diff --git "a/checkpoint-2000/trainer_state.json" "b/checkpoint-2000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2000/trainer_state.json"
@@ -0,0 +1,14033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.4153686396677051,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00020768431983385254,
+      "grad_norm": 17.47925567626953,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.1801,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004153686396677051,
+      "grad_norm": 21.894914627075195,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 2.249,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006230529595015577,
+      "grad_norm": 25.090269088745117,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 2.3123,
+      "step": 3
+    },
+    {
+      "epoch": 0.0008307372793354102,
+      "grad_norm": 18.762964248657227,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 2.1709,
+      "step": 4
+    },
+    {
+      "epoch": 0.0010384215991692627,
+      "grad_norm": 18.856515884399414,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.1823,
+      "step": 5
+    },
+    {
+      "epoch": 0.0012461059190031153,
+      "grad_norm": 12.56984806060791,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 2.0798,
+      "step": 6
+    },
+    {
+      "epoch": 0.0014537902388369677,
+      "grad_norm": 8.624156951904297,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.9833,
+      "step": 7
+    },
+    {
+      "epoch": 0.0016614745586708203,
+      "grad_norm": 7.026400089263916,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 2.0748,
+      "step": 8
+    },
+    {
+      "epoch": 0.001869158878504673,
+      "grad_norm": 4.460483074188232,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.9404,
+      "step": 9
+    },
+    {
+      "epoch": 0.0020768431983385254,
+      "grad_norm": 3.6536319255828857,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.0135,
+      "step": 10
+    },
+    {
+      "epoch": 0.002284527518172378,
+      "grad_norm": 2.5768752098083496,
+      "learning_rate": 8.8e-06,
+      "loss": 1.8598,
+      "step": 11
+    },
+    {
+      "epoch": 0.0024922118380062306,
+      "grad_norm": 2.892141103744507,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.8686,
+      "step": 12
+    },
+    {
+      "epoch": 0.0026998961578400833,
+      "grad_norm": 2.858818531036377,
+      "learning_rate": 1.04e-05,
+      "loss": 1.8634,
+      "step": 13
+    },
+    {
+      "epoch": 0.0029075804776739354,
+      "grad_norm": 2.683091163635254,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 2.0208,
+      "step": 14
+    },
+    {
+      "epoch": 0.003115264797507788,
+      "grad_norm": 2.7056639194488525,
+      "learning_rate": 1.2e-05,
+      "loss": 1.9497,
+      "step": 15
+    },
+    {
+      "epoch": 0.0033229491173416407,
+      "grad_norm": 1.5998775959014893,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.8816,
+      "step": 16
+    },
+    {
+      "epoch": 0.0035306334371754933,
+      "grad_norm": 1.1934099197387695,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.8246,
+      "step": 17
+    },
+    {
+      "epoch": 0.003738317757009346,
+      "grad_norm": 1.1701968908309937,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 1.8578,
+      "step": 18
+    },
+    {
+      "epoch": 0.003946002076843198,
+      "grad_norm": 1.1254523992538452,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 1.8861,
+      "step": 19
+    },
+    {
+      "epoch": 0.004153686396677051,
+      "grad_norm": 0.8316693305969238,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.7194,
+      "step": 20
+    },
+    {
+      "epoch": 0.004361370716510903,
+      "grad_norm": 0.8501921892166138,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.8474,
+      "step": 21
+    },
+    {
+      "epoch": 0.004569055036344756,
+      "grad_norm": 0.9284434914588928,
+      "learning_rate": 1.76e-05,
+      "loss": 1.9258,
+      "step": 22
+    },
+    {
+      "epoch": 0.004776739356178609,
+      "grad_norm": 0.9898029565811157,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 1.8441,
+      "step": 23
+    },
+    {
+      "epoch": 0.004984423676012461,
+      "grad_norm": 0.8852891325950623,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.8521,
+      "step": 24
+    },
+    {
+      "epoch": 0.005192107995846314,
+      "grad_norm": 0.9528728723526001,
+      "learning_rate": 2e-05,
+      "loss": 1.8485,
+      "step": 25
+    },
+    {
+      "epoch": 0.0053997923156801665,
+      "grad_norm": 0.928016185760498,
+      "learning_rate": 2.08e-05,
+      "loss": 1.8883,
+      "step": 26
+    },
+    {
+      "epoch": 0.005607476635514018,
+      "grad_norm": 0.9373347163200378,
+      "learning_rate": 2.1600000000000003e-05,
+      "loss": 1.891,
+      "step": 27
+    },
+    {
+      "epoch": 0.005815160955347871,
+      "grad_norm": 0.9534879326820374,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 1.8144,
+      "step": 28
+    },
+    {
+      "epoch": 0.0060228452751817235,
+      "grad_norm": 0.7779879570007324,
+      "learning_rate": 2.32e-05,
+      "loss": 1.8153,
+      "step": 29
+    },
+    {
+      "epoch": 0.006230529595015576,
+      "grad_norm": 0.8465295433998108,
+      "learning_rate": 2.4e-05,
+      "loss": 1.794,
+      "step": 30
+    },
+    {
+      "epoch": 0.006438213914849429,
+      "grad_norm": 0.797340452671051,
+      "learning_rate": 2.4800000000000003e-05,
+      "loss": 1.8324,
+      "step": 31
+    },
+    {
+      "epoch": 0.006645898234683281,
+      "grad_norm": 0.7843337655067444,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 1.729,
+      "step": 32
+    },
+    {
+      "epoch": 0.006853582554517134,
+      "grad_norm": 0.7067154049873352,
+      "learning_rate": 2.6400000000000005e-05,
+      "loss": 1.7337,
+      "step": 33
+    },
+    {
+      "epoch": 0.007061266874350987,
+      "grad_norm": 0.7495686411857605,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 1.8074,
+      "step": 34
+    },
+    {
+      "epoch": 0.007268951194184839,
+      "grad_norm": 0.7126153707504272,
+      "learning_rate": 2.8e-05,
+      "loss": 1.7377,
+      "step": 35
+    },
+    {
+      "epoch": 0.007476635514018692,
+      "grad_norm": 0.7612162828445435,
+      "learning_rate": 2.8800000000000002e-05,
+      "loss": 1.775,
+      "step": 36
+    },
+    {
+      "epoch": 0.0076843198338525445,
+      "grad_norm": 0.7538896799087524,
+      "learning_rate": 2.96e-05,
+      "loss": 1.7478,
+      "step": 37
+    },
+    {
+      "epoch": 0.007892004153686396,
+      "grad_norm": 0.8245300650596619,
+      "learning_rate": 3.0400000000000004e-05,
+      "loss": 1.769,
+      "step": 38
+    },
+    {
+      "epoch": 0.00809968847352025,
+      "grad_norm": 0.7681639790534973,
+      "learning_rate": 3.1200000000000006e-05,
+      "loss": 1.7789,
+      "step": 39
+    },
+    {
+      "epoch": 0.008307372793354102,
+      "grad_norm": 0.7439791560173035,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.8245,
+      "step": 40
+    },
+    {
+      "epoch": 0.008515057113187955,
+      "grad_norm": 0.7001054286956787,
+      "learning_rate": 3.28e-05,
+      "loss": 1.8516,
+      "step": 41
+    },
+    {
+      "epoch": 0.008722741433021807,
+      "grad_norm": 0.7122063040733337,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 1.7698,
+      "step": 42
+    },
+    {
+      "epoch": 0.008930425752855659,
+      "grad_norm": 0.6748693585395813,
+      "learning_rate": 3.44e-05,
+      "loss": 1.7525,
+      "step": 43
+    },
+    {
+      "epoch": 0.009138110072689512,
+      "grad_norm": 0.7190465927124023,
+      "learning_rate": 3.52e-05,
+      "loss": 1.7725,
+      "step": 44
+    },
+    {
+      "epoch": 0.009345794392523364,
+      "grad_norm": 0.6779818534851074,
+      "learning_rate": 3.6e-05,
+      "loss": 1.7658,
+      "step": 45
+    },
+    {
+      "epoch": 0.009553478712357217,
+      "grad_norm": 0.6692732572555542,
+      "learning_rate": 3.680000000000001e-05,
+      "loss": 1.698,
+      "step": 46
+    },
+    {
+      "epoch": 0.009761163032191069,
+      "grad_norm": 0.6866817474365234,
+      "learning_rate": 3.76e-05,
+      "loss": 1.8105,
+      "step": 47
+    },
+    {
+      "epoch": 0.009968847352024923,
+      "grad_norm": 0.6470504999160767,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 1.6455,
+      "step": 48
+    },
+    {
+      "epoch": 0.010176531671858774,
+      "grad_norm": 0.681182324886322,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 1.6234,
+      "step": 49
+    },
+    {
+      "epoch": 0.010384215991692628,
+      "grad_norm": 0.6945406198501587,
+      "learning_rate": 4e-05,
+      "loss": 1.8115,
+      "step": 50
+    },
+    {
+      "epoch": 0.01059190031152648,
+      "grad_norm": 0.635296642780304,
+      "learning_rate": 4.08e-05,
+      "loss": 1.752,
+      "step": 51
+    },
+    {
+      "epoch": 0.010799584631360333,
+      "grad_norm": 0.6995029449462891,
+      "learning_rate": 4.16e-05,
+      "loss": 1.7869,
+      "step": 52
+    },
+    {
+      "epoch": 0.011007268951194185,
+      "grad_norm": 0.6692876815795898,
+      "learning_rate": 4.240000000000001e-05,
+      "loss": 1.7784,
+      "step": 53
+    },
+    {
+      "epoch": 0.011214953271028037,
+      "grad_norm": 0.6663792133331299,
+      "learning_rate": 4.3200000000000007e-05,
+      "loss": 1.7742,
+      "step": 54
+    },
+    {
+      "epoch": 0.01142263759086189,
+      "grad_norm": 0.6469627022743225,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.7418,
+      "step": 55
+    },
+    {
+      "epoch": 0.011630321910695742,
+      "grad_norm": 0.6327781677246094,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 1.7479,
+      "step": 56
+    },
+    {
+      "epoch": 0.011838006230529595,
+      "grad_norm": 0.6531626582145691,
+      "learning_rate": 4.56e-05,
+      "loss": 1.7121,
+      "step": 57
+    },
+    {
+      "epoch": 0.012045690550363447,
+      "grad_norm": 0.6819879412651062,
+      "learning_rate": 4.64e-05,
+      "loss": 1.8296,
+      "step": 58
+    },
+    {
+      "epoch": 0.0122533748701973,
+      "grad_norm": 0.6116477847099304,
+      "learning_rate": 4.72e-05,
+      "loss": 1.6922,
+      "step": 59
+    },
+    {
+      "epoch": 0.012461059190031152,
+      "grad_norm": 0.6569270491600037,
+      "learning_rate": 4.8e-05,
+      "loss": 1.7661,
+      "step": 60
+    },
+    {
+      "epoch": 0.012668743509865006,
+      "grad_norm": 0.6796997785568237,
+      "learning_rate": 4.88e-05,
+      "loss": 1.7959,
+      "step": 61
+    },
+    {
+      "epoch": 0.012876427829698858,
+      "grad_norm": 0.7139971256256104,
+      "learning_rate": 4.9600000000000006e-05,
+      "loss": 1.7764,
+      "step": 62
+    },
+    {
+      "epoch": 0.013084112149532711,
+      "grad_norm": 0.7374599575996399,
+      "learning_rate": 5.0400000000000005e-05,
+      "loss": 1.7271,
+      "step": 63
+    },
+    {
+      "epoch": 0.013291796469366563,
+      "grad_norm": 0.6610848307609558,
+      "learning_rate": 5.1200000000000004e-05,
+      "loss": 1.8526,
+      "step": 64
+    },
+    {
+      "epoch": 0.013499480789200415,
+      "grad_norm": 0.742107093334198,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.754,
+      "step": 65
+    },
+    {
+      "epoch": 0.013707165109034268,
+      "grad_norm": 0.6820899844169617,
+      "learning_rate": 5.280000000000001e-05,
+      "loss": 1.7355,
+      "step": 66
+    },
+    {
+      "epoch": 0.01391484942886812,
+      "grad_norm": 0.649463951587677,
+      "learning_rate": 5.360000000000001e-05,
+      "loss": 1.6453,
+      "step": 67
+    },
+    {
+      "epoch": 0.014122533748701973,
+      "grad_norm": 0.7077696323394775,
+      "learning_rate": 5.440000000000001e-05,
+      "loss": 1.7962,
+      "step": 68
+    },
+    {
+      "epoch": 0.014330218068535825,
+      "grad_norm": 0.6177412867546082,
+      "learning_rate": 5.52e-05,
+      "loss": 1.6742,
+      "step": 69
+    },
+    {
+      "epoch": 0.014537902388369679,
+      "grad_norm": 0.7839769721031189,
+      "learning_rate": 5.6e-05,
+      "loss": 1.8364,
+      "step": 70
+    },
+    {
+      "epoch": 0.01474558670820353,
+      "grad_norm": 0.6374268531799316,
+      "learning_rate": 5.6800000000000005e-05,
+      "loss": 1.7423,
+      "step": 71
+    },
+    {
+      "epoch": 0.014953271028037384,
+      "grad_norm": 0.6933043599128723,
+      "learning_rate": 5.7600000000000004e-05,
+      "loss": 1.7669,
+      "step": 72
+    },
+    {
+      "epoch": 0.015160955347871236,
+      "grad_norm": 0.641764760017395,
+      "learning_rate": 5.84e-05,
+      "loss": 1.7328,
+      "step": 73
+    },
+    {
+      "epoch": 0.015368639667705089,
+      "grad_norm": 0.6187850832939148,
+      "learning_rate": 5.92e-05,
+      "loss": 1.6375,
+      "step": 74
+    },
+    {
+      "epoch": 0.01557632398753894,
+      "grad_norm": 0.6676532030105591,
+      "learning_rate": 6.000000000000001e-05,
+      "loss": 1.7784,
+      "step": 75
+    },
+    {
+      "epoch": 0.015784008307372793,
+      "grad_norm": 0.6391186118125916,
+      "learning_rate": 6.080000000000001e-05,
+      "loss": 1.6554,
+      "step": 76
+    },
+    {
+      "epoch": 0.015991692627206646,
+      "grad_norm": 0.6301543116569519,
+      "learning_rate": 6.16e-05,
+      "loss": 1.9057,
+      "step": 77
+    },
+    {
+      "epoch": 0.0161993769470405,
+      "grad_norm": 0.6136651039123535,
+      "learning_rate": 6.240000000000001e-05,
+      "loss": 1.6371,
+      "step": 78
+    },
+    {
+      "epoch": 0.01640706126687435,
+      "grad_norm": 0.6152656674385071,
+      "learning_rate": 6.32e-05,
+      "loss": 1.7339,
+      "step": 79
+    },
+    {
+      "epoch": 0.016614745586708203,
+      "grad_norm": 0.6085280179977417,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 1.7106,
+      "step": 80
+    },
+    {
+      "epoch": 0.016822429906542057,
+      "grad_norm": 0.6468527913093567,
+      "learning_rate": 6.48e-05,
+      "loss": 1.7993,
+      "step": 81
+    },
+    {
+      "epoch": 0.01703011422637591,
+      "grad_norm": 0.6175693273544312,
+      "learning_rate": 6.56e-05,
+      "loss": 1.7434,
+      "step": 82
+    },
+    {
+      "epoch": 0.01723779854620976,
+      "grad_norm": 0.5871699452400208,
+      "learning_rate": 6.64e-05,
+      "loss": 1.6838,
+      "step": 83
+    },
+    {
+      "epoch": 0.017445482866043614,
+      "grad_norm": 0.6958571672439575,
+      "learning_rate": 6.720000000000001e-05,
+      "loss": 1.7389,
+      "step": 84
+    },
+    {
+      "epoch": 0.017653167185877467,
+      "grad_norm": 0.5976506471633911,
+      "learning_rate": 6.8e-05,
+      "loss": 1.7073,
+      "step": 85
+    },
+    {
+      "epoch": 0.017860851505711317,
+      "grad_norm": 0.6843888163566589,
+      "learning_rate": 6.88e-05,
+      "loss": 1.7339,
+      "step": 86
+    },
+    {
+      "epoch": 0.01806853582554517,
+      "grad_norm": 0.5901737213134766,
+      "learning_rate": 6.960000000000001e-05,
+      "loss": 1.7577,
+      "step": 87
+    },
+    {
+      "epoch": 0.018276220145379024,
+      "grad_norm": 0.6460241675376892,
+      "learning_rate": 7.04e-05,
+      "loss": 1.7482,
+      "step": 88
+    },
+    {
+      "epoch": 0.018483904465212878,
+      "grad_norm": 0.6465390920639038,
+      "learning_rate": 7.120000000000001e-05,
+      "loss": 1.8085,
+      "step": 89
+    },
+    {
+      "epoch": 0.018691588785046728,
+      "grad_norm": 0.6187453866004944,
+      "learning_rate": 7.2e-05,
+      "loss": 1.7507,
+      "step": 90
+    },
+    {
+      "epoch": 0.01889927310488058,
+      "grad_norm": 0.6101999282836914,
+      "learning_rate": 7.280000000000001e-05,
+      "loss": 1.7488,
+      "step": 91
+    },
+    {
+      "epoch": 0.019106957424714435,
+      "grad_norm": 0.5893586874008179,
+      "learning_rate": 7.360000000000001e-05,
+      "loss": 1.6846,
+      "step": 92
+    },
+    {
+      "epoch": 0.019314641744548288,
+      "grad_norm": 0.6104580163955688,
+      "learning_rate": 7.44e-05,
+      "loss": 1.6573,
+      "step": 93
+    },
+    {
+      "epoch": 0.019522326064382138,
+      "grad_norm": 0.6451858878135681,
+      "learning_rate": 7.52e-05,
+      "loss": 1.8052,
+      "step": 94
+    },
+    {
+      "epoch": 0.01973001038421599,
+      "grad_norm": 0.5865317583084106,
+      "learning_rate": 7.6e-05,
+      "loss": 1.6534,
+      "step": 95
+    },
+    {
+      "epoch": 0.019937694704049845,
+      "grad_norm": 0.6278175711631775,
+      "learning_rate": 7.680000000000001e-05,
+      "loss": 1.6898,
+      "step": 96
+    },
+    {
+      "epoch": 0.020145379023883695,
+      "grad_norm": 0.5943008661270142,
+      "learning_rate": 7.76e-05,
+      "loss": 1.7049,
+      "step": 97
+    },
+    {
+      "epoch": 0.02035306334371755,
+      "grad_norm": 0.6195578575134277,
+      "learning_rate": 7.840000000000001e-05,
+      "loss": 1.7527,
+      "step": 98
+    },
+    {
+      "epoch": 0.020560747663551402,
+      "grad_norm": 0.6871982216835022,
+      "learning_rate": 7.92e-05,
+      "loss": 1.708,
+      "step": 99
+    },
+    {
+      "epoch": 0.020768431983385256,
+      "grad_norm": 0.6266918182373047,
+      "learning_rate": 8e-05,
+      "loss": 1.781,
+      "step": 100
+    },
+    {
+      "epoch": 0.020976116303219106,
+      "grad_norm": 0.7398074269294739,
+      "learning_rate": 8e-05,
+      "loss": 1.7166,
+      "step": 101
+    },
+    {
+      "epoch": 0.02118380062305296,
+      "grad_norm": 0.6563434600830078,
+      "learning_rate": 8e-05,
+      "loss": 1.8624,
+      "step": 102
+    },
+    {
+      "epoch": 0.021391484942886813,
+      "grad_norm": 0.7264277338981628,
+      "learning_rate": 8e-05,
+      "loss": 1.7253,
+      "step": 103
+    },
+    {
+      "epoch": 0.021599169262720666,
+      "grad_norm": 0.6289424896240234,
+      "learning_rate": 8e-05,
+      "loss": 1.7794,
+      "step": 104
+    },
+    {
+      "epoch": 0.021806853582554516,
+      "grad_norm": 0.6097893714904785,
+      "learning_rate": 8e-05,
+      "loss": 1.7288,
+      "step": 105
+    },
+    {
+      "epoch": 0.02201453790238837,
+      "grad_norm": 0.5839325785636902,
+      "learning_rate": 8e-05,
+      "loss": 1.6222,
+      "step": 106
+    },
+    {
+      "epoch": 0.022222222222222223,
+      "grad_norm": 0.5372262001037598,
+      "learning_rate": 8e-05,
+      "loss": 1.6274,
+      "step": 107
+    },
+    {
+      "epoch": 0.022429906542056073,
+      "grad_norm": 0.5895355343818665,
+      "learning_rate": 8e-05,
+      "loss": 1.7382,
+      "step": 108
+    },
+    {
+      "epoch": 0.022637590861889927,
+      "grad_norm": 0.6392791867256165,
+      "learning_rate": 8e-05,
+      "loss": 1.7517,
+      "step": 109
+    },
+    {
+      "epoch": 0.02284527518172378,
+      "grad_norm": 0.6434172987937927,
+      "learning_rate": 8e-05,
+      "loss": 1.7384,
+      "step": 110
+    },
+    {
+      "epoch": 0.023052959501557634,
+      "grad_norm": 0.5970573425292969,
+      "learning_rate": 8e-05,
+      "loss": 1.7531,
+      "step": 111
+    },
+    {
+      "epoch": 0.023260643821391484,
+      "grad_norm": 0.5773294568061829,
+      "learning_rate": 8e-05,
+      "loss": 1.6521,
+      "step": 112
+    },
+    {
+      "epoch": 0.023468328141225337,
+      "grad_norm": 0.6489471197128296,
+      "learning_rate": 8e-05,
+      "loss": 1.7838,
+      "step": 113
+    },
+    {
+      "epoch": 0.02367601246105919,
+      "grad_norm": 0.6172572374343872,
+      "learning_rate": 8e-05,
+      "loss": 1.7552,
+      "step": 114
+    },
+    {
+      "epoch": 0.023883696780893044,
+      "grad_norm": 0.5731709003448486,
+      "learning_rate": 8e-05,
+      "loss": 1.6842,
+      "step": 115
+    },
+    {
+      "epoch": 0.024091381100726894,
+      "grad_norm": 0.5704280138015747,
+      "learning_rate": 8e-05,
+      "loss": 1.7002,
+      "step": 116
+    },
+    {
+      "epoch": 0.024299065420560748,
+      "grad_norm": 0.5774348378181458,
+      "learning_rate": 8e-05,
+      "loss": 1.7487,
+      "step": 117
+    },
+    {
+      "epoch": 0.0245067497403946,
+      "grad_norm": 0.5606811046600342,
+      "learning_rate": 8e-05,
+      "loss": 1.6928,
+      "step": 118
+    },
+    {
+      "epoch": 0.02471443406022845,
+      "grad_norm": 0.5781872868537903,
+      "learning_rate": 8e-05,
+      "loss": 1.7106,
+      "step": 119
+    },
+    {
+      "epoch": 0.024922118380062305,
+      "grad_norm": 0.5436711311340332,
+      "learning_rate": 8e-05,
+      "loss": 1.6909,
+      "step": 120
+    },
+    {
+      "epoch": 0.025129802699896158,
+      "grad_norm": 0.5731891393661499,
+      "learning_rate": 8e-05,
+      "loss": 1.6783,
+      "step": 121
+    },
+    {
+      "epoch": 0.02533748701973001,
+      "grad_norm": 0.5782831907272339,
+      "learning_rate": 8e-05,
+      "loss": 1.715,
+      "step": 122
+    },
+    {
+      "epoch": 0.02554517133956386,
+      "grad_norm": 0.6293315291404724,
+      "learning_rate": 8e-05,
+      "loss": 1.7052,
+      "step": 123
+    },
+    {
+      "epoch": 0.025752855659397715,
+      "grad_norm": 0.5766347646713257,
+      "learning_rate": 8e-05,
+      "loss": 1.7466,
+      "step": 124
+    },
+    {
+      "epoch": 0.02596053997923157,
+      "grad_norm": 0.6577507257461548,
+      "learning_rate": 8e-05,
+      "loss": 1.7412,
+      "step": 125
+    },
+    {
+      "epoch": 0.026168224299065422,
+      "grad_norm": 0.5595094561576843,
+      "learning_rate": 8e-05,
+      "loss": 1.7464,
+      "step": 126
+    },
+    {
+      "epoch": 0.026375908618899272,
+      "grad_norm": 0.5962421894073486,
+      "learning_rate": 8e-05,
+      "loss": 1.6721,
+      "step": 127
+    },
+    {
+      "epoch": 0.026583592938733126,
+      "grad_norm": 0.5711417198181152,
+      "learning_rate": 8e-05,
+      "loss": 1.7608,
+      "step": 128
+    },
+    {
+      "epoch": 0.02679127725856698,
+      "grad_norm": 0.6157316565513611,
+      "learning_rate": 8e-05,
+      "loss": 1.7212,
+      "step": 129
+    },
+    {
+      "epoch": 0.02699896157840083,
+      "grad_norm": 0.5418056845664978,
+      "learning_rate": 8e-05,
+      "loss": 1.6652,
+      "step": 130
+    },
+    {
+      "epoch": 0.027206645898234683,
+      "grad_norm": 0.5753270983695984,
+      "learning_rate": 8e-05,
+      "loss": 1.7269,
+      "step": 131
+    },
+    {
+      "epoch": 0.027414330218068536,
+      "grad_norm": 0.5521407127380371,
+      "learning_rate": 8e-05,
+      "loss": 1.7053,
+      "step": 132
+    },
+    {
+      "epoch": 0.02762201453790239,
+      "grad_norm": 0.5795088410377502,
+      "learning_rate": 8e-05,
+      "loss": 1.8084,
+      "step": 133
+    },
+    {
+      "epoch": 0.02782969885773624,
+      "grad_norm": 0.5171025991439819,
+      "learning_rate": 8e-05,
+      "loss": 1.6639,
+      "step": 134
+    },
+    {
+      "epoch": 0.028037383177570093,
+      "grad_norm": 0.5902402997016907,
+      "learning_rate": 8e-05,
+      "loss": 1.6855,
+      "step": 135
+    },
+    {
+      "epoch": 0.028245067497403947,
+      "grad_norm": 0.535606861114502,
+      "learning_rate": 8e-05,
+      "loss": 1.7269,
+      "step": 136
+    },
+    {
+      "epoch": 0.0284527518172378,
+      "grad_norm": 0.61818528175354,
+      "learning_rate": 8e-05,
+      "loss": 1.7777,
+      "step": 137
+    },
+    {
+      "epoch": 0.02866043613707165,
+      "grad_norm": 0.5806450843811035,
+      "learning_rate": 8e-05,
+      "loss": 1.738,
+      "step": 138
+    },
+    {
+      "epoch": 0.028868120456905504,
+      "grad_norm": 0.5649462342262268,
+      "learning_rate": 8e-05,
+      "loss": 1.6814,
+      "step": 139
+    },
+    {
+      "epoch": 0.029075804776739357,
+      "grad_norm": 0.5286867022514343,
+      "learning_rate": 8e-05,
+      "loss": 1.6813,
+      "step": 140
+    },
+    {
+      "epoch": 0.029283489096573207,
+      "grad_norm": 0.5503925085067749,
+      "learning_rate": 8e-05,
+      "loss": 1.6656,
+      "step": 141
+    },
+    {
+      "epoch": 0.02949117341640706,
+      "grad_norm": 0.5422264933586121,
+      "learning_rate": 8e-05,
+      "loss": 1.6465,
+      "step": 142
+    },
+    {
+      "epoch": 0.029698857736240914,
+      "grad_norm": 0.5629477500915527,
+      "learning_rate": 8e-05,
+      "loss": 1.7515,
+      "step": 143
+    },
+    {
+      "epoch": 0.029906542056074768,
+      "grad_norm": 0.5476765036582947,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 144
+    },
+    {
+      "epoch": 0.030114226375908618,
+      "grad_norm": 0.5577101111412048,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 145
+    },
+    {
+      "epoch": 0.03032191069574247,
+      "grad_norm": 0.566056489944458,
+      "learning_rate": 8e-05,
+      "loss": 1.7592,
+      "step": 146
+    },
+    {
+      "epoch": 0.030529595015576325,
+      "grad_norm": 0.5325253009796143,
+      "learning_rate": 8e-05,
+      "loss": 1.7202,
+      "step": 147
+    },
+    {
+      "epoch": 0.030737279335410178,
+      "grad_norm": 0.569216787815094,
+      "learning_rate": 8e-05,
+      "loss": 1.7454,
+      "step": 148
+    },
+    {
+      "epoch": 0.030944963655244028,
+      "grad_norm": 0.5541402101516724,
+      "learning_rate": 8e-05,
+      "loss": 1.5909,
+      "step": 149
+    },
+    {
+      "epoch": 0.03115264797507788,
+      "grad_norm": 0.5762909650802612,
+      "learning_rate": 8e-05,
+      "loss": 1.7057,
+      "step": 150
+    },
+    {
+      "epoch": 0.03136033229491173,
+      "grad_norm": 0.5259131193161011,
+      "learning_rate": 8e-05,
+      "loss": 1.624,
+      "step": 151
+    },
+    {
+      "epoch": 0.031568016614745585,
+      "grad_norm": 0.5526155233383179,
+      "learning_rate": 8e-05,
+      "loss": 1.6645,
+      "step": 152
+    },
+    {
+      "epoch": 0.03177570093457944,
+      "grad_norm": 0.5420238375663757,
+      "learning_rate": 8e-05,
+      "loss": 1.6899,
+      "step": 153
+    },
+    {
+      "epoch": 0.03198338525441329,
+      "grad_norm": 0.5291162133216858,
+      "learning_rate": 8e-05,
+      "loss": 1.7062,
+      "step": 154
+    },
+    {
+      "epoch": 0.032191069574247146,
+      "grad_norm": 0.5605945587158203,
+      "learning_rate": 8e-05,
+      "loss": 1.7297,
+      "step": 155
+    },
+    {
+      "epoch": 0.032398753894081,
+      "grad_norm": 0.5243666768074036,
+      "learning_rate": 8e-05,
+      "loss": 1.6641,
+      "step": 156
+    },
+    {
+      "epoch": 0.03260643821391485,
+      "grad_norm": 0.5971183180809021,
+      "learning_rate": 8e-05,
+      "loss": 1.6927,
+      "step": 157
+    },
+    {
+      "epoch": 0.0328141225337487,
+      "grad_norm": 0.5431556105613708,
+      "learning_rate": 8e-05,
+      "loss": 1.6636,
+      "step": 158
+    },
+    {
+      "epoch": 0.03302180685358255,
+      "grad_norm": 0.5140684247016907,
+      "learning_rate": 8e-05,
+      "loss": 1.6315,
+      "step": 159
+    },
+    {
+      "epoch": 0.033229491173416406,
+      "grad_norm": 0.518284022808075,
+      "learning_rate": 8e-05,
+      "loss": 1.6613,
+      "step": 160
+    },
+    {
+      "epoch": 0.03343717549325026,
+      "grad_norm": 0.5419729948043823,
+      "learning_rate": 8e-05,
+      "loss": 1.7693,
+      "step": 161
+    },
+    {
+      "epoch": 0.03364485981308411,
+      "grad_norm": 0.5171647667884827,
+      "learning_rate": 8e-05,
+      "loss": 1.6841,
+      "step": 162
+    },
+    {
+      "epoch": 0.03385254413291797,
+      "grad_norm": 0.5692399740219116,
+      "learning_rate": 8e-05,
+      "loss": 1.7286,
+      "step": 163
+    },
+    {
+      "epoch": 0.03406022845275182,
+      "grad_norm": 0.5294033288955688,
+      "learning_rate": 8e-05,
+      "loss": 1.7233,
+      "step": 164
+    },
+    {
+      "epoch": 0.03426791277258567,
+      "grad_norm": 0.5355773568153381,
+      "learning_rate": 8e-05,
+      "loss": 1.748,
+      "step": 165
+    },
+    {
+      "epoch": 0.03447559709241952,
+      "grad_norm": 0.5389161705970764,
+      "learning_rate": 8e-05,
+      "loss": 1.7231,
+      "step": 166
+    },
+    {
+      "epoch": 0.034683281412253374,
+      "grad_norm": 0.5242868065834045,
+      "learning_rate": 8e-05,
+      "loss": 1.759,
+      "step": 167
+    },
+    {
+      "epoch": 0.03489096573208723,
+      "grad_norm": 0.5554731488227844,
+      "learning_rate": 8e-05,
+      "loss": 1.7271,
+      "step": 168
+    },
+    {
+      "epoch": 0.03509865005192108,
+      "grad_norm": 0.603803277015686,
+      "learning_rate": 8e-05,
+      "loss": 1.7537,
+      "step": 169
+    },
+    {
+      "epoch": 0.035306334371754934,
+      "grad_norm": 0.5199013948440552,
+      "learning_rate": 8e-05,
+      "loss": 1.6379,
+      "step": 170
+    },
+    {
+      "epoch": 0.03551401869158879,
+      "grad_norm": 0.5927119851112366,
+      "learning_rate": 8e-05,
+      "loss": 1.6706,
+      "step": 171
+    },
+    {
+      "epoch": 0.035721703011422634,
+      "grad_norm": 0.5246160626411438,
+      "learning_rate": 8e-05,
+      "loss": 1.6838,
+      "step": 172
+    },
+    {
+      "epoch": 0.03592938733125649,
+      "grad_norm": 0.568761944770813,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 173
+    },
+    {
+      "epoch": 0.03613707165109034,
+      "grad_norm": 0.5335773825645447,
+      "learning_rate": 8e-05,
+      "loss": 1.6861,
+      "step": 174
+    },
+    {
+      "epoch": 0.036344755970924195,
+      "grad_norm": 0.5661566257476807,
+      "learning_rate": 8e-05,
+      "loss": 1.7069,
+      "step": 175
+    },
+    {
+      "epoch": 0.03655244029075805,
+      "grad_norm": 0.550425112247467,
+      "learning_rate": 8e-05,
+      "loss": 1.7824,
+      "step": 176
+    },
+    {
+      "epoch": 0.0367601246105919,
+      "grad_norm": 0.5691404938697815,
+      "learning_rate": 8e-05,
+      "loss": 1.7234,
+      "step": 177
+    },
+    {
+      "epoch": 0.036967808930425755,
+      "grad_norm": 0.5029447078704834,
+      "learning_rate": 8e-05,
+      "loss": 1.592,
+      "step": 178
+    },
+    {
+      "epoch": 0.03717549325025961,
+      "grad_norm": 0.5434035062789917,
+      "learning_rate": 8e-05,
+      "loss": 1.76,
+      "step": 179
+    },
+    {
+      "epoch": 0.037383177570093455,
+      "grad_norm": 0.5171444416046143,
+      "learning_rate": 8e-05,
+      "loss": 1.6538,
+      "step": 180
+    },
+    {
+      "epoch": 0.03759086188992731,
+      "grad_norm": 0.5830038785934448,
+      "learning_rate": 8e-05,
+      "loss": 1.8091,
+      "step": 181
+    },
+    {
+      "epoch": 0.03779854620976116,
+      "grad_norm": 0.5217128992080688,
+      "learning_rate": 8e-05,
+      "loss": 1.7283,
+      "step": 182
+    },
+    {
+      "epoch": 0.038006230529595016,
+      "grad_norm": 0.6341305375099182,
+      "learning_rate": 8e-05,
+      "loss": 1.7947,
+      "step": 183
+    },
+    {
+      "epoch": 0.03821391484942887,
+      "grad_norm": 0.5161294341087341,
+      "learning_rate": 8e-05,
+      "loss": 1.6634,
+      "step": 184
+    },
+    {
+      "epoch": 0.03842159916926272,
+      "grad_norm": 0.6006560325622559,
+      "learning_rate": 8e-05,
+      "loss": 1.6709,
+      "step": 185
+    },
+    {
+      "epoch": 0.038629283489096576,
+      "grad_norm": 0.5234457850456238,
+      "learning_rate": 8e-05,
+      "loss": 1.695,
+      "step": 186
+    },
+    {
+      "epoch": 0.03883696780893042,
+      "grad_norm": 0.5401941537857056,
+      "learning_rate": 8e-05,
+      "loss": 1.6625,
+      "step": 187
+    },
+    {
+      "epoch": 0.039044652128764276,
+      "grad_norm": 0.5211032032966614,
+      "learning_rate": 8e-05,
+      "loss": 1.6295,
+      "step": 188
+    },
+    {
+      "epoch": 0.03925233644859813,
+      "grad_norm": 0.5172737240791321,
+      "learning_rate": 8e-05,
+      "loss": 1.5994,
+      "step": 189
+    },
+    {
+      "epoch": 0.03946002076843198,
+      "grad_norm": 0.5083449482917786,
+      "learning_rate": 8e-05,
+      "loss": 1.6803,
+      "step": 190
+    },
+    {
+      "epoch": 0.03966770508826584,
+      "grad_norm": 0.5440491437911987,
+      "learning_rate": 8e-05,
+      "loss": 1.7009,
+      "step": 191
+    },
+    {
+      "epoch": 0.03987538940809969,
+      "grad_norm": 0.6146543622016907,
+      "learning_rate": 8e-05,
+      "loss": 1.7213,
+      "step": 192
+    },
+    {
+      "epoch": 0.040083073727933544,
+      "grad_norm": 0.5229283571243286,
+      "learning_rate": 8e-05,
+      "loss": 1.6891,
+      "step": 193
+    },
+    {
+      "epoch": 0.04029075804776739,
+      "grad_norm": 0.6130139827728271,
+      "learning_rate": 8e-05,
+      "loss": 1.7485,
+      "step": 194
+    },
+    {
+      "epoch": 0.040498442367601244,
+      "grad_norm": 0.5257254242897034,
+      "learning_rate": 8e-05,
+      "loss": 1.6668,
+      "step": 195
+    },
+    {
+      "epoch": 0.0407061266874351,
+      "grad_norm": 0.5981298089027405,
+      "learning_rate": 8e-05,
+      "loss": 1.7177,
+      "step": 196
+    },
+    {
+      "epoch": 0.04091381100726895,
+      "grad_norm": 0.48841404914855957,
+      "learning_rate": 8e-05,
+      "loss": 1.6452,
+      "step": 197
+    },
+    {
+      "epoch": 0.041121495327102804,
+      "grad_norm": 0.6227656006813049,
+      "learning_rate": 8e-05,
+      "loss": 1.6978,
+      "step": 198
+    },
+    {
+      "epoch": 0.04132917964693666,
+      "grad_norm": 0.5224297046661377,
+      "learning_rate": 8e-05,
+      "loss": 1.7488,
+      "step": 199
+    },
+    {
+      "epoch": 0.04153686396677051,
+      "grad_norm": 0.6549935936927795,
+      "learning_rate": 8e-05,
+      "loss": 1.759,
+      "step": 200
+    },
+    {
+      "epoch": 0.041744548286604365,
+      "grad_norm": 0.5142560601234436,
+      "learning_rate": 8e-05,
+      "loss": 1.6678,
+      "step": 201
+    },
+    {
+      "epoch": 0.04195223260643821,
+      "grad_norm": 0.53005051612854,
+      "learning_rate": 8e-05,
+      "loss": 1.6766,
+      "step": 202
+    },
+    {
+      "epoch": 0.042159916926272065,
+      "grad_norm": 0.5193365812301636,
+      "learning_rate": 8e-05,
+      "loss": 1.6739,
+      "step": 203
+    },
+    {
+      "epoch": 0.04236760124610592,
+      "grad_norm": 0.5400128364562988,
+      "learning_rate": 8e-05,
+      "loss": 1.7213,
+      "step": 204
+    },
+    {
+      "epoch": 0.04257528556593977,
+      "grad_norm": 0.5410746932029724,
+      "learning_rate": 8e-05,
+      "loss": 1.721,
+      "step": 205
+    },
+    {
+      "epoch": 0.042782969885773625,
+      "grad_norm": 0.4970126450061798,
+      "learning_rate": 8e-05,
+      "loss": 1.6326,
+      "step": 206
+    },
+    {
+      "epoch": 0.04299065420560748,
+      "grad_norm": 0.5962429642677307,
+      "learning_rate": 8e-05,
+      "loss": 1.7096,
+      "step": 207
+    },
+    {
+      "epoch": 0.04319833852544133,
+      "grad_norm": 0.5132887363433838,
+      "learning_rate": 8e-05,
+      "loss": 1.7047,
+      "step": 208
+    },
+    {
+      "epoch": 0.04340602284527518,
+      "grad_norm": 0.5727942585945129,
+      "learning_rate": 8e-05,
+      "loss": 1.6485,
+      "step": 209
+    },
+    {
+      "epoch": 0.04361370716510903,
+      "grad_norm": 0.516907811164856,
+      "learning_rate": 8e-05,
+      "loss": 1.693,
+      "step": 210
+    },
+    {
+      "epoch": 0.043821391484942886,
+      "grad_norm": 0.5103829503059387,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 211
+    },
+    {
+      "epoch": 0.04402907580477674,
+      "grad_norm": 0.5416673421859741,
+      "learning_rate": 8e-05,
+      "loss": 1.7322,
+      "step": 212
+    },
+    {
+      "epoch": 0.04423676012461059,
+      "grad_norm": 0.4991842806339264,
+      "learning_rate": 8e-05,
+      "loss": 1.6948,
+      "step": 213
+    },
+    {
+      "epoch": 0.044444444444444446,
+      "grad_norm": 0.5187154412269592,
+      "learning_rate": 8e-05,
+      "loss": 1.5997,
+      "step": 214
+    },
+    {
+      "epoch": 0.0446521287642783,
+      "grad_norm": 0.5057274699211121,
+      "learning_rate": 8e-05,
+      "loss": 1.7124,
+      "step": 215
+    },
+    {
+      "epoch": 0.044859813084112146,
+      "grad_norm": 0.522167980670929,
+      "learning_rate": 8e-05,
+      "loss": 1.646,
+      "step": 216
+    },
+    {
+      "epoch": 0.045067497403946,
+      "grad_norm": 0.5210246443748474,
+      "learning_rate": 8e-05,
+      "loss": 1.7006,
+      "step": 217
+    },
+    {
+      "epoch": 0.04527518172377985,
+      "grad_norm": 0.5107594728469849,
+      "learning_rate": 8e-05,
+      "loss": 1.6886,
+      "step": 218
+    },
+    {
+      "epoch": 0.04548286604361371,
+      "grad_norm": 0.546195387840271,
+      "learning_rate": 8e-05,
+      "loss": 1.6922,
+      "step": 219
+    },
+    {
+      "epoch": 0.04569055036344756,
+      "grad_norm": 0.5224754214286804,
+      "learning_rate": 8e-05,
+      "loss": 1.7163,
+      "step": 220
+    },
+    {
+      "epoch": 0.045898234683281414,
+      "grad_norm": 0.538343071937561,
+      "learning_rate": 8e-05,
+      "loss": 1.6758,
+      "step": 221
+    },
+    {
+      "epoch": 0.04610591900311527,
+      "grad_norm": 0.5105881094932556,
+      "learning_rate": 8e-05,
+      "loss": 1.7446,
+      "step": 222
+    },
+    {
+      "epoch": 0.04631360332294912,
+      "grad_norm": 0.5250230431556702,
+      "learning_rate": 8e-05,
+      "loss": 1.7269,
+      "step": 223
+    },
+    {
+      "epoch": 0.04652128764278297,
+      "grad_norm": 0.5123667120933533,
+      "learning_rate": 8e-05,
+      "loss": 1.6593,
+      "step": 224
+    },
+    {
+      "epoch": 0.04672897196261682,
+      "grad_norm": 0.5394571423530579,
+      "learning_rate": 8e-05,
+      "loss": 1.6677,
+      "step": 225
+    },
+    {
+      "epoch": 0.046936656282450674,
+      "grad_norm": 0.5013480186462402,
+      "learning_rate": 8e-05,
+      "loss": 1.6936,
+      "step": 226
+    },
+    {
+      "epoch": 0.04714434060228453,
+      "grad_norm": 0.501785933971405,
+      "learning_rate": 8e-05,
+      "loss": 1.5588,
+      "step": 227
+    },
+    {
+      "epoch": 0.04735202492211838,
+      "grad_norm": 0.5234807729721069,
+      "learning_rate": 8e-05,
+      "loss": 1.7372,
+      "step": 228
+    },
+    {
+      "epoch": 0.047559709241952235,
+      "grad_norm": 0.5221506357192993,
+      "learning_rate": 8e-05,
+      "loss": 1.6345,
+      "step": 229
+    },
+    {
+      "epoch": 0.04776739356178609,
+      "grad_norm": 0.5281847715377808,
+      "learning_rate": 8e-05,
+      "loss": 1.6074,
+      "step": 230
+    },
+    {
+      "epoch": 0.047975077881619935,
+      "grad_norm": 0.5333276987075806,
+      "learning_rate": 8e-05,
+      "loss": 1.6483,
+      "step": 231
+    },
+    {
+      "epoch": 0.04818276220145379,
+      "grad_norm": 0.49470800161361694,
+      "learning_rate": 8e-05,
+      "loss": 1.6387,
+      "step": 232
+    },
+    {
+      "epoch": 0.04839044652128764,
+      "grad_norm": 0.5513780117034912,
+      "learning_rate": 8e-05,
+      "loss": 1.6956,
+      "step": 233
+    },
+    {
+      "epoch": 0.048598130841121495,
+      "grad_norm": 0.48948371410369873,
+      "learning_rate": 8e-05,
+      "loss": 1.6548,
+      "step": 234
+    },
+    {
+      "epoch": 0.04880581516095535,
+      "grad_norm": 0.5753876566886902,
+      "learning_rate": 8e-05,
+      "loss": 1.7143,
+      "step": 235
+    },
+    {
+      "epoch": 0.0490134994807892,
+      "grad_norm": 0.5774050951004028,
+      "learning_rate": 8e-05,
+      "loss": 1.6995,
+      "step": 236
+    },
+    {
+      "epoch": 0.049221183800623056,
+      "grad_norm": 0.5836580395698547,
+      "learning_rate": 8e-05,
+      "loss": 1.75,
+      "step": 237
+    },
+    {
+      "epoch": 0.0494288681204569,
+      "grad_norm": 0.6421523094177246,
+      "learning_rate": 8e-05,
+      "loss": 1.7579,
+      "step": 238
+    },
+    {
+      "epoch": 0.049636552440290756,
+      "grad_norm": 0.5229347944259644,
+      "learning_rate": 8e-05,
+      "loss": 1.6735,
+      "step": 239
+    },
+    {
+      "epoch": 0.04984423676012461,
+      "grad_norm": 0.6658037304878235,
+      "learning_rate": 8e-05,
+      "loss": 1.7808,
+      "step": 240
+    },
+    {
+      "epoch": 0.05005192107995846,
+      "grad_norm": 0.5098143219947815,
+      "learning_rate": 8e-05,
+      "loss": 1.6961,
+      "step": 241
+    },
+    {
+      "epoch": 0.050259605399792316,
+      "grad_norm": 0.5492331385612488,
+      "learning_rate": 8e-05,
+      "loss": 1.6589,
+      "step": 242
+    },
+    {
+      "epoch": 0.05046728971962617,
+      "grad_norm": 0.5122767686843872,
+      "learning_rate": 8e-05,
+      "loss": 1.688,
+      "step": 243
+    },
+    {
+      "epoch": 0.05067497403946002,
+      "grad_norm": 0.5612162351608276,
+      "learning_rate": 8e-05,
+      "loss": 1.7156,
+      "step": 244
+    },
+    {
+      "epoch": 0.05088265835929388,
+      "grad_norm": 0.5456447601318359,
+      "learning_rate": 8e-05,
+      "loss": 1.668,
+      "step": 245
+    },
+    {
+      "epoch": 0.05109034267912772,
+      "grad_norm": 0.5987516641616821,
+      "learning_rate": 8e-05,
+      "loss": 1.7434,
+      "step": 246
+    },
+    {
+      "epoch": 0.05129802699896158,
+      "grad_norm": 0.5334711074829102,
+      "learning_rate": 8e-05,
+      "loss": 1.6684,
+      "step": 247
+    },
+    {
+      "epoch": 0.05150571131879543,
+      "grad_norm": 0.5441988706588745,
+      "learning_rate": 8e-05,
+      "loss": 1.667,
+      "step": 248
+    },
+    {
+      "epoch": 0.051713395638629284,
+      "grad_norm": 0.5460416078567505,
+      "learning_rate": 8e-05,
+      "loss": 1.6615,
+      "step": 249
+    },
+    {
+      "epoch": 0.05192107995846314,
+      "grad_norm": 0.5425359010696411,
+      "learning_rate": 8e-05,
+      "loss": 1.6637,
+      "step": 250
+    },
+    {
+      "epoch": 0.05212876427829699,
+      "grad_norm": 0.5430195927619934,
+      "learning_rate": 8e-05,
+      "loss": 1.6521,
+      "step": 251
+    },
+    {
+      "epoch": 0.052336448598130844,
+      "grad_norm": 0.4934813678264618,
+      "learning_rate": 8e-05,
+      "loss": 1.7507,
+      "step": 252
+    },
+    {
+      "epoch": 0.05254413291796469,
+      "grad_norm": 0.5606212019920349,
+      "learning_rate": 8e-05,
+      "loss": 1.754,
+      "step": 253
+    },
+    {
+      "epoch": 0.052751817237798544,
+      "grad_norm": 0.5214331746101379,
+      "learning_rate": 8e-05,
+      "loss": 1.676,
+      "step": 254
+    },
+    {
+      "epoch": 0.0529595015576324,
+      "grad_norm": 0.60882568359375,
+      "learning_rate": 8e-05,
+      "loss": 1.7517,
+      "step": 255
+    },
+    {
+      "epoch": 0.05316718587746625,
+      "grad_norm": 0.5168995261192322,
+      "learning_rate": 8e-05,
+      "loss": 1.6832,
+      "step": 256
+    },
+    {
+      "epoch": 0.053374870197300105,
+      "grad_norm": 0.5288822650909424,
+      "learning_rate": 8e-05,
+      "loss": 1.6817,
+      "step": 257
+    },
+    {
+      "epoch": 0.05358255451713396,
+      "grad_norm": 0.563677966594696,
+      "learning_rate": 8e-05,
+      "loss": 1.7,
+      "step": 258
+    },
+    {
+      "epoch": 0.05379023883696781,
+      "grad_norm": 0.5071162581443787,
+      "learning_rate": 8e-05,
+      "loss": 1.72,
+      "step": 259
+    },
+    {
+      "epoch": 0.05399792315680166,
+      "grad_norm": 0.5226413607597351,
+      "learning_rate": 8e-05,
+      "loss": 1.6503,
+      "step": 260
+    },
+    {
+      "epoch": 0.05420560747663551,
+      "grad_norm": 0.5036090016365051,
+      "learning_rate": 8e-05,
+      "loss": 1.671,
+      "step": 261
+    },
+    {
+      "epoch": 0.054413291796469365,
+      "grad_norm": 0.5291279554367065,
+      "learning_rate": 8e-05,
+      "loss": 1.6547,
+      "step": 262
+    },
+    {
+      "epoch": 0.05462097611630322,
+      "grad_norm": 0.5152414441108704,
+      "learning_rate": 8e-05,
+      "loss": 1.7235,
+      "step": 263
+    },
+    {
+      "epoch": 0.05482866043613707,
+      "grad_norm": 0.5221415758132935,
+      "learning_rate": 8e-05,
+      "loss": 1.7344,
+      "step": 264
+    },
+    {
+      "epoch": 0.055036344755970926,
+      "grad_norm": 0.5111775398254395,
+      "learning_rate": 8e-05,
+      "loss": 1.6625,
+      "step": 265
+    },
+    {
+      "epoch": 0.05524402907580478,
+      "grad_norm": 0.49026650190353394,
+      "learning_rate": 8e-05,
+      "loss": 1.626,
+      "step": 266
+    },
+    {
+      "epoch": 0.05545171339563863,
+      "grad_norm": 0.5113672614097595,
+      "learning_rate": 8e-05,
+      "loss": 1.7152,
+      "step": 267
+    },
+    {
+      "epoch": 0.05565939771547248,
+      "grad_norm": 0.5143501162528992,
+      "learning_rate": 8e-05,
+      "loss": 1.6065,
+      "step": 268
+    },
+    {
+      "epoch": 0.05586708203530633,
+      "grad_norm": 0.548305094242096,
+      "learning_rate": 8e-05,
+      "loss": 1.7169,
+      "step": 269
+    },
+    {
+      "epoch": 0.056074766355140186,
+      "grad_norm": 0.5203267931938171,
+      "learning_rate": 8e-05,
+      "loss": 1.7894,
+      "step": 270
+    },
+    {
+      "epoch": 0.05628245067497404,
+      "grad_norm": 0.5775648951530457,
+      "learning_rate": 8e-05,
+      "loss": 1.6821,
+      "step": 271
+    },
+    {
+      "epoch": 0.05649013499480789,
+      "grad_norm": 0.48842960596084595,
+      "learning_rate": 8e-05,
+      "loss": 1.6095,
+      "step": 272
+    },
+    {
+      "epoch": 0.05669781931464175,
+      "grad_norm": 0.5582982301712036,
+      "learning_rate": 8e-05,
+      "loss": 1.7071,
+      "step": 273
+    },
+    {
+      "epoch": 0.0569055036344756,
+      "grad_norm": 0.4931994378566742,
+      "learning_rate": 8e-05,
+      "loss": 1.6754,
+      "step": 274
+    },
+    {
+      "epoch": 0.05711318795430945,
+      "grad_norm": 0.5403690934181213,
+      "learning_rate": 8e-05,
+      "loss": 1.6986,
+      "step": 275
+    },
+    {
+      "epoch": 0.0573208722741433,
+      "grad_norm": 0.5050585269927979,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 276
+    },
+    {
+      "epoch": 0.057528556593977154,
+      "grad_norm": 0.5011170506477356,
+      "learning_rate": 8e-05,
+      "loss": 1.6561,
+      "step": 277
+    },
+    {
+      "epoch": 0.05773624091381101,
+      "grad_norm": 0.5033249258995056,
+      "learning_rate": 8e-05,
+      "loss": 1.7003,
+      "step": 278
+    },
+    {
+      "epoch": 0.05794392523364486,
+      "grad_norm": 0.5251483917236328,
+      "learning_rate": 8e-05,
+      "loss": 1.6675,
+      "step": 279
+    },
+    {
+      "epoch": 0.058151609553478714,
+      "grad_norm": 0.5164880752563477,
+      "learning_rate": 8e-05,
+      "loss": 1.7132,
+      "step": 280
+    },
+    {
+      "epoch": 0.05835929387331257,
+      "grad_norm": 0.4889441728591919,
+      "learning_rate": 8e-05,
+      "loss": 1.7148,
+      "step": 281
+    },
+    {
+      "epoch": 0.058566978193146414,
+      "grad_norm": 0.49667251110076904,
+      "learning_rate": 8e-05,
+      "loss": 1.7136,
+      "step": 282
+    },
+    {
+      "epoch": 0.05877466251298027,
+      "grad_norm": 0.5121577382087708,
+      "learning_rate": 8e-05,
+      "loss": 1.764,
+      "step": 283
+    },
+    {
+      "epoch": 0.05898234683281412,
+      "grad_norm": 0.48562902212142944,
+      "learning_rate": 8e-05,
+      "loss": 1.7124,
+      "step": 284
+    },
+    {
+      "epoch": 0.059190031152647975,
+      "grad_norm": 0.4913182556629181,
+      "learning_rate": 8e-05,
+      "loss": 1.6498,
+      "step": 285
+    },
+    {
+      "epoch": 0.05939771547248183,
+      "grad_norm": 0.5064699053764343,
+      "learning_rate": 8e-05,
+      "loss": 1.598,
+      "step": 286
+    },
+    {
+      "epoch": 0.05960539979231568,
+      "grad_norm": 0.553032398223877,
+      "learning_rate": 8e-05,
+      "loss": 1.705,
+      "step": 287
+    },
+    {
+      "epoch": 0.059813084112149535,
+      "grad_norm": 0.5344775319099426,
+      "learning_rate": 8e-05,
+      "loss": 1.7229,
+      "step": 288
+    },
+    {
+      "epoch": 0.06002076843198339,
+      "grad_norm": 0.5126497149467468,
+      "learning_rate": 8e-05,
+      "loss": 1.7099,
+      "step": 289
+    },
+    {
+      "epoch": 0.060228452751817235,
+      "grad_norm": 0.4790019690990448,
+      "learning_rate": 8e-05,
+      "loss": 1.6468,
+      "step": 290
+    },
+    {
+      "epoch": 0.06043613707165109,
+      "grad_norm": 0.5178213715553284,
+      "learning_rate": 8e-05,
+      "loss": 1.729,
+      "step": 291
+    },
+    {
+      "epoch": 0.06064382139148494,
+      "grad_norm": 0.4989386796951294,
+      "learning_rate": 8e-05,
+      "loss": 1.7297,
+      "step": 292
+    },
+    {
+      "epoch": 0.060851505711318796,
+      "grad_norm": 0.49952176213264465,
+      "learning_rate": 8e-05,
+      "loss": 1.7531,
+      "step": 293
+    },
+    {
+      "epoch": 0.06105919003115265,
+      "grad_norm": 0.5082775354385376,
+      "learning_rate": 8e-05,
+      "loss": 1.7071,
+      "step": 294
+    },
+    {
+      "epoch": 0.0612668743509865,
+      "grad_norm": 0.4975671172142029,
+      "learning_rate": 8e-05,
+      "loss": 1.6984,
+      "step": 295
+    },
+    {
+      "epoch": 0.061474558670820356,
+      "grad_norm": 0.5003031492233276,
+      "learning_rate": 8e-05,
+      "loss": 1.657,
+      "step": 296
+    },
+    {
+      "epoch": 0.0616822429906542,
+      "grad_norm": 0.524743378162384,
+      "learning_rate": 8e-05,
+      "loss": 1.6763,
+      "step": 297
+    },
+    {
+      "epoch": 0.061889927310488056,
+      "grad_norm": 0.5136753916740417,
+      "learning_rate": 8e-05,
+      "loss": 1.6642,
+      "step": 298
+    },
+    {
+      "epoch": 0.06209761163032191,
+      "grad_norm": 0.55498868227005,
+      "learning_rate": 8e-05,
+      "loss": 1.7189,
+      "step": 299
+    },
+    {
+      "epoch": 0.06230529595015576,
+      "grad_norm": 0.468551367521286,
+      "learning_rate": 8e-05,
+      "loss": 1.6853,
+      "step": 300
+    },
+    {
+      "epoch": 0.06251298026998961,
+      "grad_norm": 0.5169484615325928,
+      "learning_rate": 8e-05,
+      "loss": 1.6934,
+      "step": 301
+    },
+    {
+      "epoch": 0.06272066458982346,
+      "grad_norm": 0.4962407052516937,
+      "learning_rate": 8e-05,
+      "loss": 1.67,
+      "step": 302
+    },
+    {
+      "epoch": 0.06292834890965732,
+      "grad_norm": 0.5252827405929565,
+      "learning_rate": 8e-05,
+      "loss": 1.7051,
+      "step": 303
+    },
+    {
+      "epoch": 0.06313603322949117,
+      "grad_norm": 0.49051743745803833,
+      "learning_rate": 8e-05,
+      "loss": 1.6435,
+      "step": 304
+    },
+    {
+      "epoch": 0.06334371754932502,
+      "grad_norm": 0.49855688214302063,
+      "learning_rate": 8e-05,
+      "loss": 1.6876,
+      "step": 305
+    },
+    {
+      "epoch": 0.06355140186915888,
+      "grad_norm": 0.5097089409828186,
+      "learning_rate": 8e-05,
+      "loss": 1.7275,
+      "step": 306
+    },
+    {
+      "epoch": 0.06375908618899273,
+      "grad_norm": 0.4844972491264343,
+      "learning_rate": 8e-05,
+      "loss": 1.596,
+      "step": 307
+    },
+    {
+      "epoch": 0.06396677050882658,
+      "grad_norm": 0.4773125946521759,
+      "learning_rate": 8e-05,
+      "loss": 1.6302,
+      "step": 308
+    },
+    {
+      "epoch": 0.06417445482866044,
+      "grad_norm": 0.4860425293445587,
+      "learning_rate": 8e-05,
+      "loss": 1.6042,
+      "step": 309
+    },
+    {
+      "epoch": 0.06438213914849429,
+      "grad_norm": 0.49147120118141174,
+      "learning_rate": 8e-05,
+      "loss": 1.6496,
+      "step": 310
+    },
+    {
+      "epoch": 0.06458982346832814,
+      "grad_norm": 0.4904959797859192,
+      "learning_rate": 8e-05,
+      "loss": 1.6267,
+      "step": 311
+    },
+    {
+      "epoch": 0.064797507788162,
+      "grad_norm": 0.5017173886299133,
+      "learning_rate": 8e-05,
+      "loss": 1.6827,
+      "step": 312
+    },
+    {
+      "epoch": 0.06500519210799585,
+      "grad_norm": 0.5119773149490356,
+      "learning_rate": 8e-05,
+      "loss": 1.7423,
+      "step": 313
+    },
+    {
+      "epoch": 0.0652128764278297,
+      "grad_norm": 0.5187619924545288,
+      "learning_rate": 8e-05,
+      "loss": 1.6919,
+      "step": 314
+    },
+    {
+      "epoch": 0.06542056074766354,
+      "grad_norm": 0.499828577041626,
+      "learning_rate": 8e-05,
+      "loss": 1.6802,
+      "step": 315
+    },
+    {
+      "epoch": 0.0656282450674974,
+      "grad_norm": 0.5605705976486206,
+      "learning_rate": 8e-05,
+      "loss": 1.7527,
+      "step": 316
+    },
+    {
+      "epoch": 0.06583592938733125,
+      "grad_norm": 0.477637380361557,
+      "learning_rate": 8e-05,
+      "loss": 1.6605,
+      "step": 317
+    },
+    {
+      "epoch": 0.0660436137071651,
+      "grad_norm": 0.5356731414794922,
+      "learning_rate": 8e-05,
+      "loss": 1.6564,
+      "step": 318
+    },
+    {
+      "epoch": 0.06625129802699896,
+      "grad_norm": 0.4828336834907532,
+      "learning_rate": 8e-05,
+      "loss": 1.5889,
+      "step": 319
+    },
+    {
+      "epoch": 0.06645898234683281,
+      "grad_norm": 0.5663031339645386,
+      "learning_rate": 8e-05,
+      "loss": 1.7158,
+      "step": 320
+    },
+    {
+      "epoch": 0.06666666666666667,
+      "grad_norm": 0.5123764872550964,
+      "learning_rate": 8e-05,
+      "loss": 1.6212,
+      "step": 321
+    },
+    {
+      "epoch": 0.06687435098650052,
+      "grad_norm": 0.5193995833396912,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 322
+    },
+    {
+      "epoch": 0.06708203530633437,
+      "grad_norm": 0.612371563911438,
+      "learning_rate": 8e-05,
+      "loss": 1.7297,
+      "step": 323
+    },
+    {
+      "epoch": 0.06728971962616823,
+      "grad_norm": 0.5306211709976196,
+      "learning_rate": 8e-05,
+      "loss": 1.5693,
+      "step": 324
+    },
+    {
+      "epoch": 0.06749740394600208,
+      "grad_norm": 0.5609772801399231,
+      "learning_rate": 8e-05,
+      "loss": 1.6011,
+      "step": 325
+    },
+    {
+      "epoch": 0.06770508826583593,
+      "grad_norm": 0.5140113234519958,
+      "learning_rate": 8e-05,
+      "loss": 1.6519,
+      "step": 326
+    },
+    {
+      "epoch": 0.06791277258566979,
+      "grad_norm": 0.5529564023017883,
+      "learning_rate": 8e-05,
+      "loss": 1.6217,
+      "step": 327
+    },
+    {
+      "epoch": 0.06812045690550364,
+      "grad_norm": 0.48979297280311584,
+      "learning_rate": 8e-05,
+      "loss": 1.6044,
+      "step": 328
+    },
+    {
+      "epoch": 0.0683281412253375,
+      "grad_norm": 0.5272225141525269,
+      "learning_rate": 8e-05,
+      "loss": 1.7409,
+      "step": 329
+    },
+    {
+      "epoch": 0.06853582554517133,
+      "grad_norm": 0.5034517645835876,
+      "learning_rate": 8e-05,
+      "loss": 1.6181,
+      "step": 330
+    },
+    {
+      "epoch": 0.06874350986500519,
+      "grad_norm": 0.4756111800670624,
+      "learning_rate": 8e-05,
+      "loss": 1.6358,
+      "step": 331
+    },
+    {
+      "epoch": 0.06895119418483904,
+      "grad_norm": 0.5277131199836731,
+      "learning_rate": 8e-05,
+      "loss": 1.7289,
+      "step": 332
+    },
+    {
+      "epoch": 0.0691588785046729,
+      "grad_norm": 0.4864356815814972,
+      "learning_rate": 8e-05,
+      "loss": 1.6573,
+      "step": 333
+    },
+    {
+      "epoch": 0.06936656282450675,
+      "grad_norm": 0.49576079845428467,
+      "learning_rate": 8e-05,
+      "loss": 1.7073,
+      "step": 334
+    },
+    {
+      "epoch": 0.0695742471443406,
+      "grad_norm": 0.4868834614753723,
+      "learning_rate": 8e-05,
+      "loss": 1.6662,
+      "step": 335
+    },
+    {
+      "epoch": 0.06978193146417445,
+      "grad_norm": 0.5043144822120667,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 336
+    },
+    {
+      "epoch": 0.06998961578400831,
+      "grad_norm": 0.49576571583747864,
+      "learning_rate": 8e-05,
+      "loss": 1.7137,
+      "step": 337
+    },
+    {
+      "epoch": 0.07019730010384216,
+      "grad_norm": 0.4854160249233246,
+      "learning_rate": 8e-05,
+      "loss": 1.6409,
+      "step": 338
+    },
+    {
+      "epoch": 0.07040498442367601,
+      "grad_norm": 0.48695647716522217,
+      "learning_rate": 8e-05,
+      "loss": 1.6114,
+      "step": 339
+    },
+    {
+      "epoch": 0.07061266874350987,
+      "grad_norm": 0.4941887855529785,
+      "learning_rate": 8e-05,
+      "loss": 1.6843,
+      "step": 340
+    },
+    {
+      "epoch": 0.07082035306334372,
+      "grad_norm": 0.49328118562698364,
+      "learning_rate": 8e-05,
+      "loss": 1.6837,
+      "step": 341
+    },
+    {
+      "epoch": 0.07102803738317758,
+      "grad_norm": 0.4951527714729309,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 342
+    },
+    {
+      "epoch": 0.07123572170301143,
+      "grad_norm": 0.49303266406059265,
+      "learning_rate": 8e-05,
+      "loss": 1.6726,
+      "step": 343
+    },
+    {
+      "epoch": 0.07144340602284527,
+      "grad_norm": 0.48809322714805603,
+      "learning_rate": 8e-05,
+      "loss": 1.6716,
+      "step": 344
+    },
+    {
+      "epoch": 0.07165109034267912,
+      "grad_norm": 0.5064594745635986,
+      "learning_rate": 8e-05,
+      "loss": 1.7361,
+      "step": 345
+    },
+    {
+      "epoch": 0.07185877466251298,
+      "grad_norm": 0.5080369710922241,
+      "learning_rate": 8e-05,
+      "loss": 1.6929,
+      "step": 346
+    },
+    {
+      "epoch": 0.07206645898234683,
+      "grad_norm": 0.4928899109363556,
+      "learning_rate": 8e-05,
+      "loss": 1.7132,
+      "step": 347
+    },
+    {
+      "epoch": 0.07227414330218068,
+      "grad_norm": 0.47712448239326477,
+      "learning_rate": 8e-05,
+      "loss": 1.6334,
+      "step": 348
+    },
+    {
+      "epoch": 0.07248182762201454,
+      "grad_norm": 0.513994038105011,
+      "learning_rate": 8e-05,
+      "loss": 1.699,
+      "step": 349
+    },
+    {
+      "epoch": 0.07268951194184839,
+      "grad_norm": 0.5016348958015442,
+      "learning_rate": 8e-05,
+      "loss": 1.634,
+      "step": 350
+    },
+    {
+      "epoch": 0.07289719626168224,
+      "grad_norm": 0.5038484334945679,
+      "learning_rate": 8e-05,
+      "loss": 1.6643,
+      "step": 351
+    },
+    {
+      "epoch": 0.0731048805815161,
+      "grad_norm": 0.4928259551525116,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 352
+    },
+    {
+      "epoch": 0.07331256490134995,
+      "grad_norm": 0.5103161334991455,
+      "learning_rate": 8e-05,
+      "loss": 1.6494,
+      "step": 353
+    },
+    {
+      "epoch": 0.0735202492211838,
+      "grad_norm": 0.47103309631347656,
+      "learning_rate": 8e-05,
+      "loss": 1.6299,
+      "step": 354
+    },
+    {
+      "epoch": 0.07372793354101766,
+      "grad_norm": 0.5037011504173279,
+      "learning_rate": 8e-05,
+      "loss": 1.6283,
+      "step": 355
+    },
+    {
+      "epoch": 0.07393561786085151,
+      "grad_norm": 0.5026373267173767,
+      "learning_rate": 8e-05,
+      "loss": 1.7157,
+      "step": 356
+    },
+    {
+      "epoch": 0.07414330218068536,
+      "grad_norm": 0.5053650140762329,
+      "learning_rate": 8e-05,
+      "loss": 1.6587,
+      "step": 357
+    },
+    {
+      "epoch": 0.07435098650051922,
+      "grad_norm": 0.496429443359375,
+      "learning_rate": 8e-05,
+      "loss": 1.7041,
+      "step": 358
+    },
+    {
+      "epoch": 0.07455867082035306,
+      "grad_norm": 0.47459688782691956,
+      "learning_rate": 8e-05,
+      "loss": 1.6136,
+      "step": 359
+    },
+    {
+      "epoch": 0.07476635514018691,
+      "grad_norm": 0.5140553116798401,
+      "learning_rate": 8e-05,
+      "loss": 1.6625,
+      "step": 360
+    },
+    {
+      "epoch": 0.07497403946002076,
+      "grad_norm": 0.515044093132019,
+      "learning_rate": 8e-05,
+      "loss": 1.711,
+      "step": 361
+    },
+    {
+      "epoch": 0.07518172377985462,
+      "grad_norm": 0.5119950771331787,
+      "learning_rate": 8e-05,
+      "loss": 1.6852,
+      "step": 362
+    },
+    {
+      "epoch": 0.07538940809968847,
+      "grad_norm": 0.5062432289123535,
+      "learning_rate": 8e-05,
+      "loss": 1.6892,
+      "step": 363
+    },
+    {
+      "epoch": 0.07559709241952232,
+      "grad_norm": 0.5517569780349731,
+      "learning_rate": 8e-05,
+      "loss": 1.6883,
+      "step": 364
+    },
+    {
+      "epoch": 0.07580477673935618,
+      "grad_norm": 0.48322999477386475,
+      "learning_rate": 8e-05,
+      "loss": 1.6383,
+      "step": 365
+    },
+    {
+      "epoch": 0.07601246105919003,
+      "grad_norm": 0.47941911220550537,
+      "learning_rate": 8e-05,
+      "loss": 1.6032,
+      "step": 366
+    },
+    {
+      "epoch": 0.07622014537902388,
+      "grad_norm": 0.4892639219760895,
+      "learning_rate": 8e-05,
+      "loss": 1.6862,
+      "step": 367
+    },
+    {
+      "epoch": 0.07642782969885774,
+      "grad_norm": 0.48457542061805725,
+      "learning_rate": 8e-05,
+      "loss": 1.6253,
+      "step": 368
+    },
+    {
+      "epoch": 0.07663551401869159,
+      "grad_norm": 0.48371532559394836,
+      "learning_rate": 8e-05,
+      "loss": 1.7072,
+      "step": 369
+    },
+    {
+      "epoch": 0.07684319833852545,
+      "grad_norm": 0.5162315964698792,
+      "learning_rate": 8e-05,
+      "loss": 1.6952,
+      "step": 370
+    },
+    {
+      "epoch": 0.0770508826583593,
+      "grad_norm": 0.494991660118103,
+      "learning_rate": 8e-05,
+      "loss": 1.7292,
+      "step": 371
+    },
+    {
+      "epoch": 0.07725856697819315,
+      "grad_norm": 0.510187029838562,
+      "learning_rate": 8e-05,
+      "loss": 1.6835,
+      "step": 372
+    },
+    {
+      "epoch": 0.077466251298027,
+      "grad_norm": 0.47694623470306396,
+      "learning_rate": 8e-05,
+      "loss": 1.6809,
+      "step": 373
+    },
+    {
+      "epoch": 0.07767393561786085,
+      "grad_norm": 0.4950973093509674,
+      "learning_rate": 8e-05,
+      "loss": 1.6691,
+      "step": 374
+    },
+    {
+      "epoch": 0.0778816199376947,
+      "grad_norm": 0.4885307848453522,
+      "learning_rate": 8e-05,
+      "loss": 1.6682,
+      "step": 375
+    },
+    {
+      "epoch": 0.07808930425752855,
+      "grad_norm": 0.5185424089431763,
+      "learning_rate": 8e-05,
+      "loss": 1.698,
+      "step": 376
+    },
+    {
+      "epoch": 0.0782969885773624,
+      "grad_norm": 0.48990127444267273,
+      "learning_rate": 8e-05,
+      "loss": 1.6877,
+      "step": 377
+    },
+    {
+      "epoch": 0.07850467289719626,
+      "grad_norm": 0.5352708101272583,
+      "learning_rate": 8e-05,
+      "loss": 1.6141,
+      "step": 378
+    },
+    {
+      "epoch": 0.07871235721703011,
+      "grad_norm": 0.48971933126449585,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 379
+    },
+    {
+      "epoch": 0.07892004153686397,
+      "grad_norm": 0.5039674639701843,
+      "learning_rate": 8e-05,
+      "loss": 1.6772,
+      "step": 380
+    },
+    {
+      "epoch": 0.07912772585669782,
+      "grad_norm": 0.5058568120002747,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 381
+    },
+    {
+      "epoch": 0.07933541017653167,
+      "grad_norm": 0.5172670483589172,
+      "learning_rate": 8e-05,
+      "loss": 1.6229,
+      "step": 382
+    },
+    {
+      "epoch": 0.07954309449636553,
+      "grad_norm": 0.4776594340801239,
+      "learning_rate": 8e-05,
+      "loss": 1.6517,
+      "step": 383
+    },
+    {
+      "epoch": 0.07975077881619938,
+      "grad_norm": 0.5346738696098328,
+      "learning_rate": 8e-05,
+      "loss": 1.6985,
+      "step": 384
+    },
+    {
+      "epoch": 0.07995846313603323,
+      "grad_norm": 0.5013498067855835,
+      "learning_rate": 8e-05,
+      "loss": 1.6297,
+      "step": 385
+    },
+    {
+      "epoch": 0.08016614745586709,
+      "grad_norm": 0.6052170991897583,
+      "learning_rate": 8e-05,
+      "loss": 1.6397,
+      "step": 386
+    },
+    {
+      "epoch": 0.08037383177570094,
+      "grad_norm": 0.4749467074871063,
+      "learning_rate": 8e-05,
+      "loss": 1.6263,
+      "step": 387
+    },
+    {
+      "epoch": 0.08058151609553478,
+      "grad_norm": 0.4913139045238495,
+      "learning_rate": 8e-05,
+      "loss": 1.6091,
+      "step": 388
+    },
+    {
+      "epoch": 0.08078920041536863,
+      "grad_norm": 0.5050566792488098,
+      "learning_rate": 8e-05,
+      "loss": 1.6801,
+      "step": 389
+    },
+    {
+      "epoch": 0.08099688473520249,
+      "grad_norm": 0.4727949798107147,
+      "learning_rate": 8e-05,
+      "loss": 1.6141,
+      "step": 390
+    },
+    {
+      "epoch": 0.08120456905503634,
+      "grad_norm": 0.4729914367198944,
+      "learning_rate": 8e-05,
+      "loss": 1.5569,
+      "step": 391
+    },
+    {
+      "epoch": 0.0814122533748702,
+      "grad_norm": 0.4925561249256134,
+      "learning_rate": 8e-05,
+      "loss": 1.7113,
+      "step": 392
+    },
+    {
+      "epoch": 0.08161993769470405,
+      "grad_norm": 0.49586695432662964,
+      "learning_rate": 8e-05,
+      "loss": 1.6699,
+      "step": 393
+    },
+    {
+      "epoch": 0.0818276220145379,
+      "grad_norm": 0.5150527954101562,
+      "learning_rate": 8e-05,
+      "loss": 1.6564,
+      "step": 394
+    },
+    {
+      "epoch": 0.08203530633437175,
+      "grad_norm": 0.5019596219062805,
+      "learning_rate": 8e-05,
+      "loss": 1.6275,
+      "step": 395
+    },
+    {
+      "epoch": 0.08224299065420561,
+      "grad_norm": 0.4876788258552551,
+      "learning_rate": 8e-05,
+      "loss": 1.5836,
+      "step": 396
+    },
+    {
+      "epoch": 0.08245067497403946,
+      "grad_norm": 0.49811601638793945,
+      "learning_rate": 8e-05,
+      "loss": 1.6608,
+      "step": 397
+    },
+    {
+      "epoch": 0.08265835929387332,
+      "grad_norm": 0.5187938809394836,
+      "learning_rate": 8e-05,
+      "loss": 1.6999,
+      "step": 398
+    },
+    {
+      "epoch": 0.08286604361370717,
+      "grad_norm": 0.4808737635612488,
+      "learning_rate": 8e-05,
+      "loss": 1.6004,
+      "step": 399
+    },
+    {
+      "epoch": 0.08307372793354102,
+      "grad_norm": 0.515282392501831,
+      "learning_rate": 8e-05,
+      "loss": 1.7455,
+      "step": 400
+    },
+    {
+      "epoch": 0.08328141225337488,
+      "grad_norm": 0.5063252449035645,
+      "learning_rate": 8e-05,
+      "loss": 1.6955,
+      "step": 401
+    },
+    {
+      "epoch": 0.08348909657320873,
+      "grad_norm": 0.5026871562004089,
+      "learning_rate": 8e-05,
+      "loss": 1.7067,
+      "step": 402
+    },
+    {
+      "epoch": 0.08369678089304257,
+      "grad_norm": 0.4903314709663391,
+      "learning_rate": 8e-05,
+      "loss": 1.6333,
+      "step": 403
+    },
+    {
+      "epoch": 0.08390446521287642,
+      "grad_norm": 0.4954453110694885,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 404
+    },
+    {
+      "epoch": 0.08411214953271028,
+      "grad_norm": 0.5026134252548218,
+      "learning_rate": 8e-05,
+      "loss": 1.6969,
+      "step": 405
+    },
+    {
+      "epoch": 0.08431983385254413,
+      "grad_norm": 0.49587976932525635,
+      "learning_rate": 8e-05,
+      "loss": 1.651,
+      "step": 406
+    },
+    {
+      "epoch": 0.08452751817237798,
+      "grad_norm": 0.5106424689292908,
+      "learning_rate": 8e-05,
+      "loss": 1.7021,
+      "step": 407
+    },
+    {
+      "epoch": 0.08473520249221184,
+      "grad_norm": 0.5195350646972656,
+      "learning_rate": 8e-05,
+      "loss": 1.7539,
+      "step": 408
+    },
+    {
+      "epoch": 0.08494288681204569,
+      "grad_norm": 0.501416027545929,
+      "learning_rate": 8e-05,
+      "loss": 1.725,
+      "step": 409
+    },
+    {
+      "epoch": 0.08515057113187954,
+      "grad_norm": 0.48874419927597046,
+      "learning_rate": 8e-05,
+      "loss": 1.7037,
+      "step": 410
+    },
+    {
+      "epoch": 0.0853582554517134,
+      "grad_norm": 0.4698718190193176,
+      "learning_rate": 8e-05,
+      "loss": 1.6509,
+      "step": 411
+    },
+    {
+      "epoch": 0.08556593977154725,
+      "grad_norm": 0.4783203899860382,
+      "learning_rate": 8e-05,
+      "loss": 1.6171,
+      "step": 412
+    },
+    {
+      "epoch": 0.0857736240913811,
+      "grad_norm": 0.481812983751297,
+      "learning_rate": 8e-05,
+      "loss": 1.6676,
+      "step": 413
+    },
+    {
+      "epoch": 0.08598130841121496,
+      "grad_norm": 0.481001615524292,
+      "learning_rate": 8e-05,
+      "loss": 1.7201,
+      "step": 414
+    },
+    {
+      "epoch": 0.08618899273104881,
+      "grad_norm": 0.48555779457092285,
+      "learning_rate": 8e-05,
+      "loss": 1.6573,
+      "step": 415
+    },
+    {
+      "epoch": 0.08639667705088266,
+      "grad_norm": 0.49405309557914734,
+      "learning_rate": 8e-05,
+      "loss": 1.6688,
+      "step": 416
+    },
+    {
+      "epoch": 0.08660436137071652,
+      "grad_norm": 0.5051506757736206,
+      "learning_rate": 8e-05,
+      "loss": 1.7261,
+      "step": 417
+    },
+    {
+      "epoch": 0.08681204569055036,
+      "grad_norm": 0.48117661476135254,
+      "learning_rate": 8e-05,
+      "loss": 1.6696,
+      "step": 418
+    },
+    {
+      "epoch": 0.08701973001038421,
+      "grad_norm": 0.4655582904815674,
+      "learning_rate": 8e-05,
+      "loss": 1.5525,
+      "step": 419
+    },
+    {
+      "epoch": 0.08722741433021806,
+      "grad_norm": 0.4815155863761902,
+      "learning_rate": 8e-05,
+      "loss": 1.6482,
+      "step": 420
+    },
+    {
+      "epoch": 0.08743509865005192,
+      "grad_norm": 0.4939725697040558,
+      "learning_rate": 8e-05,
+      "loss": 1.6752,
+      "step": 421
+    },
+    {
+      "epoch": 0.08764278296988577,
+      "grad_norm": 0.4941619634628296,
+      "learning_rate": 8e-05,
+      "loss": 1.658,
+      "step": 422
+    },
+    {
+      "epoch": 0.08785046728971962,
+      "grad_norm": 0.5332744121551514,
+      "learning_rate": 8e-05,
+      "loss": 1.6118,
+      "step": 423
+    },
+    {
+      "epoch": 0.08805815160955348,
+      "grad_norm": 0.5089534521102905,
+      "learning_rate": 8e-05,
+      "loss": 1.6835,
+      "step": 424
+    },
+    {
+      "epoch": 0.08826583592938733,
+      "grad_norm": 0.5659358501434326,
+      "learning_rate": 8e-05,
+      "loss": 1.646,
+      "step": 425
+    },
+    {
+      "epoch": 0.08847352024922119,
+      "grad_norm": 0.49219921231269836,
+      "learning_rate": 8e-05,
+      "loss": 1.6452,
+      "step": 426
+    },
+    {
+      "epoch": 0.08868120456905504,
+      "grad_norm": 0.5574254989624023,
+      "learning_rate": 8e-05,
+      "loss": 1.6968,
+      "step": 427
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 0.4780307412147522,
+      "learning_rate": 8e-05,
+      "loss": 1.6332,
+      "step": 428
+    },
+    {
+      "epoch": 0.08909657320872275,
+      "grad_norm": 0.4689270555973053,
+      "learning_rate": 8e-05,
+      "loss": 1.6342,
+      "step": 429
+    },
+    {
+      "epoch": 0.0893042575285566,
+      "grad_norm": 0.5344724655151367,
+      "learning_rate": 8e-05,
+      "loss": 1.6307,
+      "step": 430
+    },
+    {
+      "epoch": 0.08951194184839045,
+      "grad_norm": 0.4806777536869049,
+      "learning_rate": 8e-05,
+      "loss": 1.6404,
+      "step": 431
+    },
+    {
+      "epoch": 0.08971962616822429,
+      "grad_norm": 0.5253303050994873,
+      "learning_rate": 8e-05,
+      "loss": 1.6228,
+      "step": 432
+    },
+    {
+      "epoch": 0.08992731048805815,
+      "grad_norm": 0.49155253171920776,
+      "learning_rate": 8e-05,
+      "loss": 1.6423,
+      "step": 433
+    },
+    {
+      "epoch": 0.090134994807892,
+      "grad_norm": 0.5021010637283325,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 434
+    },
+    {
+      "epoch": 0.09034267912772585,
+      "grad_norm": 0.5126837491989136,
+      "learning_rate": 8e-05,
+      "loss": 1.6719,
+      "step": 435
+    },
+    {
+      "epoch": 0.0905503634475597,
+      "grad_norm": 0.48987722396850586,
+      "learning_rate": 8e-05,
+      "loss": 1.6355,
+      "step": 436
+    },
+    {
+      "epoch": 0.09075804776739356,
+      "grad_norm": 0.46434321999549866,
+      "learning_rate": 8e-05,
+      "loss": 1.6236,
+      "step": 437
+    },
+    {
+      "epoch": 0.09096573208722741,
+      "grad_norm": 0.4966569244861603,
+      "learning_rate": 8e-05,
+      "loss": 1.7529,
+      "step": 438
+    },
+    {
+      "epoch": 0.09117341640706127,
+      "grad_norm": 0.5068960189819336,
+      "learning_rate": 8e-05,
+      "loss": 1.718,
+      "step": 439
+    },
+    {
+      "epoch": 0.09138110072689512,
+      "grad_norm": 0.46872174739837646,
+      "learning_rate": 8e-05,
+      "loss": 1.5913,
+      "step": 440
+    },
+    {
+      "epoch": 0.09158878504672897,
+      "grad_norm": 0.5093649625778198,
+      "learning_rate": 8e-05,
+      "loss": 1.7232,
+      "step": 441
+    },
+    {
+      "epoch": 0.09179646936656283,
+      "grad_norm": 0.48676973581314087,
+      "learning_rate": 8e-05,
+      "loss": 1.6004,
+      "step": 442
+    },
+    {
+      "epoch": 0.09200415368639668,
+      "grad_norm": 0.5160446763038635,
+      "learning_rate": 8e-05,
+      "loss": 1.7038,
+      "step": 443
+    },
+    {
+      "epoch": 0.09221183800623053,
+      "grad_norm": 0.4958758056163788,
+      "learning_rate": 8e-05,
+      "loss": 1.6985,
+      "step": 444
+    },
+    {
+      "epoch": 0.09241952232606439,
+      "grad_norm": 0.5180954337120056,
+      "learning_rate": 8e-05,
+      "loss": 1.636,
+      "step": 445
+    },
+    {
+      "epoch": 0.09262720664589824,
+      "grad_norm": 0.5121625661849976,
+      "learning_rate": 8e-05,
+      "loss": 1.6144,
+      "step": 446
+    },
+    {
+      "epoch": 0.09283489096573208,
+      "grad_norm": 0.49673452973365784,
+      "learning_rate": 8e-05,
+      "loss": 1.5977,
+      "step": 447
+    },
+    {
+      "epoch": 0.09304257528556593,
+      "grad_norm": 0.4997375011444092,
+      "learning_rate": 8e-05,
+      "loss": 1.686,
+      "step": 448
+    },
+    {
+      "epoch": 0.09325025960539979,
+      "grad_norm": 0.49643149971961975,
+      "learning_rate": 8e-05,
+      "loss": 1.6721,
+      "step": 449
+    },
+    {
+      "epoch": 0.09345794392523364,
+      "grad_norm": 0.4948190152645111,
+      "learning_rate": 8e-05,
+      "loss": 1.657,
+      "step": 450
+    },
+    {
+      "epoch": 0.0936656282450675,
+      "grad_norm": 0.4922389090061188,
+      "learning_rate": 8e-05,
+      "loss": 1.65,
+      "step": 451
+    },
+    {
+      "epoch": 0.09387331256490135,
+      "grad_norm": 0.5069667100906372,
+      "learning_rate": 8e-05,
+      "loss": 1.7523,
+      "step": 452
+    },
+    {
+      "epoch": 0.0940809968847352,
+      "grad_norm": 0.4961357116699219,
+      "learning_rate": 8e-05,
+      "loss": 1.752,
+      "step": 453
+    },
+    {
+      "epoch": 0.09428868120456906,
+      "grad_norm": 0.5226089358329773,
+      "learning_rate": 8e-05,
+      "loss": 1.6604,
+      "step": 454
+    },
+    {
+      "epoch": 0.09449636552440291,
+      "grad_norm": 0.47510433197021484,
+      "learning_rate": 8e-05,
+      "loss": 1.7006,
+      "step": 455
+    },
+    {
+      "epoch": 0.09470404984423676,
+      "grad_norm": 0.5109766125679016,
+      "learning_rate": 8e-05,
+      "loss": 1.7016,
+      "step": 456
+    },
+    {
+      "epoch": 0.09491173416407062,
+      "grad_norm": 0.4836071729660034,
+      "learning_rate": 8e-05,
+      "loss": 1.6091,
+      "step": 457
+    },
+    {
+      "epoch": 0.09511941848390447,
+      "grad_norm": 0.523279070854187,
+      "learning_rate": 8e-05,
+      "loss": 1.6313,
+      "step": 458
+    },
+    {
+      "epoch": 0.09532710280373832,
+      "grad_norm": 0.5089949369430542,
+      "learning_rate": 8e-05,
+      "loss": 1.6241,
+      "step": 459
+    },
+    {
+      "epoch": 0.09553478712357218,
+      "grad_norm": 0.509610652923584,
+      "learning_rate": 8e-05,
+      "loss": 1.6721,
+      "step": 460
+    },
+    {
+      "epoch": 0.09574247144340603,
+      "grad_norm": 0.4768628776073456,
+      "learning_rate": 8e-05,
+      "loss": 1.5789,
+      "step": 461
+    },
+    {
+      "epoch": 0.09595015576323987,
+      "grad_norm": 0.5115273594856262,
+      "learning_rate": 8e-05,
+      "loss": 1.7081,
+      "step": 462
+    },
+    {
+      "epoch": 0.09615784008307372,
+      "grad_norm": 0.4946316182613373,
+      "learning_rate": 8e-05,
+      "loss": 1.6407,
+      "step": 463
+    },
+    {
+      "epoch": 0.09636552440290758,
+      "grad_norm": 0.5241323113441467,
+      "learning_rate": 8e-05,
+      "loss": 1.7573,
+      "step": 464
+    },
+    {
+      "epoch": 0.09657320872274143,
+      "grad_norm": 0.47573548555374146,
+      "learning_rate": 8e-05,
+      "loss": 1.5695,
+      "step": 465
+    },
+    {
+      "epoch": 0.09678089304257528,
+      "grad_norm": 0.49498170614242554,
+      "learning_rate": 8e-05,
+      "loss": 1.6743,
+      "step": 466
+    },
+    {
+      "epoch": 0.09698857736240914,
+      "grad_norm": 0.48741909861564636,
+      "learning_rate": 8e-05,
+      "loss": 1.645,
+      "step": 467
+    },
+    {
+      "epoch": 0.09719626168224299,
+      "grad_norm": 0.498971551656723,
+      "learning_rate": 8e-05,
+      "loss": 1.6154,
+      "step": 468
+    },
+    {
+      "epoch": 0.09740394600207684,
+      "grad_norm": 0.5093749165534973,
+      "learning_rate": 8e-05,
+      "loss": 1.7089,
+      "step": 469
+    },
+    {
+      "epoch": 0.0976116303219107,
+      "grad_norm": 0.5314822793006897,
+      "learning_rate": 8e-05,
+      "loss": 1.7294,
+      "step": 470
+    },
+    {
+      "epoch": 0.09781931464174455,
+      "grad_norm": 0.5050390958786011,
+      "learning_rate": 8e-05,
+      "loss": 1.6589,
+      "step": 471
+    },
+    {
+      "epoch": 0.0980269989615784,
+      "grad_norm": 0.4838104546070099,
+      "learning_rate": 8e-05,
+      "loss": 1.6221,
+      "step": 472
+    },
+    {
+      "epoch": 0.09823468328141226,
+      "grad_norm": 0.5459195375442505,
+      "learning_rate": 8e-05,
+      "loss": 1.6315,
+      "step": 473
+    },
+    {
+      "epoch": 0.09844236760124611,
+      "grad_norm": 0.511231541633606,
+      "learning_rate": 8e-05,
+      "loss": 1.7314,
+      "step": 474
+    },
+    {
+      "epoch": 0.09865005192107996,
+      "grad_norm": 0.5203128457069397,
+      "learning_rate": 8e-05,
+      "loss": 1.6929,
+      "step": 475
+    },
+    {
+      "epoch": 0.0988577362409138,
+      "grad_norm": 0.509968101978302,
+      "learning_rate": 8e-05,
+      "loss": 1.5961,
+      "step": 476
+    },
+    {
+      "epoch": 0.09906542056074766,
+      "grad_norm": 0.45741814374923706,
+      "learning_rate": 8e-05,
+      "loss": 1.5686,
+      "step": 477
+    },
+    {
+      "epoch": 0.09927310488058151,
+      "grad_norm": 0.49833589792251587,
+      "learning_rate": 8e-05,
+      "loss": 1.6054,
+      "step": 478
+    },
+    {
+      "epoch": 0.09948078920041536,
+      "grad_norm": 0.4866011142730713,
+      "learning_rate": 8e-05,
+      "loss": 1.5743,
+      "step": 479
+    },
+    {
+      "epoch": 0.09968847352024922,
+      "grad_norm": 0.48203375935554504,
+      "learning_rate": 8e-05,
+      "loss": 1.6288,
+      "step": 480
+    },
+    {
+      "epoch": 0.09989615784008307,
+      "grad_norm": 0.48031631112098694,
+      "learning_rate": 8e-05,
+      "loss": 1.6257,
+      "step": 481
+    },
+    {
+      "epoch": 0.10010384215991693,
+      "grad_norm": 0.4775748550891876,
+      "learning_rate": 8e-05,
+      "loss": 1.582,
+      "step": 482
+    },
+    {
+      "epoch": 0.10031152647975078,
+      "grad_norm": 0.5151557326316833,
+      "learning_rate": 8e-05,
+      "loss": 1.6394,
+      "step": 483
+    },
+    {
+      "epoch": 0.10051921079958463,
+      "grad_norm": 0.47847479581832886,
+      "learning_rate": 8e-05,
+      "loss": 1.6036,
+      "step": 484
+    },
+    {
+      "epoch": 0.10072689511941849,
+      "grad_norm": 0.5062543153762817,
+      "learning_rate": 8e-05,
+      "loss": 1.6405,
+      "step": 485
+    },
+    {
+      "epoch": 0.10093457943925234,
+      "grad_norm": 0.46788135170936584,
+      "learning_rate": 8e-05,
+      "loss": 1.6106,
+      "step": 486
+    },
+    {
+      "epoch": 0.10114226375908619,
+      "grad_norm": 0.47770899534225464,
+      "learning_rate": 8e-05,
+      "loss": 1.6511,
+      "step": 487
+    },
+    {
+      "epoch": 0.10134994807892005,
+      "grad_norm": 0.5121338963508606,
+      "learning_rate": 8e-05,
+      "loss": 1.6954,
+      "step": 488
+    },
+    {
+      "epoch": 0.1015576323987539,
+      "grad_norm": 0.5268594026565552,
+      "learning_rate": 8e-05,
+      "loss": 1.7471,
+      "step": 489
+    },
+    {
+      "epoch": 0.10176531671858775,
+      "grad_norm": 0.4916820526123047,
+      "learning_rate": 8e-05,
+      "loss": 1.6428,
+      "step": 490
+    },
+    {
+      "epoch": 0.10197300103842159,
+      "grad_norm": 0.4892474412918091,
+      "learning_rate": 8e-05,
+      "loss": 1.6466,
+      "step": 491
+    },
+    {
+      "epoch": 0.10218068535825545,
+      "grad_norm": 0.48951107263565063,
+      "learning_rate": 8e-05,
+      "loss": 1.7112,
+      "step": 492
+    },
+    {
+      "epoch": 0.1023883696780893,
+      "grad_norm": 0.4870230555534363,
+      "learning_rate": 8e-05,
+      "loss": 1.7277,
+      "step": 493
+    },
+    {
+      "epoch": 0.10259605399792315,
+      "grad_norm": 0.5075536370277405,
+      "learning_rate": 8e-05,
+      "loss": 1.6364,
+      "step": 494
+    },
+    {
+      "epoch": 0.102803738317757,
+      "grad_norm": 0.5025159120559692,
+      "learning_rate": 8e-05,
+      "loss": 1.6866,
+      "step": 495
+    },
+    {
+      "epoch": 0.10301142263759086,
+      "grad_norm": 0.49492985010147095,
+      "learning_rate": 8e-05,
+      "loss": 1.6144,
+      "step": 496
+    },
+    {
+      "epoch": 0.10321910695742471,
+      "grad_norm": 0.48841235041618347,
+      "learning_rate": 8e-05,
+      "loss": 1.6522,
+      "step": 497
+    },
+    {
+      "epoch": 0.10342679127725857,
+      "grad_norm": 0.47402575612068176,
+      "learning_rate": 8e-05,
+      "loss": 1.5525,
+      "step": 498
+    },
+    {
+      "epoch": 0.10363447559709242,
+      "grad_norm": 0.4916023910045624,
+      "learning_rate": 8e-05,
+      "loss": 1.6744,
+      "step": 499
+    },
+    {
+      "epoch": 0.10384215991692627,
+      "grad_norm": 0.4725121259689331,
+      "learning_rate": 8e-05,
+      "loss": 1.5623,
+      "step": 500
+    },
+    {
+      "epoch": 0.10404984423676013,
+      "grad_norm": 0.4987272322177887,
+      "learning_rate": 8e-05,
+      "loss": 1.6302,
+      "step": 501
+    },
+    {
+      "epoch": 0.10425752855659398,
+      "grad_norm": 0.48811638355255127,
+      "learning_rate": 8e-05,
+      "loss": 1.6124,
+      "step": 502
+    },
+    {
+      "epoch": 0.10446521287642783,
+      "grad_norm": 0.5044394731521606,
+      "learning_rate": 8e-05,
+      "loss": 1.7075,
+      "step": 503
+    },
+    {
+      "epoch": 0.10467289719626169,
+      "grad_norm": 0.4859680235385895,
+      "learning_rate": 8e-05,
+      "loss": 1.6975,
+      "step": 504
+    },
+    {
+      "epoch": 0.10488058151609553,
+      "grad_norm": 0.528005063533783,
+      "learning_rate": 8e-05,
+      "loss": 1.6129,
+      "step": 505
+    },
+    {
+      "epoch": 0.10508826583592938,
+      "grad_norm": 0.5075289011001587,
+      "learning_rate": 8e-05,
+      "loss": 1.7663,
+      "step": 506
+    },
+    {
+      "epoch": 0.10529595015576323,
+      "grad_norm": 0.5202328562736511,
+      "learning_rate": 8e-05,
+      "loss": 1.685,
+      "step": 507
+    },
+    {
+      "epoch": 0.10550363447559709,
+      "grad_norm": 0.47399210929870605,
+      "learning_rate": 8e-05,
+      "loss": 1.612,
+      "step": 508
+    },
+    {
+      "epoch": 0.10571131879543094,
+      "grad_norm": 0.48944804072380066,
+      "learning_rate": 8e-05,
+      "loss": 1.5956,
+      "step": 509
+    },
+    {
+      "epoch": 0.1059190031152648,
+      "grad_norm": 0.5143656730651855,
+      "learning_rate": 8e-05,
+      "loss": 1.6964,
+      "step": 510
+    },
+    {
+      "epoch": 0.10612668743509865,
+      "grad_norm": 0.4810107350349426,
+      "learning_rate": 8e-05,
+      "loss": 1.6548,
+      "step": 511
+    },
+    {
+      "epoch": 0.1063343717549325,
+      "grad_norm": 0.48855581879615784,
+      "learning_rate": 8e-05,
+      "loss": 1.637,
+      "step": 512
+    },
+    {
+      "epoch": 0.10654205607476636,
+      "grad_norm": 0.477912038564682,
+      "learning_rate": 8e-05,
+      "loss": 1.6137,
+      "step": 513
+    },
+    {
+      "epoch": 0.10674974039460021,
+      "grad_norm": 0.49473705887794495,
+      "learning_rate": 8e-05,
+      "loss": 1.6657,
+      "step": 514
+    },
+    {
+      "epoch": 0.10695742471443406,
+      "grad_norm": 0.4874957501888275,
+      "learning_rate": 8e-05,
+      "loss": 1.5984,
+      "step": 515
+    },
+    {
+      "epoch": 0.10716510903426792,
+      "grad_norm": 0.47925207018852234,
+      "learning_rate": 8e-05,
+      "loss": 1.6142,
+      "step": 516
+    },
+    {
+      "epoch": 0.10737279335410177,
+      "grad_norm": 0.49199312925338745,
+      "learning_rate": 8e-05,
+      "loss": 1.6706,
+      "step": 517
+    },
+    {
+      "epoch": 0.10758047767393562,
+      "grad_norm": 0.4763036072254181,
+      "learning_rate": 8e-05,
+      "loss": 1.6835,
+      "step": 518
+    },
+    {
+      "epoch": 0.10778816199376948,
+      "grad_norm": 0.5105186700820923,
+      "learning_rate": 8e-05,
+      "loss": 1.6513,
+      "step": 519
+    },
+    {
+      "epoch": 0.10799584631360332,
+      "grad_norm": 0.48001575469970703,
+      "learning_rate": 8e-05,
+      "loss": 1.6549,
+      "step": 520
+    },
+    {
+      "epoch": 0.10820353063343717,
+      "grad_norm": 0.5005422830581665,
+      "learning_rate": 8e-05,
+      "loss": 1.6499,
+      "step": 521
+    },
+    {
+      "epoch": 0.10841121495327102,
+      "grad_norm": 0.4944564402103424,
+      "learning_rate": 8e-05,
+      "loss": 1.6491,
+      "step": 522
+    },
+    {
+      "epoch": 0.10861889927310488,
+      "grad_norm": 0.47744297981262207,
+      "learning_rate": 8e-05,
+      "loss": 1.5596,
+      "step": 523
+    },
+    {
+      "epoch": 0.10882658359293873,
+      "grad_norm": 0.4701041877269745,
+      "learning_rate": 8e-05,
+      "loss": 1.6377,
+      "step": 524
+    },
+    {
+      "epoch": 0.10903426791277258,
+      "grad_norm": 0.5091982483863831,
+      "learning_rate": 8e-05,
+      "loss": 1.7002,
+      "step": 525
+    },
+    {
+      "epoch": 0.10924195223260644,
+      "grad_norm": 0.4894920587539673,
+      "learning_rate": 8e-05,
+      "loss": 1.6587,
+      "step": 526
+    },
+    {
+      "epoch": 0.10944963655244029,
+      "grad_norm": 0.48994290828704834,
+      "learning_rate": 8e-05,
+      "loss": 1.6763,
+      "step": 527
+    },
+    {
+      "epoch": 0.10965732087227414,
+      "grad_norm": 0.49817031621932983,
+      "learning_rate": 8e-05,
+      "loss": 1.6896,
+      "step": 528
+    },
+    {
+      "epoch": 0.109865005192108,
+      "grad_norm": 0.4927424490451813,
+      "learning_rate": 8e-05,
+      "loss": 1.634,
+      "step": 529
+    },
+    {
+      "epoch": 0.11007268951194185,
+      "grad_norm": 0.46971508860588074,
+      "learning_rate": 8e-05,
+      "loss": 1.6402,
+      "step": 530
+    },
+    {
+      "epoch": 0.1102803738317757,
+      "grad_norm": 0.5013573169708252,
+      "learning_rate": 8e-05,
+      "loss": 1.6279,
+      "step": 531
+    },
+    {
+      "epoch": 0.11048805815160956,
+      "grad_norm": 0.494057297706604,
+      "learning_rate": 8e-05,
+      "loss": 1.6284,
+      "step": 532
+    },
+    {
+      "epoch": 0.11069574247144341,
+      "grad_norm": 0.49397385120391846,
+      "learning_rate": 8e-05,
+      "loss": 1.6885,
+      "step": 533
+    },
+    {
+      "epoch": 0.11090342679127727,
+      "grad_norm": 0.5148130059242249,
+      "learning_rate": 8e-05,
+      "loss": 1.6592,
+      "step": 534
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 0.4772680401802063,
+      "learning_rate": 8e-05,
+      "loss": 1.6081,
+      "step": 535
+    },
+    {
+      "epoch": 0.11131879543094496,
+      "grad_norm": 0.5171798467636108,
+      "learning_rate": 8e-05,
+      "loss": 1.7278,
+      "step": 536
+    },
+    {
+      "epoch": 0.11152647975077881,
+      "grad_norm": 0.47712597250938416,
+      "learning_rate": 8e-05,
+      "loss": 1.6092,
+      "step": 537
+    },
+    {
+      "epoch": 0.11173416407061267,
+      "grad_norm": 0.47499656677246094,
+      "learning_rate": 8e-05,
+      "loss": 1.598,
+      "step": 538
+    },
+    {
+      "epoch": 0.11194184839044652,
+      "grad_norm": 0.4703918397426605,
+      "learning_rate": 8e-05,
+      "loss": 1.6449,
+      "step": 539
+    },
+    {
+      "epoch": 0.11214953271028037,
+      "grad_norm": 0.5092900991439819,
+      "learning_rate": 8e-05,
+      "loss": 1.6808,
+      "step": 540
+    },
+    {
+      "epoch": 0.11235721703011423,
+      "grad_norm": 0.48537686467170715,
+      "learning_rate": 8e-05,
+      "loss": 1.6026,
+      "step": 541
+    },
+    {
+      "epoch": 0.11256490134994808,
+      "grad_norm": 0.49423104524612427,
+      "learning_rate": 8e-05,
+      "loss": 1.7332,
+      "step": 542
+    },
+    {
+      "epoch": 0.11277258566978193,
+      "grad_norm": 0.48707374930381775,
+      "learning_rate": 8e-05,
+      "loss": 1.6686,
+      "step": 543
+    },
+    {
+      "epoch": 0.11298026998961579,
+      "grad_norm": 0.4906916618347168,
+      "learning_rate": 8e-05,
+      "loss": 1.6518,
+      "step": 544
+    },
+    {
+      "epoch": 0.11318795430944964,
+      "grad_norm": 0.4747893512248993,
+      "learning_rate": 8e-05,
+      "loss": 1.6452,
+      "step": 545
+    },
+    {
+      "epoch": 0.1133956386292835,
+      "grad_norm": 0.4864533245563507,
+      "learning_rate": 8e-05,
+      "loss": 1.6477,
+      "step": 546
+    },
+    {
+      "epoch": 0.11360332294911735,
+      "grad_norm": 0.469113826751709,
+      "learning_rate": 8e-05,
+      "loss": 1.5731,
+      "step": 547
+    },
+    {
+      "epoch": 0.1138110072689512,
+      "grad_norm": 0.48744997382164,
+      "learning_rate": 8e-05,
+      "loss": 1.6906,
+      "step": 548
+    },
+    {
+      "epoch": 0.11401869158878504,
+      "grad_norm": 0.4924449920654297,
+      "learning_rate": 8e-05,
+      "loss": 1.5568,
+      "step": 549
+    },
+    {
+      "epoch": 0.1142263759086189,
+      "grad_norm": 0.46387964487075806,
+      "learning_rate": 8e-05,
+      "loss": 1.6574,
+      "step": 550
+    },
+    {
+      "epoch": 0.11443406022845275,
+      "grad_norm": 0.49000611901283264,
+      "learning_rate": 8e-05,
+      "loss": 1.6969,
+      "step": 551
+    },
+    {
+      "epoch": 0.1146417445482866,
+      "grad_norm": 0.5043030977249146,
+      "learning_rate": 8e-05,
+      "loss": 1.6773,
+      "step": 552
+    },
+    {
+      "epoch": 0.11484942886812045,
+      "grad_norm": 0.5181741118431091,
+      "learning_rate": 8e-05,
+      "loss": 1.6705,
+      "step": 553
+    },
+    {
+      "epoch": 0.11505711318795431,
+      "grad_norm": 0.517162561416626,
+      "learning_rate": 8e-05,
+      "loss": 1.7016,
+      "step": 554
+    },
+    {
+      "epoch": 0.11526479750778816,
+      "grad_norm": 0.5007038116455078,
+      "learning_rate": 8e-05,
+      "loss": 1.6708,
+      "step": 555
+    },
+    {
+      "epoch": 0.11547248182762201,
+      "grad_norm": 0.49101993441581726,
+      "learning_rate": 8e-05,
+      "loss": 1.6829,
+      "step": 556
+    },
+    {
+      "epoch": 0.11568016614745587,
+      "grad_norm": 0.47777554392814636,
+      "learning_rate": 8e-05,
+      "loss": 1.6972,
+      "step": 557
+    },
+    {
+      "epoch": 0.11588785046728972,
+      "grad_norm": 0.5146732330322266,
+      "learning_rate": 8e-05,
+      "loss": 1.6615,
+      "step": 558
+    },
+    {
+      "epoch": 0.11609553478712357,
+      "grad_norm": 0.4889799654483795,
+      "learning_rate": 8e-05,
+      "loss": 1.6389,
+      "step": 559
+    },
+    {
+      "epoch": 0.11630321910695743,
+      "grad_norm": 0.48216649889945984,
+      "learning_rate": 8e-05,
+      "loss": 1.7034,
+      "step": 560
+    },
+    {
+      "epoch": 0.11651090342679128,
+      "grad_norm": 0.466871052980423,
+      "learning_rate": 8e-05,
+      "loss": 1.5976,
+      "step": 561
+    },
+    {
+      "epoch": 0.11671858774662514,
+      "grad_norm": 0.4961405396461487,
+      "learning_rate": 8e-05,
+      "loss": 1.6173,
+      "step": 562
+    },
+    {
+      "epoch": 0.11692627206645899,
+      "grad_norm": 0.5158309936523438,
+      "learning_rate": 8e-05,
+      "loss": 1.7243,
+      "step": 563
+    },
+    {
+      "epoch": 0.11713395638629283,
+      "grad_norm": 0.5011987686157227,
+      "learning_rate": 8e-05,
+      "loss": 1.6861,
+      "step": 564
+    },
+    {
+      "epoch": 0.11734164070612668,
+      "grad_norm": 0.535325825214386,
+      "learning_rate": 8e-05,
+      "loss": 1.799,
+      "step": 565
+    },
+    {
+      "epoch": 0.11754932502596054,
+      "grad_norm": 0.4968489110469818,
+      "learning_rate": 8e-05,
+      "loss": 1.7096,
+      "step": 566
+    },
+    {
+      "epoch": 0.11775700934579439,
+      "grad_norm": 0.4858616292476654,
+      "learning_rate": 8e-05,
+      "loss": 1.6617,
+      "step": 567
+    },
+    {
+      "epoch": 0.11796469366562824,
+      "grad_norm": 0.49563148617744446,
+      "learning_rate": 8e-05,
+      "loss": 1.6681,
+      "step": 568
+    },
+    {
+      "epoch": 0.1181723779854621,
+      "grad_norm": 0.49028831720352173,
+      "learning_rate": 8e-05,
+      "loss": 1.6425,
+      "step": 569
+    },
+    {
+      "epoch": 0.11838006230529595,
+      "grad_norm": 0.4926828145980835,
+      "learning_rate": 8e-05,
+      "loss": 1.6134,
+      "step": 570
+    },
+    {
+      "epoch": 0.1185877466251298,
+      "grad_norm": 0.4846639037132263,
+      "learning_rate": 8e-05,
+      "loss": 1.6194,
+      "step": 571
+    },
+    {
+      "epoch": 0.11879543094496366,
+      "grad_norm": 0.4756839871406555,
+      "learning_rate": 8e-05,
+      "loss": 1.6454,
+      "step": 572
+    },
+    {
+      "epoch": 0.11900311526479751,
+      "grad_norm": 0.4722893536090851,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 573
+    },
+    {
+      "epoch": 0.11921079958463136,
+      "grad_norm": 0.4920216202735901,
+      "learning_rate": 8e-05,
+      "loss": 1.6916,
+      "step": 574
+    },
+    {
+      "epoch": 0.11941848390446522,
+      "grad_norm": 0.48332494497299194,
+      "learning_rate": 8e-05,
+      "loss": 1.6414,
+      "step": 575
+    },
+    {
+      "epoch": 0.11962616822429907,
+      "grad_norm": 0.5009651780128479,
+      "learning_rate": 8e-05,
+      "loss": 1.6119,
+      "step": 576
+    },
+    {
+      "epoch": 0.11983385254413292,
+      "grad_norm": 0.4851033389568329,
+      "learning_rate": 8e-05,
+      "loss": 1.6018,
+      "step": 577
+    },
+    {
+      "epoch": 0.12004153686396678,
+      "grad_norm": 0.4941191077232361,
+      "learning_rate": 8e-05,
+      "loss": 1.6982,
+      "step": 578
+    },
+    {
+      "epoch": 0.12024922118380062,
+      "grad_norm": 0.4699910879135132,
+      "learning_rate": 8e-05,
+      "loss": 1.6004,
+      "step": 579
+    },
+    {
+      "epoch": 0.12045690550363447,
+      "grad_norm": 0.47666141390800476,
+      "learning_rate": 8e-05,
+      "loss": 1.6523,
+      "step": 580
+    },
+    {
+      "epoch": 0.12066458982346832,
+      "grad_norm": 0.5250875949859619,
+      "learning_rate": 8e-05,
+      "loss": 1.7131,
+      "step": 581
+    },
+    {
+      "epoch": 0.12087227414330218,
+      "grad_norm": 0.47818005084991455,
+      "learning_rate": 8e-05,
+      "loss": 1.6228,
+      "step": 582
+    },
+    {
+      "epoch": 0.12107995846313603,
+      "grad_norm": 0.46727368235588074,
+      "learning_rate": 8e-05,
+      "loss": 1.5709,
+      "step": 583
+    },
+    {
+      "epoch": 0.12128764278296988,
+      "grad_norm": 0.4783129394054413,
+      "learning_rate": 8e-05,
+      "loss": 1.6809,
+      "step": 584
+    },
+    {
+      "epoch": 0.12149532710280374,
+      "grad_norm": 0.48866942524909973,
+      "learning_rate": 8e-05,
+      "loss": 1.6795,
+      "step": 585
+    },
+    {
+      "epoch": 0.12170301142263759,
+      "grad_norm": 0.4863775968551636,
+      "learning_rate": 8e-05,
+      "loss": 1.6499,
+      "step": 586
+    },
+    {
+      "epoch": 0.12191069574247144,
+      "grad_norm": 0.48488086462020874,
+      "learning_rate": 8e-05,
+      "loss": 1.6765,
+      "step": 587
+    },
+    {
+      "epoch": 0.1221183800623053,
+      "grad_norm": 0.5019093751907349,
+      "learning_rate": 8e-05,
+      "loss": 1.6547,
+      "step": 588
+    },
+    {
+      "epoch": 0.12232606438213915,
+      "grad_norm": 0.484720379114151,
+      "learning_rate": 8e-05,
+      "loss": 1.6438,
+      "step": 589
+    },
+    {
+      "epoch": 0.122533748701973,
+      "grad_norm": 0.48393747210502625,
+      "learning_rate": 8e-05,
+      "loss": 1.6595,
+      "step": 590
+    },
+    {
+      "epoch": 0.12274143302180686,
+      "grad_norm": 0.5197237133979797,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 591
+    },
+    {
+      "epoch": 0.12294911734164071,
+      "grad_norm": 0.4648591876029968,
+      "learning_rate": 8e-05,
+      "loss": 1.6076,
+      "step": 592
+    },
+    {
+      "epoch": 0.12315680166147455,
+      "grad_norm": 0.4920279085636139,
+      "learning_rate": 8e-05,
+      "loss": 1.6787,
+      "step": 593
+    },
+    {
+      "epoch": 0.1233644859813084,
+      "grad_norm": 0.48412156105041504,
+      "learning_rate": 8e-05,
+      "loss": 1.6896,
+      "step": 594
+    },
+    {
+      "epoch": 0.12357217030114226,
+      "grad_norm": 0.4906349778175354,
+      "learning_rate": 8e-05,
+      "loss": 1.6746,
+      "step": 595
+    },
+    {
+      "epoch": 0.12377985462097611,
+      "grad_norm": 0.49094000458717346,
+      "learning_rate": 8e-05,
+      "loss": 1.6618,
+      "step": 596
+    },
+    {
+      "epoch": 0.12398753894080997,
+      "grad_norm": 0.48553135991096497,
+      "learning_rate": 8e-05,
+      "loss": 1.7193,
+      "step": 597
+    },
+    {
+      "epoch": 0.12419522326064382,
+      "grad_norm": 0.48377707600593567,
+      "learning_rate": 8e-05,
+      "loss": 1.6537,
+      "step": 598
+    },
+    {
+      "epoch": 0.12440290758047767,
+      "grad_norm": 0.47426944971084595,
+      "learning_rate": 8e-05,
+      "loss": 1.6248,
+      "step": 599
+    },
+    {
+      "epoch": 0.12461059190031153,
+      "grad_norm": 0.4826415777206421,
+      "learning_rate": 8e-05,
+      "loss": 1.6596,
+      "step": 600
+    },
+    {
+      "epoch": 0.12481827622014538,
+      "grad_norm": 0.49226781725883484,
+      "learning_rate": 8e-05,
+      "loss": 1.7157,
+      "step": 601
+    },
+    {
+      "epoch": 0.12502596053997922,
+      "grad_norm": 0.46622025966644287,
+      "learning_rate": 8e-05,
+      "loss": 1.61,
+      "step": 602
+    },
+    {
+      "epoch": 0.1252336448598131,
+      "grad_norm": 0.4895409345626831,
+      "learning_rate": 8e-05,
+      "loss": 1.6629,
+      "step": 603
+    },
+    {
+      "epoch": 0.12544132917964693,
+      "grad_norm": 0.4662224352359772,
+      "learning_rate": 8e-05,
+      "loss": 1.6188,
+      "step": 604
+    },
+    {
+      "epoch": 0.1256490134994808,
+      "grad_norm": 0.4780591130256653,
+      "learning_rate": 8e-05,
+      "loss": 1.5856,
+      "step": 605
+    },
+    {
+      "epoch": 0.12585669781931463,
+      "grad_norm": 0.49765658378601074,
+      "learning_rate": 8e-05,
+      "loss": 1.613,
+      "step": 606
+    },
+    {
+      "epoch": 0.1260643821391485,
+      "grad_norm": 0.4997657239437103,
+      "learning_rate": 8e-05,
+      "loss": 1.6758,
+      "step": 607
+    },
+    {
+      "epoch": 0.12627206645898234,
+      "grad_norm": 0.48155564069747925,
+      "learning_rate": 8e-05,
+      "loss": 1.6179,
+      "step": 608
+    },
+    {
+      "epoch": 0.1264797507788162,
+      "grad_norm": 0.4809686541557312,
+      "learning_rate": 8e-05,
+      "loss": 1.5715,
+      "step": 609
+    },
+    {
+      "epoch": 0.12668743509865005,
+      "grad_norm": 0.5093618035316467,
+      "learning_rate": 8e-05,
+      "loss": 1.7242,
+      "step": 610
+    },
+    {
+      "epoch": 0.12689511941848391,
+      "grad_norm": 0.4809294641017914,
+      "learning_rate": 8e-05,
+      "loss": 1.6881,
+      "step": 611
+    },
+    {
+      "epoch": 0.12710280373831775,
+      "grad_norm": 0.4776981472969055,
+      "learning_rate": 8e-05,
+      "loss": 1.6183,
+      "step": 612
+    },
+    {
+      "epoch": 0.12731048805815162,
+      "grad_norm": 0.4951392412185669,
+      "learning_rate": 8e-05,
+      "loss": 1.7086,
+      "step": 613
+    },
+    {
+      "epoch": 0.12751817237798546,
+      "grad_norm": 0.48191651701927185,
+      "learning_rate": 8e-05,
+      "loss": 1.6385,
+      "step": 614
+    },
+    {
+      "epoch": 0.1277258566978193,
+      "grad_norm": 0.4886009693145752,
+      "learning_rate": 8e-05,
+      "loss": 1.618,
+      "step": 615
+    },
+    {
+      "epoch": 0.12793354101765317,
+      "grad_norm": 0.49545979499816895,
+      "learning_rate": 8e-05,
+      "loss": 1.6361,
+      "step": 616
+    },
+    {
+      "epoch": 0.128141225337487,
+      "grad_norm": 0.49346309900283813,
+      "learning_rate": 8e-05,
+      "loss": 1.693,
+      "step": 617
+    },
+    {
+      "epoch": 0.12834890965732088,
+      "grad_norm": 0.5284665822982788,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 618
+    },
+    {
+      "epoch": 0.12855659397715472,
+      "grad_norm": 0.5081282258033752,
+      "learning_rate": 8e-05,
+      "loss": 1.6581,
+      "step": 619
+    },
+    {
+      "epoch": 0.12876427829698858,
+      "grad_norm": 0.5118349194526672,
+      "learning_rate": 8e-05,
+      "loss": 1.7038,
+      "step": 620
+    },
+    {
+      "epoch": 0.12897196261682242,
+      "grad_norm": 0.49056780338287354,
+      "learning_rate": 8e-05,
+      "loss": 1.7102,
+      "step": 621
+    },
+    {
+      "epoch": 0.1291796469366563,
+      "grad_norm": 0.4869656562805176,
+      "learning_rate": 8e-05,
+      "loss": 1.619,
+      "step": 622
+    },
+    {
+      "epoch": 0.12938733125649013,
+      "grad_norm": 0.5157124996185303,
+      "learning_rate": 8e-05,
+      "loss": 1.6398,
+      "step": 623
+    },
+    {
+      "epoch": 0.129595015576324,
+      "grad_norm": 0.48705726861953735,
+      "learning_rate": 8e-05,
+      "loss": 1.6655,
+      "step": 624
+    },
+    {
+      "epoch": 0.12980269989615784,
+      "grad_norm": 0.46967780590057373,
+      "learning_rate": 8e-05,
+      "loss": 1.6131,
+      "step": 625
+    },
+    {
+      "epoch": 0.1300103842159917,
+      "grad_norm": 0.47689181566238403,
+      "learning_rate": 8e-05,
+      "loss": 1.6086,
+      "step": 626
+    },
+    {
+      "epoch": 0.13021806853582554,
+      "grad_norm": 0.48610159754753113,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 627
+    },
+    {
+      "epoch": 0.1304257528556594,
+      "grad_norm": 0.4769737124443054,
+      "learning_rate": 8e-05,
+      "loss": 1.57,
+      "step": 628
+    },
+    {
+      "epoch": 0.13063343717549325,
+      "grad_norm": 0.47366201877593994,
+      "learning_rate": 8e-05,
+      "loss": 1.6771,
+      "step": 629
+    },
+    {
+      "epoch": 0.1308411214953271,
+      "grad_norm": 0.5110802054405212,
+      "learning_rate": 8e-05,
+      "loss": 1.6377,
+      "step": 630
+    },
+    {
+      "epoch": 0.13104880581516096,
+      "grad_norm": 0.5230108499526978,
+      "learning_rate": 8e-05,
+      "loss": 1.6979,
+      "step": 631
+    },
+    {
+      "epoch": 0.1312564901349948,
+      "grad_norm": 0.4913478493690491,
+      "learning_rate": 8e-05,
+      "loss": 1.6687,
+      "step": 632
+    },
+    {
+      "epoch": 0.13146417445482866,
+      "grad_norm": 0.5109069347381592,
+      "learning_rate": 8e-05,
+      "loss": 1.6655,
+      "step": 633
+    },
+    {
+      "epoch": 0.1316718587746625,
+      "grad_norm": 0.5116170644760132,
+      "learning_rate": 8e-05,
+      "loss": 1.7231,
+      "step": 634
+    },
+    {
+      "epoch": 0.13187954309449637,
+      "grad_norm": 0.5041365623474121,
+      "learning_rate": 8e-05,
+      "loss": 1.6218,
+      "step": 635
+    },
+    {
+      "epoch": 0.1320872274143302,
+      "grad_norm": 0.4839491546154022,
+      "learning_rate": 8e-05,
+      "loss": 1.6518,
+      "step": 636
+    },
+    {
+      "epoch": 0.13229491173416408,
+      "grad_norm": 0.4792550802230835,
+      "learning_rate": 8e-05,
+      "loss": 1.6506,
+      "step": 637
+    },
+    {
+      "epoch": 0.13250259605399792,
+      "grad_norm": 0.48295944929122925,
+      "learning_rate": 8e-05,
+      "loss": 1.6901,
+      "step": 638
+    },
+    {
+      "epoch": 0.13271028037383178,
+      "grad_norm": 0.4973006844520569,
+      "learning_rate": 8e-05,
+      "loss": 1.6432,
+      "step": 639
+    },
+    {
+      "epoch": 0.13291796469366562,
+      "grad_norm": 0.4931349456310272,
+      "learning_rate": 8e-05,
+      "loss": 1.6903,
+      "step": 640
+    },
+    {
+      "epoch": 0.1331256490134995,
+      "grad_norm": 0.5359208583831787,
+      "learning_rate": 8e-05,
+      "loss": 1.6107,
+      "step": 641
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 0.5108036994934082,
+      "learning_rate": 8e-05,
+      "loss": 1.743,
+      "step": 642
+    },
+    {
+      "epoch": 0.1335410176531672,
+      "grad_norm": 0.53212970495224,
+      "learning_rate": 8e-05,
+      "loss": 1.6529,
+      "step": 643
+    },
+    {
+      "epoch": 0.13374870197300104,
+      "grad_norm": 0.4618084132671356,
+      "learning_rate": 8e-05,
+      "loss": 1.6473,
+      "step": 644
+    },
+    {
+      "epoch": 0.13395638629283488,
+      "grad_norm": 0.4782806932926178,
+      "learning_rate": 8e-05,
+      "loss": 1.6267,
+      "step": 645
+    },
+    {
+      "epoch": 0.13416407061266875,
+      "grad_norm": 0.5155082941055298,
+      "learning_rate": 8e-05,
+      "loss": 1.6114,
+      "step": 646
+    },
+    {
+      "epoch": 0.13437175493250259,
+      "grad_norm": 0.5027504563331604,
+      "learning_rate": 8e-05,
+      "loss": 1.6559,
+      "step": 647
+    },
+    {
+      "epoch": 0.13457943925233645,
+      "grad_norm": 0.5325614213943481,
+      "learning_rate": 8e-05,
+      "loss": 1.5957,
+      "step": 648
+    },
+    {
+      "epoch": 0.1347871235721703,
+      "grad_norm": 0.4770219922065735,
+      "learning_rate": 8e-05,
+      "loss": 1.5635,
+      "step": 649
+    },
+    {
+      "epoch": 0.13499480789200416,
+      "grad_norm": 0.4837128818035126,
+      "learning_rate": 8e-05,
+      "loss": 1.5886,
+      "step": 650
+    },
+    {
+      "epoch": 0.135202492211838,
+      "grad_norm": 0.5071947574615479,
+      "learning_rate": 8e-05,
+      "loss": 1.7164,
+      "step": 651
+    },
+    {
+      "epoch": 0.13541017653167187,
+      "grad_norm": 0.48948243260383606,
+      "learning_rate": 8e-05,
+      "loss": 1.6695,
+      "step": 652
+    },
+    {
+      "epoch": 0.1356178608515057,
+      "grad_norm": 0.49877387285232544,
+      "learning_rate": 8e-05,
+      "loss": 1.6044,
+      "step": 653
+    },
+    {
+      "epoch": 0.13582554517133957,
+      "grad_norm": 0.48378750681877136,
+      "learning_rate": 8e-05,
+      "loss": 1.6118,
+      "step": 654
+    },
+    {
+      "epoch": 0.1360332294911734,
+      "grad_norm": 0.49697884917259216,
+      "learning_rate": 8e-05,
+      "loss": 1.6245,
+      "step": 655
+    },
+    {
+      "epoch": 0.13624091381100728,
+      "grad_norm": 0.5250710248947144,
+      "learning_rate": 8e-05,
+      "loss": 1.5699,
+      "step": 656
+    },
+    {
+      "epoch": 0.13644859813084112,
+      "grad_norm": 0.49164098501205444,
+      "learning_rate": 8e-05,
+      "loss": 1.7009,
+      "step": 657
+    },
+    {
+      "epoch": 0.136656282450675,
+      "grad_norm": 0.51181560754776,
+      "learning_rate": 8e-05,
+      "loss": 1.6749,
+      "step": 658
+    },
+    {
+      "epoch": 0.13686396677050883,
+      "grad_norm": 0.4895937442779541,
+      "learning_rate": 8e-05,
+      "loss": 1.6471,
+      "step": 659
+    },
+    {
+      "epoch": 0.13707165109034267,
+      "grad_norm": 0.500437319278717,
+      "learning_rate": 8e-05,
+      "loss": 1.6039,
+      "step": 660
+    },
+    {
+      "epoch": 0.13727933541017653,
+      "grad_norm": 0.49318140745162964,
+      "learning_rate": 8e-05,
+      "loss": 1.6222,
+      "step": 661
+    },
+    {
+      "epoch": 0.13748701973001037,
+      "grad_norm": 0.47766220569610596,
+      "learning_rate": 8e-05,
+      "loss": 1.6307,
+      "step": 662
+    },
+    {
+      "epoch": 0.13769470404984424,
+      "grad_norm": 0.526156485080719,
+      "learning_rate": 8e-05,
+      "loss": 1.6572,
+      "step": 663
+    },
+    {
+      "epoch": 0.13790238836967808,
+      "grad_norm": 0.48735764622688293,
+      "learning_rate": 8e-05,
+      "loss": 1.6393,
+      "step": 664
+    },
+    {
+      "epoch": 0.13811007268951195,
+      "grad_norm": 0.49642348289489746,
+      "learning_rate": 8e-05,
+      "loss": 1.6371,
+      "step": 665
+    },
+    {
+      "epoch": 0.1383177570093458,
+      "grad_norm": 0.5099259614944458,
+      "learning_rate": 8e-05,
+      "loss": 1.6536,
+      "step": 666
+    },
+    {
+      "epoch": 0.13852544132917965,
+      "grad_norm": 0.5075072646141052,
+      "learning_rate": 8e-05,
+      "loss": 1.6389,
+      "step": 667
+    },
+    {
+      "epoch": 0.1387331256490135,
+      "grad_norm": 0.49564462900161743,
+      "learning_rate": 8e-05,
+      "loss": 1.6303,
+      "step": 668
+    },
+    {
+      "epoch": 0.13894080996884736,
+      "grad_norm": 0.468453049659729,
+      "learning_rate": 8e-05,
+      "loss": 1.6019,
+      "step": 669
+    },
+    {
+      "epoch": 0.1391484942886812,
+      "grad_norm": 0.4859846234321594,
+      "learning_rate": 8e-05,
+      "loss": 1.6411,
+      "step": 670
+    },
+    {
+      "epoch": 0.13935617860851507,
+      "grad_norm": 0.468565970659256,
+      "learning_rate": 8e-05,
+      "loss": 1.5784,
+      "step": 671
+    },
+    {
+      "epoch": 0.1395638629283489,
+      "grad_norm": 0.4832054674625397,
+      "learning_rate": 8e-05,
+      "loss": 1.6492,
+      "step": 672
+    },
+    {
+      "epoch": 0.13977154724818278,
+      "grad_norm": 0.49340352416038513,
+      "learning_rate": 8e-05,
+      "loss": 1.6563,
+      "step": 673
+    },
+    {
+      "epoch": 0.13997923156801662,
+      "grad_norm": 0.5209794044494629,
+      "learning_rate": 8e-05,
+      "loss": 1.7233,
+      "step": 674
+    },
+    {
+      "epoch": 0.14018691588785046,
+      "grad_norm": 0.48782119154930115,
+      "learning_rate": 8e-05,
+      "loss": 1.6472,
+      "step": 675
+    },
+    {
+      "epoch": 0.14039460020768432,
+      "grad_norm": 0.5135282874107361,
+      "learning_rate": 8e-05,
+      "loss": 1.6647,
+      "step": 676
+    },
+    {
+      "epoch": 0.14060228452751816,
+      "grad_norm": 0.5062872171401978,
+      "learning_rate": 8e-05,
+      "loss": 1.6414,
+      "step": 677
+    },
+    {
+      "epoch": 0.14080996884735203,
+      "grad_norm": 0.4905490577220917,
+      "learning_rate": 8e-05,
+      "loss": 1.5628,
+      "step": 678
+    },
+    {
+      "epoch": 0.14101765316718587,
+      "grad_norm": 0.4919457733631134,
+      "learning_rate": 8e-05,
+      "loss": 1.6491,
+      "step": 679
+    },
+    {
+      "epoch": 0.14122533748701974,
+      "grad_norm": 0.5233182907104492,
+      "learning_rate": 8e-05,
+      "loss": 1.6324,
+      "step": 680
+    },
+    {
+      "epoch": 0.14143302180685358,
+      "grad_norm": 0.4853593111038208,
+      "learning_rate": 8e-05,
+      "loss": 1.6134,
+      "step": 681
+    },
+    {
+      "epoch": 0.14164070612668744,
+      "grad_norm": 0.48515936732292175,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 682
+    },
+    {
+      "epoch": 0.14184839044652128,
+      "grad_norm": 0.5354658961296082,
+      "learning_rate": 8e-05,
+      "loss": 1.5981,
+      "step": 683
+    },
+    {
+      "epoch": 0.14205607476635515,
+      "grad_norm": 0.48015713691711426,
+      "learning_rate": 8e-05,
+      "loss": 1.6279,
+      "step": 684
+    },
+    {
+      "epoch": 0.142263759086189,
+      "grad_norm": 0.5790321230888367,
+      "learning_rate": 8e-05,
+      "loss": 1.7114,
+      "step": 685
+    },
+    {
+      "epoch": 0.14247144340602286,
+      "grad_norm": 0.46386808156967163,
+      "learning_rate": 8e-05,
+      "loss": 1.6201,
+      "step": 686
+    },
+    {
+      "epoch": 0.1426791277258567,
+      "grad_norm": 0.5704835653305054,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 687
+    },
+    {
+      "epoch": 0.14288681204569054,
+      "grad_norm": 0.4924614727497101,
+      "learning_rate": 8e-05,
+      "loss": 1.7339,
+      "step": 688
+    },
+    {
+      "epoch": 0.1430944963655244,
+      "grad_norm": 0.5158697962760925,
+      "learning_rate": 8e-05,
+      "loss": 1.6294,
+      "step": 689
+    },
+    {
+      "epoch": 0.14330218068535824,
+      "grad_norm": 0.5031910538673401,
+      "learning_rate": 8e-05,
+      "loss": 1.5753,
+      "step": 690
+    },
+    {
+      "epoch": 0.1435098650051921,
+      "grad_norm": 0.4993326961994171,
+      "learning_rate": 8e-05,
+      "loss": 1.7069,
+      "step": 691
+    },
+    {
+      "epoch": 0.14371754932502595,
+      "grad_norm": 0.5596215724945068,
+      "learning_rate": 8e-05,
+      "loss": 1.6887,
+      "step": 692
+    },
+    {
+      "epoch": 0.14392523364485982,
+      "grad_norm": 0.464817613363266,
+      "learning_rate": 8e-05,
+      "loss": 1.5974,
+      "step": 693
+    },
+    {
+      "epoch": 0.14413291796469366,
+      "grad_norm": 0.5313472151756287,
+      "learning_rate": 8e-05,
+      "loss": 1.6584,
+      "step": 694
+    },
+    {
+      "epoch": 0.14434060228452752,
+      "grad_norm": 0.470864862203598,
+      "learning_rate": 8e-05,
+      "loss": 1.5668,
+      "step": 695
+    },
+    {
+      "epoch": 0.14454828660436136,
+      "grad_norm": 0.4789741039276123,
+      "learning_rate": 8e-05,
+      "loss": 1.6432,
+      "step": 696
+    },
+    {
+      "epoch": 0.14475597092419523,
+      "grad_norm": 0.4912903606891632,
+      "learning_rate": 8e-05,
+      "loss": 1.5999,
+      "step": 697
+    },
+    {
+      "epoch": 0.14496365524402907,
+      "grad_norm": 0.49190813302993774,
+      "learning_rate": 8e-05,
+      "loss": 1.6045,
+      "step": 698
+    },
+    {
+      "epoch": 0.14517133956386294,
+      "grad_norm": 0.48753032088279724,
+      "learning_rate": 8e-05,
+      "loss": 1.6548,
+      "step": 699
+    },
+    {
+      "epoch": 0.14537902388369678,
+      "grad_norm": 0.47778916358947754,
+      "learning_rate": 8e-05,
+      "loss": 1.5859,
+      "step": 700
+    },
+    {
+      "epoch": 0.14558670820353065,
+      "grad_norm": 0.48212990164756775,
+      "learning_rate": 8e-05,
+      "loss": 1.6022,
+      "step": 701
+    },
+    {
+      "epoch": 0.14579439252336449,
+      "grad_norm": 0.5078028440475464,
+      "learning_rate": 8e-05,
+      "loss": 1.7241,
+      "step": 702
+    },
+    {
+      "epoch": 0.14600207684319833,
+      "grad_norm": 0.49921220541000366,
+      "learning_rate": 8e-05,
+      "loss": 1.7156,
+      "step": 703
+    },
+    {
+      "epoch": 0.1462097611630322,
+      "grad_norm": 0.5023846626281738,
+      "learning_rate": 8e-05,
+      "loss": 1.71,
+      "step": 704
+    },
+    {
+      "epoch": 0.14641744548286603,
+      "grad_norm": 0.4875635802745819,
+      "learning_rate": 8e-05,
+      "loss": 1.6838,
+      "step": 705
+    },
+    {
+      "epoch": 0.1466251298026999,
+      "grad_norm": 0.477684885263443,
+      "learning_rate": 8e-05,
+      "loss": 1.6201,
+      "step": 706
+    },
+    {
+      "epoch": 0.14683281412253374,
+      "grad_norm": 0.4834393858909607,
+      "learning_rate": 8e-05,
+      "loss": 1.6228,
+      "step": 707
+    },
+    {
+      "epoch": 0.1470404984423676,
+      "grad_norm": 0.4797709286212921,
+      "learning_rate": 8e-05,
+      "loss": 1.589,
+      "step": 708
+    },
+    {
+      "epoch": 0.14724818276220145,
+      "grad_norm": 0.4814303517341614,
+      "learning_rate": 8e-05,
+      "loss": 1.6916,
+      "step": 709
+    },
+    {
+      "epoch": 0.1474558670820353,
+      "grad_norm": 0.48855525255203247,
+      "learning_rate": 8e-05,
+      "loss": 1.6431,
+      "step": 710
+    },
+    {
+      "epoch": 0.14766355140186915,
+      "grad_norm": 0.4643092453479767,
+      "learning_rate": 8e-05,
+      "loss": 1.6707,
+      "step": 711
+    },
+    {
+      "epoch": 0.14787123572170302,
+      "grad_norm": 0.4828472435474396,
+      "learning_rate": 8e-05,
+      "loss": 1.6574,
+      "step": 712
+    },
+    {
+      "epoch": 0.14807892004153686,
+      "grad_norm": 0.5036082863807678,
+      "learning_rate": 8e-05,
+      "loss": 1.7352,
+      "step": 713
+    },
+    {
+      "epoch": 0.14828660436137073,
+      "grad_norm": 0.49272510409355164,
+      "learning_rate": 8e-05,
+      "loss": 1.676,
+      "step": 714
+    },
+    {
+      "epoch": 0.14849428868120457,
+      "grad_norm": 0.4979461431503296,
+      "learning_rate": 8e-05,
+      "loss": 1.6097,
+      "step": 715
+    },
+    {
+      "epoch": 0.14870197300103843,
+      "grad_norm": 0.4923213720321655,
+      "learning_rate": 8e-05,
+      "loss": 1.5899,
+      "step": 716
+    },
+    {
+      "epoch": 0.14890965732087227,
+      "grad_norm": 0.48915570974349976,
+      "learning_rate": 8e-05,
+      "loss": 1.5901,
+      "step": 717
+    },
+    {
+      "epoch": 0.1491173416407061,
+      "grad_norm": 0.4918684661388397,
+      "learning_rate": 8e-05,
+      "loss": 1.6548,
+      "step": 718
+    },
+    {
+      "epoch": 0.14932502596053998,
+      "grad_norm": 0.47110480070114136,
+      "learning_rate": 8e-05,
+      "loss": 1.6423,
+      "step": 719
+    },
+    {
+      "epoch": 0.14953271028037382,
+      "grad_norm": 0.5042193531990051,
+      "learning_rate": 8e-05,
+      "loss": 1.6419,
+      "step": 720
+    },
+    {
+      "epoch": 0.1497403946002077,
+      "grad_norm": 0.4912029504776001,
+      "learning_rate": 8e-05,
+      "loss": 1.6902,
+      "step": 721
+    },
+    {
+      "epoch": 0.14994807892004153,
+      "grad_norm": 0.4772588312625885,
+      "learning_rate": 8e-05,
+      "loss": 1.572,
+      "step": 722
+    },
+    {
+      "epoch": 0.1501557632398754,
+      "grad_norm": 0.4800530970096588,
+      "learning_rate": 8e-05,
+      "loss": 1.6327,
+      "step": 723
+    },
+    {
+      "epoch": 0.15036344755970923,
+      "grad_norm": 0.5064852237701416,
+      "learning_rate": 8e-05,
+      "loss": 1.6544,
+      "step": 724
+    },
+    {
+      "epoch": 0.1505711318795431,
+      "grad_norm": 0.4814448654651642,
+      "learning_rate": 8e-05,
+      "loss": 1.6141,
+      "step": 725
+    },
+    {
+      "epoch": 0.15077881619937694,
+      "grad_norm": 0.497315376996994,
+      "learning_rate": 8e-05,
+      "loss": 1.7214,
+      "step": 726
+    },
+    {
+      "epoch": 0.1509865005192108,
+      "grad_norm": 0.4816543459892273,
+      "learning_rate": 8e-05,
+      "loss": 1.6629,
+      "step": 727
+    },
+    {
+      "epoch": 0.15119418483904465,
+      "grad_norm": 0.4759445786476135,
+      "learning_rate": 8e-05,
+      "loss": 1.5397,
+      "step": 728
+    },
+    {
+      "epoch": 0.15140186915887852,
+      "grad_norm": 0.4926449954509735,
+      "learning_rate": 8e-05,
+      "loss": 1.6486,
+      "step": 729
+    },
+    {
+      "epoch": 0.15160955347871236,
+      "grad_norm": 0.4702441990375519,
+      "learning_rate": 8e-05,
+      "loss": 1.5585,
+      "step": 730
+    },
+    {
+      "epoch": 0.15181723779854622,
+      "grad_norm": 0.48316627740859985,
+      "learning_rate": 8e-05,
+      "loss": 1.6167,
+      "step": 731
+    },
+    {
+      "epoch": 0.15202492211838006,
+      "grad_norm": 0.5125194191932678,
+      "learning_rate": 8e-05,
+      "loss": 1.6876,
+      "step": 732
+    },
+    {
+      "epoch": 0.1522326064382139,
+      "grad_norm": 0.4879029393196106,
+      "learning_rate": 8e-05,
+      "loss": 1.6086,
+      "step": 733
+    },
+    {
+      "epoch": 0.15244029075804777,
+      "grad_norm": 0.48626405000686646,
+      "learning_rate": 8e-05,
+      "loss": 1.5933,
+      "step": 734
+    },
+    {
+      "epoch": 0.1526479750778816,
+      "grad_norm": 0.504183828830719,
+      "learning_rate": 8e-05,
+      "loss": 1.6104,
+      "step": 735
+    },
+    {
+      "epoch": 0.15285565939771548,
+      "grad_norm": 0.4829785227775574,
+      "learning_rate": 8e-05,
+      "loss": 1.6874,
+      "step": 736
+    },
+    {
+      "epoch": 0.15306334371754932,
+      "grad_norm": 0.5468804240226746,
+      "learning_rate": 8e-05,
+      "loss": 1.7613,
+      "step": 737
+    },
+    {
+      "epoch": 0.15327102803738318,
+      "grad_norm": 0.5711941719055176,
+      "learning_rate": 8e-05,
+      "loss": 1.6815,
+      "step": 738
+    },
+    {
+      "epoch": 0.15347871235721702,
+      "grad_norm": 0.4895186722278595,
+      "learning_rate": 8e-05,
+      "loss": 1.6076,
+      "step": 739
+    },
+    {
+      "epoch": 0.1536863966770509,
+      "grad_norm": 0.5142952799797058,
+      "learning_rate": 8e-05,
+      "loss": 1.7081,
+      "step": 740
+    },
+    {
+      "epoch": 0.15389408099688473,
+      "grad_norm": 0.498079776763916,
+      "learning_rate": 8e-05,
+      "loss": 1.6074,
+      "step": 741
+    },
+    {
+      "epoch": 0.1541017653167186,
+      "grad_norm": 0.47368738055229187,
+      "learning_rate": 8e-05,
+      "loss": 1.6465,
+      "step": 742
+    },
+    {
+      "epoch": 0.15430944963655244,
+      "grad_norm": 0.5251184105873108,
+      "learning_rate": 8e-05,
+      "loss": 1.7324,
+      "step": 743
+    },
+    {
+      "epoch": 0.1545171339563863,
+      "grad_norm": 0.49982884526252747,
+      "learning_rate": 8e-05,
+      "loss": 1.7063,
+      "step": 744
+    },
+    {
+      "epoch": 0.15472481827622014,
+      "grad_norm": 0.4818509519100189,
+      "learning_rate": 8e-05,
+      "loss": 1.6719,
+      "step": 745
+    },
+    {
+      "epoch": 0.154932502596054,
+      "grad_norm": 0.48312175273895264,
+      "learning_rate": 8e-05,
+      "loss": 1.6511,
+      "step": 746
+    },
+    {
+      "epoch": 0.15514018691588785,
+      "grad_norm": 0.4745786190032959,
+      "learning_rate": 8e-05,
+      "loss": 1.6419,
+      "step": 747
+    },
+    {
+      "epoch": 0.1553478712357217,
+      "grad_norm": 0.5050838589668274,
+      "learning_rate": 8e-05,
+      "loss": 1.6486,
+      "step": 748
+    },
+    {
+      "epoch": 0.15555555555555556,
+      "grad_norm": 0.4608010947704315,
+      "learning_rate": 8e-05,
+      "loss": 1.5949,
+      "step": 749
+    },
+    {
+      "epoch": 0.1557632398753894,
+      "grad_norm": 0.499937504529953,
+      "learning_rate": 8e-05,
+      "loss": 1.6977,
+      "step": 750
+    },
+    {
+      "epoch": 0.15597092419522327,
+      "grad_norm": 0.48160862922668457,
+      "learning_rate": 8e-05,
+      "loss": 1.6612,
+      "step": 751
+    },
+    {
+      "epoch": 0.1561786085150571,
+      "grad_norm": 0.47431129217147827,
+      "learning_rate": 8e-05,
+      "loss": 1.5773,
+      "step": 752
+    },
+    {
+      "epoch": 0.15638629283489097,
+      "grad_norm": 0.48737314343452454,
+      "learning_rate": 8e-05,
+      "loss": 1.5839,
+      "step": 753
+    },
+    {
+      "epoch": 0.1565939771547248,
+      "grad_norm": 0.49073323607444763,
+      "learning_rate": 8e-05,
+      "loss": 1.5981,
+      "step": 754
+    },
+    {
+      "epoch": 0.15680166147455868,
+      "grad_norm": 0.4805930554866791,
+      "learning_rate": 8e-05,
+      "loss": 1.5883,
+      "step": 755
+    },
+    {
+      "epoch": 0.15700934579439252,
+      "grad_norm": 0.4898761212825775,
+      "learning_rate": 8e-05,
+      "loss": 1.6197,
+      "step": 756
+    },
+    {
+      "epoch": 0.15721703011422639,
+      "grad_norm": 0.48299679160118103,
+      "learning_rate": 8e-05,
+      "loss": 1.6251,
+      "step": 757
+    },
+    {
+      "epoch": 0.15742471443406023,
+      "grad_norm": 0.483012318611145,
+      "learning_rate": 8e-05,
+      "loss": 1.5927,
+      "step": 758
+    },
+    {
+      "epoch": 0.1576323987538941,
+      "grad_norm": 0.49116986989974976,
+      "learning_rate": 8e-05,
+      "loss": 1.6458,
+      "step": 759
+    },
+    {
+      "epoch": 0.15784008307372793,
+      "grad_norm": 0.4684026837348938,
+      "learning_rate": 8e-05,
+      "loss": 1.585,
+      "step": 760
+    },
+    {
+      "epoch": 0.1580477673935618,
+      "grad_norm": 0.4952329099178314,
+      "learning_rate": 8e-05,
+      "loss": 1.6516,
+      "step": 761
+    },
+    {
+      "epoch": 0.15825545171339564,
+      "grad_norm": 0.4795088768005371,
+      "learning_rate": 8e-05,
+      "loss": 1.6788,
+      "step": 762
+    },
+    {
+      "epoch": 0.15846313603322948,
+      "grad_norm": 0.481624960899353,
+      "learning_rate": 8e-05,
+      "loss": 1.577,
+      "step": 763
+    },
+    {
+      "epoch": 0.15867082035306335,
+      "grad_norm": 0.48516589403152466,
+      "learning_rate": 8e-05,
+      "loss": 1.6067,
+      "step": 764
+    },
+    {
+      "epoch": 0.1588785046728972,
+      "grad_norm": 0.5102306604385376,
+      "learning_rate": 8e-05,
+      "loss": 1.7154,
+      "step": 765
+    },
+    {
+      "epoch": 0.15908618899273105,
+      "grad_norm": 0.48237496614456177,
+      "learning_rate": 8e-05,
+      "loss": 1.6296,
+      "step": 766
+    },
+    {
+      "epoch": 0.1592938733125649,
+      "grad_norm": 0.5044882297515869,
+      "learning_rate": 8e-05,
+      "loss": 1.6751,
+      "step": 767
+    },
+    {
+      "epoch": 0.15950155763239876,
+      "grad_norm": 0.47297602891921997,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 768
+    },
+    {
+      "epoch": 0.1597092419522326,
+      "grad_norm": 0.5018360018730164,
+      "learning_rate": 8e-05,
+      "loss": 1.6738,
+      "step": 769
+    },
+    {
+      "epoch": 0.15991692627206647,
+      "grad_norm": 0.49589648842811584,
+      "learning_rate": 8e-05,
+      "loss": 1.6752,
+      "step": 770
+    },
+    {
+      "epoch": 0.1601246105919003,
+      "grad_norm": 0.4882931411266327,
+      "learning_rate": 8e-05,
+      "loss": 1.666,
+      "step": 771
+    },
+    {
+      "epoch": 0.16033229491173417,
+      "grad_norm": 0.4880520701408386,
+      "learning_rate": 8e-05,
+      "loss": 1.5941,
+      "step": 772
+    },
+    {
+      "epoch": 0.16053997923156801,
+      "grad_norm": 0.4973679780960083,
+      "learning_rate": 8e-05,
+      "loss": 1.6388,
+      "step": 773
+    },
+    {
+      "epoch": 0.16074766355140188,
+      "grad_norm": 0.494547039270401,
+      "learning_rate": 8e-05,
+      "loss": 1.6878,
+      "step": 774
+    },
+    {
+      "epoch": 0.16095534787123572,
+      "grad_norm": 0.4797399938106537,
+      "learning_rate": 8e-05,
+      "loss": 1.6462,
+      "step": 775
+    },
+    {
+      "epoch": 0.16116303219106956,
+      "grad_norm": 0.4973697364330292,
+      "learning_rate": 8e-05,
+      "loss": 1.6277,
+      "step": 776
+    },
+    {
+      "epoch": 0.16137071651090343,
+      "grad_norm": 0.4668048322200775,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 777
+    },
+    {
+      "epoch": 0.16157840083073727,
+      "grad_norm": 0.49718210101127625,
+      "learning_rate": 8e-05,
+      "loss": 1.6764,
+      "step": 778
+    },
+    {
+      "epoch": 0.16178608515057114,
+      "grad_norm": 0.48876386880874634,
+      "learning_rate": 8e-05,
+      "loss": 1.6308,
+      "step": 779
+    },
+    {
+      "epoch": 0.16199376947040497,
+      "grad_norm": 0.4705182909965515,
+      "learning_rate": 8e-05,
+      "loss": 1.6355,
+      "step": 780
+    },
+    {
+      "epoch": 0.16220145379023884,
+      "grad_norm": 0.48368439078330994,
+      "learning_rate": 8e-05,
+      "loss": 1.6519,
+      "step": 781
+    },
+    {
+      "epoch": 0.16240913811007268,
+      "grad_norm": 0.479199081659317,
+      "learning_rate": 8e-05,
+      "loss": 1.6046,
+      "step": 782
+    },
+    {
+      "epoch": 0.16261682242990655,
+      "grad_norm": 0.4942733943462372,
+      "learning_rate": 8e-05,
+      "loss": 1.6996,
+      "step": 783
+    },
+    {
+      "epoch": 0.1628245067497404,
+      "grad_norm": 0.5083856582641602,
+      "learning_rate": 8e-05,
+      "loss": 1.7383,
+      "step": 784
+    },
+    {
+      "epoch": 0.16303219106957426,
+      "grad_norm": 0.46334561705589294,
+      "learning_rate": 8e-05,
+      "loss": 1.6044,
+      "step": 785
+    },
+    {
+      "epoch": 0.1632398753894081,
+      "grad_norm": 0.5042205452919006,
+      "learning_rate": 8e-05,
+      "loss": 1.6773,
+      "step": 786
+    },
+    {
+      "epoch": 0.16344755970924196,
+      "grad_norm": 0.4855109453201294,
+      "learning_rate": 8e-05,
+      "loss": 1.6655,
+      "step": 787
+    },
+    {
+      "epoch": 0.1636552440290758,
+      "grad_norm": 0.4775260090827942,
+      "learning_rate": 8e-05,
+      "loss": 1.5763,
+      "step": 788
+    },
+    {
+      "epoch": 0.16386292834890967,
+      "grad_norm": 0.49258020520210266,
+      "learning_rate": 8e-05,
+      "loss": 1.6628,
+      "step": 789
+    },
+    {
+      "epoch": 0.1640706126687435,
+      "grad_norm": 0.4889313280582428,
+      "learning_rate": 8e-05,
+      "loss": 1.6056,
+      "step": 790
+    },
+    {
+      "epoch": 0.16427829698857735,
+      "grad_norm": 0.49056270718574524,
+      "learning_rate": 8e-05,
+      "loss": 1.6066,
+      "step": 791
+    },
+    {
+      "epoch": 0.16448598130841122,
+      "grad_norm": 0.47176501154899597,
+      "learning_rate": 8e-05,
+      "loss": 1.6171,
+      "step": 792
+    },
+    {
+      "epoch": 0.16469366562824506,
+      "grad_norm": 0.5094092488288879,
+      "learning_rate": 8e-05,
+      "loss": 1.6178,
+      "step": 793
+    },
+    {
+      "epoch": 0.16490134994807892,
+      "grad_norm": 0.4841829240322113,
+      "learning_rate": 8e-05,
+      "loss": 1.5962,
+      "step": 794
+    },
+    {
+      "epoch": 0.16510903426791276,
+      "grad_norm": 0.4807954728603363,
+      "learning_rate": 8e-05,
+      "loss": 1.5233,
+      "step": 795
+    },
+    {
+      "epoch": 0.16531671858774663,
+      "grad_norm": 0.4819432497024536,
+      "learning_rate": 8e-05,
+      "loss": 1.6909,
+      "step": 796
+    },
+    {
+      "epoch": 0.16552440290758047,
+      "grad_norm": 0.513128936290741,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 797
+    },
+    {
+      "epoch": 0.16573208722741434,
+      "grad_norm": 0.4752349555492401,
+      "learning_rate": 8e-05,
+      "loss": 1.5488,
+      "step": 798
+    },
+    {
+      "epoch": 0.16593977154724818,
+      "grad_norm": 0.48235711455345154,
+      "learning_rate": 8e-05,
+      "loss": 1.6281,
+      "step": 799
+    },
+    {
+      "epoch": 0.16614745586708204,
+      "grad_norm": 0.49711811542510986,
+      "learning_rate": 8e-05,
+      "loss": 1.6796,
+      "step": 800
+    },
+    {
+      "epoch": 0.16635514018691588,
+      "grad_norm": 0.5144046545028687,
+      "learning_rate": 8e-05,
+      "loss": 1.6711,
+      "step": 801
+    },
+    {
+      "epoch": 0.16656282450674975,
+      "grad_norm": 0.506710410118103,
+      "learning_rate": 8e-05,
+      "loss": 1.6507,
+      "step": 802
+    },
+    {
+      "epoch": 0.1667705088265836,
+      "grad_norm": 0.5138517618179321,
+      "learning_rate": 8e-05,
+      "loss": 1.6207,
+      "step": 803
+    },
+    {
+      "epoch": 0.16697819314641746,
+      "grad_norm": 0.480153352022171,
+      "learning_rate": 8e-05,
+      "loss": 1.6877,
+      "step": 804
+    },
+    {
+      "epoch": 0.1671858774662513,
+      "grad_norm": 0.5096755623817444,
+      "learning_rate": 8e-05,
+      "loss": 1.7058,
+      "step": 805
+    },
+    {
+      "epoch": 0.16739356178608514,
+      "grad_norm": 0.49383774399757385,
+      "learning_rate": 8e-05,
+      "loss": 1.6983,
+      "step": 806
+    },
+    {
+      "epoch": 0.167601246105919,
+      "grad_norm": 0.5019107460975647,
+      "learning_rate": 8e-05,
+      "loss": 1.6543,
+      "step": 807
+    },
+    {
+      "epoch": 0.16780893042575284,
+      "grad_norm": 0.48990175127983093,
+      "learning_rate": 8e-05,
+      "loss": 1.6206,
+      "step": 808
+    },
+    {
+      "epoch": 0.1680166147455867,
+      "grad_norm": 0.4764305055141449,
+      "learning_rate": 8e-05,
+      "loss": 1.6453,
+      "step": 809
+    },
+    {
+      "epoch": 0.16822429906542055,
+      "grad_norm": 0.5044610500335693,
+      "learning_rate": 8e-05,
+      "loss": 1.6739,
+      "step": 810
+    },
+    {
+      "epoch": 0.16843198338525442,
+      "grad_norm": 0.48134395480155945,
+      "learning_rate": 8e-05,
+      "loss": 1.6153,
+      "step": 811
+    },
+    {
+      "epoch": 0.16863966770508826,
+      "grad_norm": 0.4967174232006073,
+      "learning_rate": 8e-05,
+      "loss": 1.677,
+      "step": 812
+    },
+    {
+      "epoch": 0.16884735202492213,
+      "grad_norm": 0.47364965081214905,
+      "learning_rate": 8e-05,
+      "loss": 1.555,
+      "step": 813
+    },
+    {
+      "epoch": 0.16905503634475597,
+      "grad_norm": 0.49788594245910645,
+      "learning_rate": 8e-05,
+      "loss": 1.6635,
+      "step": 814
+    },
+    {
+      "epoch": 0.16926272066458983,
+      "grad_norm": 0.49295979738235474,
+      "learning_rate": 8e-05,
+      "loss": 1.6627,
+      "step": 815
+    },
+    {
+      "epoch": 0.16947040498442367,
+      "grad_norm": 0.5006287693977356,
+      "learning_rate": 8e-05,
+      "loss": 1.6183,
+      "step": 816
+    },
+    {
+      "epoch": 0.16967808930425754,
+      "grad_norm": 0.4745385944843292,
+      "learning_rate": 8e-05,
+      "loss": 1.585,
+      "step": 817
+    },
+    {
+      "epoch": 0.16988577362409138,
+      "grad_norm": 0.5007401704788208,
+      "learning_rate": 8e-05,
+      "loss": 1.6358,
+      "step": 818
+    },
+    {
+      "epoch": 0.17009345794392525,
+      "grad_norm": 0.4691254198551178,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 819
+    },
+    {
+      "epoch": 0.1703011422637591,
+      "grad_norm": 0.48528677225112915,
+      "learning_rate": 8e-05,
+      "loss": 1.6889,
+      "step": 820
+    },
+    {
+      "epoch": 0.17050882658359293,
+      "grad_norm": 0.48160579800605774,
+      "learning_rate": 8e-05,
+      "loss": 1.6284,
+      "step": 821
+    },
+    {
+      "epoch": 0.1707165109034268,
+      "grad_norm": 0.4866841435432434,
+      "learning_rate": 8e-05,
+      "loss": 1.635,
+      "step": 822
+    },
+    {
+      "epoch": 0.17092419522326063,
+      "grad_norm": 0.4998115003108978,
+      "learning_rate": 8e-05,
+      "loss": 1.7034,
+      "step": 823
+    },
+    {
+      "epoch": 0.1711318795430945,
+      "grad_norm": 0.49009475111961365,
+      "learning_rate": 8e-05,
+      "loss": 1.6505,
+      "step": 824
+    },
+    {
+      "epoch": 0.17133956386292834,
+      "grad_norm": 0.47515684366226196,
+      "learning_rate": 8e-05,
+      "loss": 1.6409,
+      "step": 825
+    },
+    {
+      "epoch": 0.1715472481827622,
+      "grad_norm": 0.5075555443763733,
+      "learning_rate": 8e-05,
+      "loss": 1.6623,
+      "step": 826
+    },
+    {
+      "epoch": 0.17175493250259605,
+      "grad_norm": 0.4834977984428406,
+      "learning_rate": 8e-05,
+      "loss": 1.6343,
+      "step": 827
+    },
+    {
+      "epoch": 0.17196261682242991,
+      "grad_norm": 0.4810510575771332,
+      "learning_rate": 8e-05,
+      "loss": 1.5518,
+      "step": 828
+    },
+    {
+      "epoch": 0.17217030114226375,
+      "grad_norm": 0.5000099539756775,
+      "learning_rate": 8e-05,
+      "loss": 1.6423,
+      "step": 829
+    },
+    {
+      "epoch": 0.17237798546209762,
+      "grad_norm": 0.4854555130004883,
+      "learning_rate": 8e-05,
+      "loss": 1.6345,
+      "step": 830
+    },
+    {
+      "epoch": 0.17258566978193146,
+      "grad_norm": 0.47438621520996094,
+      "learning_rate": 8e-05,
+      "loss": 1.5127,
+      "step": 831
+    },
+    {
+      "epoch": 0.17279335410176533,
+      "grad_norm": 0.4803098440170288,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 832
+    },
+    {
+      "epoch": 0.17300103842159917,
+      "grad_norm": 0.48187175393104553,
+      "learning_rate": 8e-05,
+      "loss": 1.6423,
+      "step": 833
+    },
+    {
+      "epoch": 0.17320872274143304,
+      "grad_norm": 0.4821400046348572,
+      "learning_rate": 8e-05,
+      "loss": 1.6798,
+      "step": 834
+    },
+    {
+      "epoch": 0.17341640706126688,
+      "grad_norm": 0.4715557098388672,
+      "learning_rate": 8e-05,
+      "loss": 1.6162,
+      "step": 835
+    },
+    {
+      "epoch": 0.17362409138110071,
+      "grad_norm": 0.48765188455581665,
+      "learning_rate": 8e-05,
+      "loss": 1.6846,
+      "step": 836
+    },
+    {
+      "epoch": 0.17383177570093458,
+      "grad_norm": 0.5081665515899658,
+      "learning_rate": 8e-05,
+      "loss": 1.6857,
+      "step": 837
+    },
+    {
+      "epoch": 0.17403946002076842,
+      "grad_norm": 0.49220722913742065,
+      "learning_rate": 8e-05,
+      "loss": 1.6495,
+      "step": 838
+    },
+    {
+      "epoch": 0.1742471443406023,
+      "grad_norm": 0.461845338344574,
+      "learning_rate": 8e-05,
+      "loss": 1.6098,
+      "step": 839
+    },
+    {
+      "epoch": 0.17445482866043613,
+      "grad_norm": 0.4715192914009094,
+      "learning_rate": 8e-05,
+      "loss": 1.5672,
+      "step": 840
+    },
+    {
+      "epoch": 0.17466251298027,
+      "grad_norm": 0.4969162940979004,
+      "learning_rate": 8e-05,
+      "loss": 1.6727,
+      "step": 841
+    },
+    {
+      "epoch": 0.17487019730010384,
+      "grad_norm": 0.4748702645301819,
+      "learning_rate": 8e-05,
+      "loss": 1.6493,
+      "step": 842
+    },
+    {
+      "epoch": 0.1750778816199377,
+      "grad_norm": 0.47799915075302124,
+      "learning_rate": 8e-05,
+      "loss": 1.5978,
+      "step": 843
+    },
+    {
+      "epoch": 0.17528556593977154,
+      "grad_norm": 0.4815102815628052,
+      "learning_rate": 8e-05,
+      "loss": 1.587,
+      "step": 844
+    },
+    {
+      "epoch": 0.1754932502596054,
+      "grad_norm": 0.4794101417064667,
+      "learning_rate": 8e-05,
+      "loss": 1.5926,
+      "step": 845
+    },
+    {
+      "epoch": 0.17570093457943925,
+      "grad_norm": 0.480980783700943,
+      "learning_rate": 8e-05,
+      "loss": 1.6106,
+      "step": 846
+    },
+    {
+      "epoch": 0.17590861889927312,
+      "grad_norm": 0.483927458524704,
+      "learning_rate": 8e-05,
+      "loss": 1.6169,
+      "step": 847
+    },
+    {
+      "epoch": 0.17611630321910696,
+      "grad_norm": 0.4977688193321228,
+      "learning_rate": 8e-05,
+      "loss": 1.6233,
+      "step": 848
+    },
+    {
+      "epoch": 0.1763239875389408,
+      "grad_norm": 0.4976138770580292,
+      "learning_rate": 8e-05,
+      "loss": 1.7223,
+      "step": 849
+    },
+    {
+      "epoch": 0.17653167185877466,
+      "grad_norm": 0.4990401864051819,
+      "learning_rate": 8e-05,
+      "loss": 1.6844,
+      "step": 850
+    },
+    {
+      "epoch": 0.1767393561786085,
+      "grad_norm": 0.5020548105239868,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 851
+    },
+    {
+      "epoch": 0.17694704049844237,
+      "grad_norm": 0.5174390077590942,
+      "learning_rate": 8e-05,
+      "loss": 1.634,
+      "step": 852
+    },
+    {
+      "epoch": 0.1771547248182762,
+      "grad_norm": 0.4973217248916626,
+      "learning_rate": 8e-05,
+      "loss": 1.6551,
+      "step": 853
+    },
+    {
+      "epoch": 0.17736240913811008,
+      "grad_norm": 0.504899263381958,
+      "learning_rate": 8e-05,
+      "loss": 1.6016,
+      "step": 854
+    },
+    {
+      "epoch": 0.17757009345794392,
+      "grad_norm": 0.5085044503211975,
+      "learning_rate": 8e-05,
+      "loss": 1.6352,
+      "step": 855
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 0.4816834330558777,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 856
+    },
+    {
+      "epoch": 0.17798546209761162,
+      "grad_norm": 0.5155865550041199,
+      "learning_rate": 8e-05,
+      "loss": 1.6151,
+      "step": 857
+    },
+    {
+      "epoch": 0.1781931464174455,
+      "grad_norm": 0.48705244064331055,
+      "learning_rate": 8e-05,
+      "loss": 1.648,
+      "step": 858
+    },
+    {
+      "epoch": 0.17840083073727933,
+      "grad_norm": 0.5086380243301392,
+      "learning_rate": 8e-05,
+      "loss": 1.6601,
+      "step": 859
+    },
+    {
+      "epoch": 0.1786085150571132,
+      "grad_norm": 0.46647846698760986,
+      "learning_rate": 8e-05,
+      "loss": 1.6007,
+      "step": 860
+    },
+    {
+      "epoch": 0.17881619937694704,
+      "grad_norm": 0.4949185252189636,
+      "learning_rate": 8e-05,
+      "loss": 1.6181,
+      "step": 861
+    },
+    {
+      "epoch": 0.1790238836967809,
+      "grad_norm": 0.4831303656101227,
+      "learning_rate": 8e-05,
+      "loss": 1.6037,
+      "step": 862
+    },
+    {
+      "epoch": 0.17923156801661475,
+      "grad_norm": 0.5041294693946838,
+      "learning_rate": 8e-05,
+      "loss": 1.6582,
+      "step": 863
+    },
+    {
+      "epoch": 0.17943925233644858,
+      "grad_norm": 0.5045524835586548,
+      "learning_rate": 8e-05,
+      "loss": 1.6141,
+      "step": 864
+    },
+    {
+      "epoch": 0.17964693665628245,
+      "grad_norm": 0.49055367708206177,
+      "learning_rate": 8e-05,
+      "loss": 1.6371,
+      "step": 865
+    },
+    {
+      "epoch": 0.1798546209761163,
+      "grad_norm": 0.5056166648864746,
+      "learning_rate": 8e-05,
+      "loss": 1.6949,
+      "step": 866
+    },
+    {
+      "epoch": 0.18006230529595016,
+      "grad_norm": 0.4675775170326233,
+      "learning_rate": 8e-05,
+      "loss": 1.5697,
+      "step": 867
+    },
+    {
+      "epoch": 0.180269989615784,
+      "grad_norm": 0.490346223115921,
+      "learning_rate": 8e-05,
+      "loss": 1.6092,
+      "step": 868
+    },
+    {
+      "epoch": 0.18047767393561787,
+      "grad_norm": 0.4684089720249176,
+      "learning_rate": 8e-05,
+      "loss": 1.5648,
+      "step": 869
+    },
+    {
+      "epoch": 0.1806853582554517,
+      "grad_norm": 0.46101874113082886,
+      "learning_rate": 8e-05,
+      "loss": 1.566,
+      "step": 870
+    },
+    {
+      "epoch": 0.18089304257528557,
+      "grad_norm": 0.4861421287059784,
+      "learning_rate": 8e-05,
+      "loss": 1.6511,
+      "step": 871
+    },
+    {
+      "epoch": 0.1811007268951194,
+      "grad_norm": 0.48315370082855225,
+      "learning_rate": 8e-05,
+      "loss": 1.6527,
+      "step": 872
+    },
+    {
+      "epoch": 0.18130841121495328,
+      "grad_norm": 0.48477229475975037,
+      "learning_rate": 8e-05,
+      "loss": 1.6477,
+      "step": 873
+    },
+    {
+      "epoch": 0.18151609553478712,
+      "grad_norm": 0.4834623336791992,
+      "learning_rate": 8e-05,
+      "loss": 1.5799,
+      "step": 874
+    },
+    {
+      "epoch": 0.181723779854621,
+      "grad_norm": 0.4903808534145355,
+      "learning_rate": 8e-05,
+      "loss": 1.6191,
+      "step": 875
+    },
+    {
+      "epoch": 0.18193146417445483,
+      "grad_norm": 0.4885776937007904,
+      "learning_rate": 8e-05,
+      "loss": 1.6815,
+      "step": 876
+    },
+    {
+      "epoch": 0.1821391484942887,
+      "grad_norm": 0.48751139640808105,
+      "learning_rate": 8e-05,
+      "loss": 1.6905,
+      "step": 877
+    },
+    {
+      "epoch": 0.18234683281412253,
+      "grad_norm": 0.48983898758888245,
+      "learning_rate": 8e-05,
+      "loss": 1.5902,
+      "step": 878
+    },
+    {
+      "epoch": 0.18255451713395637,
+      "grad_norm": 0.47432875633239746,
+      "learning_rate": 8e-05,
+      "loss": 1.6571,
+      "step": 879
+    },
+    {
+      "epoch": 0.18276220145379024,
+      "grad_norm": 0.4667503535747528,
+      "learning_rate": 8e-05,
+      "loss": 1.6198,
+      "step": 880
+    },
+    {
+      "epoch": 0.18296988577362408,
+      "grad_norm": 0.5021142363548279,
+      "learning_rate": 8e-05,
+      "loss": 1.7175,
+      "step": 881
+    },
+    {
+      "epoch": 0.18317757009345795,
+      "grad_norm": 0.46607160568237305,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 882
+    },
+    {
+      "epoch": 0.1833852544132918,
+      "grad_norm": 0.4718491733074188,
+      "learning_rate": 8e-05,
+      "loss": 1.5974,
+      "step": 883
+    },
+    {
+      "epoch": 0.18359293873312565,
+      "grad_norm": 0.49445825815200806,
+      "learning_rate": 8e-05,
+      "loss": 1.6271,
+      "step": 884
+    },
+    {
+      "epoch": 0.1838006230529595,
+      "grad_norm": 0.46848443150520325,
+      "learning_rate": 8e-05,
+      "loss": 1.6061,
+      "step": 885
+    },
+    {
+      "epoch": 0.18400830737279336,
+      "grad_norm": 0.46459993720054626,
+      "learning_rate": 8e-05,
+      "loss": 1.5613,
+      "step": 886
+    },
+    {
+      "epoch": 0.1842159916926272,
+      "grad_norm": 0.47281578183174133,
+      "learning_rate": 8e-05,
+      "loss": 1.704,
+      "step": 887
+    },
+    {
+      "epoch": 0.18442367601246107,
+      "grad_norm": 0.456195592880249,
+      "learning_rate": 8e-05,
+      "loss": 1.559,
+      "step": 888
+    },
+    {
+      "epoch": 0.1846313603322949,
+      "grad_norm": 0.5002410411834717,
+      "learning_rate": 8e-05,
+      "loss": 1.6179,
+      "step": 889
+    },
+    {
+      "epoch": 0.18483904465212878,
+      "grad_norm": 0.4889829158782959,
+      "learning_rate": 8e-05,
+      "loss": 1.6081,
+      "step": 890
+    },
+    {
+      "epoch": 0.18504672897196262,
+      "grad_norm": 0.5022076964378357,
+      "learning_rate": 8e-05,
+      "loss": 1.629,
+      "step": 891
+    },
+    {
+      "epoch": 0.18525441329179648,
+      "grad_norm": 0.496134489774704,
+      "learning_rate": 8e-05,
+      "loss": 1.611,
+      "step": 892
+    },
+    {
+      "epoch": 0.18546209761163032,
+      "grad_norm": 0.4800807237625122,
+      "learning_rate": 8e-05,
+      "loss": 1.68,
+      "step": 893
+    },
+    {
+      "epoch": 0.18566978193146416,
+      "grad_norm": 0.4833254814147949,
+      "learning_rate": 8e-05,
+      "loss": 1.6493,
+      "step": 894
+    },
+    {
+      "epoch": 0.18587746625129803,
+      "grad_norm": 0.47736915946006775,
+      "learning_rate": 8e-05,
+      "loss": 1.5986,
+      "step": 895
+    },
+    {
+      "epoch": 0.18608515057113187,
+      "grad_norm": 0.5091724395751953,
+      "learning_rate": 8e-05,
+      "loss": 1.6303,
+      "step": 896
+    },
+    {
+      "epoch": 0.18629283489096574,
+      "grad_norm": 0.46926015615463257,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 897
+    },
+    {
+      "epoch": 0.18650051921079958,
+      "grad_norm": 0.48769834637641907,
+      "learning_rate": 8e-05,
+      "loss": 1.6396,
+      "step": 898
+    },
+    {
+      "epoch": 0.18670820353063344,
+      "grad_norm": 0.47755125164985657,
+      "learning_rate": 8e-05,
+      "loss": 1.6374,
+      "step": 899
+    },
+    {
+      "epoch": 0.18691588785046728,
+      "grad_norm": 0.4844597280025482,
+      "learning_rate": 8e-05,
+      "loss": 1.6072,
+      "step": 900
+    },
+    {
+      "epoch": 0.18712357217030115,
+      "grad_norm": 0.5078486800193787,
+      "learning_rate": 8e-05,
+      "loss": 1.6898,
+      "step": 901
+    },
+    {
+      "epoch": 0.187331256490135,
+      "grad_norm": 0.4972101151943207,
+      "learning_rate": 8e-05,
+      "loss": 1.7502,
+      "step": 902
+    },
+    {
+      "epoch": 0.18753894080996886,
+      "grad_norm": 0.4829871952533722,
+      "learning_rate": 8e-05,
+      "loss": 1.615,
+      "step": 903
+    },
+    {
+      "epoch": 0.1877466251298027,
+      "grad_norm": 0.5250853300094604,
+      "learning_rate": 8e-05,
+      "loss": 1.6795,
+      "step": 904
+    },
+    {
+      "epoch": 0.18795430944963656,
+      "grad_norm": 0.4843866229057312,
+      "learning_rate": 8e-05,
+      "loss": 1.6664,
+      "step": 905
+    },
+    {
+      "epoch": 0.1881619937694704,
+      "grad_norm": 0.4833875894546509,
+      "learning_rate": 8e-05,
+      "loss": 1.5387,
+      "step": 906
+    },
+    {
+      "epoch": 0.18836967808930427,
+      "grad_norm": 0.5005854368209839,
+      "learning_rate": 8e-05,
+      "loss": 1.5622,
+      "step": 907
+    },
+    {
+      "epoch": 0.1885773624091381,
+      "grad_norm": 0.47214603424072266,
+      "learning_rate": 8e-05,
+      "loss": 1.6011,
+      "step": 908
+    },
+    {
+      "epoch": 0.18878504672897195,
+      "grad_norm": 0.5020946264266968,
+      "learning_rate": 8e-05,
+      "loss": 1.6588,
+      "step": 909
+    },
+    {
+      "epoch": 0.18899273104880582,
+      "grad_norm": 0.5007855892181396,
+      "learning_rate": 8e-05,
+      "loss": 1.6688,
+      "step": 910
+    },
+    {
+      "epoch": 0.18920041536863966,
+      "grad_norm": 0.5172010064125061,
+      "learning_rate": 8e-05,
+      "loss": 1.6751,
+      "step": 911
+    },
+    {
+      "epoch": 0.18940809968847352,
+      "grad_norm": 0.4866928160190582,
+      "learning_rate": 8e-05,
+      "loss": 1.56,
+      "step": 912
+    },
+    {
+      "epoch": 0.18961578400830736,
+      "grad_norm": 0.47611889243125916,
+      "learning_rate": 8e-05,
+      "loss": 1.5304,
+      "step": 913
+    },
+    {
+      "epoch": 0.18982346832814123,
+      "grad_norm": 0.4758463203907013,
+      "learning_rate": 8e-05,
+      "loss": 1.6201,
+      "step": 914
+    },
+    {
+      "epoch": 0.19003115264797507,
+      "grad_norm": 0.4987952709197998,
+      "learning_rate": 8e-05,
+      "loss": 1.6785,
+      "step": 915
+    },
+    {
+      "epoch": 0.19023883696780894,
+      "grad_norm": 0.5098257064819336,
+      "learning_rate": 8e-05,
+      "loss": 1.6758,
+      "step": 916
+    },
+    {
+      "epoch": 0.19044652128764278,
+      "grad_norm": 0.4878323972225189,
+      "learning_rate": 8e-05,
+      "loss": 1.5964,
+      "step": 917
+    },
+    {
+      "epoch": 0.19065420560747665,
+      "grad_norm": 0.49030864238739014,
+      "learning_rate": 8e-05,
+      "loss": 1.6166,
+      "step": 918
+    },
+    {
+      "epoch": 0.19086188992731049,
+      "grad_norm": 0.4641314446926117,
+      "learning_rate": 8e-05,
+      "loss": 1.6265,
+      "step": 919
+    },
+    {
+      "epoch": 0.19106957424714435,
+      "grad_norm": 0.49231404066085815,
+      "learning_rate": 8e-05,
+      "loss": 1.6691,
+      "step": 920
+    },
+    {
+      "epoch": 0.1912772585669782,
+      "grad_norm": 0.5089201331138611,
+      "learning_rate": 8e-05,
+      "loss": 1.6911,
+      "step": 921
+    },
+    {
+      "epoch": 0.19148494288681206,
+      "grad_norm": 0.4495040774345398,
+      "learning_rate": 8e-05,
+      "loss": 1.5579,
+      "step": 922
+    },
+    {
+      "epoch": 0.1916926272066459,
+      "grad_norm": 0.47652044892311096,
+      "learning_rate": 8e-05,
+      "loss": 1.5685,
+      "step": 923
+    },
+    {
+      "epoch": 0.19190031152647974,
+      "grad_norm": 0.4827858805656433,
+      "learning_rate": 8e-05,
+      "loss": 1.6168,
+      "step": 924
+    },
+    {
+      "epoch": 0.1921079958463136,
+      "grad_norm": 0.47965332865715027,
+      "learning_rate": 8e-05,
+      "loss": 1.5983,
+      "step": 925
+    },
+    {
+      "epoch": 0.19231568016614745,
+      "grad_norm": 0.4766390919685364,
+      "learning_rate": 8e-05,
+      "loss": 1.599,
+      "step": 926
+    },
+    {
+      "epoch": 0.1925233644859813,
+      "grad_norm": 0.48412805795669556,
+      "learning_rate": 8e-05,
+      "loss": 1.6749,
+      "step": 927
+    },
+    {
+      "epoch": 0.19273104880581515,
+      "grad_norm": 0.47555628418922424,
+      "learning_rate": 8e-05,
+      "loss": 1.5841,
+      "step": 928
+    },
+    {
+      "epoch": 0.19293873312564902,
+      "grad_norm": 0.48913460969924927,
+      "learning_rate": 8e-05,
+      "loss": 1.6512,
+      "step": 929
+    },
+    {
+      "epoch": 0.19314641744548286,
+      "grad_norm": 0.47544533014297485,
+      "learning_rate": 8e-05,
+      "loss": 1.6341,
+      "step": 930
+    },
+    {
+      "epoch": 0.19335410176531673,
+      "grad_norm": 0.4919319450855255,
+      "learning_rate": 8e-05,
+      "loss": 1.6381,
+      "step": 931
+    },
+    {
+      "epoch": 0.19356178608515057,
+      "grad_norm": 0.49583807587623596,
+      "learning_rate": 8e-05,
+      "loss": 1.6419,
+      "step": 932
+    },
+    {
+      "epoch": 0.19376947040498443,
+      "grad_norm": 0.5070441961288452,
+      "learning_rate": 8e-05,
+      "loss": 1.6657,
+      "step": 933
+    },
+    {
+      "epoch": 0.19397715472481827,
+      "grad_norm": 0.5061896443367004,
+      "learning_rate": 8e-05,
+      "loss": 1.6081,
+      "step": 934
+    },
+    {
+      "epoch": 0.19418483904465214,
+      "grad_norm": 0.5091527104377747,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 935
+    },
+    {
+      "epoch": 0.19439252336448598,
+      "grad_norm": 0.5117039084434509,
+      "learning_rate": 8e-05,
+      "loss": 1.632,
+      "step": 936
+    },
+    {
+      "epoch": 0.19460020768431982,
+      "grad_norm": 0.4791244864463806,
+      "learning_rate": 8e-05,
+      "loss": 1.6062,
+      "step": 937
+    },
+    {
+      "epoch": 0.1948078920041537,
+      "grad_norm": 0.5442696213722229,
+      "learning_rate": 8e-05,
+      "loss": 1.6536,
+      "step": 938
+    },
+    {
+      "epoch": 0.19501557632398753,
+      "grad_norm": 0.48307886719703674,
+      "learning_rate": 8e-05,
+      "loss": 1.5979,
+      "step": 939
+    },
+    {
+      "epoch": 0.1952232606438214,
+      "grad_norm": 0.5369094014167786,
+      "learning_rate": 8e-05,
+      "loss": 1.5968,
+      "step": 940
+    },
+    {
+      "epoch": 0.19543094496365523,
+      "grad_norm": 0.4949534237384796,
+      "learning_rate": 8e-05,
+      "loss": 1.6489,
+      "step": 941
+    },
+    {
+      "epoch": 0.1956386292834891,
+      "grad_norm": 0.5377048850059509,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 942
+    },
+    {
+      "epoch": 0.19584631360332294,
+      "grad_norm": 0.49918225407600403,
+      "learning_rate": 8e-05,
+      "loss": 1.5978,
+      "step": 943
+    },
+    {
+      "epoch": 0.1960539979231568,
+      "grad_norm": 0.5020906329154968,
+      "learning_rate": 8e-05,
+      "loss": 1.6248,
+      "step": 944
+    },
+    {
+      "epoch": 0.19626168224299065,
+      "grad_norm": 0.5110490918159485,
+      "learning_rate": 8e-05,
+      "loss": 1.5623,
+      "step": 945
+    },
+    {
+      "epoch": 0.19646936656282452,
+      "grad_norm": 0.4651744067668915,
+      "learning_rate": 8e-05,
+      "loss": 1.554,
+      "step": 946
+    },
+    {
+      "epoch": 0.19667705088265836,
+      "grad_norm": 0.5259853005409241,
+      "learning_rate": 8e-05,
+      "loss": 1.619,
+      "step": 947
+    },
+    {
+      "epoch": 0.19688473520249222,
+      "grad_norm": 0.5092069506645203,
+      "learning_rate": 8e-05,
+      "loss": 1.6075,
+      "step": 948
+    },
+    {
+      "epoch": 0.19709241952232606,
+      "grad_norm": 0.47999849915504456,
+      "learning_rate": 8e-05,
+      "loss": 1.4621,
+      "step": 949
+    },
+    {
+      "epoch": 0.19730010384215993,
+      "grad_norm": 0.5035858750343323,
+      "learning_rate": 8e-05,
+      "loss": 1.5704,
+      "step": 950
+    },
+    {
+      "epoch": 0.19750778816199377,
+      "grad_norm": 0.5029184222221375,
+      "learning_rate": 8e-05,
+      "loss": 1.691,
+      "step": 951
+    },
+    {
+      "epoch": 0.1977154724818276,
+      "grad_norm": 0.5592398047447205,
+      "learning_rate": 8e-05,
+      "loss": 1.6728,
+      "step": 952
+    },
+    {
+      "epoch": 0.19792315680166148,
+      "grad_norm": 0.5204004645347595,
+      "learning_rate": 8e-05,
+      "loss": 1.5816,
+      "step": 953
+    },
+    {
+      "epoch": 0.19813084112149532,
+      "grad_norm": 0.5047582983970642,
+      "learning_rate": 8e-05,
+      "loss": 1.6473,
+      "step": 954
+    },
+    {
+      "epoch": 0.19833852544132918,
+      "grad_norm": 0.5084308385848999,
+      "learning_rate": 8e-05,
+      "loss": 1.6297,
+      "step": 955
+    },
+    {
+      "epoch": 0.19854620976116302,
+      "grad_norm": 0.4851769506931305,
+      "learning_rate": 8e-05,
+      "loss": 1.676,
+      "step": 956
+    },
+    {
+      "epoch": 0.1987538940809969,
+      "grad_norm": 0.48817214369773865,
+      "learning_rate": 8e-05,
+      "loss": 1.6213,
+      "step": 957
+    },
+    {
+      "epoch": 0.19896157840083073,
+      "grad_norm": 0.5137952566146851,
+      "learning_rate": 8e-05,
+      "loss": 1.6807,
+      "step": 958
+    },
+    {
+      "epoch": 0.1991692627206646,
+      "grad_norm": 0.494804322719574,
+      "learning_rate": 8e-05,
+      "loss": 1.633,
+      "step": 959
+    },
+    {
+      "epoch": 0.19937694704049844,
+      "grad_norm": 0.4694535434246063,
+      "learning_rate": 8e-05,
+      "loss": 1.5782,
+      "step": 960
+    },
+    {
+      "epoch": 0.1995846313603323,
+      "grad_norm": 0.4649278223514557,
+      "learning_rate": 8e-05,
+      "loss": 1.5167,
+      "step": 961
+    },
+    {
+      "epoch": 0.19979231568016614,
+      "grad_norm": 0.49113568663597107,
+      "learning_rate": 8e-05,
+      "loss": 1.63,
+      "step": 962
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4700087308883667,
+      "learning_rate": 8e-05,
+      "loss": 1.5999,
+      "step": 963
+    },
+    {
+      "epoch": 0.20020768431983385,
+      "grad_norm": 0.4888930320739746,
+      "learning_rate": 8e-05,
+      "loss": 1.6966,
+      "step": 964
+    },
+    {
+      "epoch": 0.20041536863966772,
+      "grad_norm": 0.4866234362125397,
+      "learning_rate": 8e-05,
+      "loss": 1.6195,
+      "step": 965
+    },
+    {
+      "epoch": 0.20062305295950156,
+      "grad_norm": 0.5145866870880127,
+      "learning_rate": 8e-05,
+      "loss": 1.6721,
+      "step": 966
+    },
+    {
+      "epoch": 0.2008307372793354,
+      "grad_norm": 0.47302988171577454,
+      "learning_rate": 8e-05,
+      "loss": 1.6644,
+      "step": 967
+    },
+    {
+      "epoch": 0.20103842159916926,
+      "grad_norm": 0.48115986585617065,
+      "learning_rate": 8e-05,
+      "loss": 1.6557,
+      "step": 968
+    },
+    {
+      "epoch": 0.2012461059190031,
+      "grad_norm": 0.4676402807235718,
+      "learning_rate": 8e-05,
+      "loss": 1.5842,
+      "step": 969
+    },
+    {
+      "epoch": 0.20145379023883697,
+      "grad_norm": 0.4860267639160156,
+      "learning_rate": 8e-05,
+      "loss": 1.6519,
+      "step": 970
+    },
+    {
+      "epoch": 0.2016614745586708,
+      "grad_norm": 0.4892720580101013,
+      "learning_rate": 8e-05,
+      "loss": 1.5545,
+      "step": 971
+    },
+    {
+      "epoch": 0.20186915887850468,
+      "grad_norm": 0.46676620841026306,
+      "learning_rate": 8e-05,
+      "loss": 1.553,
+      "step": 972
+    },
+    {
+      "epoch": 0.20207684319833852,
+      "grad_norm": 0.47388455271720886,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 973
+    },
+    {
+      "epoch": 0.20228452751817239,
+      "grad_norm": 0.4880634546279907,
+      "learning_rate": 8e-05,
+      "loss": 1.5832,
+      "step": 974
+    },
+    {
+      "epoch": 0.20249221183800623,
+      "grad_norm": 0.4873626232147217,
+      "learning_rate": 8e-05,
+      "loss": 1.6224,
+      "step": 975
+    },
+    {
+      "epoch": 0.2026998961578401,
+      "grad_norm": 0.5022002458572388,
+      "learning_rate": 8e-05,
+      "loss": 1.6834,
+      "step": 976
+    },
+    {
+      "epoch": 0.20290758047767393,
+      "grad_norm": 0.47066470980644226,
+      "learning_rate": 8e-05,
+      "loss": 1.5887,
+      "step": 977
+    },
+    {
+      "epoch": 0.2031152647975078,
+      "grad_norm": 0.4675566554069519,
+      "learning_rate": 8e-05,
+      "loss": 1.6026,
+      "step": 978
+    },
+    {
+      "epoch": 0.20332294911734164,
+      "grad_norm": 0.5087757110595703,
+      "learning_rate": 8e-05,
+      "loss": 1.6192,
+      "step": 979
+    },
+    {
+      "epoch": 0.2035306334371755,
+      "grad_norm": 0.49187105894088745,
+      "learning_rate": 8e-05,
+      "loss": 1.6126,
+      "step": 980
+    },
+    {
+      "epoch": 0.20373831775700935,
+      "grad_norm": 0.4877180755138397,
+      "learning_rate": 8e-05,
+      "loss": 1.6781,
+      "step": 981
+    },
+    {
+      "epoch": 0.20394600207684319,
+      "grad_norm": 0.47267571091651917,
+      "learning_rate": 8e-05,
+      "loss": 1.646,
+      "step": 982
+    },
+    {
+      "epoch": 0.20415368639667705,
+      "grad_norm": 0.4767378270626068,
+      "learning_rate": 8e-05,
+      "loss": 1.5953,
+      "step": 983
+    },
+    {
+      "epoch": 0.2043613707165109,
+      "grad_norm": 0.48661693930625916,
+      "learning_rate": 8e-05,
+      "loss": 1.5469,
+      "step": 984
+    },
+    {
+      "epoch": 0.20456905503634476,
+      "grad_norm": 0.4928230047225952,
+      "learning_rate": 8e-05,
+      "loss": 1.7054,
+      "step": 985
+    },
+    {
+      "epoch": 0.2047767393561786,
+      "grad_norm": 0.4975184202194214,
+      "learning_rate": 8e-05,
+      "loss": 1.594,
+      "step": 986
+    },
+    {
+      "epoch": 0.20498442367601247,
+      "grad_norm": 0.4876691997051239,
+      "learning_rate": 8e-05,
+      "loss": 1.6356,
+      "step": 987
+    },
+    {
+      "epoch": 0.2051921079958463,
+      "grad_norm": 0.4785061776638031,
+      "learning_rate": 8e-05,
+      "loss": 1.6646,
+      "step": 988
+    },
+    {
+      "epoch": 0.20539979231568017,
+      "grad_norm": 0.49012821912765503,
+      "learning_rate": 8e-05,
+      "loss": 1.689,
+      "step": 989
+    },
+    {
+      "epoch": 0.205607476635514,
+      "grad_norm": 0.45855405926704407,
+      "learning_rate": 8e-05,
+      "loss": 1.6318,
+      "step": 990
+    },
+    {
+      "epoch": 0.20581516095534788,
+      "grad_norm": 0.4825817048549652,
+      "learning_rate": 8e-05,
+      "loss": 1.6189,
+      "step": 991
+    },
+    {
+      "epoch": 0.20602284527518172,
+      "grad_norm": 0.4723818004131317,
+      "learning_rate": 8e-05,
+      "loss": 1.5096,
+      "step": 992
+    },
+    {
+      "epoch": 0.2062305295950156,
+      "grad_norm": 0.46535414457321167,
+      "learning_rate": 8e-05,
+      "loss": 1.6112,
+      "step": 993
+    },
+    {
+      "epoch": 0.20643821391484943,
+      "grad_norm": 0.4890800714492798,
+      "learning_rate": 8e-05,
+      "loss": 1.6472,
+      "step": 994
+    },
+    {
+      "epoch": 0.2066458982346833,
+      "grad_norm": 0.48340532183647156,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 995
+    },
+    {
+      "epoch": 0.20685358255451713,
+      "grad_norm": 0.490472674369812,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 996
+    },
+    {
+      "epoch": 0.20706126687435097,
+      "grad_norm": 0.5247419476509094,
+      "learning_rate": 8e-05,
+      "loss": 1.7016,
+      "step": 997
+    },
+    {
+      "epoch": 0.20726895119418484,
+      "grad_norm": 0.4798457622528076,
+      "learning_rate": 8e-05,
+      "loss": 1.6581,
+      "step": 998
+    },
+    {
+      "epoch": 0.20747663551401868,
+      "grad_norm": 0.5022425651550293,
+      "learning_rate": 8e-05,
+      "loss": 1.6491,
+      "step": 999
+    },
+    {
+      "epoch": 0.20768431983385255,
+      "grad_norm": 0.5252692699432373,
+      "learning_rate": 8e-05,
+      "loss": 1.624,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2078920041536864,
+      "grad_norm": 0.4893631041049957,
+      "learning_rate": 8e-05,
+      "loss": 1.6016,
+      "step": 1001
+    },
+    {
+      "epoch": 0.20809968847352026,
+      "grad_norm": 0.49601900577545166,
+      "learning_rate": 8e-05,
+      "loss": 1.6679,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2083073727933541,
+      "grad_norm": 0.484885036945343,
+      "learning_rate": 8e-05,
+      "loss": 1.5687,
+      "step": 1003
+    },
+    {
+      "epoch": 0.20851505711318796,
+      "grad_norm": 0.4795399606227875,
+      "learning_rate": 8e-05,
+      "loss": 1.5629,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2087227414330218,
+      "grad_norm": 0.5045424699783325,
+      "learning_rate": 8e-05,
+      "loss": 1.6474,
+      "step": 1005
+    },
+    {
+      "epoch": 0.20893042575285567,
+      "grad_norm": 0.49389609694480896,
+      "learning_rate": 8e-05,
+      "loss": 1.6446,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2091381100726895,
+      "grad_norm": 0.5181095600128174,
+      "learning_rate": 8e-05,
+      "loss": 1.6578,
+      "step": 1007
+    },
+    {
+      "epoch": 0.20934579439252338,
+      "grad_norm": 0.4928913414478302,
+      "learning_rate": 8e-05,
+      "loss": 1.6385,
+      "step": 1008
+    },
+    {
+      "epoch": 0.20955347871235722,
+      "grad_norm": 0.5067183971405029,
+      "learning_rate": 8e-05,
+      "loss": 1.5462,
+      "step": 1009
+    },
+    {
+      "epoch": 0.20976116303219106,
+      "grad_norm": 0.46599841117858887,
+      "learning_rate": 8e-05,
+      "loss": 1.6113,
+      "step": 1010
+    },
+    {
+      "epoch": 0.20996884735202492,
+      "grad_norm": 0.5027761459350586,
+      "learning_rate": 8e-05,
+      "loss": 1.6009,
+      "step": 1011
+    },
+    {
+      "epoch": 0.21017653167185876,
+      "grad_norm": 0.5024246573448181,
+      "learning_rate": 8e-05,
+      "loss": 1.5819,
+      "step": 1012
+    },
+    {
+      "epoch": 0.21038421599169263,
+      "grad_norm": 0.4743015468120575,
+      "learning_rate": 8e-05,
+      "loss": 1.6115,
+      "step": 1013
+    },
+    {
+      "epoch": 0.21059190031152647,
+      "grad_norm": 0.5321120619773865,
+      "learning_rate": 8e-05,
+      "loss": 1.6409,
+      "step": 1014
+    },
+    {
+      "epoch": 0.21079958463136034,
+      "grad_norm": 0.4797392189502716,
+      "learning_rate": 8e-05,
+      "loss": 1.5711,
+      "step": 1015
+    },
+    {
+      "epoch": 0.21100726895119418,
+      "grad_norm": 0.5282681584358215,
+      "learning_rate": 8e-05,
+      "loss": 1.6369,
+      "step": 1016
+    },
+    {
+      "epoch": 0.21121495327102804,
+      "grad_norm": 0.4802643954753876,
+      "learning_rate": 8e-05,
+      "loss": 1.6363,
+      "step": 1017
+    },
+    {
+      "epoch": 0.21142263759086188,
+      "grad_norm": 0.479918897151947,
+      "learning_rate": 8e-05,
+      "loss": 1.6828,
+      "step": 1018
+    },
+    {
+      "epoch": 0.21163032191069575,
+      "grad_norm": 0.485012412071228,
+      "learning_rate": 8e-05,
+      "loss": 1.6451,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2118380062305296,
+      "grad_norm": 0.47825416922569275,
+      "learning_rate": 8e-05,
+      "loss": 1.6492,
+      "step": 1020
+    },
+    {
+      "epoch": 0.21204569055036346,
+      "grad_norm": 0.48356735706329346,
+      "learning_rate": 8e-05,
+      "loss": 1.5729,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2122533748701973,
+      "grad_norm": 0.4711926579475403,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 1022
+    },
+    {
+      "epoch": 0.21246105919003117,
+      "grad_norm": 0.4887542128562927,
+      "learning_rate": 8e-05,
+      "loss": 1.6684,
+      "step": 1023
+    },
+    {
+      "epoch": 0.212668743509865,
+      "grad_norm": 0.47914424538612366,
+      "learning_rate": 8e-05,
+      "loss": 1.5889,
+      "step": 1024
+    },
+    {
+      "epoch": 0.21287642782969884,
+      "grad_norm": 0.49408769607543945,
+      "learning_rate": 8e-05,
+      "loss": 1.663,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2130841121495327,
+      "grad_norm": 0.49186617136001587,
+      "learning_rate": 8e-05,
+      "loss": 1.657,
+      "step": 1026
+    },
+    {
+      "epoch": 0.21329179646936655,
+      "grad_norm": 0.5070930123329163,
+      "learning_rate": 8e-05,
+      "loss": 1.6457,
+      "step": 1027
+    },
+    {
+      "epoch": 0.21349948078920042,
+      "grad_norm": 0.4801136553287506,
+      "learning_rate": 8e-05,
+      "loss": 1.6029,
+      "step": 1028
+    },
+    {
+      "epoch": 0.21370716510903426,
+      "grad_norm": 0.4819825291633606,
+      "learning_rate": 8e-05,
+      "loss": 1.6109,
+      "step": 1029
+    },
+    {
+      "epoch": 0.21391484942886813,
+      "grad_norm": 0.5037396550178528,
+      "learning_rate": 8e-05,
+      "loss": 1.6871,
+      "step": 1030
+    },
+    {
+      "epoch": 0.21412253374870197,
+      "grad_norm": 0.5130371451377869,
+      "learning_rate": 8e-05,
+      "loss": 1.7099,
+      "step": 1031
+    },
+    {
+      "epoch": 0.21433021806853583,
+      "grad_norm": 0.5002063512802124,
+      "learning_rate": 8e-05,
+      "loss": 1.6737,
+      "step": 1032
+    },
+    {
+      "epoch": 0.21453790238836967,
+      "grad_norm": 0.4774034917354584,
+      "learning_rate": 8e-05,
+      "loss": 1.6461,
+      "step": 1033
+    },
+    {
+      "epoch": 0.21474558670820354,
+      "grad_norm": 0.48661088943481445,
+      "learning_rate": 8e-05,
+      "loss": 1.5844,
+      "step": 1034
+    },
+    {
+      "epoch": 0.21495327102803738,
+      "grad_norm": 0.4912685453891754,
+      "learning_rate": 8e-05,
+      "loss": 1.6028,
+      "step": 1035
+    },
+    {
+      "epoch": 0.21516095534787125,
+      "grad_norm": 0.5158385634422302,
+      "learning_rate": 8e-05,
+      "loss": 1.6254,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2153686396677051,
+      "grad_norm": 0.4866856038570404,
+      "learning_rate": 8e-05,
+      "loss": 1.654,
+      "step": 1037
+    },
+    {
+      "epoch": 0.21557632398753895,
+      "grad_norm": 0.5178074836730957,
+      "learning_rate": 8e-05,
+      "loss": 1.6212,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2157840083073728,
+      "grad_norm": 0.48483380675315857,
+      "learning_rate": 8e-05,
+      "loss": 1.6116,
+      "step": 1039
+    },
+    {
+      "epoch": 0.21599169262720663,
+      "grad_norm": 0.47141504287719727,
+      "learning_rate": 8e-05,
+      "loss": 1.6042,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2161993769470405,
+      "grad_norm": 0.5006888508796692,
+      "learning_rate": 8e-05,
+      "loss": 1.6242,
+      "step": 1041
+    },
+    {
+      "epoch": 0.21640706126687434,
+      "grad_norm": 0.47855380177497864,
+      "learning_rate": 8e-05,
+      "loss": 1.5661,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2166147455867082,
+      "grad_norm": 0.5185720324516296,
+      "learning_rate": 8e-05,
+      "loss": 1.6747,
+      "step": 1043
+    },
+    {
+      "epoch": 0.21682242990654205,
+      "grad_norm": 0.4679183065891266,
+      "learning_rate": 8e-05,
+      "loss": 1.534,
+      "step": 1044
+    },
+    {
+      "epoch": 0.21703011422637591,
+      "grad_norm": 0.5030350685119629,
+      "learning_rate": 8e-05,
+      "loss": 1.624,
+      "step": 1045
+    },
+    {
+      "epoch": 0.21723779854620975,
+      "grad_norm": 0.48393726348876953,
+      "learning_rate": 8e-05,
+      "loss": 1.6531,
+      "step": 1046
+    },
+    {
+      "epoch": 0.21744548286604362,
+      "grad_norm": 0.4825136065483093,
+      "learning_rate": 8e-05,
+      "loss": 1.6242,
+      "step": 1047
+    },
+    {
+      "epoch": 0.21765316718587746,
+      "grad_norm": 0.4966621696949005,
+      "learning_rate": 8e-05,
+      "loss": 1.6368,
+      "step": 1048
+    },
+    {
+      "epoch": 0.21786085150571133,
+      "grad_norm": 0.48175907135009766,
+      "learning_rate": 8e-05,
+      "loss": 1.6121,
+      "step": 1049
+    },
+    {
+      "epoch": 0.21806853582554517,
+      "grad_norm": 0.4931303858757019,
+      "learning_rate": 8e-05,
+      "loss": 1.6452,
+      "step": 1050
+    },
+    {
+      "epoch": 0.21827622014537904,
+      "grad_norm": 0.4742048680782318,
+      "learning_rate": 8e-05,
+      "loss": 1.6037,
+      "step": 1051
+    },
+    {
+      "epoch": 0.21848390446521287,
+      "grad_norm": 0.4972934126853943,
+      "learning_rate": 8e-05,
+      "loss": 1.6522,
+      "step": 1052
+    },
+    {
+      "epoch": 0.21869158878504674,
+      "grad_norm": 0.49021288752555847,
+      "learning_rate": 8e-05,
+      "loss": 1.6386,
+      "step": 1053
+    },
+    {
+      "epoch": 0.21889927310488058,
+      "grad_norm": 0.4739423096179962,
+      "learning_rate": 8e-05,
+      "loss": 1.6267,
+      "step": 1054
+    },
+    {
+      "epoch": 0.21910695742471442,
+      "grad_norm": 0.49233463406562805,
+      "learning_rate": 8e-05,
+      "loss": 1.6757,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2193146417445483,
+      "grad_norm": 0.4631211757659912,
+      "learning_rate": 8e-05,
+      "loss": 1.5198,
+      "step": 1056
+    },
+    {
+      "epoch": 0.21952232606438213,
+      "grad_norm": 0.5057953596115112,
+      "learning_rate": 8e-05,
+      "loss": 1.6831,
+      "step": 1057
+    },
+    {
+      "epoch": 0.219730010384216,
+      "grad_norm": 0.46444934606552124,
+      "learning_rate": 8e-05,
+      "loss": 1.6107,
+      "step": 1058
+    },
+    {
+      "epoch": 0.21993769470404984,
+      "grad_norm": 0.48006969690322876,
+      "learning_rate": 8e-05,
+      "loss": 1.6356,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2201453790238837,
+      "grad_norm": 0.47697484493255615,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 1060
+    },
+    {
+      "epoch": 0.22035306334371754,
+      "grad_norm": 0.45909324288368225,
+      "learning_rate": 8e-05,
+      "loss": 1.6047,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2205607476635514,
+      "grad_norm": 0.45758333802223206,
+      "learning_rate": 8e-05,
+      "loss": 1.5597,
+      "step": 1062
+    },
+    {
+      "epoch": 0.22076843198338525,
+      "grad_norm": 0.49077147245407104,
+      "learning_rate": 8e-05,
+      "loss": 1.6428,
+      "step": 1063
+    },
+    {
+      "epoch": 0.22097611630321912,
+      "grad_norm": 0.49720385670661926,
+      "learning_rate": 8e-05,
+      "loss": 1.6476,
+      "step": 1064
+    },
+    {
+      "epoch": 0.22118380062305296,
+      "grad_norm": 0.5079086422920227,
+      "learning_rate": 8e-05,
+      "loss": 1.6564,
+      "step": 1065
+    },
+    {
+      "epoch": 0.22139148494288682,
+      "grad_norm": 0.47034189105033875,
+      "learning_rate": 8e-05,
+      "loss": 1.5323,
+      "step": 1066
+    },
+    {
+      "epoch": 0.22159916926272066,
+      "grad_norm": 0.5000166893005371,
+      "learning_rate": 8e-05,
+      "loss": 1.6268,
+      "step": 1067
+    },
+    {
+      "epoch": 0.22180685358255453,
+      "grad_norm": 0.47734546661376953,
+      "learning_rate": 8e-05,
+      "loss": 1.6071,
+      "step": 1068
+    },
+    {
+      "epoch": 0.22201453790238837,
+      "grad_norm": 0.49549058079719543,
+      "learning_rate": 8e-05,
+      "loss": 1.6152,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.4837387502193451,
+      "learning_rate": 8e-05,
+      "loss": 1.6049,
+      "step": 1070
+    },
+    {
+      "epoch": 0.22242990654205608,
+      "grad_norm": 0.46780574321746826,
+      "learning_rate": 8e-05,
+      "loss": 1.5809,
+      "step": 1071
+    },
+    {
+      "epoch": 0.22263759086188992,
+      "grad_norm": 0.4783675968647003,
+      "learning_rate": 8e-05,
+      "loss": 1.5655,
+      "step": 1072
+    },
+    {
+      "epoch": 0.22284527518172378,
+      "grad_norm": 0.4945738613605499,
+      "learning_rate": 8e-05,
+      "loss": 1.6605,
+      "step": 1073
+    },
+    {
+      "epoch": 0.22305295950155762,
+      "grad_norm": 0.5201888084411621,
+      "learning_rate": 8e-05,
+      "loss": 1.6003,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2232606438213915,
+      "grad_norm": 0.5067545175552368,
+      "learning_rate": 8e-05,
+      "loss": 1.6575,
+      "step": 1075
+    },
+    {
+      "epoch": 0.22346832814122533,
+      "grad_norm": 0.4898606538772583,
+      "learning_rate": 8e-05,
+      "loss": 1.6554,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2236760124610592,
+      "grad_norm": 0.5530943274497986,
+      "learning_rate": 8e-05,
+      "loss": 1.6715,
+      "step": 1077
+    },
+    {
+      "epoch": 0.22388369678089304,
+      "grad_norm": 0.4874219596385956,
+      "learning_rate": 8e-05,
+      "loss": 1.6182,
+      "step": 1078
+    },
+    {
+      "epoch": 0.2240913811007269,
+      "grad_norm": 0.5321915745735168,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 1079
+    },
+    {
+      "epoch": 0.22429906542056074,
+      "grad_norm": 0.46676793694496155,
+      "learning_rate": 8e-05,
+      "loss": 1.6459,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2245067497403946,
+      "grad_norm": 0.4740728735923767,
+      "learning_rate": 8e-05,
+      "loss": 1.5671,
+      "step": 1081
+    },
+    {
+      "epoch": 0.22471443406022845,
+      "grad_norm": 0.48530635237693787,
+      "learning_rate": 8e-05,
+      "loss": 1.5294,
+      "step": 1082
+    },
+    {
+      "epoch": 0.22492211838006232,
+      "grad_norm": 0.4611906111240387,
+      "learning_rate": 8e-05,
+      "loss": 1.6403,
+      "step": 1083
+    },
+    {
+      "epoch": 0.22512980269989616,
+      "grad_norm": 0.4903087913990021,
+      "learning_rate": 8e-05,
+      "loss": 1.6264,
+      "step": 1084
+    },
+    {
+      "epoch": 0.22533748701973,
+      "grad_norm": 0.486613392829895,
+      "learning_rate": 8e-05,
+      "loss": 1.6315,
+      "step": 1085
+    },
+    {
+      "epoch": 0.22554517133956387,
+      "grad_norm": 0.5356956124305725,
+      "learning_rate": 8e-05,
+      "loss": 1.6754,
+      "step": 1086
+    },
+    {
+      "epoch": 0.2257528556593977,
+      "grad_norm": 0.5136194825172424,
+      "learning_rate": 8e-05,
+      "loss": 1.6052,
+      "step": 1087
+    },
+    {
+      "epoch": 0.22596053997923157,
+      "grad_norm": 0.49597111344337463,
+      "learning_rate": 8e-05,
+      "loss": 1.6403,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2261682242990654,
+      "grad_norm": 0.48345381021499634,
+      "learning_rate": 8e-05,
+      "loss": 1.6072,
+      "step": 1089
+    },
+    {
+      "epoch": 0.22637590861889928,
+      "grad_norm": 0.505710780620575,
+      "learning_rate": 8e-05,
+      "loss": 1.5855,
+      "step": 1090
+    },
+    {
+      "epoch": 0.22658359293873312,
+      "grad_norm": 0.5041759014129639,
+      "learning_rate": 8e-05,
+      "loss": 1.6558,
+      "step": 1091
+    },
+    {
+      "epoch": 0.226791277258567,
+      "grad_norm": 0.5124122500419617,
+      "learning_rate": 8e-05,
+      "loss": 1.6171,
+      "step": 1092
+    },
+    {
+      "epoch": 0.22699896157840083,
+      "grad_norm": 0.5028002262115479,
+      "learning_rate": 8e-05,
+      "loss": 1.667,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2272066458982347,
+      "grad_norm": 0.5096011161804199,
+      "learning_rate": 8e-05,
+      "loss": 1.6043,
+      "step": 1094
+    },
+    {
+      "epoch": 0.22741433021806853,
+      "grad_norm": 0.4749337136745453,
+      "learning_rate": 8e-05,
+      "loss": 1.6256,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2276220145379024,
+      "grad_norm": 0.4887832701206207,
+      "learning_rate": 8e-05,
+      "loss": 1.6197,
+      "step": 1096
+    },
+    {
+      "epoch": 0.22782969885773624,
+      "grad_norm": 0.4867877662181854,
+      "learning_rate": 8e-05,
+      "loss": 1.6318,
+      "step": 1097
+    },
+    {
+      "epoch": 0.22803738317757008,
+      "grad_norm": 0.45605990290641785,
+      "learning_rate": 8e-05,
+      "loss": 1.5584,
+      "step": 1098
+    },
+    {
+      "epoch": 0.22824506749740395,
+      "grad_norm": 0.48189279437065125,
+      "learning_rate": 8e-05,
+      "loss": 1.618,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2284527518172378,
+      "grad_norm": 0.510169506072998,
+      "learning_rate": 8e-05,
+      "loss": 1.6289,
+      "step": 1100
+    },
+    {
+      "epoch": 0.22866043613707165,
+      "grad_norm": 0.47950589656829834,
+      "learning_rate": 8e-05,
+      "loss": 1.6601,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2288681204569055,
+      "grad_norm": 0.5073772072792053,
+      "learning_rate": 8e-05,
+      "loss": 1.6404,
+      "step": 1102
+    },
+    {
+      "epoch": 0.22907580477673936,
+      "grad_norm": 0.47775015234947205,
+      "learning_rate": 8e-05,
+      "loss": 1.6235,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2292834890965732,
+      "grad_norm": 0.4975552558898926,
+      "learning_rate": 8e-05,
+      "loss": 1.6022,
+      "step": 1104
+    },
+    {
+      "epoch": 0.22949117341640707,
+      "grad_norm": 0.49373266100883484,
+      "learning_rate": 8e-05,
+      "loss": 1.631,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2296988577362409,
+      "grad_norm": 0.5255943536758423,
+      "learning_rate": 8e-05,
+      "loss": 1.6067,
+      "step": 1106
+    },
+    {
+      "epoch": 0.22990654205607478,
+      "grad_norm": 0.4758288264274597,
+      "learning_rate": 8e-05,
+      "loss": 1.5626,
+      "step": 1107
+    },
+    {
+      "epoch": 0.23011422637590861,
+      "grad_norm": 0.4850333631038666,
+      "learning_rate": 8e-05,
+      "loss": 1.5805,
+      "step": 1108
+    },
+    {
+      "epoch": 0.23032191069574248,
+      "grad_norm": 0.4803512990474701,
+      "learning_rate": 8e-05,
+      "loss": 1.5439,
+      "step": 1109
+    },
+    {
+      "epoch": 0.23052959501557632,
+      "grad_norm": 0.4789465665817261,
+      "learning_rate": 8e-05,
+      "loss": 1.6022,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2307372793354102,
+      "grad_norm": 0.5042217969894409,
+      "learning_rate": 8e-05,
+      "loss": 1.6692,
+      "step": 1111
+    },
+    {
+      "epoch": 0.23094496365524403,
+      "grad_norm": 0.497745543718338,
+      "learning_rate": 8e-05,
+      "loss": 1.6743,
+      "step": 1112
+    },
+    {
+      "epoch": 0.23115264797507787,
+      "grad_norm": 0.45986318588256836,
+      "learning_rate": 8e-05,
+      "loss": 1.5806,
+      "step": 1113
+    },
+    {
+      "epoch": 0.23136033229491174,
+      "grad_norm": 0.4838648736476898,
+      "learning_rate": 8e-05,
+      "loss": 1.653,
+      "step": 1114
+    },
+    {
+      "epoch": 0.23156801661474558,
+      "grad_norm": 0.4738873243331909,
+      "learning_rate": 8e-05,
+      "loss": 1.5853,
+      "step": 1115
+    },
+    {
+      "epoch": 0.23177570093457944,
+      "grad_norm": 0.48498794436454773,
+      "learning_rate": 8e-05,
+      "loss": 1.673,
+      "step": 1116
+    },
+    {
+      "epoch": 0.23198338525441328,
+      "grad_norm": 0.4791082739830017,
+      "learning_rate": 8e-05,
+      "loss": 1.5446,
+      "step": 1117
+    },
+    {
+      "epoch": 0.23219106957424715,
+      "grad_norm": 0.4974493086338043,
+      "learning_rate": 8e-05,
+      "loss": 1.5747,
+      "step": 1118
+    },
+    {
+      "epoch": 0.232398753894081,
+      "grad_norm": 0.5050761103630066,
+      "learning_rate": 8e-05,
+      "loss": 1.5965,
+      "step": 1119
+    },
+    {
+      "epoch": 0.23260643821391486,
+      "grad_norm": 0.47682154178619385,
+      "learning_rate": 8e-05,
+      "loss": 1.581,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2328141225337487,
+      "grad_norm": 0.5180225372314453,
+      "learning_rate": 8e-05,
+      "loss": 1.6094,
+      "step": 1121
+    },
+    {
+      "epoch": 0.23302180685358256,
+      "grad_norm": 0.5043385028839111,
+      "learning_rate": 8e-05,
+      "loss": 1.6338,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2332294911734164,
+      "grad_norm": 0.5274357199668884,
+      "learning_rate": 8e-05,
+      "loss": 1.6203,
+      "step": 1123
+    },
+    {
+      "epoch": 0.23343717549325027,
+      "grad_norm": 0.49157971143722534,
+      "learning_rate": 8e-05,
+      "loss": 1.5854,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2336448598130841,
+      "grad_norm": 0.5090160369873047,
+      "learning_rate": 8e-05,
+      "loss": 1.6155,
+      "step": 1125
+    },
+    {
+      "epoch": 0.23385254413291798,
+      "grad_norm": 0.4960342049598694,
+      "learning_rate": 8e-05,
+      "loss": 1.6501,
+      "step": 1126
+    },
+    {
+      "epoch": 0.23406022845275182,
+      "grad_norm": 0.49579086899757385,
+      "learning_rate": 8e-05,
+      "loss": 1.6299,
+      "step": 1127
+    },
+    {
+      "epoch": 0.23426791277258566,
+      "grad_norm": 0.5103940963745117,
+      "learning_rate": 8e-05,
+      "loss": 1.5807,
+      "step": 1128
+    },
+    {
+      "epoch": 0.23447559709241952,
+      "grad_norm": 0.47810351848602295,
+      "learning_rate": 8e-05,
+      "loss": 1.6593,
+      "step": 1129
+    },
+    {
+      "epoch": 0.23468328141225336,
+      "grad_norm": 0.48690786957740784,
+      "learning_rate": 8e-05,
+      "loss": 1.5911,
+      "step": 1130
+    },
+    {
+      "epoch": 0.23489096573208723,
+      "grad_norm": 0.5075724720954895,
+      "learning_rate": 8e-05,
+      "loss": 1.6673,
+      "step": 1131
+    },
+    {
+      "epoch": 0.23509865005192107,
+      "grad_norm": 0.5023790001869202,
+      "learning_rate": 8e-05,
+      "loss": 1.6244,
+      "step": 1132
+    },
+    {
+      "epoch": 0.23530633437175494,
+      "grad_norm": 0.49479010701179504,
+      "learning_rate": 8e-05,
+      "loss": 1.5765,
+      "step": 1133
+    },
+    {
+      "epoch": 0.23551401869158878,
+      "grad_norm": 0.47907403111457825,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 1134
+    },
+    {
+      "epoch": 0.23572170301142265,
+      "grad_norm": 0.5096808075904846,
+      "learning_rate": 8e-05,
+      "loss": 1.5931,
+      "step": 1135
+    },
+    {
+      "epoch": 0.23592938733125648,
+      "grad_norm": 0.4935376048088074,
+      "learning_rate": 8e-05,
+      "loss": 1.6679,
+      "step": 1136
+    },
+    {
+      "epoch": 0.23613707165109035,
+      "grad_norm": 0.4955846667289734,
+      "learning_rate": 8e-05,
+      "loss": 1.6786,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2363447559709242,
+      "grad_norm": 0.5047092437744141,
+      "learning_rate": 8e-05,
+      "loss": 1.5719,
+      "step": 1138
+    },
+    {
+      "epoch": 0.23655244029075806,
+      "grad_norm": 0.4767501950263977,
+      "learning_rate": 8e-05,
+      "loss": 1.5391,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2367601246105919,
+      "grad_norm": 0.49148985743522644,
+      "learning_rate": 8e-05,
+      "loss": 1.6281,
+      "step": 1140
+    },
+    {
+      "epoch": 0.23696780893042577,
+      "grad_norm": 0.4789850413799286,
+      "learning_rate": 8e-05,
+      "loss": 1.541,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2371754932502596,
+      "grad_norm": 0.47179505228996277,
+      "learning_rate": 8e-05,
+      "loss": 1.56,
+      "step": 1142
+    },
+    {
+      "epoch": 0.23738317757009345,
+      "grad_norm": 0.4852232038974762,
+      "learning_rate": 8e-05,
+      "loss": 1.6166,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2375908618899273,
+      "grad_norm": 0.4789161682128906,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 1144
+    },
+    {
+      "epoch": 0.23779854620976115,
+      "grad_norm": 0.5053293108940125,
+      "learning_rate": 8e-05,
+      "loss": 1.6531,
+      "step": 1145
+    },
+    {
+      "epoch": 0.23800623052959502,
+      "grad_norm": 0.48226839303970337,
+      "learning_rate": 8e-05,
+      "loss": 1.6138,
+      "step": 1146
+    },
+    {
+      "epoch": 0.23821391484942886,
+      "grad_norm": 0.4766741693019867,
+      "learning_rate": 8e-05,
+      "loss": 1.5757,
+      "step": 1147
+    },
+    {
+      "epoch": 0.23842159916926273,
+      "grad_norm": 0.4817132353782654,
+      "learning_rate": 8e-05,
+      "loss": 1.5919,
+      "step": 1148
+    },
+    {
+      "epoch": 0.23862928348909657,
+      "grad_norm": 0.48039570450782776,
+      "learning_rate": 8e-05,
+      "loss": 1.5237,
+      "step": 1149
+    },
+    {
+      "epoch": 0.23883696780893043,
+      "grad_norm": 0.539954662322998,
+      "learning_rate": 8e-05,
+      "loss": 1.7231,
+      "step": 1150
+    },
+    {
+      "epoch": 0.23904465212876427,
+      "grad_norm": 0.49873772263526917,
+      "learning_rate": 8e-05,
+      "loss": 1.6099,
+      "step": 1151
+    },
+    {
+      "epoch": 0.23925233644859814,
+      "grad_norm": 0.5055965185165405,
+      "learning_rate": 8e-05,
+      "loss": 1.6688,
+      "step": 1152
+    },
+    {
+      "epoch": 0.23946002076843198,
+      "grad_norm": 0.4793665409088135,
+      "learning_rate": 8e-05,
+      "loss": 1.6186,
+      "step": 1153
+    },
+    {
+      "epoch": 0.23966770508826585,
+      "grad_norm": 0.49260565638542175,
+      "learning_rate": 8e-05,
+      "loss": 1.5712,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2398753894080997,
+      "grad_norm": 0.48806118965148926,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 1155
+    },
+    {
+      "epoch": 0.24008307372793355,
+      "grad_norm": 0.4877453148365021,
+      "learning_rate": 8e-05,
+      "loss": 1.5825,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2402907580477674,
+      "grad_norm": 0.4928903877735138,
+      "learning_rate": 8e-05,
+      "loss": 1.6033,
+      "step": 1157
+    },
+    {
+      "epoch": 0.24049844236760123,
+      "grad_norm": 0.4725443124771118,
+      "learning_rate": 8e-05,
+      "loss": 1.6135,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2407061266874351,
+      "grad_norm": 0.47031083703041077,
+      "learning_rate": 8e-05,
+      "loss": 1.6,
+      "step": 1159
+    },
+    {
+      "epoch": 0.24091381100726894,
+      "grad_norm": 0.46935370564460754,
+      "learning_rate": 8e-05,
+      "loss": 1.574,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2411214953271028,
+      "grad_norm": 0.4930771291255951,
+      "learning_rate": 8e-05,
+      "loss": 1.5769,
+      "step": 1161
+    },
+    {
+      "epoch": 0.24132917964693665,
+      "grad_norm": 0.4878680408000946,
+      "learning_rate": 8e-05,
+      "loss": 1.6057,
+      "step": 1162
+    },
+    {
+      "epoch": 0.24153686396677052,
+      "grad_norm": 0.46639183163642883,
+      "learning_rate": 8e-05,
+      "loss": 1.5862,
+      "step": 1163
+    },
+    {
+      "epoch": 0.24174454828660435,
+      "grad_norm": 0.48250576853752136,
+      "learning_rate": 8e-05,
+      "loss": 1.5948,
+      "step": 1164
+    },
+    {
+      "epoch": 0.24195223260643822,
+      "grad_norm": 0.4780838191509247,
+      "learning_rate": 8e-05,
+      "loss": 1.5429,
+      "step": 1165
+    },
+    {
+      "epoch": 0.24215991692627206,
+      "grad_norm": 0.4783189296722412,
+      "learning_rate": 8e-05,
+      "loss": 1.6205,
+      "step": 1166
+    },
+    {
+      "epoch": 0.24236760124610593,
+      "grad_norm": 0.4721772372722626,
+      "learning_rate": 8e-05,
+      "loss": 1.5667,
+      "step": 1167
+    },
+    {
+      "epoch": 0.24257528556593977,
+      "grad_norm": 0.49488070607185364,
+      "learning_rate": 8e-05,
+      "loss": 1.6791,
+      "step": 1168
+    },
+    {
+      "epoch": 0.24278296988577364,
+      "grad_norm": 0.49183887243270874,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 1169
+    },
+    {
+      "epoch": 0.24299065420560748,
+      "grad_norm": 0.4702599346637726,
+      "learning_rate": 8e-05,
+      "loss": 1.5553,
+      "step": 1170
+    },
+    {
+      "epoch": 0.24319833852544132,
+      "grad_norm": 0.48021695017814636,
+      "learning_rate": 8e-05,
+      "loss": 1.5804,
+      "step": 1171
+    },
+    {
+      "epoch": 0.24340602284527518,
+      "grad_norm": 0.5404965281486511,
+      "learning_rate": 8e-05,
+      "loss": 1.6724,
+      "step": 1172
+    },
+    {
+      "epoch": 0.24361370716510902,
+      "grad_norm": 0.5167496204376221,
+      "learning_rate": 8e-05,
+      "loss": 1.5631,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2438213914849429,
+      "grad_norm": 0.4974764883518219,
+      "learning_rate": 8e-05,
+      "loss": 1.5615,
+      "step": 1174
+    },
+    {
+      "epoch": 0.24402907580477673,
+      "grad_norm": 0.4905969202518463,
+      "learning_rate": 8e-05,
+      "loss": 1.6023,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2442367601246106,
+      "grad_norm": 0.5358709692955017,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 1176
+    },
+    {
+      "epoch": 0.24444444444444444,
+      "grad_norm": 0.49605727195739746,
+      "learning_rate": 8e-05,
+      "loss": 1.6356,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2446521287642783,
+      "grad_norm": 0.4891799986362457,
+      "learning_rate": 8e-05,
+      "loss": 1.6877,
+      "step": 1178
+    },
+    {
+      "epoch": 0.24485981308411214,
+      "grad_norm": 0.4895115792751312,
+      "learning_rate": 8e-05,
+      "loss": 1.5594,
+      "step": 1179
+    },
+    {
+      "epoch": 0.245067497403946,
+      "grad_norm": 0.49354806542396545,
+      "learning_rate": 8e-05,
+      "loss": 1.6815,
+      "step": 1180
+    },
+    {
+      "epoch": 0.24527518172377985,
+      "grad_norm": 0.4897477924823761,
+      "learning_rate": 8e-05,
+      "loss": 1.6378,
+      "step": 1181
+    },
+    {
+      "epoch": 0.24548286604361372,
+      "grad_norm": 0.4873749613761902,
+      "learning_rate": 8e-05,
+      "loss": 1.5757,
+      "step": 1182
+    },
+    {
+      "epoch": 0.24569055036344756,
+      "grad_norm": 0.464875191450119,
+      "learning_rate": 8e-05,
+      "loss": 1.5587,
+      "step": 1183
+    },
+    {
+      "epoch": 0.24589823468328142,
+      "grad_norm": 0.4927029311656952,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 1184
+    },
+    {
+      "epoch": 0.24610591900311526,
+      "grad_norm": 0.4638020694255829,
+      "learning_rate": 8e-05,
+      "loss": 1.5619,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2463136033229491,
+      "grad_norm": 0.4894510805606842,
+      "learning_rate": 8e-05,
+      "loss": 1.5729,
+      "step": 1186
+    },
+    {
+      "epoch": 0.24652128764278297,
+      "grad_norm": 0.4695272743701935,
+      "learning_rate": 8e-05,
+      "loss": 1.6059,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2467289719626168,
+      "grad_norm": 0.47015345096588135,
+      "learning_rate": 8e-05,
+      "loss": 1.6093,
+      "step": 1188
+    },
+    {
+      "epoch": 0.24693665628245068,
+      "grad_norm": 0.4933643043041229,
+      "learning_rate": 8e-05,
+      "loss": 1.5702,
+      "step": 1189
+    },
+    {
+      "epoch": 0.24714434060228452,
+      "grad_norm": 0.4731082022190094,
+      "learning_rate": 8e-05,
+      "loss": 1.5177,
+      "step": 1190
+    },
+    {
+      "epoch": 0.24735202492211839,
+      "grad_norm": 0.48732051253318787,
+      "learning_rate": 8e-05,
+      "loss": 1.6784,
+      "step": 1191
+    },
+    {
+      "epoch": 0.24755970924195222,
+      "grad_norm": 0.5021098256111145,
+      "learning_rate": 8e-05,
+      "loss": 1.6564,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2477673935617861,
+      "grad_norm": 0.48108044266700745,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 1193
+    },
+    {
+      "epoch": 0.24797507788161993,
+      "grad_norm": 0.47929883003234863,
+      "learning_rate": 8e-05,
+      "loss": 1.6104,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2481827622014538,
+      "grad_norm": 0.502677321434021,
+      "learning_rate": 8e-05,
+      "loss": 1.6059,
+      "step": 1195
+    },
+    {
+      "epoch": 0.24839044652128764,
+      "grad_norm": 0.49433907866477966,
+      "learning_rate": 8e-05,
+      "loss": 1.6257,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2485981308411215,
+      "grad_norm": 0.4825282394886017,
+      "learning_rate": 8e-05,
+      "loss": 1.6134,
+      "step": 1197
+    },
+    {
+      "epoch": 0.24880581516095535,
+      "grad_norm": 0.4730031192302704,
+      "learning_rate": 8e-05,
+      "loss": 1.581,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2490134994807892,
+      "grad_norm": 0.4965001940727234,
+      "learning_rate": 8e-05,
+      "loss": 1.6166,
+      "step": 1199
+    },
+    {
+      "epoch": 0.24922118380062305,
+      "grad_norm": 0.49881839752197266,
+      "learning_rate": 8e-05,
+      "loss": 1.6278,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2494288681204569,
+      "grad_norm": 0.484718382358551,
+      "learning_rate": 8e-05,
+      "loss": 1.6499,
+      "step": 1201
+    },
+    {
+      "epoch": 0.24963655244029076,
+      "grad_norm": 0.4792778491973877,
+      "learning_rate": 8e-05,
+      "loss": 1.6444,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2498442367601246,
+      "grad_norm": 0.4959498643875122,
+      "learning_rate": 8e-05,
+      "loss": 1.6745,
+      "step": 1203
+    },
+    {
+      "epoch": 0.25005192107995844,
+      "grad_norm": 0.48365911841392517,
+      "learning_rate": 8e-05,
+      "loss": 1.5993,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2502596053997923,
+      "grad_norm": 0.5093377828598022,
+      "learning_rate": 8e-05,
+      "loss": 1.628,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2504672897196262,
+      "grad_norm": 0.486716091632843,
+      "learning_rate": 8e-05,
+      "loss": 1.6436,
+      "step": 1206
+    },
+    {
+      "epoch": 0.25067497403946004,
+      "grad_norm": 0.4940541982650757,
+      "learning_rate": 8e-05,
+      "loss": 1.5606,
+      "step": 1207
+    },
+    {
+      "epoch": 0.25088265835929385,
+      "grad_norm": 0.4985447824001312,
+      "learning_rate": 8e-05,
+      "loss": 1.7143,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2510903426791277,
+      "grad_norm": 0.49395254254341125,
+      "learning_rate": 8e-05,
+      "loss": 1.5876,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2512980269989616,
+      "grad_norm": 0.4942306876182556,
+      "learning_rate": 8e-05,
+      "loss": 1.6473,
+      "step": 1210
+    },
+    {
+      "epoch": 0.25150571131879546,
+      "grad_norm": 0.48080599308013916,
+      "learning_rate": 8e-05,
+      "loss": 1.6188,
+      "step": 1211
+    },
+    {
+      "epoch": 0.25171339563862927,
+      "grad_norm": 0.4913437068462372,
+      "learning_rate": 8e-05,
+      "loss": 1.6626,
+      "step": 1212
+    },
+    {
+      "epoch": 0.25192107995846313,
+      "grad_norm": 0.4762611985206604,
+      "learning_rate": 8e-05,
+      "loss": 1.5715,
+      "step": 1213
+    },
+    {
+      "epoch": 0.252128764278297,
+      "grad_norm": 0.49467694759368896,
+      "learning_rate": 8e-05,
+      "loss": 1.6428,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2523364485981308,
+      "grad_norm": 0.49399566650390625,
+      "learning_rate": 8e-05,
+      "loss": 1.5658,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2525441329179647,
+      "grad_norm": 0.47850465774536133,
+      "learning_rate": 8e-05,
+      "loss": 1.6057,
+      "step": 1216
+    },
+    {
+      "epoch": 0.25275181723779855,
+      "grad_norm": 0.5025778412818909,
+      "learning_rate": 8e-05,
+      "loss": 1.6746,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2529595015576324,
+      "grad_norm": 0.4746226966381073,
+      "learning_rate": 8e-05,
+      "loss": 1.5973,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2531671858774662,
+      "grad_norm": 0.490772008895874,
+      "learning_rate": 8e-05,
+      "loss": 1.6176,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2533748701973001,
+      "grad_norm": 0.477453351020813,
+      "learning_rate": 8e-05,
+      "loss": 1.6027,
+      "step": 1220
+    },
+    {
+      "epoch": 0.25358255451713396,
+      "grad_norm": 0.48569169640541077,
+      "learning_rate": 8e-05,
+      "loss": 1.6144,
+      "step": 1221
+    },
+    {
+      "epoch": 0.25379023883696783,
+      "grad_norm": 0.48016172647476196,
+      "learning_rate": 8e-05,
+      "loss": 1.612,
+      "step": 1222
+    },
+    {
+      "epoch": 0.25399792315680164,
+      "grad_norm": 0.458385169506073,
+      "learning_rate": 8e-05,
+      "loss": 1.4938,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2542056074766355,
+      "grad_norm": 0.4843856692314148,
+      "learning_rate": 8e-05,
+      "loss": 1.5878,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2544132917964694,
+      "grad_norm": 0.5237738490104675,
+      "learning_rate": 8e-05,
+      "loss": 1.6753,
+      "step": 1225
+    },
+    {
+      "epoch": 0.25462097611630324,
+      "grad_norm": 0.5184034109115601,
+      "learning_rate": 8e-05,
+      "loss": 1.5755,
+      "step": 1226
+    },
+    {
+      "epoch": 0.25482866043613706,
+      "grad_norm": 0.5014573931694031,
+      "learning_rate": 8e-05,
+      "loss": 1.6439,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2550363447559709,
+      "grad_norm": 0.4746113419532776,
+      "learning_rate": 8e-05,
+      "loss": 1.6612,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2552440290758048,
+      "grad_norm": 0.48897260427474976,
+      "learning_rate": 8e-05,
+      "loss": 1.5836,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2554517133956386,
+      "grad_norm": 0.47817346453666687,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 1230
+    },
+    {
+      "epoch": 0.25565939771547247,
+      "grad_norm": 0.5092668533325195,
+      "learning_rate": 8e-05,
+      "loss": 1.6462,
+      "step": 1231
+    },
+    {
+      "epoch": 0.25586708203530634,
+      "grad_norm": 0.4775659739971161,
+      "learning_rate": 8e-05,
+      "loss": 1.5857,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2560747663551402,
+      "grad_norm": 0.5099481344223022,
+      "learning_rate": 8e-05,
+      "loss": 1.592,
+      "step": 1233
+    },
+    {
+      "epoch": 0.256282450674974,
+      "grad_norm": 0.4921092987060547,
+      "learning_rate": 8e-05,
+      "loss": 1.5091,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2564901349948079,
+      "grad_norm": 0.5136047601699829,
+      "learning_rate": 8e-05,
+      "loss": 1.6556,
+      "step": 1235
+    },
+    {
+      "epoch": 0.25669781931464175,
+      "grad_norm": 0.4959682822227478,
+      "learning_rate": 8e-05,
+      "loss": 1.6347,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2569055036344756,
+      "grad_norm": 0.5049384236335754,
+      "learning_rate": 8e-05,
+      "loss": 1.6096,
+      "step": 1237
+    },
+    {
+      "epoch": 0.25711318795430943,
+      "grad_norm": 0.5177755951881409,
+      "learning_rate": 8e-05,
+      "loss": 1.5408,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2573208722741433,
+      "grad_norm": 0.4715554416179657,
+      "learning_rate": 8e-05,
+      "loss": 1.6184,
+      "step": 1239
+    },
+    {
+      "epoch": 0.25752855659397716,
+      "grad_norm": 0.5331630110740662,
+      "learning_rate": 8e-05,
+      "loss": 1.6457,
+      "step": 1240
+    },
+    {
+      "epoch": 0.25773624091381103,
+      "grad_norm": 0.48646849393844604,
+      "learning_rate": 8e-05,
+      "loss": 1.6077,
+      "step": 1241
+    },
+    {
+      "epoch": 0.25794392523364484,
+      "grad_norm": 0.48699328303337097,
+      "learning_rate": 8e-05,
+      "loss": 1.5695,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2581516095534787,
+      "grad_norm": 0.4788655936717987,
+      "learning_rate": 8e-05,
+      "loss": 1.6047,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2583592938733126,
+      "grad_norm": 0.49768000841140747,
+      "learning_rate": 8e-05,
+      "loss": 1.6321,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2585669781931464,
+      "grad_norm": 0.4728749096393585,
+      "learning_rate": 8e-05,
+      "loss": 1.4864,
+      "step": 1245
+    },
+    {
+      "epoch": 0.25877466251298026,
+      "grad_norm": 0.4823184311389923,
+      "learning_rate": 8e-05,
+      "loss": 1.5463,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2589823468328141,
+      "grad_norm": 0.49181875586509705,
+      "learning_rate": 8e-05,
+      "loss": 1.6372,
+      "step": 1247
+    },
+    {
+      "epoch": 0.259190031152648,
+      "grad_norm": 0.48405319452285767,
+      "learning_rate": 8e-05,
+      "loss": 1.598,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2593977154724818,
+      "grad_norm": 0.5048245787620544,
+      "learning_rate": 8e-05,
+      "loss": 1.5521,
+      "step": 1249
+    },
+    {
+      "epoch": 0.25960539979231567,
+      "grad_norm": 0.4842333495616913,
+      "learning_rate": 8e-05,
+      "loss": 1.6642,
+      "step": 1250
+    },
+    {
+      "epoch": 0.25981308411214954,
+      "grad_norm": 0.48424315452575684,
+      "learning_rate": 8e-05,
+      "loss": 1.6036,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2600207684319834,
+      "grad_norm": 0.46493804454803467,
+      "learning_rate": 8e-05,
+      "loss": 1.6679,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2602284527518172,
+      "grad_norm": 0.5107417106628418,
+      "learning_rate": 8e-05,
+      "loss": 1.6116,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2604361370716511,
+      "grad_norm": 0.48230957984924316,
+      "learning_rate": 8e-05,
+      "loss": 1.5848,
+      "step": 1254
+    },
+    {
+      "epoch": 0.26064382139148495,
+      "grad_norm": 0.4716993570327759,
+      "learning_rate": 8e-05,
+      "loss": 1.5599,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2608515057113188,
+      "grad_norm": 0.4995378255844116,
+      "learning_rate": 8e-05,
+      "loss": 1.5162,
+      "step": 1256
+    },
+    {
+      "epoch": 0.26105919003115263,
+      "grad_norm": 0.49914610385894775,
+      "learning_rate": 8e-05,
+      "loss": 1.5889,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2612668743509865,
+      "grad_norm": 0.5149269104003906,
+      "learning_rate": 8e-05,
+      "loss": 1.5693,
+      "step": 1258
+    },
+    {
+      "epoch": 0.26147455867082037,
+      "grad_norm": 0.4664236903190613,
+      "learning_rate": 8e-05,
+      "loss": 1.5911,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2616822429906542,
+      "grad_norm": 0.4691005349159241,
+      "learning_rate": 8e-05,
+      "loss": 1.5378,
+      "step": 1260
+    },
+    {
+      "epoch": 0.26188992731048805,
+      "grad_norm": 0.5125289559364319,
+      "learning_rate": 8e-05,
+      "loss": 1.6207,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2620976116303219,
+      "grad_norm": 0.5016699433326721,
+      "learning_rate": 8e-05,
+      "loss": 1.7096,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2623052959501558,
+      "grad_norm": 0.4931148290634155,
+      "learning_rate": 8e-05,
+      "loss": 1.535,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2625129802699896,
+      "grad_norm": 0.4818016588687897,
+      "learning_rate": 8e-05,
+      "loss": 1.6163,
+      "step": 1264
+    },
+    {
+      "epoch": 0.26272066458982346,
+      "grad_norm": 0.4864917993545532,
+      "learning_rate": 8e-05,
+      "loss": 1.5272,
+      "step": 1265
+    },
+    {
+      "epoch": 0.26292834890965733,
+      "grad_norm": 0.4974689483642578,
+      "learning_rate": 8e-05,
+      "loss": 1.6396,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2631360332294912,
+      "grad_norm": 0.48342207074165344,
+      "learning_rate": 8e-05,
+      "loss": 1.5803,
+      "step": 1267
+    },
+    {
+      "epoch": 0.263343717549325,
+      "grad_norm": 0.5663976073265076,
+      "learning_rate": 8e-05,
+      "loss": 1.6336,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2635514018691589,
+      "grad_norm": 0.5170464515686035,
+      "learning_rate": 8e-05,
+      "loss": 1.6435,
+      "step": 1269
+    },
+    {
+      "epoch": 0.26375908618899274,
+      "grad_norm": 0.49931710958480835,
+      "learning_rate": 8e-05,
+      "loss": 1.697,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2639667705088266,
+      "grad_norm": 0.4786977767944336,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2641744548286604,
+      "grad_norm": 0.490163117647171,
+      "learning_rate": 8e-05,
+      "loss": 1.5944,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2643821391484943,
+      "grad_norm": 0.469289630651474,
+      "learning_rate": 8e-05,
+      "loss": 1.4951,
+      "step": 1273
+    },
+    {
+      "epoch": 0.26458982346832816,
+      "grad_norm": 0.49278539419174194,
+      "learning_rate": 8e-05,
+      "loss": 1.5941,
+      "step": 1274
+    },
+    {
+      "epoch": 0.26479750778816197,
+      "grad_norm": 0.4842006266117096,
+      "learning_rate": 8e-05,
+      "loss": 1.5923,
+      "step": 1275
+    },
+    {
+      "epoch": 0.26500519210799584,
+      "grad_norm": 0.5030384659767151,
+      "learning_rate": 8e-05,
+      "loss": 1.6361,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2652128764278297,
+      "grad_norm": 0.4949890971183777,
+      "learning_rate": 8e-05,
+      "loss": 1.6389,
+      "step": 1277
+    },
+    {
+      "epoch": 0.26542056074766357,
+      "grad_norm": 0.4904261827468872,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2656282450674974,
+      "grad_norm": 0.48278945684432983,
+      "learning_rate": 8e-05,
+      "loss": 1.64,
+      "step": 1279
+    },
+    {
+      "epoch": 0.26583592938733125,
+      "grad_norm": 0.48540300130844116,
+      "learning_rate": 8e-05,
+      "loss": 1.6054,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2660436137071651,
+      "grad_norm": 0.5015822649002075,
+      "learning_rate": 8e-05,
+      "loss": 1.5868,
+      "step": 1281
+    },
+    {
+      "epoch": 0.266251298026999,
+      "grad_norm": 0.49217545986175537,
+      "learning_rate": 8e-05,
+      "loss": 1.6145,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2664589823468328,
+      "grad_norm": 0.5260308384895325,
+      "learning_rate": 8e-05,
+      "loss": 1.6333,
+      "step": 1283
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 0.5329016447067261,
+      "learning_rate": 8e-05,
+      "loss": 1.7274,
+      "step": 1284
+    },
+    {
+      "epoch": 0.26687435098650053,
+      "grad_norm": 0.5138272047042847,
+      "learning_rate": 8e-05,
+      "loss": 1.6601,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2670820353063344,
+      "grad_norm": 0.4883299171924591,
+      "learning_rate": 8e-05,
+      "loss": 1.5988,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2672897196261682,
+      "grad_norm": 0.5150203108787537,
+      "learning_rate": 8e-05,
+      "loss": 1.6732,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2674974039460021,
+      "grad_norm": 0.49123409390449524,
+      "learning_rate": 8e-05,
+      "loss": 1.5287,
+      "step": 1288
+    },
+    {
+      "epoch": 0.26770508826583594,
+      "grad_norm": 0.4945281744003296,
+      "learning_rate": 8e-05,
+      "loss": 1.6321,
+      "step": 1289
+    },
+    {
+      "epoch": 0.26791277258566976,
+      "grad_norm": 0.5029784440994263,
+      "learning_rate": 8e-05,
+      "loss": 1.6214,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2681204569055036,
+      "grad_norm": 0.46249303221702576,
+      "learning_rate": 8e-05,
+      "loss": 1.4891,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2683281412253375,
+      "grad_norm": 0.5195319652557373,
+      "learning_rate": 8e-05,
+      "loss": 1.6059,
+      "step": 1292
+    },
+    {
+      "epoch": 0.26853582554517136,
+      "grad_norm": 0.5203794836997986,
+      "learning_rate": 8e-05,
+      "loss": 1.6673,
+      "step": 1293
+    },
+    {
+      "epoch": 0.26874350986500517,
+      "grad_norm": 0.4765598773956299,
+      "learning_rate": 8e-05,
+      "loss": 1.6292,
+      "step": 1294
+    },
+    {
+      "epoch": 0.26895119418483904,
+      "grad_norm": 0.4790004789829254,
+      "learning_rate": 8e-05,
+      "loss": 1.5275,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2691588785046729,
+      "grad_norm": 0.486456036567688,
+      "learning_rate": 8e-05,
+      "loss": 1.5916,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2693665628245068,
+      "grad_norm": 0.4780852496623993,
+      "learning_rate": 8e-05,
+      "loss": 1.6143,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2695742471443406,
+      "grad_norm": 0.5063662528991699,
+      "learning_rate": 8e-05,
+      "loss": 1.598,
+      "step": 1298
+    },
+    {
+      "epoch": 0.26978193146417445,
+      "grad_norm": 0.4856896996498108,
+      "learning_rate": 8e-05,
+      "loss": 1.6217,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2699896157840083,
+      "grad_norm": 0.4855417013168335,
+      "learning_rate": 8e-05,
+      "loss": 1.5746,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2701973001038422,
+      "grad_norm": 0.4749685227870941,
+      "learning_rate": 8e-05,
+      "loss": 1.529,
+      "step": 1301
+    },
+    {
+      "epoch": 0.270404984423676,
+      "grad_norm": 0.4671476185321808,
+      "learning_rate": 8e-05,
+      "loss": 1.5717,
+      "step": 1302
+    },
+    {
+      "epoch": 0.27061266874350987,
+      "grad_norm": 0.49062880873680115,
+      "learning_rate": 8e-05,
+      "loss": 1.5799,
+      "step": 1303
+    },
+    {
+      "epoch": 0.27082035306334373,
+      "grad_norm": 0.4858238995075226,
+      "learning_rate": 8e-05,
+      "loss": 1.6249,
+      "step": 1304
+    },
+    {
+      "epoch": 0.27102803738317754,
+      "grad_norm": 0.4788179397583008,
+      "learning_rate": 8e-05,
+      "loss": 1.5742,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2712357217030114,
+      "grad_norm": 0.5083610415458679,
+      "learning_rate": 8e-05,
+      "loss": 1.5634,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2714434060228453,
+      "grad_norm": 0.4782449007034302,
+      "learning_rate": 8e-05,
+      "loss": 1.593,
+      "step": 1307
+    },
+    {
+      "epoch": 0.27165109034267915,
+      "grad_norm": 0.5111173391342163,
+      "learning_rate": 8e-05,
+      "loss": 1.5827,
+      "step": 1308
+    },
+    {
+      "epoch": 0.27185877466251296,
+      "grad_norm": 0.4896087646484375,
+      "learning_rate": 8e-05,
+      "loss": 1.5713,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2720664589823468,
+      "grad_norm": 0.50101238489151,
+      "learning_rate": 8e-05,
+      "loss": 1.5708,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2722741433021807,
+      "grad_norm": 0.4787428677082062,
+      "learning_rate": 8e-05,
+      "loss": 1.53,
+      "step": 1311
+    },
+    {
+      "epoch": 0.27248182762201456,
+      "grad_norm": 0.4917222857475281,
+      "learning_rate": 8e-05,
+      "loss": 1.6704,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2726895119418484,
+      "grad_norm": 0.4935707449913025,
+      "learning_rate": 8e-05,
+      "loss": 1.5838,
+      "step": 1313
+    },
+    {
+      "epoch": 0.27289719626168224,
+      "grad_norm": 0.48440101742744446,
+      "learning_rate": 8e-05,
+      "loss": 1.6517,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2731048805815161,
+      "grad_norm": 0.47382527589797974,
+      "learning_rate": 8e-05,
+      "loss": 1.5884,
+      "step": 1315
+    },
+    {
+      "epoch": 0.27331256490135,
+      "grad_norm": 0.5190213322639465,
+      "learning_rate": 8e-05,
+      "loss": 1.6797,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2735202492211838,
+      "grad_norm": 0.49640318751335144,
+      "learning_rate": 8e-05,
+      "loss": 1.6594,
+      "step": 1317
+    },
+    {
+      "epoch": 0.27372793354101765,
+      "grad_norm": 0.5151790380477905,
+      "learning_rate": 8e-05,
+      "loss": 1.5557,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2739356178608515,
+      "grad_norm": 0.47358784079551697,
+      "learning_rate": 8e-05,
+      "loss": 1.5639,
+      "step": 1319
+    },
+    {
+      "epoch": 0.27414330218068533,
+      "grad_norm": 0.5438617467880249,
+      "learning_rate": 8e-05,
+      "loss": 1.6362,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2743509865005192,
+      "grad_norm": 0.4804224967956543,
+      "learning_rate": 8e-05,
+      "loss": 1.5946,
+      "step": 1321
+    },
+    {
+      "epoch": 0.27455867082035307,
+      "grad_norm": 0.5258082151412964,
+      "learning_rate": 8e-05,
+      "loss": 1.5561,
+      "step": 1322
+    },
+    {
+      "epoch": 0.27476635514018694,
+      "grad_norm": 0.48843201994895935,
+      "learning_rate": 8e-05,
+      "loss": 1.5366,
+      "step": 1323
+    },
+    {
+      "epoch": 0.27497403946002075,
+      "grad_norm": 0.47955572605133057,
+      "learning_rate": 8e-05,
+      "loss": 1.6056,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2751817237798546,
+      "grad_norm": 0.4964466392993927,
+      "learning_rate": 8e-05,
+      "loss": 1.6292,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2753894080996885,
+      "grad_norm": 0.4776870608329773,
+      "learning_rate": 8e-05,
+      "loss": 1.6147,
+      "step": 1326
+    },
+    {
+      "epoch": 0.27559709241952235,
+      "grad_norm": 0.49412986636161804,
+      "learning_rate": 8e-05,
+      "loss": 1.596,
+      "step": 1327
+    },
+    {
+      "epoch": 0.27580477673935616,
+      "grad_norm": 0.49872222542762756,
+      "learning_rate": 8e-05,
+      "loss": 1.622,
+      "step": 1328
+    },
+    {
+      "epoch": 0.27601246105919003,
+      "grad_norm": 0.48729175329208374,
+      "learning_rate": 8e-05,
+      "loss": 1.6105,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2762201453790239,
+      "grad_norm": 0.4850686490535736,
+      "learning_rate": 8e-05,
+      "loss": 1.5978,
+      "step": 1330
+    },
+    {
+      "epoch": 0.27642782969885776,
+      "grad_norm": 0.49125537276268005,
+      "learning_rate": 8e-05,
+      "loss": 1.5584,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2766355140186916,
+      "grad_norm": 0.5107133984565735,
+      "learning_rate": 8e-05,
+      "loss": 1.6258,
+      "step": 1332
+    },
+    {
+      "epoch": 0.27684319833852544,
+      "grad_norm": 0.4897415339946747,
+      "learning_rate": 8e-05,
+      "loss": 1.6198,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2770508826583593,
+      "grad_norm": 0.4734158217906952,
+      "learning_rate": 8e-05,
+      "loss": 1.5732,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2772585669781931,
+      "grad_norm": 0.5000379681587219,
+      "learning_rate": 8e-05,
+      "loss": 1.6496,
+      "step": 1335
+    },
+    {
+      "epoch": 0.277466251298027,
+      "grad_norm": 0.5036786198616028,
+      "learning_rate": 8e-05,
+      "loss": 1.6218,
+      "step": 1336
+    },
+    {
+      "epoch": 0.27767393561786086,
+      "grad_norm": 0.4840192198753357,
+      "learning_rate": 8e-05,
+      "loss": 1.5044,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2778816199376947,
+      "grad_norm": 0.4741503894329071,
+      "learning_rate": 8e-05,
+      "loss": 1.5801,
+      "step": 1338
+    },
+    {
+      "epoch": 0.27808930425752854,
+      "grad_norm": 0.488294392824173,
+      "learning_rate": 8e-05,
+      "loss": 1.597,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2782969885773624,
+      "grad_norm": 0.5086274147033691,
+      "learning_rate": 8e-05,
+      "loss": 1.6069,
+      "step": 1340
+    },
+    {
+      "epoch": 0.27850467289719627,
+      "grad_norm": 0.510127604007721,
+      "learning_rate": 8e-05,
+      "loss": 1.596,
+      "step": 1341
+    },
+    {
+      "epoch": 0.27871235721703014,
+      "grad_norm": 0.46870115399360657,
+      "learning_rate": 8e-05,
+      "loss": 1.5364,
+      "step": 1342
+    },
+    {
+      "epoch": 0.27892004153686395,
+      "grad_norm": 0.45464587211608887,
+      "learning_rate": 8e-05,
+      "loss": 1.5018,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2791277258566978,
+      "grad_norm": 0.48707786202430725,
+      "learning_rate": 8e-05,
+      "loss": 1.5411,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2793354101765317,
+      "grad_norm": 0.4926038086414337,
+      "learning_rate": 8e-05,
+      "loss": 1.6213,
+      "step": 1345
+    },
+    {
+      "epoch": 0.27954309449636555,
+      "grad_norm": 0.48424461483955383,
+      "learning_rate": 8e-05,
+      "loss": 1.6463,
+      "step": 1346
+    },
+    {
+      "epoch": 0.27975077881619936,
+      "grad_norm": 0.48776450753211975,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 1347
+    },
+    {
+      "epoch": 0.27995846313603323,
+      "grad_norm": 0.4794577360153198,
+      "learning_rate": 8e-05,
+      "loss": 1.5981,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2801661474558671,
+      "grad_norm": 0.4882426857948303,
+      "learning_rate": 8e-05,
+      "loss": 1.6774,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2803738317757009,
+      "grad_norm": 0.5013483166694641,
+      "learning_rate": 8e-05,
+      "loss": 1.6659,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2805815160955348,
+      "grad_norm": 0.4758848547935486,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 1351
+    },
+    {
+      "epoch": 0.28078920041536864,
+      "grad_norm": 0.4662100374698639,
+      "learning_rate": 8e-05,
+      "loss": 1.5151,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2809968847352025,
+      "grad_norm": 0.47722524404525757,
+      "learning_rate": 8e-05,
+      "loss": 1.5534,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2812045690550363,
+      "grad_norm": 0.5073485374450684,
+      "learning_rate": 8e-05,
+      "loss": 1.6202,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2814122533748702,
+      "grad_norm": 0.49153366684913635,
+      "learning_rate": 8e-05,
+      "loss": 1.5921,
+      "step": 1355
+    },
+    {
+      "epoch": 0.28161993769470406,
+      "grad_norm": 0.4862796664237976,
+      "learning_rate": 8e-05,
+      "loss": 1.5981,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2818276220145379,
+      "grad_norm": 0.49322280287742615,
+      "learning_rate": 8e-05,
+      "loss": 1.6894,
+      "step": 1357
+    },
+    {
+      "epoch": 0.28203530633437174,
+      "grad_norm": 0.48775988817214966,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2822429906542056,
+      "grad_norm": 0.47947344183921814,
+      "learning_rate": 8e-05,
+      "loss": 1.5709,
+      "step": 1359
+    },
+    {
+      "epoch": 0.2824506749740395,
+      "grad_norm": 0.5054206848144531,
+      "learning_rate": 8e-05,
+      "loss": 1.6541,
+      "step": 1360
+    },
+    {
+      "epoch": 0.28265835929387334,
+      "grad_norm": 0.4824366569519043,
+      "learning_rate": 8e-05,
+      "loss": 1.5929,
+      "step": 1361
+    },
+    {
+      "epoch": 0.28286604361370715,
+      "grad_norm": 0.49978917837142944,
+      "learning_rate": 8e-05,
+      "loss": 1.6151,
+      "step": 1362
+    },
+    {
+      "epoch": 0.283073727933541,
+      "grad_norm": 0.4934566915035248,
+      "learning_rate": 8e-05,
+      "loss": 1.6641,
+      "step": 1363
+    },
+    {
+      "epoch": 0.2832814122533749,
+      "grad_norm": 0.4824499785900116,
+      "learning_rate": 8e-05,
+      "loss": 1.6189,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2834890965732087,
+      "grad_norm": 0.47832030057907104,
+      "learning_rate": 8e-05,
+      "loss": 1.5083,
+      "step": 1365
+    },
+    {
+      "epoch": 0.28369678089304257,
+      "grad_norm": 0.4860023260116577,
+      "learning_rate": 8e-05,
+      "loss": 1.565,
+      "step": 1366
+    },
+    {
+      "epoch": 0.28390446521287643,
+      "grad_norm": 0.48774632811546326,
+      "learning_rate": 8e-05,
+      "loss": 1.638,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2841121495327103,
+      "grad_norm": 0.47733739018440247,
+      "learning_rate": 8e-05,
+      "loss": 1.6151,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2843198338525441,
+      "grad_norm": 0.48727935552597046,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 1369
+    },
+    {
+      "epoch": 0.284527518172378,
+      "grad_norm": 0.4770117998123169,
+      "learning_rate": 8e-05,
+      "loss": 1.6759,
+      "step": 1370
+    },
+    {
+      "epoch": 0.28473520249221185,
+      "grad_norm": 0.49704787135124207,
+      "learning_rate": 8e-05,
+      "loss": 1.6116,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2849428868120457,
+      "grad_norm": 0.48451000452041626,
+      "learning_rate": 8e-05,
+      "loss": 1.5776,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2851505711318795,
+      "grad_norm": 0.49385419487953186,
+      "learning_rate": 8e-05,
+      "loss": 1.6142,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2853582554517134,
+      "grad_norm": 0.49187561869621277,
+      "learning_rate": 8e-05,
+      "loss": 1.6076,
+      "step": 1374
+    },
+    {
+      "epoch": 0.28556593977154726,
+      "grad_norm": 0.4695932865142822,
+      "learning_rate": 8e-05,
+      "loss": 1.5801,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2857736240913811,
+      "grad_norm": 0.49953845143318176,
+      "learning_rate": 8e-05,
+      "loss": 1.617,
+      "step": 1376
+    },
+    {
+      "epoch": 0.28598130841121494,
+      "grad_norm": 0.49016183614730835,
+      "learning_rate": 8e-05,
+      "loss": 1.6751,
+      "step": 1377
+    },
+    {
+      "epoch": 0.2861889927310488,
+      "grad_norm": 0.5039032101631165,
+      "learning_rate": 8e-05,
+      "loss": 1.5667,
+      "step": 1378
+    },
+    {
+      "epoch": 0.2863966770508827,
+      "grad_norm": 0.5153660178184509,
+      "learning_rate": 8e-05,
+      "loss": 1.5351,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2866043613707165,
+      "grad_norm": 0.4923986792564392,
+      "learning_rate": 8e-05,
+      "loss": 1.6022,
+      "step": 1380
+    },
+    {
+      "epoch": 0.28681204569055035,
+      "grad_norm": 0.4953305125236511,
+      "learning_rate": 8e-05,
+      "loss": 1.6172,
+      "step": 1381
+    },
+    {
+      "epoch": 0.2870197300103842,
+      "grad_norm": 0.48334407806396484,
+      "learning_rate": 8e-05,
+      "loss": 1.5416,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2872274143302181,
+      "grad_norm": 0.4850659966468811,
+      "learning_rate": 8e-05,
+      "loss": 1.6704,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2874350986500519,
+      "grad_norm": 0.5282303690910339,
+      "learning_rate": 8e-05,
+      "loss": 1.6134,
+      "step": 1384
+    },
+    {
+      "epoch": 0.28764278296988577,
+      "grad_norm": 0.4880530834197998,
+      "learning_rate": 8e-05,
+      "loss": 1.6108,
+      "step": 1385
+    },
+    {
+      "epoch": 0.28785046728971964,
+      "grad_norm": 0.5376881957054138,
+      "learning_rate": 8e-05,
+      "loss": 1.6852,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2880581516095535,
+      "grad_norm": 0.5560894012451172,
+      "learning_rate": 8e-05,
+      "loss": 1.677,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2882658359293873,
+      "grad_norm": 0.5074867606163025,
+      "learning_rate": 8e-05,
+      "loss": 1.5932,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2884735202492212,
+      "grad_norm": 0.4841192066669464,
+      "learning_rate": 8e-05,
+      "loss": 1.5258,
+      "step": 1389
+    },
+    {
+      "epoch": 0.28868120456905505,
+      "grad_norm": 0.5013852715492249,
+      "learning_rate": 8e-05,
+      "loss": 1.5784,
+      "step": 1390
+    },
+    {
+      "epoch": 0.28888888888888886,
+      "grad_norm": 0.494810551404953,
+      "learning_rate": 8e-05,
+      "loss": 1.5967,
+      "step": 1391
+    },
+    {
+      "epoch": 0.28909657320872273,
+      "grad_norm": 0.49166232347488403,
+      "learning_rate": 8e-05,
+      "loss": 1.5865,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2893042575285566,
+      "grad_norm": 0.5110693573951721,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 1393
+    },
+    {
+      "epoch": 0.28951194184839046,
+      "grad_norm": 0.542017936706543,
+      "learning_rate": 8e-05,
+      "loss": 1.6099,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2897196261682243,
+      "grad_norm": 0.5205735564231873,
+      "learning_rate": 8e-05,
+      "loss": 1.6194,
+      "step": 1395
+    },
+    {
+      "epoch": 0.28992731048805814,
+      "grad_norm": 0.542843222618103,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 1396
+    },
+    {
+      "epoch": 0.290134994807892,
+      "grad_norm": 0.5020806789398193,
+      "learning_rate": 8e-05,
+      "loss": 1.6645,
+      "step": 1397
+    },
+    {
+      "epoch": 0.2903426791277259,
+      "grad_norm": 0.48713013529777527,
+      "learning_rate": 8e-05,
+      "loss": 1.5229,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2905503634475597,
+      "grad_norm": 0.49321314692497253,
+      "learning_rate": 8e-05,
+      "loss": 1.6144,
+      "step": 1399
+    },
+    {
+      "epoch": 0.29075804776739356,
+      "grad_norm": 0.4920049011707306,
+      "learning_rate": 8e-05,
+      "loss": 1.6081,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2909657320872274,
+      "grad_norm": 0.4814404249191284,
+      "learning_rate": 8e-05,
+      "loss": 1.5424,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2911734164070613,
+      "grad_norm": 0.48672202229499817,
+      "learning_rate": 8e-05,
+      "loss": 1.5617,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2913811007268951,
+      "grad_norm": 0.4993645250797272,
+      "learning_rate": 8e-05,
+      "loss": 1.5552,
+      "step": 1403
+    },
+    {
+      "epoch": 0.29158878504672897,
+      "grad_norm": 0.493221640586853,
+      "learning_rate": 8e-05,
+      "loss": 1.6083,
+      "step": 1404
+    },
+    {
+      "epoch": 0.29179646936656284,
+      "grad_norm": 0.5242977738380432,
+      "learning_rate": 8e-05,
+      "loss": 1.6794,
+      "step": 1405
+    },
+    {
+      "epoch": 0.29200415368639665,
+      "grad_norm": 0.5010468363761902,
+      "learning_rate": 8e-05,
+      "loss": 1.6118,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2922118380062305,
+      "grad_norm": 0.4755447506904602,
+      "learning_rate": 8e-05,
+      "loss": 1.5602,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2924195223260644,
+      "grad_norm": 0.48869359493255615,
+      "learning_rate": 8e-05,
+      "loss": 1.6312,
+      "step": 1408
+    },
+    {
+      "epoch": 0.29262720664589825,
+      "grad_norm": 0.505214512348175,
+      "learning_rate": 8e-05,
+      "loss": 1.7103,
+      "step": 1409
+    },
+    {
+      "epoch": 0.29283489096573206,
+      "grad_norm": 0.4835333526134491,
+      "learning_rate": 8e-05,
+      "loss": 1.5842,
+      "step": 1410
+    },
+    {
+      "epoch": 0.29304257528556593,
+      "grad_norm": 0.47053441405296326,
+      "learning_rate": 8e-05,
+      "loss": 1.5655,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2932502596053998,
+      "grad_norm": 0.4897231161594391,
+      "learning_rate": 8e-05,
+      "loss": 1.5427,
+      "step": 1412
+    },
+    {
+      "epoch": 0.29345794392523367,
+      "grad_norm": 0.4817632734775543,
+      "learning_rate": 8e-05,
+      "loss": 1.5703,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2936656282450675,
+      "grad_norm": 0.5053437352180481,
+      "learning_rate": 8e-05,
+      "loss": 1.5884,
+      "step": 1414
+    },
+    {
+      "epoch": 0.29387331256490135,
+      "grad_norm": 0.48642995953559875,
+      "learning_rate": 8e-05,
+      "loss": 1.5685,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2940809968847352,
+      "grad_norm": 0.4815353453159332,
+      "learning_rate": 8e-05,
+      "loss": 1.6158,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2942886812045691,
+      "grad_norm": 0.49331390857696533,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2944963655244029,
+      "grad_norm": 0.5165848135948181,
+      "learning_rate": 8e-05,
+      "loss": 1.6028,
+      "step": 1418
+    },
+    {
+      "epoch": 0.29470404984423676,
+      "grad_norm": 0.4790689945220947,
+      "learning_rate": 8e-05,
+      "loss": 1.6281,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2949117341640706,
+      "grad_norm": 0.46984192728996277,
+      "learning_rate": 8e-05,
+      "loss": 1.5487,
+      "step": 1420
+    },
+    {
+      "epoch": 0.29511941848390444,
+      "grad_norm": 0.5013918876647949,
+      "learning_rate": 8e-05,
+      "loss": 1.5629,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2953271028037383,
+      "grad_norm": 0.5373260974884033,
+      "learning_rate": 8e-05,
+      "loss": 1.6564,
+      "step": 1422
+    },
+    {
+      "epoch": 0.2955347871235722,
+      "grad_norm": 0.5395689606666565,
+      "learning_rate": 8e-05,
+      "loss": 1.6307,
+      "step": 1423
+    },
+    {
+      "epoch": 0.29574247144340604,
+      "grad_norm": 0.5050875544548035,
+      "learning_rate": 8e-05,
+      "loss": 1.5577,
+      "step": 1424
+    },
+    {
+      "epoch": 0.29595015576323985,
+      "grad_norm": 0.46294352412223816,
+      "learning_rate": 8e-05,
+      "loss": 1.5705,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2961578400830737,
+      "grad_norm": 0.4974973201751709,
+      "learning_rate": 8e-05,
+      "loss": 1.5908,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2963655244029076,
+      "grad_norm": 0.5147601366043091,
+      "learning_rate": 8e-05,
+      "loss": 1.5833,
+      "step": 1427
+    },
+    {
+      "epoch": 0.29657320872274145,
+      "grad_norm": 0.5149271488189697,
+      "learning_rate": 8e-05,
+      "loss": 1.6029,
+      "step": 1428
+    },
+    {
+      "epoch": 0.29678089304257527,
+      "grad_norm": 0.48388561606407166,
+      "learning_rate": 8e-05,
+      "loss": 1.5629,
+      "step": 1429
+    },
+    {
+      "epoch": 0.29698857736240913,
+      "grad_norm": 0.4807972013950348,
+      "learning_rate": 8e-05,
+      "loss": 1.5618,
+      "step": 1430
+    },
+    {
+      "epoch": 0.297196261682243,
+      "grad_norm": 0.4944983124732971,
+      "learning_rate": 8e-05,
+      "loss": 1.6518,
+      "step": 1431
+    },
+    {
+      "epoch": 0.29740394600207687,
+      "grad_norm": 0.467385470867157,
+      "learning_rate": 8e-05,
+      "loss": 1.5768,
+      "step": 1432
+    },
+    {
+      "epoch": 0.2976116303219107,
+      "grad_norm": 0.4817611277103424,
+      "learning_rate": 8e-05,
+      "loss": 1.559,
+      "step": 1433
+    },
+    {
+      "epoch": 0.29781931464174455,
+      "grad_norm": 0.4777863919734955,
+      "learning_rate": 8e-05,
+      "loss": 1.5712,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2980269989615784,
+      "grad_norm": 0.46716588735580444,
+      "learning_rate": 8e-05,
+      "loss": 1.5301,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2982346832814122,
+      "grad_norm": 0.49335551261901855,
+      "learning_rate": 8e-05,
+      "loss": 1.633,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2984423676012461,
+      "grad_norm": 0.47975221276283264,
+      "learning_rate": 8e-05,
+      "loss": 1.5948,
+      "step": 1437
+    },
+    {
+      "epoch": 0.29865005192107996,
+      "grad_norm": 0.4699002504348755,
+      "learning_rate": 8e-05,
+      "loss": 1.5368,
+      "step": 1438
+    },
+    {
+      "epoch": 0.29885773624091383,
+      "grad_norm": 0.5026534199714661,
+      "learning_rate": 8e-05,
+      "loss": 1.6412,
+      "step": 1439
+    },
+    {
+      "epoch": 0.29906542056074764,
+      "grad_norm": 0.4809010326862335,
+      "learning_rate": 8e-05,
+      "loss": 1.5511,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2992731048805815,
+      "grad_norm": 0.5056330561637878,
+      "learning_rate": 8e-05,
+      "loss": 1.6516,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2994807892004154,
+      "grad_norm": 0.49161332845687866,
+      "learning_rate": 8e-05,
+      "loss": 1.5964,
+      "step": 1442
+    },
+    {
+      "epoch": 0.29968847352024924,
+      "grad_norm": 0.49896249175071716,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 1443
+    },
+    {
+      "epoch": 0.29989615784008306,
+      "grad_norm": 0.4822237491607666,
+      "learning_rate": 8e-05,
+      "loss": 1.6416,
+      "step": 1444
+    },
+    {
+      "epoch": 0.3001038421599169,
+      "grad_norm": 0.48241591453552246,
+      "learning_rate": 8e-05,
+      "loss": 1.6328,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3003115264797508,
+      "grad_norm": 0.48107171058654785,
+      "learning_rate": 8e-05,
+      "loss": 1.6912,
+      "step": 1446
+    },
+    {
+      "epoch": 0.30051921079958466,
+      "grad_norm": 0.4950486421585083,
+      "learning_rate": 8e-05,
+      "loss": 1.5809,
+      "step": 1447
+    },
+    {
+      "epoch": 0.30072689511941847,
+      "grad_norm": 0.516384482383728,
+      "learning_rate": 8e-05,
+      "loss": 1.5433,
+      "step": 1448
+    },
+    {
+      "epoch": 0.30093457943925234,
+      "grad_norm": 0.47670841217041016,
+      "learning_rate": 8e-05,
+      "loss": 1.5641,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3011422637590862,
+      "grad_norm": 0.4875502288341522,
+      "learning_rate": 8e-05,
+      "loss": 1.5905,
+      "step": 1450
+    },
+    {
+      "epoch": 0.30134994807892,
+      "grad_norm": 0.4830038547515869,
+      "learning_rate": 8e-05,
+      "loss": 1.5338,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3015576323987539,
+      "grad_norm": 0.4864535331726074,
+      "learning_rate": 8e-05,
+      "loss": 1.5884,
+      "step": 1452
+    },
+    {
+      "epoch": 0.30176531671858775,
+      "grad_norm": 0.48559343814849854,
+      "learning_rate": 8e-05,
+      "loss": 1.5684,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3019730010384216,
+      "grad_norm": 0.4923616051673889,
+      "learning_rate": 8e-05,
+      "loss": 1.5975,
+      "step": 1454
+    },
+    {
+      "epoch": 0.30218068535825543,
+      "grad_norm": 0.4875079095363617,
+      "learning_rate": 8e-05,
+      "loss": 1.5574,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3023883696780893,
+      "grad_norm": 0.5041059851646423,
+      "learning_rate": 8e-05,
+      "loss": 1.566,
+      "step": 1456
+    },
+    {
+      "epoch": 0.30259605399792316,
+      "grad_norm": 0.5048243403434753,
+      "learning_rate": 8e-05,
+      "loss": 1.5744,
+      "step": 1457
+    },
+    {
+      "epoch": 0.30280373831775703,
+      "grad_norm": 0.48077791929244995,
+      "learning_rate": 8e-05,
+      "loss": 1.6088,
+      "step": 1458
+    },
+    {
+      "epoch": 0.30301142263759084,
+      "grad_norm": 0.49609705805778503,
+      "learning_rate": 8e-05,
+      "loss": 1.6863,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3032191069574247,
+      "grad_norm": 0.5017523765563965,
+      "learning_rate": 8e-05,
+      "loss": 1.6013,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3034267912772586,
+      "grad_norm": 0.4952293634414673,
+      "learning_rate": 8e-05,
+      "loss": 1.6157,
+      "step": 1461
+    },
+    {
+      "epoch": 0.30363447559709245,
+      "grad_norm": 0.4877483546733856,
+      "learning_rate": 8e-05,
+      "loss": 1.5992,
+      "step": 1462
+    },
+    {
+      "epoch": 0.30384215991692626,
+      "grad_norm": 0.4839021861553192,
+      "learning_rate": 8e-05,
+      "loss": 1.66,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3040498442367601,
+      "grad_norm": 0.523611307144165,
+      "learning_rate": 8e-05,
+      "loss": 1.6295,
+      "step": 1464
+    },
+    {
+      "epoch": 0.304257528556594,
+      "grad_norm": 0.4874498248100281,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3044652128764278,
+      "grad_norm": 0.49205684661865234,
+      "learning_rate": 8e-05,
+      "loss": 1.575,
+      "step": 1466
+    },
+    {
+      "epoch": 0.30467289719626167,
+      "grad_norm": 0.5001851916313171,
+      "learning_rate": 8e-05,
+      "loss": 1.6221,
+      "step": 1467
+    },
+    {
+      "epoch": 0.30488058151609554,
+      "grad_norm": 0.5134277939796448,
+      "learning_rate": 8e-05,
+      "loss": 1.5791,
+      "step": 1468
+    },
+    {
+      "epoch": 0.3050882658359294,
+      "grad_norm": 0.4986041784286499,
+      "learning_rate": 8e-05,
+      "loss": 1.6249,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3052959501557632,
+      "grad_norm": 0.4865819811820984,
+      "learning_rate": 8e-05,
+      "loss": 1.6065,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3055036344755971,
+      "grad_norm": 0.48902952671051025,
+      "learning_rate": 8e-05,
+      "loss": 1.6383,
+      "step": 1471
+    },
+    {
+      "epoch": 0.30571131879543095,
+      "grad_norm": 0.5065701007843018,
+      "learning_rate": 8e-05,
+      "loss": 1.5784,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3059190031152648,
+      "grad_norm": 0.48749783635139465,
+      "learning_rate": 8e-05,
+      "loss": 1.5601,
+      "step": 1473
+    },
+    {
+      "epoch": 0.30612668743509863,
+      "grad_norm": 0.517052948474884,
+      "learning_rate": 8e-05,
+      "loss": 1.6655,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3063343717549325,
+      "grad_norm": 0.4751560389995575,
+      "learning_rate": 8e-05,
+      "loss": 1.5892,
+      "step": 1475
+    },
+    {
+      "epoch": 0.30654205607476637,
+      "grad_norm": 0.4731171429157257,
+      "learning_rate": 8e-05,
+      "loss": 1.5153,
+      "step": 1476
+    },
+    {
+      "epoch": 0.30674974039460023,
+      "grad_norm": 0.5009525418281555,
+      "learning_rate": 8e-05,
+      "loss": 1.6547,
+      "step": 1477
+    },
+    {
+      "epoch": 0.30695742471443405,
+      "grad_norm": 0.4930509626865387,
+      "learning_rate": 8e-05,
+      "loss": 1.513,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3071651090342679,
+      "grad_norm": 0.5279929041862488,
+      "learning_rate": 8e-05,
+      "loss": 1.6086,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3073727933541018,
+      "grad_norm": 0.4923204481601715,
+      "learning_rate": 8e-05,
+      "loss": 1.5659,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3075804776739356,
+      "grad_norm": 0.4844015836715698,
+      "learning_rate": 8e-05,
+      "loss": 1.6044,
+      "step": 1481
+    },
+    {
+      "epoch": 0.30778816199376946,
+      "grad_norm": 0.5059119462966919,
+      "learning_rate": 8e-05,
+      "loss": 1.61,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3079958463136033,
+      "grad_norm": 0.5250138640403748,
+      "learning_rate": 8e-05,
+      "loss": 1.5921,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3082035306334372,
+      "grad_norm": 0.512847363948822,
+      "learning_rate": 8e-05,
+      "loss": 1.6694,
+      "step": 1484
+    },
+    {
+      "epoch": 0.308411214953271,
+      "grad_norm": 0.5249167084693909,
+      "learning_rate": 8e-05,
+      "loss": 1.6558,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3086188992731049,
+      "grad_norm": 0.5241632461547852,
+      "learning_rate": 8e-05,
+      "loss": 1.6323,
+      "step": 1486
+    },
+    {
+      "epoch": 0.30882658359293874,
+      "grad_norm": 0.49093738198280334,
+      "learning_rate": 8e-05,
+      "loss": 1.6032,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3090342679127726,
+      "grad_norm": 0.49326688051223755,
+      "learning_rate": 8e-05,
+      "loss": 1.5726,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3092419522326064,
+      "grad_norm": 0.49778178334236145,
+      "learning_rate": 8e-05,
+      "loss": 1.5792,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3094496365524403,
+      "grad_norm": 0.49950110912323,
+      "learning_rate": 8e-05,
+      "loss": 1.5706,
+      "step": 1490
+    },
+    {
+      "epoch": 0.30965732087227416,
+      "grad_norm": 0.5080294013023376,
+      "learning_rate": 8e-05,
+      "loss": 1.64,
+      "step": 1491
+    },
+    {
+      "epoch": 0.309865005192108,
+      "grad_norm": 0.5099862217903137,
+      "learning_rate": 8e-05,
+      "loss": 1.6631,
+      "step": 1492
+    },
+    {
+      "epoch": 0.31007268951194183,
+      "grad_norm": 0.49251455068588257,
+      "learning_rate": 8e-05,
+      "loss": 1.5711,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3102803738317757,
+      "grad_norm": 0.49985072016716003,
+      "learning_rate": 8e-05,
+      "loss": 1.5286,
+      "step": 1494
+    },
+    {
+      "epoch": 0.31048805815160957,
+      "grad_norm": 0.5015414953231812,
+      "learning_rate": 8e-05,
+      "loss": 1.5992,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3106957424714434,
+      "grad_norm": 0.5101944804191589,
+      "learning_rate": 8e-05,
+      "loss": 1.5845,
+      "step": 1496
+    },
+    {
+      "epoch": 0.31090342679127725,
+      "grad_norm": 0.5153470635414124,
+      "learning_rate": 8e-05,
+      "loss": 1.7133,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3111111111111111,
+      "grad_norm": 0.5057701468467712,
+      "learning_rate": 8e-05,
+      "loss": 1.6648,
+      "step": 1498
+    },
+    {
+      "epoch": 0.311318795430945,
+      "grad_norm": 0.5079687833786011,
+      "learning_rate": 8e-05,
+      "loss": 1.6237,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3115264797507788,
+      "grad_norm": 0.475410133600235,
+      "learning_rate": 8e-05,
+      "loss": 1.6165,
+      "step": 1500
+    },
+    {
+      "epoch": 0.31173416407061266,
+      "grad_norm": 0.5307968854904175,
+      "learning_rate": 8e-05,
+      "loss": 1.5859,
+      "step": 1501
+    },
+    {
+      "epoch": 0.31194184839044653,
+      "grad_norm": 0.47486138343811035,
+      "learning_rate": 8e-05,
+      "loss": 1.5676,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3121495327102804,
+      "grad_norm": 0.5097178220748901,
+      "learning_rate": 8e-05,
+      "loss": 1.5636,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3123572170301142,
+      "grad_norm": 0.4760870039463043,
+      "learning_rate": 8e-05,
+      "loss": 1.5721,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3125649013499481,
+      "grad_norm": 0.4842042624950409,
+      "learning_rate": 8e-05,
+      "loss": 1.6645,
+      "step": 1505
+    },
+    {
+      "epoch": 0.31277258566978194,
+      "grad_norm": 0.5192430019378662,
+      "learning_rate": 8e-05,
+      "loss": 1.6495,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3129802699896158,
+      "grad_norm": 0.4800757169723511,
+      "learning_rate": 8e-05,
+      "loss": 1.5741,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3131879543094496,
+      "grad_norm": 0.4967311918735504,
+      "learning_rate": 8e-05,
+      "loss": 1.6412,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3133956386292835,
+      "grad_norm": 0.4981956481933594,
+      "learning_rate": 8e-05,
+      "loss": 1.6111,
+      "step": 1509
+    },
+    {
+      "epoch": 0.31360332294911736,
+      "grad_norm": 0.5167219042778015,
+      "learning_rate": 8e-05,
+      "loss": 1.6406,
+      "step": 1510
+    },
+    {
+      "epoch": 0.31381100726895117,
+      "grad_norm": 0.48610803484916687,
+      "learning_rate": 8e-05,
+      "loss": 1.5525,
+      "step": 1511
+    },
+    {
+      "epoch": 0.31401869158878504,
+      "grad_norm": 0.47807812690734863,
+      "learning_rate": 8e-05,
+      "loss": 1.5205,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3142263759086189,
+      "grad_norm": 0.5001601576805115,
+      "learning_rate": 8e-05,
+      "loss": 1.6265,
+      "step": 1513
+    },
+    {
+      "epoch": 0.31443406022845277,
+      "grad_norm": 0.5112875699996948,
+      "learning_rate": 8e-05,
+      "loss": 1.6357,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3146417445482866,
+      "grad_norm": 0.4997986853122711,
+      "learning_rate": 8e-05,
+      "loss": 1.6137,
+      "step": 1515
+    },
+    {
+      "epoch": 0.31484942886812045,
+      "grad_norm": 0.4741184115409851,
+      "learning_rate": 8e-05,
+      "loss": 1.6266,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3150571131879543,
+      "grad_norm": 0.5144311785697937,
+      "learning_rate": 8e-05,
+      "loss": 1.5782,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3152647975077882,
+      "grad_norm": 0.5066690444946289,
+      "learning_rate": 8e-05,
+      "loss": 1.6677,
+      "step": 1518
+    },
+    {
+      "epoch": 0.315472481827622,
+      "grad_norm": 0.5042521953582764,
+      "learning_rate": 8e-05,
+      "loss": 1.5916,
+      "step": 1519
+    },
+    {
+      "epoch": 0.31568016614745587,
+      "grad_norm": 0.4977933466434479,
+      "learning_rate": 8e-05,
+      "loss": 1.5574,
+      "step": 1520
+    },
+    {
+      "epoch": 0.31588785046728973,
+      "grad_norm": 0.5137097239494324,
+      "learning_rate": 8e-05,
+      "loss": 1.6411,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3160955347871236,
+      "grad_norm": 0.4740540385246277,
+      "learning_rate": 8e-05,
+      "loss": 1.57,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3163032191069574,
+      "grad_norm": 0.5093042254447937,
+      "learning_rate": 8e-05,
+      "loss": 1.6752,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3165109034267913,
+      "grad_norm": 0.47612494230270386,
+      "learning_rate": 8e-05,
+      "loss": 1.5722,
+      "step": 1524
+    },
+    {
+      "epoch": 0.31671858774662515,
+      "grad_norm": 0.5046824812889099,
+      "learning_rate": 8e-05,
+      "loss": 1.649,
+      "step": 1525
+    },
+    {
+      "epoch": 0.31692627206645896,
+      "grad_norm": 0.49512314796447754,
+      "learning_rate": 8e-05,
+      "loss": 1.5808,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3171339563862928,
+      "grad_norm": 0.4888535737991333,
+      "learning_rate": 8e-05,
+      "loss": 1.6138,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3173416407061267,
+      "grad_norm": 0.5067277550697327,
+      "learning_rate": 8e-05,
+      "loss": 1.626,
+      "step": 1528
+    },
+    {
+      "epoch": 0.31754932502596056,
+      "grad_norm": 0.47953104972839355,
+      "learning_rate": 8e-05,
+      "loss": 1.6125,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3177570093457944,
+      "grad_norm": 0.5111708641052246,
+      "learning_rate": 8e-05,
+      "loss": 1.5847,
+      "step": 1530
+    },
+    {
+      "epoch": 0.31796469366562824,
+      "grad_norm": 0.48254770040512085,
+      "learning_rate": 8e-05,
+      "loss": 1.5371,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3181723779854621,
+      "grad_norm": 0.6176688075065613,
+      "learning_rate": 8e-05,
+      "loss": 1.5441,
+      "step": 1532
+    },
+    {
+      "epoch": 0.318380062305296,
+      "grad_norm": 0.4850156307220459,
+      "learning_rate": 8e-05,
+      "loss": 1.573,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3185877466251298,
+      "grad_norm": 0.4983154237270355,
+      "learning_rate": 8e-05,
+      "loss": 1.5298,
+      "step": 1534
+    },
+    {
+      "epoch": 0.31879543094496365,
+      "grad_norm": 0.5193361043930054,
+      "learning_rate": 8e-05,
+      "loss": 1.6185,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3190031152647975,
+      "grad_norm": 0.483345627784729,
+      "learning_rate": 8e-05,
+      "loss": 1.5397,
+      "step": 1536
+    },
+    {
+      "epoch": 0.31921079958463133,
+      "grad_norm": 0.48317113518714905,
+      "learning_rate": 8e-05,
+      "loss": 1.6233,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3194184839044652,
+      "grad_norm": 0.4782498776912689,
+      "learning_rate": 8e-05,
+      "loss": 1.5813,
+      "step": 1538
+    },
+    {
+      "epoch": 0.31962616822429907,
+      "grad_norm": 0.47537708282470703,
+      "learning_rate": 8e-05,
+      "loss": 1.5454,
+      "step": 1539
+    },
+    {
+      "epoch": 0.31983385254413293,
+      "grad_norm": 0.4868844151496887,
+      "learning_rate": 8e-05,
+      "loss": 1.6,
+      "step": 1540
+    },
+    {
+      "epoch": 0.32004153686396675,
+      "grad_norm": 0.5081853866577148,
+      "learning_rate": 8e-05,
+      "loss": 1.5926,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3202492211838006,
+      "grad_norm": 0.48315268754959106,
+      "learning_rate": 8e-05,
+      "loss": 1.5898,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3204569055036345,
+      "grad_norm": 0.4909741282463074,
+      "learning_rate": 8e-05,
+      "loss": 1.5828,
+      "step": 1543
+    },
+    {
+      "epoch": 0.32066458982346835,
+      "grad_norm": 0.49861210584640503,
+      "learning_rate": 8e-05,
+      "loss": 1.6455,
+      "step": 1544
+    },
+    {
+      "epoch": 0.32087227414330216,
+      "grad_norm": 0.478344589471817,
+      "learning_rate": 8e-05,
+      "loss": 1.6188,
+      "step": 1545
+    },
+    {
+      "epoch": 0.32107995846313603,
+      "grad_norm": 0.5060451030731201,
+      "learning_rate": 8e-05,
+      "loss": 1.5827,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3212876427829699,
+      "grad_norm": 0.48769256472587585,
+      "learning_rate": 8e-05,
+      "loss": 1.6526,
+      "step": 1547
+    },
+    {
+      "epoch": 0.32149532710280376,
+      "grad_norm": 0.492779403924942,
+      "learning_rate": 8e-05,
+      "loss": 1.6032,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3217030114226376,
+      "grad_norm": 0.4987448453903198,
+      "learning_rate": 8e-05,
+      "loss": 1.5259,
+      "step": 1549
+    },
+    {
+      "epoch": 0.32191069574247144,
+      "grad_norm": 0.48688334226608276,
+      "learning_rate": 8e-05,
+      "loss": 1.5775,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3221183800623053,
+      "grad_norm": 0.4970977306365967,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3223260643821391,
+      "grad_norm": 0.5189974904060364,
+      "learning_rate": 8e-05,
+      "loss": 1.5758,
+      "step": 1552
+    },
+    {
+      "epoch": 0.322533748701973,
+      "grad_norm": 0.48815828561782837,
+      "learning_rate": 8e-05,
+      "loss": 1.5439,
+      "step": 1553
+    },
+    {
+      "epoch": 0.32274143302180686,
+      "grad_norm": 0.4929064214229584,
+      "learning_rate": 8e-05,
+      "loss": 1.545,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3229491173416407,
+      "grad_norm": 0.48322778940200806,
+      "learning_rate": 8e-05,
+      "loss": 1.5601,
+      "step": 1555
+    },
+    {
+      "epoch": 0.32315680166147454,
+      "grad_norm": 0.5003882646560669,
+      "learning_rate": 8e-05,
+      "loss": 1.5581,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3233644859813084,
+      "grad_norm": 0.5089275240898132,
+      "learning_rate": 8e-05,
+      "loss": 1.6147,
+      "step": 1557
+    },
+    {
+      "epoch": 0.32357217030114227,
+      "grad_norm": 0.48767319321632385,
+      "learning_rate": 8e-05,
+      "loss": 1.5953,
+      "step": 1558
+    },
+    {
+      "epoch": 0.32377985462097614,
+      "grad_norm": 0.5171458125114441,
+      "learning_rate": 8e-05,
+      "loss": 1.6395,
+      "step": 1559
+    },
+    {
+      "epoch": 0.32398753894080995,
+      "grad_norm": 0.4896124601364136,
+      "learning_rate": 8e-05,
+      "loss": 1.6219,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3241952232606438,
+      "grad_norm": 0.4953930974006653,
+      "learning_rate": 8e-05,
+      "loss": 1.6007,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3244029075804777,
+      "grad_norm": 0.5223313570022583,
+      "learning_rate": 8e-05,
+      "loss": 1.6687,
+      "step": 1562
+    },
+    {
+      "epoch": 0.32461059190031155,
+      "grad_norm": 0.50156569480896,
+      "learning_rate": 8e-05,
+      "loss": 1.6407,
+      "step": 1563
+    },
+    {
+      "epoch": 0.32481827622014536,
+      "grad_norm": 0.48758435249328613,
+      "learning_rate": 8e-05,
+      "loss": 1.5725,
+      "step": 1564
+    },
+    {
+      "epoch": 0.32502596053997923,
+      "grad_norm": 0.49146777391433716,
+      "learning_rate": 8e-05,
+      "loss": 1.5366,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3252336448598131,
+      "grad_norm": 0.5082824230194092,
+      "learning_rate": 8e-05,
+      "loss": 1.6088,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3254413291796469,
+      "grad_norm": 0.5137510299682617,
+      "learning_rate": 8e-05,
+      "loss": 1.6697,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3256490134994808,
+      "grad_norm": 0.48702284693717957,
+      "learning_rate": 8e-05,
+      "loss": 1.5483,
+      "step": 1568
+    },
+    {
+      "epoch": 0.32585669781931464,
+      "grad_norm": 0.4997518062591553,
+      "learning_rate": 8e-05,
+      "loss": 1.5455,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3260643821391485,
+      "grad_norm": 0.4797830283641815,
+      "learning_rate": 8e-05,
+      "loss": 1.6014,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3262720664589823,
+      "grad_norm": 0.4881299138069153,
+      "learning_rate": 8e-05,
+      "loss": 1.5972,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3264797507788162,
+      "grad_norm": 0.4965881407260895,
+      "learning_rate": 8e-05,
+      "loss": 1.5945,
+      "step": 1572
+    },
+    {
+      "epoch": 0.32668743509865006,
+      "grad_norm": 0.49967294931411743,
+      "learning_rate": 8e-05,
+      "loss": 1.5934,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3268951194184839,
+      "grad_norm": 0.5094643831253052,
+      "learning_rate": 8e-05,
+      "loss": 1.6762,
+      "step": 1574
+    },
+    {
+      "epoch": 0.32710280373831774,
+      "grad_norm": 0.4757131040096283,
+      "learning_rate": 8e-05,
+      "loss": 1.6471,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3273104880581516,
+      "grad_norm": 0.5002568364143372,
+      "learning_rate": 8e-05,
+      "loss": 1.538,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3275181723779855,
+      "grad_norm": 0.5113179087638855,
+      "learning_rate": 8e-05,
+      "loss": 1.5968,
+      "step": 1577
+    },
+    {
+      "epoch": 0.32772585669781934,
+      "grad_norm": 0.4950655698776245,
+      "learning_rate": 8e-05,
+      "loss": 1.6031,
+      "step": 1578
+    },
+    {
+      "epoch": 0.32793354101765315,
+      "grad_norm": 0.4921990931034088,
+      "learning_rate": 8e-05,
+      "loss": 1.5819,
+      "step": 1579
+    },
+    {
+      "epoch": 0.328141225337487,
+      "grad_norm": 0.48755714297294617,
+      "learning_rate": 8e-05,
+      "loss": 1.5708,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3283489096573209,
+      "grad_norm": 0.49338051676750183,
+      "learning_rate": 8e-05,
+      "loss": 1.6067,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3285565939771547,
+      "grad_norm": 0.4969124495983124,
+      "learning_rate": 8e-05,
+      "loss": 1.5753,
+      "step": 1582
+    },
+    {
+      "epoch": 0.32876427829698857,
+      "grad_norm": 0.49120715260505676,
+      "learning_rate": 8e-05,
+      "loss": 1.5947,
+      "step": 1583
+    },
+    {
+      "epoch": 0.32897196261682243,
+      "grad_norm": 0.47977736592292786,
+      "learning_rate": 8e-05,
+      "loss": 1.5577,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3291796469366563,
+      "grad_norm": 0.49589675664901733,
+      "learning_rate": 8e-05,
+      "loss": 1.6218,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3293873312564901,
+      "grad_norm": 0.506141185760498,
+      "learning_rate": 8e-05,
+      "loss": 1.5925,
+      "step": 1586
+    },
+    {
+      "epoch": 0.329595015576324,
+      "grad_norm": 0.5115627646446228,
+      "learning_rate": 8e-05,
+      "loss": 1.6107,
+      "step": 1587
+    },
+    {
+      "epoch": 0.32980269989615785,
+      "grad_norm": 0.5363782644271851,
+      "learning_rate": 8e-05,
+      "loss": 1.6134,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3300103842159917,
+      "grad_norm": 0.5303878784179688,
+      "learning_rate": 8e-05,
+      "loss": 1.6054,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3302180685358255,
+      "grad_norm": 0.5335170030593872,
+      "learning_rate": 8e-05,
+      "loss": 1.5758,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3304257528556594,
+      "grad_norm": 0.49111613631248474,
+      "learning_rate": 8e-05,
+      "loss": 1.6672,
+      "step": 1591
+    },
+    {
+      "epoch": 0.33063343717549326,
+      "grad_norm": 0.4870830774307251,
+      "learning_rate": 8e-05,
+      "loss": 1.6392,
+      "step": 1592
+    },
+    {
+      "epoch": 0.33084112149532713,
+      "grad_norm": 0.5026964545249939,
+      "learning_rate": 8e-05,
+      "loss": 1.546,
+      "step": 1593
+    },
+    {
+      "epoch": 0.33104880581516094,
+      "grad_norm": 0.47665566205978394,
+      "learning_rate": 8e-05,
+      "loss": 1.5447,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3312564901349948,
+      "grad_norm": 0.4723517894744873,
+      "learning_rate": 8e-05,
+      "loss": 1.5622,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3314641744548287,
+      "grad_norm": 0.48958736658096313,
+      "learning_rate": 8e-05,
+      "loss": 1.5539,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3316718587746625,
+      "grad_norm": 0.5126999616622925,
+      "learning_rate": 8e-05,
+      "loss": 1.6348,
+      "step": 1597
+    },
+    {
+      "epoch": 0.33187954309449635,
+      "grad_norm": 0.5067579746246338,
+      "learning_rate": 8e-05,
+      "loss": 1.6046,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3320872274143302,
+      "grad_norm": 0.4940493404865265,
+      "learning_rate": 8e-05,
+      "loss": 1.5553,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3322949117341641,
+      "grad_norm": 0.5140849947929382,
+      "learning_rate": 8e-05,
+      "loss": 1.6741,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3325025960539979,
+      "grad_norm": 0.47822049260139465,
+      "learning_rate": 8e-05,
+      "loss": 1.5474,
+      "step": 1601
+    },
+    {
+      "epoch": 0.33271028037383177,
+      "grad_norm": 0.4843752384185791,
+      "learning_rate": 8e-05,
+      "loss": 1.5194,
+      "step": 1602
+    },
+    {
+      "epoch": 0.33291796469366564,
+      "grad_norm": 0.5137794613838196,
+      "learning_rate": 8e-05,
+      "loss": 1.6013,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3331256490134995,
+      "grad_norm": 0.4957379698753357,
+      "learning_rate": 8e-05,
+      "loss": 1.5426,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.527511477470398,
+      "learning_rate": 8e-05,
+      "loss": 1.656,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3335410176531672,
+      "grad_norm": 0.4836502969264984,
+      "learning_rate": 8e-05,
+      "loss": 1.5422,
+      "step": 1606
+    },
+    {
+      "epoch": 0.33374870197300105,
+      "grad_norm": 0.47433504462242126,
+      "learning_rate": 8e-05,
+      "loss": 1.5434,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3339563862928349,
+      "grad_norm": 0.5102041363716125,
+      "learning_rate": 8e-05,
+      "loss": 1.7392,
+      "step": 1608
+    },
+    {
+      "epoch": 0.33416407061266873,
+      "grad_norm": 0.4976068437099457,
+      "learning_rate": 8e-05,
+      "loss": 1.6558,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3343717549325026,
+      "grad_norm": 0.49436476826667786,
+      "learning_rate": 8e-05,
+      "loss": 1.5691,
+      "step": 1610
+    },
+    {
+      "epoch": 0.33457943925233646,
+      "grad_norm": 0.4891389310359955,
+      "learning_rate": 8e-05,
+      "loss": 1.6462,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3347871235721703,
+      "grad_norm": 0.5754526853561401,
+      "learning_rate": 8e-05,
+      "loss": 1.6686,
+      "step": 1612
+    },
+    {
+      "epoch": 0.33499480789200414,
+      "grad_norm": 0.49320200085639954,
+      "learning_rate": 8e-05,
+      "loss": 1.6849,
+      "step": 1613
+    },
+    {
+      "epoch": 0.335202492211838,
+      "grad_norm": 0.4995647370815277,
+      "learning_rate": 8e-05,
+      "loss": 1.5772,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3354101765316719,
+      "grad_norm": 0.48338550329208374,
+      "learning_rate": 8e-05,
+      "loss": 1.5217,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3356178608515057,
+      "grad_norm": 0.47079768776893616,
+      "learning_rate": 8e-05,
+      "loss": 1.5064,
+      "step": 1616
+    },
+    {
+      "epoch": 0.33582554517133956,
+      "grad_norm": 0.5078921914100647,
+      "learning_rate": 8e-05,
+      "loss": 1.588,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3360332294911734,
+      "grad_norm": 0.5282722115516663,
+      "learning_rate": 8e-05,
+      "loss": 1.5059,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3362409138110073,
+      "grad_norm": 0.5180701613426208,
+      "learning_rate": 8e-05,
+      "loss": 1.5696,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3364485981308411,
+      "grad_norm": 0.5044119954109192,
+      "learning_rate": 8e-05,
+      "loss": 1.5619,
+      "step": 1620
+    },
+    {
+      "epoch": 0.33665628245067497,
+      "grad_norm": 0.5039921402931213,
+      "learning_rate": 8e-05,
+      "loss": 1.6287,
+      "step": 1621
+    },
+    {
+      "epoch": 0.33686396677050884,
+      "grad_norm": 0.5082249641418457,
+      "learning_rate": 8e-05,
+      "loss": 1.5873,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3370716510903427,
+      "grad_norm": 0.48252761363983154,
+      "learning_rate": 8e-05,
+      "loss": 1.5671,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3372793354101765,
+      "grad_norm": 0.5083025097846985,
+      "learning_rate": 8e-05,
+      "loss": 1.6022,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3374870197300104,
+      "grad_norm": 0.5168206095695496,
+      "learning_rate": 8e-05,
+      "loss": 1.6703,
+      "step": 1625
+    },
+    {
+      "epoch": 0.33769470404984425,
+      "grad_norm": 0.4816582500934601,
+      "learning_rate": 8e-05,
+      "loss": 1.5788,
+      "step": 1626
+    },
+    {
+      "epoch": 0.33790238836967806,
+      "grad_norm": 0.490945965051651,
+      "learning_rate": 8e-05,
+      "loss": 1.5626,
+      "step": 1627
+    },
+    {
+      "epoch": 0.33811007268951193,
+      "grad_norm": 0.5070871710777283,
+      "learning_rate": 8e-05,
+      "loss": 1.676,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3383177570093458,
+      "grad_norm": 0.49754798412323,
+      "learning_rate": 8e-05,
+      "loss": 1.6187,
+      "step": 1629
+    },
+    {
+      "epoch": 0.33852544132917967,
+      "grad_norm": 0.500566303730011,
+      "learning_rate": 8e-05,
+      "loss": 1.6224,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3387331256490135,
+      "grad_norm": 0.5073097944259644,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 1631
+    },
+    {
+      "epoch": 0.33894080996884735,
+      "grad_norm": 0.4712946116924286,
+      "learning_rate": 8e-05,
+      "loss": 1.5831,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3391484942886812,
+      "grad_norm": 0.49603715538978577,
+      "learning_rate": 8e-05,
+      "loss": 1.6233,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3393561786085151,
+      "grad_norm": 0.5190393328666687,
+      "learning_rate": 8e-05,
+      "loss": 1.6539,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3395638629283489,
+      "grad_norm": 0.49072492122650146,
+      "learning_rate": 8e-05,
+      "loss": 1.51,
+      "step": 1635
+    },
+    {
+      "epoch": 0.33977154724818276,
+      "grad_norm": 0.49581220746040344,
+      "learning_rate": 8e-05,
+      "loss": 1.5618,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3399792315680166,
+      "grad_norm": 0.4976690113544464,
+      "learning_rate": 8e-05,
+      "loss": 1.619,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3401869158878505,
+      "grad_norm": 0.4935644865036011,
+      "learning_rate": 8e-05,
+      "loss": 1.6651,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3403946002076843,
+      "grad_norm": 0.5480844974517822,
+      "learning_rate": 8e-05,
+      "loss": 1.7232,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3406022845275182,
+      "grad_norm": 0.5207673907279968,
+      "learning_rate": 8e-05,
+      "loss": 1.646,
+      "step": 1640
+    },
+    {
+      "epoch": 0.34080996884735204,
+      "grad_norm": 0.5026153326034546,
+      "learning_rate": 8e-05,
+      "loss": 1.5155,
+      "step": 1641
+    },
+    {
+      "epoch": 0.34101765316718585,
+      "grad_norm": 0.4776262044906616,
+      "learning_rate": 8e-05,
+      "loss": 1.5248,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3412253374870197,
+      "grad_norm": 0.5081517100334167,
+      "learning_rate": 8e-05,
+      "loss": 1.6524,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3414330218068536,
+      "grad_norm": 0.4948354661464691,
+      "learning_rate": 8e-05,
+      "loss": 1.6231,
+      "step": 1644
+    },
+    {
+      "epoch": 0.34164070612668745,
+      "grad_norm": 0.5102015733718872,
+      "learning_rate": 8e-05,
+      "loss": 1.6152,
+      "step": 1645
+    },
+    {
+      "epoch": 0.34184839044652127,
+      "grad_norm": 0.5125258564949036,
+      "learning_rate": 8e-05,
+      "loss": 1.644,
+      "step": 1646
+    },
+    {
+      "epoch": 0.34205607476635513,
+      "grad_norm": 0.49975356459617615,
+      "learning_rate": 8e-05,
+      "loss": 1.5621,
+      "step": 1647
+    },
+    {
+      "epoch": 0.342263759086189,
+      "grad_norm": 0.5100876092910767,
+      "learning_rate": 8e-05,
+      "loss": 1.5948,
+      "step": 1648
+    },
+    {
+      "epoch": 0.34247144340602287,
+      "grad_norm": 0.4993584156036377,
+      "learning_rate": 8e-05,
+      "loss": 1.5826,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3426791277258567,
+      "grad_norm": 0.4974980652332306,
+      "learning_rate": 8e-05,
+      "loss": 1.5497,
+      "step": 1650
+    },
+    {
+      "epoch": 0.34288681204569055,
+      "grad_norm": 0.5110526084899902,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3430944963655244,
+      "grad_norm": 0.4917437434196472,
+      "learning_rate": 8e-05,
+      "loss": 1.5912,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3433021806853583,
+      "grad_norm": 0.5046085119247437,
+      "learning_rate": 8e-05,
+      "loss": 1.6313,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3435098650051921,
+      "grad_norm": 0.5008227229118347,
+      "learning_rate": 8e-05,
+      "loss": 1.5603,
+      "step": 1654
+    },
+    {
+      "epoch": 0.34371754932502596,
+      "grad_norm": 0.48017212748527527,
+      "learning_rate": 8e-05,
+      "loss": 1.5851,
+      "step": 1655
+    },
+    {
+      "epoch": 0.34392523364485983,
+      "grad_norm": 0.5130458474159241,
+      "learning_rate": 8e-05,
+      "loss": 1.6015,
+      "step": 1656
+    },
+    {
+      "epoch": 0.34413291796469364,
+      "grad_norm": 0.5011090636253357,
+      "learning_rate": 8e-05,
+      "loss": 1.586,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3443406022845275,
+      "grad_norm": 0.4897594451904297,
+      "learning_rate": 8e-05,
+      "loss": 1.5021,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3445482866043614,
+      "grad_norm": 0.6089925765991211,
+      "learning_rate": 8e-05,
+      "loss": 1.6325,
+      "step": 1659
+    },
+    {
+      "epoch": 0.34475597092419524,
+      "grad_norm": 0.5230244994163513,
+      "learning_rate": 8e-05,
+      "loss": 1.6087,
+      "step": 1660
+    },
+    {
+      "epoch": 0.34496365524402905,
+      "grad_norm": 0.5000819563865662,
+      "learning_rate": 8e-05,
+      "loss": 1.6082,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3451713395638629,
+      "grad_norm": 0.5213269591331482,
+      "learning_rate": 8e-05,
+      "loss": 1.645,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3453790238836968,
+      "grad_norm": 0.4994923174381256,
+      "learning_rate": 8e-05,
+      "loss": 1.5727,
+      "step": 1663
+    },
+    {
+      "epoch": 0.34558670820353066,
+      "grad_norm": 0.5164852142333984,
+      "learning_rate": 8e-05,
+      "loss": 1.6262,
+      "step": 1664
+    },
+    {
+      "epoch": 0.34579439252336447,
+      "grad_norm": 0.5150447487831116,
+      "learning_rate": 8e-05,
+      "loss": 1.6247,
+      "step": 1665
+    },
+    {
+      "epoch": 0.34600207684319834,
+      "grad_norm": 0.5243363976478577,
+      "learning_rate": 8e-05,
+      "loss": 1.6594,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3462097611630322,
+      "grad_norm": 0.48954543471336365,
+      "learning_rate": 8e-05,
+      "loss": 1.5796,
+      "step": 1667
+    },
+    {
+      "epoch": 0.34641744548286607,
+      "grad_norm": 0.5317123532295227,
+      "learning_rate": 8e-05,
+      "loss": 1.6389,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3466251298026999,
+      "grad_norm": 0.48210567235946655,
+      "learning_rate": 8e-05,
+      "loss": 1.611,
+      "step": 1669
+    },
+    {
+      "epoch": 0.34683281412253375,
+      "grad_norm": 0.48340731859207153,
+      "learning_rate": 8e-05,
+      "loss": 1.5521,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3470404984423676,
+      "grad_norm": 0.5049550533294678,
+      "learning_rate": 8e-05,
+      "loss": 1.5751,
+      "step": 1671
+    },
+    {
+      "epoch": 0.34724818276220143,
+      "grad_norm": 0.5040217638015747,
+      "learning_rate": 8e-05,
+      "loss": 1.5263,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3474558670820353,
+      "grad_norm": 0.5133804678916931,
+      "learning_rate": 8e-05,
+      "loss": 1.651,
+      "step": 1673
+    },
+    {
+      "epoch": 0.34766355140186916,
+      "grad_norm": 0.4757126569747925,
+      "learning_rate": 8e-05,
+      "loss": 1.5168,
+      "step": 1674
+    },
+    {
+      "epoch": 0.34787123572170303,
+      "grad_norm": 0.5413657426834106,
+      "learning_rate": 8e-05,
+      "loss": 1.6296,
+      "step": 1675
+    },
+    {
+      "epoch": 0.34807892004153684,
+      "grad_norm": 0.5213454961776733,
+      "learning_rate": 8e-05,
+      "loss": 1.5978,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3482866043613707,
+      "grad_norm": 0.4883634150028229,
+      "learning_rate": 8e-05,
+      "loss": 1.6488,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3484942886812046,
+      "grad_norm": 0.5572940111160278,
+      "learning_rate": 8e-05,
+      "loss": 1.5427,
+      "step": 1678
+    },
+    {
+      "epoch": 0.34870197300103845,
+      "grad_norm": 0.5513511300086975,
+      "learning_rate": 8e-05,
+      "loss": 1.648,
+      "step": 1679
+    },
+    {
+      "epoch": 0.34890965732087226,
+      "grad_norm": 0.5172803401947021,
+      "learning_rate": 8e-05,
+      "loss": 1.6542,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3491173416407061,
+      "grad_norm": 0.5045535564422607,
+      "learning_rate": 8e-05,
+      "loss": 1.5983,
+      "step": 1681
+    },
+    {
+      "epoch": 0.34932502596054,
+      "grad_norm": 0.49136364459991455,
+      "learning_rate": 8e-05,
+      "loss": 1.5774,
+      "step": 1682
+    },
+    {
+      "epoch": 0.34953271028037386,
+      "grad_norm": 0.4870191216468811,
+      "learning_rate": 8e-05,
+      "loss": 1.5807,
+      "step": 1683
+    },
+    {
+      "epoch": 0.34974039460020767,
+      "grad_norm": 0.524327278137207,
+      "learning_rate": 8e-05,
+      "loss": 1.6471,
+      "step": 1684
+    },
+    {
+      "epoch": 0.34994807892004154,
+      "grad_norm": 0.5097550749778748,
+      "learning_rate": 8e-05,
+      "loss": 1.5905,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3501557632398754,
+      "grad_norm": 0.516825795173645,
+      "learning_rate": 8e-05,
+      "loss": 1.6539,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3503634475597092,
+      "grad_norm": 0.5085751414299011,
+      "learning_rate": 8e-05,
+      "loss": 1.6649,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3505711318795431,
+      "grad_norm": 0.49749553203582764,
+      "learning_rate": 8e-05,
+      "loss": 1.5978,
+      "step": 1688
+    },
+    {
+      "epoch": 0.35077881619937695,
+      "grad_norm": 0.49809882044792175,
+      "learning_rate": 8e-05,
+      "loss": 1.6228,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3509865005192108,
+      "grad_norm": 0.48347166180610657,
+      "learning_rate": 8e-05,
+      "loss": 1.5463,
+      "step": 1690
+    },
+    {
+      "epoch": 0.35119418483904463,
+      "grad_norm": 0.4847400486469269,
+      "learning_rate": 8e-05,
+      "loss": 1.5379,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3514018691588785,
+      "grad_norm": 0.48331671953201294,
+      "learning_rate": 8e-05,
+      "loss": 1.5704,
+      "step": 1692
+    },
+    {
+      "epoch": 0.35160955347871237,
+      "grad_norm": 0.49434080719947815,
+      "learning_rate": 8e-05,
+      "loss": 1.6055,
+      "step": 1693
+    },
+    {
+      "epoch": 0.35181723779854623,
+      "grad_norm": 0.5171917676925659,
+      "learning_rate": 8e-05,
+      "loss": 1.6155,
+      "step": 1694
+    },
+    {
+      "epoch": 0.35202492211838005,
+      "grad_norm": 0.4984574317932129,
+      "learning_rate": 8e-05,
+      "loss": 1.5811,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3522326064382139,
+      "grad_norm": 0.4882890284061432,
+      "learning_rate": 8e-05,
+      "loss": 1.5948,
+      "step": 1696
+    },
+    {
+      "epoch": 0.3524402907580478,
+      "grad_norm": 0.48760464787483215,
+      "learning_rate": 8e-05,
+      "loss": 1.5775,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3526479750778816,
+      "grad_norm": 0.48519834876060486,
+      "learning_rate": 8e-05,
+      "loss": 1.5734,
+      "step": 1698
+    },
+    {
+      "epoch": 0.35285565939771546,
+      "grad_norm": 0.518786609172821,
+      "learning_rate": 8e-05,
+      "loss": 1.5969,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3530633437175493,
+      "grad_norm": 0.4933086335659027,
+      "learning_rate": 8e-05,
+      "loss": 1.5626,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3532710280373832,
+      "grad_norm": 0.5093357563018799,
+      "learning_rate": 8e-05,
+      "loss": 1.5513,
+      "step": 1701
+    },
+    {
+      "epoch": 0.353478712357217,
+      "grad_norm": 0.5171862244606018,
+      "learning_rate": 8e-05,
+      "loss": 1.5887,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3536863966770509,
+      "grad_norm": 0.5157456994056702,
+      "learning_rate": 8e-05,
+      "loss": 1.6256,
+      "step": 1703
+    },
+    {
+      "epoch": 0.35389408099688474,
+      "grad_norm": 0.5482586622238159,
+      "learning_rate": 8e-05,
+      "loss": 1.6649,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3541017653167186,
+      "grad_norm": 0.5034733414649963,
+      "learning_rate": 8e-05,
+      "loss": 1.617,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3543094496365524,
+      "grad_norm": 0.5075942277908325,
+      "learning_rate": 8e-05,
+      "loss": 1.6087,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3545171339563863,
+      "grad_norm": 0.5159339904785156,
+      "learning_rate": 8e-05,
+      "loss": 1.5778,
+      "step": 1707
+    },
+    {
+      "epoch": 0.35472481827622016,
+      "grad_norm": 0.5066738724708557,
+      "learning_rate": 8e-05,
+      "loss": 1.6548,
+      "step": 1708
+    },
+    {
+      "epoch": 0.354932502596054,
+      "grad_norm": 0.48170992732048035,
+      "learning_rate": 8e-05,
+      "loss": 1.6448,
+      "step": 1709
+    },
+    {
+      "epoch": 0.35514018691588783,
+      "grad_norm": 0.47656309604644775,
+      "learning_rate": 8e-05,
+      "loss": 1.5139,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3553478712357217,
+      "grad_norm": 0.5038301944732666,
+      "learning_rate": 8e-05,
+      "loss": 1.5363,
+      "step": 1711
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 0.5651991963386536,
+      "learning_rate": 8e-05,
+      "loss": 1.6422,
+      "step": 1712
+    },
+    {
+      "epoch": 0.3557632398753894,
+      "grad_norm": 0.49263623356819153,
+      "learning_rate": 8e-05,
+      "loss": 1.5621,
+      "step": 1713
+    },
+    {
+      "epoch": 0.35597092419522325,
+      "grad_norm": 0.5081530809402466,
+      "learning_rate": 8e-05,
+      "loss": 1.5904,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3561786085150571,
+      "grad_norm": 0.5229151844978333,
+      "learning_rate": 8e-05,
+      "loss": 1.6046,
+      "step": 1715
+    },
+    {
+      "epoch": 0.356386292834891,
+      "grad_norm": 0.4824669361114502,
+      "learning_rate": 8e-05,
+      "loss": 1.5738,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3565939771547248,
+      "grad_norm": 0.5122133493423462,
+      "learning_rate": 8e-05,
+      "loss": 1.635,
+      "step": 1717
+    },
+    {
+      "epoch": 0.35680166147455866,
+      "grad_norm": 0.4866729974746704,
+      "learning_rate": 8e-05,
+      "loss": 1.5987,
+      "step": 1718
+    },
+    {
+      "epoch": 0.35700934579439253,
+      "grad_norm": 0.5040173530578613,
+      "learning_rate": 8e-05,
+      "loss": 1.5918,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3572170301142264,
+      "grad_norm": 0.496620774269104,
+      "learning_rate": 8e-05,
+      "loss": 1.5897,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3574247144340602,
+      "grad_norm": 0.521233081817627,
+      "learning_rate": 8e-05,
+      "loss": 1.6024,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3576323987538941,
+      "grad_norm": 0.4817935526371002,
+      "learning_rate": 8e-05,
+      "loss": 1.5798,
+      "step": 1722
+    },
+    {
+      "epoch": 0.35784008307372794,
+      "grad_norm": 0.5107322335243225,
+      "learning_rate": 8e-05,
+      "loss": 1.6237,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3580477673935618,
+      "grad_norm": 0.49109092354774475,
+      "learning_rate": 8e-05,
+      "loss": 1.5852,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3582554517133956,
+      "grad_norm": 0.49932169914245605,
+      "learning_rate": 8e-05,
+      "loss": 1.5684,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3584631360332295,
+      "grad_norm": 0.4933273494243622,
+      "learning_rate": 8e-05,
+      "loss": 1.5987,
+      "step": 1726
+    },
+    {
+      "epoch": 0.35867082035306336,
+      "grad_norm": 0.5152563452720642,
+      "learning_rate": 8e-05,
+      "loss": 1.6229,
+      "step": 1727
+    },
+    {
+      "epoch": 0.35887850467289717,
+      "grad_norm": 0.49047115445137024,
+      "learning_rate": 8e-05,
+      "loss": 1.5985,
+      "step": 1728
+    },
+    {
+      "epoch": 0.35908618899273104,
+      "grad_norm": 0.5049684643745422,
+      "learning_rate": 8e-05,
+      "loss": 1.5635,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3592938733125649,
+      "grad_norm": 0.4843440055847168,
+      "learning_rate": 8e-05,
+      "loss": 1.6021,
+      "step": 1730
+    },
+    {
+      "epoch": 0.35950155763239877,
+      "grad_norm": 0.48820793628692627,
+      "learning_rate": 8e-05,
+      "loss": 1.5598,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3597092419522326,
+      "grad_norm": 0.5228747725486755,
+      "learning_rate": 8e-05,
+      "loss": 1.5545,
+      "step": 1732
+    },
+    {
+      "epoch": 0.35991692627206645,
+      "grad_norm": 0.4973585903644562,
+      "learning_rate": 8e-05,
+      "loss": 1.5758,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3601246105919003,
+      "grad_norm": 0.5158231258392334,
+      "learning_rate": 8e-05,
+      "loss": 1.6215,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3603322949117342,
+      "grad_norm": 0.4845174551010132,
+      "learning_rate": 8e-05,
+      "loss": 1.5932,
+      "step": 1735
+    },
+    {
+      "epoch": 0.360539979231568,
+      "grad_norm": 0.505311906337738,
+      "learning_rate": 8e-05,
+      "loss": 1.6716,
+      "step": 1736
+    },
+    {
+      "epoch": 0.36074766355140186,
+      "grad_norm": 0.5100334286689758,
+      "learning_rate": 8e-05,
+      "loss": 1.6133,
+      "step": 1737
+    },
+    {
+      "epoch": 0.36095534787123573,
+      "grad_norm": 0.510661244392395,
+      "learning_rate": 8e-05,
+      "loss": 1.6705,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3611630321910696,
+      "grad_norm": 0.49372726678848267,
+      "learning_rate": 8e-05,
+      "loss": 1.6432,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3613707165109034,
+      "grad_norm": 0.4995138645172119,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3615784008307373,
+      "grad_norm": 0.518836259841919,
+      "learning_rate": 8e-05,
+      "loss": 1.5321,
+      "step": 1741
+    },
+    {
+      "epoch": 0.36178608515057115,
+      "grad_norm": 0.5100981593132019,
+      "learning_rate": 8e-05,
+      "loss": 1.573,
+      "step": 1742
+    },
+    {
+      "epoch": 0.36199376947040496,
+      "grad_norm": 0.4908148944377899,
+      "learning_rate": 8e-05,
+      "loss": 1.6022,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3622014537902388,
+      "grad_norm": 0.5129155516624451,
+      "learning_rate": 8e-05,
+      "loss": 1.5825,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3624091381100727,
+      "grad_norm": 0.4997279942035675,
+      "learning_rate": 8e-05,
+      "loss": 1.672,
+      "step": 1745
+    },
+    {
+      "epoch": 0.36261682242990656,
+      "grad_norm": 0.4920547902584076,
+      "learning_rate": 8e-05,
+      "loss": 1.5973,
+      "step": 1746
+    },
+    {
+      "epoch": 0.36282450674974037,
+      "grad_norm": 0.5037687420845032,
+      "learning_rate": 8e-05,
+      "loss": 1.4707,
+      "step": 1747
+    },
+    {
+      "epoch": 0.36303219106957424,
+      "grad_norm": 0.5036182999610901,
+      "learning_rate": 8e-05,
+      "loss": 1.6161,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3632398753894081,
+      "grad_norm": 0.5063603520393372,
+      "learning_rate": 8e-05,
+      "loss": 1.5124,
+      "step": 1749
+    },
+    {
+      "epoch": 0.363447559709242,
+      "grad_norm": 0.5060688257217407,
+      "learning_rate": 8e-05,
+      "loss": 1.553,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3636552440290758,
+      "grad_norm": 0.4944371283054352,
+      "learning_rate": 8e-05,
+      "loss": 1.572,
+      "step": 1751
+    },
+    {
+      "epoch": 0.36386292834890965,
+      "grad_norm": 0.5195178985595703,
+      "learning_rate": 8e-05,
+      "loss": 1.5257,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3640706126687435,
+      "grad_norm": 0.4922371208667755,
+      "learning_rate": 8e-05,
+      "loss": 1.5622,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3642782969885774,
+      "grad_norm": 0.5210584998130798,
+      "learning_rate": 8e-05,
+      "loss": 1.6212,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3644859813084112,
+      "grad_norm": 0.5490279197692871,
+      "learning_rate": 8e-05,
+      "loss": 1.6151,
+      "step": 1755
+    },
+    {
+      "epoch": 0.36469366562824507,
+      "grad_norm": 0.4872822165489197,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 1756
+    },
+    {
+      "epoch": 0.36490134994807893,
+      "grad_norm": 0.5005167126655579,
+      "learning_rate": 8e-05,
+      "loss": 1.5788,
+      "step": 1757
+    },
+    {
+      "epoch": 0.36510903426791275,
+      "grad_norm": 0.4953737258911133,
+      "learning_rate": 8e-05,
+      "loss": 1.6162,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3653167185877466,
+      "grad_norm": 0.5173213481903076,
+      "learning_rate": 8e-05,
+      "loss": 1.569,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3655244029075805,
+      "grad_norm": 0.5120421648025513,
+      "learning_rate": 8e-05,
+      "loss": 1.6336,
+      "step": 1760
+    },
+    {
+      "epoch": 0.36573208722741435,
+      "grad_norm": 0.5063916444778442,
+      "learning_rate": 8e-05,
+      "loss": 1.5695,
+      "step": 1761
+    },
+    {
+      "epoch": 0.36593977154724816,
+      "grad_norm": 0.5345103144645691,
+      "learning_rate": 8e-05,
+      "loss": 1.5974,
+      "step": 1762
+    },
+    {
+      "epoch": 0.36614745586708203,
+      "grad_norm": 0.47018736600875854,
+      "learning_rate": 8e-05,
+      "loss": 1.507,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3663551401869159,
+      "grad_norm": 0.5206469297409058,
+      "learning_rate": 8e-05,
+      "loss": 1.5425,
+      "step": 1764
+    },
+    {
+      "epoch": 0.36656282450674976,
+      "grad_norm": 0.49119699001312256,
+      "learning_rate": 8e-05,
+      "loss": 1.5484,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3667705088265836,
+      "grad_norm": 0.48123517632484436,
+      "learning_rate": 8e-05,
+      "loss": 1.5174,
+      "step": 1766
+    },
+    {
+      "epoch": 0.36697819314641744,
+      "grad_norm": 0.5121713280677795,
+      "learning_rate": 8e-05,
+      "loss": 1.6451,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3671858774662513,
+      "grad_norm": 0.5062276124954224,
+      "learning_rate": 8e-05,
+      "loss": 1.5787,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3673935617860852,
+      "grad_norm": 0.4850326180458069,
+      "learning_rate": 8e-05,
+      "loss": 1.605,
+      "step": 1769
+    },
+    {
+      "epoch": 0.367601246105919,
+      "grad_norm": 0.5524799823760986,
+      "learning_rate": 8e-05,
+      "loss": 1.6658,
+      "step": 1770
+    },
+    {
+      "epoch": 0.36780893042575286,
+      "grad_norm": 0.5042113661766052,
+      "learning_rate": 8e-05,
+      "loss": 1.5878,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3680166147455867,
+      "grad_norm": 0.5291147828102112,
+      "learning_rate": 8e-05,
+      "loss": 1.6076,
+      "step": 1772
+    },
+    {
+      "epoch": 0.36822429906542054,
+      "grad_norm": 0.5282319188117981,
+      "learning_rate": 8e-05,
+      "loss": 1.6024,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3684319833852544,
+      "grad_norm": 0.5120516419410706,
+      "learning_rate": 8e-05,
+      "loss": 1.5431,
+      "step": 1774
+    },
+    {
+      "epoch": 0.36863966770508827,
+      "grad_norm": 0.5158196687698364,
+      "learning_rate": 8e-05,
+      "loss": 1.6004,
+      "step": 1775
+    },
+    {
+      "epoch": 0.36884735202492214,
+      "grad_norm": 0.50716632604599,
+      "learning_rate": 8e-05,
+      "loss": 1.6615,
+      "step": 1776
+    },
+    {
+      "epoch": 0.36905503634475595,
+      "grad_norm": 0.5380442142486572,
+      "learning_rate": 8e-05,
+      "loss": 1.5219,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3692627206645898,
+      "grad_norm": 0.4817923903465271,
+      "learning_rate": 8e-05,
+      "loss": 1.5422,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3694704049844237,
+      "grad_norm": 0.509099006652832,
+      "learning_rate": 8e-05,
+      "loss": 1.5917,
+      "step": 1779
+    },
+    {
+      "epoch": 0.36967808930425755,
+      "grad_norm": 0.5473061800003052,
+      "learning_rate": 8e-05,
+      "loss": 1.6635,
+      "step": 1780
+    },
+    {
+      "epoch": 0.36988577362409136,
+      "grad_norm": 0.5307387709617615,
+      "learning_rate": 8e-05,
+      "loss": 1.6171,
+      "step": 1781
+    },
+    {
+      "epoch": 0.37009345794392523,
+      "grad_norm": 0.5154988169670105,
+      "learning_rate": 8e-05,
+      "loss": 1.5596,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3703011422637591,
+      "grad_norm": 0.4852389097213745,
+      "learning_rate": 8e-05,
+      "loss": 1.5276,
+      "step": 1783
+    },
+    {
+      "epoch": 0.37050882658359297,
+      "grad_norm": 0.4957145154476166,
+      "learning_rate": 8e-05,
+      "loss": 1.5823,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3707165109034268,
+      "grad_norm": 0.5096319317817688,
+      "learning_rate": 8e-05,
+      "loss": 1.5288,
+      "step": 1785
+    },
+    {
+      "epoch": 0.37092419522326064,
+      "grad_norm": 0.5101845860481262,
+      "learning_rate": 8e-05,
+      "loss": 1.6088,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3711318795430945,
+      "grad_norm": 0.5254088640213013,
+      "learning_rate": 8e-05,
+      "loss": 1.6396,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3713395638629283,
+      "grad_norm": 0.5071054697036743,
+      "learning_rate": 8e-05,
+      "loss": 1.569,
+      "step": 1788
+    },
+    {
+      "epoch": 0.3715472481827622,
+      "grad_norm": 0.51131272315979,
+      "learning_rate": 8e-05,
+      "loss": 1.6019,
+      "step": 1789
+    },
+    {
+      "epoch": 0.37175493250259606,
+      "grad_norm": 0.5088433623313904,
+      "learning_rate": 8e-05,
+      "loss": 1.5169,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3719626168224299,
+      "grad_norm": 0.5493926405906677,
+      "learning_rate": 8e-05,
+      "loss": 1.595,
+      "step": 1791
+    },
+    {
+      "epoch": 0.37217030114226374,
+      "grad_norm": 0.4982267916202545,
+      "learning_rate": 8e-05,
+      "loss": 1.5716,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3723779854620976,
+      "grad_norm": 0.5942201614379883,
+      "learning_rate": 8e-05,
+      "loss": 1.6425,
+      "step": 1793
+    },
+    {
+      "epoch": 0.37258566978193147,
+      "grad_norm": 0.4930921196937561,
+      "learning_rate": 8e-05,
+      "loss": 1.5942,
+      "step": 1794
+    },
+    {
+      "epoch": 0.37279335410176534,
+      "grad_norm": 0.5201108455657959,
+      "learning_rate": 8e-05,
+      "loss": 1.6025,
+      "step": 1795
+    },
+    {
+      "epoch": 0.37300103842159915,
+      "grad_norm": 0.5413525104522705,
+      "learning_rate": 8e-05,
+      "loss": 1.6227,
+      "step": 1796
+    },
+    {
+      "epoch": 0.373208722741433,
+      "grad_norm": 0.4801349937915802,
+      "learning_rate": 8e-05,
+      "loss": 1.5519,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3734164070612669,
+      "grad_norm": 0.545444905757904,
+      "learning_rate": 8e-05,
+      "loss": 1.5614,
+      "step": 1798
+    },
+    {
+      "epoch": 0.37362409138110075,
+      "grad_norm": 0.5094887614250183,
+      "learning_rate": 8e-05,
+      "loss": 1.5961,
+      "step": 1799
+    },
+    {
+      "epoch": 0.37383177570093457,
+      "grad_norm": 0.5024911761283875,
+      "learning_rate": 8e-05,
+      "loss": 1.6617,
+      "step": 1800
+    },
+    {
+      "epoch": 0.37403946002076843,
+      "grad_norm": 0.4914880692958832,
+      "learning_rate": 8e-05,
+      "loss": 1.547,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3742471443406023,
+      "grad_norm": 0.488719642162323,
+      "learning_rate": 8e-05,
+      "loss": 1.5906,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3744548286604361,
+      "grad_norm": 0.5118622779846191,
+      "learning_rate": 8e-05,
+      "loss": 1.5741,
+      "step": 1803
+    },
+    {
+      "epoch": 0.37466251298027,
+      "grad_norm": 0.5298313498497009,
+      "learning_rate": 8e-05,
+      "loss": 1.6255,
+      "step": 1804
+    },
+    {
+      "epoch": 0.37487019730010385,
+      "grad_norm": 0.46859586238861084,
+      "learning_rate": 8e-05,
+      "loss": 1.5109,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3750778816199377,
+      "grad_norm": 0.5379807949066162,
+      "learning_rate": 8e-05,
+      "loss": 1.6255,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3752855659397715,
+      "grad_norm": 0.4950341284275055,
+      "learning_rate": 8e-05,
+      "loss": 1.659,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3754932502596054,
+      "grad_norm": 0.49582797288894653,
+      "learning_rate": 8e-05,
+      "loss": 1.6734,
+      "step": 1808
+    },
+    {
+      "epoch": 0.37570093457943926,
+      "grad_norm": 0.522197961807251,
+      "learning_rate": 8e-05,
+      "loss": 1.5411,
+      "step": 1809
+    },
+    {
+      "epoch": 0.37590861889927313,
+      "grad_norm": 0.49645760655403137,
+      "learning_rate": 8e-05,
+      "loss": 1.5901,
+      "step": 1810
+    },
+    {
+      "epoch": 0.37611630321910694,
+      "grad_norm": 0.5847791433334351,
+      "learning_rate": 8e-05,
+      "loss": 1.5925,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3763239875389408,
+      "grad_norm": 0.4936416447162628,
+      "learning_rate": 8e-05,
+      "loss": 1.6004,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3765316718587747,
+      "grad_norm": 0.5401305556297302,
+      "learning_rate": 8e-05,
+      "loss": 1.6375,
+      "step": 1813
+    },
+    {
+      "epoch": 0.37673935617860854,
+      "grad_norm": 0.5169132351875305,
+      "learning_rate": 8e-05,
+      "loss": 1.6277,
+      "step": 1814
+    },
+    {
+      "epoch": 0.37694704049844235,
+      "grad_norm": 0.5267390012741089,
+      "learning_rate": 8e-05,
+      "loss": 1.595,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3771547248182762,
+      "grad_norm": 0.5062999129295349,
+      "learning_rate": 8e-05,
+      "loss": 1.6584,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3773624091381101,
+      "grad_norm": 0.4729461967945099,
+      "learning_rate": 8e-05,
+      "loss": 1.5783,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3775700934579439,
+      "grad_norm": 0.4871976375579834,
+      "learning_rate": 8e-05,
+      "loss": 1.6477,
+      "step": 1818
+    },
+    {
+      "epoch": 0.37777777777777777,
+      "grad_norm": 0.4819161593914032,
+      "learning_rate": 8e-05,
+      "loss": 1.5632,
+      "step": 1819
+    },
+    {
+      "epoch": 0.37798546209761164,
+      "grad_norm": 0.48835310339927673,
+      "learning_rate": 8e-05,
+      "loss": 1.6809,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3781931464174455,
+      "grad_norm": 0.5206708908081055,
+      "learning_rate": 8e-05,
+      "loss": 1.6919,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3784008307372793,
+      "grad_norm": 0.49859198927879333,
+      "learning_rate": 8e-05,
+      "loss": 1.624,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3786085150571132,
+      "grad_norm": 0.47674262523651123,
+      "learning_rate": 8e-05,
+      "loss": 1.4634,
+      "step": 1823
+    },
+    {
+      "epoch": 0.37881619937694705,
+      "grad_norm": 0.5091409683227539,
+      "learning_rate": 8e-05,
+      "loss": 1.5764,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3790238836967809,
+      "grad_norm": 0.4878379702568054,
+      "learning_rate": 8e-05,
+      "loss": 1.6147,
+      "step": 1825
+    },
+    {
+      "epoch": 0.37923156801661473,
+      "grad_norm": 0.4939122200012207,
+      "learning_rate": 8e-05,
+      "loss": 1.5762,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3794392523364486,
+      "grad_norm": 0.5266447067260742,
+      "learning_rate": 8e-05,
+      "loss": 1.6838,
+      "step": 1827
+    },
+    {
+      "epoch": 0.37964693665628246,
+      "grad_norm": 0.49246594309806824,
+      "learning_rate": 8e-05,
+      "loss": 1.561,
+      "step": 1828
+    },
+    {
+      "epoch": 0.37985462097611633,
+      "grad_norm": 0.4985910952091217,
+      "learning_rate": 8e-05,
+      "loss": 1.4934,
+      "step": 1829
+    },
+    {
+      "epoch": 0.38006230529595014,
+      "grad_norm": 0.49930235743522644,
+      "learning_rate": 8e-05,
+      "loss": 1.5857,
+      "step": 1830
+    },
+    {
+      "epoch": 0.380269989615784,
+      "grad_norm": 0.5117544531822205,
+      "learning_rate": 8e-05,
+      "loss": 1.5451,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3804776739356179,
+      "grad_norm": 0.5053813457489014,
+      "learning_rate": 8e-05,
+      "loss": 1.5836,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3806853582554517,
+      "grad_norm": 0.5296281576156616,
+      "learning_rate": 8e-05,
+      "loss": 1.5972,
+      "step": 1833
+    },
+    {
+      "epoch": 0.38089304257528556,
+      "grad_norm": 0.5012361407279968,
+      "learning_rate": 8e-05,
+      "loss": 1.5809,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3811007268951194,
+      "grad_norm": 0.5064777135848999,
+      "learning_rate": 8e-05,
+      "loss": 1.5584,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3813084112149533,
+      "grad_norm": 0.5177083611488342,
+      "learning_rate": 8e-05,
+      "loss": 1.5851,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3815160955347871,
+      "grad_norm": 0.5089267492294312,
+      "learning_rate": 8e-05,
+      "loss": 1.5429,
+      "step": 1837
+    },
+    {
+      "epoch": 0.38172377985462097,
+      "grad_norm": 0.49458834528923035,
+      "learning_rate": 8e-05,
+      "loss": 1.6052,
+      "step": 1838
+    },
+    {
+      "epoch": 0.38193146417445484,
+      "grad_norm": 0.4878547489643097,
+      "learning_rate": 8e-05,
+      "loss": 1.5777,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3821391484942887,
+      "grad_norm": 0.5592621564865112,
+      "learning_rate": 8e-05,
+      "loss": 1.6285,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3823468328141225,
+      "grad_norm": 0.5030962824821472,
+      "learning_rate": 8e-05,
+      "loss": 1.5928,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3825545171339564,
+      "grad_norm": 0.5183880925178528,
+      "learning_rate": 8e-05,
+      "loss": 1.6346,
+      "step": 1842
+    },
+    {
+      "epoch": 0.38276220145379025,
+      "grad_norm": 0.4975292980670929,
+      "learning_rate": 8e-05,
+      "loss": 1.5236,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3829698857736241,
+      "grad_norm": 0.5079745650291443,
+      "learning_rate": 8e-05,
+      "loss": 1.5622,
+      "step": 1844
+    },
+    {
+      "epoch": 0.38317757009345793,
+      "grad_norm": 0.49533551931381226,
+      "learning_rate": 8e-05,
+      "loss": 1.563,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3833852544132918,
+      "grad_norm": 0.5062286257743835,
+      "learning_rate": 8e-05,
+      "loss": 1.6209,
+      "step": 1846
+    },
+    {
+      "epoch": 0.38359293873312567,
+      "grad_norm": 0.49643903970718384,
+      "learning_rate": 8e-05,
+      "loss": 1.6075,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3838006230529595,
+      "grad_norm": 0.6108715534210205,
+      "learning_rate": 8e-05,
+      "loss": 1.6111,
+      "step": 1848
+    },
+    {
+      "epoch": 0.38400830737279334,
+      "grad_norm": 0.5051819682121277,
+      "learning_rate": 8e-05,
+      "loss": 1.5039,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3842159916926272,
+      "grad_norm": 0.5083017945289612,
+      "learning_rate": 8e-05,
+      "loss": 1.5674,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3844236760124611,
+      "grad_norm": 0.5128228068351746,
+      "learning_rate": 8e-05,
+      "loss": 1.6137,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3846313603322949,
+      "grad_norm": 0.5230603218078613,
+      "learning_rate": 8e-05,
+      "loss": 1.6066,
+      "step": 1852
+    },
+    {
+      "epoch": 0.38483904465212876,
+      "grad_norm": 0.527700662612915,
+      "learning_rate": 8e-05,
+      "loss": 1.6253,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3850467289719626,
+      "grad_norm": 0.567374050617218,
+      "learning_rate": 8e-05,
+      "loss": 1.5531,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3852544132917965,
+      "grad_norm": 0.4868682622909546,
+      "learning_rate": 8e-05,
+      "loss": 1.6318,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3854620976116303,
+      "grad_norm": 0.5227965116500854,
+      "learning_rate": 8e-05,
+      "loss": 1.5589,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3856697819314642,
+      "grad_norm": 0.5066382884979248,
+      "learning_rate": 8e-05,
+      "loss": 1.6554,
+      "step": 1857
+    },
+    {
+      "epoch": 0.38587746625129804,
+      "grad_norm": 0.49491891264915466,
+      "learning_rate": 8e-05,
+      "loss": 1.6015,
+      "step": 1858
+    },
+    {
+      "epoch": 0.38608515057113185,
+      "grad_norm": 0.5157635807991028,
+      "learning_rate": 8e-05,
+      "loss": 1.6216,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3862928348909657,
+      "grad_norm": 0.5314090847969055,
+      "learning_rate": 8e-05,
+      "loss": 1.5958,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3865005192107996,
+      "grad_norm": 0.48347991704940796,
+      "learning_rate": 8e-05,
+      "loss": 1.5142,
+      "step": 1861
+    },
+    {
+      "epoch": 0.38670820353063345,
+      "grad_norm": 0.5229157209396362,
+      "learning_rate": 8e-05,
+      "loss": 1.5493,
+      "step": 1862
+    },
+    {
+      "epoch": 0.38691588785046727,
+      "grad_norm": 0.5444154739379883,
+      "learning_rate": 8e-05,
+      "loss": 1.6436,
+      "step": 1863
+    },
+    {
+      "epoch": 0.38712357217030113,
+      "grad_norm": 0.5127759575843811,
+      "learning_rate": 8e-05,
+      "loss": 1.6032,
+      "step": 1864
+    },
+    {
+      "epoch": 0.387331256490135,
+      "grad_norm": 0.4927530884742737,
+      "learning_rate": 8e-05,
+      "loss": 1.6051,
+      "step": 1865
+    },
+    {
+      "epoch": 0.38753894080996887,
+      "grad_norm": 0.5002699494361877,
+      "learning_rate": 8e-05,
+      "loss": 1.6294,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3877466251298027,
+      "grad_norm": 0.5165937542915344,
+      "learning_rate": 8e-05,
+      "loss": 1.5966,
+      "step": 1867
+    },
+    {
+      "epoch": 0.38795430944963655,
+      "grad_norm": 0.5008817315101624,
+      "learning_rate": 8e-05,
+      "loss": 1.5651,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3881619937694704,
+      "grad_norm": 0.49471011757850647,
+      "learning_rate": 8e-05,
+      "loss": 1.5699,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3883696780893043,
+      "grad_norm": 0.5040736198425293,
+      "learning_rate": 8e-05,
+      "loss": 1.6638,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3885773624091381,
+      "grad_norm": 0.4965846538543701,
+      "learning_rate": 8e-05,
+      "loss": 1.5604,
+      "step": 1871
+    },
+    {
+      "epoch": 0.38878504672897196,
+      "grad_norm": 0.48028695583343506,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 1872
+    },
+    {
+      "epoch": 0.38899273104880583,
+      "grad_norm": 0.506149411201477,
+      "learning_rate": 8e-05,
+      "loss": 1.5405,
+      "step": 1873
+    },
+    {
+      "epoch": 0.38920041536863964,
+      "grad_norm": 0.4821113049983978,
+      "learning_rate": 8e-05,
+      "loss": 1.6343,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3894080996884735,
+      "grad_norm": 0.545812726020813,
+      "learning_rate": 8e-05,
+      "loss": 1.6731,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3896157840083074,
+      "grad_norm": 0.5006311535835266,
+      "learning_rate": 8e-05,
+      "loss": 1.5129,
+      "step": 1876
+    },
+    {
+      "epoch": 0.38982346832814124,
+      "grad_norm": 0.48582524061203003,
+      "learning_rate": 8e-05,
+      "loss": 1.5433,
+      "step": 1877
+    },
+    {
+      "epoch": 0.39003115264797505,
+      "grad_norm": 0.5167272090911865,
+      "learning_rate": 8e-05,
+      "loss": 1.6357,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3902388369678089,
+      "grad_norm": 0.5142748951911926,
+      "learning_rate": 8e-05,
+      "loss": 1.6098,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3904465212876428,
+      "grad_norm": 0.5022456645965576,
+      "learning_rate": 8e-05,
+      "loss": 1.5784,
+      "step": 1880
+    },
+    {
+      "epoch": 0.39065420560747666,
+      "grad_norm": 0.5111859440803528,
+      "learning_rate": 8e-05,
+      "loss": 1.5819,
+      "step": 1881
+    },
+    {
+      "epoch": 0.39086188992731047,
+      "grad_norm": 0.49327921867370605,
+      "learning_rate": 8e-05,
+      "loss": 1.5932,
+      "step": 1882
+    },
+    {
+      "epoch": 0.39106957424714434,
+      "grad_norm": 0.5264003872871399,
+      "learning_rate": 8e-05,
+      "loss": 1.615,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3912772585669782,
+      "grad_norm": 0.4997003376483917,
+      "learning_rate": 8e-05,
+      "loss": 1.5645,
+      "step": 1884
+    },
+    {
+      "epoch": 0.39148494288681207,
+      "grad_norm": 0.48707181215286255,
+      "learning_rate": 8e-05,
+      "loss": 1.566,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3916926272066459,
+      "grad_norm": 0.5190602540969849,
+      "learning_rate": 8e-05,
+      "loss": 1.5918,
+      "step": 1886
+    },
+    {
+      "epoch": 0.39190031152647975,
+      "grad_norm": 0.5049939155578613,
+      "learning_rate": 8e-05,
+      "loss": 1.5857,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3921079958463136,
+      "grad_norm": 0.5178766250610352,
+      "learning_rate": 8e-05,
+      "loss": 1.6534,
+      "step": 1888
+    },
+    {
+      "epoch": 0.39231568016614743,
+      "grad_norm": 0.510301411151886,
+      "learning_rate": 8e-05,
+      "loss": 1.6123,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3925233644859813,
+      "grad_norm": 0.48783424496650696,
+      "learning_rate": 8e-05,
+      "loss": 1.51,
+      "step": 1890
+    },
+    {
+      "epoch": 0.39273104880581516,
+      "grad_norm": 0.5162112712860107,
+      "learning_rate": 8e-05,
+      "loss": 1.5571,
+      "step": 1891
+    },
+    {
+      "epoch": 0.39293873312564903,
+      "grad_norm": 0.5092341303825378,
+      "learning_rate": 8e-05,
+      "loss": 1.6414,
+      "step": 1892
+    },
+    {
+      "epoch": 0.39314641744548284,
+      "grad_norm": 0.5020235776901245,
+      "learning_rate": 8e-05,
+      "loss": 1.626,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3933541017653167,
+      "grad_norm": 0.5041497945785522,
+      "learning_rate": 8e-05,
+      "loss": 1.5922,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3935617860851506,
+      "grad_norm": 0.5252364277839661,
+      "learning_rate": 8e-05,
+      "loss": 1.6367,
+      "step": 1895
+    },
+    {
+      "epoch": 0.39376947040498445,
+      "grad_norm": 0.5267825126647949,
+      "learning_rate": 8e-05,
+      "loss": 1.6483,
+      "step": 1896
+    },
+    {
+      "epoch": 0.39397715472481826,
+      "grad_norm": 0.47401174902915955,
+      "learning_rate": 8e-05,
+      "loss": 1.4967,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3941848390446521,
+      "grad_norm": 0.47972801327705383,
+      "learning_rate": 8e-05,
+      "loss": 1.5402,
+      "step": 1898
+    },
+    {
+      "epoch": 0.394392523364486,
+      "grad_norm": 0.5107938051223755,
+      "learning_rate": 8e-05,
+      "loss": 1.6526,
+      "step": 1899
+    },
+    {
+      "epoch": 0.39460020768431986,
+      "grad_norm": 0.4895242154598236,
+      "learning_rate": 8e-05,
+      "loss": 1.5815,
+      "step": 1900
+    },
+    {
+      "epoch": 0.39480789200415367,
+      "grad_norm": 0.5084724426269531,
+      "learning_rate": 8e-05,
+      "loss": 1.5669,
+      "step": 1901
+    },
+    {
+      "epoch": 0.39501557632398754,
+      "grad_norm": 0.4843336045742035,
+      "learning_rate": 8e-05,
+      "loss": 1.5556,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3952232606438214,
+      "grad_norm": 0.5088340044021606,
+      "learning_rate": 8e-05,
+      "loss": 1.5571,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3954309449636552,
+      "grad_norm": 0.5239298939704895,
+      "learning_rate": 8e-05,
+      "loss": 1.6224,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3956386292834891,
+      "grad_norm": 0.49526721239089966,
+      "learning_rate": 8e-05,
+      "loss": 1.6317,
+      "step": 1905
+    },
+    {
+      "epoch": 0.39584631360332295,
+      "grad_norm": 0.5118904113769531,
+      "learning_rate": 8e-05,
+      "loss": 1.6292,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3960539979231568,
+      "grad_norm": 0.5064014196395874,
+      "learning_rate": 8e-05,
+      "loss": 1.6595,
+      "step": 1907
+    },
+    {
+      "epoch": 0.39626168224299063,
+      "grad_norm": 0.5004372000694275,
+      "learning_rate": 8e-05,
+      "loss": 1.5685,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3964693665628245,
+      "grad_norm": 0.5007834434509277,
+      "learning_rate": 8e-05,
+      "loss": 1.5655,
+      "step": 1909
+    },
+    {
+      "epoch": 0.39667705088265837,
+      "grad_norm": 0.4961642920970917,
+      "learning_rate": 8e-05,
+      "loss": 1.5386,
+      "step": 1910
+    },
+    {
+      "epoch": 0.39688473520249223,
+      "grad_norm": 0.49714717268943787,
+      "learning_rate": 8e-05,
+      "loss": 1.5585,
+      "step": 1911
+    },
+    {
+      "epoch": 0.39709241952232605,
+      "grad_norm": 0.5007422566413879,
+      "learning_rate": 8e-05,
+      "loss": 1.5251,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3973001038421599,
+      "grad_norm": 0.4922661781311035,
+      "learning_rate": 8e-05,
+      "loss": 1.5312,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3975077881619938,
+      "grad_norm": 0.5031492710113525,
+      "learning_rate": 8e-05,
+      "loss": 1.6502,
+      "step": 1914
+    },
+    {
+      "epoch": 0.39771547248182765,
+      "grad_norm": 0.4995499551296234,
+      "learning_rate": 8e-05,
+      "loss": 1.5721,
+      "step": 1915
+    },
+    {
+      "epoch": 0.39792315680166146,
+      "grad_norm": 0.5230512022972107,
+      "learning_rate": 8e-05,
+      "loss": 1.6209,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3981308411214953,
+      "grad_norm": 0.4955296516418457,
+      "learning_rate": 8e-05,
+      "loss": 1.6216,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3983385254413292,
+      "grad_norm": 0.49905237555503845,
+      "learning_rate": 8e-05,
+      "loss": 1.6325,
+      "step": 1918
+    },
+    {
+      "epoch": 0.398546209761163,
+      "grad_norm": 0.49909424781799316,
+      "learning_rate": 8e-05,
+      "loss": 1.6182,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3987538940809969,
+      "grad_norm": 0.5071685910224915,
+      "learning_rate": 8e-05,
+      "loss": 1.6344,
+      "step": 1920
+    },
+    {
+      "epoch": 0.39896157840083074,
+      "grad_norm": 0.4874131977558136,
+      "learning_rate": 8e-05,
+      "loss": 1.6058,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3991692627206646,
+      "grad_norm": 0.4769895374774933,
+      "learning_rate": 8e-05,
+      "loss": 1.4976,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3993769470404984,
+      "grad_norm": 0.47417256236076355,
+      "learning_rate": 8e-05,
+      "loss": 1.4958,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3995846313603323,
+      "grad_norm": 0.5021626353263855,
+      "learning_rate": 8e-05,
+      "loss": 1.6307,
+      "step": 1924
+    },
+    {
+      "epoch": 0.39979231568016615,
+      "grad_norm": 0.48706451058387756,
+      "learning_rate": 8e-05,
+      "loss": 1.5546,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5021299719810486,
+      "learning_rate": 8e-05,
+      "loss": 1.6139,
+      "step": 1926
+    },
+    {
+      "epoch": 0.40020768431983383,
+      "grad_norm": 0.4878026247024536,
+      "learning_rate": 8e-05,
+      "loss": 1.565,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4004153686396677,
+      "grad_norm": 0.4966025650501251,
+      "learning_rate": 8e-05,
+      "loss": 1.5577,
+      "step": 1928
+    },
+    {
+      "epoch": 0.40062305295950157,
+      "grad_norm": 0.49730899930000305,
+      "learning_rate": 8e-05,
+      "loss": 1.5656,
+      "step": 1929
+    },
+    {
+      "epoch": 0.40083073727933544,
+      "grad_norm": 0.5028313994407654,
+      "learning_rate": 8e-05,
+      "loss": 1.6599,
+      "step": 1930
+    },
+    {
+      "epoch": 0.40103842159916925,
+      "grad_norm": 0.5250521898269653,
+      "learning_rate": 8e-05,
+      "loss": 1.5481,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4012461059190031,
+      "grad_norm": 0.5033625364303589,
+      "learning_rate": 8e-05,
+      "loss": 1.5811,
+      "step": 1932
+    },
+    {
+      "epoch": 0.401453790238837,
+      "grad_norm": 0.4944176971912384,
+      "learning_rate": 8e-05,
+      "loss": 1.619,
+      "step": 1933
+    },
+    {
+      "epoch": 0.4016614745586708,
+      "grad_norm": 0.5117709040641785,
+      "learning_rate": 8e-05,
+      "loss": 1.5833,
+      "step": 1934
+    },
+    {
+      "epoch": 0.40186915887850466,
+      "grad_norm": 0.50387042760849,
+      "learning_rate": 8e-05,
+      "loss": 1.5767,
+      "step": 1935
+    },
+    {
+      "epoch": 0.40207684319833853,
+      "grad_norm": 0.5241379141807556,
+      "learning_rate": 8e-05,
+      "loss": 1.5629,
+      "step": 1936
+    },
+    {
+      "epoch": 0.4022845275181724,
+      "grad_norm": 0.5203877687454224,
+      "learning_rate": 8e-05,
+      "loss": 1.5862,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4024922118380062,
+      "grad_norm": 0.49413588643074036,
+      "learning_rate": 8e-05,
+      "loss": 1.5689,
+      "step": 1938
+    },
+    {
+      "epoch": 0.4026998961578401,
+      "grad_norm": 0.5159587860107422,
+      "learning_rate": 8e-05,
+      "loss": 1.6136,
+      "step": 1939
+    },
+    {
+      "epoch": 0.40290758047767394,
+      "grad_norm": 0.5072512626647949,
+      "learning_rate": 8e-05,
+      "loss": 1.5555,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4031152647975078,
+      "grad_norm": 0.500592052936554,
+      "learning_rate": 8e-05,
+      "loss": 1.5846,
+      "step": 1941
+    },
+    {
+      "epoch": 0.4033229491173416,
+      "grad_norm": 0.4672267735004425,
+      "learning_rate": 8e-05,
+      "loss": 1.5187,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4035306334371755,
+      "grad_norm": 0.48535841703414917,
+      "learning_rate": 8e-05,
+      "loss": 1.6116,
+      "step": 1943
+    },
+    {
+      "epoch": 0.40373831775700936,
+      "grad_norm": 0.48251843452453613,
+      "learning_rate": 8e-05,
+      "loss": 1.5747,
+      "step": 1944
+    },
+    {
+      "epoch": 0.4039460020768432,
+      "grad_norm": 0.4987678825855255,
+      "learning_rate": 8e-05,
+      "loss": 1.5857,
+      "step": 1945
+    },
+    {
+      "epoch": 0.40415368639667704,
+      "grad_norm": 0.5083807110786438,
+      "learning_rate": 8e-05,
+      "loss": 1.5437,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4043613707165109,
+      "grad_norm": 0.5144785642623901,
+      "learning_rate": 8e-05,
+      "loss": 1.6153,
+      "step": 1947
+    },
+    {
+      "epoch": 0.40456905503634477,
+      "grad_norm": 0.4901933968067169,
+      "learning_rate": 8e-05,
+      "loss": 1.5634,
+      "step": 1948
+    },
+    {
+      "epoch": 0.4047767393561786,
+      "grad_norm": 0.47113847732543945,
+      "learning_rate": 8e-05,
+      "loss": 1.5647,
+      "step": 1949
+    },
+    {
+      "epoch": 0.40498442367601245,
+      "grad_norm": 0.4938381314277649,
+      "learning_rate": 8e-05,
+      "loss": 1.5668,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4051921079958463,
+      "grad_norm": 0.5108892321586609,
+      "learning_rate": 8e-05,
+      "loss": 1.6338,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4053997923156802,
+      "grad_norm": 0.46494653820991516,
+      "learning_rate": 8e-05,
+      "loss": 1.4997,
+      "step": 1952
+    },
+    {
+      "epoch": 0.405607476635514,
+      "grad_norm": 0.48271963000297546,
+      "learning_rate": 8e-05,
+      "loss": 1.5029,
+      "step": 1953
+    },
+    {
+      "epoch": 0.40581516095534786,
+      "grad_norm": 0.5015127658843994,
+      "learning_rate": 8e-05,
+      "loss": 1.5271,
+      "step": 1954
+    },
+    {
+      "epoch": 0.40602284527518173,
+      "grad_norm": 0.5078144669532776,
+      "learning_rate": 8e-05,
+      "loss": 1.5529,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4062305295950156,
+      "grad_norm": 0.5320901870727539,
+      "learning_rate": 8e-05,
+      "loss": 1.6427,
+      "step": 1956
+    },
+    {
+      "epoch": 0.4064382139148494,
+      "grad_norm": 0.501735270023346,
+      "learning_rate": 8e-05,
+      "loss": 1.5977,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4066458982346833,
+      "grad_norm": 0.4956776201725006,
+      "learning_rate": 8e-05,
+      "loss": 1.5942,
+      "step": 1958
+    },
+    {
+      "epoch": 0.40685358255451715,
+      "grad_norm": 0.48016679286956787,
+      "learning_rate": 8e-05,
+      "loss": 1.5697,
+      "step": 1959
+    },
+    {
+      "epoch": 0.407061266874351,
+      "grad_norm": 0.500160813331604,
+      "learning_rate": 8e-05,
+      "loss": 1.5724,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4072689511941848,
+      "grad_norm": 0.5022308230400085,
+      "learning_rate": 8e-05,
+      "loss": 1.5612,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4074766355140187,
+      "grad_norm": 0.5192114114761353,
+      "learning_rate": 8e-05,
+      "loss": 1.6205,
+      "step": 1962
+    },
+    {
+      "epoch": 0.40768431983385256,
+      "grad_norm": 0.470253050327301,
+      "learning_rate": 8e-05,
+      "loss": 1.5219,
+      "step": 1963
+    },
+    {
+      "epoch": 0.40789200415368637,
+      "grad_norm": 0.48591867089271545,
+      "learning_rate": 8e-05,
+      "loss": 1.6203,
+      "step": 1964
+    },
+    {
+      "epoch": 0.40809968847352024,
+      "grad_norm": 0.49583899974823,
+      "learning_rate": 8e-05,
+      "loss": 1.548,
+      "step": 1965
+    },
+    {
+      "epoch": 0.4083073727933541,
+      "grad_norm": 0.5171915292739868,
+      "learning_rate": 8e-05,
+      "loss": 1.6081,
+      "step": 1966
+    },
+    {
+      "epoch": 0.408515057113188,
+      "grad_norm": 0.49792858958244324,
+      "learning_rate": 8e-05,
+      "loss": 1.6061,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4087227414330218,
+      "grad_norm": 0.4974616467952728,
+      "learning_rate": 8e-05,
+      "loss": 1.6322,
+      "step": 1968
+    },
+    {
+      "epoch": 0.40893042575285565,
+      "grad_norm": 0.5105950832366943,
+      "learning_rate": 8e-05,
+      "loss": 1.5612,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4091381100726895,
+      "grad_norm": 0.49406614899635315,
+      "learning_rate": 8e-05,
+      "loss": 1.5887,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4093457943925234,
+      "grad_norm": 0.4994363784790039,
+      "learning_rate": 8e-05,
+      "loss": 1.6257,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4095534787123572,
+      "grad_norm": 0.5137128233909607,
+      "learning_rate": 8e-05,
+      "loss": 1.5861,
+      "step": 1972
+    },
+    {
+      "epoch": 0.40976116303219107,
+      "grad_norm": 0.47833117842674255,
+      "learning_rate": 8e-05,
+      "loss": 1.5661,
+      "step": 1973
+    },
+    {
+      "epoch": 0.40996884735202493,
+      "grad_norm": 0.5011354684829712,
+      "learning_rate": 8e-05,
+      "loss": 1.6508,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4101765316718588,
+      "grad_norm": 0.49250781536102295,
+      "learning_rate": 8e-05,
+      "loss": 1.587,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4103842159916926,
+      "grad_norm": 0.5081549882888794,
+      "learning_rate": 8e-05,
+      "loss": 1.5634,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4105919003115265,
+      "grad_norm": 0.503896176815033,
+      "learning_rate": 8e-05,
+      "loss": 1.5714,
+      "step": 1977
+    },
+    {
+      "epoch": 0.41079958463136035,
+      "grad_norm": 0.5175586938858032,
+      "learning_rate": 8e-05,
+      "loss": 1.5938,
+      "step": 1978
+    },
+    {
+      "epoch": 0.41100726895119416,
+      "grad_norm": 0.4877651631832123,
+      "learning_rate": 8e-05,
+      "loss": 1.5255,
+      "step": 1979
+    },
+    {
+      "epoch": 0.411214953271028,
+      "grad_norm": 0.5077330470085144,
+      "learning_rate": 8e-05,
+      "loss": 1.5783,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4114226375908619,
+      "grad_norm": 0.48941168189048767,
+      "learning_rate": 8e-05,
+      "loss": 1.6246,
+      "step": 1981
+    },
+    {
+      "epoch": 0.41163032191069576,
+      "grad_norm": 0.4887639582157135,
+      "learning_rate": 8e-05,
+      "loss": 1.581,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4118380062305296,
+      "grad_norm": 0.48687008023262024,
+      "learning_rate": 8e-05,
+      "loss": 1.5368,
+      "step": 1983
+    },
+    {
+      "epoch": 0.41204569055036344,
+      "grad_norm": 0.4982350468635559,
+      "learning_rate": 8e-05,
+      "loss": 1.6293,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4122533748701973,
+      "grad_norm": 0.48353442549705505,
+      "learning_rate": 8e-05,
+      "loss": 1.5929,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4124610591900312,
+      "grad_norm": 0.47105276584625244,
+      "learning_rate": 8e-05,
+      "loss": 1.5741,
+      "step": 1986
+    },
+    {
+      "epoch": 0.412668743509865,
+      "grad_norm": 0.49714410305023193,
+      "learning_rate": 8e-05,
+      "loss": 1.5607,
+      "step": 1987
+    },
+    {
+      "epoch": 0.41287642782969886,
+      "grad_norm": 0.4911838173866272,
+      "learning_rate": 8e-05,
+      "loss": 1.5755,
+      "step": 1988
+    },
+    {
+      "epoch": 0.4130841121495327,
+      "grad_norm": 0.47256773710250854,
+      "learning_rate": 8e-05,
+      "loss": 1.4882,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4132917964693666,
+      "grad_norm": 0.5255026817321777,
+      "learning_rate": 8e-05,
+      "loss": 1.586,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4134994807892004,
+      "grad_norm": 0.48072895407676697,
+      "learning_rate": 8e-05,
+      "loss": 1.541,
+      "step": 1991
+    },
+    {
+      "epoch": 0.41370716510903427,
+      "grad_norm": 0.5017218589782715,
+      "learning_rate": 8e-05,
+      "loss": 1.6442,
+      "step": 1992
+    },
+    {
+      "epoch": 0.41391484942886814,
+      "grad_norm": 0.49725833535194397,
+      "learning_rate": 8e-05,
+      "loss": 1.6044,
+      "step": 1993
+    },
+    {
+      "epoch": 0.41412253374870195,
+      "grad_norm": 0.4846404790878296,
+      "learning_rate": 8e-05,
+      "loss": 1.5728,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4143302180685358,
+      "grad_norm": 0.5051259398460388,
+      "learning_rate": 8e-05,
+      "loss": 1.5835,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4145379023883697,
+      "grad_norm": 0.504957377910614,
+      "learning_rate": 8e-05,
+      "loss": 1.5677,
+      "step": 1996
+    },
+    {
+      "epoch": 0.41474558670820355,
+      "grad_norm": 0.5060474276542664,
+      "learning_rate": 8e-05,
+      "loss": 1.5864,
+      "step": 1997
+    },
+    {
+      "epoch": 0.41495327102803736,
+      "grad_norm": 0.5011793375015259,
+      "learning_rate": 8e-05,
+      "loss": 1.6073,
+      "step": 1998
+    },
+    {
+      "epoch": 0.41516095534787123,
+      "grad_norm": 0.5134116411209106,
+      "learning_rate": 8e-05,
+      "loss": 1.59,
+      "step": 1999
+    },
+    {
+      "epoch": 0.4153686396677051,
+      "grad_norm": 0.5138752460479736,
+      "learning_rate": 8e-05,
+      "loss": 1.5954,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4815,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1353608608022528e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}