epoch": 1.1961414790996785,
+      "grad_norm": 0.5200409181500246,
+      "learning_rate": 3.767678875400235e-05,
+      "loss": 0.5109,
+      "step": 186
+    },
+    {
+      "epoch": 1.202572347266881,
+      "grad_norm": 0.5889395130209956,
+      "learning_rate": 3.76344398443698e-05,
+      "loss": 0.522,
+      "step": 187
+    },
+    {
+      "epoch": 1.2090032154340835,
+      "grad_norm": 0.6206592730523811,
+      "learning_rate": 3.7591732677375936e-05,
+      "loss": 0.4934,
+      "step": 188
+    },
+    {
+      "epoch": 1.2154340836012862,
+      "grad_norm": 0.4617930098346987,
+      "learning_rate": 3.754866812065008e-05,
+      "loss": 0.5076,
+      "step": 189
+    },
+    {
+      "epoch": 1.2218649517684887,
+      "grad_norm": 0.6234348850888431,
+      "learning_rate": 3.750524704908218e-05,
+      "loss": 0.4922,
+      "step": 190
+    },
+    {
+      "epoch": 1.2282958199356913,
+      "grad_norm": 0.4728516845115116,
+      "learning_rate": 3.746147034480508e-05,
+      "loss": 0.5071,
+      "step": 191
+    },
+    {
+      "epoch": 1.234726688102894,
+      "grad_norm": 0.6176528689395011,
+      "learning_rate": 3.741733889717659e-05,
+      "loss": 0.501,
+      "step": 192
+    },
+    {
+      "epoch": 1.2411575562700965,
+      "grad_norm": 0.4953177762013319,
+      "learning_rate": 3.737285360276138e-05,
+      "loss": 0.5211,
+      "step": 193
+    },
+    {
+      "epoch": 1.247588424437299,
+      "grad_norm": 0.517093105507378,
+      "learning_rate": 3.7328015365312815e-05,
+      "loss": 0.5143,
+      "step": 194
+    },
+    {
+      "epoch": 1.2540192926045015,
+      "grad_norm": 0.5033478462641321,
+      "learning_rate": 3.728282509575456e-05,
+      "loss": 0.5162,
+      "step": 195
+    },
+    {
+      "epoch": 1.2604501607717042,
+      "grad_norm": 0.4993533951001497,
+      "learning_rate": 3.7237283712162083e-05,
+      "loss": 0.5113,
+      "step": 196
+    },
+    {
+      "epoch": 1.2668810289389068,
+      "grad_norm": 0.6037856276302761,
+      "learning_rate": 3.719139213974403e-05,
+      "loss": 0.4998,
+      "step": 197
+    },
+    {
+      "epoch": 1.2733118971061093,
+      "grad_norm": 0.5038804366385807,
+      "learning_rate": 3.7145151310823365e-05,
+      "loss": 0.5148,
+      "step": 198
+    },
+    {
+      "epoch": 1.279742765273312,
+      "grad_norm": 0.5642289105766572,
+      "learning_rate": 3.709856216481852e-05,
+      "loss": 0.5039,
+      "step": 199
+    },
+    {
+      "epoch": 1.2861736334405145,
+      "grad_norm": 0.5624512333298962,
+      "learning_rate": 3.705162564822421e-05,
+      "loss": 0.512,
+      "step": 200
+    },
+    {
+      "epoch": 1.292604501607717,
+      "grad_norm": 0.6032524605551683,
+      "learning_rate": 3.700434271459229e-05,
+      "loss": 0.4976,
+      "step": 201
+    },
+    {
+      "epoch": 1.2990353697749195,
+      "grad_norm": 0.533859788679996,
+      "learning_rate": 3.6956714324512325e-05,
+      "loss": 0.5096,
+      "step": 202
+    },
+    {
+      "epoch": 1.3054662379421222,
+      "grad_norm": 0.4824400621436784,
+      "learning_rate": 3.6908741445592104e-05,
+      "loss": 0.4857,
+      "step": 203
+    },
+    {
+      "epoch": 1.3118971061093248,
+      "grad_norm": 0.5907188079923736,
+      "learning_rate": 3.6860425052437986e-05,
+      "loss": 0.508,
+      "step": 204
+    },
+    {
+      "epoch": 1.3183279742765273,
+      "grad_norm": 0.5704144839129195,
+      "learning_rate": 3.6811766126635064e-05,
+      "loss": 0.5209,
+      "step": 205
+    },
+    {
+      "epoch": 1.32475884244373,
+      "grad_norm": 0.4131859693587685,
+      "learning_rate": 3.6762765656727274e-05,
+      "loss": 0.5183,
+      "step": 206
+    },
+    {
+      "epoch": 1.3311897106109325,
+      "grad_norm": 0.5661980553809456,
+      "learning_rate": 3.6713424638197264e-05,
+      "loss": 0.5046,
+      "step": 207
+    },
+    {
+      "epoch": 1.337620578778135,
+      "grad_norm": 0.6444215959829735,
+      "learning_rate": 3.6663744073446206e-05,
+      "loss": 0.5097,
+      "step": 208
+    },
+    {
+      "epoch": 1.3440514469453375,
+      "grad_norm": 0.5064701270658051,
+      "learning_rate": 3.6613724971773426e-05,
+      "loss": 0.5264,
+      "step": 209
+    },
+    {
+      "epoch": 1.3504823151125402,
+      "grad_norm": 0.48096987955272896,
+      "learning_rate": 3.6563368349355855e-05,
+      "loss": 0.5418,
+      "step": 210
+    },
+    {
+      "epoch": 1.3569131832797428,
+      "grad_norm": 0.5675177783459339,
+      "learning_rate": 3.651267522922746e-05,
+      "loss": 0.4984,
+      "step": 211
+    },
+    {
+      "epoch": 1.3633440514469453,
+      "grad_norm": 0.49460123776464476,
+      "learning_rate": 3.646164664125841e-05,
+      "loss": 0.5195,
+      "step": 212
+    },
+    {
+      "epoch": 1.369774919614148,
+      "grad_norm": 0.5367405812801022,
+      "learning_rate": 3.6410283622134146e-05,
+      "loss": 0.5015,
+      "step": 213
+    },
+    {
+      "epoch": 1.3762057877813505,
+      "grad_norm": 0.6103896049600983,
+      "learning_rate": 3.6358587215334355e-05,
+      "loss": 0.5101,
+      "step": 214
+    },
+    {
+      "epoch": 1.382636655948553,
+      "grad_norm": 0.5322908731661551,
+      "learning_rate": 3.630655847111175e-05,
+      "loss": 0.5184,
+      "step": 215
+    },
+    {
+      "epoch": 1.3890675241157555,
+      "grad_norm": 0.6999762343810539,
+      "learning_rate": 3.6254198446470745e-05,
+      "loss": 0.5134,
+      "step": 216
+    },
+    {
+      "epoch": 1.3954983922829582,
+      "grad_norm": 0.5288833935336679,
+      "learning_rate": 3.6201508205145956e-05,
+      "loss": 0.5217,
+      "step": 217
+    },
+    {
+      "epoch": 1.4019292604501608,
+      "grad_norm": 0.48553809409521503,
+      "learning_rate": 3.614848881758062e-05,
+      "loss": 0.5198,
+      "step": 218
+    },
+    {
+      "epoch": 1.4083601286173635,
+      "grad_norm": 0.5508524699425467,
+      "learning_rate": 3.609514136090483e-05,
+      "loss": 0.4967,
+      "step": 219
+    },
+    {
+      "epoch": 1.414790996784566,
+      "grad_norm": 0.44795778785987206,
+      "learning_rate": 3.604146691891367e-05,
+      "loss": 0.529,
+      "step": 220
+    },
+    {
+      "epoch": 1.4212218649517685,
+      "grad_norm": 0.5653585844558164,
+      "learning_rate": 3.598746658204517e-05,
+      "loss": 0.4997,
+      "step": 221
+    },
+    {
+      "epoch": 1.427652733118971,
+      "grad_norm": 0.49053599842454126,
+      "learning_rate": 3.5933141447358176e-05,
+      "loss": 0.5109,
+      "step": 222
+    },
+    {
+      "epoch": 1.4340836012861735,
+      "grad_norm": 0.4380277791705383,
+      "learning_rate": 3.5878492618510055e-05,
+      "loss": 0.5058,
+      "step": 223
+    },
+    {
+      "epoch": 1.4405144694533762,
+      "grad_norm": 0.5009856686371109,
+      "learning_rate": 3.582352120573427e-05,
+      "loss": 0.5178,
+      "step": 224
+    },
+    {
+      "epoch": 1.4469453376205788,
+      "grad_norm": 0.4446341387675549,
+      "learning_rate": 3.576822832581783e-05,
+      "loss": 0.5221,
+      "step": 225
+    },
+    {
+      "epoch": 1.4533762057877815,
+      "grad_norm": 0.5225676833607608,
+      "learning_rate": 3.571261510207858e-05,
+      "loss": 0.5026,
+      "step": 226
+    },
+    {
+      "epoch": 1.459807073954984,
+      "grad_norm": 0.4737485519384992,
+      "learning_rate": 3.565668266434243e-05,
+      "loss": 0.5076,
+      "step": 227
+    },
+    {
+      "epoch": 1.4662379421221865,
+      "grad_norm": 0.5909989415292732,
+      "learning_rate": 3.5600432148920366e-05,
+      "loss": 0.4994,
+      "step": 228
+    },
+    {
+      "epoch": 1.472668810289389,
+      "grad_norm": 0.5661782640635898,
+      "learning_rate": 3.554386469858534e-05,
+      "loss": 0.5122,
+      "step": 229
+    },
+    {
+      "epoch": 1.4790996784565915,
+      "grad_norm": 0.5097174696672023,
+      "learning_rate": 3.548698146254911e-05,
+      "loss": 0.5239,
+      "step": 230
+    },
+    {
+      "epoch": 1.4855305466237942,
+      "grad_norm": 0.5434140128697973,
+      "learning_rate": 3.5429783596438864e-05,
+      "loss": 0.4992,
+      "step": 231
+    },
+    {
+      "epoch": 1.4919614147909968,
+      "grad_norm": 0.5799295479540262,
+      "learning_rate": 3.5372272262273734e-05,
+      "loss": 0.515,
+      "step": 232
+    },
+    {
+      "epoch": 1.4983922829581995,
+      "grad_norm": 0.5162993628654837,
+      "learning_rate": 3.531444862844121e-05,
+      "loss": 0.5225,
+      "step": 233
+    },
+    {
+      "epoch": 1.504823151125402,
+      "grad_norm": 0.6797172926898785,
+      "learning_rate": 3.5256313869673385e-05,
+      "loss": 0.5108,
+      "step": 234
+    },
+    {
+      "epoch": 1.5112540192926045,
+      "grad_norm": 0.485989540974696,
+      "learning_rate": 3.51978691670231e-05,
+      "loss": 0.5075,
+      "step": 235
+    },
+    {
+      "epoch": 1.517684887459807,
+      "grad_norm": 0.56031776669901,
+      "learning_rate": 3.5139115707839947e-05,
+      "loss": 0.509,
+      "step": 236
+    },
+    {
+      "epoch": 1.5241157556270095,
+      "grad_norm": 0.48021875320279245,
+      "learning_rate": 3.5080054685746155e-05,
+      "loss": 0.504,
+      "step": 237
+    },
+    {
+      "epoch": 1.5305466237942122,
+      "grad_norm": 0.47214897929224975,
+      "learning_rate": 3.502068730061231e-05,
+      "loss": 0.5211,
+      "step": 238
+    },
+    {
+      "epoch": 1.5369774919614148,
+      "grad_norm": 0.48294314268175204,
+      "learning_rate": 3.4961014758533025e-05,
+      "loss": 0.5021,
+      "step": 239
+    },
+    {
+      "epoch": 1.5434083601286175,
+      "grad_norm": 0.4113694799013807,
+      "learning_rate": 3.4901038271802404e-05,
+      "loss": 0.5184,
+      "step": 240
+    },
+    {
+      "epoch": 1.54983922829582,
+      "grad_norm": 0.41926168952311244,
+      "learning_rate": 3.4840759058889425e-05,
+      "loss": 0.5261,
+      "step": 241
+    },
+    {
+      "epoch": 1.5562700964630225,
+      "grad_norm": 0.509040362496011,
+      "learning_rate": 3.478017834441319e-05,
+      "loss": 0.5076,
+      "step": 242
+    },
+    {
+      "epoch": 1.562700964630225,
+      "grad_norm": 0.3932431816096667,
+      "learning_rate": 3.471929735911801e-05,
+      "loss": 0.5059,
+      "step": 243
+    },
+    {
+      "epoch": 1.5691318327974275,
+      "grad_norm": 0.41634496138466764,
+      "learning_rate": 3.4658117339848476e-05,
+      "loss": 0.5175,
+      "step": 244
+    },
+    {
+      "epoch": 1.5755627009646302,
+      "grad_norm": 0.4058902892872811,
+      "learning_rate": 3.459663952952425e-05,
+      "loss": 0.4996,
+      "step": 245
+    },
+    {
+      "epoch": 1.5819935691318328,
+      "grad_norm": 0.4588339041680799,
+      "learning_rate": 3.453486517711489e-05,
+      "loss": 0.525,
+      "step": 246
+    },
+    {
+      "epoch": 1.5884244372990355,
+      "grad_norm": 0.4045041584131534,
+      "learning_rate": 3.447279553761438e-05,
+      "loss": 0.5067,
+      "step": 247
+    },
+    {
+      "epoch": 1.594855305466238,
+      "grad_norm": 0.4516534525938604,
+      "learning_rate": 3.441043187201574e-05,
+      "loss": 0.4938,
+      "step": 248
+    },
+    {
+      "epoch": 1.6012861736334405,
+      "grad_norm": 0.43968031642345473,
+      "learning_rate": 3.434777544728535e-05,
+      "loss": 0.5073,
+      "step": 249
+    },
+    {
+      "epoch": 1.607717041800643,
+      "grad_norm": 0.4862454080541775,
+      "learning_rate": 3.428482753633719e-05,
+      "loss": 0.5015,
+      "step": 250
+    },
+    {
+      "epoch": 1.6141479099678455,
+      "grad_norm": 0.503278425755903,
+      "learning_rate": 3.422158941800706e-05,
+      "loss": 0.5063,
+      "step": 251
+    },
+    {
+      "epoch": 1.6205787781350482,
+      "grad_norm": 0.4535944512249209,
+      "learning_rate": 3.415806237702649e-05,
+      "loss": 0.514,
+      "step": 252
+    },
+    {
+      "epoch": 1.6270096463022508,
+      "grad_norm": 0.5032326783622072,
+      "learning_rate": 3.409424770399675e-05,
+      "loss": 0.4948,
+      "step": 253
+    },
+    {
+      "epoch": 1.6334405144694535,
+      "grad_norm": 0.4235714850808415,
+      "learning_rate": 3.403014669536254e-05,
+      "loss": 0.5018,
+      "step": 254
+    },
+    {
+      "epoch": 1.639871382636656,
+      "grad_norm": 0.6528530045866414,
+      "learning_rate": 3.396576065338572e-05,
+      "loss": 0.5091,
+      "step": 255
+    },
+    {
+      "epoch": 1.6463022508038585,
+      "grad_norm": 0.5888218448662212,
+      "learning_rate": 3.3901090886118814e-05,
+      "loss": 0.507,
+      "step": 256
+    },
+    {
+      "epoch": 1.652733118971061,
+      "grad_norm": 0.5622321070987689,
+      "learning_rate": 3.3836138707378434e-05,
+      "loss": 0.5049,
+      "step": 257
+    },
+    {
+      "epoch": 1.6591639871382635,
+      "grad_norm": 0.5995469349556404,
+      "learning_rate": 3.377090543671861e-05,
+      "loss": 0.5194,
+      "step": 258
+    },
+    {
+      "epoch": 1.6655948553054662,
+      "grad_norm": 0.4702549705052989,
+      "learning_rate": 3.370539239940398e-05,
+      "loss": 0.5039,
+      "step": 259
+    },
+    {
+      "epoch": 1.6720257234726688,
+      "grad_norm": 0.5530648047865966,
+      "learning_rate": 3.3639600926382854e-05,
+      "loss": 0.5107,
+      "step": 260
+    },
+    {
+      "epoch": 1.6784565916398715,
+      "grad_norm": 0.41673110351066345,
+      "learning_rate": 3.357353235426017e-05,
+      "loss": 0.5022,
+      "step": 261
+    },
+    {
+      "epoch": 1.684887459807074,
+      "grad_norm": 0.44081211863724745,
+      "learning_rate": 3.350718802527036e-05,
+      "loss": 0.5269,
+      "step": 262
+    },
+    {
+      "epoch": 1.6913183279742765,
+      "grad_norm": 0.47378262331981014,
+      "learning_rate": 3.344056928725007e-05,
+      "loss": 0.5153,
+      "step": 263
+    },
+    {
+      "epoch": 1.697749196141479,
+      "grad_norm": 0.44381911448039835,
+      "learning_rate": 3.337367749361079e-05,
+      "loss": 0.5016,
+      "step": 264
+    },
+    {
+      "epoch": 1.7041800643086815,
+      "grad_norm": 0.4794604635634858,
+      "learning_rate": 3.3306514003311305e-05,
+      "loss": 0.5242,
+      "step": 265
+    },
+    {
+      "epoch": 1.7106109324758842,
+      "grad_norm": 0.5053984638515733,
+      "learning_rate": 3.3239080180830194e-05,
+      "loss": 0.5132,
+      "step": 266
+    },
+    {
+      "epoch": 1.717041800643087,
+      "grad_norm": 0.5308453150998752,
+      "learning_rate": 3.317137739613799e-05,
+      "loss": 0.5022,
+      "step": 267
+    },
+    {
+      "epoch": 1.7234726688102895,
+      "grad_norm": 0.5698340450528304,
+      "learning_rate": 3.310340702466943e-05,
+      "loss": 0.519,
+      "step": 268
+    },
+    {
+      "epoch": 1.729903536977492,
+      "grad_norm": 0.46142254428517565,
+      "learning_rate": 3.3035170447295477e-05,
+      "loss": 0.5199,
+      "step": 269
+    },
+    {
+      "epoch": 1.7363344051446945,
+      "grad_norm": 0.520697787878528,
+      "learning_rate": 3.296666905029527e-05,
+      "loss": 0.5009,
+      "step": 270
+    },
+    {
+      "epoch": 1.742765273311897,
+      "grad_norm": 0.36612747577813315,
+      "learning_rate": 3.289790422532797e-05,
+      "loss": 0.5126,
+      "step": 271
+    },
+    {
+      "epoch": 1.7491961414790995,
+      "grad_norm": 0.4607496715577009,
+      "learning_rate": 3.282887736940448e-05,
+      "loss": 0.5147,
+      "step": 272
+    },
+    {
+      "epoch": 1.7556270096463023,
+      "grad_norm": 0.3668989653384773,
+      "learning_rate": 3.275958988485907e-05,
+      "loss": 0.5117,
+      "step": 273
+    },
+    {
+      "epoch": 1.762057877813505,
+      "grad_norm": 0.48341516601329765,
+      "learning_rate": 3.269004317932088e-05,
+      "loss": 0.5102,
+      "step": 274
+    },
+    {
+      "epoch": 1.7684887459807075,
+      "grad_norm": 0.43999263326894533,
+      "learning_rate": 3.262023866568531e-05,
+      "loss": 0.5064,
+      "step": 275
+    },
+    {
+      "epoch": 1.77491961414791,
+      "grad_norm": 0.48922643182576403,
+      "learning_rate": 3.255017776208537e-05,
+      "loss": 0.519,
+      "step": 276
+    },
+    {
+      "epoch": 1.7813504823151125,
+      "grad_norm": 0.5082242358909038,
+      "learning_rate": 3.2479861891862803e-05,
+      "loss": 0.526,
+      "step": 277
+    },
+    {
+      "epoch": 1.787781350482315,
+      "grad_norm": 0.4792654197302876,
+      "learning_rate": 3.24092924835392e-05,
+      "loss": 0.4995,
+      "step": 278
+    },
+    {
+      "epoch": 1.7942122186495175,
+      "grad_norm": 0.4522729725429338,
+      "learning_rate": 3.2338470970786975e-05,
+      "loss": 0.5043,
+      "step": 279
+    },
+    {
+      "epoch": 1.8006430868167203,
+      "grad_norm": 0.464231879470683,
+      "learning_rate": 3.2267398792400254e-05,
+      "loss": 0.5157,
+      "step": 280
+    },
+    {
+      "epoch": 1.807073954983923,
+      "grad_norm": 0.39009202659601355,
+      "learning_rate": 3.219607739226561e-05,
+      "loss": 0.5093,
+      "step": 281
+    },
+    {
+      "epoch": 1.8135048231511255,
+      "grad_norm": 0.39701513522833887,
+      "learning_rate": 3.212450821933277e-05,
+      "loss": 0.502,
+      "step": 282
+    },
+    {
+      "epoch": 1.819935691318328,
+      "grad_norm": 0.402838761134884,
+      "learning_rate": 3.205269272758513e-05,
+      "loss": 0.5039,
+      "step": 283
+    },
+    {
+      "epoch": 1.8263665594855305,
+      "grad_norm": 0.4258397971386761,
+      "learning_rate": 3.198063237601028e-05,
+      "loss": 0.5306,
+      "step": 284
+    },
+    {
+      "epoch": 1.832797427652733,
+      "grad_norm": 0.42742952212417085,
+      "learning_rate": 3.1908328628570306e-05,
+      "loss": 0.5137,
+      "step": 285
+    },
+    {
+      "epoch": 1.8392282958199357,
+      "grad_norm": 0.42229640914181593,
+      "learning_rate": 3.183578295417208e-05,
+      "loss": 0.5116,
+      "step": 286
+    },
+    {
+      "epoch": 1.8456591639871383,
+      "grad_norm": 0.420593756012977,
+      "learning_rate": 3.1762996826637386e-05,
+      "loss": 0.5046,
+      "step": 287
+    },
+    {
+      "epoch": 1.852090032154341,
+      "grad_norm": 0.5083756999230344,
+      "learning_rate": 3.168997172467303e-05,
+      "loss": 0.5143,
+      "step": 288
+    },
+    {
+      "epoch": 1.8585209003215435,
+      "grad_norm": 0.4178718551932849,
+      "learning_rate": 3.161670913184075e-05,
+      "loss": 0.5202,
+      "step": 289
+    },
+    {
+      "epoch": 1.864951768488746,
+      "grad_norm": 0.5877020292428744,
+      "learning_rate": 3.15432105365271e-05,
+      "loss": 0.5093,
+      "step": 290
+    },
+    {
+      "epoch": 1.8713826366559485,
+      "grad_norm": 0.3619257977183654,
+      "learning_rate": 3.146947743191323e-05,
+      "loss": 0.5083,
+      "step": 291
+    },
+    {
+      "epoch": 1.877813504823151,
+      "grad_norm": 0.5373446565783662,
+      "learning_rate": 3.139551131594449e-05,
+      "loss": 0.4984,
+      "step": 292
+    },
+    {
+      "epoch": 1.8842443729903537,
+      "grad_norm": 0.41947810760587556,
+      "learning_rate": 3.1321313691300055e-05,
+      "loss": 0.5035,
+      "step": 293
+    },
+    {
+      "epoch": 1.8906752411575563,
+      "grad_norm": 0.47687138554766845,
+      "learning_rate": 3.1246886065362384e-05,
+      "loss": 0.5156,
+      "step": 294
+    },
+    {
+      "epoch": 1.897106109324759,
+      "grad_norm": 0.4215865503118443,
+      "learning_rate": 3.117222995018659e-05,
+      "loss": 0.5055,
+      "step": 295
+    },
+    {
+      "epoch": 1.9035369774919615,
+      "grad_norm": 0.48107675778569436,
+      "learning_rate": 3.109734686246972e-05,
+      "loss": 0.5026,
+      "step": 296
+    },
+    {
+      "epoch": 1.909967845659164,
+      "grad_norm": 0.5392517175053669,
+      "learning_rate": 3.1022238323519945e-05,
+      "loss": 0.5009,
+      "step": 297
+    },
+    {
+      "epoch": 1.9163987138263665,
+      "grad_norm": 0.4651620328098306,
+      "learning_rate": 3.094690585922564e-05,
+      "loss": 0.5136,
+      "step": 298
+    },
+    {
+      "epoch": 1.922829581993569,
+      "grad_norm": 0.47235142578919365,
+      "learning_rate": 3.0871351000024425e-05,
+      "loss": 0.5237,
+      "step": 299
+    },
+    {
+      "epoch": 1.9292604501607717,
+      "grad_norm": 0.506545820963245,
+      "learning_rate": 3.0795575280872024e-05,
+      "loss": 0.5038,
+      "step": 300
+    },
+    {
+      "epoch": 1.9356913183279743,
+      "grad_norm": 0.4386605642519953,
+      "learning_rate": 3.071958024121111e-05,
+      "loss": 0.5417,
+      "step": 301
+    },
+    {
+      "epoch": 1.942122186495177,
+      "grad_norm": 0.5068932453220504,
+      "learning_rate": 3.0643367424940026e-05,
+      "loss": 0.4988,
+      "step": 302
+    },
+    {
+      "epoch": 1.9485530546623795,
+      "grad_norm": 0.40706958635441526,
+      "learning_rate": 3.05669383803814e-05,
+      "loss": 0.5177,
+      "step": 303
+    },
+    {
+      "epoch": 1.954983922829582,
+      "grad_norm": 0.5318892089739192,
+      "learning_rate": 3.049029466025073e-05,
+      "loss": 0.515,
+      "step": 304
+    },
+    {
+      "epoch": 1.9614147909967845,
+      "grad_norm": 0.44068399306925904,
+      "learning_rate": 3.0413437821624795e-05,
+      "loss": 0.5148,
+      "step": 305
+    },
+    {
+      "epoch": 1.967845659163987,
+      "grad_norm": 0.4528513683011978,
+      "learning_rate": 3.0336369425910058e-05,
+      "loss": 0.5071,
+      "step": 306
+    },
+    {
+      "epoch": 1.9742765273311897,
+      "grad_norm": 0.5246544586773848,
+      "learning_rate": 3.0259091038810916e-05,
+      "loss": 0.5082,
+      "step": 307
+    },
+    {
+      "epoch": 1.9807073954983923,
+      "grad_norm": 0.38871034271480154,
+      "learning_rate": 3.018160423029791e-05,
+      "loss": 0.5229,
+      "step": 308
+    },
+    {
+      "epoch": 1.987138263665595,
+      "grad_norm": 0.40273526828763445,
+      "learning_rate": 3.010391057457582e-05,
+      "loss": 0.5145,
+      "step": 309
+    },
+    {
+      "epoch": 1.9935691318327975,
+      "grad_norm": 0.44802486496553573,
+      "learning_rate": 3.0026011650051695e-05,
+      "loss": 0.4858,
+      "step": 310
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.44992613125378766,
+      "learning_rate": 2.994790903930278e-05,
+      "loss": 0.5149,
+      "step": 311
+    },
+    {
+      "epoch": 2.0064308681672025,
+      "grad_norm": 0.6647528945628535,
+      "learning_rate": 2.9869604329044345e-05,
+      "loss": 0.4287,
+      "step": 312
+    },
+    {
+      "epoch": 2.012861736334405,
+      "grad_norm": 0.7168947658958187,
+      "learning_rate": 2.9791099110097502e-05,
+      "loss": 0.4263,
+      "step": 313
+    },
+    {
+      "epoch": 2.0192926045016075,
+      "grad_norm": 0.4618213260898823,
+      "learning_rate": 2.9712394977356824e-05,
+      "loss": 0.4118,
+      "step": 314
+    },
+    {
+      "epoch": 2.0257234726688105,
+      "grad_norm": 0.6182835025602303,
+      "learning_rate": 2.9633493529757988e-05,
+      "loss": 0.4089,
+      "step": 315
+    },
+    {
+      "epoch": 2.032154340836013,
+      "grad_norm": 0.46151278794201206,
+      "learning_rate": 2.955439637024526e-05,
+      "loss": 0.4197,
+      "step": 316
+    },
+    {
+      "epoch": 2.0385852090032155,
+      "grad_norm": 0.4994231462526297,
+      "learning_rate": 2.947510510573897e-05,
+      "loss": 0.4191,
+      "step": 317
+    },
+    {
+      "epoch": 2.045016077170418,
+      "grad_norm": 0.5573932258943173,
+      "learning_rate": 2.9395621347102817e-05,
+      "loss": 0.4368,
+      "step": 318
+    },
+    {
+      "epoch": 2.0514469453376205,
+      "grad_norm": 0.44935859747981777,
+      "learning_rate": 2.931594670911119e-05,
+      "loss": 0.403,
+      "step": 319
+    },
+    {
+      "epoch": 2.057877813504823,
+      "grad_norm": 0.5944794827587898,
+      "learning_rate": 2.9236082810416316e-05,
+      "loss": 0.4228,
+      "step": 320
+    },
+    {
+      "epoch": 2.0643086816720255,
+      "grad_norm": 0.5714987763441384,
+      "learning_rate": 2.9156031273515427e-05,
+      "loss": 0.4436,
+      "step": 321
+    },
+    {
+      "epoch": 2.0707395498392285,
+      "grad_norm": 0.6002823095770211,
+      "learning_rate": 2.9075793724717755e-05,
+      "loss": 0.4002,
+      "step": 322
+    },
+    {
+      "epoch": 2.077170418006431,
+      "grad_norm": 0.5398634876829567,
+      "learning_rate": 2.8995371794111515e-05,
+      "loss": 0.4135,
+      "step": 323
+    },
+    {
+      "epoch": 2.0836012861736335,
+      "grad_norm": 0.591012932847218,
+      "learning_rate": 2.891476711553077e-05,
+      "loss": 0.3909,
+      "step": 324
+    },
+    {
+      "epoch": 2.090032154340836,
+      "grad_norm": 0.5517319879202164,
+      "learning_rate": 2.883398132652226e-05,
+      "loss": 0.4316,
+      "step": 325
+    },
+    {
+      "epoch": 2.0964630225080385,
+      "grad_norm": 0.5068351709623404,
+      "learning_rate": 2.8753016068312124e-05,
+      "loss": 0.4275,
+      "step": 326
+    },
+    {
+      "epoch": 2.102893890675241,
+      "grad_norm": 0.5931900353148353,
+      "learning_rate": 2.867187298577257e-05,
+      "loss": 0.4106,
+      "step": 327
+    },
+    {
+      "epoch": 2.1093247588424435,
+      "grad_norm": 0.3918558956713872,
+      "learning_rate": 2.859055372738841e-05,
+      "loss": 0.4031,
+      "step": 328
+    },
+    {
+      "epoch": 2.1157556270096465,
+      "grad_norm": 0.5497454667039948,
+      "learning_rate": 2.850905994522364e-05,
+      "loss": 0.4116,
+      "step": 329
+    },
+    {
+      "epoch": 2.122186495176849,
+      "grad_norm": 0.41293894559462124,
+      "learning_rate": 2.842739329488783e-05,
+      "loss": 0.4349,
+      "step": 330
+    },
+    {
+      "epoch": 2.1286173633440515,
+      "grad_norm": 0.47012857990766127,
+      "learning_rate": 2.83455554355025e-05,
+      "loss": 0.4226,
+      "step": 331
+    },
+    {
+      "epoch": 2.135048231511254,
+      "grad_norm": 0.49571483122654914,
+      "learning_rate": 2.8263548029667413e-05,
+      "loss": 0.4311,
+      "step": 332
+    },
+    {
+      "epoch": 2.1414790996784565,
+      "grad_norm": 0.4418403283009763,
+      "learning_rate": 2.8181372743426805e-05,
+      "loss": 0.4004,
+      "step": 333
+    },
+    {
+      "epoch": 2.147909967845659,
+      "grad_norm": 0.3963850265086265,
+      "learning_rate": 2.8099031246235518e-05,
+      "loss": 0.411,
+      "step": 334
+    },
+    {
+      "epoch": 2.154340836012862,
+      "grad_norm": 0.400560287536601,
+      "learning_rate": 2.8016525210925146e-05,
+      "loss": 0.4275,
+      "step": 335
+    },
+    {
+      "epoch": 2.1607717041800645,
+      "grad_norm": 0.3876396299454253,
+      "learning_rate": 2.793385631366993e-05,
+      "loss": 0.4158,
+      "step": 336
+    },
+    {
+      "epoch": 2.167202572347267,
+      "grad_norm": 0.4059506494242803,
+      "learning_rate": 2.7851026233952844e-05,
+      "loss": 0.4229,
+      "step": 337
+    },
+    {
+      "epoch": 2.1736334405144695,
+      "grad_norm": 0.4757267613591744,
+      "learning_rate": 2.7768036654531365e-05,
+      "loss": 0.4361,
+      "step": 338
+    },
+    {
+      "epoch": 2.180064308681672,
+      "grad_norm": 0.4512207921188538,
+      "learning_rate": 2.768488926140336e-05,
+      "loss": 0.432,
+      "step": 339
+    },
+    {
+      "epoch": 2.1864951768488745,
+      "grad_norm": 0.44275177188546366,
+      "learning_rate": 2.760158574377278e-05,
+      "loss": 0.4247,
+      "step": 340
+    },
+    {
+      "epoch": 2.192926045016077,
+      "grad_norm": 0.44419958942182214,
+      "learning_rate": 2.7518127794015384e-05,
+      "loss": 0.4179,
+      "step": 341
+    },
+    {
+      "epoch": 2.19935691318328,
+      "grad_norm": 0.4278691906925567,
+      "learning_rate": 2.743451710764433e-05,
+      "loss": 0.4368,
+      "step": 342
+    },
+    {
+      "epoch": 2.2057877813504825,
+      "grad_norm": 0.48103181469759326,
+      "learning_rate": 2.7350755383275757e-05,
+      "loss": 0.3995,
+      "step": 343
+    },
+    {
+      "epoch": 2.212218649517685,
+      "grad_norm": 0.4240098577929835,
+      "learning_rate": 2.7266844322594228e-05,
+      "loss": 0.4115,
+      "step": 344
+    },
+    {
+      "epoch": 2.2186495176848875,
+      "grad_norm": 0.49820215802380435,
+      "learning_rate": 2.718278563031821e-05,
+      "loss": 0.435,
+      "step": 345
+    },
+    {
+      "epoch": 2.22508038585209,
+      "grad_norm": 0.3877604432372655,
+      "learning_rate": 2.7098581014165426e-05,
+      "loss": 0.4362,
+      "step": 346
+    },
+    {
+      "epoch": 2.2315112540192925,
+      "grad_norm": 0.506445549832141,
+      "learning_rate": 2.7014232184818134e-05,
+      "loss": 0.4102,
+      "step": 347
+    },
+    {
+      "epoch": 2.237942122186495,
+      "grad_norm": 0.4098592246520702,
+      "learning_rate": 2.6929740855888416e-05,
+      "loss": 0.414,
+      "step": 348
+    },
+    {
+      "epoch": 2.244372990353698,
+      "grad_norm": 0.5663405400503388,
+      "learning_rate": 2.684510874388333e-05,
+      "loss": 0.4191,
+      "step": 349
+    },
+    {
+      "epoch": 2.2508038585209005,
+      "grad_norm": 0.4273707729795677,
+      "learning_rate": 2.6760337568170056e-05,
+      "loss": 0.4158,
+      "step": 350
+    },
+    {
+      "epoch": 2.257234726688103,
+      "grad_norm": 0.4986604561853341,
+      "learning_rate": 2.6675429050940976e-05,
+      "loss": 0.4218,
+      "step": 351
+    },
+    {
+      "epoch": 2.2636655948553055,
+      "grad_norm": 0.4675367887772543,
+      "learning_rate": 2.659038491717864e-05,
+      "loss": 0.4095,
+      "step": 352
+    },
+    {
+      "epoch": 2.270096463022508,
+      "grad_norm": 0.5057278576934203,
+      "learning_rate": 2.650520689462079e-05,
+      "loss": 0.4384,
+      "step": 353
+    },
+    {
+      "epoch": 2.2765273311897105,
+      "grad_norm": 0.41733242728535985,
+      "learning_rate": 2.6419896713725194e-05,
+      "loss": 0.4252,
+      "step": 354
+    },
+    {
+      "epoch": 2.282958199356913,
+      "grad_norm": 0.4683245477705893,
+      "learning_rate": 2.6334456107634526e-05,
+      "loss": 0.4066,
+      "step": 355
+    },
+    {
+      "epoch": 2.289389067524116,
+      "grad_norm": 0.46673184121507844,
+      "learning_rate": 2.6248886812141164e-05,
+      "loss": 0.4063,
+      "step": 356
+    },
+    {
+      "epoch": 2.2958199356913185,
+      "grad_norm": 0.42307659075383675,
+      "learning_rate": 2.6163190565651895e-05,
+      "loss": 0.4064,
+      "step": 357
+    },
+    {
+      "epoch": 2.302250803858521,
+      "grad_norm": 0.4171903856498429,
+      "learning_rate": 2.6077369109152617e-05,
+      "loss": 0.4271,
+      "step": 358
+    },
+    {
+      "epoch": 2.3086816720257235,
+      "grad_norm": 0.4990975466004433,
+      "learning_rate": 2.599142418617299e-05,
+      "loss": 0.4337,
+      "step": 359
+    },
+    {
+      "epoch": 2.315112540192926,
+      "grad_norm": 0.3959570339594964,
+      "learning_rate": 2.5905357542750937e-05,
+      "loss": 0.3997,
+      "step": 360
+    },
+    {
+      "epoch": 2.3215434083601285,
+      "grad_norm": 0.4765248478457776,
+      "learning_rate": 2.5819170927397288e-05,
+      "loss": 0.4048,
+      "step": 361
+    },
+    {
+      "epoch": 2.327974276527331,
+      "grad_norm": 0.3882612926978572,
+      "learning_rate": 2.573286609106016e-05,
+      "loss": 0.4302,
+      "step": 362
+    },
+    {
+      "epoch": 2.334405144694534,
+      "grad_norm": 0.45198493980403964,
+      "learning_rate": 2.5646444787089435e-05,
+      "loss": 0.437,
+      "step": 363
+    },
+    {
+      "epoch": 2.3408360128617365,
+      "grad_norm": 0.38217511369100665,
+      "learning_rate": 2.555990877120111e-05,
+      "loss": 0.4192,
+      "step": 364
+    },
+    {
+      "epoch": 2.347266881028939,
+      "grad_norm": 0.42262056495561456,
+      "learning_rate": 2.5473259801441663e-05,
+      "loss": 0.4191,
+      "step": 365
+    },
+    {
+      "epoch": 2.3536977491961415,
+      "grad_norm": 0.41765273157933114,
+      "learning_rate": 2.53864996381523e-05,
+      "loss": 0.4142,
+      "step": 366
+    },
+    {
+      "epoch": 2.360128617363344,
+      "grad_norm": 0.40488762786497534,
+      "learning_rate": 2.529963004393324e-05,
+      "loss": 0.4187,
+      "step": 367
+    },
+    {
+      "epoch": 2.3665594855305465,
+      "grad_norm": 0.4080271801790144,
+      "learning_rate": 2.521265278360783e-05,
+      "loss": 0.414,
+      "step": 368
+    },
+    {
+      "epoch": 2.372990353697749,
+      "grad_norm": 0.40483254655041323,
+      "learning_rate": 2.5125569624186782e-05,
+      "loss": 0.4247,
+      "step": 369
+    },
+    {
+      "epoch": 2.379421221864952,
+      "grad_norm": 0.4305982633063572,
+      "learning_rate": 2.5038382334832207e-05,
+      "loss": 0.3911,
+      "step": 370
+    },
+    {
+      "epoch": 2.3858520900321545,
+      "grad_norm": 0.4122997020523331,
+      "learning_rate": 2.4951092686821697e-05,
+      "loss": 0.39,
+      "step": 371
+    },
+    {
+      "epoch": 2.392282958199357,
+      "grad_norm": 0.40274674461442084,
+      "learning_rate": 2.486370245351235e-05,
+      "loss": 0.4198,
+      "step": 372
+    },
+    {
+      "epoch": 2.3987138263665595,
+      "grad_norm": 0.42063234047236986,
+      "learning_rate": 2.4776213410304722e-05,
+      "loss": 0.424,
+      "step": 373
+    },
+    {
+      "epoch": 2.405144694533762,
+      "grad_norm": 0.37679725326760555,
+      "learning_rate": 2.4688627334606773e-05,
+      "loss": 0.398,
+      "step": 374
+    },
+    {
+      "epoch": 2.4115755627009645,
+      "grad_norm": 0.41889135037527103,
+      "learning_rate": 2.460094600579777e-05,
+      "loss": 0.4317,
+      "step": 375
+    },
+    {
+      "epoch": 2.418006430868167,
+      "grad_norm": 0.3838772373610373,
+      "learning_rate": 2.4513171205192076e-05,
+      "loss": 0.4295,
+      "step": 376
+    },
+    {
+      "epoch": 2.42443729903537,
+      "grad_norm": 0.4215792016331523,
+      "learning_rate": 2.4425304716003065e-05,
+      "loss": 0.4293,
+      "step": 377
+    },
+    {
+      "epoch": 2.4308681672025725,
+      "grad_norm": 0.3977033456422348,
+      "learning_rate": 2.433734832330679e-05,
+      "loss": 0.4183,
+      "step": 378
+    },
+    {
+      "epoch": 2.437299035369775,
+      "grad_norm": 0.3528962937028162,
+      "learning_rate": 2.4249303814005787e-05,
+      "loss": 0.4348,
+      "step": 379
+    },
+    {
+      "epoch": 2.4437299035369775,
+      "grad_norm": 0.3444297123425228,
+      "learning_rate": 2.416117297679274e-05,
+      "loss": 0.418,
+      "step": 380
+    },
+    {
+      "epoch": 2.45016077170418,
+      "grad_norm": 0.37390979375744465,
+      "learning_rate": 2.407295760211415e-05,
+      "loss": 0.4314,
+      "step": 381
+    },
+    {
+      "epoch": 2.4565916398713825,
+      "grad_norm": 0.39952459117481703,
+      "learning_rate": 2.3984659482133964e-05,
+      "loss": 0.4139,
+      "step": 382
+    },
+    {
+      "epoch": 2.463022508038585,
+      "grad_norm": 0.3804428944868265,
+      "learning_rate": 2.3896280410697184e-05,
+      "loss": 0.3907,
+      "step": 383
+    },
+    {
+      "epoch": 2.469453376205788,
+      "grad_norm": 0.3674572711621663,
+      "learning_rate": 2.380782218329337e-05,
+      "loss": 0.4013,
+      "step": 384
+    },
+    {
+      "epoch": 2.4758842443729905,
+      "grad_norm": 0.38222903425625193,
+      "learning_rate": 2.3719286597020223e-05,
+      "loss": 0.4318,
+      "step": 385
+    },
+    {
+      "epoch": 2.482315112540193,
+      "grad_norm": 0.4061156008709529,
+      "learning_rate": 2.3630675450547035e-05,
+      "loss": 0.3992,
+      "step": 386
+    },
+    {
+      "epoch": 2.4887459807073955,
+      "grad_norm": 0.37476585596417883,
+      "learning_rate": 2.354199054407817e-05,
+      "loss": 0.4151,
+      "step": 387
+    },
+    {
+      "epoch": 2.495176848874598,
+      "grad_norm": 0.42119555154139743,
+      "learning_rate": 2.3453233679316474e-05,
+      "loss": 0.4062,
+      "step": 388
+    },
+    {
+      "epoch": 2.5016077170418005,
+      "grad_norm": 0.39495496392373747,
+      "learning_rate": 2.33644066594267e-05,
+      "loss": 0.4272,
+      "step": 389
+    },
+    {
+      "epoch": 2.508038585209003,
+      "grad_norm": 0.401750662623451,
+      "learning_rate": 2.327551128899884e-05,
+      "loss": 0.4225,
+      "step": 390
+    },
+    {
+      "epoch": 2.514469453376206,
+      "grad_norm": 0.3636214667784954,
+      "learning_rate": 2.3186549374011484e-05,
+      "loss": 0.4074,
+      "step": 391
+    },
+    {
+      "epoch": 2.5209003215434085,
+      "grad_norm": 0.409731943120677,
+      "learning_rate": 2.3097522721795135e-05,
+      "loss": 0.4138,
+      "step": 392
+    },
+    {
+      "epoch": 2.527331189710611,
+      "grad_norm": 0.41819215942803173,
+      "learning_rate": 2.3008433140995473e-05,
+      "loss": 0.4273,
+      "step": 393
+    },
+    {
+      "epoch": 2.5337620578778135,
+      "grad_norm": 0.4067917678154936,
+      "learning_rate": 2.2919282441536622e-05,
+      "loss": 0.4282,
+      "step": 394
+    },
+    {
+      "epoch": 2.540192926045016,
+      "grad_norm": 0.43999642613256557,
+      "learning_rate": 2.283007243458438e-05,
+      "loss": 0.4103,
+      "step": 395
+    },
+    {
+      "epoch": 2.5466237942122185,
+      "grad_norm": 0.37707548388284345,
+      "learning_rate": 2.2740804932509443e-05,
+      "loss": 0.4155,
+      "step": 396
+    },
+    {
+      "epoch": 2.553054662379421,
+      "grad_norm": 0.45630461317249305,
+      "learning_rate": 2.2651481748850522e-05,
+      "loss": 0.4062,
+      "step": 397
+    },
+    {
+      "epoch": 2.559485530546624,
+      "grad_norm": 0.4019638185705102,
+      "learning_rate": 2.2562104698277587e-05,
+      "loss": 0.4054,
+      "step": 398
+    },
+    {
+      "epoch": 2.5659163987138265,
+      "grad_norm": 0.452881602759346,
+      "learning_rate": 2.247267559655492e-05,
+      "loss": 0.4068,
+      "step": 399
+    },
+    {
+      "epoch": 2.572347266881029,
+      "grad_norm": 0.40471187969458244,
+      "learning_rate": 2.2383196260504297e-05,
+      "loss": 0.4289,
+      "step": 400
+    },
+    {
+      "epoch": 2.5787781350482315,
+      "grad_norm": 0.45026153483875125,
+      "learning_rate": 2.2293668507968015e-05,
+      "loss": 0.4027,
+      "step": 401
+    },
+    {
+      "epoch": 2.585209003215434,
+      "grad_norm": 0.4244819736473437,
+      "learning_rate": 2.2204094157772e-05,
+      "loss": 0.4435,
+      "step": 402
+    },
+    {
+      "epoch": 2.5916398713826365,
+      "grad_norm": 0.3793703429152497,
+      "learning_rate": 2.2114475029688845e-05,
+      "loss": 0.4186,
+      "step": 403
+    },
+    {
+      "epoch": 2.598070739549839,
+      "grad_norm": 0.48272960686216815,
+      "learning_rate": 2.202481294440086e-05,
+      "loss": 0.4194,
+      "step": 404
+    },
+    {
+      "epoch": 2.604501607717042,
+      "grad_norm": 0.354469132752044,
+      "learning_rate": 2.1935109723463018e-05,
+      "loss": 0.4222,
+      "step": 405
+    },
+    {
+      "epoch": 2.6109324758842445,
+      "grad_norm": 0.4928379051181033,
+      "learning_rate": 2.184536718926604e-05,
+      "loss": 0.4213,
+      "step": 406
+    },
+    {
+      "epoch": 2.617363344051447,
+      "grad_norm": 0.400485176897503,
+      "learning_rate": 2.1755587164999308e-05,
+      "loss": 0.4379,
+      "step": 407
+    },
+    {
+      "epoch": 2.6237942122186495,
+      "grad_norm": 0.4227671618965105,
+      "learning_rate": 2.1665771474613844e-05,
+      "loss": 0.4108,
+      "step": 408
+    },
+    {
+      "epoch": 2.630225080385852,
+      "grad_norm": 0.41665136995023616,
+      "learning_rate": 2.1575921942785247e-05,
+      "loss": 0.4414,
+      "step": 409
+    },
+    {
+      "epoch": 2.6366559485530545,
+      "grad_norm": 0.41330338591303095,
+      "learning_rate": 2.1486040394876653e-05,
+      "loss": 0.4185,
+      "step": 410
+    },
+    {
+      "epoch": 2.643086816720257,
+      "grad_norm": 0.4308954366383958,
+      "learning_rate": 2.1396128656901613e-05,
+      "loss": 0.4324,
+      "step": 411
+    },
+    {
+      "epoch": 2.64951768488746,
+      "grad_norm": 0.396755042878508,
+      "learning_rate": 2.1306188555487012e-05,
+      "loss": 0.4065,
+      "step": 412
+    },
+    {
+      "epoch": 2.6559485530546625,
+      "grad_norm": 0.4493581590494059,
+      "learning_rate": 2.1216221917835963e-05,
+      "loss": 0.4428,
+      "step": 413
+    },
+    {
+      "epoch": 2.662379421221865,
+      "grad_norm": 0.3615612042430632,
+      "learning_rate": 2.1126230571690688e-05,
+      "loss": 0.4406,
+      "step": 414
+    },
+    {
+      "epoch": 2.6688102893890675,
+      "grad_norm": 0.3730567208860946,
+      "learning_rate": 2.1036216345295384e-05,
+      "loss": 0.4056,
+      "step": 415
+    },
+    {
+      "epoch": 2.67524115755627,
+      "grad_norm": 0.3606084375386736,
+      "learning_rate": 2.0946181067359066e-05,
+      "loss": 0.4223,
+      "step": 416
+    },
+    {
+      "epoch": 2.6816720257234725,
+      "grad_norm": 0.3980629387043826,
+      "learning_rate": 2.085612656701843e-05,
+      "loss": 0.4129,
+      "step": 417
+    },
+    {
+      "epoch": 2.688102893890675,
+      "grad_norm": 0.38669866056747154,
+      "learning_rate": 2.076605467380071e-05,
+      "loss": 0.4211,
+      "step": 418
+    },
+    {
+      "epoch": 2.694533762057878,
+      "grad_norm": 0.38877719567354047,
+      "learning_rate": 2.0675967217586453e-05,
+      "loss": 0.4097,
+      "step": 419
+    },
+    {
+      "epoch": 2.7009646302250805,
+      "grad_norm": 0.3907697728479256,
+      "learning_rate": 2.058586602857242e-05,
+      "loss": 0.4232,
+      "step": 420
+    },
+    {
+      "epoch": 2.707395498392283,
+      "grad_norm": 0.4154425957737256,
+      "learning_rate": 2.049575293723433e-05,
+      "loss": 0.4226,
+      "step": 421
+    },
+    {
+      "epoch": 2.7138263665594855,
+      "grad_norm": 0.362008498405758,
+      "learning_rate": 2.0405629774289726e-05,
+      "loss": 0.4179,
+      "step": 422
+    },
+    {
+      "epoch": 2.720257234726688,
+      "grad_norm": 0.3783004284520917,
+      "learning_rate": 2.0315498370660767e-05,
+      "loss": 0.4186,
+      "step": 423
+    },
+    {
+      "epoch": 2.7266881028938905,
+      "grad_norm": 0.39210883305746097,
+      "learning_rate": 2.022536055743702e-05,
+      "loss": 0.4294,
+      "step": 424
+    },
+    {
+      "epoch": 2.733118971061093,
+      "grad_norm": 0.3306443409459496,
+      "learning_rate": 2.0135218165838256e-05,
+      "loss": 0.4203,
+      "step": 425
+    },
+    {
+      "epoch": 2.739549839228296,
+      "grad_norm": 0.3938452571698424,
+      "learning_rate": 2.004507302717729e-05,
+      "loss": 0.4045,
+      "step": 426
+    },
+    {
+      "epoch": 2.7459807073954985,
+      "grad_norm": 0.34812659623890796,
+      "learning_rate": 1.9954926972822716e-05,
+      "loss": 0.4185,
+      "step": 427
+    },
+    {
+      "epoch": 2.752411575562701,
+      "grad_norm": 0.38511842967731696,
+      "learning_rate": 1.9864781834161748e-05,
+      "loss": 0.401,
+      "step": 428
+    },
+    {
+      "epoch": 2.7588424437299035,
+      "grad_norm": 0.3664578122176016,
+      "learning_rate": 1.9774639442562994e-05,
+      "loss": 0.4296,
+      "step": 429
+    },
+    {
+      "epoch": 2.765273311897106,
+      "grad_norm": 0.3534177510621182,
+      "learning_rate": 1.968450162933924e-05,
+      "loss": 0.4191,
+      "step": 430
+    },
+    {
+      "epoch": 2.7717041800643085,
+      "grad_norm": 0.37106435448060127,
+      "learning_rate": 1.959437022571028e-05,
+      "loss": 0.4096,
+      "step": 431
+    },
+    {
+      "epoch": 2.778135048231511,
+      "grad_norm": 0.35898089463840865,
+      "learning_rate": 1.9504247062765674e-05,
+      "loss": 0.4313,
+      "step": 432
+    },
+    {
+      "epoch": 2.784565916398714,
+      "grad_norm": 0.33865317022516006,
+      "learning_rate": 1.941413397142759e-05,
+      "loss": 0.4348,
+      "step": 433
+    },
+    {
+      "epoch": 2.7909967845659165,
+      "grad_norm": 0.3367494195662646,
+      "learning_rate": 1.932403278241355e-05,
+      "loss": 0.4071,
+      "step": 434
+    },
+    {
+      "epoch": 2.797427652733119,
+      "grad_norm": 0.35917915685142043,
+      "learning_rate": 1.9233945326199295e-05,
+      "loss": 0.4139,
+      "step": 435
+    },
+    {
+      "epoch": 2.8038585209003215,
+      "grad_norm": 0.3349790261187887,
+      "learning_rate": 1.9143873432981566e-05,
+      "loss": 0.439,
+      "step": 436
+    },
+    {
+      "epoch": 2.810289389067524,
+      "grad_norm": 0.3386796606602423,
+      "learning_rate": 1.905381893264094e-05,
+      "loss": 0.4174,
+      "step": 437
+    },
+    {
+      "epoch": 2.816720257234727,
+      "grad_norm": 0.37311347044535176,
+      "learning_rate": 1.896378365470462e-05,
+      "loss": 0.424,
+      "step": 438
+    },
+    {
+      "epoch": 2.823151125401929,
+      "grad_norm": 0.3543481522948843,
+      "learning_rate": 1.8873769428309315e-05,
+      "loss": 0.4487,
+      "step": 439
+    },
+    {
+      "epoch": 2.829581993569132,
+      "grad_norm": 0.35174149614975103,
+      "learning_rate": 1.878377808216404e-05,
+      "loss": 0.3996,
+      "step": 440
+    },
+    {
+      "epoch": 2.8360128617363345,
+      "grad_norm": 0.38242028517490756,
+      "learning_rate": 1.8693811444512998e-05,
+      "loss": 0.4038,
+      "step": 441
+    },
+    {
+      "epoch": 2.842443729903537,
+      "grad_norm": 0.3564288757729943,
+      "learning_rate": 1.8603871343098397e-05,
+      "loss": 0.4244,
+      "step": 442
+    },
+    {
+      "epoch": 2.8488745980707395,
+      "grad_norm": 0.3897350965863513,
+      "learning_rate": 1.851395960512335e-05,
+      "loss": 0.4076,
+      "step": 443
+    },
+    {
+      "epoch": 2.855305466237942,
+      "grad_norm": 0.4082600108340964,
+      "learning_rate": 1.8424078057214753e-05,
+      "loss": 0.43,
+      "step": 444
+    },
+    {
+      "epoch": 2.861736334405145,
+      "grad_norm": 0.36120387049878894,
+      "learning_rate": 1.8334228525386166e-05,
+      "loss": 0.4023,
+      "step": 445
+    },
+    {
+      "epoch": 2.868167202572347,
+      "grad_norm": 0.35734988643281534,
+      "learning_rate": 1.8244412835000696e-05,
+      "loss": 0.4376,
+      "step": 446
+    },
+    {
+      "epoch": 2.87459807073955,
+      "grad_norm": 0.3681854421194748,
+      "learning_rate": 1.815463281073396e-05,
+      "loss": 0.4045,
+      "step": 447
+    },
+    {
+      "epoch": 2.8810289389067525,
+      "grad_norm": 0.9736526939562563,
+      "learning_rate": 1.8064890276536985e-05,
+      "loss": 0.4099,
+      "step": 448
+    },
+    {
+      "epoch": 2.887459807073955,
+      "grad_norm": 0.4027769144827298,
+      "learning_rate": 1.7975187055599153e-05,
+      "loss": 0.4433,
+      "step": 449
+    },
+    {
+      "epoch": 2.8938906752411575,
+      "grad_norm": 0.3751995588652852,
+      "learning_rate": 1.7885524970311158e-05,
+      "loss": 0.4248,
+      "step": 450
+    },
+    {
+      "epoch": 2.90032154340836,
+      "grad_norm": 0.42015767109305785,
+      "learning_rate": 1.7795905842228008e-05,
+      "loss": 0.388,
+      "step": 451
+    },
+    {
+      "epoch": 2.906752411575563,
+      "grad_norm": 0.34334134430703095,
+      "learning_rate": 1.7706331492031995e-05,
+      "loss": 0.4115,
+      "step": 452
+    },
+    {
+      "epoch": 2.913183279742765,
+      "grad_norm": 0.37761054940371147,
+      "learning_rate": 1.7616803739495716e-05,
+      "loss": 0.4138,
+      "step": 453
+    },
+    {
+      "epoch": 2.919614147909968,
+      "grad_norm": 0.3356430895455353,
+      "learning_rate": 1.7527324403445086e-05,
+      "loss": 0.4033,
+      "step": 454
+    },
+    {
+      "epoch": 2.9260450160771705,
+      "grad_norm": 0.3860642493958761,
+      "learning_rate": 1.7437895301722424e-05,
+      "loss": 0.4277,
+      "step": 455
+    },
+    {
+      "epoch": 2.932475884244373,
+      "grad_norm": 0.36522544257627154,
+      "learning_rate": 1.734851825114948e-05,
+      "loss": 0.4169,
+      "step": 456
+    },
+    {
+      "epoch": 2.9389067524115755,
+      "grad_norm": 0.3671503027694236,
+      "learning_rate": 1.7259195067490567e-05,
+      "loss": 0.4285,
+      "step": 457
+    },
+    {
+      "epoch": 2.945337620578778,
+      "grad_norm": 0.36388073283214833,
+      "learning_rate": 1.7169927565415624e-05,
+      "loss": 0.4139,
+      "step": 458
+    },
+    {
+      "epoch": 2.951768488745981,
+      "grad_norm": 0.3488449299032217,
+      "learning_rate": 1.708071755846338e-05,
+      "loss": 0.4199,
+      "step": 459
+    },
+    {
+      "epoch": 2.958199356913183,
+      "grad_norm": 0.35513138045091164,
+      "learning_rate": 1.6991566859004534e-05,
+      "loss": 0.4165,
+      "step": 460
+    },
+    {
+      "epoch": 2.964630225080386,
+      "grad_norm": 0.3426432631224078,
+      "learning_rate": 1.6902477278204875e-05,
+      "loss": 0.434,
+      "step": 461
+    },
+    {
+      "epoch": 2.9710610932475885,
+      "grad_norm": 0.386152174466679,
+      "learning_rate": 1.681345062598852e-05,
+      "loss": 0.4111,
+      "step": 462
+    },
+    {
+      "epoch": 2.977491961414791,
+      "grad_norm": 0.3527196772848708,
+      "learning_rate": 1.6724488711001166e-05,
+      "loss": 0.4467,
+      "step": 463
+    },
+    {
+      "epoch": 2.9839228295819935,
+      "grad_norm": 0.35001625448663354,
+      "learning_rate": 1.66355933405733e-05,
+      "loss": 0.4209,
+      "step": 464
+    },
+    {
+      "epoch": 2.990353697749196,
+      "grad_norm": 0.40721218863788,
+      "learning_rate": 1.6546766320683532e-05,
+      "loss": 0.4199,
+      "step": 465
+    },
+    {
+      "epoch": 2.996784565916399,
+      "grad_norm": 0.3655958412277545,
+      "learning_rate": 1.6458009455921842e-05,
+      "loss": 0.3981,
+      "step": 466
+    },
+    {
+      "epoch": 3.0032154340836015,
+      "grad_norm": 0.4102349209568708,
+      "learning_rate": 1.636932454945297e-05,
+      "loss": 0.3891,
+      "step": 467
+    },
+    {
+      "epoch": 3.009646302250804,
+      "grad_norm": 0.41532136993824226,
+      "learning_rate": 1.6280713402979784e-05,
+      "loss": 0.3477,
+      "step": 468
+    },
+    {
+      "epoch": 3.0160771704180065,
+      "grad_norm": 0.6169474318758771,
+      "learning_rate": 1.619217781670663e-05,
+      "loss": 0.3435,
+      "step": 469
+    },
+    {
+      "epoch": 3.022508038585209,
+      "grad_norm": 0.5037930665617163,
+      "learning_rate": 1.6103719589302826e-05,
+      "loss": 0.3488,
+      "step": 470
+    },
+    {
+      "epoch": 3.0289389067524115,
+      "grad_norm": 0.5731525233393719,
+      "learning_rate": 1.601534051786604e-05,
+      "loss": 0.3503,
+      "step": 471
+    },
+    {
+      "epoch": 3.035369774919614,
+      "grad_norm": 0.5443770556136349,
+      "learning_rate": 1.592704239788586e-05,
+      "loss": 0.3481,
+      "step": 472
+    },
+    {
+      "epoch": 3.0418006430868165,
+      "grad_norm": 0.45479991470010417,
+      "learning_rate": 1.583882702320727e-05,
+      "loss": 0.3466,
+      "step": 473
+    },
+    {
+      "epoch": 3.0482315112540195,
+      "grad_norm": 0.4660904730868544,
+      "learning_rate": 1.5750696185994226e-05,
+      "loss": 0.3348,
+      "step": 474
+    },
+    {
+      "epoch": 3.054662379421222,
+      "grad_norm": 0.43734423911973497,
+      "learning_rate": 1.5662651676693213e-05,
+      "loss": 0.3188,
+      "step": 475
+    },
+    {
+      "epoch": 3.0610932475884245,
+      "grad_norm": 0.3762681204775496,
+      "learning_rate": 1.5574695283996942e-05,
+      "loss": 0.3215,
+      "step": 476
+    },
+    {
+      "epoch": 3.067524115755627,
+      "grad_norm": 0.42787289032141385,
+      "learning_rate": 1.5486828794807923e-05,
+      "loss": 0.3421,
+      "step": 477
+    },
+    {
+      "epoch": 3.0739549839228295,
+      "grad_norm": 0.3897816235186344,
+      "learning_rate": 1.5399053994202243e-05,
+      "loss": 0.3361,
+      "step": 478
+    },
+    {
+      "epoch": 3.080385852090032,
+      "grad_norm": 0.41498121393671433,
+      "learning_rate": 1.531137266539323e-05,
+      "loss": 0.3308,
+      "step": 479
+    },
+    {
+      "epoch": 3.0868167202572345,
+      "grad_norm": 0.39646083993636266,
+      "learning_rate": 1.5223786589695281e-05,
+      "loss": 0.3425,
+      "step": 480
+    },
+    {
+      "epoch": 3.0932475884244375,
+      "grad_norm": 0.3724151354324873,
+      "learning_rate": 1.5136297546487658e-05,
+      "loss": 0.3743,
+      "step": 481
+    },
+    {
+      "epoch": 3.09967845659164,
+      "grad_norm": 0.3874304470470023,
+      "learning_rate": 1.5048907313178313e-05,
+      "loss": 0.3308,
+      "step": 482
+    },
+    {
+      "epoch": 3.1061093247588425,
+      "grad_norm": 0.3811422217949994,
+      "learning_rate": 1.49616176651678e-05,
+      "loss": 0.3153,
+      "step": 483
+    },
+    {
+      "epoch": 3.112540192926045,
+      "grad_norm": 0.3972102549577263,
+      "learning_rate": 1.4874430375813223e-05,
+      "loss": 0.3415,
+      "step": 484
+    },
+    {
+      "epoch": 3.1189710610932475,
+      "grad_norm": 0.3639595571598912,
+      "learning_rate": 1.4787347216392171e-05,
+      "loss": 0.3294,
+      "step": 485
+    },
+    {
+      "epoch": 3.12540192926045,
+      "grad_norm": 0.37095643913123144,
+      "learning_rate": 1.4700369956066771e-05,
+      "loss": 0.3294,
+      "step": 486
+    },
+    {
+      "epoch": 3.1318327974276525,
+      "grad_norm": 0.347550013269568,
+      "learning_rate": 1.4613500361847703e-05,
+      "loss": 0.3286,
+      "step": 487
+    },
+    {
+      "epoch": 3.1382636655948555,
+      "grad_norm": 0.3595568827120924,
+      "learning_rate": 1.4526740198558345e-05,
+      "loss": 0.3415,
+      "step": 488
+    },
+    {
+      "epoch": 3.144694533762058,
+      "grad_norm": 0.3325887945606145,
+      "learning_rate": 1.4440091228798896e-05,
+      "loss": 0.3258,
+      "step": 489
+    },
+    {
+      "epoch": 3.1511254019292605,
+      "grad_norm": 0.3396889942519828,
+      "learning_rate": 1.4353555212910577e-05,
+      "loss": 0.3522,
+      "step": 490
+    },
+    {
+      "epoch": 3.157556270096463,
+      "grad_norm": 0.35664644793805467,
+      "learning_rate": 1.4267133908939842e-05,
+      "loss": 0.3421,
+      "step": 491
+    },
+    {
+      "epoch": 3.1639871382636655,
+      "grad_norm": 0.34001836863101187,
+      "learning_rate": 1.4180829072602715e-05,
+      "loss": 0.335,
+      "step": 492
+    },
+    {
+      "epoch": 3.170418006430868,
+      "grad_norm": 0.3593329066143372,
+      "learning_rate": 1.4094642457249063e-05,
+      "loss": 0.3659,
+      "step": 493
+    },
+    {
+      "epoch": 3.176848874598071,
+      "grad_norm": 0.33860210246347033,
+      "learning_rate": 1.4008575813827023e-05,
+      "loss": 0.3466,
+      "step": 494
+    },
+    {
+      "epoch": 3.1832797427652735,
+      "grad_norm": 0.3693876810382247,
+      "learning_rate": 1.3922630890847387e-05,
+      "loss": 0.3505,
+      "step": 495
+    },
+    {
+      "epoch": 3.189710610932476,
+      "grad_norm": 0.3175046542412462,
+      "learning_rate": 1.3836809434348108e-05,
+      "loss": 0.3455,
+      "step": 496
+    },
+    {
+      "epoch": 3.1961414790996785,
+      "grad_norm": 0.365133352460214,
+      "learning_rate": 1.375111318785884e-05,
+      "loss": 0.3183,
+      "step": 497
+    },
+    {
+      "epoch": 3.202572347266881,
+      "grad_norm": 0.3424344738744123,
+      "learning_rate": 1.3665543892365482e-05,
+      "loss": 0.3393,
+      "step": 498
+    },
+    {
+      "epoch": 3.2090032154340835,
+      "grad_norm": 0.33023930745143787,
+      "learning_rate": 1.3580103286274816e-05,
+      "loss": 0.3643,
+      "step": 499
+    },
+    {
+      "epoch": 3.215434083601286,
+      "grad_norm": 0.33185985883046265,
+      "learning_rate": 1.3494793105379217e-05,
+      "loss": 0.3222,
+      "step": 500
+    },
+    {
+      "epoch": 3.221864951768489,
+      "grad_norm": 0.32551560148888986,
+      "learning_rate": 1.340961508282136e-05,
+      "loss": 0.3288,
+      "step": 501
+    },
+    {
+      "epoch": 3.2282958199356915,
+      "grad_norm": 0.323589627002325,
+      "learning_rate": 1.3324570949059035e-05,
+      "loss": 0.3453,
+      "step": 502
+    },
+    {
+      "epoch": 3.234726688102894,
+      "grad_norm": 0.3106749867134289,
+      "learning_rate": 1.3239662431829949e-05,
+      "loss": 0.3716,
+      "step": 503
+    },
+    {
+      "epoch": 3.2411575562700965,
+      "grad_norm": 0.34549558845546635,
+      "learning_rate": 1.3154891256116677e-05,
+      "loss": 0.3422,
+      "step": 504
+    },
+    {
+      "epoch": 3.247588424437299,
+      "grad_norm": 0.3277979638472998,
+      "learning_rate": 1.3070259144111592e-05,
+      "loss": 0.3444,
+      "step": 505
+    },
+    {
+      "epoch": 3.2540192926045015,
+      "grad_norm": 0.35789879175453515,
+      "learning_rate": 1.2985767815181878e-05,
+      "loss": 0.3181,
+      "step": 506
+    },
+    {
+      "epoch": 3.260450160771704,
+      "grad_norm": 0.3321930319977574,
+      "learning_rate": 1.2901418985834582e-05,
+      "loss": 0.3298,
+      "step": 507
+    },
+    {
+      "epoch": 3.266881028938907,
+      "grad_norm": 0.3128707110842694,
+      "learning_rate": 1.2817214369681794e-05,
+      "loss": 0.342,
+      "step": 508
+    },
+    {
+      "epoch": 3.2733118971061095,
+      "grad_norm": 0.33348093643652205,
+      "learning_rate": 1.2733155677405776e-05,
+      "loss": 0.3587,
+      "step": 509
+    },
+    {
+      "epoch": 3.279742765273312,
+      "grad_norm": 0.339207053764794,
+      "learning_rate": 1.2649244616724253e-05,
+      "loss": 0.329,
+      "step": 510
+    },
+    {
+      "epoch": 3.2861736334405145,
+      "grad_norm": 0.3422170366956234,
+      "learning_rate": 1.2565482892355674e-05,
+      "loss": 0.3457,
+      "step": 511
+    },
+    {
+      "epoch": 3.292604501607717,
+      "grad_norm": 0.2943236021955186,
+      "learning_rate": 1.248187220598462e-05,
+      "loss": 0.3493,
+      "step": 512
+    },
+    {
+      "epoch": 3.2990353697749195,
+      "grad_norm": 0.3035388624219581,
+      "learning_rate": 1.2398414256227227e-05,
+      "loss": 0.3465,
+      "step": 513
+    },
+    {
+      "epoch": 3.305466237942122,
+      "grad_norm": 0.33471632658357653,
+      "learning_rate": 1.2315110738596654e-05,
+      "loss": 0.3263,
+      "step": 514
+    },
+    {
+      "epoch": 3.311897106109325,
+      "grad_norm": 0.29768881691936494,
+      "learning_rate": 1.223196334546864e-05,
+      "loss": 0.3383,
+      "step": 515
+    },
+    {
+      "epoch": 3.3183279742765275,
+      "grad_norm": 0.31373198806666036,
+      "learning_rate": 1.2148973766047164e-05,
+      "loss": 0.3151,
+      "step": 516
+    },
+    {
+      "epoch": 3.32475884244373,
+      "grad_norm": 0.3095596372386397,
+      "learning_rate": 1.2066143686330069e-05,
+      "loss": 0.3177,
+      "step": 517
+    },
+    {
+      "epoch": 3.3311897106109325,
+      "grad_norm": 0.32616752602823745,
+      "learning_rate": 1.1983474789074867e-05,
+      "loss": 0.3255,
+      "step": 518
+    },
+    {
+      "epoch": 3.337620578778135,
+      "grad_norm": 0.30771383148143083,
+      "learning_rate": 1.1900968753764483e-05,
+      "loss": 0.3159,
+      "step": 519
+    },
+    {
+      "epoch": 3.3440514469453375,
+      "grad_norm": 0.3147026157328475,
+      "learning_rate": 1.1818627256573203e-05,
+      "loss": 0.3417,
+      "step": 520
+    },
+    {
+      "epoch": 3.35048231511254,
+      "grad_norm": 0.31316932886086846,
+      "learning_rate": 1.1736451970332594e-05,
+      "loss": 0.3359,
+      "step": 521
+    },
+    {
+      "epoch": 3.356913183279743,
+      "grad_norm": 0.3036155487748744,
+      "learning_rate": 1.1654444564497507e-05,
+      "loss": 0.3328,
+      "step": 522
+    },
+    {
+      "epoch": 3.3633440514469455,
+      "grad_norm": 0.30544167608450157,
+      "learning_rate": 1.1572606705112177e-05,
+      "loss": 0.3188,
+      "step": 523
+    },
+    {
+      "epoch": 3.369774919614148,
+      "grad_norm": 0.3226490321653266,
+      "learning_rate": 1.1490940054776365e-05,
+      "loss": 0.3489,
+      "step": 524
+    },
+    {
+      "epoch": 3.3762057877813505,
+      "grad_norm": 0.28804525279504883,
+      "learning_rate": 1.1409446272611591e-05,
+      "loss": 0.337,
+      "step": 525
+    },
+    {
+      "epoch": 3.382636655948553,
+      "grad_norm": 0.30753505734977593,
+      "learning_rate": 1.132812701422744e-05,
+      "loss": 0.337,
+      "step": 526
+    },
+    {
+      "epoch": 3.3890675241157555,
+      "grad_norm": 0.31053703098114926,
+      "learning_rate": 1.1246983931687874e-05,
+      "loss": 0.3383,
+      "step": 527
+    },
+    {
+      "epoch": 3.395498392282958,
+      "grad_norm": 0.30272870232775245,
+      "learning_rate": 1.1166018673477748e-05,
+      "loss": 0.3328,
+      "step": 528
+    },
+    {
+      "epoch": 3.401929260450161,
+      "grad_norm": 0.2823989712823126,
+      "learning_rate": 1.1085232884469236e-05,
+      "loss": 0.351,
+      "step": 529
+    },
+    {
+      "epoch": 3.4083601286173635,
+      "grad_norm": 0.30846740972932163,
+      "learning_rate": 1.1004628205888496e-05,
+      "loss": 0.334,
+      "step": 530
+    },
+    {
+      "epoch": 3.414790996784566,
+      "grad_norm": 0.3140803547062033,
+      "learning_rate": 1.0924206275282248e-05,
+      "loss": 0.3493,
+      "step": 531
+    },
+    {
+      "epoch": 3.4212218649517685,
+      "grad_norm": 0.30828616597228864,
+      "learning_rate": 1.084396872648457e-05,
+      "loss": 0.3181,
+      "step": 532
+    },
+    {
+      "epoch": 3.427652733118971,
+      "grad_norm": 0.3205196440562544,
+      "learning_rate": 1.0763917189583685e-05,
+      "loss": 0.3488,
+      "step": 533
+    },
+    {
+      "epoch": 3.4340836012861735,
+      "grad_norm": 0.3269965140193297,
+      "learning_rate": 1.0684053290888824e-05,
+      "loss": 0.3241,
+      "step": 534
+    },
+    {
+      "epoch": 3.440514469453376,
+      "grad_norm": 0.3190697695545462,
+      "learning_rate": 1.060437865289719e-05,
+      "loss": 0.3303,
+      "step": 535
+    },
+    {
+      "epoch": 3.446945337620579,
+      "grad_norm": 0.30058923975882595,
+      "learning_rate": 1.0524894894261035e-05,
+      "loss": 0.3109,
+      "step": 536
+    },
+    {
+      "epoch": 3.4533762057877815,
+      "grad_norm": 0.30880846001588463,
+      "learning_rate": 1.044560362975474e-05,
+      "loss": 0.3187,
+      "step": 537
+    },
+    {
+      "epoch": 3.459807073954984,
+      "grad_norm": 0.3246456142532528,
+      "learning_rate": 1.0366506470242019e-05,
+      "loss": 0.336,
+      "step": 538
+    },
+    {
+      "epoch": 3.4662379421221865,
+      "grad_norm": 0.3055236949162367,
+      "learning_rate": 1.0287605022643183e-05,
+      "loss": 0.3243,
+      "step": 539
+    },
+    {
+      "epoch": 3.472668810289389,
+      "grad_norm": 0.3088688923442282,
+      "learning_rate": 1.0208900889902504e-05,
+      "loss": 0.3602,
+      "step": 540
+    },
+    {
+      "epoch": 3.4790996784565915,
+      "grad_norm": 0.29948721470294226,
+      "learning_rate": 1.0130395670955655e-05,
+      "loss": 0.323,
+      "step": 541
+    },
+    {
+      "epoch": 3.485530546623794,
+      "grad_norm": 0.3074858144654124,
+      "learning_rate": 1.0052090960697231e-05,
+      "loss": 0.3368,
+      "step": 542
+    },
+    {
+      "epoch": 3.491961414790997,
+      "grad_norm": 0.3146321880179152,
+      "learning_rate": 9.97398834994831e-06,
+      "loss": 0.2997,
+      "step": 543
+    },
+    {
+      "epoch": 3.4983922829581995,
+      "grad_norm": 0.3075533658637388,
+      "learning_rate": 9.896089425424188e-06,
+      "loss": 0.3431,
+      "step": 544
+    },
+    {
+      "epoch": 3.504823151125402,
+      "grad_norm": 0.30606333212239234,
+      "learning_rate": 9.818395769702099e-06,
+      "loss": 0.338,
+      "step": 545
+    },
+    {
+      "epoch": 3.5112540192926045,
+      "grad_norm": 0.3139502509780247,
+      "learning_rate": 9.740908961189089e-06,
+      "loss": 0.343,
+      "step": 546
+    },
+    {
+      "epoch": 3.517684887459807,
+      "grad_norm": 0.29754994978817495,
+      "learning_rate": 9.663630574089948e-06,
+      "loss": 0.3577,
+      "step": 547
+    },
+    {
+      "epoch": 3.5241157556270095,
+      "grad_norm": 0.29724371650818526,
+      "learning_rate": 9.586562178375206e-06,
+      "loss": 0.3469,
+      "step": 548
+    },
+    {
+      "epoch": 3.530546623794212,
+      "grad_norm": 0.3308822180362072,
+      "learning_rate": 9.509705339749277e-06,
+      "loss": 0.336,
+      "step": 549
+    },
+    {
+      "epoch": 3.536977491961415,
+      "grad_norm": 0.29272895208721433,
+      "learning_rate": 9.433061619618604e-06,
+      "loss": 0.3414,
+      "step": 550
+    },
+    {
+      "epoch": 3.5434083601286175,
+      "grad_norm": 0.320986465865753,
+      "learning_rate": 9.356632575059986e-06,
+      "loss": 0.3386,
+      "step": 551
+    },
+    {
+      "epoch": 3.54983922829582,
+      "grad_norm": 0.31008654153647475,
+      "learning_rate": 9.280419758788894e-06,
+      "loss": 0.3486,
+      "step": 552
+    },
+    {
+      "epoch": 3.5562700964630225,
+      "grad_norm": 0.2925897227549084,
+      "learning_rate": 9.204424719127976e-06,
+      "loss": 0.3613,
+      "step": 553
+    },
+    {
+      "epoch": 3.562700964630225,
+      "grad_norm": 0.343142790220052,
+      "learning_rate": 9.12864899997558e-06,
+      "loss": 0.3073,
+      "step": 554
+    },
+    {
+      "epoch": 3.5691318327974275,
+      "grad_norm": 0.3042440507540189,
+      "learning_rate": 9.053094140774367e-06,
+      "loss": 0.3567,
+      "step": 555
+    },
+    {
+      "epoch": 3.57556270096463,
+      "grad_norm": 0.3116852171680514,
+      "learning_rate": 8.977761676480065e-06,
+      "loss": 0.3269,
+      "step": 556
+    },
+    {
+      "epoch": 3.581993569131833,
+      "grad_norm": 0.3436371078553796,
+      "learning_rate": 8.902653137530283e-06,
+      "loss": 0.3276,
+      "step": 557
+    },
+    {
+      "epoch": 3.5884244372990355,
+      "grad_norm": 0.3475488081092682,
+      "learning_rate": 8.82777004981341e-06,
+      "loss": 0.3221,
+      "step": 558
+    },
+    {
+      "epoch": 3.594855305466238,
+      "grad_norm": 0.2953633447942229,
+      "learning_rate": 8.753113934637621e-06,
+      "loss": 0.3602,
+      "step": 559
+    },
+    {
+      "epoch": 3.6012861736334405,
+      "grad_norm": 0.2947407879791063,
+      "learning_rate": 8.678686308699957e-06,
+      "loss": 0.3445,
+      "step": 560
+    },
+    {
+      "epoch": 3.607717041800643,
+      "grad_norm": 0.33981328275420586,
+      "learning_rate": 8.60448868405552e-06,
+      "loss": 0.316,
+      "step": 561
+    },
+    {
+      "epoch": 3.6141479099678455,
+      "grad_norm": 0.33917263018781646,
+      "learning_rate": 8.530522568086777e-06,
+      "loss": 0.32,
+      "step": 562
+    },
+    {
+      "epoch": 3.620578778135048,
+      "grad_norm": 0.33849255042910925,
+      "learning_rate": 8.456789463472902e-06,
+      "loss": 0.3186,
+      "step": 563
+    },
+    {
+      "epoch": 3.627009646302251,
+      "grad_norm": 0.2969530188789648,
+      "learning_rate": 8.383290868159256e-06,
+      "loss": 0.341,
+      "step": 564
+    },
+    {
+      "epoch": 3.6334405144694535,
+      "grad_norm": 0.2972013956972423,
+      "learning_rate": 8.31002827532698e-06,
+      "loss": 0.3329,
+      "step": 565
+    },
+    {
+      "epoch": 3.639871382636656,
+      "grad_norm": 0.3035743650008469,
+      "learning_rate": 8.23700317336262e-06,
+      "loss": 0.3374,
+      "step": 566
+    },
+    {
+      "epoch": 3.6463022508038585,
+      "grad_norm": 0.3081348138393716,
+      "learning_rate": 8.164217045827934e-06,
+      "loss": 0.3458,
+      "step": 567
+    },
+    {
+      "epoch": 3.652733118971061,
+      "grad_norm": 0.2876537772969051,
+      "learning_rate": 8.091671371429697e-06,
+      "loss": 0.3405,
+      "step": 568
+    },
+    {
+      "epoch": 3.6591639871382635,
+      "grad_norm": 0.3313526167727788,
+      "learning_rate": 8.01936762398972e-06,
+      "loss": 0.3483,
+      "step": 569
+    },
+    {
+      "epoch": 3.665594855305466,
+      "grad_norm": 0.3112433837967559,
+      "learning_rate": 7.947307272414874e-06,
+      "loss": 0.3243,
+      "step": 570
+    },
+    {
+      "epoch": 3.672025723472669,
+      "grad_norm": 0.30108399349218257,
+      "learning_rate": 7.875491780667246e-06,
+      "loss": 0.3198,
+      "step": 571
+    },
+    {
+      "epoch": 3.6784565916398715,
+      "grad_norm": 0.2946041397166188,
+      "learning_rate": 7.803922607734398e-06,
+      "loss": 0.3226,
+      "step": 572
+    },
+    {
+      "epoch": 3.684887459807074,
+      "grad_norm": 0.29779909506256386,
+      "learning_rate": 7.732601207599753e-06,
+      "loss": 0.3269,
+      "step": 573
+    },
+    {
+      "epoch": 3.6913183279742765,
+      "grad_norm": 0.2931702862841286,
+      "learning_rate": 7.661529029213023e-06,
+      "loss": 0.3467,
+      "step": 574
+    },
+    {
+      "epoch": 3.697749196141479,
+      "grad_norm": 0.2996030618184555,
+      "learning_rate": 7.590707516460807e-06,
+      "loss": 0.3323,
+      "step": 575
+    },
+    {
+      "epoch": 3.7041800643086815,
+      "grad_norm": 0.2888522917219981,
+      "learning_rate": 7.520138108137205e-06,
+      "loss": 0.342,
+      "step": 576
+    },
+    {
+      "epoch": 3.710610932475884,
+      "grad_norm": 0.291296782783801,
+      "learning_rate": 7.44982223791463e-06,
+      "loss": 0.3607,
+      "step": 577
+    },
+    {
+      "epoch": 3.717041800643087,
+      "grad_norm": 0.30115471015972206,
+      "learning_rate": 7.379761334314688e-06,
+      "loss": 0.3448,
+      "step": 578
+    },
+    {
+      "epoch": 3.7234726688102895,
+      "grad_norm": 0.2830094904547036,
+      "learning_rate": 7.3099568206791315e-06,
+      "loss": 0.3288,
+      "step": 579
+    },
+    {
+      "epoch": 3.729903536977492,
+      "grad_norm": 0.2694638012556624,
+      "learning_rate": 7.240410115140934e-06,
+      "loss": 0.3466,
+      "step": 580
+    },
+    {
+      "epoch": 3.7363344051446945,
+      "grad_norm": 0.2925665957791904,
+      "learning_rate": 7.171122630595526e-06,
+      "loss": 0.3202,
+      "step": 581
+    },
+    {
+      "epoch": 3.742765273311897,
+      "grad_norm": 0.2935199603841542,
+      "learning_rate": 7.102095774672035e-06,
+      "loss": 0.3502,
+      "step": 582
+    },
+    {
+      "epoch": 3.7491961414790995,
+      "grad_norm": 0.2854219231454806,
+      "learning_rate": 7.033330949704738e-06,
+      "loss": 0.3203,
+      "step": 583
+    },
+    {
+      "epoch": 3.755627009646302,
+      "grad_norm": 0.29551028789541817,
+      "learning_rate": 6.964829552704526e-06,
+      "loss": 0.3465,
+      "step": 584
+    },
+    {
+      "epoch": 3.762057877813505,
+      "grad_norm": 0.2960060979737996,
+      "learning_rate": 6.89659297533057e-06,
+      "loss": 0.3431,
+      "step": 585
+    },
+    {
+      "epoch": 3.7684887459807075,
+      "grad_norm": 0.2857182383139957,
+      "learning_rate": 6.828622603862012e-06,
+      "loss": 0.3531,
+      "step": 586
+    },
+    {
+      "epoch": 3.77491961414791,
+      "grad_norm": 0.26784654599409924,
+      "learning_rate": 6.760919819169813e-06,
+      "loss": 0.3644,
+      "step": 587
+    },
+    {
+      "epoch": 3.7813504823151125,
+      "grad_norm": 0.312552743679182,
+      "learning_rate": 6.693485996688695e-06,
+      "loss": 0.3183,
+      "step": 588
+    },
+    {
+      "epoch": 3.787781350482315,
+      "grad_norm": 0.29363220312638616,
+      "learning_rate": 6.62632250638922e-06,
+      "loss": 0.3296,
+      "step": 589
+    },
+    {
+      "epoch": 3.7942122186495175,
+      "grad_norm": 0.27923945424032065,
+      "learning_rate": 6.559430712749928e-06,
+      "loss": 0.3485,
+      "step": 590
+    },
+    {
+      "epoch": 3.80064308681672,
+      "grad_norm": 0.31593530341666676,
+      "learning_rate": 6.492811974729645e-06,
+      "loss": 0.3105,
+      "step": 591
+    },
+    {
+      "epoch": 3.807073954983923,
+      "grad_norm": 0.30886963920639826,
+      "learning_rate": 6.42646764573984e-06,
+      "loss": 0.3473,
+      "step": 592
+    },
+    {
+      "epoch": 3.8135048231511255,
+      "grad_norm": 0.2756903619698076,
+      "learning_rate": 6.360399073617152e-06,
+      "loss": 0.3421,
+      "step": 593
+    },
+    {
+      "epoch": 3.819935691318328,
+      "grad_norm": 0.29662133437844923,
+      "learning_rate": 6.2946076005960184e-06,
+      "loss": 0.3257,
+      "step": 594
+    },
+    {
+      "epoch": 3.8263665594855305,
+      "grad_norm": 0.3125323281572811,
+      "learning_rate": 6.229094563281393e-06,
+      "loss": 0.3398,
+      "step": 595
+    },
+    {
+      "epoch": 3.832797427652733,
+      "grad_norm": 0.3004202491602741,
+      "learning_rate": 6.163861292621569e-06,
+      "loss": 0.3422,
+      "step": 596
+    },
+    {
+      "epoch": 3.839228295819936,
+      "grad_norm": 0.27365799620978254,
+      "learning_rate": 6.098909113881193e-06,
+      "loss": 0.3415,
+      "step": 597
+    },
+    {
+      "epoch": 3.845659163987138,
+      "grad_norm": 0.28390963671233893,
+      "learning_rate": 6.034239346614281e-06,
+      "loss": 0.3501,
+      "step": 598
+    },
+    {
+      "epoch": 3.852090032154341,
+      "grad_norm": 0.2957088044678507,
+      "learning_rate": 5.969853304637467e-06,
+      "loss": 0.347,
+      "step": 599
+    },
+    {
+      "epoch": 3.8585209003215435,
+      "grad_norm": 0.3037260413469084,
+      "learning_rate": 5.90575229600326e-06,
+      "loss": 0.3621,
+      "step": 600
+    },
+    {
+      "epoch": 3.864951768488746,
+      "grad_norm": 0.2846873294718638,
+      "learning_rate": 5.841937622973515e-06,
+      "loss": 0.3273,
+      "step": 601
+    },
+    {
+      "epoch": 3.8713826366559485,
+      "grad_norm": 0.2734379481889535,
+      "learning_rate": 5.778410581992953e-06,
+      "loss": 0.3232,
+      "step": 602
+    },
+    {
+      "epoch": 3.877813504823151,
+      "grad_norm": 0.30102950288912395,
+      "learning_rate": 5.715172463662817e-06,
+      "loss": 0.3482,
+      "step": 603
+    },
+    {
+      "epoch": 3.884244372990354,
+      "grad_norm": 0.2873762472608675,
+      "learning_rate": 5.65222455271466e-06,
+      "loss": 0.3357,
+      "step": 604
+    },
+    {
+      "epoch": 3.890675241157556,
+      "grad_norm": 0.26966499675618716,
+      "learning_rate": 5.5895681279842615e-06,
+      "loss": 0.3377,
+      "step": 605
+    },
+    {
+      "epoch": 3.897106109324759,
+      "grad_norm": 0.28727016570076136,
+      "learning_rate": 5.52720446238562e-06,
+      "loss": 0.3297,
+      "step": 606
+    },
+    {
+      "epoch": 3.9035369774919615,
+      "grad_norm": 0.32095880409841926,
+      "learning_rate": 5.465134822885119e-06,
+      "loss": 0.3125,
+      "step": 607
+    },
+    {
+      "epoch": 3.909967845659164,
+      "grad_norm": 0.2893275192389225,
+      "learning_rate": 5.403360470475751e-06,
+      "loss": 0.2965,
+      "step": 608
+    },
+    {
+      "epoch": 3.9163987138263665,
+      "grad_norm": 0.28970930860220095,
+      "learning_rate": 5.341882660151527e-06,
+      "loss": 0.336,
+      "step": 609
+    },
+    {
+      "epoch": 3.922829581993569,
+      "grad_norm": 0.2866602239587237,
+      "learning_rate": 5.280702640881992e-06,
+      "loss": 0.3316,
+      "step": 610
+    },
+    {
+      "epoch": 3.929260450160772,
+      "grad_norm": 0.30658938426758525,
+      "learning_rate": 5.219821655586821e-06,
+      "loss": 0.3225,
+      "step": 611
+    },
+    {
+      "epoch": 3.935691318327974,
+      "grad_norm": 0.2931321320943746,
+      "learning_rate": 5.1592409411105775e-06,
+      "loss": 0.3206,
+      "step": 612
+    },
+    {
+      "epoch": 3.942122186495177,
+      "grad_norm": 0.276755969499715,
+      "learning_rate": 5.0989617281976e-06,
+      "loss": 0.3289,
+      "step": 613
+    },
+    {
+      "epoch": 3.9485530546623795,
+      "grad_norm": 0.2902682534805332,
+      "learning_rate": 5.038985241466978e-06,
+      "loss": 0.3262,
+      "step": 614
+    },
+    {
+      "epoch": 3.954983922829582,
+      "grad_norm": 0.29928353455708795,
+      "learning_rate": 4.979312699387699e-06,
+      "loss": 0.3297,
+      "step": 615
+    },
+    {
+      "epoch": 3.9614147909967845,
+      "grad_norm": 0.31093114025786367,
+      "learning_rate": 4.919945314253853e-06,
+      "loss": 0.3294,
+      "step": 616
+    },
+    {
+      "epoch": 3.967845659163987,
+      "grad_norm": 0.2733038554998018,
+      "learning_rate": 4.860884292160053e-06,
+      "loss": 0.3427,
+      "step": 617
+    },
+    {
+      "epoch": 3.97427652733119,
+      "grad_norm": 0.2743950196139407,
+      "learning_rate": 4.802130832976905e-06,
+      "loss": 0.3387,
+      "step": 618
+    },
+    {
+      "epoch": 3.980707395498392,
+      "grad_norm": 0.2826846110367078,
+      "learning_rate": 4.7436861303266255e-06,
+      "loss": 0.337,
+      "step": 619
+    },
+    {
+      "epoch": 3.987138263665595,
+      "grad_norm": 0.2817137199406774,
+      "learning_rate": 4.685551371558798e-06,
+      "loss": 0.3495,
+      "step": 620
+    },
+    {
+      "epoch": 3.9935691318327975,
+      "grad_norm": 0.28988916677018695,
+      "learning_rate": 4.6277277377262705e-06,
+      "loss": 0.3278,
+      "step": 621
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.2851529864649889,
+      "learning_rate": 4.570216403561141e-06,
+      "loss": 0.3295,
+      "step": 622
+    },
+    {
+      "epoch": 4.006430868167203,
+      "grad_norm": 0.4567824735504892,
+      "learning_rate": 4.513018537450897e-06,
+      "loss": 0.2947,
+      "step": 623
+    },
+    {
+      "epoch": 4.012861736334405,
+      "grad_norm": 0.3606957980091955,
+      "learning_rate": 4.456135301414672e-06,
+      "loss": 0.2952,
+      "step": 624
+    },
+    {
+      "epoch": 4.019292604501608,
+      "grad_norm": 0.3024652773391222,
+      "learning_rate": 4.3995678510796445e-06,
+      "loss": 0.3121,
+      "step": 625
+    },
+    {
+      "epoch": 4.02572347266881,
+      "grad_norm": 0.38469556146443795,
+      "learning_rate": 4.343317335657571e-06,
+      "loss": 0.3027,
+      "step": 626
+    },
+    {
+      "epoch": 4.032154340836013,
+      "grad_norm": 0.39470411746264217,
+      "learning_rate": 4.287384897921423e-06,
+      "loss": 0.2738,
+      "step": 627
+    },
+    {
+      "epoch": 4.038585209003215,
+      "grad_norm": 0.3673224271597665,
+      "learning_rate": 4.231771674182181e-06,
+      "loss": 0.3134,
+      "step": 628
+    },
+    {
+      "epoch": 4.045016077170418,
+      "grad_norm": 0.3110256996947676,
+      "learning_rate": 4.176478794265737e-06,
+      "loss": 0.303,
+      "step": 629
+    },
+    {
+      "epoch": 4.051446945337621,
+      "grad_norm": 0.37379175483411864,
+      "learning_rate": 4.1215073814899466e-06,
+      "loss": 0.2687,
+      "step": 630
+    },
+    {
+      "epoch": 4.057877813504823,
+      "grad_norm": 0.40656578141463195,
+      "learning_rate": 4.0668585526418235e-06,
+      "loss": 0.3086,
+      "step": 631
+    },
+    {
+      "epoch": 4.064308681672026,
+      "grad_norm": 0.32829686580036266,
+      "learning_rate": 4.012533417954834e-06,
+      "loss": 0.2677,
+      "step": 632
+    },
+    {
+      "epoch": 4.070739549839228,
+      "grad_norm": 0.2688487866613182,
+      "learning_rate": 3.958533081086331e-06,
+      "loss": 0.2869,
+      "step": 633
+    },
+    {
+      "epoch": 4.077170418006431,
+      "grad_norm": 0.32157717722125595,
+      "learning_rate": 3.904858639095174e-06,
+      "loss": 0.2826,
+      "step": 634
+    },
+    {
+      "epoch": 4.083601286173633,
+      "grad_norm": 0.37975140184420975,
+      "learning_rate": 3.851511182419385e-06,
+      "loss": 0.2645,
+      "step": 635
+    },
+    {
+      "epoch": 4.090032154340836,
+      "grad_norm": 0.33545527980626505,
+      "learning_rate": 3.7984917948540513e-06,
+      "loss": 0.2751,
+      "step": 636
+    },
+    {
+      "epoch": 4.096463022508039,
+      "grad_norm": 0.28143153218858996,
+      "learning_rate": 3.7458015535292603e-06,
+      "loss": 0.294,
+      "step": 637
+    },
+    {
+      "epoch": 4.102893890675241,
+      "grad_norm": 0.28097041325075667,
+      "learning_rate": 3.6934415288882487e-06,
+      "loss": 0.2776,
+      "step": 638
+    },
+    {
+      "epoch": 4.109324758842444,
+      "grad_norm": 0.2981092899958691,
+      "learning_rate": 3.641412784665648e-06,
+      "loss": 0.3021,
+      "step": 639
+    },
+    {
+      "epoch": 4.115755627009646,
+      "grad_norm": 0.29274552045900554,
+      "learning_rate": 3.5897163778658618e-06,
+      "loss": 0.2877,
+      "step": 640
+    },
+    {
+      "epoch": 4.122186495176849,
+      "grad_norm": 0.2987623101962567,
+      "learning_rate": 3.5383533587415954e-06,
+      "loss": 0.2804,
+      "step": 641
+    },
+    {
+      "epoch": 4.128617363344051,
+      "grad_norm": 0.27807759429925244,
+      "learning_rate": 3.4873247707725376e-06,
+      "loss": 0.2786,
+      "step": 642
+    },
+    {
+      "epoch": 4.135048231511254,
+      "grad_norm": 0.27701583824467924,
+      "learning_rate": 3.436631650644142e-06,
+      "loss": 0.2471,
+      "step": 643
+    },
+    {
+      "epoch": 4.141479099678457,
+      "grad_norm": 0.29645381215257427,
+      "learning_rate": 3.3862750282265798e-06,
+      "loss": 0.2981,
+      "step": 644
+    },
+    {
+      "epoch": 4.147909967845659,
+      "grad_norm": 0.3068453953399427,
+      "learning_rate": 3.3362559265537954e-06,
+      "loss": 0.2729,
+      "step": 645
+    },
+    {
+      "epoch": 4.154340836012862,
+      "grad_norm": 0.2744620598678601,
+      "learning_rate": 3.2865753618027398e-06,
+      "loss": 0.2702,
+      "step": 646
+    },
+    {
+      "epoch": 4.160771704180064,
+      "grad_norm": 0.2595743484197355,
+      "learning_rate": 3.2372343432727305e-06,
+      "loss": 0.2743,
+      "step": 647
+    },
+    {
+      "epoch": 4.167202572347267,
+      "grad_norm": 0.2795784931667666,
+      "learning_rate": 3.188233873364941e-06,
+      "loss": 0.2582,
+      "step": 648
+    },
+    {
+      "epoch": 4.173633440514469,
+      "grad_norm": 0.2617565961209036,
+      "learning_rate": 3.1395749475620185e-06,
+      "loss": 0.2875,
+      "step": 649
+    },
+    {
+      "epoch": 4.180064308681672,
+      "grad_norm": 0.2562729205459389,
+      "learning_rate": 3.091258554407901e-06,
+      "loss": 0.2848,
+      "step": 650
+    },
+    {
+      "epoch": 4.186495176848875,
+      "grad_norm": 0.24403415508275006,
+      "learning_rate": 3.0432856754876817e-06,
+      "loss": 0.3006,
+      "step": 651
+    },
+    {
+      "epoch": 4.192926045016077,
+      "grad_norm": 0.2430123930712371,
+      "learning_rate": 2.9956572854077205e-06,
+      "loss": 0.2901,
+      "step": 652
+    },
+    {
+      "epoch": 4.19935691318328,
+      "grad_norm": 0.262800227366545,
+      "learning_rate": 2.9483743517757956e-06,
+      "loss": 0.3014,
+      "step": 653
+    },
+    {
+      "epoch": 4.205787781350482,
+      "grad_norm": 0.2645604346480339,
+      "learning_rate": 2.9014378351814866e-06,
+      "loss": 0.263,
+      "step": 654
+    },
+    {
+      "epoch": 4.212218649517685,
+      "grad_norm": 0.2508559029138612,
+      "learning_rate": 2.854848689176637e-06,
+      "loss": 0.2769,
+      "step": 655
+    },
+    {
+      "epoch": 4.218649517684887,
+      "grad_norm": 0.26283770420408925,
+      "learning_rate": 2.808607860255981e-06,
+      "loss": 0.2734,
+      "step": 656
+    },
+    {
+      "epoch": 4.22508038585209,
+      "grad_norm": 0.25341636376648646,
+      "learning_rate": 2.762716287837921e-06,
+      "loss": 0.2866,
+      "step": 657
+    },
+    {
+      "epoch": 4.231511254019293,
+      "grad_norm": 0.26634825360743036,
+      "learning_rate": 2.717174904245445e-06,
+      "loss": 0.3012,
+      "step": 658
+    },
+    {
+      "epoch": 4.237942122186495,
+      "grad_norm": 0.26332857755436223,
+      "learning_rate": 2.671984634687186e-06,
+      "loss": 0.2775,
+      "step": 659
+    },
+    {
+      "epoch": 4.244372990353698,
+      "grad_norm": 0.2618426361007995,
+      "learning_rate": 2.6271463972386226e-06,
+      "loss": 0.2894,
+      "step": 660
+    },
+    {
+      "epoch": 4.2508038585209,
+      "grad_norm": 0.2694111472281508,
+      "learning_rate": 2.582661102823416e-06,
+      "loss": 0.2691,
+      "step": 661
+    },
+    {
+      "epoch": 4.257234726688103,
+      "grad_norm": 0.24428222525548662,
+      "learning_rate": 2.5385296551949214e-06,
+      "loss": 0.2958,
+      "step": 662
+    },
+    {
+      "epoch": 4.263665594855305,
+      "grad_norm": 0.27619057174239925,
+      "learning_rate": 2.4947529509178246e-06,
+      "loss": 0.2557,
+      "step": 663
+    },
+    {
+      "epoch": 4.270096463022508,
+      "grad_norm": 0.24779423757073077,
+      "learning_rate": 2.4513318793499274e-06,
+      "loss": 0.2754,
+      "step": 664
+    },
+    {
+      "epoch": 4.276527331189711,
+      "grad_norm": 0.2621523733317381,
+      "learning_rate": 2.4082673226240627e-06,
+      "loss": 0.293,
+      "step": 665
+    },
+    {
+      "epoch": 4.282958199356913,
+      "grad_norm": 0.2622093588974721,
+      "learning_rate": 2.365560155630202e-06,
+      "loss": 0.2677,
+      "step": 666
+    },
+    {
+      "epoch": 4.289389067524116,
+      "grad_norm": 0.2538363206827879,
+      "learning_rate": 2.3232112459976562e-06,
+      "loss": 0.3004,
+      "step": 667
+    },
+    {
+      "epoch": 4.295819935691318,
+      "grad_norm": 0.2537936416161383,
+      "learning_rate": 2.2812214540774645e-06,
+      "loss": 0.3121,
+      "step": 668
+    },
+    {
+      "epoch": 4.302250803858521,
+      "grad_norm": 0.2517193245956825,
+      "learning_rate": 2.239591632924907e-06,
+      "loss": 0.3024,
+      "step": 669
+    },
+    {
+      "epoch": 4.308681672025724,
+      "grad_norm": 0.23949967787453436,
+      "learning_rate": 2.198322628282181e-06,
+      "loss": 0.303,
+      "step": 670
+    },
+    {
+      "epoch": 4.315112540192926,
+      "grad_norm": 0.25336761111259143,
+      "learning_rate": 2.157415278561219e-06,
+      "loss": 0.2857,
+      "step": 671
+    },
+    {
+      "epoch": 4.321543408360129,
+      "grad_norm": 0.24904597925421662,
+      "learning_rate": 2.1168704148266415e-06,
+      "loss": 0.2713,
+      "step": 672
+    },
+    {
+      "epoch": 4.327974276527331,
+      "grad_norm": 0.23674410584391498,
+      "learning_rate": 2.0766888607788906e-06,
+      "loss": 0.2981,
+      "step": 673
+    },
+    {
+      "epoch": 4.334405144694534,
+      "grad_norm": 0.2412361270314516,
+      "learning_rate": 2.0368714327374905e-06,
+      "loss": 0.2989,
+      "step": 674
+    },
+    {
+      "epoch": 4.340836012861736,
+      "grad_norm": 0.25505323003799557,
+      "learning_rate": 1.9974189396244627e-06,
+      "loss": 0.2577,
+      "step": 675
+    },
+    {
+      "epoch": 4.347266881028939,
+      "grad_norm": 0.2449893223118034,
+      "learning_rate": 1.958332182947895e-06,
+      "loss": 0.301,
+      "step": 676
+    },
+    {
+      "epoch": 4.353697749196142,
+      "grad_norm": 0.257635171130121,
+      "learning_rate": 1.9196119567856455e-06,
+      "loss": 0.2595,
+      "step": 677
+    },
+    {
+      "epoch": 4.360128617363344,
+      "grad_norm": 0.23978909063369228,
+      "learning_rate": 1.8812590477692216e-06,
+      "loss": 0.2851,
+      "step": 678
+    },
+    {
+      "epoch": 4.366559485530547,
+      "grad_norm": 0.2430936237725035,
+      "learning_rate": 1.8432742350678023e-06,
+      "loss": 0.2762,
+      "step": 679
+    },
+    {
+      "epoch": 4.372990353697749,
+      "grad_norm": 0.23170095440198504,
+      "learning_rate": 1.8056582903724095e-06,
+      "loss": 0.3076,
+      "step": 680
+    },
+    {
+      "epoch": 4.379421221864952,
+      "grad_norm": 0.24390117261389707,
+      "learning_rate": 1.7684119778802066e-06,
+      "loss": 0.2627,
+      "step": 681
+    },
+    {
+      "epoch": 4.385852090032154,
+      "grad_norm": 0.24094057316775563,
+      "learning_rate": 1.7315360542790105e-06,
+      "loss": 0.2882,
+      "step": 682
+    },
+    {
+      "epoch": 4.392282958199357,
+      "grad_norm": 0.2433821314933139,
+      "learning_rate": 1.6950312687318816e-06,
+      "loss": 0.2752,
+      "step": 683
+    },
+    {
+      "epoch": 4.39871382636656,
+      "grad_norm": 0.23204030155583652,
+      "learning_rate": 1.6588983628619404e-06,
+      "loss": 0.3032,
+      "step": 684
+    },
+    {
+      "epoch": 4.405144694533762,
+      "grad_norm": 0.25451718600324474,
+      "learning_rate": 1.6231380707372668e-06,
+      "loss": 0.2749,
+      "step": 685
+    },
+    {
+      "epoch": 4.411575562700965,
+      "grad_norm": 0.24304264510143034,
+      "learning_rate": 1.5877511188560092e-06,
+      "loss": 0.3008,
+      "step": 686
+    },
+    {
+      "epoch": 4.418006430868167,
+      "grad_norm": 0.24956199178775862,
+      "learning_rate": 1.5527382261316272e-06,
+      "loss": 0.287,
+      "step": 687
+    },
+    {
+      "epoch": 4.42443729903537,
+      "grad_norm": 0.2558887040245036,
+      "learning_rate": 1.5181001038782618e-06,
+      "loss": 0.2704,
+      "step": 688
+    },
+    {
+      "epoch": 4.430868167202572,
+      "grad_norm": 0.23908309045820575,
+      "learning_rate": 1.4838374557963064e-06,
+      "loss": 0.2919,
+      "step": 689
+    },
+    {
+      "epoch": 4.437299035369775,
+      "grad_norm": 0.2513577299370825,
+      "learning_rate": 1.4499509779581078e-06,
+      "loss": 0.2752,
+      "step": 690
+    },
+    {
+      "epoch": 4.443729903536978,
+      "grad_norm": 0.24135457656926776,
+      "learning_rate": 1.4164413587938209e-06,
+      "loss": 0.2851,
+      "step": 691
+    },
+    {
+      "epoch": 4.45016077170418,
+      "grad_norm": 0.27572598203582127,
+      "learning_rate": 1.383309279077425e-06,
+      "loss": 0.2623,
+      "step": 692
+    },
+    {
+      "epoch": 4.456591639871383,
+      "grad_norm": 0.24770086378208764,
+      "learning_rate": 1.3505554119128861e-06,
+      "loss": 0.2671,
+      "step": 693
+    },
+    {
+      "epoch": 4.463022508038585,
+      "grad_norm": 0.24577054798012407,
+      "learning_rate": 1.3181804227204897e-06,
+      "loss": 0.2886,
+      "step": 694
+    },
+    {
+      "epoch": 4.469453376205788,
+      "grad_norm": 0.24372173141685605,
+      "learning_rate": 1.286184969223323e-06,
+      "loss": 0.2817,
+      "step": 695
+    },
+    {
+      "epoch": 4.47588424437299,
+      "grad_norm": 0.24760820912266293,
+      "learning_rate": 1.2545697014339141e-06,
+      "loss": 0.303,
+      "step": 696
+    },
+    {
+      "epoch": 4.482315112540193,
+      "grad_norm": 0.24697453300232203,
+      "learning_rate": 1.2233352616410122e-06,
+      "loss": 0.2779,
+      "step": 697
+    },
+    {
+      "epoch": 4.488745980707396,
+      "grad_norm": 0.24359394116807057,
+      "learning_rate": 1.192482284396561e-06,
+      "loss": 0.2886,
+      "step": 698
+    },
+    {
+      "epoch": 4.495176848874598,
+      "grad_norm": 0.25125045167829413,
+      "learning_rate": 1.1620113965027823e-06,
+      "loss": 0.2642,
+      "step": 699
+    },
+    {
+      "epoch": 4.501607717041801,
+      "grad_norm": 0.25718449388891956,
+      "learning_rate": 1.131923216999471e-06,
+      "loss": 0.2848,
+      "step": 700
+    },
+    {
+      "epoch": 4.508038585209003,
+      "grad_norm": 0.24243292389290208,
+      "learning_rate": 1.1022183571513944e-06,
+      "loss": 0.2791,
+      "step": 701
+    },
+    {
+      "epoch": 4.514469453376206,
+      "grad_norm": 0.24284338974991793,
+      "learning_rate": 1.0728974204358833e-06,
+      "loss": 0.2955,
+      "step": 702
+    },
+    {
+      "epoch": 4.520900321543408,
+      "grad_norm": 0.24769230191148933,
+      "learning_rate": 1.0439610025305735e-06,
+      "loss": 0.2924,
+      "step": 703
+    },
+    {
+      "epoch": 4.527331189710611,
+      "grad_norm": 0.2528125675058168,
+      "learning_rate": 1.0154096913012989e-06,
+      "loss": 0.2548,
+      "step": 704
+    },
+    {
+      "epoch": 4.533762057877814,
+      "grad_norm": 0.24277535112629353,
+      "learning_rate": 9.872440667901562e-07,
+      "loss": 0.2757,
+      "step": 705
+    },
+    {
+      "epoch": 4.540192926045016,
+      "grad_norm": 0.24620559766212627,
+      "learning_rate": 9.59464701203714e-07,
+      "loss": 0.2705,
+      "step": 706
+    },
+    {
+      "epoch": 4.546623794212219,
+      "grad_norm": 0.24633134617014393,
+      "learning_rate": 9.320721589013892e-07,
+      "loss": 0.2566,
+      "step": 707
+    },
+    {
+      "epoch": 4.553054662379421,
+      "grad_norm": 0.25062242066380686,
+      "learning_rate": 9.050669963839875e-07,
+      "loss": 0.2794,
+      "step": 708
+    },
+    {
+      "epoch": 4.559485530546624,
+      "grad_norm": 0.2540272108727865,
+      "learning_rate": 8.784497622823874e-07,
+      "loss": 0.2603,
+      "step": 709
+    },
+    {
+      "epoch": 4.565916398713826,
+      "grad_norm": 0.24314763337396839,
+      "learning_rate": 8.522209973464024e-07,
+      "loss": 0.304,
+      "step": 710
+    },
+    {
+      "epoch": 4.572347266881029,
+      "grad_norm": 0.23138009881362143,
+      "learning_rate": 8.263812344337907e-07,
+      "loss": 0.3007,
+      "step": 711
+    },
+    {
+      "epoch": 4.578778135048232,
+      "grad_norm": 0.24573233235679195,
+      "learning_rate": 8.009309984994363e-07,
+      "loss": 0.273,
+      "step": 712
+    },
+    {
+      "epoch": 4.585209003215434,
+      "grad_norm": 0.24642734947253456,
+      "learning_rate": 7.758708065846799e-07,
+      "loss": 0.2772,
+      "step": 713
+    },
+    {
+      "epoch": 4.591639871382637,
+      "grad_norm": 0.23747548119288367,
+      "learning_rate": 7.512011678068077e-07,
+      "loss": 0.2907,
+      "step": 714
+    },
+    {
+      "epoch": 4.598070739549839,
+      "grad_norm": 0.24266659100239624,
+      "learning_rate": 7.269225833487193e-07,
+      "loss": 0.3144,
+      "step": 715
+    },
+    {
+      "epoch": 4.604501607717042,
+      "grad_norm": 0.25290531585087117,
+      "learning_rate": 7.030355464487404e-07,
+      "loss": 0.2798,
+      "step": 716
+    },
+    {
+      "epoch": 4.610932475884244,
+      "grad_norm": 0.24362587141597356,
+      "learning_rate": 6.79540542390611e-07,
+      "loss": 0.2866,
+      "step": 717
+    },
+    {
+      "epoch": 4.617363344051447,
+      "grad_norm": 0.2491192489963348,
+      "learning_rate": 6.564380484936062e-07,
+      "loss": 0.2663,
+      "step": 718
+    },
+    {
+      "epoch": 4.62379421221865,
+      "grad_norm": 0.24506961221576087,
+      "learning_rate": 6.33728534102862e-07,
+      "loss": 0.2954,
+      "step": 719
+    },
+    {
+      "epoch": 4.630225080385852,
+      "grad_norm": 0.2499590865403639,
+      "learning_rate": 6.114124605798232e-07,
+      "loss": 0.2727,
+      "step": 720
+    },
+    {
+      "epoch": 4.636655948553055,
+      "grad_norm": 0.23646801239979942,
+      "learning_rate": 5.894902812928838e-07,
+      "loss": 0.3034,
+      "step": 721
+    },
+    {
+      "epoch": 4.643086816720257,
+      "grad_norm": 0.24262823296382405,
+      "learning_rate": 5.679624416081652e-07,
+      "loss": 0.2449,
+      "step": 722
+    },
+    {
+      "epoch": 4.64951768488746,
+      "grad_norm": 0.23918779281810207,
+      "learning_rate": 5.468293788804735e-07,
+      "loss": 0.2826,
+      "step": 723
+    },
+    {
+      "epoch": 4.655948553054662,
+      "grad_norm": 0.24592289642811546,
+      "learning_rate": 5.260915224444207e-07,
+      "loss": 0.283,
+      "step": 724
+    },
+    {
+      "epoch": 4.662379421221865,
+      "grad_norm": 0.22658994888094983,
+      "learning_rate": 5.05749293605684e-07,
+      "loss": 0.3025,
+      "step": 725
+    },
+    {
+      "epoch": 4.668810289389068,
+      "grad_norm": 0.2295461621009824,
+      "learning_rate": 4.858031056324652e-07,
+      "loss": 0.2976,
+      "step": 726
+    },
+    {
+      "epoch": 4.67524115755627,
+      "grad_norm": 0.25175092403832117,
+      "learning_rate": 4.6625336374708894e-07,
+      "loss": 0.2594,
+      "step": 727
+    },
+    {
+      "epoch": 4.681672025723473,
+      "grad_norm": 0.2483079833834222,
+      "learning_rate": 4.471004651177646e-07,
+      "loss": 0.2675,
+      "step": 728
+    },
+    {
+      "epoch": 4.688102893890675,
+      "grad_norm": 0.23873921452119407,
+      "learning_rate": 4.2834479885052846e-07,
+      "loss": 0.3021,
+      "step": 729
+    },
+    {
+      "epoch": 4.694533762057878,
+      "grad_norm": 0.23933768222340976,
+      "learning_rate": 4.099867459813256e-07,
+      "loss": 0.2879,
+      "step": 730
+    },
+    {
+      "epoch": 4.70096463022508,
+      "grad_norm": 0.23726681093169322,
+      "learning_rate": 3.9202667946828033e-07,
+      "loss": 0.2828,
+      "step": 731
+    },
+    {
+      "epoch": 4.707395498392283,
+      "grad_norm": 0.23347685236455923,
+      "learning_rate": 3.744649641841114e-07,
+      "loss": 0.2818,
+      "step": 732
+    },
+    {
+      "epoch": 4.713826366559486,
+      "grad_norm": 0.24563695951256362,
+      "learning_rate": 3.573019569087266e-07,
+      "loss": 0.2888,
+      "step": 733
+    },
+    {
+      "epoch": 4.720257234726688,
+      "grad_norm": 0.2438786827488778,
+      "learning_rate": 3.4053800632196434e-07,
+      "loss": 0.2677,
+      "step": 734
+    },
+    {
+      "epoch": 4.726688102893891,
+      "grad_norm": 0.25416910133661985,
+      "learning_rate": 3.2417345299652346e-07,
+      "loss": 0.2785,
+      "step": 735
+    },
+    {
+      "epoch": 4.733118971061093,
+      "grad_norm": 0.24625279625369642,
+      "learning_rate": 3.0820862939103137e-07,
+      "loss": 0.2882,
+      "step": 736
+    },
+    {
+      "epoch": 4.739549839228296,
+      "grad_norm": 0.24374020827153117,
+      "learning_rate": 2.926438598432957e-07,
+      "loss": 0.2678,
+      "step": 737
+    },
+    {
+      "epoch": 4.745980707395498,
+      "grad_norm": 0.24087511582687704,
+      "learning_rate": 2.7747946056371654e-07,
+      "loss": 0.2801,
+      "step": 738
+    },
+    {
+      "epoch": 4.752411575562701,
+      "grad_norm": 0.23425542846338973,
+      "learning_rate": 2.627157396288604e-07,
+      "loss": 0.2977,
+      "step": 739
+    },
+    {
+      "epoch": 4.758842443729904,
+      "grad_norm": 0.24597974749704576,
+      "learning_rate": 2.48352996975203e-07,
+      "loss": 0.2569,
+      "step": 740
+    },
+    {
+      "epoch": 4.765273311897106,
+      "grad_norm": 0.2411396540589386,
+      "learning_rate": 2.343915243930317e-07,
+      "loss": 0.2999,
+      "step": 741
+    },
+    {
+      "epoch": 4.771704180064309,
+      "grad_norm": 0.24435836074048942,
+      "learning_rate": 2.208316055205173e-07,
+      "loss": 0.2772,
+      "step": 742
+    },
+    {
+      "epoch": 4.778135048231511,
+      "grad_norm": 0.25221784434928296,
+      "learning_rate": 2.076735158379606e-07,
+      "loss": 0.2711,
+      "step": 743
+    },
+    {
+      "epoch": 4.784565916398714,
+      "grad_norm": 0.22748509132409628,
+      "learning_rate": 1.94917522662188e-07,
+      "loss": 0.2913,
+      "step": 744
+    },
+    {
+      "epoch": 4.790996784565916,
+      "grad_norm": 0.2439276165012799,
+      "learning_rate": 1.8256388514112487e-07,
+      "loss": 0.2511,
+      "step": 745
+    },
+    {
+      "epoch": 4.797427652733119,
+      "grad_norm": 0.2346306695745195,
+      "learning_rate": 1.7061285424852415e-07,
+      "loss": 0.2869,
+      "step": 746
+    },
+    {
+      "epoch": 4.803858520900322,
+      "grad_norm": 0.25042053288491783,
+      "learning_rate": 1.590646727788725e-07,
+      "loss": 0.264,
+      "step": 747
+    },
+    {
+      "epoch": 4.810289389067524,
+      "grad_norm": 0.22725635488603368,
+      "learning_rate": 1.4791957534245893e-07,
+      "loss": 0.2826,
+      "step": 748
+    },
+    {
+      "epoch": 4.816720257234727,
+      "grad_norm": 0.2322822349163901,
+      "learning_rate": 1.3717778836060735e-07,
+      "loss": 0.2583,
+      "step": 749
+    },
+    {
+      "epoch": 4.823151125401929,
+      "grad_norm": 0.23097851234408598,
+      "learning_rate": 1.2683953006107362e-07,
+      "loss": 0.3046,
+      "step": 750
+    },
+    {
+      "epoch": 4.829581993569132,
+      "grad_norm": 0.24742405287216132,
+      "learning_rate": 1.1690501047361802e-07,
+      "loss": 0.2915,
+      "step": 751
+    },
+    {
+      "epoch": 4.836012861736334,
+      "grad_norm": 0.2507464084690092,
+      "learning_rate": 1.0737443142573301e-07,
+      "loss": 0.2951,
+      "step": 752
+    },
+    {
+      "epoch": 4.842443729903537,
+      "grad_norm": 0.23157728587496582,
+      "learning_rate": 9.824798653854217e-08,
+      "loss": 0.308,
+      "step": 753
+    },
+    {
+      "epoch": 4.84887459807074,
+      "grad_norm": 0.24916843135288552,
+      "learning_rate": 8.952586122287443e-08,
+      "loss": 0.2686,
+      "step": 754
+    },
+    {
+      "epoch": 4.855305466237942,
+      "grad_norm": 0.23859222294802174,
+      "learning_rate": 8.120823267548705e-08,
+      "loss": 0.2919,
+      "step": 755
+    },
+    {
+      "epoch": 4.861736334405145,
+      "grad_norm": 0.23943142091197492,
+      "learning_rate": 7.329526987547742e-08,
+      "loss": 0.2685,
+      "step": 756
+    },
+    {
+      "epoch": 4.868167202572347,
+      "grad_norm": 0.24029342217504873,
+      "learning_rate": 6.578713358083466e-08,
+      "loss": 0.3088,
+      "step": 757
+    },
+    {
+      "epoch": 4.87459807073955,
+      "grad_norm": 0.24524980488168965,
+      "learning_rate": 5.8683976325191185e-08,
+      "loss": 0.2694,
+      "step": 758
+    },
+    {
+      "epoch": 4.881028938906752,
+      "grad_norm": 0.24853681075043993,
+      "learning_rate": 5.1985942414709556e-08,
+      "loss": 0.2744,
+      "step": 759
+    },
+    {
+      "epoch": 4.887459807073955,
+      "grad_norm": 0.23789283333331324,
+      "learning_rate": 4.569316792515599e-08,
+      "loss": 0.2934,
+      "step": 760
+    },
+    {
+      "epoch": 4.893890675241158,
+      "grad_norm": 0.23474344825282825,
+      "learning_rate": 3.9805780699144755e-08,
+      "loss": 0.2791,
+      "step": 761
+    },
+    {
+      "epoch": 4.90032154340836,
+      "grad_norm": 0.2376604880712761,
+      "learning_rate": 3.432390034352251e-08,
+      "loss": 0.2813,
+      "step": 762
+    },
+    {
+      "epoch": 4.906752411575563,
+      "grad_norm": 0.24156511566543826,
+      "learning_rate": 2.9247638226950205e-08,
+      "loss": 0.2731,
+      "step": 763
+    },
+    {
+      "epoch": 4.913183279742765,
+      "grad_norm": 0.24199298942709446,
+      "learning_rate": 2.4577097477647137e-08,
+      "loss": 0.2512,
+      "step": 764
+    },
+    {
+      "epoch": 4.919614147909968,
+      "grad_norm": 0.2512888397179236,
+      "learning_rate": 2.0312372981274864e-08,
+      "loss": 0.2872,
+      "step": 765
+    },
+    {
+      "epoch": 4.92604501607717,
+      "grad_norm": 0.24930263130547045,
+      "learning_rate": 1.645355137903426e-08,
+      "loss": 0.2864,
+      "step": 766
+    },
+    {
+      "epoch": 4.932475884244373,
+      "grad_norm": 0.24008784931651084,
+      "learning_rate": 1.3000711065882521e-08,
+      "loss": 0.2821,
+      "step": 767
+    },
+    {
+      "epoch": 4.938906752411576,
+      "grad_norm": 0.2490423786655241,
+      "learning_rate": 9.953922188958854e-09,
+      "loss": 0.2878,
+      "step": 768
+    },
+    {
+      "epoch": 4.945337620578778,
+      "grad_norm": 0.24851457857336262,
+      "learning_rate": 7.313246646150074e-09,
+      "loss": 0.2652,
+      "step": 769
+    },
+    {
+      "epoch": 4.951768488745981,
+      "grad_norm": 0.23945971674195224,
+      "learning_rate": 5.078738084831614e-09,
+      "loss": 0.2943,
+      "step": 770
+    },
+    {
+      "epoch": 4.958199356913183,
+      "grad_norm": 0.2387923205453703,
+      "learning_rate": 3.2504419007839406e-09,
+      "loss": 0.2773,
+      "step": 771
+    },
+    {
+      "epoch": 4.964630225080386,
+      "grad_norm": 0.2380336931503509,
+      "learning_rate": 1.8283952372666337e-09,
+      "loss": 0.2796,
+      "step": 772
+    },
+    {
+      "epoch": 4.971061093247588,
+      "grad_norm": 0.24152308156450342,
+      "learning_rate": 8.126269842612111e-10,
+      "loss": 0.266,
+      "step": 773
+    },
+    {
+      "epoch": 4.977491961414791,
+      "grad_norm": 0.23877498920387313,
+      "learning_rate": 2.0315777789159387e-10,
+      "loss": 0.2778,
+      "step": 774
+    },
+    {
+      "epoch": 4.983922829581994,
+      "grad_norm": 0.2352901333228021,
+      "learning_rate": 0.0,
+      "loss": 0.2972,
+      "step": 775
+    },
+    {
+      "epoch": 4.983922829581994,
+      "step": 775,
+      "total_flos": 2.321081242868515e+18,
+      "train_loss": 0.44090195780800234,
+      "train_runtime": 33916.0911,
+      "train_samples_per_second": 2.934,
+      "train_steps_per_second": 0.023
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 775,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.321081242868515e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}