diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,56138 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3503212552996197,
+  "eval_steps": 500,
+  "global_step": 8015,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 4.3708204029896414e-05,
+      "grad_norm": 23.75,
+      "learning_rate": 1e-05,
+      "loss": 5.1721,
+      "step": 1
+    },
+    {
+      "epoch": 8.741640805979283e-05,
+      "grad_norm": 25.625,
+      "learning_rate": 2e-05,
+      "loss": 5.2813,
+      "step": 2
+    },
+    {
+      "epoch": 0.00013112461208968922,
+      "grad_norm": 26.75,
+      "learning_rate": 3e-05,
+      "loss": 5.1543,
+      "step": 3
+    },
+    {
+      "epoch": 0.00017483281611958566,
+      "grad_norm": 20.0,
+      "learning_rate": 4e-05,
+      "loss": 5.2777,
+      "step": 4
+    },
+    {
+      "epoch": 0.00021854102014948207,
+      "grad_norm": 15.3125,
+      "learning_rate": 5e-05,
+      "loss": 4.738,
+      "step": 5
+    },
+    {
+      "epoch": 0.00026224922417937845,
+      "grad_norm": 15.875,
+      "learning_rate": 6e-05,
+      "loss": 4.0695,
+      "step": 6
+    },
+    {
+      "epoch": 0.0003059574282092749,
+      "grad_norm": 13.625,
+      "learning_rate": 7e-05,
+      "loss": 4.3845,
+      "step": 7
+    },
+    {
+      "epoch": 0.0003496656322391713,
+      "grad_norm": 12.3125,
+      "learning_rate": 8e-05,
+      "loss": 4.2192,
+      "step": 8
+    },
+    {
+      "epoch": 0.0003933738362690677,
+      "grad_norm": 8.5625,
+      "learning_rate": 9e-05,
+      "loss": 3.635,
+      "step": 9
+    },
+    {
+      "epoch": 0.00043708204029896413,
+      "grad_norm": 8.875,
+      "learning_rate": 0.0001,
+      "loss": 3.798,
+      "step": 10
+    },
+    {
+      "epoch": 0.0004807902443288605,
+      "grad_norm": 8.25,
+      "learning_rate": 9.999999952821362e-05,
+      "loss": 3.4472,
+      "step": 11
+    },
+    {
+      "epoch": 0.0005244984483587569,
+      "grad_norm": 9.4375,
+      "learning_rate": 9.999999811285448e-05,
+      "loss": 3.7404,
+      "step": 12
+    },
+    {
+      "epoch": 0.0005682066523886533,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.999999575392258e-05,
+      "loss": 3.5688,
+      "step": 13
+    },
+    {
+      "epoch": 0.0006119148564185498,
+      "grad_norm": 6.46875,
+      "learning_rate": 9.9999992451418e-05,
+      "loss": 3.6849,
+      "step": 14
+    },
+    {
+      "epoch": 0.0006556230604484462,
+      "grad_norm": 7.0,
+      "learning_rate": 9.99999882053408e-05,
+      "loss": 3.6462,
+      "step": 15
+    },
+    {
+      "epoch": 0.0006993312644783426,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.999998301569104e-05,
+      "loss": 3.1855,
+      "step": 16
+    },
+    {
+      "epoch": 0.000743039468508239,
+      "grad_norm": 5.75,
+      "learning_rate": 9.999997688246885e-05,
+      "loss": 3.388,
+      "step": 17
+    },
+    {
+      "epoch": 0.0007867476725381354,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.999996980567431e-05,
+      "loss": 3.2684,
+      "step": 18
+    },
+    {
+      "epoch": 0.0008304558765680318,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.999996178530757e-05,
+      "loss": 2.8991,
+      "step": 19
+    },
+    {
+      "epoch": 0.0008741640805979283,
+      "grad_norm": 6.3125,
+      "learning_rate": 9.999995282136878e-05,
+      "loss": 3.6678,
+      "step": 20
+    },
+    {
+      "epoch": 0.0009178722846278247,
+      "grad_norm": 6.4375,
+      "learning_rate": 9.99999429138581e-05,
+      "loss": 3.3619,
+      "step": 21
+    },
+    {
+      "epoch": 0.000961580488657721,
+      "grad_norm": 7.34375,
+      "learning_rate": 9.999993206277573e-05,
+      "loss": 3.2241,
+      "step": 22
+    },
+    {
+      "epoch": 0.0010052886926876176,
+      "grad_norm": 38.0,
+      "learning_rate": 9.999992026812187e-05,
+      "loss": 3.4659,
+      "step": 23
+    },
+    {
+      "epoch": 0.0010489968967175138,
+      "grad_norm": 7.03125,
+      "learning_rate": 9.999990752989675e-05,
+      "loss": 3.2605,
+      "step": 24
+    },
+    {
+      "epoch": 0.0010927051007474102,
+      "grad_norm": 14.625,
+      "learning_rate": 9.99998938481006e-05,
+      "loss": 3.3462,
+      "step": 25
+    },
+    {
+      "epoch": 0.0011364133047773067,
+      "grad_norm": 6.03125,
+      "learning_rate": 9.999987922273368e-05,
+      "loss": 3.2599,
+      "step": 26
+    },
+    {
+      "epoch": 0.001180121508807203,
+      "grad_norm": 6.4375,
+      "learning_rate": 9.999986365379628e-05,
+      "loss": 3.0645,
+      "step": 27
+    },
+    {
+      "epoch": 0.0012238297128370995,
+      "grad_norm": 5.75,
+      "learning_rate": 9.999984714128867e-05,
+      "loss": 3.1231,
+      "step": 28
+    },
+    {
+      "epoch": 0.001267537916866996,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.999982968521116e-05,
+      "loss": 2.9634,
+      "step": 29
+    },
+    {
+      "epoch": 0.0013112461208968924,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.999981128556411e-05,
+      "loss": 2.902,
+      "step": 30
+    },
+    {
+      "epoch": 0.0013549543249267888,
+      "grad_norm": 21.0,
+      "learning_rate": 9.999979194234786e-05,
+      "loss": 3.1391,
+      "step": 31
+    },
+    {
+      "epoch": 0.0013986625289566853,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.999977165556273e-05,
+      "loss": 2.9417,
+      "step": 32
+    },
+    {
+      "epoch": 0.0014423707329865815,
+      "grad_norm": 6.5,
+      "learning_rate": 9.999975042520917e-05,
+      "loss": 2.8276,
+      "step": 33
+    },
+    {
+      "epoch": 0.001486078937016478,
+      "grad_norm": 6.4375,
+      "learning_rate": 9.999972825128754e-05,
+      "loss": 3.7324,
+      "step": 34
+    },
+    {
+      "epoch": 0.0015297871410463744,
+      "grad_norm": 6.28125,
+      "learning_rate": 9.999970513379826e-05,
+      "loss": 3.2154,
+      "step": 35
+    },
+    {
+      "epoch": 0.0015734953450762708,
+      "grad_norm": 5.8125,
+      "learning_rate": 9.999968107274177e-05,
+      "loss": 3.2734,
+      "step": 36
+    },
+    {
+      "epoch": 0.0016172035491061672,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.999965606811854e-05,
+      "loss": 2.8065,
+      "step": 37
+    },
+    {
+      "epoch": 0.0016609117531360637,
+      "grad_norm": 5.59375,
+      "learning_rate": 9.999963011992902e-05,
+      "loss": 2.8479,
+      "step": 38
+    },
+    {
+      "epoch": 0.00170461995716596,
+      "grad_norm": 12.875,
+      "learning_rate": 9.99996032281737e-05,
+      "loss": 3.0187,
+      "step": 39
+    },
+    {
+      "epoch": 0.0017483281611958565,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.999957539285312e-05,
+      "loss": 3.0339,
+      "step": 40
+    },
+    {
+      "epoch": 0.001792036365225753,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.999954661396777e-05,
+      "loss": 3.1146,
+      "step": 41
+    },
+    {
+      "epoch": 0.0018357445692556494,
+      "grad_norm": 8.875,
+      "learning_rate": 9.99995168915182e-05,
+      "loss": 2.7607,
+      "step": 42
+    },
+    {
+      "epoch": 0.0018794527732855456,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.999948622550497e-05,
+      "loss": 3.1231,
+      "step": 43
+    },
+    {
+      "epoch": 0.001923160977315442,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.999945461592867e-05,
+      "loss": 3.0816,
+      "step": 44
+    },
+    {
+      "epoch": 0.0019668691813453385,
+      "grad_norm": 5.125,
+      "learning_rate": 9.99994220627899e-05,
+      "loss": 3.6489,
+      "step": 45
+    },
+    {
+      "epoch": 0.002010577385375235,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.999938856608926e-05,
+      "loss": 3.0423,
+      "step": 46
+    },
+    {
+      "epoch": 0.0020542855894051314,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.999935412582738e-05,
+      "loss": 3.1826,
+      "step": 47
+    },
+    {
+      "epoch": 0.0020979937934350276,
+      "grad_norm": 5.0,
+      "learning_rate": 9.999931874200492e-05,
+      "loss": 3.0762,
+      "step": 48
+    },
+    {
+      "epoch": 0.0021417019974649242,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.999928241462255e-05,
+      "loss": 3.4191,
+      "step": 49
+    },
+    {
+      "epoch": 0.0021854102014948204,
+      "grad_norm": 6.90625,
+      "learning_rate": 9.999924514368095e-05,
+      "loss": 3.2524,
+      "step": 50
+    },
+    {
+      "epoch": 0.002229118405524717,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.999920692918082e-05,
+      "loss": 2.9654,
+      "step": 51
+    },
+    {
+      "epoch": 0.0022728266095546133,
+      "grad_norm": 5.75,
+      "learning_rate": 9.999916777112288e-05,
+      "loss": 3.1085,
+      "step": 52
+    },
+    {
+      "epoch": 0.00231653481358451,
+      "grad_norm": 4.75,
+      "learning_rate": 9.999912766950789e-05,
+      "loss": 3.0036,
+      "step": 53
+    },
+    {
+      "epoch": 0.002360243017614406,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.999908662433657e-05,
+      "loss": 3.0041,
+      "step": 54
+    },
+    {
+      "epoch": 0.002403951221644303,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.999904463560975e-05,
+      "loss": 2.6111,
+      "step": 55
+    },
+    {
+      "epoch": 0.002447659425674199,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.999900170332814e-05,
+      "loss": 2.7588,
+      "step": 56
+    },
+    {
+      "epoch": 0.0024913676297040953,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.999895782749263e-05,
+      "loss": 3.0982,
+      "step": 57
+    },
+    {
+      "epoch": 0.002535075833733992,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.9998913008104e-05,
+      "loss": 3.2029,
+      "step": 58
+    },
+    {
+      "epoch": 0.002578784037763888,
+      "grad_norm": 6.125,
+      "learning_rate": 9.999886724516312e-05,
+      "loss": 2.601,
+      "step": 59
+    },
+    {
+      "epoch": 0.002622492241793785,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.999882053867085e-05,
+      "loss": 2.5674,
+      "step": 60
+    },
+    {
+      "epoch": 0.002666200445823681,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.999877288862806e-05,
+      "loss": 2.8234,
+      "step": 61
+    },
+    {
+      "epoch": 0.0027099086498535777,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.999872429503565e-05,
+      "loss": 2.9025,
+      "step": 62
+    },
+    {
+      "epoch": 0.002753616853883474,
+      "grad_norm": 4.625,
+      "learning_rate": 9.999867475789455e-05,
+      "loss": 2.7504,
+      "step": 63
+    },
+    {
+      "epoch": 0.0027973250579133705,
+      "grad_norm": 10.125,
+      "learning_rate": 9.999862427720568e-05,
+      "loss": 3.6081,
+      "step": 64
+    },
+    {
+      "epoch": 0.0028410332619432667,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.999857285297e-05,
+      "loss": 2.7429,
+      "step": 65
+    },
+    {
+      "epoch": 0.002884741465973163,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.999852048518849e-05,
+      "loss": 2.8513,
+      "step": 66
+    },
+    {
+      "epoch": 0.0029284496700030596,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.999846717386214e-05,
+      "loss": 2.7622,
+      "step": 67
+    },
+    {
+      "epoch": 0.002972157874032956,
+      "grad_norm": 6.90625,
+      "learning_rate": 9.999841291899193e-05,
+      "loss": 3.102,
+      "step": 68
+    },
+    {
+      "epoch": 0.0030158660780628525,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.999835772057891e-05,
+      "loss": 2.6154,
+      "step": 69
+    },
+    {
+      "epoch": 0.0030595742820927487,
+      "grad_norm": 5.9375,
+      "learning_rate": 9.99983015786241e-05,
+      "loss": 2.9899,
+      "step": 70
+    },
+    {
+      "epoch": 0.0031032824861226454,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.999824449312856e-05,
+      "loss": 2.8435,
+      "step": 71
+    },
+    {
+      "epoch": 0.0031469906901525416,
+      "grad_norm": 13.375,
+      "learning_rate": 9.999818646409339e-05,
+      "loss": 4.1482,
+      "step": 72
+    },
+    {
+      "epoch": 0.0031906988941824382,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.999812749151966e-05,
+      "loss": 2.5349,
+      "step": 73
+    },
+    {
+      "epoch": 0.0032344070982123344,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.999806757540851e-05,
+      "loss": 2.5921,
+      "step": 74
+    },
+    {
+      "epoch": 0.0032781153022422307,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.999800671576106e-05,
+      "loss": 2.6727,
+      "step": 75
+    },
+    {
+      "epoch": 0.0033218235062721273,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.999794491257845e-05,
+      "loss": 2.7172,
+      "step": 76
+    },
+    {
+      "epoch": 0.0033655317103020235,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.999788216586186e-05,
+      "loss": 3.1049,
+      "step": 77
+    },
+    {
+      "epoch": 0.00340923991433192,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.999781847561245e-05,
+      "loss": 2.8463,
+      "step": 78
+    },
+    {
+      "epoch": 0.0034529481183618164,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.999775384183143e-05,
+      "loss": 2.6217,
+      "step": 79
+    },
+    {
+      "epoch": 0.003496656322391713,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.999768826452004e-05,
+      "loss": 2.9236,
+      "step": 80
+    },
+    {
+      "epoch": 0.0035403645264216093,
+      "grad_norm": 3.875,
+      "learning_rate": 9.99976217436795e-05,
+      "loss": 2.4763,
+      "step": 81
+    },
+    {
+      "epoch": 0.003584072730451506,
+      "grad_norm": 6.46875,
+      "learning_rate": 9.999755427931107e-05,
+      "loss": 2.5611,
+      "step": 82
+    },
+    {
+      "epoch": 0.003627780934481402,
+      "grad_norm": 6.84375,
+      "learning_rate": 9.999748587141602e-05,
+      "loss": 2.4349,
+      "step": 83
+    },
+    {
+      "epoch": 0.003671489138511299,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.999741651999566e-05,
+      "loss": 2.5408,
+      "step": 84
+    },
+    {
+      "epoch": 0.003715197342541195,
+      "grad_norm": 4.875,
+      "learning_rate": 9.999734622505126e-05,
+      "loss": 2.9666,
+      "step": 85
+    },
+    {
+      "epoch": 0.0037589055465710912,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.999727498658417e-05,
+      "loss": 2.679,
+      "step": 86
+    },
+    {
+      "epoch": 0.003802613750600988,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.999720280459576e-05,
+      "loss": 2.6728,
+      "step": 87
+    },
+    {
+      "epoch": 0.003846321954630884,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.999712967908735e-05,
+      "loss": 3.1564,
+      "step": 88
+    },
+    {
+      "epoch": 0.0038900301586607807,
+      "grad_norm": 6.40625,
+      "learning_rate": 9.999705561006034e-05,
+      "loss": 2.4467,
+      "step": 89
+    },
+    {
+      "epoch": 0.003933738362690677,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.999698059751609e-05,
+      "loss": 2.8369,
+      "step": 90
+    },
+    {
+      "epoch": 0.003977446566720573,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.999690464145609e-05,
+      "loss": 2.6195,
+      "step": 91
+    },
+    {
+      "epoch": 0.00402115477075047,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.999682774188173e-05,
+      "loss": 2.7844,
+      "step": 92
+    },
+    {
+      "epoch": 0.0040648629747803665,
+      "grad_norm": 4.375,
+      "learning_rate": 9.999674989879444e-05,
+      "loss": 2.6323,
+      "step": 93
+    },
+    {
+      "epoch": 0.004108571178810263,
+      "grad_norm": 4.125,
+      "learning_rate": 9.999667111219573e-05,
+      "loss": 2.692,
+      "step": 94
+    },
+    {
+      "epoch": 0.004152279382840159,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.999659138208705e-05,
+      "loss": 2.4742,
+      "step": 95
+    },
+    {
+      "epoch": 0.004195987586870055,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.999651070846995e-05,
+      "loss": 2.5278,
+      "step": 96
+    },
+    {
+      "epoch": 0.004239695790899952,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.999642909134592e-05,
+      "loss": 2.8639,
+      "step": 97
+    },
+    {
+      "epoch": 0.0042834039949298484,
+      "grad_norm": 5.75,
+      "learning_rate": 9.99963465307165e-05,
+      "loss": 2.6517,
+      "step": 98
+    },
+    {
+      "epoch": 0.004327112198959745,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.999626302658324e-05,
+      "loss": 2.7603,
+      "step": 99
+    },
+    {
+      "epoch": 0.004370820402989641,
+      "grad_norm": 4.375,
+      "learning_rate": 9.999617857894777e-05,
+      "loss": 2.3816,
+      "step": 100
+    },
+    {
+      "epoch": 0.004414528607019538,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.99960931878116e-05,
+      "loss": 2.6044,
+      "step": 101
+    },
+    {
+      "epoch": 0.004458236811049434,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.999600685317642e-05,
+      "loss": 2.3602,
+      "step": 102
+    },
+    {
+      "epoch": 0.00450194501507933,
+      "grad_norm": 9.0,
+      "learning_rate": 9.99959195750438e-05,
+      "loss": 2.4109,
+      "step": 103
+    },
+    {
+      "epoch": 0.004545653219109227,
+      "grad_norm": 6.0625,
+      "learning_rate": 9.999583135341544e-05,
+      "loss": 2.9449,
+      "step": 104
+    },
+    {
+      "epoch": 0.004589361423139123,
+      "grad_norm": 5.96875,
+      "learning_rate": 9.999574218829295e-05,
+      "loss": 2.4216,
+      "step": 105
+    },
+    {
+      "epoch": 0.00463306962716902,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.999565207967805e-05,
+      "loss": 2.8841,
+      "step": 106
+    },
+    {
+      "epoch": 0.004676777831198916,
+      "grad_norm": 3.875,
+      "learning_rate": 9.999556102757244e-05,
+      "loss": 2.6692,
+      "step": 107
+    },
+    {
+      "epoch": 0.004720486035228812,
+      "grad_norm": 5.90625,
+      "learning_rate": 9.99954690319778e-05,
+      "loss": 2.6152,
+      "step": 108
+    },
+    {
+      "epoch": 0.004764194239258709,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.999537609289592e-05,
+      "loss": 2.6917,
+      "step": 109
+    },
+    {
+      "epoch": 0.004807902443288606,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.999528221032852e-05,
+      "loss": 2.2566,
+      "step": 110
+    },
+    {
+      "epoch": 0.004851610647318502,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.999518738427737e-05,
+      "loss": 2.2956,
+      "step": 111
+    },
+    {
+      "epoch": 0.004895318851348398,
+      "grad_norm": 6.625,
+      "learning_rate": 9.99950916147443e-05,
+      "loss": 2.4404,
+      "step": 112
+    },
+    {
+      "epoch": 0.004939027055378294,
+      "grad_norm": 5.0,
+      "learning_rate": 9.999499490173106e-05,
+      "loss": 2.5939,
+      "step": 113
+    },
+    {
+      "epoch": 0.0049827352594081905,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.999489724523951e-05,
+      "loss": 3.225,
+      "step": 114
+    },
+    {
+      "epoch": 0.005026443463438088,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.999479864527148e-05,
+      "loss": 2.8029,
+      "step": 115
+    },
+    {
+      "epoch": 0.005070151667467984,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.999469910182885e-05,
+      "loss": 2.5389,
+      "step": 116
+    },
+    {
+      "epoch": 0.00511385987149788,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.999459861491348e-05,
+      "loss": 2.9194,
+      "step": 117
+    },
+    {
+      "epoch": 0.005157568075527776,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.999449718452725e-05,
+      "loss": 2.623,
+      "step": 118
+    },
+    {
+      "epoch": 0.005201276279557673,
+      "grad_norm": 6.46875,
+      "learning_rate": 9.999439481067212e-05,
+      "loss": 3.8513,
+      "step": 119
+    },
+    {
+      "epoch": 0.00524498448358757,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.999429149334998e-05,
+      "loss": 2.4835,
+      "step": 120
+    },
+    {
+      "epoch": 0.005288692687617466,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.999418723256279e-05,
+      "loss": 2.4446,
+      "step": 121
+    },
+    {
+      "epoch": 0.005332400891647362,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.999408202831255e-05,
+      "loss": 3.1651,
+      "step": 122
+    },
+    {
+      "epoch": 0.005376109095677258,
+      "grad_norm": 3.625,
+      "learning_rate": 9.99939758806012e-05,
+      "loss": 2.908,
+      "step": 123
+    },
+    {
+      "epoch": 0.005419817299707155,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.999386878943077e-05,
+      "loss": 2.483,
+      "step": 124
+    },
+    {
+      "epoch": 0.0054635255037370515,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.999376075480327e-05,
+      "loss": 2.8223,
+      "step": 125
+    },
+    {
+      "epoch": 0.005507233707766948,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.999365177672075e-05,
+      "loss": 2.6433,
+      "step": 126
+    },
+    {
+      "epoch": 0.005550941911796844,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.999354185518525e-05,
+      "loss": 3.3299,
+      "step": 127
+    },
+    {
+      "epoch": 0.005594650115826741,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.999343099019884e-05,
+      "loss": 2.6354,
+      "step": 128
+    },
+    {
+      "epoch": 0.005638358319856637,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.999331918176365e-05,
+      "loss": 2.099,
+      "step": 129
+    },
+    {
+      "epoch": 0.0056820665238865335,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.999320642988175e-05,
+      "loss": 2.7426,
+      "step": 130
+    },
+    {
+      "epoch": 0.00572577472791643,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.999309273455528e-05,
+      "loss": 3.2764,
+      "step": 131
+    },
+    {
+      "epoch": 0.005769482931946326,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.99929780957864e-05,
+      "loss": 2.2569,
+      "step": 132
+    },
+    {
+      "epoch": 0.005813191135976223,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.999286251357727e-05,
+      "loss": 2.6164,
+      "step": 133
+    },
+    {
+      "epoch": 0.005856899340006119,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.999274598793005e-05,
+      "loss": 2.2387,
+      "step": 134
+    },
+    {
+      "epoch": 0.0059006075440360154,
+      "grad_norm": 14.125,
+      "learning_rate": 9.999262851884695e-05,
+      "loss": 2.7666,
+      "step": 135
+    },
+    {
+      "epoch": 0.005944315748065912,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.99925101063302e-05,
+      "loss": 2.6582,
+      "step": 136
+    },
+    {
+      "epoch": 0.005988023952095809,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.9992390750382e-05,
+      "loss": 2.0556,
+      "step": 137
+    },
+    {
+      "epoch": 0.006031732156125705,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.999227045100465e-05,
+      "loss": 2.7892,
+      "step": 138
+    },
+    {
+      "epoch": 0.006075440360155601,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.999214920820039e-05,
+      "loss": 2.4652,
+      "step": 139
+    },
+    {
+      "epoch": 0.006119148564185497,
+      "grad_norm": 5.25,
+      "learning_rate": 9.999202702197151e-05,
+      "loss": 3.0553,
+      "step": 140
+    },
+    {
+      "epoch": 0.006162856768215394,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.999190389232032e-05,
+      "loss": 2.8345,
+      "step": 141
+    },
+    {
+      "epoch": 0.006206564972245291,
+      "grad_norm": 9.8125,
+      "learning_rate": 9.999177981924916e-05,
+      "loss": 2.97,
+      "step": 142
+    },
+    {
+      "epoch": 0.006250273176275187,
+      "grad_norm": 6.0625,
+      "learning_rate": 9.999165480276034e-05,
+      "loss": 2.718,
+      "step": 143
+    },
+    {
+      "epoch": 0.006293981380305083,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.999152884285622e-05,
+      "loss": 2.0698,
+      "step": 144
+    },
+    {
+      "epoch": 0.006337689584334979,
+      "grad_norm": 4.875,
+      "learning_rate": 9.999140193953921e-05,
+      "loss": 2.4787,
+      "step": 145
+    },
+    {
+      "epoch": 0.0063813977883648764,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.99912740928117e-05,
+      "loss": 2.3143,
+      "step": 146
+    },
+    {
+      "epoch": 0.006425105992394773,
+      "grad_norm": 10.4375,
+      "learning_rate": 9.999114530267607e-05,
+      "loss": 2.3755,
+      "step": 147
+    },
+    {
+      "epoch": 0.006468814196424669,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.999101556913477e-05,
+      "loss": 2.5425,
+      "step": 148
+    },
+    {
+      "epoch": 0.006512522400454565,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.999088489219027e-05,
+      "loss": 2.6765,
+      "step": 149
+    },
+    {
+      "epoch": 0.006556230604484461,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.999075327184499e-05,
+      "loss": 2.6608,
+      "step": 150
+    },
+    {
+      "epoch": 0.006599938808514358,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.999062070810144e-05,
+      "loss": 2.8477,
+      "step": 151
+    },
+    {
+      "epoch": 0.006643647012544255,
+      "grad_norm": 5.375,
+      "learning_rate": 9.999048720096212e-05,
+      "loss": 2.8137,
+      "step": 152
+    },
+    {
+      "epoch": 0.006687355216574151,
+      "grad_norm": 4.375,
+      "learning_rate": 9.999035275042954e-05,
+      "loss": 2.6685,
+      "step": 153
+    },
+    {
+      "epoch": 0.006731063420604047,
+      "grad_norm": 7.09375,
+      "learning_rate": 9.999021735650627e-05,
+      "loss": 3.1262,
+      "step": 154
+    },
+    {
+      "epoch": 0.006774771624633944,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.999008101919482e-05,
+      "loss": 2.5198,
+      "step": 155
+    },
+    {
+      "epoch": 0.00681847982866384,
+      "grad_norm": 6.625,
+      "learning_rate": 9.99899437384978e-05,
+      "loss": 3.0499,
+      "step": 156
+    },
+    {
+      "epoch": 0.006862188032693737,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.998980551441776e-05,
+      "loss": 2.6218,
+      "step": 157
+    },
+    {
+      "epoch": 0.006905896236723633,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.998966634695737e-05,
+      "loss": 2.47,
+      "step": 158
+    },
+    {
+      "epoch": 0.006949604440753529,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.99895262361192e-05,
+      "loss": 2.1887,
+      "step": 159
+    },
+    {
+      "epoch": 0.006993312644783426,
+      "grad_norm": 4.25,
+      "learning_rate": 9.998938518190591e-05,
+      "loss": 3.6236,
+      "step": 160
+    },
+    {
+      "epoch": 0.007037020848813322,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.998924318432016e-05,
+      "loss": 2.6418,
+      "step": 161
+    },
+    {
+      "epoch": 0.0070807290528432185,
+      "grad_norm": 4.0,
+      "learning_rate": 9.998910024336466e-05,
+      "loss": 2.4386,
+      "step": 162
+    },
+    {
+      "epoch": 0.007124437256873115,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.998895635904205e-05,
+      "loss": 2.7331,
+      "step": 163
+    },
+    {
+      "epoch": 0.007168145460903012,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.99888115313551e-05,
+      "loss": 2.7163,
+      "step": 164
+    },
+    {
+      "epoch": 0.007211853664932908,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.998866576030651e-05,
+      "loss": 2.123,
+      "step": 165
+    },
+    {
+      "epoch": 0.007255561868962804,
+      "grad_norm": 5.375,
+      "learning_rate": 9.998851904589905e-05,
+      "loss": 2.7204,
+      "step": 166
+    },
+    {
+      "epoch": 0.0072992700729927005,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.998837138813549e-05,
+      "loss": 2.6734,
+      "step": 167
+    },
+    {
+      "epoch": 0.007342978277022598,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.998822278701858e-05,
+      "loss": 2.7616,
+      "step": 168
+    },
+    {
+      "epoch": 0.007386686481052494,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.998807324255118e-05,
+      "loss": 2.556,
+      "step": 169
+    },
+    {
+      "epoch": 0.00743039468508239,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.998792275473607e-05,
+      "loss": 2.5006,
+      "step": 170
+    },
+    {
+      "epoch": 0.007474102889112286,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.99877713235761e-05,
+      "loss": 2.8722,
+      "step": 171
+    },
+    {
+      "epoch": 0.0075178110931421824,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.998761894907414e-05,
+      "loss": 3.0658,
+      "step": 172
+    },
+    {
+      "epoch": 0.0075615192971720795,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.998746563123305e-05,
+      "loss": 2.8287,
+      "step": 173
+    },
+    {
+      "epoch": 0.007605227501201976,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.998731137005572e-05,
+      "loss": 2.5613,
+      "step": 174
+    },
+    {
+      "epoch": 0.007648935705231872,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.99871561655451e-05,
+      "loss": 2.4192,
+      "step": 175
+    },
+    {
+      "epoch": 0.007692643909261768,
+      "grad_norm": 9.875,
+      "learning_rate": 9.998700001770406e-05,
+      "loss": 3.3212,
+      "step": 176
+    },
+    {
+      "epoch": 0.007736352113291665,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.998684292653559e-05,
+      "loss": 2.8366,
+      "step": 177
+    },
+    {
+      "epoch": 0.0077800603173215615,
+      "grad_norm": 4.75,
+      "learning_rate": 9.998668489204266e-05,
+      "loss": 2.4589,
+      "step": 178
+    },
+    {
+      "epoch": 0.007823768521351457,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.998652591422821e-05,
+      "loss": 2.6667,
+      "step": 179
+    },
+    {
+      "epoch": 0.007867476725381354,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.998636599309527e-05,
+      "loss": 2.7617,
+      "step": 180
+    },
+    {
+      "epoch": 0.007911184929411251,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.998620512864686e-05,
+      "loss": 2.5969,
+      "step": 181
+    },
+    {
+      "epoch": 0.007954893133441146,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.9986043320886e-05,
+      "loss": 2.214,
+      "step": 182
+    },
+    {
+      "epoch": 0.007998601337471043,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.998588056981575e-05,
+      "loss": 2.6208,
+      "step": 183
+    },
+    {
+      "epoch": 0.00804230954150094,
+      "grad_norm": 4.375,
+      "learning_rate": 9.998571687543918e-05,
+      "loss": 2.8798,
+      "step": 184
+    },
+    {
+      "epoch": 0.008086017745530836,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.99855522377594e-05,
+      "loss": 2.4923,
+      "step": 185
+    },
+    {
+      "epoch": 0.008129725949560733,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.998538665677948e-05,
+      "loss": 2.055,
+      "step": 186
+    },
+    {
+      "epoch": 0.008173434153590628,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.998522013250257e-05,
+      "loss": 2.6016,
+      "step": 187
+    },
+    {
+      "epoch": 0.008217142357620525,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.998505266493181e-05,
+      "loss": 2.4205,
+      "step": 188
+    },
+    {
+      "epoch": 0.008260850561650422,
+      "grad_norm": 5.25,
+      "learning_rate": 9.998488425407035e-05,
+      "loss": 3.2423,
+      "step": 189
+    },
+    {
+      "epoch": 0.008304558765680318,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.998471489992138e-05,
+      "loss": 2.3305,
+      "step": 190
+    },
+    {
+      "epoch": 0.008348266969710215,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.998454460248809e-05,
+      "loss": 2.736,
+      "step": 191
+    },
+    {
+      "epoch": 0.00839197517374011,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.998437336177369e-05,
+      "loss": 2.8134,
+      "step": 192
+    },
+    {
+      "epoch": 0.008435683377770007,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.998420117778141e-05,
+      "loss": 2.3994,
+      "step": 193
+    },
+    {
+      "epoch": 0.008479391581799904,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.998402805051452e-05,
+      "loss": 2.5559,
+      "step": 194
+    },
+    {
+      "epoch": 0.0085230997858298,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.998385397997625e-05,
+      "loss": 2.3871,
+      "step": 195
+    },
+    {
+      "epoch": 0.008566807989859697,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.998367896616993e-05,
+      "loss": 2.1715,
+      "step": 196
+    },
+    {
+      "epoch": 0.008610516193889592,
+      "grad_norm": 6.15625,
+      "learning_rate": 9.998350300909883e-05,
+      "loss": 2.7056,
+      "step": 197
+    },
+    {
+      "epoch": 0.00865422439791949,
+      "grad_norm": 10.3125,
+      "learning_rate": 9.998332610876629e-05,
+      "loss": 2.2428,
+      "step": 198
+    },
+    {
+      "epoch": 0.008697932601949386,
+      "grad_norm": 6.40625,
+      "learning_rate": 9.998314826517563e-05,
+      "loss": 2.7718,
+      "step": 199
+    },
+    {
+      "epoch": 0.008741640805979282,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.998296947833021e-05,
+      "loss": 3.1103,
+      "step": 200
+    },
+    {
+      "epoch": 0.008785349010009179,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.998278974823342e-05,
+      "loss": 2.9857,
+      "step": 201
+    },
+    {
+      "epoch": 0.008829057214039076,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.998260907488863e-05,
+      "loss": 2.2535,
+      "step": 202
+    },
+    {
+      "epoch": 0.008872765418068971,
+      "grad_norm": 3.625,
+      "learning_rate": 9.998242745829927e-05,
+      "loss": 2.3132,
+      "step": 203
+    },
+    {
+      "epoch": 0.008916473622098868,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.998224489846877e-05,
+      "loss": 2.4809,
+      "step": 204
+    },
+    {
+      "epoch": 0.008960181826128764,
+      "grad_norm": 5.125,
+      "learning_rate": 9.998206139540054e-05,
+      "loss": 3.1461,
+      "step": 205
+    },
+    {
+      "epoch": 0.00900389003015866,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.998187694909807e-05,
+      "loss": 2.5521,
+      "step": 206
+    },
+    {
+      "epoch": 0.009047598234188558,
+      "grad_norm": 5.6875,
+      "learning_rate": 9.998169155956485e-05,
+      "loss": 2.6275,
+      "step": 207
+    },
+    {
+      "epoch": 0.009091306438218453,
+      "grad_norm": 5.09375,
+      "learning_rate": 9.998150522680437e-05,
+      "loss": 2.3942,
+      "step": 208
+    },
+    {
+      "epoch": 0.00913501464224835,
+      "grad_norm": 19.0,
+      "learning_rate": 9.998131795082011e-05,
+      "loss": 2.7506,
+      "step": 209
+    },
+    {
+      "epoch": 0.009178722846278246,
+      "grad_norm": 4.125,
+      "learning_rate": 9.998112973161566e-05,
+      "loss": 2.8633,
+      "step": 210
+    },
+    {
+      "epoch": 0.009222431050308143,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.998094056919454e-05,
+      "loss": 2.8713,
+      "step": 211
+    },
+    {
+      "epoch": 0.00926613925433804,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.998075046356033e-05,
+      "loss": 2.5957,
+      "step": 212
+    },
+    {
+      "epoch": 0.009309847458367935,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.998055941471662e-05,
+      "loss": 2.4896,
+      "step": 213
+    },
+    {
+      "epoch": 0.009353555662397832,
+      "grad_norm": 4.5,
+      "learning_rate": 9.998036742266701e-05,
+      "loss": 3.1921,
+      "step": 214
+    },
+    {
+      "epoch": 0.009397263866427728,
+      "grad_norm": 4.375,
+      "learning_rate": 9.998017448741512e-05,
+      "loss": 2.8811,
+      "step": 215
+    },
+    {
+      "epoch": 0.009440972070457625,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.997998060896458e-05,
+      "loss": 2.6873,
+      "step": 216
+    },
+    {
+      "epoch": 0.009484680274487522,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.997978578731908e-05,
+      "loss": 2.5835,
+      "step": 217
+    },
+    {
+      "epoch": 0.009528388478517417,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.997959002248229e-05,
+      "loss": 2.7929,
+      "step": 218
+    },
+    {
+      "epoch": 0.009572096682547314,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.997939331445787e-05,
+      "loss": 2.1056,
+      "step": 219
+    },
+    {
+      "epoch": 0.009615804886577211,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.997919566324959e-05,
+      "loss": 3.3739,
+      "step": 220
+    },
+    {
+      "epoch": 0.009659513090607107,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.99789970688611e-05,
+      "loss": 2.6079,
+      "step": 221
+    },
+    {
+      "epoch": 0.009703221294637004,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.997879753129624e-05,
+      "loss": 2.4539,
+      "step": 222
+    },
+    {
+      "epoch": 0.009746929498666899,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.99785970505587e-05,
+      "loss": 2.2522,
+      "step": 223
+    },
+    {
+      "epoch": 0.009790637702696796,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.99783956266523e-05,
+      "loss": 2.6285,
+      "step": 224
+    },
+    {
+      "epoch": 0.009834345906726693,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.997819325958085e-05,
+      "loss": 2.7673,
+      "step": 225
+    },
+    {
+      "epoch": 0.009878054110756589,
+      "grad_norm": 7.5,
+      "learning_rate": 9.997798994934813e-05,
+      "loss": 2.7084,
+      "step": 226
+    },
+    {
+      "epoch": 0.009921762314786486,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.997778569595801e-05,
+      "loss": 2.9379,
+      "step": 227
+    },
+    {
+      "epoch": 0.009965470518816381,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.997758049941435e-05,
+      "loss": 2.7648,
+      "step": 228
+    },
+    {
+      "epoch": 0.010009178722846278,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.997737435972099e-05,
+      "loss": 2.381,
+      "step": 229
+    },
+    {
+      "epoch": 0.010052886926876175,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.997716727688183e-05,
+      "loss": 2.2979,
+      "step": 230
+    },
+    {
+      "epoch": 0.01009659513090607,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.99769592509008e-05,
+      "loss": 2.7071,
+      "step": 231
+    },
+    {
+      "epoch": 0.010140303334935968,
+      "grad_norm": 4.875,
+      "learning_rate": 9.99767502817818e-05,
+      "loss": 3.5704,
+      "step": 232
+    },
+    {
+      "epoch": 0.010184011538965865,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.997654036952879e-05,
+      "loss": 1.8784,
+      "step": 233
+    },
+    {
+      "epoch": 0.01022771974299576,
+      "grad_norm": 4.5,
+      "learning_rate": 9.997632951414573e-05,
+      "loss": 2.5868,
+      "step": 234
+    },
+    {
+      "epoch": 0.010271427947025657,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.99761177156366e-05,
+      "loss": 2.1905,
+      "step": 235
+    },
+    {
+      "epoch": 0.010315136151055553,
+      "grad_norm": 4.25,
+      "learning_rate": 9.997590497400538e-05,
+      "loss": 3.011,
+      "step": 236
+    },
+    {
+      "epoch": 0.01035884435508545,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.997569128925611e-05,
+      "loss": 2.5319,
+      "step": 237
+    },
+    {
+      "epoch": 0.010402552559115347,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.997547666139279e-05,
+      "loss": 2.0043,
+      "step": 238
+    },
+    {
+      "epoch": 0.010446260763145242,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.99752610904195e-05,
+      "loss": 2.708,
+      "step": 239
+    },
+    {
+      "epoch": 0.01048996896717514,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.997504457634029e-05,
+      "loss": 2.214,
+      "step": 240
+    },
+    {
+      "epoch": 0.010533677171205034,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.997482711915927e-05,
+      "loss": 2.275,
+      "step": 241
+    },
+    {
+      "epoch": 0.010577385375234932,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.997460871888052e-05,
+      "loss": 2.7156,
+      "step": 242
+    },
+    {
+      "epoch": 0.010621093579264829,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.997438937550816e-05,
+      "loss": 3.3969,
+      "step": 243
+    },
+    {
+      "epoch": 0.010664801783294724,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.997416908904633e-05,
+      "loss": 2.7685,
+      "step": 244
+    },
+    {
+      "epoch": 0.010708509987324621,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.997394785949922e-05,
+      "loss": 2.6075,
+      "step": 245
+    },
+    {
+      "epoch": 0.010752218191354516,
+      "grad_norm": 3.625,
+      "learning_rate": 9.997372568687097e-05,
+      "loss": 2.2984,
+      "step": 246
+    },
+    {
+      "epoch": 0.010795926395384414,
+      "grad_norm": 4.375,
+      "learning_rate": 9.997350257116578e-05,
+      "loss": 2.589,
+      "step": 247
+    },
+    {
+      "epoch": 0.01083963459941431,
+      "grad_norm": 3.75,
+      "learning_rate": 9.997327851238788e-05,
+      "loss": 2.0634,
+      "step": 248
+    },
+    {
+      "epoch": 0.010883342803444206,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.997305351054146e-05,
+      "loss": 2.4742,
+      "step": 249
+    },
+    {
+      "epoch": 0.010927051007474103,
+      "grad_norm": 4.0,
+      "learning_rate": 9.99728275656308e-05,
+      "loss": 2.9405,
+      "step": 250
+    },
+    {
+      "epoch": 0.010970759211504,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.997260067766014e-05,
+      "loss": 2.304,
+      "step": 251
+    },
+    {
+      "epoch": 0.011014467415533895,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.997237284663379e-05,
+      "loss": 2.9572,
+      "step": 252
+    },
+    {
+      "epoch": 0.011058175619563793,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.997214407255602e-05,
+      "loss": 2.3134,
+      "step": 253
+    },
+    {
+      "epoch": 0.011101883823593688,
+      "grad_norm": 5.125,
+      "learning_rate": 9.997191435543117e-05,
+      "loss": 3.2421,
+      "step": 254
+    },
+    {
+      "epoch": 0.011145592027623585,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.997168369526355e-05,
+      "loss": 2.6112,
+      "step": 255
+    },
+    {
+      "epoch": 0.011189300231653482,
+      "grad_norm": 5.96875,
+      "learning_rate": 9.997145209205754e-05,
+      "loss": 2.4515,
+      "step": 256
+    },
+    {
+      "epoch": 0.011233008435683377,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.99712195458175e-05,
+      "loss": 2.8503,
+      "step": 257
+    },
+    {
+      "epoch": 0.011276716639713275,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.997098605654782e-05,
+      "loss": 2.4916,
+      "step": 258
+    },
+    {
+      "epoch": 0.01132042484374317,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.99707516242529e-05,
+      "loss": 2.2976,
+      "step": 259
+    },
+    {
+      "epoch": 0.011364133047773067,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.997051624893716e-05,
+      "loss": 2.5726,
+      "step": 260
+    },
+    {
+      "epoch": 0.011407841251802964,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.997027993060506e-05,
+      "loss": 2.0185,
+      "step": 261
+    },
+    {
+      "epoch": 0.01145154945583286,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.997004266926105e-05,
+      "loss": 3.0021,
+      "step": 262
+    },
+    {
+      "epoch": 0.011495257659862756,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.99698044649096e-05,
+      "loss": 2.7337,
+      "step": 263
+    },
+    {
+      "epoch": 0.011538965863892652,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.996956531755521e-05,
+      "loss": 2.3048,
+      "step": 264
+    },
+    {
+      "epoch": 0.011582674067922549,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.996932522720242e-05,
+      "loss": 3.3072,
+      "step": 265
+    },
+    {
+      "epoch": 0.011626382271952446,
+      "grad_norm": 8.625,
+      "learning_rate": 9.996908419385571e-05,
+      "loss": 3.0222,
+      "step": 266
+    },
+    {
+      "epoch": 0.011670090475982341,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.996884221751966e-05,
+      "loss": 2.5027,
+      "step": 267
+    },
+    {
+      "epoch": 0.011713798680012238,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.996859929819882e-05,
+      "loss": 2.2139,
+      "step": 268
+    },
+    {
+      "epoch": 0.011757506884042136,
+      "grad_norm": 62.75,
+      "learning_rate": 9.996835543589781e-05,
+      "loss": 3.957,
+      "step": 269
+    },
+    {
+      "epoch": 0.011801215088072031,
+      "grad_norm": 4.5,
+      "learning_rate": 9.996811063062119e-05,
+      "loss": 2.7251,
+      "step": 270
+    },
+    {
+      "epoch": 0.011844923292101928,
+      "grad_norm": 4.75,
+      "learning_rate": 9.99678648823736e-05,
+      "loss": 2.8835,
+      "step": 271
+    },
+    {
+      "epoch": 0.011888631496131823,
+      "grad_norm": 4.5,
+      "learning_rate": 9.996761819115968e-05,
+      "loss": 2.8296,
+      "step": 272
+    },
+    {
+      "epoch": 0.01193233970016172,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.996737055698409e-05,
+      "loss": 2.8128,
+      "step": 273
+    },
+    {
+      "epoch": 0.011976047904191617,
+      "grad_norm": 4.125,
+      "learning_rate": 9.996712197985147e-05,
+      "loss": 2.365,
+      "step": 274
+    },
+    {
+      "epoch": 0.012019756108221513,
+      "grad_norm": 6.3125,
+      "learning_rate": 9.996687245976655e-05,
+      "loss": 2.4859,
+      "step": 275
+    },
+    {
+      "epoch": 0.01206346431225141,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.996662199673401e-05,
+      "loss": 2.6919,
+      "step": 276
+    },
+    {
+      "epoch": 0.012107172516281305,
+      "grad_norm": 4.5,
+      "learning_rate": 9.996637059075861e-05,
+      "loss": 2.649,
+      "step": 277
+    },
+    {
+      "epoch": 0.012150880720311202,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.996611824184505e-05,
+      "loss": 3.3843,
+      "step": 278
+    },
+    {
+      "epoch": 0.0121945889243411,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.996586494999814e-05,
+      "loss": 2.5776,
+      "step": 279
+    },
+    {
+      "epoch": 0.012238297128370995,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.996561071522264e-05,
+      "loss": 2.43,
+      "step": 280
+    },
+    {
+      "epoch": 0.012282005332400892,
+      "grad_norm": 5.3125,
+      "learning_rate": 9.996535553752331e-05,
+      "loss": 2.3819,
+      "step": 281
+    },
+    {
+      "epoch": 0.012325713536430787,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.996509941690503e-05,
+      "loss": 2.1984,
+      "step": 282
+    },
+    {
+      "epoch": 0.012369421740460684,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.99648423533726e-05,
+      "loss": 2.7669,
+      "step": 283
+    },
+    {
+      "epoch": 0.012413129944490581,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.996458434693086e-05,
+      "loss": 2.799,
+      "step": 284
+    },
+    {
+      "epoch": 0.012456838148520477,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.99643253975847e-05,
+      "loss": 2.7216,
+      "step": 285
+    },
+    {
+      "epoch": 0.012500546352550374,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.996406550533901e-05,
+      "loss": 2.8937,
+      "step": 286
+    },
+    {
+      "epoch": 0.012544254556580271,
+      "grad_norm": 6.0625,
+      "learning_rate": 9.996380467019868e-05,
+      "loss": 2.3943,
+      "step": 287
+    },
+    {
+      "epoch": 0.012587962760610166,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.996354289216863e-05,
+      "loss": 2.9379,
+      "step": 288
+    },
+    {
+      "epoch": 0.012631670964640063,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.996328017125381e-05,
+      "loss": 2.5333,
+      "step": 289
+    },
+    {
+      "epoch": 0.012675379168669959,
+      "grad_norm": 5.09375,
+      "learning_rate": 9.996301650745917e-05,
+      "loss": 2.6082,
+      "step": 290
+    },
+    {
+      "epoch": 0.012719087372699856,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.99627519007897e-05,
+      "loss": 3.1054,
+      "step": 291
+    },
+    {
+      "epoch": 0.012762795576729753,
+      "grad_norm": 4.5,
+      "learning_rate": 9.996248635125039e-05,
+      "loss": 2.423,
+      "step": 292
+    },
+    {
+      "epoch": 0.012806503780759648,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.996221985884623e-05,
+      "loss": 2.3539,
+      "step": 293
+    },
+    {
+      "epoch": 0.012850211984789545,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.996195242358226e-05,
+      "loss": 2.8162,
+      "step": 294
+    },
+    {
+      "epoch": 0.01289392018881944,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.996168404546356e-05,
+      "loss": 2.4511,
+      "step": 295
+    },
+    {
+      "epoch": 0.012937628392849338,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.996141472449514e-05,
+      "loss": 2.3848,
+      "step": 296
+    },
+    {
+      "epoch": 0.012981336596879235,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.996114446068212e-05,
+      "loss": 2.4308,
+      "step": 297
+    },
+    {
+      "epoch": 0.01302504480090913,
+      "grad_norm": 5.0,
+      "learning_rate": 9.996087325402959e-05,
+      "loss": 2.6241,
+      "step": 298
+    },
+    {
+      "epoch": 0.013068753004939027,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.996060110454266e-05,
+      "loss": 2.1833,
+      "step": 299
+    },
+    {
+      "epoch": 0.013112461208968923,
+      "grad_norm": 3.625,
+      "learning_rate": 9.996032801222648e-05,
+      "loss": 2.782,
+      "step": 300
+    },
+    {
+      "epoch": 0.01315616941299882,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.996005397708619e-05,
+      "loss": 2.0714,
+      "step": 301
+    },
+    {
+      "epoch": 0.013199877617028717,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.995977899912697e-05,
+      "loss": 2.7152,
+      "step": 302
+    },
+    {
+      "epoch": 0.013243585821058612,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.995950307835401e-05,
+      "loss": 2.8984,
+      "step": 303
+    },
+    {
+      "epoch": 0.01328729402508851,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.995922621477252e-05,
+      "loss": 2.3674,
+      "step": 304
+    },
+    {
+      "epoch": 0.013331002229118406,
+      "grad_norm": 6.0625,
+      "learning_rate": 9.995894840838771e-05,
+      "loss": 2.7993,
+      "step": 305
+    },
+    {
+      "epoch": 0.013374710433148302,
+      "grad_norm": 5.78125,
+      "learning_rate": 9.995866965920485e-05,
+      "loss": 2.4223,
+      "step": 306
+    },
+    {
+      "epoch": 0.013418418637178199,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.995838996722914e-05,
+      "loss": 2.8455,
+      "step": 307
+    },
+    {
+      "epoch": 0.013462126841208094,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.995810933246594e-05,
+      "loss": 2.0036,
+      "step": 308
+    },
+    {
+      "epoch": 0.013505835045237991,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.995782775492048e-05,
+      "loss": 2.5858,
+      "step": 309
+    },
+    {
+      "epoch": 0.013549543249267888,
+      "grad_norm": 11.75,
+      "learning_rate": 9.995754523459813e-05,
+      "loss": 2.921,
+      "step": 310
+    },
+    {
+      "epoch": 0.013593251453297784,
+      "grad_norm": 27.875,
+      "learning_rate": 9.995726177150418e-05,
+      "loss": 2.7016,
+      "step": 311
+    },
+    {
+      "epoch": 0.01363695965732768,
+      "grad_norm": 5.09375,
+      "learning_rate": 9.995697736564397e-05,
+      "loss": 2.4968,
+      "step": 312
+    },
+    {
+      "epoch": 0.013680667861357576,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.995669201702291e-05,
+      "loss": 3.1727,
+      "step": 313
+    },
+    {
+      "epoch": 0.013724376065387473,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.995640572564635e-05,
+      "loss": 2.812,
+      "step": 314
+    },
+    {
+      "epoch": 0.01376808426941737,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.995611849151971e-05,
+      "loss": 2.3059,
+      "step": 315
+    },
+    {
+      "epoch": 0.013811792473447266,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.995583031464842e-05,
+      "loss": 3.082,
+      "step": 316
+    },
+    {
+      "epoch": 0.013855500677477163,
+      "grad_norm": 4.0,
+      "learning_rate": 9.99555411950379e-05,
+      "loss": 2.0426,
+      "step": 317
+    },
+    {
+      "epoch": 0.013899208881507058,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.99552511326936e-05,
+      "loss": 2.8893,
+      "step": 318
+    },
+    {
+      "epoch": 0.013942917085536955,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.9954960127621e-05,
+      "loss": 2.4564,
+      "step": 319
+    },
+    {
+      "epoch": 0.013986625289566852,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.995466817982562e-05,
+      "loss": 2.5178,
+      "step": 320
+    },
+    {
+      "epoch": 0.014030333493596748,
+      "grad_norm": 15.625,
+      "learning_rate": 9.995437528931293e-05,
+      "loss": 3.7793,
+      "step": 321
+    },
+    {
+      "epoch": 0.014074041697626645,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.995408145608847e-05,
+      "loss": 2.3784,
+      "step": 322
+    },
+    {
+      "epoch": 0.014117749901656542,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.99537866801578e-05,
+      "loss": 2.2287,
+      "step": 323
+    },
+    {
+      "epoch": 0.014161458105686437,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.995349096152645e-05,
+      "loss": 2.4412,
+      "step": 324
+    },
+    {
+      "epoch": 0.014205166309716334,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.995319430020003e-05,
+      "loss": 2.3395,
+      "step": 325
+    },
+    {
+      "epoch": 0.01424887451374623,
+      "grad_norm": 4.5,
+      "learning_rate": 9.995289669618415e-05,
+      "loss": 2.7861,
+      "step": 326
+    },
+    {
+      "epoch": 0.014292582717776127,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.995259814948439e-05,
+      "loss": 2.4553,
+      "step": 327
+    },
+    {
+      "epoch": 0.014336290921806024,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.99522986601064e-05,
+      "loss": 2.9104,
+      "step": 328
+    },
+    {
+      "epoch": 0.014379999125835919,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.995199822805583e-05,
+      "loss": 2.8891,
+      "step": 329
+    },
+    {
+      "epoch": 0.014423707329865816,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.995169685333836e-05,
+      "loss": 2.8756,
+      "step": 330
+    },
+    {
+      "epoch": 0.014467415533895711,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.995139453595968e-05,
+      "loss": 2.123,
+      "step": 331
+    },
+    {
+      "epoch": 0.014511123737925609,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.995109127592546e-05,
+      "loss": 2.7999,
+      "step": 332
+    },
+    {
+      "epoch": 0.014554831941955506,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.995078707324146e-05,
+      "loss": 2.3765,
+      "step": 333
+    },
+    {
+      "epoch": 0.014598540145985401,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.99504819279134e-05,
+      "loss": 2.0792,
+      "step": 334
+    },
+    {
+      "epoch": 0.014642248350015298,
+      "grad_norm": 4.125,
+      "learning_rate": 9.995017583994706e-05,
+      "loss": 2.5463,
+      "step": 335
+    },
+    {
+      "epoch": 0.014685956554045195,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.99498688093482e-05,
+      "loss": 3.4456,
+      "step": 336
+    },
+    {
+      "epoch": 0.01472966475807509,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.994956083612261e-05,
+      "loss": 2.3138,
+      "step": 337
+    },
+    {
+      "epoch": 0.014773372962104988,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.99492519202761e-05,
+      "loss": 2.4555,
+      "step": 338
+    },
+    {
+      "epoch": 0.014817081166134883,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.994894206181452e-05,
+      "loss": 2.4877,
+      "step": 339
+    },
+    {
+      "epoch": 0.01486078937016478,
+      "grad_norm": 4.875,
+      "learning_rate": 9.994863126074371e-05,
+      "loss": 2.9761,
+      "step": 340
+    },
+    {
+      "epoch": 0.014904497574194677,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.994831951706953e-05,
+      "loss": 2.5688,
+      "step": 341
+    },
+    {
+      "epoch": 0.014948205778224572,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.994800683079786e-05,
+      "loss": 2.6589,
+      "step": 342
+    },
+    {
+      "epoch": 0.01499191398225447,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.99476932019346e-05,
+      "loss": 2.4292,
+      "step": 343
+    },
+    {
+      "epoch": 0.015035622186284365,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.994737863048567e-05,
+      "loss": 2.5704,
+      "step": 344
+    },
+    {
+      "epoch": 0.015079330390314262,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.994706311645703e-05,
+      "loss": 2.5487,
+      "step": 345
+    },
+    {
+      "epoch": 0.015123038594344159,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.994674665985461e-05,
+      "loss": 2.1273,
+      "step": 346
+    },
+    {
+      "epoch": 0.015166746798374054,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.994642926068438e-05,
+      "loss": 2.0096,
+      "step": 347
+    },
+    {
+      "epoch": 0.015210455002403951,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.994611091895234e-05,
+      "loss": 2.841,
+      "step": 348
+    },
+    {
+      "epoch": 0.015254163206433847,
+      "grad_norm": 3.375,
+      "learning_rate": 9.994579163466448e-05,
+      "loss": 2.3199,
+      "step": 349
+    },
+    {
+      "epoch": 0.015297871410463744,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.994547140782686e-05,
+      "loss": 2.1122,
+      "step": 350
+    },
+    {
+      "epoch": 0.015341579614493641,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.994515023844548e-05,
+      "loss": 2.705,
+      "step": 351
+    },
+    {
+      "epoch": 0.015385287818523536,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.994482812652645e-05,
+      "loss": 1.9516,
+      "step": 352
+    },
+    {
+      "epoch": 0.015428996022553433,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.99445050720758e-05,
+      "loss": 2.2709,
+      "step": 353
+    },
+    {
+      "epoch": 0.01547270422658333,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.994418107509966e-05,
+      "loss": 2.2912,
+      "step": 354
+    },
+    {
+      "epoch": 0.015516412430613226,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.994385613560413e-05,
+      "loss": 2.6541,
+      "step": 355
+    },
+    {
+      "epoch": 0.015560120634643123,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.994353025359535e-05,
+      "loss": 2.1626,
+      "step": 356
+    },
+    {
+      "epoch": 0.015603828838673018,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.994320342907945e-05,
+      "loss": 2.693,
+      "step": 357
+    },
+    {
+      "epoch": 0.015647537042702914,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.994287566206264e-05,
+      "loss": 2.8311,
+      "step": 358
+    },
+    {
+      "epoch": 0.01569124524673281,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.994254695255105e-05,
+      "loss": 2.4746,
+      "step": 359
+    },
+    {
+      "epoch": 0.015734953450762708,
+      "grad_norm": 3.375,
+      "learning_rate": 9.994221730055091e-05,
+      "loss": 2.3147,
+      "step": 360
+    },
+    {
+      "epoch": 0.015778661654792605,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.994188670606846e-05,
+      "loss": 2.7883,
+      "step": 361
+    },
+    {
+      "epoch": 0.015822369858822502,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.994155516910991e-05,
+      "loss": 2.2414,
+      "step": 362
+    },
+    {
+      "epoch": 0.0158660780628524,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.994122268968154e-05,
+      "loss": 2.3983,
+      "step": 363
+    },
+    {
+      "epoch": 0.015909786266882293,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.99408892677896e-05,
+      "loss": 2.6232,
+      "step": 364
+    },
+    {
+      "epoch": 0.01595349447091219,
+      "grad_norm": 6.09375,
+      "learning_rate": 9.99405549034404e-05,
+      "loss": 2.3536,
+      "step": 365
+    },
+    {
+      "epoch": 0.015997202674942087,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.994021959664024e-05,
+      "loss": 2.8419,
+      "step": 366
+    },
+    {
+      "epoch": 0.016040910878971984,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.993988334739544e-05,
+      "loss": 2.9319,
+      "step": 367
+    },
+    {
+      "epoch": 0.01608461908300188,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.993954615571238e-05,
+      "loss": 2.4233,
+      "step": 368
+    },
+    {
+      "epoch": 0.016128327287031775,
+      "grad_norm": 5.84375,
+      "learning_rate": 9.993920802159739e-05,
+      "loss": 3.2623,
+      "step": 369
+    },
+    {
+      "epoch": 0.016172035491061672,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.993886894505686e-05,
+      "loss": 2.6416,
+      "step": 370
+    },
+    {
+      "epoch": 0.01621574369509157,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.993852892609718e-05,
+      "loss": 2.7184,
+      "step": 371
+    },
+    {
+      "epoch": 0.016259451899121466,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.99381879647248e-05,
+      "loss": 2.9044,
+      "step": 372
+    },
+    {
+      "epoch": 0.016303160103151363,
+      "grad_norm": 4.5,
+      "learning_rate": 9.993784606094612e-05,
+      "loss": 2.1336,
+      "step": 373
+    },
+    {
+      "epoch": 0.016346868307181257,
+      "grad_norm": 6.34375,
+      "learning_rate": 9.99375032147676e-05,
+      "loss": 2.2281,
+      "step": 374
+    },
+    {
+      "epoch": 0.016390576511211154,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.993715942619573e-05,
+      "loss": 2.3788,
+      "step": 375
+    },
+    {
+      "epoch": 0.01643428471524105,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.993681469523697e-05,
+      "loss": 1.9586,
+      "step": 376
+    },
+    {
+      "epoch": 0.016477992919270948,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.993646902189784e-05,
+      "loss": 3.2296,
+      "step": 377
+    },
+    {
+      "epoch": 0.016521701123300845,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.993612240618485e-05,
+      "loss": 2.4422,
+      "step": 378
+    },
+    {
+      "epoch": 0.01656540932733074,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.993577484810455e-05,
+      "loss": 2.8639,
+      "step": 379
+    },
+    {
+      "epoch": 0.016609117531360636,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.993542634766352e-05,
+      "loss": 2.9119,
+      "step": 380
+    },
+    {
+      "epoch": 0.016652825735390533,
+      "grad_norm": 4.25,
+      "learning_rate": 9.993507690486831e-05,
+      "loss": 2.5102,
+      "step": 381
+    },
+    {
+      "epoch": 0.01669653393942043,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.99347265197255e-05,
+      "loss": 2.5379,
+      "step": 382
+    },
+    {
+      "epoch": 0.016740242143450327,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.993437519224176e-05,
+      "loss": 2.2643,
+      "step": 383
+    },
+    {
+      "epoch": 0.01678395034748022,
+      "grad_norm": 3.625,
+      "learning_rate": 9.993402292242367e-05,
+      "loss": 2.3072,
+      "step": 384
+    },
+    {
+      "epoch": 0.016827658551510118,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.993366971027788e-05,
+      "loss": 2.592,
+      "step": 385
+    },
+    {
+      "epoch": 0.016871366755540015,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.993331555581108e-05,
+      "loss": 2.6759,
+      "step": 386
+    },
+    {
+      "epoch": 0.016915074959569912,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.993296045902994e-05,
+      "loss": 2.2537,
+      "step": 387
+    },
+    {
+      "epoch": 0.01695878316359981,
+      "grad_norm": 4.625,
+      "learning_rate": 9.993260441994116e-05,
+      "loss": 2.7636,
+      "step": 388
+    },
+    {
+      "epoch": 0.017002491367629703,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.993224743855145e-05,
+      "loss": 2.5618,
+      "step": 389
+    },
+    {
+      "epoch": 0.0170461995716596,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.993188951486758e-05,
+      "loss": 2.3234,
+      "step": 390
+    },
+    {
+      "epoch": 0.017089907775689497,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.993153064889626e-05,
+      "loss": 3.1043,
+      "step": 391
+    },
+    {
+      "epoch": 0.017133615979719394,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.99311708406443e-05,
+      "loss": 2.9755,
+      "step": 392
+    },
+    {
+      "epoch": 0.01717732418374929,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.993081009011847e-05,
+      "loss": 2.5243,
+      "step": 393
+    },
+    {
+      "epoch": 0.017221032387779184,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.993044839732559e-05,
+      "loss": 2.388,
+      "step": 394
+    },
+    {
+      "epoch": 0.01726474059180908,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.993008576227247e-05,
+      "loss": 2.1978,
+      "step": 395
+    },
+    {
+      "epoch": 0.01730844879583898,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.992972218496597e-05,
+      "loss": 2.2283,
+      "step": 396
+    },
+    {
+      "epoch": 0.017352156999868876,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.992935766541294e-05,
+      "loss": 2.2622,
+      "step": 397
+    },
+    {
+      "epoch": 0.017395865203898773,
+      "grad_norm": 3.625,
+      "learning_rate": 9.992899220362025e-05,
+      "loss": 2.4432,
+      "step": 398
+    },
+    {
+      "epoch": 0.01743957340792867,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.992862579959481e-05,
+      "loss": 2.408,
+      "step": 399
+    },
+    {
+      "epoch": 0.017483281611958564,
+      "grad_norm": 3.625,
+      "learning_rate": 9.992825845334355e-05,
+      "loss": 1.867,
+      "step": 400
+    },
+    {
+      "epoch": 0.01752698981598846,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.992789016487337e-05,
+      "loss": 2.1599,
+      "step": 401
+    },
+    {
+      "epoch": 0.017570698020018358,
+      "grad_norm": 7.4375,
+      "learning_rate": 9.992752093419124e-05,
+      "loss": 2.4313,
+      "step": 402
+    },
+    {
+      "epoch": 0.017614406224048255,
+      "grad_norm": 6.5,
+      "learning_rate": 9.992715076130414e-05,
+      "loss": 2.2794,
+      "step": 403
+    },
+    {
+      "epoch": 0.017658114428078152,
+      "grad_norm": 4.25,
+      "learning_rate": 9.992677964621901e-05,
+      "loss": 2.8487,
+      "step": 404
+    },
+    {
+      "epoch": 0.017701822632108045,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.992640758894292e-05,
+      "loss": 2.3595,
+      "step": 405
+    },
+    {
+      "epoch": 0.017745530836137943,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.992603458948281e-05,
+      "loss": 2.5555,
+      "step": 406
+    },
+    {
+      "epoch": 0.01778923904016784,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.992566064784581e-05,
+      "loss": 2.0888,
+      "step": 407
+    },
+    {
+      "epoch": 0.017832947244197737,
+      "grad_norm": 4.5,
+      "learning_rate": 9.99252857640389e-05,
+      "loss": 2.4566,
+      "step": 408
+    },
+    {
+      "epoch": 0.017876655448227634,
+      "grad_norm": 3.125,
+      "learning_rate": 9.99249099380692e-05,
+      "loss": 2.1311,
+      "step": 409
+    },
+    {
+      "epoch": 0.017920363652257527,
+      "grad_norm": 3.875,
+      "learning_rate": 9.992453316994377e-05,
+      "loss": 2.4535,
+      "step": 410
+    },
+    {
+      "epoch": 0.017964071856287425,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.992415545966976e-05,
+      "loss": 2.1546,
+      "step": 411
+    },
+    {
+      "epoch": 0.01800778006031732,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.992377680725425e-05,
+      "loss": 2.4082,
+      "step": 412
+    },
+    {
+      "epoch": 0.01805148826434722,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.992339721270443e-05,
+      "loss": 2.7536,
+      "step": 413
+    },
+    {
+      "epoch": 0.018095196468377116,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.992301667602743e-05,
+      "loss": 2.5332,
+      "step": 414
+    },
+    {
+      "epoch": 0.01813890467240701,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.992263519723046e-05,
+      "loss": 2.2629,
+      "step": 415
+    },
+    {
+      "epoch": 0.018182612876436906,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.99222527763207e-05,
+      "loss": 2.08,
+      "step": 416
+    },
+    {
+      "epoch": 0.018226321080466804,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.992186941330537e-05,
+      "loss": 2.2829,
+      "step": 417
+    },
+    {
+      "epoch": 0.0182700292844967,
+      "grad_norm": 3.0,
+      "learning_rate": 9.99214851081917e-05,
+      "loss": 2.5898,
+      "step": 418
+    },
+    {
+      "epoch": 0.018313737488526598,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.992109986098696e-05,
+      "loss": 2.425,
+      "step": 419
+    },
+    {
+      "epoch": 0.01835744569255649,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.992071367169839e-05,
+      "loss": 2.2076,
+      "step": 420
+    },
+    {
+      "epoch": 0.01840115389658639,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.992032654033333e-05,
+      "loss": 2.2598,
+      "step": 421
+    },
+    {
+      "epoch": 0.018444862100616286,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.991993846689902e-05,
+      "loss": 2.3534,
+      "step": 422
+    },
+    {
+      "epoch": 0.018488570304646183,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.991954945140284e-05,
+      "loss": 2.023,
+      "step": 423
+    },
+    {
+      "epoch": 0.01853227850867608,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.99191594938521e-05,
+      "loss": 2.5662,
+      "step": 424
+    },
+    {
+      "epoch": 0.018575986712705973,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.991876859425415e-05,
+      "loss": 2.2721,
+      "step": 425
+    },
+    {
+      "epoch": 0.01861969491673587,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.991837675261641e-05,
+      "loss": 2.6638,
+      "step": 426
+    },
+    {
+      "epoch": 0.018663403120765767,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.991798396894622e-05,
+      "loss": 2.6768,
+      "step": 427
+    },
+    {
+      "epoch": 0.018707111324795665,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.991759024325104e-05,
+      "loss": 3.2616,
+      "step": 428
+    },
+    {
+      "epoch": 0.01875081952882556,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.99171955755383e-05,
+      "loss": 2.2708,
+      "step": 429
+    },
+    {
+      "epoch": 0.018794527732855455,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.991679996581539e-05,
+      "loss": 2.6224,
+      "step": 430
+    },
+    {
+      "epoch": 0.018838235936885352,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.991640341408984e-05,
+      "loss": 1.7391,
+      "step": 431
+    },
+    {
+      "epoch": 0.01888194414091525,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.991600592036908e-05,
+      "loss": 2.494,
+      "step": 432
+    },
+    {
+      "epoch": 0.018925652344945147,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.991560748466067e-05,
+      "loss": 2.703,
+      "step": 433
+    },
+    {
+      "epoch": 0.018969360548975044,
+      "grad_norm": 8.625,
+      "learning_rate": 9.991520810697208e-05,
+      "loss": 2.121,
+      "step": 434
+    },
+    {
+      "epoch": 0.01901306875300494,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.991480778731086e-05,
+      "loss": 1.7441,
+      "step": 435
+    },
+    {
+      "epoch": 0.019056776957034834,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.991440652568458e-05,
+      "loss": 1.911,
+      "step": 436
+    },
+    {
+      "epoch": 0.01910048516106473,
+      "grad_norm": 3.0,
+      "learning_rate": 9.99140043221008e-05,
+      "loss": 2.441,
+      "step": 437
+    },
+    {
+      "epoch": 0.01914419336509463,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.991360117656712e-05,
+      "loss": 2.7907,
+      "step": 438
+    },
+    {
+      "epoch": 0.019187901569124526,
+      "grad_norm": 4.125,
+      "learning_rate": 9.991319708909113e-05,
+      "loss": 2.7946,
+      "step": 439
+    },
+    {
+      "epoch": 0.019231609773154423,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.991279205968046e-05,
+      "loss": 2.985,
+      "step": 440
+    },
+    {
+      "epoch": 0.019275317977184316,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.991238608834276e-05,
+      "loss": 2.6155,
+      "step": 441
+    },
+    {
+      "epoch": 0.019319026181214213,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.99119791750857e-05,
+      "loss": 2.3012,
+      "step": 442
+    },
+    {
+      "epoch": 0.01936273438524411,
+      "grad_norm": 6.09375,
+      "learning_rate": 9.991157131991695e-05,
+      "loss": 3.8634,
+      "step": 443
+    },
+    {
+      "epoch": 0.019406442589274008,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.991116252284421e-05,
+      "loss": 2.4776,
+      "step": 444
+    },
+    {
+      "epoch": 0.019450150793303905,
+      "grad_norm": 6.625,
+      "learning_rate": 9.991075278387518e-05,
+      "loss": 2.2461,
+      "step": 445
+    },
+    {
+      "epoch": 0.019493858997333798,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.99103421030176e-05,
+      "loss": 2.2593,
+      "step": 446
+    },
+    {
+      "epoch": 0.019537567201363695,
+      "grad_norm": 10.375,
+      "learning_rate": 9.990993048027923e-05,
+      "loss": 2.8026,
+      "step": 447
+    },
+    {
+      "epoch": 0.019581275405393592,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.990951791566784e-05,
+      "loss": 2.234,
+      "step": 448
+    },
+    {
+      "epoch": 0.01962498360942349,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.99091044091912e-05,
+      "loss": 2.7442,
+      "step": 449
+    },
+    {
+      "epoch": 0.019668691813453387,
+      "grad_norm": 4.125,
+      "learning_rate": 9.990868996085712e-05,
+      "loss": 2.5416,
+      "step": 450
+    },
+    {
+      "epoch": 0.01971240001748328,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.990827457067343e-05,
+      "loss": 2.4102,
+      "step": 451
+    },
+    {
+      "epoch": 0.019756108221513177,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.990785823864795e-05,
+      "loss": 2.8615,
+      "step": 452
+    },
+    {
+      "epoch": 0.019799816425543074,
+      "grad_norm": 3.125,
+      "learning_rate": 9.990744096478855e-05,
+      "loss": 2.4941,
+      "step": 453
+    },
+    {
+      "epoch": 0.01984352462957297,
+      "grad_norm": 5.0,
+      "learning_rate": 9.990702274910309e-05,
+      "loss": 3.4405,
+      "step": 454
+    },
+    {
+      "epoch": 0.01988723283360287,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.990660359159949e-05,
+      "loss": 2.8826,
+      "step": 455
+    },
+    {
+      "epoch": 0.019930941037632762,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.990618349228564e-05,
+      "loss": 2.3629,
+      "step": 456
+    },
+    {
+      "epoch": 0.01997464924166266,
+      "grad_norm": 5.96875,
+      "learning_rate": 9.990576245116947e-05,
+      "loss": 3.5398,
+      "step": 457
+    },
+    {
+      "epoch": 0.020018357445692556,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.990534046825893e-05,
+      "loss": 2.4665,
+      "step": 458
+    },
+    {
+      "epoch": 0.020062065649722453,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.990491754356199e-05,
+      "loss": 2.6201,
+      "step": 459
+    },
+    {
+      "epoch": 0.02010577385375235,
+      "grad_norm": 4.0,
+      "learning_rate": 9.990449367708661e-05,
+      "loss": 2.3883,
+      "step": 460
+    },
+    {
+      "epoch": 0.020149482057782244,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.99040688688408e-05,
+      "loss": 2.013,
+      "step": 461
+    },
+    {
+      "epoch": 0.02019319026181214,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.99036431188326e-05,
+      "loss": 2.3463,
+      "step": 462
+    },
+    {
+      "epoch": 0.020236898465842038,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.990321642707001e-05,
+      "loss": 1.838,
+      "step": 463
+    },
+    {
+      "epoch": 0.020280606669871935,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.99027887935611e-05,
+      "loss": 2.5029,
+      "step": 464
+    },
+    {
+      "epoch": 0.020324314873901832,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.990236021831391e-05,
+      "loss": 2.2701,
+      "step": 465
+    },
+    {
+      "epoch": 0.02036802307793173,
+      "grad_norm": 4.125,
+      "learning_rate": 9.990193070133659e-05,
+      "loss": 2.177,
+      "step": 466
+    },
+    {
+      "epoch": 0.020411731281961623,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.99015002426372e-05,
+      "loss": 2.4276,
+      "step": 467
+    },
+    {
+      "epoch": 0.02045543948599152,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.990106884222385e-05,
+      "loss": 2.2515,
+      "step": 468
+    },
+    {
+      "epoch": 0.020499147690021417,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.990063650010473e-05,
+      "loss": 2.238,
+      "step": 469
+    },
+    {
+      "epoch": 0.020542855894051314,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.990020321628794e-05,
+      "loss": 2.5933,
+      "step": 470
+    },
+    {
+      "epoch": 0.02058656409808121,
+      "grad_norm": 10.1875,
+      "learning_rate": 9.989976899078172e-05,
+      "loss": 3.4384,
+      "step": 471
+    },
+    {
+      "epoch": 0.020630272302111105,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.989933382359422e-05,
+      "loss": 2.2785,
+      "step": 472
+    },
+    {
+      "epoch": 0.020673980506141002,
+      "grad_norm": 4.125,
+      "learning_rate": 9.989889771473367e-05,
+      "loss": 2.3461,
+      "step": 473
+    },
+    {
+      "epoch": 0.0207176887101709,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.989846066420829e-05,
+      "loss": 2.3127,
+      "step": 474
+    },
+    {
+      "epoch": 0.020761396914200796,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.989802267202635e-05,
+      "loss": 3.1751,
+      "step": 475
+    },
+    {
+      "epoch": 0.020805105118230693,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.989758373819608e-05,
+      "loss": 2.5285,
+      "step": 476
+    },
+    {
+      "epoch": 0.020848813322260587,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.989714386272579e-05,
+      "loss": 2.064,
+      "step": 477
+    },
+    {
+      "epoch": 0.020892521526290484,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.989670304562377e-05,
+      "loss": 2.2734,
+      "step": 478
+    },
+    {
+      "epoch": 0.02093622973032038,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.989626128689835e-05,
+      "loss": 2.8476,
+      "step": 479
+    },
+    {
+      "epoch": 0.02097993793435028,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.989581858655785e-05,
+      "loss": 2.5702,
+      "step": 480
+    },
+    {
+      "epoch": 0.021023646138380175,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.989537494461064e-05,
+      "loss": 2.4645,
+      "step": 481
+    },
+    {
+      "epoch": 0.02106735434241007,
+      "grad_norm": 7.40625,
+      "learning_rate": 9.989493036106507e-05,
+      "loss": 3.0028,
+      "step": 482
+    },
+    {
+      "epoch": 0.021111062546439966,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.989448483592957e-05,
+      "loss": 2.9433,
+      "step": 483
+    },
+    {
+      "epoch": 0.021154770750469863,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.989403836921251e-05,
+      "loss": 2.474,
+      "step": 484
+    },
+    {
+      "epoch": 0.02119847895449976,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.989359096092233e-05,
+      "loss": 2.9873,
+      "step": 485
+    },
+    {
+      "epoch": 0.021242187158529657,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.989314261106749e-05,
+      "loss": 2.3552,
+      "step": 486
+    },
+    {
+      "epoch": 0.02128589536255955,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.98926933196564e-05,
+      "loss": 2.4021,
+      "step": 487
+    },
+    {
+      "epoch": 0.021329603566589448,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.989224308669758e-05,
+      "loss": 2.3827,
+      "step": 488
+    },
+    {
+      "epoch": 0.021373311770619345,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.989179191219952e-05,
+      "loss": 2.1144,
+      "step": 489
+    },
+    {
+      "epoch": 0.021417019974649242,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.989133979617074e-05,
+      "loss": 2.1834,
+      "step": 490
+    },
+    {
+      "epoch": 0.02146072817867914,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.989088673861977e-05,
+      "loss": 2.3981,
+      "step": 491
+    },
+    {
+      "epoch": 0.021504436382709033,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.989043273955513e-05,
+      "loss": 2.8218,
+      "step": 492
+    },
+    {
+      "epoch": 0.02154814458673893,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.988997779898545e-05,
+      "loss": 2.0585,
+      "step": 493
+    },
+    {
+      "epoch": 0.021591852790768827,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.988952191691925e-05,
+      "loss": 2.5417,
+      "step": 494
+    },
+    {
+      "epoch": 0.021635560994798724,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.988906509336518e-05,
+      "loss": 1.9448,
+      "step": 495
+    },
+    {
+      "epoch": 0.02167926919882862,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.988860732833182e-05,
+      "loss": 2.2873,
+      "step": 496
+    },
+    {
+      "epoch": 0.021722977402858515,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.988814862182783e-05,
+      "loss": 2.8405,
+      "step": 497
+    },
+    {
+      "epoch": 0.021766685606888412,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.988768897386188e-05,
+      "loss": 2.4951,
+      "step": 498
+    },
+    {
+      "epoch": 0.02181039381091831,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.988722838444262e-05,
+      "loss": 2.1855,
+      "step": 499
+    },
+    {
+      "epoch": 0.021854102014948206,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.988676685357876e-05,
+      "loss": 1.9974,
+      "step": 500
+    },
+    {
+      "epoch": 0.021897810218978103,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.988630438127901e-05,
+      "loss": 1.9175,
+      "step": 501
+    },
+    {
+      "epoch": 0.021941518423008,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.988584096755208e-05,
+      "loss": 2.3125,
+      "step": 502
+    },
+    {
+      "epoch": 0.021985226627037894,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.988537661240673e-05,
+      "loss": 2.1997,
+      "step": 503
+    },
+    {
+      "epoch": 0.02202893483106779,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.988491131585171e-05,
+      "loss": 2.3069,
+      "step": 504
+    },
+    {
+      "epoch": 0.022072643035097688,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.988444507789582e-05,
+      "loss": 2.2246,
+      "step": 505
+    },
+    {
+      "epoch": 0.022116351239127585,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.988397789854784e-05,
+      "loss": 2.7865,
+      "step": 506
+    },
+    {
+      "epoch": 0.022160059443157482,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.98835097778166e-05,
+      "loss": 2.2723,
+      "step": 507
+    },
+    {
+      "epoch": 0.022203767647187376,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.988304071571093e-05,
+      "loss": 2.3632,
+      "step": 508
+    },
+    {
+      "epoch": 0.022247475851217273,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.988257071223968e-05,
+      "loss": 2.6076,
+      "step": 509
+    },
+    {
+      "epoch": 0.02229118405524717,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.988209976741172e-05,
+      "loss": 2.2625,
+      "step": 510
+    },
+    {
+      "epoch": 0.022334892259277067,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.988162788123594e-05,
+      "loss": 2.5188,
+      "step": 511
+    },
+    {
+      "epoch": 0.022378600463306964,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.988115505372123e-05,
+      "loss": 2.7811,
+      "step": 512
+    },
+    {
+      "epoch": 0.022422308667336858,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.988068128487654e-05,
+      "loss": 2.3825,
+      "step": 513
+    },
+    {
+      "epoch": 0.022466016871366755,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.988020657471077e-05,
+      "loss": 2.4349,
+      "step": 514
+    },
+    {
+      "epoch": 0.022509725075396652,
+      "grad_norm": 3.875,
+      "learning_rate": 9.987973092323293e-05,
+      "loss": 2.6618,
+      "step": 515
+    },
+    {
+      "epoch": 0.02255343327942655,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.987925433045197e-05,
+      "loss": 1.955,
+      "step": 516
+    },
+    {
+      "epoch": 0.022597141483456446,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.987877679637688e-05,
+      "loss": 2.9896,
+      "step": 517
+    },
+    {
+      "epoch": 0.02264084968748634,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.987829832101667e-05,
+      "loss": 2.0962,
+      "step": 518
+    },
+    {
+      "epoch": 0.022684557891516237,
+      "grad_norm": 4.25,
+      "learning_rate": 9.987781890438039e-05,
+      "loss": 2.0316,
+      "step": 519
+    },
+    {
+      "epoch": 0.022728266095546134,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.987733854647707e-05,
+      "loss": 2.8837,
+      "step": 520
+    },
+    {
+      "epoch": 0.02277197429957603,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.987685724731577e-05,
+      "loss": 2.0682,
+      "step": 521
+    },
+    {
+      "epoch": 0.022815682503605928,
+      "grad_norm": 4.125,
+      "learning_rate": 9.987637500690559e-05,
+      "loss": 2.4832,
+      "step": 522
+    },
+    {
+      "epoch": 0.02285939070763582,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.987589182525561e-05,
+      "loss": 3.3895,
+      "step": 523
+    },
+    {
+      "epoch": 0.02290309891166572,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.987540770237498e-05,
+      "loss": 2.2143,
+      "step": 524
+    },
+    {
+      "epoch": 0.022946807115695616,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.98749226382728e-05,
+      "loss": 2.1937,
+      "step": 525
+    },
+    {
+      "epoch": 0.022990515319725513,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.987443663295825e-05,
+      "loss": 3.2734,
+      "step": 526
+    },
+    {
+      "epoch": 0.02303422352375541,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.987394968644049e-05,
+      "loss": 2.7082,
+      "step": 527
+    },
+    {
+      "epoch": 0.023077931727785304,
+      "grad_norm": 3.25,
+      "learning_rate": 9.987346179872869e-05,
+      "loss": 2.4288,
+      "step": 528
+    },
+    {
+      "epoch": 0.0231216399318152,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.987297296983211e-05,
+      "loss": 2.2455,
+      "step": 529
+    },
+    {
+      "epoch": 0.023165348135845098,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.987248319975993e-05,
+      "loss": 2.1247,
+      "step": 530
+    },
+    {
+      "epoch": 0.023209056339874995,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.987199248852141e-05,
+      "loss": 2.9593,
+      "step": 531
+    },
+    {
+      "epoch": 0.023252764543904892,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.987150083612579e-05,
+      "loss": 2.4991,
+      "step": 532
+    },
+    {
+      "epoch": 0.023296472747934786,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.987100824258239e-05,
+      "loss": 2.68,
+      "step": 533
+    },
+    {
+      "epoch": 0.023340180951964683,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.987051470790048e-05,
+      "loss": 2.7721,
+      "step": 534
+    },
+    {
+      "epoch": 0.02338388915599458,
+      "grad_norm": 3.5,
+      "learning_rate": 9.987002023208935e-05,
+      "loss": 2.1934,
+      "step": 535
+    },
+    {
+      "epoch": 0.023427597360024477,
+      "grad_norm": 3.5,
+      "learning_rate": 9.986952481515836e-05,
+      "loss": 2.1404,
+      "step": 536
+    },
+    {
+      "epoch": 0.023471305564054374,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.986902845711687e-05,
+      "loss": 2.2057,
+      "step": 537
+    },
+    {
+      "epoch": 0.02351501376808427,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.986853115797423e-05,
+      "loss": 2.0918,
+      "step": 538
+    },
+    {
+      "epoch": 0.023558721972114165,
+      "grad_norm": 3.25,
+      "learning_rate": 9.986803291773982e-05,
+      "loss": 2.3491,
+      "step": 539
+    },
+    {
+      "epoch": 0.023602430176144062,
+      "grad_norm": 3.25,
+      "learning_rate": 9.986753373642306e-05,
+      "loss": 3.0529,
+      "step": 540
+    },
+    {
+      "epoch": 0.02364613838017396,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.986703361403335e-05,
+      "loss": 2.1784,
+      "step": 541
+    },
+    {
+      "epoch": 0.023689846584203856,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.986653255058014e-05,
+      "loss": 2.5431,
+      "step": 542
+    },
+    {
+      "epoch": 0.023733554788233753,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.986603054607288e-05,
+      "loss": 2.2682,
+      "step": 543
+    },
+    {
+      "epoch": 0.023777262992263647,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.986552760052105e-05,
+      "loss": 2.5113,
+      "step": 544
+    },
+    {
+      "epoch": 0.023820971196293544,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.986502371393413e-05,
+      "loss": 2.7172,
+      "step": 545
+    },
+    {
+      "epoch": 0.02386467940032344,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.986451888632165e-05,
+      "loss": 2.6705,
+      "step": 546
+    },
+    {
+      "epoch": 0.023908387604353338,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.986401311769312e-05,
+      "loss": 2.2219,
+      "step": 547
+    },
+    {
+      "epoch": 0.023952095808383235,
+      "grad_norm": 3.875,
+      "learning_rate": 9.98635064080581e-05,
+      "loss": 2.3801,
+      "step": 548
+    },
+    {
+      "epoch": 0.02399580401241313,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.986299875742613e-05,
+      "loss": 2.2103,
+      "step": 549
+    },
+    {
+      "epoch": 0.024039512216443026,
+      "grad_norm": 4.75,
+      "learning_rate": 9.98624901658068e-05,
+      "loss": 2.581,
+      "step": 550
+    },
+    {
+      "epoch": 0.024083220420472923,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.986198063320971e-05,
+      "loss": 2.2305,
+      "step": 551
+    },
+    {
+      "epoch": 0.02412692862450282,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.986147015964446e-05,
+      "loss": 2.5085,
+      "step": 552
+    },
+    {
+      "epoch": 0.024170636828532717,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.986095874512072e-05,
+      "loss": 2.3899,
+      "step": 553
+    },
+    {
+      "epoch": 0.02421434503256261,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.986044638964811e-05,
+      "loss": 2.2853,
+      "step": 554
+    },
+    {
+      "epoch": 0.024258053236592508,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.985993309323631e-05,
+      "loss": 2.8576,
+      "step": 555
+    },
+    {
+      "epoch": 0.024301761440622405,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.985941885589502e-05,
+      "loss": 2.8423,
+      "step": 556
+    },
+    {
+      "epoch": 0.024345469644652302,
+      "grad_norm": 6.40625,
+      "learning_rate": 9.985890367763391e-05,
+      "loss": 2.9808,
+      "step": 557
+    },
+    {
+      "epoch": 0.0243891778486822,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.985838755846273e-05,
+      "loss": 2.2293,
+      "step": 558
+    },
+    {
+      "epoch": 0.024432886052712093,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.98578704983912e-05,
+      "loss": 3.302,
+      "step": 559
+    },
+    {
+      "epoch": 0.02447659425674199,
+      "grad_norm": 13.1875,
+      "learning_rate": 9.98573524974291e-05,
+      "loss": 3.9046,
+      "step": 560
+    },
+    {
+      "epoch": 0.024520302460771887,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.98568335555862e-05,
+      "loss": 2.1789,
+      "step": 561
+    },
+    {
+      "epoch": 0.024564010664801784,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.985631367287226e-05,
+      "loss": 2.3703,
+      "step": 562
+    },
+    {
+      "epoch": 0.02460771886883168,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.985579284929715e-05,
+      "loss": 2.5617,
+      "step": 563
+    },
+    {
+      "epoch": 0.024651427072861574,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.985527108487065e-05,
+      "loss": 2.1361,
+      "step": 564
+    },
+    {
+      "epoch": 0.02469513527689147,
+      "grad_norm": 4.125,
+      "learning_rate": 9.985474837960263e-05,
+      "loss": 2.3338,
+      "step": 565
+    },
+    {
+      "epoch": 0.02473884348092137,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.985422473350295e-05,
+      "loss": 2.0618,
+      "step": 566
+    },
+    {
+      "epoch": 0.024782551684951266,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.985370014658148e-05,
+      "loss": 2.133,
+      "step": 567
+    },
+    {
+      "epoch": 0.024826259888981163,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.985317461884814e-05,
+      "loss": 2.9443,
+      "step": 568
+    },
+    {
+      "epoch": 0.02486996809301106,
+      "grad_norm": 3.25,
+      "learning_rate": 9.985264815031283e-05,
+      "loss": 2.6085,
+      "step": 569
+    },
+    {
+      "epoch": 0.024913676297040954,
+      "grad_norm": 4.625,
+      "learning_rate": 9.98521207409855e-05,
+      "loss": 2.6604,
+      "step": 570
+    },
+    {
+      "epoch": 0.02495738450107085,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.985159239087609e-05,
+      "loss": 2.368,
+      "step": 571
+    },
+    {
+      "epoch": 0.025001092705100748,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.985106309999458e-05,
+      "loss": 2.2466,
+      "step": 572
+    },
+    {
+      "epoch": 0.025044800909130645,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.985053286835095e-05,
+      "loss": 2.3146,
+      "step": 573
+    },
+    {
+      "epoch": 0.025088509113160542,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.985000169595521e-05,
+      "loss": 2.3523,
+      "step": 574
+    },
+    {
+      "epoch": 0.025132217317190435,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.984946958281739e-05,
+      "loss": 2.6499,
+      "step": 575
+    },
+    {
+      "epoch": 0.025175925521220333,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.984893652894753e-05,
+      "loss": 1.8698,
+      "step": 576
+    },
+    {
+      "epoch": 0.02521963372525023,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.984840253435568e-05,
+      "loss": 2.1902,
+      "step": 577
+    },
+    {
+      "epoch": 0.025263341929280127,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.984786759905191e-05,
+      "loss": 2.1288,
+      "step": 578
+    },
+    {
+      "epoch": 0.025307050133310024,
+      "grad_norm": 3.25,
+      "learning_rate": 9.984733172304634e-05,
+      "loss": 2.2317,
+      "step": 579
+    },
+    {
+      "epoch": 0.025350758337339917,
+      "grad_norm": 10.75,
+      "learning_rate": 9.984679490634907e-05,
+      "loss": 2.4892,
+      "step": 580
+    },
+    {
+      "epoch": 0.025394466541369815,
+      "grad_norm": 4.75,
+      "learning_rate": 9.984625714897024e-05,
+      "loss": 2.6579,
+      "step": 581
+    },
+    {
+      "epoch": 0.02543817474539971,
+      "grad_norm": 3.375,
+      "learning_rate": 9.984571845091999e-05,
+      "loss": 2.2262,
+      "step": 582
+    },
+    {
+      "epoch": 0.02548188294942961,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.984517881220848e-05,
+      "loss": 2.4475,
+      "step": 583
+    },
+    {
+      "epoch": 0.025525591153459506,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.984463823284589e-05,
+      "loss": 2.2006,
+      "step": 584
+    },
+    {
+      "epoch": 0.0255692993574894,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.984409671284243e-05,
+      "loss": 2.0782,
+      "step": 585
+    },
+    {
+      "epoch": 0.025613007561519296,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.984355425220835e-05,
+      "loss": 2.7302,
+      "step": 586
+    },
+    {
+      "epoch": 0.025656715765549194,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.984301085095382e-05,
+      "loss": 2.5203,
+      "step": 587
+    },
+    {
+      "epoch": 0.02570042396957909,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.984246650908915e-05,
+      "loss": 2.5343,
+      "step": 588
+    },
+    {
+      "epoch": 0.025744132173608988,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.98419212266246e-05,
+      "loss": 2.0457,
+      "step": 589
+    },
+    {
+      "epoch": 0.02578784037763888,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.984137500357044e-05,
+      "loss": 2.2221,
+      "step": 590
+    },
+    {
+      "epoch": 0.02583154858166878,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.984082783993703e-05,
+      "loss": 2.0484,
+      "step": 591
+    },
+    {
+      "epoch": 0.025875256785698676,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.984027973573462e-05,
+      "loss": 2.1899,
+      "step": 592
+    },
+    {
+      "epoch": 0.025918964989728573,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.983973069097359e-05,
+      "loss": 2.5412,
+      "step": 593
+    },
+    {
+      "epoch": 0.02596267319375847,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.983918070566433e-05,
+      "loss": 2.7712,
+      "step": 594
+    },
+    {
+      "epoch": 0.026006381397788363,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.983862977981718e-05,
+      "loss": 2.521,
+      "step": 595
+    },
+    {
+      "epoch": 0.02605008960181826,
+      "grad_norm": 3.5,
+      "learning_rate": 9.983807791344255e-05,
+      "loss": 2.3225,
+      "step": 596
+    },
+    {
+      "epoch": 0.026093797805848157,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.983752510655084e-05,
+      "loss": 2.5047,
+      "step": 597
+    },
+    {
+      "epoch": 0.026137506009878055,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.983697135915252e-05,
+      "loss": 2.0369,
+      "step": 598
+    },
+    {
+      "epoch": 0.02618121421390795,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.9836416671258e-05,
+      "loss": 2.3224,
+      "step": 599
+    },
+    {
+      "epoch": 0.026224922417937845,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.983586104287778e-05,
+      "loss": 1.9921,
+      "step": 600
+    },
+    {
+      "epoch": 0.026268630621967742,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.983530447402231e-05,
+      "loss": 1.903,
+      "step": 601
+    },
+    {
+      "epoch": 0.02631233882599764,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.983474696470212e-05,
+      "loss": 2.197,
+      "step": 602
+    },
+    {
+      "epoch": 0.026356047030027537,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.983418851492773e-05,
+      "loss": 2.5317,
+      "step": 603
+    },
+    {
+      "epoch": 0.026399755234057434,
+      "grad_norm": 4.375,
+      "learning_rate": 9.983362912470966e-05,
+      "loss": 2.7071,
+      "step": 604
+    },
+    {
+      "epoch": 0.02644346343808733,
+      "grad_norm": 6.6875,
+      "learning_rate": 9.98330687940585e-05,
+      "loss": 2.3881,
+      "step": 605
+    },
+    {
+      "epoch": 0.026487171642117224,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.983250752298478e-05,
+      "loss": 2.201,
+      "step": 606
+    },
+    {
+      "epoch": 0.02653087984614712,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.983194531149914e-05,
+      "loss": 2.5151,
+      "step": 607
+    },
+    {
+      "epoch": 0.02657458805017702,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.983138215961214e-05,
+      "loss": 2.4005,
+      "step": 608
+    },
+    {
+      "epoch": 0.026618296254206916,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.983081806733444e-05,
+      "loss": 2.1401,
+      "step": 609
+    },
+    {
+      "epoch": 0.026662004458236813,
+      "grad_norm": 3.25,
+      "learning_rate": 9.983025303467668e-05,
+      "loss": 2.2363,
+      "step": 610
+    },
+    {
+      "epoch": 0.026705712662266706,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.982968706164953e-05,
+      "loss": 1.9678,
+      "step": 611
+    },
+    {
+      "epoch": 0.026749420866296603,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.982912014826365e-05,
+      "loss": 2.2302,
+      "step": 612
+    },
+    {
+      "epoch": 0.0267931290703265,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.982855229452975e-05,
+      "loss": 2.2146,
+      "step": 613
+    },
+    {
+      "epoch": 0.026836837274356398,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.982798350045854e-05,
+      "loss": 2.1537,
+      "step": 614
+    },
+    {
+      "epoch": 0.026880545478386295,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.982741376606078e-05,
+      "loss": 2.3415,
+      "step": 615
+    },
+    {
+      "epoch": 0.026924253682416188,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.982684309134719e-05,
+      "loss": 1.9102,
+      "step": 616
+    },
+    {
+      "epoch": 0.026967961886446085,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.982627147632855e-05,
+      "loss": 2.2455,
+      "step": 617
+    },
+    {
+      "epoch": 0.027011670090475982,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.982569892101565e-05,
+      "loss": 2.1046,
+      "step": 618
+    },
+    {
+      "epoch": 0.02705537829450588,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.982512542541929e-05,
+      "loss": 2.8151,
+      "step": 619
+    },
+    {
+      "epoch": 0.027099086498535777,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.98245509895503e-05,
+      "loss": 2.7252,
+      "step": 620
+    },
+    {
+      "epoch": 0.02714279470256567,
+      "grad_norm": 2.875,
+      "learning_rate": 9.982397561341952e-05,
+      "loss": 1.9278,
+      "step": 621
+    },
+    {
+      "epoch": 0.027186502906595567,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.982339929703781e-05,
+      "loss": 2.1459,
+      "step": 622
+    },
+    {
+      "epoch": 0.027230211110625464,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.982282204041604e-05,
+      "loss": 2.5963,
+      "step": 623
+    },
+    {
+      "epoch": 0.02727391931465536,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.982224384356508e-05,
+      "loss": 2.058,
+      "step": 624
+    },
+    {
+      "epoch": 0.02731762751868526,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.98216647064959e-05,
+      "loss": 2.4754,
+      "step": 625
+    },
+    {
+      "epoch": 0.027361335722715152,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.982108462921937e-05,
+      "loss": 2.964,
+      "step": 626
+    },
+    {
+      "epoch": 0.02740504392674505,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.982050361174647e-05,
+      "loss": 2.1477,
+      "step": 627
+    },
+    {
+      "epoch": 0.027448752130774946,
+      "grad_norm": 2.75,
+      "learning_rate": 9.981992165408816e-05,
+      "loss": 2.2316,
+      "step": 628
+    },
+    {
+      "epoch": 0.027492460334804843,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.981933875625542e-05,
+      "loss": 2.4692,
+      "step": 629
+    },
+    {
+      "epoch": 0.02753616853883474,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.981875491825924e-05,
+      "loss": 2.175,
+      "step": 630
+    },
+    {
+      "epoch": 0.027579876742864634,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.981817014011066e-05,
+      "loss": 2.2929,
+      "step": 631
+    },
+    {
+      "epoch": 0.02762358494689453,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.981758442182068e-05,
+      "loss": 2.5712,
+      "step": 632
+    },
+    {
+      "epoch": 0.027667293150924428,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.981699776340039e-05,
+      "loss": 1.9267,
+      "step": 633
+    },
+    {
+      "epoch": 0.027711001354954325,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.981641016486085e-05,
+      "loss": 2.7868,
+      "step": 634
+    },
+    {
+      "epoch": 0.027754709558984222,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.981582162621314e-05,
+      "loss": 2.5112,
+      "step": 635
+    },
+    {
+      "epoch": 0.027798417763014116,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.981523214746837e-05,
+      "loss": 2.2265,
+      "step": 636
+    },
+    {
+      "epoch": 0.027842125967044013,
+      "grad_norm": 4.125,
+      "learning_rate": 9.981464172863768e-05,
+      "loss": 2.7183,
+      "step": 637
+    },
+    {
+      "epoch": 0.02788583417107391,
+      "grad_norm": 5.0,
+      "learning_rate": 9.981405036973219e-05,
+      "loss": 2.5362,
+      "step": 638
+    },
+    {
+      "epoch": 0.027929542375103807,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.981345807076307e-05,
+      "loss": 3.1845,
+      "step": 639
+    },
+    {
+      "epoch": 0.027973250579133704,
+      "grad_norm": 3.25,
+      "learning_rate": 9.98128648317415e-05,
+      "loss": 2.0682,
+      "step": 640
+    },
+    {
+      "epoch": 0.0280169587831636,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.981227065267867e-05,
+      "loss": 2.0779,
+      "step": 641
+    },
+    {
+      "epoch": 0.028060666987193495,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.981167553358579e-05,
+      "loss": 2.5607,
+      "step": 642
+    },
+    {
+      "epoch": 0.028104375191223392,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.981107947447409e-05,
+      "loss": 2.2148,
+      "step": 643
+    },
+    {
+      "epoch": 0.02814808339525329,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.981048247535483e-05,
+      "loss": 2.6086,
+      "step": 644
+    },
+    {
+      "epoch": 0.028191791599283186,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.980988453623928e-05,
+      "loss": 2.6902,
+      "step": 645
+    },
+    {
+      "epoch": 0.028235499803313083,
+      "grad_norm": 4.9375,
+      "learning_rate": 9.98092856571387e-05,
+      "loss": 2.5818,
+      "step": 646
+    },
+    {
+      "epoch": 0.028279208007342977,
+      "grad_norm": 3.5,
+      "learning_rate": 9.98086858380644e-05,
+      "loss": 2.1797,
+      "step": 647
+    },
+    {
+      "epoch": 0.028322916211372874,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.980808507902773e-05,
+      "loss": 2.1088,
+      "step": 648
+    },
+    {
+      "epoch": 0.02836662441540277,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.980748338003998e-05,
+      "loss": 2.5253,
+      "step": 649
+    },
+    {
+      "epoch": 0.02841033261943267,
+      "grad_norm": 4.75,
+      "learning_rate": 9.980688074111253e-05,
+      "loss": 2.7148,
+      "step": 650
+    },
+    {
+      "epoch": 0.028454040823462565,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.980627716225675e-05,
+      "loss": 2.6032,
+      "step": 651
+    },
+    {
+      "epoch": 0.02849774902749246,
+      "grad_norm": 15.375,
+      "learning_rate": 9.980567264348404e-05,
+      "loss": 2.927,
+      "step": 652
+    },
+    {
+      "epoch": 0.028541457231522356,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.98050671848058e-05,
+      "loss": 2.7096,
+      "step": 653
+    },
+    {
+      "epoch": 0.028585165435552253,
+      "grad_norm": 6.0625,
+      "learning_rate": 9.980446078623345e-05,
+      "loss": 1.9591,
+      "step": 654
+    },
+    {
+      "epoch": 0.02862887363958215,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.980385344777842e-05,
+      "loss": 2.5283,
+      "step": 655
+    },
+    {
+      "epoch": 0.028672581843612047,
+      "grad_norm": 3.125,
+      "learning_rate": 9.980324516945221e-05,
+      "loss": 1.8712,
+      "step": 656
+    },
+    {
+      "epoch": 0.02871629004764194,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.980263595126629e-05,
+      "loss": 2.1822,
+      "step": 657
+    },
+    {
+      "epoch": 0.028759998251671838,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.980202579323212e-05,
+      "loss": 2.3666,
+      "step": 658
+    },
+    {
+      "epoch": 0.028803706455701735,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.980141469536125e-05,
+      "loss": 2.4666,
+      "step": 659
+    },
+    {
+      "epoch": 0.028847414659731632,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.98008026576652e-05,
+      "loss": 2.176,
+      "step": 660
+    },
+    {
+      "epoch": 0.02889112286376153,
+      "grad_norm": 3.25,
+      "learning_rate": 9.980018968015552e-05,
+      "loss": 1.8907,
+      "step": 661
+    },
+    {
+      "epoch": 0.028934831067791423,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.979957576284379e-05,
+      "loss": 1.7983,
+      "step": 662
+    },
+    {
+      "epoch": 0.02897853927182132,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.979896090574157e-05,
+      "loss": 3.0609,
+      "step": 663
+    },
+    {
+      "epoch": 0.029022247475851217,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.97983451088605e-05,
+      "loss": 2.1672,
+      "step": 664
+    },
+    {
+      "epoch": 0.029065955679881114,
+      "grad_norm": 10.0625,
+      "learning_rate": 9.979772837221216e-05,
+      "loss": 2.3753,
+      "step": 665
+    },
+    {
+      "epoch": 0.02910966388391101,
+      "grad_norm": 2.875,
+      "learning_rate": 9.979711069580821e-05,
+      "loss": 2.6742,
+      "step": 666
+    },
+    {
+      "epoch": 0.029153372087940905,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.979649207966031e-05,
+      "loss": 2.415,
+      "step": 667
+    },
+    {
+      "epoch": 0.029197080291970802,
+      "grad_norm": 5.78125,
+      "learning_rate": 9.979587252378013e-05,
+      "loss": 2.6886,
+      "step": 668
+    },
+    {
+      "epoch": 0.0292407884960007,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.979525202817936e-05,
+      "loss": 2.5185,
+      "step": 669
+    },
+    {
+      "epoch": 0.029284496700030596,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.979463059286972e-05,
+      "loss": 2.3531,
+      "step": 670
+    },
+    {
+      "epoch": 0.029328204904060493,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.979400821786291e-05,
+      "loss": 2.2063,
+      "step": 671
+    },
+    {
+      "epoch": 0.02937191310809039,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.979338490317072e-05,
+      "loss": 2.5651,
+      "step": 672
+    },
+    {
+      "epoch": 0.029415621312120284,
+      "grad_norm": 2.75,
+      "learning_rate": 9.979276064880486e-05,
+      "loss": 2.0795,
+      "step": 673
+    },
+    {
+      "epoch": 0.02945932951615018,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.979213545477715e-05,
+      "loss": 2.2151,
+      "step": 674
+    },
+    {
+      "epoch": 0.029503037720180078,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.979150932109937e-05,
+      "loss": 2.5187,
+      "step": 675
+    },
+    {
+      "epoch": 0.029546745924209975,
+      "grad_norm": 2.875,
+      "learning_rate": 9.979088224778335e-05,
+      "loss": 2.2578,
+      "step": 676
+    },
+    {
+      "epoch": 0.029590454128239872,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.97902542348409e-05,
+      "loss": 2.3404,
+      "step": 677
+    },
+    {
+      "epoch": 0.029634162332269766,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.978962528228388e-05,
+      "loss": 2.2639,
+      "step": 678
+    },
+    {
+      "epoch": 0.029677870536299663,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.978899539012418e-05,
+      "loss": 1.9094,
+      "step": 679
+    },
+    {
+      "epoch": 0.02972157874032956,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.978836455837368e-05,
+      "loss": 2.413,
+      "step": 680
+    },
+    {
+      "epoch": 0.029765286944359457,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.978773278704426e-05,
+      "loss": 2.5859,
+      "step": 681
+    },
+    {
+      "epoch": 0.029808995148389354,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.978710007614786e-05,
+      "loss": 2.1896,
+      "step": 682
+    },
+    {
+      "epoch": 0.029852703352419248,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.978646642569644e-05,
+      "loss": 2.4116,
+      "step": 683
+    },
+    {
+      "epoch": 0.029896411556449145,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.978583183570193e-05,
+      "loss": 2.1242,
+      "step": 684
+    },
+    {
+      "epoch": 0.029940119760479042,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.97851963061763e-05,
+      "loss": 2.1152,
+      "step": 685
+    },
+    {
+      "epoch": 0.02998382796450894,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.978455983713157e-05,
+      "loss": 2.6089,
+      "step": 686
+    },
+    {
+      "epoch": 0.030027536168538836,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.978392242857973e-05,
+      "loss": 2.4736,
+      "step": 687
+    },
+    {
+      "epoch": 0.03007124437256873,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.978328408053282e-05,
+      "loss": 2.5537,
+      "step": 688
+    },
+    {
+      "epoch": 0.030114952576598627,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.978264479300289e-05,
+      "loss": 1.7285,
+      "step": 689
+    },
+    {
+      "epoch": 0.030158660780628524,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.978200456600198e-05,
+      "loss": 2.3336,
+      "step": 690
+    },
+    {
+      "epoch": 0.03020236898465842,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.97813633995422e-05,
+      "loss": 1.8666,
+      "step": 691
+    },
+    {
+      "epoch": 0.030246077188688318,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.978072129363564e-05,
+      "loss": 2.2779,
+      "step": 692
+    },
+    {
+      "epoch": 0.030289785392718212,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.978007824829442e-05,
+      "loss": 2.6795,
+      "step": 693
+    },
+    {
+      "epoch": 0.03033349359674811,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.977943426353067e-05,
+      "loss": 2.7468,
+      "step": 694
+    },
+    {
+      "epoch": 0.030377201800778006,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.977878933935655e-05,
+      "loss": 2.453,
+      "step": 695
+    },
+    {
+      "epoch": 0.030420910004807903,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.977814347578421e-05,
+      "loss": 2.216,
+      "step": 696
+    },
+    {
+      "epoch": 0.0304646182088378,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.977749667282588e-05,
+      "loss": 2.5925,
+      "step": 697
+    },
+    {
+      "epoch": 0.030508326412867694,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.977684893049371e-05,
+      "loss": 2.7342,
+      "step": 698
+    },
+    {
+      "epoch": 0.03055203461689759,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.977620024879997e-05,
+      "loss": 2.2558,
+      "step": 699
+    },
+    {
+      "epoch": 0.030595742820927488,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.977555062775688e-05,
+      "loss": 2.322,
+      "step": 700
+    },
+    {
+      "epoch": 0.030639451024957385,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.97749000673767e-05,
+      "loss": 2.3314,
+      "step": 701
+    },
+    {
+      "epoch": 0.030683159228987282,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.977424856767172e-05,
+      "loss": 2.0297,
+      "step": 702
+    },
+    {
+      "epoch": 0.030726867433017176,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.977359612865423e-05,
+      "loss": 2.2477,
+      "step": 703
+    },
+    {
+      "epoch": 0.030770575637047073,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.977294275033654e-05,
+      "loss": 2.28,
+      "step": 704
+    },
+    {
+      "epoch": 0.03081428384107697,
+      "grad_norm": 11.4375,
+      "learning_rate": 9.977228843273098e-05,
+      "loss": 6.5621,
+      "step": 705
+    },
+    {
+      "epoch": 0.030857992045106867,
+      "grad_norm": 4.125,
+      "learning_rate": 9.977163317584988e-05,
+      "loss": 2.6921,
+      "step": 706
+    },
+    {
+      "epoch": 0.030901700249136764,
+      "grad_norm": 6.71875,
+      "learning_rate": 9.977097697970563e-05,
+      "loss": 2.1848,
+      "step": 707
+    },
+    {
+      "epoch": 0.03094540845316666,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.977031984431063e-05,
+      "loss": 2.3677,
+      "step": 708
+    },
+    {
+      "epoch": 0.030989116657196555,
+      "grad_norm": 4.0,
+      "learning_rate": 9.976966176967722e-05,
+      "loss": 2.5439,
+      "step": 709
+    },
+    {
+      "epoch": 0.031032824861226452,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.976900275581789e-05,
+      "loss": 2.305,
+      "step": 710
+    },
+    {
+      "epoch": 0.03107653306525635,
+      "grad_norm": 3.375,
+      "learning_rate": 9.976834280274502e-05,
+      "loss": 2.5969,
+      "step": 711
+    },
+    {
+      "epoch": 0.031120241269286246,
+      "grad_norm": 4.75,
+      "learning_rate": 9.976768191047109e-05,
+      "loss": 2.6965,
+      "step": 712
+    },
+    {
+      "epoch": 0.031163949473316143,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.976702007900857e-05,
+      "loss": 1.9579,
+      "step": 713
+    },
+    {
+      "epoch": 0.031207657677346037,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.976635730836995e-05,
+      "loss": 1.9571,
+      "step": 714
+    },
+    {
+      "epoch": 0.031251365881375934,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.976569359856773e-05,
+      "loss": 2.0944,
+      "step": 715
+    },
+    {
+      "epoch": 0.03129507408540583,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.976502894961445e-05,
+      "loss": 2.0834,
+      "step": 716
+    },
+    {
+      "epoch": 0.03133878228943573,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.976436336152265e-05,
+      "loss": 2.4847,
+      "step": 717
+    },
+    {
+      "epoch": 0.03138249049346562,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.976369683430487e-05,
+      "loss": 2.2834,
+      "step": 718
+    },
+    {
+      "epoch": 0.03142619869749552,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.976302936797371e-05,
+      "loss": 2.1861,
+      "step": 719
+    },
+    {
+      "epoch": 0.031469906901525416,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.976236096254177e-05,
+      "loss": 3.4581,
+      "step": 720
+    },
+    {
+      "epoch": 0.03151361510555531,
+      "grad_norm": 5.09375,
+      "learning_rate": 9.976169161802164e-05,
+      "loss": 2.4288,
+      "step": 721
+    },
+    {
+      "epoch": 0.03155732330958521,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.976102133442596e-05,
+      "loss": 2.3548,
+      "step": 722
+    },
+    {
+      "epoch": 0.031601031513615103,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.976035011176738e-05,
+      "loss": 2.1473,
+      "step": 723
+    },
+    {
+      "epoch": 0.031644739717645004,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.975967795005859e-05,
+      "loss": 2.6153,
+      "step": 724
+    },
+    {
+      "epoch": 0.0316884479216749,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.975900484931225e-05,
+      "loss": 2.9447,
+      "step": 725
+    },
+    {
+      "epoch": 0.0317321561257048,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.975833080954107e-05,
+      "loss": 2.2077,
+      "step": 726
+    },
+    {
+      "epoch": 0.03177586432973469,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.975765583075776e-05,
+      "loss": 2.2345,
+      "step": 727
+    },
+    {
+      "epoch": 0.031819572533764585,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.975697991297506e-05,
+      "loss": 2.1691,
+      "step": 728
+    },
+    {
+      "epoch": 0.031863280737794486,
+      "grad_norm": 3.25,
+      "learning_rate": 9.975630305620574e-05,
+      "loss": 2.5123,
+      "step": 729
+    },
+    {
+      "epoch": 0.03190698894182438,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.975562526046256e-05,
+      "loss": 2.519,
+      "step": 730
+    },
+    {
+      "epoch": 0.03195069714585428,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.975494652575832e-05,
+      "loss": 2.0585,
+      "step": 731
+    },
+    {
+      "epoch": 0.031994405349884174,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.975426685210582e-05,
+      "loss": 2.1076,
+      "step": 732
+    },
+    {
+      "epoch": 0.03203811355391407,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.975358623951789e-05,
+      "loss": 2.4889,
+      "step": 733
+    },
+    {
+      "epoch": 0.03208182175794397,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.975290468800739e-05,
+      "loss": 2.576,
+      "step": 734
+    },
+    {
+      "epoch": 0.03212552996197386,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.975222219758716e-05,
+      "loss": 2.4601,
+      "step": 735
+    },
+    {
+      "epoch": 0.03216923816600376,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.975153876827008e-05,
+      "loss": 2.4984,
+      "step": 736
+    },
+    {
+      "epoch": 0.032212946370033656,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.975085440006904e-05,
+      "loss": 2.5157,
+      "step": 737
+    },
+    {
+      "epoch": 0.03225665457406355,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.9750169092997e-05,
+      "loss": 1.9031,
+      "step": 738
+    },
+    {
+      "epoch": 0.03230036277809345,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.974948284706685e-05,
+      "loss": 2.3383,
+      "step": 739
+    },
+    {
+      "epoch": 0.032344070982123344,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.974879566229154e-05,
+      "loss": 2.1407,
+      "step": 740
+    },
+    {
+      "epoch": 0.032387779186153244,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.974810753868405e-05,
+      "loss": 2.9169,
+      "step": 741
+    },
+    {
+      "epoch": 0.03243148739018314,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.974741847625737e-05,
+      "loss": 2.4408,
+      "step": 742
+    },
+    {
+      "epoch": 0.03247519559421303,
+      "grad_norm": 4.0,
+      "learning_rate": 9.974672847502451e-05,
+      "loss": 2.501,
+      "step": 743
+    },
+    {
+      "epoch": 0.03251890379824293,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.974603753499847e-05,
+      "loss": 2.9058,
+      "step": 744
+    },
+    {
+      "epoch": 0.032562612002272825,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.97453456561923e-05,
+      "loss": 2.0979,
+      "step": 745
+    },
+    {
+      "epoch": 0.032606320206302726,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.974465283861905e-05,
+      "loss": 2.2657,
+      "step": 746
+    },
+    {
+      "epoch": 0.03265002841033262,
+      "grad_norm": 3.125,
+      "learning_rate": 9.974395908229181e-05,
+      "loss": 2.202,
+      "step": 747
+    },
+    {
+      "epoch": 0.03269373661436251,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.974326438722367e-05,
+      "loss": 2.4637,
+      "step": 748
+    },
+    {
+      "epoch": 0.032737444818392414,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.974256875342772e-05,
+      "loss": 2.0341,
+      "step": 749
+    },
+    {
+      "epoch": 0.03278115302242231,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.974187218091711e-05,
+      "loss": 2.1654,
+      "step": 750
+    },
+    {
+      "epoch": 0.03282486122645221,
+      "grad_norm": 5.125,
+      "learning_rate": 9.974117466970496e-05,
+      "loss": 3.255,
+      "step": 751
+    },
+    {
+      "epoch": 0.0328685694304821,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.974047621980447e-05,
+      "loss": 2.4065,
+      "step": 752
+    },
+    {
+      "epoch": 0.032912277634511995,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.97397768312288e-05,
+      "loss": 2.6438,
+      "step": 753
+    },
+    {
+      "epoch": 0.032955985838541896,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.973907650399113e-05,
+      "loss": 2.4801,
+      "step": 754
+    },
+    {
+      "epoch": 0.03299969404257179,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.973837523810471e-05,
+      "loss": 2.4131,
+      "step": 755
+    },
+    {
+      "epoch": 0.03304340224660169,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.973767303358275e-05,
+      "loss": 2.4873,
+      "step": 756
+    },
+    {
+      "epoch": 0.033087110450631584,
+      "grad_norm": 3.125,
+      "learning_rate": 9.973696989043852e-05,
+      "loss": 2.5531,
+      "step": 757
+    },
+    {
+      "epoch": 0.03313081865466148,
+      "grad_norm": 3.5,
+      "learning_rate": 9.973626580868527e-05,
+      "loss": 2.2931,
+      "step": 758
+    },
+    {
+      "epoch": 0.03317452685869138,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.97355607883363e-05,
+      "loss": 1.8566,
+      "step": 759
+    },
+    {
+      "epoch": 0.03321823506272127,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.97348548294049e-05,
+      "loss": 3.2503,
+      "step": 760
+    },
+    {
+      "epoch": 0.03326194326675117,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.973414793190441e-05,
+      "loss": 2.9089,
+      "step": 761
+    },
+    {
+      "epoch": 0.033305651470781066,
+      "grad_norm": 3.5,
+      "learning_rate": 9.973344009584818e-05,
+      "loss": 2.7956,
+      "step": 762
+    },
+    {
+      "epoch": 0.03334935967481096,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.973273132124954e-05,
+      "loss": 1.8721,
+      "step": 763
+    },
+    {
+      "epoch": 0.03339306787884086,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.973202160812187e-05,
+      "loss": 2.4582,
+      "step": 764
+    },
+    {
+      "epoch": 0.03343677608287075,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.973131095647858e-05,
+      "loss": 1.904,
+      "step": 765
+    },
+    {
+      "epoch": 0.033480484286900654,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.973059936633306e-05,
+      "loss": 2.0774,
+      "step": 766
+    },
+    {
+      "epoch": 0.03352419249093055,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.972988683769877e-05,
+      "loss": 2.5509,
+      "step": 767
+    },
+    {
+      "epoch": 0.03356790069496044,
+      "grad_norm": 4.15625,
+      "learning_rate": 9.972917337058913e-05,
+      "loss": 2.0877,
+      "step": 768
+    },
+    {
+      "epoch": 0.03361160889899034,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.972845896501761e-05,
+      "loss": 2.6669,
+      "step": 769
+    },
+    {
+      "epoch": 0.033655317103020235,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.972774362099768e-05,
+      "loss": 2.1274,
+      "step": 770
+    },
+    {
+      "epoch": 0.033699025307050136,
+      "grad_norm": 6.4375,
+      "learning_rate": 9.972702733854286e-05,
+      "loss": 2.6224,
+      "step": 771
+    },
+    {
+      "epoch": 0.03374273351108003,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.972631011766668e-05,
+      "loss": 2.4328,
+      "step": 772
+    },
+    {
+      "epoch": 0.03378644171510992,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.972559195838263e-05,
+      "loss": 2.3105,
+      "step": 773
+    },
+    {
+      "epoch": 0.033830149919139824,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.97248728607043e-05,
+      "loss": 1.9875,
+      "step": 774
+    },
+    {
+      "epoch": 0.03387385812316972,
+      "grad_norm": 6.21875,
+      "learning_rate": 9.972415282464524e-05,
+      "loss": 3.3821,
+      "step": 775
+    },
+    {
+      "epoch": 0.03391756632719962,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.972343185021906e-05,
+      "loss": 2.2215,
+      "step": 776
+    },
+    {
+      "epoch": 0.03396127453122951,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.972270993743934e-05,
+      "loss": 1.8742,
+      "step": 777
+    },
+    {
+      "epoch": 0.034004982735259405,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.972198708631972e-05,
+      "loss": 2.1591,
+      "step": 778
+    },
+    {
+      "epoch": 0.034048690939289306,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.972126329687384e-05,
+      "loss": 3.2081,
+      "step": 779
+    },
+    {
+      "epoch": 0.0340923991433192,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.972053856911534e-05,
+      "loss": 1.9423,
+      "step": 780
+    },
+    {
+      "epoch": 0.0341361073473491,
+      "grad_norm": 9.4375,
+      "learning_rate": 9.971981290305792e-05,
+      "loss": 1.6246,
+      "step": 781
+    },
+    {
+      "epoch": 0.03417981555137899,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.971908629871527e-05,
+      "loss": 2.1293,
+      "step": 782
+    },
+    {
+      "epoch": 0.03422352375540889,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.971835875610109e-05,
+      "loss": 2.006,
+      "step": 783
+    },
+    {
+      "epoch": 0.03426723195943879,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.971763027522914e-05,
+      "loss": 2.1156,
+      "step": 784
+    },
+    {
+      "epoch": 0.03431094016346868,
+      "grad_norm": 4.375,
+      "learning_rate": 9.971690085611314e-05,
+      "loss": 3.3214,
+      "step": 785
+    },
+    {
+      "epoch": 0.03435464836749858,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.971617049876684e-05,
+      "loss": 2.8245,
+      "step": 786
+    },
+    {
+      "epoch": 0.034398356571528475,
+      "grad_norm": 3.125,
+      "learning_rate": 9.971543920320407e-05,
+      "loss": 2.7465,
+      "step": 787
+    },
+    {
+      "epoch": 0.03444206477555837,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.971470696943859e-05,
+      "loss": 2.3266,
+      "step": 788
+    },
+    {
+      "epoch": 0.03448577297958827,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.971397379748424e-05,
+      "loss": 1.9464,
+      "step": 789
+    },
+    {
+      "epoch": 0.03452948118361816,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.971323968735484e-05,
+      "loss": 2.7519,
+      "step": 790
+    },
+    {
+      "epoch": 0.034573189387648064,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.971250463906426e-05,
+      "loss": 2.2776,
+      "step": 791
+    },
+    {
+      "epoch": 0.03461689759167796,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.971176865262635e-05,
+      "loss": 1.9055,
+      "step": 792
+    },
+    {
+      "epoch": 0.03466060579570786,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.971103172805503e-05,
+      "loss": 2.7596,
+      "step": 793
+    },
+    {
+      "epoch": 0.03470431399973775,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.971029386536419e-05,
+      "loss": 2.2475,
+      "step": 794
+    },
+    {
+      "epoch": 0.034748022203767645,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.970955506456776e-05,
+      "loss": 2.4918,
+      "step": 795
+    },
+    {
+      "epoch": 0.034791730407797546,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.970881532567967e-05,
+      "loss": 3.1343,
+      "step": 796
+    },
+    {
+      "epoch": 0.03483543861182744,
+      "grad_norm": 2.875,
+      "learning_rate": 9.970807464871387e-05,
+      "loss": 1.9314,
+      "step": 797
+    },
+    {
+      "epoch": 0.03487914681585734,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.970733303368438e-05,
+      "loss": 2.3322,
+      "step": 798
+    },
+    {
+      "epoch": 0.03492285501988723,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.970659048060515e-05,
+      "loss": 2.1469,
+      "step": 799
+    },
+    {
+      "epoch": 0.03496656322391713,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.970584698949023e-05,
+      "loss": 2.778,
+      "step": 800
+    },
+    {
+      "epoch": 0.03501027142794703,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.970510256035364e-05,
+      "loss": 2.5041,
+      "step": 801
+    },
+    {
+      "epoch": 0.03505397963197692,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.970435719320941e-05,
+      "loss": 2.6473,
+      "step": 802
+    },
+    {
+      "epoch": 0.03509768783600682,
+      "grad_norm": 5.0,
+      "learning_rate": 9.97036108880716e-05,
+      "loss": 2.6472,
+      "step": 803
+    },
+    {
+      "epoch": 0.035141396040036715,
+      "grad_norm": 3.5,
+      "learning_rate": 9.970286364495434e-05,
+      "loss": 2.0429,
+      "step": 804
+    },
+    {
+      "epoch": 0.03518510424406661,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.970211546387169e-05,
+      "loss": 2.8479,
+      "step": 805
+    },
+    {
+      "epoch": 0.03522881244809651,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.970136634483779e-05,
+      "loss": 2.3924,
+      "step": 806
+    },
+    {
+      "epoch": 0.0352725206521264,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.970061628786678e-05,
+      "loss": 2.8506,
+      "step": 807
+    },
+    {
+      "epoch": 0.035316228856156304,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.96998652929728e-05,
+      "loss": 1.8322,
+      "step": 808
+    },
+    {
+      "epoch": 0.0353599370601862,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.969911336017e-05,
+      "loss": 2.0667,
+      "step": 809
+    },
+    {
+      "epoch": 0.03540364526421609,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.969836048947263e-05,
+      "loss": 2.4905,
+      "step": 810
+    },
+    {
+      "epoch": 0.03544735346824599,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.969760668089486e-05,
+      "loss": 2.3651,
+      "step": 811
+    },
+    {
+      "epoch": 0.035491061672275885,
+      "grad_norm": 3.375,
+      "learning_rate": 9.969685193445091e-05,
+      "loss": 2.1256,
+      "step": 812
+    },
+    {
+      "epoch": 0.035534769876305786,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.969609625015502e-05,
+      "loss": 2.6686,
+      "step": 813
+    },
+    {
+      "epoch": 0.03557847808033568,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.96953396280215e-05,
+      "loss": 2.7127,
+      "step": 814
+    },
+    {
+      "epoch": 0.03562218628436557,
+      "grad_norm": 2.75,
+      "learning_rate": 9.969458206806456e-05,
+      "loss": 1.9587,
+      "step": 815
+    },
+    {
+      "epoch": 0.03566589448839547,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.969382357029856e-05,
+      "loss": 2.139,
+      "step": 816
+    },
+    {
+      "epoch": 0.03570960269242537,
+      "grad_norm": 5.84375,
+      "learning_rate": 9.969306413473776e-05,
+      "loss": 2.1272,
+      "step": 817
+    },
+    {
+      "epoch": 0.03575331089645527,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.969230376139651e-05,
+      "loss": 2.8055,
+      "step": 818
+    },
+    {
+      "epoch": 0.03579701910048516,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.969154245028917e-05,
+      "loss": 1.7156,
+      "step": 819
+    },
+    {
+      "epoch": 0.035840727304515055,
+      "grad_norm": 6.0,
+      "learning_rate": 9.969078020143012e-05,
+      "loss": 2.3962,
+      "step": 820
+    },
+    {
+      "epoch": 0.035884435508544955,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.969001701483372e-05,
+      "loss": 1.9315,
+      "step": 821
+    },
+    {
+      "epoch": 0.03592814371257485,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.968925289051436e-05,
+      "loss": 2.0727,
+      "step": 822
+    },
+    {
+      "epoch": 0.03597185191660475,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.96884878284865e-05,
+      "loss": 1.8745,
+      "step": 823
+    },
+    {
+      "epoch": 0.03601556012063464,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.968772182876454e-05,
+      "loss": 2.4207,
+      "step": 824
+    },
+    {
+      "epoch": 0.03605926832466454,
+      "grad_norm": 2.875,
+      "learning_rate": 9.968695489136296e-05,
+      "loss": 1.7213,
+      "step": 825
+    },
+    {
+      "epoch": 0.03610297652869444,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.968618701629623e-05,
+      "loss": 2.442,
+      "step": 826
+    },
+    {
+      "epoch": 0.03614668473272433,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.968541820357883e-05,
+      "loss": 2.593,
+      "step": 827
+    },
+    {
+      "epoch": 0.03619039293675423,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.968464845322527e-05,
+      "loss": 2.513,
+      "step": 828
+    },
+    {
+      "epoch": 0.036234101140784125,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.96838777652501e-05,
+      "loss": 2.9353,
+      "step": 829
+    },
+    {
+      "epoch": 0.03627780934481402,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.968310613966783e-05,
+      "loss": 2.7721,
+      "step": 830
+    },
+    {
+      "epoch": 0.03632151754884392,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.968233357649303e-05,
+      "loss": 1.9949,
+      "step": 831
+    },
+    {
+      "epoch": 0.03636522575287381,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.96815600757403e-05,
+      "loss": 1.9257,
+      "step": 832
+    },
+    {
+      "epoch": 0.036408933956903713,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.968078563742421e-05,
+      "loss": 2.6905,
+      "step": 833
+    },
+    {
+      "epoch": 0.03645264216093361,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.968001026155939e-05,
+      "loss": 1.9883,
+      "step": 834
+    },
+    {
+      "epoch": 0.0364963503649635,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.967923394816049e-05,
+      "loss": 3.019,
+      "step": 835
+    },
+    {
+      "epoch": 0.0365400585689934,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.967845669724212e-05,
+      "loss": 2.1879,
+      "step": 836
+    },
+    {
+      "epoch": 0.036583766773023295,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.967767850881898e-05,
+      "loss": 2.161,
+      "step": 837
+    },
+    {
+      "epoch": 0.036627474977053195,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.967689938290574e-05,
+      "loss": 2.8632,
+      "step": 838
+    },
+    {
+      "epoch": 0.03667118318108309,
+      "grad_norm": 5.1875,
+      "learning_rate": 9.96761193195171e-05,
+      "loss": 2.971,
+      "step": 839
+    },
+    {
+      "epoch": 0.03671489138511298,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.96753383186678e-05,
+      "loss": 2.5535,
+      "step": 840
+    },
+    {
+      "epoch": 0.03675859958914288,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.967455638037257e-05,
+      "loss": 2.0311,
+      "step": 841
+    },
+    {
+      "epoch": 0.03680230779317278,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.967377350464615e-05,
+      "loss": 1.9492,
+      "step": 842
+    },
+    {
+      "epoch": 0.03684601599720268,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.967298969150334e-05,
+      "loss": 2.2931,
+      "step": 843
+    },
+    {
+      "epoch": 0.03688972420123257,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.96722049409589e-05,
+      "loss": 1.7512,
+      "step": 844
+    },
+    {
+      "epoch": 0.036933432405262465,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.967141925302768e-05,
+      "loss": 2.3825,
+      "step": 845
+    },
+    {
+      "epoch": 0.036977140609292365,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.967063262772447e-05,
+      "loss": 2.2133,
+      "step": 846
+    },
+    {
+      "epoch": 0.03702084881332226,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.966984506506413e-05,
+      "loss": 2.1161,
+      "step": 847
+    },
+    {
+      "epoch": 0.03706455701735216,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.966905656506154e-05,
+      "loss": 2.2078,
+      "step": 848
+    },
+    {
+      "epoch": 0.03710826522138205,
+      "grad_norm": 3.625,
+      "learning_rate": 9.966826712773155e-05,
+      "loss": 2.5632,
+      "step": 849
+    },
+    {
+      "epoch": 0.03715197342541195,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.966747675308907e-05,
+      "loss": 2.3147,
+      "step": 850
+    },
+    {
+      "epoch": 0.03719568162944185,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.966668544114902e-05,
+      "loss": 2.1258,
+      "step": 851
+    },
+    {
+      "epoch": 0.03723938983347174,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.966589319192633e-05,
+      "loss": 2.3756,
+      "step": 852
+    },
+    {
+      "epoch": 0.03728309803750164,
+      "grad_norm": 4.125,
+      "learning_rate": 9.966510000543594e-05,
+      "loss": 2.3031,
+      "step": 853
+    },
+    {
+      "epoch": 0.037326806241531535,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.966430588169285e-05,
+      "loss": 2.0828,
+      "step": 854
+    },
+    {
+      "epoch": 0.03737051444556143,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.966351082071201e-05,
+      "loss": 2.4494,
+      "step": 855
+    },
+    {
+      "epoch": 0.03741422264959133,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.966271482250845e-05,
+      "loss": 2.3533,
+      "step": 856
+    },
+    {
+      "epoch": 0.03745793085362122,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.966191788709716e-05,
+      "loss": 3.1356,
+      "step": 857
+    },
+    {
+      "epoch": 0.03750163905765112,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.966112001449321e-05,
+      "loss": 2.1622,
+      "step": 858
+    },
+    {
+      "epoch": 0.03754534726168102,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.966032120471165e-05,
+      "loss": 2.6896,
+      "step": 859
+    },
+    {
+      "epoch": 0.03758905546571091,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.965952145776754e-05,
+      "loss": 2.0189,
+      "step": 860
+    },
+    {
+      "epoch": 0.03763276366974081,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.9658720773676e-05,
+      "loss": 2.3372,
+      "step": 861
+    },
+    {
+      "epoch": 0.037676471873770705,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.96579191524521e-05,
+      "loss": 2.0823,
+      "step": 862
+    },
+    {
+      "epoch": 0.037720180077800605,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.965711659411102e-05,
+      "loss": 2.822,
+      "step": 863
+    },
+    {
+      "epoch": 0.0377638882818305,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.965631309866788e-05,
+      "loss": 2.695,
+      "step": 864
+    },
+    {
+      "epoch": 0.0378075964858604,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.965550866613782e-05,
+      "loss": 2.7255,
+      "step": 865
+    },
+    {
+      "epoch": 0.03785130468989029,
+      "grad_norm": 3.25,
+      "learning_rate": 9.965470329653604e-05,
+      "loss": 2.4268,
+      "step": 866
+    },
+    {
+      "epoch": 0.03789501289392019,
+      "grad_norm": 4.78125,
+      "learning_rate": 9.965389698987775e-05,
+      "loss": 2.2879,
+      "step": 867
+    },
+    {
+      "epoch": 0.03793872109795009,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.965308974617816e-05,
+      "loss": 2.4958,
+      "step": 868
+    },
+    {
+      "epoch": 0.03798242930197998,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.96522815654525e-05,
+      "loss": 2.387,
+      "step": 869
+    },
+    {
+      "epoch": 0.03802613750600988,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.965147244771601e-05,
+      "loss": 1.918,
+      "step": 870
+    },
+    {
+      "epoch": 0.038069845710039775,
+      "grad_norm": 3.25,
+      "learning_rate": 9.965066239298398e-05,
+      "loss": 2.7065,
+      "step": 871
+    },
+    {
+      "epoch": 0.03811355391406967,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.964985140127168e-05,
+      "loss": 2.3908,
+      "step": 872
+    },
+    {
+      "epoch": 0.03815726211809957,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.964903947259443e-05,
+      "loss": 1.9557,
+      "step": 873
+    },
+    {
+      "epoch": 0.03820097032212946,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.964822660696753e-05,
+      "loss": 2.1748,
+      "step": 874
+    },
+    {
+      "epoch": 0.03824467852615936,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.964741280440634e-05,
+      "loss": 1.9483,
+      "step": 875
+    },
+    {
+      "epoch": 0.03828838673018926,
+      "grad_norm": 8.25,
+      "learning_rate": 9.964659806492624e-05,
+      "loss": 4.1064,
+      "step": 876
+    },
+    {
+      "epoch": 0.03833209493421915,
+      "grad_norm": 4.625,
+      "learning_rate": 9.964578238854254e-05,
+      "loss": 2.6809,
+      "step": 877
+    },
+    {
+      "epoch": 0.03837580313824905,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.964496577527069e-05,
+      "loss": 1.873,
+      "step": 878
+    },
+    {
+      "epoch": 0.038419511342278945,
+      "grad_norm": 4.875,
+      "learning_rate": 9.964414822512607e-05,
+      "loss": 2.5123,
+      "step": 879
+    },
+    {
+      "epoch": 0.038463219546308845,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.964332973812414e-05,
+      "loss": 2.3094,
+      "step": 880
+    },
+    {
+      "epoch": 0.03850692775033874,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.96425103142803e-05,
+      "loss": 1.6306,
+      "step": 881
+    },
+    {
+      "epoch": 0.03855063595436863,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.964168995361005e-05,
+      "loss": 2.1723,
+      "step": 882
+    },
+    {
+      "epoch": 0.03859434415839853,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.964086865612887e-05,
+      "loss": 2.1385,
+      "step": 883
+    },
+    {
+      "epoch": 0.03863805236242843,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.964004642185223e-05,
+      "loss": 2.2899,
+      "step": 884
+    },
+    {
+      "epoch": 0.03868176056645833,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.963922325079567e-05,
+      "loss": 2.6071,
+      "step": 885
+    },
+    {
+      "epoch": 0.03872546877048822,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.963839914297473e-05,
+      "loss": 2.1812,
+      "step": 886
+    },
+    {
+      "epoch": 0.038769176974518114,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.963757409840495e-05,
+      "loss": 2.4299,
+      "step": 887
+    },
+    {
+      "epoch": 0.038812885178548015,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.963674811710191e-05,
+      "loss": 2.3144,
+      "step": 888
+    },
+    {
+      "epoch": 0.03885659338257791,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.963592119908119e-05,
+      "loss": 2.3799,
+      "step": 889
+    },
+    {
+      "epoch": 0.03890030158660781,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.963509334435838e-05,
+      "loss": 2.4469,
+      "step": 890
+    },
+    {
+      "epoch": 0.0389440097906377,
+      "grad_norm": 3.875,
+      "learning_rate": 9.963426455294913e-05,
+      "loss": 2.425,
+      "step": 891
+    },
+    {
+      "epoch": 0.038987717994667596,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.963343482486906e-05,
+      "loss": 2.833,
+      "step": 892
+    },
+    {
+      "epoch": 0.0390314261986975,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.963260416013385e-05,
+      "loss": 2.5376,
+      "step": 893
+    },
+    {
+      "epoch": 0.03907513440272739,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.963177255875915e-05,
+      "loss": 2.5258,
+      "step": 894
+    },
+    {
+      "epoch": 0.03911884260675729,
+      "grad_norm": 7.4375,
+      "learning_rate": 9.963094002076066e-05,
+      "loss": 3.2763,
+      "step": 895
+    },
+    {
+      "epoch": 0.039162550810787185,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.963010654615411e-05,
+      "loss": 3.1825,
+      "step": 896
+    },
+    {
+      "epoch": 0.03920625901481708,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.962927213495521e-05,
+      "loss": 2.2391,
+      "step": 897
+    },
+    {
+      "epoch": 0.03924996721884698,
+      "grad_norm": 3.125,
+      "learning_rate": 9.962843678717972e-05,
+      "loss": 2.2213,
+      "step": 898
+    },
+    {
+      "epoch": 0.03929367542287687,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.96276005028434e-05,
+      "loss": 2.3569,
+      "step": 899
+    },
+    {
+      "epoch": 0.03933738362690677,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.962676328196202e-05,
+      "loss": 1.9611,
+      "step": 900
+    },
+    {
+      "epoch": 0.03938109183093667,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.962592512455138e-05,
+      "loss": 1.869,
+      "step": 901
+    },
+    {
+      "epoch": 0.03942480003496656,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.962508603062732e-05,
+      "loss": 2.3317,
+      "step": 902
+    },
+    {
+      "epoch": 0.03946850823899646,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.962424600020567e-05,
+      "loss": 1.9824,
+      "step": 903
+    },
+    {
+      "epoch": 0.039512216443026354,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.962340503330226e-05,
+      "loss": 2.6484,
+      "step": 904
+    },
+    {
+      "epoch": 0.039555924647056255,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.962256312993297e-05,
+      "loss": 2.3466,
+      "step": 905
+    },
+    {
+      "epoch": 0.03959963285108615,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.96217202901137e-05,
+      "loss": 2.0264,
+      "step": 906
+    },
+    {
+      "epoch": 0.03964334105511604,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.962087651386034e-05,
+      "loss": 2.5601,
+      "step": 907
+    },
+    {
+      "epoch": 0.03968704925914594,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.962003180118882e-05,
+      "loss": 2.2833,
+      "step": 908
+    },
+    {
+      "epoch": 0.039730757463175836,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.96191861521151e-05,
+      "loss": 2.6963,
+      "step": 909
+    },
+    {
+      "epoch": 0.03977446566720574,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.961833956665509e-05,
+      "loss": 2.1049,
+      "step": 910
+    },
+    {
+      "epoch": 0.03981817387123563,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.961749204482483e-05,
+      "loss": 3.2084,
+      "step": 911
+    },
+    {
+      "epoch": 0.039861882075265524,
+      "grad_norm": 2.75,
+      "learning_rate": 9.961664358664024e-05,
+      "loss": 2.6011,
+      "step": 912
+    },
+    {
+      "epoch": 0.039905590279295425,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.961579419211741e-05,
+      "loss": 2.6555,
+      "step": 913
+    },
+    {
+      "epoch": 0.03994929848332532,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.961494386127231e-05,
+      "loss": 1.9832,
+      "step": 914
+    },
+    {
+      "epoch": 0.03999300668735522,
+      "grad_norm": 4.5,
+      "learning_rate": 9.961409259412102e-05,
+      "loss": 2.9627,
+      "step": 915
+    },
+    {
+      "epoch": 0.04003671489138511,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.961324039067958e-05,
+      "loss": 2.9962,
+      "step": 916
+    },
+    {
+      "epoch": 0.040080423095415006,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.961238725096409e-05,
+      "loss": 2.2526,
+      "step": 917
+    },
+    {
+      "epoch": 0.04012413129944491,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.961153317499065e-05,
+      "loss": 2.5092,
+      "step": 918
+    },
+    {
+      "epoch": 0.0401678395034748,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.961067816277536e-05,
+      "loss": 1.9513,
+      "step": 919
+    },
+    {
+      "epoch": 0.0402115477075047,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.960982221433439e-05,
+      "loss": 2.3512,
+      "step": 920
+    },
+    {
+      "epoch": 0.040255255911534595,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.960896532968385e-05,
+      "loss": 2.5303,
+      "step": 921
+    },
+    {
+      "epoch": 0.04029896411556449,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.960810750883995e-05,
+      "loss": 1.8781,
+      "step": 922
+    },
+    {
+      "epoch": 0.04034267231959439,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.960724875181885e-05,
+      "loss": 2.3986,
+      "step": 923
+    },
+    {
+      "epoch": 0.04038638052362428,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.960638905863676e-05,
+      "loss": 2.0327,
+      "step": 924
+    },
+    {
+      "epoch": 0.04043008872765418,
+      "grad_norm": 6.15625,
+      "learning_rate": 9.960552842930992e-05,
+      "loss": 1.6378,
+      "step": 925
+    },
+    {
+      "epoch": 0.040473796931684076,
+      "grad_norm": 9.0,
+      "learning_rate": 9.960466686385456e-05,
+      "loss": 2.1269,
+      "step": 926
+    },
+    {
+      "epoch": 0.04051750513571397,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.960380436228693e-05,
+      "loss": 2.3509,
+      "step": 927
+    },
+    {
+      "epoch": 0.04056121333974387,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.960294092462332e-05,
+      "loss": 2.1872,
+      "step": 928
+    },
+    {
+      "epoch": 0.040604921543773764,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.960207655088003e-05,
+      "loss": 2.3827,
+      "step": 929
+    },
+    {
+      "epoch": 0.040648629747803665,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.960121124107336e-05,
+      "loss": 1.9355,
+      "step": 930
+    },
+    {
+      "epoch": 0.04069233795183356,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.960034499521964e-05,
+      "loss": 1.8716,
+      "step": 931
+    },
+    {
+      "epoch": 0.04073604615586346,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.959947781333522e-05,
+      "loss": 1.9923,
+      "step": 932
+    },
+    {
+      "epoch": 0.04077975435989335,
+      "grad_norm": 3.0,
+      "learning_rate": 9.959860969543648e-05,
+      "loss": 2.2197,
+      "step": 933
+    },
+    {
+      "epoch": 0.040823462563923246,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.959774064153977e-05,
+      "loss": 2.2632,
+      "step": 934
+    },
+    {
+      "epoch": 0.04086717076795315,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.959687065166151e-05,
+      "loss": 2.3649,
+      "step": 935
+    },
+    {
+      "epoch": 0.04091087897198304,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.959599972581812e-05,
+      "loss": 2.6295,
+      "step": 936
+    },
+    {
+      "epoch": 0.04095458717601294,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.959512786402603e-05,
+      "loss": 2.1709,
+      "step": 937
+    },
+    {
+      "epoch": 0.040998295380042835,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.95942550663017e-05,
+      "loss": 2.1453,
+      "step": 938
+    },
+    {
+      "epoch": 0.04104200358407273,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.959338133266158e-05,
+      "loss": 2.5455,
+      "step": 939
+    },
+    {
+      "epoch": 0.04108571178810263,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.959250666312219e-05,
+      "loss": 2.0837,
+      "step": 940
+    },
+    {
+      "epoch": 0.04112941999213252,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.959163105770002e-05,
+      "loss": 2.2131,
+      "step": 941
+    },
+    {
+      "epoch": 0.04117312819616242,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.959075451641159e-05,
+      "loss": 2.1649,
+      "step": 942
+    },
+    {
+      "epoch": 0.04121683640019232,
+      "grad_norm": 3.5,
+      "learning_rate": 9.958987703927345e-05,
+      "loss": 2.245,
+      "step": 943
+    },
+    {
+      "epoch": 0.04126054460422221,
+      "grad_norm": 3.0,
+      "learning_rate": 9.958899862630216e-05,
+      "loss": 2.1328,
+      "step": 944
+    },
+    {
+      "epoch": 0.04130425280825211,
+      "grad_norm": 3.25,
+      "learning_rate": 9.958811927751428e-05,
+      "loss": 2.9265,
+      "step": 945
+    },
+    {
+      "epoch": 0.041347961012282004,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.958723899292641e-05,
+      "loss": 2.4016,
+      "step": 946
+    },
+    {
+      "epoch": 0.041391669216311905,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.958635777255518e-05,
+      "loss": 2.2132,
+      "step": 947
+    },
+    {
+      "epoch": 0.0414353774203418,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.958547561641722e-05,
+      "loss": 2.1233,
+      "step": 948
+    },
+    {
+      "epoch": 0.04147908562437169,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.958459252452916e-05,
+      "loss": 2.8399,
+      "step": 949
+    },
+    {
+      "epoch": 0.04152279382840159,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.958370849690767e-05,
+      "loss": 2.7092,
+      "step": 950
+    },
+    {
+      "epoch": 0.041566502032431486,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.958282353356943e-05,
+      "loss": 2.0668,
+      "step": 951
+    },
+    {
+      "epoch": 0.04161021023646139,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.958193763453116e-05,
+      "loss": 2.2635,
+      "step": 952
+    },
+    {
+      "epoch": 0.04165391844049128,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.958105079980954e-05,
+      "loss": 2.0128,
+      "step": 953
+    },
+    {
+      "epoch": 0.041697626644521174,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.958016302942135e-05,
+      "loss": 2.1925,
+      "step": 954
+    },
+    {
+      "epoch": 0.041741334848551075,
+      "grad_norm": 3.25,
+      "learning_rate": 9.957927432338332e-05,
+      "loss": 2.0548,
+      "step": 955
+    },
+    {
+      "epoch": 0.04178504305258097,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.957838468171223e-05,
+      "loss": 2.3297,
+      "step": 956
+    },
+    {
+      "epoch": 0.04182875125661087,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.957749410442485e-05,
+      "loss": 2.3113,
+      "step": 957
+    },
+    {
+      "epoch": 0.04187245946064076,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.9576602591538e-05,
+      "loss": 2.4494,
+      "step": 958
+    },
+    {
+      "epoch": 0.041916167664670656,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.957571014306852e-05,
+      "loss": 2.2834,
+      "step": 959
+    },
+    {
+      "epoch": 0.04195987586870056,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.957481675903321e-05,
+      "loss": 1.9757,
+      "step": 960
+    },
+    {
+      "epoch": 0.04200358407273045,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.957392243944898e-05,
+      "loss": 1.9881,
+      "step": 961
+    },
+    {
+      "epoch": 0.04204729227676035,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.957302718433266e-05,
+      "loss": 2.0498,
+      "step": 962
+    },
+    {
+      "epoch": 0.042091000480790244,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.957213099370117e-05,
+      "loss": 2.2652,
+      "step": 963
+    },
+    {
+      "epoch": 0.04213470868482014,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.957123386757144e-05,
+      "loss": 2.2596,
+      "step": 964
+    },
+    {
+      "epoch": 0.04217841688885004,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.957033580596036e-05,
+      "loss": 2.7782,
+      "step": 965
+    },
+    {
+      "epoch": 0.04222212509287993,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.95694368088849e-05,
+      "loss": 2.0644,
+      "step": 966
+    },
+    {
+      "epoch": 0.04226583329690983,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.956853687636203e-05,
+      "loss": 2.149,
+      "step": 967
+    },
+    {
+      "epoch": 0.042309541500939726,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.956763600840873e-05,
+      "loss": 2.8036,
+      "step": 968
+    },
+    {
+      "epoch": 0.04235324970496962,
+      "grad_norm": 3.125,
+      "learning_rate": 9.9566734205042e-05,
+      "loss": 1.844,
+      "step": 969
+    },
+    {
+      "epoch": 0.04239695790899952,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.956583146627883e-05,
+      "loss": 3.511,
+      "step": 970
+    },
+    {
+      "epoch": 0.042440666113029414,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.956492779213629e-05,
+      "loss": 2.1949,
+      "step": 971
+    },
+    {
+      "epoch": 0.042484374317059315,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.956402318263143e-05,
+      "loss": 2.4102,
+      "step": 972
+    },
+    {
+      "epoch": 0.04252808252108921,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.95631176377813e-05,
+      "loss": 2.4406,
+      "step": 973
+    },
+    {
+      "epoch": 0.0425717907251191,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.956221115760302e-05,
+      "loss": 2.7419,
+      "step": 974
+    },
+    {
+      "epoch": 0.042615498929149,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.956130374211369e-05,
+      "loss": 2.8455,
+      "step": 975
+    },
+    {
+      "epoch": 0.042659207133178896,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.956039539133042e-05,
+      "loss": 2.433,
+      "step": 976
+    },
+    {
+      "epoch": 0.0427029153372088,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.955948610527035e-05,
+      "loss": 2.2266,
+      "step": 977
+    },
+    {
+      "epoch": 0.04274662354123869,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.955857588395065e-05,
+      "loss": 2.2863,
+      "step": 978
+    },
+    {
+      "epoch": 0.042790331745268584,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.955766472738847e-05,
+      "loss": 1.9107,
+      "step": 979
+    },
+    {
+      "epoch": 0.042834039949298484,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.955675263560107e-05,
+      "loss": 1.9685,
+      "step": 980
+    },
+    {
+      "epoch": 0.04287774815332838,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.95558396086056e-05,
+      "loss": 1.9,
+      "step": 981
+    },
+    {
+      "epoch": 0.04292145635735828,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.955492564641931e-05,
+      "loss": 1.9955,
+      "step": 982
+    },
+    {
+      "epoch": 0.04296516456138817,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.955401074905945e-05,
+      "loss": 2.3646,
+      "step": 983
+    },
+    {
+      "epoch": 0.043008872765418066,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.95530949165433e-05,
+      "loss": 1.7669,
+      "step": 984
+    },
+    {
+      "epoch": 0.043052580969447966,
+      "grad_norm": 3.125,
+      "learning_rate": 9.955217814888811e-05,
+      "loss": 2.2699,
+      "step": 985
+    },
+    {
+      "epoch": 0.04309628917347786,
+      "grad_norm": 6.0625,
+      "learning_rate": 9.955126044611121e-05,
+      "loss": 2.6444,
+      "step": 986
+    },
+    {
+      "epoch": 0.04313999737750776,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.955034180822988e-05,
+      "loss": 2.1632,
+      "step": 987
+    },
+    {
+      "epoch": 0.043183705581537654,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.954942223526152e-05,
+      "loss": 1.9997,
+      "step": 988
+    },
+    {
+      "epoch": 0.04322741378556755,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.954850172722344e-05,
+      "loss": 2.6184,
+      "step": 989
+    },
+    {
+      "epoch": 0.04327112198959745,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.9547580284133e-05,
+      "loss": 2.7852,
+      "step": 990
+    },
+    {
+      "epoch": 0.04331483019362734,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.954665790600761e-05,
+      "loss": 2.9001,
+      "step": 991
+    },
+    {
+      "epoch": 0.04335853839765724,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.954573459286468e-05,
+      "loss": 2.6454,
+      "step": 992
+    },
+    {
+      "epoch": 0.043402246601687136,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.954481034472163e-05,
+      "loss": 2.3826,
+      "step": 993
+    },
+    {
+      "epoch": 0.04344595480571703,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.95438851615959e-05,
+      "loss": 2.5778,
+      "step": 994
+    },
+    {
+      "epoch": 0.04348966300974693,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.954295904350495e-05,
+      "loss": 2.5694,
+      "step": 995
+    },
+    {
+      "epoch": 0.043533371213776824,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.954203199046624e-05,
+      "loss": 2.0787,
+      "step": 996
+    },
+    {
+      "epoch": 0.043577079417806724,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.95411040024973e-05,
+      "loss": 2.1673,
+      "step": 997
+    },
+    {
+      "epoch": 0.04362078762183662,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.954017507961561e-05,
+      "loss": 2.8185,
+      "step": 998
+    },
+    {
+      "epoch": 0.04366449582586651,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.953924522183872e-05,
+      "loss": 2.495,
+      "step": 999
+    },
+    {
+      "epoch": 0.04370820402989641,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.953831442918418e-05,
+      "loss": 1.975,
+      "step": 1000
+    },
+    {
+      "epoch": 0.043751912233926306,
+      "grad_norm": 3.375,
+      "learning_rate": 9.953738270166954e-05,
+      "loss": 2.072,
+      "step": 1001
+    },
+    {
+      "epoch": 0.043795620437956206,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.953645003931239e-05,
+      "loss": 2.2948,
+      "step": 1002
+    },
+    {
+      "epoch": 0.0438393286419861,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.953551644213033e-05,
+      "loss": 2.4571,
+      "step": 1003
+    },
+    {
+      "epoch": 0.043883036846016,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.953458191014098e-05,
+      "loss": 2.5195,
+      "step": 1004
+    },
+    {
+      "epoch": 0.043926745050045894,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.953364644336199e-05,
+      "loss": 2.5952,
+      "step": 1005
+    },
+    {
+      "epoch": 0.04397045325407579,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.953271004181097e-05,
+      "loss": 2.1683,
+      "step": 1006
+    },
+    {
+      "epoch": 0.04401416145810569,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.953177270550564e-05,
+      "loss": 1.818,
+      "step": 1007
+    },
+    {
+      "epoch": 0.04405786966213558,
+      "grad_norm": 3.375,
+      "learning_rate": 9.953083443446366e-05,
+      "loss": 2.5001,
+      "step": 1008
+    },
+    {
+      "epoch": 0.04410157786616548,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.952989522870275e-05,
+      "loss": 2.2333,
+      "step": 1009
+    },
+    {
+      "epoch": 0.044145286070195376,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.952895508824063e-05,
+      "loss": 2.1663,
+      "step": 1010
+    },
+    {
+      "epoch": 0.04418899427422527,
+      "grad_norm": 16.25,
+      "learning_rate": 9.952801401309503e-05,
+      "loss": 2.8942,
+      "step": 1011
+    },
+    {
+      "epoch": 0.04423270247825517,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.952707200328374e-05,
+      "loss": 1.8845,
+      "step": 1012
+    },
+    {
+      "epoch": 0.044276410682285064,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.952612905882451e-05,
+      "loss": 2.5532,
+      "step": 1013
+    },
+    {
+      "epoch": 0.044320118886314964,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.952518517973515e-05,
+      "loss": 2.4217,
+      "step": 1014
+    },
+    {
+      "epoch": 0.04436382709034486,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.952424036603345e-05,
+      "loss": 1.7717,
+      "step": 1015
+    },
+    {
+      "epoch": 0.04440753529437475,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.952329461773726e-05,
+      "loss": 2.0673,
+      "step": 1016
+    },
+    {
+      "epoch": 0.04445124349840465,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.952234793486443e-05,
+      "loss": 1.9221,
+      "step": 1017
+    },
+    {
+      "epoch": 0.044494951702434546,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.95214003174328e-05,
+      "loss": 2.2079,
+      "step": 1018
+    },
+    {
+      "epoch": 0.044538659906464446,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.95204517654603e-05,
+      "loss": 2.5225,
+      "step": 1019
+    },
+    {
+      "epoch": 0.04458236811049434,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.951950227896478e-05,
+      "loss": 2.343,
+      "step": 1020
+    },
+    {
+      "epoch": 0.044626076314524234,
+      "grad_norm": 7.15625,
+      "learning_rate": 9.95185518579642e-05,
+      "loss": 2.4092,
+      "step": 1021
+    },
+    {
+      "epoch": 0.044669784518554134,
+      "grad_norm": 3.75,
+      "learning_rate": 9.951760050247646e-05,
+      "loss": 2.3665,
+      "step": 1022
+    },
+    {
+      "epoch": 0.04471349272258403,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.951664821251952e-05,
+      "loss": 2.2941,
+      "step": 1023
+    },
+    {
+      "epoch": 0.04475720092661393,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.951569498811137e-05,
+      "loss": 2.8114,
+      "step": 1024
+    },
+    {
+      "epoch": 0.04480090913064382,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.951474082927e-05,
+      "loss": 2.7321,
+      "step": 1025
+    },
+    {
+      "epoch": 0.044844617334673716,
+      "grad_norm": 3.0,
+      "learning_rate": 9.951378573601338e-05,
+      "loss": 2.2982,
+      "step": 1026
+    },
+    {
+      "epoch": 0.044888325538703616,
+      "grad_norm": 3.25,
+      "learning_rate": 9.951282970835957e-05,
+      "loss": 2.0281,
+      "step": 1027
+    },
+    {
+      "epoch": 0.04493203374273351,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.951187274632661e-05,
+      "loss": 2.1427,
+      "step": 1028
+    },
+    {
+      "epoch": 0.04497574194676341,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.951091484993256e-05,
+      "loss": 2.1392,
+      "step": 1029
+    },
+    {
+      "epoch": 0.045019450150793304,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.950995601919546e-05,
+      "loss": 2.1285,
+      "step": 1030
+    },
+    {
+      "epoch": 0.0450631583548232,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.950899625413345e-05,
+      "loss": 1.7744,
+      "step": 1031
+    },
+    {
+      "epoch": 0.0451068665588531,
+      "grad_norm": 4.875,
+      "learning_rate": 9.950803555476463e-05,
+      "loss": 3.7282,
+      "step": 1032
+    },
+    {
+      "epoch": 0.04515057476288299,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.950707392110709e-05,
+      "loss": 3.1688,
+      "step": 1033
+    },
+    {
+      "epoch": 0.04519428296691289,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.950611135317904e-05,
+      "loss": 1.9093,
+      "step": 1034
+    },
+    {
+      "epoch": 0.045237991170942786,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.95051478509986e-05,
+      "loss": 2.6153,
+      "step": 1035
+    },
+    {
+      "epoch": 0.04528169937497268,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.950418341458398e-05,
+      "loss": 2.0785,
+      "step": 1036
+    },
+    {
+      "epoch": 0.04532540757900258,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.950321804395338e-05,
+      "loss": 2.2215,
+      "step": 1037
+    },
+    {
+      "epoch": 0.045369115783032474,
+      "grad_norm": 3.625,
+      "learning_rate": 9.950225173912499e-05,
+      "loss": 3.471,
+      "step": 1038
+    },
+    {
+      "epoch": 0.045412823987062374,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.950128450011706e-05,
+      "loss": 2.7834,
+      "step": 1039
+    },
+    {
+      "epoch": 0.04545653219109227,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.950031632694785e-05,
+      "loss": 2.182,
+      "step": 1040
+    },
+    {
+      "epoch": 0.04550024039512216,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.949934721963563e-05,
+      "loss": 2.1175,
+      "step": 1041
+    },
+    {
+      "epoch": 0.04554394859915206,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.949837717819868e-05,
+      "loss": 2.6425,
+      "step": 1042
+    },
+    {
+      "epoch": 0.045587656803181956,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.949740620265532e-05,
+      "loss": 2.0681,
+      "step": 1043
+    },
+    {
+      "epoch": 0.045631365007211856,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.949643429302386e-05,
+      "loss": 3.4625,
+      "step": 1044
+    },
+    {
+      "epoch": 0.04567507321124175,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.949546144932265e-05,
+      "loss": 1.965,
+      "step": 1045
+    },
+    {
+      "epoch": 0.04571878141527164,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.949448767157003e-05,
+      "loss": 2.3129,
+      "step": 1046
+    },
+    {
+      "epoch": 0.045762489619301544,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.949351295978441e-05,
+      "loss": 1.81,
+      "step": 1047
+    },
+    {
+      "epoch": 0.04580619782333144,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.949253731398416e-05,
+      "loss": 2.5846,
+      "step": 1048
+    },
+    {
+      "epoch": 0.04584990602736134,
+      "grad_norm": 4.25,
+      "learning_rate": 9.949156073418769e-05,
+      "loss": 2.8508,
+      "step": 1049
+    },
+    {
+      "epoch": 0.04589361423139123,
+      "grad_norm": 2.75,
+      "learning_rate": 9.949058322041345e-05,
+      "loss": 2.2982,
+      "step": 1050
+    },
+    {
+      "epoch": 0.045937322435421125,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.948960477267986e-05,
+      "loss": 3.0965,
+      "step": 1051
+    },
+    {
+      "epoch": 0.045981030639451026,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.948862539100541e-05,
+      "loss": 2.16,
+      "step": 1052
+    },
+    {
+      "epoch": 0.04602473884348092,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.948764507540858e-05,
+      "loss": 2.1271,
+      "step": 1053
+    },
+    {
+      "epoch": 0.04606844704751082,
+      "grad_norm": 2.875,
+      "learning_rate": 9.948666382590785e-05,
+      "loss": 2.0429,
+      "step": 1054
+    },
+    {
+      "epoch": 0.046112155251540714,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.948568164252174e-05,
+      "loss": 2.3033,
+      "step": 1055
+    },
+    {
+      "epoch": 0.04615586345557061,
+      "grad_norm": 7.3125,
+      "learning_rate": 9.948469852526881e-05,
+      "loss": 2.4171,
+      "step": 1056
+    },
+    {
+      "epoch": 0.04619957165960051,
+      "grad_norm": 3.5,
+      "learning_rate": 9.948371447416758e-05,
+      "loss": 2.5804,
+      "step": 1057
+    },
+    {
+      "epoch": 0.0462432798636304,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.948272948923666e-05,
+      "loss": 2.2676,
+      "step": 1058
+    },
+    {
+      "epoch": 0.0462869880676603,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.94817435704946e-05,
+      "loss": 2.327,
+      "step": 1059
+    },
+    {
+      "epoch": 0.046330696271690196,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.948075671796004e-05,
+      "loss": 1.9911,
+      "step": 1060
+    },
+    {
+      "epoch": 0.04637440447572009,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.947976893165156e-05,
+      "loss": 2.0437,
+      "step": 1061
+    },
+    {
+      "epoch": 0.04641811267974999,
+      "grad_norm": 3.0,
+      "learning_rate": 9.947878021158784e-05,
+      "loss": 2.4868,
+      "step": 1062
+    },
+    {
+      "epoch": 0.046461820883779884,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.947779055778752e-05,
+      "loss": 2.313,
+      "step": 1063
+    },
+    {
+      "epoch": 0.046505529087809784,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.947679997026929e-05,
+      "loss": 2.0247,
+      "step": 1064
+    },
+    {
+      "epoch": 0.04654923729183968,
+      "grad_norm": 12.3125,
+      "learning_rate": 9.947580844905182e-05,
+      "loss": 3.1176,
+      "step": 1065
+    },
+    {
+      "epoch": 0.04659294549586957,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.947481599415384e-05,
+      "loss": 2.522,
+      "step": 1066
+    },
+    {
+      "epoch": 0.04663665369989947,
+      "grad_norm": 3.5,
+      "learning_rate": 9.947382260559408e-05,
+      "loss": 2.9242,
+      "step": 1067
+    },
+    {
+      "epoch": 0.046680361903929365,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.947282828339128e-05,
+      "loss": 1.9872,
+      "step": 1068
+    },
+    {
+      "epoch": 0.046724070107959266,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.947183302756422e-05,
+      "loss": 2.146,
+      "step": 1069
+    },
+    {
+      "epoch": 0.04676777831198916,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.947083683813165e-05,
+      "loss": 2.3677,
+      "step": 1070
+    },
+    {
+      "epoch": 0.04681148651601906,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.946983971511239e-05,
+      "loss": 2.5506,
+      "step": 1071
+    },
+    {
+      "epoch": 0.046855194720048954,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.946884165852525e-05,
+      "loss": 2.0168,
+      "step": 1072
+    },
+    {
+      "epoch": 0.04689890292407885,
+      "grad_norm": 2.625,
+      "learning_rate": 9.946784266838909e-05,
+      "loss": 1.9957,
+      "step": 1073
+    },
+    {
+      "epoch": 0.04694261112810875,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.94668427447227e-05,
+      "loss": 2.315,
+      "step": 1074
+    },
+    {
+      "epoch": 0.04698631933213864,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.946584188754504e-05,
+      "loss": 2.1152,
+      "step": 1075
+    },
+    {
+      "epoch": 0.04703002753616854,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.946484009687493e-05,
+      "loss": 2.0469,
+      "step": 1076
+    },
+    {
+      "epoch": 0.047073735740198436,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.946383737273129e-05,
+      "loss": 2.7682,
+      "step": 1077
+    },
+    {
+      "epoch": 0.04711744394422833,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.946283371513305e-05,
+      "loss": 3.337,
+      "step": 1078
+    },
+    {
+      "epoch": 0.04716115214825823,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.946182912409915e-05,
+      "loss": 2.2733,
+      "step": 1079
+    },
+    {
+      "epoch": 0.047204860352288124,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.946082359964855e-05,
+      "loss": 2.0077,
+      "step": 1080
+    },
+    {
+      "epoch": 0.047248568556318024,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.945981714180021e-05,
+      "loss": 2.0198,
+      "step": 1081
+    },
+    {
+      "epoch": 0.04729227676034792,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.945880975057315e-05,
+      "loss": 1.9136,
+      "step": 1082
+    },
+    {
+      "epoch": 0.04733598496437781,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.945780142598636e-05,
+      "loss": 2.7276,
+      "step": 1083
+    },
+    {
+      "epoch": 0.04737969316840771,
+      "grad_norm": 3.875,
+      "learning_rate": 9.945679216805887e-05,
+      "loss": 2.2242,
+      "step": 1084
+    },
+    {
+      "epoch": 0.047423401372437606,
+      "grad_norm": 3.0,
+      "learning_rate": 9.945578197680974e-05,
+      "loss": 2.5395,
+      "step": 1085
+    },
+    {
+      "epoch": 0.047467109576467506,
+      "grad_norm": 2.75,
+      "learning_rate": 9.9454770852258e-05,
+      "loss": 2.159,
+      "step": 1086
+    },
+    {
+      "epoch": 0.0475108177804974,
+      "grad_norm": 3.25,
+      "learning_rate": 9.945375879442277e-05,
+      "loss": 1.941,
+      "step": 1087
+    },
+    {
+      "epoch": 0.04755452598452729,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.945274580332316e-05,
+      "loss": 2.7804,
+      "step": 1088
+    },
+    {
+      "epoch": 0.047598234188557194,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.945173187897823e-05,
+      "loss": 2.6826,
+      "step": 1089
+    },
+    {
+      "epoch": 0.04764194239258709,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.945071702140716e-05,
+      "loss": 2.7705,
+      "step": 1090
+    },
+    {
+      "epoch": 0.04768565059661699,
+      "grad_norm": 3.375,
+      "learning_rate": 9.94497012306291e-05,
+      "loss": 2.5417,
+      "step": 1091
+    },
+    {
+      "epoch": 0.04772935880064688,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.944868450666318e-05,
+      "loss": 3.0292,
+      "step": 1092
+    },
+    {
+      "epoch": 0.047773067004676775,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.944766684952863e-05,
+      "loss": 2.1641,
+      "step": 1093
+    },
+    {
+      "epoch": 0.047816775208706676,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.944664825924463e-05,
+      "loss": 2.5355,
+      "step": 1094
+    },
+    {
+      "epoch": 0.04786048341273657,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.944562873583042e-05,
+      "loss": 2.6627,
+      "step": 1095
+    },
+    {
+      "epoch": 0.04790419161676647,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.944460827930521e-05,
+      "loss": 2.0701,
+      "step": 1096
+    },
+    {
+      "epoch": 0.047947899820796364,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.944358688968831e-05,
+      "loss": 2.3189,
+      "step": 1097
+    },
+    {
+      "epoch": 0.04799160802482626,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.944256456699895e-05,
+      "loss": 2.0858,
+      "step": 1098
+    },
+    {
+      "epoch": 0.04803531622885616,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.944154131125642e-05,
+      "loss": 2.1744,
+      "step": 1099
+    },
+    {
+      "epoch": 0.04807902443288605,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.944051712248009e-05,
+      "loss": 2.2945,
+      "step": 1100
+    },
+    {
+      "epoch": 0.04812273263691595,
+      "grad_norm": 4.0,
+      "learning_rate": 9.943949200068921e-05,
+      "loss": 3.715,
+      "step": 1101
+    },
+    {
+      "epoch": 0.048166440840945846,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.943846594590316e-05,
+      "loss": 2.1754,
+      "step": 1102
+    },
+    {
+      "epoch": 0.04821014904497574,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.943743895814131e-05,
+      "loss": 2.2889,
+      "step": 1103
+    },
+    {
+      "epoch": 0.04825385724900564,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.943641103742303e-05,
+      "loss": 2.7616,
+      "step": 1104
+    },
+    {
+      "epoch": 0.04829756545303553,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.943538218376773e-05,
+      "loss": 2.4096,
+      "step": 1105
+    },
+    {
+      "epoch": 0.048341273657065434,
+      "grad_norm": 2.75,
+      "learning_rate": 9.943435239719482e-05,
+      "loss": 2.522,
+      "step": 1106
+    },
+    {
+      "epoch": 0.04838498186109533,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.943332167772372e-05,
+      "loss": 2.0895,
+      "step": 1107
+    },
+    {
+      "epoch": 0.04842869006512522,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.943229002537391e-05,
+      "loss": 2.8614,
+      "step": 1108
+    },
+    {
+      "epoch": 0.04847239826915512,
+      "grad_norm": 3.125,
+      "learning_rate": 9.943125744016483e-05,
+      "loss": 2.4255,
+      "step": 1109
+    },
+    {
+      "epoch": 0.048516106473185015,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.943022392211599e-05,
+      "loss": 2.4272,
+      "step": 1110
+    },
+    {
+      "epoch": 0.048559814677214916,
+      "grad_norm": 3.625,
+      "learning_rate": 9.942918947124686e-05,
+      "loss": 3.2011,
+      "step": 1111
+    },
+    {
+      "epoch": 0.04860352288124481,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.942815408757699e-05,
+      "loss": 1.8,
+      "step": 1112
+    },
+    {
+      "epoch": 0.0486472310852747,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.942711777112594e-05,
+      "loss": 2.1547,
+      "step": 1113
+    },
+    {
+      "epoch": 0.048690939289304604,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.942608052191321e-05,
+      "loss": 2.4241,
+      "step": 1114
+    },
+    {
+      "epoch": 0.0487346474933345,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.942504233995842e-05,
+      "loss": 1.774,
+      "step": 1115
+    },
+    {
+      "epoch": 0.0487783556973644,
+      "grad_norm": 7.46875,
+      "learning_rate": 9.942400322528114e-05,
+      "loss": 2.6383,
+      "step": 1116
+    },
+    {
+      "epoch": 0.04882206390139429,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.942296317790099e-05,
+      "loss": 1.941,
+      "step": 1117
+    },
+    {
+      "epoch": 0.048865772105424185,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.942192219783758e-05,
+      "loss": 1.9,
+      "step": 1118
+    },
+    {
+      "epoch": 0.048909480309454086,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.942088028511059e-05,
+      "loss": 2.382,
+      "step": 1119
+    },
+    {
+      "epoch": 0.04895318851348398,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.941983743973964e-05,
+      "loss": 2.1542,
+      "step": 1120
+    },
+    {
+      "epoch": 0.04899689671751388,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.941879366174444e-05,
+      "loss": 2.5014,
+      "step": 1121
+    },
+    {
+      "epoch": 0.04904060492154377,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.941774895114467e-05,
+      "loss": 2.2071,
+      "step": 1122
+    },
+    {
+      "epoch": 0.04908431312557367,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.941670330796007e-05,
+      "loss": 2.6761,
+      "step": 1123
+    },
+    {
+      "epoch": 0.04912802132960357,
+      "grad_norm": 6.0625,
+      "learning_rate": 9.941565673221034e-05,
+      "loss": 2.7923,
+      "step": 1124
+    },
+    {
+      "epoch": 0.04917172953363346,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.941460922391526e-05,
+      "loss": 2.3441,
+      "step": 1125
+    },
+    {
+      "epoch": 0.04921543773766336,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.941356078309457e-05,
+      "loss": 2.1081,
+      "step": 1126
+    },
+    {
+      "epoch": 0.049259145941693255,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.941251140976807e-05,
+      "loss": 1.9443,
+      "step": 1127
+    },
+    {
+      "epoch": 0.04930285414572315,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.941146110395557e-05,
+      "loss": 2.0367,
+      "step": 1128
+    },
+    {
+      "epoch": 0.04934656234975305,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.941040986567689e-05,
+      "loss": 2.5013,
+      "step": 1129
+    },
+    {
+      "epoch": 0.04939027055378294,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.940935769495186e-05,
+      "loss": 1.9259,
+      "step": 1130
+    },
+    {
+      "epoch": 0.049433978757812844,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.940830459180033e-05,
+      "loss": 1.9416,
+      "step": 1131
+    },
+    {
+      "epoch": 0.04947768696184274,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.940725055624218e-05,
+      "loss": 2.3391,
+      "step": 1132
+    },
+    {
+      "epoch": 0.04952139516587263,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.940619558829731e-05,
+      "loss": 2.6132,
+      "step": 1133
+    },
+    {
+      "epoch": 0.04956510336990253,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.940513968798563e-05,
+      "loss": 2.4921,
+      "step": 1134
+    },
+    {
+      "epoch": 0.049608811573932425,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.940408285532705e-05,
+      "loss": 2.0902,
+      "step": 1135
+    },
+    {
+      "epoch": 0.049652519777962326,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.940302509034152e-05,
+      "loss": 2.2618,
+      "step": 1136
+    },
+    {
+      "epoch": 0.04969622798199222,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.9401966393049e-05,
+      "loss": 2.1943,
+      "step": 1137
+    },
+    {
+      "epoch": 0.04973993618602212,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.940090676346948e-05,
+      "loss": 2.2639,
+      "step": 1138
+    },
+    {
+      "epoch": 0.04978364439005201,
+      "grad_norm": 3.0,
+      "learning_rate": 9.939984620162295e-05,
+      "loss": 2.3306,
+      "step": 1139
+    },
+    {
+      "epoch": 0.04982735259408191,
+      "grad_norm": 5.375,
+      "learning_rate": 9.939878470752942e-05,
+      "loss": 1.9926,
+      "step": 1140
+    },
+    {
+      "epoch": 0.04987106079811181,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.939772228120893e-05,
+      "loss": 2.2294,
+      "step": 1141
+    },
+    {
+      "epoch": 0.0499147690021417,
+      "grad_norm": 2.75,
+      "learning_rate": 9.939665892268152e-05,
+      "loss": 1.99,
+      "step": 1142
+    },
+    {
+      "epoch": 0.0499584772061716,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.939559463196727e-05,
+      "loss": 2.5241,
+      "step": 1143
+    },
+    {
+      "epoch": 0.050002185410201495,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.939452940908626e-05,
+      "loss": 2.1868,
+      "step": 1144
+    },
+    {
+      "epoch": 0.05004589361423139,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.939346325405858e-05,
+      "loss": 3.2205,
+      "step": 1145
+    },
+    {
+      "epoch": 0.05008960181826129,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.939239616690436e-05,
+      "loss": 3.0481,
+      "step": 1146
+    },
+    {
+      "epoch": 0.05013331002229118,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.939132814764375e-05,
+      "loss": 2.1522,
+      "step": 1147
+    },
+    {
+      "epoch": 0.050177018226321084,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.939025919629688e-05,
+      "loss": 2.2517,
+      "step": 1148
+    },
+    {
+      "epoch": 0.05022072643035098,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.938918931288395e-05,
+      "loss": 1.72,
+      "step": 1149
+    },
+    {
+      "epoch": 0.05026443463438087,
+      "grad_norm": 3.125,
+      "learning_rate": 9.93881184974251e-05,
+      "loss": 2.559,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05030814283841077,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.938704674994062e-05,
+      "loss": 2.2228,
+      "step": 1151
+    },
+    {
+      "epoch": 0.050351851042440665,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.938597407045065e-05,
+      "loss": 2.6145,
+      "step": 1152
+    },
+    {
+      "epoch": 0.050395559246470566,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.93849004589755e-05,
+      "loss": 2.0967,
+      "step": 1153
+    },
+    {
+      "epoch": 0.05043926745050046,
+      "grad_norm": 4.125,
+      "learning_rate": 9.938382591553538e-05,
+      "loss": 2.0268,
+      "step": 1154
+    },
+    {
+      "epoch": 0.05048297565453035,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.938275044015059e-05,
+      "loss": 2.994,
+      "step": 1155
+    },
+    {
+      "epoch": 0.05052668385856025,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.938167403284144e-05,
+      "loss": 3.0414,
+      "step": 1156
+    },
+    {
+      "epoch": 0.05057039206259015,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.938059669362822e-05,
+      "loss": 2.5288,
+      "step": 1157
+    },
+    {
+      "epoch": 0.05061410026662005,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.937951842253127e-05,
+      "loss": 1.7965,
+      "step": 1158
+    },
+    {
+      "epoch": 0.05065780847064994,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.937843921957095e-05,
+      "loss": 2.2143,
+      "step": 1159
+    },
+    {
+      "epoch": 0.050701516674679835,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.937735908476762e-05,
+      "loss": 2.3242,
+      "step": 1160
+    },
+    {
+      "epoch": 0.050745224878709735,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.937627801814165e-05,
+      "loss": 1.9423,
+      "step": 1161
+    },
+    {
+      "epoch": 0.05078893308273963,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.937519601971343e-05,
+      "loss": 2.6635,
+      "step": 1162
+    },
+    {
+      "epoch": 0.05083264128676953,
+      "grad_norm": 6.1875,
+      "learning_rate": 9.937411308950342e-05,
+      "loss": 2.6874,
+      "step": 1163
+    },
+    {
+      "epoch": 0.05087634949079942,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.937302922753203e-05,
+      "loss": 2.2795,
+      "step": 1164
+    },
+    {
+      "epoch": 0.05092005769482932,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.937194443381972e-05,
+      "loss": 1.9133,
+      "step": 1165
+    },
+    {
+      "epoch": 0.05096376589885922,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.937085870838695e-05,
+      "loss": 1.9703,
+      "step": 1166
+    },
+    {
+      "epoch": 0.05100747410288911,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.936977205125424e-05,
+      "loss": 1.8399,
+      "step": 1167
+    },
+    {
+      "epoch": 0.05105118230691901,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.936868446244208e-05,
+      "loss": 2.7845,
+      "step": 1168
+    },
+    {
+      "epoch": 0.051094890510948905,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.936759594197098e-05,
+      "loss": 2.5411,
+      "step": 1169
+    },
+    {
+      "epoch": 0.0511385987149788,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.936650648986148e-05,
+      "loss": 2.3818,
+      "step": 1170
+    },
+    {
+      "epoch": 0.0511823069190087,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.936541610613416e-05,
+      "loss": 2.4754,
+      "step": 1171
+    },
+    {
+      "epoch": 0.05122601512303859,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.936432479080961e-05,
+      "loss": 1.9356,
+      "step": 1172
+    },
+    {
+      "epoch": 0.051269723327068494,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.93632325439084e-05,
+      "loss": 2.1584,
+      "step": 1173
+    },
+    {
+      "epoch": 0.05131343153109839,
+      "grad_norm": 2.5,
+      "learning_rate": 9.936213936545113e-05,
+      "loss": 2.1896,
+      "step": 1174
+    },
+    {
+      "epoch": 0.05135713973512828,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.936104525545846e-05,
+      "loss": 2.0162,
+      "step": 1175
+    },
+    {
+      "epoch": 0.05140084793915818,
+      "grad_norm": 2.625,
+      "learning_rate": 9.935995021395102e-05,
+      "loss": 2.0853,
+      "step": 1176
+    },
+    {
+      "epoch": 0.051444556143188075,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.935885424094948e-05,
+      "loss": 1.9196,
+      "step": 1177
+    },
+    {
+      "epoch": 0.051488264347217975,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.935775733647452e-05,
+      "loss": 2.0876,
+      "step": 1178
+    },
+    {
+      "epoch": 0.05153197255124787,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.935665950054684e-05,
+      "loss": 2.6171,
+      "step": 1179
+    },
+    {
+      "epoch": 0.05157568075527776,
+      "grad_norm": 2.875,
+      "learning_rate": 9.935556073318716e-05,
+      "loss": 2.4073,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05161938895930766,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.935446103441623e-05,
+      "loss": 2.6037,
+      "step": 1181
+    },
+    {
+      "epoch": 0.05166309716333756,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.935336040425478e-05,
+      "loss": 2.2674,
+      "step": 1182
+    },
+    {
+      "epoch": 0.05170680536736746,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.935225884272359e-05,
+      "loss": 3.6372,
+      "step": 1183
+    },
+    {
+      "epoch": 0.05175051357139735,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.935115634984345e-05,
+      "loss": 2.1184,
+      "step": 1184
+    },
+    {
+      "epoch": 0.051794221775427245,
+      "grad_norm": 6.875,
+      "learning_rate": 9.935005292563515e-05,
+      "loss": 2.7781,
+      "step": 1185
+    },
+    {
+      "epoch": 0.051837929979457145,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.934894857011953e-05,
+      "loss": 2.1325,
+      "step": 1186
+    },
+    {
+      "epoch": 0.05188163818348704,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.934784328331743e-05,
+      "loss": 2.4299,
+      "step": 1187
+    },
+    {
+      "epoch": 0.05192534638751694,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.934673706524969e-05,
+      "loss": 2.3853,
+      "step": 1188
+    },
+    {
+      "epoch": 0.05196905459154683,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.934562991593722e-05,
+      "loss": 1.9624,
+      "step": 1189
+    },
+    {
+      "epoch": 0.05201276279557673,
+      "grad_norm": 8.625,
+      "learning_rate": 9.934452183540089e-05,
+      "loss": 2.052,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05205647099960663,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.934341282366162e-05,
+      "loss": 2.0976,
+      "step": 1191
+    },
+    {
+      "epoch": 0.05210017920363652,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.934230288074032e-05,
+      "loss": 2.397,
+      "step": 1192
+    },
+    {
+      "epoch": 0.05214388740766642,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.934119200665795e-05,
+      "loss": 2.1238,
+      "step": 1193
+    },
+    {
+      "epoch": 0.052187595611696315,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.934008020143548e-05,
+      "loss": 2.1589,
+      "step": 1194
+    },
+    {
+      "epoch": 0.05223130381572621,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.933896746509391e-05,
+      "loss": 2.0372,
+      "step": 1195
+    },
+    {
+      "epoch": 0.05227501201975611,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.933785379765417e-05,
+      "loss": 2.1988,
+      "step": 1196
+    },
+    {
+      "epoch": 0.052318720223786,
+      "grad_norm": 3.25,
+      "learning_rate": 9.933673919913735e-05,
+      "loss": 2.2642,
+      "step": 1197
+    },
+    {
+      "epoch": 0.0523624284278159,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.933562366956445e-05,
+      "loss": 1.705,
+      "step": 1198
+    },
+    {
+      "epoch": 0.0524061366318458,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.933450720895651e-05,
+      "loss": 2.5751,
+      "step": 1199
+    },
+    {
+      "epoch": 0.05244984483587569,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.933338981733464e-05,
+      "loss": 2.3266,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05249355303990559,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.933227149471991e-05,
+      "loss": 2.8926,
+      "step": 1201
+    },
+    {
+      "epoch": 0.052537261243935485,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.933115224113338e-05,
+      "loss": 2.0426,
+      "step": 1202
+    },
+    {
+      "epoch": 0.052580969447965385,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.933003205659623e-05,
+      "loss": 2.0112,
+      "step": 1203
+    },
+    {
+      "epoch": 0.05262467765199528,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.93289109411296e-05,
+      "loss": 1.9949,
+      "step": 1204
+    },
+    {
+      "epoch": 0.05266838585602517,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.93277888947546e-05,
+      "loss": 2.1734,
+      "step": 1205
+    },
+    {
+      "epoch": 0.05271209406005507,
+      "grad_norm": 3.25,
+      "learning_rate": 9.932666591749242e-05,
+      "loss": 3.1198,
+      "step": 1206
+    },
+    {
+      "epoch": 0.05275580226408497,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.932554200936429e-05,
+      "loss": 2.1966,
+      "step": 1207
+    },
+    {
+      "epoch": 0.05279951046811487,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.932441717039138e-05,
+      "loss": 2.2443,
+      "step": 1208
+    },
+    {
+      "epoch": 0.05284321867214476,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.932329140059494e-05,
+      "loss": 2.2764,
+      "step": 1209
+    },
+    {
+      "epoch": 0.05288692687617466,
+      "grad_norm": 3.375,
+      "learning_rate": 9.932216469999618e-05,
+      "loss": 2.5796,
+      "step": 1210
+    },
+    {
+      "epoch": 0.052930635080204555,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.93210370686164e-05,
+      "loss": 2.1995,
+      "step": 1211
+    },
+    {
+      "epoch": 0.05297434328423445,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.931990850647688e-05,
+      "loss": 2.3949,
+      "step": 1212
+    },
+    {
+      "epoch": 0.05301805148826435,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.931877901359888e-05,
+      "loss": 1.9585,
+      "step": 1213
+    },
+    {
+      "epoch": 0.05306175969229424,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.931764859000375e-05,
+      "loss": 2.6056,
+      "step": 1214
+    },
+    {
+      "epoch": 0.05310546789632414,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.931651723571282e-05,
+      "loss": 2.7266,
+      "step": 1215
+    },
+    {
+      "epoch": 0.05314917610035404,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.931538495074743e-05,
+      "loss": 2.2353,
+      "step": 1216
+    },
+    {
+      "epoch": 0.05319288430438393,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.931425173512895e-05,
+      "loss": 2.2246,
+      "step": 1217
+    },
+    {
+      "epoch": 0.05323659250841383,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.931311758887877e-05,
+      "loss": 1.8312,
+      "step": 1218
+    },
+    {
+      "epoch": 0.053280300712443725,
+      "grad_norm": 2.625,
+      "learning_rate": 9.931198251201828e-05,
+      "loss": 1.859,
+      "step": 1219
+    },
+    {
+      "epoch": 0.053324008916473625,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.931084650456892e-05,
+      "loss": 2.4562,
+      "step": 1220
+    },
+    {
+      "epoch": 0.05336771712050352,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.930970956655212e-05,
+      "loss": 3.3246,
+      "step": 1221
+    },
+    {
+      "epoch": 0.05341142532453341,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.930857169798931e-05,
+      "loss": 2.1604,
+      "step": 1222
+    },
+    {
+      "epoch": 0.05345513352856331,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.9307432898902e-05,
+      "loss": 2.7602,
+      "step": 1223
+    },
+    {
+      "epoch": 0.05349884173259321,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.930629316931168e-05,
+      "loss": 2.5866,
+      "step": 1224
+    },
+    {
+      "epoch": 0.05354254993662311,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.930515250923984e-05,
+      "loss": 1.9332,
+      "step": 1225
+    },
+    {
+      "epoch": 0.053586258140653,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.9304010918708e-05,
+      "loss": 1.9111,
+      "step": 1226
+    },
+    {
+      "epoch": 0.053629966344682894,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.930286839773773e-05,
+      "loss": 2.2201,
+      "step": 1227
+    },
+    {
+      "epoch": 0.053673674548712795,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.930172494635057e-05,
+      "loss": 2.6325,
+      "step": 1228
+    },
+    {
+      "epoch": 0.05371738275274269,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.93005805645681e-05,
+      "loss": 2.6803,
+      "step": 1229
+    },
+    {
+      "epoch": 0.05376109095677259,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.929943525241194e-05,
+      "loss": 3.2208,
+      "step": 1230
+    },
+    {
+      "epoch": 0.05380479916080248,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.929828900990367e-05,
+      "loss": 2.3413,
+      "step": 1231
+    },
+    {
+      "epoch": 0.053848507364832376,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.929714183706493e-05,
+      "loss": 2.2428,
+      "step": 1232
+    },
+    {
+      "epoch": 0.05389221556886228,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.929599373391738e-05,
+      "loss": 2.9937,
+      "step": 1233
+    },
+    {
+      "epoch": 0.05393592377289217,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.92948447004827e-05,
+      "loss": 2.2362,
+      "step": 1234
+    },
+    {
+      "epoch": 0.05397963197692207,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.929369473678253e-05,
+      "loss": 3.1684,
+      "step": 1235
+    },
+    {
+      "epoch": 0.054023340180951965,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.92925438428386e-05,
+      "loss": 2.0654,
+      "step": 1236
+    },
+    {
+      "epoch": 0.05406704838498186,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.929139201867263e-05,
+      "loss": 2.3591,
+      "step": 1237
+    },
+    {
+      "epoch": 0.05411075658901176,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.929023926430636e-05,
+      "loss": 2.0845,
+      "step": 1238
+    },
+    {
+      "epoch": 0.05415446479304165,
+      "grad_norm": 2.75,
+      "learning_rate": 9.928908557976153e-05,
+      "loss": 2.1563,
+      "step": 1239
+    },
+    {
+      "epoch": 0.05419817299707155,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.928793096505992e-05,
+      "loss": 1.9936,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05424188120110145,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.928677542022331e-05,
+      "loss": 2.7263,
+      "step": 1241
+    },
+    {
+      "epoch": 0.05428558940513134,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.928561894527353e-05,
+      "loss": 1.5295,
+      "step": 1242
+    },
+    {
+      "epoch": 0.05432929760916124,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.928446154023238e-05,
+      "loss": 2.5526,
+      "step": 1243
+    },
+    {
+      "epoch": 0.054373005813191135,
+      "grad_norm": 3.0,
+      "learning_rate": 9.928330320512171e-05,
+      "loss": 2.418,
+      "step": 1244
+    },
+    {
+      "epoch": 0.054416714017221035,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.928214393996339e-05,
+      "loss": 2.2359,
+      "step": 1245
+    },
+    {
+      "epoch": 0.05446042222125093,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.928098374477927e-05,
+      "loss": 2.3314,
+      "step": 1246
+    },
+    {
+      "epoch": 0.05450413042528082,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.927982261959127e-05,
+      "loss": 1.9215,
+      "step": 1247
+    },
+    {
+      "epoch": 0.05454783862931072,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.927866056442128e-05,
+      "loss": 1.9057,
+      "step": 1248
+    },
+    {
+      "epoch": 0.054591546833340616,
+      "grad_norm": 3.125,
+      "learning_rate": 9.927749757929125e-05,
+      "loss": 2.2604,
+      "step": 1249
+    },
+    {
+      "epoch": 0.05463525503737052,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.927633366422314e-05,
+      "loss": 2.3643,
+      "step": 1250
+    },
+    {
+      "epoch": 0.05467896324140041,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.927516881923889e-05,
+      "loss": 2.77,
+      "step": 1251
+    },
+    {
+      "epoch": 0.054722671445430304,
+      "grad_norm": 6.625,
+      "learning_rate": 9.927400304436047e-05,
+      "loss": 1.8398,
+      "step": 1252
+    },
+    {
+      "epoch": 0.054766379649460205,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.92728363396099e-05,
+      "loss": 2.0013,
+      "step": 1253
+    },
+    {
+      "epoch": 0.0548100878534901,
+      "grad_norm": 3.125,
+      "learning_rate": 9.927166870500922e-05,
+      "loss": 2.3079,
+      "step": 1254
+    },
+    {
+      "epoch": 0.05485379605752,
+      "grad_norm": 2.875,
+      "learning_rate": 9.927050014058042e-05,
+      "loss": 2.1234,
+      "step": 1255
+    },
+    {
+      "epoch": 0.05489750426154989,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.926933064634558e-05,
+      "loss": 2.1321,
+      "step": 1256
+    },
+    {
+      "epoch": 0.054941212465579786,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.926816022232675e-05,
+      "loss": 2.8561,
+      "step": 1257
+    },
+    {
+      "epoch": 0.05498492066960969,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.926698886854604e-05,
+      "loss": 2.4861,
+      "step": 1258
+    },
+    {
+      "epoch": 0.05502862887363958,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.926581658502554e-05,
+      "loss": 2.2771,
+      "step": 1259
+    },
+    {
+      "epoch": 0.05507233707766948,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.926464337178738e-05,
+      "loss": 2.2922,
+      "step": 1260
+    },
+    {
+      "epoch": 0.055116045281699375,
+      "grad_norm": 3.5,
+      "learning_rate": 9.92634692288537e-05,
+      "loss": 2.449,
+      "step": 1261
+    },
+    {
+      "epoch": 0.05515975348572927,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.926229415624666e-05,
+      "loss": 2.2916,
+      "step": 1262
+    },
+    {
+      "epoch": 0.05520346168975917,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.926111815398843e-05,
+      "loss": 2.6731,
+      "step": 1263
+    },
+    {
+      "epoch": 0.05524716989378906,
+      "grad_norm": 32.25,
+      "learning_rate": 9.92599412221012e-05,
+      "loss": 2.2843,
+      "step": 1264
+    },
+    {
+      "epoch": 0.05529087809781896,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.925876336060719e-05,
+      "loss": 1.9131,
+      "step": 1265
+    },
+    {
+      "epoch": 0.055334586301848857,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.925758456952862e-05,
+      "loss": 1.8637,
+      "step": 1266
+    },
+    {
+      "epoch": 0.05537829450587875,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.925640484888774e-05,
+      "loss": 2.5708,
+      "step": 1267
+    },
+    {
+      "epoch": 0.05542200270990865,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.925522419870681e-05,
+      "loss": 2.1795,
+      "step": 1268
+    },
+    {
+      "epoch": 0.055465710913938544,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.92540426190081e-05,
+      "loss": 2.3094,
+      "step": 1269
+    },
+    {
+      "epoch": 0.055509419117968445,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.925286010981394e-05,
+      "loss": 2.4622,
+      "step": 1270
+    },
+    {
+      "epoch": 0.05555312732199834,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.925167667114661e-05,
+      "loss": 1.9271,
+      "step": 1271
+    },
+    {
+      "epoch": 0.05559683552602823,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.925049230302846e-05,
+      "loss": 2.2545,
+      "step": 1272
+    },
+    {
+      "epoch": 0.05564054373005813,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.924930700548185e-05,
+      "loss": 2.7818,
+      "step": 1273
+    },
+    {
+      "epoch": 0.055684251934088026,
+      "grad_norm": 2.875,
+      "learning_rate": 9.924812077852913e-05,
+      "loss": 1.8465,
+      "step": 1274
+    },
+    {
+      "epoch": 0.05572796013811793,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.924693362219269e-05,
+      "loss": 2.0316,
+      "step": 1275
+    },
+    {
+      "epoch": 0.05577166834214782,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.924574553649496e-05,
+      "loss": 1.9583,
+      "step": 1276
+    },
+    {
+      "epoch": 0.05581537654617772,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.924455652145831e-05,
+      "loss": 2.0286,
+      "step": 1277
+    },
+    {
+      "epoch": 0.055859084750207615,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.924336657710522e-05,
+      "loss": 1.9079,
+      "step": 1278
+    },
+    {
+      "epoch": 0.05590279295423751,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.924217570345813e-05,
+      "loss": 2.3216,
+      "step": 1279
+    },
+    {
+      "epoch": 0.05594650115826741,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.924098390053951e-05,
+      "loss": 2.2232,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0559902093622973,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.923979116837185e-05,
+      "loss": 2.5149,
+      "step": 1281
+    },
+    {
+      "epoch": 0.0560339175663272,
+      "grad_norm": 5.0,
+      "learning_rate": 9.923859750697768e-05,
+      "loss": 2.9284,
+      "step": 1282
+    },
+    {
+      "epoch": 0.0560776257703571,
+      "grad_norm": 24.25,
+      "learning_rate": 9.923740291637951e-05,
+      "loss": 0.6024,
+      "step": 1283
+    },
+    {
+      "epoch": 0.05612133397438699,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.923620739659989e-05,
+      "loss": 3.0348,
+      "step": 1284
+    },
+    {
+      "epoch": 0.05616504217841689,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.923501094766136e-05,
+      "loss": 2.269,
+      "step": 1285
+    },
+    {
+      "epoch": 0.056208750382446784,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.923381356958654e-05,
+      "loss": 2.0336,
+      "step": 1286
+    },
+    {
+      "epoch": 0.056252458586476685,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.923261526239798e-05,
+      "loss": 2.3973,
+      "step": 1287
+    },
+    {
+      "epoch": 0.05629616679050658,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.923141602611834e-05,
+      "loss": 2.2979,
+      "step": 1288
+    },
+    {
+      "epoch": 0.05633987499453647,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.92302158607702e-05,
+      "loss": 2.3606,
+      "step": 1289
+    },
+    {
+      "epoch": 0.05638358319856637,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.922901476637625e-05,
+      "loss": 2.5858,
+      "step": 1290
+    },
+    {
+      "epoch": 0.056427291402596266,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.922781274295913e-05,
+      "loss": 2.6504,
+      "step": 1291
+    },
+    {
+      "epoch": 0.05647099960662617,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.922660979054155e-05,
+      "loss": 2.3354,
+      "step": 1292
+    },
+    {
+      "epoch": 0.05651470781065606,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.922540590914619e-05,
+      "loss": 1.7236,
+      "step": 1293
+    },
+    {
+      "epoch": 0.056558416014685954,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.922420109879578e-05,
+      "loss": 1.9761,
+      "step": 1294
+    },
+    {
+      "epoch": 0.056602124218715855,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.922299535951305e-05,
+      "loss": 2.3798,
+      "step": 1295
+    },
+    {
+      "epoch": 0.05664583242274575,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.922178869132075e-05,
+      "loss": 2.1449,
+      "step": 1296
+    },
+    {
+      "epoch": 0.05668954062677565,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.922058109424167e-05,
+      "loss": 2.036,
+      "step": 1297
+    },
+    {
+      "epoch": 0.05673324883080554,
+      "grad_norm": 3.0,
+      "learning_rate": 9.921937256829859e-05,
+      "loss": 2.653,
+      "step": 1298
+    },
+    {
+      "epoch": 0.056776957034835436,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.921816311351431e-05,
+      "loss": 2.4909,
+      "step": 1299
+    },
+    {
+      "epoch": 0.05682066523886534,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.921695272991165e-05,
+      "loss": 2.1007,
+      "step": 1300
+    },
+    {
+      "epoch": 0.05686437344289523,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.921574141751346e-05,
+      "loss": 1.91,
+      "step": 1301
+    },
+    {
+      "epoch": 0.05690808164692513,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.921452917634261e-05,
+      "loss": 2.7349,
+      "step": 1302
+    },
+    {
+      "epoch": 0.056951789850955024,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.921331600642196e-05,
+      "loss": 1.8035,
+      "step": 1303
+    },
+    {
+      "epoch": 0.05699549805498492,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.921210190777441e-05,
+      "loss": 2.4165,
+      "step": 1304
+    },
+    {
+      "epoch": 0.05703920625901482,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.921088688042287e-05,
+      "loss": 2.0328,
+      "step": 1305
+    },
+    {
+      "epoch": 0.05708291446304471,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.920967092439027e-05,
+      "loss": 2.5423,
+      "step": 1306
+    },
+    {
+      "epoch": 0.05712662266707461,
+      "grad_norm": 6.59375,
+      "learning_rate": 9.920845403969957e-05,
+      "loss": 2.742,
+      "step": 1307
+    },
+    {
+      "epoch": 0.057170330871104506,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.920723622637371e-05,
+      "loss": 2.1735,
+      "step": 1308
+    },
+    {
+      "epoch": 0.0572140390751344,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.92060174844357e-05,
+      "loss": 2.5959,
+      "step": 1309
+    },
+    {
+      "epoch": 0.0572577472791643,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.920479781390852e-05,
+      "loss": 2.1141,
+      "step": 1310
+    },
+    {
+      "epoch": 0.057301455483194194,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.920357721481518e-05,
+      "loss": 1.6989,
+      "step": 1311
+    },
+    {
+      "epoch": 0.057345163687224095,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.920235568717873e-05,
+      "loss": 1.9488,
+      "step": 1312
+    },
+    {
+      "epoch": 0.05738887189125399,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.920113323102223e-05,
+      "loss": 2.0413,
+      "step": 1313
+    },
+    {
+      "epoch": 0.05743258009528388,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.919990984636871e-05,
+      "loss": 2.6842,
+      "step": 1314
+    },
+    {
+      "epoch": 0.05747628829931378,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.91986855332413e-05,
+      "loss": 2.3465,
+      "step": 1315
+    },
+    {
+      "epoch": 0.057519996503343676,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.919746029166311e-05,
+      "loss": 2.2191,
+      "step": 1316
+    },
+    {
+      "epoch": 0.05756370470737358,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.91962341216572e-05,
+      "loss": 2.654,
+      "step": 1317
+    },
+    {
+      "epoch": 0.05760741291140347,
+      "grad_norm": 2.625,
+      "learning_rate": 9.919500702324677e-05,
+      "loss": 1.9608,
+      "step": 1318
+    },
+    {
+      "epoch": 0.057651121115433364,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.919377899645497e-05,
+      "loss": 1.9015,
+      "step": 1319
+    },
+    {
+      "epoch": 0.057694829319463264,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.919255004130494e-05,
+      "loss": 2.0449,
+      "step": 1320
+    },
+    {
+      "epoch": 0.05773853752349316,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.919132015781991e-05,
+      "loss": 2.9302,
+      "step": 1321
+    },
+    {
+      "epoch": 0.05778224572752306,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.919008934602307e-05,
+      "loss": 1.5977,
+      "step": 1322
+    },
+    {
+      "epoch": 0.05782595393155295,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.918885760593764e-05,
+      "loss": 2.4274,
+      "step": 1323
+    },
+    {
+      "epoch": 0.057869662135582846,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.918762493758689e-05,
+      "loss": 1.7857,
+      "step": 1324
+    },
+    {
+      "epoch": 0.057913370339612746,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.918639134099407e-05,
+      "loss": 2.3214,
+      "step": 1325
+    },
+    {
+      "epoch": 0.05795707854364264,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.918515681618246e-05,
+      "loss": 2.0117,
+      "step": 1326
+    },
+    {
+      "epoch": 0.05800078674767254,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.918392136317533e-05,
+      "loss": 2.0574,
+      "step": 1327
+    },
+    {
+      "epoch": 0.058044494951702434,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.918268498199604e-05,
+      "loss": 1.8831,
+      "step": 1328
+    },
+    {
+      "epoch": 0.05808820315573233,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.918144767266791e-05,
+      "loss": 1.9005,
+      "step": 1329
+    },
+    {
+      "epoch": 0.05813191135976223,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.918020943521427e-05,
+      "loss": 1.7483,
+      "step": 1330
+    },
+    {
+      "epoch": 0.05817561956379212,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.91789702696585e-05,
+      "loss": 1.9886,
+      "step": 1331
+    },
+    {
+      "epoch": 0.05821932776782202,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.917773017602399e-05,
+      "loss": 2.307,
+      "step": 1332
+    },
+    {
+      "epoch": 0.058263035971851916,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.917648915433413e-05,
+      "loss": 2.0734,
+      "step": 1333
+    },
+    {
+      "epoch": 0.05830674417588181,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.917524720461234e-05,
+      "loss": 2.1241,
+      "step": 1334
+    },
+    {
+      "epoch": 0.05835045237991171,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.917400432688208e-05,
+      "loss": 1.9862,
+      "step": 1335
+    },
+    {
+      "epoch": 0.058394160583941604,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.917276052116677e-05,
+      "loss": 2.1588,
+      "step": 1336
+    },
+    {
+      "epoch": 0.058437868787971504,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.917151578748994e-05,
+      "loss": 2.2892,
+      "step": 1337
+    },
+    {
+      "epoch": 0.0584815769920014,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.9170270125875e-05,
+      "loss": 2.1376,
+      "step": 1338
+    },
+    {
+      "epoch": 0.05852528519603129,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.916902353634552e-05,
+      "loss": 2.2808,
+      "step": 1339
+    },
+    {
+      "epoch": 0.05856899340006119,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.916777601892499e-05,
+      "loss": 1.8809,
+      "step": 1340
+    },
+    {
+      "epoch": 0.058612701604091086,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.916652757363698e-05,
+      "loss": 3.0737,
+      "step": 1341
+    },
+    {
+      "epoch": 0.058656409808120986,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.916527820050504e-05,
+      "loss": 2.0382,
+      "step": 1342
+    },
+    {
+      "epoch": 0.05870011801215088,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.916402789955272e-05,
+      "loss": 2.1752,
+      "step": 1343
+    },
+    {
+      "epoch": 0.05874382621618078,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.916277667080365e-05,
+      "loss": 2.1868,
+      "step": 1344
+    },
+    {
+      "epoch": 0.058787534420210674,
+      "grad_norm": 2.625,
+      "learning_rate": 9.916152451428144e-05,
+      "loss": 1.9418,
+      "step": 1345
+    },
+    {
+      "epoch": 0.05883124262424057,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.91602714300097e-05,
+      "loss": 1.4852,
+      "step": 1346
+    },
+    {
+      "epoch": 0.05887495082827047,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.91590174180121e-05,
+      "loss": 1.6127,
+      "step": 1347
+    },
+    {
+      "epoch": 0.05891865903230036,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.915776247831227e-05,
+      "loss": 2.0565,
+      "step": 1348
+    },
+    {
+      "epoch": 0.05896236723633026,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.915650661093395e-05,
+      "loss": 2.064,
+      "step": 1349
+    },
+    {
+      "epoch": 0.059006075440360156,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.915524981590079e-05,
+      "loss": 1.8135,
+      "step": 1350
+    },
+    {
+      "epoch": 0.05904978364439005,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.915399209323652e-05,
+      "loss": 2.0698,
+      "step": 1351
+    },
+    {
+      "epoch": 0.05909349184841995,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.915273344296488e-05,
+      "loss": 2.1185,
+      "step": 1352
+    },
+    {
+      "epoch": 0.059137200052449844,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.915147386510964e-05,
+      "loss": 2.5517,
+      "step": 1353
+    },
+    {
+      "epoch": 0.059180908256479745,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.915021335969452e-05,
+      "loss": 2.5902,
+      "step": 1354
+    },
+    {
+      "epoch": 0.05922461646050964,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.914895192674336e-05,
+      "loss": 2.0517,
+      "step": 1355
+    },
+    {
+      "epoch": 0.05926832466453953,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.914768956627994e-05,
+      "loss": 2.1344,
+      "step": 1356
+    },
+    {
+      "epoch": 0.05931203286856943,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.914642627832808e-05,
+      "loss": 2.0166,
+      "step": 1357
+    },
+    {
+      "epoch": 0.059355741072599326,
+      "grad_norm": 3.0,
+      "learning_rate": 9.914516206291165e-05,
+      "loss": 1.9574,
+      "step": 1358
+    },
+    {
+      "epoch": 0.059399449276629226,
+      "grad_norm": 3.5,
+      "learning_rate": 9.914389692005446e-05,
+      "loss": 2.3096,
+      "step": 1359
+    },
+    {
+      "epoch": 0.05944315748065912,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.914263084978042e-05,
+      "loss": 1.9114,
+      "step": 1360
+    },
+    {
+      "epoch": 0.059486865684689014,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.914136385211341e-05,
+      "loss": 1.9104,
+      "step": 1361
+    },
+    {
+      "epoch": 0.059530573888718914,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.914009592707733e-05,
+      "loss": 2.0492,
+      "step": 1362
+    },
+    {
+      "epoch": 0.05957428209274881,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.913882707469614e-05,
+      "loss": 2.3943,
+      "step": 1363
+    },
+    {
+      "epoch": 0.05961799029677871,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.913755729499376e-05,
+      "loss": 2.0374,
+      "step": 1364
+    },
+    {
+      "epoch": 0.0596616985008086,
+      "grad_norm": 2.375,
+      "learning_rate": 9.913628658799415e-05,
+      "loss": 2.1342,
+      "step": 1365
+    },
+    {
+      "epoch": 0.059705406704838496,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.91350149537213e-05,
+      "loss": 2.3227,
+      "step": 1366
+    },
+    {
+      "epoch": 0.059749114908868396,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.913374239219922e-05,
+      "loss": 2.3847,
+      "step": 1367
+    },
+    {
+      "epoch": 0.05979282311289829,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.913246890345189e-05,
+      "loss": 1.9194,
+      "step": 1368
+    },
+    {
+      "epoch": 0.05983653131692819,
+      "grad_norm": 3.125,
+      "learning_rate": 9.913119448750337e-05,
+      "loss": 2.4296,
+      "step": 1369
+    },
+    {
+      "epoch": 0.059880239520958084,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.912991914437771e-05,
+      "loss": 1.7472,
+      "step": 1370
+    },
+    {
+      "epoch": 0.05992394772498798,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.912864287409896e-05,
+      "loss": 1.9631,
+      "step": 1371
+    },
+    {
+      "epoch": 0.05996765592901788,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.912736567669121e-05,
+      "loss": 1.94,
+      "step": 1372
+    },
+    {
+      "epoch": 0.06001136413304777,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.912608755217859e-05,
+      "loss": 2.1358,
+      "step": 1373
+    },
+    {
+      "epoch": 0.06005507233707767,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.912480850058516e-05,
+      "loss": 1.6588,
+      "step": 1374
+    },
+    {
+      "epoch": 0.060098780541107566,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.912352852193514e-05,
+      "loss": 2.0139,
+      "step": 1375
+    },
+    {
+      "epoch": 0.06014248874513746,
+      "grad_norm": 4.125,
+      "learning_rate": 9.912224761625262e-05,
+      "loss": 2.7426,
+      "step": 1376
+    },
+    {
+      "epoch": 0.06018619694916736,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.912096578356179e-05,
+      "loss": 2.7015,
+      "step": 1377
+    },
+    {
+      "epoch": 0.060229905153197254,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.911968302388685e-05,
+      "loss": 2.041,
+      "step": 1378
+    },
+    {
+      "epoch": 0.060273613357227154,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.9118399337252e-05,
+      "loss": 2.2014,
+      "step": 1379
+    },
+    {
+      "epoch": 0.06031732156125705,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.911711472368148e-05,
+      "loss": 1.9344,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06036102976528694,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.91158291831995e-05,
+      "loss": 1.7271,
+      "step": 1381
+    },
+    {
+      "epoch": 0.06040473796931684,
+      "grad_norm": 7.03125,
+      "learning_rate": 9.911454271583034e-05,
+      "loss": 2.765,
+      "step": 1382
+    },
+    {
+      "epoch": 0.060448446173346736,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.911325532159828e-05,
+      "loss": 2.8126,
+      "step": 1383
+    },
+    {
+      "epoch": 0.060492154377376636,
+      "grad_norm": 2.625,
+      "learning_rate": 9.91119670005276e-05,
+      "loss": 2.4261,
+      "step": 1384
+    },
+    {
+      "epoch": 0.06053586258140653,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.911067775264264e-05,
+      "loss": 1.7767,
+      "step": 1385
+    },
+    {
+      "epoch": 0.060579570785436423,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.91093875779677e-05,
+      "loss": 2.9243,
+      "step": 1386
+    },
+    {
+      "epoch": 0.060623278989466324,
+      "grad_norm": 4.0,
+      "learning_rate": 9.910809647652715e-05,
+      "loss": 3.1425,
+      "step": 1387
+    },
+    {
+      "epoch": 0.06066698719349622,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.910680444834535e-05,
+      "loss": 1.9814,
+      "step": 1388
+    },
+    {
+      "epoch": 0.06071069539752612,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.910551149344669e-05,
+      "loss": 2.4808,
+      "step": 1389
+    },
+    {
+      "epoch": 0.06075440360155601,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.910421761185553e-05,
+      "loss": 1.9118,
+      "step": 1390
+    },
+    {
+      "epoch": 0.060798111805585905,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.910292280359631e-05,
+      "loss": 2.3266,
+      "step": 1391
+    },
+    {
+      "epoch": 0.060841820009615806,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.91016270686935e-05,
+      "loss": 2.2481,
+      "step": 1392
+    },
+    {
+      "epoch": 0.0608855282136457,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.910033040717152e-05,
+      "loss": 1.938,
+      "step": 1393
+    },
+    {
+      "epoch": 0.0609292364176756,
+      "grad_norm": 3.25,
+      "learning_rate": 9.909903281905484e-05,
+      "loss": 2.3012,
+      "step": 1394
+    },
+    {
+      "epoch": 0.060972944621705494,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.909773430436794e-05,
+      "loss": 2.2996,
+      "step": 1395
+    },
+    {
+      "epoch": 0.06101665282573539,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.909643486313533e-05,
+      "loss": 2.4493,
+      "step": 1396
+    },
+    {
+      "epoch": 0.06106036102976529,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.909513449538156e-05,
+      "loss": 2.1811,
+      "step": 1397
+    },
+    {
+      "epoch": 0.06110406923379518,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.909383320113113e-05,
+      "loss": 1.8874,
+      "step": 1398
+    },
+    {
+      "epoch": 0.06114777743782508,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.909253098040863e-05,
+      "loss": 2.762,
+      "step": 1399
+    },
+    {
+      "epoch": 0.061191485641854976,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.909122783323861e-05,
+      "loss": 2.0106,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06123519384588487,
+      "grad_norm": 3.5,
+      "learning_rate": 9.908992375964568e-05,
+      "loss": 1.8271,
+      "step": 1401
+    },
+    {
+      "epoch": 0.06127890204991477,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.908861875965443e-05,
+      "loss": 2.2767,
+      "step": 1402
+    },
+    {
+      "epoch": 0.061322610253944664,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.908731283328949e-05,
+      "loss": 2.3544,
+      "step": 1403
+    },
+    {
+      "epoch": 0.061366318457974564,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.908600598057554e-05,
+      "loss": 1.7036,
+      "step": 1404
+    },
+    {
+      "epoch": 0.06141002666200446,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.90846982015372e-05,
+      "loss": 2.2406,
+      "step": 1405
+    },
+    {
+      "epoch": 0.06145373486603435,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.908338949619917e-05,
+      "loss": 2.2038,
+      "step": 1406
+    },
+    {
+      "epoch": 0.06149744307006425,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.908207986458613e-05,
+      "loss": 2.0234,
+      "step": 1407
+    },
+    {
+      "epoch": 0.061541151274094145,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.908076930672282e-05,
+      "loss": 2.0903,
+      "step": 1408
+    },
+    {
+      "epoch": 0.061584859478124046,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.907945782263396e-05,
+      "loss": 2.2869,
+      "step": 1409
+    },
+    {
+      "epoch": 0.06162856768215394,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.907814541234429e-05,
+      "loss": 2.3249,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06167227588618383,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.907683207587859e-05,
+      "loss": 2.0903,
+      "step": 1411
+    },
+    {
+      "epoch": 0.061715984090213734,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.907551781326165e-05,
+      "loss": 1.7549,
+      "step": 1412
+    },
+    {
+      "epoch": 0.06175969229424363,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.907420262451826e-05,
+      "loss": 2.3239,
+      "step": 1413
+    },
+    {
+      "epoch": 0.06180340049827353,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.907288650967324e-05,
+      "loss": 2.0301,
+      "step": 1414
+    },
+    {
+      "epoch": 0.06184710870230342,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.907156946875142e-05,
+      "loss": 1.8346,
+      "step": 1415
+    },
+    {
+      "epoch": 0.06189081690633332,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.907025150177768e-05,
+      "loss": 2.1143,
+      "step": 1416
+    },
+    {
+      "epoch": 0.061934525110363216,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.906893260877686e-05,
+      "loss": 2.0177,
+      "step": 1417
+    },
+    {
+      "epoch": 0.06197823331439311,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.906761278977387e-05,
+      "loss": 3.0162,
+      "step": 1418
+    },
+    {
+      "epoch": 0.06202194151842301,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.906629204479362e-05,
+      "loss": 2.8375,
+      "step": 1419
+    },
+    {
+      "epoch": 0.062065649722452904,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.906497037386102e-05,
+      "loss": 1.6367,
+      "step": 1420
+    },
+    {
+      "epoch": 0.062109357926482804,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.906364777700104e-05,
+      "loss": 2.1294,
+      "step": 1421
+    },
+    {
+      "epoch": 0.0621530661305127,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.906232425423858e-05,
+      "loss": 2.181,
+      "step": 1422
+    },
+    {
+      "epoch": 0.06219677433454259,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.906099980559868e-05,
+      "loss": 1.8974,
+      "step": 1423
+    },
+    {
+      "epoch": 0.06224048253857249,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.90596744311063e-05,
+      "loss": 2.1287,
+      "step": 1424
+    },
+    {
+      "epoch": 0.062284190742602386,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.905834813078646e-05,
+      "loss": 2.0806,
+      "step": 1425
+    },
+    {
+      "epoch": 0.062327898946632286,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.905702090466419e-05,
+      "loss": 2.4359,
+      "step": 1426
+    },
+    {
+      "epoch": 0.06237160715066218,
+      "grad_norm": 3.5,
+      "learning_rate": 9.905569275276454e-05,
+      "loss": 2.2533,
+      "step": 1427
+    },
+    {
+      "epoch": 0.06241531535469207,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.905436367511256e-05,
+      "loss": 1.8447,
+      "step": 1428
+    },
+    {
+      "epoch": 0.062459023558721974,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.905303367173336e-05,
+      "loss": 2.7585,
+      "step": 1429
+    },
+    {
+      "epoch": 0.06250273176275187,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.9051702742652e-05,
+      "loss": 1.7898,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06254643996678176,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.905037088789363e-05,
+      "loss": 2.5768,
+      "step": 1431
+    },
+    {
+      "epoch": 0.06259014817081165,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.904903810748339e-05,
+      "loss": 2.9688,
+      "step": 1432
+    },
+    {
+      "epoch": 0.06263385637484156,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.904770440144638e-05,
+      "loss": 1.9799,
+      "step": 1433
+    },
+    {
+      "epoch": 0.06267756457887146,
+      "grad_norm": 3.125,
+      "learning_rate": 9.904636976980782e-05,
+      "loss": 2.2924,
+      "step": 1434
+    },
+    {
+      "epoch": 0.06272127278290135,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.904503421259288e-05,
+      "loss": 2.0099,
+      "step": 1435
+    },
+    {
+      "epoch": 0.06276498098693124,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.904369772982676e-05,
+      "loss": 2.7724,
+      "step": 1436
+    },
+    {
+      "epoch": 0.06280868919096114,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.904236032153469e-05,
+      "loss": 2.749,
+      "step": 1437
+    },
+    {
+      "epoch": 0.06285239739499104,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.904102198774188e-05,
+      "loss": 1.9447,
+      "step": 1438
+    },
+    {
+      "epoch": 0.06289610559902094,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.903968272847363e-05,
+      "loss": 2.5455,
+      "step": 1439
+    },
+    {
+      "epoch": 0.06293981380305083,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.90383425437552e-05,
+      "loss": 2.74,
+      "step": 1440
+    },
+    {
+      "epoch": 0.06298352200708073,
+      "grad_norm": 2.5,
+      "learning_rate": 9.903700143361185e-05,
+      "loss": 1.8687,
+      "step": 1441
+    },
+    {
+      "epoch": 0.06302723021111062,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.903565939806893e-05,
+      "loss": 1.8802,
+      "step": 1442
+    },
+    {
+      "epoch": 0.06307093841514053,
+      "grad_norm": 2.5,
+      "learning_rate": 9.903431643715175e-05,
+      "loss": 1.7479,
+      "step": 1443
+    },
+    {
+      "epoch": 0.06311464661917042,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.903297255088563e-05,
+      "loss": 2.1058,
+      "step": 1444
+    },
+    {
+      "epoch": 0.06315835482320031,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.903162773929599e-05,
+      "loss": 2.3713,
+      "step": 1445
+    },
+    {
+      "epoch": 0.06320206302723021,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.903028200240815e-05,
+      "loss": 2.5098,
+      "step": 1446
+    },
+    {
+      "epoch": 0.0632457712312601,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.902893534024753e-05,
+      "loss": 2.1725,
+      "step": 1447
+    },
+    {
+      "epoch": 0.06328947943529001,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.902758775283955e-05,
+      "loss": 2.1446,
+      "step": 1448
+    },
+    {
+      "epoch": 0.0633331876393199,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.902623924020962e-05,
+      "loss": 1.8664,
+      "step": 1449
+    },
+    {
+      "epoch": 0.0633768958433498,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.902488980238322e-05,
+      "loss": 2.2319,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06342060404737969,
+      "grad_norm": 2.5,
+      "learning_rate": 9.902353943938578e-05,
+      "loss": 1.9501,
+      "step": 1451
+    },
+    {
+      "epoch": 0.0634643122514096,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.90221881512428e-05,
+      "loss": 1.8332,
+      "step": 1452
+    },
+    {
+      "epoch": 0.06350802045543949,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.902083593797979e-05,
+      "loss": 2.0205,
+      "step": 1453
+    },
+    {
+      "epoch": 0.06355172865946938,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.901948279962226e-05,
+      "loss": 2.2535,
+      "step": 1454
+    },
+    {
+      "epoch": 0.06359543686349928,
+      "grad_norm": 3.625,
+      "learning_rate": 9.901812873619574e-05,
+      "loss": 2.2449,
+      "step": 1455
+    },
+    {
+      "epoch": 0.06363914506752917,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.901677374772579e-05,
+      "loss": 2.5408,
+      "step": 1456
+    },
+    {
+      "epoch": 0.06368285327155908,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.901541783423798e-05,
+      "loss": 2.0815,
+      "step": 1457
+    },
+    {
+      "epoch": 0.06372656147558897,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.90140609957579e-05,
+      "loss": 2.0225,
+      "step": 1458
+    },
+    {
+      "epoch": 0.06377026967961887,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.901270323231115e-05,
+      "loss": 1.7129,
+      "step": 1459
+    },
+    {
+      "epoch": 0.06381397788364876,
+      "grad_norm": 3.734375,
+      "learning_rate": 9.901134454392334e-05,
+      "loss": 2.7482,
+      "step": 1460
+    },
+    {
+      "epoch": 0.06385768608767865,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.900998493062015e-05,
+      "loss": 2.3469,
+      "step": 1461
+    },
+    {
+      "epoch": 0.06390139429170856,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.900862439242719e-05,
+      "loss": 1.9815,
+      "step": 1462
+    },
+    {
+      "epoch": 0.06394510249573845,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.900726292937018e-05,
+      "loss": 2.0003,
+      "step": 1463
+    },
+    {
+      "epoch": 0.06398881069976835,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.900590054147478e-05,
+      "loss": 2.2793,
+      "step": 1464
+    },
+    {
+      "epoch": 0.06403251890379824,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.900453722876672e-05,
+      "loss": 1.8329,
+      "step": 1465
+    },
+    {
+      "epoch": 0.06407622710782813,
+      "grad_norm": 2.75,
+      "learning_rate": 9.900317299127171e-05,
+      "loss": 1.8097,
+      "step": 1466
+    },
+    {
+      "epoch": 0.06411993531185804,
+      "grad_norm": 3.5,
+      "learning_rate": 9.900180782901551e-05,
+      "loss": 2.5161,
+      "step": 1467
+    },
+    {
+      "epoch": 0.06416364351588794,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.900044174202388e-05,
+      "loss": 1.8867,
+      "step": 1468
+    },
+    {
+      "epoch": 0.06420735171991783,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.899907473032259e-05,
+      "loss": 2.3157,
+      "step": 1469
+    },
+    {
+      "epoch": 0.06425105992394772,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.899770679393747e-05,
+      "loss": 1.8642,
+      "step": 1470
+    },
+    {
+      "epoch": 0.06429476812797762,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.899633793289427e-05,
+      "loss": 1.941,
+      "step": 1471
+    },
+    {
+      "epoch": 0.06433847633200752,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.89949681472189e-05,
+      "loss": 2.7477,
+      "step": 1472
+    },
+    {
+      "epoch": 0.06438218453603742,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.899359743693714e-05,
+      "loss": 2.2268,
+      "step": 1473
+    },
+    {
+      "epoch": 0.06442589274006731,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.899222580207492e-05,
+      "loss": 2.1666,
+      "step": 1474
+    },
+    {
+      "epoch": 0.0644696009440972,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.899085324265807e-05,
+      "loss": 2.0582,
+      "step": 1475
+    },
+    {
+      "epoch": 0.0645133091481271,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.898947975871253e-05,
+      "loss": 1.966,
+      "step": 1476
+    },
+    {
+      "epoch": 0.064557017352157,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.89881053502642e-05,
+      "loss": 2.102,
+      "step": 1477
+    },
+    {
+      "epoch": 0.0646007255561869,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.898673001733902e-05,
+      "loss": 2.1326,
+      "step": 1478
+    },
+    {
+      "epoch": 0.0646444337602168,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.898535375996296e-05,
+      "loss": 1.7869,
+      "step": 1479
+    },
+    {
+      "epoch": 0.06468814196424669,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.898397657816198e-05,
+      "loss": 2.1237,
+      "step": 1480
+    },
+    {
+      "epoch": 0.06473185016827658,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.898259847196205e-05,
+      "loss": 2.1493,
+      "step": 1481
+    },
+    {
+      "epoch": 0.06477555837230649,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.89812194413892e-05,
+      "loss": 2.0955,
+      "step": 1482
+    },
+    {
+      "epoch": 0.06481926657633638,
+      "grad_norm": 2.875,
+      "learning_rate": 9.897983948646948e-05,
+      "loss": 1.7992,
+      "step": 1483
+    },
+    {
+      "epoch": 0.06486297478036628,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.897845860722888e-05,
+      "loss": 2.1167,
+      "step": 1484
+    },
+    {
+      "epoch": 0.06490668298439617,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.897707680369348e-05,
+      "loss": 2.0633,
+      "step": 1485
+    },
+    {
+      "epoch": 0.06495039118842606,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.897569407588935e-05,
+      "loss": 3.077,
+      "step": 1486
+    },
+    {
+      "epoch": 0.06499409939245597,
+      "grad_norm": 3.25,
+      "learning_rate": 9.897431042384261e-05,
+      "loss": 2.4853,
+      "step": 1487
+    },
+    {
+      "epoch": 0.06503780759648586,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.897292584757934e-05,
+      "loss": 2.4062,
+      "step": 1488
+    },
+    {
+      "epoch": 0.06508151580051576,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.897154034712568e-05,
+      "loss": 1.9607,
+      "step": 1489
+    },
+    {
+      "epoch": 0.06512522400454565,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.897015392250779e-05,
+      "loss": 2.6984,
+      "step": 1490
+    },
+    {
+      "epoch": 0.06516893220857554,
+      "grad_norm": 3.0,
+      "learning_rate": 9.896876657375183e-05,
+      "loss": 2.8004,
+      "step": 1491
+    },
+    {
+      "epoch": 0.06521264041260545,
+      "grad_norm": 3.75,
+      "learning_rate": 9.896737830088396e-05,
+      "loss": 1.97,
+      "step": 1492
+    },
+    {
+      "epoch": 0.06525634861663535,
+      "grad_norm": 4.625,
+      "learning_rate": 9.89659891039304e-05,
+      "loss": 2.434,
+      "step": 1493
+    },
+    {
+      "epoch": 0.06530005682066524,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.896459898291734e-05,
+      "loss": 1.8298,
+      "step": 1494
+    },
+    {
+      "epoch": 0.06534376502469513,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.896320793787106e-05,
+      "loss": 2.0294,
+      "step": 1495
+    },
+    {
+      "epoch": 0.06538747322872503,
+      "grad_norm": 4.875,
+      "learning_rate": 9.896181596881777e-05,
+      "loss": 1.8582,
+      "step": 1496
+    },
+    {
+      "epoch": 0.06543118143275493,
+      "grad_norm": 3.0,
+      "learning_rate": 9.896042307578376e-05,
+      "loss": 2.4271,
+      "step": 1497
+    },
+    {
+      "epoch": 0.06547488963678483,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.89590292587953e-05,
+      "loss": 2.879,
+      "step": 1498
+    },
+    {
+      "epoch": 0.06551859784081472,
+      "grad_norm": 2.875,
+      "learning_rate": 9.895763451787869e-05,
+      "loss": 2.1706,
+      "step": 1499
+    },
+    {
+      "epoch": 0.06556230604484461,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.895623885306029e-05,
+      "loss": 1.8151,
+      "step": 1500
+    },
+    {
+      "epoch": 0.06560601424887451,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.89548422643664e-05,
+      "loss": 2.4741,
+      "step": 1501
+    },
+    {
+      "epoch": 0.06564972245290442,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.895344475182338e-05,
+      "loss": 1.8052,
+      "step": 1502
+    },
+    {
+      "epoch": 0.06569343065693431,
+      "grad_norm": 3.625,
+      "learning_rate": 9.89520463154576e-05,
+      "loss": 2.1158,
+      "step": 1503
+    },
+    {
+      "epoch": 0.0657371388609642,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.895064695529548e-05,
+      "loss": 1.9689,
+      "step": 1504
+    },
+    {
+      "epoch": 0.0657808470649941,
+      "grad_norm": 4.0,
+      "learning_rate": 9.89492466713634e-05,
+      "loss": 2.3131,
+      "step": 1505
+    },
+    {
+      "epoch": 0.06582455526902399,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.894784546368779e-05,
+      "loss": 2.4752,
+      "step": 1506
+    },
+    {
+      "epoch": 0.0658682634730539,
+      "grad_norm": 3.125,
+      "learning_rate": 9.894644333229511e-05,
+      "loss": 2.025,
+      "step": 1507
+    },
+    {
+      "epoch": 0.06591197167708379,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.894504027721179e-05,
+      "loss": 2.1173,
+      "step": 1508
+    },
+    {
+      "epoch": 0.06595567988111369,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.894363629846432e-05,
+      "loss": 2.4434,
+      "step": 1509
+    },
+    {
+      "epoch": 0.06599938808514358,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.894223139607921e-05,
+      "loss": 1.9026,
+      "step": 1510
+    },
+    {
+      "epoch": 0.06604309628917347,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.894082557008296e-05,
+      "loss": 1.9825,
+      "step": 1511
+    },
+    {
+      "epoch": 0.06608680449320338,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.893941882050209e-05,
+      "loss": 2.8117,
+      "step": 1512
+    },
+    {
+      "epoch": 0.06613051269723327,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.893801114736318e-05,
+      "loss": 2.2047,
+      "step": 1513
+    },
+    {
+      "epoch": 0.06617422090126317,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.893660255069275e-05,
+      "loss": 2.2197,
+      "step": 1514
+    },
+    {
+      "epoch": 0.06621792910529306,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.893519303051742e-05,
+      "loss": 2.2841,
+      "step": 1515
+    },
+    {
+      "epoch": 0.06626163730932295,
+      "grad_norm": 6.5625,
+      "learning_rate": 9.893378258686377e-05,
+      "loss": 2.3869,
+      "step": 1516
+    },
+    {
+      "epoch": 0.06630534551335286,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.893237121975843e-05,
+      "loss": 2.418,
+      "step": 1517
+    },
+    {
+      "epoch": 0.06634905371738276,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.893095892922803e-05,
+      "loss": 2.641,
+      "step": 1518
+    },
+    {
+      "epoch": 0.06639276192141265,
+      "grad_norm": 7.0,
+      "learning_rate": 9.89295457152992e-05,
+      "loss": 2.23,
+      "step": 1519
+    },
+    {
+      "epoch": 0.06643647012544254,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.892813157799864e-05,
+      "loss": 1.9,
+      "step": 1520
+    },
+    {
+      "epoch": 0.06648017832947244,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.892671651735304e-05,
+      "loss": 2.3359,
+      "step": 1521
+    },
+    {
+      "epoch": 0.06652388653350234,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.892530053338909e-05,
+      "loss": 2.4135,
+      "step": 1522
+    },
+    {
+      "epoch": 0.06656759473753224,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.89238836261335e-05,
+      "loss": 2.7742,
+      "step": 1523
+    },
+    {
+      "epoch": 0.06661130294156213,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.892246579561302e-05,
+      "loss": 1.8957,
+      "step": 1524
+    },
+    {
+      "epoch": 0.06665501114559202,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.89210470418544e-05,
+      "loss": 2.5296,
+      "step": 1525
+    },
+    {
+      "epoch": 0.06669871934962192,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.891962736488443e-05,
+      "loss": 2.2513,
+      "step": 1526
+    },
+    {
+      "epoch": 0.06674242755365183,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.89182067647299e-05,
+      "loss": 2.5814,
+      "step": 1527
+    },
+    {
+      "epoch": 0.06678613575768172,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.891678524141758e-05,
+      "loss": 1.811,
+      "step": 1528
+    },
+    {
+      "epoch": 0.06682984396171161,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.891536279497436e-05,
+      "loss": 2.0597,
+      "step": 1529
+    },
+    {
+      "epoch": 0.0668735521657415,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.891393942542704e-05,
+      "loss": 2.3774,
+      "step": 1530
+    },
+    {
+      "epoch": 0.0669172603697714,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.891251513280248e-05,
+      "loss": 2.5181,
+      "step": 1531
+    },
+    {
+      "epoch": 0.06696096857380131,
+      "grad_norm": 2.375,
+      "learning_rate": 9.891108991712759e-05,
+      "loss": 2.0023,
+      "step": 1532
+    },
+    {
+      "epoch": 0.0670046767778312,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.890966377842925e-05,
+      "loss": 1.9709,
+      "step": 1533
+    },
+    {
+      "epoch": 0.0670483849818611,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.890823671673436e-05,
+      "loss": 2.2626,
+      "step": 1534
+    },
+    {
+      "epoch": 0.06709209318589099,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.890680873206986e-05,
+      "loss": 2.0042,
+      "step": 1535
+    },
+    {
+      "epoch": 0.06713580138992088,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.89053798244627e-05,
+      "loss": 2.0635,
+      "step": 1536
+    },
+    {
+      "epoch": 0.06717950959395079,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.890394999393984e-05,
+      "loss": 2.7742,
+      "step": 1537
+    },
+    {
+      "epoch": 0.06722321779798068,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.890251924052827e-05,
+      "loss": 2.0707,
+      "step": 1538
+    },
+    {
+      "epoch": 0.06726692600201058,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.8901087564255e-05,
+      "loss": 2.774,
+      "step": 1539
+    },
+    {
+      "epoch": 0.06731063420604047,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.889965496514702e-05,
+      "loss": 3.196,
+      "step": 1540
+    },
+    {
+      "epoch": 0.06735434241007036,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.889822144323137e-05,
+      "loss": 2.3068,
+      "step": 1541
+    },
+    {
+      "epoch": 0.06739805061410027,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.889678699853514e-05,
+      "loss": 2.4428,
+      "step": 1542
+    },
+    {
+      "epoch": 0.06744175881813017,
+      "grad_norm": 2.5,
+      "learning_rate": 9.889535163108537e-05,
+      "loss": 2.0466,
+      "step": 1543
+    },
+    {
+      "epoch": 0.06748546702216006,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.889391534090912e-05,
+      "loss": 1.907,
+      "step": 1544
+    },
+    {
+      "epoch": 0.06752917522618995,
+      "grad_norm": 12.0625,
+      "learning_rate": 9.889247812803356e-05,
+      "loss": 6.6027,
+      "step": 1545
+    },
+    {
+      "epoch": 0.06757288343021985,
+      "grad_norm": 3.125,
+      "learning_rate": 9.889103999248576e-05,
+      "loss": 2.2651,
+      "step": 1546
+    },
+    {
+      "epoch": 0.06761659163424975,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.88896009342929e-05,
+      "loss": 1.9441,
+      "step": 1547
+    },
+    {
+      "epoch": 0.06766029983827965,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.888816095348209e-05,
+      "loss": 2.6584,
+      "step": 1548
+    },
+    {
+      "epoch": 0.06770400804230954,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.888672005008054e-05,
+      "loss": 2.5327,
+      "step": 1549
+    },
+    {
+      "epoch": 0.06774771624633943,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.888527822411543e-05,
+      "loss": 2.1271,
+      "step": 1550
+    },
+    {
+      "epoch": 0.06779142445036933,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.888383547561398e-05,
+      "loss": 2.3199,
+      "step": 1551
+    },
+    {
+      "epoch": 0.06783513265439924,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.888239180460339e-05,
+      "loss": 2.153,
+      "step": 1552
+    },
+    {
+      "epoch": 0.06787884085842913,
+      "grad_norm": 2.625,
+      "learning_rate": 9.888094721111093e-05,
+      "loss": 2.5418,
+      "step": 1553
+    },
+    {
+      "epoch": 0.06792254906245902,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.887950169516386e-05,
+      "loss": 2.1519,
+      "step": 1554
+    },
+    {
+      "epoch": 0.06796625726648892,
+      "grad_norm": 3.25,
+      "learning_rate": 9.887805525678943e-05,
+      "loss": 2.2746,
+      "step": 1555
+    },
+    {
+      "epoch": 0.06800996547051881,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.887660789601499e-05,
+      "loss": 1.8309,
+      "step": 1556
+    },
+    {
+      "epoch": 0.06805367367454872,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.88751596128678e-05,
+      "loss": 2.4382,
+      "step": 1557
+    },
+    {
+      "epoch": 0.06809738187857861,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.887371040737523e-05,
+      "loss": 2.1003,
+      "step": 1558
+    },
+    {
+      "epoch": 0.0681410900826085,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.88722602795646e-05,
+      "loss": 2.383,
+      "step": 1559
+    },
+    {
+      "epoch": 0.0681847982866384,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.887080922946329e-05,
+      "loss": 2.1193,
+      "step": 1560
+    },
+    {
+      "epoch": 0.06822850649066829,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.886935725709868e-05,
+      "loss": 2.1573,
+      "step": 1561
+    },
+    {
+      "epoch": 0.0682722146946982,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.886790436249818e-05,
+      "loss": 2.4051,
+      "step": 1562
+    },
+    {
+      "epoch": 0.0683159228987281,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.886645054568919e-05,
+      "loss": 1.9849,
+      "step": 1563
+    },
+    {
+      "epoch": 0.06835963110275799,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.886499580669917e-05,
+      "loss": 1.915,
+      "step": 1564
+    },
+    {
+      "epoch": 0.06840333930678788,
+      "grad_norm": 2.75,
+      "learning_rate": 9.886354014555554e-05,
+      "loss": 2.0232,
+      "step": 1565
+    },
+    {
+      "epoch": 0.06844704751081777,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.886208356228581e-05,
+      "loss": 2.0822,
+      "step": 1566
+    },
+    {
+      "epoch": 0.06849075571484768,
+      "grad_norm": 2.75,
+      "learning_rate": 9.886062605691743e-05,
+      "loss": 1.6554,
+      "step": 1567
+    },
+    {
+      "epoch": 0.06853446391887758,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.885916762947795e-05,
+      "loss": 2.097,
+      "step": 1568
+    },
+    {
+      "epoch": 0.06857817212290747,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.885770827999484e-05,
+      "loss": 1.9294,
+      "step": 1569
+    },
+    {
+      "epoch": 0.06862188032693736,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.885624800849567e-05,
+      "loss": 2.3477,
+      "step": 1570
+    },
+    {
+      "epoch": 0.06866558853096726,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.8854786815008e-05,
+      "loss": 2.2106,
+      "step": 1571
+    },
+    {
+      "epoch": 0.06870929673499716,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.88533246995594e-05,
+      "loss": 1.9569,
+      "step": 1572
+    },
+    {
+      "epoch": 0.06875300493902706,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.885186166217746e-05,
+      "loss": 1.7803,
+      "step": 1573
+    },
+    {
+      "epoch": 0.06879671314305695,
+      "grad_norm": 2.75,
+      "learning_rate": 9.885039770288979e-05,
+      "loss": 2.2029,
+      "step": 1574
+    },
+    {
+      "epoch": 0.06884042134708684,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.884893282172401e-05,
+      "loss": 2.0134,
+      "step": 1575
+    },
+    {
+      "epoch": 0.06888412955111674,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.884746701870777e-05,
+      "loss": 2.7702,
+      "step": 1576
+    },
+    {
+      "epoch": 0.06892783775514665,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.884600029386875e-05,
+      "loss": 2.3747,
+      "step": 1577
+    },
+    {
+      "epoch": 0.06897154595917654,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.884453264723459e-05,
+      "loss": 2.3949,
+      "step": 1578
+    },
+    {
+      "epoch": 0.06901525416320643,
+      "grad_norm": 5.78125,
+      "learning_rate": 9.884306407883301e-05,
+      "loss": 2.6275,
+      "step": 1579
+    },
+    {
+      "epoch": 0.06905896236723633,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.884159458869173e-05,
+      "loss": 2.6411,
+      "step": 1580
+    },
+    {
+      "epoch": 0.06910267057126622,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.884012417683849e-05,
+      "loss": 2.56,
+      "step": 1581
+    },
+    {
+      "epoch": 0.06914637877529613,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.8838652843301e-05,
+      "loss": 1.8407,
+      "step": 1582
+    },
+    {
+      "epoch": 0.06919008697932602,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.883718058810707e-05,
+      "loss": 2.2617,
+      "step": 1583
+    },
+    {
+      "epoch": 0.06923379518335591,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.883570741128446e-05,
+      "loss": 2.1872,
+      "step": 1584
+    },
+    {
+      "epoch": 0.06927750338738581,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.883423331286096e-05,
+      "loss": 3.165,
+      "step": 1585
+    },
+    {
+      "epoch": 0.06932121159141572,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.88327582928644e-05,
+      "loss": 2.0316,
+      "step": 1586
+    },
+    {
+      "epoch": 0.06936491979544561,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.883128235132264e-05,
+      "loss": 2.3265,
+      "step": 1587
+    },
+    {
+      "epoch": 0.0694086279994755,
+      "grad_norm": 2.75,
+      "learning_rate": 9.882980548826349e-05,
+      "loss": 2.2151,
+      "step": 1588
+    },
+    {
+      "epoch": 0.0694523362035054,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.882832770371487e-05,
+      "loss": 2.3292,
+      "step": 1589
+    },
+    {
+      "epoch": 0.06949604440753529,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.88268489977046e-05,
+      "loss": 2.4083,
+      "step": 1590
+    },
+    {
+      "epoch": 0.0695397526115652,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.882536937026066e-05,
+      "loss": 2.4895,
+      "step": 1591
+    },
+    {
+      "epoch": 0.06958346081559509,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.882388882141092e-05,
+      "loss": 2.478,
+      "step": 1592
+    },
+    {
+      "epoch": 0.06962716901962498,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.882240735118334e-05,
+      "loss": 1.8026,
+      "step": 1593
+    },
+    {
+      "epoch": 0.06967087722365488,
+      "grad_norm": 2.375,
+      "learning_rate": 9.882092495960589e-05,
+      "loss": 1.8172,
+      "step": 1594
+    },
+    {
+      "epoch": 0.06971458542768477,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.881944164670651e-05,
+      "loss": 2.2879,
+      "step": 1595
+    },
+    {
+      "epoch": 0.06975829363171468,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.881795741251323e-05,
+      "loss": 2.2479,
+      "step": 1596
+    },
+    {
+      "epoch": 0.06980200183574457,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.881647225705403e-05,
+      "loss": 2.1813,
+      "step": 1597
+    },
+    {
+      "epoch": 0.06984571003977447,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.881498618035695e-05,
+      "loss": 2.2546,
+      "step": 1598
+    },
+    {
+      "epoch": 0.06988941824380436,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.881349918245005e-05,
+      "loss": 2.346,
+      "step": 1599
+    },
+    {
+      "epoch": 0.06993312644783425,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.881201126336135e-05,
+      "loss": 2.2346,
+      "step": 1600
+    },
+    {
+      "epoch": 0.06997683465186416,
+      "grad_norm": 2.375,
+      "learning_rate": 9.881052242311896e-05,
+      "loss": 1.7978,
+      "step": 1601
+    },
+    {
+      "epoch": 0.07002054285589406,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.880903266175098e-05,
+      "loss": 1.906,
+      "step": 1602
+    },
+    {
+      "epoch": 0.07006425105992395,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.880754197928553e-05,
+      "loss": 2.1055,
+      "step": 1603
+    },
+    {
+      "epoch": 0.07010795926395384,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.88060503757507e-05,
+      "loss": 2.0959,
+      "step": 1604
+    },
+    {
+      "epoch": 0.07015166746798374,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.880455785117469e-05,
+      "loss": 2.2334,
+      "step": 1605
+    },
+    {
+      "epoch": 0.07019537567201364,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.880306440558562e-05,
+      "loss": 2.0596,
+      "step": 1606
+    },
+    {
+      "epoch": 0.07023908387604354,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.880157003901171e-05,
+      "loss": 2.3092,
+      "step": 1607
+    },
+    {
+      "epoch": 0.07028279208007343,
+      "grad_norm": 2.75,
+      "learning_rate": 9.880007475148114e-05,
+      "loss": 1.8754,
+      "step": 1608
+    },
+    {
+      "epoch": 0.07032650028410332,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.879857854302214e-05,
+      "loss": 1.8996,
+      "step": 1609
+    },
+    {
+      "epoch": 0.07037020848813322,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.879708141366293e-05,
+      "loss": 2.2173,
+      "step": 1610
+    },
+    {
+      "epoch": 0.07041391669216313,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.879558336343177e-05,
+      "loss": 2.2584,
+      "step": 1611
+    },
+    {
+      "epoch": 0.07045762489619302,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.879408439235696e-05,
+      "loss": 1.9219,
+      "step": 1612
+    },
+    {
+      "epoch": 0.07050133310022291,
+      "grad_norm": 2.25,
+      "learning_rate": 9.879258450046673e-05,
+      "loss": 1.9287,
+      "step": 1613
+    },
+    {
+      "epoch": 0.0705450413042528,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.879108368778943e-05,
+      "loss": 2.1835,
+      "step": 1614
+    },
+    {
+      "epoch": 0.0705887495082827,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.878958195435338e-05,
+      "loss": 2.613,
+      "step": 1615
+    },
+    {
+      "epoch": 0.07063245771231261,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.878807930018689e-05,
+      "loss": 2.775,
+      "step": 1616
+    },
+    {
+      "epoch": 0.0706761659163425,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.878657572531833e-05,
+      "loss": 2.0104,
+      "step": 1617
+    },
+    {
+      "epoch": 0.0707198741203724,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.878507122977609e-05,
+      "loss": 2.1273,
+      "step": 1618
+    },
+    {
+      "epoch": 0.07076358232440229,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.878356581358856e-05,
+      "loss": 2.3979,
+      "step": 1619
+    },
+    {
+      "epoch": 0.07080729052843218,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.878205947678414e-05,
+      "loss": 2.65,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07085099873246209,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.878055221939127e-05,
+      "loss": 2.9008,
+      "step": 1621
+    },
+    {
+      "epoch": 0.07089470693649198,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.877904404143837e-05,
+      "loss": 1.728,
+      "step": 1622
+    },
+    {
+      "epoch": 0.07093841514052188,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.87775349429539e-05,
+      "loss": 2.2248,
+      "step": 1623
+    },
+    {
+      "epoch": 0.07098212334455177,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.877602492396636e-05,
+      "loss": 2.0064,
+      "step": 1624
+    },
+    {
+      "epoch": 0.07102583154858166,
+      "grad_norm": 3.25,
+      "learning_rate": 9.877451398450426e-05,
+      "loss": 2.1671,
+      "step": 1625
+    },
+    {
+      "epoch": 0.07106953975261157,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.877300212459608e-05,
+      "loss": 2.1163,
+      "step": 1626
+    },
+    {
+      "epoch": 0.07111324795664146,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.877148934427037e-05,
+      "loss": 2.1899,
+      "step": 1627
+    },
+    {
+      "epoch": 0.07115695616067136,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.876997564355565e-05,
+      "loss": 2.778,
+      "step": 1628
+    },
+    {
+      "epoch": 0.07120066436470125,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.876846102248053e-05,
+      "loss": 2.4448,
+      "step": 1629
+    },
+    {
+      "epoch": 0.07124437256873115,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.876694548107357e-05,
+      "loss": 2.8018,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07128808077276105,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.876542901936336e-05,
+      "loss": 2.128,
+      "step": 1631
+    },
+    {
+      "epoch": 0.07133178897679095,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.876391163737853e-05,
+      "loss": 2.2933,
+      "step": 1632
+    },
+    {
+      "epoch": 0.07137549718082084,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.876239333514772e-05,
+      "loss": 1.9584,
+      "step": 1633
+    },
+    {
+      "epoch": 0.07141920538485073,
+      "grad_norm": 3.75,
+      "learning_rate": 9.876087411269959e-05,
+      "loss": 2.2731,
+      "step": 1634
+    },
+    {
+      "epoch": 0.07146291358888063,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.875935397006278e-05,
+      "loss": 2.5422,
+      "step": 1635
+    },
+    {
+      "epoch": 0.07150662179291054,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.875783290726601e-05,
+      "loss": 2.1556,
+      "step": 1636
+    },
+    {
+      "epoch": 0.07155032999694043,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.875631092433795e-05,
+      "loss": 2.452,
+      "step": 1637
+    },
+    {
+      "epoch": 0.07159403820097032,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.875478802130736e-05,
+      "loss": 1.9157,
+      "step": 1638
+    },
+    {
+      "epoch": 0.07163774640500022,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.875326419820296e-05,
+      "loss": 2.4768,
+      "step": 1639
+    },
+    {
+      "epoch": 0.07168145460903011,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.87517394550535e-05,
+      "loss": 2.0934,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07172516281306002,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.875021379188776e-05,
+      "loss": 2.2172,
+      "step": 1641
+    },
+    {
+      "epoch": 0.07176887101708991,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.874868720873454e-05,
+      "loss": 2.0909,
+      "step": 1642
+    },
+    {
+      "epoch": 0.0718125792211198,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.874715970562262e-05,
+      "loss": 2.1855,
+      "step": 1643
+    },
+    {
+      "epoch": 0.0718562874251497,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.874563128258087e-05,
+      "loss": 2.1557,
+      "step": 1644
+    },
+    {
+      "epoch": 0.07189999562917959,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.874410193963813e-05,
+      "loss": 2.1594,
+      "step": 1645
+    },
+    {
+      "epoch": 0.0719437038332095,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.874257167682321e-05,
+      "loss": 2.2683,
+      "step": 1646
+    },
+    {
+      "epoch": 0.07198741203723939,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.874104049416502e-05,
+      "loss": 2.4891,
+      "step": 1647
+    },
+    {
+      "epoch": 0.07203112024126929,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.873950839169248e-05,
+      "loss": 2.486,
+      "step": 1648
+    },
+    {
+      "epoch": 0.07207482844529918,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.873797536943447e-05,
+      "loss": 2.0965,
+      "step": 1649
+    },
+    {
+      "epoch": 0.07211853664932907,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.873644142741992e-05,
+      "loss": 2.0246,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07216224485335898,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.87349065656778e-05,
+      "loss": 2.8423,
+      "step": 1651
+    },
+    {
+      "epoch": 0.07220595305738887,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.873337078423706e-05,
+      "loss": 2.1673,
+      "step": 1652
+    },
+    {
+      "epoch": 0.07224966126141877,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.873183408312668e-05,
+      "loss": 2.2967,
+      "step": 1653
+    },
+    {
+      "epoch": 0.07229336946544866,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.873029646237567e-05,
+      "loss": 1.8027,
+      "step": 1654
+    },
+    {
+      "epoch": 0.07233707766947856,
+      "grad_norm": 2.625,
+      "learning_rate": 9.872875792201304e-05,
+      "loss": 2.0913,
+      "step": 1655
+    },
+    {
+      "epoch": 0.07238078587350846,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.872721846206783e-05,
+      "loss": 2.0012,
+      "step": 1656
+    },
+    {
+      "epoch": 0.07242449407753836,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.872567808256909e-05,
+      "loss": 2.3273,
+      "step": 1657
+    },
+    {
+      "epoch": 0.07246820228156825,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.872413678354589e-05,
+      "loss": 2.4328,
+      "step": 1658
+    },
+    {
+      "epoch": 0.07251191048559814,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.87225945650273e-05,
+      "loss": 2.1735,
+      "step": 1659
+    },
+    {
+      "epoch": 0.07255561868962804,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.872105142704244e-05,
+      "loss": 2.0066,
+      "step": 1660
+    },
+    {
+      "epoch": 0.07259932689365795,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.871950736962044e-05,
+      "loss": 2.249,
+      "step": 1661
+    },
+    {
+      "epoch": 0.07264303509768784,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.871796239279043e-05,
+      "loss": 2.347,
+      "step": 1662
+    },
+    {
+      "epoch": 0.07268674330171773,
+      "grad_norm": 4.375,
+      "learning_rate": 9.871641649658155e-05,
+      "loss": 2.0257,
+      "step": 1663
+    },
+    {
+      "epoch": 0.07273045150574763,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.871486968102299e-05,
+      "loss": 2.1046,
+      "step": 1664
+    },
+    {
+      "epoch": 0.07277415970977752,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.871332194614395e-05,
+      "loss": 2.882,
+      "step": 1665
+    },
+    {
+      "epoch": 0.07281786791380743,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.871177329197362e-05,
+      "loss": 1.7936,
+      "step": 1666
+    },
+    {
+      "epoch": 0.07286157611783732,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.871022371854123e-05,
+      "loss": 1.7715,
+      "step": 1667
+    },
+    {
+      "epoch": 0.07290528432186721,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.870867322587602e-05,
+      "loss": 2.7403,
+      "step": 1668
+    },
+    {
+      "epoch": 0.07294899252589711,
+      "grad_norm": 3.0,
+      "learning_rate": 9.870712181400726e-05,
+      "loss": 2.8774,
+      "step": 1669
+    },
+    {
+      "epoch": 0.072992700729927,
+      "grad_norm": 3.375,
+      "learning_rate": 9.870556948296423e-05,
+      "loss": 2.4485,
+      "step": 1670
+    },
+    {
+      "epoch": 0.07303640893395691,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.87040162327762e-05,
+      "loss": 1.856,
+      "step": 1671
+    },
+    {
+      "epoch": 0.0730801171379868,
+      "grad_norm": 3.0,
+      "learning_rate": 9.870246206347252e-05,
+      "loss": 2.0979,
+      "step": 1672
+    },
+    {
+      "epoch": 0.0731238253420167,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.870090697508248e-05,
+      "loss": 1.9364,
+      "step": 1673
+    },
+    {
+      "epoch": 0.07316753354604659,
+      "grad_norm": 4.96875,
+      "learning_rate": 9.869935096763543e-05,
+      "loss": 2.922,
+      "step": 1674
+    },
+    {
+      "epoch": 0.07321124175007648,
+      "grad_norm": 2.625,
+      "learning_rate": 9.869779404116078e-05,
+      "loss": 2.2812,
+      "step": 1675
+    },
+    {
+      "epoch": 0.07325494995410639,
+      "grad_norm": 3.0,
+      "learning_rate": 9.869623619568786e-05,
+      "loss": 2.1977,
+      "step": 1676
+    },
+    {
+      "epoch": 0.07329865815813628,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.86946774312461e-05,
+      "loss": 1.9256,
+      "step": 1677
+    },
+    {
+      "epoch": 0.07334236636216618,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.86931177478649e-05,
+      "loss": 2.031,
+      "step": 1678
+    },
+    {
+      "epoch": 0.07338607456619607,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.86915571455737e-05,
+      "loss": 2.2388,
+      "step": 1679
+    },
+    {
+      "epoch": 0.07342978277022597,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.868999562440194e-05,
+      "loss": 2.7069,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07347349097425587,
+      "grad_norm": 2.875,
+      "learning_rate": 9.86884331843791e-05,
+      "loss": 2.6101,
+      "step": 1681
+    },
+    {
+      "epoch": 0.07351719917828577,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.868686982553468e-05,
+      "loss": 2.801,
+      "step": 1682
+    },
+    {
+      "epoch": 0.07356090738231566,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.868530554789815e-05,
+      "loss": 2.0419,
+      "step": 1683
+    },
+    {
+      "epoch": 0.07360461558634555,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.868374035149905e-05,
+      "loss": 2.0969,
+      "step": 1684
+    },
+    {
+      "epoch": 0.07364832379037545,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.868217423636693e-05,
+      "loss": 2.2815,
+      "step": 1685
+    },
+    {
+      "epoch": 0.07369203199440535,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.86806072025313e-05,
+      "loss": 2.4339,
+      "step": 1686
+    },
+    {
+      "epoch": 0.07373574019843525,
+      "grad_norm": 11.3125,
+      "learning_rate": 9.867903925002178e-05,
+      "loss": 2.8798,
+      "step": 1687
+    },
+    {
+      "epoch": 0.07377944840246514,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.867747037886793e-05,
+      "loss": 2.0418,
+      "step": 1688
+    },
+    {
+      "epoch": 0.07382315660649504,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.867590058909936e-05,
+      "loss": 2.1559,
+      "step": 1689
+    },
+    {
+      "epoch": 0.07386686481052493,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.867432988074572e-05,
+      "loss": 2.198,
+      "step": 1690
+    },
+    {
+      "epoch": 0.07391057301455484,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.867275825383664e-05,
+      "loss": 1.8589,
+      "step": 1691
+    },
+    {
+      "epoch": 0.07395428121858473,
+      "grad_norm": 2.875,
+      "learning_rate": 9.867118570840175e-05,
+      "loss": 2.4594,
+      "step": 1692
+    },
+    {
+      "epoch": 0.07399798942261462,
+      "grad_norm": 2.625,
+      "learning_rate": 9.866961224447075e-05,
+      "loss": 2.5562,
+      "step": 1693
+    },
+    {
+      "epoch": 0.07404169762664452,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.866803786207335e-05,
+      "loss": 2.3999,
+      "step": 1694
+    },
+    {
+      "epoch": 0.07408540583067441,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.866646256123922e-05,
+      "loss": 1.985,
+      "step": 1695
+    },
+    {
+      "epoch": 0.07412911403470432,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.866488634199813e-05,
+      "loss": 2.6449,
+      "step": 1696
+    },
+    {
+      "epoch": 0.07417282223873421,
+      "grad_norm": 2.75,
+      "learning_rate": 9.866330920437979e-05,
+      "loss": 2.1322,
+      "step": 1697
+    },
+    {
+      "epoch": 0.0742165304427641,
+      "grad_norm": 2.25,
+      "learning_rate": 9.8661731148414e-05,
+      "loss": 1.5896,
+      "step": 1698
+    },
+    {
+      "epoch": 0.074260238646794,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.86601521741305e-05,
+      "loss": 2.0734,
+      "step": 1699
+    },
+    {
+      "epoch": 0.0743039468508239,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.865857228155911e-05,
+      "loss": 2.4698,
+      "step": 1700
+    },
+    {
+      "epoch": 0.0743476550548538,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.865699147072964e-05,
+      "loss": 2.0935,
+      "step": 1701
+    },
+    {
+      "epoch": 0.0743913632588837,
+      "grad_norm": 3.0,
+      "learning_rate": 9.865540974167193e-05,
+      "loss": 2.4411,
+      "step": 1702
+    },
+    {
+      "epoch": 0.07443507146291359,
+      "grad_norm": 2.875,
+      "learning_rate": 9.865382709441584e-05,
+      "loss": 2.3764,
+      "step": 1703
+    },
+    {
+      "epoch": 0.07447877966694348,
+      "grad_norm": 2.75,
+      "learning_rate": 9.865224352899119e-05,
+      "loss": 2.1394,
+      "step": 1704
+    },
+    {
+      "epoch": 0.07452248787097338,
+      "grad_norm": 3.125,
+      "learning_rate": 9.865065904542792e-05,
+      "loss": 2.3577,
+      "step": 1705
+    },
+    {
+      "epoch": 0.07456619607500328,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.864907364375589e-05,
+      "loss": 2.5026,
+      "step": 1706
+    },
+    {
+      "epoch": 0.07460990427903318,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.864748732400504e-05,
+      "loss": 2.5778,
+      "step": 1707
+    },
+    {
+      "epoch": 0.07465361248306307,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.86459000862053e-05,
+      "loss": 2.1946,
+      "step": 1708
+    },
+    {
+      "epoch": 0.07469732068709296,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.864431193038662e-05,
+      "loss": 2.1463,
+      "step": 1709
+    },
+    {
+      "epoch": 0.07474102889112286,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.864272285657898e-05,
+      "loss": 2.2934,
+      "step": 1710
+    },
+    {
+      "epoch": 0.07478473709515276,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.864113286481237e-05,
+      "loss": 2.0675,
+      "step": 1711
+    },
+    {
+      "epoch": 0.07482844529918266,
+      "grad_norm": 2.5,
+      "learning_rate": 9.863954195511677e-05,
+      "loss": 2.022,
+      "step": 1712
+    },
+    {
+      "epoch": 0.07487215350321255,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.863795012752224e-05,
+      "loss": 2.0623,
+      "step": 1713
+    },
+    {
+      "epoch": 0.07491586170724245,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.863635738205881e-05,
+      "loss": 2.5347,
+      "step": 1714
+    },
+    {
+      "epoch": 0.07495956991127234,
+      "grad_norm": 6.125,
+      "learning_rate": 9.863476371875651e-05,
+      "loss": 2.5752,
+      "step": 1715
+    },
+    {
+      "epoch": 0.07500327811530225,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.863316913764545e-05,
+      "loss": 2.5674,
+      "step": 1716
+    },
+    {
+      "epoch": 0.07504698631933214,
+      "grad_norm": 3.5,
+      "learning_rate": 9.863157363875568e-05,
+      "loss": 2.4708,
+      "step": 1717
+    },
+    {
+      "epoch": 0.07509069452336203,
+      "grad_norm": 21.0,
+      "learning_rate": 9.862997722211735e-05,
+      "loss": 7.6248,
+      "step": 1718
+    },
+    {
+      "epoch": 0.07513440272739193,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.862837988776059e-05,
+      "loss": 1.906,
+      "step": 1719
+    },
+    {
+      "epoch": 0.07517811093142182,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.86267816357155e-05,
+      "loss": 2.6795,
+      "step": 1720
+    },
+    {
+      "epoch": 0.07522181913545173,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.86251824660123e-05,
+      "loss": 3.0829,
+      "step": 1721
+    },
+    {
+      "epoch": 0.07526552733948162,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.86235823786811e-05,
+      "loss": 2.1595,
+      "step": 1722
+    },
+    {
+      "epoch": 0.07530923554351152,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.862198137375215e-05,
+      "loss": 2.2193,
+      "step": 1723
+    },
+    {
+      "epoch": 0.07535294374754141,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.862037945125564e-05,
+      "loss": 2.0731,
+      "step": 1724
+    },
+    {
+      "epoch": 0.07539665195157132,
+      "grad_norm": 2.875,
+      "learning_rate": 9.86187766112218e-05,
+      "loss": 1.9634,
+      "step": 1725
+    },
+    {
+      "epoch": 0.07544036015560121,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.861717285368091e-05,
+      "loss": 2.0719,
+      "step": 1726
+    },
+    {
+      "epoch": 0.0754840683596311,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.861556817866318e-05,
+      "loss": 2.5894,
+      "step": 1727
+    },
+    {
+      "epoch": 0.075527776563661,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.861396258619894e-05,
+      "loss": 2.2967,
+      "step": 1728
+    },
+    {
+      "epoch": 0.07557148476769089,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.861235607631847e-05,
+      "loss": 1.8993,
+      "step": 1729
+    },
+    {
+      "epoch": 0.0756151929717208,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.861074864905207e-05,
+      "loss": 1.9114,
+      "step": 1730
+    },
+    {
+      "epoch": 0.07565890117575069,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.860914030443012e-05,
+      "loss": 1.9079,
+      "step": 1731
+    },
+    {
+      "epoch": 0.07570260937978059,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.860753104248292e-05,
+      "loss": 2.5769,
+      "step": 1732
+    },
+    {
+      "epoch": 0.07574631758381048,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.860592086324088e-05,
+      "loss": 2.4568,
+      "step": 1733
+    },
+    {
+      "epoch": 0.07579002578784037,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.860430976673436e-05,
+      "loss": 2.1262,
+      "step": 1734
+    },
+    {
+      "epoch": 0.07583373399187028,
+      "grad_norm": 2.625,
+      "learning_rate": 9.86026977529938e-05,
+      "loss": 1.8477,
+      "step": 1735
+    },
+    {
+      "epoch": 0.07587744219590017,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.860108482204957e-05,
+      "loss": 2.0501,
+      "step": 1736
+    },
+    {
+      "epoch": 0.07592115039993007,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.859947097393215e-05,
+      "loss": 2.7278,
+      "step": 1737
+    },
+    {
+      "epoch": 0.07596485860395996,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.859785620867197e-05,
+      "loss": 2.0078,
+      "step": 1738
+    },
+    {
+      "epoch": 0.07600856680798986,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.859624052629951e-05,
+      "loss": 2.1657,
+      "step": 1739
+    },
+    {
+      "epoch": 0.07605227501201976,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.859462392684526e-05,
+      "loss": 2.1615,
+      "step": 1740
+    },
+    {
+      "epoch": 0.07609598321604966,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.859300641033974e-05,
+      "loss": 2.0125,
+      "step": 1741
+    },
+    {
+      "epoch": 0.07613969142007955,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.859138797681347e-05,
+      "loss": 2.6558,
+      "step": 1742
+    },
+    {
+      "epoch": 0.07618339962410944,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.858976862629698e-05,
+      "loss": 2.9906,
+      "step": 1743
+    },
+    {
+      "epoch": 0.07622710782813934,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.858814835882085e-05,
+      "loss": 2.4754,
+      "step": 1744
+    },
+    {
+      "epoch": 0.07627081603216924,
+      "grad_norm": 3.0,
+      "learning_rate": 9.85865271744156e-05,
+      "loss": 2.1856,
+      "step": 1745
+    },
+    {
+      "epoch": 0.07631452423619914,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.85849050731119e-05,
+      "loss": 2.4571,
+      "step": 1746
+    },
+    {
+      "epoch": 0.07635823244022903,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.858328205494035e-05,
+      "loss": 1.9271,
+      "step": 1747
+    },
+    {
+      "epoch": 0.07640194064425893,
+      "grad_norm": 2.75,
+      "learning_rate": 9.858165811993153e-05,
+      "loss": 2.4178,
+      "step": 1748
+    },
+    {
+      "epoch": 0.07644564884828882,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.858003326811611e-05,
+      "loss": 2.3161,
+      "step": 1749
+    },
+    {
+      "epoch": 0.07648935705231873,
+      "grad_norm": 5.15625,
+      "learning_rate": 9.857840749952478e-05,
+      "loss": 1.7724,
+      "step": 1750
+    },
+    {
+      "epoch": 0.07653306525634862,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.857678081418818e-05,
+      "loss": 2.2475,
+      "step": 1751
+    },
+    {
+      "epoch": 0.07657677346037851,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.857515321213704e-05,
+      "loss": 1.9093,
+      "step": 1752
+    },
+    {
+      "epoch": 0.07662048166440841,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.857352469340204e-05,
+      "loss": 2.1066,
+      "step": 1753
+    },
+    {
+      "epoch": 0.0766641898684383,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.857189525801396e-05,
+      "loss": 1.8703,
+      "step": 1754
+    },
+    {
+      "epoch": 0.07670789807246821,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.857026490600349e-05,
+      "loss": 2.1322,
+      "step": 1755
+    },
+    {
+      "epoch": 0.0767516062764981,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.856863363740145e-05,
+      "loss": 2.6255,
+      "step": 1756
+    },
+    {
+      "epoch": 0.076795314480528,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.856700145223862e-05,
+      "loss": 1.8698,
+      "step": 1757
+    },
+    {
+      "epoch": 0.07683902268455789,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.856536835054577e-05,
+      "loss": 3.1263,
+      "step": 1758
+    },
+    {
+      "epoch": 0.07688273088858778,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.856373433235373e-05,
+      "loss": 1.8888,
+      "step": 1759
+    },
+    {
+      "epoch": 0.07692643909261769,
+      "grad_norm": 2.5,
+      "learning_rate": 9.856209939769335e-05,
+      "loss": 2.0877,
+      "step": 1760
+    },
+    {
+      "epoch": 0.07697014729664758,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.856046354659547e-05,
+      "loss": 2.0977,
+      "step": 1761
+    },
+    {
+      "epoch": 0.07701385550067748,
+      "grad_norm": 2.625,
+      "learning_rate": 9.855882677909099e-05,
+      "loss": 2.3225,
+      "step": 1762
+    },
+    {
+      "epoch": 0.07705756370470737,
+      "grad_norm": 2.375,
+      "learning_rate": 9.855718909521075e-05,
+      "loss": 1.7009,
+      "step": 1763
+    },
+    {
+      "epoch": 0.07710127190873726,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.855555049498568e-05,
+      "loss": 2.2034,
+      "step": 1764
+    },
+    {
+      "epoch": 0.07714498011276717,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.85539109784467e-05,
+      "loss": 1.5595,
+      "step": 1765
+    },
+    {
+      "epoch": 0.07718868831679707,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.855227054562476e-05,
+      "loss": 2.9596,
+      "step": 1766
+    },
+    {
+      "epoch": 0.07723239652082696,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.855062919655083e-05,
+      "loss": 1.9394,
+      "step": 1767
+    },
+    {
+      "epoch": 0.07727610472485685,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.854898693125586e-05,
+      "loss": 2.1817,
+      "step": 1768
+    },
+    {
+      "epoch": 0.07731981292888675,
+      "grad_norm": 3.8125,
+      "learning_rate": 9.854734374977081e-05,
+      "loss": 2.0313,
+      "step": 1769
+    },
+    {
+      "epoch": 0.07736352113291665,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.854569965212676e-05,
+      "loss": 1.8892,
+      "step": 1770
+    },
+    {
+      "epoch": 0.07740722933694655,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.854405463835468e-05,
+      "loss": 2.2463,
+      "step": 1771
+    },
+    {
+      "epoch": 0.07745093754097644,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.854240870848565e-05,
+      "loss": 1.8787,
+      "step": 1772
+    },
+    {
+      "epoch": 0.07749464574500634,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.854076186255072e-05,
+      "loss": 2.0774,
+      "step": 1773
+    },
+    {
+      "epoch": 0.07753835394903623,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.853911410058097e-05,
+      "loss": 2.1188,
+      "step": 1774
+    },
+    {
+      "epoch": 0.07758206215306614,
+      "grad_norm": 5.28125,
+      "learning_rate": 9.853746542260749e-05,
+      "loss": 2.0926,
+      "step": 1775
+    },
+    {
+      "epoch": 0.07762577035709603,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.853581582866139e-05,
+      "loss": 2.7162,
+      "step": 1776
+    },
+    {
+      "epoch": 0.07766947856112592,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.85341653187738e-05,
+      "loss": 2.7719,
+      "step": 1777
+    },
+    {
+      "epoch": 0.07771318676515582,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.853251389297587e-05,
+      "loss": 2.4834,
+      "step": 1778
+    },
+    {
+      "epoch": 0.07775689496918571,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.853086155129878e-05,
+      "loss": 2.0547,
+      "step": 1779
+    },
+    {
+      "epoch": 0.07780060317321562,
+      "grad_norm": 2.5,
+      "learning_rate": 9.852920829377369e-05,
+      "loss": 2.4811,
+      "step": 1780
+    },
+    {
+      "epoch": 0.07784431137724551,
+      "grad_norm": 3.25,
+      "learning_rate": 9.85275541204318e-05,
+      "loss": 2.6087,
+      "step": 1781
+    },
+    {
+      "epoch": 0.0778880195812754,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.852589903130435e-05,
+      "loss": 2.869,
+      "step": 1782
+    },
+    {
+      "epoch": 0.0779317277853053,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.852424302642256e-05,
+      "loss": 1.683,
+      "step": 1783
+    },
+    {
+      "epoch": 0.07797543598933519,
+      "grad_norm": 2.75,
+      "learning_rate": 9.852258610581768e-05,
+      "loss": 2.5439,
+      "step": 1784
+    },
+    {
+      "epoch": 0.0780191441933651,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.852092826952097e-05,
+      "loss": 1.7083,
+      "step": 1785
+    },
+    {
+      "epoch": 0.078062852397395,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.851926951756374e-05,
+      "loss": 2.2147,
+      "step": 1786
+    },
+    {
+      "epoch": 0.07810656060142489,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.851760984997727e-05,
+      "loss": 1.9388,
+      "step": 1787
+    },
+    {
+      "epoch": 0.07815026880545478,
+      "grad_norm": 4.125,
+      "learning_rate": 9.851594926679287e-05,
+      "loss": 2.1677,
+      "step": 1788
+    },
+    {
+      "epoch": 0.07819397700948467,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.851428776804191e-05,
+      "loss": 3.2133,
+      "step": 1789
+    },
+    {
+      "epoch": 0.07823768521351458,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.851262535375574e-05,
+      "loss": 2.3131,
+      "step": 1790
+    },
+    {
+      "epoch": 0.07828139341754448,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.851096202396572e-05,
+      "loss": 2.0723,
+      "step": 1791
+    },
+    {
+      "epoch": 0.07832510162157437,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.850929777870324e-05,
+      "loss": 2.3914,
+      "step": 1792
+    },
+    {
+      "epoch": 0.07836880982560426,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.850763261799969e-05,
+      "loss": 1.8671,
+      "step": 1793
+    },
+    {
+      "epoch": 0.07841251802963416,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.850596654188653e-05,
+      "loss": 2.0209,
+      "step": 1794
+    },
+    {
+      "epoch": 0.07845622623366406,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.850429955039518e-05,
+      "loss": 2.0369,
+      "step": 1795
+    },
+    {
+      "epoch": 0.07849993443769396,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.85026316435571e-05,
+      "loss": 2.1397,
+      "step": 1796
+    },
+    {
+      "epoch": 0.07854364264172385,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.850096282140379e-05,
+      "loss": 1.6577,
+      "step": 1797
+    },
+    {
+      "epoch": 0.07858735084575375,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.84992930839667e-05,
+      "loss": 2.19,
+      "step": 1798
+    },
+    {
+      "epoch": 0.07863105904978364,
+      "grad_norm": 3.375,
+      "learning_rate": 9.849762243127737e-05,
+      "loss": 1.9441,
+      "step": 1799
+    },
+    {
+      "epoch": 0.07867476725381355,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.849595086336732e-05,
+      "loss": 2.5172,
+      "step": 1800
+    },
+    {
+      "epoch": 0.07871847545784344,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.84942783802681e-05,
+      "loss": 1.7464,
+      "step": 1801
+    },
+    {
+      "epoch": 0.07876218366187333,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.849260498201126e-05,
+      "loss": 2.0358,
+      "step": 1802
+    },
+    {
+      "epoch": 0.07880589186590323,
+      "grad_norm": 2.75,
+      "learning_rate": 9.849093066862837e-05,
+      "loss": 2.3709,
+      "step": 1803
+    },
+    {
+      "epoch": 0.07884960006993312,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.848925544015106e-05,
+      "loss": 2.3652,
+      "step": 1804
+    },
+    {
+      "epoch": 0.07889330827396303,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.848757929661095e-05,
+      "loss": 1.8576,
+      "step": 1805
+    },
+    {
+      "epoch": 0.07893701647799292,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.848590223803961e-05,
+      "loss": 1.9136,
+      "step": 1806
+    },
+    {
+      "epoch": 0.07898072468202282,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.848422426446875e-05,
+      "loss": 2.979,
+      "step": 1807
+    },
+    {
+      "epoch": 0.07902443288605271,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.848254537593e-05,
+      "loss": 2.4617,
+      "step": 1808
+    },
+    {
+      "epoch": 0.0790681410900826,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.848086557245507e-05,
+      "loss": 1.9116,
+      "step": 1809
+    },
+    {
+      "epoch": 0.07911184929411251,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.847918485407563e-05,
+      "loss": 2.1013,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0791555574981424,
+      "grad_norm": 2.75,
+      "learning_rate": 9.847750322082341e-05,
+      "loss": 2.1538,
+      "step": 1811
+    },
+    {
+      "epoch": 0.0791992657021723,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.847582067273015e-05,
+      "loss": 2.1033,
+      "step": 1812
+    },
+    {
+      "epoch": 0.07924297390620219,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.847413720982763e-05,
+      "loss": 2.1661,
+      "step": 1813
+    },
+    {
+      "epoch": 0.07928668211023208,
+      "grad_norm": 2.375,
+      "learning_rate": 9.847245283214757e-05,
+      "loss": 1.8955,
+      "step": 1814
+    },
+    {
+      "epoch": 0.07933039031426199,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.847076753972176e-05,
+      "loss": 1.9043,
+      "step": 1815
+    },
+    {
+      "epoch": 0.07937409851829189,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.846908133258204e-05,
+      "loss": 1.8656,
+      "step": 1816
+    },
+    {
+      "epoch": 0.07941780672232178,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.846739421076022e-05,
+      "loss": 2.4502,
+      "step": 1817
+    },
+    {
+      "epoch": 0.07946151492635167,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.846570617428811e-05,
+      "loss": 2.2399,
+      "step": 1818
+    },
+    {
+      "epoch": 0.07950522313038157,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.84640172231976e-05,
+      "loss": 2.2902,
+      "step": 1819
+    },
+    {
+      "epoch": 0.07954893133441147,
+      "grad_norm": 5.0625,
+      "learning_rate": 9.846232735752055e-05,
+      "loss": 1.7842,
+      "step": 1820
+    },
+    {
+      "epoch": 0.07959263953844137,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.846063657728884e-05,
+      "loss": 1.8978,
+      "step": 1821
+    },
+    {
+      "epoch": 0.07963634774247126,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.845894488253438e-05,
+      "loss": 2.3265,
+      "step": 1822
+    },
+    {
+      "epoch": 0.07968005594650115,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.84572522732891e-05,
+      "loss": 1.7457,
+      "step": 1823
+    },
+    {
+      "epoch": 0.07972376415053105,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.845555874958496e-05,
+      "loss": 2.2468,
+      "step": 1824
+    },
+    {
+      "epoch": 0.07976747235456096,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.84538643114539e-05,
+      "loss": 2.579,
+      "step": 1825
+    },
+    {
+      "epoch": 0.07981118055859085,
+      "grad_norm": 12.9375,
+      "learning_rate": 9.84521689589279e-05,
+      "loss": 1.3026,
+      "step": 1826
+    },
+    {
+      "epoch": 0.07985488876262074,
+      "grad_norm": 3.625,
+      "learning_rate": 9.845047269203895e-05,
+      "loss": 2.2743,
+      "step": 1827
+    },
+    {
+      "epoch": 0.07989859696665064,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.844877551081906e-05,
+      "loss": 1.8684,
+      "step": 1828
+    },
+    {
+      "epoch": 0.07994230517068053,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.844707741530026e-05,
+      "loss": 2.2146,
+      "step": 1829
+    },
+    {
+      "epoch": 0.07998601337471044,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.844537840551462e-05,
+      "loss": 2.2717,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08002972157874033,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.844367848149417e-05,
+      "loss": 1.9394,
+      "step": 1831
+    },
+    {
+      "epoch": 0.08007342978277023,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.844197764327097e-05,
+      "loss": 2.0863,
+      "step": 1832
+    },
+    {
+      "epoch": 0.08011713798680012,
+      "grad_norm": 2.75,
+      "learning_rate": 9.844027589087719e-05,
+      "loss": 2.6755,
+      "step": 1833
+    },
+    {
+      "epoch": 0.08016084619083001,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.843857322434487e-05,
+      "loss": 2.1548,
+      "step": 1834
+    },
+    {
+      "epoch": 0.08020455439485992,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.84368696437062e-05,
+      "loss": 2.247,
+      "step": 1835
+    },
+    {
+      "epoch": 0.08024826259888981,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.843516514899329e-05,
+      "loss": 1.8082,
+      "step": 1836
+    },
+    {
+      "epoch": 0.08029197080291971,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.843345974023832e-05,
+      "loss": 2.3728,
+      "step": 1837
+    },
+    {
+      "epoch": 0.0803356790069496,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.843175341747348e-05,
+      "loss": 2.3455,
+      "step": 1838
+    },
+    {
+      "epoch": 0.0803793872109795,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.843004618073096e-05,
+      "loss": 2.3092,
+      "step": 1839
+    },
+    {
+      "epoch": 0.0804230954150094,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.842833803004298e-05,
+      "loss": 1.8681,
+      "step": 1840
+    },
+    {
+      "epoch": 0.0804668036190393,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.842662896544176e-05,
+      "loss": 1.9617,
+      "step": 1841
+    },
+    {
+      "epoch": 0.08051051182306919,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.84249189869596e-05,
+      "loss": 2.3176,
+      "step": 1842
+    },
+    {
+      "epoch": 0.08055422002709908,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.842320809462873e-05,
+      "loss": 2.0864,
+      "step": 1843
+    },
+    {
+      "epoch": 0.08059792823112898,
+      "grad_norm": 3.125,
+      "learning_rate": 9.842149628848145e-05,
+      "loss": 1.9941,
+      "step": 1844
+    },
+    {
+      "epoch": 0.08064163643515888,
+      "grad_norm": 4.25,
+      "learning_rate": 9.841978356855005e-05,
+      "loss": 2.5347,
+      "step": 1845
+    },
+    {
+      "epoch": 0.08068534463918878,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.841806993486686e-05,
+      "loss": 2.5548,
+      "step": 1846
+    },
+    {
+      "epoch": 0.08072905284321867,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.841635538746424e-05,
+      "loss": 2.0899,
+      "step": 1847
+    },
+    {
+      "epoch": 0.08077276104724856,
+      "grad_norm": 2.625,
+      "learning_rate": 9.841463992637451e-05,
+      "loss": 2.5336,
+      "step": 1848
+    },
+    {
+      "epoch": 0.08081646925127846,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.841292355163008e-05,
+      "loss": 1.8719,
+      "step": 1849
+    },
+    {
+      "epoch": 0.08086017745530837,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.84112062632633e-05,
+      "loss": 2.7897,
+      "step": 1850
+    },
+    {
+      "epoch": 0.08090388565933826,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.84094880613066e-05,
+      "loss": 3.1195,
+      "step": 1851
+    },
+    {
+      "epoch": 0.08094759386336815,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.840776894579241e-05,
+      "loss": 1.8838,
+      "step": 1852
+    },
+    {
+      "epoch": 0.08099130206739805,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.840604891675317e-05,
+      "loss": 1.8534,
+      "step": 1853
+    },
+    {
+      "epoch": 0.08103501027142794,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.840432797422132e-05,
+      "loss": 1.9414,
+      "step": 1854
+    },
+    {
+      "epoch": 0.08107871847545785,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.840260611822938e-05,
+      "loss": 2.4429,
+      "step": 1855
+    },
+    {
+      "epoch": 0.08112242667948774,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.84008833488098e-05,
+      "loss": 2.4672,
+      "step": 1856
+    },
+    {
+      "epoch": 0.08116613488351763,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.83991596659951e-05,
+      "loss": 2.3259,
+      "step": 1857
+    },
+    {
+      "epoch": 0.08120984308754753,
+      "grad_norm": 5.65625,
+      "learning_rate": 9.839743506981782e-05,
+      "loss": 1.7876,
+      "step": 1858
+    },
+    {
+      "epoch": 0.08125355129157742,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.83957095603105e-05,
+      "loss": 2.2488,
+      "step": 1859
+    },
+    {
+      "epoch": 0.08129725949560733,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.839398313750571e-05,
+      "loss": 2.2605,
+      "step": 1860
+    },
+    {
+      "epoch": 0.08134096769963722,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.839225580143601e-05,
+      "loss": 1.7144,
+      "step": 1861
+    },
+    {
+      "epoch": 0.08138467590366712,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.839052755213403e-05,
+      "loss": 2.2738,
+      "step": 1862
+    },
+    {
+      "epoch": 0.08142838410769701,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.838879838963235e-05,
+      "loss": 1.904,
+      "step": 1863
+    },
+    {
+      "epoch": 0.08147209231172692,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.838706831396362e-05,
+      "loss": 2.7066,
+      "step": 1864
+    },
+    {
+      "epoch": 0.08151580051575681,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.838533732516051e-05,
+      "loss": 1.9853,
+      "step": 1865
+    },
+    {
+      "epoch": 0.0815595087197867,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.838360542325565e-05,
+      "loss": 2.0883,
+      "step": 1866
+    },
+    {
+      "epoch": 0.0816032169238166,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.838187260828173e-05,
+      "loss": 2.198,
+      "step": 1867
+    },
+    {
+      "epoch": 0.08164692512784649,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.838013888027145e-05,
+      "loss": 2.0494,
+      "step": 1868
+    },
+    {
+      "epoch": 0.0816906333318764,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.837840423925755e-05,
+      "loss": 1.6796,
+      "step": 1869
+    },
+    {
+      "epoch": 0.0817343415359063,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.837666868527274e-05,
+      "loss": 2.2668,
+      "step": 1870
+    },
+    {
+      "epoch": 0.08177804973993619,
+      "grad_norm": 2.375,
+      "learning_rate": 9.837493221834979e-05,
+      "loss": 2.0788,
+      "step": 1871
+    },
+    {
+      "epoch": 0.08182175794396608,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.837319483852147e-05,
+      "loss": 1.9986,
+      "step": 1872
+    },
+    {
+      "epoch": 0.08186546614799597,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.837145654582054e-05,
+      "loss": 2.0753,
+      "step": 1873
+    },
+    {
+      "epoch": 0.08190917435202588,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.836971734027985e-05,
+      "loss": 2.1466,
+      "step": 1874
+    },
+    {
+      "epoch": 0.08195288255605578,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.836797722193217e-05,
+      "loss": 2.1172,
+      "step": 1875
+    },
+    {
+      "epoch": 0.08199659076008567,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.836623619081039e-05,
+      "loss": 2.0186,
+      "step": 1876
+    },
+    {
+      "epoch": 0.08204029896411556,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.836449424694732e-05,
+      "loss": 1.7065,
+      "step": 1877
+    },
+    {
+      "epoch": 0.08208400716814546,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.836275139037585e-05,
+      "loss": 2.7346,
+      "step": 1878
+    },
+    {
+      "epoch": 0.08212771537217536,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.836100762112888e-05,
+      "loss": 2.1246,
+      "step": 1879
+    },
+    {
+      "epoch": 0.08217142357620526,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.83592629392393e-05,
+      "loss": 2.2013,
+      "step": 1880
+    },
+    {
+      "epoch": 0.08221513178023515,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.835751734474005e-05,
+      "loss": 2.2353,
+      "step": 1881
+    },
+    {
+      "epoch": 0.08225883998426504,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.835577083766407e-05,
+      "loss": 2.8191,
+      "step": 1882
+    },
+    {
+      "epoch": 0.08230254818829494,
+      "grad_norm": 3.25,
+      "learning_rate": 9.835402341804432e-05,
+      "loss": 2.6574,
+      "step": 1883
+    },
+    {
+      "epoch": 0.08234625639232485,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.835227508591376e-05,
+      "loss": 2.3453,
+      "step": 1884
+    },
+    {
+      "epoch": 0.08238996459635474,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.835052584130541e-05,
+      "loss": 1.847,
+      "step": 1885
+    },
+    {
+      "epoch": 0.08243367280038463,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.834877568425225e-05,
+      "loss": 1.968,
+      "step": 1886
+    },
+    {
+      "epoch": 0.08247738100441453,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.834702461478733e-05,
+      "loss": 1.6705,
+      "step": 1887
+    },
+    {
+      "epoch": 0.08252108920844442,
+      "grad_norm": 3.125,
+      "learning_rate": 9.834527263294371e-05,
+      "loss": 2.1591,
+      "step": 1888
+    },
+    {
+      "epoch": 0.08256479741247433,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.834351973875441e-05,
+      "loss": 1.9738,
+      "step": 1889
+    },
+    {
+      "epoch": 0.08260850561650422,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.834176593225254e-05,
+      "loss": 2.5599,
+      "step": 1890
+    },
+    {
+      "epoch": 0.08265221382053412,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.834001121347119e-05,
+      "loss": 1.9454,
+      "step": 1891
+    },
+    {
+      "epoch": 0.08269592202456401,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.833825558244347e-05,
+      "loss": 1.8525,
+      "step": 1892
+    },
+    {
+      "epoch": 0.0827396302285939,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.83364990392025e-05,
+      "loss": 2.792,
+      "step": 1893
+    },
+    {
+      "epoch": 0.08278333843262381,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.833474158378147e-05,
+      "loss": 2.171,
+      "step": 1894
+    },
+    {
+      "epoch": 0.0828270466366537,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.833298321621349e-05,
+      "loss": 2.5549,
+      "step": 1895
+    },
+    {
+      "epoch": 0.0828707548406836,
+      "grad_norm": 2.625,
+      "learning_rate": 9.833122393653178e-05,
+      "loss": 2.6365,
+      "step": 1896
+    },
+    {
+      "epoch": 0.08291446304471349,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.832946374476954e-05,
+      "loss": 2.2036,
+      "step": 1897
+    },
+    {
+      "epoch": 0.08295817124874338,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.832770264095998e-05,
+      "loss": 2.0949,
+      "step": 1898
+    },
+    {
+      "epoch": 0.08300187945277329,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.832594062513634e-05,
+      "loss": 1.9104,
+      "step": 1899
+    },
+    {
+      "epoch": 0.08304558765680319,
+      "grad_norm": 3.25,
+      "learning_rate": 9.832417769733185e-05,
+      "loss": 2.0378,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08308929586083308,
+      "grad_norm": 2.625,
+      "learning_rate": 9.832241385757981e-05,
+      "loss": 2.1519,
+      "step": 1901
+    },
+    {
+      "epoch": 0.08313300406486297,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.832064910591348e-05,
+      "loss": 2.804,
+      "step": 1902
+    },
+    {
+      "epoch": 0.08317671226889287,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.831888344236617e-05,
+      "loss": 1.7018,
+      "step": 1903
+    },
+    {
+      "epoch": 0.08322042047292277,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.83171168669712e-05,
+      "loss": 2.1303,
+      "step": 1904
+    },
+    {
+      "epoch": 0.08326412867695267,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.831534937976192e-05,
+      "loss": 2.5991,
+      "step": 1905
+    },
+    {
+      "epoch": 0.08330783688098256,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.831358098077168e-05,
+      "loss": 2.0354,
+      "step": 1906
+    },
+    {
+      "epoch": 0.08335154508501245,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.831181167003385e-05,
+      "loss": 1.8323,
+      "step": 1907
+    },
+    {
+      "epoch": 0.08339525328904235,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.831004144758183e-05,
+      "loss": 1.5327,
+      "step": 1908
+    },
+    {
+      "epoch": 0.08343896149307226,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.8308270313449e-05,
+      "loss": 1.8952,
+      "step": 1909
+    },
+    {
+      "epoch": 0.08348266969710215,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.83064982676688e-05,
+      "loss": 2.3476,
+      "step": 1910
+    },
+    {
+      "epoch": 0.08352637790113204,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.830472531027468e-05,
+      "loss": 2.3885,
+      "step": 1911
+    },
+    {
+      "epoch": 0.08357008610516194,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.830295144130009e-05,
+      "loss": 2.0377,
+      "step": 1912
+    },
+    {
+      "epoch": 0.08361379430919183,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.830117666077849e-05,
+      "loss": 3.0738,
+      "step": 1913
+    },
+    {
+      "epoch": 0.08365750251322174,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.82994009687434e-05,
+      "loss": 2.0192,
+      "step": 1914
+    },
+    {
+      "epoch": 0.08370121071725163,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.829762436522831e-05,
+      "loss": 2.0367,
+      "step": 1915
+    },
+    {
+      "epoch": 0.08374491892128152,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.829584685026676e-05,
+      "loss": 2.0879,
+      "step": 1916
+    },
+    {
+      "epoch": 0.08378862712531142,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.829406842389229e-05,
+      "loss": 3.2253,
+      "step": 1917
+    },
+    {
+      "epoch": 0.08383233532934131,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.829228908613845e-05,
+      "loss": 1.9372,
+      "step": 1918
+    },
+    {
+      "epoch": 0.08387604353337122,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.829050883703882e-05,
+      "loss": 1.9538,
+      "step": 1919
+    },
+    {
+      "epoch": 0.08391975173740111,
+      "grad_norm": 2.75,
+      "learning_rate": 9.828872767662705e-05,
+      "loss": 2.3358,
+      "step": 1920
+    },
+    {
+      "epoch": 0.083963459941431,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.828694560493667e-05,
+      "loss": 2.8138,
+      "step": 1921
+    },
+    {
+      "epoch": 0.0840071681454609,
+      "grad_norm": 8.0,
+      "learning_rate": 9.828516262200135e-05,
+      "loss": 2.7766,
+      "step": 1922
+    },
+    {
+      "epoch": 0.0840508763494908,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.828337872785475e-05,
+      "loss": 2.4589,
+      "step": 1923
+    },
+    {
+      "epoch": 0.0840945845535207,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.828159392253052e-05,
+      "loss": 2.168,
+      "step": 1924
+    },
+    {
+      "epoch": 0.0841382927575506,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.827980820606232e-05,
+      "loss": 2.1371,
+      "step": 1925
+    },
+    {
+      "epoch": 0.08418200096158049,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.82780215784839e-05,
+      "loss": 2.5149,
+      "step": 1926
+    },
+    {
+      "epoch": 0.08422570916561038,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.827623403982892e-05,
+      "loss": 2.0682,
+      "step": 1927
+    },
+    {
+      "epoch": 0.08426941736964028,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.827444559013115e-05,
+      "loss": 2.2317,
+      "step": 1928
+    },
+    {
+      "epoch": 0.08431312557367018,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.827265622942434e-05,
+      "loss": 1.63,
+      "step": 1929
+    },
+    {
+      "epoch": 0.08435683377770008,
+      "grad_norm": 2.75,
+      "learning_rate": 9.827086595774225e-05,
+      "loss": 1.8615,
+      "step": 1930
+    },
+    {
+      "epoch": 0.08440054198172997,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.826907477511865e-05,
+      "loss": 2.2187,
+      "step": 1931
+    },
+    {
+      "epoch": 0.08444425018575986,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.826728268158737e-05,
+      "loss": 2.1593,
+      "step": 1932
+    },
+    {
+      "epoch": 0.08448795838978976,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.826548967718221e-05,
+      "loss": 2.1498,
+      "step": 1933
+    },
+    {
+      "epoch": 0.08453166659381967,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.826369576193701e-05,
+      "loss": 2.1045,
+      "step": 1934
+    },
+    {
+      "epoch": 0.08457537479784956,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.826190093588563e-05,
+      "loss": 2.108,
+      "step": 1935
+    },
+    {
+      "epoch": 0.08461908300187945,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.826010519906194e-05,
+      "loss": 2.1962,
+      "step": 1936
+    },
+    {
+      "epoch": 0.08466279120590935,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.825830855149983e-05,
+      "loss": 2.0458,
+      "step": 1937
+    },
+    {
+      "epoch": 0.08470649940993924,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.82565109932332e-05,
+      "loss": 2.0363,
+      "step": 1938
+    },
+    {
+      "epoch": 0.08475020761396915,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.825471252429596e-05,
+      "loss": 1.9519,
+      "step": 1939
+    },
+    {
+      "epoch": 0.08479391581799904,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.825291314472208e-05,
+      "loss": 2.0237,
+      "step": 1940
+    },
+    {
+      "epoch": 0.08483762402202893,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.82511128545455e-05,
+      "loss": 2.245,
+      "step": 1941
+    },
+    {
+      "epoch": 0.08488133222605883,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.824931165380018e-05,
+      "loss": 3.7645,
+      "step": 1942
+    },
+    {
+      "epoch": 0.08492504043008872,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.824750954252014e-05,
+      "loss": 2.355,
+      "step": 1943
+    },
+    {
+      "epoch": 0.08496874863411863,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.824570652073935e-05,
+      "loss": 2.4083,
+      "step": 1944
+    },
+    {
+      "epoch": 0.08501245683814852,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.824390258849188e-05,
+      "loss": 2.2051,
+      "step": 1945
+    },
+    {
+      "epoch": 0.08505616504217842,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.824209774581174e-05,
+      "loss": 2.7646,
+      "step": 1946
+    },
+    {
+      "epoch": 0.08509987324620831,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.824029199273302e-05,
+      "loss": 3.1808,
+      "step": 1947
+    },
+    {
+      "epoch": 0.0851435814502382,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.823848532928978e-05,
+      "loss": 2.0116,
+      "step": 1948
+    },
+    {
+      "epoch": 0.08518728965426811,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.823667775551611e-05,
+      "loss": 2.1517,
+      "step": 1949
+    },
+    {
+      "epoch": 0.085230997858298,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.823486927144612e-05,
+      "loss": 2.1716,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0852747060623279,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.823305987711393e-05,
+      "loss": 1.7987,
+      "step": 1951
+    },
+    {
+      "epoch": 0.08531841426635779,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.823124957255372e-05,
+      "loss": 1.7863,
+      "step": 1952
+    },
+    {
+      "epoch": 0.08536212247038769,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.822943835779963e-05,
+      "loss": 1.958,
+      "step": 1953
+    },
+    {
+      "epoch": 0.0854058306744176,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.822762623288584e-05,
+      "loss": 2.1187,
+      "step": 1954
+    },
+    {
+      "epoch": 0.08544953887844749,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.822581319784656e-05,
+      "loss": 1.9922,
+      "step": 1955
+    },
+    {
+      "epoch": 0.08549324708247738,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.822399925271598e-05,
+      "loss": 2.26,
+      "step": 1956
+    },
+    {
+      "epoch": 0.08553695528650727,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.822218439752834e-05,
+      "loss": 1.9014,
+      "step": 1957
+    },
+    {
+      "epoch": 0.08558066349053717,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.822036863231793e-05,
+      "loss": 1.8039,
+      "step": 1958
+    },
+    {
+      "epoch": 0.08562437169456708,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.821855195711897e-05,
+      "loss": 2.4106,
+      "step": 1959
+    },
+    {
+      "epoch": 0.08566807989859697,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.821673437196574e-05,
+      "loss": 1.9691,
+      "step": 1960
+    },
+    {
+      "epoch": 0.08571178810262686,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.821491587689257e-05,
+      "loss": 1.9763,
+      "step": 1961
+    },
+    {
+      "epoch": 0.08575549630665676,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.821309647193376e-05,
+      "loss": 2.3648,
+      "step": 1962
+    },
+    {
+      "epoch": 0.08579920451068665,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.821127615712364e-05,
+      "loss": 2.5316,
+      "step": 1963
+    },
+    {
+      "epoch": 0.08584291271471656,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.820945493249659e-05,
+      "loss": 1.8494,
+      "step": 1964
+    },
+    {
+      "epoch": 0.08588662091874645,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.820763279808695e-05,
+      "loss": 2.0713,
+      "step": 1965
+    },
+    {
+      "epoch": 0.08593032912277634,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.820580975392912e-05,
+      "loss": 1.7838,
+      "step": 1966
+    },
+    {
+      "epoch": 0.08597403732680624,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.820398580005749e-05,
+      "loss": 3.0606,
+      "step": 1967
+    },
+    {
+      "epoch": 0.08601774553083613,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.820216093650649e-05,
+      "loss": 1.7928,
+      "step": 1968
+    },
+    {
+      "epoch": 0.08606145373486604,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.820033516331057e-05,
+      "loss": 1.9501,
+      "step": 1969
+    },
+    {
+      "epoch": 0.08610516193889593,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.819850848050419e-05,
+      "loss": 2.061,
+      "step": 1970
+    },
+    {
+      "epoch": 0.08614887014292583,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.819668088812177e-05,
+      "loss": 1.9728,
+      "step": 1971
+    },
+    {
+      "epoch": 0.08619257834695572,
+      "grad_norm": 3.375,
+      "learning_rate": 9.819485238619784e-05,
+      "loss": 2.0728,
+      "step": 1972
+    },
+    {
+      "epoch": 0.08623628655098561,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.819302297476692e-05,
+      "loss": 1.947,
+      "step": 1973
+    },
+    {
+      "epoch": 0.08627999475501552,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.819119265386349e-05,
+      "loss": 2.1996,
+      "step": 1974
+    },
+    {
+      "epoch": 0.08632370295904541,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.818936142352212e-05,
+      "loss": 2.351,
+      "step": 1975
+    },
+    {
+      "epoch": 0.08636741116307531,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.81875292837774e-05,
+      "loss": 1.9119,
+      "step": 1976
+    },
+    {
+      "epoch": 0.0864111193671052,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.818569623466383e-05,
+      "loss": 2.0526,
+      "step": 1977
+    },
+    {
+      "epoch": 0.0864548275711351,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.818386227621605e-05,
+      "loss": 2.6655,
+      "step": 1978
+    },
+    {
+      "epoch": 0.086498535775165,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.818202740846867e-05,
+      "loss": 2.1808,
+      "step": 1979
+    },
+    {
+      "epoch": 0.0865422439791949,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.818019163145631e-05,
+      "loss": 2.8019,
+      "step": 1980
+    },
+    {
+      "epoch": 0.08658595218322479,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.81783549452136e-05,
+      "loss": 2.1154,
+      "step": 1981
+    },
+    {
+      "epoch": 0.08662966038725468,
+      "grad_norm": 2.5,
+      "learning_rate": 9.81765173497752e-05,
+      "loss": 2.0534,
+      "step": 1982
+    },
+    {
+      "epoch": 0.08667336859128458,
+      "grad_norm": 2.75,
+      "learning_rate": 9.817467884517582e-05,
+      "loss": 2.0711,
+      "step": 1983
+    },
+    {
+      "epoch": 0.08671707679531448,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.817283943145013e-05,
+      "loss": 2.0547,
+      "step": 1984
+    },
+    {
+      "epoch": 0.08676078499934438,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.817099910863285e-05,
+      "loss": 2.4254,
+      "step": 1985
+    },
+    {
+      "epoch": 0.08680449320337427,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.81691578767587e-05,
+      "loss": 2.117,
+      "step": 1986
+    },
+    {
+      "epoch": 0.08684820140740417,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.816731573586245e-05,
+      "loss": 2.1401,
+      "step": 1987
+    },
+    {
+      "epoch": 0.08689190961143406,
+      "grad_norm": 2.75,
+      "learning_rate": 9.816547268597882e-05,
+      "loss": 2.3913,
+      "step": 1988
+    },
+    {
+      "epoch": 0.08693561781546397,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.816362872714264e-05,
+      "loss": 2.3797,
+      "step": 1989
+    },
+    {
+      "epoch": 0.08697932601949386,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.816178385938868e-05,
+      "loss": 2.1,
+      "step": 1990
+    },
+    {
+      "epoch": 0.08702303422352375,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.815993808275177e-05,
+      "loss": 2.0036,
+      "step": 1991
+    },
+    {
+      "epoch": 0.08706674242755365,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.815809139726673e-05,
+      "loss": 1.5592,
+      "step": 1992
+    },
+    {
+      "epoch": 0.08711045063158354,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.815624380296841e-05,
+      "loss": 2.726,
+      "step": 1993
+    },
+    {
+      "epoch": 0.08715415883561345,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.815439529989168e-05,
+      "loss": 2.5015,
+      "step": 1994
+    },
+    {
+      "epoch": 0.08719786703964334,
+      "grad_norm": 3.125,
+      "learning_rate": 9.815254588807143e-05,
+      "loss": 1.9638,
+      "step": 1995
+    },
+    {
+      "epoch": 0.08724157524367324,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.815069556754256e-05,
+      "loss": 2.2552,
+      "step": 1996
+    },
+    {
+      "epoch": 0.08728528344770313,
+      "grad_norm": 2.625,
+      "learning_rate": 9.814884433833999e-05,
+      "loss": 2.0396,
+      "step": 1997
+    },
+    {
+      "epoch": 0.08732899165173302,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.814699220049863e-05,
+      "loss": 1.9973,
+      "step": 1998
+    },
+    {
+      "epoch": 0.08737269985576293,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.814513915405346e-05,
+      "loss": 2.2719,
+      "step": 1999
+    },
+    {
+      "epoch": 0.08741640805979282,
+      "grad_norm": 3.625,
+      "learning_rate": 9.814328519903946e-05,
+      "loss": 2.6787,
+      "step": 2000
+    },
+    {
+      "epoch": 0.08746011626382272,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.814143033549157e-05,
+      "loss": 2.5491,
+      "step": 2001
+    },
+    {
+      "epoch": 0.08750382446785261,
+      "grad_norm": 3.125,
+      "learning_rate": 9.813957456344485e-05,
+      "loss": 3.0051,
+      "step": 2002
+    },
+    {
+      "epoch": 0.08754753267188252,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.813771788293429e-05,
+      "loss": 1.8316,
+      "step": 2003
+    },
+    {
+      "epoch": 0.08759124087591241,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.813586029399492e-05,
+      "loss": 1.6633,
+      "step": 2004
+    },
+    {
+      "epoch": 0.0876349490799423,
+      "grad_norm": 4.375,
+      "learning_rate": 9.813400179666181e-05,
+      "loss": 2.5481,
+      "step": 2005
+    },
+    {
+      "epoch": 0.0876786572839722,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.813214239097004e-05,
+      "loss": 2.0448,
+      "step": 2006
+    },
+    {
+      "epoch": 0.0877223654880021,
+      "grad_norm": 3.25,
+      "learning_rate": 9.813028207695467e-05,
+      "loss": 1.9342,
+      "step": 2007
+    },
+    {
+      "epoch": 0.087766073692032,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.812842085465085e-05,
+      "loss": 2.5093,
+      "step": 2008
+    },
+    {
+      "epoch": 0.0878097818960619,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.812655872409366e-05,
+      "loss": 2.2058,
+      "step": 2009
+    },
+    {
+      "epoch": 0.08785349010009179,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.812469568531828e-05,
+      "loss": 2.2205,
+      "step": 2010
+    },
+    {
+      "epoch": 0.08789719830412168,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.812283173835985e-05,
+      "loss": 2.2833,
+      "step": 2011
+    },
+    {
+      "epoch": 0.08794090650815158,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.812096688325354e-05,
+      "loss": 2.1205,
+      "step": 2012
+    },
+    {
+      "epoch": 0.08798461471218148,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.811910112003455e-05,
+      "loss": 1.9461,
+      "step": 2013
+    },
+    {
+      "epoch": 0.08802832291621138,
+      "grad_norm": 42.5,
+      "learning_rate": 9.811723444873809e-05,
+      "loss": 2.3175,
+      "step": 2014
+    },
+    {
+      "epoch": 0.08807203112024127,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.811536686939937e-05,
+      "loss": 1.8189,
+      "step": 2015
+    },
+    {
+      "epoch": 0.08811573932427116,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.811349838205366e-05,
+      "loss": 2.5017,
+      "step": 2016
+    },
+    {
+      "epoch": 0.08815944752830106,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.81116289867362e-05,
+      "loss": 1.8558,
+      "step": 2017
+    },
+    {
+      "epoch": 0.08820315573233097,
+      "grad_norm": 2.375,
+      "learning_rate": 9.81097586834823e-05,
+      "loss": 2.0018,
+      "step": 2018
+    },
+    {
+      "epoch": 0.08824686393636086,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.810788747232721e-05,
+      "loss": 2.0362,
+      "step": 2019
+    },
+    {
+      "epoch": 0.08829057214039075,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.810601535330627e-05,
+      "loss": 2.3536,
+      "step": 2020
+    },
+    {
+      "epoch": 0.08833428034442065,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.810414232645482e-05,
+      "loss": 2.0757,
+      "step": 2021
+    },
+    {
+      "epoch": 0.08837798854845054,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.810226839180817e-05,
+      "loss": 2.1589,
+      "step": 2022
+    },
+    {
+      "epoch": 0.08842169675248045,
+      "grad_norm": 3.5,
+      "learning_rate": 9.810039354940173e-05,
+      "loss": 2.8814,
+      "step": 2023
+    },
+    {
+      "epoch": 0.08846540495651034,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.809851779927084e-05,
+      "loss": 2.5291,
+      "step": 2024
+    },
+    {
+      "epoch": 0.08850911316054023,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.809664114145091e-05,
+      "loss": 2.4861,
+      "step": 2025
+    },
+    {
+      "epoch": 0.08855282136457013,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.809476357597738e-05,
+      "loss": 2.0149,
+      "step": 2026
+    },
+    {
+      "epoch": 0.08859652956860002,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.809288510288566e-05,
+      "loss": 3.2423,
+      "step": 2027
+    },
+    {
+      "epoch": 0.08864023777262993,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.809100572221118e-05,
+      "loss": 1.4191,
+      "step": 2028
+    },
+    {
+      "epoch": 0.08868394597665982,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.808912543398945e-05,
+      "loss": 2.0518,
+      "step": 2029
+    },
+    {
+      "epoch": 0.08872765418068972,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.808724423825592e-05,
+      "loss": 1.9862,
+      "step": 2030
+    },
+    {
+      "epoch": 0.08877136238471961,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.80853621350461e-05,
+      "loss": 1.978,
+      "step": 2031
+    },
+    {
+      "epoch": 0.0888150705887495,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.808347912439554e-05,
+      "loss": 2.0013,
+      "step": 2032
+    },
+    {
+      "epoch": 0.08885877879277941,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.808159520633973e-05,
+      "loss": 2.1639,
+      "step": 2033
+    },
+    {
+      "epoch": 0.0889024869968093,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.807971038091423e-05,
+      "loss": 2.9512,
+      "step": 2034
+    },
+    {
+      "epoch": 0.0889461952008392,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.807782464815463e-05,
+      "loss": 2.0199,
+      "step": 2035
+    },
+    {
+      "epoch": 0.08898990340486909,
+      "grad_norm": 2.5,
+      "learning_rate": 9.80759380080965e-05,
+      "loss": 2.0716,
+      "step": 2036
+    },
+    {
+      "epoch": 0.08903361160889899,
+      "grad_norm": 2.5,
+      "learning_rate": 9.807405046077545e-05,
+      "loss": 2.1774,
+      "step": 2037
+    },
+    {
+      "epoch": 0.08907731981292889,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.80721620062271e-05,
+      "loss": 2.5221,
+      "step": 2038
+    },
+    {
+      "epoch": 0.08912102801695879,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.807027264448708e-05,
+      "loss": 2.0483,
+      "step": 2039
+    },
+    {
+      "epoch": 0.08916473622098868,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.806838237559107e-05,
+      "loss": 1.8315,
+      "step": 2040
+    },
+    {
+      "epoch": 0.08920844442501857,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.80664911995747e-05,
+      "loss": 2.7907,
+      "step": 2041
+    },
+    {
+      "epoch": 0.08925215262904847,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.80645991164737e-05,
+      "loss": 2.3077,
+      "step": 2042
+    },
+    {
+      "epoch": 0.08929586083307837,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.806270612632375e-05,
+      "loss": 2.032,
+      "step": 2043
+    },
+    {
+      "epoch": 0.08933956903710827,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.806081222916059e-05,
+      "loss": 2.1644,
+      "step": 2044
+    },
+    {
+      "epoch": 0.08938327724113816,
+      "grad_norm": 5.5625,
+      "learning_rate": 9.805891742501996e-05,
+      "loss": 2.7271,
+      "step": 2045
+    },
+    {
+      "epoch": 0.08942698544516806,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.80570217139376e-05,
+      "loss": 1.9969,
+      "step": 2046
+    },
+    {
+      "epoch": 0.08947069364919795,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.80551250959493e-05,
+      "loss": 2.655,
+      "step": 2047
+    },
+    {
+      "epoch": 0.08951440185322786,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.805322757109086e-05,
+      "loss": 2.6689,
+      "step": 2048
+    },
+    {
+      "epoch": 0.08955811005725775,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.805132913939807e-05,
+      "loss": 2.1874,
+      "step": 2049
+    },
+    {
+      "epoch": 0.08960181826128764,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.804942980090676e-05,
+      "loss": 2.3881,
+      "step": 2050
+    },
+    {
+      "epoch": 0.08964552646531754,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.804752955565278e-05,
+      "loss": 2.0989,
+      "step": 2051
+    },
+    {
+      "epoch": 0.08968923466934743,
+      "grad_norm": 3.125,
+      "learning_rate": 9.804562840367198e-05,
+      "loss": 2.2578,
+      "step": 2052
+    },
+    {
+      "epoch": 0.08973294287337734,
+      "grad_norm": 2.625,
+      "learning_rate": 9.804372634500026e-05,
+      "loss": 1.8339,
+      "step": 2053
+    },
+    {
+      "epoch": 0.08977665107740723,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.804182337967349e-05,
+      "loss": 2.4403,
+      "step": 2054
+    },
+    {
+      "epoch": 0.08982035928143713,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.80399195077276e-05,
+      "loss": 2.6908,
+      "step": 2055
+    },
+    {
+      "epoch": 0.08986406748546702,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.803801472919849e-05,
+      "loss": 2.0668,
+      "step": 2056
+    },
+    {
+      "epoch": 0.08990777568949691,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.803610904412214e-05,
+      "loss": 2.2085,
+      "step": 2057
+    },
+    {
+      "epoch": 0.08995148389352682,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.803420245253449e-05,
+      "loss": 2.0336,
+      "step": 2058
+    },
+    {
+      "epoch": 0.08999519209755671,
+      "grad_norm": 3.125,
+      "learning_rate": 9.803229495447154e-05,
+      "loss": 2.9318,
+      "step": 2059
+    },
+    {
+      "epoch": 0.09003890030158661,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.803038654996927e-05,
+      "loss": 1.6366,
+      "step": 2060
+    },
+    {
+      "epoch": 0.0900826085056165,
+      "grad_norm": 2.375,
+      "learning_rate": 9.802847723906371e-05,
+      "loss": 1.9645,
+      "step": 2061
+    },
+    {
+      "epoch": 0.0901263167096464,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.802656702179088e-05,
+      "loss": 2.3127,
+      "step": 2062
+    },
+    {
+      "epoch": 0.0901700249136763,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.802465589818683e-05,
+      "loss": 1.8978,
+      "step": 2063
+    },
+    {
+      "epoch": 0.0902137331177062,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.802274386828762e-05,
+      "loss": 1.9738,
+      "step": 2064
+    },
+    {
+      "epoch": 0.09025744132173609,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.802083093212935e-05,
+      "loss": 2.8582,
+      "step": 2065
+    },
+    {
+      "epoch": 0.09030114952576598,
+      "grad_norm": 3.125,
+      "learning_rate": 9.801891708974809e-05,
+      "loss": 2.021,
+      "step": 2066
+    },
+    {
+      "epoch": 0.09034485772979588,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.801700234117999e-05,
+      "loss": 2.0541,
+      "step": 2067
+    },
+    {
+      "epoch": 0.09038856593382578,
+      "grad_norm": 4.03125,
+      "learning_rate": 9.801508668646118e-05,
+      "loss": 1.8612,
+      "step": 2068
+    },
+    {
+      "epoch": 0.09043227413785568,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.801317012562779e-05,
+      "loss": 2.3819,
+      "step": 2069
+    },
+    {
+      "epoch": 0.09047598234188557,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.8011252658716e-05,
+      "loss": 2.1188,
+      "step": 2070
+    },
+    {
+      "epoch": 0.09051969054591547,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.8009334285762e-05,
+      "loss": 2.0454,
+      "step": 2071
+    },
+    {
+      "epoch": 0.09056339874994536,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.800741500680197e-05,
+      "loss": 2.1398,
+      "step": 2072
+    },
+    {
+      "epoch": 0.09060710695397527,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.800549482187216e-05,
+      "loss": 1.7601,
+      "step": 2073
+    },
+    {
+      "epoch": 0.09065081515800516,
+      "grad_norm": 6.40625,
+      "learning_rate": 9.800357373100882e-05,
+      "loss": 2.1453,
+      "step": 2074
+    },
+    {
+      "epoch": 0.09069452336203505,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.800165173424814e-05,
+      "loss": 2.3561,
+      "step": 2075
+    },
+    {
+      "epoch": 0.09073823156606495,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.799972883162645e-05,
+      "loss": 1.8398,
+      "step": 2076
+    },
+    {
+      "epoch": 0.09078193977009484,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.799780502318e-05,
+      "loss": 2.2264,
+      "step": 2077
+    },
+    {
+      "epoch": 0.09082564797412475,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.799588030894512e-05,
+      "loss": 2.0077,
+      "step": 2078
+    },
+    {
+      "epoch": 0.09086935617815464,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.799395468895812e-05,
+      "loss": 2.0867,
+      "step": 2079
+    },
+    {
+      "epoch": 0.09091306438218454,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.799202816325534e-05,
+      "loss": 2.1293,
+      "step": 2080
+    },
+    {
+      "epoch": 0.09095677258621443,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.799010073187316e-05,
+      "loss": 2.2106,
+      "step": 2081
+    },
+    {
+      "epoch": 0.09100048079024432,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.798817239484792e-05,
+      "loss": 1.7821,
+      "step": 2082
+    },
+    {
+      "epoch": 0.09104418899427423,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.798624315221603e-05,
+      "loss": 2.1652,
+      "step": 2083
+    },
+    {
+      "epoch": 0.09108789719830412,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.798431300401388e-05,
+      "loss": 1.9395,
+      "step": 2084
+    },
+    {
+      "epoch": 0.09113160540233402,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.79823819502779e-05,
+      "loss": 2.8928,
+      "step": 2085
+    },
+    {
+      "epoch": 0.09117531360636391,
+      "grad_norm": 2.375,
+      "learning_rate": 9.798044999104456e-05,
+      "loss": 2.069,
+      "step": 2086
+    },
+    {
+      "epoch": 0.0912190218103938,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.797851712635028e-05,
+      "loss": 2.0174,
+      "step": 2087
+    },
+    {
+      "epoch": 0.09126273001442371,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.797658335623155e-05,
+      "loss": 2.1882,
+      "step": 2088
+    },
+    {
+      "epoch": 0.0913064382184536,
+      "grad_norm": 2.75,
+      "learning_rate": 9.797464868072488e-05,
+      "loss": 2.2737,
+      "step": 2089
+    },
+    {
+      "epoch": 0.0913501464224835,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.797271309986676e-05,
+      "loss": 2.0482,
+      "step": 2090
+    },
+    {
+      "epoch": 0.0913938546265134,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.797077661369372e-05,
+      "loss": 2.1115,
+      "step": 2091
+    },
+    {
+      "epoch": 0.09143756283054329,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.79688392222423e-05,
+      "loss": 2.2927,
+      "step": 2092
+    },
+    {
+      "epoch": 0.0914812710345732,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.796690092554908e-05,
+      "loss": 1.8428,
+      "step": 2093
+    },
+    {
+      "epoch": 0.09152497923860309,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.796496172365061e-05,
+      "loss": 1.911,
+      "step": 2094
+    },
+    {
+      "epoch": 0.09156868744263298,
+      "grad_norm": 3.75,
+      "learning_rate": 9.796302161658353e-05,
+      "loss": 2.5428,
+      "step": 2095
+    },
+    {
+      "epoch": 0.09161239564666288,
+      "grad_norm": 2.875,
+      "learning_rate": 9.79610806043844e-05,
+      "loss": 1.9935,
+      "step": 2096
+    },
+    {
+      "epoch": 0.09165610385069277,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.795913868708989e-05,
+      "loss": 2.322,
+      "step": 2097
+    },
+    {
+      "epoch": 0.09169981205472268,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.795719586473663e-05,
+      "loss": 2.0394,
+      "step": 2098
+    },
+    {
+      "epoch": 0.09174352025875257,
+      "grad_norm": 5.03125,
+      "learning_rate": 9.795525213736128e-05,
+      "loss": 2.4226,
+      "step": 2099
+    },
+    {
+      "epoch": 0.09178722846278246,
+      "grad_norm": 12.25,
+      "learning_rate": 9.795330750500054e-05,
+      "loss": 1.4187,
+      "step": 2100
+    },
+    {
+      "epoch": 0.09183093666681236,
+      "grad_norm": 2.5,
+      "learning_rate": 9.795136196769107e-05,
+      "loss": 1.8698,
+      "step": 2101
+    },
+    {
+      "epoch": 0.09187464487084225,
+      "grad_norm": 3.0,
+      "learning_rate": 9.794941552546963e-05,
+      "loss": 2.5359,
+      "step": 2102
+    },
+    {
+      "epoch": 0.09191835307487216,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.794746817837293e-05,
+      "loss": 2.3038,
+      "step": 2103
+    },
+    {
+      "epoch": 0.09196206127890205,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.794551992643772e-05,
+      "loss": 2.0825,
+      "step": 2104
+    },
+    {
+      "epoch": 0.09200576948293195,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.794357076970076e-05,
+      "loss": 2.0303,
+      "step": 2105
+    },
+    {
+      "epoch": 0.09204947768696184,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.794162070819885e-05,
+      "loss": 1.7951,
+      "step": 2106
+    },
+    {
+      "epoch": 0.09209318589099173,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.793966974196878e-05,
+      "loss": 2.0509,
+      "step": 2107
+    },
+    {
+      "epoch": 0.09213689409502164,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.793771787104735e-05,
+      "loss": 2.2638,
+      "step": 2108
+    },
+    {
+      "epoch": 0.09218060229905153,
+      "grad_norm": 2.875,
+      "learning_rate": 9.793576509547144e-05,
+      "loss": 2.2946,
+      "step": 2109
+    },
+    {
+      "epoch": 0.09222431050308143,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.793381141527786e-05,
+      "loss": 2.5537,
+      "step": 2110
+    },
+    {
+      "epoch": 0.09226801870711132,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.79318568305035e-05,
+      "loss": 2.0841,
+      "step": 2111
+    },
+    {
+      "epoch": 0.09231172691114121,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.792990134118525e-05,
+      "loss": 2.1249,
+      "step": 2112
+    },
+    {
+      "epoch": 0.09235543511517112,
+      "grad_norm": 4.125,
+      "learning_rate": 9.792794494736e-05,
+      "loss": 2.4162,
+      "step": 2113
+    },
+    {
+      "epoch": 0.09239914331920102,
+      "grad_norm": 2.625,
+      "learning_rate": 9.792598764906466e-05,
+      "loss": 2.1053,
+      "step": 2114
+    },
+    {
+      "epoch": 0.09244285152323091,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.792402944633619e-05,
+      "loss": 2.5305,
+      "step": 2115
+    },
+    {
+      "epoch": 0.0924865597272608,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.792207033921152e-05,
+      "loss": 2.1751,
+      "step": 2116
+    },
+    {
+      "epoch": 0.0925302679312907,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.792011032772765e-05,
+      "loss": 1.9627,
+      "step": 2117
+    },
+    {
+      "epoch": 0.0925739761353206,
+      "grad_norm": 2.75,
+      "learning_rate": 9.791814941192155e-05,
+      "loss": 2.2857,
+      "step": 2118
+    },
+    {
+      "epoch": 0.0926176843393505,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.791618759183023e-05,
+      "loss": 2.2022,
+      "step": 2119
+    },
+    {
+      "epoch": 0.09266139254338039,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.791422486749071e-05,
+      "loss": 2.8579,
+      "step": 2120
+    },
+    {
+      "epoch": 0.09270510074741029,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.791226123894003e-05,
+      "loss": 2.7253,
+      "step": 2121
+    },
+    {
+      "epoch": 0.09274880895144018,
+      "grad_norm": 3.0,
+      "learning_rate": 9.791029670621525e-05,
+      "loss": 1.9731,
+      "step": 2122
+    },
+    {
+      "epoch": 0.09279251715547009,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.790833126935343e-05,
+      "loss": 1.9945,
+      "step": 2123
+    },
+    {
+      "epoch": 0.09283622535949998,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.79063649283917e-05,
+      "loss": 1.9241,
+      "step": 2124
+    },
+    {
+      "epoch": 0.09287993356352987,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.79043976833671e-05,
+      "loss": 2.9551,
+      "step": 2125
+    },
+    {
+      "epoch": 0.09292364176755977,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.790242953431682e-05,
+      "loss": 1.9754,
+      "step": 2126
+    },
+    {
+      "epoch": 0.09296734997158966,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.790046048127797e-05,
+      "loss": 1.9089,
+      "step": 2127
+    },
+    {
+      "epoch": 0.09301105817561957,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.789849052428772e-05,
+      "loss": 2.0825,
+      "step": 2128
+    },
+    {
+      "epoch": 0.09305476637964946,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.789651966338323e-05,
+      "loss": 2.5024,
+      "step": 2129
+    },
+    {
+      "epoch": 0.09309847458367936,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.789454789860171e-05,
+      "loss": 1.8926,
+      "step": 2130
+    },
+    {
+      "epoch": 0.09314218278770925,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.789257522998037e-05,
+      "loss": 2.1313,
+      "step": 2131
+    },
+    {
+      "epoch": 0.09318589099173914,
+      "grad_norm": 2.625,
+      "learning_rate": 9.789060165755643e-05,
+      "loss": 2.2453,
+      "step": 2132
+    },
+    {
+      "epoch": 0.09322959919576905,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.788862718136713e-05,
+      "loss": 2.0841,
+      "step": 2133
+    },
+    {
+      "epoch": 0.09327330739979894,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.788665180144973e-05,
+      "loss": 2.1431,
+      "step": 2134
+    },
+    {
+      "epoch": 0.09331701560382884,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.788467551784153e-05,
+      "loss": 2.3447,
+      "step": 2135
+    },
+    {
+      "epoch": 0.09336072380785873,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.788269833057979e-05,
+      "loss": 2.0081,
+      "step": 2136
+    },
+    {
+      "epoch": 0.09340443201188864,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.788072023970187e-05,
+      "loss": 2.434,
+      "step": 2137
+    },
+    {
+      "epoch": 0.09344814021591853,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.787874124524505e-05,
+      "loss": 1.61,
+      "step": 2138
+    },
+    {
+      "epoch": 0.09349184841994843,
+      "grad_norm": 2.25,
+      "learning_rate": 9.78767613472467e-05,
+      "loss": 1.7895,
+      "step": 2139
+    },
+    {
+      "epoch": 0.09353555662397832,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.787478054574419e-05,
+      "loss": 1.8193,
+      "step": 2140
+    },
+    {
+      "epoch": 0.09357926482800821,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.787279884077489e-05,
+      "loss": 2.2462,
+      "step": 2141
+    },
+    {
+      "epoch": 0.09362297303203812,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.78708162323762e-05,
+      "loss": 1.8915,
+      "step": 2142
+    },
+    {
+      "epoch": 0.09366668123606801,
+      "grad_norm": 2.375,
+      "learning_rate": 9.786883272058554e-05,
+      "loss": 2.1082,
+      "step": 2143
+    },
+    {
+      "epoch": 0.09371038944009791,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.786684830544033e-05,
+      "loss": 1.9747,
+      "step": 2144
+    },
+    {
+      "epoch": 0.0937540976441278,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.786486298697803e-05,
+      "loss": 1.9774,
+      "step": 2145
+    },
+    {
+      "epoch": 0.0937978058481577,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.786287676523609e-05,
+      "loss": 2.7382,
+      "step": 2146
+    },
+    {
+      "epoch": 0.0938415140521876,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.786088964025201e-05,
+      "loss": 1.9554,
+      "step": 2147
+    },
+    {
+      "epoch": 0.0938852222562175,
+      "grad_norm": 2.875,
+      "learning_rate": 9.78589016120633e-05,
+      "loss": 1.9873,
+      "step": 2148
+    },
+    {
+      "epoch": 0.09392893046024739,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.785691268070745e-05,
+      "loss": 2.1718,
+      "step": 2149
+    },
+    {
+      "epoch": 0.09397263866427728,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.7854922846222e-05,
+      "loss": 2.5165,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09401634686830718,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.785293210864451e-05,
+      "loss": 1.8265,
+      "step": 2151
+    },
+    {
+      "epoch": 0.09406005507233708,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.785094046801256e-05,
+      "loss": 2.2153,
+      "step": 2152
+    },
+    {
+      "epoch": 0.09410376327636698,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.78489479243637e-05,
+      "loss": 1.8579,
+      "step": 2153
+    },
+    {
+      "epoch": 0.09414747148039687,
+      "grad_norm": 2.875,
+      "learning_rate": 9.784695447773555e-05,
+      "loss": 3.1056,
+      "step": 2154
+    },
+    {
+      "epoch": 0.09419117968442677,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.784496012816573e-05,
+      "loss": 1.8621,
+      "step": 2155
+    },
+    {
+      "epoch": 0.09423488788845666,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.78429648756919e-05,
+      "loss": 1.3941,
+      "step": 2156
+    },
+    {
+      "epoch": 0.09427859609248657,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.784096872035167e-05,
+      "loss": 2.677,
+      "step": 2157
+    },
+    {
+      "epoch": 0.09432230429651646,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.783897166218273e-05,
+      "loss": 2.3034,
+      "step": 2158
+    },
+    {
+      "epoch": 0.09436601250054635,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.783697370122278e-05,
+      "loss": 2.7365,
+      "step": 2159
+    },
+    {
+      "epoch": 0.09440972070457625,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.78349748375095e-05,
+      "loss": 2.675,
+      "step": 2160
+    },
+    {
+      "epoch": 0.09445342890860614,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.783297507108065e-05,
+      "loss": 2.2508,
+      "step": 2161
+    },
+    {
+      "epoch": 0.09449713711263605,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.78309744019739e-05,
+      "loss": 2.5425,
+      "step": 2162
+    },
+    {
+      "epoch": 0.09454084531666594,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.78289728302271e-05,
+      "loss": 1.9132,
+      "step": 2163
+    },
+    {
+      "epoch": 0.09458455352069584,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.782697035587793e-05,
+      "loss": 1.9358,
+      "step": 2164
+    },
+    {
+      "epoch": 0.09462826172472573,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.782496697896424e-05,
+      "loss": 2.3742,
+      "step": 2165
+    },
+    {
+      "epoch": 0.09467196992875562,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.78229626995238e-05,
+      "loss": 2.5638,
+      "step": 2166
+    },
+    {
+      "epoch": 0.09471567813278553,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.782095751759448e-05,
+      "loss": 2.0113,
+      "step": 2167
+    },
+    {
+      "epoch": 0.09475938633681542,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.781895143321406e-05,
+      "loss": 2.1018,
+      "step": 2168
+    },
+    {
+      "epoch": 0.09480309454084532,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.781694444642044e-05,
+      "loss": 1.7566,
+      "step": 2169
+    },
+    {
+      "epoch": 0.09484680274487521,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.781493655725149e-05,
+      "loss": 2.4728,
+      "step": 2170
+    },
+    {
+      "epoch": 0.0948905109489051,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.781292776574509e-05,
+      "loss": 1.4731,
+      "step": 2171
+    },
+    {
+      "epoch": 0.09493421915293501,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.781091807193916e-05,
+      "loss": 2.251,
+      "step": 2172
+    },
+    {
+      "epoch": 0.0949779273569649,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.780890747587164e-05,
+      "loss": 1.8029,
+      "step": 2173
+    },
+    {
+      "epoch": 0.0950216355609948,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.780689597758041e-05,
+      "loss": 1.749,
+      "step": 2174
+    },
+    {
+      "epoch": 0.09506534376502469,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.78048835771035e-05,
+      "loss": 2.6273,
+      "step": 2175
+    },
+    {
+      "epoch": 0.09510905196905459,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.780287027447886e-05,
+      "loss": 2.1231,
+      "step": 2176
+    },
+    {
+      "epoch": 0.0951527601730845,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.780085606974448e-05,
+      "loss": 2.3221,
+      "step": 2177
+    },
+    {
+      "epoch": 0.09519646837711439,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.779884096293838e-05,
+      "loss": 2.2185,
+      "step": 2178
+    },
+    {
+      "epoch": 0.09524017658114428,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.779682495409857e-05,
+      "loss": 2.3901,
+      "step": 2179
+    },
+    {
+      "epoch": 0.09528388478517417,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.779480804326313e-05,
+      "loss": 2.7555,
+      "step": 2180
+    },
+    {
+      "epoch": 0.09532759298920407,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.779279023047008e-05,
+      "loss": 2.0593,
+      "step": 2181
+    },
+    {
+      "epoch": 0.09537130119323398,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.779077151575753e-05,
+      "loss": 2.1978,
+      "step": 2182
+    },
+    {
+      "epoch": 0.09541500939726387,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.778875189916356e-05,
+      "loss": 2.2172,
+      "step": 2183
+    },
+    {
+      "epoch": 0.09545871760129376,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.778673138072629e-05,
+      "loss": 1.9492,
+      "step": 2184
+    },
+    {
+      "epoch": 0.09550242580532366,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.778470996048383e-05,
+      "loss": 2.308,
+      "step": 2185
+    },
+    {
+      "epoch": 0.09554613400935355,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.778268763847438e-05,
+      "loss": 2.4126,
+      "step": 2186
+    },
+    {
+      "epoch": 0.09558984221338346,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.778066441473604e-05,
+      "loss": 1.7932,
+      "step": 2187
+    },
+    {
+      "epoch": 0.09563355041741335,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.777864028930705e-05,
+      "loss": 2.7622,
+      "step": 2188
+    },
+    {
+      "epoch": 0.09567725862144325,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.777661526222556e-05,
+      "loss": 2.0042,
+      "step": 2189
+    },
+    {
+      "epoch": 0.09572096682547314,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.777458933352981e-05,
+      "loss": 2.0112,
+      "step": 2190
+    },
+    {
+      "epoch": 0.09576467502950303,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.777256250325801e-05,
+      "loss": 2.1095,
+      "step": 2191
+    },
+    {
+      "epoch": 0.09580838323353294,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.777053477144844e-05,
+      "loss": 2.1284,
+      "step": 2192
+    },
+    {
+      "epoch": 0.09585209143756283,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.776850613813936e-05,
+      "loss": 2.2209,
+      "step": 2193
+    },
+    {
+      "epoch": 0.09589579964159273,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.776647660336903e-05,
+      "loss": 2.518,
+      "step": 2194
+    },
+    {
+      "epoch": 0.09593950784562262,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.776444616717578e-05,
+      "loss": 2.4591,
+      "step": 2195
+    },
+    {
+      "epoch": 0.09598321604965251,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.776241482959791e-05,
+      "loss": 2.046,
+      "step": 2196
+    },
+    {
+      "epoch": 0.09602692425368242,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.776038259067375e-05,
+      "loss": 2.0352,
+      "step": 2197
+    },
+    {
+      "epoch": 0.09607063245771232,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.775834945044167e-05,
+      "loss": 2.0531,
+      "step": 2198
+    },
+    {
+      "epoch": 0.09611434066174221,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.775631540894002e-05,
+      "loss": 2.5601,
+      "step": 2199
+    },
+    {
+      "epoch": 0.0961580488657721,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.775428046620718e-05,
+      "loss": 2.8106,
+      "step": 2200
+    },
+    {
+      "epoch": 0.096201757069802,
+      "grad_norm": 2.75,
+      "learning_rate": 9.775224462228159e-05,
+      "loss": 2.3545,
+      "step": 2201
+    },
+    {
+      "epoch": 0.0962454652738319,
+      "grad_norm": 6.71875,
+      "learning_rate": 9.775020787720162e-05,
+      "loss": 1.8256,
+      "step": 2202
+    },
+    {
+      "epoch": 0.0962891734778618,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.774817023100575e-05,
+      "loss": 2.4899,
+      "step": 2203
+    },
+    {
+      "epoch": 0.09633288168189169,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.77461316837324e-05,
+      "loss": 2.3327,
+      "step": 2204
+    },
+    {
+      "epoch": 0.09637658988592158,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.774409223542006e-05,
+      "loss": 2.6428,
+      "step": 2205
+    },
+    {
+      "epoch": 0.09642029808995148,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.774205188610721e-05,
+      "loss": 1.8124,
+      "step": 2206
+    },
+    {
+      "epoch": 0.09646400629398139,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.774001063583235e-05,
+      "loss": 2.2997,
+      "step": 2207
+    },
+    {
+      "epoch": 0.09650771449801128,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.773796848463402e-05,
+      "loss": 2.1743,
+      "step": 2208
+    },
+    {
+      "epoch": 0.09655142270204117,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.773592543255074e-05,
+      "loss": 2.0742,
+      "step": 2209
+    },
+    {
+      "epoch": 0.09659513090607107,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.773388147962106e-05,
+      "loss": 1.9118,
+      "step": 2210
+    },
+    {
+      "epoch": 0.09663883911010096,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.773183662588356e-05,
+      "loss": 1.9707,
+      "step": 2211
+    },
+    {
+      "epoch": 0.09668254731413087,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.772979087137686e-05,
+      "loss": 2.2365,
+      "step": 2212
+    },
+    {
+      "epoch": 0.09672625551816076,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.772774421613951e-05,
+      "loss": 1.9471,
+      "step": 2213
+    },
+    {
+      "epoch": 0.09676996372219066,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.772569666021018e-05,
+      "loss": 2.2223,
+      "step": 2214
+    },
+    {
+      "epoch": 0.09681367192622055,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.772364820362749e-05,
+      "loss": 2.2746,
+      "step": 2215
+    },
+    {
+      "epoch": 0.09685738013025044,
+      "grad_norm": 2.375,
+      "learning_rate": 9.77215988464301e-05,
+      "loss": 1.7533,
+      "step": 2216
+    },
+    {
+      "epoch": 0.09690108833428035,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.771954858865668e-05,
+      "loss": 2.0964,
+      "step": 2217
+    },
+    {
+      "epoch": 0.09694479653831024,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.771749743034592e-05,
+      "loss": 2.1936,
+      "step": 2218
+    },
+    {
+      "epoch": 0.09698850474234014,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.771544537153653e-05,
+      "loss": 2.7054,
+      "step": 2219
+    },
+    {
+      "epoch": 0.09703221294637003,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.771339241226726e-05,
+      "loss": 2.2081,
+      "step": 2220
+    },
+    {
+      "epoch": 0.09707592115039992,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.771133855257684e-05,
+      "loss": 2.1181,
+      "step": 2221
+    },
+    {
+      "epoch": 0.09711962935442983,
+      "grad_norm": 13.4375,
+      "learning_rate": 9.770928379250399e-05,
+      "loss": 2.4265,
+      "step": 2222
+    },
+    {
+      "epoch": 0.09716333755845973,
+      "grad_norm": 10.25,
+      "learning_rate": 9.770722813208754e-05,
+      "loss": 6.5293,
+      "step": 2223
+    },
+    {
+      "epoch": 0.09720704576248962,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.770517157136625e-05,
+      "loss": 2.4392,
+      "step": 2224
+    },
+    {
+      "epoch": 0.09725075396651951,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.770311411037894e-05,
+      "loss": 2.3015,
+      "step": 2225
+    },
+    {
+      "epoch": 0.0972944621705494,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.770105574916443e-05,
+      "loss": 2.1963,
+      "step": 2226
+    },
+    {
+      "epoch": 0.09733817037457931,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.76989964877616e-05,
+      "loss": 1.7115,
+      "step": 2227
+    },
+    {
+      "epoch": 0.09738187857860921,
+      "grad_norm": 4.0,
+      "learning_rate": 9.769693632620926e-05,
+      "loss": 1.7496,
+      "step": 2228
+    },
+    {
+      "epoch": 0.0974255867826391,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.769487526454631e-05,
+      "loss": 2.2873,
+      "step": 2229
+    },
+    {
+      "epoch": 0.097469294986669,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.769281330281165e-05,
+      "loss": 2.1804,
+      "step": 2230
+    },
+    {
+      "epoch": 0.09751300319069889,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.76907504410442e-05,
+      "loss": 2.0584,
+      "step": 2231
+    },
+    {
+      "epoch": 0.0975567113947288,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.768868667928288e-05,
+      "loss": 1.7148,
+      "step": 2232
+    },
+    {
+      "epoch": 0.09760041959875869,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.768662201756662e-05,
+      "loss": 1.9899,
+      "step": 2233
+    },
+    {
+      "epoch": 0.09764412780278858,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.768455645593441e-05,
+      "loss": 2.4444,
+      "step": 2234
+    },
+    {
+      "epoch": 0.09768783600681848,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.768248999442522e-05,
+      "loss": 1.8971,
+      "step": 2235
+    },
+    {
+      "epoch": 0.09773154421084837,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.768042263307804e-05,
+      "loss": 2.4553,
+      "step": 2236
+    },
+    {
+      "epoch": 0.09777525241487828,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.767835437193187e-05,
+      "loss": 1.7158,
+      "step": 2237
+    },
+    {
+      "epoch": 0.09781896061890817,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.767628521102578e-05,
+      "loss": 2.3434,
+      "step": 2238
+    },
+    {
+      "epoch": 0.09786266882293806,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.767421515039881e-05,
+      "loss": 1.9139,
+      "step": 2239
+    },
+    {
+      "epoch": 0.09790637702696796,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.767214419009e-05,
+      "loss": 2.0438,
+      "step": 2240
+    },
+    {
+      "epoch": 0.09795008523099785,
+      "grad_norm": 2.375,
+      "learning_rate": 9.767007233013845e-05,
+      "loss": 1.9096,
+      "step": 2241
+    },
+    {
+      "epoch": 0.09799379343502776,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.766799957058324e-05,
+      "loss": 2.0401,
+      "step": 2242
+    },
+    {
+      "epoch": 0.09803750163905765,
+      "grad_norm": 3.796875,
+      "learning_rate": 9.766592591146352e-05,
+      "loss": 3.3079,
+      "step": 2243
+    },
+    {
+      "epoch": 0.09808120984308755,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.766385135281839e-05,
+      "loss": 2.1826,
+      "step": 2244
+    },
+    {
+      "epoch": 0.09812491804711744,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.766177589468701e-05,
+      "loss": 1.8902,
+      "step": 2245
+    },
+    {
+      "epoch": 0.09816862625114733,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.765969953710857e-05,
+      "loss": 1.9289,
+      "step": 2246
+    },
+    {
+      "epoch": 0.09821233445517724,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.765762228012222e-05,
+      "loss": 1.9364,
+      "step": 2247
+    },
+    {
+      "epoch": 0.09825604265920714,
+      "grad_norm": 18.75,
+      "learning_rate": 9.765554412376719e-05,
+      "loss": 6.6763,
+      "step": 2248
+    },
+    {
+      "epoch": 0.09829975086323703,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.765346506808266e-05,
+      "loss": 1.8456,
+      "step": 2249
+    },
+    {
+      "epoch": 0.09834345906726692,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.765138511310791e-05,
+      "loss": 2.7566,
+      "step": 2250
+    },
+    {
+      "epoch": 0.09838716727129682,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.764930425888215e-05,
+      "loss": 1.8674,
+      "step": 2251
+    },
+    {
+      "epoch": 0.09843087547532672,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.764722250544469e-05,
+      "loss": 2.1064,
+      "step": 2252
+    },
+    {
+      "epoch": 0.09847458367935662,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.764513985283478e-05,
+      "loss": 2.2086,
+      "step": 2253
+    },
+    {
+      "epoch": 0.09851829188338651,
+      "grad_norm": 2.625,
+      "learning_rate": 9.764305630109175e-05,
+      "loss": 2.2642,
+      "step": 2254
+    },
+    {
+      "epoch": 0.0985620000874164,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.764097185025489e-05,
+      "loss": 2.1661,
+      "step": 2255
+    },
+    {
+      "epoch": 0.0986057082914463,
+      "grad_norm": 2.375,
+      "learning_rate": 9.763888650036354e-05,
+      "loss": 1.985,
+      "step": 2256
+    },
+    {
+      "epoch": 0.0986494164954762,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.76368002514571e-05,
+      "loss": 2.8063,
+      "step": 2257
+    },
+    {
+      "epoch": 0.0986931246995061,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.763471310357488e-05,
+      "loss": 1.8268,
+      "step": 2258
+    },
+    {
+      "epoch": 0.09873683290353599,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.763262505675632e-05,
+      "loss": 2.2688,
+      "step": 2259
+    },
+    {
+      "epoch": 0.09878054110756589,
+      "grad_norm": 6.71875,
+      "learning_rate": 9.763053611104079e-05,
+      "loss": 2.1481,
+      "step": 2260
+    },
+    {
+      "epoch": 0.09882424931159578,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.76284462664677e-05,
+      "loss": 1.79,
+      "step": 2261
+    },
+    {
+      "epoch": 0.09886795751562569,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.762635552307653e-05,
+      "loss": 2.7382,
+      "step": 2262
+    },
+    {
+      "epoch": 0.09891166571965558,
+      "grad_norm": 2.5,
+      "learning_rate": 9.762426388090671e-05,
+      "loss": 2.0873,
+      "step": 2263
+    },
+    {
+      "epoch": 0.09895537392368547,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.762217133999771e-05,
+      "loss": 2.393,
+      "step": 2264
+    },
+    {
+      "epoch": 0.09899908212771537,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.762007790038904e-05,
+      "loss": 1.5162,
+      "step": 2265
+    },
+    {
+      "epoch": 0.09904279033174526,
+      "grad_norm": 3.875,
+      "learning_rate": 9.761798356212019e-05,
+      "loss": 3.1219,
+      "step": 2266
+    },
+    {
+      "epoch": 0.09908649853577517,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.761588832523067e-05,
+      "loss": 2.3465,
+      "step": 2267
+    },
+    {
+      "epoch": 0.09913020673980506,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.761379218976005e-05,
+      "loss": 1.912,
+      "step": 2268
+    },
+    {
+      "epoch": 0.09917391494383496,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.761169515574786e-05,
+      "loss": 1.8608,
+      "step": 2269
+    },
+    {
+      "epoch": 0.09921762314786485,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.760959722323371e-05,
+      "loss": 1.9141,
+      "step": 2270
+    },
+    {
+      "epoch": 0.09926133135189474,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.760749839225714e-05,
+      "loss": 1.937,
+      "step": 2271
+    },
+    {
+      "epoch": 0.09930503955592465,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.760539866285781e-05,
+      "loss": 2.1976,
+      "step": 2272
+    },
+    {
+      "epoch": 0.09934874775995454,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.76032980350753e-05,
+      "loss": 1.4496,
+      "step": 2273
+    },
+    {
+      "epoch": 0.09939245596398444,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.760119650894929e-05,
+      "loss": 2.3222,
+      "step": 2274
+    },
+    {
+      "epoch": 0.09943616416801433,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.759909408451942e-05,
+      "loss": 2.3326,
+      "step": 2275
+    },
+    {
+      "epoch": 0.09947987237204424,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.759699076182536e-05,
+      "loss": 1.8843,
+      "step": 2276
+    },
+    {
+      "epoch": 0.09952358057607413,
+      "grad_norm": 6.25,
+      "learning_rate": 9.759488654090681e-05,
+      "loss": 2.3823,
+      "step": 2277
+    },
+    {
+      "epoch": 0.09956728878010403,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.759278142180348e-05,
+      "loss": 2.1267,
+      "step": 2278
+    },
+    {
+      "epoch": 0.09961099698413392,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.759067540455511e-05,
+      "loss": 1.9924,
+      "step": 2279
+    },
+    {
+      "epoch": 0.09965470518816381,
+      "grad_norm": 6.0625,
+      "learning_rate": 9.758856848920142e-05,
+      "loss": 2.2601,
+      "step": 2280
+    },
+    {
+      "epoch": 0.09969841339219372,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.758646067578216e-05,
+      "loss": 2.5403,
+      "step": 2281
+    },
+    {
+      "epoch": 0.09974212159622362,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.758435196433716e-05,
+      "loss": 2.2937,
+      "step": 2282
+    },
+    {
+      "epoch": 0.09978582980025351,
+      "grad_norm": 5.125,
+      "learning_rate": 9.758224235490618e-05,
+      "loss": 1.7025,
+      "step": 2283
+    },
+    {
+      "epoch": 0.0998295380042834,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.758013184752901e-05,
+      "loss": 2.2129,
+      "step": 2284
+    },
+    {
+      "epoch": 0.0998732462083133,
+      "grad_norm": 2.5,
+      "learning_rate": 9.757802044224553e-05,
+      "loss": 2.5303,
+      "step": 2285
+    },
+    {
+      "epoch": 0.0999169544123432,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.757590813909554e-05,
+      "loss": 1.7859,
+      "step": 2286
+    },
+    {
+      "epoch": 0.0999606626163731,
+      "grad_norm": 2.375,
+      "learning_rate": 9.757379493811892e-05,
+      "loss": 2.1649,
+      "step": 2287
+    },
+    {
+      "epoch": 0.10000437082040299,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.757168083935556e-05,
+      "loss": 1.764,
+      "step": 2288
+    },
+    {
+      "epoch": 0.10004807902443288,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.756956584284533e-05,
+      "loss": 1.8986,
+      "step": 2289
+    },
+    {
+      "epoch": 0.10009178722846278,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.756744994862817e-05,
+      "loss": 2.2794,
+      "step": 2290
+    },
+    {
+      "epoch": 0.10013549543249269,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.7565333156744e-05,
+      "loss": 1.987,
+      "step": 2291
+    },
+    {
+      "epoch": 0.10017920363652258,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.756321546723277e-05,
+      "loss": 2.2158,
+      "step": 2292
+    },
+    {
+      "epoch": 0.10022291184055247,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.756109688013442e-05,
+      "loss": 1.4978,
+      "step": 2293
+    },
+    {
+      "epoch": 0.10026662004458237,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.755897739548896e-05,
+      "loss": 2.0564,
+      "step": 2294
+    },
+    {
+      "epoch": 0.10031032824861226,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.75568570133364e-05,
+      "loss": 1.9511,
+      "step": 2295
+    },
+    {
+      "epoch": 0.10035403645264217,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.75547357337167e-05,
+      "loss": 2.2555,
+      "step": 2296
+    },
+    {
+      "epoch": 0.10039774465667206,
+      "grad_norm": 2.375,
+      "learning_rate": 9.755261355666994e-05,
+      "loss": 1.7196,
+      "step": 2297
+    },
+    {
+      "epoch": 0.10044145286070195,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.755049048223615e-05,
+      "loss": 1.6631,
+      "step": 2298
+    },
+    {
+      "epoch": 0.10048516106473185,
+      "grad_norm": 2.625,
+      "learning_rate": 9.754836651045538e-05,
+      "loss": 1.9797,
+      "step": 2299
+    },
+    {
+      "epoch": 0.10052886926876174,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.754624164136774e-05,
+      "loss": 2.3664,
+      "step": 2300
+    },
+    {
+      "epoch": 0.10057257747279165,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.754411587501333e-05,
+      "loss": 1.7274,
+      "step": 2301
+    },
+    {
+      "epoch": 0.10061628567682154,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.754198921143226e-05,
+      "loss": 1.8973,
+      "step": 2302
+    },
+    {
+      "epoch": 0.10065999388085144,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.753986165066464e-05,
+      "loss": 2.0499,
+      "step": 2303
+    },
+    {
+      "epoch": 0.10070370208488133,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.753773319275065e-05,
+      "loss": 1.7278,
+      "step": 2304
+    },
+    {
+      "epoch": 0.10074741028891122,
+      "grad_norm": 2.875,
+      "learning_rate": 9.753560383773046e-05,
+      "loss": 2.5053,
+      "step": 2305
+    },
+    {
+      "epoch": 0.10079111849294113,
+      "grad_norm": 14.0,
+      "learning_rate": 9.753347358564423e-05,
+      "loss": 5.9202,
+      "step": 2306
+    },
+    {
+      "epoch": 0.10083482669697102,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.753134243653217e-05,
+      "loss": 1.9736,
+      "step": 2307
+    },
+    {
+      "epoch": 0.10087853490100092,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.75292103904345e-05,
+      "loss": 2.354,
+      "step": 2308
+    },
+    {
+      "epoch": 0.10092224310503081,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.752707744739145e-05,
+      "loss": 3.0874,
+      "step": 2309
+    },
+    {
+      "epoch": 0.1009659513090607,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.752494360744329e-05,
+      "loss": 2.251,
+      "step": 2310
+    },
+    {
+      "epoch": 0.10100965951309061,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.752280887063026e-05,
+      "loss": 2.0361,
+      "step": 2311
+    },
+    {
+      "epoch": 0.1010533677171205,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.752067323699267e-05,
+      "loss": 2.2617,
+      "step": 2312
+    },
+    {
+      "epoch": 0.1010970759211504,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.751853670657081e-05,
+      "loss": 1.9537,
+      "step": 2313
+    },
+    {
+      "epoch": 0.1011407841251803,
+      "grad_norm": 2.875,
+      "learning_rate": 9.7516399279405e-05,
+      "loss": 2.3131,
+      "step": 2314
+    },
+    {
+      "epoch": 0.10118449232921019,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.751426095553557e-05,
+      "loss": 3.22,
+      "step": 2315
+    },
+    {
+      "epoch": 0.1012282005332401,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.751212173500291e-05,
+      "loss": 1.9288,
+      "step": 2316
+    },
+    {
+      "epoch": 0.10127190873726999,
+      "grad_norm": 2.625,
+      "learning_rate": 9.750998161784734e-05,
+      "loss": 2.2181,
+      "step": 2317
+    },
+    {
+      "epoch": 0.10131561694129988,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.750784060410927e-05,
+      "loss": 2.2256,
+      "step": 2318
+    },
+    {
+      "epoch": 0.10135932514532978,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.750569869382911e-05,
+      "loss": 2.3126,
+      "step": 2319
+    },
+    {
+      "epoch": 0.10140303334935967,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.750355588704727e-05,
+      "loss": 2.6601,
+      "step": 2320
+    },
+    {
+      "epoch": 0.10144674155338958,
+      "grad_norm": 3.375,
+      "learning_rate": 9.750141218380419e-05,
+      "loss": 2.0413,
+      "step": 2321
+    },
+    {
+      "epoch": 0.10149044975741947,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.749926758414035e-05,
+      "loss": 2.0241,
+      "step": 2322
+    },
+    {
+      "epoch": 0.10153415796144936,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.749712208809618e-05,
+      "loss": 2.1908,
+      "step": 2323
+    },
+    {
+      "epoch": 0.10157786616547926,
+      "grad_norm": 3.25,
+      "learning_rate": 9.749497569571217e-05,
+      "loss": 1.5666,
+      "step": 2324
+    },
+    {
+      "epoch": 0.10162157436950915,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.749282840702887e-05,
+      "loss": 2.0106,
+      "step": 2325
+    },
+    {
+      "epoch": 0.10166528257353906,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.749068022208676e-05,
+      "loss": 2.8774,
+      "step": 2326
+    },
+    {
+      "epoch": 0.10170899077756895,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.748853114092639e-05,
+      "loss": 1.7896,
+      "step": 2327
+    },
+    {
+      "epoch": 0.10175269898159885,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.748638116358834e-05,
+      "loss": 2.3381,
+      "step": 2328
+    },
+    {
+      "epoch": 0.10179640718562874,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.748423029011317e-05,
+      "loss": 1.9284,
+      "step": 2329
+    },
+    {
+      "epoch": 0.10184011538965863,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.748207852054144e-05,
+      "loss": 1.5755,
+      "step": 2330
+    },
+    {
+      "epoch": 0.10188382359368854,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.747992585491379e-05,
+      "loss": 2.5287,
+      "step": 2331
+    },
+    {
+      "epoch": 0.10192753179771843,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.747777229327084e-05,
+      "loss": 2.5582,
+      "step": 2332
+    },
+    {
+      "epoch": 0.10197124000174833,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.747561783565323e-05,
+      "loss": 1.9647,
+      "step": 2333
+    },
+    {
+      "epoch": 0.10201494820577822,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.747346248210161e-05,
+      "loss": 2.9669,
+      "step": 2334
+    },
+    {
+      "epoch": 0.10205865640980812,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.747130623265665e-05,
+      "loss": 2.9084,
+      "step": 2335
+    },
+    {
+      "epoch": 0.10210236461383802,
+      "grad_norm": 3.375,
+      "learning_rate": 9.746914908735906e-05,
+      "loss": 1.8825,
+      "step": 2336
+    },
+    {
+      "epoch": 0.10214607281786792,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.746699104624953e-05,
+      "loss": 2.0626,
+      "step": 2337
+    },
+    {
+      "epoch": 0.10218978102189781,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.746483210936881e-05,
+      "loss": 2.3356,
+      "step": 2338
+    },
+    {
+      "epoch": 0.1022334892259277,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.74626722767576e-05,
+      "loss": 2.2161,
+      "step": 2339
+    },
+    {
+      "epoch": 0.1022771974299576,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.74605115484567e-05,
+      "loss": 2.2453,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1023209056339875,
+      "grad_norm": 2.5,
+      "learning_rate": 9.745834992450689e-05,
+      "loss": 1.6678,
+      "step": 2341
+    },
+    {
+      "epoch": 0.1023646138380174,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.745618740494892e-05,
+      "loss": 1.775,
+      "step": 2342
+    },
+    {
+      "epoch": 0.10240832204204729,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.745402398982363e-05,
+      "loss": 1.9103,
+      "step": 2343
+    },
+    {
+      "epoch": 0.10245203024607719,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.745185967917184e-05,
+      "loss": 1.9037,
+      "step": 2344
+    },
+    {
+      "epoch": 0.10249573845010708,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.74496944730344e-05,
+      "loss": 2.4936,
+      "step": 2345
+    },
+    {
+      "epoch": 0.10253944665413699,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.744752837145217e-05,
+      "loss": 1.9646,
+      "step": 2346
+    },
+    {
+      "epoch": 0.10258315485816688,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.744536137446601e-05,
+      "loss": 1.7475,
+      "step": 2347
+    },
+    {
+      "epoch": 0.10262686306219677,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.744319348211684e-05,
+      "loss": 2.0058,
+      "step": 2348
+    },
+    {
+      "epoch": 0.10267057126622667,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.744102469444555e-05,
+      "loss": 2.3312,
+      "step": 2349
+    },
+    {
+      "epoch": 0.10271427947025656,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.743885501149308e-05,
+      "loss": 2.1554,
+      "step": 2350
+    },
+    {
+      "epoch": 0.10275798767428647,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.743668443330037e-05,
+      "loss": 2.3322,
+      "step": 2351
+    },
+    {
+      "epoch": 0.10280169587831636,
+      "grad_norm": 2.375,
+      "learning_rate": 9.743451295990837e-05,
+      "loss": 1.9326,
+      "step": 2352
+    },
+    {
+      "epoch": 0.10284540408234626,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.743234059135811e-05,
+      "loss": 2.1341,
+      "step": 2353
+    },
+    {
+      "epoch": 0.10288911228637615,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.743016732769053e-05,
+      "loss": 2.1587,
+      "step": 2354
+    },
+    {
+      "epoch": 0.10293282049040604,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.742799316894663e-05,
+      "loss": 2.2016,
+      "step": 2355
+    },
+    {
+      "epoch": 0.10297652869443595,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.742581811516751e-05,
+      "loss": 1.9144,
+      "step": 2356
+    },
+    {
+      "epoch": 0.10302023689846584,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.742364216639416e-05,
+      "loss": 2.4872,
+      "step": 2357
+    },
+    {
+      "epoch": 0.10306394510249574,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.742146532266767e-05,
+      "loss": 1.7991,
+      "step": 2358
+    },
+    {
+      "epoch": 0.10310765330652563,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.741928758402912e-05,
+      "loss": 2.3372,
+      "step": 2359
+    },
+    {
+      "epoch": 0.10315136151055553,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.741710895051958e-05,
+      "loss": 1.8571,
+      "step": 2360
+    },
+    {
+      "epoch": 0.10319506971458543,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.741492942218018e-05,
+      "loss": 1.9904,
+      "step": 2361
+    },
+    {
+      "epoch": 0.10323877791861533,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.741274899905207e-05,
+      "loss": 2.087,
+      "step": 2362
+    },
+    {
+      "epoch": 0.10328248612264522,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.741056768117636e-05,
+      "loss": 2.236,
+      "step": 2363
+    },
+    {
+      "epoch": 0.10332619432667511,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.740838546859426e-05,
+      "loss": 1.9287,
+      "step": 2364
+    },
+    {
+      "epoch": 0.10336990253070501,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.74062023613469e-05,
+      "loss": 2.3908,
+      "step": 2365
+    },
+    {
+      "epoch": 0.10341361073473491,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.740401835947551e-05,
+      "loss": 1.9419,
+      "step": 2366
+    },
+    {
+      "epoch": 0.10345731893876481,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.740183346302131e-05,
+      "loss": 2.2898,
+      "step": 2367
+    },
+    {
+      "epoch": 0.1035010271427947,
+      "grad_norm": 2.375,
+      "learning_rate": 9.739964767202552e-05,
+      "loss": 2.1998,
+      "step": 2368
+    },
+    {
+      "epoch": 0.1035447353468246,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.739746098652939e-05,
+      "loss": 2.2489,
+      "step": 2369
+    },
+    {
+      "epoch": 0.10358844355085449,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.73952734065742e-05,
+      "loss": 1.9166,
+      "step": 2370
+    },
+    {
+      "epoch": 0.1036321517548844,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.73930849322012e-05,
+      "loss": 1.8404,
+      "step": 2371
+    },
+    {
+      "epoch": 0.10367585995891429,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.739089556345171e-05,
+      "loss": 2.4283,
+      "step": 2372
+    },
+    {
+      "epoch": 0.10371956816294418,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.738870530036706e-05,
+      "loss": 1.9826,
+      "step": 2373
+    },
+    {
+      "epoch": 0.10376327636697408,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.738651414298857e-05,
+      "loss": 2.3027,
+      "step": 2374
+    },
+    {
+      "epoch": 0.10380698457100397,
+      "grad_norm": 2.625,
+      "learning_rate": 9.738432209135757e-05,
+      "loss": 1.8125,
+      "step": 2375
+    },
+    {
+      "epoch": 0.10385069277503388,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.738212914551547e-05,
+      "loss": 1.6889,
+      "step": 2376
+    },
+    {
+      "epoch": 0.10389440097906377,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.737993530550362e-05,
+      "loss": 2.5132,
+      "step": 2377
+    },
+    {
+      "epoch": 0.10393810918309367,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.737774057136344e-05,
+      "loss": 2.1089,
+      "step": 2378
+    },
+    {
+      "epoch": 0.10398181738712356,
+      "grad_norm": 5.0,
+      "learning_rate": 9.737554494313635e-05,
+      "loss": 2.1395,
+      "step": 2379
+    },
+    {
+      "epoch": 0.10402552559115345,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.737334842086374e-05,
+      "loss": 2.4614,
+      "step": 2380
+    },
+    {
+      "epoch": 0.10406923379518336,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.737115100458713e-05,
+      "loss": 1.8017,
+      "step": 2381
+    },
+    {
+      "epoch": 0.10411294199921325,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.736895269434794e-05,
+      "loss": 2.0097,
+      "step": 2382
+    },
+    {
+      "epoch": 0.10415665020324315,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.736675349018767e-05,
+      "loss": 1.8963,
+      "step": 2383
+    },
+    {
+      "epoch": 0.10420035840727304,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.736455339214783e-05,
+      "loss": 2.386,
+      "step": 2384
+    },
+    {
+      "epoch": 0.10424406661130294,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.736235240026993e-05,
+      "loss": 2.0695,
+      "step": 2385
+    },
+    {
+      "epoch": 0.10428777481533284,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.73601505145955e-05,
+      "loss": 2.3432,
+      "step": 2386
+    },
+    {
+      "epoch": 0.10433148301936274,
+      "grad_norm": 3.25,
+      "learning_rate": 9.735794773516611e-05,
+      "loss": 2.3339,
+      "step": 2387
+    },
+    {
+      "epoch": 0.10437519122339263,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.735574406202332e-05,
+      "loss": 1.8667,
+      "step": 2388
+    },
+    {
+      "epoch": 0.10441889942742252,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.735353949520871e-05,
+      "loss": 2.2035,
+      "step": 2389
+    },
+    {
+      "epoch": 0.10446260763145242,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.73513340347639e-05,
+      "loss": 1.8563,
+      "step": 2390
+    },
+    {
+      "epoch": 0.10450631583548232,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.73491276807305e-05,
+      "loss": 1.6536,
+      "step": 2391
+    },
+    {
+      "epoch": 0.10455002403951222,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.734692043315012e-05,
+      "loss": 2.2219,
+      "step": 2392
+    },
+    {
+      "epoch": 0.10459373224354211,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.734471229206448e-05,
+      "loss": 2.2594,
+      "step": 2393
+    },
+    {
+      "epoch": 0.104637440447572,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.73425032575152e-05,
+      "loss": 2.0679,
+      "step": 2394
+    },
+    {
+      "epoch": 0.1046811486516019,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.734029332954395e-05,
+      "loss": 2.0133,
+      "step": 2395
+    },
+    {
+      "epoch": 0.1047248568556318,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.73380825081925e-05,
+      "loss": 1.9773,
+      "step": 2396
+    },
+    {
+      "epoch": 0.1047685650596617,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.733587079350252e-05,
+      "loss": 2.7695,
+      "step": 2397
+    },
+    {
+      "epoch": 0.1048122732636916,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.733365818551576e-05,
+      "loss": 1.9399,
+      "step": 2398
+    },
+    {
+      "epoch": 0.10485598146772149,
+      "grad_norm": 2.625,
+      "learning_rate": 9.7331444684274e-05,
+      "loss": 2.1683,
+      "step": 2399
+    },
+    {
+      "epoch": 0.10489968967175138,
+      "grad_norm": 3.0,
+      "learning_rate": 9.732923028981897e-05,
+      "loss": 2.5887,
+      "step": 2400
+    },
+    {
+      "epoch": 0.10494339787578129,
+      "grad_norm": 2.75,
+      "learning_rate": 9.732701500219251e-05,
+      "loss": 1.808,
+      "step": 2401
+    },
+    {
+      "epoch": 0.10498710607981118,
+      "grad_norm": 4.8125,
+      "learning_rate": 9.732479882143636e-05,
+      "loss": 3.8407,
+      "step": 2402
+    },
+    {
+      "epoch": 0.10503081428384108,
+      "grad_norm": 7.09375,
+      "learning_rate": 9.732258174759239e-05,
+      "loss": 3.2358,
+      "step": 2403
+    },
+    {
+      "epoch": 0.10507452248787097,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.732036378070243e-05,
+      "loss": 1.934,
+      "step": 2404
+    },
+    {
+      "epoch": 0.10511823069190086,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.731814492080832e-05,
+      "loss": 2.7517,
+      "step": 2405
+    },
+    {
+      "epoch": 0.10516193889593077,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.731592516795197e-05,
+      "loss": 2.9847,
+      "step": 2406
+    },
+    {
+      "epoch": 0.10520564709996066,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.731370452217524e-05,
+      "loss": 1.8667,
+      "step": 2407
+    },
+    {
+      "epoch": 0.10524935530399056,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.731148298352004e-05,
+      "loss": 2.0742,
+      "step": 2408
+    },
+    {
+      "epoch": 0.10529306350802045,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.73092605520283e-05,
+      "loss": 1.9423,
+      "step": 2409
+    },
+    {
+      "epoch": 0.10533677171205034,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.730703722774196e-05,
+      "loss": 2.7559,
+      "step": 2410
+    },
+    {
+      "epoch": 0.10538047991608025,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.730481301070298e-05,
+      "loss": 2.0713,
+      "step": 2411
+    },
+    {
+      "epoch": 0.10542418812011015,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.730258790095331e-05,
+      "loss": 2.1342,
+      "step": 2412
+    },
+    {
+      "epoch": 0.10546789632414004,
+      "grad_norm": 2.875,
+      "learning_rate": 9.730036189853498e-05,
+      "loss": 2.3037,
+      "step": 2413
+    },
+    {
+      "epoch": 0.10551160452816993,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.729813500348997e-05,
+      "loss": 2.0718,
+      "step": 2414
+    },
+    {
+      "epoch": 0.10555531273219984,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.72959072158603e-05,
+      "loss": 3.004,
+      "step": 2415
+    },
+    {
+      "epoch": 0.10559902093622973,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.729367853568805e-05,
+      "loss": 1.9165,
+      "step": 2416
+    },
+    {
+      "epoch": 0.10564272914025963,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.729144896301524e-05,
+      "loss": 1.8318,
+      "step": 2417
+    },
+    {
+      "epoch": 0.10568643734428952,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.728921849788397e-05,
+      "loss": 2.5869,
+      "step": 2418
+    },
+    {
+      "epoch": 0.10573014554831942,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.72869871403363e-05,
+      "loss": 2.5118,
+      "step": 2419
+    },
+    {
+      "epoch": 0.10577385375234932,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.728475489041438e-05,
+      "loss": 2.0763,
+      "step": 2420
+    },
+    {
+      "epoch": 0.10581756195637922,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.728252174816031e-05,
+      "loss": 2.224,
+      "step": 2421
+    },
+    {
+      "epoch": 0.10586127016040911,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.728028771361624e-05,
+      "loss": 2.1799,
+      "step": 2422
+    },
+    {
+      "epoch": 0.105904978364439,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.727805278682431e-05,
+      "loss": 1.9667,
+      "step": 2423
+    },
+    {
+      "epoch": 0.1059486865684689,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.727581696782673e-05,
+      "loss": 2.0938,
+      "step": 2424
+    },
+    {
+      "epoch": 0.1059923947724988,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.727358025666568e-05,
+      "loss": 2.1156,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1060361029765287,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.727134265338335e-05,
+      "loss": 1.8436,
+      "step": 2426
+    },
+    {
+      "epoch": 0.10607981118055859,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.7269104158022e-05,
+      "loss": 2.3787,
+      "step": 2427
+    },
+    {
+      "epoch": 0.10612351938458849,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.726686477062386e-05,
+      "loss": 1.9247,
+      "step": 2428
+    },
+    {
+      "epoch": 0.10616722758861838,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.726462449123117e-05,
+      "loss": 2.2084,
+      "step": 2429
+    },
+    {
+      "epoch": 0.10621093579264829,
+      "grad_norm": 2.375,
+      "learning_rate": 9.726238331988624e-05,
+      "loss": 2.3032,
+      "step": 2430
+    },
+    {
+      "epoch": 0.10625464399667818,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.726014125663135e-05,
+      "loss": 1.9042,
+      "step": 2431
+    },
+    {
+      "epoch": 0.10629835220070807,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.725789830150882e-05,
+      "loss": 2.1058,
+      "step": 2432
+    },
+    {
+      "epoch": 0.10634206040473797,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.725565445456095e-05,
+      "loss": 1.7417,
+      "step": 2433
+    },
+    {
+      "epoch": 0.10638576860876786,
+      "grad_norm": 7.0,
+      "learning_rate": 9.72534097158301e-05,
+      "loss": 2.8365,
+      "step": 2434
+    },
+    {
+      "epoch": 0.10642947681279777,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.725116408535864e-05,
+      "loss": 2.0676,
+      "step": 2435
+    },
+    {
+      "epoch": 0.10647318501682766,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.724891756318895e-05,
+      "loss": 2.0782,
+      "step": 2436
+    },
+    {
+      "epoch": 0.10651689322085756,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.724667014936342e-05,
+      "loss": 1.8681,
+      "step": 2437
+    },
+    {
+      "epoch": 0.10656060142488745,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.724442184392445e-05,
+      "loss": 2.1118,
+      "step": 2438
+    },
+    {
+      "epoch": 0.10660430962891734,
+      "grad_norm": 2.625,
+      "learning_rate": 9.724217264691448e-05,
+      "loss": 2.16,
+      "step": 2439
+    },
+    {
+      "epoch": 0.10664801783294725,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.723992255837596e-05,
+      "loss": 1.5494,
+      "step": 2440
+    },
+    {
+      "epoch": 0.10669172603697714,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.723767157835135e-05,
+      "loss": 2.2222,
+      "step": 2441
+    },
+    {
+      "epoch": 0.10673543424100704,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.723541970688311e-05,
+      "loss": 2.3099,
+      "step": 2442
+    },
+    {
+      "epoch": 0.10677914244503693,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.723316694401377e-05,
+      "loss": 2.4099,
+      "step": 2443
+    },
+    {
+      "epoch": 0.10682285064906683,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.723091328978581e-05,
+      "loss": 2.0523,
+      "step": 2444
+    },
+    {
+      "epoch": 0.10686655885309673,
+      "grad_norm": 2.75,
+      "learning_rate": 9.722865874424178e-05,
+      "loss": 2.5832,
+      "step": 2445
+    },
+    {
+      "epoch": 0.10691026705712663,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.722640330742423e-05,
+      "loss": 2.1906,
+      "step": 2446
+    },
+    {
+      "epoch": 0.10695397526115652,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.722414697937572e-05,
+      "loss": 2.4322,
+      "step": 2447
+    },
+    {
+      "epoch": 0.10699768346518641,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.72218897601388e-05,
+      "loss": 1.8567,
+      "step": 2448
+    },
+    {
+      "epoch": 0.10704139166921631,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.721963164975612e-05,
+      "loss": 1.9813,
+      "step": 2449
+    },
+    {
+      "epoch": 0.10708509987324621,
+      "grad_norm": 2.625,
+      "learning_rate": 9.721737264827025e-05,
+      "loss": 2.5686,
+      "step": 2450
+    },
+    {
+      "epoch": 0.10712880807727611,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.721511275572384e-05,
+      "loss": 1.8549,
+      "step": 2451
+    },
+    {
+      "epoch": 0.107172516281306,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.721285197215952e-05,
+      "loss": 2.0145,
+      "step": 2452
+    },
+    {
+      "epoch": 0.1072162244853359,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.721059029761999e-05,
+      "loss": 1.6011,
+      "step": 2453
+    },
+    {
+      "epoch": 0.10725993268936579,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.720832773214789e-05,
+      "loss": 1.7277,
+      "step": 2454
+    },
+    {
+      "epoch": 0.1073036408933957,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.720606427578595e-05,
+      "loss": 2.0673,
+      "step": 2455
+    },
+    {
+      "epoch": 0.10734734909742559,
+      "grad_norm": 2.75,
+      "learning_rate": 9.720379992857687e-05,
+      "loss": 2.2455,
+      "step": 2456
+    },
+    {
+      "epoch": 0.10739105730145548,
+      "grad_norm": 2.375,
+      "learning_rate": 9.720153469056338e-05,
+      "loss": 2.2157,
+      "step": 2457
+    },
+    {
+      "epoch": 0.10743476550548538,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.719926856178823e-05,
+      "loss": 1.8761,
+      "step": 2458
+    },
+    {
+      "epoch": 0.10747847370951527,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.71970015422942e-05,
+      "loss": 2.5707,
+      "step": 2459
+    },
+    {
+      "epoch": 0.10752218191354518,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.719473363212405e-05,
+      "loss": 2.0183,
+      "step": 2460
+    },
+    {
+      "epoch": 0.10756589011757507,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.719246483132058e-05,
+      "loss": 2.1769,
+      "step": 2461
+    },
+    {
+      "epoch": 0.10760959832160497,
+      "grad_norm": 2.25,
+      "learning_rate": 9.719019513992662e-05,
+      "loss": 1.8736,
+      "step": 2462
+    },
+    {
+      "epoch": 0.10765330652563486,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.7187924557985e-05,
+      "loss": 2.0864,
+      "step": 2463
+    },
+    {
+      "epoch": 0.10769701472966475,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.718565308553857e-05,
+      "loss": 1.8813,
+      "step": 2464
+    },
+    {
+      "epoch": 0.10774072293369466,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.718338072263017e-05,
+      "loss": 1.9003,
+      "step": 2465
+    },
+    {
+      "epoch": 0.10778443113772455,
+      "grad_norm": 3.75,
+      "learning_rate": 9.718110746930272e-05,
+      "loss": 1.6998,
+      "step": 2466
+    },
+    {
+      "epoch": 0.10782813934175445,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.71788333255991e-05,
+      "loss": 1.8179,
+      "step": 2467
+    },
+    {
+      "epoch": 0.10787184754578434,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.717655829156222e-05,
+      "loss": 1.872,
+      "step": 2468
+    },
+    {
+      "epoch": 0.10791555574981423,
+      "grad_norm": 6.28125,
+      "learning_rate": 9.717428236723505e-05,
+      "loss": 2.1897,
+      "step": 2469
+    },
+    {
+      "epoch": 0.10795926395384414,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.717200555266049e-05,
+      "loss": 1.6204,
+      "step": 2470
+    },
+    {
+      "epoch": 0.10800297215787404,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.716972784788152e-05,
+      "loss": 2.1799,
+      "step": 2471
+    },
+    {
+      "epoch": 0.10804668036190393,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.716744925294116e-05,
+      "loss": 2.0839,
+      "step": 2472
+    },
+    {
+      "epoch": 0.10809038856593382,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.716516976788236e-05,
+      "loss": 2.656,
+      "step": 2473
+    },
+    {
+      "epoch": 0.10813409676996372,
+      "grad_norm": 2.625,
+      "learning_rate": 9.716288939274819e-05,
+      "loss": 2.4926,
+      "step": 2474
+    },
+    {
+      "epoch": 0.10817780497399362,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.716060812758163e-05,
+      "loss": 2.2936,
+      "step": 2475
+    },
+    {
+      "epoch": 0.10822151317802352,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.715832597242576e-05,
+      "loss": 1.5035,
+      "step": 2476
+    },
+    {
+      "epoch": 0.10826522138205341,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.715604292732366e-05,
+      "loss": 2.7127,
+      "step": 2477
+    },
+    {
+      "epoch": 0.1083089295860833,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.715375899231837e-05,
+      "loss": 2.8202,
+      "step": 2478
+    },
+    {
+      "epoch": 0.1083526377901132,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.715147416745303e-05,
+      "loss": 1.6242,
+      "step": 2479
+    },
+    {
+      "epoch": 0.1083963459941431,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.714918845277075e-05,
+      "loss": 2.0765,
+      "step": 2480
+    },
+    {
+      "epoch": 0.108440054198173,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.714690184831465e-05,
+      "loss": 2.0368,
+      "step": 2481
+    },
+    {
+      "epoch": 0.1084837624022029,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.714461435412792e-05,
+      "loss": 2.2531,
+      "step": 2482
+    },
+    {
+      "epoch": 0.10852747060623279,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.714232597025368e-05,
+      "loss": 1.8642,
+      "step": 2483
+    },
+    {
+      "epoch": 0.10857117881026268,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.714003669673515e-05,
+      "loss": 2.1971,
+      "step": 2484
+    },
+    {
+      "epoch": 0.10861488701429259,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.713774653361549e-05,
+      "loss": 1.6596,
+      "step": 2485
+    },
+    {
+      "epoch": 0.10865859521832248,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.713545548093797e-05,
+      "loss": 2.9122,
+      "step": 2486
+    },
+    {
+      "epoch": 0.10870230342235238,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.713316353874581e-05,
+      "loss": 1.8774,
+      "step": 2487
+    },
+    {
+      "epoch": 0.10874601162638227,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.713087070708224e-05,
+      "loss": 2.1811,
+      "step": 2488
+    },
+    {
+      "epoch": 0.10878971983041216,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.712857698599054e-05,
+      "loss": 1.9505,
+      "step": 2489
+    },
+    {
+      "epoch": 0.10883342803444207,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.712628237551402e-05,
+      "loss": 1.8161,
+      "step": 2490
+    },
+    {
+      "epoch": 0.10887713623847196,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.712398687569595e-05,
+      "loss": 1.9934,
+      "step": 2491
+    },
+    {
+      "epoch": 0.10892084444250186,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.712169048657966e-05,
+      "loss": 2.0326,
+      "step": 2492
+    },
+    {
+      "epoch": 0.10896455264653175,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.711939320820848e-05,
+      "loss": 1.9797,
+      "step": 2493
+    },
+    {
+      "epoch": 0.10900826085056164,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.71170950406258e-05,
+      "loss": 1.9381,
+      "step": 2494
+    },
+    {
+      "epoch": 0.10905196905459155,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.711479598387494e-05,
+      "loss": 2.5,
+      "step": 2495
+    },
+    {
+      "epoch": 0.10909567725862145,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.71124960379993e-05,
+      "loss": 1.9666,
+      "step": 2496
+    },
+    {
+      "epoch": 0.10913938546265134,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.711019520304231e-05,
+      "loss": 2.0122,
+      "step": 2497
+    },
+    {
+      "epoch": 0.10918309366668123,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.710789347904736e-05,
+      "loss": 2.2053,
+      "step": 2498
+    },
+    {
+      "epoch": 0.10922680187071113,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.71055908660579e-05,
+      "loss": 2.1717,
+      "step": 2499
+    },
+    {
+      "epoch": 0.10927051007474103,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.710328736411737e-05,
+      "loss": 2.2256,
+      "step": 2500
+    },
+    {
+      "epoch": 0.10931421827877093,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.710098297326928e-05,
+      "loss": 1.8244,
+      "step": 2501
+    },
+    {
+      "epoch": 0.10935792648280082,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.709867769355707e-05,
+      "loss": 2.2914,
+      "step": 2502
+    },
+    {
+      "epoch": 0.10940163468683071,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.709637152502427e-05,
+      "loss": 2.2509,
+      "step": 2503
+    },
+    {
+      "epoch": 0.10944534289086061,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.709406446771439e-05,
+      "loss": 2.1208,
+      "step": 2504
+    },
+    {
+      "epoch": 0.10948905109489052,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.709175652167096e-05,
+      "loss": 1.6196,
+      "step": 2505
+    },
+    {
+      "epoch": 0.10953275929892041,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.708944768693755e-05,
+      "loss": 2.3912,
+      "step": 2506
+    },
+    {
+      "epoch": 0.1095764675029503,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.708713796355773e-05,
+      "loss": 2.3145,
+      "step": 2507
+    },
+    {
+      "epoch": 0.1096201757069802,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.708482735157509e-05,
+      "loss": 2.4361,
+      "step": 2508
+    },
+    {
+      "epoch": 0.10966388391101009,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.708251585103322e-05,
+      "loss": 2.2467,
+      "step": 2509
+    },
+    {
+      "epoch": 0.10970759211504,
+      "grad_norm": 4.375,
+      "learning_rate": 9.708020346197577e-05,
+      "loss": 2.3246,
+      "step": 2510
+    },
+    {
+      "epoch": 0.10975130031906989,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.707789018444636e-05,
+      "loss": 2.2745,
+      "step": 2511
+    },
+    {
+      "epoch": 0.10979500852309979,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.707557601848862e-05,
+      "loss": 2.3604,
+      "step": 2512
+    },
+    {
+      "epoch": 0.10983871672712968,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.707326096414625e-05,
+      "loss": 1.7883,
+      "step": 2513
+    },
+    {
+      "epoch": 0.10988242493115957,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.707094502146294e-05,
+      "loss": 1.9555,
+      "step": 2514
+    },
+    {
+      "epoch": 0.10992613313518948,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.706862819048239e-05,
+      "loss": 2.4865,
+      "step": 2515
+    },
+    {
+      "epoch": 0.10996984133921937,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.706631047124833e-05,
+      "loss": 2.1589,
+      "step": 2516
+    },
+    {
+      "epoch": 0.11001354954324927,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.706399186380446e-05,
+      "loss": 1.9848,
+      "step": 2517
+    },
+    {
+      "epoch": 0.11005725774727916,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.706167236819459e-05,
+      "loss": 2.1828,
+      "step": 2518
+    },
+    {
+      "epoch": 0.11010096595130905,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.705935198446246e-05,
+      "loss": 1.9501,
+      "step": 2519
+    },
+    {
+      "epoch": 0.11014467415533896,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.705703071265187e-05,
+      "loss": 2.0098,
+      "step": 2520
+    },
+    {
+      "epoch": 0.11018838235936886,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.70547085528066e-05,
+      "loss": 1.8923,
+      "step": 2521
+    },
+    {
+      "epoch": 0.11023209056339875,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.705238550497053e-05,
+      "loss": 2.8409,
+      "step": 2522
+    },
+    {
+      "epoch": 0.11027579876742864,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.705006156918744e-05,
+      "loss": 1.8483,
+      "step": 2523
+    },
+    {
+      "epoch": 0.11031950697145854,
+      "grad_norm": 2.625,
+      "learning_rate": 9.704773674550123e-05,
+      "loss": 2.0112,
+      "step": 2524
+    },
+    {
+      "epoch": 0.11036321517548844,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.704541103395574e-05,
+      "loss": 2.122,
+      "step": 2525
+    },
+    {
+      "epoch": 0.11040692337951834,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.704308443459487e-05,
+      "loss": 2.662,
+      "step": 2526
+    },
+    {
+      "epoch": 0.11045063158354823,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.704075694746253e-05,
+      "loss": 2.0194,
+      "step": 2527
+    },
+    {
+      "epoch": 0.11049433978757812,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.703842857260263e-05,
+      "loss": 1.7063,
+      "step": 2528
+    },
+    {
+      "epoch": 0.11053804799160802,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.703609931005914e-05,
+      "loss": 1.9463,
+      "step": 2529
+    },
+    {
+      "epoch": 0.11058175619563793,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.703376915987601e-05,
+      "loss": 1.9706,
+      "step": 2530
+    },
+    {
+      "epoch": 0.11062546439966782,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.703143812209718e-05,
+      "loss": 2.4519,
+      "step": 2531
+    },
+    {
+      "epoch": 0.11066917260369771,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.702910619676667e-05,
+      "loss": 2.0152,
+      "step": 2532
+    },
+    {
+      "epoch": 0.1107128808077276,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.702677338392847e-05,
+      "loss": 2.2778,
+      "step": 2533
+    },
+    {
+      "epoch": 0.1107565890117575,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.702443968362662e-05,
+      "loss": 1.9025,
+      "step": 2534
+    },
+    {
+      "epoch": 0.11080029721578741,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.702210509590514e-05,
+      "loss": 2.1866,
+      "step": 2535
+    },
+    {
+      "epoch": 0.1108440054198173,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.701976962080812e-05,
+      "loss": 1.7364,
+      "step": 2536
+    },
+    {
+      "epoch": 0.1108877136238472,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.70174332583796e-05,
+      "loss": 2.5829,
+      "step": 2537
+    },
+    {
+      "epoch": 0.11093142182787709,
+      "grad_norm": 5.65625,
+      "learning_rate": 9.701509600866368e-05,
+      "loss": 2.1956,
+      "step": 2538
+    },
+    {
+      "epoch": 0.11097513003190698,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.701275787170448e-05,
+      "loss": 2.4632,
+      "step": 2539
+    },
+    {
+      "epoch": 0.11101883823593689,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.701041884754612e-05,
+      "loss": 2.389,
+      "step": 2540
+    },
+    {
+      "epoch": 0.11106254643996678,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.700807893623272e-05,
+      "loss": 2.3511,
+      "step": 2541
+    },
+    {
+      "epoch": 0.11110625464399668,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.700573813780847e-05,
+      "loss": 2.1516,
+      "step": 2542
+    },
+    {
+      "epoch": 0.11114996284802657,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.700339645231751e-05,
+      "loss": 2.1829,
+      "step": 2543
+    },
+    {
+      "epoch": 0.11119367105205646,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.700105387980406e-05,
+      "loss": 2.7425,
+      "step": 2544
+    },
+    {
+      "epoch": 0.11123737925608637,
+      "grad_norm": 3.0,
+      "learning_rate": 9.699871042031232e-05,
+      "loss": 2.3237,
+      "step": 2545
+    },
+    {
+      "epoch": 0.11128108746011627,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.69963660738865e-05,
+      "loss": 1.8203,
+      "step": 2546
+    },
+    {
+      "epoch": 0.11132479566414616,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.699402084057086e-05,
+      "loss": 1.8776,
+      "step": 2547
+    },
+    {
+      "epoch": 0.11136850386817605,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.699167472040964e-05,
+      "loss": 2.8036,
+      "step": 2548
+    },
+    {
+      "epoch": 0.11141221207220595,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.698932771344715e-05,
+      "loss": 2.2721,
+      "step": 2549
+    },
+    {
+      "epoch": 0.11145592027623585,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.698697981972763e-05,
+      "loss": 1.9911,
+      "step": 2550
+    },
+    {
+      "epoch": 0.11149962848026575,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.698463103929542e-05,
+      "loss": 2.914,
+      "step": 2551
+    },
+    {
+      "epoch": 0.11154333668429564,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.698228137219485e-05,
+      "loss": 2.3899,
+      "step": 2552
+    },
+    {
+      "epoch": 0.11158704488832553,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.697993081847024e-05,
+      "loss": 2.4362,
+      "step": 2553
+    },
+    {
+      "epoch": 0.11163075309235544,
+      "grad_norm": 2.375,
+      "learning_rate": 9.697757937816596e-05,
+      "loss": 2.0962,
+      "step": 2554
+    },
+    {
+      "epoch": 0.11167446129638534,
+      "grad_norm": 11.1875,
+      "learning_rate": 9.69752270513264e-05,
+      "loss": 2.4154,
+      "step": 2555
+    },
+    {
+      "epoch": 0.11171816950041523,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.697287383799592e-05,
+      "loss": 1.3796,
+      "step": 2556
+    },
+    {
+      "epoch": 0.11176187770444512,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.697051973821895e-05,
+      "loss": 1.9024,
+      "step": 2557
+    },
+    {
+      "epoch": 0.11180558590847502,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.696816475203992e-05,
+      "loss": 2.4041,
+      "step": 2558
+    },
+    {
+      "epoch": 0.11184929411250492,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.696580887950324e-05,
+      "loss": 2.0977,
+      "step": 2559
+    },
+    {
+      "epoch": 0.11189300231653482,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.69634521206534e-05,
+      "loss": 2.501,
+      "step": 2560
+    },
+    {
+      "epoch": 0.11193671052056471,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.696109447553488e-05,
+      "loss": 1.9393,
+      "step": 2561
+    },
+    {
+      "epoch": 0.1119804187245946,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.695873594419213e-05,
+      "loss": 1.7693,
+      "step": 2562
+    },
+    {
+      "epoch": 0.1120241269286245,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.695637652666972e-05,
+      "loss": 2.4633,
+      "step": 2563
+    },
+    {
+      "epoch": 0.1120678351326544,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.695401622301212e-05,
+      "loss": 1.7753,
+      "step": 2564
+    },
+    {
+      "epoch": 0.1121115433366843,
+      "grad_norm": 3.75,
+      "learning_rate": 9.69516550332639e-05,
+      "loss": 3.1476,
+      "step": 2565
+    },
+    {
+      "epoch": 0.1121552515407142,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.694929295746963e-05,
+      "loss": 2.3226,
+      "step": 2566
+    },
+    {
+      "epoch": 0.11219895974474409,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.694692999567386e-05,
+      "loss": 2.6497,
+      "step": 2567
+    },
+    {
+      "epoch": 0.11224266794877398,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.694456614792119e-05,
+      "loss": 1.9439,
+      "step": 2568
+    },
+    {
+      "epoch": 0.11228637615280389,
+      "grad_norm": 2.5,
+      "learning_rate": 9.694220141425623e-05,
+      "loss": 2.1219,
+      "step": 2569
+    },
+    {
+      "epoch": 0.11233008435683378,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.69398357947236e-05,
+      "loss": 1.9452,
+      "step": 2570
+    },
+    {
+      "epoch": 0.11237379256086368,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.693746928936798e-05,
+      "loss": 1.7742,
+      "step": 2571
+    },
+    {
+      "epoch": 0.11241750076489357,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.693510189823398e-05,
+      "loss": 2.0195,
+      "step": 2572
+    },
+    {
+      "epoch": 0.11246120896892346,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.69327336213663e-05,
+      "loss": 2.213,
+      "step": 2573
+    },
+    {
+      "epoch": 0.11250491717295337,
+      "grad_norm": 3.125,
+      "learning_rate": 9.693036445880963e-05,
+      "loss": 1.6811,
+      "step": 2574
+    },
+    {
+      "epoch": 0.11254862537698326,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.692799441060868e-05,
+      "loss": 2.2884,
+      "step": 2575
+    },
+    {
+      "epoch": 0.11259233358101316,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.692562347680817e-05,
+      "loss": 2.2061,
+      "step": 2576
+    },
+    {
+      "epoch": 0.11263604178504305,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.692325165745285e-05,
+      "loss": 1.8484,
+      "step": 2577
+    },
+    {
+      "epoch": 0.11267974998907294,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.692087895258748e-05,
+      "loss": 2.6623,
+      "step": 2578
+    },
+    {
+      "epoch": 0.11272345819310285,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.691850536225684e-05,
+      "loss": 2.2441,
+      "step": 2579
+    },
+    {
+      "epoch": 0.11276716639713275,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.691613088650571e-05,
+      "loss": 1.9522,
+      "step": 2580
+    },
+    {
+      "epoch": 0.11281087460116264,
+      "grad_norm": 2.75,
+      "learning_rate": 9.69137555253789e-05,
+      "loss": 2.5621,
+      "step": 2581
+    },
+    {
+      "epoch": 0.11285458280519253,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.691137927892125e-05,
+      "loss": 2.6895,
+      "step": 2582
+    },
+    {
+      "epoch": 0.11289829100922243,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.69090021471776e-05,
+      "loss": 1.9568,
+      "step": 2583
+    },
+    {
+      "epoch": 0.11294199921325233,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.69066241301928e-05,
+      "loss": 1.9114,
+      "step": 2584
+    },
+    {
+      "epoch": 0.11298570741728223,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.690424522801173e-05,
+      "loss": 2.1984,
+      "step": 2585
+    },
+    {
+      "epoch": 0.11302941562131212,
+      "grad_norm": 2.25,
+      "learning_rate": 9.69018654406793e-05,
+      "loss": 2.0646,
+      "step": 2586
+    },
+    {
+      "epoch": 0.11307312382534201,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.68994847682404e-05,
+      "loss": 1.6406,
+      "step": 2587
+    },
+    {
+      "epoch": 0.11311683202937191,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.689710321073997e-05,
+      "loss": 2.687,
+      "step": 2588
+    },
+    {
+      "epoch": 0.11316054023340182,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.689472076822295e-05,
+      "loss": 1.8021,
+      "step": 2589
+    },
+    {
+      "epoch": 0.11320424843743171,
+      "grad_norm": 2.875,
+      "learning_rate": 9.689233744073427e-05,
+      "loss": 2.06,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1132479566414616,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.688995322831895e-05,
+      "loss": 2.283,
+      "step": 2591
+    },
+    {
+      "epoch": 0.1132916648454915,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.688756813102197e-05,
+      "loss": 2.036,
+      "step": 2592
+    },
+    {
+      "epoch": 0.11333537304952139,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.688518214888836e-05,
+      "loss": 2.0899,
+      "step": 2593
+    },
+    {
+      "epoch": 0.1133790812535513,
+      "grad_norm": 2.5,
+      "learning_rate": 9.688279528196309e-05,
+      "loss": 2.0154,
+      "step": 2594
+    },
+    {
+      "epoch": 0.11342278945758119,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.688040753029125e-05,
+      "loss": 1.6289,
+      "step": 2595
+    },
+    {
+      "epoch": 0.11346649766161108,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.687801889391789e-05,
+      "loss": 2.3163,
+      "step": 2596
+    },
+    {
+      "epoch": 0.11351020586564098,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.687562937288807e-05,
+      "loss": 2.118,
+      "step": 2597
+    },
+    {
+      "epoch": 0.11355391406967087,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.687323896724693e-05,
+      "loss": 2.3258,
+      "step": 2598
+    },
+    {
+      "epoch": 0.11359762227370078,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.687084767703954e-05,
+      "loss": 2.6265,
+      "step": 2599
+    },
+    {
+      "epoch": 0.11364133047773067,
+      "grad_norm": 2.875,
+      "learning_rate": 9.686845550231102e-05,
+      "loss": 2.2793,
+      "step": 2600
+    },
+    {
+      "epoch": 0.11368503868176057,
+      "grad_norm": 2.5,
+      "learning_rate": 9.686606244310654e-05,
+      "loss": 1.9895,
+      "step": 2601
+    },
+    {
+      "epoch": 0.11372874688579046,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.686366849947126e-05,
+      "loss": 2.4364,
+      "step": 2602
+    },
+    {
+      "epoch": 0.11377245508982035,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.686127367145034e-05,
+      "loss": 2.4419,
+      "step": 2603
+    },
+    {
+      "epoch": 0.11381616329385026,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.685887795908899e-05,
+      "loss": 1.8057,
+      "step": 2604
+    },
+    {
+      "epoch": 0.11385987149788016,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.68564813624324e-05,
+      "loss": 1.5621,
+      "step": 2605
+    },
+    {
+      "epoch": 0.11390357970191005,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.685408388152581e-05,
+      "loss": 1.7561,
+      "step": 2606
+    },
+    {
+      "epoch": 0.11394728790593994,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.685168551641448e-05,
+      "loss": 2.0413,
+      "step": 2607
+    },
+    {
+      "epoch": 0.11399099610996984,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.684928626714365e-05,
+      "loss": 1.4987,
+      "step": 2608
+    },
+    {
+      "epoch": 0.11403470431399974,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.68468861337586e-05,
+      "loss": 2.4914,
+      "step": 2609
+    },
+    {
+      "epoch": 0.11407841251802964,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.684448511630461e-05,
+      "loss": 2.1296,
+      "step": 2610
+    },
+    {
+      "epoch": 0.11412212072205953,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.684208321482704e-05,
+      "loss": 1.8615,
+      "step": 2611
+    },
+    {
+      "epoch": 0.11416582892608942,
+      "grad_norm": 2.875,
+      "learning_rate": 9.683968042937117e-05,
+      "loss": 2.1629,
+      "step": 2612
+    },
+    {
+      "epoch": 0.11420953713011932,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.683727675998236e-05,
+      "loss": 1.9825,
+      "step": 2613
+    },
+    {
+      "epoch": 0.11425324533414923,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.683487220670595e-05,
+      "loss": 2.6845,
+      "step": 2614
+    },
+    {
+      "epoch": 0.11429695353817912,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.683246676958735e-05,
+      "loss": 2.0385,
+      "step": 2615
+    },
+    {
+      "epoch": 0.11434066174220901,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.683006044867194e-05,
+      "loss": 2.3356,
+      "step": 2616
+    },
+    {
+      "epoch": 0.1143843699462389,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.682765324400514e-05,
+      "loss": 1.81,
+      "step": 2617
+    },
+    {
+      "epoch": 0.1144280781502688,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.682524515563236e-05,
+      "loss": 2.5127,
+      "step": 2618
+    },
+    {
+      "epoch": 0.11447178635429871,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.682283618359905e-05,
+      "loss": 2.0217,
+      "step": 2619
+    },
+    {
+      "epoch": 0.1145154945583286,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.682042632795067e-05,
+      "loss": 1.9385,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1145592027623585,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.681801558873272e-05,
+      "loss": 2.1403,
+      "step": 2621
+    },
+    {
+      "epoch": 0.11460291096638839,
+      "grad_norm": 3.0,
+      "learning_rate": 9.681560396599068e-05,
+      "loss": 1.5651,
+      "step": 2622
+    },
+    {
+      "epoch": 0.11464661917041828,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.681319145977003e-05,
+      "loss": 2.7997,
+      "step": 2623
+    },
+    {
+      "epoch": 0.11469032737444819,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.681077807011634e-05,
+      "loss": 2.129,
+      "step": 2624
+    },
+    {
+      "epoch": 0.11473403557847808,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.680836379707513e-05,
+      "loss": 2.0857,
+      "step": 2625
+    },
+    {
+      "epoch": 0.11477774378250798,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.680594864069197e-05,
+      "loss": 1.9064,
+      "step": 2626
+    },
+    {
+      "epoch": 0.11482145198653787,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.680353260101245e-05,
+      "loss": 1.6006,
+      "step": 2627
+    },
+    {
+      "epoch": 0.11486516019056776,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.680111567808213e-05,
+      "loss": 1.6344,
+      "step": 2628
+    },
+    {
+      "epoch": 0.11490886839459767,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.679869787194664e-05,
+      "loss": 2.1607,
+      "step": 2629
+    },
+    {
+      "epoch": 0.11495257659862756,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.679627918265163e-05,
+      "loss": 2.1579,
+      "step": 2630
+    },
+    {
+      "epoch": 0.11499628480265746,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.679385961024271e-05,
+      "loss": 2.589,
+      "step": 2631
+    },
+    {
+      "epoch": 0.11503999300668735,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.679143915476556e-05,
+      "loss": 3.4489,
+      "step": 2632
+    },
+    {
+      "epoch": 0.11508370121071725,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.678901781626584e-05,
+      "loss": 1.725,
+      "step": 2633
+    },
+    {
+      "epoch": 0.11512740941474715,
+      "grad_norm": 5.25,
+      "learning_rate": 9.678659559478926e-05,
+      "loss": 2.3209,
+      "step": 2634
+    },
+    {
+      "epoch": 0.11517111761877705,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.678417249038154e-05,
+      "loss": 2.0092,
+      "step": 2635
+    },
+    {
+      "epoch": 0.11521482582280694,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.678174850308839e-05,
+      "loss": 2.1935,
+      "step": 2636
+    },
+    {
+      "epoch": 0.11525853402683683,
+      "grad_norm": 3.25,
+      "learning_rate": 9.677932363295555e-05,
+      "loss": 2.3839,
+      "step": 2637
+    },
+    {
+      "epoch": 0.11530224223086673,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.677689788002879e-05,
+      "loss": 2.3866,
+      "step": 2638
+    },
+    {
+      "epoch": 0.11534595043489664,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.677447124435389e-05,
+      "loss": 1.7951,
+      "step": 2639
+    },
+    {
+      "epoch": 0.11538965863892653,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.677204372597663e-05,
+      "loss": 2.4154,
+      "step": 2640
+    },
+    {
+      "epoch": 0.11543336684295642,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.676961532494284e-05,
+      "loss": 2.1352,
+      "step": 2641
+    },
+    {
+      "epoch": 0.11547707504698632,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.676718604129832e-05,
+      "loss": 2.3192,
+      "step": 2642
+    },
+    {
+      "epoch": 0.11552078325101621,
+      "grad_norm": 2.125,
+      "learning_rate": 9.676475587508897e-05,
+      "loss": 2.0761,
+      "step": 2643
+    },
+    {
+      "epoch": 0.11556449145504612,
+      "grad_norm": 2.375,
+      "learning_rate": 9.67623248263606e-05,
+      "loss": 1.6954,
+      "step": 2644
+    },
+    {
+      "epoch": 0.11560819965907601,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.675989289515908e-05,
+      "loss": 2.0882,
+      "step": 2645
+    },
+    {
+      "epoch": 0.1156519078631059,
+      "grad_norm": 3.25,
+      "learning_rate": 9.675746008153035e-05,
+      "loss": 2.324,
+      "step": 2646
+    },
+    {
+      "epoch": 0.1156956160671358,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.675502638552029e-05,
+      "loss": 2.7426,
+      "step": 2647
+    },
+    {
+      "epoch": 0.11573932427116569,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.675259180717482e-05,
+      "loss": 2.0731,
+      "step": 2648
+    },
+    {
+      "epoch": 0.1157830324751956,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.675015634653992e-05,
+      "loss": 2.5737,
+      "step": 2649
+    },
+    {
+      "epoch": 0.11582674067922549,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.674772000366151e-05,
+      "loss": 2.3047,
+      "step": 2650
+    },
+    {
+      "epoch": 0.11587044888325539,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.674528277858559e-05,
+      "loss": 1.3378,
+      "step": 2651
+    },
+    {
+      "epoch": 0.11591415708728528,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.674284467135816e-05,
+      "loss": 2.2169,
+      "step": 2652
+    },
+    {
+      "epoch": 0.11595786529131517,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.67404056820252e-05,
+      "loss": 2.0575,
+      "step": 2653
+    },
+    {
+      "epoch": 0.11600157349534508,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.673796581063278e-05,
+      "loss": 2.0379,
+      "step": 2654
+    },
+    {
+      "epoch": 0.11604528169937497,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.67355250572269e-05,
+      "loss": 1.7116,
+      "step": 2655
+    },
+    {
+      "epoch": 0.11608898990340487,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.673308342185365e-05,
+      "loss": 1.9168,
+      "step": 2656
+    },
+    {
+      "epoch": 0.11613269810743476,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.673064090455911e-05,
+      "loss": 2.8693,
+      "step": 2657
+    },
+    {
+      "epoch": 0.11617640631146466,
+      "grad_norm": 2.375,
+      "learning_rate": 9.672819750538935e-05,
+      "loss": 2.0332,
+      "step": 2658
+    },
+    {
+      "epoch": 0.11622011451549456,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.67257532243905e-05,
+      "loss": 2.4218,
+      "step": 2659
+    },
+    {
+      "epoch": 0.11626382271952446,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.672330806160868e-05,
+      "loss": 1.8975,
+      "step": 2660
+    },
+    {
+      "epoch": 0.11630753092355435,
+      "grad_norm": 3.0,
+      "learning_rate": 9.672086201709003e-05,
+      "loss": 2.4923,
+      "step": 2661
+    },
+    {
+      "epoch": 0.11635123912758424,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.671841509088073e-05,
+      "loss": 2.2724,
+      "step": 2662
+    },
+    {
+      "epoch": 0.11639494733161414,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.671596728302692e-05,
+      "loss": 2.2291,
+      "step": 2663
+    },
+    {
+      "epoch": 0.11643865553564405,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.671351859357483e-05,
+      "loss": 2.263,
+      "step": 2664
+    },
+    {
+      "epoch": 0.11648236373967394,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.671106902257065e-05,
+      "loss": 2.7241,
+      "step": 2665
+    },
+    {
+      "epoch": 0.11652607194370383,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.670861857006061e-05,
+      "loss": 2.1171,
+      "step": 2666
+    },
+    {
+      "epoch": 0.11656978014773373,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.670616723609096e-05,
+      "loss": 2.266,
+      "step": 2667
+    },
+    {
+      "epoch": 0.11661348835176362,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.670371502070795e-05,
+      "loss": 1.9497,
+      "step": 2668
+    },
+    {
+      "epoch": 0.11665719655579353,
+      "grad_norm": 2.875,
+      "learning_rate": 9.670126192395787e-05,
+      "loss": 2.612,
+      "step": 2669
+    },
+    {
+      "epoch": 0.11670090475982342,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.669880794588701e-05,
+      "loss": 2.1332,
+      "step": 2670
+    },
+    {
+      "epoch": 0.11674461296385331,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.669635308654166e-05,
+      "loss": 2.334,
+      "step": 2671
+    },
+    {
+      "epoch": 0.11678832116788321,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.669389734596819e-05,
+      "loss": 2.0459,
+      "step": 2672
+    },
+    {
+      "epoch": 0.1168320293719131,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.66914407242129e-05,
+      "loss": 2.9586,
+      "step": 2673
+    },
+    {
+      "epoch": 0.11687573757594301,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.668898322132218e-05,
+      "loss": 1.893,
+      "step": 2674
+    },
+    {
+      "epoch": 0.1169194457799729,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.668652483734237e-05,
+      "loss": 2.3674,
+      "step": 2675
+    },
+    {
+      "epoch": 0.1169631539840028,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.668406557231991e-05,
+      "loss": 2.0931,
+      "step": 2676
+    },
+    {
+      "epoch": 0.11700686218803269,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.668160542630118e-05,
+      "loss": 1.92,
+      "step": 2677
+    },
+    {
+      "epoch": 0.11705057039206258,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.667914439933262e-05,
+      "loss": 1.6033,
+      "step": 2678
+    },
+    {
+      "epoch": 0.11709427859609249,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.667668249146067e-05,
+      "loss": 2.6922,
+      "step": 2679
+    },
+    {
+      "epoch": 0.11713798680012238,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.667421970273177e-05,
+      "loss": 1.8012,
+      "step": 2680
+    },
+    {
+      "epoch": 0.11718169500415228,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.667175603319243e-05,
+      "loss": 2.2011,
+      "step": 2681
+    },
+    {
+      "epoch": 0.11722540320818217,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.66692914828891e-05,
+      "loss": 2.2007,
+      "step": 2682
+    },
+    {
+      "epoch": 0.11726911141221207,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.666682605186835e-05,
+      "loss": 2.1641,
+      "step": 2683
+    },
+    {
+      "epoch": 0.11731281961624197,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.666435974017665e-05,
+      "loss": 2.4392,
+      "step": 2684
+    },
+    {
+      "epoch": 0.11735652782027187,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.666189254786056e-05,
+      "loss": 1.8101,
+      "step": 2685
+    },
+    {
+      "epoch": 0.11740023602430176,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.665942447496666e-05,
+      "loss": 2.6438,
+      "step": 2686
+    },
+    {
+      "epoch": 0.11744394422833165,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.66569555215415e-05,
+      "loss": 2.9357,
+      "step": 2687
+    },
+    {
+      "epoch": 0.11748765243236156,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.665448568763169e-05,
+      "loss": 2.0782,
+      "step": 2688
+    },
+    {
+      "epoch": 0.11753136063639145,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.665201497328384e-05,
+      "loss": 2.1359,
+      "step": 2689
+    },
+    {
+      "epoch": 0.11757506884042135,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.664954337854455e-05,
+      "loss": 1.7294,
+      "step": 2690
+    },
+    {
+      "epoch": 0.11761877704445124,
+      "grad_norm": 2.625,
+      "learning_rate": 9.66470709034605e-05,
+      "loss": 2.2099,
+      "step": 2691
+    },
+    {
+      "epoch": 0.11766248524848114,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.664459754807832e-05,
+      "loss": 3.201,
+      "step": 2692
+    },
+    {
+      "epoch": 0.11770619345251104,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.66421233124447e-05,
+      "loss": 2.0315,
+      "step": 2693
+    },
+    {
+      "epoch": 0.11774990165654094,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.663964819660633e-05,
+      "loss": 1.792,
+      "step": 2694
+    },
+    {
+      "epoch": 0.11779360986057083,
+      "grad_norm": 2.75,
+      "learning_rate": 9.663717220060991e-05,
+      "loss": 2.191,
+      "step": 2695
+    },
+    {
+      "epoch": 0.11783731806460072,
+      "grad_norm": 2.5,
+      "learning_rate": 9.663469532450218e-05,
+      "loss": 2.0232,
+      "step": 2696
+    },
+    {
+      "epoch": 0.11788102626863062,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.663221756832988e-05,
+      "loss": 1.938,
+      "step": 2697
+    },
+    {
+      "epoch": 0.11792473447266053,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.662973893213976e-05,
+      "loss": 2.7432,
+      "step": 2698
+    },
+    {
+      "epoch": 0.11796844267669042,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.662725941597859e-05,
+      "loss": 1.9639,
+      "step": 2699
+    },
+    {
+      "epoch": 0.11801215088072031,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.662477901989318e-05,
+      "loss": 1.9061,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1180558590847502,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.662229774393032e-05,
+      "loss": 2.5382,
+      "step": 2701
+    },
+    {
+      "epoch": 0.1180995672887801,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.661981558813687e-05,
+      "loss": 2.3339,
+      "step": 2702
+    },
+    {
+      "epoch": 0.11814327549281001,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.661733255255963e-05,
+      "loss": 1.7774,
+      "step": 2703
+    },
+    {
+      "epoch": 0.1181869836968399,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.661484863724549e-05,
+      "loss": 1.8731,
+      "step": 2704
+    },
+    {
+      "epoch": 0.1182306919008698,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.661236384224129e-05,
+      "loss": 2.4805,
+      "step": 2705
+    },
+    {
+      "epoch": 0.11827440010489969,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.660987816759396e-05,
+      "loss": 1.7,
+      "step": 2706
+    },
+    {
+      "epoch": 0.11831810830892958,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.66073916133504e-05,
+      "loss": 2.4671,
+      "step": 2707
+    },
+    {
+      "epoch": 0.11836181651295949,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.660490417955749e-05,
+      "loss": 2.5451,
+      "step": 2708
+    },
+    {
+      "epoch": 0.11840552471698938,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.660241586626224e-05,
+      "loss": 2.3707,
+      "step": 2709
+    },
+    {
+      "epoch": 0.11844923292101928,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.659992667351157e-05,
+      "loss": 2.45,
+      "step": 2710
+    },
+    {
+      "epoch": 0.11849294112504917,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.659743660135245e-05,
+      "loss": 2.5622,
+      "step": 2711
+    },
+    {
+      "epoch": 0.11853664932907906,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.659494564983191e-05,
+      "loss": 2.4197,
+      "step": 2712
+    },
+    {
+      "epoch": 0.11858035753310897,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.659245381899691e-05,
+      "loss": 2.1187,
+      "step": 2713
+    },
+    {
+      "epoch": 0.11862406573713886,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.658996110889449e-05,
+      "loss": 2.2612,
+      "step": 2714
+    },
+    {
+      "epoch": 0.11866777394116876,
+      "grad_norm": 2.25,
+      "learning_rate": 9.65874675195717e-05,
+      "loss": 1.9677,
+      "step": 2715
+    },
+    {
+      "epoch": 0.11871148214519865,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.658497305107558e-05,
+      "loss": 1.9515,
+      "step": 2716
+    },
+    {
+      "epoch": 0.11875519034922855,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.658247770345323e-05,
+      "loss": 1.9262,
+      "step": 2717
+    },
+    {
+      "epoch": 0.11879889855325845,
+      "grad_norm": 2.875,
+      "learning_rate": 9.657998147675173e-05,
+      "loss": 2.469,
+      "step": 2718
+    },
+    {
+      "epoch": 0.11884260675728835,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.65774843710182e-05,
+      "loss": 2.1998,
+      "step": 2719
+    },
+    {
+      "epoch": 0.11888631496131824,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.657498638629972e-05,
+      "loss": 1.9848,
+      "step": 2720
+    },
+    {
+      "epoch": 0.11893002316534813,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.657248752264348e-05,
+      "loss": 2.4081,
+      "step": 2721
+    },
+    {
+      "epoch": 0.11897373136937803,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.656998778009661e-05,
+      "loss": 1.5549,
+      "step": 2722
+    },
+    {
+      "epoch": 0.11901743957340793,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.656748715870629e-05,
+      "loss": 1.8315,
+      "step": 2723
+    },
+    {
+      "epoch": 0.11906114777743783,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.656498565851972e-05,
+      "loss": 1.9119,
+      "step": 2724
+    },
+    {
+      "epoch": 0.11910485598146772,
+      "grad_norm": 2.625,
+      "learning_rate": 9.656248327958409e-05,
+      "loss": 1.7329,
+      "step": 2725
+    },
+    {
+      "epoch": 0.11914856418549762,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.655998002194663e-05,
+      "loss": 2.5823,
+      "step": 2726
+    },
+    {
+      "epoch": 0.11919227238952751,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.65574758856546e-05,
+      "loss": 1.5248,
+      "step": 2727
+    },
+    {
+      "epoch": 0.11923598059355742,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.655497087075522e-05,
+      "loss": 2.3173,
+      "step": 2728
+    },
+    {
+      "epoch": 0.11927968879758731,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.655246497729578e-05,
+      "loss": 1.7577,
+      "step": 2729
+    },
+    {
+      "epoch": 0.1193233970016172,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.654995820532359e-05,
+      "loss": 1.9014,
+      "step": 2730
+    },
+    {
+      "epoch": 0.1193671052056471,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.654745055488592e-05,
+      "loss": 1.8937,
+      "step": 2731
+    },
+    {
+      "epoch": 0.11941081340967699,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.654494202603013e-05,
+      "loss": 1.7926,
+      "step": 2732
+    },
+    {
+      "epoch": 0.1194545216137069,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.654243261880353e-05,
+      "loss": 1.5646,
+      "step": 2733
+    },
+    {
+      "epoch": 0.11949822981773679,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.653992233325348e-05,
+      "loss": 3.3764,
+      "step": 2734
+    },
+    {
+      "epoch": 0.11954193802176669,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.653741116942738e-05,
+      "loss": 2.5024,
+      "step": 2735
+    },
+    {
+      "epoch": 0.11958564622579658,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.653489912737258e-05,
+      "loss": 2.5474,
+      "step": 2736
+    },
+    {
+      "epoch": 0.11962935442982647,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.653238620713652e-05,
+      "loss": 2.585,
+      "step": 2737
+    },
+    {
+      "epoch": 0.11967306263385638,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.652987240876659e-05,
+      "loss": 2.2581,
+      "step": 2738
+    },
+    {
+      "epoch": 0.11971677083788627,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.652735773231025e-05,
+      "loss": 2.341,
+      "step": 2739
+    },
+    {
+      "epoch": 0.11976047904191617,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.652484217781497e-05,
+      "loss": 2.4574,
+      "step": 2740
+    },
+    {
+      "epoch": 0.11980418724594606,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.65223257453282e-05,
+      "loss": 2.1156,
+      "step": 2741
+    },
+    {
+      "epoch": 0.11984789544997596,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.651980843489742e-05,
+      "loss": 1.846,
+      "step": 2742
+    },
+    {
+      "epoch": 0.11989160365400586,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.651729024657014e-05,
+      "loss": 2.0689,
+      "step": 2743
+    },
+    {
+      "epoch": 0.11993531185803576,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.651477118039391e-05,
+      "loss": 1.9795,
+      "step": 2744
+    },
+    {
+      "epoch": 0.11997902006206565,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.651225123641625e-05,
+      "loss": 2.5617,
+      "step": 2745
+    },
+    {
+      "epoch": 0.12002272826609554,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.65097304146847e-05,
+      "loss": 2.7336,
+      "step": 2746
+    },
+    {
+      "epoch": 0.12006643647012544,
+      "grad_norm": 2.875,
+      "learning_rate": 9.650720871524686e-05,
+      "loss": 1.7494,
+      "step": 2747
+    },
+    {
+      "epoch": 0.12011014467415534,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.65046861381503e-05,
+      "loss": 1.8763,
+      "step": 2748
+    },
+    {
+      "epoch": 0.12015385287818524,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.650216268344262e-05,
+      "loss": 1.9197,
+      "step": 2749
+    },
+    {
+      "epoch": 0.12019756108221513,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.649963835117147e-05,
+      "loss": 2.1772,
+      "step": 2750
+    },
+    {
+      "epoch": 0.12024126928624503,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.649711314138443e-05,
+      "loss": 1.8606,
+      "step": 2751
+    },
+    {
+      "epoch": 0.12028497749027492,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.649458705412921e-05,
+      "loss": 1.6259,
+      "step": 2752
+    },
+    {
+      "epoch": 0.12032868569430483,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.649206008945348e-05,
+      "loss": 1.7527,
+      "step": 2753
+    },
+    {
+      "epoch": 0.12037239389833472,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.648953224740489e-05,
+      "loss": 3.3193,
+      "step": 2754
+    },
+    {
+      "epoch": 0.12041610210236461,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.648700352803118e-05,
+      "loss": 2.3352,
+      "step": 2755
+    },
+    {
+      "epoch": 0.12045981030639451,
+      "grad_norm": 2.875,
+      "learning_rate": 9.648447393138005e-05,
+      "loss": 2.2762,
+      "step": 2756
+    },
+    {
+      "epoch": 0.1205035185104244,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.648194345749923e-05,
+      "loss": 2.3559,
+      "step": 2757
+    },
+    {
+      "epoch": 0.12054722671445431,
+      "grad_norm": 2.625,
+      "learning_rate": 9.64794121064365e-05,
+      "loss": 2.6152,
+      "step": 2758
+    },
+    {
+      "epoch": 0.1205909349184842,
+      "grad_norm": 2.25,
+      "learning_rate": 9.64768798782396e-05,
+      "loss": 1.7972,
+      "step": 2759
+    },
+    {
+      "epoch": 0.1206346431225141,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.647434677295635e-05,
+      "loss": 1.8208,
+      "step": 2760
+    },
+    {
+      "epoch": 0.12067835132654399,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.647181279063453e-05,
+      "loss": 2.6227,
+      "step": 2761
+    },
+    {
+      "epoch": 0.12072205953057388,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.646927793132195e-05,
+      "loss": 2.2009,
+      "step": 2762
+    },
+    {
+      "epoch": 0.12076576773460379,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.646674219506648e-05,
+      "loss": 2.5652,
+      "step": 2763
+    },
+    {
+      "epoch": 0.12080947593863368,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.646420558191596e-05,
+      "loss": 2.5864,
+      "step": 2764
+    },
+    {
+      "epoch": 0.12085318414266358,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.646166809191824e-05,
+      "loss": 2.542,
+      "step": 2765
+    },
+    {
+      "epoch": 0.12089689234669347,
+      "grad_norm": 2.375,
+      "learning_rate": 9.645912972512124e-05,
+      "loss": 1.6604,
+      "step": 2766
+    },
+    {
+      "epoch": 0.12094060055072337,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.645659048157282e-05,
+      "loss": 1.5808,
+      "step": 2767
+    },
+    {
+      "epoch": 0.12098430875475327,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.645405036132093e-05,
+      "loss": 1.7339,
+      "step": 2768
+    },
+    {
+      "epoch": 0.12102801695878317,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.645150936441351e-05,
+      "loss": 2.326,
+      "step": 2769
+    },
+    {
+      "epoch": 0.12107172516281306,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.64489674908985e-05,
+      "loss": 1.763,
+      "step": 2770
+    },
+    {
+      "epoch": 0.12111543336684295,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.644642474082386e-05,
+      "loss": 1.7941,
+      "step": 2771
+    },
+    {
+      "epoch": 0.12115914157087285,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.644388111423759e-05,
+      "loss": 1.76,
+      "step": 2772
+    },
+    {
+      "epoch": 0.12120284977490275,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.644133661118769e-05,
+      "loss": 2.1463,
+      "step": 2773
+    },
+    {
+      "epoch": 0.12124655797893265,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.643879123172218e-05,
+      "loss": 1.7819,
+      "step": 2774
+    },
+    {
+      "epoch": 0.12129026618296254,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.643624497588908e-05,
+      "loss": 2.1013,
+      "step": 2775
+    },
+    {
+      "epoch": 0.12133397438699244,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.643369784373645e-05,
+      "loss": 2.0237,
+      "step": 2776
+    },
+    {
+      "epoch": 0.12137768259102233,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.643114983531238e-05,
+      "loss": 1.6875,
+      "step": 2777
+    },
+    {
+      "epoch": 0.12142139079505224,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.642860095066493e-05,
+      "loss": 2.4743,
+      "step": 2778
+    },
+    {
+      "epoch": 0.12146509899908213,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.64260511898422e-05,
+      "loss": 2.577,
+      "step": 2779
+    },
+    {
+      "epoch": 0.12150880720311202,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.642350055289232e-05,
+      "loss": 2.5573,
+      "step": 2780
+    },
+    {
+      "epoch": 0.12155251540714192,
+      "grad_norm": 2.25,
+      "learning_rate": 9.642094903986341e-05,
+      "loss": 1.871,
+      "step": 2781
+    },
+    {
+      "epoch": 0.12159622361117181,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.641839665080363e-05,
+      "loss": 2.1364,
+      "step": 2782
+    },
+    {
+      "epoch": 0.12163993181520172,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.641584338576115e-05,
+      "loss": 1.9107,
+      "step": 2783
+    },
+    {
+      "epoch": 0.12168364001923161,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.641328924478416e-05,
+      "loss": 1.7113,
+      "step": 2784
+    },
+    {
+      "epoch": 0.1217273482232615,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.641073422792085e-05,
+      "loss": 1.885,
+      "step": 2785
+    },
+    {
+      "epoch": 0.1217710564272914,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.640817833521941e-05,
+      "loss": 2.057,
+      "step": 2786
+    },
+    {
+      "epoch": 0.12181476463132129,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.640562156672812e-05,
+      "loss": 2.3443,
+      "step": 2787
+    },
+    {
+      "epoch": 0.1218584728353512,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.64030639224952e-05,
+      "loss": 1.3211,
+      "step": 2788
+    },
+    {
+      "epoch": 0.1219021810393811,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.640050540256896e-05,
+      "loss": 1.985,
+      "step": 2789
+    },
+    {
+      "epoch": 0.12194588924341099,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.639794600699761e-05,
+      "loss": 1.9536,
+      "step": 2790
+    },
+    {
+      "epoch": 0.12198959744744088,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.639538573582951e-05,
+      "loss": 1.9697,
+      "step": 2791
+    },
+    {
+      "epoch": 0.12203330565147077,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.639282458911297e-05,
+      "loss": 1.9432,
+      "step": 2792
+    },
+    {
+      "epoch": 0.12207701385550068,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.639026256689628e-05,
+      "loss": 1.8395,
+      "step": 2793
+    },
+    {
+      "epoch": 0.12212072205953058,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.638769966922783e-05,
+      "loss": 2.3836,
+      "step": 2794
+    },
+    {
+      "epoch": 0.12216443026356047,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.638513589615596e-05,
+      "loss": 2.228,
+      "step": 2795
+    },
+    {
+      "epoch": 0.12220813846759036,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.638257124772909e-05,
+      "loss": 2.281,
+      "step": 2796
+    },
+    {
+      "epoch": 0.12225184667162026,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.638000572399559e-05,
+      "loss": 2.0127,
+      "step": 2797
+    },
+    {
+      "epoch": 0.12229555487565016,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.637743932500385e-05,
+      "loss": 2.5281,
+      "step": 2798
+    },
+    {
+      "epoch": 0.12233926307968006,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.637487205080236e-05,
+      "loss": 2.0413,
+      "step": 2799
+    },
+    {
+      "epoch": 0.12238297128370995,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.637230390143953e-05,
+      "loss": 2.1581,
+      "step": 2800
+    },
+    {
+      "epoch": 0.12242667948773985,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.636973487696383e-05,
+      "loss": 2.5069,
+      "step": 2801
+    },
+    {
+      "epoch": 0.12247038769176974,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.636716497742375e-05,
+      "loss": 1.5927,
+      "step": 2802
+    },
+    {
+      "epoch": 0.12251409589579965,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.636459420286779e-05,
+      "loss": 1.7826,
+      "step": 2803
+    },
+    {
+      "epoch": 0.12255780409982954,
+      "grad_norm": 3.71875,
+      "learning_rate": 9.636202255334444e-05,
+      "loss": 2.4786,
+      "step": 2804
+    },
+    {
+      "epoch": 0.12260151230385943,
+      "grad_norm": 2.625,
+      "learning_rate": 9.635945002890225e-05,
+      "loss": 2.1374,
+      "step": 2805
+    },
+    {
+      "epoch": 0.12264522050788933,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.635687662958978e-05,
+      "loss": 1.6501,
+      "step": 2806
+    },
+    {
+      "epoch": 0.12268892871191922,
+      "grad_norm": 2.625,
+      "learning_rate": 9.635430235545557e-05,
+      "loss": 2.1622,
+      "step": 2807
+    },
+    {
+      "epoch": 0.12273263691594913,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.635172720654822e-05,
+      "loss": 1.9891,
+      "step": 2808
+    },
+    {
+      "epoch": 0.12277634511997902,
+      "grad_norm": 2.375,
+      "learning_rate": 9.634915118291629e-05,
+      "loss": 1.9462,
+      "step": 2809
+    },
+    {
+      "epoch": 0.12282005332400892,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.634657428460844e-05,
+      "loss": 2.5531,
+      "step": 2810
+    },
+    {
+      "epoch": 0.12286376152803881,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.634399651167328e-05,
+      "loss": 1.9049,
+      "step": 2811
+    },
+    {
+      "epoch": 0.1229074697320687,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.634141786415944e-05,
+      "loss": 2.4264,
+      "step": 2812
+    },
+    {
+      "epoch": 0.12295117793609861,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.633883834211562e-05,
+      "loss": 2.2442,
+      "step": 2813
+    },
+    {
+      "epoch": 0.1229948861401285,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.633625794559045e-05,
+      "loss": 2.1397,
+      "step": 2814
+    },
+    {
+      "epoch": 0.1230385943441584,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.633367667463267e-05,
+      "loss": 2.8998,
+      "step": 2815
+    },
+    {
+      "epoch": 0.12308230254818829,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.633109452929097e-05,
+      "loss": 1.9054,
+      "step": 2816
+    },
+    {
+      "epoch": 0.12312601075221818,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.632851150961409e-05,
+      "loss": 1.9339,
+      "step": 2817
+    },
+    {
+      "epoch": 0.12316971895624809,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.632592761565077e-05,
+      "loss": 1.8701,
+      "step": 2818
+    },
+    {
+      "epoch": 0.12321342716027799,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.632334284744978e-05,
+      "loss": 2.0784,
+      "step": 2819
+    },
+    {
+      "epoch": 0.12325713536430788,
+      "grad_norm": 7.4375,
+      "learning_rate": 9.632075720505987e-05,
+      "loss": 1.8226,
+      "step": 2820
+    },
+    {
+      "epoch": 0.12330084356833777,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.631817068852986e-05,
+      "loss": 1.7425,
+      "step": 2821
+    },
+    {
+      "epoch": 0.12334455177236767,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.631558329790857e-05,
+      "loss": 1.9114,
+      "step": 2822
+    },
+    {
+      "epoch": 0.12338825997639757,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.63129950332448e-05,
+      "loss": 2.1165,
+      "step": 2823
+    },
+    {
+      "epoch": 0.12343196818042747,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.631040589458741e-05,
+      "loss": 1.8682,
+      "step": 2824
+    },
+    {
+      "epoch": 0.12347567638445736,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.630781588198526e-05,
+      "loss": 1.7066,
+      "step": 2825
+    },
+    {
+      "epoch": 0.12351938458848725,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.630522499548723e-05,
+      "loss": 2.5323,
+      "step": 2826
+    },
+    {
+      "epoch": 0.12356309279251716,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.630263323514222e-05,
+      "loss": 2.2879,
+      "step": 2827
+    },
+    {
+      "epoch": 0.12360680099654706,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.630004060099911e-05,
+      "loss": 2.0478,
+      "step": 2828
+    },
+    {
+      "epoch": 0.12365050920057695,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.629744709310686e-05,
+      "loss": 1.6938,
+      "step": 2829
+    },
+    {
+      "epoch": 0.12369421740460684,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.629485271151439e-05,
+      "loss": 1.9106,
+      "step": 2830
+    },
+    {
+      "epoch": 0.12373792560863674,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.629225745627069e-05,
+      "loss": 2.8505,
+      "step": 2831
+    },
+    {
+      "epoch": 0.12378163381266664,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.628966132742469e-05,
+      "loss": 2.4035,
+      "step": 2832
+    },
+    {
+      "epoch": 0.12382534201669654,
+      "grad_norm": 2.75,
+      "learning_rate": 9.628706432502543e-05,
+      "loss": 1.9975,
+      "step": 2833
+    },
+    {
+      "epoch": 0.12386905022072643,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.62844664491219e-05,
+      "loss": 1.9244,
+      "step": 2834
+    },
+    {
+      "epoch": 0.12391275842475633,
+      "grad_norm": 2.25,
+      "learning_rate": 9.628186769976311e-05,
+      "loss": 1.9185,
+      "step": 2835
+    },
+    {
+      "epoch": 0.12395646662878622,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.627926807699812e-05,
+      "loss": 1.9349,
+      "step": 2836
+    },
+    {
+      "epoch": 0.12400017483281613,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.6276667580876e-05,
+      "loss": 1.671,
+      "step": 2837
+    },
+    {
+      "epoch": 0.12404388303684602,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.627406621144578e-05,
+      "loss": 1.6789,
+      "step": 2838
+    },
+    {
+      "epoch": 0.12408759124087591,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.62714639687566e-05,
+      "loss": 1.9697,
+      "step": 2839
+    },
+    {
+      "epoch": 0.12413129944490581,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.626886085285755e-05,
+      "loss": 2.0127,
+      "step": 2840
+    },
+    {
+      "epoch": 0.1241750076489357,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.626625686379776e-05,
+      "loss": 1.8073,
+      "step": 2841
+    },
+    {
+      "epoch": 0.12421871585296561,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.626365200162636e-05,
+      "loss": 1.6166,
+      "step": 2842
+    },
+    {
+      "epoch": 0.1242624240569955,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.626104626639252e-05,
+      "loss": 1.9251,
+      "step": 2843
+    },
+    {
+      "epoch": 0.1243061322610254,
+      "grad_norm": 3.875,
+      "learning_rate": 9.625843965814539e-05,
+      "loss": 2.5869,
+      "step": 2844
+    },
+    {
+      "epoch": 0.12434984046505529,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.62558321769342e-05,
+      "loss": 3.0161,
+      "step": 2845
+    },
+    {
+      "epoch": 0.12439354866908518,
+      "grad_norm": 2.375,
+      "learning_rate": 9.625322382280811e-05,
+      "loss": 1.901,
+      "step": 2846
+    },
+    {
+      "epoch": 0.12443725687311509,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.625061459581639e-05,
+      "loss": 1.9044,
+      "step": 2847
+    },
+    {
+      "epoch": 0.12448096507714498,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.624800449600825e-05,
+      "loss": 2.3202,
+      "step": 2848
+    },
+    {
+      "epoch": 0.12452467328117488,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.624539352343295e-05,
+      "loss": 2.5673,
+      "step": 2849
+    },
+    {
+      "epoch": 0.12456838148520477,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.624278167813977e-05,
+      "loss": 2.3917,
+      "step": 2850
+    },
+    {
+      "epoch": 0.12461208968923466,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.6240168960178e-05,
+      "loss": 3.0122,
+      "step": 2851
+    },
+    {
+      "epoch": 0.12465579789326457,
+      "grad_norm": 2.125,
+      "learning_rate": 9.623755536959693e-05,
+      "loss": 1.8228,
+      "step": 2852
+    },
+    {
+      "epoch": 0.12469950609729447,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.623494090644591e-05,
+      "loss": 3.0652,
+      "step": 2853
+    },
+    {
+      "epoch": 0.12474321430132436,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.623232557077426e-05,
+      "loss": 2.1147,
+      "step": 2854
+    },
+    {
+      "epoch": 0.12478692250535425,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.622970936263134e-05,
+      "loss": 2.8506,
+      "step": 2855
+    },
+    {
+      "epoch": 0.12483063070938415,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.622709228206651e-05,
+      "loss": 2.3544,
+      "step": 2856
+    },
+    {
+      "epoch": 0.12487433891341405,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.622447432912918e-05,
+      "loss": 1.9846,
+      "step": 2857
+    },
+    {
+      "epoch": 0.12491804711744395,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.622185550386873e-05,
+      "loss": 1.9084,
+      "step": 2858
+    },
+    {
+      "epoch": 0.12496175532147384,
+      "grad_norm": 2.25,
+      "learning_rate": 9.62192358063346e-05,
+      "loss": 1.8372,
+      "step": 2859
+    },
+    {
+      "epoch": 0.12500546352550373,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.621661523657623e-05,
+      "loss": 2.4741,
+      "step": 2860
+    },
+    {
+      "epoch": 0.12504917172953364,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.621399379464306e-05,
+      "loss": 2.4071,
+      "step": 2861
+    },
+    {
+      "epoch": 0.12509287993356352,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.621137148058457e-05,
+      "loss": 2.3149,
+      "step": 2862
+    },
+    {
+      "epoch": 0.12513658813759343,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.620874829445023e-05,
+      "loss": 2.3921,
+      "step": 2863
+    },
+    {
+      "epoch": 0.1251802963416233,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.620612423628956e-05,
+      "loss": 1.9683,
+      "step": 2864
+    },
+    {
+      "epoch": 0.12522400454565322,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.620349930615207e-05,
+      "loss": 2.6925,
+      "step": 2865
+    },
+    {
+      "epoch": 0.12526771274968312,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.620087350408732e-05,
+      "loss": 2.2013,
+      "step": 2866
+    },
+    {
+      "epoch": 0.125311420953713,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.619824683014484e-05,
+      "loss": 2.3714,
+      "step": 2867
+    },
+    {
+      "epoch": 0.1253551291577429,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.61956192843742e-05,
+      "loss": 1.9963,
+      "step": 2868
+    },
+    {
+      "epoch": 0.1253988373617728,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.619299086682498e-05,
+      "loss": 1.7972,
+      "step": 2869
+    },
+    {
+      "epoch": 0.1254425455658027,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.61903615775468e-05,
+      "loss": 2.12,
+      "step": 2870
+    },
+    {
+      "epoch": 0.1254862537698326,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.618773141658927e-05,
+      "loss": 2.1438,
+      "step": 2871
+    },
+    {
+      "epoch": 0.12552996197386249,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.618510038400203e-05,
+      "loss": 2.0048,
+      "step": 2872
+    },
+    {
+      "epoch": 0.1255736701778924,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.618246847983471e-05,
+      "loss": 2.315,
+      "step": 2873
+    },
+    {
+      "epoch": 0.12561737838192227,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.617983570413702e-05,
+      "loss": 1.9862,
+      "step": 2874
+    },
+    {
+      "epoch": 0.12566108658595218,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.617720205695862e-05,
+      "loss": 1.7937,
+      "step": 2875
+    },
+    {
+      "epoch": 0.1257047947899821,
+      "grad_norm": 2.75,
+      "learning_rate": 9.617456753834919e-05,
+      "loss": 2.5493,
+      "step": 2876
+    },
+    {
+      "epoch": 0.12574850299401197,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.617193214835847e-05,
+      "loss": 2.4217,
+      "step": 2877
+    },
+    {
+      "epoch": 0.12579221119804188,
+      "grad_norm": 8.375,
+      "learning_rate": 9.616929588703618e-05,
+      "loss": 2.2522,
+      "step": 2878
+    },
+    {
+      "epoch": 0.12583591940207176,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.61666587544321e-05,
+      "loss": 1.9273,
+      "step": 2879
+    },
+    {
+      "epoch": 0.12587962760610166,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.616402075059597e-05,
+      "loss": 2.637,
+      "step": 2880
+    },
+    {
+      "epoch": 0.12592333581013157,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.616138187557758e-05,
+      "loss": 2.0555,
+      "step": 2881
+    },
+    {
+      "epoch": 0.12596704401416145,
+      "grad_norm": 3.953125,
+      "learning_rate": 9.615874212942673e-05,
+      "loss": 2.0235,
+      "step": 2882
+    },
+    {
+      "epoch": 0.12601075221819136,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.615610151219323e-05,
+      "loss": 2.136,
+      "step": 2883
+    },
+    {
+      "epoch": 0.12605446042222124,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.615346002392692e-05,
+      "loss": 2.1353,
+      "step": 2884
+    },
+    {
+      "epoch": 0.12609816862625114,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.615081766467764e-05,
+      "loss": 2.4363,
+      "step": 2885
+    },
+    {
+      "epoch": 0.12614187683028105,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.614817443449529e-05,
+      "loss": 1.9039,
+      "step": 2886
+    },
+    {
+      "epoch": 0.12618558503431093,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.614553033342969e-05,
+      "loss": 2.5831,
+      "step": 2887
+    },
+    {
+      "epoch": 0.12622929323834084,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.614288536153078e-05,
+      "loss": 2.0155,
+      "step": 2888
+    },
+    {
+      "epoch": 0.12627300144237072,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.614023951884848e-05,
+      "loss": 1.7467,
+      "step": 2889
+    },
+    {
+      "epoch": 0.12631670964640063,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.613759280543269e-05,
+      "loss": 1.8073,
+      "step": 2890
+    },
+    {
+      "epoch": 0.12636041785043053,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.613494522133337e-05,
+      "loss": 3.2485,
+      "step": 2891
+    },
+    {
+      "epoch": 0.12640412605446041,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.613229676660049e-05,
+      "loss": 2.328,
+      "step": 2892
+    },
+    {
+      "epoch": 0.12644783425849032,
+      "grad_norm": 2.0,
+      "learning_rate": 9.612964744128404e-05,
+      "loss": 1.8794,
+      "step": 2893
+    },
+    {
+      "epoch": 0.1264915424625202,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.6126997245434e-05,
+      "loss": 2.0111,
+      "step": 2894
+    },
+    {
+      "epoch": 0.1265352506665501,
+      "grad_norm": 2.25,
+      "learning_rate": 9.612434617910038e-05,
+      "loss": 2.0805,
+      "step": 2895
+    },
+    {
+      "epoch": 0.12657895887058002,
+      "grad_norm": 2.375,
+      "learning_rate": 9.612169424233323e-05,
+      "loss": 2.1638,
+      "step": 2896
+    },
+    {
+      "epoch": 0.1266226670746099,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.611904143518257e-05,
+      "loss": 2.546,
+      "step": 2897
+    },
+    {
+      "epoch": 0.1266663752786398,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.611638775769848e-05,
+      "loss": 2.0348,
+      "step": 2898
+    },
+    {
+      "epoch": 0.1267100834826697,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.611373320993104e-05,
+      "loss": 1.8323,
+      "step": 2899
+    },
+    {
+      "epoch": 0.1267537916866996,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.611107779193033e-05,
+      "loss": 1.6366,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1267974998907295,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.610842150374647e-05,
+      "loss": 1.679,
+      "step": 2901
+    },
+    {
+      "epoch": 0.12684120809475938,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.610576434542959e-05,
+      "loss": 1.9237,
+      "step": 2902
+    },
+    {
+      "epoch": 0.12688491629878929,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.610310631702983e-05,
+      "loss": 2.8621,
+      "step": 2903
+    },
+    {
+      "epoch": 0.1269286245028192,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.610044741859736e-05,
+      "loss": 2.1277,
+      "step": 2904
+    },
+    {
+      "epoch": 0.12697233270684907,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.609778765018235e-05,
+      "loss": 1.9442,
+      "step": 2905
+    },
+    {
+      "epoch": 0.12701604091087898,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.609512701183499e-05,
+      "loss": 2.3559,
+      "step": 2906
+    },
+    {
+      "epoch": 0.12705974911490886,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.609246550360551e-05,
+      "loss": 2.4651,
+      "step": 2907
+    },
+    {
+      "epoch": 0.12710345731893877,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.60898031255441e-05,
+      "loss": 2.1018,
+      "step": 2908
+    },
+    {
+      "epoch": 0.12714716552296867,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.608713987770103e-05,
+      "loss": 1.988,
+      "step": 2909
+    },
+    {
+      "epoch": 0.12719087372699855,
+      "grad_norm": 2.375,
+      "learning_rate": 9.608447576012656e-05,
+      "loss": 2.4612,
+      "step": 2910
+    },
+    {
+      "epoch": 0.12723458193102846,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.608181077287098e-05,
+      "loss": 2.0709,
+      "step": 2911
+    },
+    {
+      "epoch": 0.12727829013505834,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.607914491598453e-05,
+      "loss": 1.7109,
+      "step": 2912
+    },
+    {
+      "epoch": 0.12732199833908825,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.607647818951756e-05,
+      "loss": 2.0656,
+      "step": 2913
+    },
+    {
+      "epoch": 0.12736570654311816,
+      "grad_norm": 3.125,
+      "learning_rate": 9.607381059352038e-05,
+      "loss": 3.0216,
+      "step": 2914
+    },
+    {
+      "epoch": 0.12740941474714804,
+      "grad_norm": 2.25,
+      "learning_rate": 9.607114212804335e-05,
+      "loss": 1.9264,
+      "step": 2915
+    },
+    {
+      "epoch": 0.12745312295117794,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.606847279313681e-05,
+      "loss": 2.1365,
+      "step": 2916
+    },
+    {
+      "epoch": 0.12749683115520782,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.606580258885114e-05,
+      "loss": 1.7999,
+      "step": 2917
+    },
+    {
+      "epoch": 0.12754053935923773,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.606313151523672e-05,
+      "loss": 1.9917,
+      "step": 2918
+    },
+    {
+      "epoch": 0.12758424756326764,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.606045957234398e-05,
+      "loss": 2.5768,
+      "step": 2919
+    },
+    {
+      "epoch": 0.12762795576729752,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.605778676022333e-05,
+      "loss": 1.913,
+      "step": 2920
+    },
+    {
+      "epoch": 0.12767166397132743,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.605511307892519e-05,
+      "loss": 2.1415,
+      "step": 2921
+    },
+    {
+      "epoch": 0.1277153721753573,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.605243852850006e-05,
+      "loss": 1.9214,
+      "step": 2922
+    },
+    {
+      "epoch": 0.1277590803793872,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.604976310899837e-05,
+      "loss": 2.1233,
+      "step": 2923
+    },
+    {
+      "epoch": 0.12780278858341712,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.604708682047064e-05,
+      "loss": 2.0758,
+      "step": 2924
+    },
+    {
+      "epoch": 0.127846496787447,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.604440966296734e-05,
+      "loss": 2.0466,
+      "step": 2925
+    },
+    {
+      "epoch": 0.1278902049914769,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.604173163653904e-05,
+      "loss": 2.3425,
+      "step": 2926
+    },
+    {
+      "epoch": 0.1279339131955068,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.603905274123626e-05,
+      "loss": 2.4204,
+      "step": 2927
+    },
+    {
+      "epoch": 0.1279776213995367,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.603637297710954e-05,
+      "loss": 2.0726,
+      "step": 2928
+    },
+    {
+      "epoch": 0.1280213296035666,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.603369234420945e-05,
+      "loss": 1.755,
+      "step": 2929
+    },
+    {
+      "epoch": 0.12806503780759648,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.603101084258658e-05,
+      "loss": 2.0132,
+      "step": 2930
+    },
+    {
+      "epoch": 0.1281087460116264,
+      "grad_norm": 2.5,
+      "learning_rate": 9.602832847229156e-05,
+      "loss": 1.9451,
+      "step": 2931
+    },
+    {
+      "epoch": 0.12815245421565627,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.602564523337498e-05,
+      "loss": 1.7434,
+      "step": 2932
+    },
+    {
+      "epoch": 0.12819616241968618,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.602296112588749e-05,
+      "loss": 2.1285,
+      "step": 2933
+    },
+    {
+      "epoch": 0.12823987062371608,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.602027614987974e-05,
+      "loss": 2.3753,
+      "step": 2934
+    },
+    {
+      "epoch": 0.12828357882774596,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.60175903054024e-05,
+      "loss": 1.763,
+      "step": 2935
+    },
+    {
+      "epoch": 0.12832728703177587,
+      "grad_norm": 2.125,
+      "learning_rate": 9.601490359250615e-05,
+      "loss": 1.7454,
+      "step": 2936
+    },
+    {
+      "epoch": 0.12837099523580575,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.601221601124172e-05,
+      "loss": 1.8542,
+      "step": 2937
+    },
+    {
+      "epoch": 0.12841470343983566,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.600952756165979e-05,
+      "loss": 2.1361,
+      "step": 2938
+    },
+    {
+      "epoch": 0.12845841164386557,
+      "grad_norm": 2.75,
+      "learning_rate": 9.600683824381112e-05,
+      "loss": 1.8827,
+      "step": 2939
+    },
+    {
+      "epoch": 0.12850211984789545,
+      "grad_norm": 4.40625,
+      "learning_rate": 9.600414805774643e-05,
+      "loss": 2.8128,
+      "step": 2940
+    },
+    {
+      "epoch": 0.12854582805192535,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.600145700351652e-05,
+      "loss": 1.7991,
+      "step": 2941
+    },
+    {
+      "epoch": 0.12858953625595523,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.599876508117219e-05,
+      "loss": 2.6316,
+      "step": 2942
+    },
+    {
+      "epoch": 0.12863324445998514,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.599607229076418e-05,
+      "loss": 1.8649,
+      "step": 2943
+    },
+    {
+      "epoch": 0.12867695266401505,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.599337863234335e-05,
+      "loss": 2.396,
+      "step": 2944
+    },
+    {
+      "epoch": 0.12872066086804493,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.599068410596053e-05,
+      "loss": 2.2869,
+      "step": 2945
+    },
+    {
+      "epoch": 0.12876436907207484,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.598798871166656e-05,
+      "loss": 2.4572,
+      "step": 2946
+    },
+    {
+      "epoch": 0.12880807727610472,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.598529244951232e-05,
+      "loss": 2.1541,
+      "step": 2947
+    },
+    {
+      "epoch": 0.12885178548013462,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.598259531954868e-05,
+      "loss": 2.0215,
+      "step": 2948
+    },
+    {
+      "epoch": 0.12889549368416453,
+      "grad_norm": 2.625,
+      "learning_rate": 9.597989732182654e-05,
+      "loss": 2.2977,
+      "step": 2949
+    },
+    {
+      "epoch": 0.1289392018881944,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.597719845639682e-05,
+      "loss": 1.8592,
+      "step": 2950
+    },
+    {
+      "epoch": 0.12898291009222432,
+      "grad_norm": 2.375,
+      "learning_rate": 9.597449872331045e-05,
+      "loss": 2.1149,
+      "step": 2951
+    },
+    {
+      "epoch": 0.1290266182962542,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.597179812261836e-05,
+      "loss": 1.9048,
+      "step": 2952
+    },
+    {
+      "epoch": 0.1290703265002841,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.596909665437155e-05,
+      "loss": 2.1005,
+      "step": 2953
+    },
+    {
+      "epoch": 0.129114034704314,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.596639431862098e-05,
+      "loss": 1.4811,
+      "step": 2954
+    },
+    {
+      "epoch": 0.1291577429083439,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.596369111541764e-05,
+      "loss": 2.0181,
+      "step": 2955
+    },
+    {
+      "epoch": 0.1292014511123738,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.596098704481255e-05,
+      "loss": 2.2414,
+      "step": 2956
+    },
+    {
+      "epoch": 0.12924515931640368,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.595828210685675e-05,
+      "loss": 1.9929,
+      "step": 2957
+    },
+    {
+      "epoch": 0.1292888675204336,
+      "grad_norm": 2.375,
+      "learning_rate": 9.595557630160127e-05,
+      "loss": 2.1108,
+      "step": 2958
+    },
+    {
+      "epoch": 0.1293325757244635,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.595286962909717e-05,
+      "loss": 1.8243,
+      "step": 2959
+    },
+    {
+      "epoch": 0.12937628392849337,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.595016208939555e-05,
+      "loss": 2.0417,
+      "step": 2960
+    },
+    {
+      "epoch": 0.12941999213252328,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.594745368254751e-05,
+      "loss": 2.0429,
+      "step": 2961
+    },
+    {
+      "epoch": 0.12946370033655316,
+      "grad_norm": 4.0,
+      "learning_rate": 9.594474440860412e-05,
+      "loss": 2.2023,
+      "step": 2962
+    },
+    {
+      "epoch": 0.12950740854058307,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.594203426761656e-05,
+      "loss": 2.6341,
+      "step": 2963
+    },
+    {
+      "epoch": 0.12955111674461298,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.593932325963593e-05,
+      "loss": 2.112,
+      "step": 2964
+    },
+    {
+      "epoch": 0.12959482494864286,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.593661138471342e-05,
+      "loss": 1.9946,
+      "step": 2965
+    },
+    {
+      "epoch": 0.12963853315267276,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.593389864290018e-05,
+      "loss": 2.502,
+      "step": 2966
+    },
+    {
+      "epoch": 0.12968224135670264,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.593118503424743e-05,
+      "loss": 2.1195,
+      "step": 2967
+    },
+    {
+      "epoch": 0.12972594956073255,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.592847055880636e-05,
+      "loss": 2.0636,
+      "step": 2968
+    },
+    {
+      "epoch": 0.12976965776476246,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.592575521662821e-05,
+      "loss": 2.1656,
+      "step": 2969
+    },
+    {
+      "epoch": 0.12981336596879234,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.592303900776422e-05,
+      "loss": 2.2107,
+      "step": 2970
+    },
+    {
+      "epoch": 0.12985707417282225,
+      "grad_norm": 3.25,
+      "learning_rate": 9.592032193226564e-05,
+      "loss": 2.9833,
+      "step": 2971
+    },
+    {
+      "epoch": 0.12990078237685213,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.591760399018375e-05,
+      "loss": 2.4552,
+      "step": 2972
+    },
+    {
+      "epoch": 0.12994449058088203,
+      "grad_norm": 2.25,
+      "learning_rate": 9.591488518156985e-05,
+      "loss": 2.3015,
+      "step": 2973
+    },
+    {
+      "epoch": 0.12998819878491194,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.591216550647524e-05,
+      "loss": 1.8688,
+      "step": 2974
+    },
+    {
+      "epoch": 0.13003190698894182,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.590944496495124e-05,
+      "loss": 1.8787,
+      "step": 2975
+    },
+    {
+      "epoch": 0.13007561519297173,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.59067235570492e-05,
+      "loss": 2.0342,
+      "step": 2976
+    },
+    {
+      "epoch": 0.1301193233970016,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.590400128282047e-05,
+      "loss": 2.1873,
+      "step": 2977
+    },
+    {
+      "epoch": 0.13016303160103151,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.590127814231642e-05,
+      "loss": 3.4453,
+      "step": 2978
+    },
+    {
+      "epoch": 0.13020673980506142,
+      "grad_norm": 2.125,
+      "learning_rate": 9.589855413558846e-05,
+      "loss": 1.825,
+      "step": 2979
+    },
+    {
+      "epoch": 0.1302504480090913,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.589582926268798e-05,
+      "loss": 1.9386,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1302941562131212,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.589310352366639e-05,
+      "loss": 2.2409,
+      "step": 2981
+    },
+    {
+      "epoch": 0.1303378644171511,
+      "grad_norm": 2.875,
+      "learning_rate": 9.589037691857515e-05,
+      "loss": 2.7937,
+      "step": 2982
+    },
+    {
+      "epoch": 0.130381572621181,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.588764944746571e-05,
+      "loss": 2.0267,
+      "step": 2983
+    },
+    {
+      "epoch": 0.1304252808252109,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.588492111038953e-05,
+      "loss": 2.0755,
+      "step": 2984
+    },
+    {
+      "epoch": 0.13046898902924078,
+      "grad_norm": 2.875,
+      "learning_rate": 9.588219190739811e-05,
+      "loss": 2.3942,
+      "step": 2985
+    },
+    {
+      "epoch": 0.1305126972332707,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.587946183854295e-05,
+      "loss": 1.7443,
+      "step": 2986
+    },
+    {
+      "epoch": 0.13055640543730057,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.587673090387558e-05,
+      "loss": 1.7173,
+      "step": 2987
+    },
+    {
+      "epoch": 0.13060011364133048,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.587399910344753e-05,
+      "loss": 2.3119,
+      "step": 2988
+    },
+    {
+      "epoch": 0.13064382184536039,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.587126643731033e-05,
+      "loss": 1.8181,
+      "step": 2989
+    },
+    {
+      "epoch": 0.13068753004939027,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.586853290551558e-05,
+      "loss": 2.0235,
+      "step": 2990
+    },
+    {
+      "epoch": 0.13073123825342017,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.586579850811486e-05,
+      "loss": 1.7663,
+      "step": 2991
+    },
+    {
+      "epoch": 0.13077494645745005,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.586306324515976e-05,
+      "loss": 2.0949,
+      "step": 2992
+    },
+    {
+      "epoch": 0.13081865466147996,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.58603271167019e-05,
+      "loss": 2.186,
+      "step": 2993
+    },
+    {
+      "epoch": 0.13086236286550987,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.585759012279294e-05,
+      "loss": 1.7462,
+      "step": 2994
+    },
+    {
+      "epoch": 0.13090607106953975,
+      "grad_norm": 3.984375,
+      "learning_rate": 9.58548522634845e-05,
+      "loss": 2.6174,
+      "step": 2995
+    },
+    {
+      "epoch": 0.13094977927356966,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.585211353882826e-05,
+      "loss": 2.5182,
+      "step": 2996
+    },
+    {
+      "epoch": 0.13099348747759954,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.58493739488759e-05,
+      "loss": 1.7231,
+      "step": 2997
+    },
+    {
+      "epoch": 0.13103719568162944,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.584663349367912e-05,
+      "loss": 2.4883,
+      "step": 2998
+    },
+    {
+      "epoch": 0.13108090388565935,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.584389217328966e-05,
+      "loss": 1.7723,
+      "step": 2999
+    },
+    {
+      "epoch": 0.13112461208968923,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.584114998775921e-05,
+      "loss": 1.8153,
+      "step": 3000
+    },
+    {
+      "epoch": 0.13116832029371914,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.583840693713954e-05,
+      "loss": 2.0988,
+      "step": 3001
+    },
+    {
+      "epoch": 0.13121202849774902,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.583566302148244e-05,
+      "loss": 2.2539,
+      "step": 3002
+    },
+    {
+      "epoch": 0.13125573670177892,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.583291824083965e-05,
+      "loss": 1.6684,
+      "step": 3003
+    },
+    {
+      "epoch": 0.13129944490580883,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.583017259526299e-05,
+      "loss": 2.04,
+      "step": 3004
+    },
+    {
+      "epoch": 0.1313431531098387,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.582742608480428e-05,
+      "loss": 2.2517,
+      "step": 3005
+    },
+    {
+      "epoch": 0.13138686131386862,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.582467870951533e-05,
+      "loss": 1.9857,
+      "step": 3006
+    },
+    {
+      "epoch": 0.1314305695178985,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.5821930469448e-05,
+      "loss": 2.4797,
+      "step": 3007
+    },
+    {
+      "epoch": 0.1314742777219284,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.581918136465416e-05,
+      "loss": 2.4298,
+      "step": 3008
+    },
+    {
+      "epoch": 0.13151798592595831,
+      "grad_norm": 2.375,
+      "learning_rate": 9.581643139518565e-05,
+      "loss": 1.7604,
+      "step": 3009
+    },
+    {
+      "epoch": 0.1315616941299882,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.581368056109443e-05,
+      "loss": 2.3003,
+      "step": 3010
+    },
+    {
+      "epoch": 0.1316054023340181,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.581092886243237e-05,
+      "loss": 1.9386,
+      "step": 3011
+    },
+    {
+      "epoch": 0.13164911053804798,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.58081762992514e-05,
+      "loss": 2.2748,
+      "step": 3012
+    },
+    {
+      "epoch": 0.1316928187420779,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.580542287160348e-05,
+      "loss": 2.9969,
+      "step": 3013
+    },
+    {
+      "epoch": 0.1317365269461078,
+      "grad_norm": 2.625,
+      "learning_rate": 9.580266857954057e-05,
+      "loss": 2.2068,
+      "step": 3014
+    },
+    {
+      "epoch": 0.13178023515013768,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.579991342311463e-05,
+      "loss": 1.8651,
+      "step": 3015
+    },
+    {
+      "epoch": 0.13182394335416758,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.579715740237766e-05,
+      "loss": 1.7948,
+      "step": 3016
+    },
+    {
+      "epoch": 0.13186765155819746,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.579440051738168e-05,
+      "loss": 2.0184,
+      "step": 3017
+    },
+    {
+      "epoch": 0.13191135976222737,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.579164276817873e-05,
+      "loss": 2.3655,
+      "step": 3018
+    },
+    {
+      "epoch": 0.13195506796625728,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.578888415482082e-05,
+      "loss": 1.9905,
+      "step": 3019
+    },
+    {
+      "epoch": 0.13199877617028716,
+      "grad_norm": 3.65625,
+      "learning_rate": 9.578612467736004e-05,
+      "loss": 2.846,
+      "step": 3020
+    },
+    {
+      "epoch": 0.13204248437431707,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.578336433584842e-05,
+      "loss": 2.0329,
+      "step": 3021
+    },
+    {
+      "epoch": 0.13208619257834694,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.57806031303381e-05,
+      "loss": 2.0556,
+      "step": 3022
+    },
+    {
+      "epoch": 0.13212990078237685,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.577784106088115e-05,
+      "loss": 2.2177,
+      "step": 3023
+    },
+    {
+      "epoch": 0.13217360898640676,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.577507812752972e-05,
+      "loss": 1.7264,
+      "step": 3024
+    },
+    {
+      "epoch": 0.13221731719043664,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.577231433033596e-05,
+      "loss": 1.7551,
+      "step": 3025
+    },
+    {
+      "epoch": 0.13226102539446655,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.576954966935198e-05,
+      "loss": 1.812,
+      "step": 3026
+    },
+    {
+      "epoch": 0.13230473359849643,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.576678414463001e-05,
+      "loss": 3.2237,
+      "step": 3027
+    },
+    {
+      "epoch": 0.13234844180252633,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.576401775622222e-05,
+      "loss": 1.7513,
+      "step": 3028
+    },
+    {
+      "epoch": 0.13239215000655624,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.576125050418077e-05,
+      "loss": 2.0373,
+      "step": 3029
+    },
+    {
+      "epoch": 0.13243585821058612,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.575848238855796e-05,
+      "loss": 2.1212,
+      "step": 3030
+    },
+    {
+      "epoch": 0.13247956641461603,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.575571340940597e-05,
+      "loss": 1.8877,
+      "step": 3031
+    },
+    {
+      "epoch": 0.1325232746186459,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.575294356677707e-05,
+      "loss": 2.2786,
+      "step": 3032
+    },
+    {
+      "epoch": 0.13256698282267582,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.575017286072355e-05,
+      "loss": 1.922,
+      "step": 3033
+    },
+    {
+      "epoch": 0.13261069102670572,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.574740129129767e-05,
+      "loss": 2.1338,
+      "step": 3034
+    },
+    {
+      "epoch": 0.1326543992307356,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.574462885855174e-05,
+      "loss": 2.4,
+      "step": 3035
+    },
+    {
+      "epoch": 0.1326981074347655,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.57418555625381e-05,
+      "loss": 2.177,
+      "step": 3036
+    },
+    {
+      "epoch": 0.1327418156387954,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.573908140330905e-05,
+      "loss": 2.1712,
+      "step": 3037
+    },
+    {
+      "epoch": 0.1327855238428253,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.573630638091698e-05,
+      "loss": 1.9705,
+      "step": 3038
+    },
+    {
+      "epoch": 0.1328292320468552,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.573353049541425e-05,
+      "loss": 1.9445,
+      "step": 3039
+    },
+    {
+      "epoch": 0.13287294025088509,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.573075374685323e-05,
+      "loss": 2.0247,
+      "step": 3040
+    },
+    {
+      "epoch": 0.132916648454915,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.572797613528633e-05,
+      "loss": 1.9356,
+      "step": 3041
+    },
+    {
+      "epoch": 0.13296035665894487,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.572519766076595e-05,
+      "loss": 2.2053,
+      "step": 3042
+    },
+    {
+      "epoch": 0.13300406486297478,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.572241832334457e-05,
+      "loss": 2.8439,
+      "step": 3043
+    },
+    {
+      "epoch": 0.1330477730670047,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.571963812307459e-05,
+      "loss": 2.3801,
+      "step": 3044
+    },
+    {
+      "epoch": 0.13309148127103457,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.57168570600085e-05,
+      "loss": 2.0442,
+      "step": 3045
+    },
+    {
+      "epoch": 0.13313518947506447,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.571407513419877e-05,
+      "loss": 1.8136,
+      "step": 3046
+    },
+    {
+      "epoch": 0.13317889767909435,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.571129234569792e-05,
+      "loss": 2.2411,
+      "step": 3047
+    },
+    {
+      "epoch": 0.13322260588312426,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.570850869455845e-05,
+      "loss": 2.2694,
+      "step": 3048
+    },
+    {
+      "epoch": 0.13326631408715417,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.57057241808329e-05,
+      "loss": 1.9991,
+      "step": 3049
+    },
+    {
+      "epoch": 0.13331002229118405,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.570293880457382e-05,
+      "loss": 2.7739,
+      "step": 3050
+    },
+    {
+      "epoch": 0.13335373049521396,
+      "grad_norm": 2.875,
+      "learning_rate": 9.570015256583375e-05,
+      "loss": 2.8135,
+      "step": 3051
+    },
+    {
+      "epoch": 0.13339743869924384,
+      "grad_norm": 2.5,
+      "learning_rate": 9.56973654646653e-05,
+      "loss": 1.9088,
+      "step": 3052
+    },
+    {
+      "epoch": 0.13344114690327374,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.569457750112106e-05,
+      "loss": 2.6223,
+      "step": 3053
+    },
+    {
+      "epoch": 0.13348485510730365,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.569178867525362e-05,
+      "loss": 2.0422,
+      "step": 3054
+    },
+    {
+      "epoch": 0.13352856331133353,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.568899898711563e-05,
+      "loss": 2.0285,
+      "step": 3055
+    },
+    {
+      "epoch": 0.13357227151536344,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.568620843675975e-05,
+      "loss": 2.091,
+      "step": 3056
+    },
+    {
+      "epoch": 0.13361597971939332,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.56834170242386e-05,
+      "loss": 2.0307,
+      "step": 3057
+    },
+    {
+      "epoch": 0.13365968792342323,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.568062474960489e-05,
+      "loss": 2.0899,
+      "step": 3058
+    },
+    {
+      "epoch": 0.13370339612745313,
+      "grad_norm": 2.625,
+      "learning_rate": 9.56778316129113e-05,
+      "loss": 1.8432,
+      "step": 3059
+    },
+    {
+      "epoch": 0.133747104331483,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.567503761421057e-05,
+      "loss": 1.4322,
+      "step": 3060
+    },
+    {
+      "epoch": 0.13379081253551292,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.567224275355539e-05,
+      "loss": 2.1823,
+      "step": 3061
+    },
+    {
+      "epoch": 0.1338345207395428,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.566944703099852e-05,
+      "loss": 1.718,
+      "step": 3062
+    },
+    {
+      "epoch": 0.1338782289435727,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.56666504465927e-05,
+      "loss": 2.4914,
+      "step": 3063
+    },
+    {
+      "epoch": 0.13392193714760262,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.566385300039074e-05,
+      "loss": 2.0368,
+      "step": 3064
+    },
+    {
+      "epoch": 0.1339656453516325,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.56610546924454e-05,
+      "loss": 1.9711,
+      "step": 3065
+    },
+    {
+      "epoch": 0.1340093535556624,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.56582555228095e-05,
+      "loss": 2.2761,
+      "step": 3066
+    },
+    {
+      "epoch": 0.13405306175969228,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.565545549153588e-05,
+      "loss": 1.7822,
+      "step": 3067
+    },
+    {
+      "epoch": 0.1340967699637222,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.565265459867736e-05,
+      "loss": 2.0829,
+      "step": 3068
+    },
+    {
+      "epoch": 0.1341404781677521,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.564985284428679e-05,
+      "loss": 1.8644,
+      "step": 3069
+    },
+    {
+      "epoch": 0.13418418637178198,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.564705022841706e-05,
+      "loss": 1.8726,
+      "step": 3070
+    },
+    {
+      "epoch": 0.13422789457581188,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.564424675112106e-05,
+      "loss": 2.0443,
+      "step": 3071
+    },
+    {
+      "epoch": 0.13427160277984176,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.564144241245173e-05,
+      "loss": 2.0608,
+      "step": 3072
+    },
+    {
+      "epoch": 0.13431531098387167,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.563863721246191e-05,
+      "loss": 1.9922,
+      "step": 3073
+    },
+    {
+      "epoch": 0.13435901918790158,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.563583115120458e-05,
+      "loss": 2.3749,
+      "step": 3074
+    },
+    {
+      "epoch": 0.13440272739193146,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.563302422873272e-05,
+      "loss": 1.9774,
+      "step": 3075
+    },
+    {
+      "epoch": 0.13444643559596137,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.563021644509926e-05,
+      "loss": 2.1169,
+      "step": 3076
+    },
+    {
+      "epoch": 0.13449014379999125,
+      "grad_norm": 2.875,
+      "learning_rate": 9.562740780035721e-05,
+      "loss": 2.4444,
+      "step": 3077
+    },
+    {
+      "epoch": 0.13453385200402115,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.562459829455957e-05,
+      "loss": 1.5088,
+      "step": 3078
+    },
+    {
+      "epoch": 0.13457756020805106,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.562178792775936e-05,
+      "loss": 2.3153,
+      "step": 3079
+    },
+    {
+      "epoch": 0.13462126841208094,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.561897670000958e-05,
+      "loss": 2.1347,
+      "step": 3080
+    },
+    {
+      "epoch": 0.13466497661611085,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.561616461136336e-05,
+      "loss": 2.0139,
+      "step": 3081
+    },
+    {
+      "epoch": 0.13470868482014073,
+      "grad_norm": 2.375,
+      "learning_rate": 9.56133516618737e-05,
+      "loss": 1.9502,
+      "step": 3082
+    },
+    {
+      "epoch": 0.13475239302417064,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.561053785159371e-05,
+      "loss": 1.7509,
+      "step": 3083
+    },
+    {
+      "epoch": 0.13479610122820054,
+      "grad_norm": 2.875,
+      "learning_rate": 9.56077231805765e-05,
+      "loss": 2.5232,
+      "step": 3084
+    },
+    {
+      "epoch": 0.13483980943223042,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.560490764887516e-05,
+      "loss": 2.7191,
+      "step": 3085
+    },
+    {
+      "epoch": 0.13488351763626033,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.560209125654282e-05,
+      "loss": 1.7089,
+      "step": 3086
+    },
+    {
+      "epoch": 0.1349272258402902,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.559927400363268e-05,
+      "loss": 1.6867,
+      "step": 3087
+    },
+    {
+      "epoch": 0.13497093404432012,
+      "grad_norm": 2.875,
+      "learning_rate": 9.559645589019785e-05,
+      "loss": 1.9889,
+      "step": 3088
+    },
+    {
+      "epoch": 0.13501464224835003,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.559363691629155e-05,
+      "loss": 1.6799,
+      "step": 3089
+    },
+    {
+      "epoch": 0.1350583504523799,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.559081708196696e-05,
+      "loss": 2.2714,
+      "step": 3090
+    },
+    {
+      "epoch": 0.1351020586564098,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.55879963872773e-05,
+      "loss": 2.3897,
+      "step": 3091
+    },
+    {
+      "epoch": 0.1351457668604397,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.558517483227579e-05,
+      "loss": 2.3887,
+      "step": 3092
+    },
+    {
+      "epoch": 0.1351894750644696,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.55823524170157e-05,
+      "loss": 1.8025,
+      "step": 3093
+    },
+    {
+      "epoch": 0.1352331832684995,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.557952914155027e-05,
+      "loss": 2.3162,
+      "step": 3094
+    },
+    {
+      "epoch": 0.1352768914725294,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.557670500593276e-05,
+      "loss": 1.6999,
+      "step": 3095
+    },
+    {
+      "epoch": 0.1353205996765593,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.557388001021653e-05,
+      "loss": 2.1748,
+      "step": 3096
+    },
+    {
+      "epoch": 0.13536430788058917,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.557105415445484e-05,
+      "loss": 2.0518,
+      "step": 3097
+    },
+    {
+      "epoch": 0.13540801608461908,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.556822743870104e-05,
+      "loss": 2.3651,
+      "step": 3098
+    },
+    {
+      "epoch": 0.135451724288649,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.556539986300845e-05,
+      "loss": 1.8946,
+      "step": 3099
+    },
+    {
+      "epoch": 0.13549543249267887,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.556257142743046e-05,
+      "loss": 2.0344,
+      "step": 3100
+    },
+    {
+      "epoch": 0.13553914069670878,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.555974213202044e-05,
+      "loss": 2.6906,
+      "step": 3101
+    },
+    {
+      "epoch": 0.13558284890073866,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.555691197683177e-05,
+      "loss": 2.0669,
+      "step": 3102
+    },
+    {
+      "epoch": 0.13562655710476856,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.555408096191786e-05,
+      "loss": 1.9788,
+      "step": 3103
+    },
+    {
+      "epoch": 0.13567026530879847,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.555124908733215e-05,
+      "loss": 1.8438,
+      "step": 3104
+    },
+    {
+      "epoch": 0.13571397351282835,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.554841635312805e-05,
+      "loss": 2.1955,
+      "step": 3105
+    },
+    {
+      "epoch": 0.13575768171685826,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.554558275935907e-05,
+      "loss": 1.9396,
+      "step": 3106
+    },
+    {
+      "epoch": 0.13580138992088814,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.554274830607866e-05,
+      "loss": 1.7357,
+      "step": 3107
+    },
+    {
+      "epoch": 0.13584509812491805,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.553991299334028e-05,
+      "loss": 1.8541,
+      "step": 3108
+    },
+    {
+      "epoch": 0.13588880632894795,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.553707682119746e-05,
+      "loss": 1.9001,
+      "step": 3109
+    },
+    {
+      "epoch": 0.13593251453297783,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.553423978970376e-05,
+      "loss": 2.2973,
+      "step": 3110
+    },
+    {
+      "epoch": 0.13597622273700774,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.553140189891266e-05,
+      "loss": 2.4227,
+      "step": 3111
+    },
+    {
+      "epoch": 0.13601993094103762,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.552856314887772e-05,
+      "loss": 1.8651,
+      "step": 3112
+    },
+    {
+      "epoch": 0.13606363914506753,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.552572353965254e-05,
+      "loss": 1.8486,
+      "step": 3113
+    },
+    {
+      "epoch": 0.13610734734909744,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.552288307129072e-05,
+      "loss": 2.4455,
+      "step": 3114
+    },
+    {
+      "epoch": 0.13615105555312731,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.552004174384583e-05,
+      "loss": 2.3039,
+      "step": 3115
+    },
+    {
+      "epoch": 0.13619476375715722,
+      "grad_norm": 2.5,
+      "learning_rate": 9.551719955737148e-05,
+      "loss": 2.0296,
+      "step": 3116
+    },
+    {
+      "epoch": 0.1362384719611871,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.551435651192135e-05,
+      "loss": 1.6363,
+      "step": 3117
+    },
+    {
+      "epoch": 0.136282180165217,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.551151260754907e-05,
+      "loss": 1.8366,
+      "step": 3118
+    },
+    {
+      "epoch": 0.13632588836924692,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.550866784430829e-05,
+      "loss": 1.8424,
+      "step": 3119
+    },
+    {
+      "epoch": 0.1363695965732768,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.550582222225273e-05,
+      "loss": 2.4484,
+      "step": 3120
+    },
+    {
+      "epoch": 0.1364133047773067,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.550297574143608e-05,
+      "loss": 2.1218,
+      "step": 3121
+    },
+    {
+      "epoch": 0.13645701298133658,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.550012840191203e-05,
+      "loss": 1.9442,
+      "step": 3122
+    },
+    {
+      "epoch": 0.1365007211853665,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.549728020373434e-05,
+      "loss": 2.4977,
+      "step": 3123
+    },
+    {
+      "epoch": 0.1365444293893964,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.549443114695676e-05,
+      "loss": 1.6271,
+      "step": 3124
+    },
+    {
+      "epoch": 0.13658813759342628,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.549158123163305e-05,
+      "loss": 1.5913,
+      "step": 3125
+    },
+    {
+      "epoch": 0.1366318457974562,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.5488730457817e-05,
+      "loss": 2.578,
+      "step": 3126
+    },
+    {
+      "epoch": 0.13667555400148607,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.54858788255624e-05,
+      "loss": 3.1502,
+      "step": 3127
+    },
+    {
+      "epoch": 0.13671926220551597,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.548302633492306e-05,
+      "loss": 1.9293,
+      "step": 3128
+    },
+    {
+      "epoch": 0.13676297040954588,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.548017298595279e-05,
+      "loss": 1.9183,
+      "step": 3129
+    },
+    {
+      "epoch": 0.13680667861357576,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.54773187787055e-05,
+      "loss": 2.0055,
+      "step": 3130
+    },
+    {
+      "epoch": 0.13685038681760567,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.547446371323501e-05,
+      "loss": 2.2359,
+      "step": 3131
+    },
+    {
+      "epoch": 0.13689409502163555,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.547160778959519e-05,
+      "loss": 2.2192,
+      "step": 3132
+    },
+    {
+      "epoch": 0.13693780322566546,
+      "grad_norm": 2.625,
+      "learning_rate": 9.546875100783996e-05,
+      "loss": 2.0813,
+      "step": 3133
+    },
+    {
+      "epoch": 0.13698151142969536,
+      "grad_norm": 3.609375,
+      "learning_rate": 9.54658933680232e-05,
+      "loss": 2.0326,
+      "step": 3134
+    },
+    {
+      "epoch": 0.13702521963372524,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.546303487019888e-05,
+      "loss": 2.1513,
+      "step": 3135
+    },
+    {
+      "epoch": 0.13706892783775515,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.546017551442092e-05,
+      "loss": 2.4596,
+      "step": 3136
+    },
+    {
+      "epoch": 0.13711263604178503,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.545731530074328e-05,
+      "loss": 2.2997,
+      "step": 3137
+    },
+    {
+      "epoch": 0.13715634424581494,
+      "grad_norm": 2.625,
+      "learning_rate": 9.545445422921996e-05,
+      "loss": 2.3707,
+      "step": 3138
+    },
+    {
+      "epoch": 0.13720005244984484,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.545159229990493e-05,
+      "loss": 2.7255,
+      "step": 3139
+    },
+    {
+      "epoch": 0.13724376065387472,
+      "grad_norm": 3.90625,
+      "learning_rate": 9.544872951285217e-05,
+      "loss": 3.7178,
+      "step": 3140
+    },
+    {
+      "epoch": 0.13728746885790463,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.544586586811576e-05,
+      "loss": 1.2338,
+      "step": 3141
+    },
+    {
+      "epoch": 0.1373311770619345,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.544300136574973e-05,
+      "loss": 2.3403,
+      "step": 3142
+    },
+    {
+      "epoch": 0.13737488526596442,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.54401360058081e-05,
+      "loss": 2.3907,
+      "step": 3143
+    },
+    {
+      "epoch": 0.13741859346999433,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.543726978834497e-05,
+      "loss": 2.0025,
+      "step": 3144
+    },
+    {
+      "epoch": 0.1374623016740242,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.543440271341444e-05,
+      "loss": 1.6868,
+      "step": 3145
+    },
+    {
+      "epoch": 0.13750600987805411,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.543153478107061e-05,
+      "loss": 1.867,
+      "step": 3146
+    },
+    {
+      "epoch": 0.137549718082084,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.542866599136759e-05,
+      "loss": 2.2322,
+      "step": 3147
+    },
+    {
+      "epoch": 0.1375934262861139,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.54257963443595e-05,
+      "loss": 2.1401,
+      "step": 3148
+    },
+    {
+      "epoch": 0.1376371344901438,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.542292584010056e-05,
+      "loss": 1.5096,
+      "step": 3149
+    },
+    {
+      "epoch": 0.1376808426941737,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.542005447864488e-05,
+      "loss": 2.3567,
+      "step": 3150
+    },
+    {
+      "epoch": 0.1377245508982036,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.541718226004665e-05,
+      "loss": 2.2019,
+      "step": 3151
+    },
+    {
+      "epoch": 0.13776825910223348,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.541430918436011e-05,
+      "loss": 1.9213,
+      "step": 3152
+    },
+    {
+      "epoch": 0.13781196730626338,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.541143525163946e-05,
+      "loss": 1.7277,
+      "step": 3153
+    },
+    {
+      "epoch": 0.1378556755102933,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.540856046193894e-05,
+      "loss": 2.1676,
+      "step": 3154
+    },
+    {
+      "epoch": 0.13789938371432317,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.540568481531277e-05,
+      "loss": 1.7101,
+      "step": 3155
+    },
+    {
+      "epoch": 0.13794309191835308,
+      "grad_norm": 2.125,
+      "learning_rate": 9.540280831181525e-05,
+      "loss": 1.8688,
+      "step": 3156
+    },
+    {
+      "epoch": 0.13798680012238296,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.539993095150066e-05,
+      "loss": 1.7798,
+      "step": 3157
+    },
+    {
+      "epoch": 0.13803050832641287,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.53970527344233e-05,
+      "loss": 2.4279,
+      "step": 3158
+    },
+    {
+      "epoch": 0.13807421653044277,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.539417366063748e-05,
+      "loss": 2.1791,
+      "step": 3159
+    },
+    {
+      "epoch": 0.13811792473447265,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.539129373019754e-05,
+      "loss": 1.9027,
+      "step": 3160
+    },
+    {
+      "epoch": 0.13816163293850256,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.53884129431578e-05,
+      "loss": 1.9052,
+      "step": 3161
+    },
+    {
+      "epoch": 0.13820534114253244,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.538553129957268e-05,
+      "loss": 2.1048,
+      "step": 3162
+    },
+    {
+      "epoch": 0.13824904934656235,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.538264879949652e-05,
+      "loss": 1.6879,
+      "step": 3163
+    },
+    {
+      "epoch": 0.13829275755059225,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.537976544298373e-05,
+      "loss": 2.0213,
+      "step": 3164
+    },
+    {
+      "epoch": 0.13833646575462213,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.53768812300887e-05,
+      "loss": 2.1036,
+      "step": 3165
+    },
+    {
+      "epoch": 0.13838017395865204,
+      "grad_norm": 2.25,
+      "learning_rate": 9.537399616086588e-05,
+      "loss": 2.0177,
+      "step": 3166
+    },
+    {
+      "epoch": 0.13842388216268192,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.537111023536973e-05,
+      "loss": 1.8663,
+      "step": 3167
+    },
+    {
+      "epoch": 0.13846759036671183,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.53682234536547e-05,
+      "loss": 1.6764,
+      "step": 3168
+    },
+    {
+      "epoch": 0.13851129857074174,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.536533581577525e-05,
+      "loss": 2.2184,
+      "step": 3169
+    },
+    {
+      "epoch": 0.13855500677477162,
+      "grad_norm": 2.625,
+      "learning_rate": 9.536244732178588e-05,
+      "loss": 1.9081,
+      "step": 3170
+    },
+    {
+      "epoch": 0.13859871497880152,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.535955797174112e-05,
+      "loss": 2.1123,
+      "step": 3171
+    },
+    {
+      "epoch": 0.13864242318283143,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.535666776569547e-05,
+      "loss": 2.1118,
+      "step": 3172
+    },
+    {
+      "epoch": 0.1386861313868613,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.53537767037035e-05,
+      "loss": 2.053,
+      "step": 3173
+    },
+    {
+      "epoch": 0.13872983959089122,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.535088478581975e-05,
+      "loss": 2.8949,
+      "step": 3174
+    },
+    {
+      "epoch": 0.1387735477949211,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.53479920120988e-05,
+      "loss": 1.5815,
+      "step": 3175
+    },
+    {
+      "epoch": 0.138817255998951,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.534509838259523e-05,
+      "loss": 2.0353,
+      "step": 3176
+    },
+    {
+      "epoch": 0.1388609642029809,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.534220389736367e-05,
+      "loss": 1.7666,
+      "step": 3177
+    },
+    {
+      "epoch": 0.1389046724070108,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.533930855645872e-05,
+      "loss": 3.2732,
+      "step": 3178
+    },
+    {
+      "epoch": 0.1389483806110407,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.533641235993504e-05,
+      "loss": 2.1203,
+      "step": 3179
+    },
+    {
+      "epoch": 0.13899208881507058,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.533351530784726e-05,
+      "loss": 2.1847,
+      "step": 3180
+    },
+    {
+      "epoch": 0.1390357970191005,
+      "grad_norm": 11.3125,
+      "learning_rate": 9.533061740025008e-05,
+      "loss": 3.1752,
+      "step": 3181
+    },
+    {
+      "epoch": 0.1390795052231304,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.532771863719816e-05,
+      "loss": 2.4082,
+      "step": 3182
+    },
+    {
+      "epoch": 0.13912321342716027,
+      "grad_norm": 2.125,
+      "learning_rate": 9.532481901874624e-05,
+      "loss": 2.0876,
+      "step": 3183
+    },
+    {
+      "epoch": 0.13916692163119018,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.532191854494901e-05,
+      "loss": 2.1728,
+      "step": 3184
+    },
+    {
+      "epoch": 0.13921062983522006,
+      "grad_norm": 3.0,
+      "learning_rate": 9.531901721586121e-05,
+      "loss": 2.6058,
+      "step": 3185
+    },
+    {
+      "epoch": 0.13925433803924997,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.531611503153759e-05,
+      "loss": 1.9939,
+      "step": 3186
+    },
+    {
+      "epoch": 0.13929804624327988,
+      "grad_norm": 2.75,
+      "learning_rate": 9.531321199203292e-05,
+      "loss": 2.3874,
+      "step": 3187
+    },
+    {
+      "epoch": 0.13934175444730976,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.5310308097402e-05,
+      "loss": 1.8303,
+      "step": 3188
+    },
+    {
+      "epoch": 0.13938546265133966,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.530740334769963e-05,
+      "loss": 2.1354,
+      "step": 3189
+    },
+    {
+      "epoch": 0.13942917085536954,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.53044977429806e-05,
+      "loss": 2.4122,
+      "step": 3190
+    },
+    {
+      "epoch": 0.13947287905939945,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.530159128329976e-05,
+      "loss": 1.836,
+      "step": 3191
+    },
+    {
+      "epoch": 0.13951658726342936,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.529868396871197e-05,
+      "loss": 1.8899,
+      "step": 3192
+    },
+    {
+      "epoch": 0.13956029546745924,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.529577579927209e-05,
+      "loss": 2.129,
+      "step": 3193
+    },
+    {
+      "epoch": 0.13960400367148915,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.529286677503499e-05,
+      "loss": 2.3418,
+      "step": 3194
+    },
+    {
+      "epoch": 0.13964771187551903,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.528995689605556e-05,
+      "loss": 1.5699,
+      "step": 3195
+    },
+    {
+      "epoch": 0.13969142007954893,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.528704616238874e-05,
+      "loss": 1.5867,
+      "step": 3196
+    },
+    {
+      "epoch": 0.13973512828357884,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.528413457408944e-05,
+      "loss": 1.8923,
+      "step": 3197
+    },
+    {
+      "epoch": 0.13977883648760872,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.528122213121262e-05,
+      "loss": 2.0186,
+      "step": 3198
+    },
+    {
+      "epoch": 0.13982254469163863,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.527830883381324e-05,
+      "loss": 1.7945,
+      "step": 3199
+    },
+    {
+      "epoch": 0.1398662528956685,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.527539468194625e-05,
+      "loss": 1.8891,
+      "step": 3200
+    },
+    {
+      "epoch": 0.13990996109969842,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.527247967566668e-05,
+      "loss": 2.1507,
+      "step": 3201
+    },
+    {
+      "epoch": 0.13995366930372832,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.526956381502953e-05,
+      "loss": 2.117,
+      "step": 3202
+    },
+    {
+      "epoch": 0.1399973775077582,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.526664710008983e-05,
+      "loss": 2.8609,
+      "step": 3203
+    },
+    {
+      "epoch": 0.1400410857117881,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.52637295309026e-05,
+      "loss": 2.5578,
+      "step": 3204
+    },
+    {
+      "epoch": 0.140084793915818,
+      "grad_norm": 5.8125,
+      "learning_rate": 9.526081110752292e-05,
+      "loss": 2.2185,
+      "step": 3205
+    },
+    {
+      "epoch": 0.1401285021198479,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.525789183000588e-05,
+      "loss": 1.6714,
+      "step": 3206
+    },
+    {
+      "epoch": 0.1401722103238778,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.525497169840653e-05,
+      "loss": 2.083,
+      "step": 3207
+    },
+    {
+      "epoch": 0.14021591852790768,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.525205071278e-05,
+      "loss": 2.0151,
+      "step": 3208
+    },
+    {
+      "epoch": 0.1402596267319376,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.524912887318142e-05,
+      "loss": 1.9866,
+      "step": 3209
+    },
+    {
+      "epoch": 0.14030333493596747,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.524620617966593e-05,
+      "loss": 2.3579,
+      "step": 3210
+    },
+    {
+      "epoch": 0.14034704313999738,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.524328263228865e-05,
+      "loss": 1.9456,
+      "step": 3211
+    },
+    {
+      "epoch": 0.1403907513440273,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.52403582311048e-05,
+      "loss": 1.9813,
+      "step": 3212
+    },
+    {
+      "epoch": 0.14043445954805717,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.523743297616954e-05,
+      "loss": 2.962,
+      "step": 3213
+    },
+    {
+      "epoch": 0.14047816775208707,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.523450686753807e-05,
+      "loss": 1.8951,
+      "step": 3214
+    },
+    {
+      "epoch": 0.14052187595611695,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.523157990526564e-05,
+      "loss": 2.1464,
+      "step": 3215
+    },
+    {
+      "epoch": 0.14056558416014686,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.522865208940745e-05,
+      "loss": 2.693,
+      "step": 3216
+    },
+    {
+      "epoch": 0.14060929236417677,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.522572342001876e-05,
+      "loss": 2.5182,
+      "step": 3217
+    },
+    {
+      "epoch": 0.14065300056820665,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.522279389715488e-05,
+      "loss": 2.3654,
+      "step": 3218
+    },
+    {
+      "epoch": 0.14069670877223656,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.521986352087102e-05,
+      "loss": 2.6586,
+      "step": 3219
+    },
+    {
+      "epoch": 0.14074041697626644,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.521693229122255e-05,
+      "loss": 2.3786,
+      "step": 3220
+    },
+    {
+      "epoch": 0.14078412518029634,
+      "grad_norm": 2.375,
+      "learning_rate": 9.521400020826475e-05,
+      "loss": 2.2397,
+      "step": 3221
+    },
+    {
+      "epoch": 0.14082783338432625,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.521106727205295e-05,
+      "loss": 1.7273,
+      "step": 3222
+    },
+    {
+      "epoch": 0.14087154158835613,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.520813348264252e-05,
+      "loss": 1.6895,
+      "step": 3223
+    },
+    {
+      "epoch": 0.14091524979238604,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.520519884008881e-05,
+      "loss": 2.1822,
+      "step": 3224
+    },
+    {
+      "epoch": 0.14095895799641592,
+      "grad_norm": 2.375,
+      "learning_rate": 9.52022633444472e-05,
+      "loss": 2.0736,
+      "step": 3225
+    },
+    {
+      "epoch": 0.14100266620044583,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.519932699577309e-05,
+      "loss": 2.0192,
+      "step": 3226
+    },
+    {
+      "epoch": 0.14104637440447573,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.519638979412191e-05,
+      "loss": 2.1896,
+      "step": 3227
+    },
+    {
+      "epoch": 0.1410900826085056,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.519345173954907e-05,
+      "loss": 2.3782,
+      "step": 3228
+    },
+    {
+      "epoch": 0.14113379081253552,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.519051283211002e-05,
+      "loss": 2.3049,
+      "step": 3229
+    },
+    {
+      "epoch": 0.1411774990165654,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.518757307186021e-05,
+      "loss": 1.8682,
+      "step": 3230
+    },
+    {
+      "epoch": 0.1412212072205953,
+      "grad_norm": 10.5,
+      "learning_rate": 9.518463245885513e-05,
+      "loss": 3.0038,
+      "step": 3231
+    },
+    {
+      "epoch": 0.14126491542462521,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.518169099315028e-05,
+      "loss": 1.859,
+      "step": 3232
+    },
+    {
+      "epoch": 0.1413086236286551,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.517874867480117e-05,
+      "loss": 1.7138,
+      "step": 3233
+    },
+    {
+      "epoch": 0.141352331832685,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.517580550386331e-05,
+      "loss": 3.028,
+      "step": 3234
+    },
+    {
+      "epoch": 0.14139604003671488,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.517286148039223e-05,
+      "loss": 2.257,
+      "step": 3235
+    },
+    {
+      "epoch": 0.1414397482407448,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.516991660444355e-05,
+      "loss": 2.1228,
+      "step": 3236
+    },
+    {
+      "epoch": 0.1414834564447747,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.516697087607276e-05,
+      "loss": 2.25,
+      "step": 3237
+    },
+    {
+      "epoch": 0.14152716464880458,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.516402429533552e-05,
+      "loss": 2.3626,
+      "step": 3238
+    },
+    {
+      "epoch": 0.14157087285283448,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.51610768622874e-05,
+      "loss": 1.8227,
+      "step": 3239
+    },
+    {
+      "epoch": 0.14161458105686436,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.515812857698403e-05,
+      "loss": 2.5844,
+      "step": 3240
+    },
+    {
+      "epoch": 0.14165828926089427,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.515517943948105e-05,
+      "loss": 1.8612,
+      "step": 3241
+    },
+    {
+      "epoch": 0.14170199746492418,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.51522294498341e-05,
+      "loss": 2.173,
+      "step": 3242
+    },
+    {
+      "epoch": 0.14174570566895406,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.514927860809888e-05,
+      "loss": 2.1819,
+      "step": 3243
+    },
+    {
+      "epoch": 0.14178941387298397,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.514632691433107e-05,
+      "loss": 2.0039,
+      "step": 3244
+    },
+    {
+      "epoch": 0.14183312207701385,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.514337436858635e-05,
+      "loss": 2.5136,
+      "step": 3245
+    },
+    {
+      "epoch": 0.14187683028104375,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.514042097092045e-05,
+      "loss": 1.9573,
+      "step": 3246
+    },
+    {
+      "epoch": 0.14192053848507366,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.513746672138911e-05,
+      "loss": 2.0925,
+      "step": 3247
+    },
+    {
+      "epoch": 0.14196424668910354,
+      "grad_norm": 2.5,
+      "learning_rate": 9.513451162004809e-05,
+      "loss": 1.9035,
+      "step": 3248
+    },
+    {
+      "epoch": 0.14200795489313345,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.513155566695313e-05,
+      "loss": 2.0891,
+      "step": 3249
+    },
+    {
+      "epoch": 0.14205166309716333,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.512859886216003e-05,
+      "loss": 2.0599,
+      "step": 3250
+    },
+    {
+      "epoch": 0.14209537130119324,
+      "grad_norm": 2.625,
+      "learning_rate": 9.51256412057246e-05,
+      "loss": 2.3514,
+      "step": 3251
+    },
+    {
+      "epoch": 0.14213907950522314,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.512268269770264e-05,
+      "loss": 2.2689,
+      "step": 3252
+    },
+    {
+      "epoch": 0.14218278770925302,
+      "grad_norm": 2.625,
+      "learning_rate": 9.511972333814998e-05,
+      "loss": 2.5667,
+      "step": 3253
+    },
+    {
+      "epoch": 0.14222649591328293,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.511676312712246e-05,
+      "loss": 1.9769,
+      "step": 3254
+    },
+    {
+      "epoch": 0.1422702041173128,
+      "grad_norm": 2.375,
+      "learning_rate": 9.511380206467597e-05,
+      "loss": 2.144,
+      "step": 3255
+    },
+    {
+      "epoch": 0.14231391232134272,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.511084015086637e-05,
+      "loss": 2.1737,
+      "step": 3256
+    },
+    {
+      "epoch": 0.14235762052537262,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.510787738574958e-05,
+      "loss": 1.7987,
+      "step": 3257
+    },
+    {
+      "epoch": 0.1424013287294025,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.510491376938147e-05,
+      "loss": 2.2213,
+      "step": 3258
+    },
+    {
+      "epoch": 0.1424450369334324,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.510194930181799e-05,
+      "loss": 1.9827,
+      "step": 3259
+    },
+    {
+      "epoch": 0.1424887451374623,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.50989839831151e-05,
+      "loss": 2.4605,
+      "step": 3260
+    },
+    {
+      "epoch": 0.1425324533414922,
+      "grad_norm": 2.25,
+      "learning_rate": 9.509601781332873e-05,
+      "loss": 1.6873,
+      "step": 3261
+    },
+    {
+      "epoch": 0.1425761615455221,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.509305079251487e-05,
+      "loss": 1.9218,
+      "step": 3262
+    },
+    {
+      "epoch": 0.142619869749552,
+      "grad_norm": 2.5,
+      "learning_rate": 9.509008292072951e-05,
+      "loss": 1.5666,
+      "step": 3263
+    },
+    {
+      "epoch": 0.1426635779535819,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.508711419802867e-05,
+      "loss": 1.6043,
+      "step": 3264
+    },
+    {
+      "epoch": 0.14270728615761177,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.508414462446835e-05,
+      "loss": 1.8474,
+      "step": 3265
+    },
+    {
+      "epoch": 0.14275099436164168,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.508117420010462e-05,
+      "loss": 2.0662,
+      "step": 3266
+    },
+    {
+      "epoch": 0.1427947025656716,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.507820292499353e-05,
+      "loss": 1.7683,
+      "step": 3267
+    },
+    {
+      "epoch": 0.14283841076970147,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.507523079919111e-05,
+      "loss": 1.8974,
+      "step": 3268
+    },
+    {
+      "epoch": 0.14288211897373138,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.507225782275349e-05,
+      "loss": 1.9533,
+      "step": 3269
+    },
+    {
+      "epoch": 0.14292582717776126,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.506928399573678e-05,
+      "loss": 2.3724,
+      "step": 3270
+    },
+    {
+      "epoch": 0.14296953538179116,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.506630931819707e-05,
+      "loss": 2.1634,
+      "step": 3271
+    },
+    {
+      "epoch": 0.14301324358582107,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.506333379019052e-05,
+      "loss": 2.2706,
+      "step": 3272
+    },
+    {
+      "epoch": 0.14305695178985095,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.506035741177329e-05,
+      "loss": 1.7177,
+      "step": 3273
+    },
+    {
+      "epoch": 0.14310065999388086,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.50573801830015e-05,
+      "loss": 1.9437,
+      "step": 3274
+    },
+    {
+      "epoch": 0.14314436819791074,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.50544021039314e-05,
+      "loss": 1.9555,
+      "step": 3275
+    },
+    {
+      "epoch": 0.14318807640194064,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.505142317461915e-05,
+      "loss": 2.1584,
+      "step": 3276
+    },
+    {
+      "epoch": 0.14323178460597055,
+      "grad_norm": 3.125,
+      "learning_rate": 9.504844339512095e-05,
+      "loss": 3.2088,
+      "step": 3277
+    },
+    {
+      "epoch": 0.14327549281000043,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.504546276549309e-05,
+      "loss": 1.9269,
+      "step": 3278
+    },
+    {
+      "epoch": 0.14331920101403034,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.504248128579177e-05,
+      "loss": 2.091,
+      "step": 3279
+    },
+    {
+      "epoch": 0.14336290921806022,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.503949895607329e-05,
+      "loss": 2.1584,
+      "step": 3280
+    },
+    {
+      "epoch": 0.14340661742209013,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.50365157763939e-05,
+      "loss": 1.7713,
+      "step": 3281
+    },
+    {
+      "epoch": 0.14345032562612003,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.503353174680991e-05,
+      "loss": 2.0846,
+      "step": 3282
+    },
+    {
+      "epoch": 0.14349403383014991,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.503054686737763e-05,
+      "loss": 2.443,
+      "step": 3283
+    },
+    {
+      "epoch": 0.14353774203417982,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.502756113815338e-05,
+      "loss": 2.3072,
+      "step": 3284
+    },
+    {
+      "epoch": 0.1435814502382097,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.502457455919355e-05,
+      "loss": 1.807,
+      "step": 3285
+    },
+    {
+      "epoch": 0.1436251584422396,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.502158713055444e-05,
+      "loss": 2.1623,
+      "step": 3286
+    },
+    {
+      "epoch": 0.14366886664626952,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.501859885229248e-05,
+      "loss": 1.5188,
+      "step": 3287
+    },
+    {
+      "epoch": 0.1437125748502994,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.501560972446402e-05,
+      "loss": 2.3192,
+      "step": 3288
+    },
+    {
+      "epoch": 0.1437562830543293,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.501261974712548e-05,
+      "loss": 1.9327,
+      "step": 3289
+    },
+    {
+      "epoch": 0.14379999125835918,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.50096289203333e-05,
+      "loss": 2.3821,
+      "step": 3290
+    },
+    {
+      "epoch": 0.1438436994623891,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.500663724414392e-05,
+      "loss": 1.9964,
+      "step": 3291
+    },
+    {
+      "epoch": 0.143887407666419,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.500364471861378e-05,
+      "loss": 1.928,
+      "step": 3292
+    },
+    {
+      "epoch": 0.14393111587044888,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.500065134379939e-05,
+      "loss": 2.0895,
+      "step": 3293
+    },
+    {
+      "epoch": 0.14397482407447879,
+      "grad_norm": 3.375,
+      "learning_rate": 9.49976571197572e-05,
+      "loss": 1.7263,
+      "step": 3294
+    },
+    {
+      "epoch": 0.14401853227850867,
+      "grad_norm": 3.0,
+      "learning_rate": 9.499466204654372e-05,
+      "loss": 2.0133,
+      "step": 3295
+    },
+    {
+      "epoch": 0.14406224048253857,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.499166612421548e-05,
+      "loss": 2.2996,
+      "step": 3296
+    },
+    {
+      "epoch": 0.14410594868656848,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.498866935282902e-05,
+      "loss": 2.0241,
+      "step": 3297
+    },
+    {
+      "epoch": 0.14414965689059836,
+      "grad_norm": 2.125,
+      "learning_rate": 9.49856717324409e-05,
+      "loss": 2.0614,
+      "step": 3298
+    },
+    {
+      "epoch": 0.14419336509462827,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.498267326310768e-05,
+      "loss": 1.7679,
+      "step": 3299
+    },
+    {
+      "epoch": 0.14423707329865815,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.497967394488594e-05,
+      "loss": 2.318,
+      "step": 3300
+    },
+    {
+      "epoch": 0.14428078150268805,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.497667377783228e-05,
+      "loss": 2.6375,
+      "step": 3301
+    },
+    {
+      "epoch": 0.14432448970671796,
+      "grad_norm": 6.15625,
+      "learning_rate": 9.497367276200335e-05,
+      "loss": 1.3413,
+      "step": 3302
+    },
+    {
+      "epoch": 0.14436819791074784,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.497067089745572e-05,
+      "loss": 2.1526,
+      "step": 3303
+    },
+    {
+      "epoch": 0.14441190611477775,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.496766818424612e-05,
+      "loss": 2.5314,
+      "step": 3304
+    },
+    {
+      "epoch": 0.14445561431880763,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.496466462243115e-05,
+      "loss": 2.562,
+      "step": 3305
+    },
+    {
+      "epoch": 0.14449932252283754,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.496166021206753e-05,
+      "loss": 1.9449,
+      "step": 3306
+    },
+    {
+      "epoch": 0.14454303072686744,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.495865495321194e-05,
+      "loss": 1.7941,
+      "step": 3307
+    },
+    {
+      "epoch": 0.14458673893089732,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.495564884592109e-05,
+      "loss": 3.1137,
+      "step": 3308
+    },
+    {
+      "epoch": 0.14463044713492723,
+      "grad_norm": 2.25,
+      "learning_rate": 9.495264189025172e-05,
+      "loss": 1.586,
+      "step": 3309
+    },
+    {
+      "epoch": 0.1446741553389571,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.494963408626056e-05,
+      "loss": 1.9154,
+      "step": 3310
+    },
+    {
+      "epoch": 0.14471786354298702,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.49466254340044e-05,
+      "loss": 1.7515,
+      "step": 3311
+    },
+    {
+      "epoch": 0.14476157174701693,
+      "grad_norm": 10.5,
+      "learning_rate": 9.494361593354e-05,
+      "loss": 6.3968,
+      "step": 3312
+    },
+    {
+      "epoch": 0.1448052799510468,
+      "grad_norm": 2.625,
+      "learning_rate": 9.494060558492415e-05,
+      "loss": 2.0223,
+      "step": 3313
+    },
+    {
+      "epoch": 0.1448489881550767,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.493759438821366e-05,
+      "loss": 1.9596,
+      "step": 3314
+    },
+    {
+      "epoch": 0.1448926963591066,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.493458234346537e-05,
+      "loss": 1.9956,
+      "step": 3315
+    },
+    {
+      "epoch": 0.1449364045631365,
+      "grad_norm": 2.875,
+      "learning_rate": 9.493156945073611e-05,
+      "loss": 2.5605,
+      "step": 3316
+    },
+    {
+      "epoch": 0.1449801127671664,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.492855571008275e-05,
+      "loss": 2.2362,
+      "step": 3317
+    },
+    {
+      "epoch": 0.1450238209711963,
+      "grad_norm": 2.625,
+      "learning_rate": 9.492554112156214e-05,
+      "loss": 1.9502,
+      "step": 3318
+    },
+    {
+      "epoch": 0.1450675291752262,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.492252568523117e-05,
+      "loss": 1.599,
+      "step": 3319
+    },
+    {
+      "epoch": 0.14511123737925608,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.491950940114678e-05,
+      "loss": 1.7509,
+      "step": 3320
+    },
+    {
+      "epoch": 0.14515494558328598,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.491649226936585e-05,
+      "loss": 3.2139,
+      "step": 3321
+    },
+    {
+      "epoch": 0.1451986537873159,
+      "grad_norm": 4.75,
+      "learning_rate": 9.491347428994536e-05,
+      "loss": 1.9854,
+      "step": 3322
+    },
+    {
+      "epoch": 0.14524236199134577,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.491045546294223e-05,
+      "loss": 2.3345,
+      "step": 3323
+    },
+    {
+      "epoch": 0.14528607019537568,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.490743578841344e-05,
+      "loss": 2.7734,
+      "step": 3324
+    },
+    {
+      "epoch": 0.14532977839940556,
+      "grad_norm": 2.75,
+      "learning_rate": 9.490441526641599e-05,
+      "loss": 1.8867,
+      "step": 3325
+    },
+    {
+      "epoch": 0.14537348660343546,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.490139389700685e-05,
+      "loss": 2.8756,
+      "step": 3326
+    },
+    {
+      "epoch": 0.14541719480746537,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.489837168024307e-05,
+      "loss": 2.0287,
+      "step": 3327
+    },
+    {
+      "epoch": 0.14546090301149525,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.489534861618166e-05,
+      "loss": 2.3343,
+      "step": 3328
+    },
+    {
+      "epoch": 0.14550461121552516,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.48923247048797e-05,
+      "loss": 2.1087,
+      "step": 3329
+    },
+    {
+      "epoch": 0.14554831941955504,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.488929994639421e-05,
+      "loss": 2.1728,
+      "step": 3330
+    },
+    {
+      "epoch": 0.14559202762358495,
+      "grad_norm": 2.5,
+      "learning_rate": 9.488627434078232e-05,
+      "loss": 2.136,
+      "step": 3331
+    },
+    {
+      "epoch": 0.14563573582761485,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.488324788810108e-05,
+      "loss": 2.0132,
+      "step": 3332
+    },
+    {
+      "epoch": 0.14567944403164473,
+      "grad_norm": 2.5,
+      "learning_rate": 9.488022058840765e-05,
+      "loss": 1.7678,
+      "step": 3333
+    },
+    {
+      "epoch": 0.14572315223567464,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.487719244175912e-05,
+      "loss": 1.7634,
+      "step": 3334
+    },
+    {
+      "epoch": 0.14576686043970452,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.487416344821267e-05,
+      "loss": 1.8089,
+      "step": 3335
+    },
+    {
+      "epoch": 0.14581056864373443,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.487113360782543e-05,
+      "loss": 1.9985,
+      "step": 3336
+    },
+    {
+      "epoch": 0.14585427684776434,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.48681029206546e-05,
+      "loss": 1.8382,
+      "step": 3337
+    },
+    {
+      "epoch": 0.14589798505179422,
+      "grad_norm": 2.875,
+      "learning_rate": 9.486507138675735e-05,
+      "loss": 2.4906,
+      "step": 3338
+    },
+    {
+      "epoch": 0.14594169325582412,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.486203900619092e-05,
+      "loss": 2.1798,
+      "step": 3339
+    },
+    {
+      "epoch": 0.145985401459854,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.485900577901252e-05,
+      "loss": 2.698,
+      "step": 3340
+    },
+    {
+      "epoch": 0.1460291096638839,
+      "grad_norm": 2.25,
+      "learning_rate": 9.485597170527939e-05,
+      "loss": 2.2058,
+      "step": 3341
+    },
+    {
+      "epoch": 0.14607281786791382,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.485293678504879e-05,
+      "loss": 2.5278,
+      "step": 3342
+    },
+    {
+      "epoch": 0.1461165260719437,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.484990101837799e-05,
+      "loss": 1.7709,
+      "step": 3343
+    },
+    {
+      "epoch": 0.1461602342759736,
+      "grad_norm": 2.25,
+      "learning_rate": 9.484686440532429e-05,
+      "loss": 1.9356,
+      "step": 3344
+    },
+    {
+      "epoch": 0.14620394248000348,
+      "grad_norm": 3.25,
+      "learning_rate": 9.484382694594498e-05,
+      "loss": 1.9257,
+      "step": 3345
+    },
+    {
+      "epoch": 0.1462476506840334,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.484078864029739e-05,
+      "loss": 1.8498,
+      "step": 3346
+    },
+    {
+      "epoch": 0.1462913588880633,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.483774948843884e-05,
+      "loss": 2.0713,
+      "step": 3347
+    },
+    {
+      "epoch": 0.14633506709209318,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.483470949042672e-05,
+      "loss": 1.8746,
+      "step": 3348
+    },
+    {
+      "epoch": 0.1463787752961231,
+      "grad_norm": 3.25,
+      "learning_rate": 9.483166864631837e-05,
+      "loss": 2.3901,
+      "step": 3349
+    },
+    {
+      "epoch": 0.14642248350015297,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.482862695617119e-05,
+      "loss": 2.7898,
+      "step": 3350
+    },
+    {
+      "epoch": 0.14646619170418287,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.482558442004257e-05,
+      "loss": 2.0902,
+      "step": 3351
+    },
+    {
+      "epoch": 0.14650989990821278,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.482254103798993e-05,
+      "loss": 1.9117,
+      "step": 3352
+    },
+    {
+      "epoch": 0.14655360811224266,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.481949681007069e-05,
+      "loss": 2.7748,
+      "step": 3353
+    },
+    {
+      "epoch": 0.14659731631627257,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.481645173634234e-05,
+      "loss": 1.8882,
+      "step": 3354
+    },
+    {
+      "epoch": 0.14664102452030245,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.48134058168623e-05,
+      "loss": 2.5552,
+      "step": 3355
+    },
+    {
+      "epoch": 0.14668473272433236,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.481035905168808e-05,
+      "loss": 1.67,
+      "step": 3356
+    },
+    {
+      "epoch": 0.14672844092836226,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.480731144087716e-05,
+      "loss": 1.8939,
+      "step": 3357
+    },
+    {
+      "epoch": 0.14677214913239214,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.480426298448706e-05,
+      "loss": 1.988,
+      "step": 3358
+    },
+    {
+      "epoch": 0.14681585733642205,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.480121368257531e-05,
+      "loss": 1.9197,
+      "step": 3359
+    },
+    {
+      "epoch": 0.14685956554045193,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.479816353519946e-05,
+      "loss": 1.9942,
+      "step": 3360
+    },
+    {
+      "epoch": 0.14690327374448184,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.479511254241704e-05,
+      "loss": 2.4083,
+      "step": 3361
+    },
+    {
+      "epoch": 0.14694698194851175,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.479206070428568e-05,
+      "loss": 2.3581,
+      "step": 3362
+    },
+    {
+      "epoch": 0.14699069015254163,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.478900802086292e-05,
+      "loss": 1.7141,
+      "step": 3363
+    },
+    {
+      "epoch": 0.14703439835657153,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.478595449220639e-05,
+      "loss": 1.9283,
+      "step": 3364
+    },
+    {
+      "epoch": 0.1470781065606014,
+      "grad_norm": 2.125,
+      "learning_rate": 9.478290011837375e-05,
+      "loss": 1.9525,
+      "step": 3365
+    },
+    {
+      "epoch": 0.14712181476463132,
+      "grad_norm": 2.25,
+      "learning_rate": 9.477984489942258e-05,
+      "loss": 2.0672,
+      "step": 3366
+    },
+    {
+      "epoch": 0.14716552296866123,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.477678883541055e-05,
+      "loss": 1.9176,
+      "step": 3367
+    },
+    {
+      "epoch": 0.1472092311726911,
+      "grad_norm": 2.25,
+      "learning_rate": 9.477373192639536e-05,
+      "loss": 2.0645,
+      "step": 3368
+    },
+    {
+      "epoch": 0.14725293937672101,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.477067417243468e-05,
+      "loss": 1.8324,
+      "step": 3369
+    },
+    {
+      "epoch": 0.1472966475807509,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.476761557358623e-05,
+      "loss": 2.603,
+      "step": 3370
+    },
+    {
+      "epoch": 0.1473403557847808,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.476455612990771e-05,
+      "loss": 2.2431,
+      "step": 3371
+    },
+    {
+      "epoch": 0.1473840639888107,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.476149584145687e-05,
+      "loss": 1.86,
+      "step": 3372
+    },
+    {
+      "epoch": 0.1474277721928406,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.475843470829145e-05,
+      "loss": 1.998,
+      "step": 3373
+    },
+    {
+      "epoch": 0.1474714803968705,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.475537273046922e-05,
+      "loss": 2.1818,
+      "step": 3374
+    },
+    {
+      "epoch": 0.14751518860090038,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.475230990804797e-05,
+      "loss": 1.7166,
+      "step": 3375
+    },
+    {
+      "epoch": 0.14755889680493028,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.47492462410855e-05,
+      "loss": 1.8999,
+      "step": 3376
+    },
+    {
+      "epoch": 0.1476026050089602,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.474618172963963e-05,
+      "loss": 2.1685,
+      "step": 3377
+    },
+    {
+      "epoch": 0.14764631321299007,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.474311637376818e-05,
+      "loss": 1.6009,
+      "step": 3378
+    },
+    {
+      "epoch": 0.14769002141701998,
+      "grad_norm": 3.0,
+      "learning_rate": 9.474005017352899e-05,
+      "loss": 2.7177,
+      "step": 3379
+    },
+    {
+      "epoch": 0.14773372962104986,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.473698312897997e-05,
+      "loss": 2.1018,
+      "step": 3380
+    },
+    {
+      "epoch": 0.14777743782507977,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.473391524017894e-05,
+      "loss": 2.058,
+      "step": 3381
+    },
+    {
+      "epoch": 0.14782114602910967,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.473084650718382e-05,
+      "loss": 2.4579,
+      "step": 3382
+    },
+    {
+      "epoch": 0.14786485423313955,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.472777693005254e-05,
+      "loss": 2.458,
+      "step": 3383
+    },
+    {
+      "epoch": 0.14790856243716946,
+      "grad_norm": 2.5,
+      "learning_rate": 9.4724706508843e-05,
+      "loss": 2.1446,
+      "step": 3384
+    },
+    {
+      "epoch": 0.14795227064119934,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.472163524361315e-05,
+      "loss": 2.2017,
+      "step": 3385
+    },
+    {
+      "epoch": 0.14799597884522925,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.471856313442098e-05,
+      "loss": 2.6495,
+      "step": 3386
+    },
+    {
+      "epoch": 0.14803968704925916,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.471549018132442e-05,
+      "loss": 1.7599,
+      "step": 3387
+    },
+    {
+      "epoch": 0.14808339525328904,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.471241638438148e-05,
+      "loss": 2.5642,
+      "step": 3388
+    },
+    {
+      "epoch": 0.14812710345731894,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.470934174365016e-05,
+      "loss": 1.8783,
+      "step": 3389
+    },
+    {
+      "epoch": 0.14817081166134882,
+      "grad_norm": 5.21875,
+      "learning_rate": 9.470626625918851e-05,
+      "loss": 2.3672,
+      "step": 3390
+    },
+    {
+      "epoch": 0.14821451986537873,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.470318993105453e-05,
+      "loss": 2.1336,
+      "step": 3391
+    },
+    {
+      "epoch": 0.14825822806940864,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.47001127593063e-05,
+      "loss": 1.8938,
+      "step": 3392
+    },
+    {
+      "epoch": 0.14830193627343852,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.469703474400188e-05,
+      "loss": 2.27,
+      "step": 3393
+    },
+    {
+      "epoch": 0.14834564447746842,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.469395588519939e-05,
+      "loss": 2.4541,
+      "step": 3394
+    },
+    {
+      "epoch": 0.1483893526814983,
+      "grad_norm": 2.375,
+      "learning_rate": 9.469087618295687e-05,
+      "loss": 1.814,
+      "step": 3395
+    },
+    {
+      "epoch": 0.1484330608855282,
+      "grad_norm": 2.125,
+      "learning_rate": 9.468779563733248e-05,
+      "loss": 1.6851,
+      "step": 3396
+    },
+    {
+      "epoch": 0.14847676908955812,
+      "grad_norm": 5.375,
+      "learning_rate": 9.468471424838434e-05,
+      "loss": 1.4152,
+      "step": 3397
+    },
+    {
+      "epoch": 0.148520477293588,
+      "grad_norm": 5.53125,
+      "learning_rate": 9.468163201617062e-05,
+      "loss": 1.3591,
+      "step": 3398
+    },
+    {
+      "epoch": 0.1485641854976179,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.467854894074945e-05,
+      "loss": 2.6621,
+      "step": 3399
+    },
+    {
+      "epoch": 0.1486078937016478,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.467546502217907e-05,
+      "loss": 1.9949,
+      "step": 3400
+    },
+    {
+      "epoch": 0.1486516019056777,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.467238026051762e-05,
+      "loss": 2.0779,
+      "step": 3401
+    },
+    {
+      "epoch": 0.1486953101097076,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.466929465582335e-05,
+      "loss": 1.8255,
+      "step": 3402
+    },
+    {
+      "epoch": 0.14873901831373748,
+      "grad_norm": 2.875,
+      "learning_rate": 9.466620820815446e-05,
+      "loss": 2.032,
+      "step": 3403
+    },
+    {
+      "epoch": 0.1487827265177674,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.466312091756922e-05,
+      "loss": 2.1223,
+      "step": 3404
+    },
+    {
+      "epoch": 0.14882643472179727,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.46600327841259e-05,
+      "loss": 2.3961,
+      "step": 3405
+    },
+    {
+      "epoch": 0.14887014292582718,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.465694380788274e-05,
+      "loss": 2.0143,
+      "step": 3406
+    },
+    {
+      "epoch": 0.14891385112985708,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.465385398889806e-05,
+      "loss": 2.1623,
+      "step": 3407
+    },
+    {
+      "epoch": 0.14895755933388696,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.465076332723017e-05,
+      "loss": 2.1627,
+      "step": 3408
+    },
+    {
+      "epoch": 0.14900126753791687,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.464767182293739e-05,
+      "loss": 1.8762,
+      "step": 3409
+    },
+    {
+      "epoch": 0.14904497574194675,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.464457947607805e-05,
+      "loss": 2.4482,
+      "step": 3410
+    },
+    {
+      "epoch": 0.14908868394597666,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.464148628671053e-05,
+      "loss": 1.5772,
+      "step": 3411
+    },
+    {
+      "epoch": 0.14913239215000657,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.46383922548932e-05,
+      "loss": 2.1683,
+      "step": 3412
+    },
+    {
+      "epoch": 0.14917610035403644,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.463529738068441e-05,
+      "loss": 2.185,
+      "step": 3413
+    },
+    {
+      "epoch": 0.14921980855806635,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.463220166414262e-05,
+      "loss": 2.1126,
+      "step": 3414
+    },
+    {
+      "epoch": 0.14926351676209623,
+      "grad_norm": 2.5,
+      "learning_rate": 9.462910510532621e-05,
+      "loss": 2.6235,
+      "step": 3415
+    },
+    {
+      "epoch": 0.14930722496612614,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.462600770429364e-05,
+      "loss": 1.4827,
+      "step": 3416
+    },
+    {
+      "epoch": 0.14935093317015605,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.462290946110335e-05,
+      "loss": 2.3322,
+      "step": 3417
+    },
+    {
+      "epoch": 0.14939464137418593,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.461981037581383e-05,
+      "loss": 2.0577,
+      "step": 3418
+    },
+    {
+      "epoch": 0.14943834957821583,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.461671044848352e-05,
+      "loss": 1.8846,
+      "step": 3419
+    },
+    {
+      "epoch": 0.14948205778224571,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.461360967917098e-05,
+      "loss": 1.9623,
+      "step": 3420
+    },
+    {
+      "epoch": 0.14952576598627562,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.461050806793468e-05,
+      "loss": 1.7624,
+      "step": 3421
+    },
+    {
+      "epoch": 0.14956947419030553,
+      "grad_norm": 2.25,
+      "learning_rate": 9.460740561483314e-05,
+      "loss": 2.1842,
+      "step": 3422
+    },
+    {
+      "epoch": 0.1496131823943354,
+      "grad_norm": 5.875,
+      "learning_rate": 9.460430231992496e-05,
+      "loss": 1.1398,
+      "step": 3423
+    },
+    {
+      "epoch": 0.14965689059836532,
+      "grad_norm": 2.375,
+      "learning_rate": 9.460119818326866e-05,
+      "loss": 1.8948,
+      "step": 3424
+    },
+    {
+      "epoch": 0.1497005988023952,
+      "grad_norm": 2.625,
+      "learning_rate": 9.459809320492286e-05,
+      "loss": 1.7911,
+      "step": 3425
+    },
+    {
+      "epoch": 0.1497443070064251,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.459498738494613e-05,
+      "loss": 2.3553,
+      "step": 3426
+    },
+    {
+      "epoch": 0.149788015210455,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.459188072339706e-05,
+      "loss": 1.7611,
+      "step": 3427
+    },
+    {
+      "epoch": 0.1498317234144849,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.458877322033431e-05,
+      "loss": 2.0372,
+      "step": 3428
+    },
+    {
+      "epoch": 0.1498754316185148,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.458566487581653e-05,
+      "loss": 1.9606,
+      "step": 3429
+    },
+    {
+      "epoch": 0.14991913982254468,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.458255568990235e-05,
+      "loss": 2.0393,
+      "step": 3430
+    },
+    {
+      "epoch": 0.14996284802657459,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.457944566265045e-05,
+      "loss": 2.1543,
+      "step": 3431
+    },
+    {
+      "epoch": 0.1500065562306045,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.457633479411952e-05,
+      "loss": 2.36,
+      "step": 3432
+    },
+    {
+      "epoch": 0.15005026443463437,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.457322308436828e-05,
+      "loss": 1.8532,
+      "step": 3433
+    },
+    {
+      "epoch": 0.15009397263866428,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.457011053345547e-05,
+      "loss": 2.2903,
+      "step": 3434
+    },
+    {
+      "epoch": 0.15013768084269416,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.45669971414398e-05,
+      "loss": 2.2657,
+      "step": 3435
+    },
+    {
+      "epoch": 0.15018138904672407,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.456388290838e-05,
+      "loss": 1.8505,
+      "step": 3436
+    },
+    {
+      "epoch": 0.15022509725075398,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.45607678343349e-05,
+      "loss": 1.9809,
+      "step": 3437
+    },
+    {
+      "epoch": 0.15026880545478385,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.455765191936326e-05,
+      "loss": 1.7168,
+      "step": 3438
+    },
+    {
+      "epoch": 0.15031251365881376,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.455453516352385e-05,
+      "loss": 2.7301,
+      "step": 3439
+    },
+    {
+      "epoch": 0.15035622186284364,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.455141756687554e-05,
+      "loss": 2.3094,
+      "step": 3440
+    },
+    {
+      "epoch": 0.15039993006687355,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.454829912947712e-05,
+      "loss": 1.5946,
+      "step": 3441
+    },
+    {
+      "epoch": 0.15044363827090346,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.454517985138747e-05,
+      "loss": 1.9504,
+      "step": 3442
+    },
+    {
+      "epoch": 0.15048734647493334,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.454205973266543e-05,
+      "loss": 2.2709,
+      "step": 3443
+    },
+    {
+      "epoch": 0.15053105467896324,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.453893877336991e-05,
+      "loss": 1.9691,
+      "step": 3444
+    },
+    {
+      "epoch": 0.15057476288299312,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.453581697355978e-05,
+      "loss": 2.2119,
+      "step": 3445
+    },
+    {
+      "epoch": 0.15061847108702303,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.453269433329398e-05,
+      "loss": 1.9717,
+      "step": 3446
+    },
+    {
+      "epoch": 0.15066217929105294,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.452957085263142e-05,
+      "loss": 1.919,
+      "step": 3447
+    },
+    {
+      "epoch": 0.15070588749508282,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.452644653163104e-05,
+      "loss": 2.4109,
+      "step": 3448
+    },
+    {
+      "epoch": 0.15074959569911273,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.452332137035181e-05,
+      "loss": 1.6877,
+      "step": 3449
+    },
+    {
+      "epoch": 0.15079330390314263,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.452019536885271e-05,
+      "loss": 2.4079,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1508370121071725,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.451706852719273e-05,
+      "loss": 1.923,
+      "step": 3451
+    },
+    {
+      "epoch": 0.15088072031120242,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.451394084543087e-05,
+      "loss": 1.6903,
+      "step": 3452
+    },
+    {
+      "epoch": 0.1509244285152323,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.451081232362616e-05,
+      "loss": 1.7523,
+      "step": 3453
+    },
+    {
+      "epoch": 0.1509681367192622,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.450768296183765e-05,
+      "loss": 2.4434,
+      "step": 3454
+    },
+    {
+      "epoch": 0.15101184492329212,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.450455276012435e-05,
+      "loss": 1.9919,
+      "step": 3455
+    },
+    {
+      "epoch": 0.151055553127322,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.45014217185454e-05,
+      "loss": 3.1905,
+      "step": 3456
+    },
+    {
+      "epoch": 0.1510992613313519,
+      "grad_norm": 2.25,
+      "learning_rate": 9.449828983715985e-05,
+      "loss": 1.9003,
+      "step": 3457
+    },
+    {
+      "epoch": 0.15114296953538178,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.44951571160268e-05,
+      "loss": 1.8811,
+      "step": 3458
+    },
+    {
+      "epoch": 0.1511866777394117,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.449202355520537e-05,
+      "loss": 2.2725,
+      "step": 3459
+    },
+    {
+      "epoch": 0.1512303859434416,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.448888915475471e-05,
+      "loss": 1.641,
+      "step": 3460
+    },
+    {
+      "epoch": 0.15127409414747148,
+      "grad_norm": 4.875,
+      "learning_rate": 9.448575391473396e-05,
+      "loss": 1.6954,
+      "step": 3461
+    },
+    {
+      "epoch": 0.15131780235150138,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.448261783520228e-05,
+      "loss": 1.897,
+      "step": 3462
+    },
+    {
+      "epoch": 0.15136151055553126,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.447948091621886e-05,
+      "loss": 2.1477,
+      "step": 3463
+    },
+    {
+      "epoch": 0.15140521875956117,
+      "grad_norm": 2.125,
+      "learning_rate": 9.44763431578429e-05,
+      "loss": 1.9178,
+      "step": 3464
+    },
+    {
+      "epoch": 0.15144892696359108,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.447320456013362e-05,
+      "loss": 2.9479,
+      "step": 3465
+    },
+    {
+      "epoch": 0.15149263516762096,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.447006512315025e-05,
+      "loss": 2.2641,
+      "step": 3466
+    },
+    {
+      "epoch": 0.15153634337165087,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.4466924846952e-05,
+      "loss": 1.8904,
+      "step": 3467
+    },
+    {
+      "epoch": 0.15158005157568075,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.446378373159818e-05,
+      "loss": 1.7885,
+      "step": 3468
+    },
+    {
+      "epoch": 0.15162375977971065,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.446064177714804e-05,
+      "loss": 3.0216,
+      "step": 3469
+    },
+    {
+      "epoch": 0.15166746798374056,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.445749898366089e-05,
+      "loss": 2.1373,
+      "step": 3470
+    },
+    {
+      "epoch": 0.15171117618777044,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.445435535119602e-05,
+      "loss": 1.9214,
+      "step": 3471
+    },
+    {
+      "epoch": 0.15175488439180035,
+      "grad_norm": 3.96875,
+      "learning_rate": 9.445121087981277e-05,
+      "loss": 2.1844,
+      "step": 3472
+    },
+    {
+      "epoch": 0.15179859259583023,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.444806556957047e-05,
+      "loss": 1.8605,
+      "step": 3473
+    },
+    {
+      "epoch": 0.15184230079986014,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.444491942052849e-05,
+      "loss": 1.8039,
+      "step": 3474
+    },
+    {
+      "epoch": 0.15188600900389004,
+      "grad_norm": 3.1875,
+      "learning_rate": 9.444177243274618e-05,
+      "loss": 2.4046,
+      "step": 3475
+    },
+    {
+      "epoch": 0.15192971720791992,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.443862460628295e-05,
+      "loss": 2.4799,
+      "step": 3476
+    },
+    {
+      "epoch": 0.15197342541194983,
+      "grad_norm": 2.375,
+      "learning_rate": 9.44354759411982e-05,
+      "loss": 1.925,
+      "step": 3477
+    },
+    {
+      "epoch": 0.1520171336159797,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.443232643755133e-05,
+      "loss": 2.4786,
+      "step": 3478
+    },
+    {
+      "epoch": 0.15206084182000962,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.442917609540181e-05,
+      "loss": 1.7428,
+      "step": 3479
+    },
+    {
+      "epoch": 0.15210455002403953,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.442602491480906e-05,
+      "loss": 2.5866,
+      "step": 3480
+    },
+    {
+      "epoch": 0.1521482582280694,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.442287289583259e-05,
+      "loss": 1.6681,
+      "step": 3481
+    },
+    {
+      "epoch": 0.1521919664320993,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.441972003853181e-05,
+      "loss": 2.4017,
+      "step": 3482
+    },
+    {
+      "epoch": 0.1522356746361292,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.44165663429663e-05,
+      "loss": 1.6264,
+      "step": 3483
+    },
+    {
+      "epoch": 0.1522793828401591,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.441341180919551e-05,
+      "loss": 1.6506,
+      "step": 3484
+    },
+    {
+      "epoch": 0.152323091044189,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.4410256437279e-05,
+      "loss": 1.826,
+      "step": 3485
+    },
+    {
+      "epoch": 0.1523667992482189,
+      "grad_norm": 2.75,
+      "learning_rate": 9.440710022727634e-05,
+      "loss": 2.2919,
+      "step": 3486
+    },
+    {
+      "epoch": 0.1524105074522488,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.440394317924706e-05,
+      "loss": 1.9157,
+      "step": 3487
+    },
+    {
+      "epoch": 0.15245421565627867,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.440078529325073e-05,
+      "loss": 1.9245,
+      "step": 3488
+    },
+    {
+      "epoch": 0.15249792386030858,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.439762656934698e-05,
+      "loss": 1.9615,
+      "step": 3489
+    },
+    {
+      "epoch": 0.1525416320643385,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.439446700759537e-05,
+      "loss": 2.1101,
+      "step": 3490
+    },
+    {
+      "epoch": 0.15258534026836837,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.439130660805558e-05,
+      "loss": 2.4986,
+      "step": 3491
+    },
+    {
+      "epoch": 0.15262904847239828,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.438814537078722e-05,
+      "loss": 1.9813,
+      "step": 3492
+    },
+    {
+      "epoch": 0.15267275667642816,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.438498329584995e-05,
+      "loss": 1.785,
+      "step": 3493
+    },
+    {
+      "epoch": 0.15271646488045806,
+      "grad_norm": 2.125,
+      "learning_rate": 9.438182038330345e-05,
+      "loss": 2.1659,
+      "step": 3494
+    },
+    {
+      "epoch": 0.15276017308448797,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.43786566332074e-05,
+      "loss": 1.9646,
+      "step": 3495
+    },
+    {
+      "epoch": 0.15280388128851785,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.437549204562151e-05,
+      "loss": 1.9267,
+      "step": 3496
+    },
+    {
+      "epoch": 0.15284758949254776,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.43723266206055e-05,
+      "loss": 1.9642,
+      "step": 3497
+    },
+    {
+      "epoch": 0.15289129769657764,
+      "grad_norm": 2.0,
+      "learning_rate": 9.43691603582191e-05,
+      "loss": 1.8211,
+      "step": 3498
+    },
+    {
+      "epoch": 0.15293500590060755,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.436599325852208e-05,
+      "loss": 1.7812,
+      "step": 3499
+    },
+    {
+      "epoch": 0.15297871410463745,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.436282532157419e-05,
+      "loss": 1.7518,
+      "step": 3500
+    },
+    {
+      "epoch": 0.15302242230866733,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.435965654743522e-05,
+      "loss": 1.9779,
+      "step": 3501
+    },
+    {
+      "epoch": 0.15306613051269724,
+      "grad_norm": 19.5,
+      "learning_rate": 9.435648693616496e-05,
+      "loss": 0.4233,
+      "step": 3502
+    },
+    {
+      "epoch": 0.15310983871672712,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.435331648782324e-05,
+      "loss": 2.0467,
+      "step": 3503
+    },
+    {
+      "epoch": 0.15315354692075703,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.43501452024699e-05,
+      "loss": 1.829,
+      "step": 3504
+    },
+    {
+      "epoch": 0.15319725512478694,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.434697308016475e-05,
+      "loss": 2.502,
+      "step": 3505
+    },
+    {
+      "epoch": 0.15324096332881681,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.434380012096768e-05,
+      "loss": 2.0048,
+      "step": 3506
+    },
+    {
+      "epoch": 0.15328467153284672,
+      "grad_norm": 3.640625,
+      "learning_rate": 9.434062632493856e-05,
+      "loss": 1.4008,
+      "step": 3507
+    },
+    {
+      "epoch": 0.1533283797368766,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.43374516921373e-05,
+      "loss": 2.3127,
+      "step": 3508
+    },
+    {
+      "epoch": 0.1533720879409065,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.433427622262379e-05,
+      "loss": 2.145,
+      "step": 3509
+    },
+    {
+      "epoch": 0.15341579614493642,
+      "grad_norm": 2.375,
+      "learning_rate": 9.433109991645795e-05,
+      "loss": 1.7612,
+      "step": 3510
+    },
+    {
+      "epoch": 0.1534595043489663,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.432792277369974e-05,
+      "loss": 2.0052,
+      "step": 3511
+    },
+    {
+      "epoch": 0.1535032125529962,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.432474479440912e-05,
+      "loss": 1.7988,
+      "step": 3512
+    },
+    {
+      "epoch": 0.15354692075702608,
+      "grad_norm": 4.25,
+      "learning_rate": 9.432156597864604e-05,
+      "loss": 1.9816,
+      "step": 3513
+    },
+    {
+      "epoch": 0.153590628961056,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.431838632647052e-05,
+      "loss": 1.6767,
+      "step": 3514
+    },
+    {
+      "epoch": 0.1536343371650859,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.431520583794254e-05,
+      "loss": 1.8451,
+      "step": 3515
+    },
+    {
+      "epoch": 0.15367804536911578,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.431202451312211e-05,
+      "loss": 2.6531,
+      "step": 3516
+    },
+    {
+      "epoch": 0.1537217535731457,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.43088423520693e-05,
+      "loss": 2.4231,
+      "step": 3517
+    },
+    {
+      "epoch": 0.15376546177717557,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.430565935484416e-05,
+      "loss": 2.4119,
+      "step": 3518
+    },
+    {
+      "epoch": 0.15380916998120547,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.430247552150673e-05,
+      "loss": 2.4021,
+      "step": 3519
+    },
+    {
+      "epoch": 0.15385287818523538,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.42992908521171e-05,
+      "loss": 1.9986,
+      "step": 3520
+    },
+    {
+      "epoch": 0.15389658638926526,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.429610534673538e-05,
+      "loss": 1.8919,
+      "step": 3521
+    },
+    {
+      "epoch": 0.15394029459329517,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.42929190054217e-05,
+      "loss": 1.787,
+      "step": 3522
+    },
+    {
+      "epoch": 0.15398400279732505,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.428973182823616e-05,
+      "loss": 2.095,
+      "step": 3523
+    },
+    {
+      "epoch": 0.15402771100135496,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.428654381523892e-05,
+      "loss": 1.0479,
+      "step": 3524
+    },
+    {
+      "epoch": 0.15407141920538486,
+      "grad_norm": 3.5,
+      "learning_rate": 9.428335496649014e-05,
+      "loss": 2.9287,
+      "step": 3525
+    },
+    {
+      "epoch": 0.15411512740941474,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.428016528205001e-05,
+      "loss": 2.4471,
+      "step": 3526
+    },
+    {
+      "epoch": 0.15415883561344465,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.42769747619787e-05,
+      "loss": 1.9132,
+      "step": 3527
+    },
+    {
+      "epoch": 0.15420254381747453,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.427378340633645e-05,
+      "loss": 1.7993,
+      "step": 3528
+    },
+    {
+      "epoch": 0.15424625202150444,
+      "grad_norm": 3.859375,
+      "learning_rate": 9.427059121518346e-05,
+      "loss": 2.7769,
+      "step": 3529
+    },
+    {
+      "epoch": 0.15428996022553434,
+      "grad_norm": 3.625,
+      "learning_rate": 9.426739818857998e-05,
+      "loss": 1.8669,
+      "step": 3530
+    },
+    {
+      "epoch": 0.15433366842956422,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.426420432658627e-05,
+      "loss": 2.2612,
+      "step": 3531
+    },
+    {
+      "epoch": 0.15437737663359413,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.426100962926261e-05,
+      "loss": 1.631,
+      "step": 3532
+    },
+    {
+      "epoch": 0.154421084837624,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.425781409666926e-05,
+      "loss": 1.6795,
+      "step": 3533
+    },
+    {
+      "epoch": 0.15446479304165392,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.425461772886656e-05,
+      "loss": 3.0124,
+      "step": 3534
+    },
+    {
+      "epoch": 0.15450850124568383,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.42514205259148e-05,
+      "loss": 1.6248,
+      "step": 3535
+    },
+    {
+      "epoch": 0.1545522094497137,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.424822248787435e-05,
+      "loss": 2.2115,
+      "step": 3536
+    },
+    {
+      "epoch": 0.15459591765374361,
+      "grad_norm": 2.375,
+      "learning_rate": 9.424502361480552e-05,
+      "loss": 2.0061,
+      "step": 3537
+    },
+    {
+      "epoch": 0.1546396258577735,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.424182390676872e-05,
+      "loss": 1.5383,
+      "step": 3538
+    },
+    {
+      "epoch": 0.1546833340618034,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.423862336382429e-05,
+      "loss": 1.6646,
+      "step": 3539
+    },
+    {
+      "epoch": 0.1547270422658333,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.423542198603267e-05,
+      "loss": 1.7199,
+      "step": 3540
+    },
+    {
+      "epoch": 0.1547707504698632,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.423221977345424e-05,
+      "loss": 1.8018,
+      "step": 3541
+    },
+    {
+      "epoch": 0.1548144586738931,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.422901672614946e-05,
+      "loss": 2.5242,
+      "step": 3542
+    },
+    {
+      "epoch": 0.15485816687792298,
+      "grad_norm": 2.875,
+      "learning_rate": 9.422581284417875e-05,
+      "loss": 2.0725,
+      "step": 3543
+    },
+    {
+      "epoch": 0.15490187508195288,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.422260812760259e-05,
+      "loss": 2.1044,
+      "step": 3544
+    },
+    {
+      "epoch": 0.1549455832859828,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.421940257648146e-05,
+      "loss": 1.7116,
+      "step": 3545
+    },
+    {
+      "epoch": 0.15498929149001267,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.421619619087582e-05,
+      "loss": 1.9174,
+      "step": 3546
+    },
+    {
+      "epoch": 0.15503299969404258,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.421298897084623e-05,
+      "loss": 1.6679,
+      "step": 3547
+    },
+    {
+      "epoch": 0.15507670789807246,
+      "grad_norm": 3.25,
+      "learning_rate": 9.420978091645318e-05,
+      "loss": 2.2195,
+      "step": 3548
+    },
+    {
+      "epoch": 0.15512041610210237,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.420657202775722e-05,
+      "loss": 1.989,
+      "step": 3549
+    },
+    {
+      "epoch": 0.15516412430613227,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.42033623048189e-05,
+      "loss": 1.9881,
+      "step": 3550
+    },
+    {
+      "epoch": 0.15520783251016215,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.420015174769881e-05,
+      "loss": 2.7604,
+      "step": 3551
+    },
+    {
+      "epoch": 0.15525154071419206,
+      "grad_norm": 2.375,
+      "learning_rate": 9.419694035645751e-05,
+      "loss": 1.9209,
+      "step": 3552
+    },
+    {
+      "epoch": 0.15529524891822194,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.419372813115563e-05,
+      "loss": 2.8501,
+      "step": 3553
+    },
+    {
+      "epoch": 0.15533895712225185,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.419051507185378e-05,
+      "loss": 2.2907,
+      "step": 3554
+    },
+    {
+      "epoch": 0.15538266532628175,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.418730117861259e-05,
+      "loss": 2.0245,
+      "step": 3555
+    },
+    {
+      "epoch": 0.15542637353031163,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.418408645149273e-05,
+      "loss": 1.9516,
+      "step": 3556
+    },
+    {
+      "epoch": 0.15547008173434154,
+      "grad_norm": 3.125,
+      "learning_rate": 9.418087089055484e-05,
+      "loss": 2.2403,
+      "step": 3557
+    },
+    {
+      "epoch": 0.15551378993837142,
+      "grad_norm": 2.375,
+      "learning_rate": 9.417765449585961e-05,
+      "loss": 2.1481,
+      "step": 3558
+    },
+    {
+      "epoch": 0.15555749814240133,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.417443726746776e-05,
+      "loss": 2.1226,
+      "step": 3559
+    },
+    {
+      "epoch": 0.15560120634643124,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.417121920543996e-05,
+      "loss": 2.2953,
+      "step": 3560
+    },
+    {
+      "epoch": 0.15564491455046112,
+      "grad_norm": 3.515625,
+      "learning_rate": 9.416800030983699e-05,
+      "loss": 1.9059,
+      "step": 3561
+    },
+    {
+      "epoch": 0.15568862275449102,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.416478058071956e-05,
+      "loss": 2.1399,
+      "step": 3562
+    },
+    {
+      "epoch": 0.1557323309585209,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.416156001814843e-05,
+      "loss": 2.1246,
+      "step": 3563
+    },
+    {
+      "epoch": 0.1557760391625508,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.41583386221844e-05,
+      "loss": 1.8416,
+      "step": 3564
+    },
+    {
+      "epoch": 0.15581974736658072,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.415511639288826e-05,
+      "loss": 2.3951,
+      "step": 3565
+    },
+    {
+      "epoch": 0.1558634555706106,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.41518933303208e-05,
+      "loss": 2.3691,
+      "step": 3566
+    },
+    {
+      "epoch": 0.1559071637746405,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.414866943454284e-05,
+      "loss": 2.3471,
+      "step": 3567
+    },
+    {
+      "epoch": 0.15595087197867039,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.414544470561524e-05,
+      "loss": 2.5009,
+      "step": 3568
+    },
+    {
+      "epoch": 0.1559945801827003,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.414221914359886e-05,
+      "loss": 2.6902,
+      "step": 3569
+    },
+    {
+      "epoch": 0.1560382883867302,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.413899274855454e-05,
+      "loss": 2.5366,
+      "step": 3570
+    },
+    {
+      "epoch": 0.15608199659076008,
+      "grad_norm": 2.375,
+      "learning_rate": 9.41357655205432e-05,
+      "loss": 1.9691,
+      "step": 3571
+    },
+    {
+      "epoch": 0.15612570479479,
+      "grad_norm": 4.6875,
+      "learning_rate": 9.413253745962573e-05,
+      "loss": 2.2237,
+      "step": 3572
+    },
+    {
+      "epoch": 0.15616941299881987,
+      "grad_norm": 2.125,
+      "learning_rate": 9.412930856586304e-05,
+      "loss": 2.151,
+      "step": 3573
+    },
+    {
+      "epoch": 0.15621312120284978,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.412607883931607e-05,
+      "loss": 1.9646,
+      "step": 3574
+    },
+    {
+      "epoch": 0.15625682940687968,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.412284828004577e-05,
+      "loss": 1.9428,
+      "step": 3575
+    },
+    {
+      "epoch": 0.15630053761090956,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.41196168881131e-05,
+      "loss": 2.2769,
+      "step": 3576
+    },
+    {
+      "epoch": 0.15634424581493947,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.411638466357906e-05,
+      "loss": 1.6582,
+      "step": 3577
+    },
+    {
+      "epoch": 0.15638795401896935,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.411315160650462e-05,
+      "loss": 1.6979,
+      "step": 3578
+    },
+    {
+      "epoch": 0.15643166222299926,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.410991771695082e-05,
+      "loss": 2.0957,
+      "step": 3579
+    },
+    {
+      "epoch": 0.15647537042702916,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.410668299497864e-05,
+      "loss": 1.7496,
+      "step": 3580
+    },
+    {
+      "epoch": 0.15651907863105904,
+      "grad_norm": 2.25,
+      "learning_rate": 9.410344744064919e-05,
+      "loss": 1.6589,
+      "step": 3581
+    },
+    {
+      "epoch": 0.15656278683508895,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.410021105402348e-05,
+      "loss": 3.1007,
+      "step": 3582
+    },
+    {
+      "epoch": 0.15660649503911883,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.409697383516263e-05,
+      "loss": 2.4073,
+      "step": 3583
+    },
+    {
+      "epoch": 0.15665020324314874,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.409373578412769e-05,
+      "loss": 2.0086,
+      "step": 3584
+    },
+    {
+      "epoch": 0.15669391144717865,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.409049690097977e-05,
+      "loss": 1.5228,
+      "step": 3585
+    },
+    {
+      "epoch": 0.15673761965120853,
+      "grad_norm": 3.0,
+      "learning_rate": 9.408725718578e-05,
+      "loss": 2.3846,
+      "step": 3586
+    },
+    {
+      "epoch": 0.15678132785523843,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.408401663858953e-05,
+      "loss": 1.5352,
+      "step": 3587
+    },
+    {
+      "epoch": 0.1568250360592683,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.408077525946952e-05,
+      "loss": 2.0641,
+      "step": 3588
+    },
+    {
+      "epoch": 0.15686874426329822,
+      "grad_norm": 2.375,
+      "learning_rate": 9.40775330484811e-05,
+      "loss": 2.0999,
+      "step": 3589
+    },
+    {
+      "epoch": 0.15691245246732813,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.407429000568549e-05,
+      "loss": 1.557,
+      "step": 3590
+    },
+    {
+      "epoch": 0.156956160671358,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.407104613114388e-05,
+      "loss": 1.735,
+      "step": 3591
+    },
+    {
+      "epoch": 0.15699986887538792,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.406780142491748e-05,
+      "loss": 1.8904,
+      "step": 3592
+    },
+    {
+      "epoch": 0.1570435770794178,
+      "grad_norm": 11.9375,
+      "learning_rate": 9.406455588706752e-05,
+      "loss": 2.161,
+      "step": 3593
+    },
+    {
+      "epoch": 0.1570872852834477,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.406130951765529e-05,
+      "loss": 2.2186,
+      "step": 3594
+    },
+    {
+      "epoch": 0.1571309934874776,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.405806231674202e-05,
+      "loss": 1.8745,
+      "step": 3595
+    },
+    {
+      "epoch": 0.1571747016915075,
+      "grad_norm": 2.25,
+      "learning_rate": 9.405481428438896e-05,
+      "loss": 1.8984,
+      "step": 3596
+    },
+    {
+      "epoch": 0.1572184098955374,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.405156542065745e-05,
+      "loss": 1.8932,
+      "step": 3597
+    },
+    {
+      "epoch": 0.15726211809956728,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.404831572560879e-05,
+      "loss": 2.3254,
+      "step": 3598
+    },
+    {
+      "epoch": 0.15730582630359718,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.40450651993043e-05,
+      "loss": 2.3937,
+      "step": 3599
+    },
+    {
+      "epoch": 0.1573495345076271,
+      "grad_norm": 2.375,
+      "learning_rate": 9.404181384180532e-05,
+      "loss": 2.4118,
+      "step": 3600
+    },
+    {
+      "epoch": 0.15739324271165697,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.403856165317321e-05,
+      "loss": 2.095,
+      "step": 3601
+    },
+    {
+      "epoch": 0.15743695091568688,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.403530863346937e-05,
+      "loss": 2.1934,
+      "step": 3602
+    },
+    {
+      "epoch": 0.15748065911971676,
+      "grad_norm": 2.875,
+      "learning_rate": 9.403205478275514e-05,
+      "loss": 1.9076,
+      "step": 3603
+    },
+    {
+      "epoch": 0.15752436732374667,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.402880010109196e-05,
+      "loss": 2.0982,
+      "step": 3604
+    },
+    {
+      "epoch": 0.15756807552777657,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.402554458854125e-05,
+      "loss": 2.1694,
+      "step": 3605
+    },
+    {
+      "epoch": 0.15761178373180645,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.402228824516442e-05,
+      "loss": 2.0995,
+      "step": 3606
+    },
+    {
+      "epoch": 0.15765549193583636,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.401903107102296e-05,
+      "loss": 2.2792,
+      "step": 3607
+    },
+    {
+      "epoch": 0.15769920013986624,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.40157730661783e-05,
+      "loss": 1.9848,
+      "step": 3608
+    },
+    {
+      "epoch": 0.15774290834389615,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.401251423069194e-05,
+      "loss": 1.9047,
+      "step": 3609
+    },
+    {
+      "epoch": 0.15778661654792606,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.400925456462539e-05,
+      "loss": 1.9244,
+      "step": 3610
+    },
+    {
+      "epoch": 0.15783032475195594,
+      "grad_norm": 3.375,
+      "learning_rate": 9.400599406804016e-05,
+      "loss": 2.0467,
+      "step": 3611
+    },
+    {
+      "epoch": 0.15787403295598584,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.400273274099776e-05,
+      "loss": 1.9429,
+      "step": 3612
+    },
+    {
+      "epoch": 0.15791774116001572,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.399947058355976e-05,
+      "loss": 1.8686,
+      "step": 3613
+    },
+    {
+      "epoch": 0.15796144936404563,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.399620759578769e-05,
+      "loss": 1.9697,
+      "step": 3614
+    },
+    {
+      "epoch": 0.15800515756807554,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.399294377774318e-05,
+      "loss": 2.0929,
+      "step": 3615
+    },
+    {
+      "epoch": 0.15804886577210542,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.398967912948778e-05,
+      "loss": 2.5104,
+      "step": 3616
+    },
+    {
+      "epoch": 0.15809257397613533,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.398641365108309e-05,
+      "loss": 2.3617,
+      "step": 3617
+    },
+    {
+      "epoch": 0.1581362821801652,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.398314734259078e-05,
+      "loss": 2.0578,
+      "step": 3618
+    },
+    {
+      "epoch": 0.1581799903841951,
+      "grad_norm": 6.15625,
+      "learning_rate": 9.397988020407246e-05,
+      "loss": 2.7893,
+      "step": 3619
+    },
+    {
+      "epoch": 0.15822369858822502,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.397661223558979e-05,
+      "loss": 1.8977,
+      "step": 3620
+    },
+    {
+      "epoch": 0.1582674067922549,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.397334343720445e-05,
+      "loss": 1.6987,
+      "step": 3621
+    },
+    {
+      "epoch": 0.1583111149962848,
+      "grad_norm": 3.0,
+      "learning_rate": 9.39700738089781e-05,
+      "loss": 2.5893,
+      "step": 3622
+    },
+    {
+      "epoch": 0.1583548232003147,
+      "grad_norm": 2.25,
+      "learning_rate": 9.396680335097247e-05,
+      "loss": 1.7377,
+      "step": 3623
+    },
+    {
+      "epoch": 0.1583985314043446,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.396353206324929e-05,
+      "loss": 2.4146,
+      "step": 3624
+    },
+    {
+      "epoch": 0.1584422396083745,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.396025994587024e-05,
+      "loss": 1.7168,
+      "step": 3625
+    },
+    {
+      "epoch": 0.15848594781240438,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.395698699889713e-05,
+      "loss": 1.5608,
+      "step": 3626
+    },
+    {
+      "epoch": 0.1585296560164343,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.395371322239168e-05,
+      "loss": 1.6012,
+      "step": 3627
+    },
+    {
+      "epoch": 0.15857336422046417,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.395043861641571e-05,
+      "loss": 2.1641,
+      "step": 3628
+    },
+    {
+      "epoch": 0.15861707242449408,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.394716318103098e-05,
+      "loss": 2.3077,
+      "step": 3629
+    },
+    {
+      "epoch": 0.15866078062852398,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.394388691629932e-05,
+      "loss": 1.7936,
+      "step": 3630
+    },
+    {
+      "epoch": 0.15870448883255386,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.394060982228257e-05,
+      "loss": 1.8558,
+      "step": 3631
+    },
+    {
+      "epoch": 0.15874819703658377,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.393733189904254e-05,
+      "loss": 1.8328,
+      "step": 3632
+    },
+    {
+      "epoch": 0.15879190524061365,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.393405314664113e-05,
+      "loss": 2.028,
+      "step": 3633
+    },
+    {
+      "epoch": 0.15883561344464356,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.393077356514018e-05,
+      "loss": 1.8092,
+      "step": 3634
+    },
+    {
+      "epoch": 0.15887932164867347,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.392749315460161e-05,
+      "loss": 1.8547,
+      "step": 3635
+    },
+    {
+      "epoch": 0.15892302985270335,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.392421191508729e-05,
+      "loss": 1.7204,
+      "step": 3636
+    },
+    {
+      "epoch": 0.15896673805673325,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.392092984665918e-05,
+      "loss": 3.101,
+      "step": 3637
+    },
+    {
+      "epoch": 0.15901044626076313,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.391764694937919e-05,
+      "loss": 2.1922,
+      "step": 3638
+    },
+    {
+      "epoch": 0.15905415446479304,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.391436322330928e-05,
+      "loss": 1.7696,
+      "step": 3639
+    },
+    {
+      "epoch": 0.15909786266882295,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.391107866851143e-05,
+      "loss": 2.4794,
+      "step": 3640
+    },
+    {
+      "epoch": 0.15914157087285283,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.390779328504762e-05,
+      "loss": 2.1426,
+      "step": 3641
+    },
+    {
+      "epoch": 0.15918527907688274,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.390450707297984e-05,
+      "loss": 1.9995,
+      "step": 3642
+    },
+    {
+      "epoch": 0.15922898728091261,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.39012200323701e-05,
+      "loss": 2.6311,
+      "step": 3643
+    },
+    {
+      "epoch": 0.15927269548494252,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.389793216328047e-05,
+      "loss": 1.9958,
+      "step": 3644
+    },
+    {
+      "epoch": 0.15931640368897243,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.389464346577295e-05,
+      "loss": 1.8677,
+      "step": 3645
+    },
+    {
+      "epoch": 0.1593601118930023,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.389135393990962e-05,
+      "loss": 2.6298,
+      "step": 3646
+    },
+    {
+      "epoch": 0.15940382009703222,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.388806358575256e-05,
+      "loss": 2.1751,
+      "step": 3647
+    },
+    {
+      "epoch": 0.1594475283010621,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.388477240336387e-05,
+      "loss": 2.0166,
+      "step": 3648
+    },
+    {
+      "epoch": 0.159491236505092,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.388148039280566e-05,
+      "loss": 2.4338,
+      "step": 3649
+    },
+    {
+      "epoch": 0.1595349447091219,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.387818755414004e-05,
+      "loss": 1.426,
+      "step": 3650
+    },
+    {
+      "epoch": 0.1595786529131518,
+      "grad_norm": 2.25,
+      "learning_rate": 9.387489388742917e-05,
+      "loss": 1.8721,
+      "step": 3651
+    },
+    {
+      "epoch": 0.1596223611171817,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.387159939273518e-05,
+      "loss": 1.9787,
+      "step": 3652
+    },
+    {
+      "epoch": 0.15966606932121158,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.386830407012026e-05,
+      "loss": 2.2221,
+      "step": 3653
+    },
+    {
+      "epoch": 0.1597097775252415,
+      "grad_norm": 4.90625,
+      "learning_rate": 9.386500791964661e-05,
+      "loss": 2.0422,
+      "step": 3654
+    },
+    {
+      "epoch": 0.1597534857292714,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.38617109413764e-05,
+      "loss": 2.3686,
+      "step": 3655
+    },
+    {
+      "epoch": 0.15979719393330127,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.385841313537187e-05,
+      "loss": 2.6774,
+      "step": 3656
+    },
+    {
+      "epoch": 0.15984090213733118,
+      "grad_norm": 12.375,
+      "learning_rate": 9.385511450169525e-05,
+      "loss": 2.3355,
+      "step": 3657
+    },
+    {
+      "epoch": 0.15988461034136106,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.385181504040881e-05,
+      "loss": 2.1482,
+      "step": 3658
+    },
+    {
+      "epoch": 0.15992831854539097,
+      "grad_norm": 2.375,
+      "learning_rate": 9.384851475157477e-05,
+      "loss": 2.1601,
+      "step": 3659
+    },
+    {
+      "epoch": 0.15997202674942088,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.384521363525546e-05,
+      "loss": 2.2922,
+      "step": 3660
+    },
+    {
+      "epoch": 0.16001573495345076,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.384191169151313e-05,
+      "loss": 2.2924,
+      "step": 3661
+    },
+    {
+      "epoch": 0.16005944315748066,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.383860892041014e-05,
+      "loss": 2.3593,
+      "step": 3662
+    },
+    {
+      "epoch": 0.16010315136151054,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.383530532200879e-05,
+      "loss": 2.0018,
+      "step": 3663
+    },
+    {
+      "epoch": 0.16014685956554045,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.383200089637143e-05,
+      "loss": 1.9835,
+      "step": 3664
+    },
+    {
+      "epoch": 0.16019056776957036,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.382869564356043e-05,
+      "loss": 2.4303,
+      "step": 3665
+    },
+    {
+      "epoch": 0.16023427597360024,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.382538956363813e-05,
+      "loss": 2.014,
+      "step": 3666
+    },
+    {
+      "epoch": 0.16027798417763015,
+      "grad_norm": 2.125,
+      "learning_rate": 9.382208265666695e-05,
+      "loss": 1.8586,
+      "step": 3667
+    },
+    {
+      "epoch": 0.16032169238166002,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.38187749227093e-05,
+      "loss": 2.2283,
+      "step": 3668
+    },
+    {
+      "epoch": 0.16036540058568993,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.381546636182758e-05,
+      "loss": 2.8335,
+      "step": 3669
+    },
+    {
+      "epoch": 0.16040910878971984,
+      "grad_norm": 3.5625,
+      "learning_rate": 9.381215697408426e-05,
+      "loss": 1.6542,
+      "step": 3670
+    },
+    {
+      "epoch": 0.16045281699374972,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.380884675954176e-05,
+      "loss": 2.2729,
+      "step": 3671
+    },
+    {
+      "epoch": 0.16049652519777963,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.380553571826256e-05,
+      "loss": 1.8695,
+      "step": 3672
+    },
+    {
+      "epoch": 0.1605402334018095,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.380222385030915e-05,
+      "loss": 1.9721,
+      "step": 3673
+    },
+    {
+      "epoch": 0.16058394160583941,
+      "grad_norm": 2.25,
+      "learning_rate": 9.379891115574402e-05,
+      "loss": 2.1319,
+      "step": 3674
+    },
+    {
+      "epoch": 0.16062764980986932,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.379559763462968e-05,
+      "loss": 2.5423,
+      "step": 3675
+    },
+    {
+      "epoch": 0.1606713580138992,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.379228328702868e-05,
+      "loss": 1.8878,
+      "step": 3676
+    },
+    {
+      "epoch": 0.1607150662179291,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.378896811300356e-05,
+      "loss": 2.2672,
+      "step": 3677
+    },
+    {
+      "epoch": 0.160758774421959,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.378565211261687e-05,
+      "loss": 1.9762,
+      "step": 3678
+    },
+    {
+      "epoch": 0.1608024826259889,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.378233528593121e-05,
+      "loss": 2.1997,
+      "step": 3679
+    },
+    {
+      "epoch": 0.1608461908300188,
+      "grad_norm": 2.375,
+      "learning_rate": 9.377901763300916e-05,
+      "loss": 1.9741,
+      "step": 3680
+    },
+    {
+      "epoch": 0.16088989903404868,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.377569915391333e-05,
+      "loss": 2.4164,
+      "step": 3681
+    },
+    {
+      "epoch": 0.1609336072380786,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.377237984870634e-05,
+      "loss": 2.2117,
+      "step": 3682
+    },
+    {
+      "epoch": 0.16097731544210847,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.376905971745085e-05,
+      "loss": 2.0477,
+      "step": 3683
+    },
+    {
+      "epoch": 0.16102102364613838,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.376573876020948e-05,
+      "loss": 1.9321,
+      "step": 3684
+    },
+    {
+      "epoch": 0.16106473185016829,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.376241697704493e-05,
+      "loss": 2.2536,
+      "step": 3685
+    },
+    {
+      "epoch": 0.16110844005419817,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.375909436801988e-05,
+      "loss": 1.8819,
+      "step": 3686
+    },
+    {
+      "epoch": 0.16115214825822807,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.375577093319701e-05,
+      "loss": 1.461,
+      "step": 3687
+    },
+    {
+      "epoch": 0.16119585646225795,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.37524466726391e-05,
+      "loss": 2.1925,
+      "step": 3688
+    },
+    {
+      "epoch": 0.16123956466628786,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.37491215864088e-05,
+      "loss": 1.9754,
+      "step": 3689
+    },
+    {
+      "epoch": 0.16128327287031777,
+      "grad_norm": 2.625,
+      "learning_rate": 9.374579567456892e-05,
+      "loss": 1.9294,
+      "step": 3690
+    },
+    {
+      "epoch": 0.16132698107434765,
+      "grad_norm": 2.5,
+      "learning_rate": 9.374246893718221e-05,
+      "loss": 1.5906,
+      "step": 3691
+    },
+    {
+      "epoch": 0.16137068927837755,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.373914137431146e-05,
+      "loss": 2.2835,
+      "step": 3692
+    },
+    {
+      "epoch": 0.16141439748240743,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.373581298601943e-05,
+      "loss": 2.0546,
+      "step": 3693
+    },
+    {
+      "epoch": 0.16145810568643734,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.373248377236896e-05,
+      "loss": 1.9919,
+      "step": 3694
+    },
+    {
+      "epoch": 0.16150181389046725,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.372915373342288e-05,
+      "loss": 1.7252,
+      "step": 3695
+    },
+    {
+      "epoch": 0.16154552209449713,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.3725822869244e-05,
+      "loss": 1.6739,
+      "step": 3696
+    },
+    {
+      "epoch": 0.16158923029852704,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.372249117989523e-05,
+      "loss": 2.5599,
+      "step": 3697
+    },
+    {
+      "epoch": 0.16163293850255692,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.37191586654394e-05,
+      "loss": 1.8535,
+      "step": 3698
+    },
+    {
+      "epoch": 0.16167664670658682,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.371582532593943e-05,
+      "loss": 2.326,
+      "step": 3699
+    },
+    {
+      "epoch": 0.16172035491061673,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.37124911614582e-05,
+      "loss": 1.3687,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1617640631146466,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.370915617205865e-05,
+      "loss": 1.9865,
+      "step": 3701
+    },
+    {
+      "epoch": 0.16180777131867652,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.370582035780371e-05,
+      "loss": 2.1977,
+      "step": 3702
+    },
+    {
+      "epoch": 0.1618514795227064,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.370248371875631e-05,
+      "loss": 2.3067,
+      "step": 3703
+    },
+    {
+      "epoch": 0.1618951877267363,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.369914625497945e-05,
+      "loss": 1.9207,
+      "step": 3704
+    },
+    {
+      "epoch": 0.1619388959307662,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.36958079665361e-05,
+      "loss": 2.5128,
+      "step": 3705
+    },
+    {
+      "epoch": 0.1619826041347961,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.369246885348926e-05,
+      "loss": 2.426,
+      "step": 3706
+    },
+    {
+      "epoch": 0.162026312338826,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.368912891590192e-05,
+      "loss": 1.9204,
+      "step": 3707
+    },
+    {
+      "epoch": 0.16207002054285588,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.368578815383715e-05,
+      "loss": 1.9233,
+      "step": 3708
+    },
+    {
+      "epoch": 0.1621137287468858,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.368244656735798e-05,
+      "loss": 1.6875,
+      "step": 3709
+    },
+    {
+      "epoch": 0.1621574369509157,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.367910415652745e-05,
+      "loss": 2.2647,
+      "step": 3710
+    },
+    {
+      "epoch": 0.16220114515494558,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.367576092140866e-05,
+      "loss": 2.1075,
+      "step": 3711
+    },
+    {
+      "epoch": 0.16224485335897548,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.367241686206469e-05,
+      "loss": 1.6637,
+      "step": 3712
+    },
+    {
+      "epoch": 0.16228856156300536,
+      "grad_norm": 4.84375,
+      "learning_rate": 9.366907197855868e-05,
+      "loss": 2.8638,
+      "step": 3713
+    },
+    {
+      "epoch": 0.16233226976703527,
+      "grad_norm": 3.84375,
+      "learning_rate": 9.366572627095369e-05,
+      "loss": 1.9139,
+      "step": 3714
+    },
+    {
+      "epoch": 0.16237597797106518,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.366237973931291e-05,
+      "loss": 2.1883,
+      "step": 3715
+    },
+    {
+      "epoch": 0.16241968617509506,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.365903238369946e-05,
+      "loss": 1.3408,
+      "step": 3716
+    },
+    {
+      "epoch": 0.16246339437912496,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.365568420417655e-05,
+      "loss": 2.2424,
+      "step": 3717
+    },
+    {
+      "epoch": 0.16250710258315484,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.365233520080731e-05,
+      "loss": 2.0933,
+      "step": 3718
+    },
+    {
+      "epoch": 0.16255081078718475,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.364898537365501e-05,
+      "loss": 1.9398,
+      "step": 3719
+    },
+    {
+      "epoch": 0.16259451899121466,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.36456347227828e-05,
+      "loss": 1.7613,
+      "step": 3720
+    },
+    {
+      "epoch": 0.16263822719524454,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.364228324825395e-05,
+      "loss": 1.9097,
+      "step": 3721
+    },
+    {
+      "epoch": 0.16268193539927445,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.36389309501317e-05,
+      "loss": 2.2346,
+      "step": 3722
+    },
+    {
+      "epoch": 0.16272564360330435,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.36355778284793e-05,
+      "loss": 2.4336,
+      "step": 3723
+    },
+    {
+      "epoch": 0.16276935180733423,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.363222388336004e-05,
+      "loss": 2.0523,
+      "step": 3724
+    },
+    {
+      "epoch": 0.16281306001136414,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.362886911483722e-05,
+      "loss": 1.5397,
+      "step": 3725
+    },
+    {
+      "epoch": 0.16285676821539402,
+      "grad_norm": 5.84375,
+      "learning_rate": 9.362551352297413e-05,
+      "loss": 1.9769,
+      "step": 3726
+    },
+    {
+      "epoch": 0.16290047641942393,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.362215710783411e-05,
+      "loss": 1.7988,
+      "step": 3727
+    },
+    {
+      "epoch": 0.16294418462345384,
+      "grad_norm": 1.8984375,
+      "learning_rate": 9.36187998694805e-05,
+      "loss": 1.6174,
+      "step": 3728
+    },
+    {
+      "epoch": 0.16298789282748372,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.361544180797665e-05,
+      "loss": 2.2354,
+      "step": 3729
+    },
+    {
+      "epoch": 0.16303160103151362,
+      "grad_norm": 2.125,
+      "learning_rate": 9.361208292338593e-05,
+      "loss": 2.0043,
+      "step": 3730
+    },
+    {
+      "epoch": 0.1630753092355435,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.360872321577174e-05,
+      "loss": 2.1913,
+      "step": 3731
+    },
+    {
+      "epoch": 0.1631190174395734,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.360536268519746e-05,
+      "loss": 1.9443,
+      "step": 3732
+    },
+    {
+      "epoch": 0.16316272564360332,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.360200133172655e-05,
+      "loss": 2.0476,
+      "step": 3733
+    },
+    {
+      "epoch": 0.1632064338476332,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.359863915542238e-05,
+      "loss": 1.656,
+      "step": 3734
+    },
+    {
+      "epoch": 0.1632501420516631,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.359527615634844e-05,
+      "loss": 1.7408,
+      "step": 3735
+    },
+    {
+      "epoch": 0.16329385025569298,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.359191233456821e-05,
+      "loss": 1.8783,
+      "step": 3736
+    },
+    {
+      "epoch": 0.1633375584597229,
+      "grad_norm": 4.53125,
+      "learning_rate": 9.358854769014513e-05,
+      "loss": 2.5117,
+      "step": 3737
+    },
+    {
+      "epoch": 0.1633812666637528,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.358518222314272e-05,
+      "loss": 2.0793,
+      "step": 3738
+    },
+    {
+      "epoch": 0.16342497486778268,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.35818159336245e-05,
+      "loss": 1.7235,
+      "step": 3739
+    },
+    {
+      "epoch": 0.1634686830718126,
+      "grad_norm": 2.375,
+      "learning_rate": 9.357844882165396e-05,
+      "loss": 2.2602,
+      "step": 3740
+    },
+    {
+      "epoch": 0.16351239127584247,
+      "grad_norm": 2.75,
+      "learning_rate": 9.357508088729468e-05,
+      "loss": 1.9349,
+      "step": 3741
+    },
+    {
+      "epoch": 0.16355609947987237,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.357171213061021e-05,
+      "loss": 1.54,
+      "step": 3742
+    },
+    {
+      "epoch": 0.16359980768390228,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.356834255166409e-05,
+      "loss": 2.17,
+      "step": 3743
+    },
+    {
+      "epoch": 0.16364351588793216,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.356497215051996e-05,
+      "loss": 2.0412,
+      "step": 3744
+    },
+    {
+      "epoch": 0.16368722409196207,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.356160092724138e-05,
+      "loss": 2.0443,
+      "step": 3745
+    },
+    {
+      "epoch": 0.16373093229599195,
+      "grad_norm": 5.4375,
+      "learning_rate": 9.355822888189201e-05,
+      "loss": 2.7347,
+      "step": 3746
+    },
+    {
+      "epoch": 0.16377464050002186,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.355485601453545e-05,
+      "loss": 2.0591,
+      "step": 3747
+    },
+    {
+      "epoch": 0.16381834870405176,
+      "grad_norm": 2.75,
+      "learning_rate": 9.355148232523537e-05,
+      "loss": 1.9002,
+      "step": 3748
+    },
+    {
+      "epoch": 0.16386205690808164,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.354810781405543e-05,
+      "loss": 1.8313,
+      "step": 3749
+    },
+    {
+      "epoch": 0.16390576511211155,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.354473248105932e-05,
+      "loss": 1.8867,
+      "step": 3750
+    },
+    {
+      "epoch": 0.16394947331614143,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.354135632631073e-05,
+      "loss": 2.0169,
+      "step": 3751
+    },
+    {
+      "epoch": 0.16399318152017134,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.353797934987338e-05,
+      "loss": 1.8563,
+      "step": 3752
+    },
+    {
+      "epoch": 0.16403688972420125,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.353460155181098e-05,
+      "loss": 3.3005,
+      "step": 3753
+    },
+    {
+      "epoch": 0.16408059792823113,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.35312229321873e-05,
+      "loss": 1.9956,
+      "step": 3754
+    },
+    {
+      "epoch": 0.16412430613226103,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.352784349106608e-05,
+      "loss": 1.6531,
+      "step": 3755
+    },
+    {
+      "epoch": 0.1641680143362909,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.35244632285111e-05,
+      "loss": 1.8427,
+      "step": 3756
+    },
+    {
+      "epoch": 0.16421172254032082,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.352108214458616e-05,
+      "loss": 2.2344,
+      "step": 3757
+    },
+    {
+      "epoch": 0.16425543074435073,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.351770023935506e-05,
+      "loss": 1.8037,
+      "step": 3758
+    },
+    {
+      "epoch": 0.1642991389483806,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.35143175128816e-05,
+      "loss": 1.6607,
+      "step": 3759
+    },
+    {
+      "epoch": 0.16434284715241051,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.351093396522965e-05,
+      "loss": 2.2265,
+      "step": 3760
+    },
+    {
+      "epoch": 0.1643865553564404,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.350754959646306e-05,
+      "loss": 1.9707,
+      "step": 3761
+    },
+    {
+      "epoch": 0.1644302635604703,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.350416440664566e-05,
+      "loss": 2.1641,
+      "step": 3762
+    },
+    {
+      "epoch": 0.1644739717645002,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.350077839584138e-05,
+      "loss": 2.1195,
+      "step": 3763
+    },
+    {
+      "epoch": 0.1645176799685301,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.34973915641141e-05,
+      "loss": 2.0303,
+      "step": 3764
+    },
+    {
+      "epoch": 0.16456138817256,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.349400391152773e-05,
+      "loss": 2.0424,
+      "step": 3765
+    },
+    {
+      "epoch": 0.16460509637658988,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.34906154381462e-05,
+      "loss": 1.924,
+      "step": 3766
+    },
+    {
+      "epoch": 0.16464880458061978,
+      "grad_norm": 2.25,
+      "learning_rate": 9.348722614403345e-05,
+      "loss": 2.0095,
+      "step": 3767
+    },
+    {
+      "epoch": 0.1646925127846497,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.348383602925347e-05,
+      "loss": 1.8124,
+      "step": 3768
+    },
+    {
+      "epoch": 0.16473622098867957,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.34804450938702e-05,
+      "loss": 1.9358,
+      "step": 3769
+    },
+    {
+      "epoch": 0.16477992919270948,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.347705333794767e-05,
+      "loss": 2.5346,
+      "step": 3770
+    },
+    {
+      "epoch": 0.16482363739673936,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.347366076154984e-05,
+      "loss": 1.8833,
+      "step": 3771
+    },
+    {
+      "epoch": 0.16486734560076927,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.347026736474078e-05,
+      "loss": 1.9519,
+      "step": 3772
+    },
+    {
+      "epoch": 0.16491105380479917,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.346687314758448e-05,
+      "loss": 2.0777,
+      "step": 3773
+    },
+    {
+      "epoch": 0.16495476200882905,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.346347811014504e-05,
+      "loss": 2.6569,
+      "step": 3774
+    },
+    {
+      "epoch": 0.16499847021285896,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.346008225248651e-05,
+      "loss": 2.0684,
+      "step": 3775
+    },
+    {
+      "epoch": 0.16504217841688884,
+      "grad_norm": 2.25,
+      "learning_rate": 9.345668557467298e-05,
+      "loss": 2.0878,
+      "step": 3776
+    },
+    {
+      "epoch": 0.16508588662091875,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.345328807676854e-05,
+      "loss": 2.1552,
+      "step": 3777
+    },
+    {
+      "epoch": 0.16512959482494866,
+      "grad_norm": 2.625,
+      "learning_rate": 9.34498897588373e-05,
+      "loss": 1.7176,
+      "step": 3778
+    },
+    {
+      "epoch": 0.16517330302897854,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.344649062094342e-05,
+      "loss": 2.4368,
+      "step": 3779
+    },
+    {
+      "epoch": 0.16521701123300844,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.344309066315101e-05,
+      "loss": 2.6929,
+      "step": 3780
+    },
+    {
+      "epoch": 0.16526071943703832,
+      "grad_norm": 2.375,
+      "learning_rate": 9.343968988552426e-05,
+      "loss": 2.2669,
+      "step": 3781
+    },
+    {
+      "epoch": 0.16530442764106823,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.343628828812734e-05,
+      "loss": 2.6719,
+      "step": 3782
+    },
+    {
+      "epoch": 0.16534813584509814,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.343288587102443e-05,
+      "loss": 2.3055,
+      "step": 3783
+    },
+    {
+      "epoch": 0.16539184404912802,
+      "grad_norm": 11.5625,
+      "learning_rate": 9.342948263427977e-05,
+      "loss": 5.9702,
+      "step": 3784
+    },
+    {
+      "epoch": 0.16543555225315792,
+      "grad_norm": 3.828125,
+      "learning_rate": 9.342607857795756e-05,
+      "loss": 2.4608,
+      "step": 3785
+    },
+    {
+      "epoch": 0.1654792604571878,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.342267370212203e-05,
+      "loss": 2.3413,
+      "step": 3786
+    },
+    {
+      "epoch": 0.1655229686612177,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.341926800683745e-05,
+      "loss": 1.8671,
+      "step": 3787
+    },
+    {
+      "epoch": 0.16556667686524762,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.34158614921681e-05,
+      "loss": 2.4157,
+      "step": 3788
+    },
+    {
+      "epoch": 0.1656103850692775,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.341245415817825e-05,
+      "loss": 2.5063,
+      "step": 3789
+    },
+    {
+      "epoch": 0.1656540932733074,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.34090460049322e-05,
+      "loss": 1.7339,
+      "step": 3790
+    },
+    {
+      "epoch": 0.1656978014773373,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.340563703249428e-05,
+      "loss": 2.6966,
+      "step": 3791
+    },
+    {
+      "epoch": 0.1657415096813672,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.34022272409288e-05,
+      "loss": 2.1976,
+      "step": 3792
+    },
+    {
+      "epoch": 0.1657852178853971,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.339881663030014e-05,
+      "loss": 1.7816,
+      "step": 3793
+    },
+    {
+      "epoch": 0.16582892608942698,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.339540520067265e-05,
+      "loss": 1.6969,
+      "step": 3794
+    },
+    {
+      "epoch": 0.1658726342934569,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.33919929521107e-05,
+      "loss": 1.4362,
+      "step": 3795
+    },
+    {
+      "epoch": 0.16591634249748677,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.338857988467868e-05,
+      "loss": 1.5953,
+      "step": 3796
+    },
+    {
+      "epoch": 0.16596005070151668,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.338516599844101e-05,
+      "loss": 2.5578,
+      "step": 3797
+    },
+    {
+      "epoch": 0.16600375890554658,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.338175129346213e-05,
+      "loss": 2.6275,
+      "step": 3798
+    },
+    {
+      "epoch": 0.16604746710957646,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.337833576980646e-05,
+      "loss": 2.0663,
+      "step": 3799
+    },
+    {
+      "epoch": 0.16609117531360637,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.337491942753845e-05,
+      "loss": 2.4615,
+      "step": 3800
+    },
+    {
+      "epoch": 0.16613488351763625,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.337150226672258e-05,
+      "loss": 1.79,
+      "step": 3801
+    },
+    {
+      "epoch": 0.16617859172166616,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.336808428742335e-05,
+      "loss": 2.3343,
+      "step": 3802
+    },
+    {
+      "epoch": 0.16622229992569607,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.336466548970522e-05,
+      "loss": 2.0193,
+      "step": 3803
+    },
+    {
+      "epoch": 0.16626600812972595,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.336124587363278e-05,
+      "loss": 1.7385,
+      "step": 3804
+    },
+    {
+      "epoch": 0.16630971633375585,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.335782543927049e-05,
+      "loss": 1.5442,
+      "step": 3805
+    },
+    {
+      "epoch": 0.16635342453778573,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.335440418668294e-05,
+      "loss": 2.1318,
+      "step": 3806
+    },
+    {
+      "epoch": 0.16639713274181564,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.335098211593469e-05,
+      "loss": 1.9159,
+      "step": 3807
+    },
+    {
+      "epoch": 0.16644084094584555,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.334755922709031e-05,
+      "loss": 1.7075,
+      "step": 3808
+    },
+    {
+      "epoch": 0.16648454914987543,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.334413552021439e-05,
+      "loss": 2.5683,
+      "step": 3809
+    },
+    {
+      "epoch": 0.16652825735390533,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.334071099537154e-05,
+      "loss": 1.9669,
+      "step": 3810
+    },
+    {
+      "epoch": 0.16657196555793521,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.333728565262642e-05,
+      "loss": 2.0282,
+      "step": 3811
+    },
+    {
+      "epoch": 0.16661567376196512,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.333385949204363e-05,
+      "loss": 1.8504,
+      "step": 3812
+    },
+    {
+      "epoch": 0.16665938196599503,
+      "grad_norm": 2.125,
+      "learning_rate": 9.333043251368784e-05,
+      "loss": 1.7398,
+      "step": 3813
+    },
+    {
+      "epoch": 0.1667030901700249,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.332700471762374e-05,
+      "loss": 1.6821,
+      "step": 3814
+    },
+    {
+      "epoch": 0.16674679837405482,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.332357610391598e-05,
+      "loss": 2.1855,
+      "step": 3815
+    },
+    {
+      "epoch": 0.1667905065780847,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.332014667262929e-05,
+      "loss": 2.1317,
+      "step": 3816
+    },
+    {
+      "epoch": 0.1668342147821146,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.33167164238284e-05,
+      "loss": 1.906,
+      "step": 3817
+    },
+    {
+      "epoch": 0.1668779229861445,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.331328535757801e-05,
+      "loss": 1.8553,
+      "step": 3818
+    },
+    {
+      "epoch": 0.1669216311901744,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.33098534739429e-05,
+      "loss": 1.9632,
+      "step": 3819
+    },
+    {
+      "epoch": 0.1669653393942043,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.33064207729878e-05,
+      "loss": 2.1928,
+      "step": 3820
+    },
+    {
+      "epoch": 0.16700904759823418,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.330298725477753e-05,
+      "loss": 2.4237,
+      "step": 3821
+    },
+    {
+      "epoch": 0.16705275580226409,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.329955291937684e-05,
+      "loss": 3.0202,
+      "step": 3822
+    },
+    {
+      "epoch": 0.167096464006294,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.329611776685059e-05,
+      "loss": 2.1526,
+      "step": 3823
+    },
+    {
+      "epoch": 0.16714017221032387,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.329268179726359e-05,
+      "loss": 2.0835,
+      "step": 3824
+    },
+    {
+      "epoch": 0.16718388041435378,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.328924501068066e-05,
+      "loss": 1.9372,
+      "step": 3825
+    },
+    {
+      "epoch": 0.16722758861838366,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.328580740716666e-05,
+      "loss": 2.1097,
+      "step": 3826
+    },
+    {
+      "epoch": 0.16727129682241357,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.32823689867865e-05,
+      "loss": 1.9521,
+      "step": 3827
+    },
+    {
+      "epoch": 0.16731500502644348,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.327892974960503e-05,
+      "loss": 1.81,
+      "step": 3828
+    },
+    {
+      "epoch": 0.16735871323047335,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.327548969568716e-05,
+      "loss": 1.8508,
+      "step": 3829
+    },
+    {
+      "epoch": 0.16740242143450326,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.327204882509784e-05,
+      "loss": 3.2697,
+      "step": 3830
+    },
+    {
+      "epoch": 0.16744612963853314,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.326860713790195e-05,
+      "loss": 1.9622,
+      "step": 3831
+    },
+    {
+      "epoch": 0.16748983784256305,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.326516463416448e-05,
+      "loss": 2.7874,
+      "step": 3832
+    },
+    {
+      "epoch": 0.16753354604659296,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.32617213139504e-05,
+      "loss": 1.9411,
+      "step": 3833
+    },
+    {
+      "epoch": 0.16757725425062284,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.325827717732465e-05,
+      "loss": 2.0798,
+      "step": 3834
+    },
+    {
+      "epoch": 0.16762096245465274,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.325483222435226e-05,
+      "loss": 2.3578,
+      "step": 3835
+    },
+    {
+      "epoch": 0.16766467065868262,
+      "grad_norm": 2.25,
+      "learning_rate": 9.325138645509823e-05,
+      "loss": 1.6908,
+      "step": 3836
+    },
+    {
+      "epoch": 0.16770837886271253,
+      "grad_norm": 2.25,
+      "learning_rate": 9.324793986962758e-05,
+      "loss": 1.8728,
+      "step": 3837
+    },
+    {
+      "epoch": 0.16775208706674244,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.324449246800537e-05,
+      "loss": 1.8805,
+      "step": 3838
+    },
+    {
+      "epoch": 0.16779579527077232,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.324104425029665e-05,
+      "loss": 1.7632,
+      "step": 3839
+    },
+    {
+      "epoch": 0.16783950347480223,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.32375952165665e-05,
+      "loss": 2.3108,
+      "step": 3840
+    },
+    {
+      "epoch": 0.1678832116788321,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.323414536687997e-05,
+      "loss": 2.0386,
+      "step": 3841
+    },
+    {
+      "epoch": 0.167926919882862,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.323069470130221e-05,
+      "loss": 2.6539,
+      "step": 3842
+    },
+    {
+      "epoch": 0.16797062808689192,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.322724321989833e-05,
+      "loss": 2.3568,
+      "step": 3843
+    },
+    {
+      "epoch": 0.1680143362909218,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.322379092273345e-05,
+      "loss": 2.3111,
+      "step": 3844
+    },
+    {
+      "epoch": 0.1680580444949517,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.322033780987272e-05,
+      "loss": 1.8544,
+      "step": 3845
+    },
+    {
+      "epoch": 0.1681017526989816,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.321688388138132e-05,
+      "loss": 2.1158,
+      "step": 3846
+    },
+    {
+      "epoch": 0.1681454609030115,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.321342913732441e-05,
+      "loss": 2.156,
+      "step": 3847
+    },
+    {
+      "epoch": 0.1681891691070414,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.32099735777672e-05,
+      "loss": 1.8919,
+      "step": 3848
+    },
+    {
+      "epoch": 0.16823287731107128,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.320651720277491e-05,
+      "loss": 1.758,
+      "step": 3849
+    },
+    {
+      "epoch": 0.1682765855151012,
+      "grad_norm": 3.125,
+      "learning_rate": 9.320306001241275e-05,
+      "loss": 2.3838,
+      "step": 3850
+    },
+    {
+      "epoch": 0.16832029371913107,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.319960200674597e-05,
+      "loss": 2.0288,
+      "step": 3851
+    },
+    {
+      "epoch": 0.16836400192316098,
+      "grad_norm": 2.25,
+      "learning_rate": 9.319614318583982e-05,
+      "loss": 2.3708,
+      "step": 3852
+    },
+    {
+      "epoch": 0.16840771012719088,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.319268354975959e-05,
+      "loss": 2.506,
+      "step": 3853
+    },
+    {
+      "epoch": 0.16845141833122076,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.318922309857055e-05,
+      "loss": 1.9185,
+      "step": 3854
+    },
+    {
+      "epoch": 0.16849512653525067,
+      "grad_norm": 2.75,
+      "learning_rate": 9.3185761832338e-05,
+      "loss": 2.2522,
+      "step": 3855
+    },
+    {
+      "epoch": 0.16853883473928055,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.318229975112728e-05,
+      "loss": 2.6273,
+      "step": 3856
+    },
+    {
+      "epoch": 0.16858254294331046,
+      "grad_norm": 2.375,
+      "learning_rate": 9.317883685500373e-05,
+      "loss": 2.1665,
+      "step": 3857
+    },
+    {
+      "epoch": 0.16862625114734037,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.317537314403267e-05,
+      "loss": 1.6587,
+      "step": 3858
+    },
+    {
+      "epoch": 0.16866995935137025,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.317190861827949e-05,
+      "loss": 1.9373,
+      "step": 3859
+    },
+    {
+      "epoch": 0.16871366755540015,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.316844327780955e-05,
+      "loss": 1.7105,
+      "step": 3860
+    },
+    {
+      "epoch": 0.16875737575943003,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.316497712268825e-05,
+      "loss": 1.3369,
+      "step": 3861
+    },
+    {
+      "epoch": 0.16880108396345994,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.316151015298103e-05,
+      "loss": 1.8242,
+      "step": 3862
+    },
+    {
+      "epoch": 0.16884479216748985,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.315804236875327e-05,
+      "loss": 1.7831,
+      "step": 3863
+    },
+    {
+      "epoch": 0.16888850037151973,
+      "grad_norm": 2.625,
+      "learning_rate": 9.315457377007046e-05,
+      "loss": 2.3505,
+      "step": 3864
+    },
+    {
+      "epoch": 0.16893220857554964,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.315110435699804e-05,
+      "loss": 2.0785,
+      "step": 3865
+    },
+    {
+      "epoch": 0.16897591677957952,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.314763412960144e-05,
+      "loss": 1.6823,
+      "step": 3866
+    },
+    {
+      "epoch": 0.16901962498360942,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.314416308794621e-05,
+      "loss": 2.4492,
+      "step": 3867
+    },
+    {
+      "epoch": 0.16906333318763933,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.314069123209784e-05,
+      "loss": 1.8297,
+      "step": 3868
+    },
+    {
+      "epoch": 0.1691070413916692,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.313721856212181e-05,
+      "loss": 2.1315,
+      "step": 3869
+    },
+    {
+      "epoch": 0.16915074959569912,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.313374507808371e-05,
+      "loss": 2.3644,
+      "step": 3870
+    },
+    {
+      "epoch": 0.169194457799729,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.313027078004903e-05,
+      "loss": 2.1034,
+      "step": 3871
+    },
+    {
+      "epoch": 0.1692381660037589,
+      "grad_norm": 2.625,
+      "learning_rate": 9.31267956680834e-05,
+      "loss": 1.766,
+      "step": 3872
+    },
+    {
+      "epoch": 0.1692818742077888,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.312331974225235e-05,
+      "loss": 2.1167,
+      "step": 3873
+    },
+    {
+      "epoch": 0.1693255824118187,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.31198430026215e-05,
+      "loss": 2.5551,
+      "step": 3874
+    },
+    {
+      "epoch": 0.1693692906158486,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.311636544925645e-05,
+      "loss": 1.8521,
+      "step": 3875
+    },
+    {
+      "epoch": 0.16941299881987848,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.311288708222284e-05,
+      "loss": 2.3467,
+      "step": 3876
+    },
+    {
+      "epoch": 0.1694567070239084,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.310940790158629e-05,
+      "loss": 1.8095,
+      "step": 3877
+    },
+    {
+      "epoch": 0.1695004152279383,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.310592790741248e-05,
+      "loss": 2.0915,
+      "step": 3878
+    },
+    {
+      "epoch": 0.16954412343196817,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.310244709976707e-05,
+      "loss": 1.654,
+      "step": 3879
+    },
+    {
+      "epoch": 0.16958783163599808,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.309896547871576e-05,
+      "loss": 2.3766,
+      "step": 3880
+    },
+    {
+      "epoch": 0.16963153984002796,
+      "grad_norm": 3.5,
+      "learning_rate": 9.309548304432421e-05,
+      "loss": 2.7146,
+      "step": 3881
+    },
+    {
+      "epoch": 0.16967524804405787,
+      "grad_norm": 3.125,
+      "learning_rate": 9.30919997966582e-05,
+      "loss": 1.8983,
+      "step": 3882
+    },
+    {
+      "epoch": 0.16971895624808778,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.308851573578344e-05,
+      "loss": 1.8319,
+      "step": 3883
+    },
+    {
+      "epoch": 0.16976266445211766,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.308503086176565e-05,
+      "loss": 1.8928,
+      "step": 3884
+    },
+    {
+      "epoch": 0.16980637265614756,
+      "grad_norm": 3.765625,
+      "learning_rate": 9.308154517467065e-05,
+      "loss": 1.9515,
+      "step": 3885
+    },
+    {
+      "epoch": 0.16985008086017744,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.307805867456418e-05,
+      "loss": 2.2746,
+      "step": 3886
+    },
+    {
+      "epoch": 0.16989378906420735,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.307457136151204e-05,
+      "loss": 1.6149,
+      "step": 3887
+    },
+    {
+      "epoch": 0.16993749726823726,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.307108323558005e-05,
+      "loss": 1.6674,
+      "step": 3888
+    },
+    {
+      "epoch": 0.16998120547226714,
+      "grad_norm": 4.0625,
+      "learning_rate": 9.306759429683404e-05,
+      "loss": 1.9972,
+      "step": 3889
+    },
+    {
+      "epoch": 0.17002491367629705,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.306410454533982e-05,
+      "loss": 2.2144,
+      "step": 3890
+    },
+    {
+      "epoch": 0.17006862188032693,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.306061398116331e-05,
+      "loss": 1.7609,
+      "step": 3891
+    },
+    {
+      "epoch": 0.17011233008435683,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.305712260437031e-05,
+      "loss": 2.6109,
+      "step": 3892
+    },
+    {
+      "epoch": 0.17015603828838674,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.305363041502675e-05,
+      "loss": 2.5016,
+      "step": 3893
+    },
+    {
+      "epoch": 0.17019974649241662,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.305013741319852e-05,
+      "loss": 2.1102,
+      "step": 3894
+    },
+    {
+      "epoch": 0.17024345469644653,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.304664359895155e-05,
+      "loss": 2.5799,
+      "step": 3895
+    },
+    {
+      "epoch": 0.1702871629004764,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.304314897235176e-05,
+      "loss": 1.9134,
+      "step": 3896
+    },
+    {
+      "epoch": 0.17033087110450632,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.303965353346508e-05,
+      "loss": 2.659,
+      "step": 3897
+    },
+    {
+      "epoch": 0.17037457930853622,
+      "grad_norm": 2.5,
+      "learning_rate": 9.303615728235753e-05,
+      "loss": 2.0285,
+      "step": 3898
+    },
+    {
+      "epoch": 0.1704182875125661,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.303266021909504e-05,
+      "loss": 2.3058,
+      "step": 3899
+    },
+    {
+      "epoch": 0.170461995716596,
+      "grad_norm": 2.625,
+      "learning_rate": 9.302916234374361e-05,
+      "loss": 2.6567,
+      "step": 3900
+    },
+    {
+      "epoch": 0.1705057039206259,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.302566365636928e-05,
+      "loss": 2.077,
+      "step": 3901
+    },
+    {
+      "epoch": 0.1705494121246558,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.302216415703805e-05,
+      "loss": 2.523,
+      "step": 3902
+    },
+    {
+      "epoch": 0.1705931203286857,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.301866384581597e-05,
+      "loss": 1.9372,
+      "step": 3903
+    },
+    {
+      "epoch": 0.17063682853271558,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.301516272276907e-05,
+      "loss": 2.5272,
+      "step": 3904
+    },
+    {
+      "epoch": 0.1706805367367455,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.301166078796347e-05,
+      "loss": 1.977,
+      "step": 3905
+    },
+    {
+      "epoch": 0.17072424494077537,
+      "grad_norm": 2.625,
+      "learning_rate": 9.300815804146522e-05,
+      "loss": 2.5269,
+      "step": 3906
+    },
+    {
+      "epoch": 0.17076795314480528,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.300465448334044e-05,
+      "loss": 2.3774,
+      "step": 3907
+    },
+    {
+      "epoch": 0.1708116613488352,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.300115011365522e-05,
+      "loss": 2.1276,
+      "step": 3908
+    },
+    {
+      "epoch": 0.17085536955286507,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.299764493247574e-05,
+      "loss": 1.9362,
+      "step": 3909
+    },
+    {
+      "epoch": 0.17089907775689497,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.29941389398681e-05,
+      "loss": 2.0708,
+      "step": 3910
+    },
+    {
+      "epoch": 0.17094278596092485,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.299063213589849e-05,
+      "loss": 2.2451,
+      "step": 3911
+    },
+    {
+      "epoch": 0.17098649416495476,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.298712452063309e-05,
+      "loss": 1.7547,
+      "step": 3912
+    },
+    {
+      "epoch": 0.17103020236898467,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.298361609413805e-05,
+      "loss": 2.5509,
+      "step": 3913
+    },
+    {
+      "epoch": 0.17107391057301455,
+      "grad_norm": 2.75,
+      "learning_rate": 9.298010685647966e-05,
+      "loss": 1.9587,
+      "step": 3914
+    },
+    {
+      "epoch": 0.17111761877704446,
+      "grad_norm": 2.125,
+      "learning_rate": 9.297659680772408e-05,
+      "loss": 1.6686,
+      "step": 3915
+    },
+    {
+      "epoch": 0.17116132698107434,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.297308594793756e-05,
+      "loss": 2.1915,
+      "step": 3916
+    },
+    {
+      "epoch": 0.17120503518510424,
+      "grad_norm": 2.125,
+      "learning_rate": 9.296957427718638e-05,
+      "loss": 2.1469,
+      "step": 3917
+    },
+    {
+      "epoch": 0.17124874338913415,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.296606179553679e-05,
+      "loss": 2.1142,
+      "step": 3918
+    },
+    {
+      "epoch": 0.17129245159316403,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.296254850305506e-05,
+      "loss": 1.9486,
+      "step": 3919
+    },
+    {
+      "epoch": 0.17133615979719394,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.295903439980755e-05,
+      "loss": 1.8862,
+      "step": 3920
+    },
+    {
+      "epoch": 0.17137986800122382,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.29555194858605e-05,
+      "loss": 3.0025,
+      "step": 3921
+    },
+    {
+      "epoch": 0.17142357620525372,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.295200376128031e-05,
+      "loss": 2.0601,
+      "step": 3922
+    },
+    {
+      "epoch": 0.17146728440928363,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.294848722613326e-05,
+      "loss": 1.4115,
+      "step": 3923
+    },
+    {
+      "epoch": 0.1715109926133135,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.294496988048578e-05,
+      "loss": 1.6362,
+      "step": 3924
+    },
+    {
+      "epoch": 0.17155470081734342,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.29414517244042e-05,
+      "loss": 1.8668,
+      "step": 3925
+    },
+    {
+      "epoch": 0.1715984090213733,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.293793275795492e-05,
+      "loss": 1.9503,
+      "step": 3926
+    },
+    {
+      "epoch": 0.1716421172254032,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.293441298120436e-05,
+      "loss": 1.7978,
+      "step": 3927
+    },
+    {
+      "epoch": 0.17168582542943311,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.293089239421895e-05,
+      "loss": 2.4183,
+      "step": 3928
+    },
+    {
+      "epoch": 0.171729533633463,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.29273709970651e-05,
+      "loss": 1.8488,
+      "step": 3929
+    },
+    {
+      "epoch": 0.1717732418374929,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.29238487898093e-05,
+      "loss": 2.9994,
+      "step": 3930
+    },
+    {
+      "epoch": 0.17181695004152278,
+      "grad_norm": 7.125,
+      "learning_rate": 9.2920325772518e-05,
+      "loss": 2.1471,
+      "step": 3931
+    },
+    {
+      "epoch": 0.1718606582455527,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.291680194525767e-05,
+      "loss": 1.7098,
+      "step": 3932
+    },
+    {
+      "epoch": 0.1719043664495826,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.291327730809483e-05,
+      "loss": 1.9738,
+      "step": 3933
+    },
+    {
+      "epoch": 0.17194807465361248,
+      "grad_norm": 2.875,
+      "learning_rate": 9.2909751861096e-05,
+      "loss": 1.9738,
+      "step": 3934
+    },
+    {
+      "epoch": 0.17199178285764238,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.29062256043277e-05,
+      "loss": 2.3958,
+      "step": 3935
+    },
+    {
+      "epoch": 0.17203549106167226,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.290269853785645e-05,
+      "loss": 2.3367,
+      "step": 3936
+    },
+    {
+      "epoch": 0.17207919926570217,
+      "grad_norm": 5.375,
+      "learning_rate": 9.289917066174886e-05,
+      "loss": 2.0566,
+      "step": 3937
+    },
+    {
+      "epoch": 0.17212290746973208,
+      "grad_norm": 5.46875,
+      "learning_rate": 9.289564197607148e-05,
+      "loss": 1.1127,
+      "step": 3938
+    },
+    {
+      "epoch": 0.17216661567376196,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.28921124808909e-05,
+      "loss": 2.359,
+      "step": 3939
+    },
+    {
+      "epoch": 0.17221032387779187,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.288858217627374e-05,
+      "loss": 3.0929,
+      "step": 3940
+    },
+    {
+      "epoch": 0.17225403208182175,
+      "grad_norm": 2.625,
+      "learning_rate": 9.28850510622866e-05,
+      "loss": 1.8901,
+      "step": 3941
+    },
+    {
+      "epoch": 0.17229774028585165,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.288151913899614e-05,
+      "loss": 1.8007,
+      "step": 3942
+    },
+    {
+      "epoch": 0.17234144848988156,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.287798640646898e-05,
+      "loss": 3.0978,
+      "step": 3943
+    },
+    {
+      "epoch": 0.17238515669391144,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.287445286477184e-05,
+      "loss": 2.0779,
+      "step": 3944
+    },
+    {
+      "epoch": 0.17242886489794135,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.287091851397137e-05,
+      "loss": 1.9458,
+      "step": 3945
+    },
+    {
+      "epoch": 0.17247257310197123,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.286738335413425e-05,
+      "loss": 1.902,
+      "step": 3946
+    },
+    {
+      "epoch": 0.17251628130600113,
+      "grad_norm": 3.0,
+      "learning_rate": 9.286384738532723e-05,
+      "loss": 2.942,
+      "step": 3947
+    },
+    {
+      "epoch": 0.17255998951003104,
+      "grad_norm": 4.1875,
+      "learning_rate": 9.286031060761703e-05,
+      "loss": 1.6749,
+      "step": 3948
+    },
+    {
+      "epoch": 0.17260369771406092,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.285677302107039e-05,
+      "loss": 2.1474,
+      "step": 3949
+    },
+    {
+      "epoch": 0.17264740591809083,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.285323462575406e-05,
+      "loss": 2.5917,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1726911141221207,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.284969542173482e-05,
+      "loss": 2.1621,
+      "step": 3951
+    },
+    {
+      "epoch": 0.17273482232615062,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.284615540907947e-05,
+      "loss": 2.9806,
+      "step": 3952
+    },
+    {
+      "epoch": 0.17277853053018052,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.28426145878548e-05,
+      "loss": 1.9889,
+      "step": 3953
+    },
+    {
+      "epoch": 0.1728222387342104,
+      "grad_norm": 2.875,
+      "learning_rate": 9.283907295812765e-05,
+      "loss": 2.4204,
+      "step": 3954
+    },
+    {
+      "epoch": 0.1728659469382403,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.283553051996483e-05,
+      "loss": 2.1639,
+      "step": 3955
+    },
+    {
+      "epoch": 0.1729096551422702,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.283198727343322e-05,
+      "loss": 3.293,
+      "step": 3956
+    },
+    {
+      "epoch": 0.1729533633463001,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.282844321859965e-05,
+      "loss": 2.632,
+      "step": 3957
+    },
+    {
+      "epoch": 0.17299707155033,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.282489835553106e-05,
+      "loss": 1.9452,
+      "step": 3958
+    },
+    {
+      "epoch": 0.17304077975435989,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.282135268429427e-05,
+      "loss": 1.9549,
+      "step": 3959
+    },
+    {
+      "epoch": 0.1730844879583898,
+      "grad_norm": 2.75,
+      "learning_rate": 9.281780620495624e-05,
+      "loss": 2.5024,
+      "step": 3960
+    },
+    {
+      "epoch": 0.17312819616241967,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.28142589175839e-05,
+      "loss": 3.4426,
+      "step": 3961
+    },
+    {
+      "epoch": 0.17317190436644958,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.281071082224418e-05,
+      "loss": 1.8448,
+      "step": 3962
+    },
+    {
+      "epoch": 0.1732156125704795,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.280716191900404e-05,
+      "loss": 1.6593,
+      "step": 3963
+    },
+    {
+      "epoch": 0.17325932077450937,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.280361220793044e-05,
+      "loss": 2.7787,
+      "step": 3964
+    },
+    {
+      "epoch": 0.17330302897853928,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.280006168909039e-05,
+      "loss": 2.3553,
+      "step": 3965
+    },
+    {
+      "epoch": 0.17334673718256915,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.279651036255088e-05,
+      "loss": 1.7812,
+      "step": 3966
+    },
+    {
+      "epoch": 0.17339044538659906,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.279295822837893e-05,
+      "loss": 2.0827,
+      "step": 3967
+    },
+    {
+      "epoch": 0.17343415359062897,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.278940528664158e-05,
+      "loss": 2.0041,
+      "step": 3968
+    },
+    {
+      "epoch": 0.17347786179465885,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.278585153740587e-05,
+      "loss": 2.1094,
+      "step": 3969
+    },
+    {
+      "epoch": 0.17352156999868876,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.278229698073888e-05,
+      "loss": 1.5974,
+      "step": 3970
+    },
+    {
+      "epoch": 0.17356527820271864,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.277874161670766e-05,
+      "loss": 2.0873,
+      "step": 3971
+    },
+    {
+      "epoch": 0.17360898640674854,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.277518544537934e-05,
+      "loss": 1.7993,
+      "step": 3972
+    },
+    {
+      "epoch": 0.17365269461077845,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.277162846682102e-05,
+      "loss": 1.8639,
+      "step": 3973
+    },
+    {
+      "epoch": 0.17369640281480833,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.276807068109981e-05,
+      "loss": 2.4488,
+      "step": 3974
+    },
+    {
+      "epoch": 0.17374011101883824,
+      "grad_norm": 2.125,
+      "learning_rate": 9.276451208828285e-05,
+      "loss": 1.8302,
+      "step": 3975
+    },
+    {
+      "epoch": 0.17378381922286812,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.276095268843732e-05,
+      "loss": 2.6344,
+      "step": 3976
+    },
+    {
+      "epoch": 0.17382752742689803,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.275739248163037e-05,
+      "loss": 2.0113,
+      "step": 3977
+    },
+    {
+      "epoch": 0.17387123563092793,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.27538314679292e-05,
+      "loss": 2.4077,
+      "step": 3978
+    },
+    {
+      "epoch": 0.1739149438349578,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.275026964740101e-05,
+      "loss": 1.8287,
+      "step": 3979
+    },
+    {
+      "epoch": 0.17395865203898772,
+      "grad_norm": 4.4375,
+      "learning_rate": 9.2746707020113e-05,
+      "loss": 1.3134,
+      "step": 3980
+    },
+    {
+      "epoch": 0.1740023602430176,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.274314358613241e-05,
+      "loss": 2.0166,
+      "step": 3981
+    },
+    {
+      "epoch": 0.1740460684470475,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.27395793455265e-05,
+      "loss": 2.2178,
+      "step": 3982
+    },
+    {
+      "epoch": 0.17408977665107742,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.273601429836253e-05,
+      "loss": 2.4367,
+      "step": 3983
+    },
+    {
+      "epoch": 0.1741334848551073,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.273244844470777e-05,
+      "loss": 2.2315,
+      "step": 3984
+    },
+    {
+      "epoch": 0.1741771930591372,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.27288817846295e-05,
+      "loss": 2.0803,
+      "step": 3985
+    },
+    {
+      "epoch": 0.17422090126316708,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.272531431819504e-05,
+      "loss": 2.5731,
+      "step": 3986
+    },
+    {
+      "epoch": 0.174264609467197,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.272174604547172e-05,
+      "loss": 2.1792,
+      "step": 3987
+    },
+    {
+      "epoch": 0.1743083176712269,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.271817696652688e-05,
+      "loss": 1.598,
+      "step": 3988
+    },
+    {
+      "epoch": 0.17435202587525678,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.271460708142787e-05,
+      "loss": 2.1333,
+      "step": 3989
+    },
+    {
+      "epoch": 0.17439573407928669,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.271103639024204e-05,
+      "loss": 2.2194,
+      "step": 3990
+    },
+    {
+      "epoch": 0.17443944228331656,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.27074648930368e-05,
+      "loss": 2.4873,
+      "step": 3991
+    },
+    {
+      "epoch": 0.17448315048734647,
+      "grad_norm": 2.375,
+      "learning_rate": 9.270389258987956e-05,
+      "loss": 1.8827,
+      "step": 3992
+    },
+    {
+      "epoch": 0.17452685869137638,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.270031948083769e-05,
+      "loss": 2.8426,
+      "step": 3993
+    },
+    {
+      "epoch": 0.17457056689540626,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.269674556597865e-05,
+      "loss": 2.0483,
+      "step": 3994
+    },
+    {
+      "epoch": 0.17461427509943617,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.269317084536988e-05,
+      "loss": 2.9369,
+      "step": 3995
+    },
+    {
+      "epoch": 0.17465798330346605,
+      "grad_norm": 2.0,
+      "learning_rate": 9.268959531907883e-05,
+      "loss": 1.7993,
+      "step": 3996
+    },
+    {
+      "epoch": 0.17470169150749595,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.268601898717299e-05,
+      "loss": 2.6471,
+      "step": 3997
+    },
+    {
+      "epoch": 0.17474539971152586,
+      "grad_norm": 2.0,
+      "learning_rate": 9.268244184971984e-05,
+      "loss": 1.9124,
+      "step": 3998
+    },
+    {
+      "epoch": 0.17478910791555574,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.267886390678691e-05,
+      "loss": 1.875,
+      "step": 3999
+    },
+    {
+      "epoch": 0.17483281611958565,
+      "grad_norm": 4.75,
+      "learning_rate": 9.267528515844168e-05,
+      "loss": 1.9821,
+      "step": 4000
+    },
+    {
+      "epoch": 0.17487652432361556,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.267170560475172e-05,
+      "loss": 1.5852,
+      "step": 4001
+    },
+    {
+      "epoch": 0.17492023252764544,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.266812524578457e-05,
+      "loss": 2.142,
+      "step": 4002
+    },
+    {
+      "epoch": 0.17496394073167534,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.266454408160779e-05,
+      "loss": 1.7585,
+      "step": 4003
+    },
+    {
+      "epoch": 0.17500764893570522,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.266096211228896e-05,
+      "loss": 2.1145,
+      "step": 4004
+    },
+    {
+      "epoch": 0.17505135713973513,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.265737933789571e-05,
+      "loss": 1.99,
+      "step": 4005
+    },
+    {
+      "epoch": 0.17509506534376504,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.26537957584956e-05,
+      "loss": 2.156,
+      "step": 4006
+    },
+    {
+      "epoch": 0.17513877354779492,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.26502113741563e-05,
+      "loss": 1.9717,
+      "step": 4007
+    },
+    {
+      "epoch": 0.17518248175182483,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.264662618494544e-05,
+      "loss": 2.2891,
+      "step": 4008
+    },
+    {
+      "epoch": 0.1752261899558547,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.264304019093066e-05,
+      "loss": 2.4795,
+      "step": 4009
+    },
+    {
+      "epoch": 0.1752698981598846,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.263945339217967e-05,
+      "loss": 2.0334,
+      "step": 4010
+    },
+    {
+      "epoch": 0.17531360636391452,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.263586578876011e-05,
+      "loss": 1.9267,
+      "step": 4011
+    },
+    {
+      "epoch": 0.1753573145679444,
+      "grad_norm": 1.8984375,
+      "learning_rate": 9.263227738073973e-05,
+      "loss": 1.7599,
+      "step": 4012
+    },
+    {
+      "epoch": 0.1754010227719743,
+      "grad_norm": 6.71875,
+      "learning_rate": 9.262868816818622e-05,
+      "loss": 2.0304,
+      "step": 4013
+    },
+    {
+      "epoch": 0.1754447309760042,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.262509815116732e-05,
+      "loss": 1.9665,
+      "step": 4014
+    },
+    {
+      "epoch": 0.1754884391800341,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.262150732975078e-05,
+      "loss": 2.1552,
+      "step": 4015
+    },
+    {
+      "epoch": 0.175532147384064,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.261791570400436e-05,
+      "loss": 2.0909,
+      "step": 4016
+    },
+    {
+      "epoch": 0.17557585558809388,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.261432327399583e-05,
+      "loss": 2.7963,
+      "step": 4017
+    },
+    {
+      "epoch": 0.1756195637921238,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.261073003979303e-05,
+      "loss": 1.868,
+      "step": 4018
+    },
+    {
+      "epoch": 0.17566327199615367,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.260713600146373e-05,
+      "loss": 2.179,
+      "step": 4019
+    },
+    {
+      "epoch": 0.17570698020018358,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.260354115907574e-05,
+      "loss": 2.1183,
+      "step": 4020
+    },
+    {
+      "epoch": 0.17575068840421348,
+      "grad_norm": 2.5,
+      "learning_rate": 9.259994551269694e-05,
+      "loss": 1.9912,
+      "step": 4021
+    },
+    {
+      "epoch": 0.17579439660824336,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.259634906239516e-05,
+      "loss": 1.9105,
+      "step": 4022
+    },
+    {
+      "epoch": 0.17583810481227327,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.259275180823829e-05,
+      "loss": 1.8473,
+      "step": 4023
+    },
+    {
+      "epoch": 0.17588181301630315,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.258915375029418e-05,
+      "loss": 1.7336,
+      "step": 4024
+    },
+    {
+      "epoch": 0.17592552122033306,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.258555488863078e-05,
+      "loss": 2.0827,
+      "step": 4025
+    },
+    {
+      "epoch": 0.17596922942436297,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.258195522331596e-05,
+      "loss": 2.2803,
+      "step": 4026
+    },
+    {
+      "epoch": 0.17601293762839285,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.257835475441768e-05,
+      "loss": 2.1713,
+      "step": 4027
+    },
+    {
+      "epoch": 0.17605664583242275,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.257475348200387e-05,
+      "loss": 2.4612,
+      "step": 4028
+    },
+    {
+      "epoch": 0.17610035403645263,
+      "grad_norm": 2.0,
+      "learning_rate": 9.25711514061425e-05,
+      "loss": 1.716,
+      "step": 4029
+    },
+    {
+      "epoch": 0.17614406224048254,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.256754852690152e-05,
+      "loss": 2.4852,
+      "step": 4030
+    },
+    {
+      "epoch": 0.17618777044451245,
+      "grad_norm": 3.703125,
+      "learning_rate": 9.256394484434899e-05,
+      "loss": 2.4831,
+      "step": 4031
+    },
+    {
+      "epoch": 0.17623147864854233,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.256034035855283e-05,
+      "loss": 2.3615,
+      "step": 4032
+    },
+    {
+      "epoch": 0.17627518685257224,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.255673506958113e-05,
+      "loss": 2.2626,
+      "step": 4033
+    },
+    {
+      "epoch": 0.17631889505660212,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.255312897750189e-05,
+      "loss": 1.6607,
+      "step": 4034
+    },
+    {
+      "epoch": 0.17636260326063202,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.254952208238318e-05,
+      "loss": 1.9265,
+      "step": 4035
+    },
+    {
+      "epoch": 0.17640631146466193,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.254591438429306e-05,
+      "loss": 2.1257,
+      "step": 4036
+    },
+    {
+      "epoch": 0.1764500196686918,
+      "grad_norm": 5.40625,
+      "learning_rate": 9.25423058832996e-05,
+      "loss": 2.5334,
+      "step": 4037
+    },
+    {
+      "epoch": 0.17649372787272172,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.253869657947092e-05,
+      "loss": 2.8139,
+      "step": 4038
+    },
+    {
+      "epoch": 0.1765374360767516,
+      "grad_norm": 2.125,
+      "learning_rate": 9.253508647287512e-05,
+      "loss": 1.7314,
+      "step": 4039
+    },
+    {
+      "epoch": 0.1765811442807815,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.253147556358034e-05,
+      "loss": 1.8548,
+      "step": 4040
+    },
+    {
+      "epoch": 0.1766248524848114,
+      "grad_norm": 2.375,
+      "learning_rate": 9.252786385165471e-05,
+      "loss": 2.5766,
+      "step": 4041
+    },
+    {
+      "epoch": 0.1766685606888413,
+      "grad_norm": 2.625,
+      "learning_rate": 9.25242513371664e-05,
+      "loss": 1.997,
+      "step": 4042
+    },
+    {
+      "epoch": 0.1767122688928712,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.252063802018356e-05,
+      "loss": 2.0024,
+      "step": 4043
+    },
+    {
+      "epoch": 0.17675597709690108,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.251702390077441e-05,
+      "loss": 2.189,
+      "step": 4044
+    },
+    {
+      "epoch": 0.176799685300931,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.251340897900713e-05,
+      "loss": 2.05,
+      "step": 4045
+    },
+    {
+      "epoch": 0.1768433935049609,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.250979325494995e-05,
+      "loss": 1.6734,
+      "step": 4046
+    },
+    {
+      "epoch": 0.17688710170899077,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.250617672867108e-05,
+      "loss": 2.1073,
+      "step": 4047
+    },
+    {
+      "epoch": 0.17693080991302068,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.250255940023882e-05,
+      "loss": 1.8738,
+      "step": 4048
+    },
+    {
+      "epoch": 0.17697451811705056,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.24989412697214e-05,
+      "loss": 1.796,
+      "step": 4049
+    },
+    {
+      "epoch": 0.17701822632108047,
+      "grad_norm": 7.3125,
+      "learning_rate": 9.24953223371871e-05,
+      "loss": 1.5202,
+      "step": 4050
+    },
+    {
+      "epoch": 0.17706193452511038,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.249170260270421e-05,
+      "loss": 3.0731,
+      "step": 4051
+    },
+    {
+      "epoch": 0.17710564272914026,
+      "grad_norm": 2.25,
+      "learning_rate": 9.248808206634105e-05,
+      "loss": 1.5085,
+      "step": 4052
+    },
+    {
+      "epoch": 0.17714935093317016,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.248446072816595e-05,
+      "loss": 1.917,
+      "step": 4053
+    },
+    {
+      "epoch": 0.17719305913720004,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.248083858824725e-05,
+      "loss": 2.1638,
+      "step": 4054
+    },
+    {
+      "epoch": 0.17723676734122995,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.247721564665329e-05,
+      "loss": 1.7179,
+      "step": 4055
+    },
+    {
+      "epoch": 0.17728047554525986,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.247359190345243e-05,
+      "loss": 1.8676,
+      "step": 4056
+    },
+    {
+      "epoch": 0.17732418374928974,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.24699673587131e-05,
+      "loss": 1.9322,
+      "step": 4057
+    },
+    {
+      "epoch": 0.17736789195331965,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.246634201250366e-05,
+      "loss": 1.9405,
+      "step": 4058
+    },
+    {
+      "epoch": 0.17741160015734952,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.246271586489255e-05,
+      "loss": 1.8649,
+      "step": 4059
+    },
+    {
+      "epoch": 0.17745530836137943,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.245908891594818e-05,
+      "loss": 2.0666,
+      "step": 4060
+    },
+    {
+      "epoch": 0.17749901656540934,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.245546116573901e-05,
+      "loss": 1.7037,
+      "step": 4061
+    },
+    {
+      "epoch": 0.17754272476943922,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.245183261433349e-05,
+      "loss": 1.4171,
+      "step": 4062
+    },
+    {
+      "epoch": 0.17758643297346913,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.244820326180011e-05,
+      "loss": 1.9897,
+      "step": 4063
+    },
+    {
+      "epoch": 0.177630141177499,
+      "grad_norm": 4.59375,
+      "learning_rate": 9.244457310820736e-05,
+      "loss": 1.8852,
+      "step": 4064
+    },
+    {
+      "epoch": 0.17767384938152891,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.244094215362373e-05,
+      "loss": 1.8307,
+      "step": 4065
+    },
+    {
+      "epoch": 0.17771755758555882,
+      "grad_norm": 3.140625,
+      "learning_rate": 9.243731039811775e-05,
+      "loss": 3.5239,
+      "step": 4066
+    },
+    {
+      "epoch": 0.1777612657895887,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.243367784175796e-05,
+      "loss": 2.6461,
+      "step": 4067
+    },
+    {
+      "epoch": 0.1778049739936186,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.243004448461293e-05,
+      "loss": 2.6207,
+      "step": 4068
+    },
+    {
+      "epoch": 0.1778486821976485,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.242641032675117e-05,
+      "loss": 1.9232,
+      "step": 4069
+    },
+    {
+      "epoch": 0.1778923904016784,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.242277536824134e-05,
+      "loss": 1.782,
+      "step": 4070
+    },
+    {
+      "epoch": 0.1779360986057083,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.241913960915197e-05,
+      "loss": 1.5369,
+      "step": 4071
+    },
+    {
+      "epoch": 0.17797980680973818,
+      "grad_norm": 5.34375,
+      "learning_rate": 9.241550304955168e-05,
+      "loss": 2.1154,
+      "step": 4072
+    },
+    {
+      "epoch": 0.1780235150137681,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.241186568950915e-05,
+      "loss": 2.0227,
+      "step": 4073
+    },
+    {
+      "epoch": 0.17806722321779797,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.240822752909298e-05,
+      "loss": 2.1406,
+      "step": 4074
+    },
+    {
+      "epoch": 0.17811093142182788,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.240458856837182e-05,
+      "loss": 2.0179,
+      "step": 4075
+    },
+    {
+      "epoch": 0.17815463962585779,
+      "grad_norm": 2.625,
+      "learning_rate": 9.240094880741437e-05,
+      "loss": 2.2995,
+      "step": 4076
+    },
+    {
+      "epoch": 0.17819834782988767,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.23973082462893e-05,
+      "loss": 1.9567,
+      "step": 4077
+    },
+    {
+      "epoch": 0.17824205603391757,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.23936668850653e-05,
+      "loss": 2.1861,
+      "step": 4078
+    },
+    {
+      "epoch": 0.17828576423794745,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.239002472381113e-05,
+      "loss": 1.9636,
+      "step": 4079
+    },
+    {
+      "epoch": 0.17832947244197736,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.238638176259549e-05,
+      "loss": 1.7762,
+      "step": 4080
+    },
+    {
+      "epoch": 0.17837318064600727,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.238273800148712e-05,
+      "loss": 1.7434,
+      "step": 4081
+    },
+    {
+      "epoch": 0.17841688885003715,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.237909344055482e-05,
+      "loss": 1.8037,
+      "step": 4082
+    },
+    {
+      "epoch": 0.17846059705406705,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.237544807986733e-05,
+      "loss": 2.0798,
+      "step": 4083
+    },
+    {
+      "epoch": 0.17850430525809693,
+      "grad_norm": 2.25,
+      "learning_rate": 9.237180191949347e-05,
+      "loss": 2.1024,
+      "step": 4084
+    },
+    {
+      "epoch": 0.17854801346212684,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.236815495950204e-05,
+      "loss": 2.6516,
+      "step": 4085
+    },
+    {
+      "epoch": 0.17859172166615675,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.236450719996185e-05,
+      "loss": 1.988,
+      "step": 4086
+    },
+    {
+      "epoch": 0.17863542987018663,
+      "grad_norm": 2.25,
+      "learning_rate": 9.236085864094177e-05,
+      "loss": 2.3926,
+      "step": 4087
+    },
+    {
+      "epoch": 0.17867913807421654,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.235720928251063e-05,
+      "loss": 2.2052,
+      "step": 4088
+    },
+    {
+      "epoch": 0.17872284627824642,
+      "grad_norm": 2.375,
+      "learning_rate": 9.235355912473729e-05,
+      "loss": 1.9098,
+      "step": 4089
+    },
+    {
+      "epoch": 0.17876655448227632,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.234990816769065e-05,
+      "loss": 2.6546,
+      "step": 4090
+    },
+    {
+      "epoch": 0.17881026268630623,
+      "grad_norm": 4.5625,
+      "learning_rate": 9.23462564114396e-05,
+      "loss": 2.1178,
+      "step": 4091
+    },
+    {
+      "epoch": 0.1788539708903361,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.234260385605308e-05,
+      "loss": 2.0182,
+      "step": 4092
+    },
+    {
+      "epoch": 0.17889767909436602,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.233895050159999e-05,
+      "loss": 2.1849,
+      "step": 4093
+    },
+    {
+      "epoch": 0.1789413872983959,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.233529634814928e-05,
+      "loss": 2.3742,
+      "step": 4094
+    },
+    {
+      "epoch": 0.1789850955024258,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.23316413957699e-05,
+      "loss": 2.0213,
+      "step": 4095
+    },
+    {
+      "epoch": 0.1790288037064557,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.232798564453086e-05,
+      "loss": 1.5629,
+      "step": 4096
+    },
+    {
+      "epoch": 0.1790725119104856,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.23243290945011e-05,
+      "loss": 1.5609,
+      "step": 4097
+    },
+    {
+      "epoch": 0.1791162201145155,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.232067174574968e-05,
+      "loss": 2.1387,
+      "step": 4098
+    },
+    {
+      "epoch": 0.17915992831854538,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.231701359834557e-05,
+      "loss": 2.1409,
+      "step": 4099
+    },
+    {
+      "epoch": 0.1792036365225753,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.231335465235782e-05,
+      "loss": 2.0755,
+      "step": 4100
+    },
+    {
+      "epoch": 0.1792473447266052,
+      "grad_norm": 3.75,
+      "learning_rate": 9.230969490785549e-05,
+      "loss": 2.0952,
+      "step": 4101
+    },
+    {
+      "epoch": 0.17929105293063508,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.230603436490763e-05,
+      "loss": 1.9135,
+      "step": 4102
+    },
+    {
+      "epoch": 0.17933476113466498,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.230237302358336e-05,
+      "loss": 2.0219,
+      "step": 4103
+    },
+    {
+      "epoch": 0.17937846933869486,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.22987108839517e-05,
+      "loss": 1.9435,
+      "step": 4104
+    },
+    {
+      "epoch": 0.17942217754272477,
+      "grad_norm": 3.25,
+      "learning_rate": 9.229504794608182e-05,
+      "loss": 1.6264,
+      "step": 4105
+    },
+    {
+      "epoch": 0.17946588574675468,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.229138421004284e-05,
+      "loss": 2.1426,
+      "step": 4106
+    },
+    {
+      "epoch": 0.17950959395078456,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.228771967590388e-05,
+      "loss": 1.7699,
+      "step": 4107
+    },
+    {
+      "epoch": 0.17955330215481446,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.228405434373409e-05,
+      "loss": 2.0044,
+      "step": 4108
+    },
+    {
+      "epoch": 0.17959701035884434,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.228038821360268e-05,
+      "loss": 1.7695,
+      "step": 4109
+    },
+    {
+      "epoch": 0.17964071856287425,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.227672128557879e-05,
+      "loss": 2.1878,
+      "step": 4110
+    },
+    {
+      "epoch": 0.17968442676690416,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.227305355973163e-05,
+      "loss": 2.2419,
+      "step": 4111
+    },
+    {
+      "epoch": 0.17972813497093404,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.226938503613043e-05,
+      "loss": 1.7856,
+      "step": 4112
+    },
+    {
+      "epoch": 0.17977184317496395,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.226571571484442e-05,
+      "loss": 1.968,
+      "step": 4113
+    },
+    {
+      "epoch": 0.17981555137899383,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.226204559594284e-05,
+      "loss": 2.1826,
+      "step": 4114
+    },
+    {
+      "epoch": 0.17985925958302373,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.225837467949495e-05,
+      "loss": 2.0379,
+      "step": 4115
+    },
+    {
+      "epoch": 0.17990296778705364,
+      "grad_norm": 2.375,
+      "learning_rate": 9.225470296557002e-05,
+      "loss": 1.8321,
+      "step": 4116
+    },
+    {
+      "epoch": 0.17994667599108352,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.225103045423735e-05,
+      "loss": 1.9019,
+      "step": 4117
+    },
+    {
+      "epoch": 0.17999038419511343,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.224735714556624e-05,
+      "loss": 2.7345,
+      "step": 4118
+    },
+    {
+      "epoch": 0.1800340923991433,
+      "grad_norm": 2.375,
+      "learning_rate": 9.2243683039626e-05,
+      "loss": 2.2188,
+      "step": 4119
+    },
+    {
+      "epoch": 0.18007780060317322,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.224000813648602e-05,
+      "loss": 1.8773,
+      "step": 4120
+    },
+    {
+      "epoch": 0.18012150880720312,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.223633243621556e-05,
+      "loss": 1.8854,
+      "step": 4121
+    },
+    {
+      "epoch": 0.180165217011233,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.223265593888405e-05,
+      "loss": 1.7813,
+      "step": 4122
+    },
+    {
+      "epoch": 0.1802089252152629,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.222897864456088e-05,
+      "loss": 1.6252,
+      "step": 4123
+    },
+    {
+      "epoch": 0.1802526334192928,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.22253005533154e-05,
+      "loss": 1.9214,
+      "step": 4124
+    },
+    {
+      "epoch": 0.1802963416233227,
+      "grad_norm": 2.5,
+      "learning_rate": 9.222162166521704e-05,
+      "loss": 2.6064,
+      "step": 4125
+    },
+    {
+      "epoch": 0.1803400498273526,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.221794198033525e-05,
+      "loss": 1.7756,
+      "step": 4126
+    },
+    {
+      "epoch": 0.18038375803138249,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.221426149873942e-05,
+      "loss": 1.7808,
+      "step": 4127
+    },
+    {
+      "epoch": 0.1804274662354124,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.221058022049906e-05,
+      "loss": 2.3092,
+      "step": 4128
+    },
+    {
+      "epoch": 0.18047117443944227,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.220689814568359e-05,
+      "loss": 1.8663,
+      "step": 4129
+    },
+    {
+      "epoch": 0.18051488264347218,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.220321527436256e-05,
+      "loss": 2.9884,
+      "step": 4130
+    },
+    {
+      "epoch": 0.1805585908475021,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.21995316066054e-05,
+      "loss": 1.7863,
+      "step": 4131
+    },
+    {
+      "epoch": 0.18060229905153197,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.219584714248167e-05,
+      "loss": 1.6175,
+      "step": 4132
+    },
+    {
+      "epoch": 0.18064600725556187,
+      "grad_norm": 1.8984375,
+      "learning_rate": 9.21921618820609e-05,
+      "loss": 1.6722,
+      "step": 4133
+    },
+    {
+      "epoch": 0.18068971545959175,
+      "grad_norm": 4.21875,
+      "learning_rate": 9.21884758254126e-05,
+      "loss": 1.5489,
+      "step": 4134
+    },
+    {
+      "epoch": 0.18073342366362166,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.21847889726064e-05,
+      "loss": 2.5853,
+      "step": 4135
+    },
+    {
+      "epoch": 0.18077713186765157,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.218110132371182e-05,
+      "loss": 2.0443,
+      "step": 4136
+    },
+    {
+      "epoch": 0.18082084007168145,
+      "grad_norm": 3.390625,
+      "learning_rate": 9.217741287879846e-05,
+      "loss": 1.8777,
+      "step": 4137
+    },
+    {
+      "epoch": 0.18086454827571136,
+      "grad_norm": 2.5,
+      "learning_rate": 9.217372363793592e-05,
+      "loss": 1.9894,
+      "step": 4138
+    },
+    {
+      "epoch": 0.18090825647974124,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.217003360119386e-05,
+      "loss": 1.6585,
+      "step": 4139
+    },
+    {
+      "epoch": 0.18095196468377114,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.216634276864188e-05,
+      "loss": 2.5412,
+      "step": 4140
+    },
+    {
+      "epoch": 0.18099567288780105,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.216265114034964e-05,
+      "loss": 2.1771,
+      "step": 4141
+    },
+    {
+      "epoch": 0.18103938109183093,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.21589587163868e-05,
+      "loss": 1.8523,
+      "step": 4142
+    },
+    {
+      "epoch": 0.18108308929586084,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.215526549682307e-05,
+      "loss": 2.1015,
+      "step": 4143
+    },
+    {
+      "epoch": 0.18112679749989072,
+      "grad_norm": 3.4375,
+      "learning_rate": 9.21515714817281e-05,
+      "loss": 2.06,
+      "step": 4144
+    },
+    {
+      "epoch": 0.18117050570392063,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.214787667117165e-05,
+      "loss": 2.2972,
+      "step": 4145
+    },
+    {
+      "epoch": 0.18121421390795053,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.214418106522342e-05,
+      "loss": 2.0532,
+      "step": 4146
+    },
+    {
+      "epoch": 0.1812579221119804,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.214048466395316e-05,
+      "loss": 2.1041,
+      "step": 4147
+    },
+    {
+      "epoch": 0.18130163031601032,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.21367874674306e-05,
+      "loss": 1.9628,
+      "step": 4148
+    },
+    {
+      "epoch": 0.1813453385200402,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.213308947572554e-05,
+      "loss": 2.5201,
+      "step": 4149
+    },
+    {
+      "epoch": 0.1813890467240701,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.212939068890778e-05,
+      "loss": 1.8226,
+      "step": 4150
+    },
+    {
+      "epoch": 0.18143275492810002,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.212569110704708e-05,
+      "loss": 2.3515,
+      "step": 4151
+    },
+    {
+      "epoch": 0.1814764631321299,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.212199073021329e-05,
+      "loss": 1.9236,
+      "step": 4152
+    },
+    {
+      "epoch": 0.1815201713361598,
+      "grad_norm": 2.125,
+      "learning_rate": 9.211828955847622e-05,
+      "loss": 1.9031,
+      "step": 4153
+    },
+    {
+      "epoch": 0.18156387954018968,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.211458759190573e-05,
+      "loss": 1.7923,
+      "step": 4154
+    },
+    {
+      "epoch": 0.1816075877442196,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.211088483057168e-05,
+      "loss": 2.4361,
+      "step": 4155
+    },
+    {
+      "epoch": 0.1816512959482495,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.210718127454394e-05,
+      "loss": 2.2487,
+      "step": 4156
+    },
+    {
+      "epoch": 0.18169500415227938,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.210347692389241e-05,
+      "loss": 1.8312,
+      "step": 4157
+    },
+    {
+      "epoch": 0.18173871235630928,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.209977177868698e-05,
+      "loss": 2.6821,
+      "step": 4158
+    },
+    {
+      "epoch": 0.18178242056033916,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.209606583899759e-05,
+      "loss": 1.7224,
+      "step": 4159
+    },
+    {
+      "epoch": 0.18182612876436907,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.209235910489418e-05,
+      "loss": 2.8291,
+      "step": 4160
+    },
+    {
+      "epoch": 0.18186983696839898,
+      "grad_norm": 2.5,
+      "learning_rate": 9.208865157644668e-05,
+      "loss": 1.6297,
+      "step": 4161
+    },
+    {
+      "epoch": 0.18191354517242886,
+      "grad_norm": 2.5,
+      "learning_rate": 9.208494325372508e-05,
+      "loss": 2.0138,
+      "step": 4162
+    },
+    {
+      "epoch": 0.18195725337645877,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.208123413679933e-05,
+      "loss": 2.2733,
+      "step": 4163
+    },
+    {
+      "epoch": 0.18200096158048865,
+      "grad_norm": 3.125,
+      "learning_rate": 9.207752422573946e-05,
+      "loss": 1.977,
+      "step": 4164
+    },
+    {
+      "epoch": 0.18204466978451855,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.207381352061545e-05,
+      "loss": 2.1237,
+      "step": 4165
+    },
+    {
+      "epoch": 0.18208837798854846,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.207010202149736e-05,
+      "loss": 2.2442,
+      "step": 4166
+    },
+    {
+      "epoch": 0.18213208619257834,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.206638972845522e-05,
+      "loss": 1.8178,
+      "step": 4167
+    },
+    {
+      "epoch": 0.18217579439660825,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.206267664155907e-05,
+      "loss": 2.3336,
+      "step": 4168
+    },
+    {
+      "epoch": 0.18221950260063813,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.205896276087899e-05,
+      "loss": 1.9392,
+      "step": 4169
+    },
+    {
+      "epoch": 0.18226321080466804,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.205524808648507e-05,
+      "loss": 1.5458,
+      "step": 4170
+    },
+    {
+      "epoch": 0.18230691900869794,
+      "grad_norm": 3.53125,
+      "learning_rate": 9.205153261844741e-05,
+      "loss": 2.8738,
+      "step": 4171
+    },
+    {
+      "epoch": 0.18235062721272782,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.204781635683613e-05,
+      "loss": 2.0505,
+      "step": 4172
+    },
+    {
+      "epoch": 0.18239433541675773,
+      "grad_norm": 16.25,
+      "learning_rate": 9.204409930172135e-05,
+      "loss": 2.1514,
+      "step": 4173
+    },
+    {
+      "epoch": 0.1824380436207876,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.204038145317324e-05,
+      "loss": 1.4737,
+      "step": 4174
+    },
+    {
+      "epoch": 0.18248175182481752,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.203666281126193e-05,
+      "loss": 1.922,
+      "step": 4175
+    },
+    {
+      "epoch": 0.18252546002884742,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.203294337605761e-05,
+      "loss": 2.0385,
+      "step": 4176
+    },
+    {
+      "epoch": 0.1825691682328773,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.202922314763048e-05,
+      "loss": 1.6214,
+      "step": 4177
+    },
+    {
+      "epoch": 0.1826128764369072,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.202550212605074e-05,
+      "loss": 1.7539,
+      "step": 4178
+    },
+    {
+      "epoch": 0.1826565846409371,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.202178031138862e-05,
+      "loss": 1.9389,
+      "step": 4179
+    },
+    {
+      "epoch": 0.182700292844967,
+      "grad_norm": 2.625,
+      "learning_rate": 9.201805770371432e-05,
+      "loss": 2.0368,
+      "step": 4180
+    },
+    {
+      "epoch": 0.1827440010489969,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.201433430309813e-05,
+      "loss": 1.7877,
+      "step": 4181
+    },
+    {
+      "epoch": 0.1827877092530268,
+      "grad_norm": 2.625,
+      "learning_rate": 9.20106101096103e-05,
+      "loss": 2.8653,
+      "step": 4182
+    },
+    {
+      "epoch": 0.1828314174570567,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.200688512332111e-05,
+      "loss": 2.528,
+      "step": 4183
+    },
+    {
+      "epoch": 0.18287512566108657,
+      "grad_norm": 3.671875,
+      "learning_rate": 9.200315934430088e-05,
+      "loss": 1.9375,
+      "step": 4184
+    },
+    {
+      "epoch": 0.18291883386511648,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.199943277261989e-05,
+      "loss": 1.5832,
+      "step": 4185
+    },
+    {
+      "epoch": 0.1829625420691464,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.199570540834846e-05,
+      "loss": 2.6069,
+      "step": 4186
+    },
+    {
+      "epoch": 0.18300625027317627,
+      "grad_norm": 3.375,
+      "learning_rate": 9.199197725155697e-05,
+      "loss": 2.8072,
+      "step": 4187
+    },
+    {
+      "epoch": 0.18304995847720618,
+      "grad_norm": 2.5,
+      "learning_rate": 9.198824830231573e-05,
+      "loss": 2.2208,
+      "step": 4188
+    },
+    {
+      "epoch": 0.18309366668123606,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.198451856069515e-05,
+      "loss": 2.1902,
+      "step": 4189
+    },
+    {
+      "epoch": 0.18313737488526596,
+      "grad_norm": 3.59375,
+      "learning_rate": 9.19807880267656e-05,
+      "loss": 2.2622,
+      "step": 4190
+    },
+    {
+      "epoch": 0.18318108308929587,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.197705670059747e-05,
+      "loss": 2.5455,
+      "step": 4191
+    },
+    {
+      "epoch": 0.18322479129332575,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.197332458226118e-05,
+      "loss": 1.8177,
+      "step": 4192
+    },
+    {
+      "epoch": 0.18326849949735566,
+      "grad_norm": 2.125,
+      "learning_rate": 9.196959167182719e-05,
+      "loss": 2.3912,
+      "step": 4193
+    },
+    {
+      "epoch": 0.18331220770138554,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.19658579693659e-05,
+      "loss": 1.7873,
+      "step": 4194
+    },
+    {
+      "epoch": 0.18335591590541545,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.196212347494781e-05,
+      "loss": 2.2754,
+      "step": 4195
+    },
+    {
+      "epoch": 0.18339962410944535,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.195838818864337e-05,
+      "loss": 2.2531,
+      "step": 4196
+    },
+    {
+      "epoch": 0.18344333231347523,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.195465211052306e-05,
+      "loss": 1.9503,
+      "step": 4197
+    },
+    {
+      "epoch": 0.18348704051750514,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.195091524065742e-05,
+      "loss": 1.4529,
+      "step": 4198
+    },
+    {
+      "epoch": 0.18353074872153502,
+      "grad_norm": 1.96875,
+      "learning_rate": 9.194717757911694e-05,
+      "loss": 1.9256,
+      "step": 4199
+    },
+    {
+      "epoch": 0.18357445692556493,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.194343912597218e-05,
+      "loss": 2.1218,
+      "step": 4200
+    },
+    {
+      "epoch": 0.18361816512959483,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.193969988129367e-05,
+      "loss": 2.0301,
+      "step": 4201
+    },
+    {
+      "epoch": 0.18366187333362471,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.1935959845152e-05,
+      "loss": 2.0649,
+      "step": 4202
+    },
+    {
+      "epoch": 0.18370558153765462,
+      "grad_norm": 5.375,
+      "learning_rate": 9.193221901761772e-05,
+      "loss": 3.2402,
+      "step": 4203
+    },
+    {
+      "epoch": 0.1837492897416845,
+      "grad_norm": 7.09375,
+      "learning_rate": 9.192847739876142e-05,
+      "loss": 2.1486,
+      "step": 4204
+    },
+    {
+      "epoch": 0.1837929979457144,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.192473498865376e-05,
+      "loss": 1.6001,
+      "step": 4205
+    },
+    {
+      "epoch": 0.18383670614974432,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.192099178736532e-05,
+      "loss": 1.9152,
+      "step": 4206
+    },
+    {
+      "epoch": 0.1838804143537742,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.191724779496675e-05,
+      "loss": 1.463,
+      "step": 4207
+    },
+    {
+      "epoch": 0.1839241225578041,
+      "grad_norm": 2.875,
+      "learning_rate": 9.19135030115287e-05,
+      "loss": 1.755,
+      "step": 4208
+    },
+    {
+      "epoch": 0.18396783076183398,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.190975743712184e-05,
+      "loss": 1.781,
+      "step": 4209
+    },
+    {
+      "epoch": 0.1840115389658639,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.19060110718169e-05,
+      "loss": 2.0729,
+      "step": 4210
+    },
+    {
+      "epoch": 0.1840552471698938,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.19022639156845e-05,
+      "loss": 1.388,
+      "step": 4211
+    },
+    {
+      "epoch": 0.18409895537392368,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.18985159687954e-05,
+      "loss": 1.774,
+      "step": 4212
+    },
+    {
+      "epoch": 0.18414266357795359,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.189476723122034e-05,
+      "loss": 1.8746,
+      "step": 4213
+    },
+    {
+      "epoch": 0.18418637178198347,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.189101770303003e-05,
+      "loss": 1.9607,
+      "step": 4214
+    },
+    {
+      "epoch": 0.18423007998601337,
+      "grad_norm": 2.125,
+      "learning_rate": 9.188726738429526e-05,
+      "loss": 1.4984,
+      "step": 4215
+    },
+    {
+      "epoch": 0.18427378819004328,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.188351627508678e-05,
+      "loss": 1.7226,
+      "step": 4216
+    },
+    {
+      "epoch": 0.18431749639407316,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.187976437547538e-05,
+      "loss": 2.1207,
+      "step": 4217
+    },
+    {
+      "epoch": 0.18436120459810307,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.187601168553191e-05,
+      "loss": 2.0887,
+      "step": 4218
+    },
+    {
+      "epoch": 0.18440491280213295,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.187225820532712e-05,
+      "loss": 2.0781,
+      "step": 4219
+    },
+    {
+      "epoch": 0.18444862100616286,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.186850393493188e-05,
+      "loss": 2.0583,
+      "step": 4220
+    },
+    {
+      "epoch": 0.18449232921019276,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.186474887441704e-05,
+      "loss": 2.3583,
+      "step": 4221
+    },
+    {
+      "epoch": 0.18453603741422264,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.186099302385344e-05,
+      "loss": 2.019,
+      "step": 4222
+    },
+    {
+      "epoch": 0.18457974561825255,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.185723638331201e-05,
+      "loss": 2.1852,
+      "step": 4223
+    },
+    {
+      "epoch": 0.18462345382228243,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.185347895286358e-05,
+      "loss": 1.8512,
+      "step": 4224
+    },
+    {
+      "epoch": 0.18466716202631234,
+      "grad_norm": 1.859375,
+      "learning_rate": 9.184972073257911e-05,
+      "loss": 1.7002,
+      "step": 4225
+    },
+    {
+      "epoch": 0.18471087023034224,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.184596172252948e-05,
+      "loss": 1.9231,
+      "step": 4226
+    },
+    {
+      "epoch": 0.18475457843437212,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.184220192278565e-05,
+      "loss": 1.7761,
+      "step": 4227
+    },
+    {
+      "epoch": 0.18479828663840203,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.183844133341859e-05,
+      "loss": 1.9647,
+      "step": 4228
+    },
+    {
+      "epoch": 0.1848419948424319,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.183467995449924e-05,
+      "loss": 1.9758,
+      "step": 4229
+    },
+    {
+      "epoch": 0.18488570304646182,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.183091778609858e-05,
+      "loss": 2.1707,
+      "step": 4230
+    },
+    {
+      "epoch": 0.18492941125049173,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.182715482828763e-05,
+      "loss": 2.0485,
+      "step": 4231
+    },
+    {
+      "epoch": 0.1849731194545216,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.182339108113738e-05,
+      "loss": 1.9269,
+      "step": 4232
+    },
+    {
+      "epoch": 0.1850168276585515,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.181962654471888e-05,
+      "loss": 2.4354,
+      "step": 4233
+    },
+    {
+      "epoch": 0.1850605358625814,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.181586121910317e-05,
+      "loss": 2.3555,
+      "step": 4234
+    },
+    {
+      "epoch": 0.1851042440666113,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.181209510436128e-05,
+      "loss": 2.3343,
+      "step": 4235
+    },
+    {
+      "epoch": 0.1851479522706412,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.180832820056431e-05,
+      "loss": 1.7184,
+      "step": 4236
+    },
+    {
+      "epoch": 0.1851916604746711,
+      "grad_norm": 2.0,
+      "learning_rate": 9.180456050778334e-05,
+      "loss": 1.6113,
+      "step": 4237
+    },
+    {
+      "epoch": 0.185235368678701,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.180079202608947e-05,
+      "loss": 1.749,
+      "step": 4238
+    },
+    {
+      "epoch": 0.18527907688273088,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.179702275555382e-05,
+      "loss": 2.1112,
+      "step": 4239
+    },
+    {
+      "epoch": 0.18532278508676078,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.17932526962475e-05,
+      "loss": 1.8853,
+      "step": 4240
+    },
+    {
+      "epoch": 0.1853664932907907,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.178948184824168e-05,
+      "loss": 2.6173,
+      "step": 4241
+    },
+    {
+      "epoch": 0.18541020149482057,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.178571021160753e-05,
+      "loss": 2.3146,
+      "step": 4242
+    },
+    {
+      "epoch": 0.18545390969885048,
+      "grad_norm": 3.125,
+      "learning_rate": 9.17819377864162e-05,
+      "loss": 2.4007,
+      "step": 4243
+    },
+    {
+      "epoch": 0.18549761790288036,
+      "grad_norm": 2.875,
+      "learning_rate": 9.17781645727389e-05,
+      "loss": 2.725,
+      "step": 4244
+    },
+    {
+      "epoch": 0.18554132610691026,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.177439057064683e-05,
+      "loss": 2.5454,
+      "step": 4245
+    },
+    {
+      "epoch": 0.18558503431094017,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.17706157802112e-05,
+      "loss": 1.9567,
+      "step": 4246
+    },
+    {
+      "epoch": 0.18562874251497005,
+      "grad_norm": 3.0,
+      "learning_rate": 9.176684020150326e-05,
+      "loss": 2.2343,
+      "step": 4247
+    },
+    {
+      "epoch": 0.18567245071899996,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.176306383459426e-05,
+      "loss": 2.1333,
+      "step": 4248
+    },
+    {
+      "epoch": 0.18571615892302984,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.175928667955546e-05,
+      "loss": 1.7027,
+      "step": 4249
+    },
+    {
+      "epoch": 0.18575986712705975,
+      "grad_norm": 3.125,
+      "learning_rate": 9.175550873645816e-05,
+      "loss": 2.0228,
+      "step": 4250
+    },
+    {
+      "epoch": 0.18580357533108965,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.175173000537361e-05,
+      "loss": 2.6521,
+      "step": 4251
+    },
+    {
+      "epoch": 0.18584728353511953,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.174795048637316e-05,
+      "loss": 2.2313,
+      "step": 4252
+    },
+    {
+      "epoch": 0.18589099173914944,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.174417017952812e-05,
+      "loss": 2.1287,
+      "step": 4253
+    },
+    {
+      "epoch": 0.18593469994317932,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.174038908490984e-05,
+      "loss": 1.7118,
+      "step": 4254
+    },
+    {
+      "epoch": 0.18597840814720923,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.173660720258966e-05,
+      "loss": 1.589,
+      "step": 4255
+    },
+    {
+      "epoch": 0.18602211635123914,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.173282453263897e-05,
+      "loss": 1.8921,
+      "step": 4256
+    },
+    {
+      "epoch": 0.18606582455526902,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.172904107512911e-05,
+      "loss": 1.9779,
+      "step": 4257
+    },
+    {
+      "epoch": 0.18610953275929892,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.172525683013155e-05,
+      "loss": 2.7427,
+      "step": 4258
+    },
+    {
+      "epoch": 0.1861532409633288,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.172147179771765e-05,
+      "loss": 2.3742,
+      "step": 4259
+    },
+    {
+      "epoch": 0.1861969491673587,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.171768597795885e-05,
+      "loss": 2.1021,
+      "step": 4260
+    },
+    {
+      "epoch": 0.18624065737138862,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.17138993709266e-05,
+      "loss": 2.4938,
+      "step": 4261
+    },
+    {
+      "epoch": 0.1862843655754185,
+      "grad_norm": 20.25,
+      "learning_rate": 9.171011197669236e-05,
+      "loss": 1.6165,
+      "step": 4262
+    },
+    {
+      "epoch": 0.1863280737794484,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.170632379532759e-05,
+      "loss": 1.7445,
+      "step": 4263
+    },
+    {
+      "epoch": 0.18637178198347829,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.17025348269038e-05,
+      "loss": 1.8762,
+      "step": 4264
+    },
+    {
+      "epoch": 0.1864154901875082,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.16987450714925e-05,
+      "loss": 2.254,
+      "step": 4265
+    },
+    {
+      "epoch": 0.1864591983915381,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.169495452916516e-05,
+      "loss": 3.2139,
+      "step": 4266
+    },
+    {
+      "epoch": 0.18650290659556798,
+      "grad_norm": 2.25,
+      "learning_rate": 9.169116319999336e-05,
+      "loss": 2.2422,
+      "step": 4267
+    },
+    {
+      "epoch": 0.1865466147995979,
+      "grad_norm": 3.34375,
+      "learning_rate": 9.168737108404864e-05,
+      "loss": 1.9404,
+      "step": 4268
+    },
+    {
+      "epoch": 0.18659032300362777,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.168357818140255e-05,
+      "loss": 1.9761,
+      "step": 4269
+    },
+    {
+      "epoch": 0.18663403120765767,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.167978449212666e-05,
+      "loss": 2.1656,
+      "step": 4270
+    },
+    {
+      "epoch": 0.18667773941168758,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.167599001629257e-05,
+      "loss": 2.4656,
+      "step": 4271
+    },
+    {
+      "epoch": 0.18672144761571746,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.167219475397191e-05,
+      "loss": 2.2688,
+      "step": 4272
+    },
+    {
+      "epoch": 0.18676515581974737,
+      "grad_norm": 2.875,
+      "learning_rate": 9.166839870523627e-05,
+      "loss": 1.9322,
+      "step": 4273
+    },
+    {
+      "epoch": 0.18680886402377728,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.16646018701573e-05,
+      "loss": 1.6503,
+      "step": 4274
+    },
+    {
+      "epoch": 0.18685257222780716,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.166080424880666e-05,
+      "loss": 1.9766,
+      "step": 4275
+    },
+    {
+      "epoch": 0.18689628043183706,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.165700584125601e-05,
+      "loss": 2.035,
+      "step": 4276
+    },
+    {
+      "epoch": 0.18693998863586694,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.165320664757705e-05,
+      "loss": 2.16,
+      "step": 4277
+    },
+    {
+      "epoch": 0.18698369683989685,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.164940666784143e-05,
+      "loss": 1.9903,
+      "step": 4278
+    },
+    {
+      "epoch": 0.18702740504392676,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.164560590212088e-05,
+      "loss": 1.7399,
+      "step": 4279
+    },
+    {
+      "epoch": 0.18707111324795664,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.164180435048715e-05,
+      "loss": 2.0793,
+      "step": 4280
+    },
+    {
+      "epoch": 0.18711482145198655,
+      "grad_norm": 6.625,
+      "learning_rate": 9.163800201301197e-05,
+      "loss": 2.8651,
+      "step": 4281
+    },
+    {
+      "epoch": 0.18715852965601643,
+      "grad_norm": 2.625,
+      "learning_rate": 9.163419888976708e-05,
+      "loss": 2.162,
+      "step": 4282
+    },
+    {
+      "epoch": 0.18720223786004633,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.163039498082428e-05,
+      "loss": 1.6775,
+      "step": 4283
+    },
+    {
+      "epoch": 0.18724594606407624,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.162659028625531e-05,
+      "loss": 1.9262,
+      "step": 4284
+    },
+    {
+      "epoch": 0.18728965426810612,
+      "grad_norm": 2.625,
+      "learning_rate": 9.162278480613203e-05,
+      "loss": 1.9239,
+      "step": 4285
+    },
+    {
+      "epoch": 0.18733336247213603,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.161897854052619e-05,
+      "loss": 1.6802,
+      "step": 4286
+    },
+    {
+      "epoch": 0.1873770706761659,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.161517148950967e-05,
+      "loss": 1.9088,
+      "step": 4287
+    },
+    {
+      "epoch": 0.18742077888019582,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.161136365315428e-05,
+      "loss": 1.9722,
+      "step": 4288
+    },
+    {
+      "epoch": 0.18746448708422572,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.160755503153192e-05,
+      "loss": 1.692,
+      "step": 4289
+    },
+    {
+      "epoch": 0.1875081952882556,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.16037456247144e-05,
+      "loss": 1.5951,
+      "step": 4290
+    },
+    {
+      "epoch": 0.1875519034922855,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.159993543277368e-05,
+      "loss": 2.0796,
+      "step": 4291
+    },
+    {
+      "epoch": 0.1875956116963154,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.159612445578163e-05,
+      "loss": 1.3354,
+      "step": 4292
+    },
+    {
+      "epoch": 0.1876393199003453,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.159231269381016e-05,
+      "loss": 2.0206,
+      "step": 4293
+    },
+    {
+      "epoch": 0.1876830281043752,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.158850014693123e-05,
+      "loss": 1.9006,
+      "step": 4294
+    },
+    {
+      "epoch": 0.18772673630840508,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.158468681521676e-05,
+      "loss": 2.1584,
+      "step": 4295
+    },
+    {
+      "epoch": 0.187770444512435,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.158087269873871e-05,
+      "loss": 1.8828,
+      "step": 4296
+    },
+    {
+      "epoch": 0.18781415271646487,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.15770577975691e-05,
+      "loss": 1.9622,
+      "step": 4297
+    },
+    {
+      "epoch": 0.18785786092049478,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.157324211177991e-05,
+      "loss": 1.7722,
+      "step": 4298
+    },
+    {
+      "epoch": 0.1879015691245247,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.15694256414431e-05,
+      "loss": 2.0762,
+      "step": 4299
+    },
+    {
+      "epoch": 0.18794527732855457,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.156560838663076e-05,
+      "loss": 2.1648,
+      "step": 4300
+    },
+    {
+      "epoch": 0.18798898553258447,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.156179034741486e-05,
+      "loss": 1.9935,
+      "step": 4301
+    },
+    {
+      "epoch": 0.18803269373661435,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.155797152386752e-05,
+      "loss": 1.7464,
+      "step": 4302
+    },
+    {
+      "epoch": 0.18807640194064426,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.155415191606074e-05,
+      "loss": 1.6097,
+      "step": 4303
+    },
+    {
+      "epoch": 0.18812011014467417,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.155033152406665e-05,
+      "loss": 1.9678,
+      "step": 4304
+    },
+    {
+      "epoch": 0.18816381834870405,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.154651034795734e-05,
+      "loss": 1.6927,
+      "step": 4305
+    },
+    {
+      "epoch": 0.18820752655273396,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.154268838780489e-05,
+      "loss": 2.6745,
+      "step": 4306
+    },
+    {
+      "epoch": 0.18825123475676384,
+      "grad_norm": 2.375,
+      "learning_rate": 9.153886564368145e-05,
+      "loss": 2.1593,
+      "step": 4307
+    },
+    {
+      "epoch": 0.18829494296079374,
+      "grad_norm": 2.0,
+      "learning_rate": 9.153504211565917e-05,
+      "loss": 1.9622,
+      "step": 4308
+    },
+    {
+      "epoch": 0.18833865116482365,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.15312178038102e-05,
+      "loss": 1.9805,
+      "step": 4309
+    },
+    {
+      "epoch": 0.18838235936885353,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.15273927082067e-05,
+      "loss": 1.6238,
+      "step": 4310
+    },
+    {
+      "epoch": 0.18842606757288344,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.152356682892085e-05,
+      "loss": 1.8822,
+      "step": 4311
+    },
+    {
+      "epoch": 0.18846977577691332,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.151974016602485e-05,
+      "loss": 1.5081,
+      "step": 4312
+    },
+    {
+      "epoch": 0.18851348398094323,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.151591271959094e-05,
+      "loss": 1.7578,
+      "step": 4313
+    },
+    {
+      "epoch": 0.18855719218497313,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.151208448969134e-05,
+      "loss": 1.9479,
+      "step": 4314
+    },
+    {
+      "epoch": 0.188600900389003,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.150825547639827e-05,
+      "loss": 2.4475,
+      "step": 4315
+    },
+    {
+      "epoch": 0.18864460859303292,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.150442567978402e-05,
+      "loss": 1.4966,
+      "step": 4316
+    },
+    {
+      "epoch": 0.1886883167970628,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.150059509992085e-05,
+      "loss": 1.818,
+      "step": 4317
+    },
+    {
+      "epoch": 0.1887320250010927,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.149676373688105e-05,
+      "loss": 1.8557,
+      "step": 4318
+    },
+    {
+      "epoch": 0.18877573320512261,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.149293159073691e-05,
+      "loss": 1.636,
+      "step": 4319
+    },
+    {
+      "epoch": 0.1888194414091525,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.148909866156076e-05,
+      "loss": 2.0637,
+      "step": 4320
+    },
+    {
+      "epoch": 0.1888631496131824,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.148526494942496e-05,
+      "loss": 1.9292,
+      "step": 4321
+    },
+    {
+      "epoch": 0.18890685781721228,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.14814304544018e-05,
+      "loss": 1.9628,
+      "step": 4322
+    },
+    {
+      "epoch": 0.1889505660212422,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.147759517656369e-05,
+      "loss": 2.0658,
+      "step": 4323
+    },
+    {
+      "epoch": 0.1889942742252721,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.147375911598299e-05,
+      "loss": 1.3896,
+      "step": 4324
+    },
+    {
+      "epoch": 0.18903798242930198,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.14699222727321e-05,
+      "loss": 2.1023,
+      "step": 4325
+    },
+    {
+      "epoch": 0.18908169063333188,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.146608464688341e-05,
+      "loss": 2.0631,
+      "step": 4326
+    },
+    {
+      "epoch": 0.18912539883736176,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.146224623850936e-05,
+      "loss": 2.255,
+      "step": 4327
+    },
+    {
+      "epoch": 0.18916910704139167,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.145840704768238e-05,
+      "loss": 1.578,
+      "step": 4328
+    },
+    {
+      "epoch": 0.18921281524542158,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.145456707447491e-05,
+      "loss": 1.5111,
+      "step": 4329
+    },
+    {
+      "epoch": 0.18925652344945146,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.145072631895942e-05,
+      "loss": 1.6645,
+      "step": 4330
+    },
+    {
+      "epoch": 0.18930023165348137,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.14468847812084e-05,
+      "loss": 1.8262,
+      "step": 4331
+    },
+    {
+      "epoch": 0.18934393985751125,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.144304246129434e-05,
+      "loss": 1.8456,
+      "step": 4332
+    },
+    {
+      "epoch": 0.18938764806154115,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.143919935928975e-05,
+      "loss": 2.0993,
+      "step": 4333
+    },
+    {
+      "epoch": 0.18943135626557106,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.143535547526716e-05,
+      "loss": 2.0291,
+      "step": 4334
+    },
+    {
+      "epoch": 0.18947506446960094,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.143151080929911e-05,
+      "loss": 2.1035,
+      "step": 4335
+    },
+    {
+      "epoch": 0.18951877267363085,
+      "grad_norm": 3.234375,
+      "learning_rate": 9.142766536145815e-05,
+      "loss": 2.0885,
+      "step": 4336
+    },
+    {
+      "epoch": 0.18956248087766073,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.142381913181684e-05,
+      "loss": 1.8083,
+      "step": 4337
+    },
+    {
+      "epoch": 0.18960618908169063,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.141997212044779e-05,
+      "loss": 1.5989,
+      "step": 4338
+    },
+    {
+      "epoch": 0.18964989728572054,
+      "grad_norm": 3.625,
+      "learning_rate": 9.141612432742357e-05,
+      "loss": 2.7321,
+      "step": 4339
+    },
+    {
+      "epoch": 0.18969360548975042,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.14122757528168e-05,
+      "loss": 2.4181,
+      "step": 4340
+    },
+    {
+      "epoch": 0.18973731369378033,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.140842639670014e-05,
+      "loss": 1.8314,
+      "step": 4341
+    },
+    {
+      "epoch": 0.1897810218978102,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.140457625914618e-05,
+      "loss": 2.2134,
+      "step": 4342
+    },
+    {
+      "epoch": 0.18982473010184012,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.14007253402276e-05,
+      "loss": 2.1502,
+      "step": 4343
+    },
+    {
+      "epoch": 0.18986843830587002,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.13968736400171e-05,
+      "loss": 2.2266,
+      "step": 4344
+    },
+    {
+      "epoch": 0.1899121465098999,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.139302115858733e-05,
+      "loss": 1.8241,
+      "step": 4345
+    },
+    {
+      "epoch": 0.1899558547139298,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.138916789601102e-05,
+      "loss": 2.5854,
+      "step": 4346
+    },
+    {
+      "epoch": 0.1899995629179597,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.138531385236086e-05,
+      "loss": 2.0389,
+      "step": 4347
+    },
+    {
+      "epoch": 0.1900432711219896,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.13814590277096e-05,
+      "loss": 2.3824,
+      "step": 4348
+    },
+    {
+      "epoch": 0.1900869793260195,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.137760342212997e-05,
+      "loss": 1.815,
+      "step": 4349
+    },
+    {
+      "epoch": 0.19013068753004939,
+      "grad_norm": 2.375,
+      "learning_rate": 9.137374703569475e-05,
+      "loss": 2.382,
+      "step": 4350
+    },
+    {
+      "epoch": 0.1901743957340793,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.13698898684767e-05,
+      "loss": 1.7854,
+      "step": 4351
+    },
+    {
+      "epoch": 0.19021810393810917,
+      "grad_norm": 2.0,
+      "learning_rate": 9.136603192054862e-05,
+      "loss": 1.8529,
+      "step": 4352
+    },
+    {
+      "epoch": 0.19026181214213908,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.136217319198332e-05,
+      "loss": 1.661,
+      "step": 4353
+    },
+    {
+      "epoch": 0.190305520346169,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.135831368285362e-05,
+      "loss": 1.901,
+      "step": 4354
+    },
+    {
+      "epoch": 0.19034922855019887,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.135445339323232e-05,
+      "loss": 1.9566,
+      "step": 4355
+    },
+    {
+      "epoch": 0.19039293675422878,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.135059232319232e-05,
+      "loss": 2.8984,
+      "step": 4356
+    },
+    {
+      "epoch": 0.19043664495825866,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.134673047280645e-05,
+      "loss": 2.7567,
+      "step": 4357
+    },
+    {
+      "epoch": 0.19048035316228856,
+      "grad_norm": 2.75,
+      "learning_rate": 9.134286784214759e-05,
+      "loss": 2.2439,
+      "step": 4358
+    },
+    {
+      "epoch": 0.19052406136631847,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.133900443128864e-05,
+      "loss": 1.7812,
+      "step": 4359
+    },
+    {
+      "epoch": 0.19056776957034835,
+      "grad_norm": 2.5,
+      "learning_rate": 9.133514024030252e-05,
+      "loss": 2.3759,
+      "step": 4360
+    },
+    {
+      "epoch": 0.19061147777437826,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.133127526926215e-05,
+      "loss": 2.232,
+      "step": 4361
+    },
+    {
+      "epoch": 0.19065518597840814,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.132740951824046e-05,
+      "loss": 1.6692,
+      "step": 4362
+    },
+    {
+      "epoch": 0.19069889418243804,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.13235429873104e-05,
+      "loss": 2.4393,
+      "step": 4363
+    },
+    {
+      "epoch": 0.19074260238646795,
+      "grad_norm": 2.5,
+      "learning_rate": 9.131967567654493e-05,
+      "loss": 2.0248,
+      "step": 4364
+    },
+    {
+      "epoch": 0.19078631059049783,
+      "grad_norm": 2.25,
+      "learning_rate": 9.131580758601705e-05,
+      "loss": 1.9735,
+      "step": 4365
+    },
+    {
+      "epoch": 0.19083001879452774,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.131193871579975e-05,
+      "loss": 2.3554,
+      "step": 4366
+    },
+    {
+      "epoch": 0.19087372699855762,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.130806906596603e-05,
+      "loss": 2.4335,
+      "step": 4367
+    },
+    {
+      "epoch": 0.19091743520258753,
+      "grad_norm": 1.9765625,
+      "learning_rate": 9.130419863658894e-05,
+      "loss": 1.9289,
+      "step": 4368
+    },
+    {
+      "epoch": 0.19096114340661743,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.13003274277415e-05,
+      "loss": 2.2863,
+      "step": 4369
+    },
+    {
+      "epoch": 0.1910048516106473,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.129645543949676e-05,
+      "loss": 2.1855,
+      "step": 4370
+    },
+    {
+      "epoch": 0.19104855981467722,
+      "grad_norm": 2.0,
+      "learning_rate": 9.129258267192783e-05,
+      "loss": 1.8109,
+      "step": 4371
+    },
+    {
+      "epoch": 0.1910922680187071,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.128870912510774e-05,
+      "loss": 1.9122,
+      "step": 4372
+    },
+    {
+      "epoch": 0.191135976222737,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.128483479910963e-05,
+      "loss": 1.843,
+      "step": 4373
+    },
+    {
+      "epoch": 0.19117968442676692,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.12809596940066e-05,
+      "loss": 2.258,
+      "step": 4374
+    },
+    {
+      "epoch": 0.1912233926307968,
+      "grad_norm": 2.5,
+      "learning_rate": 9.127708380987176e-05,
+      "loss": 1.8263,
+      "step": 4375
+    },
+    {
+      "epoch": 0.1912671008348267,
+      "grad_norm": 2.625,
+      "learning_rate": 9.12732071467783e-05,
+      "loss": 1.9707,
+      "step": 4376
+    },
+    {
+      "epoch": 0.19131080903885658,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.126932970479933e-05,
+      "loss": 2.1258,
+      "step": 4377
+    },
+    {
+      "epoch": 0.1913545172428865,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.126545148400807e-05,
+      "loss": 1.8641,
+      "step": 4378
+    },
+    {
+      "epoch": 0.1913982254469164,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.126157248447766e-05,
+      "loss": 1.7545,
+      "step": 4379
+    },
+    {
+      "epoch": 0.19144193365094628,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.125769270628133e-05,
+      "loss": 1.8299,
+      "step": 4380
+    },
+    {
+      "epoch": 0.19148564185497619,
+      "grad_norm": 2.125,
+      "learning_rate": 9.125381214949229e-05,
+      "loss": 2.2791,
+      "step": 4381
+    },
+    {
+      "epoch": 0.19152935005900606,
+      "grad_norm": 3.890625,
+      "learning_rate": 9.124993081418375e-05,
+      "loss": 2.0627,
+      "step": 4382
+    },
+    {
+      "epoch": 0.19157305826303597,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.124604870042901e-05,
+      "loss": 2.8771,
+      "step": 4383
+    },
+    {
+      "epoch": 0.19161676646706588,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.12421658083013e-05,
+      "loss": 1.8637,
+      "step": 4384
+    },
+    {
+      "epoch": 0.19166047467109576,
+      "grad_norm": 1.890625,
+      "learning_rate": 9.123828213787389e-05,
+      "loss": 2.0092,
+      "step": 4385
+    },
+    {
+      "epoch": 0.19170418287512567,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.123439768922006e-05,
+      "loss": 1.5351,
+      "step": 4386
+    },
+    {
+      "epoch": 0.19174789107915555,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.123051246241314e-05,
+      "loss": 2.1595,
+      "step": 4387
+    },
+    {
+      "epoch": 0.19179159928318545,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.122662645752646e-05,
+      "loss": 1.3175,
+      "step": 4388
+    },
+    {
+      "epoch": 0.19183530748721536,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.122273967463331e-05,
+      "loss": 1.9585,
+      "step": 4389
+    },
+    {
+      "epoch": 0.19187901569124524,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.121885211380708e-05,
+      "loss": 1.9942,
+      "step": 4390
+    },
+    {
+      "epoch": 0.19192272389527515,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.121496377512114e-05,
+      "loss": 1.6896,
+      "step": 4391
+    },
+    {
+      "epoch": 0.19196643209930503,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.121107465864882e-05,
+      "loss": 2.2896,
+      "step": 4392
+    },
+    {
+      "epoch": 0.19201014030333494,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.120718476446354e-05,
+      "loss": 2.1606,
+      "step": 4393
+    },
+    {
+      "epoch": 0.19205384850736484,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.120329409263871e-05,
+      "loss": 1.8183,
+      "step": 4394
+    },
+    {
+      "epoch": 0.19209755671139472,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.119940264324776e-05,
+      "loss": 1.8949,
+      "step": 4395
+    },
+    {
+      "epoch": 0.19214126491542463,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.119551041636412e-05,
+      "loss": 2.3729,
+      "step": 4396
+    },
+    {
+      "epoch": 0.1921849731194545,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.119161741206123e-05,
+      "loss": 1.9106,
+      "step": 4397
+    },
+    {
+      "epoch": 0.19222868132348442,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.11877236304126e-05,
+      "loss": 2.1901,
+      "step": 4398
+    },
+    {
+      "epoch": 0.19227238952751433,
+      "grad_norm": 2.25,
+      "learning_rate": 9.118382907149165e-05,
+      "loss": 2.2365,
+      "step": 4399
+    },
+    {
+      "epoch": 0.1923160977315442,
+      "grad_norm": 1.9140625,
+      "learning_rate": 9.11799337353719e-05,
+      "loss": 1.8551,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1923598059355741,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.117603762212687e-05,
+      "loss": 1.8847,
+      "step": 4401
+    },
+    {
+      "epoch": 0.192403514139604,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.11721407318301e-05,
+      "loss": 3.003,
+      "step": 4402
+    },
+    {
+      "epoch": 0.1924472223436339,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.116824306455509e-05,
+      "loss": 1.9676,
+      "step": 4403
+    },
+    {
+      "epoch": 0.1924909305476638,
+      "grad_norm": 2.125,
+      "learning_rate": 9.116434462037545e-05,
+      "loss": 1.9876,
+      "step": 4404
+    },
+    {
+      "epoch": 0.1925346387516937,
+      "grad_norm": 3.40625,
+      "learning_rate": 9.116044539936468e-05,
+      "loss": 1.6742,
+      "step": 4405
+    },
+    {
+      "epoch": 0.1925783469557236,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.115654540159641e-05,
+      "loss": 2.2874,
+      "step": 4406
+    },
+    {
+      "epoch": 0.19262205515975347,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.115264462714421e-05,
+      "loss": 1.6118,
+      "step": 4407
+    },
+    {
+      "epoch": 0.19266576336378338,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.114874307608175e-05,
+      "loss": 1.7595,
+      "step": 4408
+    },
+    {
+      "epoch": 0.1927094715678133,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.114484074848259e-05,
+      "loss": 2.242,
+      "step": 4409
+    },
+    {
+      "epoch": 0.19275317977184317,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.11409376444204e-05,
+      "loss": 1.8576,
+      "step": 4410
+    },
+    {
+      "epoch": 0.19279688797587308,
+      "grad_norm": 3.46875,
+      "learning_rate": 9.113703376396885e-05,
+      "loss": 2.2897,
+      "step": 4411
+    },
+    {
+      "epoch": 0.19284059617990296,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.11331291072016e-05,
+      "loss": 2.1501,
+      "step": 4412
+    },
+    {
+      "epoch": 0.19288430438393286,
+      "grad_norm": 2.125,
+      "learning_rate": 9.112922367419234e-05,
+      "loss": 2.0837,
+      "step": 4413
+    },
+    {
+      "epoch": 0.19292801258796277,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.112531746501476e-05,
+      "loss": 1.7882,
+      "step": 4414
+    },
+    {
+      "epoch": 0.19297172079199265,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.112141047974259e-05,
+      "loss": 2.9361,
+      "step": 4415
+    },
+    {
+      "epoch": 0.19301542899602256,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.111750271844957e-05,
+      "loss": 2.1117,
+      "step": 4416
+    },
+    {
+      "epoch": 0.19305913720005244,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.11135941812094e-05,
+      "loss": 2.7356,
+      "step": 4417
+    },
+    {
+      "epoch": 0.19310284540408235,
+      "grad_norm": 2.125,
+      "learning_rate": 9.110968486809588e-05,
+      "loss": 1.5091,
+      "step": 4418
+    },
+    {
+      "epoch": 0.19314655360811225,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.110577477918279e-05,
+      "loss": 2.2072,
+      "step": 4419
+    },
+    {
+      "epoch": 0.19319026181214213,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.110186391454389e-05,
+      "loss": 1.8448,
+      "step": 4420
+    },
+    {
+      "epoch": 0.19323397001617204,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.1097952274253e-05,
+      "loss": 1.9389,
+      "step": 4421
+    },
+    {
+      "epoch": 0.19327767822020192,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.109403985838392e-05,
+      "loss": 2.0713,
+      "step": 4422
+    },
+    {
+      "epoch": 0.19332138642423183,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.109012666701051e-05,
+      "loss": 2.0326,
+      "step": 4423
+    },
+    {
+      "epoch": 0.19336509462826174,
+      "grad_norm": 2.5,
+      "learning_rate": 9.108621270020661e-05,
+      "loss": 2.5237,
+      "step": 4424
+    },
+    {
+      "epoch": 0.19340880283229162,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.108229795804609e-05,
+      "loss": 1.8756,
+      "step": 4425
+    },
+    {
+      "epoch": 0.19345251103632152,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.10783824406028e-05,
+      "loss": 1.8804,
+      "step": 4426
+    },
+    {
+      "epoch": 0.1934962192403514,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.107446614795063e-05,
+      "loss": 2.3902,
+      "step": 4427
+    },
+    {
+      "epoch": 0.1935399274443813,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.107054908016352e-05,
+      "loss": 2.1118,
+      "step": 4428
+    },
+    {
+      "epoch": 0.19358363564841122,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.106663123731539e-05,
+      "loss": 1.6956,
+      "step": 4429
+    },
+    {
+      "epoch": 0.1936273438524411,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.106271261948014e-05,
+      "loss": 2.6526,
+      "step": 4430
+    },
+    {
+      "epoch": 0.193671052056471,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.105879322673174e-05,
+      "loss": 1.9012,
+      "step": 4431
+    },
+    {
+      "epoch": 0.19371476026050088,
+      "grad_norm": 2.25,
+      "learning_rate": 9.105487305914416e-05,
+      "loss": 1.5692,
+      "step": 4432
+    },
+    {
+      "epoch": 0.1937584684645308,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.105095211679135e-05,
+      "loss": 2.1344,
+      "step": 4433
+    },
+    {
+      "epoch": 0.1938021766685607,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.104703039974736e-05,
+      "loss": 2.1873,
+      "step": 4434
+    },
+    {
+      "epoch": 0.19384588487259058,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.104310790808614e-05,
+      "loss": 2.5386,
+      "step": 4435
+    },
+    {
+      "epoch": 0.1938895930766205,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.103918464188175e-05,
+      "loss": 2.2476,
+      "step": 4436
+    },
+    {
+      "epoch": 0.19393330128065037,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.103526060120821e-05,
+      "loss": 1.8868,
+      "step": 4437
+    },
+    {
+      "epoch": 0.19397700948468027,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.103133578613959e-05,
+      "loss": 1.8679,
+      "step": 4438
+    },
+    {
+      "epoch": 0.19402071768871018,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.102741019674993e-05,
+      "loss": 1.5364,
+      "step": 4439
+    },
+    {
+      "epoch": 0.19406442589274006,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.102348383311334e-05,
+      "loss": 1.9936,
+      "step": 4440
+    },
+    {
+      "epoch": 0.19410813409676997,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.101955669530391e-05,
+      "loss": 2.3603,
+      "step": 4441
+    },
+    {
+      "epoch": 0.19415184230079985,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.101562878339572e-05,
+      "loss": 1.5734,
+      "step": 4442
+    },
+    {
+      "epoch": 0.19419555050482976,
+      "grad_norm": 2.125,
+      "learning_rate": 9.101170009746294e-05,
+      "loss": 1.8535,
+      "step": 4443
+    },
+    {
+      "epoch": 0.19423925870885966,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.100777063757969e-05,
+      "loss": 2.6233,
+      "step": 4444
+    },
+    {
+      "epoch": 0.19428296691288954,
+      "grad_norm": 3.453125,
+      "learning_rate": 9.100384040382011e-05,
+      "loss": 3.0462,
+      "step": 4445
+    },
+    {
+      "epoch": 0.19432667511691945,
+      "grad_norm": 2.75,
+      "learning_rate": 9.099990939625838e-05,
+      "loss": 2.0341,
+      "step": 4446
+    },
+    {
+      "epoch": 0.19437038332094933,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.099597761496869e-05,
+      "loss": 2.5488,
+      "step": 4447
+    },
+    {
+      "epoch": 0.19441409152497924,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.099204506002525e-05,
+      "loss": 1.9444,
+      "step": 4448
+    },
+    {
+      "epoch": 0.19445779972900915,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.098811173150225e-05,
+      "loss": 1.9015,
+      "step": 4449
+    },
+    {
+      "epoch": 0.19450150793303903,
+      "grad_norm": 2.625,
+      "learning_rate": 9.098417762947394e-05,
+      "loss": 1.7492,
+      "step": 4450
+    },
+    {
+      "epoch": 0.19454521613706893,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.098024275401454e-05,
+      "loss": 2.0373,
+      "step": 4451
+    },
+    {
+      "epoch": 0.1945889243410988,
+      "grad_norm": 2.25,
+      "learning_rate": 9.097630710519829e-05,
+      "loss": 2.2535,
+      "step": 4452
+    },
+    {
+      "epoch": 0.19463263254512872,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.097237068309951e-05,
+      "loss": 1.9008,
+      "step": 4453
+    },
+    {
+      "epoch": 0.19467634074915863,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.096843348779247e-05,
+      "loss": 2.2965,
+      "step": 4454
+    },
+    {
+      "epoch": 0.1947200489531885,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.096449551935144e-05,
+      "loss": 2.1173,
+      "step": 4455
+    },
+    {
+      "epoch": 0.19476375715721841,
+      "grad_norm": 2.0,
+      "learning_rate": 9.096055677785078e-05,
+      "loss": 1.7159,
+      "step": 4456
+    },
+    {
+      "epoch": 0.1948074653612483,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.09566172633648e-05,
+      "loss": 1.6598,
+      "step": 4457
+    },
+    {
+      "epoch": 0.1948511735652782,
+      "grad_norm": 2.125,
+      "learning_rate": 9.095267697596782e-05,
+      "loss": 1.8245,
+      "step": 4458
+    },
+    {
+      "epoch": 0.1948948817693081,
+      "grad_norm": 2.375,
+      "learning_rate": 9.094873591573423e-05,
+      "loss": 2.1265,
+      "step": 4459
+    },
+    {
+      "epoch": 0.194938589973338,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.094479408273841e-05,
+      "loss": 2.1805,
+      "step": 4460
+    },
+    {
+      "epoch": 0.1949822981773679,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.094085147705472e-05,
+      "loss": 2.1034,
+      "step": 4461
+    },
+    {
+      "epoch": 0.19502600638139778,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.093690809875758e-05,
+      "loss": 1.6653,
+      "step": 4462
+    },
+    {
+      "epoch": 0.19506971458542768,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.09329639479214e-05,
+      "loss": 1.5786,
+      "step": 4463
+    },
+    {
+      "epoch": 0.1951134227894576,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.092901902462062e-05,
+      "loss": 1.8306,
+      "step": 4464
+    },
+    {
+      "epoch": 0.19515713099348747,
+      "grad_norm": 3.484375,
+      "learning_rate": 9.092507332892968e-05,
+      "loss": 2.9709,
+      "step": 4465
+    },
+    {
+      "epoch": 0.19520083919751738,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.092112686092304e-05,
+      "loss": 2.0711,
+      "step": 4466
+    },
+    {
+      "epoch": 0.19524454740154726,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.091717962067518e-05,
+      "loss": 2.3881,
+      "step": 4467
+    },
+    {
+      "epoch": 0.19528825560557717,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.09132316082606e-05,
+      "loss": 2.128,
+      "step": 4468
+    },
+    {
+      "epoch": 0.19533196380960707,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.090928282375378e-05,
+      "loss": 1.5185,
+      "step": 4469
+    },
+    {
+      "epoch": 0.19537567201363695,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.090533326722924e-05,
+      "loss": 2.0833,
+      "step": 4470
+    },
+    {
+      "epoch": 0.19541938021766686,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.090138293876155e-05,
+      "loss": 2.3409,
+      "step": 4471
+    },
+    {
+      "epoch": 0.19546308842169674,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.089743183842523e-05,
+      "loss": 1.8522,
+      "step": 4472
+    },
+    {
+      "epoch": 0.19550679662572665,
+      "grad_norm": 2.0,
+      "learning_rate": 9.089347996629484e-05,
+      "loss": 1.7542,
+      "step": 4473
+    },
+    {
+      "epoch": 0.19555050482975656,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.088952732244495e-05,
+      "loss": 1.8944,
+      "step": 4474
+    },
+    {
+      "epoch": 0.19559421303378643,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.088557390695021e-05,
+      "loss": 1.8428,
+      "step": 4475
+    },
+    {
+      "epoch": 0.19563792123781634,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.088161971988516e-05,
+      "loss": 1.6356,
+      "step": 4476
+    },
+    {
+      "epoch": 0.19568162944184622,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.087766476132444e-05,
+      "loss": 1.7155,
+      "step": 4477
+    },
+    {
+      "epoch": 0.19572533764587613,
+      "grad_norm": 5.75,
+      "learning_rate": 9.08737090313427e-05,
+      "loss": 2.7836,
+      "step": 4478
+    },
+    {
+      "epoch": 0.19576904584990604,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.086975253001457e-05,
+      "loss": 1.6453,
+      "step": 4479
+    },
+    {
+      "epoch": 0.19581275405393592,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.086579525741475e-05,
+      "loss": 1.6787,
+      "step": 4480
+    },
+    {
+      "epoch": 0.19585646225796582,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.086183721361787e-05,
+      "loss": 2.1751,
+      "step": 4481
+    },
+    {
+      "epoch": 0.1959001704619957,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.085787839869866e-05,
+      "loss": 2.2441,
+      "step": 4482
+    },
+    {
+      "epoch": 0.1959438786660256,
+      "grad_norm": 2.125,
+      "learning_rate": 9.085391881273182e-05,
+      "loss": 1.4872,
+      "step": 4483
+    },
+    {
+      "epoch": 0.19598758687005552,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.084995845579208e-05,
+      "loss": 1.9066,
+      "step": 4484
+    },
+    {
+      "epoch": 0.1960312950740854,
+      "grad_norm": 2.25,
+      "learning_rate": 9.084599732795415e-05,
+      "loss": 2.2245,
+      "step": 4485
+    },
+    {
+      "epoch": 0.1960750032781153,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.084203542929282e-05,
+      "loss": 2.049,
+      "step": 4486
+    },
+    {
+      "epoch": 0.19611871148214519,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.083807275988284e-05,
+      "loss": 2.4972,
+      "step": 4487
+    },
+    {
+      "epoch": 0.1961624196861751,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.083410931979899e-05,
+      "loss": 1.5624,
+      "step": 4488
+    },
+    {
+      "epoch": 0.196206127890205,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.083014510911604e-05,
+      "loss": 1.945,
+      "step": 4489
+    },
+    {
+      "epoch": 0.19624983609423488,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.082618012790886e-05,
+      "loss": 2.0894,
+      "step": 4490
+    },
+    {
+      "epoch": 0.1962935442982648,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.082221437625223e-05,
+      "loss": 2.0207,
+      "step": 4491
+    },
+    {
+      "epoch": 0.19633725250229467,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.081824785422099e-05,
+      "loss": 1.7351,
+      "step": 4492
+    },
+    {
+      "epoch": 0.19638096070632458,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.081428056189e-05,
+      "loss": 3.121,
+      "step": 4493
+    },
+    {
+      "epoch": 0.19642466891035448,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.081031249933416e-05,
+      "loss": 2.1267,
+      "step": 4494
+    },
+    {
+      "epoch": 0.19646837711438436,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.080634366662832e-05,
+      "loss": 1.6989,
+      "step": 4495
+    },
+    {
+      "epoch": 0.19651208531841427,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.080237406384738e-05,
+      "loss": 1.9422,
+      "step": 4496
+    },
+    {
+      "epoch": 0.19655579352244415,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.079840369106625e-05,
+      "loss": 2.0947,
+      "step": 4497
+    },
+    {
+      "epoch": 0.19659950172647406,
+      "grad_norm": 3.421875,
+      "learning_rate": 9.079443254835987e-05,
+      "loss": 1.5288,
+      "step": 4498
+    },
+    {
+      "epoch": 0.19664320993050396,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.079046063580316e-05,
+      "loss": 1.7854,
+      "step": 4499
+    },
+    {
+      "epoch": 0.19668691813453384,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.07864879534711e-05,
+      "loss": 1.9261,
+      "step": 4500
+    },
+    {
+      "epoch": 0.19673062633856375,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.078251450143866e-05,
+      "loss": 1.7282,
+      "step": 4501
+    },
+    {
+      "epoch": 0.19677433454259363,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.077854027978078e-05,
+      "loss": 1.9388,
+      "step": 4502
+    },
+    {
+      "epoch": 0.19681804274662354,
+      "grad_norm": 2.875,
+      "learning_rate": 9.077456528857254e-05,
+      "loss": 1.6096,
+      "step": 4503
+    },
+    {
+      "epoch": 0.19686175095065345,
+      "grad_norm": 3.625,
+      "learning_rate": 9.077058952788888e-05,
+      "loss": 1.4145,
+      "step": 4504
+    },
+    {
+      "epoch": 0.19690545915468333,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.076661299780486e-05,
+      "loss": 2.0563,
+      "step": 4505
+    },
+    {
+      "epoch": 0.19694916735871323,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.076263569839553e-05,
+      "loss": 1.9754,
+      "step": 4506
+    },
+    {
+      "epoch": 0.1969928755627431,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.075865762973595e-05,
+      "loss": 1.9317,
+      "step": 4507
+    },
+    {
+      "epoch": 0.19703658376677302,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.075467879190117e-05,
+      "loss": 1.9321,
+      "step": 4508
+    },
+    {
+      "epoch": 0.19708029197080293,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.075069918496625e-05,
+      "loss": 2.1764,
+      "step": 4509
+    },
+    {
+      "epoch": 0.1971240001748328,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.074671880900636e-05,
+      "loss": 1.9645,
+      "step": 4510
+    },
+    {
+      "epoch": 0.19716770837886272,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.074273766409657e-05,
+      "loss": 1.7994,
+      "step": 4511
+    },
+    {
+      "epoch": 0.1972114165828926,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.073875575031203e-05,
+      "loss": 2.1177,
+      "step": 4512
+    },
+    {
+      "epoch": 0.1972551247869225,
+      "grad_norm": 2.375,
+      "learning_rate": 9.073477306772789e-05,
+      "loss": 1.6883,
+      "step": 4513
+    },
+    {
+      "epoch": 0.1972988329909524,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.073078961641928e-05,
+      "loss": 2.5953,
+      "step": 4514
+    },
+    {
+      "epoch": 0.1973425411949823,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.07268053964614e-05,
+      "loss": 2.3189,
+      "step": 4515
+    },
+    {
+      "epoch": 0.1973862493990122,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.072282040792939e-05,
+      "loss": 2.5728,
+      "step": 4516
+    },
+    {
+      "epoch": 0.19742995760304208,
+      "grad_norm": 2.75,
+      "learning_rate": 9.071883465089852e-05,
+      "loss": 2.2928,
+      "step": 4517
+    },
+    {
+      "epoch": 0.19747366580707199,
+      "grad_norm": 5.96875,
+      "learning_rate": 9.071484812544398e-05,
+      "loss": 1.8055,
+      "step": 4518
+    },
+    {
+      "epoch": 0.1975173740111019,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.071086083164099e-05,
+      "loss": 3.7166,
+      "step": 4519
+    },
+    {
+      "epoch": 0.19756108221513177,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.07068727695648e-05,
+      "loss": 1.6634,
+      "step": 4520
+    },
+    {
+      "epoch": 0.19760479041916168,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.07028839392907e-05,
+      "loss": 2.0254,
+      "step": 4521
+    },
+    {
+      "epoch": 0.19764849862319156,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.069889434089391e-05,
+      "loss": 1.9842,
+      "step": 4522
+    },
+    {
+      "epoch": 0.19769220682722147,
+      "grad_norm": 2.125,
+      "learning_rate": 9.069490397444975e-05,
+      "loss": 1.9428,
+      "step": 4523
+    },
+    {
+      "epoch": 0.19773591503125137,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.069091284003354e-05,
+      "loss": 1.9117,
+      "step": 4524
+    },
+    {
+      "epoch": 0.19777962323528125,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.068692093772058e-05,
+      "loss": 2.1338,
+      "step": 4525
+    },
+    {
+      "epoch": 0.19782333143931116,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.06829282675862e-05,
+      "loss": 2.0547,
+      "step": 4526
+    },
+    {
+      "epoch": 0.19786703964334104,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.067893482970575e-05,
+      "loss": 1.9351,
+      "step": 4527
+    },
+    {
+      "epoch": 0.19791074784737095,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.06749406241546e-05,
+      "loss": 2.3361,
+      "step": 4528
+    },
+    {
+      "epoch": 0.19795445605140086,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.067094565100811e-05,
+      "loss": 1.8335,
+      "step": 4529
+    },
+    {
+      "epoch": 0.19799816425543074,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.066694991034169e-05,
+      "loss": 2.0857,
+      "step": 4530
+    },
+    {
+      "epoch": 0.19804187245946064,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.066295340223073e-05,
+      "loss": 2.4841,
+      "step": 4531
+    },
+    {
+      "epoch": 0.19808558066349052,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.065895612675066e-05,
+      "loss": 2.5671,
+      "step": 4532
+    },
+    {
+      "epoch": 0.19812928886752043,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.065495808397693e-05,
+      "loss": 1.6286,
+      "step": 4533
+    },
+    {
+      "epoch": 0.19817299707155034,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.065095927398495e-05,
+      "loss": 1.6137,
+      "step": 4534
+    },
+    {
+      "epoch": 0.19821670527558022,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.064695969685021e-05,
+      "loss": 1.7371,
+      "step": 4535
+    },
+    {
+      "epoch": 0.19826041347961013,
+      "grad_norm": 2.5,
+      "learning_rate": 9.064295935264818e-05,
+      "loss": 1.8502,
+      "step": 4536
+    },
+    {
+      "epoch": 0.19830412168364,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.063895824145435e-05,
+      "loss": 2.0257,
+      "step": 4537
+    },
+    {
+      "epoch": 0.1983478298876699,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.063495636334423e-05,
+      "loss": 1.7634,
+      "step": 4538
+    },
+    {
+      "epoch": 0.19839153809169982,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.063095371839337e-05,
+      "loss": 1.9927,
+      "step": 4539
+    },
+    {
+      "epoch": 0.1984352462957297,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.062695030667724e-05,
+      "loss": 2.03,
+      "step": 4540
+    },
+    {
+      "epoch": 0.1984789544997596,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.062294612827145e-05,
+      "loss": 1.8626,
+      "step": 4541
+    },
+    {
+      "epoch": 0.1985226627037895,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.061894118325154e-05,
+      "loss": 2.0393,
+      "step": 4542
+    },
+    {
+      "epoch": 0.1985663709078194,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.061493547169308e-05,
+      "loss": 1.5743,
+      "step": 4543
+    },
+    {
+      "epoch": 0.1986100791118493,
+      "grad_norm": 2.125,
+      "learning_rate": 9.061092899367169e-05,
+      "loss": 2.0798,
+      "step": 4544
+    },
+    {
+      "epoch": 0.19865378731587918,
+      "grad_norm": 8.375,
+      "learning_rate": 9.060692174926296e-05,
+      "loss": 2.4696,
+      "step": 4545
+    },
+    {
+      "epoch": 0.1986974955199091,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.060291373854251e-05,
+      "loss": 1.7727,
+      "step": 4546
+    },
+    {
+      "epoch": 0.19874120372393897,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.059890496158599e-05,
+      "loss": 2.1714,
+      "step": 4547
+    },
+    {
+      "epoch": 0.19878491192796888,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.059489541846903e-05,
+      "loss": 1.4298,
+      "step": 4548
+    },
+    {
+      "epoch": 0.19882862013199878,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.059088510926732e-05,
+      "loss": 2.2674,
+      "step": 4549
+    },
+    {
+      "epoch": 0.19887232833602866,
+      "grad_norm": 25.75,
+      "learning_rate": 9.058687403405653e-05,
+      "loss": 2.4088,
+      "step": 4550
+    },
+    {
+      "epoch": 0.19891603654005857,
+      "grad_norm": 1.90625,
+      "learning_rate": 9.058286219291237e-05,
+      "loss": 1.5939,
+      "step": 4551
+    },
+    {
+      "epoch": 0.19895974474408848,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.057884958591052e-05,
+      "loss": 2.0231,
+      "step": 4552
+    },
+    {
+      "epoch": 0.19900345294811836,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.057483621312671e-05,
+      "loss": 1.5278,
+      "step": 4553
+    },
+    {
+      "epoch": 0.19904716115214827,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.05708220746367e-05,
+      "loss": 2.2683,
+      "step": 4554
+    },
+    {
+      "epoch": 0.19909086935617815,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.056680717051622e-05,
+      "loss": 1.9358,
+      "step": 4555
+    },
+    {
+      "epoch": 0.19913457756020805,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.056279150084106e-05,
+      "loss": 2.4731,
+      "step": 4556
+    },
+    {
+      "epoch": 0.19917828576423796,
+      "grad_norm": 2.0625,
+      "learning_rate": 9.055877506568699e-05,
+      "loss": 1.8777,
+      "step": 4557
+    },
+    {
+      "epoch": 0.19922199396826784,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.05547578651298e-05,
+      "loss": 1.6332,
+      "step": 4558
+    },
+    {
+      "epoch": 0.19926570217229775,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.05507398992453e-05,
+      "loss": 2.3121,
+      "step": 4559
+    },
+    {
+      "epoch": 0.19930941037632763,
+      "grad_norm": 2.625,
+      "learning_rate": 9.054672116810932e-05,
+      "loss": 2.1761,
+      "step": 4560
+    },
+    {
+      "epoch": 0.19935311858035754,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.054270167179768e-05,
+      "loss": 2.2086,
+      "step": 4561
+    },
+    {
+      "epoch": 0.19939682678438744,
+      "grad_norm": 2.125,
+      "learning_rate": 9.053868141038628e-05,
+      "loss": 1.5267,
+      "step": 4562
+    },
+    {
+      "epoch": 0.19944053498841732,
+      "grad_norm": 2.25,
+      "learning_rate": 9.053466038395096e-05,
+      "loss": 1.4548,
+      "step": 4563
+    },
+    {
+      "epoch": 0.19948424319244723,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.053063859256758e-05,
+      "loss": 1.987,
+      "step": 4564
+    },
+    {
+      "epoch": 0.1995279513964771,
+      "grad_norm": 6.9375,
+      "learning_rate": 9.052661603631207e-05,
+      "loss": 1.8233,
+      "step": 4565
+    },
+    {
+      "epoch": 0.19957165960050702,
+      "grad_norm": 6.25,
+      "learning_rate": 9.052259271526033e-05,
+      "loss": 3.0517,
+      "step": 4566
+    },
+    {
+      "epoch": 0.19961536780453693,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.05185686294883e-05,
+      "loss": 1.7139,
+      "step": 4567
+    },
+    {
+      "epoch": 0.1996590760085668,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.051454377907191e-05,
+      "loss": 2.6329,
+      "step": 4568
+    },
+    {
+      "epoch": 0.1997027842125967,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.05105181640871e-05,
+      "loss": 2.0316,
+      "step": 4569
+    },
+    {
+      "epoch": 0.1997464924166266,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.050649178460987e-05,
+      "loss": 2.1056,
+      "step": 4570
+    },
+    {
+      "epoch": 0.1997902006206565,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.050246464071616e-05,
+      "loss": 2.4229,
+      "step": 4571
+    },
+    {
+      "epoch": 0.1998339088246864,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.049843673248199e-05,
+      "loss": 1.7415,
+      "step": 4572
+    },
+    {
+      "epoch": 0.1998776170287163,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.04944080599834e-05,
+      "loss": 2.1538,
+      "step": 4573
+    },
+    {
+      "epoch": 0.1999213252327462,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.04903786232964e-05,
+      "loss": 1.956,
+      "step": 4574
+    },
+    {
+      "epoch": 0.19996503343677607,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.048634842249699e-05,
+      "loss": 2.8857,
+      "step": 4575
+    },
+    {
+      "epoch": 0.20000874164080598,
+      "grad_norm": 2.109375,
+      "learning_rate": 9.048231745766129e-05,
+      "loss": 1.7488,
+      "step": 4576
+    },
+    {
+      "epoch": 0.2000524498448359,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.047828572886532e-05,
+      "loss": 2.6818,
+      "step": 4577
+    },
+    {
+      "epoch": 0.20009615804886577,
+      "grad_norm": 1.84375,
+      "learning_rate": 9.04742532361852e-05,
+      "loss": 1.7045,
+      "step": 4578
+    },
+    {
+      "epoch": 0.20013986625289568,
+      "grad_norm": 2.625,
+      "learning_rate": 9.047021997969701e-05,
+      "loss": 1.9284,
+      "step": 4579
+    },
+    {
+      "epoch": 0.20018357445692556,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.046618595947687e-05,
+      "loss": 2.3077,
+      "step": 4580
+    },
+    {
+      "epoch": 0.20022728266095546,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.04621511756009e-05,
+      "loss": 1.7883,
+      "step": 4581
+    },
+    {
+      "epoch": 0.20027099086498537,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.045811562814525e-05,
+      "loss": 1.8529,
+      "step": 4582
+    },
+    {
+      "epoch": 0.20031469906901525,
+      "grad_norm": 2.125,
+      "learning_rate": 9.045407931718608e-05,
+      "loss": 1.7938,
+      "step": 4583
+    },
+    {
+      "epoch": 0.20035840727304516,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.045004224279954e-05,
+      "loss": 1.9549,
+      "step": 4584
+    },
+    {
+      "epoch": 0.20040211547707504,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.044600440506184e-05,
+      "loss": 1.5256,
+      "step": 4585
+    },
+    {
+      "epoch": 0.20044582368110495,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.044196580404917e-05,
+      "loss": 2.6624,
+      "step": 4586
+    },
+    {
+      "epoch": 0.20048953188513485,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.043792643983774e-05,
+      "loss": 3.1266,
+      "step": 4587
+    },
+    {
+      "epoch": 0.20053324008916473,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.04338863125038e-05,
+      "loss": 1.8653,
+      "step": 4588
+    },
+    {
+      "epoch": 0.20057694829319464,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.042984542212355e-05,
+      "loss": 2.1311,
+      "step": 4589
+    },
+    {
+      "epoch": 0.20062065649722452,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.042580376877329e-05,
+      "loss": 1.9296,
+      "step": 4590
+    },
+    {
+      "epoch": 0.20066436470125443,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.042176135252926e-05,
+      "loss": 2.8345,
+      "step": 4591
+    },
+    {
+      "epoch": 0.20070807290528433,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.041771817346778e-05,
+      "loss": 1.6019,
+      "step": 4592
+    },
+    {
+      "epoch": 0.20075178110931421,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.041367423166513e-05,
+      "loss": 1.8668,
+      "step": 4593
+    },
+    {
+      "epoch": 0.20079548931334412,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.040962952719763e-05,
+      "loss": 2.3199,
+      "step": 4594
+    },
+    {
+      "epoch": 0.200839197517374,
+      "grad_norm": 3.125,
+      "learning_rate": 9.040558406014161e-05,
+      "loss": 1.8228,
+      "step": 4595
+    },
+    {
+      "epoch": 0.2008829057214039,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.040153783057342e-05,
+      "loss": 2.0666,
+      "step": 4596
+    },
+    {
+      "epoch": 0.20092661392543382,
+      "grad_norm": 3.203125,
+      "learning_rate": 9.039749083856938e-05,
+      "loss": 2.0252,
+      "step": 4597
+    },
+    {
+      "epoch": 0.2009703221294637,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.039344308420591e-05,
+      "loss": 2.374,
+      "step": 4598
+    },
+    {
+      "epoch": 0.2010140303334936,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.038939456755938e-05,
+      "loss": 2.3647,
+      "step": 4599
+    },
+    {
+      "epoch": 0.20105773853752348,
+      "grad_norm": 2.375,
+      "learning_rate": 9.038534528870618e-05,
+      "loss": 1.8677,
+      "step": 4600
+    },
+    {
+      "epoch": 0.2011014467415534,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.038129524772274e-05,
+      "loss": 1.8486,
+      "step": 4601
+    },
+    {
+      "epoch": 0.2011451549455833,
+      "grad_norm": 1.9921875,
+      "learning_rate": 9.03772444446855e-05,
+      "loss": 1.3506,
+      "step": 4602
+    },
+    {
+      "epoch": 0.20118886314961318,
+      "grad_norm": 1.875,
+      "learning_rate": 9.037319287967088e-05,
+      "loss": 1.6946,
+      "step": 4603
+    },
+    {
+      "epoch": 0.20123257135364309,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.036914055275534e-05,
+      "loss": 1.5738,
+      "step": 4604
+    },
+    {
+      "epoch": 0.20127627955767297,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.036508746401538e-05,
+      "loss": 1.9645,
+      "step": 4605
+    },
+    {
+      "epoch": 0.20131998776170287,
+      "grad_norm": 2.375,
+      "learning_rate": 9.036103361352746e-05,
+      "loss": 1.9012,
+      "step": 4606
+    },
+    {
+      "epoch": 0.20136369596573278,
+      "grad_norm": 3.3125,
+      "learning_rate": 9.03569790013681e-05,
+      "loss": 2.5034,
+      "step": 4607
+    },
+    {
+      "epoch": 0.20140740416976266,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.035292362761381e-05,
+      "loss": 1.9332,
+      "step": 4608
+    },
+    {
+      "epoch": 0.20145111237379257,
+      "grad_norm": 2.1875,
+      "learning_rate": 9.034886749234111e-05,
+      "loss": 1.7966,
+      "step": 4609
+    },
+    {
+      "epoch": 0.20149482057782245,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.034481059562657e-05,
+      "loss": 2.015,
+      "step": 4610
+    },
+    {
+      "epoch": 0.20153852878185236,
+      "grad_norm": 2.25,
+      "learning_rate": 9.034075293754672e-05,
+      "loss": 2.0806,
+      "step": 4611
+    },
+    {
+      "epoch": 0.20158223698588226,
+      "grad_norm": 2.375,
+      "learning_rate": 9.033669451817817e-05,
+      "loss": 2.1522,
+      "step": 4612
+    },
+    {
+      "epoch": 0.20162594518991214,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.033263533759748e-05,
+      "loss": 1.8821,
+      "step": 4613
+    },
+    {
+      "epoch": 0.20166965339394205,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.032857539588124e-05,
+      "loss": 2.1767,
+      "step": 4614
+    },
+    {
+      "epoch": 0.20171336159797193,
+      "grad_norm": 2.125,
+      "learning_rate": 9.03245146931061e-05,
+      "loss": 2.1203,
+      "step": 4615
+    },
+    {
+      "epoch": 0.20175706980200184,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.032045322934868e-05,
+      "loss": 2.2441,
+      "step": 4616
+    },
+    {
+      "epoch": 0.20180077800603174,
+      "grad_norm": 4.71875,
+      "learning_rate": 9.031639100468563e-05,
+      "loss": 1.45,
+      "step": 4617
+    },
+    {
+      "epoch": 0.20184448621006162,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.031232801919359e-05,
+      "loss": 1.8462,
+      "step": 4618
+    },
+    {
+      "epoch": 0.20188819441409153,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.030826427294924e-05,
+      "loss": 2.1721,
+      "step": 4619
+    },
+    {
+      "epoch": 0.2019319026181214,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.03041997660293e-05,
+      "loss": 1.9742,
+      "step": 4620
+    },
+    {
+      "epoch": 0.20197561082215132,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.030013449851045e-05,
+      "loss": 1.8034,
+      "step": 4621
+    },
+    {
+      "epoch": 0.20201931902618123,
+      "grad_norm": 2.375,
+      "learning_rate": 9.029606847046941e-05,
+      "loss": 2.3671,
+      "step": 4622
+    },
+    {
+      "epoch": 0.2020630272302111,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.029200168198289e-05,
+      "loss": 2.5159,
+      "step": 4623
+    },
+    {
+      "epoch": 0.202106735434241,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.028793413312767e-05,
+      "loss": 2.3314,
+      "step": 4624
+    },
+    {
+      "epoch": 0.2021504436382709,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.02838658239805e-05,
+      "loss": 1.9898,
+      "step": 4625
+    },
+    {
+      "epoch": 0.2021941518423008,
+      "grad_norm": 2.5,
+      "learning_rate": 9.027979675461814e-05,
+      "loss": 2.2,
+      "step": 4626
+    },
+    {
+      "epoch": 0.2022378600463307,
+      "grad_norm": 2.125,
+      "learning_rate": 9.027572692511739e-05,
+      "loss": 2.3884,
+      "step": 4627
+    },
+    {
+      "epoch": 0.2022815682503606,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.027165633555507e-05,
+      "loss": 2.0405,
+      "step": 4628
+    },
+    {
+      "epoch": 0.2023252764543905,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.026758498600797e-05,
+      "loss": 2.6007,
+      "step": 4629
+    },
+    {
+      "epoch": 0.20236898465842038,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.026351287655294e-05,
+      "loss": 2.4496,
+      "step": 4630
+    },
+    {
+      "epoch": 0.20241269286245028,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.02594400072668e-05,
+      "loss": 2.2956,
+      "step": 4631
+    },
+    {
+      "epoch": 0.2024564010664802,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.025536637822647e-05,
+      "loss": 1.6348,
+      "step": 4632
+    },
+    {
+      "epoch": 0.20250010927051007,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.025129198950877e-05,
+      "loss": 2.1001,
+      "step": 4633
+    },
+    {
+      "epoch": 0.20254381747453998,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.02472168411906e-05,
+      "loss": 1.516,
+      "step": 4634
+    },
+    {
+      "epoch": 0.20258752567856986,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.024314093334886e-05,
+      "loss": 2.2612,
+      "step": 4635
+    },
+    {
+      "epoch": 0.20263123388259976,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.02390642660605e-05,
+      "loss": 1.5756,
+      "step": 4636
+    },
+    {
+      "epoch": 0.20267494208662967,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.023498683940243e-05,
+      "loss": 1.7007,
+      "step": 4637
+    },
+    {
+      "epoch": 0.20271865029065955,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.02309086534516e-05,
+      "loss": 1.9744,
+      "step": 4638
+    },
+    {
+      "epoch": 0.20276235849468946,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.022682970828497e-05,
+      "loss": 2.396,
+      "step": 4639
+    },
+    {
+      "epoch": 0.20280606669871934,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.022275000397951e-05,
+      "loss": 1.851,
+      "step": 4640
+    },
+    {
+      "epoch": 0.20284977490274925,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.021866954061223e-05,
+      "loss": 1.8836,
+      "step": 4641
+    },
+    {
+      "epoch": 0.20289348310677915,
+      "grad_norm": 2.625,
+      "learning_rate": 9.02145883182601e-05,
+      "loss": 2.2614,
+      "step": 4642
+    },
+    {
+      "epoch": 0.20293719131080903,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.021050633700018e-05,
+      "loss": 2.3415,
+      "step": 4643
+    },
+    {
+      "epoch": 0.20298089951483894,
+      "grad_norm": 2.75,
+      "learning_rate": 9.020642359690947e-05,
+      "loss": 1.7504,
+      "step": 4644
+    },
+    {
+      "epoch": 0.20302460771886882,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.020234009806503e-05,
+      "loss": 1.7237,
+      "step": 4645
+    },
+    {
+      "epoch": 0.20306831592289873,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.01982558405439e-05,
+      "loss": 2.115,
+      "step": 4646
+    },
+    {
+      "epoch": 0.20311202412692864,
+      "grad_norm": 2.875,
+      "learning_rate": 9.019417082442321e-05,
+      "loss": 2.3209,
+      "step": 4647
+    },
+    {
+      "epoch": 0.20315573233095852,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.019008504978e-05,
+      "loss": 2.3012,
+      "step": 4648
+    },
+    {
+      "epoch": 0.20319944053498842,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.01859985166914e-05,
+      "loss": 1.8486,
+      "step": 4649
+    },
+    {
+      "epoch": 0.2032431487390183,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.018191122523452e-05,
+      "loss": 1.7963,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2032868569430482,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.017782317548649e-05,
+      "loss": 1.8977,
+      "step": 4651
+    },
+    {
+      "epoch": 0.20333056514707812,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.017373436752445e-05,
+      "loss": 2.0669,
+      "step": 4652
+    },
+    {
+      "epoch": 0.203374273351108,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.016964480142557e-05,
+      "loss": 2.3295,
+      "step": 4653
+    },
+    {
+      "epoch": 0.2034179815551379,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.016555447726704e-05,
+      "loss": 1.9676,
+      "step": 4654
+    },
+    {
+      "epoch": 0.20346168975916779,
+      "grad_norm": 3.921875,
+      "learning_rate": 9.016146339512606e-05,
+      "loss": 1.961,
+      "step": 4655
+    },
+    {
+      "epoch": 0.2035053979631977,
+      "grad_norm": 2.5,
+      "learning_rate": 9.015737155507978e-05,
+      "loss": 2.3247,
+      "step": 4656
+    },
+    {
+      "epoch": 0.2035491061672276,
+      "grad_norm": 2.046875,
+      "learning_rate": 9.015327895720547e-05,
+      "loss": 1.9737,
+      "step": 4657
+    },
+    {
+      "epoch": 0.20359281437125748,
+      "grad_norm": 2.125,
+      "learning_rate": 9.014918560158035e-05,
+      "loss": 2.0781,
+      "step": 4658
+    },
+    {
+      "epoch": 0.2036365225752874,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.014509148828164e-05,
+      "loss": 2.504,
+      "step": 4659
+    },
+    {
+      "epoch": 0.20368023077931727,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.014099661738664e-05,
+      "loss": 1.7876,
+      "step": 4660
+    },
+    {
+      "epoch": 0.20372393898334717,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.01369009889726e-05,
+      "loss": 1.7947,
+      "step": 4661
+    },
+    {
+      "epoch": 0.20376764718737708,
+      "grad_norm": 3.546875,
+      "learning_rate": 9.013280460311684e-05,
+      "loss": 3.0699,
+      "step": 4662
+    },
+    {
+      "epoch": 0.20381135539140696,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.012870745989663e-05,
+      "loss": 2.0769,
+      "step": 4663
+    },
+    {
+      "epoch": 0.20385506359543687,
+      "grad_norm": 3.78125,
+      "learning_rate": 9.012460955938933e-05,
+      "loss": 1.6031,
+      "step": 4664
+    },
+    {
+      "epoch": 0.20389877179946675,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.012051090167222e-05,
+      "loss": 2.1118,
+      "step": 4665
+    },
+    {
+      "epoch": 0.20394248000349666,
+      "grad_norm": 4.28125,
+      "learning_rate": 9.01164114868227e-05,
+      "loss": 1.6598,
+      "step": 4666
+    },
+    {
+      "epoch": 0.20398618820752656,
+      "grad_norm": 2.25,
+      "learning_rate": 9.011231131491808e-05,
+      "loss": 2.263,
+      "step": 4667
+    },
+    {
+      "epoch": 0.20402989641155644,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.01082103860358e-05,
+      "loss": 1.7258,
+      "step": 4668
+    },
+    {
+      "epoch": 0.20407360461558635,
+      "grad_norm": 2.625,
+      "learning_rate": 9.010410870025319e-05,
+      "loss": 2.1402,
+      "step": 4669
+    },
+    {
+      "epoch": 0.20411731281961623,
+      "grad_norm": 1.9140625,
+      "learning_rate": 9.01000062576477e-05,
+      "loss": 1.9602,
+      "step": 4670
+    },
+    {
+      "epoch": 0.20416102102364614,
+      "grad_norm": 3.6875,
+      "learning_rate": 9.009590305829672e-05,
+      "loss": 2.8867,
+      "step": 4671
+    },
+    {
+      "epoch": 0.20420472922767605,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.009179910227768e-05,
+      "loss": 1.8176,
+      "step": 4672
+    },
+    {
+      "epoch": 0.20424843743170593,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.008769438966805e-05,
+      "loss": 2.1523,
+      "step": 4673
+    },
+    {
+      "epoch": 0.20429214563573583,
+      "grad_norm": 1.953125,
+      "learning_rate": 9.008358892054528e-05,
+      "loss": 1.8026,
+      "step": 4674
+    },
+    {
+      "epoch": 0.2043358538397657,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.007948269498685e-05,
+      "loss": 1.8456,
+      "step": 4675
+    },
+    {
+      "epoch": 0.20437956204379562,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.007537571307025e-05,
+      "loss": 1.8779,
+      "step": 4676
+    },
+    {
+      "epoch": 0.20442327024782553,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.007126797487298e-05,
+      "loss": 2.1508,
+      "step": 4677
+    },
+    {
+      "epoch": 0.2044669784518554,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.006715948047257e-05,
+      "loss": 2.1961,
+      "step": 4678
+    },
+    {
+      "epoch": 0.20451068665588532,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.006305022994654e-05,
+      "loss": 1.837,
+      "step": 4679
+    },
+    {
+      "epoch": 0.2045543948599152,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.005894022337245e-05,
+      "loss": 2.0313,
+      "step": 4680
+    },
+    {
+      "epoch": 0.2045981030639451,
+      "grad_norm": 6.875,
+      "learning_rate": 9.005482946082784e-05,
+      "loss": 1.9768,
+      "step": 4681
+    },
+    {
+      "epoch": 0.204641811267975,
+      "grad_norm": 2.09375,
+      "learning_rate": 9.00507179423903e-05,
+      "loss": 1.8533,
+      "step": 4682
+    },
+    {
+      "epoch": 0.2046855194720049,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.004660566813744e-05,
+      "loss": 1.7265,
+      "step": 4683
+    },
+    {
+      "epoch": 0.2047292276760348,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.004249263814683e-05,
+      "loss": 1.9114,
+      "step": 4684
+    },
+    {
+      "epoch": 0.20477293588006468,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.003837885249612e-05,
+      "loss": 1.8118,
+      "step": 4685
+    },
+    {
+      "epoch": 0.20481664408409458,
+      "grad_norm": 4.09375,
+      "learning_rate": 9.003426431126291e-05,
+      "loss": 1.7946,
+      "step": 4686
+    },
+    {
+      "epoch": 0.2048603522881245,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.003014901452488e-05,
+      "loss": 2.2468,
+      "step": 4687
+    },
+    {
+      "epoch": 0.20490406049215437,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.002603296235967e-05,
+      "loss": 1.8963,
+      "step": 4688
+    },
+    {
+      "epoch": 0.20494776869618428,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.002191615484496e-05,
+      "loss": 2.3479,
+      "step": 4689
+    },
+    {
+      "epoch": 0.20499147690021416,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.001779859205845e-05,
+      "loss": 1.614,
+      "step": 4690
+    },
+    {
+      "epoch": 0.20503518510424407,
+      "grad_norm": 1.9296875,
+      "learning_rate": 9.001368027407784e-05,
+      "loss": 1.6517,
+      "step": 4691
+    },
+    {
+      "epoch": 0.20507889330827397,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.000956120098084e-05,
+      "loss": 2.1384,
+      "step": 4692
+    },
+    {
+      "epoch": 0.20512260151230385,
+      "grad_norm": 1.9609375,
+      "learning_rate": 9.000544137284519e-05,
+      "loss": 1.7031,
+      "step": 4693
+    },
+    {
+      "epoch": 0.20516630971633376,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.000132078974863e-05,
+      "loss": 1.6414,
+      "step": 4694
+    },
+    {
+      "epoch": 0.20521001792036364,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.999719945176894e-05,
+      "loss": 2.316,
+      "step": 4695
+    },
+    {
+      "epoch": 0.20525372612439355,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.999307735898387e-05,
+      "loss": 1.9569,
+      "step": 4696
+    },
+    {
+      "epoch": 0.20529743432842346,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.998895451147125e-05,
+      "loss": 1.6369,
+      "step": 4697
+    },
+    {
+      "epoch": 0.20534114253245334,
+      "grad_norm": 2.25,
+      "learning_rate": 8.998483090930883e-05,
+      "loss": 1.8258,
+      "step": 4698
+    },
+    {
+      "epoch": 0.20538485073648324,
+      "grad_norm": 2.375,
+      "learning_rate": 8.998070655257447e-05,
+      "loss": 2.0003,
+      "step": 4699
+    },
+    {
+      "epoch": 0.20542855894051312,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.997658144134598e-05,
+      "loss": 1.6348,
+      "step": 4700
+    },
+    {
+      "epoch": 0.20547226714454303,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.997245557570121e-05,
+      "loss": 1.7787,
+      "step": 4701
+    },
+    {
+      "epoch": 0.20551597534857294,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.996832895571803e-05,
+      "loss": 1.9775,
+      "step": 4702
+    },
+    {
+      "epoch": 0.20555968355260282,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.996420158147431e-05,
+      "loss": 1.8802,
+      "step": 4703
+    },
+    {
+      "epoch": 0.20560339175663273,
+      "grad_norm": 29.75,
+      "learning_rate": 8.996007345304795e-05,
+      "loss": 3.9084,
+      "step": 4704
+    },
+    {
+      "epoch": 0.2056470999606626,
+      "grad_norm": 18.125,
+      "learning_rate": 8.995594457051684e-05,
+      "loss": 1.2003,
+      "step": 4705
+    },
+    {
+      "epoch": 0.2056908081646925,
+      "grad_norm": 2.75,
+      "learning_rate": 8.99518149339589e-05,
+      "loss": 2.1783,
+      "step": 4706
+    },
+    {
+      "epoch": 0.20573451636872242,
+      "grad_norm": 2.75,
+      "learning_rate": 8.994768454345206e-05,
+      "loss": 1.7874,
+      "step": 4707
+    },
+    {
+      "epoch": 0.2057782245727523,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.994355339907429e-05,
+      "loss": 2.057,
+      "step": 4708
+    },
+    {
+      "epoch": 0.2058219327767822,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.993942150090352e-05,
+      "loss": 2.2157,
+      "step": 4709
+    },
+    {
+      "epoch": 0.2058656409808121,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.993528884901774e-05,
+      "loss": 2.5268,
+      "step": 4710
+    },
+    {
+      "epoch": 0.205909349184842,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.993115544349493e-05,
+      "loss": 2.3586,
+      "step": 4711
+    },
+    {
+      "epoch": 0.2059530573888719,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.992702128441311e-05,
+      "loss": 1.8512,
+      "step": 4712
+    },
+    {
+      "epoch": 0.20599676559290178,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.992288637185028e-05,
+      "loss": 1.7177,
+      "step": 4713
+    },
+    {
+      "epoch": 0.2060404737969317,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.991875070588447e-05,
+      "loss": 1.7863,
+      "step": 4714
+    },
+    {
+      "epoch": 0.20608418200096157,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.991461428659375e-05,
+      "loss": 1.9259,
+      "step": 4715
+    },
+    {
+      "epoch": 0.20612789020499148,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.991047711405617e-05,
+      "loss": 1.8682,
+      "step": 4716
+    },
+    {
+      "epoch": 0.20617159840902138,
+      "grad_norm": 2.75,
+      "learning_rate": 8.990633918834979e-05,
+      "loss": 1.8183,
+      "step": 4717
+    },
+    {
+      "epoch": 0.20621530661305126,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.990220050955271e-05,
+      "loss": 2.2547,
+      "step": 4718
+    },
+    {
+      "epoch": 0.20625901481708117,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.989806107774304e-05,
+      "loss": 1.9872,
+      "step": 4719
+    },
+    {
+      "epoch": 0.20630272302111105,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.989392089299888e-05,
+      "loss": 1.8414,
+      "step": 4720
+    },
+    {
+      "epoch": 0.20634643122514096,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.988977995539837e-05,
+      "loss": 2.3367,
+      "step": 4721
+    },
+    {
+      "epoch": 0.20639013942917087,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.988563826501965e-05,
+      "loss": 2.3742,
+      "step": 4722
+    },
+    {
+      "epoch": 0.20643384763320075,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.98814958219409e-05,
+      "loss": 2.4695,
+      "step": 4723
+    },
+    {
+      "epoch": 0.20647755583723065,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.987735262624027e-05,
+      "loss": 2.1642,
+      "step": 4724
+    },
+    {
+      "epoch": 0.20652126404126053,
+      "grad_norm": 1.84375,
+      "learning_rate": 8.987320867799594e-05,
+      "loss": 1.6123,
+      "step": 4725
+    },
+    {
+      "epoch": 0.20656497224529044,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.986906397728616e-05,
+      "loss": 1.9332,
+      "step": 4726
+    },
+    {
+      "epoch": 0.20660868044932035,
+      "grad_norm": 4.75,
+      "learning_rate": 8.98649185241891e-05,
+      "loss": 1.7614,
+      "step": 4727
+    },
+    {
+      "epoch": 0.20665238865335023,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.9860772318783e-05,
+      "loss": 2.1411,
+      "step": 4728
+    },
+    {
+      "epoch": 0.20669609685738013,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.985662536114613e-05,
+      "loss": 1.8783,
+      "step": 4729
+    },
+    {
+      "epoch": 0.20673980506141001,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.985247765135672e-05,
+      "loss": 1.7308,
+      "step": 4730
+    },
+    {
+      "epoch": 0.20678351326543992,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.984832918949306e-05,
+      "loss": 2.3303,
+      "step": 4731
+    },
+    {
+      "epoch": 0.20682722146946983,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.984417997563342e-05,
+      "loss": 1.6741,
+      "step": 4732
+    },
+    {
+      "epoch": 0.2068709296734997,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.984003000985613e-05,
+      "loss": 1.9943,
+      "step": 4733
+    },
+    {
+      "epoch": 0.20691463787752962,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.983587929223948e-05,
+      "loss": 2.332,
+      "step": 4734
+    },
+    {
+      "epoch": 0.2069583460815595,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.98317278228618e-05,
+      "loss": 2.201,
+      "step": 4735
+    },
+    {
+      "epoch": 0.2070020542855894,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.982757560180146e-05,
+      "loss": 1.8238,
+      "step": 4736
+    },
+    {
+      "epoch": 0.2070457624896193,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.982342262913679e-05,
+      "loss": 1.7964,
+      "step": 4737
+    },
+    {
+      "epoch": 0.2070894706936492,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.981926890494619e-05,
+      "loss": 1.4018,
+      "step": 4738
+    },
+    {
+      "epoch": 0.2071331788976791,
+      "grad_norm": 2.75,
+      "learning_rate": 8.981511442930802e-05,
+      "loss": 1.6008,
+      "step": 4739
+    },
+    {
+      "epoch": 0.20717688710170898,
+      "grad_norm": 2.375,
+      "learning_rate": 8.98109592023007e-05,
+      "loss": 2.0941,
+      "step": 4740
+    },
+    {
+      "epoch": 0.20722059530573889,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.980680322400264e-05,
+      "loss": 1.9683,
+      "step": 4741
+    },
+    {
+      "epoch": 0.2072643035097688,
+      "grad_norm": 2.0,
+      "learning_rate": 8.980264649449225e-05,
+      "loss": 1.7002,
+      "step": 4742
+    },
+    {
+      "epoch": 0.20730801171379867,
+      "grad_norm": 6.375,
+      "learning_rate": 8.9798489013848e-05,
+      "loss": 1.7523,
+      "step": 4743
+    },
+    {
+      "epoch": 0.20735171991782858,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.979433078214834e-05,
+      "loss": 3.0292,
+      "step": 4744
+    },
+    {
+      "epoch": 0.20739542812185846,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.979017179947174e-05,
+      "loss": 2.0299,
+      "step": 4745
+    },
+    {
+      "epoch": 0.20743913632588837,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.978601206589669e-05,
+      "loss": 2.5303,
+      "step": 4746
+    },
+    {
+      "epoch": 0.20748284452991828,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.978185158150167e-05,
+      "loss": 1.5541,
+      "step": 4747
+    },
+    {
+      "epoch": 0.20752655273394816,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.977769034636523e-05,
+      "loss": 1.9323,
+      "step": 4748
+    },
+    {
+      "epoch": 0.20757026093797806,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.977352836056587e-05,
+      "loss": 1.5497,
+      "step": 4749
+    },
+    {
+      "epoch": 0.20761396914200794,
+      "grad_norm": 2.25,
+      "learning_rate": 8.976936562418215e-05,
+      "loss": 2.4203,
+      "step": 4750
+    },
+    {
+      "epoch": 0.20765767734603785,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.97652021372926e-05,
+      "loss": 1.8565,
+      "step": 4751
+    },
+    {
+      "epoch": 0.20770138555006776,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.976103789997582e-05,
+      "loss": 1.4862,
+      "step": 4752
+    },
+    {
+      "epoch": 0.20774509375409764,
+      "grad_norm": 5.6875,
+      "learning_rate": 8.975687291231041e-05,
+      "loss": 2.5187,
+      "step": 4753
+    },
+    {
+      "epoch": 0.20778880195812754,
+      "grad_norm": 4.03125,
+      "learning_rate": 8.975270717437492e-05,
+      "loss": 2.9507,
+      "step": 4754
+    },
+    {
+      "epoch": 0.20783251016215742,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.974854068624798e-05,
+      "loss": 1.4123,
+      "step": 4755
+    },
+    {
+      "epoch": 0.20787621836618733,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.974437344800825e-05,
+      "loss": 2.8158,
+      "step": 4756
+    },
+    {
+      "epoch": 0.20791992657021724,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.974020545973433e-05,
+      "loss": 1.6974,
+      "step": 4757
+    },
+    {
+      "epoch": 0.20796363477424712,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.97360367215049e-05,
+      "loss": 1.9125,
+      "step": 4758
+    },
+    {
+      "epoch": 0.20800734297827703,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.973186723339863e-05,
+      "loss": 2.0247,
+      "step": 4759
+    },
+    {
+      "epoch": 0.2080510511823069,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.972769699549419e-05,
+      "loss": 1.7753,
+      "step": 4760
+    },
+    {
+      "epoch": 0.20809475938633681,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.972352600787028e-05,
+      "loss": 2.3908,
+      "step": 4761
+    },
+    {
+      "epoch": 0.20813846759036672,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.971935427060562e-05,
+      "loss": 2.2623,
+      "step": 4762
+    },
+    {
+      "epoch": 0.2081821757943966,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.971518178377895e-05,
+      "loss": 1.909,
+      "step": 4763
+    },
+    {
+      "epoch": 0.2082258839984265,
+      "grad_norm": 3.125,
+      "learning_rate": 8.9711008547469e-05,
+      "loss": 1.8331,
+      "step": 4764
+    },
+    {
+      "epoch": 0.2082695922024564,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.970683456175451e-05,
+      "loss": 1.4373,
+      "step": 4765
+    },
+    {
+      "epoch": 0.2083133004064863,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.970265982671427e-05,
+      "loss": 2.1941,
+      "step": 4766
+    },
+    {
+      "epoch": 0.2083570086105162,
+      "grad_norm": 3.3125,
+      "learning_rate": 8.969848434242705e-05,
+      "loss": 2.371,
+      "step": 4767
+    },
+    {
+      "epoch": 0.20840071681454608,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.969430810897166e-05,
+      "loss": 2.0571,
+      "step": 4768
+    },
+    {
+      "epoch": 0.208444425018576,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.969013112642689e-05,
+      "loss": 1.9345,
+      "step": 4769
+    },
+    {
+      "epoch": 0.20848813322260587,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.968595339487157e-05,
+      "loss": 1.3695,
+      "step": 4770
+    },
+    {
+      "epoch": 0.20853184142663578,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.968177491438457e-05,
+      "loss": 1.992,
+      "step": 4771
+    },
+    {
+      "epoch": 0.20857554963066569,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.967759568504472e-05,
+      "loss": 2.2224,
+      "step": 4772
+    },
+    {
+      "epoch": 0.20861925783469557,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.967341570693088e-05,
+      "loss": 1.8914,
+      "step": 4773
+    },
+    {
+      "epoch": 0.20866296603872547,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.966923498012196e-05,
+      "loss": 1.6996,
+      "step": 4774
+    },
+    {
+      "epoch": 0.20870667424275535,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.966505350469682e-05,
+      "loss": 2.8157,
+      "step": 4775
+    },
+    {
+      "epoch": 0.20875038244678526,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.966087128073441e-05,
+      "loss": 1.5552,
+      "step": 4776
+    },
+    {
+      "epoch": 0.20879409065081517,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.965668830831364e-05,
+      "loss": 1.7169,
+      "step": 4777
+    },
+    {
+      "epoch": 0.20883779885484505,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.965250458751343e-05,
+      "loss": 2.0599,
+      "step": 4778
+    },
+    {
+      "epoch": 0.20888150705887495,
+      "grad_norm": 3.125,
+      "learning_rate": 8.964832011841275e-05,
+      "loss": 2.4304,
+      "step": 4779
+    },
+    {
+      "epoch": 0.20892521526290483,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.964413490109055e-05,
+      "loss": 1.6481,
+      "step": 4780
+    },
+    {
+      "epoch": 0.20896892346693474,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.963994893562585e-05,
+      "loss": 2.0656,
+      "step": 4781
+    },
+    {
+      "epoch": 0.20901263167096465,
+      "grad_norm": 2.25,
+      "learning_rate": 8.96357622220976e-05,
+      "loss": 1.93,
+      "step": 4782
+    },
+    {
+      "epoch": 0.20905633987499453,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.963157476058485e-05,
+      "loss": 2.4348,
+      "step": 4783
+    },
+    {
+      "epoch": 0.20910004807902444,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.962738655116658e-05,
+      "loss": 2.1908,
+      "step": 4784
+    },
+    {
+      "epoch": 0.20914375628305432,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.962319759392188e-05,
+      "loss": 1.6403,
+      "step": 4785
+    },
+    {
+      "epoch": 0.20918746448708422,
+      "grad_norm": 20.5,
+      "learning_rate": 8.961900788892974e-05,
+      "loss": 1.6297,
+      "step": 4786
+    },
+    {
+      "epoch": 0.20923117269111413,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.961481743626928e-05,
+      "loss": 2.1056,
+      "step": 4787
+    },
+    {
+      "epoch": 0.209274880895144,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.961062623601955e-05,
+      "loss": 1.9578,
+      "step": 4788
+    },
+    {
+      "epoch": 0.20931858909917392,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.960643428825966e-05,
+      "loss": 1.7978,
+      "step": 4789
+    },
+    {
+      "epoch": 0.2093622973032038,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.96022415930687e-05,
+      "loss": 1.7017,
+      "step": 4790
+    },
+    {
+      "epoch": 0.2094060055072337,
+      "grad_norm": 2.0,
+      "learning_rate": 8.959804815052582e-05,
+      "loss": 1.9484,
+      "step": 4791
+    },
+    {
+      "epoch": 0.2094497137112636,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.959385396071012e-05,
+      "loss": 1.9809,
+      "step": 4792
+    },
+    {
+      "epoch": 0.2094934219152935,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.958965902370078e-05,
+      "loss": 1.4298,
+      "step": 4793
+    },
+    {
+      "epoch": 0.2095371301193234,
+      "grad_norm": 2.375,
+      "learning_rate": 8.958546333957694e-05,
+      "loss": 1.766,
+      "step": 4794
+    },
+    {
+      "epoch": 0.20958083832335328,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.958126690841781e-05,
+      "loss": 2.1259,
+      "step": 4795
+    },
+    {
+      "epoch": 0.2096245465273832,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.957706973030256e-05,
+      "loss": 1.6469,
+      "step": 4796
+    },
+    {
+      "epoch": 0.2096682547314131,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.95728718053104e-05,
+      "loss": 1.6772,
+      "step": 4797
+    },
+    {
+      "epoch": 0.20971196293544297,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.956867313352056e-05,
+      "loss": 2.1811,
+      "step": 4798
+    },
+    {
+      "epoch": 0.20975567113947288,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.956447371501227e-05,
+      "loss": 2.0043,
+      "step": 4799
+    },
+    {
+      "epoch": 0.20979937934350276,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.956027354986476e-05,
+      "loss": 2.1774,
+      "step": 4800
+    },
+    {
+      "epoch": 0.20984308754753267,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.955607263815732e-05,
+      "loss": 1.8232,
+      "step": 4801
+    },
+    {
+      "epoch": 0.20988679575156258,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.955187097996922e-05,
+      "loss": 2.3588,
+      "step": 4802
+    },
+    {
+      "epoch": 0.20993050395559246,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.954766857537975e-05,
+      "loss": 1.8239,
+      "step": 4803
+    },
+    {
+      "epoch": 0.20997421215962236,
+      "grad_norm": 12.1875,
+      "learning_rate": 8.95434654244682e-05,
+      "loss": 1.6482,
+      "step": 4804
+    },
+    {
+      "epoch": 0.21001792036365224,
+      "grad_norm": 2.25,
+      "learning_rate": 8.953926152731394e-05,
+      "loss": 2.0697,
+      "step": 4805
+    },
+    {
+      "epoch": 0.21006162856768215,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.953505688399624e-05,
+      "loss": 2.2319,
+      "step": 4806
+    },
+    {
+      "epoch": 0.21010533677171206,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.95308514945945e-05,
+      "loss": 1.8125,
+      "step": 4807
+    },
+    {
+      "epoch": 0.21014904497574194,
+      "grad_norm": 5.875,
+      "learning_rate": 8.952664535918803e-05,
+      "loss": 1.6745,
+      "step": 4808
+    },
+    {
+      "epoch": 0.21019275317977185,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.952243847785624e-05,
+      "loss": 2.3218,
+      "step": 4809
+    },
+    {
+      "epoch": 0.21023646138380173,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.951823085067852e-05,
+      "loss": 2.0469,
+      "step": 4810
+    },
+    {
+      "epoch": 0.21028016958783163,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.951402247773428e-05,
+      "loss": 2.3366,
+      "step": 4811
+    },
+    {
+      "epoch": 0.21032387779186154,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.950981335910291e-05,
+      "loss": 1.7435,
+      "step": 4812
+    },
+    {
+      "epoch": 0.21036758599589142,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.950560349486386e-05,
+      "loss": 1.5361,
+      "step": 4813
+    },
+    {
+      "epoch": 0.21041129419992133,
+      "grad_norm": 2.875,
+      "learning_rate": 8.950139288509658e-05,
+      "loss": 2.4291,
+      "step": 4814
+    },
+    {
+      "epoch": 0.2104550024039512,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.949718152988051e-05,
+      "loss": 2.2256,
+      "step": 4815
+    },
+    {
+      "epoch": 0.21049871060798112,
+      "grad_norm": 2.25,
+      "learning_rate": 8.949296942929514e-05,
+      "loss": 1.9177,
+      "step": 4816
+    },
+    {
+      "epoch": 0.21054241881201102,
+      "grad_norm": 3.703125,
+      "learning_rate": 8.948875658341997e-05,
+      "loss": 2.6135,
+      "step": 4817
+    },
+    {
+      "epoch": 0.2105861270160409,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.94845429923345e-05,
+      "loss": 2.557,
+      "step": 4818
+    },
+    {
+      "epoch": 0.2106298352200708,
+      "grad_norm": 2.625,
+      "learning_rate": 8.948032865611822e-05,
+      "loss": 2.2347,
+      "step": 4819
+    },
+    {
+      "epoch": 0.2106735434241007,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.947611357485068e-05,
+      "loss": 2.5274,
+      "step": 4820
+    },
+    {
+      "epoch": 0.2107172516281306,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.947189774861142e-05,
+      "loss": 1.4707,
+      "step": 4821
+    },
+    {
+      "epoch": 0.2107609598321605,
+      "grad_norm": 5.3125,
+      "learning_rate": 8.946768117748001e-05,
+      "loss": 1.3507,
+      "step": 4822
+    },
+    {
+      "epoch": 0.21080466803619038,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.946346386153601e-05,
+      "loss": 2.4313,
+      "step": 4823
+    },
+    {
+      "epoch": 0.2108483762402203,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.945924580085901e-05,
+      "loss": 1.7597,
+      "step": 4824
+    },
+    {
+      "epoch": 0.2108920844442502,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.945502699552862e-05,
+      "loss": 1.9471,
+      "step": 4825
+    },
+    {
+      "epoch": 0.21093579264828008,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.945080744562442e-05,
+      "loss": 2.3347,
+      "step": 4826
+    },
+    {
+      "epoch": 0.21097950085231,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.944658715122609e-05,
+      "loss": 2.9457,
+      "step": 4827
+    },
+    {
+      "epoch": 0.21102320905633987,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.944236611241323e-05,
+      "loss": 1.7127,
+      "step": 4828
+    },
+    {
+      "epoch": 0.21106691726036977,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.943814432926553e-05,
+      "loss": 2.4287,
+      "step": 4829
+    },
+    {
+      "epoch": 0.21111062546439968,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.943392180186265e-05,
+      "loss": 2.2223,
+      "step": 4830
+    },
+    {
+      "epoch": 0.21115433366842956,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.942969853028426e-05,
+      "loss": 1.7299,
+      "step": 4831
+    },
+    {
+      "epoch": 0.21119804187245947,
+      "grad_norm": 3.71875,
+      "learning_rate": 8.942547451461008e-05,
+      "loss": 1.8223,
+      "step": 4832
+    },
+    {
+      "epoch": 0.21124175007648935,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.942124975491981e-05,
+      "loss": 2.1694,
+      "step": 4833
+    },
+    {
+      "epoch": 0.21128545828051926,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.94170242512932e-05,
+      "loss": 1.9498,
+      "step": 4834
+    },
+    {
+      "epoch": 0.21132916648454916,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.941279800380995e-05,
+      "loss": 1.9834,
+      "step": 4835
+    },
+    {
+      "epoch": 0.21137287468857904,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.940857101254985e-05,
+      "loss": 2.5328,
+      "step": 4836
+    },
+    {
+      "epoch": 0.21141658289260895,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.940434327759264e-05,
+      "loss": 2.3951,
+      "step": 4837
+    },
+    {
+      "epoch": 0.21146029109663883,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.940011479901816e-05,
+      "loss": 2.0517,
+      "step": 4838
+    },
+    {
+      "epoch": 0.21150399930066874,
+      "grad_norm": 2.25,
+      "learning_rate": 8.939588557690614e-05,
+      "loss": 2.1932,
+      "step": 4839
+    },
+    {
+      "epoch": 0.21154770750469865,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.939165561133642e-05,
+      "loss": 2.1056,
+      "step": 4840
+    },
+    {
+      "epoch": 0.21159141570872853,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.938742490238884e-05,
+      "loss": 2.1849,
+      "step": 4841
+    },
+    {
+      "epoch": 0.21163512391275843,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.938319345014321e-05,
+      "loss": 2.4694,
+      "step": 4842
+    },
+    {
+      "epoch": 0.2116788321167883,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.937896125467942e-05,
+      "loss": 1.8354,
+      "step": 4843
+    },
+    {
+      "epoch": 0.21172254032081822,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.937472831607732e-05,
+      "loss": 2.216,
+      "step": 4844
+    },
+    {
+      "epoch": 0.21176624852484813,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.937049463441678e-05,
+      "loss": 1.7439,
+      "step": 4845
+    },
+    {
+      "epoch": 0.211809956728878,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.93662602097777e-05,
+      "loss": 2.1265,
+      "step": 4846
+    },
+    {
+      "epoch": 0.21185366493290791,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.936202504224e-05,
+      "loss": 1.944,
+      "step": 4847
+    },
+    {
+      "epoch": 0.2118973731369378,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.93577891318836e-05,
+      "loss": 2.1815,
+      "step": 4848
+    },
+    {
+      "epoch": 0.2119410813409677,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.935355247878842e-05,
+      "loss": 2.2721,
+      "step": 4849
+    },
+    {
+      "epoch": 0.2119847895449976,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.934931508303445e-05,
+      "loss": 2.313,
+      "step": 4850
+    },
+    {
+      "epoch": 0.2120284977490275,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.934507694470162e-05,
+      "loss": 2.0388,
+      "step": 4851
+    },
+    {
+      "epoch": 0.2120722059530574,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.934083806386995e-05,
+      "loss": 1.6507,
+      "step": 4852
+    },
+    {
+      "epoch": 0.21211591415708728,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.933659844061938e-05,
+      "loss": 1.8782,
+      "step": 4853
+    },
+    {
+      "epoch": 0.21215962236111718,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.933235807502996e-05,
+      "loss": 1.8664,
+      "step": 4854
+    },
+    {
+      "epoch": 0.2122033305651471,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.932811696718169e-05,
+      "loss": 1.7721,
+      "step": 4855
+    },
+    {
+      "epoch": 0.21224703876917697,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.93238751171546e-05,
+      "loss": 1.5299,
+      "step": 4856
+    },
+    {
+      "epoch": 0.21229074697320688,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.931963252502878e-05,
+      "loss": 1.7576,
+      "step": 4857
+    },
+    {
+      "epoch": 0.21233445517723676,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.931538919088425e-05,
+      "loss": 2.125,
+      "step": 4858
+    },
+    {
+      "epoch": 0.21237816338126667,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.931114511480112e-05,
+      "loss": 1.7627,
+      "step": 4859
+    },
+    {
+      "epoch": 0.21242187158529657,
+      "grad_norm": 2.25,
+      "learning_rate": 8.930690029685946e-05,
+      "loss": 1.9135,
+      "step": 4860
+    },
+    {
+      "epoch": 0.21246557978932645,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.930265473713938e-05,
+      "loss": 2.209,
+      "step": 4861
+    },
+    {
+      "epoch": 0.21250928799335636,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.9298408435721e-05,
+      "loss": 1.8138,
+      "step": 4862
+    },
+    {
+      "epoch": 0.21255299619738624,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.929416139268446e-05,
+      "loss": 2.499,
+      "step": 4863
+    },
+    {
+      "epoch": 0.21259670440141615,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.92899136081099e-05,
+      "loss": 2.1922,
+      "step": 4864
+    },
+    {
+      "epoch": 0.21264041260544606,
+      "grad_norm": 2.75,
+      "learning_rate": 8.92856650820775e-05,
+      "loss": 1.584,
+      "step": 4865
+    },
+    {
+      "epoch": 0.21268412080947594,
+      "grad_norm": 2.375,
+      "learning_rate": 8.928141581466742e-05,
+      "loss": 1.9576,
+      "step": 4866
+    },
+    {
+      "epoch": 0.21272782901350584,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.927716580595984e-05,
+      "loss": 1.8619,
+      "step": 4867
+    },
+    {
+      "epoch": 0.21277153721753572,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.9272915056035e-05,
+      "loss": 2.3122,
+      "step": 4868
+    },
+    {
+      "epoch": 0.21281524542156563,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.926866356497307e-05,
+      "loss": 1.3723,
+      "step": 4869
+    },
+    {
+      "epoch": 0.21285895362559554,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.92644113328543e-05,
+      "loss": 1.8232,
+      "step": 4870
+    },
+    {
+      "epoch": 0.21290266182962542,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.926015835975896e-05,
+      "loss": 2.2606,
+      "step": 4871
+    },
+    {
+      "epoch": 0.21294637003365532,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.925590464576727e-05,
+      "loss": 1.2309,
+      "step": 4872
+    },
+    {
+      "epoch": 0.2129900782376852,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.925165019095956e-05,
+      "loss": 1.7613,
+      "step": 4873
+    },
+    {
+      "epoch": 0.2130337864417151,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.924739499541606e-05,
+      "loss": 1.8071,
+      "step": 4874
+    },
+    {
+      "epoch": 0.21307749464574502,
+      "grad_norm": 5.375,
+      "learning_rate": 8.924313905921709e-05,
+      "loss": 2.6352,
+      "step": 4875
+    },
+    {
+      "epoch": 0.2131212028497749,
+      "grad_norm": 3.25,
+      "learning_rate": 8.923888238244298e-05,
+      "loss": 2.1339,
+      "step": 4876
+    },
+    {
+      "epoch": 0.2131649110538048,
+      "grad_norm": 3.484375,
+      "learning_rate": 8.923462496517405e-05,
+      "loss": 2.0866,
+      "step": 4877
+    },
+    {
+      "epoch": 0.2132086192578347,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.923036680749064e-05,
+      "loss": 1.9792,
+      "step": 4878
+    },
+    {
+      "epoch": 0.2132523274618646,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.922610790947311e-05,
+      "loss": 2.6828,
+      "step": 4879
+    },
+    {
+      "epoch": 0.2132960356658945,
+      "grad_norm": 2.875,
+      "learning_rate": 8.922184827120183e-05,
+      "loss": 1.9287,
+      "step": 4880
+    },
+    {
+      "epoch": 0.21333974386992438,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.92175878927572e-05,
+      "loss": 2.0443,
+      "step": 4881
+    },
+    {
+      "epoch": 0.2133834520739543,
+      "grad_norm": 3.375,
+      "learning_rate": 8.921332677421961e-05,
+      "loss": 3.1954,
+      "step": 4882
+    },
+    {
+      "epoch": 0.21342716027798417,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.920906491566944e-05,
+      "loss": 2.149,
+      "step": 4883
+    },
+    {
+      "epoch": 0.21347086848201408,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.92048023171872e-05,
+      "loss": 2.1222,
+      "step": 4884
+    },
+    {
+      "epoch": 0.21351457668604398,
+      "grad_norm": 4.125,
+      "learning_rate": 8.920053897885325e-05,
+      "loss": 1.8552,
+      "step": 4885
+    },
+    {
+      "epoch": 0.21355828489007386,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.919627490074807e-05,
+      "loss": 1.9297,
+      "step": 4886
+    },
+    {
+      "epoch": 0.21360199309410377,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.919201008295214e-05,
+      "loss": 2.3196,
+      "step": 4887
+    },
+    {
+      "epoch": 0.21364570129813365,
+      "grad_norm": 3.0,
+      "learning_rate": 8.918774452554595e-05,
+      "loss": 2.5424,
+      "step": 4888
+    },
+    {
+      "epoch": 0.21368940950216356,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.918347822860997e-05,
+      "loss": 3.2891,
+      "step": 4889
+    },
+    {
+      "epoch": 0.21373311770619347,
+      "grad_norm": 2.375,
+      "learning_rate": 8.917921119222474e-05,
+      "loss": 1.8642,
+      "step": 4890
+    },
+    {
+      "epoch": 0.21377682591022334,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.917494341647077e-05,
+      "loss": 2.4109,
+      "step": 4891
+    },
+    {
+      "epoch": 0.21382053411425325,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.917067490142858e-05,
+      "loss": 1.8142,
+      "step": 4892
+    },
+    {
+      "epoch": 0.21386424231828313,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.916640564717878e-05,
+      "loss": 1.7431,
+      "step": 4893
+    },
+    {
+      "epoch": 0.21390795052231304,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.916213565380188e-05,
+      "loss": 1.7604,
+      "step": 4894
+    },
+    {
+      "epoch": 0.21395165872634295,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.915786492137848e-05,
+      "loss": 2.2038,
+      "step": 4895
+    },
+    {
+      "epoch": 0.21399536693037283,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.915359344998919e-05,
+      "loss": 1.6342,
+      "step": 4896
+    },
+    {
+      "epoch": 0.21403907513440273,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.91493212397146e-05,
+      "loss": 1.4095,
+      "step": 4897
+    },
+    {
+      "epoch": 0.21408278333843261,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.914504829063535e-05,
+      "loss": 1.7478,
+      "step": 4898
+    },
+    {
+      "epoch": 0.21412649154246252,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.914077460283205e-05,
+      "loss": 1.538,
+      "step": 4899
+    },
+    {
+      "epoch": 0.21417019974649243,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.913650017638537e-05,
+      "loss": 2.1669,
+      "step": 4900
+    },
+    {
+      "epoch": 0.2142139079505223,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.913222501137597e-05,
+      "loss": 1.8726,
+      "step": 4901
+    },
+    {
+      "epoch": 0.21425761615455222,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.912794910788453e-05,
+      "loss": 2.4547,
+      "step": 4902
+    },
+    {
+      "epoch": 0.2143013243585821,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.912367246599175e-05,
+      "loss": 1.9188,
+      "step": 4903
+    },
+    {
+      "epoch": 0.214345032562612,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.911939508577833e-05,
+      "loss": 2.116,
+      "step": 4904
+    },
+    {
+      "epoch": 0.2143887407666419,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.911511696732498e-05,
+      "loss": 2.0171,
+      "step": 4905
+    },
+    {
+      "epoch": 0.2144324489706718,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.911083811071244e-05,
+      "loss": 2.2515,
+      "step": 4906
+    },
+    {
+      "epoch": 0.2144761571747017,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.910655851602146e-05,
+      "loss": 2.4468,
+      "step": 4907
+    },
+    {
+      "epoch": 0.21451986537873158,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.910227818333282e-05,
+      "loss": 2.6137,
+      "step": 4908
+    },
+    {
+      "epoch": 0.21456357358276149,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.909799711272727e-05,
+      "loss": 2.7769,
+      "step": 4909
+    },
+    {
+      "epoch": 0.2146072817867914,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.909371530428561e-05,
+      "loss": 1.7773,
+      "step": 4910
+    },
+    {
+      "epoch": 0.21465098999082127,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.908943275808866e-05,
+      "loss": 1.6918,
+      "step": 4911
+    },
+    {
+      "epoch": 0.21469469819485118,
+      "grad_norm": 2.25,
+      "learning_rate": 8.90851494742172e-05,
+      "loss": 2.4438,
+      "step": 4912
+    },
+    {
+      "epoch": 0.21473840639888106,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.908086545275209e-05,
+      "loss": 2.461,
+      "step": 4913
+    },
+    {
+      "epoch": 0.21478211460291097,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.907658069377418e-05,
+      "loss": 1.8086,
+      "step": 4914
+    },
+    {
+      "epoch": 0.21482582280694087,
+      "grad_norm": 3.59375,
+      "learning_rate": 8.90722951973643e-05,
+      "loss": 1.9816,
+      "step": 4915
+    },
+    {
+      "epoch": 0.21486953101097075,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.906800896360336e-05,
+      "loss": 2.106,
+      "step": 4916
+    },
+    {
+      "epoch": 0.21491323921500066,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.906372199257223e-05,
+      "loss": 1.485,
+      "step": 4917
+    },
+    {
+      "epoch": 0.21495694741903054,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.90594342843518e-05,
+      "loss": 3.2665,
+      "step": 4918
+    },
+    {
+      "epoch": 0.21500065562306045,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.9055145839023e-05,
+      "loss": 1.7803,
+      "step": 4919
+    },
+    {
+      "epoch": 0.21504436382709036,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.905085665666674e-05,
+      "loss": 2.6265,
+      "step": 4920
+    },
+    {
+      "epoch": 0.21508807203112024,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.9046566737364e-05,
+      "loss": 2.2594,
+      "step": 4921
+    },
+    {
+      "epoch": 0.21513178023515014,
+      "grad_norm": 2.625,
+      "learning_rate": 8.904227608119571e-05,
+      "loss": 1.9987,
+      "step": 4922
+    },
+    {
+      "epoch": 0.21517548843918002,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.903798468824286e-05,
+      "loss": 2.3512,
+      "step": 4923
+    },
+    {
+      "epoch": 0.21521919664320993,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.90336925585864e-05,
+      "loss": 2.123,
+      "step": 4924
+    },
+    {
+      "epoch": 0.21526290484723984,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.902939969230737e-05,
+      "loss": 1.8981,
+      "step": 4925
+    },
+    {
+      "epoch": 0.21530661305126972,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.902510608948676e-05,
+      "loss": 2.0691,
+      "step": 4926
+    },
+    {
+      "epoch": 0.21535032125529963,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.902081175020559e-05,
+      "loss": 2.0388,
+      "step": 4927
+    },
+    {
+      "epoch": 0.2153940294593295,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.901651667454492e-05,
+      "loss": 1.8893,
+      "step": 4928
+    },
+    {
+      "epoch": 0.2154377376633594,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.901222086258578e-05,
+      "loss": 1.5117,
+      "step": 4929
+    },
+    {
+      "epoch": 0.21548144586738932,
+      "grad_norm": 2.375,
+      "learning_rate": 8.900792431440927e-05,
+      "loss": 2.1315,
+      "step": 4930
+    },
+    {
+      "epoch": 0.2155251540714192,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.900362703009644e-05,
+      "loss": 1.8519,
+      "step": 4931
+    },
+    {
+      "epoch": 0.2155688622754491,
+      "grad_norm": 2.125,
+      "learning_rate": 8.89993290097284e-05,
+      "loss": 2.1304,
+      "step": 4932
+    },
+    {
+      "epoch": 0.215612570479479,
+      "grad_norm": 3.265625,
+      "learning_rate": 8.899503025338627e-05,
+      "loss": 1.9898,
+      "step": 4933
+    },
+    {
+      "epoch": 0.2156562786835089,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.899073076115116e-05,
+      "loss": 1.602,
+      "step": 4934
+    },
+    {
+      "epoch": 0.2156999868875388,
+      "grad_norm": 2.125,
+      "learning_rate": 8.898643053310422e-05,
+      "loss": 1.7423,
+      "step": 4935
+    },
+    {
+      "epoch": 0.21574369509156868,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.898212956932659e-05,
+      "loss": 1.9523,
+      "step": 4936
+    },
+    {
+      "epoch": 0.2157874032955986,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.897782786989944e-05,
+      "loss": 1.8911,
+      "step": 4937
+    },
+    {
+      "epoch": 0.21583111149962847,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.897352543490395e-05,
+      "loss": 1.5745,
+      "step": 4938
+    },
+    {
+      "epoch": 0.21587481970365838,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.89692222644213e-05,
+      "loss": 2.2996,
+      "step": 4939
+    },
+    {
+      "epoch": 0.21591852790768828,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.896491835853272e-05,
+      "loss": 2.591,
+      "step": 4940
+    },
+    {
+      "epoch": 0.21596223611171816,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.896061371731943e-05,
+      "loss": 2.5338,
+      "step": 4941
+    },
+    {
+      "epoch": 0.21600594431574807,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.895630834086264e-05,
+      "loss": 1.6313,
+      "step": 4942
+    },
+    {
+      "epoch": 0.21604965251977795,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.89520022292436e-05,
+      "loss": 1.9688,
+      "step": 4943
+    },
+    {
+      "epoch": 0.21609336072380786,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.894769538254362e-05,
+      "loss": 2.121,
+      "step": 4944
+    },
+    {
+      "epoch": 0.21613706892783777,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.894338780084392e-05,
+      "loss": 1.9814,
+      "step": 4945
+    },
+    {
+      "epoch": 0.21618077713186765,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.893907948422581e-05,
+      "loss": 1.8047,
+      "step": 4946
+    },
+    {
+      "epoch": 0.21622448533589755,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.893477043277061e-05,
+      "loss": 2.0301,
+      "step": 4947
+    },
+    {
+      "epoch": 0.21626819353992743,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.893046064655961e-05,
+      "loss": 1.6069,
+      "step": 4948
+    },
+    {
+      "epoch": 0.21631190174395734,
+      "grad_norm": 2.75,
+      "learning_rate": 8.892615012567416e-05,
+      "loss": 1.8555,
+      "step": 4949
+    },
+    {
+      "epoch": 0.21635560994798725,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.892183887019562e-05,
+      "loss": 2.1679,
+      "step": 4950
+    },
+    {
+      "epoch": 0.21639931815201713,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.891752688020532e-05,
+      "loss": 2.19,
+      "step": 4951
+    },
+    {
+      "epoch": 0.21644302635604704,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.891321415578464e-05,
+      "loss": 2.3977,
+      "step": 4952
+    },
+    {
+      "epoch": 0.21648673456007692,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.890890069701499e-05,
+      "loss": 1.8516,
+      "step": 4953
+    },
+    {
+      "epoch": 0.21653044276410682,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.890458650397774e-05,
+      "loss": 1.684,
+      "step": 4954
+    },
+    {
+      "epoch": 0.21657415096813673,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.890027157675432e-05,
+      "loss": 2.0473,
+      "step": 4955
+    },
+    {
+      "epoch": 0.2166178591721666,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.889595591542617e-05,
+      "loss": 1.8999,
+      "step": 4956
+    },
+    {
+      "epoch": 0.21666156737619652,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.889163952007471e-05,
+      "loss": 1.7533,
+      "step": 4957
+    },
+    {
+      "epoch": 0.2167052755802264,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.888732239078141e-05,
+      "loss": 1.7618,
+      "step": 4958
+    },
+    {
+      "epoch": 0.2167489837842563,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.888300452762774e-05,
+      "loss": 2.773,
+      "step": 4959
+    },
+    {
+      "epoch": 0.2167926919882862,
+      "grad_norm": 2.0,
+      "learning_rate": 8.887868593069519e-05,
+      "loss": 1.9146,
+      "step": 4960
+    },
+    {
+      "epoch": 0.2168364001923161,
+      "grad_norm": 3.125,
+      "learning_rate": 8.887436660006525e-05,
+      "loss": 1.0823,
+      "step": 4961
+    },
+    {
+      "epoch": 0.216880108396346,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.887004653581942e-05,
+      "loss": 1.4368,
+      "step": 4962
+    },
+    {
+      "epoch": 0.21692381660037588,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.886572573803926e-05,
+      "loss": 2.3,
+      "step": 4963
+    },
+    {
+      "epoch": 0.2169675248044058,
+      "grad_norm": 2.125,
+      "learning_rate": 8.886140420680627e-05,
+      "loss": 1.8728,
+      "step": 4964
+    },
+    {
+      "epoch": 0.2170112330084357,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.885708194220204e-05,
+      "loss": 2.7391,
+      "step": 4965
+    },
+    {
+      "epoch": 0.21705494121246557,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.88527589443081e-05,
+      "loss": 2.1869,
+      "step": 4966
+    },
+    {
+      "epoch": 0.21709864941649548,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.884843521320606e-05,
+      "loss": 1.664,
+      "step": 4967
+    },
+    {
+      "epoch": 0.21714235762052536,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.884411074897751e-05,
+      "loss": 1.9771,
+      "step": 4968
+    },
+    {
+      "epoch": 0.21718606582455527,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.883978555170404e-05,
+      "loss": 1.2271,
+      "step": 4969
+    },
+    {
+      "epoch": 0.21722977402858518,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.883545962146731e-05,
+      "loss": 2.1806,
+      "step": 4970
+    },
+    {
+      "epoch": 0.21727348223261506,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.883113295834892e-05,
+      "loss": 2.4585,
+      "step": 4971
+    },
+    {
+      "epoch": 0.21731719043664496,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.882680556243054e-05,
+      "loss": 1.8355,
+      "step": 4972
+    },
+    {
+      "epoch": 0.21736089864067484,
+      "grad_norm": 17.5,
+      "learning_rate": 8.882247743379383e-05,
+      "loss": 4.6282,
+      "step": 4973
+    },
+    {
+      "epoch": 0.21740460684470475,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.881814857252046e-05,
+      "loss": 1.9054,
+      "step": 4974
+    },
+    {
+      "epoch": 0.21744831504873466,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.881381897869215e-05,
+      "loss": 2.76,
+      "step": 4975
+    },
+    {
+      "epoch": 0.21749202325276454,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.880948865239057e-05,
+      "loss": 1.9151,
+      "step": 4976
+    },
+    {
+      "epoch": 0.21753573145679445,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.880515759369745e-05,
+      "loss": 1.7111,
+      "step": 4977
+    },
+    {
+      "epoch": 0.21757943966082433,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.880082580269453e-05,
+      "loss": 2.0375,
+      "step": 4978
+    },
+    {
+      "epoch": 0.21762314786485423,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.879649327946356e-05,
+      "loss": 2.5731,
+      "step": 4979
+    },
+    {
+      "epoch": 0.21766685606888414,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.879216002408631e-05,
+      "loss": 1.8861,
+      "step": 4980
+    },
+    {
+      "epoch": 0.21771056427291402,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.878782603664452e-05,
+      "loss": 1.8373,
+      "step": 4981
+    },
+    {
+      "epoch": 0.21775427247694393,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.878349131722e-05,
+      "loss": 1.8506,
+      "step": 4982
+    },
+    {
+      "epoch": 0.2177979806809738,
+      "grad_norm": 2.125,
+      "learning_rate": 8.877915586589456e-05,
+      "loss": 1.7731,
+      "step": 4983
+    },
+    {
+      "epoch": 0.21784168888500371,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.877481968275001e-05,
+      "loss": 2.25,
+      "step": 4984
+    },
+    {
+      "epoch": 0.21788539708903362,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.877048276786817e-05,
+      "loss": 1.6804,
+      "step": 4985
+    },
+    {
+      "epoch": 0.2179291052930635,
+      "grad_norm": 2.375,
+      "learning_rate": 8.87661451213309e-05,
+      "loss": 2.3068,
+      "step": 4986
+    },
+    {
+      "epoch": 0.2179728134970934,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.876180674322005e-05,
+      "loss": 2.9461,
+      "step": 4987
+    },
+    {
+      "epoch": 0.2180165217011233,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.87574676336175e-05,
+      "loss": 1.837,
+      "step": 4988
+    },
+    {
+      "epoch": 0.2180602299051532,
+      "grad_norm": 11.1875,
+      "learning_rate": 8.87531277926051e-05,
+      "loss": 1.5664,
+      "step": 4989
+    },
+    {
+      "epoch": 0.2181039381091831,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.874878722026479e-05,
+      "loss": 1.7533,
+      "step": 4990
+    },
+    {
+      "epoch": 0.21814764631321298,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.874444591667848e-05,
+      "loss": 1.566,
+      "step": 4991
+    },
+    {
+      "epoch": 0.2181913545172429,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.874010388192808e-05,
+      "loss": 1.5187,
+      "step": 4992
+    },
+    {
+      "epoch": 0.21823506272127277,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.873576111609553e-05,
+      "loss": 2.2533,
+      "step": 4993
+    },
+    {
+      "epoch": 0.21827877092530268,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.87314176192628e-05,
+      "loss": 2.0866,
+      "step": 4994
+    },
+    {
+      "epoch": 0.2183224791293326,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.872707339151183e-05,
+      "loss": 2.0056,
+      "step": 4995
+    },
+    {
+      "epoch": 0.21836618733336247,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.872272843292464e-05,
+      "loss": 1.6985,
+      "step": 4996
+    },
+    {
+      "epoch": 0.21840989553739237,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.871838274358319e-05,
+      "loss": 1.3936,
+      "step": 4997
+    },
+    {
+      "epoch": 0.21845360374142225,
+      "grad_norm": 2.5,
+      "learning_rate": 8.871403632356951e-05,
+      "loss": 1.8284,
+      "step": 4998
+    },
+    {
+      "epoch": 0.21849731194545216,
+      "grad_norm": 3.125,
+      "learning_rate": 8.870968917296562e-05,
+      "loss": 2.3903,
+      "step": 4999
+    },
+    {
+      "epoch": 0.21854102014948207,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.870534129185357e-05,
+      "loss": 1.67,
+      "step": 5000
+    },
+    {
+      "epoch": 0.21858472835351195,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.87009926803154e-05,
+      "loss": 1.8593,
+      "step": 5001
+    },
+    {
+      "epoch": 0.21862843655754186,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.869664333843315e-05,
+      "loss": 2.3788,
+      "step": 5002
+    },
+    {
+      "epoch": 0.21867214476157174,
+      "grad_norm": 3.4375,
+      "learning_rate": 8.869229326628892e-05,
+      "loss": 1.2086,
+      "step": 5003
+    },
+    {
+      "epoch": 0.21871585296560164,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.868794246396481e-05,
+      "loss": 2.5474,
+      "step": 5004
+    },
+    {
+      "epoch": 0.21875956116963155,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.868359093154292e-05,
+      "loss": 2.0143,
+      "step": 5005
+    },
+    {
+      "epoch": 0.21880326937366143,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.867923866910536e-05,
+      "loss": 2.6357,
+      "step": 5006
+    },
+    {
+      "epoch": 0.21884697757769134,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.867488567673429e-05,
+      "loss": 1.5781,
+      "step": 5007
+    },
+    {
+      "epoch": 0.21889068578172122,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.867053195451183e-05,
+      "loss": 1.9196,
+      "step": 5008
+    },
+    {
+      "epoch": 0.21893439398575112,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.866617750252014e-05,
+      "loss": 1.9003,
+      "step": 5009
+    },
+    {
+      "epoch": 0.21897810218978103,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.866182232084141e-05,
+      "loss": 2.0773,
+      "step": 5010
+    },
+    {
+      "epoch": 0.2190218103938109,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.865746640955783e-05,
+      "loss": 2.4597,
+      "step": 5011
+    },
+    {
+      "epoch": 0.21906551859784082,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.865310976875159e-05,
+      "loss": 1.7359,
+      "step": 5012
+    },
+    {
+      "epoch": 0.2191092268018707,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.864875239850489e-05,
+      "loss": 1.7675,
+      "step": 5013
+    },
+    {
+      "epoch": 0.2191529350059006,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.864439429890001e-05,
+      "loss": 1.7439,
+      "step": 5014
+    },
+    {
+      "epoch": 0.21919664320993051,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.864003547001915e-05,
+      "loss": 1.7313,
+      "step": 5015
+    },
+    {
+      "epoch": 0.2192403514139604,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.86356759119446e-05,
+      "loss": 1.7512,
+      "step": 5016
+    },
+    {
+      "epoch": 0.2192840596179903,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.86313156247586e-05,
+      "loss": 1.7813,
+      "step": 5017
+    },
+    {
+      "epoch": 0.21932776782202018,
+      "grad_norm": 2.0,
+      "learning_rate": 8.862695460854347e-05,
+      "loss": 1.8943,
+      "step": 5018
+    },
+    {
+      "epoch": 0.2193714760260501,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.862259286338145e-05,
+      "loss": 1.486,
+      "step": 5019
+    },
+    {
+      "epoch": 0.21941518423008,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.861823038935493e-05,
+      "loss": 1.7871,
+      "step": 5020
+    },
+    {
+      "epoch": 0.21945889243410988,
+      "grad_norm": 2.0,
+      "learning_rate": 8.861386718654618e-05,
+      "loss": 1.8623,
+      "step": 5021
+    },
+    {
+      "epoch": 0.21950260063813978,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.860950325503754e-05,
+      "loss": 2.0073,
+      "step": 5022
+    },
+    {
+      "epoch": 0.21954630884216966,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.86051385949114e-05,
+      "loss": 1.7694,
+      "step": 5023
+    },
+    {
+      "epoch": 0.21959001704619957,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.86007732062501e-05,
+      "loss": 2.0825,
+      "step": 5024
+    },
+    {
+      "epoch": 0.21963372525022948,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.859640708913603e-05,
+      "loss": 2.2887,
+      "step": 5025
+    },
+    {
+      "epoch": 0.21967743345425936,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.859204024365159e-05,
+      "loss": 1.7923,
+      "step": 5026
+    },
+    {
+      "epoch": 0.21972114165828927,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.858767266987917e-05,
+      "loss": 2.2799,
+      "step": 5027
+    },
+    {
+      "epoch": 0.21976484986231914,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.85833043679012e-05,
+      "loss": 2.2855,
+      "step": 5028
+    },
+    {
+      "epoch": 0.21980855806634905,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.857893533780015e-05,
+      "loss": 2.3333,
+      "step": 5029
+    },
+    {
+      "epoch": 0.21985226627037896,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.857456557965842e-05,
+      "loss": 1.4897,
+      "step": 5030
+    },
+    {
+      "epoch": 0.21989597447440884,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.857019509355851e-05,
+      "loss": 2.6878,
+      "step": 5031
+    },
+    {
+      "epoch": 0.21993968267843875,
+      "grad_norm": 6.40625,
+      "learning_rate": 8.856582387958286e-05,
+      "loss": 1.858,
+      "step": 5032
+    },
+    {
+      "epoch": 0.21998339088246863,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.8561451937814e-05,
+      "loss": 2.4977,
+      "step": 5033
+    },
+    {
+      "epoch": 0.22002709908649853,
+      "grad_norm": 2.0,
+      "learning_rate": 8.855707926833441e-05,
+      "loss": 1.8504,
+      "step": 5034
+    },
+    {
+      "epoch": 0.22007080729052844,
+      "grad_norm": 2.25,
+      "learning_rate": 8.855270587122661e-05,
+      "loss": 2.276,
+      "step": 5035
+    },
+    {
+      "epoch": 0.22011451549455832,
+      "grad_norm": 2.5,
+      "learning_rate": 8.854833174657317e-05,
+      "loss": 1.9943,
+      "step": 5036
+    },
+    {
+      "epoch": 0.22015822369858823,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.854395689445658e-05,
+      "loss": 2.0906,
+      "step": 5037
+    },
+    {
+      "epoch": 0.2202019319026181,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.853958131495943e-05,
+      "loss": 1.8033,
+      "step": 5038
+    },
+    {
+      "epoch": 0.22024564010664802,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.853520500816429e-05,
+      "loss": 1.8322,
+      "step": 5039
+    },
+    {
+      "epoch": 0.22028934831067792,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.853082797415374e-05,
+      "loss": 1.8526,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2203330565147078,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.85264502130104e-05,
+      "loss": 2.4182,
+      "step": 5041
+    },
+    {
+      "epoch": 0.2203767647187377,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.852207172481686e-05,
+      "loss": 1.7747,
+      "step": 5042
+    },
+    {
+      "epoch": 0.2204204729227676,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.851769250965577e-05,
+      "loss": 2.3158,
+      "step": 5043
+    },
+    {
+      "epoch": 0.2204641811267975,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.851331256760975e-05,
+      "loss": 1.5608,
+      "step": 5044
+    },
+    {
+      "epoch": 0.2205078893308274,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.850893189876149e-05,
+      "loss": 2.1841,
+      "step": 5045
+    },
+    {
+      "epoch": 0.22055159753485729,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.850455050319361e-05,
+      "loss": 1.779,
+      "step": 5046
+    },
+    {
+      "epoch": 0.2205953057388872,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.850016838098885e-05,
+      "loss": 1.5091,
+      "step": 5047
+    },
+    {
+      "epoch": 0.22063901394291707,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.849578553222985e-05,
+      "loss": 2.0306,
+      "step": 5048
+    },
+    {
+      "epoch": 0.22068272214694698,
+      "grad_norm": 1.859375,
+      "learning_rate": 8.849140195699936e-05,
+      "loss": 1.5791,
+      "step": 5049
+    },
+    {
+      "epoch": 0.2207264303509769,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.84870176553801e-05,
+      "loss": 1.8366,
+      "step": 5050
+    },
+    {
+      "epoch": 0.22077013855500677,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.84826326274548e-05,
+      "loss": 2.0644,
+      "step": 5051
+    },
+    {
+      "epoch": 0.22081384675903667,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.847824687330621e-05,
+      "loss": 1.9944,
+      "step": 5052
+    },
+    {
+      "epoch": 0.22085755496306655,
+      "grad_norm": 2.75,
+      "learning_rate": 8.84738603930171e-05,
+      "loss": 2.3579,
+      "step": 5053
+    },
+    {
+      "epoch": 0.22090126316709646,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.846947318667025e-05,
+      "loss": 1.8134,
+      "step": 5054
+    },
+    {
+      "epoch": 0.22094497137112637,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.846508525434845e-05,
+      "loss": 2.1187,
+      "step": 5055
+    },
+    {
+      "epoch": 0.22098867957515625,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.846069659613451e-05,
+      "loss": 2.0825,
+      "step": 5056
+    },
+    {
+      "epoch": 0.22103238777918616,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.845630721211124e-05,
+      "loss": 2.2968,
+      "step": 5057
+    },
+    {
+      "epoch": 0.22107609598321604,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.84519171023615e-05,
+      "loss": 1.684,
+      "step": 5058
+    },
+    {
+      "epoch": 0.22111980418724594,
+      "grad_norm": 2.875,
+      "learning_rate": 8.84475262669681e-05,
+      "loss": 1.9728,
+      "step": 5059
+    },
+    {
+      "epoch": 0.22116351239127585,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.844313470601393e-05,
+      "loss": 2.5474,
+      "step": 5060
+    },
+    {
+      "epoch": 0.22120722059530573,
+      "grad_norm": 4.28125,
+      "learning_rate": 8.843874241958186e-05,
+      "loss": 1.8955,
+      "step": 5061
+    },
+    {
+      "epoch": 0.22125092879933564,
+      "grad_norm": 5.65625,
+      "learning_rate": 8.843434940775476e-05,
+      "loss": 1.4155,
+      "step": 5062
+    },
+    {
+      "epoch": 0.22129463700336552,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.842995567061558e-05,
+      "loss": 2.2181,
+      "step": 5063
+    },
+    {
+      "epoch": 0.22133834520739543,
+      "grad_norm": 2.375,
+      "learning_rate": 8.842556120824719e-05,
+      "loss": 1.9322,
+      "step": 5064
+    },
+    {
+      "epoch": 0.22138205341142533,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.842116602073252e-05,
+      "loss": 1.7316,
+      "step": 5065
+    },
+    {
+      "epoch": 0.2214257616154552,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.841677010815455e-05,
+      "loss": 2.1623,
+      "step": 5066
+    },
+    {
+      "epoch": 0.22146946981948512,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.841237347059618e-05,
+      "loss": 1.8282,
+      "step": 5067
+    },
+    {
+      "epoch": 0.221513178023515,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.840797610814045e-05,
+      "loss": 2.1581,
+      "step": 5068
+    },
+    {
+      "epoch": 0.2215568862275449,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.84035780208703e-05,
+      "loss": 2.4111,
+      "step": 5069
+    },
+    {
+      "epoch": 0.22160059443157482,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.839917920886874e-05,
+      "loss": 2.1053,
+      "step": 5070
+    },
+    {
+      "epoch": 0.2216443026356047,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.839477967221879e-05,
+      "loss": 1.8067,
+      "step": 5071
+    },
+    {
+      "epoch": 0.2216880108396346,
+      "grad_norm": 2.125,
+      "learning_rate": 8.839037941100344e-05,
+      "loss": 2.0825,
+      "step": 5072
+    },
+    {
+      "epoch": 0.22173171904366448,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.838597842530578e-05,
+      "loss": 2.5282,
+      "step": 5073
+    },
+    {
+      "epoch": 0.2217754272476944,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.838157671520884e-05,
+      "loss": 1.9524,
+      "step": 5074
+    },
+    {
+      "epoch": 0.2218191354517243,
+      "grad_norm": 2.125,
+      "learning_rate": 8.837717428079566e-05,
+      "loss": 2.0709,
+      "step": 5075
+    },
+    {
+      "epoch": 0.22186284365575418,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.837277112214937e-05,
+      "loss": 1.9269,
+      "step": 5076
+    },
+    {
+      "epoch": 0.22190655185978408,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.836836723935303e-05,
+      "loss": 1.8649,
+      "step": 5077
+    },
+    {
+      "epoch": 0.22195026006381396,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.836396263248976e-05,
+      "loss": 3.6734,
+      "step": 5078
+    },
+    {
+      "epoch": 0.22199396826784387,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.835955730164269e-05,
+      "loss": 2.067,
+      "step": 5079
+    },
+    {
+      "epoch": 0.22203767647187378,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.835515124689494e-05,
+      "loss": 2.0929,
+      "step": 5080
+    },
+    {
+      "epoch": 0.22208138467590366,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.835074446832965e-05,
+      "loss": 1.9263,
+      "step": 5081
+    },
+    {
+      "epoch": 0.22212509287993357,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.834633696603e-05,
+      "loss": 1.5586,
+      "step": 5082
+    },
+    {
+      "epoch": 0.22216880108396345,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.834192874007916e-05,
+      "loss": 2.14,
+      "step": 5083
+    },
+    {
+      "epoch": 0.22221250928799335,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.833751979056032e-05,
+      "loss": 2.221,
+      "step": 5084
+    },
+    {
+      "epoch": 0.22225621749202326,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.833311011755668e-05,
+      "loss": 1.5307,
+      "step": 5085
+    },
+    {
+      "epoch": 0.22229992569605314,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.832869972115148e-05,
+      "loss": 1.4644,
+      "step": 5086
+    },
+    {
+      "epoch": 0.22234363390008305,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.832428860142792e-05,
+      "loss": 1.6832,
+      "step": 5087
+    },
+    {
+      "epoch": 0.22238734210411293,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.831987675846924e-05,
+      "loss": 2.2758,
+      "step": 5088
+    },
+    {
+      "epoch": 0.22243105030814284,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.831546419235873e-05,
+      "loss": 1.977,
+      "step": 5089
+    },
+    {
+      "epoch": 0.22247475851217274,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.831105090317965e-05,
+      "loss": 1.7981,
+      "step": 5090
+    },
+    {
+      "epoch": 0.22251846671620262,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.830663689101529e-05,
+      "loss": 1.8568,
+      "step": 5091
+    },
+    {
+      "epoch": 0.22256217492023253,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.83022221559489e-05,
+      "loss": 2.2899,
+      "step": 5092
+    },
+    {
+      "epoch": 0.2226058831242624,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.829780669806387e-05,
+      "loss": 1.7343,
+      "step": 5093
+    },
+    {
+      "epoch": 0.22264959132829232,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.829339051744346e-05,
+      "loss": 2.4551,
+      "step": 5094
+    },
+    {
+      "epoch": 0.22269329953232223,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.828897361417106e-05,
+      "loss": 1.7362,
+      "step": 5095
+    },
+    {
+      "epoch": 0.2227370077363521,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.828455598832998e-05,
+      "loss": 2.0821,
+      "step": 5096
+    },
+    {
+      "epoch": 0.222780715940382,
+      "grad_norm": 2.0,
+      "learning_rate": 8.828013764000362e-05,
+      "loss": 1.9559,
+      "step": 5097
+    },
+    {
+      "epoch": 0.2228244241444119,
+      "grad_norm": 2.25,
+      "learning_rate": 8.827571856927535e-05,
+      "loss": 1.6318,
+      "step": 5098
+    },
+    {
+      "epoch": 0.2228681323484418,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.827129877622857e-05,
+      "loss": 1.8371,
+      "step": 5099
+    },
+    {
+      "epoch": 0.2229118405524717,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.826687826094666e-05,
+      "loss": 1.8739,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2229555487565016,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.826245702351309e-05,
+      "loss": 2.28,
+      "step": 5101
+    },
+    {
+      "epoch": 0.2229992569605315,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.825803506401125e-05,
+      "loss": 1.8682,
+      "step": 5102
+    },
+    {
+      "epoch": 0.2230429651645614,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.82536123825246e-05,
+      "loss": 1.5349,
+      "step": 5103
+    },
+    {
+      "epoch": 0.22308667336859128,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.824918897913661e-05,
+      "loss": 1.752,
+      "step": 5104
+    },
+    {
+      "epoch": 0.2231303815726212,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.824476485393076e-05,
+      "loss": 1.6159,
+      "step": 5105
+    },
+    {
+      "epoch": 0.22317408977665107,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.824034000699055e-05,
+      "loss": 1.7539,
+      "step": 5106
+    },
+    {
+      "epoch": 0.22321779798068098,
+      "grad_norm": 2.0,
+      "learning_rate": 8.823591443839944e-05,
+      "loss": 1.666,
+      "step": 5107
+    },
+    {
+      "epoch": 0.22326150618471088,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.8231488148241e-05,
+      "loss": 2.2305,
+      "step": 5108
+    },
+    {
+      "epoch": 0.22330521438874076,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.822706113659872e-05,
+      "loss": 2.1343,
+      "step": 5109
+    },
+    {
+      "epoch": 0.22334892259277067,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.822263340355616e-05,
+      "loss": 1.7487,
+      "step": 5110
+    },
+    {
+      "epoch": 0.22339263079680055,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.82182049491969e-05,
+      "loss": 2.2869,
+      "step": 5111
+    },
+    {
+      "epoch": 0.22343633900083046,
+      "grad_norm": 2.0,
+      "learning_rate": 8.821377577360446e-05,
+      "loss": 2.0224,
+      "step": 5112
+    },
+    {
+      "epoch": 0.22348004720486037,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.820934587686247e-05,
+      "loss": 2.3331,
+      "step": 5113
+    },
+    {
+      "epoch": 0.22352375540889025,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.82049152590545e-05,
+      "loss": 1.472,
+      "step": 5114
+    },
+    {
+      "epoch": 0.22356746361292015,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.820048392026417e-05,
+      "loss": 1.8046,
+      "step": 5115
+    },
+    {
+      "epoch": 0.22361117181695003,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.819605186057514e-05,
+      "loss": 1.7677,
+      "step": 5116
+    },
+    {
+      "epoch": 0.22365488002097994,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.819161908007099e-05,
+      "loss": 1.8259,
+      "step": 5117
+    },
+    {
+      "epoch": 0.22369858822500985,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.818718557883541e-05,
+      "loss": 1.3694,
+      "step": 5118
+    },
+    {
+      "epoch": 0.22374229642903973,
+      "grad_norm": 2.0,
+      "learning_rate": 8.818275135695207e-05,
+      "loss": 1.6956,
+      "step": 5119
+    },
+    {
+      "epoch": 0.22378600463306964,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.817831641450462e-05,
+      "loss": 1.988,
+      "step": 5120
+    },
+    {
+      "epoch": 0.22382971283709951,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.81738807515768e-05,
+      "loss": 1.4667,
+      "step": 5121
+    },
+    {
+      "epoch": 0.22387342104112942,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.816944436825228e-05,
+      "loss": 1.899,
+      "step": 5122
+    },
+    {
+      "epoch": 0.22391712924515933,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.816500726461478e-05,
+      "loss": 2.2279,
+      "step": 5123
+    },
+    {
+      "epoch": 0.2239608374491892,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.816056944074805e-05,
+      "loss": 2.3851,
+      "step": 5124
+    },
+    {
+      "epoch": 0.22400454565321912,
+      "grad_norm": 2.375,
+      "learning_rate": 8.815613089673583e-05,
+      "loss": 1.4785,
+      "step": 5125
+    },
+    {
+      "epoch": 0.224048253857249,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.81516916326619e-05,
+      "loss": 2.0775,
+      "step": 5126
+    },
+    {
+      "epoch": 0.2240919620612789,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.814725164861001e-05,
+      "loss": 2.5111,
+      "step": 5127
+    },
+    {
+      "epoch": 0.2241356702653088,
+      "grad_norm": 2.125,
+      "learning_rate": 8.814281094466398e-05,
+      "loss": 1.7167,
+      "step": 5128
+    },
+    {
+      "epoch": 0.2241793784693387,
+      "grad_norm": 2.375,
+      "learning_rate": 8.813836952090758e-05,
+      "loss": 2.0692,
+      "step": 5129
+    },
+    {
+      "epoch": 0.2242230866733686,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.813392737742463e-05,
+      "loss": 1.7218,
+      "step": 5130
+    },
+    {
+      "epoch": 0.22426679487739848,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.812948451429898e-05,
+      "loss": 2.4234,
+      "step": 5131
+    },
+    {
+      "epoch": 0.2243105030814284,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.812504093161446e-05,
+      "loss": 1.8171,
+      "step": 5132
+    },
+    {
+      "epoch": 0.2243542112854583,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.812059662945494e-05,
+      "loss": 1.8073,
+      "step": 5133
+    },
+    {
+      "epoch": 0.22439791948948817,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.811615160790427e-05,
+      "loss": 1.7733,
+      "step": 5134
+    },
+    {
+      "epoch": 0.22444162769351808,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.811170586704633e-05,
+      "loss": 2.5059,
+      "step": 5135
+    },
+    {
+      "epoch": 0.22448533589754796,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.810725940696505e-05,
+      "loss": 1.9448,
+      "step": 5136
+    },
+    {
+      "epoch": 0.22452904410157787,
+      "grad_norm": 3.625,
+      "learning_rate": 8.810281222774432e-05,
+      "loss": 2.2897,
+      "step": 5137
+    },
+    {
+      "epoch": 0.22457275230560778,
+      "grad_norm": 3.921875,
+      "learning_rate": 8.809836432946808e-05,
+      "loss": 1.8743,
+      "step": 5138
+    },
+    {
+      "epoch": 0.22461646050963766,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.809391571222023e-05,
+      "loss": 1.9626,
+      "step": 5139
+    },
+    {
+      "epoch": 0.22466016871366756,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.808946637608477e-05,
+      "loss": 1.7792,
+      "step": 5140
+    },
+    {
+      "epoch": 0.22470387691769744,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.808501632114563e-05,
+      "loss": 1.9027,
+      "step": 5141
+    },
+    {
+      "epoch": 0.22474758512172735,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.808056554748681e-05,
+      "loss": 2.5154,
+      "step": 5142
+    },
+    {
+      "epoch": 0.22479129332575726,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.807611405519227e-05,
+      "loss": 1.9519,
+      "step": 5143
+    },
+    {
+      "epoch": 0.22483500152978714,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.807166184434607e-05,
+      "loss": 2.3366,
+      "step": 5144
+    },
+    {
+      "epoch": 0.22487870973381704,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.806720891503219e-05,
+      "loss": 1.8174,
+      "step": 5145
+    },
+    {
+      "epoch": 0.22492241793784692,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.806275526733466e-05,
+      "loss": 1.7592,
+      "step": 5146
+    },
+    {
+      "epoch": 0.22496612614187683,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.805830090133755e-05,
+      "loss": 2.2325,
+      "step": 5147
+    },
+    {
+      "epoch": 0.22500983434590674,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.805384581712492e-05,
+      "loss": 1.9405,
+      "step": 5148
+    },
+    {
+      "epoch": 0.22505354254993662,
+      "grad_norm": 2.125,
+      "learning_rate": 8.804939001478083e-05,
+      "loss": 1.8371,
+      "step": 5149
+    },
+    {
+      "epoch": 0.22509725075396653,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.804493349438937e-05,
+      "loss": 1.6809,
+      "step": 5150
+    },
+    {
+      "epoch": 0.2251409589579964,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.804047625603464e-05,
+      "loss": 2.0279,
+      "step": 5151
+    },
+    {
+      "epoch": 0.22518466716202631,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.803601829980076e-05,
+      "loss": 2.1587,
+      "step": 5152
+    },
+    {
+      "epoch": 0.22522837536605622,
+      "grad_norm": 2.125,
+      "learning_rate": 8.803155962577186e-05,
+      "loss": 1.6492,
+      "step": 5153
+    },
+    {
+      "epoch": 0.2252720835700861,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.802710023403208e-05,
+      "loss": 2.2723,
+      "step": 5154
+    },
+    {
+      "epoch": 0.225315791774116,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.802264012466557e-05,
+      "loss": 2.0675,
+      "step": 5155
+    },
+    {
+      "epoch": 0.2253594999781459,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.801817929775649e-05,
+      "loss": 1.8635,
+      "step": 5156
+    },
+    {
+      "epoch": 0.2254032081821758,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.801371775338904e-05,
+      "loss": 2.1729,
+      "step": 5157
+    },
+    {
+      "epoch": 0.2254469163862057,
+      "grad_norm": 2.25,
+      "learning_rate": 8.800925549164741e-05,
+      "loss": 1.7483,
+      "step": 5158
+    },
+    {
+      "epoch": 0.22549062459023558,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.800479251261581e-05,
+      "loss": 1.9852,
+      "step": 5159
+    },
+    {
+      "epoch": 0.2255343327942655,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.800032881637847e-05,
+      "loss": 1.364,
+      "step": 5160
+    },
+    {
+      "epoch": 0.22557804099829537,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.799586440301961e-05,
+      "loss": 2.4219,
+      "step": 5161
+    },
+    {
+      "epoch": 0.22562174920232528,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.79913992726235e-05,
+      "loss": 2.3594,
+      "step": 5162
+    },
+    {
+      "epoch": 0.22566545740635519,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.798693342527438e-05,
+      "loss": 1.6779,
+      "step": 5163
+    },
+    {
+      "epoch": 0.22570916561038507,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.798246686105654e-05,
+      "loss": 3.4992,
+      "step": 5164
+    },
+    {
+      "epoch": 0.22575287381441497,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.797799958005426e-05,
+      "loss": 1.3351,
+      "step": 5165
+    },
+    {
+      "epoch": 0.22579658201844485,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.797353158235186e-05,
+      "loss": 1.9769,
+      "step": 5166
+    },
+    {
+      "epoch": 0.22584029022247476,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.796906286803365e-05,
+      "loss": 1.8765,
+      "step": 5167
+    },
+    {
+      "epoch": 0.22588399842650467,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.796459343718397e-05,
+      "loss": 2.3319,
+      "step": 5168
+    },
+    {
+      "epoch": 0.22592770663053455,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.796012328988716e-05,
+      "loss": 1.7823,
+      "step": 5169
+    },
+    {
+      "epoch": 0.22597141483456445,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.795565242622758e-05,
+      "loss": 1.6866,
+      "step": 5170
+    },
+    {
+      "epoch": 0.22601512303859433,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.795118084628959e-05,
+      "loss": 1.9945,
+      "step": 5171
+    },
+    {
+      "epoch": 0.22605883124262424,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.794670855015757e-05,
+      "loss": 1.8617,
+      "step": 5172
+    },
+    {
+      "epoch": 0.22610253944665415,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.794223553791595e-05,
+      "loss": 2.4788,
+      "step": 5173
+    },
+    {
+      "epoch": 0.22614624765068403,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.79377618096491e-05,
+      "loss": 2.0072,
+      "step": 5174
+    },
+    {
+      "epoch": 0.22618995585471394,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.793328736544149e-05,
+      "loss": 2.0756,
+      "step": 5175
+    },
+    {
+      "epoch": 0.22623366405874382,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.792881220537751e-05,
+      "loss": 2.3657,
+      "step": 5176
+    },
+    {
+      "epoch": 0.22627737226277372,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.792433632954167e-05,
+      "loss": 2.5119,
+      "step": 5177
+    },
+    {
+      "epoch": 0.22632108046680363,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.791985973801839e-05,
+      "loss": 2.3355,
+      "step": 5178
+    },
+    {
+      "epoch": 0.2263647886708335,
+      "grad_norm": 4.0,
+      "learning_rate": 8.791538243089219e-05,
+      "loss": 2.1918,
+      "step": 5179
+    },
+    {
+      "epoch": 0.22640849687486342,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.791090440824752e-05,
+      "loss": 1.9486,
+      "step": 5180
+    },
+    {
+      "epoch": 0.2264522050788933,
+      "grad_norm": 2.0,
+      "learning_rate": 8.790642567016891e-05,
+      "loss": 2.2691,
+      "step": 5181
+    },
+    {
+      "epoch": 0.2264959132829232,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.790194621674088e-05,
+      "loss": 2.2397,
+      "step": 5182
+    },
+    {
+      "epoch": 0.2265396214869531,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.789746604804796e-05,
+      "loss": 1.9892,
+      "step": 5183
+    },
+    {
+      "epoch": 0.226583329690983,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.78929851641747e-05,
+      "loss": 1.6987,
+      "step": 5184
+    },
+    {
+      "epoch": 0.2266270378950129,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.788850356520566e-05,
+      "loss": 1.6268,
+      "step": 5185
+    },
+    {
+      "epoch": 0.22667074609904278,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.788402125122542e-05,
+      "loss": 2.4155,
+      "step": 5186
+    },
+    {
+      "epoch": 0.2267144543030727,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.787953822231855e-05,
+      "loss": 2.1763,
+      "step": 5187
+    },
+    {
+      "epoch": 0.2267581625071026,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.787505447856967e-05,
+      "loss": 1.7471,
+      "step": 5188
+    },
+    {
+      "epoch": 0.22680187071113247,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.787057002006337e-05,
+      "loss": 2.0579,
+      "step": 5189
+    },
+    {
+      "epoch": 0.22684557891516238,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.786608484688432e-05,
+      "loss": 1.7905,
+      "step": 5190
+    },
+    {
+      "epoch": 0.22688928711919226,
+      "grad_norm": 13.9375,
+      "learning_rate": 8.786159895911712e-05,
+      "loss": 2.8267,
+      "step": 5191
+    },
+    {
+      "epoch": 0.22693299532322217,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.785711235684647e-05,
+      "loss": 2.1309,
+      "step": 5192
+    },
+    {
+      "epoch": 0.22697670352725208,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.785262504015696e-05,
+      "loss": 1.6625,
+      "step": 5193
+    },
+    {
+      "epoch": 0.22702041173128196,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.784813700913337e-05,
+      "loss": 1.8822,
+      "step": 5194
+    },
+    {
+      "epoch": 0.22706411993531186,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.784364826386034e-05,
+      "loss": 2.059,
+      "step": 5195
+    },
+    {
+      "epoch": 0.22710782813934174,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.783915880442257e-05,
+      "loss": 2.0635,
+      "step": 5196
+    },
+    {
+      "epoch": 0.22715153634337165,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.783466863090482e-05,
+      "loss": 2.2928,
+      "step": 5197
+    },
+    {
+      "epoch": 0.22719524454740156,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.78301777433918e-05,
+      "loss": 2.1231,
+      "step": 5198
+    },
+    {
+      "epoch": 0.22723895275143144,
+      "grad_norm": 3.9375,
+      "learning_rate": 8.782568614196827e-05,
+      "loss": 3.9269,
+      "step": 5199
+    },
+    {
+      "epoch": 0.22728266095546135,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.782119382671898e-05,
+      "loss": 1.8587,
+      "step": 5200
+    },
+    {
+      "epoch": 0.22732636915949123,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.781670079772873e-05,
+      "loss": 1.5783,
+      "step": 5201
+    },
+    {
+      "epoch": 0.22737007736352113,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.781220705508229e-05,
+      "loss": 1.6649,
+      "step": 5202
+    },
+    {
+      "epoch": 0.22741378556755104,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.780771259886447e-05,
+      "loss": 2.1052,
+      "step": 5203
+    },
+    {
+      "epoch": 0.22745749377158092,
+      "grad_norm": 2.75,
+      "learning_rate": 8.780321742916008e-05,
+      "loss": 2.1247,
+      "step": 5204
+    },
+    {
+      "epoch": 0.22750120197561083,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.779872154605397e-05,
+      "loss": 1.8571,
+      "step": 5205
+    },
+    {
+      "epoch": 0.2275449101796407,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.779422494963096e-05,
+      "loss": 2.2196,
+      "step": 5206
+    },
+    {
+      "epoch": 0.22758861838367062,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.778972763997592e-05,
+      "loss": 2.0445,
+      "step": 5207
+    },
+    {
+      "epoch": 0.22763232658770052,
+      "grad_norm": 3.484375,
+      "learning_rate": 8.778522961717373e-05,
+      "loss": 2.1614,
+      "step": 5208
+    },
+    {
+      "epoch": 0.2276760347917304,
+      "grad_norm": 2.5,
+      "learning_rate": 8.778073088130925e-05,
+      "loss": 2.4693,
+      "step": 5209
+    },
+    {
+      "epoch": 0.2277197429957603,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.777623143246741e-05,
+      "loss": 2.0724,
+      "step": 5210
+    },
+    {
+      "epoch": 0.2277634511997902,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.777173127073308e-05,
+      "loss": 1.5359,
+      "step": 5211
+    },
+    {
+      "epoch": 0.2278071594038201,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.776723039619121e-05,
+      "loss": 1.9064,
+      "step": 5212
+    },
+    {
+      "epoch": 0.22785086760785,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.776272880892675e-05,
+      "loss": 1.9944,
+      "step": 5213
+    },
+    {
+      "epoch": 0.22789457581187988,
+      "grad_norm": 2.5,
+      "learning_rate": 8.775822650902463e-05,
+      "loss": 2.0785,
+      "step": 5214
+    },
+    {
+      "epoch": 0.2279382840159098,
+      "grad_norm": 4.84375,
+      "learning_rate": 8.775372349656981e-05,
+      "loss": 2.0266,
+      "step": 5215
+    },
+    {
+      "epoch": 0.22798199221993967,
+      "grad_norm": 3.796875,
+      "learning_rate": 8.77492197716473e-05,
+      "loss": 2.3301,
+      "step": 5216
+    },
+    {
+      "epoch": 0.22802570042396958,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.774471533434206e-05,
+      "loss": 1.6443,
+      "step": 5217
+    },
+    {
+      "epoch": 0.2280694086279995,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.77402101847391e-05,
+      "loss": 1.6125,
+      "step": 5218
+    },
+    {
+      "epoch": 0.22811311683202937,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.773570432292344e-05,
+      "loss": 1.7902,
+      "step": 5219
+    },
+    {
+      "epoch": 0.22815682503605927,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.773119774898013e-05,
+      "loss": 2.0626,
+      "step": 5220
+    },
+    {
+      "epoch": 0.22820053324008915,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.77266904629942e-05,
+      "loss": 2.1337,
+      "step": 5221
+    },
+    {
+      "epoch": 0.22824424144411906,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.772218246505073e-05,
+      "loss": 2.8284,
+      "step": 5222
+    },
+    {
+      "epoch": 0.22828794964814897,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.771767375523475e-05,
+      "loss": 1.8774,
+      "step": 5223
+    },
+    {
+      "epoch": 0.22833165785217885,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.771316433363138e-05,
+      "loss": 1.9789,
+      "step": 5224
+    },
+    {
+      "epoch": 0.22837536605620876,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.770865420032571e-05,
+      "loss": 2.3812,
+      "step": 5225
+    },
+    {
+      "epoch": 0.22841907426023864,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.770414335540285e-05,
+      "loss": 2.1789,
+      "step": 5226
+    },
+    {
+      "epoch": 0.22846278246426854,
+      "grad_norm": 2.125,
+      "learning_rate": 8.769963179894793e-05,
+      "loss": 1.7856,
+      "step": 5227
+    },
+    {
+      "epoch": 0.22850649066829845,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.76951195310461e-05,
+      "loss": 2.0435,
+      "step": 5228
+    },
+    {
+      "epoch": 0.22855019887232833,
+      "grad_norm": 3.78125,
+      "learning_rate": 8.76906065517825e-05,
+      "loss": 2.341,
+      "step": 5229
+    },
+    {
+      "epoch": 0.22859390707635824,
+      "grad_norm": 2.125,
+      "learning_rate": 8.768609286124228e-05,
+      "loss": 1.9607,
+      "step": 5230
+    },
+    {
+      "epoch": 0.22863761528038812,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.768157845951065e-05,
+      "loss": 2.5474,
+      "step": 5231
+    },
+    {
+      "epoch": 0.22868132348441803,
+      "grad_norm": 2.25,
+      "learning_rate": 8.767706334667279e-05,
+      "loss": 1.9932,
+      "step": 5232
+    },
+    {
+      "epoch": 0.22872503168844793,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.767254752281392e-05,
+      "loss": 1.9982,
+      "step": 5233
+    },
+    {
+      "epoch": 0.2287687398924778,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.766803098801924e-05,
+      "loss": 1.7068,
+      "step": 5234
+    },
+    {
+      "epoch": 0.22881244809650772,
+      "grad_norm": 2.375,
+      "learning_rate": 8.766351374237398e-05,
+      "loss": 2.1922,
+      "step": 5235
+    },
+    {
+      "epoch": 0.2288561563005376,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.765899578596342e-05,
+      "loss": 2.0894,
+      "step": 5236
+    },
+    {
+      "epoch": 0.2288998645045675,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.765447711887278e-05,
+      "loss": 2.5275,
+      "step": 5237
+    },
+    {
+      "epoch": 0.22894357270859741,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.764995774118736e-05,
+      "loss": 2.0998,
+      "step": 5238
+    },
+    {
+      "epoch": 0.2289872809126273,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.764543765299245e-05,
+      "loss": 1.943,
+      "step": 5239
+    },
+    {
+      "epoch": 0.2290309891166572,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.764091685437333e-05,
+      "loss": 1.6759,
+      "step": 5240
+    },
+    {
+      "epoch": 0.22907469732068708,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.763639534541533e-05,
+      "loss": 2.0194,
+      "step": 5241
+    },
+    {
+      "epoch": 0.229118405524717,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.763187312620377e-05,
+      "loss": 1.8292,
+      "step": 5242
+    },
+    {
+      "epoch": 0.2291621137287469,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.762735019682399e-05,
+      "loss": 1.873,
+      "step": 5243
+    },
+    {
+      "epoch": 0.22920582193277678,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.762282655736136e-05,
+      "loss": 1.5856,
+      "step": 5244
+    },
+    {
+      "epoch": 0.22924953013680668,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.761830220790123e-05,
+      "loss": 2.082,
+      "step": 5245
+    },
+    {
+      "epoch": 0.22929323834083656,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.761377714852899e-05,
+      "loss": 2.1958,
+      "step": 5246
+    },
+    {
+      "epoch": 0.22933694654486647,
+      "grad_norm": 2.25,
+      "learning_rate": 8.760925137933001e-05,
+      "loss": 2.0435,
+      "step": 5247
+    },
+    {
+      "epoch": 0.22938065474889638,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.760472490038974e-05,
+      "loss": 2.3615,
+      "step": 5248
+    },
+    {
+      "epoch": 0.22942436295292626,
+      "grad_norm": 1.875,
+      "learning_rate": 8.760019771179358e-05,
+      "loss": 1.7286,
+      "step": 5249
+    },
+    {
+      "epoch": 0.22946807115695617,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.759566981362694e-05,
+      "loss": 1.8911,
+      "step": 5250
+    },
+    {
+      "epoch": 0.22951177936098605,
+      "grad_norm": 9.9375,
+      "learning_rate": 8.759114120597532e-05,
+      "loss": 1.535,
+      "step": 5251
+    },
+    {
+      "epoch": 0.22955548756501595,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.758661188892414e-05,
+      "loss": 2.203,
+      "step": 5252
+    },
+    {
+      "epoch": 0.22959919576904586,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.75820818625589e-05,
+      "loss": 2.2292,
+      "step": 5253
+    },
+    {
+      "epoch": 0.22964290397307574,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.757755112696509e-05,
+      "loss": 2.431,
+      "step": 5254
+    },
+    {
+      "epoch": 0.22968661217710565,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.757301968222817e-05,
+      "loss": 3.2358,
+      "step": 5255
+    },
+    {
+      "epoch": 0.22973032038113553,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.756848752843369e-05,
+      "loss": 1.7219,
+      "step": 5256
+    },
+    {
+      "epoch": 0.22977402858516544,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.756395466566719e-05,
+      "loss": 1.6914,
+      "step": 5257
+    },
+    {
+      "epoch": 0.22981773678919534,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.755942109401417e-05,
+      "loss": 2.3572,
+      "step": 5258
+    },
+    {
+      "epoch": 0.22986144499322522,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.755488681356022e-05,
+      "loss": 3.0781,
+      "step": 5259
+    },
+    {
+      "epoch": 0.22990515319725513,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.755035182439088e-05,
+      "loss": 1.7704,
+      "step": 5260
+    },
+    {
+      "epoch": 0.229948861401285,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.754581612659178e-05,
+      "loss": 1.8192,
+      "step": 5261
+    },
+    {
+      "epoch": 0.22999256960531492,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.754127972024846e-05,
+      "loss": 2.4582,
+      "step": 5262
+    },
+    {
+      "epoch": 0.23003627780934482,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.753674260544657e-05,
+      "loss": 2.6562,
+      "step": 5263
+    },
+    {
+      "epoch": 0.2300799860133747,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.753220478227171e-05,
+      "loss": 1.5956,
+      "step": 5264
+    },
+    {
+      "epoch": 0.2301236942174046,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.752766625080952e-05,
+      "loss": 1.8347,
+      "step": 5265
+    },
+    {
+      "epoch": 0.2301674024214345,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.752312701114564e-05,
+      "loss": 1.8427,
+      "step": 5266
+    },
+    {
+      "epoch": 0.2302111106254644,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.751858706336576e-05,
+      "loss": 3.1018,
+      "step": 5267
+    },
+    {
+      "epoch": 0.2302548188294943,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.751404640755552e-05,
+      "loss": 1.7006,
+      "step": 5268
+    },
+    {
+      "epoch": 0.2302985270335242,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.750950504380062e-05,
+      "loss": 1.9282,
+      "step": 5269
+    },
+    {
+      "epoch": 0.2303422352375541,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.750496297218678e-05,
+      "loss": 1.5637,
+      "step": 5270
+    },
+    {
+      "epoch": 0.23038594344158397,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.75004201927997e-05,
+      "loss": 1.9573,
+      "step": 5271
+    },
+    {
+      "epoch": 0.23042965164561388,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.749587670572511e-05,
+      "loss": 2.4692,
+      "step": 5272
+    },
+    {
+      "epoch": 0.2304733598496438,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.749133251104876e-05,
+      "loss": 1.6436,
+      "step": 5273
+    },
+    {
+      "epoch": 0.23051706805367367,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.748678760885638e-05,
+      "loss": 2.428,
+      "step": 5274
+    },
+    {
+      "epoch": 0.23056077625770358,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.748224199923378e-05,
+      "loss": 1.7959,
+      "step": 5275
+    },
+    {
+      "epoch": 0.23060448446173346,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.747769568226672e-05,
+      "loss": 1.7832,
+      "step": 5276
+    },
+    {
+      "epoch": 0.23064819266576336,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.747314865804098e-05,
+      "loss": 1.8904,
+      "step": 5277
+    },
+    {
+      "epoch": 0.23069190086979327,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.74686009266424e-05,
+      "loss": 1.6946,
+      "step": 5278
+    },
+    {
+      "epoch": 0.23073560907382315,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.74640524881568e-05,
+      "loss": 2.447,
+      "step": 5279
+    },
+    {
+      "epoch": 0.23077931727785306,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.745950334266997e-05,
+      "loss": 1.8461,
+      "step": 5280
+    },
+    {
+      "epoch": 0.23082302548188294,
+      "grad_norm": 2.25,
+      "learning_rate": 8.745495349026781e-05,
+      "loss": 1.5948,
+      "step": 5281
+    },
+    {
+      "epoch": 0.23086673368591284,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.745040293103616e-05,
+      "loss": 1.9506,
+      "step": 5282
+    },
+    {
+      "epoch": 0.23091044188994275,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.744585166506089e-05,
+      "loss": 1.9788,
+      "step": 5283
+    },
+    {
+      "epoch": 0.23095415009397263,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.744129969242793e-05,
+      "loss": 2.6072,
+      "step": 5284
+    },
+    {
+      "epoch": 0.23099785829800254,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.743674701322312e-05,
+      "loss": 1.6111,
+      "step": 5285
+    },
+    {
+      "epoch": 0.23104156650203242,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.743219362753244e-05,
+      "loss": 1.7419,
+      "step": 5286
+    },
+    {
+      "epoch": 0.23108527470606233,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.742763953544175e-05,
+      "loss": 2.2248,
+      "step": 5287
+    },
+    {
+      "epoch": 0.23112898291009223,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.742308473703706e-05,
+      "loss": 1.6965,
+      "step": 5288
+    },
+    {
+      "epoch": 0.23117269111412211,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.741852923240427e-05,
+      "loss": 1.9022,
+      "step": 5289
+    },
+    {
+      "epoch": 0.23121639931815202,
+      "grad_norm": 4.21875,
+      "learning_rate": 8.741397302162939e-05,
+      "loss": 1.2419,
+      "step": 5290
+    },
+    {
+      "epoch": 0.2312601075221819,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.740941610479838e-05,
+      "loss": 1.4853,
+      "step": 5291
+    },
+    {
+      "epoch": 0.2313038157262118,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.740485848199723e-05,
+      "loss": 2.8512,
+      "step": 5292
+    },
+    {
+      "epoch": 0.23134752393024172,
+      "grad_norm": 2.125,
+      "learning_rate": 8.740030015331198e-05,
+      "loss": 1.8565,
+      "step": 5293
+    },
+    {
+      "epoch": 0.2313912321342716,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.739574111882862e-05,
+      "loss": 2.0053,
+      "step": 5294
+    },
+    {
+      "epoch": 0.2314349403383015,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.73911813786332e-05,
+      "loss": 1.7054,
+      "step": 5295
+    },
+    {
+      "epoch": 0.23147864854233138,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.738662093281177e-05,
+      "loss": 1.8426,
+      "step": 5296
+    },
+    {
+      "epoch": 0.2315223567463613,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.738205978145038e-05,
+      "loss": 2.0349,
+      "step": 5297
+    },
+    {
+      "epoch": 0.2315660649503912,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.737749792463512e-05,
+      "loss": 1.8219,
+      "step": 5298
+    },
+    {
+      "epoch": 0.23160977315442108,
+      "grad_norm": 2.5,
+      "learning_rate": 8.737293536245207e-05,
+      "loss": 3.0215,
+      "step": 5299
+    },
+    {
+      "epoch": 0.23165348135845099,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.736837209498736e-05,
+      "loss": 1.6017,
+      "step": 5300
+    },
+    {
+      "epoch": 0.23169718956248087,
+      "grad_norm": 2.125,
+      "learning_rate": 8.736380812232706e-05,
+      "loss": 2.0057,
+      "step": 5301
+    },
+    {
+      "epoch": 0.23174089776651077,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.735924344455732e-05,
+      "loss": 2.6932,
+      "step": 5302
+    },
+    {
+      "epoch": 0.23178460597054068,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.735467806176429e-05,
+      "loss": 2.12,
+      "step": 5303
+    },
+    {
+      "epoch": 0.23182831417457056,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.735011197403411e-05,
+      "loss": 2.4922,
+      "step": 5304
+    },
+    {
+      "epoch": 0.23187202237860047,
+      "grad_norm": 3.546875,
+      "learning_rate": 8.734554518145297e-05,
+      "loss": 2.1247,
+      "step": 5305
+    },
+    {
+      "epoch": 0.23191573058263035,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.734097768410703e-05,
+      "loss": 2.5513,
+      "step": 5306
+    },
+    {
+      "epoch": 0.23195943878666025,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.733640948208248e-05,
+      "loss": 2.6246,
+      "step": 5307
+    },
+    {
+      "epoch": 0.23200314699069016,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.733184057546558e-05,
+      "loss": 2.9116,
+      "step": 5308
+    },
+    {
+      "epoch": 0.23204685519472004,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.732727096434247e-05,
+      "loss": 2.7499,
+      "step": 5309
+    },
+    {
+      "epoch": 0.23209056339874995,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.732270064879947e-05,
+      "loss": 1.8118,
+      "step": 5310
+    },
+    {
+      "epoch": 0.23213427160277983,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.731812962892277e-05,
+      "loss": 1.5574,
+      "step": 5311
+    },
+    {
+      "epoch": 0.23217797980680974,
+      "grad_norm": 2.875,
+      "learning_rate": 8.731355790479862e-05,
+      "loss": 2.3718,
+      "step": 5312
+    },
+    {
+      "epoch": 0.23222168801083964,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.730898547651337e-05,
+      "loss": 1.4868,
+      "step": 5313
+    },
+    {
+      "epoch": 0.23226539621486952,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.730441234415324e-05,
+      "loss": 2.3289,
+      "step": 5314
+    },
+    {
+      "epoch": 0.23230910441889943,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.729983850780456e-05,
+      "loss": 1.4998,
+      "step": 5315
+    },
+    {
+      "epoch": 0.2323528126229293,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.729526396755365e-05,
+      "loss": 2.767,
+      "step": 5316
+    },
+    {
+      "epoch": 0.23239652082695922,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.729068872348681e-05,
+      "loss": 1.6264,
+      "step": 5317
+    },
+    {
+      "epoch": 0.23244022903098913,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.728611277569042e-05,
+      "loss": 1.6083,
+      "step": 5318
+    },
+    {
+      "epoch": 0.232483937235019,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.72815361242508e-05,
+      "loss": 1.8636,
+      "step": 5319
+    },
+    {
+      "epoch": 0.2325276454390489,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.727695876925433e-05,
+      "loss": 1.5848,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2325713536430788,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.72723807107874e-05,
+      "loss": 2.8408,
+      "step": 5321
+    },
+    {
+      "epoch": 0.2326150618471087,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.72678019489364e-05,
+      "loss": 1.5735,
+      "step": 5322
+    },
+    {
+      "epoch": 0.2326587700511386,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.726322248378775e-05,
+      "loss": 1.8741,
+      "step": 5323
+    },
+    {
+      "epoch": 0.2327024782551685,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.725864231542784e-05,
+      "loss": 1.8679,
+      "step": 5324
+    },
+    {
+      "epoch": 0.2327461864591984,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.725406144394313e-05,
+      "loss": 2.4191,
+      "step": 5325
+    },
+    {
+      "epoch": 0.23278989466322828,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.724947986942004e-05,
+      "loss": 1.9748,
+      "step": 5326
+    },
+    {
+      "epoch": 0.23283360286725818,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.724489759194507e-05,
+      "loss": 1.6349,
+      "step": 5327
+    },
+    {
+      "epoch": 0.2328773110712881,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.724031461160467e-05,
+      "loss": 2.3706,
+      "step": 5328
+    },
+    {
+      "epoch": 0.23292101927531797,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.723573092848534e-05,
+      "loss": 1.7051,
+      "step": 5329
+    },
+    {
+      "epoch": 0.23296472747934788,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.723114654267356e-05,
+      "loss": 1.9907,
+      "step": 5330
+    },
+    {
+      "epoch": 0.23300843568337776,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.722656145425587e-05,
+      "loss": 1.867,
+      "step": 5331
+    },
+    {
+      "epoch": 0.23305214388740766,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.722197566331878e-05,
+      "loss": 2.5761,
+      "step": 5332
+    },
+    {
+      "epoch": 0.23309585209143757,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.721738916994883e-05,
+      "loss": 1.9586,
+      "step": 5333
+    },
+    {
+      "epoch": 0.23313956029546745,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.721280197423258e-05,
+      "loss": 1.9132,
+      "step": 5334
+    },
+    {
+      "epoch": 0.23318326849949736,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.720821407625661e-05,
+      "loss": 2.1847,
+      "step": 5335
+    },
+    {
+      "epoch": 0.23322697670352724,
+      "grad_norm": 4.625,
+      "learning_rate": 8.720362547610747e-05,
+      "loss": 2.6595,
+      "step": 5336
+    },
+    {
+      "epoch": 0.23327068490755715,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.719903617387178e-05,
+      "loss": 2.1804,
+      "step": 5337
+    },
+    {
+      "epoch": 0.23331439311158705,
+      "grad_norm": 2.375,
+      "learning_rate": 8.719444616963613e-05,
+      "loss": 2.4745,
+      "step": 5338
+    },
+    {
+      "epoch": 0.23335810131561693,
+      "grad_norm": 1.8359375,
+      "learning_rate": 8.718985546348715e-05,
+      "loss": 1.5742,
+      "step": 5339
+    },
+    {
+      "epoch": 0.23340180951964684,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.718526405551146e-05,
+      "loss": 2.6613,
+      "step": 5340
+    },
+    {
+      "epoch": 0.23344551772367672,
+      "grad_norm": 2.25,
+      "learning_rate": 8.718067194579573e-05,
+      "loss": 2.6931,
+      "step": 5341
+    },
+    {
+      "epoch": 0.23348922592770663,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.71760791344266e-05,
+      "loss": 1.6553,
+      "step": 5342
+    },
+    {
+      "epoch": 0.23353293413173654,
+      "grad_norm": 2.125,
+      "learning_rate": 8.717148562149076e-05,
+      "loss": 1.8587,
+      "step": 5343
+    },
+    {
+      "epoch": 0.23357664233576642,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.716689140707488e-05,
+      "loss": 2.2146,
+      "step": 5344
+    },
+    {
+      "epoch": 0.23362035053979632,
+      "grad_norm": 2.125,
+      "learning_rate": 8.716229649126566e-05,
+      "loss": 1.5433,
+      "step": 5345
+    },
+    {
+      "epoch": 0.2336640587438262,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.715770087414983e-05,
+      "loss": 1.7082,
+      "step": 5346
+    },
+    {
+      "epoch": 0.2337077669478561,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.715310455581409e-05,
+      "loss": 1.8389,
+      "step": 5347
+    },
+    {
+      "epoch": 0.23375147515188602,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.71485075363452e-05,
+      "loss": 1.7461,
+      "step": 5348
+    },
+    {
+      "epoch": 0.2337951833559159,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.714390981582991e-05,
+      "loss": 1.8022,
+      "step": 5349
+    },
+    {
+      "epoch": 0.2338388915599458,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.713931139435497e-05,
+      "loss": 1.8482,
+      "step": 5350
+    },
+    {
+      "epoch": 0.23388259976397568,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.713471227200719e-05,
+      "loss": 1.8543,
+      "step": 5351
+    },
+    {
+      "epoch": 0.2339263079680056,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.713011244887331e-05,
+      "loss": 1.3542,
+      "step": 5352
+    },
+    {
+      "epoch": 0.2339700161720355,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.712551192504019e-05,
+      "loss": 1.8759,
+      "step": 5353
+    },
+    {
+      "epoch": 0.23401372437606538,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.712091070059463e-05,
+      "loss": 2.6063,
+      "step": 5354
+    },
+    {
+      "epoch": 0.2340574325800953,
+      "grad_norm": 4.8125,
+      "learning_rate": 8.711630877562345e-05,
+      "loss": 2.0466,
+      "step": 5355
+    },
+    {
+      "epoch": 0.23410114078412517,
+      "grad_norm": 3.859375,
+      "learning_rate": 8.71117061502135e-05,
+      "loss": 2.1461,
+      "step": 5356
+    },
+    {
+      "epoch": 0.23414484898815507,
+      "grad_norm": 2.75,
+      "learning_rate": 8.710710282445165e-05,
+      "loss": 2.7146,
+      "step": 5357
+    },
+    {
+      "epoch": 0.23418855719218498,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.710249879842476e-05,
+      "loss": 1.7998,
+      "step": 5358
+    },
+    {
+      "epoch": 0.23423226539621486,
+      "grad_norm": 7.21875,
+      "learning_rate": 8.709789407221971e-05,
+      "loss": 1.5454,
+      "step": 5359
+    },
+    {
+      "epoch": 0.23427597360024477,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.709328864592341e-05,
+      "loss": 2.0275,
+      "step": 5360
+    },
+    {
+      "epoch": 0.23431968180427465,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.708868251962277e-05,
+      "loss": 2.6636,
+      "step": 5361
+    },
+    {
+      "epoch": 0.23436339000830456,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.70840756934047e-05,
+      "loss": 1.9007,
+      "step": 5362
+    },
+    {
+      "epoch": 0.23440709821233446,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.707946816735617e-05,
+      "loss": 2.0047,
+      "step": 5363
+    },
+    {
+      "epoch": 0.23445080641636434,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.70748599415641e-05,
+      "loss": 1.9084,
+      "step": 5364
+    },
+    {
+      "epoch": 0.23449451462039425,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.707025101611545e-05,
+      "loss": 2.4486,
+      "step": 5365
+    },
+    {
+      "epoch": 0.23453822282442413,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.706564139109722e-05,
+      "loss": 1.7819,
+      "step": 5366
+    },
+    {
+      "epoch": 0.23458193102845404,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.706103106659638e-05,
+      "loss": 1.785,
+      "step": 5367
+    },
+    {
+      "epoch": 0.23462563923248395,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.705642004269996e-05,
+      "loss": 2.4596,
+      "step": 5368
+    },
+    {
+      "epoch": 0.23466934743651383,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.705180831949496e-05,
+      "loss": 2.7992,
+      "step": 5369
+    },
+    {
+      "epoch": 0.23471305564054373,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.70471958970684e-05,
+      "loss": 1.4907,
+      "step": 5370
+    },
+    {
+      "epoch": 0.2347567638445736,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.704258277550735e-05,
+      "loss": 1.842,
+      "step": 5371
+    },
+    {
+      "epoch": 0.23480047204860352,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.703796895489883e-05,
+      "loss": 1.6824,
+      "step": 5372
+    },
+    {
+      "epoch": 0.23484418025263343,
+      "grad_norm": 2.75,
+      "learning_rate": 8.703335443532994e-05,
+      "loss": 2.8193,
+      "step": 5373
+    },
+    {
+      "epoch": 0.2348878884566633,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.702873921688776e-05,
+      "loss": 2.1645,
+      "step": 5374
+    },
+    {
+      "epoch": 0.23493159666069321,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.702412329965937e-05,
+      "loss": 2.1561,
+      "step": 5375
+    },
+    {
+      "epoch": 0.23497530486472312,
+      "grad_norm": 1.875,
+      "learning_rate": 8.701950668373189e-05,
+      "loss": 1.7574,
+      "step": 5376
+    },
+    {
+      "epoch": 0.235019013068753,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.701488936919242e-05,
+      "loss": 2.1245,
+      "step": 5377
+    },
+    {
+      "epoch": 0.2350627212727829,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.701027135612814e-05,
+      "loss": 1.968,
+      "step": 5378
+    },
+    {
+      "epoch": 0.2351064294768128,
+      "grad_norm": 2.375,
+      "learning_rate": 8.700565264462617e-05,
+      "loss": 1.6649,
+      "step": 5379
+    },
+    {
+      "epoch": 0.2351501376808427,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.700103323477366e-05,
+      "loss": 1.9941,
+      "step": 5380
+    },
+    {
+      "epoch": 0.2351938458848726,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.699641312665782e-05,
+      "loss": 2.0117,
+      "step": 5381
+    },
+    {
+      "epoch": 0.23523755408890248,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.69917923203658e-05,
+      "loss": 1.7337,
+      "step": 5382
+    },
+    {
+      "epoch": 0.2352812622929324,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.698717081598484e-05,
+      "loss": 1.8028,
+      "step": 5383
+    },
+    {
+      "epoch": 0.23532497049696227,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.698254861360211e-05,
+      "loss": 1.703,
+      "step": 5384
+    },
+    {
+      "epoch": 0.23536867870099218,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.697792571330487e-05,
+      "loss": 2.4212,
+      "step": 5385
+    },
+    {
+      "epoch": 0.2354123869050221,
+      "grad_norm": 2.5,
+      "learning_rate": 8.697330211518038e-05,
+      "loss": 2.0185,
+      "step": 5386
+    },
+    {
+      "epoch": 0.23545609510905197,
+      "grad_norm": 2.875,
+      "learning_rate": 8.696867781931584e-05,
+      "loss": 3.0804,
+      "step": 5387
+    },
+    {
+      "epoch": 0.23549980331308187,
+      "grad_norm": 2.375,
+      "learning_rate": 8.696405282579855e-05,
+      "loss": 1.683,
+      "step": 5388
+    },
+    {
+      "epoch": 0.23554351151711175,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.695942713471577e-05,
+      "loss": 2.1921,
+      "step": 5389
+    },
+    {
+      "epoch": 0.23558721972114166,
+      "grad_norm": 3.125,
+      "learning_rate": 8.695480074615483e-05,
+      "loss": 1.9542,
+      "step": 5390
+    },
+    {
+      "epoch": 0.23563092792517157,
+      "grad_norm": 2.5,
+      "learning_rate": 8.695017366020301e-05,
+      "loss": 1.7134,
+      "step": 5391
+    },
+    {
+      "epoch": 0.23567463612920145,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.694554587694764e-05,
+      "loss": 1.7987,
+      "step": 5392
+    },
+    {
+      "epoch": 0.23571834433323136,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.694091739647602e-05,
+      "loss": 2.0596,
+      "step": 5393
+    },
+    {
+      "epoch": 0.23576205253726124,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.693628821887556e-05,
+      "loss": 2.1124,
+      "step": 5394
+    },
+    {
+      "epoch": 0.23580576074129114,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.693165834423355e-05,
+      "loss": 2.8182,
+      "step": 5395
+    },
+    {
+      "epoch": 0.23584946894532105,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.692702777263742e-05,
+      "loss": 1.3823,
+      "step": 5396
+    },
+    {
+      "epoch": 0.23589317714935093,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.692239650417452e-05,
+      "loss": 1.7818,
+      "step": 5397
+    },
+    {
+      "epoch": 0.23593688535338084,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.691776453893227e-05,
+      "loss": 2.2617,
+      "step": 5398
+    },
+    {
+      "epoch": 0.23598059355741072,
+      "grad_norm": 2.25,
+      "learning_rate": 8.691313187699805e-05,
+      "loss": 2.4695,
+      "step": 5399
+    },
+    {
+      "epoch": 0.23602430176144062,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.690849851845933e-05,
+      "loss": 1.8407,
+      "step": 5400
+    },
+    {
+      "epoch": 0.23606800996547053,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.690386446340353e-05,
+      "loss": 1.6898,
+      "step": 5401
+    },
+    {
+      "epoch": 0.2361117181695004,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.689922971191809e-05,
+      "loss": 1.4087,
+      "step": 5402
+    },
+    {
+      "epoch": 0.23615542637353032,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.689459426409045e-05,
+      "loss": 2.578,
+      "step": 5403
+    },
+    {
+      "epoch": 0.2361991345775602,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.688995812000815e-05,
+      "loss": 1.958,
+      "step": 5404
+    },
+    {
+      "epoch": 0.2362428427815901,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.688532127975865e-05,
+      "loss": 2.3085,
+      "step": 5405
+    },
+    {
+      "epoch": 0.23628655098562001,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.688068374342944e-05,
+      "loss": 1.7854,
+      "step": 5406
+    },
+    {
+      "epoch": 0.2363302591896499,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.687604551110807e-05,
+      "loss": 2.3475,
+      "step": 5407
+    },
+    {
+      "epoch": 0.2363739673936798,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.687140658288202e-05,
+      "loss": 1.7633,
+      "step": 5408
+    },
+    {
+      "epoch": 0.23641767559770968,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.686676695883889e-05,
+      "loss": 2.5959,
+      "step": 5409
+    },
+    {
+      "epoch": 0.2364613838017396,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.68621266390662e-05,
+      "loss": 1.6729,
+      "step": 5410
+    },
+    {
+      "epoch": 0.2365050920057695,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.685748562365153e-05,
+      "loss": 1.8097,
+      "step": 5411
+    },
+    {
+      "epoch": 0.23654880020979938,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.685284391268247e-05,
+      "loss": 1.8138,
+      "step": 5412
+    },
+    {
+      "epoch": 0.23659250841382928,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.684820150624659e-05,
+      "loss": 1.8495,
+      "step": 5413
+    },
+    {
+      "epoch": 0.23663621661785916,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.684355840443155e-05,
+      "loss": 2.0667,
+      "step": 5414
+    },
+    {
+      "epoch": 0.23667992482188907,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.683891460732492e-05,
+      "loss": 1.48,
+      "step": 5415
+    },
+    {
+      "epoch": 0.23672363302591898,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.683427011501434e-05,
+      "loss": 1.7694,
+      "step": 5416
+    },
+    {
+      "epoch": 0.23676734122994886,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.68296249275875e-05,
+      "loss": 1.4667,
+      "step": 5417
+    },
+    {
+      "epoch": 0.23681104943397877,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.682497904513203e-05,
+      "loss": 1.8178,
+      "step": 5418
+    },
+    {
+      "epoch": 0.23685475763800865,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.68203324677356e-05,
+      "loss": 1.4938,
+      "step": 5419
+    },
+    {
+      "epoch": 0.23689846584203855,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.681568519548591e-05,
+      "loss": 2.7004,
+      "step": 5420
+    },
+    {
+      "epoch": 0.23694217404606846,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.681103722847065e-05,
+      "loss": 2.2187,
+      "step": 5421
+    },
+    {
+      "epoch": 0.23698588225009834,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.680638856677754e-05,
+      "loss": 2.0843,
+      "step": 5422
+    },
+    {
+      "epoch": 0.23702959045412825,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.680173921049433e-05,
+      "loss": 1.6284,
+      "step": 5423
+    },
+    {
+      "epoch": 0.23707329865815813,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.679708915970873e-05,
+      "loss": 1.8602,
+      "step": 5424
+    },
+    {
+      "epoch": 0.23711700686218803,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.679243841450848e-05,
+      "loss": 2.6321,
+      "step": 5425
+    },
+    {
+      "epoch": 0.23716071506621794,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.67877869749814e-05,
+      "loss": 2.7493,
+      "step": 5426
+    },
+    {
+      "epoch": 0.23720442327024782,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.678313484121522e-05,
+      "loss": 1.7742,
+      "step": 5427
+    },
+    {
+      "epoch": 0.23724813147427773,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.677848201329774e-05,
+      "loss": 2.1923,
+      "step": 5428
+    },
+    {
+      "epoch": 0.2372918396783076,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.677382849131682e-05,
+      "loss": 1.8479,
+      "step": 5429
+    },
+    {
+      "epoch": 0.23733554788233752,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.676917427536017e-05,
+      "loss": 2.7202,
+      "step": 5430
+    },
+    {
+      "epoch": 0.23737925608636742,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.676451936551573e-05,
+      "loss": 1.9664,
+      "step": 5431
+    },
+    {
+      "epoch": 0.2374229642903973,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.67598637618713e-05,
+      "loss": 1.6204,
+      "step": 5432
+    },
+    {
+      "epoch": 0.2374666724944272,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.675520746451473e-05,
+      "loss": 1.4899,
+      "step": 5433
+    },
+    {
+      "epoch": 0.2375103806984571,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.675055047353391e-05,
+      "loss": 1.8132,
+      "step": 5434
+    },
+    {
+      "epoch": 0.237554088902487,
+      "grad_norm": 3.484375,
+      "learning_rate": 8.67458927890167e-05,
+      "loss": 1.4917,
+      "step": 5435
+    },
+    {
+      "epoch": 0.2375977971065169,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.674123441105102e-05,
+      "loss": 1.318,
+      "step": 5436
+    },
+    {
+      "epoch": 0.23764150531054679,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.673657533972477e-05,
+      "loss": 1.9401,
+      "step": 5437
+    },
+    {
+      "epoch": 0.2376852135145767,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.673191557512588e-05,
+      "loss": 2.696,
+      "step": 5438
+    },
+    {
+      "epoch": 0.23772892171860657,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.672725511734228e-05,
+      "loss": 2.0011,
+      "step": 5439
+    },
+    {
+      "epoch": 0.23777262992263648,
+      "grad_norm": 2.25,
+      "learning_rate": 8.672259396646191e-05,
+      "loss": 1.6778,
+      "step": 5440
+    },
+    {
+      "epoch": 0.2378163381266664,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.671793212257276e-05,
+      "loss": 2.2002,
+      "step": 5441
+    },
+    {
+      "epoch": 0.23786004633069627,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.671326958576279e-05,
+      "loss": 1.8127,
+      "step": 5442
+    },
+    {
+      "epoch": 0.23790375453472618,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.670860635611997e-05,
+      "loss": 2.4345,
+      "step": 5443
+    },
+    {
+      "epoch": 0.23794746273875605,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.670394243373236e-05,
+      "loss": 2.1367,
+      "step": 5444
+    },
+    {
+      "epoch": 0.23799117094278596,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.66992778186879e-05,
+      "loss": 1.5326,
+      "step": 5445
+    },
+    {
+      "epoch": 0.23803487914681587,
+      "grad_norm": 2.0,
+      "learning_rate": 8.669461251107466e-05,
+      "loss": 2.1858,
+      "step": 5446
+    },
+    {
+      "epoch": 0.23807858735084575,
+      "grad_norm": 2.625,
+      "learning_rate": 8.66899465109807e-05,
+      "loss": 2.0247,
+      "step": 5447
+    },
+    {
+      "epoch": 0.23812229555487566,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.668527981849402e-05,
+      "loss": 2.1743,
+      "step": 5448
+    },
+    {
+      "epoch": 0.23816600375890554,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.668061243370274e-05,
+      "loss": 1.9896,
+      "step": 5449
+    },
+    {
+      "epoch": 0.23820971196293544,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.66759443566949e-05,
+      "loss": 2.7154,
+      "step": 5450
+    },
+    {
+      "epoch": 0.23825342016696535,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.667127558755862e-05,
+      "loss": 2.4444,
+      "step": 5451
+    },
+    {
+      "epoch": 0.23829712837099523,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.6666606126382e-05,
+      "loss": 2.4881,
+      "step": 5452
+    },
+    {
+      "epoch": 0.23834083657502514,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.666193597325314e-05,
+      "loss": 1.6791,
+      "step": 5453
+    },
+    {
+      "epoch": 0.23838454477905502,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.66572651282602e-05,
+      "loss": 2.1325,
+      "step": 5454
+    },
+    {
+      "epoch": 0.23842825298308493,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.665259359149132e-05,
+      "loss": 2.0048,
+      "step": 5455
+    },
+    {
+      "epoch": 0.23847196118711483,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.664792136303465e-05,
+      "loss": 2.0438,
+      "step": 5456
+    },
+    {
+      "epoch": 0.2385156693911447,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.664324844297837e-05,
+      "loss": 1.7064,
+      "step": 5457
+    },
+    {
+      "epoch": 0.23855937759517462,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.663857483141064e-05,
+      "loss": 2.4179,
+      "step": 5458
+    },
+    {
+      "epoch": 0.2386030857992045,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.66339005284197e-05,
+      "loss": 1.6632,
+      "step": 5459
+    },
+    {
+      "epoch": 0.2386467940032344,
+      "grad_norm": 3.5,
+      "learning_rate": 8.662922553409373e-05,
+      "loss": 1.39,
+      "step": 5460
+    },
+    {
+      "epoch": 0.23869050220726432,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.662454984852098e-05,
+      "loss": 1.7344,
+      "step": 5461
+    },
+    {
+      "epoch": 0.2387342104112942,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.661987347178964e-05,
+      "loss": 1.8208,
+      "step": 5462
+    },
+    {
+      "epoch": 0.2387779186153241,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.661519640398801e-05,
+      "loss": 1.9327,
+      "step": 5463
+    },
+    {
+      "epoch": 0.23882162681935398,
+      "grad_norm": 2.5,
+      "learning_rate": 8.661051864520432e-05,
+      "loss": 2.1962,
+      "step": 5464
+    },
+    {
+      "epoch": 0.2388653350233839,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.660584019552687e-05,
+      "loss": 1.6932,
+      "step": 5465
+    },
+    {
+      "epoch": 0.2389090432274138,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.660116105504393e-05,
+      "loss": 1.9655,
+      "step": 5466
+    },
+    {
+      "epoch": 0.23895275143144368,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.659648122384382e-05,
+      "loss": 2.2509,
+      "step": 5467
+    },
+    {
+      "epoch": 0.23899645963547358,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.659180070201484e-05,
+      "loss": 2.0024,
+      "step": 5468
+    },
+    {
+      "epoch": 0.23904016783950346,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.658711948964533e-05,
+      "loss": 2.3215,
+      "step": 5469
+    },
+    {
+      "epoch": 0.23908387604353337,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.658243758682361e-05,
+      "loss": 1.5055,
+      "step": 5470
+    },
+    {
+      "epoch": 0.23912758424756328,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.657775499363806e-05,
+      "loss": 1.6403,
+      "step": 5471
+    },
+    {
+      "epoch": 0.23917129245159316,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.657307171017703e-05,
+      "loss": 1.699,
+      "step": 5472
+    },
+    {
+      "epoch": 0.23921500065562307,
+      "grad_norm": 1.875,
+      "learning_rate": 8.656838773652891e-05,
+      "loss": 2.1911,
+      "step": 5473
+    },
+    {
+      "epoch": 0.23925870885965295,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.65637030727821e-05,
+      "loss": 2.0156,
+      "step": 5474
+    },
+    {
+      "epoch": 0.23930241706368285,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.655901771902499e-05,
+      "loss": 1.7335,
+      "step": 5475
+    },
+    {
+      "epoch": 0.23934612526771276,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.655433167534601e-05,
+      "loss": 2.5339,
+      "step": 5476
+    },
+    {
+      "epoch": 0.23938983347174264,
+      "grad_norm": 1.84375,
+      "learning_rate": 8.654964494183358e-05,
+      "loss": 1.634,
+      "step": 5477
+    },
+    {
+      "epoch": 0.23943354167577255,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.654495751857616e-05,
+      "loss": 1.8699,
+      "step": 5478
+    },
+    {
+      "epoch": 0.23947724987980243,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.654026940566222e-05,
+      "loss": 2.0441,
+      "step": 5479
+    },
+    {
+      "epoch": 0.23952095808383234,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.653558060318018e-05,
+      "loss": 2.1452,
+      "step": 5480
+    },
+    {
+      "epoch": 0.23956466628786224,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.653089111121859e-05,
+      "loss": 2.3463,
+      "step": 5481
+    },
+    {
+      "epoch": 0.23960837449189212,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.65262009298659e-05,
+      "loss": 1.9638,
+      "step": 5482
+    },
+    {
+      "epoch": 0.23965208269592203,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.652151005921064e-05,
+      "loss": 2.3107,
+      "step": 5483
+    },
+    {
+      "epoch": 0.2396957908999519,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.651681849934134e-05,
+      "loss": 1.8512,
+      "step": 5484
+    },
+    {
+      "epoch": 0.23973949910398182,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.651212625034651e-05,
+      "loss": 1.7393,
+      "step": 5485
+    },
+    {
+      "epoch": 0.23978320730801173,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.650743331231472e-05,
+      "loss": 1.7654,
+      "step": 5486
+    },
+    {
+      "epoch": 0.2398269155120416,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.650273968533454e-05,
+      "loss": 2.1793,
+      "step": 5487
+    },
+    {
+      "epoch": 0.2398706237160715,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.649804536949454e-05,
+      "loss": 1.9098,
+      "step": 5488
+    },
+    {
+      "epoch": 0.2399143319201014,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.649335036488329e-05,
+      "loss": 2.5594,
+      "step": 5489
+    },
+    {
+      "epoch": 0.2399580401241313,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.64886546715894e-05,
+      "loss": 2.2916,
+      "step": 5490
+    },
+    {
+      "epoch": 0.2400017483281612,
+      "grad_norm": 2.625,
+      "learning_rate": 8.64839582897015e-05,
+      "loss": 2.9065,
+      "step": 5491
+    },
+    {
+      "epoch": 0.2400454565321911,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.647926121930821e-05,
+      "loss": 2.2818,
+      "step": 5492
+    },
+    {
+      "epoch": 0.240089164736221,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.647456346049816e-05,
+      "loss": 1.6659,
+      "step": 5493
+    },
+    {
+      "epoch": 0.24013287294025087,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.646986501336001e-05,
+      "loss": 2.4522,
+      "step": 5494
+    },
+    {
+      "epoch": 0.24017658114428078,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.646516587798243e-05,
+      "loss": 2.2336,
+      "step": 5495
+    },
+    {
+      "epoch": 0.2402202893483107,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.64604660544541e-05,
+      "loss": 1.8014,
+      "step": 5496
+    },
+    {
+      "epoch": 0.24026399755234057,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.645576554286372e-05,
+      "loss": 1.5453,
+      "step": 5497
+    },
+    {
+      "epoch": 0.24030770575637048,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.645106434329996e-05,
+      "loss": 2.1022,
+      "step": 5498
+    },
+    {
+      "epoch": 0.24035141396040036,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.644636245585159e-05,
+      "loss": 1.9023,
+      "step": 5499
+    },
+    {
+      "epoch": 0.24039512216443026,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.64416598806073e-05,
+      "loss": 1.5248,
+      "step": 5500
+    },
+    {
+      "epoch": 0.24043883036846017,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.643695661765585e-05,
+      "loss": 1.7943,
+      "step": 5501
+    },
+    {
+      "epoch": 0.24048253857249005,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.643225266708601e-05,
+      "loss": 2.9711,
+      "step": 5502
+    },
+    {
+      "epoch": 0.24052624677651996,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.642754802898652e-05,
+      "loss": 1.6168,
+      "step": 5503
+    },
+    {
+      "epoch": 0.24056995498054984,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.64228427034462e-05,
+      "loss": 1.6666,
+      "step": 5504
+    },
+    {
+      "epoch": 0.24061366318457975,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.641813669055381e-05,
+      "loss": 2.1982,
+      "step": 5505
+    },
+    {
+      "epoch": 0.24065737138860965,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.641342999039819e-05,
+      "loss": 1.7612,
+      "step": 5506
+    },
+    {
+      "epoch": 0.24070107959263953,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.640872260306814e-05,
+      "loss": 1.2477,
+      "step": 5507
+    },
+    {
+      "epoch": 0.24074478779666944,
+      "grad_norm": 1.8359375,
+      "learning_rate": 8.64040145286525e-05,
+      "loss": 1.8345,
+      "step": 5508
+    },
+    {
+      "epoch": 0.24078849600069932,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.639930576724014e-05,
+      "loss": 1.8334,
+      "step": 5509
+    },
+    {
+      "epoch": 0.24083220420472923,
+      "grad_norm": 2.0,
+      "learning_rate": 8.639459631891989e-05,
+      "loss": 2.0725,
+      "step": 5510
+    },
+    {
+      "epoch": 0.24087591240875914,
+      "grad_norm": 2.0,
+      "learning_rate": 8.638988618378063e-05,
+      "loss": 2.0752,
+      "step": 5511
+    },
+    {
+      "epoch": 0.24091962061278901,
+      "grad_norm": 2.625,
+      "learning_rate": 8.638517536191127e-05,
+      "loss": 2.3952,
+      "step": 5512
+    },
+    {
+      "epoch": 0.24096332881681892,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.638046385340069e-05,
+      "loss": 1.9946,
+      "step": 5513
+    },
+    {
+      "epoch": 0.2410070370208488,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.63757516583378e-05,
+      "loss": 1.7343,
+      "step": 5514
+    },
+    {
+      "epoch": 0.2410507452248787,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.637103877681155e-05,
+      "loss": 2.0783,
+      "step": 5515
+    },
+    {
+      "epoch": 0.24109445342890862,
+      "grad_norm": 5.4375,
+      "learning_rate": 8.636632520891085e-05,
+      "loss": 2.0024,
+      "step": 5516
+    },
+    {
+      "epoch": 0.2411381616329385,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.636161095472466e-05,
+      "loss": 1.9169,
+      "step": 5517
+    },
+    {
+      "epoch": 0.2411818698369684,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.635689601434196e-05,
+      "loss": 2.361,
+      "step": 5518
+    },
+    {
+      "epoch": 0.24122557804099828,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.635218038785171e-05,
+      "loss": 2.0534,
+      "step": 5519
+    },
+    {
+      "epoch": 0.2412692862450282,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.634746407534292e-05,
+      "loss": 1.9767,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2413129944490581,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.634274707690458e-05,
+      "loss": 1.6981,
+      "step": 5521
+    },
+    {
+      "epoch": 0.24135670265308798,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.63380293926257e-05,
+      "loss": 1.6489,
+      "step": 5522
+    },
+    {
+      "epoch": 0.2414004108571179,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.633331102259532e-05,
+      "loss": 1.6282,
+      "step": 5523
+    },
+    {
+      "epoch": 0.24144411906114777,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.632859196690249e-05,
+      "loss": 1.904,
+      "step": 5524
+    },
+    {
+      "epoch": 0.24148782726517767,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.632387222563622e-05,
+      "loss": 2.4687,
+      "step": 5525
+    },
+    {
+      "epoch": 0.24153153546920758,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.631915179888566e-05,
+      "loss": 2.9449,
+      "step": 5526
+    },
+    {
+      "epoch": 0.24157524367323746,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.631443068673983e-05,
+      "loss": 1.87,
+      "step": 5527
+    },
+    {
+      "epoch": 0.24161895187726737,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.630970888928784e-05,
+      "loss": 1.7336,
+      "step": 5528
+    },
+    {
+      "epoch": 0.24166266008129725,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.630498640661879e-05,
+      "loss": 2.7522,
+      "step": 5529
+    },
+    {
+      "epoch": 0.24170636828532716,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.630026323882181e-05,
+      "loss": 1.2049,
+      "step": 5530
+    },
+    {
+      "epoch": 0.24175007648935706,
+      "grad_norm": 2.25,
+      "learning_rate": 8.629553938598603e-05,
+      "loss": 2.6773,
+      "step": 5531
+    },
+    {
+      "epoch": 0.24179378469338694,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.629081484820058e-05,
+      "loss": 1.9445,
+      "step": 5532
+    },
+    {
+      "epoch": 0.24183749289741685,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.628608962555467e-05,
+      "loss": 1.8289,
+      "step": 5533
+    },
+    {
+      "epoch": 0.24188120110144673,
+      "grad_norm": 1.8359375,
+      "learning_rate": 8.62813637181374e-05,
+      "loss": 1.6484,
+      "step": 5534
+    },
+    {
+      "epoch": 0.24192490930547664,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.627663712603802e-05,
+      "loss": 1.7932,
+      "step": 5535
+    },
+    {
+      "epoch": 0.24196861750950655,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.627190984934569e-05,
+      "loss": 2.0733,
+      "step": 5536
+    },
+    {
+      "epoch": 0.24201232571353642,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.626718188814964e-05,
+      "loss": 2.6501,
+      "step": 5537
+    },
+    {
+      "epoch": 0.24205603391756633,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.626245324253906e-05,
+      "loss": 1.7905,
+      "step": 5538
+    },
+    {
+      "epoch": 0.2420997421215962,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.625772391260323e-05,
+      "loss": 1.7842,
+      "step": 5539
+    },
+    {
+      "epoch": 0.24214345032562612,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.625299389843137e-05,
+      "loss": 3.0661,
+      "step": 5540
+    },
+    {
+      "epoch": 0.24218715852965603,
+      "grad_norm": 2.25,
+      "learning_rate": 8.624826320011275e-05,
+      "loss": 2.2763,
+      "step": 5541
+    },
+    {
+      "epoch": 0.2422308667336859,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.624353181773664e-05,
+      "loss": 1.3571,
+      "step": 5542
+    },
+    {
+      "epoch": 0.24227457493771581,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.623879975139235e-05,
+      "loss": 1.9099,
+      "step": 5543
+    },
+    {
+      "epoch": 0.2423182831417457,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.623406700116917e-05,
+      "loss": 2.2006,
+      "step": 5544
+    },
+    {
+      "epoch": 0.2423619913457756,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.622933356715639e-05,
+      "loss": 1.6884,
+      "step": 5545
+    },
+    {
+      "epoch": 0.2424056995498055,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.622459944944336e-05,
+      "loss": 2.199,
+      "step": 5546
+    },
+    {
+      "epoch": 0.2424494077538354,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.621986464811943e-05,
+      "loss": 1.8362,
+      "step": 5547
+    },
+    {
+      "epoch": 0.2424931159578653,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.621512916327394e-05,
+      "loss": 2.5015,
+      "step": 5548
+    },
+    {
+      "epoch": 0.24253682416189518,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.621039299499624e-05,
+      "loss": 2.1437,
+      "step": 5549
+    },
+    {
+      "epoch": 0.24258053236592508,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.620565614337572e-05,
+      "loss": 1.5535,
+      "step": 5550
+    },
+    {
+      "epoch": 0.242624240569955,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.620091860850177e-05,
+      "loss": 1.5493,
+      "step": 5551
+    },
+    {
+      "epoch": 0.24266794877398487,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.619618039046382e-05,
+      "loss": 1.6928,
+      "step": 5552
+    },
+    {
+      "epoch": 0.24271165697801478,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.619144148935125e-05,
+      "loss": 2.0614,
+      "step": 5553
+    },
+    {
+      "epoch": 0.24275536518204466,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.618670190525352e-05,
+      "loss": 1.4767,
+      "step": 5554
+    },
+    {
+      "epoch": 0.24279907338607457,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.618196163826005e-05,
+      "loss": 3.169,
+      "step": 5555
+    },
+    {
+      "epoch": 0.24284278159010447,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.617722068846028e-05,
+      "loss": 1.8736,
+      "step": 5556
+    },
+    {
+      "epoch": 0.24288648979413435,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.617247905594372e-05,
+      "loss": 2.5414,
+      "step": 5557
+    },
+    {
+      "epoch": 0.24293019799816426,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.616773674079985e-05,
+      "loss": 2.1125,
+      "step": 5558
+    },
+    {
+      "epoch": 0.24297390620219414,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.616299374311814e-05,
+      "loss": 2.6338,
+      "step": 5559
+    },
+    {
+      "epoch": 0.24301761440622405,
+      "grad_norm": 2.125,
+      "learning_rate": 8.615825006298808e-05,
+      "loss": 1.9468,
+      "step": 5560
+    },
+    {
+      "epoch": 0.24306132261025395,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.615350570049924e-05,
+      "loss": 1.7916,
+      "step": 5561
+    },
+    {
+      "epoch": 0.24310503081428383,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.614876065574113e-05,
+      "loss": 1.6609,
+      "step": 5562
+    },
+    {
+      "epoch": 0.24314873901831374,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.614401492880329e-05,
+      "loss": 2.25,
+      "step": 5563
+    },
+    {
+      "epoch": 0.24319244722234362,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.613926851977527e-05,
+      "loss": 1.9206,
+      "step": 5564
+    },
+    {
+      "epoch": 0.24323615542637353,
+      "grad_norm": 2.625,
+      "learning_rate": 8.613452142874667e-05,
+      "loss": 2.1226,
+      "step": 5565
+    },
+    {
+      "epoch": 0.24327986363040344,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.612977365580705e-05,
+      "loss": 1.3526,
+      "step": 5566
+    },
+    {
+      "epoch": 0.24332357183443332,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.612502520104602e-05,
+      "loss": 2.4662,
+      "step": 5567
+    },
+    {
+      "epoch": 0.24336728003846322,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.61202760645532e-05,
+      "loss": 1.9605,
+      "step": 5568
+    },
+    {
+      "epoch": 0.2434109882424931,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.611552624641818e-05,
+      "loss": 1.9366,
+      "step": 5569
+    },
+    {
+      "epoch": 0.243454696446523,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.611077574673063e-05,
+      "loss": 2.3338,
+      "step": 5570
+    },
+    {
+      "epoch": 0.24349840465055292,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.610602456558018e-05,
+      "loss": 2.3373,
+      "step": 5571
+    },
+    {
+      "epoch": 0.2435421128545828,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.61012727030565e-05,
+      "loss": 2.1172,
+      "step": 5572
+    },
+    {
+      "epoch": 0.2435858210586127,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.609652015924925e-05,
+      "loss": 1.8434,
+      "step": 5573
+    },
+    {
+      "epoch": 0.24362952926264259,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.609176693424814e-05,
+      "loss": 2.0251,
+      "step": 5574
+    },
+    {
+      "epoch": 0.2436732374666725,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.608701302814286e-05,
+      "loss": 1.6486,
+      "step": 5575
+    },
+    {
+      "epoch": 0.2437169456707024,
+      "grad_norm": 3.765625,
+      "learning_rate": 8.60822584410231e-05,
+      "loss": 1.7621,
+      "step": 5576
+    },
+    {
+      "epoch": 0.24376065387473228,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.607750317297863e-05,
+      "loss": 1.6235,
+      "step": 5577
+    },
+    {
+      "epoch": 0.2438043620787622,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.607274722409918e-05,
+      "loss": 2.2202,
+      "step": 5578
+    },
+    {
+      "epoch": 0.24384807028279207,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.606799059447445e-05,
+      "loss": 1.8622,
+      "step": 5579
+    },
+    {
+      "epoch": 0.24389177848682198,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.606323328419428e-05,
+      "loss": 1.903,
+      "step": 5580
+    },
+    {
+      "epoch": 0.24393548669085188,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.605847529334839e-05,
+      "loss": 2.0046,
+      "step": 5581
+    },
+    {
+      "epoch": 0.24397919489488176,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.60537166220266e-05,
+      "loss": 1.0194,
+      "step": 5582
+    },
+    {
+      "epoch": 0.24402290309891167,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.604895727031869e-05,
+      "loss": 2.1291,
+      "step": 5583
+    },
+    {
+      "epoch": 0.24406661130294155,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.60441972383145e-05,
+      "loss": 1.747,
+      "step": 5584
+    },
+    {
+      "epoch": 0.24411031950697146,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.603943652610385e-05,
+      "loss": 2.0577,
+      "step": 5585
+    },
+    {
+      "epoch": 0.24415402771100136,
+      "grad_norm": 3.71875,
+      "learning_rate": 8.603467513377657e-05,
+      "loss": 1.9055,
+      "step": 5586
+    },
+    {
+      "epoch": 0.24419773591503124,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.602991306142252e-05,
+      "loss": 2.5746,
+      "step": 5587
+    },
+    {
+      "epoch": 0.24424144411906115,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.602515030913159e-05,
+      "loss": 2.2653,
+      "step": 5588
+    },
+    {
+      "epoch": 0.24428515232309103,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.602038687699364e-05,
+      "loss": 2.5107,
+      "step": 5589
+    },
+    {
+      "epoch": 0.24432886052712094,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.601562276509855e-05,
+      "loss": 1.8985,
+      "step": 5590
+    },
+    {
+      "epoch": 0.24437256873115085,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.601085797353625e-05,
+      "loss": 1.976,
+      "step": 5591
+    },
+    {
+      "epoch": 0.24441627693518073,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.600609250239664e-05,
+      "loss": 1.7231,
+      "step": 5592
+    },
+    {
+      "epoch": 0.24445998513921063,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.600132635176967e-05,
+      "loss": 1.9487,
+      "step": 5593
+    },
+    {
+      "epoch": 0.2445036933432405,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.599655952174527e-05,
+      "loss": 2.0027,
+      "step": 5594
+    },
+    {
+      "epoch": 0.24454740154727042,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.59917920124134e-05,
+      "loss": 1.7015,
+      "step": 5595
+    },
+    {
+      "epoch": 0.24459110975130033,
+      "grad_norm": 5.15625,
+      "learning_rate": 8.598702382386403e-05,
+      "loss": 2.4159,
+      "step": 5596
+    },
+    {
+      "epoch": 0.2446348179553302,
+      "grad_norm": 1.8359375,
+      "learning_rate": 8.598225495618715e-05,
+      "loss": 1.7125,
+      "step": 5597
+    },
+    {
+      "epoch": 0.24467852615936012,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.597748540947274e-05,
+      "loss": 1.8233,
+      "step": 5598
+    },
+    {
+      "epoch": 0.24472223436339,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.597271518381082e-05,
+      "loss": 1.7276,
+      "step": 5599
+    },
+    {
+      "epoch": 0.2447659425674199,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.596794427929142e-05,
+      "loss": 2.5536,
+      "step": 5600
+    },
+    {
+      "epoch": 0.2448096507714498,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.596317269600455e-05,
+      "loss": 2.0302,
+      "step": 5601
+    },
+    {
+      "epoch": 0.2448533589754797,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.595840043404028e-05,
+      "loss": 1.6446,
+      "step": 5602
+    },
+    {
+      "epoch": 0.2448970671795096,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.595362749348866e-05,
+      "loss": 2.3099,
+      "step": 5603
+    },
+    {
+      "epoch": 0.24494077538353948,
+      "grad_norm": 2.25,
+      "learning_rate": 8.594885387443974e-05,
+      "loss": 2.1294,
+      "step": 5604
+    },
+    {
+      "epoch": 0.24498448358756938,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.594407957698365e-05,
+      "loss": 1.8185,
+      "step": 5605
+    },
+    {
+      "epoch": 0.2450281917915993,
+      "grad_norm": 2.0,
+      "learning_rate": 8.593930460121044e-05,
+      "loss": 2.2778,
+      "step": 5606
+    },
+    {
+      "epoch": 0.24507189999562917,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.593452894721027e-05,
+      "loss": 1.0981,
+      "step": 5607
+    },
+    {
+      "epoch": 0.24511560819965908,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.592975261507323e-05,
+      "loss": 1.7008,
+      "step": 5608
+    },
+    {
+      "epoch": 0.24515931640368896,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.592497560488946e-05,
+      "loss": 1.9112,
+      "step": 5609
+    },
+    {
+      "epoch": 0.24520302460771887,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.592019791674913e-05,
+      "loss": 1.4231,
+      "step": 5610
+    },
+    {
+      "epoch": 0.24524673281174877,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.591541955074235e-05,
+      "loss": 2.1189,
+      "step": 5611
+    },
+    {
+      "epoch": 0.24529044101577865,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.591064050695936e-05,
+      "loss": 2.0184,
+      "step": 5612
+    },
+    {
+      "epoch": 0.24533414921980856,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.590586078549032e-05,
+      "loss": 2.7024,
+      "step": 5613
+    },
+    {
+      "epoch": 0.24537785742383844,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.590108038642541e-05,
+      "loss": 1.9898,
+      "step": 5614
+    },
+    {
+      "epoch": 0.24542156562786835,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.589629930985487e-05,
+      "loss": 1.8514,
+      "step": 5615
+    },
+    {
+      "epoch": 0.24546527383189826,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.589151755586892e-05,
+      "loss": 1.9361,
+      "step": 5616
+    },
+    {
+      "epoch": 0.24550898203592814,
+      "grad_norm": 2.0,
+      "learning_rate": 8.588673512455781e-05,
+      "loss": 1.7724,
+      "step": 5617
+    },
+    {
+      "epoch": 0.24555269023995804,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.588195201601177e-05,
+      "loss": 1.6943,
+      "step": 5618
+    },
+    {
+      "epoch": 0.24559639844398792,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.587716823032106e-05,
+      "loss": 1.6803,
+      "step": 5619
+    },
+    {
+      "epoch": 0.24564010664801783,
+      "grad_norm": 2.0,
+      "learning_rate": 8.587238376757597e-05,
+      "loss": 2.5287,
+      "step": 5620
+    },
+    {
+      "epoch": 0.24568381485204774,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.58675986278668e-05,
+      "loss": 2.0384,
+      "step": 5621
+    },
+    {
+      "epoch": 0.24572752305607762,
+      "grad_norm": 2.375,
+      "learning_rate": 8.586281281128383e-05,
+      "loss": 1.9584,
+      "step": 5622
+    },
+    {
+      "epoch": 0.24577123126010753,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.585802631791739e-05,
+      "loss": 2.0015,
+      "step": 5623
+    },
+    {
+      "epoch": 0.2458149394641374,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.58532391478578e-05,
+      "loss": 2.0059,
+      "step": 5624
+    },
+    {
+      "epoch": 0.2458586476681673,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.584845130119541e-05,
+      "loss": 1.6853,
+      "step": 5625
+    },
+    {
+      "epoch": 0.24590235587219722,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.584366277802057e-05,
+      "loss": 1.7735,
+      "step": 5626
+    },
+    {
+      "epoch": 0.2459460640762271,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.583887357842364e-05,
+      "loss": 2.4116,
+      "step": 5627
+    },
+    {
+      "epoch": 0.245989772280257,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.583408370249501e-05,
+      "loss": 1.9734,
+      "step": 5628
+    },
+    {
+      "epoch": 0.2460334804842869,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.582929315032507e-05,
+      "loss": 1.5228,
+      "step": 5629
+    },
+    {
+      "epoch": 0.2460771886883168,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.582450192200421e-05,
+      "loss": 1.5607,
+      "step": 5630
+    },
+    {
+      "epoch": 0.2461208968923467,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.581971001762286e-05,
+      "loss": 2.2077,
+      "step": 5631
+    },
+    {
+      "epoch": 0.24616460509637658,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.581491743727146e-05,
+      "loss": 1.8515,
+      "step": 5632
+    },
+    {
+      "epoch": 0.2462083133004065,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.581012418104044e-05,
+      "loss": 1.7113,
+      "step": 5633
+    },
+    {
+      "epoch": 0.24625202150443637,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.580533024902024e-05,
+      "loss": 1.9744,
+      "step": 5634
+    },
+    {
+      "epoch": 0.24629572970846628,
+      "grad_norm": 2.375,
+      "learning_rate": 8.580053564130137e-05,
+      "loss": 2.5002,
+      "step": 5635
+    },
+    {
+      "epoch": 0.24633943791249618,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.579574035797427e-05,
+      "loss": 2.1414,
+      "step": 5636
+    },
+    {
+      "epoch": 0.24638314611652606,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.579094439912946e-05,
+      "loss": 1.9267,
+      "step": 5637
+    },
+    {
+      "epoch": 0.24642685432055597,
+      "grad_norm": 2.25,
+      "learning_rate": 8.578614776485743e-05,
+      "loss": 1.9097,
+      "step": 5638
+    },
+    {
+      "epoch": 0.24647056252458585,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.578135045524872e-05,
+      "loss": 2.1637,
+      "step": 5639
+    },
+    {
+      "epoch": 0.24651427072861576,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.577655247039384e-05,
+      "loss": 2.1543,
+      "step": 5640
+    },
+    {
+      "epoch": 0.24655797893264567,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.577175381038337e-05,
+      "loss": 2.2151,
+      "step": 5641
+    },
+    {
+      "epoch": 0.24660168713667555,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.576695447530781e-05,
+      "loss": 1.4585,
+      "step": 5642
+    },
+    {
+      "epoch": 0.24664539534070545,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.576215446525776e-05,
+      "loss": 2.1276,
+      "step": 5643
+    },
+    {
+      "epoch": 0.24668910354473533,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.575735378032382e-05,
+      "loss": 1.7072,
+      "step": 5644
+    },
+    {
+      "epoch": 0.24673281174876524,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.575255242059656e-05,
+      "loss": 1.6291,
+      "step": 5645
+    },
+    {
+      "epoch": 0.24677651995279515,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.574775038616662e-05,
+      "loss": 2.2993,
+      "step": 5646
+    },
+    {
+      "epoch": 0.24682022815682503,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.574294767712458e-05,
+      "loss": 1.3124,
+      "step": 5647
+    },
+    {
+      "epoch": 0.24686393636085494,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.573814429356113e-05,
+      "loss": 1.6663,
+      "step": 5648
+    },
+    {
+      "epoch": 0.24690764456488482,
+      "grad_norm": 2.125,
+      "learning_rate": 8.573334023556685e-05,
+      "loss": 1.9524,
+      "step": 5649
+    },
+    {
+      "epoch": 0.24695135276891472,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.572853550323245e-05,
+      "loss": 1.5047,
+      "step": 5650
+    },
+    {
+      "epoch": 0.24699506097294463,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.572373009664858e-05,
+      "loss": 2.7538,
+      "step": 5651
+    },
+    {
+      "epoch": 0.2470387691769745,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.571892401590591e-05,
+      "loss": 1.896,
+      "step": 5652
+    },
+    {
+      "epoch": 0.24708247738100442,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.571411726109519e-05,
+      "loss": 1.6372,
+      "step": 5653
+    },
+    {
+      "epoch": 0.24712618558503432,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.570930983230707e-05,
+      "loss": 1.9993,
+      "step": 5654
+    },
+    {
+      "epoch": 0.2471698937890642,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.570450172963232e-05,
+      "loss": 1.7876,
+      "step": 5655
+    },
+    {
+      "epoch": 0.2472136019930941,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.569969295316164e-05,
+      "loss": 1.8467,
+      "step": 5656
+    },
+    {
+      "epoch": 0.247257310197124,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.569488350298583e-05,
+      "loss": 2.269,
+      "step": 5657
+    },
+    {
+      "epoch": 0.2473010184011539,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.569007337919558e-05,
+      "loss": 1.9019,
+      "step": 5658
+    },
+    {
+      "epoch": 0.2473447266051838,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.568526258188172e-05,
+      "loss": 1.8272,
+      "step": 5659
+    },
+    {
+      "epoch": 0.2473884348092137,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.568045111113501e-05,
+      "loss": 1.8297,
+      "step": 5660
+    },
+    {
+      "epoch": 0.2474321430132436,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.567563896704626e-05,
+      "loss": 2.0034,
+      "step": 5661
+    },
+    {
+      "epoch": 0.24747585121727347,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.567082614970627e-05,
+      "loss": 1.7031,
+      "step": 5662
+    },
+    {
+      "epoch": 0.24751955942130338,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.566601265920588e-05,
+      "loss": 1.8154,
+      "step": 5663
+    },
+    {
+      "epoch": 0.2475632676253333,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.566119849563594e-05,
+      "loss": 1.7957,
+      "step": 5664
+    },
+    {
+      "epoch": 0.24760697582936317,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.565638365908726e-05,
+      "loss": 2.5451,
+      "step": 5665
+    },
+    {
+      "epoch": 0.24765068403339308,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.565156814965074e-05,
+      "loss": 1.937,
+      "step": 5666
+    },
+    {
+      "epoch": 0.24769439223742296,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.564675196741722e-05,
+      "loss": 1.8272,
+      "step": 5667
+    },
+    {
+      "epoch": 0.24773810044145286,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.564193511247762e-05,
+      "loss": 1.7934,
+      "step": 5668
+    },
+    {
+      "epoch": 0.24778180864548277,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.563711758492283e-05,
+      "loss": 2.6983,
+      "step": 5669
+    },
+    {
+      "epoch": 0.24782551684951265,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.563229938484376e-05,
+      "loss": 1.7599,
+      "step": 5670
+    },
+    {
+      "epoch": 0.24786922505354256,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.562748051233135e-05,
+      "loss": 1.7215,
+      "step": 5671
+    },
+    {
+      "epoch": 0.24791293325757244,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.562266096747653e-05,
+      "loss": 2.0875,
+      "step": 5672
+    },
+    {
+      "epoch": 0.24795664146160235,
+      "grad_norm": 2.625,
+      "learning_rate": 8.561784075037023e-05,
+      "loss": 2.0816,
+      "step": 5673
+    },
+    {
+      "epoch": 0.24800034966563225,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.561301986110347e-05,
+      "loss": 1.7168,
+      "step": 5674
+    },
+    {
+      "epoch": 0.24804405786966213,
+      "grad_norm": 1.890625,
+      "learning_rate": 8.560819829976716e-05,
+      "loss": 1.9503,
+      "step": 5675
+    },
+    {
+      "epoch": 0.24808776607369204,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.560337606645234e-05,
+      "loss": 2.4694,
+      "step": 5676
+    },
+    {
+      "epoch": 0.24813147427772192,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.559855316124999e-05,
+      "loss": 2.046,
+      "step": 5677
+    },
+    {
+      "epoch": 0.24817518248175183,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.559372958425113e-05,
+      "loss": 2.6653,
+      "step": 5678
+    },
+    {
+      "epoch": 0.24821889068578173,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.55889053355468e-05,
+      "loss": 1.7663,
+      "step": 5679
+    },
+    {
+      "epoch": 0.24826259888981161,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.558408041522801e-05,
+      "loss": 2.7306,
+      "step": 5680
+    },
+    {
+      "epoch": 0.24830630709384152,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.557925482338585e-05,
+      "loss": 1.6388,
+      "step": 5681
+    },
+    {
+      "epoch": 0.2483500152978714,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.557442856011135e-05,
+      "loss": 1.9075,
+      "step": 5682
+    },
+    {
+      "epoch": 0.2483937235019013,
+      "grad_norm": 2.0,
+      "learning_rate": 8.556960162549563e-05,
+      "loss": 1.7638,
+      "step": 5683
+    },
+    {
+      "epoch": 0.24843743170593122,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.556477401962974e-05,
+      "loss": 1.8392,
+      "step": 5684
+    },
+    {
+      "epoch": 0.2484811399099611,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.55599457426048e-05,
+      "loss": 1.5737,
+      "step": 5685
+    },
+    {
+      "epoch": 0.248524848113991,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.555511679451196e-05,
+      "loss": 2.7735,
+      "step": 5686
+    },
+    {
+      "epoch": 0.24856855631802088,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.555028717544227e-05,
+      "loss": 1.7439,
+      "step": 5687
+    },
+    {
+      "epoch": 0.2486122645220508,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.554545688548696e-05,
+      "loss": 2.0171,
+      "step": 5688
+    },
+    {
+      "epoch": 0.2486559727260807,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.554062592473713e-05,
+      "loss": 2.0584,
+      "step": 5689
+    },
+    {
+      "epoch": 0.24869968093011058,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.553579429328395e-05,
+      "loss": 2.2466,
+      "step": 5690
+    },
+    {
+      "epoch": 0.24874338913414049,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.553096199121863e-05,
+      "loss": 2.3407,
+      "step": 5691
+    },
+    {
+      "epoch": 0.24878709733817037,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.552612901863233e-05,
+      "loss": 2.0368,
+      "step": 5692
+    },
+    {
+      "epoch": 0.24883080554220027,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.552129537561628e-05,
+      "loss": 1.7908,
+      "step": 5693
+    },
+    {
+      "epoch": 0.24887451374623018,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.551646106226169e-05,
+      "loss": 1.6307,
+      "step": 5694
+    },
+    {
+      "epoch": 0.24891822195026006,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.55116260786598e-05,
+      "loss": 1.781,
+      "step": 5695
+    },
+    {
+      "epoch": 0.24896193015428997,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.550679042490182e-05,
+      "loss": 2.1163,
+      "step": 5696
+    },
+    {
+      "epoch": 0.24900563835831985,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.550195410107902e-05,
+      "loss": 1.4579,
+      "step": 5697
+    },
+    {
+      "epoch": 0.24904934656234975,
+      "grad_norm": 2.0,
+      "learning_rate": 8.549711710728269e-05,
+      "loss": 1.4266,
+      "step": 5698
+    },
+    {
+      "epoch": 0.24909305476637966,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.54922794436041e-05,
+      "loss": 1.961,
+      "step": 5699
+    },
+    {
+      "epoch": 0.24913676297040954,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.548744111013452e-05,
+      "loss": 2.7949,
+      "step": 5700
+    },
+    {
+      "epoch": 0.24918047117443945,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.54826021069653e-05,
+      "loss": 2.0042,
+      "step": 5701
+    },
+    {
+      "epoch": 0.24922417937846933,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.547776243418772e-05,
+      "loss": 1.7539,
+      "step": 5702
+    },
+    {
+      "epoch": 0.24926788758249924,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.547292209189314e-05,
+      "loss": 2.0249,
+      "step": 5703
+    },
+    {
+      "epoch": 0.24931159578652914,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.546808108017288e-05,
+      "loss": 2.0008,
+      "step": 5704
+    },
+    {
+      "epoch": 0.24935530399055902,
+      "grad_norm": 2.5,
+      "learning_rate": 8.54632393991183e-05,
+      "loss": 2.4508,
+      "step": 5705
+    },
+    {
+      "epoch": 0.24939901219458893,
+      "grad_norm": 2.25,
+      "learning_rate": 8.545839704882082e-05,
+      "loss": 2.152,
+      "step": 5706
+    },
+    {
+      "epoch": 0.2494427203986188,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.545355402937175e-05,
+      "loss": 2.0039,
+      "step": 5707
+    },
+    {
+      "epoch": 0.24948642860264872,
+      "grad_norm": 3.46875,
+      "learning_rate": 8.54487103408625e-05,
+      "loss": 2.6012,
+      "step": 5708
+    },
+    {
+      "epoch": 0.24953013680667863,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.544386598338452e-05,
+      "loss": 1.8128,
+      "step": 5709
+    },
+    {
+      "epoch": 0.2495738450107085,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.54390209570292e-05,
+      "loss": 2.1566,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2496175532147384,
+      "grad_norm": 2.125,
+      "learning_rate": 8.543417526188797e-05,
+      "loss": 1.9152,
+      "step": 5711
+    },
+    {
+      "epoch": 0.2496612614187683,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.542932889805226e-05,
+      "loss": 1.7658,
+      "step": 5712
+    },
+    {
+      "epoch": 0.2497049696227982,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.542448186561359e-05,
+      "loss": 2.0908,
+      "step": 5713
+    },
+    {
+      "epoch": 0.2497486778268281,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.541963416466336e-05,
+      "loss": 1.8004,
+      "step": 5714
+    },
+    {
+      "epoch": 0.249792386030858,
+      "grad_norm": 1.8359375,
+      "learning_rate": 8.541478579529308e-05,
+      "loss": 1.5534,
+      "step": 5715
+    },
+    {
+      "epoch": 0.2498360942348879,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.540993675759427e-05,
+      "loss": 1.7388,
+      "step": 5716
+    },
+    {
+      "epoch": 0.24987980243891778,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.540508705165839e-05,
+      "loss": 1.371,
+      "step": 5717
+    },
+    {
+      "epoch": 0.24992351064294768,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.540023667757702e-05,
+      "loss": 1.7817,
+      "step": 5718
+    },
+    {
+      "epoch": 0.2499672188469776,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.539538563544163e-05,
+      "loss": 1.8297,
+      "step": 5719
+    },
+    {
+      "epoch": 0.25001092705100747,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.539053392534382e-05,
+      "loss": 2.3024,
+      "step": 5720
+    },
+    {
+      "epoch": 0.25005463525503735,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.538568154737512e-05,
+      "loss": 1.5528,
+      "step": 5721
+    },
+    {
+      "epoch": 0.2500983434590673,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.53808285016271e-05,
+      "loss": 1.8575,
+      "step": 5722
+    },
+    {
+      "epoch": 0.25014205166309716,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.537597478819136e-05,
+      "loss": 1.7515,
+      "step": 5723
+    },
+    {
+      "epoch": 0.25018575986712704,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.537112040715948e-05,
+      "loss": 1.4989,
+      "step": 5724
+    },
+    {
+      "epoch": 0.250229468071157,
+      "grad_norm": 16.25,
+      "learning_rate": 8.536626535862309e-05,
+      "loss": 5.8871,
+      "step": 5725
+    },
+    {
+      "epoch": 0.25027317627518686,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.53614096426738e-05,
+      "loss": 1.7411,
+      "step": 5726
+    },
+    {
+      "epoch": 0.25031688447921674,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.535655325940324e-05,
+      "loss": 2.1619,
+      "step": 5727
+    },
+    {
+      "epoch": 0.2503605926832466,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.535169620890306e-05,
+      "loss": 1.8591,
+      "step": 5728
+    },
+    {
+      "epoch": 0.25040430088727655,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.53468384912649e-05,
+      "loss": 1.9312,
+      "step": 5729
+    },
+    {
+      "epoch": 0.25044800909130643,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.534198010658047e-05,
+      "loss": 2.3886,
+      "step": 5730
+    },
+    {
+      "epoch": 0.2504917172953363,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.533712105494145e-05,
+      "loss": 1.7222,
+      "step": 5731
+    },
+    {
+      "epoch": 0.25053542549936625,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.533226133643951e-05,
+      "loss": 1.8809,
+      "step": 5732
+    },
+    {
+      "epoch": 0.25057913370339613,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.532740095116638e-05,
+      "loss": 1.8495,
+      "step": 5733
+    },
+    {
+      "epoch": 0.250622841907426,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.532253989921378e-05,
+      "loss": 1.7284,
+      "step": 5734
+    },
+    {
+      "epoch": 0.25066655011145594,
+      "grad_norm": 2.25,
+      "learning_rate": 8.531767818067343e-05,
+      "loss": 1.6024,
+      "step": 5735
+    },
+    {
+      "epoch": 0.2507102583154858,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.53128157956371e-05,
+      "loss": 1.7684,
+      "step": 5736
+    },
+    {
+      "epoch": 0.2507539665195157,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.530795274419654e-05,
+      "loss": 2.55,
+      "step": 5737
+    },
+    {
+      "epoch": 0.2507976747235456,
+      "grad_norm": 2.625,
+      "learning_rate": 8.530308902644353e-05,
+      "loss": 2.0154,
+      "step": 5738
+    },
+    {
+      "epoch": 0.2508413829275755,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.529822464246984e-05,
+      "loss": 1.9612,
+      "step": 5739
+    },
+    {
+      "epoch": 0.2508850911316054,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.529335959236729e-05,
+      "loss": 1.4551,
+      "step": 5740
+    },
+    {
+      "epoch": 0.2509287993356353,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.528849387622766e-05,
+      "loss": 1.9448,
+      "step": 5741
+    },
+    {
+      "epoch": 0.2509725075396652,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.528362749414279e-05,
+      "loss": 2.8076,
+      "step": 5742
+    },
+    {
+      "epoch": 0.2510162157436951,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.527876044620453e-05,
+      "loss": 2.5786,
+      "step": 5743
+    },
+    {
+      "epoch": 0.25105992394772497,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.52738927325047e-05,
+      "loss": 1.8251,
+      "step": 5744
+    },
+    {
+      "epoch": 0.2511036321517549,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.526902435313519e-05,
+      "loss": 2.0876,
+      "step": 5745
+    },
+    {
+      "epoch": 0.2511473403557848,
+      "grad_norm": 2.875,
+      "learning_rate": 8.526415530818785e-05,
+      "loss": 1.8256,
+      "step": 5746
+    },
+    {
+      "epoch": 0.25119104855981467,
+      "grad_norm": 2.875,
+      "learning_rate": 8.525928559775458e-05,
+      "loss": 1.682,
+      "step": 5747
+    },
+    {
+      "epoch": 0.25123475676384455,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.525441522192727e-05,
+      "loss": 1.7598,
+      "step": 5748
+    },
+    {
+      "epoch": 0.2512784649678745,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.524954418079782e-05,
+      "loss": 1.87,
+      "step": 5749
+    },
+    {
+      "epoch": 0.25132217317190436,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.52446724744582e-05,
+      "loss": 2.6512,
+      "step": 5750
+    },
+    {
+      "epoch": 0.25136588137593424,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.52398001030003e-05,
+      "loss": 2.5546,
+      "step": 5751
+    },
+    {
+      "epoch": 0.2514095895799642,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.523492706651607e-05,
+      "loss": 1.8665,
+      "step": 5752
+    },
+    {
+      "epoch": 0.25145329778399406,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.52300533650975e-05,
+      "loss": 1.7692,
+      "step": 5753
+    },
+    {
+      "epoch": 0.25149700598802394,
+      "grad_norm": 2.0,
+      "learning_rate": 8.522517899883654e-05,
+      "loss": 1.8136,
+      "step": 5754
+    },
+    {
+      "epoch": 0.25154071419205387,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.522030396782518e-05,
+      "loss": 1.8562,
+      "step": 5755
+    },
+    {
+      "epoch": 0.25158442239608375,
+      "grad_norm": 4.71875,
+      "learning_rate": 8.521542827215544e-05,
+      "loss": 1.8953,
+      "step": 5756
+    },
+    {
+      "epoch": 0.25162813060011363,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.52105519119193e-05,
+      "loss": 1.6913,
+      "step": 5757
+    },
+    {
+      "epoch": 0.2516718388041435,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.52056748872088e-05,
+      "loss": 1.3204,
+      "step": 5758
+    },
+    {
+      "epoch": 0.25171554700817345,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.520079719811598e-05,
+      "loss": 2.2793,
+      "step": 5759
+    },
+    {
+      "epoch": 0.2517592552122033,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.519591884473288e-05,
+      "loss": 1.7225,
+      "step": 5760
+    },
+    {
+      "epoch": 0.2518029634162332,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.519103982715158e-05,
+      "loss": 2.0289,
+      "step": 5761
+    },
+    {
+      "epoch": 0.25184667162026314,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.518616014546413e-05,
+      "loss": 1.7984,
+      "step": 5762
+    },
+    {
+      "epoch": 0.251890379824293,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.518127979976262e-05,
+      "loss": 1.9239,
+      "step": 5763
+    },
+    {
+      "epoch": 0.2519340880283229,
+      "grad_norm": 6.21875,
+      "learning_rate": 8.517639879013916e-05,
+      "loss": 1.6648,
+      "step": 5764
+    },
+    {
+      "epoch": 0.25197779623235284,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.517151711668587e-05,
+      "loss": 1.3681,
+      "step": 5765
+    },
+    {
+      "epoch": 0.2520215044363827,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.516663477949485e-05,
+      "loss": 1.8568,
+      "step": 5766
+    },
+    {
+      "epoch": 0.2520652126404126,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.516175177865827e-05,
+      "loss": 1.7729,
+      "step": 5767
+    },
+    {
+      "epoch": 0.2521089208444425,
+      "grad_norm": 3.5,
+      "learning_rate": 8.515686811426824e-05,
+      "loss": 1.5158,
+      "step": 5768
+    },
+    {
+      "epoch": 0.2521526290484724,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.515198378641694e-05,
+      "loss": 2.2034,
+      "step": 5769
+    },
+    {
+      "epoch": 0.2521963372525023,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.514709879519653e-05,
+      "loss": 1.7099,
+      "step": 5770
+    },
+    {
+      "epoch": 0.25224004545653217,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.514221314069923e-05,
+      "loss": 2.1182,
+      "step": 5771
+    },
+    {
+      "epoch": 0.2522837536605621,
+      "grad_norm": 2.625,
+      "learning_rate": 8.513732682301723e-05,
+      "loss": 2.1552,
+      "step": 5772
+    },
+    {
+      "epoch": 0.252327461864592,
+      "grad_norm": 2.5,
+      "learning_rate": 8.513243984224273e-05,
+      "loss": 2.3363,
+      "step": 5773
+    },
+    {
+      "epoch": 0.25237117006862186,
+      "grad_norm": 2.125,
+      "learning_rate": 8.512755219846793e-05,
+      "loss": 1.8566,
+      "step": 5774
+    },
+    {
+      "epoch": 0.2524148782726518,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.512266389178511e-05,
+      "loss": 1.7914,
+      "step": 5775
+    },
+    {
+      "epoch": 0.2524585864766817,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.511777492228651e-05,
+      "loss": 1.6455,
+      "step": 5776
+    },
+    {
+      "epoch": 0.25250229468071156,
+      "grad_norm": 2.125,
+      "learning_rate": 8.51128852900644e-05,
+      "loss": 1.6525,
+      "step": 5777
+    },
+    {
+      "epoch": 0.25254600288474144,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.510799499521103e-05,
+      "loss": 2.3558,
+      "step": 5778
+    },
+    {
+      "epoch": 0.2525897110887714,
+      "grad_norm": 2.375,
+      "learning_rate": 8.510310403781867e-05,
+      "loss": 1.8693,
+      "step": 5779
+    },
+    {
+      "epoch": 0.25263341929280125,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.509821241797967e-05,
+      "loss": 1.9012,
+      "step": 5780
+    },
+    {
+      "epoch": 0.25267712749683113,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.509332013578632e-05,
+      "loss": 1.8172,
+      "step": 5781
+    },
+    {
+      "epoch": 0.25272083570086107,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.508842719133094e-05,
+      "loss": 1.3929,
+      "step": 5782
+    },
+    {
+      "epoch": 0.25276454390489095,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.508353358470586e-05,
+      "loss": 2.3934,
+      "step": 5783
+    },
+    {
+      "epoch": 0.25280825210892083,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.507863931600346e-05,
+      "loss": 1.8745,
+      "step": 5784
+    },
+    {
+      "epoch": 0.25285196031295076,
+      "grad_norm": 1.859375,
+      "learning_rate": 8.507374438531607e-05,
+      "loss": 1.8801,
+      "step": 5785
+    },
+    {
+      "epoch": 0.25289566851698064,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.506884879273608e-05,
+      "loss": 1.8671,
+      "step": 5786
+    },
+    {
+      "epoch": 0.2529393767210105,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.506395253835586e-05,
+      "loss": 1.6834,
+      "step": 5787
+    },
+    {
+      "epoch": 0.2529830849250404,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.505905562226783e-05,
+      "loss": 2.4975,
+      "step": 5788
+    },
+    {
+      "epoch": 0.25302679312907034,
+      "grad_norm": 3.125,
+      "learning_rate": 8.50541580445644e-05,
+      "loss": 1.4975,
+      "step": 5789
+    },
+    {
+      "epoch": 0.2530705013331002,
+      "grad_norm": 2.25,
+      "learning_rate": 8.504925980533797e-05,
+      "loss": 2.0253,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2531142095371301,
+      "grad_norm": 2.375,
+      "learning_rate": 8.504436090468103e-05,
+      "loss": 1.482,
+      "step": 5791
+    },
+    {
+      "epoch": 0.25315791774116003,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.503946134268596e-05,
+      "loss": 2.5413,
+      "step": 5792
+    },
+    {
+      "epoch": 0.2532016259451899,
+      "grad_norm": 2.75,
+      "learning_rate": 8.503456111944529e-05,
+      "loss": 1.794,
+      "step": 5793
+    },
+    {
+      "epoch": 0.2532453341492198,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.502966023505143e-05,
+      "loss": 1.9426,
+      "step": 5794
+    },
+    {
+      "epoch": 0.2532890423532497,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.502475868959692e-05,
+      "loss": 1.6883,
+      "step": 5795
+    },
+    {
+      "epoch": 0.2533327505572796,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.501985648317423e-05,
+      "loss": 1.847,
+      "step": 5796
+    },
+    {
+      "epoch": 0.2533764587613095,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.501495361587588e-05,
+      "loss": 1.9255,
+      "step": 5797
+    },
+    {
+      "epoch": 0.2534201669653394,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.501005008779439e-05,
+      "loss": 2.6771,
+      "step": 5798
+    },
+    {
+      "epoch": 0.2534638751693693,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.50051458990223e-05,
+      "loss": 1.9616,
+      "step": 5799
+    },
+    {
+      "epoch": 0.2535075833733992,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.500024104965217e-05,
+      "loss": 1.777,
+      "step": 5800
+    },
+    {
+      "epoch": 0.25355129157742906,
+      "grad_norm": 4.46875,
+      "learning_rate": 8.499533553977654e-05,
+      "loss": 1.363,
+      "step": 5801
+    },
+    {
+      "epoch": 0.253594999781459,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.4990429369488e-05,
+      "loss": 1.8625,
+      "step": 5802
+    },
+    {
+      "epoch": 0.2536387079854889,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.498552253887912e-05,
+      "loss": 1.6072,
+      "step": 5803
+    },
+    {
+      "epoch": 0.25368241618951876,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.498061504804253e-05,
+      "loss": 1.8609,
+      "step": 5804
+    },
+    {
+      "epoch": 0.2537261243935487,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.49757068970708e-05,
+      "loss": 1.9611,
+      "step": 5805
+    },
+    {
+      "epoch": 0.25376983259757857,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.497079808605659e-05,
+      "loss": 1.9539,
+      "step": 5806
+    },
+    {
+      "epoch": 0.25381354080160845,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.496588861509253e-05,
+      "loss": 2.2939,
+      "step": 5807
+    },
+    {
+      "epoch": 0.2538572490056384,
+      "grad_norm": 2.375,
+      "learning_rate": 8.496097848427124e-05,
+      "loss": 2.2952,
+      "step": 5808
+    },
+    {
+      "epoch": 0.25390095720966827,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.495606769368543e-05,
+      "loss": 2.087,
+      "step": 5809
+    },
+    {
+      "epoch": 0.25394466541369815,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.495115624342772e-05,
+      "loss": 1.9989,
+      "step": 5810
+    },
+    {
+      "epoch": 0.253988373617728,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.494624413359083e-05,
+      "loss": 2.0572,
+      "step": 5811
+    },
+    {
+      "epoch": 0.25403208182175796,
+      "grad_norm": 18.25,
+      "learning_rate": 8.494133136426746e-05,
+      "loss": 3.0295,
+      "step": 5812
+    },
+    {
+      "epoch": 0.25407579002578784,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.493641793555032e-05,
+      "loss": 1.7027,
+      "step": 5813
+    },
+    {
+      "epoch": 0.2541194982298177,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.49315038475321e-05,
+      "loss": 2.3696,
+      "step": 5814
+    },
+    {
+      "epoch": 0.25416320643384765,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.492658910030557e-05,
+      "loss": 1.951,
+      "step": 5815
+    },
+    {
+      "epoch": 0.25420691463787753,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.492167369396349e-05,
+      "loss": 1.4611,
+      "step": 5816
+    },
+    {
+      "epoch": 0.2542506228419074,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.491675762859858e-05,
+      "loss": 2.0407,
+      "step": 5817
+    },
+    {
+      "epoch": 0.25429433104593735,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.491184090430364e-05,
+      "loss": 2.1088,
+      "step": 5818
+    },
+    {
+      "epoch": 0.25433803924996723,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.490692352117147e-05,
+      "loss": 1.9155,
+      "step": 5819
+    },
+    {
+      "epoch": 0.2543817474539971,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.490200547929481e-05,
+      "loss": 1.7446,
+      "step": 5820
+    },
+    {
+      "epoch": 0.254425455658027,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.489708677876653e-05,
+      "loss": 1.8726,
+      "step": 5821
+    },
+    {
+      "epoch": 0.2544691638620569,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.489216741967945e-05,
+      "loss": 2.0559,
+      "step": 5822
+    },
+    {
+      "epoch": 0.2545128720660868,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.488724740212636e-05,
+      "loss": 1.9535,
+      "step": 5823
+    },
+    {
+      "epoch": 0.2545565802701167,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.488232672620015e-05,
+      "loss": 2.1477,
+      "step": 5824
+    },
+    {
+      "epoch": 0.2546002884741466,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.487740539199366e-05,
+      "loss": 1.7686,
+      "step": 5825
+    },
+    {
+      "epoch": 0.2546439966781765,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.487248339959976e-05,
+      "loss": 1.6782,
+      "step": 5826
+    },
+    {
+      "epoch": 0.2546877048822064,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.486756074911137e-05,
+      "loss": 1.6152,
+      "step": 5827
+    },
+    {
+      "epoch": 0.2547314130862363,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.486263744062134e-05,
+      "loss": 1.6807,
+      "step": 5828
+    },
+    {
+      "epoch": 0.2547751212902662,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.48577134742226e-05,
+      "loss": 1.3611,
+      "step": 5829
+    },
+    {
+      "epoch": 0.2548188294942961,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.485278885000808e-05,
+      "loss": 2.2076,
+      "step": 5830
+    },
+    {
+      "epoch": 0.25486253769832595,
+      "grad_norm": 2.375,
+      "learning_rate": 8.484786356807071e-05,
+      "loss": 1.8441,
+      "step": 5831
+    },
+    {
+      "epoch": 0.2549062459023559,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.484293762850344e-05,
+      "loss": 2.0972,
+      "step": 5832
+    },
+    {
+      "epoch": 0.25494995410638577,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.483801103139923e-05,
+      "loss": 1.763,
+      "step": 5833
+    },
+    {
+      "epoch": 0.25499366231041565,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.483308377685104e-05,
+      "loss": 1.9297,
+      "step": 5834
+    },
+    {
+      "epoch": 0.2550373705144456,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.482815586495184e-05,
+      "loss": 1.6853,
+      "step": 5835
+    },
+    {
+      "epoch": 0.25508107871847546,
+      "grad_norm": 2.75,
+      "learning_rate": 8.482322729579468e-05,
+      "loss": 2.3892,
+      "step": 5836
+    },
+    {
+      "epoch": 0.25512478692250534,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.481829806947252e-05,
+      "loss": 1.6897,
+      "step": 5837
+    },
+    {
+      "epoch": 0.2551684951265353,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.481336818607842e-05,
+      "loss": 1.6136,
+      "step": 5838
+    },
+    {
+      "epoch": 0.25521220333056516,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.480843764570537e-05,
+      "loss": 2.6346,
+      "step": 5839
+    },
+    {
+      "epoch": 0.25525591153459504,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.480350644844645e-05,
+      "loss": 1.8048,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2552996197386249,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.479857459439471e-05,
+      "loss": 1.9963,
+      "step": 5841
+    },
+    {
+      "epoch": 0.25534332794265485,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.479364208364323e-05,
+      "loss": 1.7453,
+      "step": 5842
+    },
+    {
+      "epoch": 0.25538703614668473,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.478870891628507e-05,
+      "loss": 1.8611,
+      "step": 5843
+    },
+    {
+      "epoch": 0.2554307443507146,
+      "grad_norm": 3.984375,
+      "learning_rate": 8.478377509241334e-05,
+      "loss": 1.5761,
+      "step": 5844
+    },
+    {
+      "epoch": 0.25547445255474455,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.477884061212115e-05,
+      "loss": 3.9752,
+      "step": 5845
+    },
+    {
+      "epoch": 0.2555181607587744,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.477390547550162e-05,
+      "loss": 3.0617,
+      "step": 5846
+    },
+    {
+      "epoch": 0.2555618689628043,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.47689696826479e-05,
+      "loss": 1.7095,
+      "step": 5847
+    },
+    {
+      "epoch": 0.25560557716683424,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.47640332336531e-05,
+      "loss": 2.1366,
+      "step": 5848
+    },
+    {
+      "epoch": 0.2556492853708641,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.47590961286104e-05,
+      "loss": 2.36,
+      "step": 5849
+    },
+    {
+      "epoch": 0.255692993574894,
+      "grad_norm": 4.09375,
+      "learning_rate": 8.475415836761295e-05,
+      "loss": 1.0162,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2557367017789239,
+      "grad_norm": 4.25,
+      "learning_rate": 8.474921995075398e-05,
+      "loss": 1.045,
+      "step": 5851
+    },
+    {
+      "epoch": 0.2557804099829538,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.474428087812664e-05,
+      "loss": 2.409,
+      "step": 5852
+    },
+    {
+      "epoch": 0.2558241181869837,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.473934114982416e-05,
+      "loss": 2.0687,
+      "step": 5853
+    },
+    {
+      "epoch": 0.2558678263910136,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.473440076593973e-05,
+      "loss": 2.1305,
+      "step": 5854
+    },
+    {
+      "epoch": 0.2559115345950435,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.472945972656662e-05,
+      "loss": 1.7336,
+      "step": 5855
+    },
+    {
+      "epoch": 0.2559552427990734,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.472451803179807e-05,
+      "loss": 1.7277,
+      "step": 5856
+    },
+    {
+      "epoch": 0.25599895100310327,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.47195756817273e-05,
+      "loss": 1.8274,
+      "step": 5857
+    },
+    {
+      "epoch": 0.2560426592071332,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.471463267644761e-05,
+      "loss": 2.0623,
+      "step": 5858
+    },
+    {
+      "epoch": 0.2560863674111631,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.47096890160523e-05,
+      "loss": 1.7609,
+      "step": 5859
+    },
+    {
+      "epoch": 0.25613007561519296,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.470474470063461e-05,
+      "loss": 1.7962,
+      "step": 5860
+    },
+    {
+      "epoch": 0.25617378381922284,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.46997997302879e-05,
+      "loss": 2.4369,
+      "step": 5861
+    },
+    {
+      "epoch": 0.2562174920232528,
+      "grad_norm": 2.625,
+      "learning_rate": 8.469485410510545e-05,
+      "loss": 2.3581,
+      "step": 5862
+    },
+    {
+      "epoch": 0.25626120022728266,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.468990782518063e-05,
+      "loss": 2.239,
+      "step": 5863
+    },
+    {
+      "epoch": 0.25630490843131254,
+      "grad_norm": 2.5,
+      "learning_rate": 8.468496089060674e-05,
+      "loss": 1.8326,
+      "step": 5864
+    },
+    {
+      "epoch": 0.2563486166353425,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.468001330147714e-05,
+      "loss": 2.0595,
+      "step": 5865
+    },
+    {
+      "epoch": 0.25639232483937235,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.467506505788525e-05,
+      "loss": 2.3059,
+      "step": 5866
+    },
+    {
+      "epoch": 0.25643603304340223,
+      "grad_norm": 2.25,
+      "learning_rate": 8.46701161599244e-05,
+      "loss": 1.7416,
+      "step": 5867
+    },
+    {
+      "epoch": 0.25647974124743217,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.4665166607688e-05,
+      "loss": 1.5773,
+      "step": 5868
+    },
+    {
+      "epoch": 0.25652344945146205,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.466021640126945e-05,
+      "loss": 2.0622,
+      "step": 5869
+    },
+    {
+      "epoch": 0.25656715765549193,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.465526554076217e-05,
+      "loss": 1.6875,
+      "step": 5870
+    },
+    {
+      "epoch": 0.2566108658595218,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.46503140262596e-05,
+      "loss": 1.9292,
+      "step": 5871
+    },
+    {
+      "epoch": 0.25665457406355174,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.464536185785516e-05,
+      "loss": 1.3767,
+      "step": 5872
+    },
+    {
+      "epoch": 0.2566982822675816,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.464040903564234e-05,
+      "loss": 2.645,
+      "step": 5873
+    },
+    {
+      "epoch": 0.2567419904716115,
+      "grad_norm": 2.375,
+      "learning_rate": 8.463545555971458e-05,
+      "loss": 1.9756,
+      "step": 5874
+    },
+    {
+      "epoch": 0.25678569867564144,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.463050143016533e-05,
+      "loss": 1.9329,
+      "step": 5875
+    },
+    {
+      "epoch": 0.2568294068796713,
+      "grad_norm": 2.75,
+      "learning_rate": 8.462554664708815e-05,
+      "loss": 2.5044,
+      "step": 5876
+    },
+    {
+      "epoch": 0.2568731150837012,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.462059121057651e-05,
+      "loss": 2.1592,
+      "step": 5877
+    },
+    {
+      "epoch": 0.25691682328773113,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.461563512072392e-05,
+      "loss": 1.6989,
+      "step": 5878
+    },
+    {
+      "epoch": 0.256960531491761,
+      "grad_norm": 2.125,
+      "learning_rate": 8.461067837762391e-05,
+      "loss": 1.9492,
+      "step": 5879
+    },
+    {
+      "epoch": 0.2570042396957909,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.460572098137e-05,
+      "loss": 1.9093,
+      "step": 5880
+    },
+    {
+      "epoch": 0.2570479478998208,
+      "grad_norm": 3.578125,
+      "learning_rate": 8.460076293205581e-05,
+      "loss": 1.6803,
+      "step": 5881
+    },
+    {
+      "epoch": 0.2570916561038507,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.459580422977484e-05,
+      "loss": 2.7429,
+      "step": 5882
+    },
+    {
+      "epoch": 0.2571353643078806,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.459084487462072e-05,
+      "loss": 1.4708,
+      "step": 5883
+    },
+    {
+      "epoch": 0.25717907251191047,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.4585884866687e-05,
+      "loss": 2.23,
+      "step": 5884
+    },
+    {
+      "epoch": 0.2572227807159404,
+      "grad_norm": 2.75,
+      "learning_rate": 8.458092420606727e-05,
+      "loss": 1.7679,
+      "step": 5885
+    },
+    {
+      "epoch": 0.2572664889199703,
+      "grad_norm": 2.0,
+      "learning_rate": 8.45759628928552e-05,
+      "loss": 1.1209,
+      "step": 5886
+    },
+    {
+      "epoch": 0.25731019712400016,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.457100092714438e-05,
+      "loss": 2.1544,
+      "step": 5887
+    },
+    {
+      "epoch": 0.2573539053280301,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.456603830902845e-05,
+      "loss": 1.6905,
+      "step": 5888
+    },
+    {
+      "epoch": 0.25739761353206,
+      "grad_norm": 3.125,
+      "learning_rate": 8.456107503860107e-05,
+      "loss": 1.5867,
+      "step": 5889
+    },
+    {
+      "epoch": 0.25744132173608986,
+      "grad_norm": 2.125,
+      "learning_rate": 8.455611111595591e-05,
+      "loss": 2.0275,
+      "step": 5890
+    },
+    {
+      "epoch": 0.25748502994011974,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.455114654118663e-05,
+      "loss": 1.8815,
+      "step": 5891
+    },
+    {
+      "epoch": 0.25752873814414967,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.454618131438693e-05,
+      "loss": 1.4841,
+      "step": 5892
+    },
+    {
+      "epoch": 0.25757244634817955,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.454121543565052e-05,
+      "loss": 1.9968,
+      "step": 5893
+    },
+    {
+      "epoch": 0.25761615455220943,
+      "grad_norm": 4.6875,
+      "learning_rate": 8.453624890507109e-05,
+      "loss": 1.7452,
+      "step": 5894
+    },
+    {
+      "epoch": 0.25765986275623937,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.453128172274238e-05,
+      "loss": 2.609,
+      "step": 5895
+    },
+    {
+      "epoch": 0.25770357096026925,
+      "grad_norm": 2.0,
+      "learning_rate": 8.452631388875814e-05,
+      "loss": 1.4952,
+      "step": 5896
+    },
+    {
+      "epoch": 0.2577472791642991,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.452134540321208e-05,
+      "loss": 1.5478,
+      "step": 5897
+    },
+    {
+      "epoch": 0.25779098736832906,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.451637626619801e-05,
+      "loss": 2.7189,
+      "step": 5898
+    },
+    {
+      "epoch": 0.25783469557235894,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.451140647780967e-05,
+      "loss": 2.0582,
+      "step": 5899
+    },
+    {
+      "epoch": 0.2578784037763888,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.450643603814086e-05,
+      "loss": 1.5491,
+      "step": 5900
+    },
+    {
+      "epoch": 0.2579221119804187,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.45014649472854e-05,
+      "loss": 1.8259,
+      "step": 5901
+    },
+    {
+      "epoch": 0.25796582018444864,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.449649320533706e-05,
+      "loss": 2.0901,
+      "step": 5902
+    },
+    {
+      "epoch": 0.2580095283884785,
+      "grad_norm": 2.5,
+      "learning_rate": 8.44915208123897e-05,
+      "loss": 1.1867,
+      "step": 5903
+    },
+    {
+      "epoch": 0.2580532365925084,
+      "grad_norm": 2.125,
+      "learning_rate": 8.448654776853714e-05,
+      "loss": 2.0911,
+      "step": 5904
+    },
+    {
+      "epoch": 0.25809694479653833,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.448157407387323e-05,
+      "loss": 1.4067,
+      "step": 5905
+    },
+    {
+      "epoch": 0.2581406530005682,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.447659972849183e-05,
+      "loss": 1.938,
+      "step": 5906
+    },
+    {
+      "epoch": 0.2581843612045981,
+      "grad_norm": 2.25,
+      "learning_rate": 8.447162473248682e-05,
+      "loss": 1.7084,
+      "step": 5907
+    },
+    {
+      "epoch": 0.258228069408628,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.446664908595207e-05,
+      "loss": 2.4166,
+      "step": 5908
+    },
+    {
+      "epoch": 0.2582717776126579,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.44616727889815e-05,
+      "loss": 2.0318,
+      "step": 5909
+    },
+    {
+      "epoch": 0.2583154858166878,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.4456695841669e-05,
+      "loss": 1.4643,
+      "step": 5910
+    },
+    {
+      "epoch": 0.25835919402071766,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.445171824410848e-05,
+      "loss": 1.8736,
+      "step": 5911
+    },
+    {
+      "epoch": 0.2584029022247476,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.444673999639393e-05,
+      "loss": 2.0095,
+      "step": 5912
+    },
+    {
+      "epoch": 0.2584466104287775,
+      "grad_norm": 3.875,
+      "learning_rate": 8.444176109861926e-05,
+      "loss": 1.1565,
+      "step": 5913
+    },
+    {
+      "epoch": 0.25849031863280736,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.44367815508784e-05,
+      "loss": 2.3151,
+      "step": 5914
+    },
+    {
+      "epoch": 0.2585340268368373,
+      "grad_norm": 3.296875,
+      "learning_rate": 8.443180135326538e-05,
+      "loss": 2.2622,
+      "step": 5915
+    },
+    {
+      "epoch": 0.2585777350408672,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.442682050587414e-05,
+      "loss": 1.0598,
+      "step": 5916
+    },
+    {
+      "epoch": 0.25862144324489705,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.44218390087987e-05,
+      "loss": 1.5134,
+      "step": 5917
+    },
+    {
+      "epoch": 0.258665151448927,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.441685686213306e-05,
+      "loss": 1.896,
+      "step": 5918
+    },
+    {
+      "epoch": 0.25870885965295687,
+      "grad_norm": 2.25,
+      "learning_rate": 8.441187406597123e-05,
+      "loss": 3.0446,
+      "step": 5919
+    },
+    {
+      "epoch": 0.25875256785698675,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.440689062040725e-05,
+      "loss": 1.9769,
+      "step": 5920
+    },
+    {
+      "epoch": 0.25879627606101663,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.440190652553517e-05,
+      "loss": 2.0879,
+      "step": 5921
+    },
+    {
+      "epoch": 0.25883998426504656,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.439692178144906e-05,
+      "loss": 2.3422,
+      "step": 5922
+    },
+    {
+      "epoch": 0.25888369246907644,
+      "grad_norm": 3.859375,
+      "learning_rate": 8.439193638824296e-05,
+      "loss": 2.7385,
+      "step": 5923
+    },
+    {
+      "epoch": 0.2589274006731063,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.438695034601097e-05,
+      "loss": 1.7118,
+      "step": 5924
+    },
+    {
+      "epoch": 0.25897110887713626,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.438196365484718e-05,
+      "loss": 2.2815,
+      "step": 5925
+    },
+    {
+      "epoch": 0.25901481708116614,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.43769763148457e-05,
+      "loss": 1.369,
+      "step": 5926
+    },
+    {
+      "epoch": 0.259058525285196,
+      "grad_norm": 2.984375,
+      "learning_rate": 8.437198832610063e-05,
+      "loss": 2.1702,
+      "step": 5927
+    },
+    {
+      "epoch": 0.25910223348922595,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.436699968870612e-05,
+      "loss": 1.6627,
+      "step": 5928
+    },
+    {
+      "epoch": 0.25914594169325583,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.43620104027563e-05,
+      "loss": 1.8602,
+      "step": 5929
+    },
+    {
+      "epoch": 0.2591896498972857,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.435702046834535e-05,
+      "loss": 1.3718,
+      "step": 5930
+    },
+    {
+      "epoch": 0.2592333581013156,
+      "grad_norm": 5.0625,
+      "learning_rate": 8.435202988556739e-05,
+      "loss": 2.0221,
+      "step": 5931
+    },
+    {
+      "epoch": 0.2592770663053455,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.434703865451665e-05,
+      "loss": 2.0702,
+      "step": 5932
+    },
+    {
+      "epoch": 0.2593207745093754,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.43420467752873e-05,
+      "loss": 1.816,
+      "step": 5933
+    },
+    {
+      "epoch": 0.2593644827134053,
+      "grad_norm": 2.5,
+      "learning_rate": 8.433705424797355e-05,
+      "loss": 2.138,
+      "step": 5934
+    },
+    {
+      "epoch": 0.2594081909174352,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.43320610726696e-05,
+      "loss": 2.6959,
+      "step": 5935
+    },
+    {
+      "epoch": 0.2594518991214651,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.43270672494697e-05,
+      "loss": 1.9112,
+      "step": 5936
+    },
+    {
+      "epoch": 0.259495607325495,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.432207277846806e-05,
+      "loss": 1.7615,
+      "step": 5937
+    },
+    {
+      "epoch": 0.2595393155295249,
+      "grad_norm": 2.375,
+      "learning_rate": 8.431707765975898e-05,
+      "loss": 2.2894,
+      "step": 5938
+    },
+    {
+      "epoch": 0.2595830237335548,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.43120818934367e-05,
+      "loss": 1.6917,
+      "step": 5939
+    },
+    {
+      "epoch": 0.2596267319375847,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.430708547959547e-05,
+      "loss": 1.6675,
+      "step": 5940
+    },
+    {
+      "epoch": 0.25967044014161456,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.43020884183296e-05,
+      "loss": 2.4967,
+      "step": 5941
+    },
+    {
+      "epoch": 0.2597141483456445,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.429709070973342e-05,
+      "loss": 2.0623,
+      "step": 5942
+    },
+    {
+      "epoch": 0.25975785654967437,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.429209235390123e-05,
+      "loss": 1.6179,
+      "step": 5943
+    },
+    {
+      "epoch": 0.25980156475370425,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.428709335092733e-05,
+      "loss": 1.2945,
+      "step": 5944
+    },
+    {
+      "epoch": 0.2598452729577342,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.428209370090607e-05,
+      "loss": 2.2274,
+      "step": 5945
+    },
+    {
+      "epoch": 0.25988898116176407,
+      "grad_norm": 3.125,
+      "learning_rate": 8.427709340393181e-05,
+      "loss": 1.5624,
+      "step": 5946
+    },
+    {
+      "epoch": 0.25993268936579395,
+      "grad_norm": 2.625,
+      "learning_rate": 8.427209246009893e-05,
+      "loss": 1.8742,
+      "step": 5947
+    },
+    {
+      "epoch": 0.2599763975698239,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.426709086950178e-05,
+      "loss": 2.013,
+      "step": 5948
+    },
+    {
+      "epoch": 0.26002010577385376,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.426208863223473e-05,
+      "loss": 2.0844,
+      "step": 5949
+    },
+    {
+      "epoch": 0.26006381397788364,
+      "grad_norm": 2.25,
+      "learning_rate": 8.425708574839222e-05,
+      "loss": 1.6669,
+      "step": 5950
+    },
+    {
+      "epoch": 0.2601075221819135,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.425208221806862e-05,
+      "loss": 1.0572,
+      "step": 5951
+    },
+    {
+      "epoch": 0.26015123038594345,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.42470780413584e-05,
+      "loss": 1.8001,
+      "step": 5952
+    },
+    {
+      "epoch": 0.26019493858997333,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.424207321835598e-05,
+      "loss": 1.5684,
+      "step": 5953
+    },
+    {
+      "epoch": 0.2602386467940032,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.423706774915579e-05,
+      "loss": 1.7722,
+      "step": 5954
+    },
+    {
+      "epoch": 0.26028235499803315,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.42320616338523e-05,
+      "loss": 1.5773,
+      "step": 5955
+    },
+    {
+      "epoch": 0.26032606320206303,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.422705487253996e-05,
+      "loss": 1.4819,
+      "step": 5956
+    },
+    {
+      "epoch": 0.2603697714060929,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.422204746531332e-05,
+      "loss": 2.35,
+      "step": 5957
+    },
+    {
+      "epoch": 0.26041347961012284,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.421703941226682e-05,
+      "loss": 1.896,
+      "step": 5958
+    },
+    {
+      "epoch": 0.2604571878141527,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.421203071349498e-05,
+      "loss": 1.5529,
+      "step": 5959
+    },
+    {
+      "epoch": 0.2605008960181826,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.420702136909234e-05,
+      "loss": 1.3801,
+      "step": 5960
+    },
+    {
+      "epoch": 0.2605446042222125,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.420201137915342e-05,
+      "loss": 1.6023,
+      "step": 5961
+    },
+    {
+      "epoch": 0.2605883124262424,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.419700074377275e-05,
+      "loss": 1.5803,
+      "step": 5962
+    },
+    {
+      "epoch": 0.2606320206302723,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.419198946304491e-05,
+      "loss": 1.7009,
+      "step": 5963
+    },
+    {
+      "epoch": 0.2606757288343022,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.418697753706449e-05,
+      "loss": 2.4918,
+      "step": 5964
+    },
+    {
+      "epoch": 0.2607194370383321,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.418196496592603e-05,
+      "loss": 2.4145,
+      "step": 5965
+    },
+    {
+      "epoch": 0.260763145242362,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.417695174972413e-05,
+      "loss": 2.5977,
+      "step": 5966
+    },
+    {
+      "epoch": 0.2608068534463919,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.41719378885534e-05,
+      "loss": 2.1141,
+      "step": 5967
+    },
+    {
+      "epoch": 0.2608505616504218,
+      "grad_norm": 2.125,
+      "learning_rate": 8.416692338250848e-05,
+      "loss": 2.1758,
+      "step": 5968
+    },
+    {
+      "epoch": 0.2608942698544517,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.416190823168402e-05,
+      "loss": 1.8151,
+      "step": 5969
+    },
+    {
+      "epoch": 0.26093797805848157,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.41568924361746e-05,
+      "loss": 1.7852,
+      "step": 5970
+    },
+    {
+      "epoch": 0.26098168626251145,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.41518759960749e-05,
+      "loss": 1.7464,
+      "step": 5971
+    },
+    {
+      "epoch": 0.2610253944665414,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.41468589114796e-05,
+      "loss": 2.5288,
+      "step": 5972
+    },
+    {
+      "epoch": 0.26106910267057126,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.414184118248339e-05,
+      "loss": 1.4174,
+      "step": 5973
+    },
+    {
+      "epoch": 0.26111281087460114,
+      "grad_norm": 2.75,
+      "learning_rate": 8.413682280918093e-05,
+      "loss": 2.227,
+      "step": 5974
+    },
+    {
+      "epoch": 0.2611565190786311,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.413180379166694e-05,
+      "loss": 2.214,
+      "step": 5975
+    },
+    {
+      "epoch": 0.26120022728266096,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.412678413003614e-05,
+      "loss": 1.6059,
+      "step": 5976
+    },
+    {
+      "epoch": 0.26124393548669084,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.412176382438326e-05,
+      "loss": 1.8957,
+      "step": 5977
+    },
+    {
+      "epoch": 0.26128764369072077,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.411674287480303e-05,
+      "loss": 1.9365,
+      "step": 5978
+    },
+    {
+      "epoch": 0.26133135189475065,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.411172128139021e-05,
+      "loss": 2.0909,
+      "step": 5979
+    },
+    {
+      "epoch": 0.26137506009878053,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.410669904423955e-05,
+      "loss": 1.919,
+      "step": 5980
+    },
+    {
+      "epoch": 0.2614187683028104,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.410167616344586e-05,
+      "loss": 2.6475,
+      "step": 5981
+    },
+    {
+      "epoch": 0.26146247650684035,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.40966526391039e-05,
+      "loss": 1.7789,
+      "step": 5982
+    },
+    {
+      "epoch": 0.2615061847108702,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.409162847130847e-05,
+      "loss": 1.9418,
+      "step": 5983
+    },
+    {
+      "epoch": 0.2615498929149001,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.40866036601544e-05,
+      "loss": 1.7559,
+      "step": 5984
+    },
+    {
+      "epoch": 0.26159360111893004,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.40815782057365e-05,
+      "loss": 2.1114,
+      "step": 5985
+    },
+    {
+      "epoch": 0.2616373093229599,
+      "grad_norm": 1.875,
+      "learning_rate": 8.407655210814962e-05,
+      "loss": 1.7708,
+      "step": 5986
+    },
+    {
+      "epoch": 0.2616810175269898,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.407152536748861e-05,
+      "loss": 1.7084,
+      "step": 5987
+    },
+    {
+      "epoch": 0.26172472573101974,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.406649798384834e-05,
+      "loss": 1.8533,
+      "step": 5988
+    },
+    {
+      "epoch": 0.2617684339350496,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.406146995732365e-05,
+      "loss": 1.8596,
+      "step": 5989
+    },
+    {
+      "epoch": 0.2618121421390795,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.405644128800945e-05,
+      "loss": 1.7747,
+      "step": 5990
+    },
+    {
+      "epoch": 0.2618558503431094,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.405141197600064e-05,
+      "loss": 1.559,
+      "step": 5991
+    },
+    {
+      "epoch": 0.2618995585471393,
+      "grad_norm": 2.375,
+      "learning_rate": 8.404638202139213e-05,
+      "loss": 1.9718,
+      "step": 5992
+    },
+    {
+      "epoch": 0.2619432667511692,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.404135142427884e-05,
+      "loss": 1.9266,
+      "step": 5993
+    },
+    {
+      "epoch": 0.26198697495519907,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.40363201847557e-05,
+      "loss": 1.9553,
+      "step": 5994
+    },
+    {
+      "epoch": 0.262030683159229,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.403128830291767e-05,
+      "loss": 2.5839,
+      "step": 5995
+    },
+    {
+      "epoch": 0.2620743913632589,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.402625577885968e-05,
+      "loss": 2.0484,
+      "step": 5996
+    },
+    {
+      "epoch": 0.26211809956728876,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.402122261267673e-05,
+      "loss": 1.694,
+      "step": 5997
+    },
+    {
+      "epoch": 0.2621618077713187,
+      "grad_norm": 2.625,
+      "learning_rate": 8.40161888044638e-05,
+      "loss": 2.2003,
+      "step": 5998
+    },
+    {
+      "epoch": 0.2622055159753486,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.401115435431587e-05,
+      "loss": 2.1018,
+      "step": 5999
+    },
+    {
+      "epoch": 0.26224922417937846,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.400611926232797e-05,
+      "loss": 2.1926,
+      "step": 6000
+    },
+    {
+      "epoch": 0.26229293238340834,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.400108352859508e-05,
+      "loss": 2.101,
+      "step": 6001
+    },
+    {
+      "epoch": 0.2623366405874383,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.399604715321227e-05,
+      "loss": 1.8448,
+      "step": 6002
+    },
+    {
+      "epoch": 0.26238034879146815,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.399101013627458e-05,
+      "loss": 2.1145,
+      "step": 6003
+    },
+    {
+      "epoch": 0.26242405699549803,
+      "grad_norm": 2.125,
+      "learning_rate": 8.398597247787705e-05,
+      "loss": 1.993,
+      "step": 6004
+    },
+    {
+      "epoch": 0.26246776519952797,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.398093417811477e-05,
+      "loss": 1.7932,
+      "step": 6005
+    },
+    {
+      "epoch": 0.26251147340355785,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.397589523708278e-05,
+      "loss": 1.4415,
+      "step": 6006
+    },
+    {
+      "epoch": 0.26255518160758773,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.397085565487621e-05,
+      "loss": 2.1253,
+      "step": 6007
+    },
+    {
+      "epoch": 0.26259888981161766,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.396581543159017e-05,
+      "loss": 2.4571,
+      "step": 6008
+    },
+    {
+      "epoch": 0.26264259801564754,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.396077456731974e-05,
+      "loss": 1.5223,
+      "step": 6009
+    },
+    {
+      "epoch": 0.2626863062196774,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.395573306216005e-05,
+      "loss": 1.9146,
+      "step": 6010
+    },
+    {
+      "epoch": 0.2627300144237073,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.395069091620628e-05,
+      "loss": 2.2909,
+      "step": 6011
+    },
+    {
+      "epoch": 0.26277372262773724,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.394564812955355e-05,
+      "loss": 1.7254,
+      "step": 6012
+    },
+    {
+      "epoch": 0.2628174308317671,
+      "grad_norm": 1.875,
+      "learning_rate": 8.394060470229704e-05,
+      "loss": 2.082,
+      "step": 6013
+    },
+    {
+      "epoch": 0.262861139035797,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.393556063453192e-05,
+      "loss": 1.4722,
+      "step": 6014
+    },
+    {
+      "epoch": 0.26290484723982693,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.393051592635337e-05,
+      "loss": 1.8483,
+      "step": 6015
+    },
+    {
+      "epoch": 0.2629485554438568,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.392547057785661e-05,
+      "loss": 1.9085,
+      "step": 6016
+    },
+    {
+      "epoch": 0.2629922636478867,
+      "grad_norm": 2.5,
+      "learning_rate": 8.392042458913685e-05,
+      "loss": 2.4344,
+      "step": 6017
+    },
+    {
+      "epoch": 0.26303597185191663,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.39153779602893e-05,
+      "loss": 2.4289,
+      "step": 6018
+    },
+    {
+      "epoch": 0.2630796800559465,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.391033069140921e-05,
+      "loss": 2.3144,
+      "step": 6019
+    },
+    {
+      "epoch": 0.2631233882599764,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.390528278259182e-05,
+      "loss": 1.7527,
+      "step": 6020
+    },
+    {
+      "epoch": 0.26316709646400627,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.39002342339324e-05,
+      "loss": 2.7173,
+      "step": 6021
+    },
+    {
+      "epoch": 0.2632108046680362,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.389518504552622e-05,
+      "loss": 1.7396,
+      "step": 6022
+    },
+    {
+      "epoch": 0.2632545128720661,
+      "grad_norm": 2.625,
+      "learning_rate": 8.389013521746857e-05,
+      "loss": 2.6851,
+      "step": 6023
+    },
+    {
+      "epoch": 0.26329822107609596,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.388508474985474e-05,
+      "loss": 1.8855,
+      "step": 6024
+    },
+    {
+      "epoch": 0.2633419292801259,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.388003364278005e-05,
+      "loss": 1.9291,
+      "step": 6025
+    },
+    {
+      "epoch": 0.2633856374841558,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.387498189633979e-05,
+      "loss": 1.743,
+      "step": 6026
+    },
+    {
+      "epoch": 0.26342934568818566,
+      "grad_norm": 3.40625,
+      "learning_rate": 8.386992951062935e-05,
+      "loss": 2.0112,
+      "step": 6027
+    },
+    {
+      "epoch": 0.2634730538922156,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.386487648574403e-05,
+      "loss": 1.7297,
+      "step": 6028
+    },
+    {
+      "epoch": 0.26351676209624547,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.385982282177922e-05,
+      "loss": 1.628,
+      "step": 6029
+    },
+    {
+      "epoch": 0.26356047030027535,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.385476851883025e-05,
+      "loss": 1.2578,
+      "step": 6030
+    },
+    {
+      "epoch": 0.26360417850430523,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.384971357699254e-05,
+      "loss": 2.0018,
+      "step": 6031
+    },
+    {
+      "epoch": 0.26364788670833517,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.384465799636145e-05,
+      "loss": 1.8163,
+      "step": 6032
+    },
+    {
+      "epoch": 0.26369159491236505,
+      "grad_norm": 2.25,
+      "learning_rate": 8.383960177703243e-05,
+      "loss": 1.2395,
+      "step": 6033
+    },
+    {
+      "epoch": 0.2637353031163949,
+      "grad_norm": 2.125,
+      "learning_rate": 8.383454491910086e-05,
+      "loss": 1.9073,
+      "step": 6034
+    },
+    {
+      "epoch": 0.26377901132042486,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.38294874226622e-05,
+      "loss": 1.8201,
+      "step": 6035
+    },
+    {
+      "epoch": 0.26382271952445474,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.382442928781184e-05,
+      "loss": 2.0795,
+      "step": 6036
+    },
+    {
+      "epoch": 0.2638664277284846,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.38193705146453e-05,
+      "loss": 1.7793,
+      "step": 6037
+    },
+    {
+      "epoch": 0.26391013593251456,
+      "grad_norm": 2.0,
+      "learning_rate": 8.3814311103258e-05,
+      "loss": 1.4628,
+      "step": 6038
+    },
+    {
+      "epoch": 0.26395384413654444,
+      "grad_norm": 2.0,
+      "learning_rate": 8.380925105374544e-05,
+      "loss": 1.5489,
+      "step": 6039
+    },
+    {
+      "epoch": 0.2639975523405743,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.380419036620312e-05,
+      "loss": 1.9486,
+      "step": 6040
+    },
+    {
+      "epoch": 0.2640412605446042,
+      "grad_norm": 2.0,
+      "learning_rate": 8.379912904072651e-05,
+      "loss": 1.6631,
+      "step": 6041
+    },
+    {
+      "epoch": 0.26408496874863413,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.379406707741115e-05,
+      "loss": 2.0659,
+      "step": 6042
+    },
+    {
+      "epoch": 0.264128676952664,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.378900447635256e-05,
+      "loss": 1.9999,
+      "step": 6043
+    },
+    {
+      "epoch": 0.2641723851566939,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.378394123764628e-05,
+      "loss": 2.2885,
+      "step": 6044
+    },
+    {
+      "epoch": 0.2642160933607238,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.377887736138786e-05,
+      "loss": 1.6849,
+      "step": 6045
+    },
+    {
+      "epoch": 0.2642598015647537,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.377381284767285e-05,
+      "loss": 1.595,
+      "step": 6046
+    },
+    {
+      "epoch": 0.2643035097687836,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.376874769659684e-05,
+      "loss": 1.8376,
+      "step": 6047
+    },
+    {
+      "epoch": 0.2643472179728135,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.376368190825541e-05,
+      "loss": 1.3498,
+      "step": 6048
+    },
+    {
+      "epoch": 0.2643909261768434,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.375861548274418e-05,
+      "loss": 1.5076,
+      "step": 6049
+    },
+    {
+      "epoch": 0.2644346343808733,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.375354842015873e-05,
+      "loss": 1.6232,
+      "step": 6050
+    },
+    {
+      "epoch": 0.26447834258490316,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.374848072059469e-05,
+      "loss": 1.8066,
+      "step": 6051
+    },
+    {
+      "epoch": 0.2645220507889331,
+      "grad_norm": 3.21875,
+      "learning_rate": 8.374341238414769e-05,
+      "loss": 1.2586,
+      "step": 6052
+    },
+    {
+      "epoch": 0.264565758992963,
+      "grad_norm": 3.609375,
+      "learning_rate": 8.37383434109134e-05,
+      "loss": 1.9279,
+      "step": 6053
+    },
+    {
+      "epoch": 0.26460946719699285,
+      "grad_norm": 4.375,
+      "learning_rate": 8.373327380098748e-05,
+      "loss": 1.6612,
+      "step": 6054
+    },
+    {
+      "epoch": 0.2646531754010228,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.372820355446558e-05,
+      "loss": 1.921,
+      "step": 6055
+    },
+    {
+      "epoch": 0.26469688360505267,
+      "grad_norm": 2.25,
+      "learning_rate": 8.372313267144338e-05,
+      "loss": 2.0424,
+      "step": 6056
+    },
+    {
+      "epoch": 0.26474059180908255,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.37180611520166e-05,
+      "loss": 2.2897,
+      "step": 6057
+    },
+    {
+      "epoch": 0.2647843000131125,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.371298899628091e-05,
+      "loss": 1.8176,
+      "step": 6058
+    },
+    {
+      "epoch": 0.26482800821714236,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.370791620433206e-05,
+      "loss": 1.5037,
+      "step": 6059
+    },
+    {
+      "epoch": 0.26487171642117224,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.370284277626577e-05,
+      "loss": 2.4598,
+      "step": 6060
+    },
+    {
+      "epoch": 0.2649154246252021,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.369776871217781e-05,
+      "loss": 1.8672,
+      "step": 6061
+    },
+    {
+      "epoch": 0.26495913282923206,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.369269401216387e-05,
+      "loss": 2.3062,
+      "step": 6062
+    },
+    {
+      "epoch": 0.26500284103326194,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.368761867631978e-05,
+      "loss": 2.6771,
+      "step": 6063
+    },
+    {
+      "epoch": 0.2650465492372918,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.368254270474128e-05,
+      "loss": 2.3499,
+      "step": 6064
+    },
+    {
+      "epoch": 0.26509025744132175,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.367746609752419e-05,
+      "loss": 2.033,
+      "step": 6065
+    },
+    {
+      "epoch": 0.26513396564535163,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.36723888547643e-05,
+      "loss": 1.5885,
+      "step": 6066
+    },
+    {
+      "epoch": 0.2651776738493815,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.366731097655742e-05,
+      "loss": 2.0933,
+      "step": 6067
+    },
+    {
+      "epoch": 0.26522138205341145,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.366223246299938e-05,
+      "loss": 1.838,
+      "step": 6068
+    },
+    {
+      "epoch": 0.2652650902574413,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.365715331418602e-05,
+      "loss": 2.1014,
+      "step": 6069
+    },
+    {
+      "epoch": 0.2653087984614712,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.36520735302132e-05,
+      "loss": 1.5907,
+      "step": 6070
+    },
+    {
+      "epoch": 0.26535250666550114,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.364699311117677e-05,
+      "loss": 1.9376,
+      "step": 6071
+    },
+    {
+      "epoch": 0.265396214869531,
+      "grad_norm": 1.890625,
+      "learning_rate": 8.36419120571726e-05,
+      "loss": 1.7767,
+      "step": 6072
+    },
+    {
+      "epoch": 0.2654399230735609,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.36368303682966e-05,
+      "loss": 1.4563,
+      "step": 6073
+    },
+    {
+      "epoch": 0.2654836312775908,
+      "grad_norm": 12.125,
+      "learning_rate": 8.363174804464465e-05,
+      "loss": 2.3386,
+      "step": 6074
+    },
+    {
+      "epoch": 0.2655273394816207,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.362666508631264e-05,
+      "loss": 1.813,
+      "step": 6075
+    },
+    {
+      "epoch": 0.2655710476856506,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.362158149339656e-05,
+      "loss": 1.9771,
+      "step": 6076
+    },
+    {
+      "epoch": 0.2656147558896805,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.361649726599228e-05,
+      "loss": 2.1627,
+      "step": 6077
+    },
+    {
+      "epoch": 0.2656584640937104,
+      "grad_norm": 2.125,
+      "learning_rate": 8.361141240419578e-05,
+      "loss": 1.6052,
+      "step": 6078
+    },
+    {
+      "epoch": 0.2657021722977403,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.3606326908103e-05,
+      "loss": 2.0792,
+      "step": 6079
+    },
+    {
+      "epoch": 0.26574588050177017,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.360124077780994e-05,
+      "loss": 1.6047,
+      "step": 6080
+    },
+    {
+      "epoch": 0.2657895887058001,
+      "grad_norm": 2.25,
+      "learning_rate": 8.359615401341254e-05,
+      "loss": 2.0099,
+      "step": 6081
+    },
+    {
+      "epoch": 0.26583329690983,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.359106661500683e-05,
+      "loss": 1.9349,
+      "step": 6082
+    },
+    {
+      "epoch": 0.26587700511385987,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.358597858268878e-05,
+      "loss": 1.8877,
+      "step": 6083
+    },
+    {
+      "epoch": 0.26592071331788975,
+      "grad_norm": 2.0,
+      "learning_rate": 8.358088991655447e-05,
+      "loss": 1.7101,
+      "step": 6084
+    },
+    {
+      "epoch": 0.2659644215219197,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.357580061669985e-05,
+      "loss": 1.8284,
+      "step": 6085
+    },
+    {
+      "epoch": 0.26600812972594956,
+      "grad_norm": 2.5,
+      "learning_rate": 8.357071068322104e-05,
+      "loss": 2.4833,
+      "step": 6086
+    },
+    {
+      "epoch": 0.26605183792997944,
+      "grad_norm": 2.125,
+      "learning_rate": 8.356562011621404e-05,
+      "loss": 1.9072,
+      "step": 6087
+    },
+    {
+      "epoch": 0.2660955461340094,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.356052891577494e-05,
+      "loss": 1.7125,
+      "step": 6088
+    },
+    {
+      "epoch": 0.26613925433803926,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.355543708199982e-05,
+      "loss": 1.396,
+      "step": 6089
+    },
+    {
+      "epoch": 0.26618296254206913,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.355034461498477e-05,
+      "loss": 1.6723,
+      "step": 6090
+    },
+    {
+      "epoch": 0.26622667074609907,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.354525151482587e-05,
+      "loss": 1.8009,
+      "step": 6091
+    },
+    {
+      "epoch": 0.26627037895012895,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.354015778161925e-05,
+      "loss": 2.198,
+      "step": 6092
+    },
+    {
+      "epoch": 0.26631408715415883,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.353506341546104e-05,
+      "loss": 1.7902,
+      "step": 6093
+    },
+    {
+      "epoch": 0.2663577953581887,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.352996841644741e-05,
+      "loss": 1.8954,
+      "step": 6094
+    },
+    {
+      "epoch": 0.26640150356221864,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.352487278467443e-05,
+      "loss": 1.6802,
+      "step": 6095
+    },
+    {
+      "epoch": 0.2664452117662485,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.351977652023833e-05,
+      "loss": 2.098,
+      "step": 6096
+    },
+    {
+      "epoch": 0.2664889199702784,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.351467962323525e-05,
+      "loss": 1.4421,
+      "step": 6097
+    },
+    {
+      "epoch": 0.26653262817430834,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.350958209376138e-05,
+      "loss": 1.5341,
+      "step": 6098
+    },
+    {
+      "epoch": 0.2665763363783382,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.350448393191295e-05,
+      "loss": 1.7144,
+      "step": 6099
+    },
+    {
+      "epoch": 0.2666200445823681,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.349938513778613e-05,
+      "loss": 1.6064,
+      "step": 6100
+    },
+    {
+      "epoch": 0.26666375278639803,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.349428571147717e-05,
+      "loss": 1.8313,
+      "step": 6101
+    },
+    {
+      "epoch": 0.2667074609904279,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.348918565308226e-05,
+      "loss": 3.0004,
+      "step": 6102
+    },
+    {
+      "epoch": 0.2667511691944578,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.34840849626977e-05,
+      "loss": 1.8625,
+      "step": 6103
+    },
+    {
+      "epoch": 0.2667948773984877,
+      "grad_norm": 3.140625,
+      "learning_rate": 8.347898364041973e-05,
+      "loss": 2.0085,
+      "step": 6104
+    },
+    {
+      "epoch": 0.2668385856025176,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.34738816863446e-05,
+      "loss": 2.2929,
+      "step": 6105
+    },
+    {
+      "epoch": 0.2668822938065475,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.34687791005686e-05,
+      "loss": 2.3305,
+      "step": 6106
+    },
+    {
+      "epoch": 0.26692600201057737,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.346367588318804e-05,
+      "loss": 1.8152,
+      "step": 6107
+    },
+    {
+      "epoch": 0.2669697102146073,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.345857203429919e-05,
+      "loss": 2.0315,
+      "step": 6108
+    },
+    {
+      "epoch": 0.2670134184186372,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.345346755399841e-05,
+      "loss": 2.0062,
+      "step": 6109
+    },
+    {
+      "epoch": 0.26705712662266706,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.344836244238199e-05,
+      "loss": 2.1459,
+      "step": 6110
+    },
+    {
+      "epoch": 0.267100834826697,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.344325669954631e-05,
+      "loss": 1.3544,
+      "step": 6111
+    },
+    {
+      "epoch": 0.2671445430307269,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.343815032558768e-05,
+      "loss": 2.2826,
+      "step": 6112
+    },
+    {
+      "epoch": 0.26718825123475676,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.34330433206025e-05,
+      "loss": 1.2367,
+      "step": 6113
+    },
+    {
+      "epoch": 0.26723195943878664,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.342793568468713e-05,
+      "loss": 1.9825,
+      "step": 6114
+    },
+    {
+      "epoch": 0.2672756676428166,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.342282741793796e-05,
+      "loss": 1.8248,
+      "step": 6115
+    },
+    {
+      "epoch": 0.26731937584684645,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.34177185204514e-05,
+      "loss": 1.5443,
+      "step": 6116
+    },
+    {
+      "epoch": 0.26736308405087633,
+      "grad_norm": 3.0,
+      "learning_rate": 8.341260899232383e-05,
+      "loss": 2.6857,
+      "step": 6117
+    },
+    {
+      "epoch": 0.26740679225490627,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.340749883365174e-05,
+      "loss": 1.6376,
+      "step": 6118
+    },
+    {
+      "epoch": 0.26745050045893615,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.340238804453148e-05,
+      "loss": 2.387,
+      "step": 6119
+    },
+    {
+      "epoch": 0.267494208662966,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.339727662505957e-05,
+      "loss": 1.637,
+      "step": 6120
+    },
+    {
+      "epoch": 0.26753791686699596,
+      "grad_norm": 2.625,
+      "learning_rate": 8.339216457533244e-05,
+      "loss": 1.8343,
+      "step": 6121
+    },
+    {
+      "epoch": 0.26758162507102584,
+      "grad_norm": 2.375,
+      "learning_rate": 8.338705189544655e-05,
+      "loss": 2.1124,
+      "step": 6122
+    },
+    {
+      "epoch": 0.2676253332750557,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.338193858549842e-05,
+      "loss": 1.2943,
+      "step": 6123
+    },
+    {
+      "epoch": 0.2676690414790856,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.337682464558452e-05,
+      "loss": 1.8715,
+      "step": 6124
+    },
+    {
+      "epoch": 0.26771274968311554,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.337171007580135e-05,
+      "loss": 1.9677,
+      "step": 6125
+    },
+    {
+      "epoch": 0.2677564578871454,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.336659487624545e-05,
+      "loss": 2.3621,
+      "step": 6126
+    },
+    {
+      "epoch": 0.2678001660911753,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.336147904701332e-05,
+      "loss": 1.705,
+      "step": 6127
+    },
+    {
+      "epoch": 0.26784387429520523,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.335636258820155e-05,
+      "loss": 1.5326,
+      "step": 6128
+    },
+    {
+      "epoch": 0.2678875824992351,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.335124549990667e-05,
+      "loss": 2.4204,
+      "step": 6129
+    },
+    {
+      "epoch": 0.267931290703265,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.334612778222524e-05,
+      "loss": 1.5847,
+      "step": 6130
+    },
+    {
+      "epoch": 0.2679749989072949,
+      "grad_norm": 2.125,
+      "learning_rate": 8.334100943525385e-05,
+      "loss": 1.3181,
+      "step": 6131
+    },
+    {
+      "epoch": 0.2680187071113248,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.333589045908907e-05,
+      "loss": 2.1265,
+      "step": 6132
+    },
+    {
+      "epoch": 0.2680624153153547,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.333077085382752e-05,
+      "loss": 1.4887,
+      "step": 6133
+    },
+    {
+      "epoch": 0.26810612351938456,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.332565061956584e-05,
+      "loss": 1.9686,
+      "step": 6134
+    },
+    {
+      "epoch": 0.2681498317234145,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.332052975640061e-05,
+      "loss": 1.8269,
+      "step": 6135
+    },
+    {
+      "epoch": 0.2681935399274444,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.33154082644285e-05,
+      "loss": 1.8044,
+      "step": 6136
+    },
+    {
+      "epoch": 0.26823724813147426,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.331028614374612e-05,
+      "loss": 1.8124,
+      "step": 6137
+    },
+    {
+      "epoch": 0.2682809563355042,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.330516339445018e-05,
+      "loss": 2.2855,
+      "step": 6138
+    },
+    {
+      "epoch": 0.2683246645395341,
+      "grad_norm": 2.25,
+      "learning_rate": 8.330004001663733e-05,
+      "loss": 1.9794,
+      "step": 6139
+    },
+    {
+      "epoch": 0.26836837274356395,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.329491601040427e-05,
+      "loss": 2.4759,
+      "step": 6140
+    },
+    {
+      "epoch": 0.2684120809475939,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.328979137584767e-05,
+      "loss": 1.6447,
+      "step": 6141
+    },
+    {
+      "epoch": 0.26845578915162377,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.328466611306427e-05,
+      "loss": 2.2443,
+      "step": 6142
+    },
+    {
+      "epoch": 0.26849949735565365,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.327954022215076e-05,
+      "loss": 1.942,
+      "step": 6143
+    },
+    {
+      "epoch": 0.26854320555968353,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.32744137032039e-05,
+      "loss": 1.8281,
+      "step": 6144
+    },
+    {
+      "epoch": 0.26858691376371346,
+      "grad_norm": 3.6875,
+      "learning_rate": 8.326928655632043e-05,
+      "loss": 1.2905,
+      "step": 6145
+    },
+    {
+      "epoch": 0.26863062196774334,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.32641587815971e-05,
+      "loss": 1.9108,
+      "step": 6146
+    },
+    {
+      "epoch": 0.2686743301717732,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.325903037913067e-05,
+      "loss": 2.4243,
+      "step": 6147
+    },
+    {
+      "epoch": 0.26871803837580316,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.325390134901794e-05,
+      "loss": 1.801,
+      "step": 6148
+    },
+    {
+      "epoch": 0.26876174657983304,
+      "grad_norm": 2.125,
+      "learning_rate": 8.324877169135569e-05,
+      "loss": 1.8057,
+      "step": 6149
+    },
+    {
+      "epoch": 0.2688054547838629,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.324364140624073e-05,
+      "loss": 2.0626,
+      "step": 6150
+    },
+    {
+      "epoch": 0.26884916298789285,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.323851049376987e-05,
+      "loss": 1.8675,
+      "step": 6151
+    },
+    {
+      "epoch": 0.26889287119192273,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.323337895403994e-05,
+      "loss": 1.7426,
+      "step": 6152
+    },
+    {
+      "epoch": 0.2689365793959526,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.322824678714776e-05,
+      "loss": 1.7979,
+      "step": 6153
+    },
+    {
+      "epoch": 0.2689802875999825,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.322311399319023e-05,
+      "loss": 1.97,
+      "step": 6154
+    },
+    {
+      "epoch": 0.26902399580401243,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.321798057226417e-05,
+      "loss": 1.6476,
+      "step": 6155
+    },
+    {
+      "epoch": 0.2690677040080423,
+      "grad_norm": 2.0,
+      "learning_rate": 8.32128465244665e-05,
+      "loss": 1.8328,
+      "step": 6156
+    },
+    {
+      "epoch": 0.2691114122120722,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.320771184989404e-05,
+      "loss": 1.6689,
+      "step": 6157
+    },
+    {
+      "epoch": 0.2691551204161021,
+      "grad_norm": 2.75,
+      "learning_rate": 8.320257654864374e-05,
+      "loss": 1.8429,
+      "step": 6158
+    },
+    {
+      "epoch": 0.269198828620132,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.31974406208125e-05,
+      "loss": 1.3874,
+      "step": 6159
+    },
+    {
+      "epoch": 0.2692425368241619,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.319230406649723e-05,
+      "loss": 2.851,
+      "step": 6160
+    },
+    {
+      "epoch": 0.2692862450281918,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.318716688579488e-05,
+      "loss": 1.8165,
+      "step": 6161
+    },
+    {
+      "epoch": 0.2693299532322217,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.31820290788024e-05,
+      "loss": 1.682,
+      "step": 6162
+    },
+    {
+      "epoch": 0.2693736614362516,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.317689064561671e-05,
+      "loss": 1.5811,
+      "step": 6163
+    },
+    {
+      "epoch": 0.26941736964028146,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.317175158633483e-05,
+      "loss": 1.8027,
+      "step": 6164
+    },
+    {
+      "epoch": 0.2694610778443114,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.316661190105371e-05,
+      "loss": 1.8132,
+      "step": 6165
+    },
+    {
+      "epoch": 0.26950478604834127,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.316147158987036e-05,
+      "loss": 1.6619,
+      "step": 6166
+    },
+    {
+      "epoch": 0.26954849425237115,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.315633065288176e-05,
+      "loss": 1.9372,
+      "step": 6167
+    },
+    {
+      "epoch": 0.2695922024564011,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.315118909018495e-05,
+      "loss": 2.3095,
+      "step": 6168
+    },
+    {
+      "epoch": 0.26963591066043097,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.314604690187695e-05,
+      "loss": 1.9157,
+      "step": 6169
+    },
+    {
+      "epoch": 0.26967961886446085,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.314090408805482e-05,
+      "loss": 2.1959,
+      "step": 6170
+    },
+    {
+      "epoch": 0.2697233270684908,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.313576064881558e-05,
+      "loss": 1.7974,
+      "step": 6171
+    },
+    {
+      "epoch": 0.26976703527252066,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.313061658425632e-05,
+      "loss": 2.098,
+      "step": 6172
+    },
+    {
+      "epoch": 0.26981074347655054,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.312547189447409e-05,
+      "loss": 1.4758,
+      "step": 6173
+    },
+    {
+      "epoch": 0.2698544516805804,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.312032657956599e-05,
+      "loss": 2.3724,
+      "step": 6174
+    },
+    {
+      "epoch": 0.26989815988461036,
+      "grad_norm": 2.875,
+      "learning_rate": 8.311518063962914e-05,
+      "loss": 2.2493,
+      "step": 6175
+    },
+    {
+      "epoch": 0.26994186808864024,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.311003407476064e-05,
+      "loss": 2.4625,
+      "step": 6176
+    },
+    {
+      "epoch": 0.2699855762926701,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.31048868850576e-05,
+      "loss": 1.5971,
+      "step": 6177
+    },
+    {
+      "epoch": 0.27002928449670005,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.309973907061715e-05,
+      "loss": 2.2358,
+      "step": 6178
+    },
+    {
+      "epoch": 0.27007299270072993,
+      "grad_norm": 2.25,
+      "learning_rate": 8.309459063153646e-05,
+      "loss": 1.8094,
+      "step": 6179
+    },
+    {
+      "epoch": 0.2701167009047598,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.308944156791268e-05,
+      "loss": 1.8803,
+      "step": 6180
+    },
+    {
+      "epoch": 0.27016040910878975,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.308429187984297e-05,
+      "loss": 1.9824,
+      "step": 6181
+    },
+    {
+      "epoch": 0.2702041173128196,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.307914156742454e-05,
+      "loss": 1.7535,
+      "step": 6182
+    },
+    {
+      "epoch": 0.2702478255168495,
+      "grad_norm": 4.3125,
+      "learning_rate": 8.307399063075453e-05,
+      "loss": 1.4322,
+      "step": 6183
+    },
+    {
+      "epoch": 0.2702915337208794,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.306883906993022e-05,
+      "loss": 1.9071,
+      "step": 6184
+    },
+    {
+      "epoch": 0.2703352419249093,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.306368688504876e-05,
+      "loss": 2.7365,
+      "step": 6185
+    },
+    {
+      "epoch": 0.2703789501289392,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.305853407620742e-05,
+      "loss": 1.8138,
+      "step": 6186
+    },
+    {
+      "epoch": 0.2704226583329691,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.305338064350342e-05,
+      "loss": 2.2551,
+      "step": 6187
+    },
+    {
+      "epoch": 0.270466366536999,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.304822658703402e-05,
+      "loss": 1.7694,
+      "step": 6188
+    },
+    {
+      "epoch": 0.2705100747410289,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.304307190689649e-05,
+      "loss": 1.9035,
+      "step": 6189
+    },
+    {
+      "epoch": 0.2705537829450588,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.30379166031881e-05,
+      "loss": 2.906,
+      "step": 6190
+    },
+    {
+      "epoch": 0.2705974911490887,
+      "grad_norm": 2.5,
+      "learning_rate": 8.303276067600614e-05,
+      "loss": 1.8274,
+      "step": 6191
+    },
+    {
+      "epoch": 0.2706411993531186,
+      "grad_norm": 4.8125,
+      "learning_rate": 8.30276041254479e-05,
+      "loss": 1.8758,
+      "step": 6192
+    },
+    {
+      "epoch": 0.27068490755714847,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.30224469516107e-05,
+      "loss": 1.784,
+      "step": 6193
+    },
+    {
+      "epoch": 0.27072861576117835,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.301728915459188e-05,
+      "loss": 1.81,
+      "step": 6194
+    },
+    {
+      "epoch": 0.2707723239652083,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.301213073448874e-05,
+      "loss": 1.76,
+      "step": 6195
+    },
+    {
+      "epoch": 0.27081603216923816,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.300697169139867e-05,
+      "loss": 1.9981,
+      "step": 6196
+    },
+    {
+      "epoch": 0.27085974037326804,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.300181202541899e-05,
+      "loss": 2.1087,
+      "step": 6197
+    },
+    {
+      "epoch": 0.270903448577298,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.299665173664708e-05,
+      "loss": 2.4448,
+      "step": 6198
+    },
+    {
+      "epoch": 0.27094715678132786,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.299149082518034e-05,
+      "loss": 2.4269,
+      "step": 6199
+    },
+    {
+      "epoch": 0.27099086498535774,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.298632929111614e-05,
+      "loss": 2.4778,
+      "step": 6200
+    },
+    {
+      "epoch": 0.2710345731893877,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.298116713455191e-05,
+      "loss": 1.7866,
+      "step": 6201
+    },
+    {
+      "epoch": 0.27107828139341755,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.297600435558505e-05,
+      "loss": 1.9275,
+      "step": 6202
+    },
+    {
+      "epoch": 0.27112198959744743,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.297084095431298e-05,
+      "loss": 2.4905,
+      "step": 6203
+    },
+    {
+      "epoch": 0.2711656978014773,
+      "grad_norm": 2.625,
+      "learning_rate": 8.296567693083317e-05,
+      "loss": 2.0117,
+      "step": 6204
+    },
+    {
+      "epoch": 0.27120940600550725,
+      "grad_norm": 1.859375,
+      "learning_rate": 8.296051228524305e-05,
+      "loss": 1.6825,
+      "step": 6205
+    },
+    {
+      "epoch": 0.2712531142095371,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.29553470176401e-05,
+      "loss": 2.1916,
+      "step": 6206
+    },
+    {
+      "epoch": 0.271296822413567,
+      "grad_norm": 1.859375,
+      "learning_rate": 8.295018112812179e-05,
+      "loss": 1.6755,
+      "step": 6207
+    },
+    {
+      "epoch": 0.27134053061759694,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.294501461678559e-05,
+      "loss": 1.9042,
+      "step": 6208
+    },
+    {
+      "epoch": 0.2713842388216268,
+      "grad_norm": 4.15625,
+      "learning_rate": 8.293984748372901e-05,
+      "loss": 2.5244,
+      "step": 6209
+    },
+    {
+      "epoch": 0.2714279470256567,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.293467972904958e-05,
+      "loss": 2.1528,
+      "step": 6210
+    },
+    {
+      "epoch": 0.27147165522968664,
+      "grad_norm": 2.125,
+      "learning_rate": 8.292951135284481e-05,
+      "loss": 1.8469,
+      "step": 6211
+    },
+    {
+      "epoch": 0.2715153634337165,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.292434235521222e-05,
+      "loss": 1.7206,
+      "step": 6212
+    },
+    {
+      "epoch": 0.2715590716377464,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.291917273624939e-05,
+      "loss": 1.7422,
+      "step": 6213
+    },
+    {
+      "epoch": 0.2716027798417763,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.291400249605386e-05,
+      "loss": 2.3142,
+      "step": 6214
+    },
+    {
+      "epoch": 0.2716464880458062,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.290883163472318e-05,
+      "loss": 2.354,
+      "step": 6215
+    },
+    {
+      "epoch": 0.2716901962498361,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.290366015235497e-05,
+      "loss": 1.3597,
+      "step": 6216
+    },
+    {
+      "epoch": 0.27173390445386597,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.289848804904678e-05,
+      "loss": 2.4125,
+      "step": 6217
+    },
+    {
+      "epoch": 0.2717776126578959,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.289331532489626e-05,
+      "loss": 1.9229,
+      "step": 6218
+    },
+    {
+      "epoch": 0.2718213208619258,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.2888141980001e-05,
+      "loss": 1.9399,
+      "step": 6219
+    },
+    {
+      "epoch": 0.27186502906595567,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.288296801445863e-05,
+      "loss": 2.0839,
+      "step": 6220
+    },
+    {
+      "epoch": 0.2719087372699856,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.287779342836682e-05,
+      "loss": 1.7229,
+      "step": 6221
+    },
+    {
+      "epoch": 0.2719524454740155,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.287261822182316e-05,
+      "loss": 1.7666,
+      "step": 6222
+    },
+    {
+      "epoch": 0.27199615367804536,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.286744239492537e-05,
+      "loss": 1.9596,
+      "step": 6223
+    },
+    {
+      "epoch": 0.27203986188207524,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.28622659477711e-05,
+      "loss": 1.8435,
+      "step": 6224
+    },
+    {
+      "epoch": 0.2720835700861052,
+      "grad_norm": 2.25,
+      "learning_rate": 8.285708888045803e-05,
+      "loss": 1.8474,
+      "step": 6225
+    },
+    {
+      "epoch": 0.27212727829013506,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.285191119308389e-05,
+      "loss": 2.607,
+      "step": 6226
+    },
+    {
+      "epoch": 0.27217098649416493,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.284673288574637e-05,
+      "loss": 1.9533,
+      "step": 6227
+    },
+    {
+      "epoch": 0.27221469469819487,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.284155395854318e-05,
+      "loss": 1.8539,
+      "step": 6228
+    },
+    {
+      "epoch": 0.27225840290222475,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.283637441157209e-05,
+      "loss": 1.8384,
+      "step": 6229
+    },
+    {
+      "epoch": 0.27230211110625463,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.283119424493082e-05,
+      "loss": 1.7905,
+      "step": 6230
+    },
+    {
+      "epoch": 0.27234581931028456,
+      "grad_norm": 2.0,
+      "learning_rate": 8.282601345871713e-05,
+      "loss": 2.0853,
+      "step": 6231
+    },
+    {
+      "epoch": 0.27238952751431444,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.282083205302877e-05,
+      "loss": 2.1163,
+      "step": 6232
+    },
+    {
+      "epoch": 0.2724332357183443,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.281565002796356e-05,
+      "loss": 1.4279,
+      "step": 6233
+    },
+    {
+      "epoch": 0.2724769439223742,
+      "grad_norm": 2.125,
+      "learning_rate": 8.281046738361926e-05,
+      "loss": 2.0867,
+      "step": 6234
+    },
+    {
+      "epoch": 0.27252065212640414,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.28052841200937e-05,
+      "loss": 2.2017,
+      "step": 6235
+    },
+    {
+      "epoch": 0.272564360330434,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.280010023748467e-05,
+      "loss": 2.464,
+      "step": 6236
+    },
+    {
+      "epoch": 0.2726080685344639,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.279491573589004e-05,
+      "loss": 1.7016,
+      "step": 6237
+    },
+    {
+      "epoch": 0.27265177673849383,
+      "grad_norm": 2.125,
+      "learning_rate": 8.278973061540758e-05,
+      "loss": 1.849,
+      "step": 6238
+    },
+    {
+      "epoch": 0.2726954849425237,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.27845448761352e-05,
+      "loss": 2.3198,
+      "step": 6239
+    },
+    {
+      "epoch": 0.2727391931465536,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.277935851817075e-05,
+      "loss": 2.0965,
+      "step": 6240
+    },
+    {
+      "epoch": 0.27278290135058353,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.277417154161208e-05,
+      "loss": 1.7996,
+      "step": 6241
+    },
+    {
+      "epoch": 0.2728266095546134,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.27689839465571e-05,
+      "loss": 1.7935,
+      "step": 6242
+    },
+    {
+      "epoch": 0.2728703177586433,
+      "grad_norm": 2.5,
+      "learning_rate": 8.27637957331037e-05,
+      "loss": 2.6535,
+      "step": 6243
+    },
+    {
+      "epoch": 0.27291402596267317,
+      "grad_norm": 2.5,
+      "learning_rate": 8.275860690134978e-05,
+      "loss": 1.6348,
+      "step": 6244
+    },
+    {
+      "epoch": 0.2729577341667031,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.275341745139329e-05,
+      "loss": 2.5961,
+      "step": 6245
+    },
+    {
+      "epoch": 0.273001442370733,
+      "grad_norm": 2.25,
+      "learning_rate": 8.274822738333214e-05,
+      "loss": 1.988,
+      "step": 6246
+    },
+    {
+      "epoch": 0.27304515057476286,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.274303669726426e-05,
+      "loss": 1.7639,
+      "step": 6247
+    },
+    {
+      "epoch": 0.2730888587787928,
+      "grad_norm": 1.890625,
+      "learning_rate": 8.273784539328763e-05,
+      "loss": 1.5711,
+      "step": 6248
+    },
+    {
+      "epoch": 0.2731325669828227,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.273265347150021e-05,
+      "loss": 1.4728,
+      "step": 6249
+    },
+    {
+      "epoch": 0.27317627518685256,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.272746093199996e-05,
+      "loss": 1.6939,
+      "step": 6250
+    },
+    {
+      "epoch": 0.2732199833908825,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.272226777488492e-05,
+      "loss": 1.7364,
+      "step": 6251
+    },
+    {
+      "epoch": 0.2732636915949124,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.271707400025306e-05,
+      "loss": 1.8905,
+      "step": 6252
+    },
+    {
+      "epoch": 0.27330739979894225,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.271187960820237e-05,
+      "loss": 2.2283,
+      "step": 6253
+    },
+    {
+      "epoch": 0.27335110800297213,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.270668459883093e-05,
+      "loss": 2.0876,
+      "step": 6254
+    },
+    {
+      "epoch": 0.27339481620700207,
+      "grad_norm": 2.0,
+      "learning_rate": 8.270148897223673e-05,
+      "loss": 1.9225,
+      "step": 6255
+    },
+    {
+      "epoch": 0.27343852441103195,
+      "grad_norm": 2.0,
+      "learning_rate": 8.269629272851785e-05,
+      "loss": 1.451,
+      "step": 6256
+    },
+    {
+      "epoch": 0.2734822326150618,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.269109586777234e-05,
+      "loss": 1.8739,
+      "step": 6257
+    },
+    {
+      "epoch": 0.27352594081909176,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.268589839009826e-05,
+      "loss": 1.9135,
+      "step": 6258
+    },
+    {
+      "epoch": 0.27356964902312164,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.268070029559372e-05,
+      "loss": 1.818,
+      "step": 6259
+    },
+    {
+      "epoch": 0.2736133572271515,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.267550158435679e-05,
+      "loss": 2.6027,
+      "step": 6260
+    },
+    {
+      "epoch": 0.27365706543118146,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.26703022564856e-05,
+      "loss": 2.0903,
+      "step": 6261
+    },
+    {
+      "epoch": 0.27370077363521134,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.266510231207824e-05,
+      "loss": 2.4899,
+      "step": 6262
+    },
+    {
+      "epoch": 0.2737444818392412,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.265990175123286e-05,
+      "loss": 1.5422,
+      "step": 6263
+    },
+    {
+      "epoch": 0.2737881900432711,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.265470057404761e-05,
+      "loss": 2.1211,
+      "step": 6264
+    },
+    {
+      "epoch": 0.27383189824730103,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.264949878062062e-05,
+      "loss": 2.0299,
+      "step": 6265
+    },
+    {
+      "epoch": 0.2738756064513309,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.264429637105009e-05,
+      "loss": 1.7817,
+      "step": 6266
+    },
+    {
+      "epoch": 0.2739193146553608,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.263909334543416e-05,
+      "loss": 1.8252,
+      "step": 6267
+    },
+    {
+      "epoch": 0.2739630228593907,
+      "grad_norm": 2.375,
+      "learning_rate": 8.263388970387102e-05,
+      "loss": 2.439,
+      "step": 6268
+    },
+    {
+      "epoch": 0.2740067310634206,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.262868544645892e-05,
+      "loss": 1.74,
+      "step": 6269
+    },
+    {
+      "epoch": 0.2740504392674505,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.262348057329601e-05,
+      "loss": 1.8924,
+      "step": 6270
+    },
+    {
+      "epoch": 0.2740941474714804,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.261827508448056e-05,
+      "loss": 3.097,
+      "step": 6271
+    },
+    {
+      "epoch": 0.2741378556755103,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.261306898011077e-05,
+      "loss": 1.4289,
+      "step": 6272
+    },
+    {
+      "epoch": 0.2741815638795402,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.26078622602849e-05,
+      "loss": 2.0246,
+      "step": 6273
+    },
+    {
+      "epoch": 0.27422527208357006,
+      "grad_norm": 2.375,
+      "learning_rate": 8.260265492510122e-05,
+      "loss": 2.4173,
+      "step": 6274
+    },
+    {
+      "epoch": 0.2742689802876,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.259744697465799e-05,
+      "loss": 1.5699,
+      "step": 6275
+    },
+    {
+      "epoch": 0.2743126884916299,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.25922384090535e-05,
+      "loss": 1.3334,
+      "step": 6276
+    },
+    {
+      "epoch": 0.27435639669565975,
+      "grad_norm": 2.25,
+      "learning_rate": 8.258702922838603e-05,
+      "loss": 1.9971,
+      "step": 6277
+    },
+    {
+      "epoch": 0.2744001048996897,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.258181943275387e-05,
+      "loss": 1.7544,
+      "step": 6278
+    },
+    {
+      "epoch": 0.27444381310371957,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.257660902225539e-05,
+      "loss": 2.1136,
+      "step": 6279
+    },
+    {
+      "epoch": 0.27448752130774945,
+      "grad_norm": 2.25,
+      "learning_rate": 8.257139799698886e-05,
+      "loss": 1.9613,
+      "step": 6280
+    },
+    {
+      "epoch": 0.2745312295117794,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.256618635705267e-05,
+      "loss": 3.2578,
+      "step": 6281
+    },
+    {
+      "epoch": 0.27457493771580926,
+      "grad_norm": 6.0,
+      "learning_rate": 8.256097410254512e-05,
+      "loss": 1.0849,
+      "step": 6282
+    },
+    {
+      "epoch": 0.27461864591983914,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.25557612335646e-05,
+      "loss": 2.0473,
+      "step": 6283
+    },
+    {
+      "epoch": 0.274662354123869,
+      "grad_norm": 3.125,
+      "learning_rate": 8.255054775020949e-05,
+      "loss": 1.511,
+      "step": 6284
+    },
+    {
+      "epoch": 0.27470606232789896,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.254533365257817e-05,
+      "loss": 1.7947,
+      "step": 6285
+    },
+    {
+      "epoch": 0.27474977053192884,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.254011894076904e-05,
+      "loss": 2.7416,
+      "step": 6286
+    },
+    {
+      "epoch": 0.2747934787359587,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.25349036148805e-05,
+      "loss": 2.099,
+      "step": 6287
+    },
+    {
+      "epoch": 0.27483718693998865,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.252968767501098e-05,
+      "loss": 2.0382,
+      "step": 6288
+    },
+    {
+      "epoch": 0.27488089514401853,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.252447112125889e-05,
+      "loss": 1.3828,
+      "step": 6289
+    },
+    {
+      "epoch": 0.2749246033480484,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.251925395372272e-05,
+      "loss": 1.9086,
+      "step": 6290
+    },
+    {
+      "epoch": 0.27496831155207835,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.251403617250088e-05,
+      "loss": 1.6746,
+      "step": 6291
+    },
+    {
+      "epoch": 0.27501201975610823,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.250881777769187e-05,
+      "loss": 1.8238,
+      "step": 6292
+    },
+    {
+      "epoch": 0.2750557279601381,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.250359876939415e-05,
+      "loss": 2.2168,
+      "step": 6293
+    },
+    {
+      "epoch": 0.275099436164168,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.249837914770621e-05,
+      "loss": 1.8462,
+      "step": 6294
+    },
+    {
+      "epoch": 0.2751431443681979,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.249315891272655e-05,
+      "loss": 1.9032,
+      "step": 6295
+    },
+    {
+      "epoch": 0.2751868525722278,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.24879380645537e-05,
+      "loss": 1.4899,
+      "step": 6296
+    },
+    {
+      "epoch": 0.2752305607762577,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.248271660328618e-05,
+      "loss": 2.1009,
+      "step": 6297
+    },
+    {
+      "epoch": 0.2752742689802876,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.247749452902251e-05,
+      "loss": 1.9969,
+      "step": 6298
+    },
+    {
+      "epoch": 0.2753179771843175,
+      "grad_norm": 2.0,
+      "learning_rate": 8.247227184186126e-05,
+      "loss": 2.2424,
+      "step": 6299
+    },
+    {
+      "epoch": 0.2753616853883474,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.246704854190098e-05,
+      "loss": 1.373,
+      "step": 6300
+    },
+    {
+      "epoch": 0.2754053935923773,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.246182462924023e-05,
+      "loss": 2.7404,
+      "step": 6301
+    },
+    {
+      "epoch": 0.2754491017964072,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.24566001039776e-05,
+      "loss": 1.9392,
+      "step": 6302
+    },
+    {
+      "epoch": 0.27549281000043707,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.245137496621169e-05,
+      "loss": 1.6657,
+      "step": 6303
+    },
+    {
+      "epoch": 0.27553651820446695,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.244614921604111e-05,
+      "loss": 1.6642,
+      "step": 6304
+    },
+    {
+      "epoch": 0.2755802264084969,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.244092285356446e-05,
+      "loss": 1.851,
+      "step": 6305
+    },
+    {
+      "epoch": 0.27562393461252677,
+      "grad_norm": 2.875,
+      "learning_rate": 8.24356958788804e-05,
+      "loss": 1.5946,
+      "step": 6306
+    },
+    {
+      "epoch": 0.27566764281655665,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.243046829208754e-05,
+      "loss": 1.6248,
+      "step": 6307
+    },
+    {
+      "epoch": 0.2757113510205866,
+      "grad_norm": 2.25,
+      "learning_rate": 8.242524009328454e-05,
+      "loss": 2.4631,
+      "step": 6308
+    },
+    {
+      "epoch": 0.27575505922461646,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.242001128257007e-05,
+      "loss": 1.7864,
+      "step": 6309
+    },
+    {
+      "epoch": 0.27579876742864634,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.24147818600428e-05,
+      "loss": 1.8237,
+      "step": 6310
+    },
+    {
+      "epoch": 0.2758424756326763,
+      "grad_norm": 2.125,
+      "learning_rate": 8.240955182580143e-05,
+      "loss": 1.5502,
+      "step": 6311
+    },
+    {
+      "epoch": 0.27588618383670616,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.240432117994464e-05,
+      "loss": 2.0371,
+      "step": 6312
+    },
+    {
+      "epoch": 0.27592989204073604,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.239908992257115e-05,
+      "loss": 1.7174,
+      "step": 6313
+    },
+    {
+      "epoch": 0.2759736002447659,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.239385805377966e-05,
+      "loss": 1.4633,
+      "step": 6314
+    },
+    {
+      "epoch": 0.27601730844879585,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.238862557366893e-05,
+      "loss": 1.6023,
+      "step": 6315
+    },
+    {
+      "epoch": 0.27606101665282573,
+      "grad_norm": 3.65625,
+      "learning_rate": 8.23833924823377e-05,
+      "loss": 3.2681,
+      "step": 6316
+    },
+    {
+      "epoch": 0.2761047248568556,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.237815877988472e-05,
+      "loss": 1.6448,
+      "step": 6317
+    },
+    {
+      "epoch": 0.27614843306088555,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.237292446640877e-05,
+      "loss": 1.7115,
+      "step": 6318
+    },
+    {
+      "epoch": 0.2761921412649154,
+      "grad_norm": 2.125,
+      "learning_rate": 8.236768954200862e-05,
+      "loss": 1.8672,
+      "step": 6319
+    },
+    {
+      "epoch": 0.2762358494689453,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.236245400678304e-05,
+      "loss": 2.6833,
+      "step": 6320
+    },
+    {
+      "epoch": 0.27627955767297524,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.235721786083087e-05,
+      "loss": 2.5507,
+      "step": 6321
+    },
+    {
+      "epoch": 0.2763232658770051,
+      "grad_norm": 2.125,
+      "learning_rate": 8.23519811042509e-05,
+      "loss": 1.4608,
+      "step": 6322
+    },
+    {
+      "epoch": 0.276366974081035,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.234674373714196e-05,
+      "loss": 1.4405,
+      "step": 6323
+    },
+    {
+      "epoch": 0.2764106822850649,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.234150575960288e-05,
+      "loss": 1.7559,
+      "step": 6324
+    },
+    {
+      "epoch": 0.2764543904890948,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.233626717173251e-05,
+      "loss": 2.0372,
+      "step": 6325
+    },
+    {
+      "epoch": 0.2764980986931247,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.233102797362974e-05,
+      "loss": 2.3649,
+      "step": 6326
+    },
+    {
+      "epoch": 0.2765418068971546,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.23257881653934e-05,
+      "loss": 1.9945,
+      "step": 6327
+    },
+    {
+      "epoch": 0.2765855151011845,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.232054774712238e-05,
+      "loss": 1.8691,
+      "step": 6328
+    },
+    {
+      "epoch": 0.2766292233052144,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.23153067189156e-05,
+      "loss": 1.6846,
+      "step": 6329
+    },
+    {
+      "epoch": 0.27667293150924427,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.231006508087196e-05,
+      "loss": 1.7247,
+      "step": 6330
+    },
+    {
+      "epoch": 0.2767166397132742,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.230482283309035e-05,
+      "loss": 2.3358,
+      "step": 6331
+    },
+    {
+      "epoch": 0.2767603479173041,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.229957997566971e-05,
+      "loss": 1.9732,
+      "step": 6332
+    },
+    {
+      "epoch": 0.27680405612133396,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.2294336508709e-05,
+      "loss": 1.8837,
+      "step": 6333
+    },
+    {
+      "epoch": 0.27684776432536384,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.228909243230714e-05,
+      "loss": 1.7376,
+      "step": 6334
+    },
+    {
+      "epoch": 0.2768914725293938,
+      "grad_norm": 2.25,
+      "learning_rate": 8.228384774656312e-05,
+      "loss": 1.9227,
+      "step": 6335
+    },
+    {
+      "epoch": 0.27693518073342366,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.227860245157593e-05,
+      "loss": 1.8045,
+      "step": 6336
+    },
+    {
+      "epoch": 0.27697888893745354,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.22733565474445e-05,
+      "loss": 2.3304,
+      "step": 6337
+    },
+    {
+      "epoch": 0.2770225971414835,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.226811003426788e-05,
+      "loss": 1.5063,
+      "step": 6338
+    },
+    {
+      "epoch": 0.27706630534551335,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.226286291214504e-05,
+      "loss": 1.6672,
+      "step": 6339
+    },
+    {
+      "epoch": 0.27711001354954323,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.225761518117503e-05,
+      "loss": 1.7877,
+      "step": 6340
+    },
+    {
+      "epoch": 0.27715372175357317,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.225236684145688e-05,
+      "loss": 1.9904,
+      "step": 6341
+    },
+    {
+      "epoch": 0.27719742995760305,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.224711789308963e-05,
+      "loss": 1.7372,
+      "step": 6342
+    },
+    {
+      "epoch": 0.2772411381616329,
+      "grad_norm": 3.390625,
+      "learning_rate": 8.224186833617231e-05,
+      "loss": 2.7278,
+      "step": 6343
+    },
+    {
+      "epoch": 0.27728484636566286,
+      "grad_norm": 3.53125,
+      "learning_rate": 8.223661817080403e-05,
+      "loss": 2.5373,
+      "step": 6344
+    },
+    {
+      "epoch": 0.27732855456969274,
+      "grad_norm": 2.125,
+      "learning_rate": 8.223136739708383e-05,
+      "loss": 1.7289,
+      "step": 6345
+    },
+    {
+      "epoch": 0.2773722627737226,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.222611601511083e-05,
+      "loss": 1.7691,
+      "step": 6346
+    },
+    {
+      "epoch": 0.2774159709777525,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.222086402498412e-05,
+      "loss": 2.8211,
+      "step": 6347
+    },
+    {
+      "epoch": 0.27745967918178244,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.221561142680281e-05,
+      "loss": 1.4104,
+      "step": 6348
+    },
+    {
+      "epoch": 0.2775033873858123,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.221035822066601e-05,
+      "loss": 2.357,
+      "step": 6349
+    },
+    {
+      "epoch": 0.2775470955898422,
+      "grad_norm": 2.125,
+      "learning_rate": 8.220510440667289e-05,
+      "loss": 1.7921,
+      "step": 6350
+    },
+    {
+      "epoch": 0.27759080379387213,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.219984998492256e-05,
+      "loss": 1.8645,
+      "step": 6351
+    },
+    {
+      "epoch": 0.277634511997902,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.219459495551421e-05,
+      "loss": 1.8275,
+      "step": 6352
+    },
+    {
+      "epoch": 0.2776782202019319,
+      "grad_norm": 2.0,
+      "learning_rate": 8.218933931854697e-05,
+      "loss": 1.5563,
+      "step": 6353
+    },
+    {
+      "epoch": 0.2777219284059618,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.218408307412006e-05,
+      "loss": 1.6998,
+      "step": 6354
+    },
+    {
+      "epoch": 0.2777656366099917,
+      "grad_norm": 2.125,
+      "learning_rate": 8.217882622233268e-05,
+      "loss": 1.4991,
+      "step": 6355
+    },
+    {
+      "epoch": 0.2778093448140216,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.217356876328402e-05,
+      "loss": 2.0588,
+      "step": 6356
+    },
+    {
+      "epoch": 0.27785305301805147,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.216831069707326e-05,
+      "loss": 2.2262,
+      "step": 6357
+    },
+    {
+      "epoch": 0.2778967612220814,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.216305202379968e-05,
+      "loss": 1.8249,
+      "step": 6358
+    },
+    {
+      "epoch": 0.2779404694261113,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.215779274356248e-05,
+      "loss": 1.7044,
+      "step": 6359
+    },
+    {
+      "epoch": 0.27798417763014116,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.215253285646094e-05,
+      "loss": 1.7547,
+      "step": 6360
+    },
+    {
+      "epoch": 0.2780278858341711,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.214727236259431e-05,
+      "loss": 1.8101,
+      "step": 6361
+    },
+    {
+      "epoch": 0.278071594038201,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.214201126206184e-05,
+      "loss": 2.1843,
+      "step": 6362
+    },
+    {
+      "epoch": 0.27811530224223086,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.213674955496287e-05,
+      "loss": 2.5314,
+      "step": 6363
+    },
+    {
+      "epoch": 0.2781590104462608,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.213148724139665e-05,
+      "loss": 1.6071,
+      "step": 6364
+    },
+    {
+      "epoch": 0.27820271865029067,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.212622432146248e-05,
+      "loss": 1.8472,
+      "step": 6365
+    },
+    {
+      "epoch": 0.27824642685432055,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.212096079525974e-05,
+      "loss": 1.2409,
+      "step": 6366
+    },
+    {
+      "epoch": 0.27829013505835043,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.211569666288769e-05,
+      "loss": 1.8642,
+      "step": 6367
+    },
+    {
+      "epoch": 0.27833384326238036,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.211043192444572e-05,
+      "loss": 2.009,
+      "step": 6368
+    },
+    {
+      "epoch": 0.27837755146641024,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.210516658003316e-05,
+      "loss": 1.6596,
+      "step": 6369
+    },
+    {
+      "epoch": 0.2784212596704401,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.209990062974936e-05,
+      "loss": 1.956,
+      "step": 6370
+    },
+    {
+      "epoch": 0.27846496787447006,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.209463407369373e-05,
+      "loss": 2.223,
+      "step": 6371
+    },
+    {
+      "epoch": 0.27850867607849994,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.208936691196565e-05,
+      "loss": 1.4611,
+      "step": 6372
+    },
+    {
+      "epoch": 0.2785523842825298,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.20840991446645e-05,
+      "loss": 1.7562,
+      "step": 6373
+    },
+    {
+      "epoch": 0.27859609248655975,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.207883077188971e-05,
+      "loss": 1.8233,
+      "step": 6374
+    },
+    {
+      "epoch": 0.27863980069058963,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.20735617937407e-05,
+      "loss": 2.423,
+      "step": 6375
+    },
+    {
+      "epoch": 0.2786835088946195,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.20682922103169e-05,
+      "loss": 1.5783,
+      "step": 6376
+    },
+    {
+      "epoch": 0.2787272170986494,
+      "grad_norm": 2.25,
+      "learning_rate": 8.206302202171775e-05,
+      "loss": 1.7583,
+      "step": 6377
+    },
+    {
+      "epoch": 0.27877092530267933,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.20577512280427e-05,
+      "loss": 1.9529,
+      "step": 6378
+    },
+    {
+      "epoch": 0.2788146335067092,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.205247982939123e-05,
+      "loss": 2.1829,
+      "step": 6379
+    },
+    {
+      "epoch": 0.2788583417107391,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.204720782586281e-05,
+      "loss": 1.5143,
+      "step": 6380
+    },
+    {
+      "epoch": 0.278902049914769,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.204193521755694e-05,
+      "loss": 1.2953,
+      "step": 6381
+    },
+    {
+      "epoch": 0.2789457581187989,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.203666200457313e-05,
+      "loss": 2.1489,
+      "step": 6382
+    },
+    {
+      "epoch": 0.2789894663228288,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.203138818701087e-05,
+      "loss": 2.1649,
+      "step": 6383
+    },
+    {
+      "epoch": 0.2790331745268587,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.20261137649697e-05,
+      "loss": 1.788,
+      "step": 6384
+    },
+    {
+      "epoch": 0.2790768827308886,
+      "grad_norm": 1.859375,
+      "learning_rate": 8.202083873854914e-05,
+      "loss": 1.6652,
+      "step": 6385
+    },
+    {
+      "epoch": 0.2791205909349185,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.201556310784877e-05,
+      "loss": 2.2077,
+      "step": 6386
+    },
+    {
+      "epoch": 0.27916429913894836,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.20102868729681e-05,
+      "loss": 2.984,
+      "step": 6387
+    },
+    {
+      "epoch": 0.2792080073429783,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.200501003400676e-05,
+      "loss": 2.2747,
+      "step": 6388
+    },
+    {
+      "epoch": 0.2792517155470082,
+      "grad_norm": 2.0,
+      "learning_rate": 8.199973259106427e-05,
+      "loss": 1.7683,
+      "step": 6389
+    },
+    {
+      "epoch": 0.27929542375103805,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.199445454424026e-05,
+      "loss": 1.855,
+      "step": 6390
+    },
+    {
+      "epoch": 0.279339131955068,
+      "grad_norm": 2.0,
+      "learning_rate": 8.198917589363432e-05,
+      "loss": 1.5856,
+      "step": 6391
+    },
+    {
+      "epoch": 0.27938284015909787,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.198389663934611e-05,
+      "loss": 1.5465,
+      "step": 6392
+    },
+    {
+      "epoch": 0.27942654836312775,
+      "grad_norm": 2.25,
+      "learning_rate": 8.197861678147518e-05,
+      "loss": 1.8482,
+      "step": 6393
+    },
+    {
+      "epoch": 0.2794702565671577,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.197333632012123e-05,
+      "loss": 2.3386,
+      "step": 6394
+    },
+    {
+      "epoch": 0.27951396477118756,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.196805525538389e-05,
+      "loss": 2.0954,
+      "step": 6395
+    },
+    {
+      "epoch": 0.27955767297521744,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.19627735873628e-05,
+      "loss": 1.5945,
+      "step": 6396
+    },
+    {
+      "epoch": 0.2796013811792473,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.195749131615767e-05,
+      "loss": 1.8132,
+      "step": 6397
+    },
+    {
+      "epoch": 0.27964508938327726,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.195220844186817e-05,
+      "loss": 1.6905,
+      "step": 6398
+    },
+    {
+      "epoch": 0.27968879758730714,
+      "grad_norm": 2.25,
+      "learning_rate": 8.194692496459398e-05,
+      "loss": 2.4612,
+      "step": 6399
+    },
+    {
+      "epoch": 0.279732505791337,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.194164088443482e-05,
+      "loss": 1.7586,
+      "step": 6400
+    },
+    {
+      "epoch": 0.27977621399536695,
+      "grad_norm": 5.59375,
+      "learning_rate": 8.19363562014904e-05,
+      "loss": 2.1648,
+      "step": 6401
+    },
+    {
+      "epoch": 0.27981992219939683,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.193107091586048e-05,
+      "loss": 1.8121,
+      "step": 6402
+    },
+    {
+      "epoch": 0.2798636304034267,
+      "grad_norm": 3.125,
+      "learning_rate": 8.192578502764476e-05,
+      "loss": 2.687,
+      "step": 6403
+    },
+    {
+      "epoch": 0.27990733860745665,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.192049853694301e-05,
+      "loss": 1.8532,
+      "step": 6404
+    },
+    {
+      "epoch": 0.2799510468114865,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.191521144385501e-05,
+      "loss": 1.6634,
+      "step": 6405
+    },
+    {
+      "epoch": 0.2799947550155164,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.190992374848052e-05,
+      "loss": 1.8239,
+      "step": 6406
+    },
+    {
+      "epoch": 0.2800384632195463,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.190463545091931e-05,
+      "loss": 1.9444,
+      "step": 6407
+    },
+    {
+      "epoch": 0.2800821714235762,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.189934655127121e-05,
+      "loss": 1.6436,
+      "step": 6408
+    },
+    {
+      "epoch": 0.2801258796276061,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.1894057049636e-05,
+      "loss": 1.5293,
+      "step": 6409
+    },
+    {
+      "epoch": 0.280169587831636,
+      "grad_norm": 2.25,
+      "learning_rate": 8.188876694611351e-05,
+      "loss": 2.2065,
+      "step": 6410
+    },
+    {
+      "epoch": 0.2802132960356659,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.188347624080359e-05,
+      "loss": 1.4536,
+      "step": 6411
+    },
+    {
+      "epoch": 0.2802570042396958,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.187818493380607e-05,
+      "loss": 1.959,
+      "step": 6412
+    },
+    {
+      "epoch": 0.2803007124437257,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.18728930252208e-05,
+      "loss": 1.874,
+      "step": 6413
+    },
+    {
+      "epoch": 0.2803444206477556,
+      "grad_norm": 12.0625,
+      "learning_rate": 8.186760051514765e-05,
+      "loss": 1.8555,
+      "step": 6414
+    },
+    {
+      "epoch": 0.2803881288517855,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.186230740368649e-05,
+      "loss": 1.4012,
+      "step": 6415
+    },
+    {
+      "epoch": 0.28043183705581537,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.185701369093722e-05,
+      "loss": 1.9157,
+      "step": 6416
+    },
+    {
+      "epoch": 0.28047554525984525,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.185171937699974e-05,
+      "loss": 2.0665,
+      "step": 6417
+    },
+    {
+      "epoch": 0.2805192534638752,
+      "grad_norm": 2.5,
+      "learning_rate": 8.184642446197395e-05,
+      "loss": 2.6483,
+      "step": 6418
+    },
+    {
+      "epoch": 0.28056296166790506,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.184112894595979e-05,
+      "loss": 1.6593,
+      "step": 6419
+    },
+    {
+      "epoch": 0.28060666987193494,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.183583282905717e-05,
+      "loss": 2.048,
+      "step": 6420
+    },
+    {
+      "epoch": 0.2806503780759649,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.183053611136607e-05,
+      "loss": 1.9933,
+      "step": 6421
+    },
+    {
+      "epoch": 0.28069408627999476,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.18252387929864e-05,
+      "loss": 1.734,
+      "step": 6422
+    },
+    {
+      "epoch": 0.28073779448402464,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.181994087401819e-05,
+      "loss": 2.8626,
+      "step": 6423
+    },
+    {
+      "epoch": 0.2807815026880546,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.181464235456136e-05,
+      "loss": 1.482,
+      "step": 6424
+    },
+    {
+      "epoch": 0.28082521089208445,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.180934323471592e-05,
+      "loss": 1.6419,
+      "step": 6425
+    },
+    {
+      "epoch": 0.28086891909611433,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.180404351458189e-05,
+      "loss": 1.9609,
+      "step": 6426
+    },
+    {
+      "epoch": 0.2809126273001442,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.179874319425926e-05,
+      "loss": 1.8158,
+      "step": 6427
+    },
+    {
+      "epoch": 0.28095633550417415,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.179344227384807e-05,
+      "loss": 2.282,
+      "step": 6428
+    },
+    {
+      "epoch": 0.28100004370820403,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.178814075344836e-05,
+      "loss": 1.9564,
+      "step": 6429
+    },
+    {
+      "epoch": 0.2810437519122339,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.178283863316015e-05,
+      "loss": 2.1462,
+      "step": 6430
+    },
+    {
+      "epoch": 0.28108746011626384,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.177753591308352e-05,
+      "loss": 2.547,
+      "step": 6431
+    },
+    {
+      "epoch": 0.2811311683202937,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.177223259331855e-05,
+      "loss": 1.6894,
+      "step": 6432
+    },
+    {
+      "epoch": 0.2811748765243236,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.176692867396531e-05,
+      "loss": 1.7359,
+      "step": 6433
+    },
+    {
+      "epoch": 0.28121858472835354,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.176162415512388e-05,
+      "loss": 1.8399,
+      "step": 6434
+    },
+    {
+      "epoch": 0.2812622929323834,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.175631903689438e-05,
+      "loss": 1.7448,
+      "step": 6435
+    },
+    {
+      "epoch": 0.2813060011364133,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.175101331937693e-05,
+      "loss": 2.505,
+      "step": 6436
+    },
+    {
+      "epoch": 0.2813497093404432,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.174570700267163e-05,
+      "loss": 2.1827,
+      "step": 6437
+    },
+    {
+      "epoch": 0.2813934175444731,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.174040008687864e-05,
+      "loss": 1.9409,
+      "step": 6438
+    },
+    {
+      "epoch": 0.281437125748503,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.17350925720981e-05,
+      "loss": 1.7053,
+      "step": 6439
+    },
+    {
+      "epoch": 0.28148083395253287,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.17297844584302e-05,
+      "loss": 1.9401,
+      "step": 6440
+    },
+    {
+      "epoch": 0.2815245421565628,
+      "grad_norm": 2.75,
+      "learning_rate": 8.172447574597506e-05,
+      "loss": 2.2327,
+      "step": 6441
+    },
+    {
+      "epoch": 0.2815682503605927,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.171916643483291e-05,
+      "loss": 1.9431,
+      "step": 6442
+    },
+    {
+      "epoch": 0.28161195856462257,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.17138565251039e-05,
+      "loss": 2.0892,
+      "step": 6443
+    },
+    {
+      "epoch": 0.2816556667686525,
+      "grad_norm": 2.625,
+      "learning_rate": 8.170854601688828e-05,
+      "loss": 1.9925,
+      "step": 6444
+    },
+    {
+      "epoch": 0.2816993749726824,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.170323491028624e-05,
+      "loss": 1.9526,
+      "step": 6445
+    },
+    {
+      "epoch": 0.28174308317671226,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.169792320539802e-05,
+      "loss": 1.7848,
+      "step": 6446
+    },
+    {
+      "epoch": 0.28178679138074214,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.169261090232385e-05,
+      "loss": 1.9374,
+      "step": 6447
+    },
+    {
+      "epoch": 0.2818304995847721,
+      "grad_norm": 2.125,
+      "learning_rate": 8.168729800116398e-05,
+      "loss": 2.2083,
+      "step": 6448
+    },
+    {
+      "epoch": 0.28187420778880196,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.168198450201869e-05,
+      "loss": 1.8002,
+      "step": 6449
+    },
+    {
+      "epoch": 0.28191791599283184,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.167667040498823e-05,
+      "loss": 1.5561,
+      "step": 6450
+    },
+    {
+      "epoch": 0.28196162419686177,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.16713557101729e-05,
+      "loss": 2.1003,
+      "step": 6451
+    },
+    {
+      "epoch": 0.28200533240089165,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.166604041767299e-05,
+      "loss": 2.0644,
+      "step": 6452
+    },
+    {
+      "epoch": 0.28204904060492153,
+      "grad_norm": 2.0,
+      "learning_rate": 8.166072452758883e-05,
+      "loss": 1.5807,
+      "step": 6453
+    },
+    {
+      "epoch": 0.28209274880895147,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.16554080400207e-05,
+      "loss": 2.0718,
+      "step": 6454
+    },
+    {
+      "epoch": 0.28213645701298135,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.165009095506895e-05,
+      "loss": 1.9122,
+      "step": 6455
+    },
+    {
+      "epoch": 0.2821801652170112,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.164477327283391e-05,
+      "loss": 1.8248,
+      "step": 6456
+    },
+    {
+      "epoch": 0.2822238734210411,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.163945499341596e-05,
+      "loss": 2.3692,
+      "step": 6457
+    },
+    {
+      "epoch": 0.28226758162507104,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.163413611691544e-05,
+      "loss": 1.3578,
+      "step": 6458
+    },
+    {
+      "epoch": 0.2823112898291009,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.162881664343271e-05,
+      "loss": 1.7039,
+      "step": 6459
+    },
+    {
+      "epoch": 0.2823549980331308,
+      "grad_norm": 2.0,
+      "learning_rate": 8.16234965730682e-05,
+      "loss": 2.7822,
+      "step": 6460
+    },
+    {
+      "epoch": 0.28239870623716073,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.161817590592228e-05,
+      "loss": 1.3216,
+      "step": 6461
+    },
+    {
+      "epoch": 0.2824424144411906,
+      "grad_norm": 1.8046875,
+      "learning_rate": 8.161285464209537e-05,
+      "loss": 1.5041,
+      "step": 6462
+    },
+    {
+      "epoch": 0.2824861226452205,
+      "grad_norm": 2.25,
+      "learning_rate": 8.160753278168787e-05,
+      "loss": 1.7797,
+      "step": 6463
+    },
+    {
+      "epoch": 0.28252983084925043,
+      "grad_norm": 1.8984375,
+      "learning_rate": 8.160221032480021e-05,
+      "loss": 1.8841,
+      "step": 6464
+    },
+    {
+      "epoch": 0.2825735390532803,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.159688727153287e-05,
+      "loss": 1.8538,
+      "step": 6465
+    },
+    {
+      "epoch": 0.2826172472573102,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.159156362198628e-05,
+      "loss": 1.6499,
+      "step": 6466
+    },
+    {
+      "epoch": 0.28266095546134007,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.15862393762609e-05,
+      "loss": 1.8777,
+      "step": 6467
+    },
+    {
+      "epoch": 0.28270466366537,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.15809145344572e-05,
+      "loss": 1.7798,
+      "step": 6468
+    },
+    {
+      "epoch": 0.2827483718693999,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.157558909667569e-05,
+      "loss": 1.5544,
+      "step": 6469
+    },
+    {
+      "epoch": 0.28279208007342976,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.157026306301685e-05,
+      "loss": 1.9482,
+      "step": 6470
+    },
+    {
+      "epoch": 0.2828357882774597,
+      "grad_norm": 3.453125,
+      "learning_rate": 8.156493643358121e-05,
+      "loss": 1.892,
+      "step": 6471
+    },
+    {
+      "epoch": 0.2828794964814896,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.155960920846926e-05,
+      "loss": 1.6708,
+      "step": 6472
+    },
+    {
+      "epoch": 0.28292320468551946,
+      "grad_norm": 2.25,
+      "learning_rate": 8.155428138778158e-05,
+      "loss": 1.9615,
+      "step": 6473
+    },
+    {
+      "epoch": 0.2829669128895494,
+      "grad_norm": 1.796875,
+      "learning_rate": 8.154895297161866e-05,
+      "loss": 1.6599,
+      "step": 6474
+    },
+    {
+      "epoch": 0.2830106210935793,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.154362396008109e-05,
+      "loss": 2.2279,
+      "step": 6475
+    },
+    {
+      "epoch": 0.28305432929760915,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.153829435326942e-05,
+      "loss": 1.5279,
+      "step": 6476
+    },
+    {
+      "epoch": 0.28309803750163903,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.153296415128425e-05,
+      "loss": 1.9643,
+      "step": 6477
+    },
+    {
+      "epoch": 0.28314174570566897,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.152763335422613e-05,
+      "loss": 1.7157,
+      "step": 6478
+    },
+    {
+      "epoch": 0.28318545390969885,
+      "grad_norm": 2.125,
+      "learning_rate": 8.15223019621957e-05,
+      "loss": 1.7339,
+      "step": 6479
+    },
+    {
+      "epoch": 0.2832291621137287,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.151696997529354e-05,
+      "loss": 1.2496,
+      "step": 6480
+    },
+    {
+      "epoch": 0.28327287031775866,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.151163739362029e-05,
+      "loss": 1.8114,
+      "step": 6481
+    },
+    {
+      "epoch": 0.28331657852178854,
+      "grad_norm": 2.375,
+      "learning_rate": 8.150630421727659e-05,
+      "loss": 1.8059,
+      "step": 6482
+    },
+    {
+      "epoch": 0.2833602867258184,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.150097044636307e-05,
+      "loss": 1.8667,
+      "step": 6483
+    },
+    {
+      "epoch": 0.28340399492984836,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.149563608098037e-05,
+      "loss": 1.6471,
+      "step": 6484
+    },
+    {
+      "epoch": 0.28344770313387824,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.14903011212292e-05,
+      "loss": 1.4635,
+      "step": 6485
+    },
+    {
+      "epoch": 0.2834914113379081,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.148496556721022e-05,
+      "loss": 1.7774,
+      "step": 6486
+    },
+    {
+      "epoch": 0.283535119541938,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.14796294190241e-05,
+      "loss": 1.8177,
+      "step": 6487
+    },
+    {
+      "epoch": 0.28357882774596793,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.147429267677156e-05,
+      "loss": 1.8345,
+      "step": 6488
+    },
+    {
+      "epoch": 0.2836225359499978,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.146895534055332e-05,
+      "loss": 2.1024,
+      "step": 6489
+    },
+    {
+      "epoch": 0.2836662441540277,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.146361741047006e-05,
+      "loss": 2.1015,
+      "step": 6490
+    },
+    {
+      "epoch": 0.2837099523580576,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.145827888662257e-05,
+      "loss": 1.6696,
+      "step": 6491
+    },
+    {
+      "epoch": 0.2837536605620875,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.145293976911158e-05,
+      "loss": 1.7858,
+      "step": 6492
+    },
+    {
+      "epoch": 0.2837973687661174,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.144760005803783e-05,
+      "loss": 1.94,
+      "step": 6493
+    },
+    {
+      "epoch": 0.2838410769701473,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.14422597535021e-05,
+      "loss": 1.595,
+      "step": 6494
+    },
+    {
+      "epoch": 0.2838847851741772,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.143691885560515e-05,
+      "loss": 2.0082,
+      "step": 6495
+    },
+    {
+      "epoch": 0.2839284933782071,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.14315773644478e-05,
+      "loss": 1.6122,
+      "step": 6496
+    },
+    {
+      "epoch": 0.28397220158223696,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.142623528013084e-05,
+      "loss": 1.9811,
+      "step": 6497
+    },
+    {
+      "epoch": 0.2840159097862669,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.142089260275509e-05,
+      "loss": 1.9579,
+      "step": 6498
+    },
+    {
+      "epoch": 0.2840596179902968,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.141554933242135e-05,
+      "loss": 2.491,
+      "step": 6499
+    },
+    {
+      "epoch": 0.28410332619432666,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.141020546923048e-05,
+      "loss": 1.4218,
+      "step": 6500
+    },
+    {
+      "epoch": 0.2841470343983566,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.14048610132833e-05,
+      "loss": 1.2199,
+      "step": 6501
+    },
+    {
+      "epoch": 0.28419074260238647,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.13995159646807e-05,
+      "loss": 1.6209,
+      "step": 6502
+    },
+    {
+      "epoch": 0.28423445080641635,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.139417032352354e-05,
+      "loss": 3.7028,
+      "step": 6503
+    },
+    {
+      "epoch": 0.2842781590104463,
+      "grad_norm": 2.375,
+      "learning_rate": 8.138882408991268e-05,
+      "loss": 1.2836,
+      "step": 6504
+    },
+    {
+      "epoch": 0.28432186721447616,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.138347726394904e-05,
+      "loss": 1.7913,
+      "step": 6505
+    },
+    {
+      "epoch": 0.28436557541850604,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.13781298457335e-05,
+      "loss": 1.7721,
+      "step": 6506
+    },
+    {
+      "epoch": 0.2844092836225359,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.137278183536699e-05,
+      "loss": 2.4412,
+      "step": 6507
+    },
+    {
+      "epoch": 0.28445299182656586,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.13674332329504e-05,
+      "loss": 2.2241,
+      "step": 6508
+    },
+    {
+      "epoch": 0.28449670003059574,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.136208403858472e-05,
+      "loss": 2.3641,
+      "step": 6509
+    },
+    {
+      "epoch": 0.2845404082346256,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.135673425237084e-05,
+      "loss": 1.7108,
+      "step": 6510
+    },
+    {
+      "epoch": 0.28458411643865555,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.135138387440977e-05,
+      "loss": 1.385,
+      "step": 6511
+    },
+    {
+      "epoch": 0.28462782464268543,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.134603290480245e-05,
+      "loss": 2.3765,
+      "step": 6512
+    },
+    {
+      "epoch": 0.2846715328467153,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.134068134364987e-05,
+      "loss": 2.4618,
+      "step": 6513
+    },
+    {
+      "epoch": 0.28471524105074525,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.133532919105302e-05,
+      "loss": 1.9383,
+      "step": 6514
+    },
+    {
+      "epoch": 0.28475894925477513,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.13299764471129e-05,
+      "loss": 1.7269,
+      "step": 6515
+    },
+    {
+      "epoch": 0.284802657458805,
+      "grad_norm": 2.0,
+      "learning_rate": 8.132462311193053e-05,
+      "loss": 1.7784,
+      "step": 6516
+    },
+    {
+      "epoch": 0.2848463656628349,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.131926918560692e-05,
+      "loss": 1.8252,
+      "step": 6517
+    },
+    {
+      "epoch": 0.2848900738668648,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.131391466824312e-05,
+      "loss": 1.5615,
+      "step": 6518
+    },
+    {
+      "epoch": 0.2849337820708947,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.130855955994019e-05,
+      "loss": 2.0726,
+      "step": 6519
+    },
+    {
+      "epoch": 0.2849774902749246,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.130320386079915e-05,
+      "loss": 1.6181,
+      "step": 6520
+    },
+    {
+      "epoch": 0.2850211984789545,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.129784757092111e-05,
+      "loss": 1.9783,
+      "step": 6521
+    },
+    {
+      "epoch": 0.2850649066829844,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.129249069040712e-05,
+      "loss": 3.0229,
+      "step": 6522
+    },
+    {
+      "epoch": 0.2851086148870143,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.12871332193583e-05,
+      "loss": 1.6179,
+      "step": 6523
+    },
+    {
+      "epoch": 0.2851523230910442,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.128177515787574e-05,
+      "loss": 2.1017,
+      "step": 6524
+    },
+    {
+      "epoch": 0.2851960312950741,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.127641650606054e-05,
+      "loss": 1.5302,
+      "step": 6525
+    },
+    {
+      "epoch": 0.285239739499104,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.127105726401386e-05,
+      "loss": 1.546,
+      "step": 6526
+    },
+    {
+      "epoch": 0.28528344770313385,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.126569743183681e-05,
+      "loss": 1.6359,
+      "step": 6527
+    },
+    {
+      "epoch": 0.2853271559071638,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.126033700963053e-05,
+      "loss": 1.6768,
+      "step": 6528
+    },
+    {
+      "epoch": 0.28537086411119367,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.12549759974962e-05,
+      "loss": 1.6905,
+      "step": 6529
+    },
+    {
+      "epoch": 0.28541457231522355,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.1249614395535e-05,
+      "loss": 1.8334,
+      "step": 6530
+    },
+    {
+      "epoch": 0.2854582805192535,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.124425220384808e-05,
+      "loss": 2.195,
+      "step": 6531
+    },
+    {
+      "epoch": 0.28550198872328336,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.123888942253666e-05,
+      "loss": 2.0734,
+      "step": 6532
+    },
+    {
+      "epoch": 0.28554569692731324,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.123352605170191e-05,
+      "loss": 1.9242,
+      "step": 6533
+    },
+    {
+      "epoch": 0.2855894051313432,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.122816209144509e-05,
+      "loss": 2.0874,
+      "step": 6534
+    },
+    {
+      "epoch": 0.28563311333537306,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.122279754186736e-05,
+      "loss": 1.4013,
+      "step": 6535
+    },
+    {
+      "epoch": 0.28567682153940294,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.121743240307003e-05,
+      "loss": 1.9932,
+      "step": 6536
+    },
+    {
+      "epoch": 0.2857205297434328,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.121206667515431e-05,
+      "loss": 1.5798,
+      "step": 6537
+    },
+    {
+      "epoch": 0.28576423794746275,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.120670035822147e-05,
+      "loss": 2.2536,
+      "step": 6538
+    },
+    {
+      "epoch": 0.28580794615149263,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.120133345237278e-05,
+      "loss": 1.8946,
+      "step": 6539
+    },
+    {
+      "epoch": 0.2858516543555225,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.11959659577095e-05,
+      "loss": 1.5591,
+      "step": 6540
+    },
+    {
+      "epoch": 0.28589536255955245,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.119059787433294e-05,
+      "loss": 1.8271,
+      "step": 6541
+    },
+    {
+      "epoch": 0.2859390707635823,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.118522920234442e-05,
+      "loss": 1.7642,
+      "step": 6542
+    },
+    {
+      "epoch": 0.2859827789676122,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.117985994184522e-05,
+      "loss": 1.6098,
+      "step": 6543
+    },
+    {
+      "epoch": 0.28602648717164214,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.117449009293668e-05,
+      "loss": 2.3942,
+      "step": 6544
+    },
+    {
+      "epoch": 0.286070195375672,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.116911965572016e-05,
+      "loss": 2.513,
+      "step": 6545
+    },
+    {
+      "epoch": 0.2861139035797019,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.116374863029696e-05,
+      "loss": 2.7693,
+      "step": 6546
+    },
+    {
+      "epoch": 0.2861576117837318,
+      "grad_norm": 2.0,
+      "learning_rate": 8.115837701676848e-05,
+      "loss": 2.2292,
+      "step": 6547
+    },
+    {
+      "epoch": 0.2862013199877617,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.115300481523609e-05,
+      "loss": 1.8436,
+      "step": 6548
+    },
+    {
+      "epoch": 0.2862450281917916,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.114763202580113e-05,
+      "loss": 1.6627,
+      "step": 6549
+    },
+    {
+      "epoch": 0.2862887363958215,
+      "grad_norm": 6.625,
+      "learning_rate": 8.114225864856504e-05,
+      "loss": 2.6498,
+      "step": 6550
+    },
+    {
+      "epoch": 0.2863324445998514,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.11368846836292e-05,
+      "loss": 1.5364,
+      "step": 6551
+    },
+    {
+      "epoch": 0.2863761528038813,
+      "grad_norm": 2.375,
+      "learning_rate": 8.113151013109503e-05,
+      "loss": 2.3289,
+      "step": 6552
+    },
+    {
+      "epoch": 0.28641986100791117,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.112613499106396e-05,
+      "loss": 1.7235,
+      "step": 6553
+    },
+    {
+      "epoch": 0.2864635692119411,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.11207592636374e-05,
+      "loss": 2.1643,
+      "step": 6554
+    },
+    {
+      "epoch": 0.286507277415971,
+      "grad_norm": 2.875,
+      "learning_rate": 8.111538294891684e-05,
+      "loss": 1.8422,
+      "step": 6555
+    },
+    {
+      "epoch": 0.28655098562000086,
+      "grad_norm": 2.0,
+      "learning_rate": 8.11100060470037e-05,
+      "loss": 1.6211,
+      "step": 6556
+    },
+    {
+      "epoch": 0.28659469382403074,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.110462855799949e-05,
+      "loss": 1.5814,
+      "step": 6557
+    },
+    {
+      "epoch": 0.2866384020280607,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.109925048200565e-05,
+      "loss": 1.3296,
+      "step": 6558
+    },
+    {
+      "epoch": 0.28668211023209056,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.109387181912369e-05,
+      "loss": 2.0455,
+      "step": 6559
+    },
+    {
+      "epoch": 0.28672581843612044,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.108849256945513e-05,
+      "loss": 1.9988,
+      "step": 6560
+    },
+    {
+      "epoch": 0.2867695266401504,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.108311273310146e-05,
+      "loss": 1.8771,
+      "step": 6561
+    },
+    {
+      "epoch": 0.28681323484418025,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.10777323101642e-05,
+      "loss": 2.4142,
+      "step": 6562
+    },
+    {
+      "epoch": 0.28685694304821013,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.107235130074492e-05,
+      "loss": 1.7797,
+      "step": 6563
+    },
+    {
+      "epoch": 0.28690065125224007,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.106696970494514e-05,
+      "loss": 1.9409,
+      "step": 6564
+    },
+    {
+      "epoch": 0.28694435945626995,
+      "grad_norm": 1.953125,
+      "learning_rate": 8.106158752286642e-05,
+      "loss": 1.4793,
+      "step": 6565
+    },
+    {
+      "epoch": 0.28698806766029983,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.105620475461033e-05,
+      "loss": 1.6698,
+      "step": 6566
+    },
+    {
+      "epoch": 0.2870317758643297,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.105082140027846e-05,
+      "loss": 2.1729,
+      "step": 6567
+    },
+    {
+      "epoch": 0.28707548406835964,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.104543745997242e-05,
+      "loss": 1.7732,
+      "step": 6568
+    },
+    {
+      "epoch": 0.2871191922723895,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.104005293379378e-05,
+      "loss": 1.779,
+      "step": 6569
+    },
+    {
+      "epoch": 0.2871629004764194,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.103466782184415e-05,
+      "loss": 1.9623,
+      "step": 6570
+    },
+    {
+      "epoch": 0.28720660868044934,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.102928212422519e-05,
+      "loss": 2.3689,
+      "step": 6571
+    },
+    {
+      "epoch": 0.2872503168844792,
+      "grad_norm": 2.25,
+      "learning_rate": 8.102389584103849e-05,
+      "loss": 1.5722,
+      "step": 6572
+    },
+    {
+      "epoch": 0.2872940250885091,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.101850897238574e-05,
+      "loss": 2.7088,
+      "step": 6573
+    },
+    {
+      "epoch": 0.28733773329253903,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.101312151836857e-05,
+      "loss": 2.5903,
+      "step": 6574
+    },
+    {
+      "epoch": 0.2873814414965689,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.100773347908868e-05,
+      "loss": 2.3079,
+      "step": 6575
+    },
+    {
+      "epoch": 0.2874251497005988,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.100234485464771e-05,
+      "loss": 2.0066,
+      "step": 6576
+    },
+    {
+      "epoch": 0.28746885790462867,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.099695564514737e-05,
+      "loss": 2.1999,
+      "step": 6577
+    },
+    {
+      "epoch": 0.2875125661086586,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.099156585068937e-05,
+      "loss": 1.8645,
+      "step": 6578
+    },
+    {
+      "epoch": 0.2875562743126885,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.098617547137541e-05,
+      "loss": 1.7175,
+      "step": 6579
+    },
+    {
+      "epoch": 0.28759998251671837,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.098078450730723e-05,
+      "loss": 1.8844,
+      "step": 6580
+    },
+    {
+      "epoch": 0.2876436907207483,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.097539295858656e-05,
+      "loss": 1.7426,
+      "step": 6581
+    },
+    {
+      "epoch": 0.2876873989247782,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.097000082531512e-05,
+      "loss": 2.1152,
+      "step": 6582
+    },
+    {
+      "epoch": 0.28773110712880806,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.096460810759472e-05,
+      "loss": 1.8683,
+      "step": 6583
+    },
+    {
+      "epoch": 0.287774815332838,
+      "grad_norm": 2.0,
+      "learning_rate": 8.095921480552707e-05,
+      "loss": 1.4677,
+      "step": 6584
+    },
+    {
+      "epoch": 0.2878185235368679,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.095382091921399e-05,
+      "loss": 1.6928,
+      "step": 6585
+    },
+    {
+      "epoch": 0.28786223174089776,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.094842644875726e-05,
+      "loss": 1.963,
+      "step": 6586
+    },
+    {
+      "epoch": 0.28790593994492764,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.094303139425867e-05,
+      "loss": 1.5391,
+      "step": 6587
+    },
+    {
+      "epoch": 0.28794964814895757,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.093763575582006e-05,
+      "loss": 2.0163,
+      "step": 6588
+    },
+    {
+      "epoch": 0.28799335635298745,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.093223953354323e-05,
+      "loss": 1.5258,
+      "step": 6589
+    },
+    {
+      "epoch": 0.28803706455701733,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.092684272753002e-05,
+      "loss": 2.2255,
+      "step": 6590
+    },
+    {
+      "epoch": 0.28808077276104727,
+      "grad_norm": 2.5,
+      "learning_rate": 8.092144533788226e-05,
+      "loss": 1.9112,
+      "step": 6591
+    },
+    {
+      "epoch": 0.28812448096507715,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.091604736470184e-05,
+      "loss": 2.0743,
+      "step": 6592
+    },
+    {
+      "epoch": 0.288168189169107,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.091064880809061e-05,
+      "loss": 1.6284,
+      "step": 6593
+    },
+    {
+      "epoch": 0.28821189737313696,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.090524966815042e-05,
+      "loss": 2.0639,
+      "step": 6594
+    },
+    {
+      "epoch": 0.28825560557716684,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.089984994498324e-05,
+      "loss": 1.5172,
+      "step": 6595
+    },
+    {
+      "epoch": 0.2882993137811967,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.089444963869088e-05,
+      "loss": 1.8543,
+      "step": 6596
+    },
+    {
+      "epoch": 0.2883430219852266,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.088904874937528e-05,
+      "loss": 2.9156,
+      "step": 6597
+    },
+    {
+      "epoch": 0.28838673018925653,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.088364727713841e-05,
+      "loss": 1.9629,
+      "step": 6598
+    },
+    {
+      "epoch": 0.2884304383932864,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.087824522208215e-05,
+      "loss": 2.4238,
+      "step": 6599
+    },
+    {
+      "epoch": 0.2884741465973163,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.087284258430847e-05,
+      "loss": 2.4762,
+      "step": 6600
+    },
+    {
+      "epoch": 0.28851785480134623,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.08674393639193e-05,
+      "loss": 1.8678,
+      "step": 6601
+    },
+    {
+      "epoch": 0.2885615630053761,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.086203556101664e-05,
+      "loss": 3.2367,
+      "step": 6602
+    },
+    {
+      "epoch": 0.288605271209406,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.085663117570243e-05,
+      "loss": 2.9074,
+      "step": 6603
+    },
+    {
+      "epoch": 0.2886489794134359,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.08512262080787e-05,
+      "loss": 2.416,
+      "step": 6604
+    },
+    {
+      "epoch": 0.2886926876174658,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.084582065824743e-05,
+      "loss": 1.755,
+      "step": 6605
+    },
+    {
+      "epoch": 0.2887363958214957,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.084041452631062e-05,
+      "loss": 1.7571,
+      "step": 6606
+    },
+    {
+      "epoch": 0.28878010402552556,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.08350078123703e-05,
+      "loss": 2.1158,
+      "step": 6607
+    },
+    {
+      "epoch": 0.2888238122295555,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.082960051652852e-05,
+      "loss": 1.8783,
+      "step": 6608
+    },
+    {
+      "epoch": 0.2888675204335854,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.08241926388873e-05,
+      "loss": 2.3248,
+      "step": 6609
+    },
+    {
+      "epoch": 0.28891122863761526,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.08187841795487e-05,
+      "loss": 1.7596,
+      "step": 6610
+    },
+    {
+      "epoch": 0.2889549368416452,
+      "grad_norm": 1.9140625,
+      "learning_rate": 8.081337513861478e-05,
+      "loss": 1.851,
+      "step": 6611
+    },
+    {
+      "epoch": 0.2889986450456751,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.080796551618763e-05,
+      "loss": 1.7286,
+      "step": 6612
+    },
+    {
+      "epoch": 0.28904235324970495,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.080255531236935e-05,
+      "loss": 1.975,
+      "step": 6613
+    },
+    {
+      "epoch": 0.2890860614537349,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.079714452726199e-05,
+      "loss": 1.9174,
+      "step": 6614
+    },
+    {
+      "epoch": 0.28912976965776477,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.079173316096772e-05,
+      "loss": 2.6944,
+      "step": 6615
+    },
+    {
+      "epoch": 0.28917347786179465,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.07863212135886e-05,
+      "loss": 2.5183,
+      "step": 6616
+    },
+    {
+      "epoch": 0.2892171860658246,
+      "grad_norm": 2.5,
+      "learning_rate": 8.07809086852268e-05,
+      "loss": 2.2211,
+      "step": 6617
+    },
+    {
+      "epoch": 0.28926089426985446,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.077549557598448e-05,
+      "loss": 1.8159,
+      "step": 6618
+    },
+    {
+      "epoch": 0.28930460247388434,
+      "grad_norm": 2.25,
+      "learning_rate": 8.077008188596375e-05,
+      "loss": 1.7229,
+      "step": 6619
+    },
+    {
+      "epoch": 0.2893483106779142,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.076466761526678e-05,
+      "loss": 1.8865,
+      "step": 6620
+    },
+    {
+      "epoch": 0.28939201888194416,
+      "grad_norm": 1.890625,
+      "learning_rate": 8.075925276399576e-05,
+      "loss": 1.7476,
+      "step": 6621
+    },
+    {
+      "epoch": 0.28943572708597404,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.075383733225288e-05,
+      "loss": 1.7712,
+      "step": 6622
+    },
+    {
+      "epoch": 0.2894794352900039,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.074842132014034e-05,
+      "loss": 1.715,
+      "step": 6623
+    },
+    {
+      "epoch": 0.28952314349403385,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.074300472776031e-05,
+      "loss": 1.8801,
+      "step": 6624
+    },
+    {
+      "epoch": 0.28956685169806373,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.073758755521505e-05,
+      "loss": 1.5784,
+      "step": 6625
+    },
+    {
+      "epoch": 0.2896105599020936,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.073216980260678e-05,
+      "loss": 1.6752,
+      "step": 6626
+    },
+    {
+      "epoch": 0.28965426810612355,
+      "grad_norm": 3.84375,
+      "learning_rate": 8.072675147003773e-05,
+      "loss": 1.2782,
+      "step": 6627
+    },
+    {
+      "epoch": 0.2896979763101534,
+      "grad_norm": 6.90625,
+      "learning_rate": 8.072133255761017e-05,
+      "loss": 2.8148,
+      "step": 6628
+    },
+    {
+      "epoch": 0.2897416845141833,
+      "grad_norm": 1.9765625,
+      "learning_rate": 8.071591306542634e-05,
+      "loss": 1.8114,
+      "step": 6629
+    },
+    {
+      "epoch": 0.2897853927182132,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.071049299358853e-05,
+      "loss": 1.2275,
+      "step": 6630
+    },
+    {
+      "epoch": 0.2898291009222431,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.070507234219901e-05,
+      "loss": 2.4481,
+      "step": 6631
+    },
+    {
+      "epoch": 0.289872809126273,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.06996511113601e-05,
+      "loss": 2.0297,
+      "step": 6632
+    },
+    {
+      "epoch": 0.2899165173303029,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.06942293011741e-05,
+      "loss": 1.6154,
+      "step": 6633
+    },
+    {
+      "epoch": 0.2899602255343328,
+      "grad_norm": 2.375,
+      "learning_rate": 8.06888069117433e-05,
+      "loss": 2.2716,
+      "step": 6634
+    },
+    {
+      "epoch": 0.2900039337383627,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.068338394317005e-05,
+      "loss": 2.1387,
+      "step": 6635
+    },
+    {
+      "epoch": 0.2900476419423926,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.06779603955567e-05,
+      "loss": 1.6537,
+      "step": 6636
+    },
+    {
+      "epoch": 0.2900913501464225,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.067253626900558e-05,
+      "loss": 2.2958,
+      "step": 6637
+    },
+    {
+      "epoch": 0.2901350583504524,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.066711156361905e-05,
+      "loss": 2.7007,
+      "step": 6638
+    },
+    {
+      "epoch": 0.29017876655448227,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.066168627949952e-05,
+      "loss": 2.0117,
+      "step": 6639
+    },
+    {
+      "epoch": 0.29022247475851215,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.065626041674932e-05,
+      "loss": 2.6418,
+      "step": 6640
+    },
+    {
+      "epoch": 0.2902661829625421,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.065083397547086e-05,
+      "loss": 1.6984,
+      "step": 6641
+    },
+    {
+      "epoch": 0.29030989116657197,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.064540695576657e-05,
+      "loss": 2.5482,
+      "step": 6642
+    },
+    {
+      "epoch": 0.29035359937060184,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.063997935773885e-05,
+      "loss": 1.8375,
+      "step": 6643
+    },
+    {
+      "epoch": 0.2903973075746318,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.063455118149013e-05,
+      "loss": 2.1303,
+      "step": 6644
+    },
+    {
+      "epoch": 0.29044101577866166,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.062912242712282e-05,
+      "loss": 1.3989,
+      "step": 6645
+    },
+    {
+      "epoch": 0.29048472398269154,
+      "grad_norm": 1.90625,
+      "learning_rate": 8.06236930947394e-05,
+      "loss": 1.541,
+      "step": 6646
+    },
+    {
+      "epoch": 0.2905284321867215,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.061826318444232e-05,
+      "loss": 1.709,
+      "step": 6647
+    },
+    {
+      "epoch": 0.29057214039075135,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.061283269633407e-05,
+      "loss": 1.6461,
+      "step": 6648
+    },
+    {
+      "epoch": 0.29061584859478123,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.06074016305171e-05,
+      "loss": 1.5956,
+      "step": 6649
+    },
+    {
+      "epoch": 0.2906595567988111,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.060196998709391e-05,
+      "loss": 1.9987,
+      "step": 6650
+    },
+    {
+      "epoch": 0.29070326500284105,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.0596537766167e-05,
+      "loss": 2.2364,
+      "step": 6651
+    },
+    {
+      "epoch": 0.29074697320687093,
+      "grad_norm": 6.1875,
+      "learning_rate": 8.05911049678389e-05,
+      "loss": 1.6748,
+      "step": 6652
+    },
+    {
+      "epoch": 0.2907906814109008,
+      "grad_norm": 14.9375,
+      "learning_rate": 8.058567159221213e-05,
+      "loss": 3.4629,
+      "step": 6653
+    },
+    {
+      "epoch": 0.29083438961493074,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.058023763938922e-05,
+      "loss": 1.752,
+      "step": 6654
+    },
+    {
+      "epoch": 0.2908780978189606,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.057480310947271e-05,
+      "loss": 2.5652,
+      "step": 6655
+    },
+    {
+      "epoch": 0.2909218060229905,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.056936800256517e-05,
+      "loss": 2.0146,
+      "step": 6656
+    },
+    {
+      "epoch": 0.29096551422702044,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.056393231876918e-05,
+      "loss": 1.7524,
+      "step": 6657
+    },
+    {
+      "epoch": 0.2910092224310503,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.055849605818728e-05,
+      "loss": 1.9087,
+      "step": 6658
+    },
+    {
+      "epoch": 0.2910529306350802,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.055305922092208e-05,
+      "loss": 2.1884,
+      "step": 6659
+    },
+    {
+      "epoch": 0.2910966388391101,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.05476218070762e-05,
+      "loss": 1.6765,
+      "step": 6660
+    },
+    {
+      "epoch": 0.29114034704314,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.054218381675225e-05,
+      "loss": 1.5884,
+      "step": 6661
+    },
+    {
+      "epoch": 0.2911840552471699,
+      "grad_norm": 2.125,
+      "learning_rate": 8.053674525005282e-05,
+      "loss": 1.8013,
+      "step": 6662
+    },
+    {
+      "epoch": 0.2912277634511998,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.053130610708057e-05,
+      "loss": 1.8213,
+      "step": 6663
+    },
+    {
+      "epoch": 0.2912714716552297,
+      "grad_norm": 2.25,
+      "learning_rate": 8.052586638793814e-05,
+      "loss": 2.4342,
+      "step": 6664
+    },
+    {
+      "epoch": 0.2913151798592596,
+      "grad_norm": 1.9375,
+      "learning_rate": 8.052042609272817e-05,
+      "loss": 1.9834,
+      "step": 6665
+    },
+    {
+      "epoch": 0.29135888806328947,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.051498522155334e-05,
+      "loss": 1.7118,
+      "step": 6666
+    },
+    {
+      "epoch": 0.2914025962673194,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.050954377451634e-05,
+      "loss": 1.9769,
+      "step": 6667
+    },
+    {
+      "epoch": 0.2914463044713493,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.050410175171983e-05,
+      "loss": 1.5139,
+      "step": 6668
+    },
+    {
+      "epoch": 0.29149001267537916,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.049865915326653e-05,
+      "loss": 2.0808,
+      "step": 6669
+    },
+    {
+      "epoch": 0.29153372087940904,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.049321597925914e-05,
+      "loss": 1.6965,
+      "step": 6670
+    },
+    {
+      "epoch": 0.291577429083439,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.04877722298004e-05,
+      "loss": 2.0674,
+      "step": 6671
+    },
+    {
+      "epoch": 0.29162113728746886,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.0482327904993e-05,
+      "loss": 1.6745,
+      "step": 6672
+    },
+    {
+      "epoch": 0.29166484549149874,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.047688300493972e-05,
+      "loss": 1.9184,
+      "step": 6673
+    },
+    {
+      "epoch": 0.29170855369552867,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.047143752974331e-05,
+      "loss": 1.4173,
+      "step": 6674
+    },
+    {
+      "epoch": 0.29175226189955855,
+      "grad_norm": 3.96875,
+      "learning_rate": 8.046599147950651e-05,
+      "loss": 2.3369,
+      "step": 6675
+    },
+    {
+      "epoch": 0.29179597010358843,
+      "grad_norm": 1.890625,
+      "learning_rate": 8.046054485433211e-05,
+      "loss": 1.6808,
+      "step": 6676
+    },
+    {
+      "epoch": 0.29183967830761837,
+      "grad_norm": 3.25,
+      "learning_rate": 8.04550976543229e-05,
+      "loss": 1.5491,
+      "step": 6677
+    },
+    {
+      "epoch": 0.29188338651164825,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.044964987958168e-05,
+      "loss": 1.711,
+      "step": 6678
+    },
+    {
+      "epoch": 0.2919270947156781,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.044420153021124e-05,
+      "loss": 2.976,
+      "step": 6679
+    },
+    {
+      "epoch": 0.291970802919708,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.043875260631442e-05,
+      "loss": 1.2881,
+      "step": 6680
+    },
+    {
+      "epoch": 0.29201451112373794,
+      "grad_norm": 2.0,
+      "learning_rate": 8.043330310799402e-05,
+      "loss": 1.7619,
+      "step": 6681
+    },
+    {
+      "epoch": 0.2920582193277678,
+      "grad_norm": 2.984375,
+      "learning_rate": 8.042785303535289e-05,
+      "loss": 2.2613,
+      "step": 6682
+    },
+    {
+      "epoch": 0.2921019275317977,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.04224023884939e-05,
+      "loss": 1.8949,
+      "step": 6683
+    },
+    {
+      "epoch": 0.29214563573582764,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.041695116751991e-05,
+      "loss": 1.5188,
+      "step": 6684
+    },
+    {
+      "epoch": 0.2921893439398575,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.041149937253378e-05,
+      "loss": 1.3007,
+      "step": 6685
+    },
+    {
+      "epoch": 0.2922330521438874,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.040604700363838e-05,
+      "loss": 2.9892,
+      "step": 6686
+    },
+    {
+      "epoch": 0.29227676034791733,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.040059406093662e-05,
+      "loss": 2.1024,
+      "step": 6687
+    },
+    {
+      "epoch": 0.2923204685519472,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.039514054453141e-05,
+      "loss": 1.5767,
+      "step": 6688
+    },
+    {
+      "epoch": 0.2923641767559771,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.038968645452567e-05,
+      "loss": 1.3219,
+      "step": 6689
+    },
+    {
+      "epoch": 0.29240788496000697,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.03842317910223e-05,
+      "loss": 1.9124,
+      "step": 6690
+    },
+    {
+      "epoch": 0.2924515931640369,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.037877655412426e-05,
+      "loss": 1.6807,
+      "step": 6691
+    },
+    {
+      "epoch": 0.2924953013680668,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.037332074393449e-05,
+      "loss": 2.4909,
+      "step": 6692
+    },
+    {
+      "epoch": 0.29253900957209666,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.036786436055595e-05,
+      "loss": 1.5007,
+      "step": 6693
+    },
+    {
+      "epoch": 0.2925827177761266,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.036240740409162e-05,
+      "loss": 2.127,
+      "step": 6694
+    },
+    {
+      "epoch": 0.2926264259801565,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.035694987464446e-05,
+      "loss": 2.2236,
+      "step": 6695
+    },
+    {
+      "epoch": 0.29267013418418636,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.035149177231749e-05,
+      "loss": 1.577,
+      "step": 6696
+    },
+    {
+      "epoch": 0.2927138423882163,
+      "grad_norm": 1.8515625,
+      "learning_rate": 8.034603309721368e-05,
+      "loss": 1.6475,
+      "step": 6697
+    },
+    {
+      "epoch": 0.2927575505922462,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.034057384943606e-05,
+      "loss": 1.9149,
+      "step": 6698
+    },
+    {
+      "epoch": 0.29280125879627605,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.033511402908767e-05,
+      "loss": 1.9467,
+      "step": 6699
+    },
+    {
+      "epoch": 0.29284496700030593,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.03296536362715e-05,
+      "loss": 3.2761,
+      "step": 6700
+    },
+    {
+      "epoch": 0.29288867520433587,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.032419267109066e-05,
+      "loss": 1.9798,
+      "step": 6701
+    },
+    {
+      "epoch": 0.29293238340836575,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.031873113364814e-05,
+      "loss": 2.787,
+      "step": 6702
+    },
+    {
+      "epoch": 0.29297609161239563,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.031326902404703e-05,
+      "loss": 2.9573,
+      "step": 6703
+    },
+    {
+      "epoch": 0.29301979981642556,
+      "grad_norm": 1.9453125,
+      "learning_rate": 8.030780634239043e-05,
+      "loss": 1.6533,
+      "step": 6704
+    },
+    {
+      "epoch": 0.29306350802045544,
+      "grad_norm": 2.25,
+      "learning_rate": 8.030234308878142e-05,
+      "loss": 2.3842,
+      "step": 6705
+    },
+    {
+      "epoch": 0.2931072162244853,
+      "grad_norm": 17.625,
+      "learning_rate": 8.02968792633231e-05,
+      "loss": 2.1034,
+      "step": 6706
+    },
+    {
+      "epoch": 0.29315092442851526,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.029141486611856e-05,
+      "loss": 1.9293,
+      "step": 6707
+    },
+    {
+      "epoch": 0.29319463263254514,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.028594989727092e-05,
+      "loss": 1.9474,
+      "step": 6708
+    },
+    {
+      "epoch": 0.293238340836575,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.028048435688333e-05,
+      "loss": 1.9978,
+      "step": 6709
+    },
+    {
+      "epoch": 0.2932820490406049,
+      "grad_norm": 1.8828125,
+      "learning_rate": 8.027501824505895e-05,
+      "loss": 1.4904,
+      "step": 6710
+    },
+    {
+      "epoch": 0.29332575724463483,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.02695515619009e-05,
+      "loss": 2.2193,
+      "step": 6711
+    },
+    {
+      "epoch": 0.2933694654486647,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.026408430751235e-05,
+      "loss": 2.0577,
+      "step": 6712
+    },
+    {
+      "epoch": 0.2934131736526946,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.025861648199649e-05,
+      "loss": 1.9362,
+      "step": 6713
+    },
+    {
+      "epoch": 0.2934568818567245,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.02531480854565e-05,
+      "loss": 1.6689,
+      "step": 6714
+    },
+    {
+      "epoch": 0.2935005900607544,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.024767911799558e-05,
+      "loss": 2.1455,
+      "step": 6715
+    },
+    {
+      "epoch": 0.2935442982647843,
+      "grad_norm": 1.9921875,
+      "learning_rate": 8.024220957971693e-05,
+      "loss": 2.7358,
+      "step": 6716
+    },
+    {
+      "epoch": 0.2935880064688142,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.023673947072376e-05,
+      "loss": 2.3428,
+      "step": 6717
+    },
+    {
+      "epoch": 0.2936317146728441,
+      "grad_norm": 1.921875,
+      "learning_rate": 8.023126879111931e-05,
+      "loss": 2.0114,
+      "step": 6718
+    },
+    {
+      "epoch": 0.293675422876874,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.022579754100681e-05,
+      "loss": 1.714,
+      "step": 6719
+    },
+    {
+      "epoch": 0.29371913108090386,
+      "grad_norm": 1.9609375,
+      "learning_rate": 8.022032572048954e-05,
+      "loss": 1.6105,
+      "step": 6720
+    },
+    {
+      "epoch": 0.2937628392849338,
+      "grad_norm": 3.328125,
+      "learning_rate": 8.021485332967072e-05,
+      "loss": 2.8173,
+      "step": 6721
+    },
+    {
+      "epoch": 0.2938065474889637,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.020938036865365e-05,
+      "loss": 2.0372,
+      "step": 6722
+    },
+    {
+      "epoch": 0.29385025569299356,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.020390683754161e-05,
+      "loss": 1.6673,
+      "step": 6723
+    },
+    {
+      "epoch": 0.2938939638970235,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.019843273643788e-05,
+      "loss": 1.8072,
+      "step": 6724
+    },
+    {
+      "epoch": 0.29393767210105337,
+      "grad_norm": 1.984375,
+      "learning_rate": 8.019295806544578e-05,
+      "loss": 1.9004,
+      "step": 6725
+    },
+    {
+      "epoch": 0.29398138030508325,
+      "grad_norm": 2.015625,
+      "learning_rate": 8.018748282466862e-05,
+      "loss": 1.6449,
+      "step": 6726
+    },
+    {
+      "epoch": 0.2940250885091132,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.018200701420971e-05,
+      "loss": 2.4514,
+      "step": 6727
+    },
+    {
+      "epoch": 0.29406879671314307,
+      "grad_norm": 4.28125,
+      "learning_rate": 8.017653063417241e-05,
+      "loss": 1.6911,
+      "step": 6728
+    },
+    {
+      "epoch": 0.29411250491717295,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.017105368466006e-05,
+      "loss": 1.3129,
+      "step": 6729
+    },
+    {
+      "epoch": 0.2941562131212028,
+      "grad_norm": 2.5,
+      "learning_rate": 8.016557616577601e-05,
+      "loss": 2.1801,
+      "step": 6730
+    },
+    {
+      "epoch": 0.29419992132523276,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.016009807762364e-05,
+      "loss": 1.9602,
+      "step": 6731
+    },
+    {
+      "epoch": 0.29424362952926264,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.015461942030631e-05,
+      "loss": 2.5422,
+      "step": 6732
+    },
+    {
+      "epoch": 0.2942873377332925,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.014914019392743e-05,
+      "loss": 1.9976,
+      "step": 6733
+    },
+    {
+      "epoch": 0.29433104593732246,
+      "grad_norm": 2.078125,
+      "learning_rate": 8.01436603985904e-05,
+      "loss": 1.7281,
+      "step": 6734
+    },
+    {
+      "epoch": 0.29437475414135233,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.013818003439861e-05,
+      "loss": 1.6319,
+      "step": 6735
+    },
+    {
+      "epoch": 0.2944184623453822,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.013269910145552e-05,
+      "loss": 1.9,
+      "step": 6736
+    },
+    {
+      "epoch": 0.29446217054941215,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.012721759986452e-05,
+      "loss": 2.5719,
+      "step": 6737
+    },
+    {
+      "epoch": 0.29450587875344203,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.01217355297291e-05,
+      "loss": 2.2032,
+      "step": 6738
+    },
+    {
+      "epoch": 0.2945495869574719,
+      "grad_norm": 4.0,
+      "learning_rate": 8.011625289115267e-05,
+      "loss": 2.4784,
+      "step": 6739
+    },
+    {
+      "epoch": 0.2945932951615018,
+      "grad_norm": 1.875,
+      "learning_rate": 8.011076968423872e-05,
+      "loss": 1.816,
+      "step": 6740
+    },
+    {
+      "epoch": 0.2946370033655317,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.010528590909073e-05,
+      "loss": 2.3117,
+      "step": 6741
+    },
+    {
+      "epoch": 0.2946807115695616,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.009980156581217e-05,
+      "loss": 1.533,
+      "step": 6742
+    },
+    {
+      "epoch": 0.2947244197735915,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.009431665450655e-05,
+      "loss": 2.4178,
+      "step": 6743
+    },
+    {
+      "epoch": 0.2947681279776214,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.008883117527738e-05,
+      "loss": 1.5071,
+      "step": 6744
+    },
+    {
+      "epoch": 0.2948118361816513,
+      "grad_norm": 2.03125,
+      "learning_rate": 8.008334512822817e-05,
+      "loss": 1.9468,
+      "step": 6745
+    },
+    {
+      "epoch": 0.2948555443856812,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.007785851346245e-05,
+      "loss": 1.5961,
+      "step": 6746
+    },
+    {
+      "epoch": 0.2948992525897111,
+      "grad_norm": 3.5625,
+      "learning_rate": 8.007237133108376e-05,
+      "loss": 2.2648,
+      "step": 6747
+    },
+    {
+      "epoch": 0.294942960793741,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.006688358119568e-05,
+      "loss": 1.8828,
+      "step": 6748
+    },
+    {
+      "epoch": 0.2949866689977709,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.006139526390172e-05,
+      "loss": 2.2068,
+      "step": 6749
+    },
+    {
+      "epoch": 0.29503037720180075,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.005590637930548e-05,
+      "loss": 2.3664,
+      "step": 6750
+    },
+    {
+      "epoch": 0.2950740854058307,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.005041692751055e-05,
+      "loss": 1.5957,
+      "step": 6751
+    },
+    {
+      "epoch": 0.29511779360986057,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.00449269086205e-05,
+      "loss": 1.9969,
+      "step": 6752
+    },
+    {
+      "epoch": 0.29516150181389045,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.003943632273898e-05,
+      "loss": 2.201,
+      "step": 6753
+    },
+    {
+      "epoch": 0.2952052100179204,
+      "grad_norm": 2.171875,
+      "learning_rate": 8.003394516996956e-05,
+      "loss": 2.0344,
+      "step": 6754
+    },
+    {
+      "epoch": 0.29524891822195026,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.002845345041589e-05,
+      "loss": 2.9183,
+      "step": 6755
+    },
+    {
+      "epoch": 0.29529262642598014,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.00229611641816e-05,
+      "loss": 1.9433,
+      "step": 6756
+    },
+    {
+      "epoch": 0.2953363346300101,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.001746831137032e-05,
+      "loss": 1.548,
+      "step": 6757
+    },
+    {
+      "epoch": 0.29538004283403996,
+      "grad_norm": 2.125,
+      "learning_rate": 8.001197489208572e-05,
+      "loss": 2.0139,
+      "step": 6758
+    },
+    {
+      "epoch": 0.29542375103806984,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.00064809064315e-05,
+      "loss": 1.7309,
+      "step": 6759
+    },
+    {
+      "epoch": 0.2954674592420997,
+      "grad_norm": 2.5,
+      "learning_rate": 8.00009863545113e-05,
+      "loss": 1.7731,
+      "step": 6760
+    },
+    {
+      "epoch": 0.29551116744612965,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.999549123642882e-05,
+      "loss": 1.9047,
+      "step": 6761
+    },
+    {
+      "epoch": 0.29555487565015953,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.998999555228777e-05,
+      "loss": 1.9589,
+      "step": 6762
+    },
+    {
+      "epoch": 0.2955985838541894,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.998449930219185e-05,
+      "loss": 3.0505,
+      "step": 6763
+    },
+    {
+      "epoch": 0.29564229205821935,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.997900248624479e-05,
+      "loss": 1.55,
+      "step": 6764
+    },
+    {
+      "epoch": 0.2956860002622492,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.997350510455032e-05,
+      "loss": 2.1873,
+      "step": 6765
+    },
+    {
+      "epoch": 0.2957297084662791,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.99680071572122e-05,
+      "loss": 1.4288,
+      "step": 6766
+    },
+    {
+      "epoch": 0.29577341667030904,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.996250864433415e-05,
+      "loss": 1.6355,
+      "step": 6767
+    },
+    {
+      "epoch": 0.2958171248743389,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.995700956601995e-05,
+      "loss": 1.9442,
+      "step": 6768
+    },
+    {
+      "epoch": 0.2958608330783688,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.995150992237339e-05,
+      "loss": 1.9044,
+      "step": 6769
+    },
+    {
+      "epoch": 0.2959045412823987,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.994600971349825e-05,
+      "loss": 2.1128,
+      "step": 6770
+    },
+    {
+      "epoch": 0.2959482494864286,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.994050893949832e-05,
+      "loss": 2.3428,
+      "step": 6771
+    },
+    {
+      "epoch": 0.2959919576904585,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.993500760047739e-05,
+      "loss": 2.0457,
+      "step": 6772
+    },
+    {
+      "epoch": 0.2960356658944884,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.992950569653932e-05,
+      "loss": 1.9142,
+      "step": 6773
+    },
+    {
+      "epoch": 0.2960793740985183,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.992400322778791e-05,
+      "loss": 1.6232,
+      "step": 6774
+    },
+    {
+      "epoch": 0.2961230823025482,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.9918500194327e-05,
+      "loss": 2.474,
+      "step": 6775
+    },
+    {
+      "epoch": 0.29616679050657807,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.991299659626046e-05,
+      "loss": 1.8029,
+      "step": 6776
+    },
+    {
+      "epoch": 0.296210498710608,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.990749243369214e-05,
+      "loss": 2.3675,
+      "step": 6777
+    },
+    {
+      "epoch": 0.2962542069146379,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.99019877067259e-05,
+      "loss": 1.9141,
+      "step": 6778
+    },
+    {
+      "epoch": 0.29629791511866777,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.989648241546563e-05,
+      "loss": 1.9746,
+      "step": 6779
+    },
+    {
+      "epoch": 0.29634162332269764,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.989097656001524e-05,
+      "loss": 2.2656,
+      "step": 6780
+    },
+    {
+      "epoch": 0.2963853315267276,
+      "grad_norm": 1.875,
+      "learning_rate": 7.98854701404786e-05,
+      "loss": 1.8483,
+      "step": 6781
+    },
+    {
+      "epoch": 0.29642903973075746,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.987996315695965e-05,
+      "loss": 2.1934,
+      "step": 6782
+    },
+    {
+      "epoch": 0.29647274793478734,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.98744556095623e-05,
+      "loss": 1.8816,
+      "step": 6783
+    },
+    {
+      "epoch": 0.2965164561388173,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.986894749839049e-05,
+      "loss": 1.7686,
+      "step": 6784
+    },
+    {
+      "epoch": 0.29656016434284715,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.986343882354818e-05,
+      "loss": 2.356,
+      "step": 6785
+    },
+    {
+      "epoch": 0.29660387254687703,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.985792958513931e-05,
+      "loss": 1.9254,
+      "step": 6786
+    },
+    {
+      "epoch": 0.29664758075090697,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.985241978326786e-05,
+      "loss": 2.4049,
+      "step": 6787
+    },
+    {
+      "epoch": 0.29669128895493685,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.984690941803779e-05,
+      "loss": 1.249,
+      "step": 6788
+    },
+    {
+      "epoch": 0.29673499715896673,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.984139848955309e-05,
+      "loss": 2.1914,
+      "step": 6789
+    },
+    {
+      "epoch": 0.2967787053629966,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.98358869979178e-05,
+      "loss": 1.8875,
+      "step": 6790
+    },
+    {
+      "epoch": 0.29682241356702654,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.983037494323588e-05,
+      "loss": 1.6059,
+      "step": 6791
+    },
+    {
+      "epoch": 0.2968661217710564,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.982486232561138e-05,
+      "loss": 1.8311,
+      "step": 6792
+    },
+    {
+      "epoch": 0.2969098299750863,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.981934914514829e-05,
+      "loss": 2.5336,
+      "step": 6793
+    },
+    {
+      "epoch": 0.29695353817911624,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.98138354019507e-05,
+      "loss": 1.7158,
+      "step": 6794
+    },
+    {
+      "epoch": 0.2969972463831461,
+      "grad_norm": 1.9375,
+      "learning_rate": 7.980832109612265e-05,
+      "loss": 1.6328,
+      "step": 6795
+    },
+    {
+      "epoch": 0.297040954587176,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.980280622776819e-05,
+      "loss": 1.5029,
+      "step": 6796
+    },
+    {
+      "epoch": 0.29708466279120593,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.979729079699141e-05,
+      "loss": 1.9137,
+      "step": 6797
+    },
+    {
+      "epoch": 0.2971283709952358,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.979177480389637e-05,
+      "loss": 1.9737,
+      "step": 6798
+    },
+    {
+      "epoch": 0.2971720791992657,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.978625824858719e-05,
+      "loss": 2.503,
+      "step": 6799
+    },
+    {
+      "epoch": 0.2972157874032956,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.978074113116796e-05,
+      "loss": 1.8997,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2972594956073255,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.977522345174281e-05,
+      "loss": 1.8808,
+      "step": 6801
+    },
+    {
+      "epoch": 0.2973032038113554,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.976970521041585e-05,
+      "loss": 2.4508,
+      "step": 6802
+    },
+    {
+      "epoch": 0.29734691201538527,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.976418640729122e-05,
+      "loss": 1.4971,
+      "step": 6803
+    },
+    {
+      "epoch": 0.2973906202194152,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.975866704247307e-05,
+      "loss": 1.9584,
+      "step": 6804
+    },
+    {
+      "epoch": 0.2974343284234451,
+      "grad_norm": 2.25,
+      "learning_rate": 7.975314711606558e-05,
+      "loss": 2.1464,
+      "step": 6805
+    },
+    {
+      "epoch": 0.29747803662747496,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.974762662817289e-05,
+      "loss": 1.552,
+      "step": 6806
+    },
+    {
+      "epoch": 0.2975217448315049,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.974210557889919e-05,
+      "loss": 2.6815,
+      "step": 6807
+    },
+    {
+      "epoch": 0.2975654530355348,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.973658396834866e-05,
+      "loss": 2.0794,
+      "step": 6808
+    },
+    {
+      "epoch": 0.29760916123956466,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.973106179662553e-05,
+      "loss": 2.1026,
+      "step": 6809
+    },
+    {
+      "epoch": 0.29765286944359454,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.972553906383398e-05,
+      "loss": 1.8778,
+      "step": 6810
+    },
+    {
+      "epoch": 0.29769657764762447,
+      "grad_norm": 2.375,
+      "learning_rate": 7.972001577007825e-05,
+      "loss": 2.0974,
+      "step": 6811
+    },
+    {
+      "epoch": 0.29774028585165435,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.971449191546256e-05,
+      "loss": 2.1475,
+      "step": 6812
+    },
+    {
+      "epoch": 0.29778399405568423,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.970896750009115e-05,
+      "loss": 1.5038,
+      "step": 6813
+    },
+    {
+      "epoch": 0.29782770225971417,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.970344252406831e-05,
+      "loss": 2.0435,
+      "step": 6814
+    },
+    {
+      "epoch": 0.29787141046374405,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.969791698749827e-05,
+      "loss": 1.9473,
+      "step": 6815
+    },
+    {
+      "epoch": 0.2979151186677739,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.969239089048531e-05,
+      "loss": 2.4137,
+      "step": 6816
+    },
+    {
+      "epoch": 0.29795882687180386,
+      "grad_norm": 2.125,
+      "learning_rate": 7.968686423313372e-05,
+      "loss": 2.2156,
+      "step": 6817
+    },
+    {
+      "epoch": 0.29800253507583374,
+      "grad_norm": 2.125,
+      "learning_rate": 7.968133701554779e-05,
+      "loss": 1.4572,
+      "step": 6818
+    },
+    {
+      "epoch": 0.2980462432798636,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.967580923783184e-05,
+      "loss": 2.623,
+      "step": 6819
+    },
+    {
+      "epoch": 0.2980899514838935,
+      "grad_norm": 1.890625,
+      "learning_rate": 7.967028090009016e-05,
+      "loss": 1.7034,
+      "step": 6820
+    },
+    {
+      "epoch": 0.29813365968792344,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.966475200242713e-05,
+      "loss": 1.4808,
+      "step": 6821
+    },
+    {
+      "epoch": 0.2981773678919533,
+      "grad_norm": 4.71875,
+      "learning_rate": 7.965922254494702e-05,
+      "loss": 1.6554,
+      "step": 6822
+    },
+    {
+      "epoch": 0.2982210760959832,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.965369252775424e-05,
+      "loss": 1.5935,
+      "step": 6823
+    },
+    {
+      "epoch": 0.29826478430001313,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.96481619509531e-05,
+      "loss": 1.3847,
+      "step": 6824
+    },
+    {
+      "epoch": 0.298308492504043,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.9642630814648e-05,
+      "loss": 2.2384,
+      "step": 6825
+    },
+    {
+      "epoch": 0.2983522007080729,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.963709911894333e-05,
+      "loss": 1.6818,
+      "step": 6826
+    },
+    {
+      "epoch": 0.2983959089121028,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.963156686394345e-05,
+      "loss": 1.734,
+      "step": 6827
+    },
+    {
+      "epoch": 0.2984396171161327,
+      "grad_norm": 2.375,
+      "learning_rate": 7.962603404975278e-05,
+      "loss": 2.9876,
+      "step": 6828
+    },
+    {
+      "epoch": 0.2984833253201626,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.962050067647573e-05,
+      "loss": 2.0571,
+      "step": 6829
+    },
+    {
+      "epoch": 0.29852703352419246,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.961496674421672e-05,
+      "loss": 1.4679,
+      "step": 6830
+    },
+    {
+      "epoch": 0.2985707417282224,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.960943225308019e-05,
+      "loss": 1.9087,
+      "step": 6831
+    },
+    {
+      "epoch": 0.2986144499322523,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.960389720317057e-05,
+      "loss": 1.6903,
+      "step": 6832
+    },
+    {
+      "epoch": 0.29865815813628216,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.959836159459231e-05,
+      "loss": 1.7672,
+      "step": 6833
+    },
+    {
+      "epoch": 0.2987018663403121,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.959282542744992e-05,
+      "loss": 1.6609,
+      "step": 6834
+    },
+    {
+      "epoch": 0.298745574544342,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.958728870184782e-05,
+      "loss": 1.8973,
+      "step": 6835
+    },
+    {
+      "epoch": 0.29878928274837185,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.958175141789054e-05,
+      "loss": 1.6439,
+      "step": 6836
+    },
+    {
+      "epoch": 0.2988329909524018,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.957621357568254e-05,
+      "loss": 1.9569,
+      "step": 6837
+    },
+    {
+      "epoch": 0.29887669915643167,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.957067517532835e-05,
+      "loss": 2.3333,
+      "step": 6838
+    },
+    {
+      "epoch": 0.29892040736046155,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.956513621693248e-05,
+      "loss": 1.7887,
+      "step": 6839
+    },
+    {
+      "epoch": 0.29896411556449143,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.955959670059947e-05,
+      "loss": 1.6348,
+      "step": 6840
+    },
+    {
+      "epoch": 0.29900782376852136,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.955405662643384e-05,
+      "loss": 2.1351,
+      "step": 6841
+    },
+    {
+      "epoch": 0.29905153197255124,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.954851599454014e-05,
+      "loss": 2.4259,
+      "step": 6842
+    },
+    {
+      "epoch": 0.2990952401765811,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.954297480502293e-05,
+      "loss": 1.6529,
+      "step": 6843
+    },
+    {
+      "epoch": 0.29913894838061106,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.953743305798682e-05,
+      "loss": 1.8452,
+      "step": 6844
+    },
+    {
+      "epoch": 0.29918265658464094,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.953189075353633e-05,
+      "loss": 1.8695,
+      "step": 6845
+    },
+    {
+      "epoch": 0.2992263647886708,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.95263478917761e-05,
+      "loss": 2.2973,
+      "step": 6846
+    },
+    {
+      "epoch": 0.29927007299270075,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.95208044728107e-05,
+      "loss": 1.5313,
+      "step": 6847
+    },
+    {
+      "epoch": 0.29931378119673063,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.951526049674475e-05,
+      "loss": 1.9208,
+      "step": 6848
+    },
+    {
+      "epoch": 0.2993574894007605,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.950971596368289e-05,
+      "loss": 1.8555,
+      "step": 6849
+    },
+    {
+      "epoch": 0.2994011976047904,
+      "grad_norm": 2.75,
+      "learning_rate": 7.950417087372972e-05,
+      "loss": 1.3852,
+      "step": 6850
+    },
+    {
+      "epoch": 0.2994449058088203,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.949862522698992e-05,
+      "loss": 2.5552,
+      "step": 6851
+    },
+    {
+      "epoch": 0.2994886140128502,
+      "grad_norm": 2.0,
+      "learning_rate": 7.949307902356813e-05,
+      "loss": 1.6658,
+      "step": 6852
+    },
+    {
+      "epoch": 0.2995323222168801,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.9487532263569e-05,
+      "loss": 2.0919,
+      "step": 6853
+    },
+    {
+      "epoch": 0.29957603042091,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.948198494709724e-05,
+      "loss": 1.8746,
+      "step": 6854
+    },
+    {
+      "epoch": 0.2996197386249399,
+      "grad_norm": 2.125,
+      "learning_rate": 7.947643707425749e-05,
+      "loss": 1.9121,
+      "step": 6855
+    },
+    {
+      "epoch": 0.2996634468289698,
+      "grad_norm": 2.875,
+      "learning_rate": 7.94708886451545e-05,
+      "loss": 2.0367,
+      "step": 6856
+    },
+    {
+      "epoch": 0.2997071550329997,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.946533965989293e-05,
+      "loss": 1.6858,
+      "step": 6857
+    },
+    {
+      "epoch": 0.2997508632370296,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.945979011857751e-05,
+      "loss": 2.0082,
+      "step": 6858
+    },
+    {
+      "epoch": 0.2997945714410595,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.945424002131298e-05,
+      "loss": 1.92,
+      "step": 6859
+    },
+    {
+      "epoch": 0.29983827964508936,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.944868936820408e-05,
+      "loss": 1.7948,
+      "step": 6860
+    },
+    {
+      "epoch": 0.2998819878491193,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.944313815935556e-05,
+      "loss": 1.986,
+      "step": 6861
+    },
+    {
+      "epoch": 0.29992569605314917,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.943758639487216e-05,
+      "loss": 1.5364,
+      "step": 6862
+    },
+    {
+      "epoch": 0.29996940425717905,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.943203407485864e-05,
+      "loss": 2.095,
+      "step": 6863
+    },
+    {
+      "epoch": 0.300013112461209,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.942648119941982e-05,
+      "loss": 1.8194,
+      "step": 6864
+    },
+    {
+      "epoch": 0.30005682066523887,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.942092776866048e-05,
+      "loss": 1.7115,
+      "step": 6865
+    },
+    {
+      "epoch": 0.30010052886926875,
+      "grad_norm": 2.5,
+      "learning_rate": 7.94153737826854e-05,
+      "loss": 1.4567,
+      "step": 6866
+    },
+    {
+      "epoch": 0.3001442370732987,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.94098192415994e-05,
+      "loss": 2.0076,
+      "step": 6867
+    },
+    {
+      "epoch": 0.30018794527732856,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.940426414550732e-05,
+      "loss": 1.7571,
+      "step": 6868
+    },
+    {
+      "epoch": 0.30023165348135844,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.939870849451398e-05,
+      "loss": 1.8405,
+      "step": 6869
+    },
+    {
+      "epoch": 0.3002753616853883,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.939315228872421e-05,
+      "loss": 2.0807,
+      "step": 6870
+    },
+    {
+      "epoch": 0.30031906988941826,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.938759552824288e-05,
+      "loss": 2.4804,
+      "step": 6871
+    },
+    {
+      "epoch": 0.30036277809344814,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.938203821317487e-05,
+      "loss": 2.3102,
+      "step": 6872
+    },
+    {
+      "epoch": 0.300406486297478,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.937648034362502e-05,
+      "loss": 1.4459,
+      "step": 6873
+    },
+    {
+      "epoch": 0.30045019450150795,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.937092191969821e-05,
+      "loss": 1.3184,
+      "step": 6874
+    },
+    {
+      "epoch": 0.30049390270553783,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.936536294149939e-05,
+      "loss": 1.9091,
+      "step": 6875
+    },
+    {
+      "epoch": 0.3005376109095677,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.935980340913342e-05,
+      "loss": 1.5597,
+      "step": 6876
+    },
+    {
+      "epoch": 0.30058131911359764,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.935424332270522e-05,
+      "loss": 1.9584,
+      "step": 6877
+    },
+    {
+      "epoch": 0.3006250273176275,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.934868268231973e-05,
+      "loss": 1.6071,
+      "step": 6878
+    },
+    {
+      "epoch": 0.3006687355216574,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.93431214880819e-05,
+      "loss": 2.0378,
+      "step": 6879
+    },
+    {
+      "epoch": 0.3007124437256873,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.933755974009663e-05,
+      "loss": 2.7808,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3007561519297172,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.933199743846893e-05,
+      "loss": 1.7805,
+      "step": 6881
+    },
+    {
+      "epoch": 0.3007998601337471,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.932643458330374e-05,
+      "loss": 2.1663,
+      "step": 6882
+    },
+    {
+      "epoch": 0.300843568337777,
+      "grad_norm": 2.25,
+      "learning_rate": 7.932087117470606e-05,
+      "loss": 1.845,
+      "step": 6883
+    },
+    {
+      "epoch": 0.3008872765418069,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.931530721278084e-05,
+      "loss": 1.5432,
+      "step": 6884
+    },
+    {
+      "epoch": 0.3009309847458368,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.930974269763313e-05,
+      "loss": 2.3063,
+      "step": 6885
+    },
+    {
+      "epoch": 0.3009746929498667,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.93041776293679e-05,
+      "loss": 2.1164,
+      "step": 6886
+    },
+    {
+      "epoch": 0.3010184011538966,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.929861200809021e-05,
+      "loss": 2.5617,
+      "step": 6887
+    },
+    {
+      "epoch": 0.3010621093579265,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.929304583390505e-05,
+      "loss": 2.2743,
+      "step": 6888
+    },
+    {
+      "epoch": 0.30110581756195637,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.92874791069175e-05,
+      "loss": 2.0906,
+      "step": 6889
+    },
+    {
+      "epoch": 0.30114952576598625,
+      "grad_norm": 2.0,
+      "learning_rate": 7.928191182723256e-05,
+      "loss": 1.8454,
+      "step": 6890
+    },
+    {
+      "epoch": 0.3011932339700162,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.927634399495536e-05,
+      "loss": 1.8599,
+      "step": 6891
+    },
+    {
+      "epoch": 0.30123694217404606,
+      "grad_norm": 1.875,
+      "learning_rate": 7.927077561019092e-05,
+      "loss": 1.3247,
+      "step": 6892
+    },
+    {
+      "epoch": 0.30128065037807594,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.926520667304434e-05,
+      "loss": 1.5736,
+      "step": 6893
+    },
+    {
+      "epoch": 0.3013243585821059,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.925963718362073e-05,
+      "loss": 1.8178,
+      "step": 6894
+    },
+    {
+      "epoch": 0.30136806678613576,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.925406714202517e-05,
+      "loss": 1.571,
+      "step": 6895
+    },
+    {
+      "epoch": 0.30141177499016564,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.924849654836281e-05,
+      "loss": 2.1494,
+      "step": 6896
+    },
+    {
+      "epoch": 0.3014554831941956,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.924292540273872e-05,
+      "loss": 1.5882,
+      "step": 6897
+    },
+    {
+      "epoch": 0.30149919139822545,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.923735370525809e-05,
+      "loss": 1.4472,
+      "step": 6898
+    },
+    {
+      "epoch": 0.30154289960225533,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.923178145602603e-05,
+      "loss": 2.0544,
+      "step": 6899
+    },
+    {
+      "epoch": 0.30158660780628527,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.922620865514772e-05,
+      "loss": 2.23,
+      "step": 6900
+    },
+    {
+      "epoch": 0.30163031601031515,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.92206353027283e-05,
+      "loss": 1.5023,
+      "step": 6901
+    },
+    {
+      "epoch": 0.301674024214345,
+      "grad_norm": 1.9375,
+      "learning_rate": 7.921506139887297e-05,
+      "loss": 1.8787,
+      "step": 6902
+    },
+    {
+      "epoch": 0.3017177324183749,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.92094869436869e-05,
+      "loss": 1.8779,
+      "step": 6903
+    },
+    {
+      "epoch": 0.30176144062240484,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.920391193727532e-05,
+      "loss": 2.8937,
+      "step": 6904
+    },
+    {
+      "epoch": 0.3018051488264347,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.91983363797434e-05,
+      "loss": 2.0161,
+      "step": 6905
+    },
+    {
+      "epoch": 0.3018488570304646,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.91927602711964e-05,
+      "loss": 1.6358,
+      "step": 6906
+    },
+    {
+      "epoch": 0.30189256523449454,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.91871836117395e-05,
+      "loss": 1.476,
+      "step": 6907
+    },
+    {
+      "epoch": 0.3019362734385244,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.918160640147798e-05,
+      "loss": 2.321,
+      "step": 6908
+    },
+    {
+      "epoch": 0.3019799816425543,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.917602864051706e-05,
+      "loss": 2.1206,
+      "step": 6909
+    },
+    {
+      "epoch": 0.30202368984658423,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.917045032896202e-05,
+      "loss": 2.1465,
+      "step": 6910
+    },
+    {
+      "epoch": 0.3020673980506141,
+      "grad_norm": 2.0,
+      "learning_rate": 7.916487146691815e-05,
+      "loss": 1.9398,
+      "step": 6911
+    },
+    {
+      "epoch": 0.302111106254644,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.915929205449069e-05,
+      "loss": 1.5202,
+      "step": 6912
+    },
+    {
+      "epoch": 0.30215481445867387,
+      "grad_norm": 1.75,
+      "learning_rate": 7.915371209178494e-05,
+      "loss": 1.6357,
+      "step": 6913
+    },
+    {
+      "epoch": 0.3021985226627038,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.914813157890623e-05,
+      "loss": 1.6771,
+      "step": 6914
+    },
+    {
+      "epoch": 0.3022422308667337,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.914255051595984e-05,
+      "loss": 2.1166,
+      "step": 6915
+    },
+    {
+      "epoch": 0.30228593907076357,
+      "grad_norm": 2.125,
+      "learning_rate": 7.913696890305112e-05,
+      "loss": 1.9647,
+      "step": 6916
+    },
+    {
+      "epoch": 0.3023296472747935,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.913138674028537e-05,
+      "loss": 1.7212,
+      "step": 6917
+    },
+    {
+      "epoch": 0.3023733554788234,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.912580402776797e-05,
+      "loss": 1.934,
+      "step": 6918
+    },
+    {
+      "epoch": 0.30241706368285326,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.912022076560426e-05,
+      "loss": 2.1029,
+      "step": 6919
+    },
+    {
+      "epoch": 0.3024607718868832,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.911463695389959e-05,
+      "loss": 2.3138,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3025044800909131,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.910905259275936e-05,
+      "loss": 1.9484,
+      "step": 6921
+    },
+    {
+      "epoch": 0.30254818829494295,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.910346768228894e-05,
+      "loss": 2.4706,
+      "step": 6922
+    },
+    {
+      "epoch": 0.30259189649897283,
+      "grad_norm": 2.25,
+      "learning_rate": 7.909788222259372e-05,
+      "loss": 2.3833,
+      "step": 6923
+    },
+    {
+      "epoch": 0.30263560470300277,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.90922962137791e-05,
+      "loss": 1.9278,
+      "step": 6924
+    },
+    {
+      "epoch": 0.30267931290703265,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.908670965595052e-05,
+      "loss": 2.091,
+      "step": 6925
+    },
+    {
+      "epoch": 0.30272302111106253,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.908112254921341e-05,
+      "loss": 1.9959,
+      "step": 6926
+    },
+    {
+      "epoch": 0.30276672931509246,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.907553489367316e-05,
+      "loss": 1.8171,
+      "step": 6927
+    },
+    {
+      "epoch": 0.30281043751912234,
+      "grad_norm": 2.5,
+      "learning_rate": 7.906994668943528e-05,
+      "loss": 1.744,
+      "step": 6928
+    },
+    {
+      "epoch": 0.3028541457231522,
+      "grad_norm": 3.0,
+      "learning_rate": 7.906435793660519e-05,
+      "loss": 2.5186,
+      "step": 6929
+    },
+    {
+      "epoch": 0.30289785392718216,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.905876863528834e-05,
+      "loss": 2.2941,
+      "step": 6930
+    },
+    {
+      "epoch": 0.30294156213121204,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.905317878559026e-05,
+      "loss": 1.5411,
+      "step": 6931
+    },
+    {
+      "epoch": 0.3029852703352419,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.90475883876164e-05,
+      "loss": 2.3315,
+      "step": 6932
+    },
+    {
+      "epoch": 0.3030289785392718,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.904199744147228e-05,
+      "loss": 2.2387,
+      "step": 6933
+    },
+    {
+      "epoch": 0.30307268674330173,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.903640594726339e-05,
+      "loss": 2.0719,
+      "step": 6934
+    },
+    {
+      "epoch": 0.3031163949473316,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.903081390509525e-05,
+      "loss": 1.5947,
+      "step": 6935
+    },
+    {
+      "epoch": 0.3031601031513615,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.902522131507341e-05,
+      "loss": 1.4249,
+      "step": 6936
+    },
+    {
+      "epoch": 0.30320381135539143,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.901962817730341e-05,
+      "loss": 1.5904,
+      "step": 6937
+    },
+    {
+      "epoch": 0.3032475195594213,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.901403449189077e-05,
+      "loss": 1.9359,
+      "step": 6938
+    },
+    {
+      "epoch": 0.3032912277634512,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.900844025894109e-05,
+      "loss": 2.1124,
+      "step": 6939
+    },
+    {
+      "epoch": 0.3033349359674811,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.900284547855991e-05,
+      "loss": 2.2528,
+      "step": 6940
+    },
+    {
+      "epoch": 0.303378644171511,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.899725015085285e-05,
+      "loss": 2.0718,
+      "step": 6941
+    },
+    {
+      "epoch": 0.3034223523755409,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.899165427592543e-05,
+      "loss": 2.1333,
+      "step": 6942
+    },
+    {
+      "epoch": 0.30346606057957076,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.898605785388334e-05,
+      "loss": 1.8171,
+      "step": 6943
+    },
+    {
+      "epoch": 0.3035097687836007,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.898046088483214e-05,
+      "loss": 2.5504,
+      "step": 6944
+    },
+    {
+      "epoch": 0.3035534769876306,
+      "grad_norm": 2.5,
+      "learning_rate": 7.897486336887746e-05,
+      "loss": 2.1916,
+      "step": 6945
+    },
+    {
+      "epoch": 0.30359718519166046,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.896926530612492e-05,
+      "loss": 1.616,
+      "step": 6946
+    },
+    {
+      "epoch": 0.3036408933956904,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.89636666966802e-05,
+      "loss": 1.3841,
+      "step": 6947
+    },
+    {
+      "epoch": 0.30368460159972027,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.895806754064893e-05,
+      "loss": 1.758,
+      "step": 6948
+    },
+    {
+      "epoch": 0.30372830980375015,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.895246783813677e-05,
+      "loss": 2.1847,
+      "step": 6949
+    },
+    {
+      "epoch": 0.3037720180077801,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.894686758924942e-05,
+      "loss": 1.5626,
+      "step": 6950
+    },
+    {
+      "epoch": 0.30381572621180997,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.894126679409254e-05,
+      "loss": 1.8753,
+      "step": 6951
+    },
+    {
+      "epoch": 0.30385943441583985,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.893566545277184e-05,
+      "loss": 1.9157,
+      "step": 6952
+    },
+    {
+      "epoch": 0.3039031426198697,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.893006356539303e-05,
+      "loss": 1.4586,
+      "step": 6953
+    },
+    {
+      "epoch": 0.30394685082389966,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.89244611320618e-05,
+      "loss": 2.8266,
+      "step": 6954
+    },
+    {
+      "epoch": 0.30399055902792954,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.891885815288388e-05,
+      "loss": 2.0882,
+      "step": 6955
+    },
+    {
+      "epoch": 0.3040342672319594,
+      "grad_norm": 3.0,
+      "learning_rate": 7.891325462796503e-05,
+      "loss": 3.2572,
+      "step": 6956
+    },
+    {
+      "epoch": 0.30407797543598936,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.890765055741098e-05,
+      "loss": 2.1291,
+      "step": 6957
+    },
+    {
+      "epoch": 0.30412168364001924,
+      "grad_norm": 12.75,
+      "learning_rate": 7.89020459413275e-05,
+      "loss": 2.1205,
+      "step": 6958
+    },
+    {
+      "epoch": 0.3041653918440491,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.889644077982033e-05,
+      "loss": 1.6894,
+      "step": 6959
+    },
+    {
+      "epoch": 0.30420910004807905,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.889083507299529e-05,
+      "loss": 1.7101,
+      "step": 6960
+    },
+    {
+      "epoch": 0.30425280825210893,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.888522882095813e-05,
+      "loss": 1.3514,
+      "step": 6961
+    },
+    {
+      "epoch": 0.3042965164561388,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.887962202381465e-05,
+      "loss": 1.7874,
+      "step": 6962
+    },
+    {
+      "epoch": 0.3043402246601687,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.887401468167068e-05,
+      "loss": 1.7584,
+      "step": 6963
+    },
+    {
+      "epoch": 0.3043839328641986,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.886840679463203e-05,
+      "loss": 1.8778,
+      "step": 6964
+    },
+    {
+      "epoch": 0.3044276410682285,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.886279836280454e-05,
+      "loss": 1.6752,
+      "step": 6965
+    },
+    {
+      "epoch": 0.3044713492722584,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.885718938629402e-05,
+      "loss": 1.838,
+      "step": 6966
+    },
+    {
+      "epoch": 0.3045150574762883,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.885157986520634e-05,
+      "loss": 2.1676,
+      "step": 6967
+    },
+    {
+      "epoch": 0.3045587656803182,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.884596979964736e-05,
+      "loss": 1.8915,
+      "step": 6968
+    },
+    {
+      "epoch": 0.3046024738843481,
+      "grad_norm": 1.875,
+      "learning_rate": 7.884035918972295e-05,
+      "loss": 1.6817,
+      "step": 6969
+    },
+    {
+      "epoch": 0.304646182088378,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.883474803553899e-05,
+      "loss": 1.9612,
+      "step": 6970
+    },
+    {
+      "epoch": 0.3046898902924079,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.882913633720135e-05,
+      "loss": 1.4549,
+      "step": 6971
+    },
+    {
+      "epoch": 0.3047335984964378,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.882352409481597e-05,
+      "loss": 1.7752,
+      "step": 6972
+    },
+    {
+      "epoch": 0.30477730670046765,
+      "grad_norm": 3.9375,
+      "learning_rate": 7.881791130848873e-05,
+      "loss": 1.3633,
+      "step": 6973
+    },
+    {
+      "epoch": 0.3048210149044976,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.881229797832554e-05,
+      "loss": 1.5878,
+      "step": 6974
+    },
+    {
+      "epoch": 0.30486472310852747,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.880668410443238e-05,
+      "loss": 2.0573,
+      "step": 6975
+    },
+    {
+      "epoch": 0.30490843131255735,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.880106968691517e-05,
+      "loss": 1.8925,
+      "step": 6976
+    },
+    {
+      "epoch": 0.3049521395165873,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.879545472587984e-05,
+      "loss": 1.6229,
+      "step": 6977
+    },
+    {
+      "epoch": 0.30499584772061716,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.878983922143237e-05,
+      "loss": 1.9097,
+      "step": 6978
+    },
+    {
+      "epoch": 0.30503955592464704,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.878422317367873e-05,
+      "loss": 1.9204,
+      "step": 6979
+    },
+    {
+      "epoch": 0.305083264128677,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.877860658272491e-05,
+      "loss": 2.0385,
+      "step": 6980
+    },
+    {
+      "epoch": 0.30512697233270686,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.87729894486769e-05,
+      "loss": 1.7698,
+      "step": 6981
+    },
+    {
+      "epoch": 0.30517068053673674,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.876737177164071e-05,
+      "loss": 2.0247,
+      "step": 6982
+    },
+    {
+      "epoch": 0.3052143887407666,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.876175355172234e-05,
+      "loss": 2.0221,
+      "step": 6983
+    },
+    {
+      "epoch": 0.30525809694479655,
+      "grad_norm": 2.0,
+      "learning_rate": 7.875613478902782e-05,
+      "loss": 1.7424,
+      "step": 6984
+    },
+    {
+      "epoch": 0.30530180514882643,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.875051548366316e-05,
+      "loss": 1.5619,
+      "step": 6985
+    },
+    {
+      "epoch": 0.3053455133528563,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.874489563573446e-05,
+      "loss": 1.806,
+      "step": 6986
+    },
+    {
+      "epoch": 0.30538922155688625,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.873927524534775e-05,
+      "loss": 2.3417,
+      "step": 6987
+    },
+    {
+      "epoch": 0.3054329297609161,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.873365431260906e-05,
+      "loss": 1.8671,
+      "step": 6988
+    },
+    {
+      "epoch": 0.305476637964946,
+      "grad_norm": 1.875,
+      "learning_rate": 7.87280328376245e-05,
+      "loss": 1.7924,
+      "step": 6989
+    },
+    {
+      "epoch": 0.30552034616897594,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.872241082050016e-05,
+      "loss": 1.9036,
+      "step": 6990
+    },
+    {
+      "epoch": 0.3055640543730058,
+      "grad_norm": 1.875,
+      "learning_rate": 7.871678826134211e-05,
+      "loss": 1.6127,
+      "step": 6991
+    },
+    {
+      "epoch": 0.3056077625770357,
+      "grad_norm": 2.5,
+      "learning_rate": 7.871116516025647e-05,
+      "loss": 1.8068,
+      "step": 6992
+    },
+    {
+      "epoch": 0.3056514707810656,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.870554151734937e-05,
+      "loss": 2.2059,
+      "step": 6993
+    },
+    {
+      "epoch": 0.3056951789850955,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.869991733272692e-05,
+      "loss": 1.8673,
+      "step": 6994
+    },
+    {
+      "epoch": 0.3057388871891254,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.869429260649526e-05,
+      "loss": 1.5146,
+      "step": 6995
+    },
+    {
+      "epoch": 0.3057825953931553,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.868866733876052e-05,
+      "loss": 1.1441,
+      "step": 6996
+    },
+    {
+      "epoch": 0.3058263035971852,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.868304152962889e-05,
+      "loss": 1.926,
+      "step": 6997
+    },
+    {
+      "epoch": 0.3058700118012151,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.867741517920653e-05,
+      "loss": 1.4671,
+      "step": 6998
+    },
+    {
+      "epoch": 0.30591372000524497,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.867178828759958e-05,
+      "loss": 2.3758,
+      "step": 6999
+    },
+    {
+      "epoch": 0.3059574282092749,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.866616085491426e-05,
+      "loss": 1.7675,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3060011364133048,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.866053288125678e-05,
+      "loss": 1.6345,
+      "step": 7001
+    },
+    {
+      "epoch": 0.30604484461733467,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.865490436673331e-05,
+      "loss": 1.8859,
+      "step": 7002
+    },
+    {
+      "epoch": 0.30608855282136455,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.864927531145011e-05,
+      "loss": 1.9555,
+      "step": 7003
+    },
+    {
+      "epoch": 0.3061322610253945,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.864364571551337e-05,
+      "loss": 1.4031,
+      "step": 7004
+    },
+    {
+      "epoch": 0.30617596922942436,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.863801557902936e-05,
+      "loss": 1.9749,
+      "step": 7005
+    },
+    {
+      "epoch": 0.30621967743345424,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.863238490210432e-05,
+      "loss": 1.5911,
+      "step": 7006
+    },
+    {
+      "epoch": 0.3062633856374842,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.862675368484449e-05,
+      "loss": 2.1096,
+      "step": 7007
+    },
+    {
+      "epoch": 0.30630709384151406,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.862112192735616e-05,
+      "loss": 1.6069,
+      "step": 7008
+    },
+    {
+      "epoch": 0.30635080204554394,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.86154896297456e-05,
+      "loss": 2.3773,
+      "step": 7009
+    },
+    {
+      "epoch": 0.30639451024957387,
+      "grad_norm": 2.125,
+      "learning_rate": 7.86098567921191e-05,
+      "loss": 1.9284,
+      "step": 7010
+    },
+    {
+      "epoch": 0.30643821845360375,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.860422341458298e-05,
+      "loss": 1.9852,
+      "step": 7011
+    },
+    {
+      "epoch": 0.30648192665763363,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.859858949724351e-05,
+      "loss": 1.8512,
+      "step": 7012
+    },
+    {
+      "epoch": 0.3065256348616635,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.859295504020706e-05,
+      "loss": 2.174,
+      "step": 7013
+    },
+    {
+      "epoch": 0.30656934306569344,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.858732004357989e-05,
+      "loss": 1.7676,
+      "step": 7014
+    },
+    {
+      "epoch": 0.3066130512697233,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.85816845074684e-05,
+      "loss": 1.7165,
+      "step": 7015
+    },
+    {
+      "epoch": 0.3066567594737532,
+      "grad_norm": 3.40625,
+      "learning_rate": 7.857604843197896e-05,
+      "loss": 2.595,
+      "step": 7016
+    },
+    {
+      "epoch": 0.30670046767778314,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.857041181721787e-05,
+      "loss": 3.0552,
+      "step": 7017
+    },
+    {
+      "epoch": 0.306744175881813,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.856477466329152e-05,
+      "loss": 1.7754,
+      "step": 7018
+    },
+    {
+      "epoch": 0.3067878840858429,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.85591369703063e-05,
+      "loss": 1.7161,
+      "step": 7019
+    },
+    {
+      "epoch": 0.30683159228987283,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.855349873836862e-05,
+      "loss": 2.4942,
+      "step": 7020
+    },
+    {
+      "epoch": 0.3068753004939027,
+      "grad_norm": 1.875,
+      "learning_rate": 7.854785996758485e-05,
+      "loss": 1.9154,
+      "step": 7021
+    },
+    {
+      "epoch": 0.3069190086979326,
+      "grad_norm": 1.8359375,
+      "learning_rate": 7.854222065806141e-05,
+      "loss": 1.6528,
+      "step": 7022
+    },
+    {
+      "epoch": 0.3069627169019625,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.853658080990471e-05,
+      "loss": 1.9866,
+      "step": 7023
+    },
+    {
+      "epoch": 0.3070064251059924,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.853094042322121e-05,
+      "loss": 1.3854,
+      "step": 7024
+    },
+    {
+      "epoch": 0.3070501333100223,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.852529949811734e-05,
+      "loss": 2.4263,
+      "step": 7025
+    },
+    {
+      "epoch": 0.30709384151405217,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.851965803469956e-05,
+      "loss": 1.3436,
+      "step": 7026
+    },
+    {
+      "epoch": 0.3071375497180821,
+      "grad_norm": 10.875,
+      "learning_rate": 7.851401603307431e-05,
+      "loss": 5.592,
+      "step": 7027
+    },
+    {
+      "epoch": 0.307181257922112,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.85083734933481e-05,
+      "loss": 1.9178,
+      "step": 7028
+    },
+    {
+      "epoch": 0.30722496612614186,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.850273041562737e-05,
+      "loss": 2.4896,
+      "step": 7029
+    },
+    {
+      "epoch": 0.3072686743301718,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.849708680001862e-05,
+      "loss": 1.8103,
+      "step": 7030
+    },
+    {
+      "epoch": 0.3073123825342017,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.84914426466284e-05,
+      "loss": 2.4037,
+      "step": 7031
+    },
+    {
+      "epoch": 0.30735609073823156,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.848579795556316e-05,
+      "loss": 1.6429,
+      "step": 7032
+    },
+    {
+      "epoch": 0.30739979894226144,
+      "grad_norm": 1.84375,
+      "learning_rate": 7.848015272692947e-05,
+      "loss": 1.5081,
+      "step": 7033
+    },
+    {
+      "epoch": 0.3074435071462914,
+      "grad_norm": 3.5625,
+      "learning_rate": 7.847450696083385e-05,
+      "loss": 2.2542,
+      "step": 7034
+    },
+    {
+      "epoch": 0.30748721535032125,
+      "grad_norm": 2.5,
+      "learning_rate": 7.846886065738284e-05,
+      "loss": 1.5479,
+      "step": 7035
+    },
+    {
+      "epoch": 0.30753092355435113,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.846321381668298e-05,
+      "loss": 2.4093,
+      "step": 7036
+    },
+    {
+      "epoch": 0.30757463175838107,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.845756643884087e-05,
+      "loss": 1.6931,
+      "step": 7037
+    },
+    {
+      "epoch": 0.30761833996241095,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.845191852396305e-05,
+      "loss": 2.3507,
+      "step": 7038
+    },
+    {
+      "epoch": 0.3076620481664408,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.844627007215613e-05,
+      "loss": 1.7684,
+      "step": 7039
+    },
+    {
+      "epoch": 0.30770575637047076,
+      "grad_norm": 2.25,
+      "learning_rate": 7.844062108352668e-05,
+      "loss": 1.9573,
+      "step": 7040
+    },
+    {
+      "epoch": 0.30774946457450064,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.843497155818132e-05,
+      "loss": 2.1727,
+      "step": 7041
+    },
+    {
+      "epoch": 0.3077931727785305,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.842932149622666e-05,
+      "loss": 1.7293,
+      "step": 7042
+    },
+    {
+      "epoch": 0.3078368809825604,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.842367089776932e-05,
+      "loss": 1.8516,
+      "step": 7043
+    },
+    {
+      "epoch": 0.30788058918659034,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.841801976291595e-05,
+      "loss": 1.437,
+      "step": 7044
+    },
+    {
+      "epoch": 0.3079242973906202,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.841236809177317e-05,
+      "loss": 1.8,
+      "step": 7045
+    },
+    {
+      "epoch": 0.3079680055946501,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.840671588444768e-05,
+      "loss": 1.8159,
+      "step": 7046
+    },
+    {
+      "epoch": 0.30801171379868003,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.84010631410461e-05,
+      "loss": 2.0276,
+      "step": 7047
+    },
+    {
+      "epoch": 0.3080554220027099,
+      "grad_norm": 2.25,
+      "learning_rate": 7.839540986167514e-05,
+      "loss": 1.6176,
+      "step": 7048
+    },
+    {
+      "epoch": 0.3080991302067398,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.838975604644146e-05,
+      "loss": 2.1571,
+      "step": 7049
+    },
+    {
+      "epoch": 0.3081428384107697,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.838410169545176e-05,
+      "loss": 1.8845,
+      "step": 7050
+    },
+    {
+      "epoch": 0.3081865466147996,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.837844680881274e-05,
+      "loss": 1.7899,
+      "step": 7051
+    },
+    {
+      "epoch": 0.3082302548188295,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.837279138663114e-05,
+      "loss": 2.2053,
+      "step": 7052
+    },
+    {
+      "epoch": 0.30827396302285937,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.836713542901366e-05,
+      "loss": 2.0172,
+      "step": 7053
+    },
+    {
+      "epoch": 0.3083176712268893,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.836147893606707e-05,
+      "loss": 2.0153,
+      "step": 7054
+    },
+    {
+      "epoch": 0.3083613794309192,
+      "grad_norm": 3.25,
+      "learning_rate": 7.835582190789807e-05,
+      "loss": 2.5625,
+      "step": 7055
+    },
+    {
+      "epoch": 0.30840508763494906,
+      "grad_norm": 10.0625,
+      "learning_rate": 7.835016434461345e-05,
+      "loss": 1.7711,
+      "step": 7056
+    },
+    {
+      "epoch": 0.308448795838979,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.834450624631996e-05,
+      "loss": 1.8224,
+      "step": 7057
+    },
+    {
+      "epoch": 0.3084925040430089,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.83388476131244e-05,
+      "loss": 1.7466,
+      "step": 7058
+    },
+    {
+      "epoch": 0.30853621224703875,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.833318844513353e-05,
+      "loss": 2.2057,
+      "step": 7059
+    },
+    {
+      "epoch": 0.3085799204510687,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.832752874245415e-05,
+      "loss": 1.783,
+      "step": 7060
+    },
+    {
+      "epoch": 0.30862362865509857,
+      "grad_norm": 1.890625,
+      "learning_rate": 7.83218685051931e-05,
+      "loss": 1.6596,
+      "step": 7061
+    },
+    {
+      "epoch": 0.30866733685912845,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.831620773345715e-05,
+      "loss": 1.66,
+      "step": 7062
+    },
+    {
+      "epoch": 0.30871104506315833,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.831054642735315e-05,
+      "loss": 1.8626,
+      "step": 7063
+    },
+    {
+      "epoch": 0.30875475326718826,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.830488458698794e-05,
+      "loss": 1.6357,
+      "step": 7064
+    },
+    {
+      "epoch": 0.30879846147121814,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.829922221246835e-05,
+      "loss": 1.7975,
+      "step": 7065
+    },
+    {
+      "epoch": 0.308842169675248,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.829355930390125e-05,
+      "loss": 1.8929,
+      "step": 7066
+    },
+    {
+      "epoch": 0.30888587787927796,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.828789586139352e-05,
+      "loss": 1.6512,
+      "step": 7067
+    },
+    {
+      "epoch": 0.30892958608330784,
+      "grad_norm": 1.890625,
+      "learning_rate": 7.828223188505202e-05,
+      "loss": 2.0205,
+      "step": 7068
+    },
+    {
+      "epoch": 0.3089732942873377,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.827656737498365e-05,
+      "loss": 1.6325,
+      "step": 7069
+    },
+    {
+      "epoch": 0.30901700249136765,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.827090233129528e-05,
+      "loss": 1.7217,
+      "step": 7070
+    },
+    {
+      "epoch": 0.30906071069539753,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.826523675409385e-05,
+      "loss": 1.4903,
+      "step": 7071
+    },
+    {
+      "epoch": 0.3091044188994274,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.825957064348625e-05,
+      "loss": 1.8773,
+      "step": 7072
+    },
+    {
+      "epoch": 0.3091481271034573,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.825390399957944e-05,
+      "loss": 2.1115,
+      "step": 7073
+    },
+    {
+      "epoch": 0.30919183530748723,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.824823682248033e-05,
+      "loss": 1.6687,
+      "step": 7074
+    },
+    {
+      "epoch": 0.3092355435115171,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.824256911229588e-05,
+      "loss": 1.6351,
+      "step": 7075
+    },
+    {
+      "epoch": 0.309279251715547,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.823690086913305e-05,
+      "loss": 1.7782,
+      "step": 7076
+    },
+    {
+      "epoch": 0.3093229599195769,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.82312320930988e-05,
+      "loss": 1.6454,
+      "step": 7077
+    },
+    {
+      "epoch": 0.3093666681236068,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.822556278430011e-05,
+      "loss": 1.8831,
+      "step": 7078
+    },
+    {
+      "epoch": 0.3094103763276367,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.821989294284397e-05,
+      "loss": 2.0986,
+      "step": 7079
+    },
+    {
+      "epoch": 0.3094540845316666,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.821422256883736e-05,
+      "loss": 1.6004,
+      "step": 7080
+    },
+    {
+      "epoch": 0.3094977927356965,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.820855166238734e-05,
+      "loss": 1.667,
+      "step": 7081
+    },
+    {
+      "epoch": 0.3095415009397264,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.820288022360087e-05,
+      "loss": 1.7269,
+      "step": 7082
+    },
+    {
+      "epoch": 0.30958520914375626,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.819720825258501e-05,
+      "loss": 2.1683,
+      "step": 7083
+    },
+    {
+      "epoch": 0.3096289173477862,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.81915357494468e-05,
+      "loss": 1.5796,
+      "step": 7084
+    },
+    {
+      "epoch": 0.30967262555181607,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.818586271429327e-05,
+      "loss": 1.6388,
+      "step": 7085
+    },
+    {
+      "epoch": 0.30971633375584595,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.818018914723149e-05,
+      "loss": 1.7634,
+      "step": 7086
+    },
+    {
+      "epoch": 0.3097600419598759,
+      "grad_norm": 5.40625,
+      "learning_rate": 7.817451504836852e-05,
+      "loss": 1.5701,
+      "step": 7087
+    },
+    {
+      "epoch": 0.30980375016390577,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.816884041781148e-05,
+      "loss": 1.5761,
+      "step": 7088
+    },
+    {
+      "epoch": 0.30984745836793565,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.816316525566738e-05,
+      "loss": 1.9034,
+      "step": 7089
+    },
+    {
+      "epoch": 0.3098911665719656,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.815748956204337e-05,
+      "loss": 1.9308,
+      "step": 7090
+    },
+    {
+      "epoch": 0.30993487477599546,
+      "grad_norm": 2.0,
+      "learning_rate": 7.815181333704656e-05,
+      "loss": 1.768,
+      "step": 7091
+    },
+    {
+      "epoch": 0.30997858298002534,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.814613658078407e-05,
+      "loss": 2.7116,
+      "step": 7092
+    },
+    {
+      "epoch": 0.3100222911840552,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.814045929336299e-05,
+      "loss": 1.9427,
+      "step": 7093
+    },
+    {
+      "epoch": 0.31006599938808516,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.813478147489052e-05,
+      "loss": 2.0114,
+      "step": 7094
+    },
+    {
+      "epoch": 0.31010970759211504,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.812910312547375e-05,
+      "loss": 2.1606,
+      "step": 7095
+    },
+    {
+      "epoch": 0.3101534157961449,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.812342424521988e-05,
+      "loss": 2.1392,
+      "step": 7096
+    },
+    {
+      "epoch": 0.31019712400017485,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.811774483423605e-05,
+      "loss": 1.721,
+      "step": 7097
+    },
+    {
+      "epoch": 0.31024083220420473,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.811206489262945e-05,
+      "loss": 1.7087,
+      "step": 7098
+    },
+    {
+      "epoch": 0.3102845404082346,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.810638442050728e-05,
+      "loss": 1.376,
+      "step": 7099
+    },
+    {
+      "epoch": 0.31032824861226455,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.810070341797673e-05,
+      "loss": 1.9291,
+      "step": 7100
+    },
+    {
+      "epoch": 0.3103719568162944,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.8095021885145e-05,
+      "loss": 2.0947,
+      "step": 7101
+    },
+    {
+      "epoch": 0.3104156650203243,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.808933982211933e-05,
+      "loss": 1.6669,
+      "step": 7102
+    },
+    {
+      "epoch": 0.3104593732243542,
+      "grad_norm": 1.890625,
+      "learning_rate": 7.808365722900693e-05,
+      "loss": 1.8462,
+      "step": 7103
+    },
+    {
+      "epoch": 0.3105030814283841,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.807797410591504e-05,
+      "loss": 1.6853,
+      "step": 7104
+    },
+    {
+      "epoch": 0.310546789632414,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.80722904529509e-05,
+      "loss": 1.6934,
+      "step": 7105
+    },
+    {
+      "epoch": 0.3105904978364439,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.80666062702218e-05,
+      "loss": 1.8148,
+      "step": 7106
+    },
+    {
+      "epoch": 0.3106342060404738,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.806092155783497e-05,
+      "loss": 2.3154,
+      "step": 7107
+    },
+    {
+      "epoch": 0.3106779142445037,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.805523631589774e-05,
+      "loss": 2.4815,
+      "step": 7108
+    },
+    {
+      "epoch": 0.3107216224485336,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.804955054451735e-05,
+      "loss": 2.6464,
+      "step": 7109
+    },
+    {
+      "epoch": 0.3107653306525635,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.804386424380113e-05,
+      "loss": 1.8508,
+      "step": 7110
+    },
+    {
+      "epoch": 0.3108090388565934,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.803817741385635e-05,
+      "loss": 1.4178,
+      "step": 7111
+    },
+    {
+      "epoch": 0.31085274706062327,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.803249005479037e-05,
+      "loss": 1.8201,
+      "step": 7112
+    },
+    {
+      "epoch": 0.31089645526465315,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.802680216671053e-05,
+      "loss": 2.3145,
+      "step": 7113
+    },
+    {
+      "epoch": 0.3109401634686831,
+      "grad_norm": 3.390625,
+      "learning_rate": 7.80211137497241e-05,
+      "loss": 1.7484,
+      "step": 7114
+    },
+    {
+      "epoch": 0.31098387167271296,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.801542480393849e-05,
+      "loss": 2.0886,
+      "step": 7115
+    },
+    {
+      "epoch": 0.31102757987674284,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.800973532946104e-05,
+      "loss": 1.5976,
+      "step": 7116
+    },
+    {
+      "epoch": 0.3110712880807728,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.800404532639911e-05,
+      "loss": 2.2827,
+      "step": 7117
+    },
+    {
+      "epoch": 0.31111499628480266,
+      "grad_norm": 2.0,
+      "learning_rate": 7.799835479486008e-05,
+      "loss": 1.6772,
+      "step": 7118
+    },
+    {
+      "epoch": 0.31115870448883254,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.799266373495137e-05,
+      "loss": 1.6108,
+      "step": 7119
+    },
+    {
+      "epoch": 0.3112024126928625,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.798697214678032e-05,
+      "loss": 2.0943,
+      "step": 7120
+    },
+    {
+      "epoch": 0.31124612089689235,
+      "grad_norm": 2.0,
+      "learning_rate": 7.79812800304544e-05,
+      "loss": 1.4641,
+      "step": 7121
+    },
+    {
+      "epoch": 0.31128982910092223,
+      "grad_norm": 2.0,
+      "learning_rate": 7.797558738608099e-05,
+      "loss": 1.6022,
+      "step": 7122
+    },
+    {
+      "epoch": 0.3113335373049521,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.796989421376755e-05,
+      "loss": 1.8272,
+      "step": 7123
+    },
+    {
+      "epoch": 0.31137724550898205,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.796420051362148e-05,
+      "loss": 2.2746,
+      "step": 7124
+    },
+    {
+      "epoch": 0.3114209537130119,
+      "grad_norm": 2.5,
+      "learning_rate": 7.795850628575024e-05,
+      "loss": 1.6013,
+      "step": 7125
+    },
+    {
+      "epoch": 0.3114646619170418,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.79528115302613e-05,
+      "loss": 2.7004,
+      "step": 7126
+    },
+    {
+      "epoch": 0.31150837012107174,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.794711624726213e-05,
+      "loss": 1.7298,
+      "step": 7127
+    },
+    {
+      "epoch": 0.3115520783251016,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.79414204368602e-05,
+      "loss": 1.7595,
+      "step": 7128
+    },
+    {
+      "epoch": 0.3115957865291315,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.7935724099163e-05,
+      "loss": 1.7804,
+      "step": 7129
+    },
+    {
+      "epoch": 0.31163949473316144,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.793002723427802e-05,
+      "loss": 1.7247,
+      "step": 7130
+    },
+    {
+      "epoch": 0.3116832029371913,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.792432984231277e-05,
+      "loss": 1.3776,
+      "step": 7131
+    },
+    {
+      "epoch": 0.3117269111412212,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.791863192337479e-05,
+      "loss": 2.9035,
+      "step": 7132
+    },
+    {
+      "epoch": 0.3117706193452511,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.791293347757159e-05,
+      "loss": 2.0015,
+      "step": 7133
+    },
+    {
+      "epoch": 0.311814327549281,
+      "grad_norm": 2.25,
+      "learning_rate": 7.79072345050107e-05,
+      "loss": 1.801,
+      "step": 7134
+    },
+    {
+      "epoch": 0.3118580357533109,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.790153500579968e-05,
+      "loss": 1.9849,
+      "step": 7135
+    },
+    {
+      "epoch": 0.31190174395734077,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.78958349800461e-05,
+      "loss": 1.977,
+      "step": 7136
+    },
+    {
+      "epoch": 0.3119454521613707,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.789013442785749e-05,
+      "loss": 1.6839,
+      "step": 7137
+    },
+    {
+      "epoch": 0.3119891603654006,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.788443334934148e-05,
+      "loss": 1.6571,
+      "step": 7138
+    },
+    {
+      "epoch": 0.31203286856943047,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.78787317446056e-05,
+      "loss": 2.0375,
+      "step": 7139
+    },
+    {
+      "epoch": 0.3120765767734604,
+      "grad_norm": 1.90625,
+      "learning_rate": 7.787302961375748e-05,
+      "loss": 1.9974,
+      "step": 7140
+    },
+    {
+      "epoch": 0.3121202849774903,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.786732695690475e-05,
+      "loss": 2.2131,
+      "step": 7141
+    },
+    {
+      "epoch": 0.31216399318152016,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.786162377415497e-05,
+      "loss": 2.4709,
+      "step": 7142
+    },
+    {
+      "epoch": 0.31220770138555004,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.785592006561582e-05,
+      "loss": 1.3958,
+      "step": 7143
+    },
+    {
+      "epoch": 0.31225140958958,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.78502158313949e-05,
+      "loss": 1.7984,
+      "step": 7144
+    },
+    {
+      "epoch": 0.31229511779360986,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.784451107159988e-05,
+      "loss": 1.6409,
+      "step": 7145
+    },
+    {
+      "epoch": 0.31233882599763974,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.783880578633843e-05,
+      "loss": 2.3524,
+      "step": 7146
+    },
+    {
+      "epoch": 0.31238253420166967,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.783309997571819e-05,
+      "loss": 1.5622,
+      "step": 7147
+    },
+    {
+      "epoch": 0.31242624240569955,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.782739363984683e-05,
+      "loss": 1.5084,
+      "step": 7148
+    },
+    {
+      "epoch": 0.31246995060972943,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.782168677883206e-05,
+      "loss": 1.8476,
+      "step": 7149
+    },
+    {
+      "epoch": 0.31251365881375937,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.781597939278156e-05,
+      "loss": 2.3668,
+      "step": 7150
+    },
+    {
+      "epoch": 0.31255736701778924,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.781027148180304e-05,
+      "loss": 2.4507,
+      "step": 7151
+    },
+    {
+      "epoch": 0.3126010752218191,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.780456304600423e-05,
+      "loss": 2.6835,
+      "step": 7152
+    },
+    {
+      "epoch": 0.312644783425849,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.779885408549286e-05,
+      "loss": 1.8557,
+      "step": 7153
+    },
+    {
+      "epoch": 0.31268849162987894,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.779314460037663e-05,
+      "loss": 1.6768,
+      "step": 7154
+    },
+    {
+      "epoch": 0.3127321998339088,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.778743459076333e-05,
+      "loss": 2.9159,
+      "step": 7155
+    },
+    {
+      "epoch": 0.3127759080379387,
+      "grad_norm": 3.375,
+      "learning_rate": 7.778172405676068e-05,
+      "loss": 1.871,
+      "step": 7156
+    },
+    {
+      "epoch": 0.31281961624196863,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.777601299847648e-05,
+      "loss": 1.6686,
+      "step": 7157
+    },
+    {
+      "epoch": 0.3128633244459985,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.777030141601848e-05,
+      "loss": 2.1794,
+      "step": 7158
+    },
+    {
+      "epoch": 0.3129070326500284,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.776458930949446e-05,
+      "loss": 1.9874,
+      "step": 7159
+    },
+    {
+      "epoch": 0.31295074085405833,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.775887667901225e-05,
+      "loss": 2.0633,
+      "step": 7160
+    },
+    {
+      "epoch": 0.3129944490580882,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.775316352467962e-05,
+      "loss": 1.8188,
+      "step": 7161
+    },
+    {
+      "epoch": 0.3130381572621181,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.774744984660442e-05,
+      "loss": 1.6672,
+      "step": 7162
+    },
+    {
+      "epoch": 0.31308186546614797,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.774173564489445e-05,
+      "loss": 2.5734,
+      "step": 7163
+    },
+    {
+      "epoch": 0.3131255736701779,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.773602091965754e-05,
+      "loss": 2.3279,
+      "step": 7164
+    },
+    {
+      "epoch": 0.3131692818742078,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.773030567100157e-05,
+      "loss": 1.6927,
+      "step": 7165
+    },
+    {
+      "epoch": 0.31321299007823766,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.772458989903437e-05,
+      "loss": 2.3808,
+      "step": 7166
+    },
+    {
+      "epoch": 0.3132566982822676,
+      "grad_norm": 2.75,
+      "learning_rate": 7.771887360386379e-05,
+      "loss": 2.2934,
+      "step": 7167
+    },
+    {
+      "epoch": 0.3133004064862975,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.771315678559774e-05,
+      "loss": 2.1141,
+      "step": 7168
+    },
+    {
+      "epoch": 0.31334411469032736,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.770743944434407e-05,
+      "loss": 1.8697,
+      "step": 7169
+    },
+    {
+      "epoch": 0.3133878228943573,
+      "grad_norm": 2.0,
+      "learning_rate": 7.77017215802107e-05,
+      "loss": 2.1161,
+      "step": 7170
+    },
+    {
+      "epoch": 0.3134315310983872,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.769600319330552e-05,
+      "loss": 2.0299,
+      "step": 7171
+    },
+    {
+      "epoch": 0.31347523930241705,
+      "grad_norm": 2.5,
+      "learning_rate": 7.769028428373645e-05,
+      "loss": 2.1765,
+      "step": 7172
+    },
+    {
+      "epoch": 0.313518947506447,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.768456485161142e-05,
+      "loss": 1.7941,
+      "step": 7173
+    },
+    {
+      "epoch": 0.31356265571047687,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.767884489703836e-05,
+      "loss": 2.2785,
+      "step": 7174
+    },
+    {
+      "epoch": 0.31360636391450675,
+      "grad_norm": 1.875,
+      "learning_rate": 7.76731244201252e-05,
+      "loss": 1.3323,
+      "step": 7175
+    },
+    {
+      "epoch": 0.3136500721185366,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.766740342097992e-05,
+      "loss": 1.9032,
+      "step": 7176
+    },
+    {
+      "epoch": 0.31369378032256656,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.766168189971046e-05,
+      "loss": 1.7193,
+      "step": 7177
+    },
+    {
+      "epoch": 0.31373748852659644,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.765595985642483e-05,
+      "loss": 2.1914,
+      "step": 7178
+    },
+    {
+      "epoch": 0.3137811967306263,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.765023729123095e-05,
+      "loss": 1.8275,
+      "step": 7179
+    },
+    {
+      "epoch": 0.31382490493465626,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.764451420423687e-05,
+      "loss": 2.1466,
+      "step": 7180
+    },
+    {
+      "epoch": 0.31386861313868614,
+      "grad_norm": 2.25,
+      "learning_rate": 7.763879059555055e-05,
+      "loss": 2.0121,
+      "step": 7181
+    },
+    {
+      "epoch": 0.313912321342716,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.763306646528004e-05,
+      "loss": 2.1565,
+      "step": 7182
+    },
+    {
+      "epoch": 0.31395602954674595,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.762734181353335e-05,
+      "loss": 2.0406,
+      "step": 7183
+    },
+    {
+      "epoch": 0.31399973775077583,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.762161664041852e-05,
+      "loss": 1.7566,
+      "step": 7184
+    },
+    {
+      "epoch": 0.3140434459548057,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.761589094604357e-05,
+      "loss": 1.4868,
+      "step": 7185
+    },
+    {
+      "epoch": 0.3140871541588356,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.761016473051655e-05,
+      "loss": 1.4814,
+      "step": 7186
+    },
+    {
+      "epoch": 0.3141308623628655,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.760443799394557e-05,
+      "loss": 1.6667,
+      "step": 7187
+    },
+    {
+      "epoch": 0.3141745705668954,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.759871073643865e-05,
+      "loss": 2.9772,
+      "step": 7188
+    },
+    {
+      "epoch": 0.3142182787709253,
+      "grad_norm": 2.25,
+      "learning_rate": 7.75929829581039e-05,
+      "loss": 1.7509,
+      "step": 7189
+    },
+    {
+      "epoch": 0.3142619869749552,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.758725465904938e-05,
+      "loss": 1.8446,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3143056951789851,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.758152583938323e-05,
+      "loss": 2.0265,
+      "step": 7191
+    },
+    {
+      "epoch": 0.314349403383015,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.757579649921354e-05,
+      "loss": 1.84,
+      "step": 7192
+    },
+    {
+      "epoch": 0.3143931115870449,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.757006663864843e-05,
+      "loss": 1.6102,
+      "step": 7193
+    },
+    {
+      "epoch": 0.3144368197910748,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.756433625779604e-05,
+      "loss": 2.7256,
+      "step": 7194
+    },
+    {
+      "epoch": 0.3144805279951047,
+      "grad_norm": 2.25,
+      "learning_rate": 7.755860535676452e-05,
+      "loss": 2.0873,
+      "step": 7195
+    },
+    {
+      "epoch": 0.31452423619913455,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.755287393566199e-05,
+      "loss": 2.017,
+      "step": 7196
+    },
+    {
+      "epoch": 0.3145679444031645,
+      "grad_norm": 2.125,
+      "learning_rate": 7.754714199459663e-05,
+      "loss": 1.8984,
+      "step": 7197
+    },
+    {
+      "epoch": 0.31461165260719437,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.75414095336766e-05,
+      "loss": 1.4577,
+      "step": 7198
+    },
+    {
+      "epoch": 0.31465536081122425,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.753567655301012e-05,
+      "loss": 1.9013,
+      "step": 7199
+    },
+    {
+      "epoch": 0.3146990690152542,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.752994305270534e-05,
+      "loss": 1.9996,
+      "step": 7200
+    },
+    {
+      "epoch": 0.31474277721928406,
+      "grad_norm": 2.75,
+      "learning_rate": 7.752420903287044e-05,
+      "loss": 2.3113,
+      "step": 7201
+    },
+    {
+      "epoch": 0.31478648542331394,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.751847449361367e-05,
+      "loss": 1.899,
+      "step": 7202
+    },
+    {
+      "epoch": 0.3148301936273439,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.751273943504322e-05,
+      "loss": 3.2488,
+      "step": 7203
+    },
+    {
+      "epoch": 0.31487390183137376,
+      "grad_norm": 1.859375,
+      "learning_rate": 7.750700385726736e-05,
+      "loss": 1.6907,
+      "step": 7204
+    },
+    {
+      "epoch": 0.31491761003540364,
+      "grad_norm": 1.875,
+      "learning_rate": 7.75012677603943e-05,
+      "loss": 1.556,
+      "step": 7205
+    },
+    {
+      "epoch": 0.3149613182394335,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.749553114453228e-05,
+      "loss": 2.0231,
+      "step": 7206
+    },
+    {
+      "epoch": 0.31500502644346345,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.748979400978956e-05,
+      "loss": 2.0842,
+      "step": 7207
+    },
+    {
+      "epoch": 0.31504873464749333,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.748405635627444e-05,
+      "loss": 1.6753,
+      "step": 7208
+    },
+    {
+      "epoch": 0.3150924428515232,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.747831818409517e-05,
+      "loss": 1.8329,
+      "step": 7209
+    },
+    {
+      "epoch": 0.31513615105555315,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.747257949336003e-05,
+      "loss": 1.687,
+      "step": 7210
+    },
+    {
+      "epoch": 0.31517985925958303,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.746684028417733e-05,
+      "loss": 1.944,
+      "step": 7211
+    },
+    {
+      "epoch": 0.3152235674636129,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.746110055665539e-05,
+      "loss": 1.9002,
+      "step": 7212
+    },
+    {
+      "epoch": 0.31526727566764284,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.745536031090252e-05,
+      "loss": 1.7212,
+      "step": 7213
+    },
+    {
+      "epoch": 0.3153109838716727,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.744961954702703e-05,
+      "loss": 1.6796,
+      "step": 7214
+    },
+    {
+      "epoch": 0.3153546920757026,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.744387826513726e-05,
+      "loss": 2.0101,
+      "step": 7215
+    },
+    {
+      "epoch": 0.3153984002797325,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.743813646534158e-05,
+      "loss": 1.7057,
+      "step": 7216
+    },
+    {
+      "epoch": 0.3154421084837624,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.743239414774832e-05,
+      "loss": 1.9942,
+      "step": 7217
+    },
+    {
+      "epoch": 0.3154858166877923,
+      "grad_norm": 1.90625,
+      "learning_rate": 7.742665131246587e-05,
+      "loss": 1.5873,
+      "step": 7218
+    },
+    {
+      "epoch": 0.3155295248918222,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.742090795960259e-05,
+      "loss": 2.4747,
+      "step": 7219
+    },
+    {
+      "epoch": 0.3155732330958521,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.741516408926686e-05,
+      "loss": 1.9626,
+      "step": 7220
+    },
+    {
+      "epoch": 0.315616941299882,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.740941970156707e-05,
+      "loss": 1.7351,
+      "step": 7221
+    },
+    {
+      "epoch": 0.31566064950391187,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.740367479661166e-05,
+      "loss": 1.2704,
+      "step": 7222
+    },
+    {
+      "epoch": 0.3157043577079418,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.739792937450901e-05,
+      "loss": 1.7378,
+      "step": 7223
+    },
+    {
+      "epoch": 0.3157480659119717,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.739218343536757e-05,
+      "loss": 2.3897,
+      "step": 7224
+    },
+    {
+      "epoch": 0.31579177411600157,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.738643697929575e-05,
+      "loss": 2.4272,
+      "step": 7225
+    },
+    {
+      "epoch": 0.31583548232003145,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.7380690006402e-05,
+      "loss": 2.2493,
+      "step": 7226
+    },
+    {
+      "epoch": 0.3158791905240614,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.737494251679479e-05,
+      "loss": 2.0007,
+      "step": 7227
+    },
+    {
+      "epoch": 0.31592289872809126,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.736919451058258e-05,
+      "loss": 1.9244,
+      "step": 7228
+    },
+    {
+      "epoch": 0.31596660693212114,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.736344598787381e-05,
+      "loss": 1.9202,
+      "step": 7229
+    },
+    {
+      "epoch": 0.3160103151361511,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.735769694877701e-05,
+      "loss": 1.7687,
+      "step": 7230
+    },
+    {
+      "epoch": 0.31605402334018096,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.735194739340064e-05,
+      "loss": 1.5402,
+      "step": 7231
+    },
+    {
+      "epoch": 0.31609773154421084,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.734619732185322e-05,
+      "loss": 1.9211,
+      "step": 7232
+    },
+    {
+      "epoch": 0.31614143974824077,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.734044673424325e-05,
+      "loss": 2.1806,
+      "step": 7233
+    },
+    {
+      "epoch": 0.31618514795227065,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.733469563067928e-05,
+      "loss": 1.5808,
+      "step": 7234
+    },
+    {
+      "epoch": 0.31622885615630053,
+      "grad_norm": 2.5,
+      "learning_rate": 7.73289440112698e-05,
+      "loss": 2.3715,
+      "step": 7235
+    },
+    {
+      "epoch": 0.3162725643603304,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.732319187612335e-05,
+      "loss": 2.3467,
+      "step": 7236
+    },
+    {
+      "epoch": 0.31631627256436035,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.731743922534853e-05,
+      "loss": 1.9838,
+      "step": 7237
+    },
+    {
+      "epoch": 0.3163599807683902,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.731168605905388e-05,
+      "loss": 2.6055,
+      "step": 7238
+    },
+    {
+      "epoch": 0.3164036889724201,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.730593237734796e-05,
+      "loss": 3.3343,
+      "step": 7239
+    },
+    {
+      "epoch": 0.31644739717645004,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.730017818033935e-05,
+      "loss": 2.0111,
+      "step": 7240
+    },
+    {
+      "epoch": 0.3164911053804799,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.729442346813662e-05,
+      "loss": 2.0082,
+      "step": 7241
+    },
+    {
+      "epoch": 0.3165348135845098,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.728866824084842e-05,
+      "loss": 2.1414,
+      "step": 7242
+    },
+    {
+      "epoch": 0.31657852178853974,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.728291249858332e-05,
+      "loss": 1.7971,
+      "step": 7243
+    },
+    {
+      "epoch": 0.3166222299925696,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.727715624144998e-05,
+      "loss": 2.4157,
+      "step": 7244
+    },
+    {
+      "epoch": 0.3166659381965995,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.727139946955697e-05,
+      "loss": 1.6075,
+      "step": 7245
+    },
+    {
+      "epoch": 0.3167096464006294,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.726564218301297e-05,
+      "loss": 2.1659,
+      "step": 7246
+    },
+    {
+      "epoch": 0.3167533546046593,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.725988438192662e-05,
+      "loss": 1.6702,
+      "step": 7247
+    },
+    {
+      "epoch": 0.3167970628086892,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.725412606640658e-05,
+      "loss": 1.9162,
+      "step": 7248
+    },
+    {
+      "epoch": 0.31684077101271907,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.724836723656153e-05,
+      "loss": 1.8255,
+      "step": 7249
+    },
+    {
+      "epoch": 0.316884479216749,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.724260789250011e-05,
+      "loss": 1.9185,
+      "step": 7250
+    },
+    {
+      "epoch": 0.3169281874207789,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.723684803433102e-05,
+      "loss": 1.7484,
+      "step": 7251
+    },
+    {
+      "epoch": 0.31697189562480876,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.723108766216298e-05,
+      "loss": 1.7841,
+      "step": 7252
+    },
+    {
+      "epoch": 0.3170156038288387,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.72253267761047e-05,
+      "loss": 3.1336,
+      "step": 7253
+    },
+    {
+      "epoch": 0.3170593120328686,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.721956537626487e-05,
+      "loss": 1.9293,
+      "step": 7254
+    },
+    {
+      "epoch": 0.31710302023689846,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.721380346275222e-05,
+      "loss": 1.7513,
+      "step": 7255
+    },
+    {
+      "epoch": 0.31714672844092834,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.720804103567546e-05,
+      "loss": 2.3873,
+      "step": 7256
+    },
+    {
+      "epoch": 0.3171904366449583,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.720227809514343e-05,
+      "loss": 1.8492,
+      "step": 7257
+    },
+    {
+      "epoch": 0.31723414484898815,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.719651464126475e-05,
+      "loss": 2.1024,
+      "step": 7258
+    },
+    {
+      "epoch": 0.31727785305301803,
+      "grad_norm": 1.859375,
+      "learning_rate": 7.719075067414831e-05,
+      "loss": 1.4269,
+      "step": 7259
+    },
+    {
+      "epoch": 0.31732156125704797,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.718498619390283e-05,
+      "loss": 2.1015,
+      "step": 7260
+    },
+    {
+      "epoch": 0.31736526946107785,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.717922120063706e-05,
+      "loss": 1.6376,
+      "step": 7261
+    },
+    {
+      "epoch": 0.3174089776651077,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.717345569445986e-05,
+      "loss": 1.0115,
+      "step": 7262
+    },
+    {
+      "epoch": 0.31745268586913766,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.716768967547998e-05,
+      "loss": 2.3164,
+      "step": 7263
+    },
+    {
+      "epoch": 0.31749639407316754,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.716192314380626e-05,
+      "loss": 1.6065,
+      "step": 7264
+    },
+    {
+      "epoch": 0.3175401022771974,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.715615609954752e-05,
+      "loss": 2.0752,
+      "step": 7265
+    },
+    {
+      "epoch": 0.3175838104812273,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.71503885428126e-05,
+      "loss": 1.8567,
+      "step": 7266
+    },
+    {
+      "epoch": 0.31762751868525724,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.714462047371031e-05,
+      "loss": 2.532,
+      "step": 7267
+    },
+    {
+      "epoch": 0.3176712268892871,
+      "grad_norm": 2.125,
+      "learning_rate": 7.713885189234956e-05,
+      "loss": 1.9574,
+      "step": 7268
+    },
+    {
+      "epoch": 0.317714935093317,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.713308279883915e-05,
+      "loss": 2.0208,
+      "step": 7269
+    },
+    {
+      "epoch": 0.31775864329734693,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.712731319328798e-05,
+      "loss": 1.8329,
+      "step": 7270
+    },
+    {
+      "epoch": 0.3178023515013768,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.712154307580493e-05,
+      "loss": 2.0871,
+      "step": 7271
+    },
+    {
+      "epoch": 0.3178460597054067,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.711577244649888e-05,
+      "loss": 2.0626,
+      "step": 7272
+    },
+    {
+      "epoch": 0.3178897679094366,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.711000130547875e-05,
+      "loss": 2.0854,
+      "step": 7273
+    },
+    {
+      "epoch": 0.3179334761134665,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.710422965285344e-05,
+      "loss": 1.7566,
+      "step": 7274
+    },
+    {
+      "epoch": 0.3179771843174964,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.709845748873187e-05,
+      "loss": 1.6372,
+      "step": 7275
+    },
+    {
+      "epoch": 0.31802089252152627,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.709268481322296e-05,
+      "loss": 1.9832,
+      "step": 7276
+    },
+    {
+      "epoch": 0.3180646007255562,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.708691162643565e-05,
+      "loss": 1.6387,
+      "step": 7277
+    },
+    {
+      "epoch": 0.3181083089295861,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.70811379284789e-05,
+      "loss": 2.0553,
+      "step": 7278
+    },
+    {
+      "epoch": 0.31815201713361596,
+      "grad_norm": 1.90625,
+      "learning_rate": 7.707536371946167e-05,
+      "loss": 1.6551,
+      "step": 7279
+    },
+    {
+      "epoch": 0.3181957253376459,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.706958899949293e-05,
+      "loss": 1.7686,
+      "step": 7280
+    },
+    {
+      "epoch": 0.3182394335416758,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.706381376868162e-05,
+      "loss": 1.6572,
+      "step": 7281
+    },
+    {
+      "epoch": 0.31828314174570566,
+      "grad_norm": 3.921875,
+      "learning_rate": 7.705803802713677e-05,
+      "loss": 2.0216,
+      "step": 7282
+    },
+    {
+      "epoch": 0.3183268499497356,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.705226177496736e-05,
+      "loss": 2.2726,
+      "step": 7283
+    },
+    {
+      "epoch": 0.31837055815376547,
+      "grad_norm": 2.875,
+      "learning_rate": 7.70464850122824e-05,
+      "loss": 2.1551,
+      "step": 7284
+    },
+    {
+      "epoch": 0.31841426635779535,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.70407077391909e-05,
+      "loss": 1.7063,
+      "step": 7285
+    },
+    {
+      "epoch": 0.31845797456182523,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.703492995580188e-05,
+      "loss": 1.7317,
+      "step": 7286
+    },
+    {
+      "epoch": 0.31850168276585517,
+      "grad_norm": 2.125,
+      "learning_rate": 7.70291516622244e-05,
+      "loss": 1.7335,
+      "step": 7287
+    },
+    {
+      "epoch": 0.31854539096988504,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.702337285856748e-05,
+      "loss": 1.8964,
+      "step": 7288
+    },
+    {
+      "epoch": 0.3185890991739149,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.701759354494018e-05,
+      "loss": 1.7829,
+      "step": 7289
+    },
+    {
+      "epoch": 0.31863280737794486,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.701181372145159e-05,
+      "loss": 1.8129,
+      "step": 7290
+    },
+    {
+      "epoch": 0.31867651558197474,
+      "grad_norm": 3.6875,
+      "learning_rate": 7.700603338821074e-05,
+      "loss": 2.084,
+      "step": 7291
+    },
+    {
+      "epoch": 0.3187202237860046,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.700025254532673e-05,
+      "loss": 2.0998,
+      "step": 7292
+    },
+    {
+      "epoch": 0.31876393199003455,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.699447119290867e-05,
+      "loss": 2.103,
+      "step": 7293
+    },
+    {
+      "epoch": 0.31880764019406443,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.698868933106565e-05,
+      "loss": 2.3017,
+      "step": 7294
+    },
+    {
+      "epoch": 0.3188513483980943,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.698290695990677e-05,
+      "loss": 2.3595,
+      "step": 7295
+    },
+    {
+      "epoch": 0.3188950566021242,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.697712407954119e-05,
+      "loss": 2.0153,
+      "step": 7296
+    },
+    {
+      "epoch": 0.31893876480615413,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.697134069007799e-05,
+      "loss": 2.3144,
+      "step": 7297
+    },
+    {
+      "epoch": 0.318982473010184,
+      "grad_norm": 2.5,
+      "learning_rate": 7.696555679162635e-05,
+      "loss": 2.4251,
+      "step": 7298
+    },
+    {
+      "epoch": 0.3190261812142139,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.695977238429539e-05,
+      "loss": 2.1105,
+      "step": 7299
+    },
+    {
+      "epoch": 0.3190698894182438,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.695398746819431e-05,
+      "loss": 2.094,
+      "step": 7300
+    },
+    {
+      "epoch": 0.3191135976222737,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.694820204343223e-05,
+      "loss": 1.9124,
+      "step": 7301
+    },
+    {
+      "epoch": 0.3191573058263036,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.694241611011838e-05,
+      "loss": 1.5197,
+      "step": 7302
+    },
+    {
+      "epoch": 0.3192010140303335,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.693662966836191e-05,
+      "loss": 1.8983,
+      "step": 7303
+    },
+    {
+      "epoch": 0.3192447222343634,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.693084271827205e-05,
+      "loss": 1.5317,
+      "step": 7304
+    },
+    {
+      "epoch": 0.3192884304383933,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.692505525995799e-05,
+      "loss": 2.3226,
+      "step": 7305
+    },
+    {
+      "epoch": 0.31933213864242316,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.691926729352894e-05,
+      "loss": 1.7196,
+      "step": 7306
+    },
+    {
+      "epoch": 0.3193758468464531,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.691347881909412e-05,
+      "loss": 2.3532,
+      "step": 7307
+    },
+    {
+      "epoch": 0.319419555050483,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.690768983676281e-05,
+      "loss": 1.8094,
+      "step": 7308
+    },
+    {
+      "epoch": 0.31946326325451285,
+      "grad_norm": 1.90625,
+      "learning_rate": 7.690190034664423e-05,
+      "loss": 2.125,
+      "step": 7309
+    },
+    {
+      "epoch": 0.3195069714585428,
+      "grad_norm": 10.0,
+      "learning_rate": 7.689611034884763e-05,
+      "loss": 2.7774,
+      "step": 7310
+    },
+    {
+      "epoch": 0.31955067966257267,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.689031984348227e-05,
+      "loss": 1.3217,
+      "step": 7311
+    },
+    {
+      "epoch": 0.31959438786660255,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.688452883065745e-05,
+      "loss": 1.6019,
+      "step": 7312
+    },
+    {
+      "epoch": 0.3196380960706325,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.687873731048245e-05,
+      "loss": 1.9432,
+      "step": 7313
+    },
+    {
+      "epoch": 0.31968180427466236,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.687294528306655e-05,
+      "loss": 2.1253,
+      "step": 7314
+    },
+    {
+      "epoch": 0.31972551247869224,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.686715274851906e-05,
+      "loss": 1.5275,
+      "step": 7315
+    },
+    {
+      "epoch": 0.3197692206827221,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.68613597069493e-05,
+      "loss": 2.0593,
+      "step": 7316
+    },
+    {
+      "epoch": 0.31981292888675206,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.685556615846657e-05,
+      "loss": 1.4496,
+      "step": 7317
+    },
+    {
+      "epoch": 0.31985663709078194,
+      "grad_norm": 3.28125,
+      "learning_rate": 7.684977210318024e-05,
+      "loss": 2.216,
+      "step": 7318
+    },
+    {
+      "epoch": 0.3199003452948118,
+      "grad_norm": 1.8359375,
+      "learning_rate": 7.684397754119964e-05,
+      "loss": 1.4906,
+      "step": 7319
+    },
+    {
+      "epoch": 0.31994405349884175,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.683818247263407e-05,
+      "loss": 1.724,
+      "step": 7320
+    },
+    {
+      "epoch": 0.31998776170287163,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.683238689759298e-05,
+      "loss": 1.9679,
+      "step": 7321
+    },
+    {
+      "epoch": 0.3200314699069015,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.682659081618567e-05,
+      "loss": 1.907,
+      "step": 7322
+    },
+    {
+      "epoch": 0.32007517811093145,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.682079422852156e-05,
+      "loss": 2.0567,
+      "step": 7323
+    },
+    {
+      "epoch": 0.3201188863149613,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.681499713471002e-05,
+      "loss": 1.9915,
+      "step": 7324
+    },
+    {
+      "epoch": 0.3201625945189912,
+      "grad_norm": 4.25,
+      "learning_rate": 7.680919953486048e-05,
+      "loss": 2.7251,
+      "step": 7325
+    },
+    {
+      "epoch": 0.3202063027230211,
+      "grad_norm": 5.28125,
+      "learning_rate": 7.680340142908231e-05,
+      "loss": 2.0837,
+      "step": 7326
+    },
+    {
+      "epoch": 0.320250010927051,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.679760281748491e-05,
+      "loss": 1.6771,
+      "step": 7327
+    },
+    {
+      "epoch": 0.3202937191310809,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.67918037001778e-05,
+      "loss": 2.085,
+      "step": 7328
+    },
+    {
+      "epoch": 0.3203374273351108,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.678600407727032e-05,
+      "loss": 1.6378,
+      "step": 7329
+    },
+    {
+      "epoch": 0.3203811355391407,
+      "grad_norm": 2.25,
+      "learning_rate": 7.678020394887197e-05,
+      "loss": 2.031,
+      "step": 7330
+    },
+    {
+      "epoch": 0.3204248437431706,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.67744033150922e-05,
+      "loss": 1.9841,
+      "step": 7331
+    },
+    {
+      "epoch": 0.3204685519472005,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.676860217604047e-05,
+      "loss": 2.5098,
+      "step": 7332
+    },
+    {
+      "epoch": 0.3205122601512304,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.676280053182626e-05,
+      "loss": 1.766,
+      "step": 7333
+    },
+    {
+      "epoch": 0.3205559683552603,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.675699838255905e-05,
+      "loss": 2.0977,
+      "step": 7334
+    },
+    {
+      "epoch": 0.32059967655929017,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.675119572834835e-05,
+      "loss": 2.1568,
+      "step": 7335
+    },
+    {
+      "epoch": 0.32064338476332005,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.674539256930363e-05,
+      "loss": 1.7092,
+      "step": 7336
+    },
+    {
+      "epoch": 0.32068709296735,
+      "grad_norm": 1.8359375,
+      "learning_rate": 7.673958890553443e-05,
+      "loss": 1.755,
+      "step": 7337
+    },
+    {
+      "epoch": 0.32073080117137986,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.673378473715027e-05,
+      "loss": 1.5277,
+      "step": 7338
+    },
+    {
+      "epoch": 0.32077450937540974,
+      "grad_norm": 8.5,
+      "learning_rate": 7.672798006426069e-05,
+      "loss": 2.4689,
+      "step": 7339
+    },
+    {
+      "epoch": 0.3208182175794397,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.672217488697522e-05,
+      "loss": 1.7042,
+      "step": 7340
+    },
+    {
+      "epoch": 0.32086192578346956,
+      "grad_norm": 2.125,
+      "learning_rate": 7.671636920540342e-05,
+      "loss": 2.0158,
+      "step": 7341
+    },
+    {
+      "epoch": 0.32090563398749944,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.671056301965484e-05,
+      "loss": 1.4584,
+      "step": 7342
+    },
+    {
+      "epoch": 0.3209493421915294,
+      "grad_norm": 11.25,
+      "learning_rate": 7.670475632983909e-05,
+      "loss": 1.9689,
+      "step": 7343
+    },
+    {
+      "epoch": 0.32099305039555925,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.669894913606568e-05,
+      "loss": 1.9686,
+      "step": 7344
+    },
+    {
+      "epoch": 0.32103675859958913,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.669314143844428e-05,
+      "loss": 1.7838,
+      "step": 7345
+    },
+    {
+      "epoch": 0.321080466803619,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.668733323708443e-05,
+      "loss": 1.416,
+      "step": 7346
+    },
+    {
+      "epoch": 0.32112417500764895,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.668152453209576e-05,
+      "loss": 2.0266,
+      "step": 7347
+    },
+    {
+      "epoch": 0.32116788321167883,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.66757153235879e-05,
+      "loss": 1.6594,
+      "step": 7348
+    },
+    {
+      "epoch": 0.3212115914157087,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.666990561167046e-05,
+      "loss": 1.8492,
+      "step": 7349
+    },
+    {
+      "epoch": 0.32125529961973864,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.666409539645308e-05,
+      "loss": 1.8732,
+      "step": 7350
+    },
+    {
+      "epoch": 0.3212990078237685,
+      "grad_norm": 2.5,
+      "learning_rate": 7.665828467804542e-05,
+      "loss": 2.4108,
+      "step": 7351
+    },
+    {
+      "epoch": 0.3213427160277984,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.665247345655713e-05,
+      "loss": 1.5882,
+      "step": 7352
+    },
+    {
+      "epoch": 0.32138642423182834,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.664666173209787e-05,
+      "loss": 1.9882,
+      "step": 7353
+    },
+    {
+      "epoch": 0.3214301324358582,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.664084950477731e-05,
+      "loss": 1.8138,
+      "step": 7354
+    },
+    {
+      "epoch": 0.3214738406398881,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.663503677470516e-05,
+      "loss": 2.143,
+      "step": 7355
+    },
+    {
+      "epoch": 0.321517548843918,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.66292235419911e-05,
+      "loss": 1.7407,
+      "step": 7356
+    },
+    {
+      "epoch": 0.3215612570479479,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.662340980674483e-05,
+      "loss": 1.5849,
+      "step": 7357
+    },
+    {
+      "epoch": 0.3216049652519778,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.661759556907607e-05,
+      "loss": 2.7416,
+      "step": 7358
+    },
+    {
+      "epoch": 0.32164867345600767,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.661178082909455e-05,
+      "loss": 1.7122,
+      "step": 7359
+    },
+    {
+      "epoch": 0.3216923816600376,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.660596558690998e-05,
+      "loss": 2.0861,
+      "step": 7360
+    },
+    {
+      "epoch": 0.3217360898640675,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.660014984263214e-05,
+      "loss": 2.1306,
+      "step": 7361
+    },
+    {
+      "epoch": 0.32177979806809737,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.659433359637072e-05,
+      "loss": 1.573,
+      "step": 7362
+    },
+    {
+      "epoch": 0.3218235062721273,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.658851684823553e-05,
+      "loss": 1.3277,
+      "step": 7363
+    },
+    {
+      "epoch": 0.3218672144761572,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.658269959833635e-05,
+      "loss": 1.7538,
+      "step": 7364
+    },
+    {
+      "epoch": 0.32191092268018706,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.657688184678293e-05,
+      "loss": 2.4755,
+      "step": 7365
+    },
+    {
+      "epoch": 0.32195463088421694,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.657106359368507e-05,
+      "loss": 1.7868,
+      "step": 7366
+    },
+    {
+      "epoch": 0.3219983390882469,
+      "grad_norm": 1.90625,
+      "learning_rate": 7.656524483915256e-05,
+      "loss": 1.634,
+      "step": 7367
+    },
+    {
+      "epoch": 0.32204204729227676,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.655942558329523e-05,
+      "loss": 1.6302,
+      "step": 7368
+    },
+    {
+      "epoch": 0.32208575549630664,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.655360582622286e-05,
+      "loss": 1.8055,
+      "step": 7369
+    },
+    {
+      "epoch": 0.32212946370033657,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.654778556804533e-05,
+      "loss": 2.1191,
+      "step": 7370
+    },
+    {
+      "epoch": 0.32217317190436645,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.654196480887244e-05,
+      "loss": 1.8245,
+      "step": 7371
+    },
+    {
+      "epoch": 0.32221688010839633,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.653614354881402e-05,
+      "loss": 1.6235,
+      "step": 7372
+    },
+    {
+      "epoch": 0.32226058831242627,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.653032178797996e-05,
+      "loss": 2.6314,
+      "step": 7373
+    },
+    {
+      "epoch": 0.32230429651645615,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.652449952648013e-05,
+      "loss": 3.2089,
+      "step": 7374
+    },
+    {
+      "epoch": 0.322348004720486,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.65186767644244e-05,
+      "loss": 1.5702,
+      "step": 7375
+    },
+    {
+      "epoch": 0.3223917129245159,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.651285350192261e-05,
+      "loss": 1.951,
+      "step": 7376
+    },
+    {
+      "epoch": 0.32243542112854584,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.650702973908471e-05,
+      "loss": 1.9708,
+      "step": 7377
+    },
+    {
+      "epoch": 0.3224791293325757,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.650120547602056e-05,
+      "loss": 1.8858,
+      "step": 7378
+    },
+    {
+      "epoch": 0.3225228375366056,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.64953807128401e-05,
+      "loss": 1.9008,
+      "step": 7379
+    },
+    {
+      "epoch": 0.32256654574063554,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.648955544965326e-05,
+      "loss": 1.4871,
+      "step": 7380
+    },
+    {
+      "epoch": 0.3226102539446654,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.648372968656993e-05,
+      "loss": 1.5375,
+      "step": 7381
+    },
+    {
+      "epoch": 0.3226539621486953,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.647790342370009e-05,
+      "loss": 3.0121,
+      "step": 7382
+    },
+    {
+      "epoch": 0.32269767035272523,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.647207666115368e-05,
+      "loss": 1.7063,
+      "step": 7383
+    },
+    {
+      "epoch": 0.3227413785567551,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.646624939904064e-05,
+      "loss": 2.0754,
+      "step": 7384
+    },
+    {
+      "epoch": 0.322785086760785,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.646042163747097e-05,
+      "loss": 1.5163,
+      "step": 7385
+    },
+    {
+      "epoch": 0.32282879496481487,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.645459337655463e-05,
+      "loss": 2.555,
+      "step": 7386
+    },
+    {
+      "epoch": 0.3228725031688448,
+      "grad_norm": 2.625,
+      "learning_rate": 7.644876461640158e-05,
+      "loss": 2.0501,
+      "step": 7387
+    },
+    {
+      "epoch": 0.3229162113728747,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.644293535712189e-05,
+      "loss": 1.6365,
+      "step": 7388
+    },
+    {
+      "epoch": 0.32295991957690456,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.643710559882551e-05,
+      "loss": 2.3091,
+      "step": 7389
+    },
+    {
+      "epoch": 0.3230036277809345,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.643127534162247e-05,
+      "loss": 2.0301,
+      "step": 7390
+    },
+    {
+      "epoch": 0.3230473359849644,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.642544458562278e-05,
+      "loss": 2.1748,
+      "step": 7391
+    },
+    {
+      "epoch": 0.32309104418899426,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.64196133309365e-05,
+      "loss": 1.9963,
+      "step": 7392
+    },
+    {
+      "epoch": 0.3231347523930242,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.641378157767368e-05,
+      "loss": 1.8105,
+      "step": 7393
+    },
+    {
+      "epoch": 0.3231784605970541,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.640794932594433e-05,
+      "loss": 1.945,
+      "step": 7394
+    },
+    {
+      "epoch": 0.32322216880108395,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.640211657585856e-05,
+      "loss": 1.8537,
+      "step": 7395
+    },
+    {
+      "epoch": 0.32326587700511383,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.639628332752642e-05,
+      "loss": 2.0916,
+      "step": 7396
+    },
+    {
+      "epoch": 0.32330958520914377,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.639044958105799e-05,
+      "loss": 1.7224,
+      "step": 7397
+    },
+    {
+      "epoch": 0.32335329341317365,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.638461533656338e-05,
+      "loss": 1.7498,
+      "step": 7398
+    },
+    {
+      "epoch": 0.32339700161720353,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.637878059415266e-05,
+      "loss": 1.8257,
+      "step": 7399
+    },
+    {
+      "epoch": 0.32344070982123346,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.637294535393598e-05,
+      "loss": 1.925,
+      "step": 7400
+    },
+    {
+      "epoch": 0.32348441802526334,
+      "grad_norm": 2.625,
+      "learning_rate": 7.636710961602341e-05,
+      "loss": 2.1079,
+      "step": 7401
+    },
+    {
+      "epoch": 0.3235281262292932,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.636127338052512e-05,
+      "loss": 1.7199,
+      "step": 7402
+    },
+    {
+      "epoch": 0.32357183443332316,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.635543664755124e-05,
+      "loss": 1.651,
+      "step": 7403
+    },
+    {
+      "epoch": 0.32361554263735304,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.634959941721191e-05,
+      "loss": 2.3902,
+      "step": 7404
+    },
+    {
+      "epoch": 0.3236592508413829,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.634376168961729e-05,
+      "loss": 2.8142,
+      "step": 7405
+    },
+    {
+      "epoch": 0.3237029590454128,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.633792346487754e-05,
+      "loss": 1.6277,
+      "step": 7406
+    },
+    {
+      "epoch": 0.32374666724944273,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.633208474310283e-05,
+      "loss": 1.9859,
+      "step": 7407
+    },
+    {
+      "epoch": 0.3237903754534726,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.632624552440337e-05,
+      "loss": 1.6204,
+      "step": 7408
+    },
+    {
+      "epoch": 0.3238340836575025,
+      "grad_norm": 1.75,
+      "learning_rate": 7.632040580888936e-05,
+      "loss": 1.6116,
+      "step": 7409
+    },
+    {
+      "epoch": 0.3238777918615324,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.631456559667095e-05,
+      "loss": 1.6195,
+      "step": 7410
+    },
+    {
+      "epoch": 0.3239215000655623,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.630872488785841e-05,
+      "loss": 2.1594,
+      "step": 7411
+    },
+    {
+      "epoch": 0.3239652082695922,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.630288368256193e-05,
+      "loss": 1.6741,
+      "step": 7412
+    },
+    {
+      "epoch": 0.3240089164736221,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.629704198089175e-05,
+      "loss": 1.9449,
+      "step": 7413
+    },
+    {
+      "epoch": 0.324052624677652,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.629119978295811e-05,
+      "loss": 1.716,
+      "step": 7414
+    },
+    {
+      "epoch": 0.3240963328816819,
+      "grad_norm": 2.5,
+      "learning_rate": 7.62853570888713e-05,
+      "loss": 1.6255,
+      "step": 7415
+    },
+    {
+      "epoch": 0.32414004108571176,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.62795138987415e-05,
+      "loss": 1.7113,
+      "step": 7416
+    },
+    {
+      "epoch": 0.3241837492897417,
+      "grad_norm": 1.890625,
+      "learning_rate": 7.627367021267906e-05,
+      "loss": 1.3831,
+      "step": 7417
+    },
+    {
+      "epoch": 0.3242274574937716,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.626782603079421e-05,
+      "loss": 1.4712,
+      "step": 7418
+    },
+    {
+      "epoch": 0.32427116569780146,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.626198135319724e-05,
+      "loss": 1.3958,
+      "step": 7419
+    },
+    {
+      "epoch": 0.3243148739018314,
+      "grad_norm": 2.125,
+      "learning_rate": 7.625613617999847e-05,
+      "loss": 1.7865,
+      "step": 7420
+    },
+    {
+      "epoch": 0.32435858210586127,
+      "grad_norm": 5.65625,
+      "learning_rate": 7.62502905113082e-05,
+      "loss": 1.3227,
+      "step": 7421
+    },
+    {
+      "epoch": 0.32440229030989115,
+      "grad_norm": 1.8125,
+      "learning_rate": 7.624444434723674e-05,
+      "loss": 1.6127,
+      "step": 7422
+    },
+    {
+      "epoch": 0.3244459985139211,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.623859768789441e-05,
+      "loss": 1.8083,
+      "step": 7423
+    },
+    {
+      "epoch": 0.32448970671795097,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.623275053339156e-05,
+      "loss": 1.8015,
+      "step": 7424
+    },
+    {
+      "epoch": 0.32453341492198085,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.622690288383853e-05,
+      "loss": 1.9107,
+      "step": 7425
+    },
+    {
+      "epoch": 0.3245771231260107,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.62210547393457e-05,
+      "loss": 2.0738,
+      "step": 7426
+    },
+    {
+      "epoch": 0.32462083133004066,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.621520610002335e-05,
+      "loss": 1.9741,
+      "step": 7427
+    },
+    {
+      "epoch": 0.32466453953407054,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.620935696598192e-05,
+      "loss": 1.5479,
+      "step": 7428
+    },
+    {
+      "epoch": 0.3247082477381004,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.620350733733179e-05,
+      "loss": 1.9824,
+      "step": 7429
+    },
+    {
+      "epoch": 0.32475195594213035,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.619765721418335e-05,
+      "loss": 1.6375,
+      "step": 7430
+    },
+    {
+      "epoch": 0.32479566414616023,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.619180659664698e-05,
+      "loss": 2.0936,
+      "step": 7431
+    },
+    {
+      "epoch": 0.3248393723501901,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.618595548483309e-05,
+      "loss": 2.2806,
+      "step": 7432
+    },
+    {
+      "epoch": 0.32488308055422005,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.61801038788521e-05,
+      "loss": 1.6695,
+      "step": 7433
+    },
+    {
+      "epoch": 0.32492678875824993,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.617425177881446e-05,
+      "loss": 1.7873,
+      "step": 7434
+    },
+    {
+      "epoch": 0.3249704969622798,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.616839918483061e-05,
+      "loss": 1.407,
+      "step": 7435
+    },
+    {
+      "epoch": 0.3250142051663097,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.616254609701096e-05,
+      "loss": 2.6228,
+      "step": 7436
+    },
+    {
+      "epoch": 0.3250579133703396,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.6156692515466e-05,
+      "loss": 1.7178,
+      "step": 7437
+    },
+    {
+      "epoch": 0.3251016215743695,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.615083844030618e-05,
+      "loss": 2.5461,
+      "step": 7438
+    },
+    {
+      "epoch": 0.3251453297783994,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.614498387164198e-05,
+      "loss": 1.7624,
+      "step": 7439
+    },
+    {
+      "epoch": 0.3251890379824293,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.613912880958386e-05,
+      "loss": 1.7311,
+      "step": 7440
+    },
+    {
+      "epoch": 0.3252327461864592,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.613327325424235e-05,
+      "loss": 1.9983,
+      "step": 7441
+    },
+    {
+      "epoch": 0.3252764543904891,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.612741720572794e-05,
+      "loss": 1.957,
+      "step": 7442
+    },
+    {
+      "epoch": 0.325320162594519,
+      "grad_norm": 2.375,
+      "learning_rate": 7.612156066415113e-05,
+      "loss": 1.513,
+      "step": 7443
+    },
+    {
+      "epoch": 0.3253638707985489,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.611570362962248e-05,
+      "loss": 2.1919,
+      "step": 7444
+    },
+    {
+      "epoch": 0.3254075790025788,
+      "grad_norm": 1.8125,
+      "learning_rate": 7.610984610225247e-05,
+      "loss": 1.4103,
+      "step": 7445
+    },
+    {
+      "epoch": 0.3254512872066087,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.610398808215166e-05,
+      "loss": 2.1165,
+      "step": 7446
+    },
+    {
+      "epoch": 0.3254949954106386,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.609812956943063e-05,
+      "loss": 1.7011,
+      "step": 7447
+    },
+    {
+      "epoch": 0.32553870361466847,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.609227056419989e-05,
+      "loss": 1.6094,
+      "step": 7448
+    },
+    {
+      "epoch": 0.32558241181869835,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.608641106657001e-05,
+      "loss": 2.8938,
+      "step": 7449
+    },
+    {
+      "epoch": 0.3256261200227283,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.608055107665161e-05,
+      "loss": 1.9487,
+      "step": 7450
+    },
+    {
+      "epoch": 0.32566982822675816,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.607469059455526e-05,
+      "loss": 1.7912,
+      "step": 7451
+    },
+    {
+      "epoch": 0.32571353643078804,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.606882962039154e-05,
+      "loss": 1.9667,
+      "step": 7452
+    },
+    {
+      "epoch": 0.325757244634818,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.606296815427106e-05,
+      "loss": 1.8214,
+      "step": 7453
+    },
+    {
+      "epoch": 0.32580095283884786,
+      "grad_norm": 1.90625,
+      "learning_rate": 7.605710619630444e-05,
+      "loss": 2.1281,
+      "step": 7454
+    },
+    {
+      "epoch": 0.32584466104287774,
+      "grad_norm": 2.5,
+      "learning_rate": 7.605124374660231e-05,
+      "loss": 1.7559,
+      "step": 7455
+    },
+    {
+      "epoch": 0.32588836924690767,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.604538080527527e-05,
+      "loss": 1.7102,
+      "step": 7456
+    },
+    {
+      "epoch": 0.32593207745093755,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.603951737243402e-05,
+      "loss": 1.8114,
+      "step": 7457
+    },
+    {
+      "epoch": 0.32597578565496743,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.603365344818916e-05,
+      "loss": 0.9383,
+      "step": 7458
+    },
+    {
+      "epoch": 0.3260194938589973,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.602778903265137e-05,
+      "loss": 2.0806,
+      "step": 7459
+    },
+    {
+      "epoch": 0.32606320206302725,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.602192412593132e-05,
+      "loss": 2.2013,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3261069102670571,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.601605872813969e-05,
+      "loss": 2.2884,
+      "step": 7461
+    },
+    {
+      "epoch": 0.326150618471087,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.601019283938717e-05,
+      "loss": 2.3615,
+      "step": 7462
+    },
+    {
+      "epoch": 0.32619432667511694,
+      "grad_norm": 2.125,
+      "learning_rate": 7.600432645978444e-05,
+      "loss": 1.6707,
+      "step": 7463
+    },
+    {
+      "epoch": 0.3262380348791468,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.599845958944224e-05,
+      "loss": 1.7099,
+      "step": 7464
+    },
+    {
+      "epoch": 0.3262817430831767,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.599259222847127e-05,
+      "loss": 1.9748,
+      "step": 7465
+    },
+    {
+      "epoch": 0.32632545128720664,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.598672437698224e-05,
+      "loss": 1.7586,
+      "step": 7466
+    },
+    {
+      "epoch": 0.3263691594912365,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.598085603508592e-05,
+      "loss": 2.448,
+      "step": 7467
+    },
+    {
+      "epoch": 0.3264128676952664,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.597498720289302e-05,
+      "loss": 1.597,
+      "step": 7468
+    },
+    {
+      "epoch": 0.3264565758992963,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.59691178805143e-05,
+      "loss": 1.6784,
+      "step": 7469
+    },
+    {
+      "epoch": 0.3265002841033262,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.596324806806052e-05,
+      "loss": 1.8834,
+      "step": 7470
+    },
+    {
+      "epoch": 0.3265439923073561,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.595737776564249e-05,
+      "loss": 1.5806,
+      "step": 7471
+    },
+    {
+      "epoch": 0.32658770051138597,
+      "grad_norm": 1.90625,
+      "learning_rate": 7.595150697337095e-05,
+      "loss": 1.6759,
+      "step": 7472
+    },
+    {
+      "epoch": 0.3266314087154159,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.594563569135668e-05,
+      "loss": 1.974,
+      "step": 7473
+    },
+    {
+      "epoch": 0.3266751169194458,
+      "grad_norm": 2.625,
+      "learning_rate": 7.593976391971054e-05,
+      "loss": 1.97,
+      "step": 7474
+    },
+    {
+      "epoch": 0.32671882512347566,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.593389165854329e-05,
+      "loss": 2.3795,
+      "step": 7475
+    },
+    {
+      "epoch": 0.3267625333275056,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.592801890796575e-05,
+      "loss": 2.0789,
+      "step": 7476
+    },
+    {
+      "epoch": 0.3268062415315355,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.592214566808877e-05,
+      "loss": 1.7114,
+      "step": 7477
+    },
+    {
+      "epoch": 0.32684994973556536,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.591627193902315e-05,
+      "loss": 2.268,
+      "step": 7478
+    },
+    {
+      "epoch": 0.32689365793959524,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.591039772087977e-05,
+      "loss": 1.96,
+      "step": 7479
+    },
+    {
+      "epoch": 0.3269373661436252,
+      "grad_norm": 3.4375,
+      "learning_rate": 7.59045230137695e-05,
+      "loss": 2.1119,
+      "step": 7480
+    },
+    {
+      "epoch": 0.32698107434765505,
+      "grad_norm": 2.25,
+      "learning_rate": 7.589864781780314e-05,
+      "loss": 1.6881,
+      "step": 7481
+    },
+    {
+      "epoch": 0.32702478255168493,
+      "grad_norm": 1.890625,
+      "learning_rate": 7.589277213309163e-05,
+      "loss": 1.9424,
+      "step": 7482
+    },
+    {
+      "epoch": 0.32706849075571487,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.58868959597458e-05,
+      "loss": 1.2206,
+      "step": 7483
+    },
+    {
+      "epoch": 0.32711219895974475,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.588101929787658e-05,
+      "loss": 1.8753,
+      "step": 7484
+    },
+    {
+      "epoch": 0.32715590716377463,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.587514214759487e-05,
+      "loss": 1.7069,
+      "step": 7485
+    },
+    {
+      "epoch": 0.32719961536780456,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.586926450901155e-05,
+      "loss": 1.8258,
+      "step": 7486
+    },
+    {
+      "epoch": 0.32724332357183444,
+      "grad_norm": 2.375,
+      "learning_rate": 7.586338638223757e-05,
+      "loss": 1.7908,
+      "step": 7487
+    },
+    {
+      "epoch": 0.3272870317758643,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.585750776738383e-05,
+      "loss": 2.2592,
+      "step": 7488
+    },
+    {
+      "epoch": 0.3273307399798942,
+      "grad_norm": 1.875,
+      "learning_rate": 7.58516286645613e-05,
+      "loss": 1.6792,
+      "step": 7489
+    },
+    {
+      "epoch": 0.32737444818392414,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.584574907388092e-05,
+      "loss": 2.2359,
+      "step": 7490
+    },
+    {
+      "epoch": 0.327418156387954,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.583986899545362e-05,
+      "loss": 1.7483,
+      "step": 7491
+    },
+    {
+      "epoch": 0.3274618645919839,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.58339884293904e-05,
+      "loss": 1.6944,
+      "step": 7492
+    },
+    {
+      "epoch": 0.32750557279601383,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.58281073758022e-05,
+      "loss": 2.4001,
+      "step": 7493
+    },
+    {
+      "epoch": 0.3275492810000437,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.58222258348e-05,
+      "loss": 1.9724,
+      "step": 7494
+    },
+    {
+      "epoch": 0.3275929892040736,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.581634380649488e-05,
+      "loss": 1.8521,
+      "step": 7495
+    },
+    {
+      "epoch": 0.3276366974081035,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.581046129099773e-05,
+      "loss": 2.0944,
+      "step": 7496
+    },
+    {
+      "epoch": 0.3276804056121334,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.580457828841963e-05,
+      "loss": 1.59,
+      "step": 7497
+    },
+    {
+      "epoch": 0.3277241138161633,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.579869479887158e-05,
+      "loss": 2.4358,
+      "step": 7498
+    },
+    {
+      "epoch": 0.32776782202019317,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.57928108224646e-05,
+      "loss": 1.5962,
+      "step": 7499
+    },
+    {
+      "epoch": 0.3278115302242231,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.578692635930975e-05,
+      "loss": 1.5579,
+      "step": 7500
+    },
+    {
+      "epoch": 0.327855238428253,
+      "grad_norm": 15.0,
+      "learning_rate": 7.578104140951807e-05,
+      "loss": 5.9767,
+      "step": 7501
+    },
+    {
+      "epoch": 0.32789894663228286,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.577515597320062e-05,
+      "loss": 2.0851,
+      "step": 7502
+    },
+    {
+      "epoch": 0.3279426548363128,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.576927005046844e-05,
+      "loss": 2.4982,
+      "step": 7503
+    },
+    {
+      "epoch": 0.3279863630403427,
+      "grad_norm": 2.25,
+      "learning_rate": 7.576338364143264e-05,
+      "loss": 1.5677,
+      "step": 7504
+    },
+    {
+      "epoch": 0.32803007124437256,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.575749674620431e-05,
+      "loss": 1.7917,
+      "step": 7505
+    },
+    {
+      "epoch": 0.3280737794484025,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.575160936489452e-05,
+      "loss": 1.7448,
+      "step": 7506
+    },
+    {
+      "epoch": 0.32811748765243237,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.574572149761437e-05,
+      "loss": 1.9557,
+      "step": 7507
+    },
+    {
+      "epoch": 0.32816119585646225,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.573983314447499e-05,
+      "loss": 1.9883,
+      "step": 7508
+    },
+    {
+      "epoch": 0.32820490406049213,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.573394430558749e-05,
+      "loss": 1.3002,
+      "step": 7509
+    },
+    {
+      "epoch": 0.32824861226452207,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.572805498106301e-05,
+      "loss": 2.2083,
+      "step": 7510
+    },
+    {
+      "epoch": 0.32829232046855195,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.57221651710127e-05,
+      "loss": 1.7499,
+      "step": 7511
+    },
+    {
+      "epoch": 0.3283360286725818,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.571627487554769e-05,
+      "loss": 1.98,
+      "step": 7512
+    },
+    {
+      "epoch": 0.32837973687661176,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.571038409477913e-05,
+      "loss": 2.192,
+      "step": 7513
+    },
+    {
+      "epoch": 0.32842344508064164,
+      "grad_norm": 2.75,
+      "learning_rate": 7.570449282881822e-05,
+      "loss": 1.4564,
+      "step": 7514
+    },
+    {
+      "epoch": 0.3284671532846715,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.569860107777613e-05,
+      "loss": 1.8013,
+      "step": 7515
+    },
+    {
+      "epoch": 0.32851086148870146,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.5692708841764e-05,
+      "loss": 2.2148,
+      "step": 7516
+    },
+    {
+      "epoch": 0.32855456969273134,
+      "grad_norm": 2.0,
+      "learning_rate": 7.56868161208931e-05,
+      "loss": 1.547,
+      "step": 7517
+    },
+    {
+      "epoch": 0.3285982778967612,
+      "grad_norm": 2.25,
+      "learning_rate": 7.568092291527455e-05,
+      "loss": 2.1837,
+      "step": 7518
+    },
+    {
+      "epoch": 0.3286419861007911,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.567502922501963e-05,
+      "loss": 2.1615,
+      "step": 7519
+    },
+    {
+      "epoch": 0.32868569430482103,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.566913505023956e-05,
+      "loss": 1.499,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3287294025088509,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.566324039104553e-05,
+      "loss": 2.5199,
+      "step": 7521
+    },
+    {
+      "epoch": 0.3287731107128808,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.565734524754882e-05,
+      "loss": 1.5872,
+      "step": 7522
+    },
+    {
+      "epoch": 0.3288168189169107,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.565144961986064e-05,
+      "loss": 1.7103,
+      "step": 7523
+    },
+    {
+      "epoch": 0.3288605271209406,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.564555350809226e-05,
+      "loss": 2.1791,
+      "step": 7524
+    },
+    {
+      "epoch": 0.3289042353249705,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.5639656912355e-05,
+      "loss": 1.6546,
+      "step": 7525
+    },
+    {
+      "epoch": 0.3289479435290004,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.563375983276008e-05,
+      "loss": 1.3833,
+      "step": 7526
+    },
+    {
+      "epoch": 0.3289916517330303,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.56278622694188e-05,
+      "loss": 2.4949,
+      "step": 7527
+    },
+    {
+      "epoch": 0.3290353599370602,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.562196422244245e-05,
+      "loss": 1.7161,
+      "step": 7528
+    },
+    {
+      "epoch": 0.32907906814109006,
+      "grad_norm": 1.9375,
+      "learning_rate": 7.561606569194237e-05,
+      "loss": 1.6621,
+      "step": 7529
+    },
+    {
+      "epoch": 0.32912277634512,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.561016667802982e-05,
+      "loss": 1.9472,
+      "step": 7530
+    },
+    {
+      "epoch": 0.3291664845491499,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.560426718081617e-05,
+      "loss": 2.2656,
+      "step": 7531
+    },
+    {
+      "epoch": 0.32921019275317975,
+      "grad_norm": 2.375,
+      "learning_rate": 7.559836720041274e-05,
+      "loss": 1.8794,
+      "step": 7532
+    },
+    {
+      "epoch": 0.3292539009572097,
+      "grad_norm": 2.375,
+      "learning_rate": 7.559246673693085e-05,
+      "loss": 2.2259,
+      "step": 7533
+    },
+    {
+      "epoch": 0.32929760916123957,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.558656579048185e-05,
+      "loss": 1.6449,
+      "step": 7534
+    },
+    {
+      "epoch": 0.32934131736526945,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.558066436117715e-05,
+      "loss": 2.2579,
+      "step": 7535
+    },
+    {
+      "epoch": 0.3293850255692994,
+      "grad_norm": 4.875,
+      "learning_rate": 7.557476244912805e-05,
+      "loss": 2.4239,
+      "step": 7536
+    },
+    {
+      "epoch": 0.32942873377332926,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.556886005444597e-05,
+      "loss": 1.8257,
+      "step": 7537
+    },
+    {
+      "epoch": 0.32947244197735914,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.55629571772423e-05,
+      "loss": 1.2687,
+      "step": 7538
+    },
+    {
+      "epoch": 0.329516150181389,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.555705381762841e-05,
+      "loss": 2.1488,
+      "step": 7539
+    },
+    {
+      "epoch": 0.32955985838541896,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.555114997571572e-05,
+      "loss": 1.7148,
+      "step": 7540
+    },
+    {
+      "epoch": 0.32960356658944884,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.554524565161565e-05,
+      "loss": 1.781,
+      "step": 7541
+    },
+    {
+      "epoch": 0.3296472747934787,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.553934084543961e-05,
+      "loss": 1.5344,
+      "step": 7542
+    },
+    {
+      "epoch": 0.32969098299750865,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.553343555729903e-05,
+      "loss": 1.5419,
+      "step": 7543
+    },
+    {
+      "epoch": 0.32973469120153853,
+      "grad_norm": 3.828125,
+      "learning_rate": 7.552752978730536e-05,
+      "loss": 1.7144,
+      "step": 7544
+    },
+    {
+      "epoch": 0.3297783994055684,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.552162353557006e-05,
+      "loss": 2.0587,
+      "step": 7545
+    },
+    {
+      "epoch": 0.32982210760959835,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.551571680220457e-05,
+      "loss": 1.6627,
+      "step": 7546
+    },
+    {
+      "epoch": 0.3298658158136282,
+      "grad_norm": 1.75,
+      "learning_rate": 7.550980958732037e-05,
+      "loss": 1.9505,
+      "step": 7547
+    },
+    {
+      "epoch": 0.3299095240176581,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.550390189102894e-05,
+      "loss": 1.7785,
+      "step": 7548
+    },
+    {
+      "epoch": 0.329953232221688,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.549799371344175e-05,
+      "loss": 1.792,
+      "step": 7549
+    },
+    {
+      "epoch": 0.3299969404257179,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.549208505467033e-05,
+      "loss": 1.8559,
+      "step": 7550
+    },
+    {
+      "epoch": 0.3300406486297478,
+      "grad_norm": 2.0,
+      "learning_rate": 7.548617591482614e-05,
+      "loss": 1.8498,
+      "step": 7551
+    },
+    {
+      "epoch": 0.3300843568337777,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.548026629402075e-05,
+      "loss": 2.5397,
+      "step": 7552
+    },
+    {
+      "epoch": 0.3301280650378076,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.547435619236562e-05,
+      "loss": 1.9982,
+      "step": 7553
+    },
+    {
+      "epoch": 0.3301717732418375,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.54684456099723e-05,
+      "loss": 1.7053,
+      "step": 7554
+    },
+    {
+      "epoch": 0.3302154814458674,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.546253454695237e-05,
+      "loss": 1.6925,
+      "step": 7555
+    },
+    {
+      "epoch": 0.3302591896498973,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.545662300341736e-05,
+      "loss": 1.8186,
+      "step": 7556
+    },
+    {
+      "epoch": 0.3303028978539272,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.54507109794788e-05,
+      "loss": 2.1328,
+      "step": 7557
+    },
+    {
+      "epoch": 0.33034660605795707,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.544479847524829e-05,
+      "loss": 2.549,
+      "step": 7558
+    },
+    {
+      "epoch": 0.33039031426198695,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.54388854908374e-05,
+      "loss": 2.1843,
+      "step": 7559
+    },
+    {
+      "epoch": 0.3304340224660169,
+      "grad_norm": 2.0,
+      "learning_rate": 7.543297202635772e-05,
+      "loss": 2.3384,
+      "step": 7560
+    },
+    {
+      "epoch": 0.33047773067004677,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.542705808192085e-05,
+      "loss": 1.8247,
+      "step": 7561
+    },
+    {
+      "epoch": 0.33052143887407665,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.542114365763837e-05,
+      "loss": 2.4422,
+      "step": 7562
+    },
+    {
+      "epoch": 0.3305651470781066,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.541522875362193e-05,
+      "loss": 2.1137,
+      "step": 7563
+    },
+    {
+      "epoch": 0.33060885528213646,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.540931336998312e-05,
+      "loss": 1.5262,
+      "step": 7564
+    },
+    {
+      "epoch": 0.33065256348616634,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.540339750683358e-05,
+      "loss": 1.9562,
+      "step": 7565
+    },
+    {
+      "epoch": 0.3306962716901963,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.539748116428495e-05,
+      "loss": 2.2275,
+      "step": 7566
+    },
+    {
+      "epoch": 0.33073997989422615,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.539156434244892e-05,
+      "loss": 1.2463,
+      "step": 7567
+    },
+    {
+      "epoch": 0.33078368809825603,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.53856470414371e-05,
+      "loss": 1.8308,
+      "step": 7568
+    },
+    {
+      "epoch": 0.3308273963022859,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.537972926136115e-05,
+      "loss": 1.3167,
+      "step": 7569
+    },
+    {
+      "epoch": 0.33087110450631585,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.537381100233278e-05,
+      "loss": 1.7028,
+      "step": 7570
+    },
+    {
+      "epoch": 0.33091481271034573,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.536789226446367e-05,
+      "loss": 1.7325,
+      "step": 7571
+    },
+    {
+      "epoch": 0.3309585209143756,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.536197304786555e-05,
+      "loss": 1.6386,
+      "step": 7572
+    },
+    {
+      "epoch": 0.33100222911840554,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.535605335265003e-05,
+      "loss": 1.6623,
+      "step": 7573
+    },
+    {
+      "epoch": 0.3310459373224354,
+      "grad_norm": 2.25,
+      "learning_rate": 7.535013317892889e-05,
+      "loss": 1.4363,
+      "step": 7574
+    },
+    {
+      "epoch": 0.3310896455264653,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.534421252681387e-05,
+      "loss": 2.1629,
+      "step": 7575
+    },
+    {
+      "epoch": 0.33113335373049524,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.533829139641664e-05,
+      "loss": 1.831,
+      "step": 7576
+    },
+    {
+      "epoch": 0.3311770619345251,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.5332369787849e-05,
+      "loss": 1.9585,
+      "step": 7577
+    },
+    {
+      "epoch": 0.331220770138555,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.532644770122266e-05,
+      "loss": 2.5307,
+      "step": 7578
+    },
+    {
+      "epoch": 0.3312644783425849,
+      "grad_norm": 1.9375,
+      "learning_rate": 7.532052513664939e-05,
+      "loss": 1.8876,
+      "step": 7579
+    },
+    {
+      "epoch": 0.3313081865466148,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.531460209424096e-05,
+      "loss": 1.8074,
+      "step": 7580
+    },
+    {
+      "epoch": 0.3313518947506447,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.530867857410915e-05,
+      "loss": 1.6012,
+      "step": 7581
+    },
+    {
+      "epoch": 0.3313956029546746,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.530275457636574e-05,
+      "loss": 1.5269,
+      "step": 7582
+    },
+    {
+      "epoch": 0.3314393111587045,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.529683010112252e-05,
+      "loss": 1.6858,
+      "step": 7583
+    },
+    {
+      "epoch": 0.3314830193627344,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.529090514849128e-05,
+      "loss": 2.4866,
+      "step": 7584
+    },
+    {
+      "epoch": 0.33152672756676427,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.528497971858388e-05,
+      "loss": 1.7441,
+      "step": 7585
+    },
+    {
+      "epoch": 0.3315704357707942,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.52790538115121e-05,
+      "loss": 1.7589,
+      "step": 7586
+    },
+    {
+      "epoch": 0.3316141439748241,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.52731274273878e-05,
+      "loss": 1.9113,
+      "step": 7587
+    },
+    {
+      "epoch": 0.33165785217885396,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.526720056632277e-05,
+      "loss": 1.9697,
+      "step": 7588
+    },
+    {
+      "epoch": 0.33170156038288384,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.52612732284289e-05,
+      "loss": 2.1639,
+      "step": 7589
+    },
+    {
+      "epoch": 0.3317452685869138,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.525534541381806e-05,
+      "loss": 1.6989,
+      "step": 7590
+    },
+    {
+      "epoch": 0.33178897679094366,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.524941712260207e-05,
+      "loss": 1.7631,
+      "step": 7591
+    },
+    {
+      "epoch": 0.33183268499497354,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.524348835489286e-05,
+      "loss": 1.7935,
+      "step": 7592
+    },
+    {
+      "epoch": 0.33187639319900347,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.523755911080226e-05,
+      "loss": 1.505,
+      "step": 7593
+    },
+    {
+      "epoch": 0.33192010140303335,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.523162939044219e-05,
+      "loss": 2.3742,
+      "step": 7594
+    },
+    {
+      "epoch": 0.33196380960706323,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.522569919392455e-05,
+      "loss": 2.1225,
+      "step": 7595
+    },
+    {
+      "epoch": 0.33200751781109317,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.521976852136125e-05,
+      "loss": 1.9097,
+      "step": 7596
+    },
+    {
+      "epoch": 0.33205122601512305,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.521383737286423e-05,
+      "loss": 1.6101,
+      "step": 7597
+    },
+    {
+      "epoch": 0.3320949342191529,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.520790574854538e-05,
+      "loss": 1.6832,
+      "step": 7598
+    },
+    {
+      "epoch": 0.3321386424231828,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.520197364851667e-05,
+      "loss": 1.3584,
+      "step": 7599
+    },
+    {
+      "epoch": 0.33218235062721274,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.519604107289003e-05,
+      "loss": 1.5448,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3322260588312426,
+      "grad_norm": 3.875,
+      "learning_rate": 7.519010802177744e-05,
+      "loss": 2.2089,
+      "step": 7601
+    },
+    {
+      "epoch": 0.3322697670352725,
+      "grad_norm": 1.890625,
+      "learning_rate": 7.518417449529085e-05,
+      "loss": 1.6202,
+      "step": 7602
+    },
+    {
+      "epoch": 0.33231347523930244,
+      "grad_norm": 2.25,
+      "learning_rate": 7.517824049354221e-05,
+      "loss": 1.6155,
+      "step": 7603
+    },
+    {
+      "epoch": 0.3323571834433323,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.517230601664354e-05,
+      "loss": 2.1604,
+      "step": 7604
+    },
+    {
+      "epoch": 0.3324008916473622,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.516637106470683e-05,
+      "loss": 2.687,
+      "step": 7605
+    },
+    {
+      "epoch": 0.33244459985139213,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.516043563784405e-05,
+      "loss": 2.0484,
+      "step": 7606
+    },
+    {
+      "epoch": 0.332488308055422,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.515449973616723e-05,
+      "loss": 1.984,
+      "step": 7607
+    },
+    {
+      "epoch": 0.3325320162594519,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.514856335978842e-05,
+      "loss": 2.0978,
+      "step": 7608
+    },
+    {
+      "epoch": 0.33257572446348177,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.514262650881958e-05,
+      "loss": 1.5738,
+      "step": 7609
+    },
+    {
+      "epoch": 0.3326194326675117,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.51366891833728e-05,
+      "loss": 2.2538,
+      "step": 7610
+    },
+    {
+      "epoch": 0.3326631408715416,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.513075138356012e-05,
+      "loss": 2.524,
+      "step": 7611
+    },
+    {
+      "epoch": 0.33270684907557146,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.512481310949358e-05,
+      "loss": 1.6862,
+      "step": 7612
+    },
+    {
+      "epoch": 0.3327505572796014,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.511887436128525e-05,
+      "loss": 1.6454,
+      "step": 7613
+    },
+    {
+      "epoch": 0.3327942654836313,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.511293513904718e-05,
+      "loss": 2.314,
+      "step": 7614
+    },
+    {
+      "epoch": 0.33283797368766116,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.510699544289151e-05,
+      "loss": 1.6821,
+      "step": 7615
+    },
+    {
+      "epoch": 0.3328816818916911,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.510105527293026e-05,
+      "loss": 1.4252,
+      "step": 7616
+    },
+    {
+      "epoch": 0.332925390095721,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.509511462927559e-05,
+      "loss": 2.2525,
+      "step": 7617
+    },
+    {
+      "epoch": 0.33296909829975085,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.508917351203957e-05,
+      "loss": 1.7308,
+      "step": 7618
+    },
+    {
+      "epoch": 0.33301280650378073,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.508323192133432e-05,
+      "loss": 1.7358,
+      "step": 7619
+    },
+    {
+      "epoch": 0.33305651470781067,
+      "grad_norm": 2.375,
+      "learning_rate": 7.507728985727199e-05,
+      "loss": 2.3117,
+      "step": 7620
+    },
+    {
+      "epoch": 0.33310022291184055,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.50713473199647e-05,
+      "loss": 1.3512,
+      "step": 7621
+    },
+    {
+      "epoch": 0.33314393111587043,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.506540430952461e-05,
+      "loss": 1.5071,
+      "step": 7622
+    },
+    {
+      "epoch": 0.33318763931990036,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.505946082606386e-05,
+      "loss": 1.7641,
+      "step": 7623
+    },
+    {
+      "epoch": 0.33323134752393024,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.505351686969457e-05,
+      "loss": 2.113,
+      "step": 7624
+    },
+    {
+      "epoch": 0.3332750557279601,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.504757244052901e-05,
+      "loss": 1.863,
+      "step": 7625
+    },
+    {
+      "epoch": 0.33331876393199006,
+      "grad_norm": 3.75,
+      "learning_rate": 7.504162753867927e-05,
+      "loss": 2.0576,
+      "step": 7626
+    },
+    {
+      "epoch": 0.33336247213601994,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.503568216425757e-05,
+      "loss": 1.4004,
+      "step": 7627
+    },
+    {
+      "epoch": 0.3334061803400498,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.502973631737612e-05,
+      "loss": 2.2958,
+      "step": 7628
+    },
+    {
+      "epoch": 0.3334498885440797,
+      "grad_norm": 1.828125,
+      "learning_rate": 7.50237899981471e-05,
+      "loss": 1.5195,
+      "step": 7629
+    },
+    {
+      "epoch": 0.33349359674810963,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.501784320668277e-05,
+      "loss": 2.7067,
+      "step": 7630
+    },
+    {
+      "epoch": 0.3335373049521395,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.501189594309531e-05,
+      "loss": 0.8452,
+      "step": 7631
+    },
+    {
+      "epoch": 0.3335810131561694,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.500594820749698e-05,
+      "loss": 2.0745,
+      "step": 7632
+    },
+    {
+      "epoch": 0.33362472136019933,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.715,
+      "step": 7633
+    },
+    {
+      "epoch": 0.3336684295642292,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.499405132071665e-05,
+      "loss": 1.7119,
+      "step": 7634
+    },
+    {
+      "epoch": 0.3337121377682591,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.498810216975917e-05,
+      "loss": 2.5392,
+      "step": 7635
+    },
+    {
+      "epoch": 0.333755845972289,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.498215254723982e-05,
+      "loss": 1.5535,
+      "step": 7636
+    },
+    {
+      "epoch": 0.3337995541763189,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.49762024532709e-05,
+      "loss": 2.2684,
+      "step": 7637
+    },
+    {
+      "epoch": 0.3338432623803488,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.497025188796469e-05,
+      "loss": 1.5744,
+      "step": 7638
+    },
+    {
+      "epoch": 0.33388697058437866,
+      "grad_norm": 1.828125,
+      "learning_rate": 7.496430085143348e-05,
+      "loss": 1.4891,
+      "step": 7639
+    },
+    {
+      "epoch": 0.3339306787884086,
+      "grad_norm": 2.0,
+      "learning_rate": 7.495834934378958e-05,
+      "loss": 1.5024,
+      "step": 7640
+    },
+    {
+      "epoch": 0.3339743869924385,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.495239736514531e-05,
+      "loss": 1.609,
+      "step": 7641
+    },
+    {
+      "epoch": 0.33401809519646836,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.494644491561299e-05,
+      "loss": 2.082,
+      "step": 7642
+    },
+    {
+      "epoch": 0.3340618034004983,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.494049199530494e-05,
+      "loss": 1.2633,
+      "step": 7643
+    },
+    {
+      "epoch": 0.33410551160452817,
+      "grad_norm": 1.90625,
+      "learning_rate": 7.49345386043335e-05,
+      "loss": 1.6481,
+      "step": 7644
+    },
+    {
+      "epoch": 0.33414921980855805,
+      "grad_norm": 1.9375,
+      "learning_rate": 7.492858474281103e-05,
+      "loss": 1.5163,
+      "step": 7645
+    },
+    {
+      "epoch": 0.334192928012588,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.492263041084988e-05,
+      "loss": 1.8762,
+      "step": 7646
+    },
+    {
+      "epoch": 0.33423663621661787,
+      "grad_norm": 4.84375,
+      "learning_rate": 7.491667560856242e-05,
+      "loss": 1.7709,
+      "step": 7647
+    },
+    {
+      "epoch": 0.33428034442064775,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.491072033606104e-05,
+      "loss": 1.7079,
+      "step": 7648
+    },
+    {
+      "epoch": 0.3343240526246776,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.49047645934581e-05,
+      "loss": 1.9067,
+      "step": 7649
+    },
+    {
+      "epoch": 0.33436776082870756,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.4898808380866e-05,
+      "loss": 2.0671,
+      "step": 7650
+    },
+    {
+      "epoch": 0.33441146903273744,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.489285169839717e-05,
+      "loss": 1.4923,
+      "step": 7651
+    },
+    {
+      "epoch": 0.3344551772367673,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.488689454616399e-05,
+      "loss": 2.8122,
+      "step": 7652
+    },
+    {
+      "epoch": 0.33449888544079726,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.488093692427887e-05,
+      "loss": 2.0358,
+      "step": 7653
+    },
+    {
+      "epoch": 0.33454259364482714,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.487497883285428e-05,
+      "loss": 2.0688,
+      "step": 7654
+    },
+    {
+      "epoch": 0.334586301848857,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.486902027200263e-05,
+      "loss": 1.5415,
+      "step": 7655
+    },
+    {
+      "epoch": 0.33463001005288695,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.486306124183637e-05,
+      "loss": 1.4973,
+      "step": 7656
+    },
+    {
+      "epoch": 0.33467371825691683,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.485710174246794e-05,
+      "loss": 1.6159,
+      "step": 7657
+    },
+    {
+      "epoch": 0.3347174264609467,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.485114177400984e-05,
+      "loss": 1.4921,
+      "step": 7658
+    },
+    {
+      "epoch": 0.3347611346649766,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.484518133657455e-05,
+      "loss": 1.5661,
+      "step": 7659
+    },
+    {
+      "epoch": 0.3348048428690065,
+      "grad_norm": 4.96875,
+      "learning_rate": 7.483922043027448e-05,
+      "loss": 1.9808,
+      "step": 7660
+    },
+    {
+      "epoch": 0.3348485510730364,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.48332590552222e-05,
+      "loss": 2.1961,
+      "step": 7661
+    },
+    {
+      "epoch": 0.3348922592770663,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.482729721153016e-05,
+      "loss": 1.6836,
+      "step": 7662
+    },
+    {
+      "epoch": 0.3349359674810962,
+      "grad_norm": 1.828125,
+      "learning_rate": 7.482133489931091e-05,
+      "loss": 1.5001,
+      "step": 7663
+    },
+    {
+      "epoch": 0.3349796756851261,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.481537211867693e-05,
+      "loss": 1.6789,
+      "step": 7664
+    },
+    {
+      "epoch": 0.335023383889156,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.480940886974077e-05,
+      "loss": 1.4361,
+      "step": 7665
+    },
+    {
+      "epoch": 0.3350670920931859,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.480344515261495e-05,
+      "loss": 2.1453,
+      "step": 7666
+    },
+    {
+      "epoch": 0.3351108002972158,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.479748096741201e-05,
+      "loss": 1.6566,
+      "step": 7667
+    },
+    {
+      "epoch": 0.3351545085012457,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.479151631424453e-05,
+      "loss": 1.9521,
+      "step": 7668
+    },
+    {
+      "epoch": 0.33519821670527555,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.478555119322505e-05,
+      "loss": 2.6565,
+      "step": 7669
+    },
+    {
+      "epoch": 0.3352419249093055,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.477958560446613e-05,
+      "loss": 2.3781,
+      "step": 7670
+    },
+    {
+      "epoch": 0.33528563311333537,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.477361954808037e-05,
+      "loss": 1.7341,
+      "step": 7671
+    },
+    {
+      "epoch": 0.33532934131736525,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.476765302418037e-05,
+      "loss": 1.6521,
+      "step": 7672
+    },
+    {
+      "epoch": 0.3353730495213952,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.47616860328787e-05,
+      "loss": 2.4899,
+      "step": 7673
+    },
+    {
+      "epoch": 0.33541675772542506,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.475571857428797e-05,
+      "loss": 1.8374,
+      "step": 7674
+    },
+    {
+      "epoch": 0.33546046592945494,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.474975064852081e-05,
+      "loss": 1.6771,
+      "step": 7675
+    },
+    {
+      "epoch": 0.3355041741334849,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.474378225568983e-05,
+      "loss": 2.0606,
+      "step": 7676
+    },
+    {
+      "epoch": 0.33554788233751476,
+      "grad_norm": 2.25,
+      "learning_rate": 7.473781339590766e-05,
+      "loss": 1.8691,
+      "step": 7677
+    },
+    {
+      "epoch": 0.33559159054154464,
+      "grad_norm": 3.3125,
+      "learning_rate": 7.473184406928696e-05,
+      "loss": 2.2992,
+      "step": 7678
+    },
+    {
+      "epoch": 0.3356352987455745,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.472587427594037e-05,
+      "loss": 1.6181,
+      "step": 7679
+    },
+    {
+      "epoch": 0.33567900694960445,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.471990401598052e-05,
+      "loss": 1.5764,
+      "step": 7680
+    },
+    {
+      "epoch": 0.33572271515363433,
+      "grad_norm": 2.0,
+      "learning_rate": 7.471393328952012e-05,
+      "loss": 1.7561,
+      "step": 7681
+    },
+    {
+      "epoch": 0.3357664233576642,
+      "grad_norm": 2.0,
+      "learning_rate": 7.470796209667184e-05,
+      "loss": 1.7098,
+      "step": 7682
+    },
+    {
+      "epoch": 0.33581013156169415,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.470199043754833e-05,
+      "loss": 1.8006,
+      "step": 7683
+    },
+    {
+      "epoch": 0.335853839765724,
+      "grad_norm": 2.875,
+      "learning_rate": 7.469601831226233e-05,
+      "loss": 1.7551,
+      "step": 7684
+    },
+    {
+      "epoch": 0.3358975479697539,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.469004572092651e-05,
+      "loss": 1.9204,
+      "step": 7685
+    },
+    {
+      "epoch": 0.33594125617378384,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.46840726636536e-05,
+      "loss": 1.3732,
+      "step": 7686
+    },
+    {
+      "epoch": 0.3359849643778137,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.46780991405563e-05,
+      "loss": 1.8284,
+      "step": 7687
+    },
+    {
+      "epoch": 0.3360286725818436,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.467212515174736e-05,
+      "loss": 1.9712,
+      "step": 7688
+    },
+    {
+      "epoch": 0.3360723807858735,
+      "grad_norm": 1.859375,
+      "learning_rate": 7.466615069733951e-05,
+      "loss": 1.4069,
+      "step": 7689
+    },
+    {
+      "epoch": 0.3361160889899034,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.466017577744549e-05,
+      "loss": 2.1121,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3361597971939333,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.465420039217806e-05,
+      "loss": 1.5186,
+      "step": 7691
+    },
+    {
+      "epoch": 0.3362035053979632,
+      "grad_norm": 2.125,
+      "learning_rate": 7.464822454165e-05,
+      "loss": 1.6417,
+      "step": 7692
+    },
+    {
+      "epoch": 0.3362472136019931,
+      "grad_norm": 4.90625,
+      "learning_rate": 7.464224822597407e-05,
+      "loss": 1.5733,
+      "step": 7693
+    },
+    {
+      "epoch": 0.336290921806023,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.463627144526304e-05,
+      "loss": 2.7127,
+      "step": 7694
+    },
+    {
+      "epoch": 0.33633463001005287,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.463029419962971e-05,
+      "loss": 1.4256,
+      "step": 7695
+    },
+    {
+      "epoch": 0.3363783382140828,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.462431648918689e-05,
+      "loss": 1.8544,
+      "step": 7696
+    },
+    {
+      "epoch": 0.3364220464181127,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.461833831404737e-05,
+      "loss": 2.4548,
+      "step": 7697
+    },
+    {
+      "epoch": 0.33646575462214257,
+      "grad_norm": 4.78125,
+      "learning_rate": 7.461235967432398e-05,
+      "loss": 2.2272,
+      "step": 7698
+    },
+    {
+      "epoch": 0.33650946282617245,
+      "grad_norm": 5.53125,
+      "learning_rate": 7.460638057012955e-05,
+      "loss": 2.3309,
+      "step": 7699
+    },
+    {
+      "epoch": 0.3365531710302024,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.46004010015769e-05,
+      "loss": 2.1454,
+      "step": 7700
+    },
+    {
+      "epoch": 0.33659687923423226,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.459442096877886e-05,
+      "loss": 1.3449,
+      "step": 7701
+    },
+    {
+      "epoch": 0.33664058743826214,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.458844047184832e-05,
+      "loss": 2.5731,
+      "step": 7702
+    },
+    {
+      "epoch": 0.3366842956422921,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.458245951089813e-05,
+      "loss": 1.9036,
+      "step": 7703
+    },
+    {
+      "epoch": 0.33672800384632195,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.457647808604113e-05,
+      "loss": 2.333,
+      "step": 7704
+    },
+    {
+      "epoch": 0.33677171205035183,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.457049619739022e-05,
+      "loss": 1.6744,
+      "step": 7705
+    },
+    {
+      "epoch": 0.33681542025438177,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.45645138450583e-05,
+      "loss": 1.5279,
+      "step": 7706
+    },
+    {
+      "epoch": 0.33685912845841165,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.455853102915825e-05,
+      "loss": 1.5336,
+      "step": 7707
+    },
+    {
+      "epoch": 0.33690283666244153,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.455254774980297e-05,
+      "loss": 1.969,
+      "step": 7708
+    },
+    {
+      "epoch": 0.3369465448664714,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.45465640071054e-05,
+      "loss": 2.3044,
+      "step": 7709
+    },
+    {
+      "epoch": 0.33699025307050134,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.454057980117841e-05,
+      "loss": 1.9823,
+      "step": 7710
+    },
+    {
+      "epoch": 0.3370339612745312,
+      "grad_norm": 1.875,
+      "learning_rate": 7.453459513213498e-05,
+      "loss": 1.6091,
+      "step": 7711
+    },
+    {
+      "epoch": 0.3370776694785611,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.452861000008803e-05,
+      "loss": 2.207,
+      "step": 7712
+    },
+    {
+      "epoch": 0.33712137768259104,
+      "grad_norm": 1.859375,
+      "learning_rate": 7.45226244051505e-05,
+      "loss": 1.6165,
+      "step": 7713
+    },
+    {
+      "epoch": 0.3371650858866209,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.451663834743537e-05,
+      "loss": 1.8244,
+      "step": 7714
+    },
+    {
+      "epoch": 0.3372087940906508,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.451065182705558e-05,
+      "loss": 1.6859,
+      "step": 7715
+    },
+    {
+      "epoch": 0.33725250229468073,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.450466484412413e-05,
+      "loss": 2.2727,
+      "step": 7716
+    },
+    {
+      "epoch": 0.3372962104987106,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.449867739875397e-05,
+      "loss": 1.7578,
+      "step": 7717
+    },
+    {
+      "epoch": 0.3373399187027405,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.449268949105812e-05,
+      "loss": 1.8431,
+      "step": 7718
+    },
+    {
+      "epoch": 0.33738362690677043,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.448670112114959e-05,
+      "loss": 2.004,
+      "step": 7719
+    },
+    {
+      "epoch": 0.3374273351108003,
+      "grad_norm": 2.125,
+      "learning_rate": 7.448071228914134e-05,
+      "loss": 1.9159,
+      "step": 7720
+    },
+    {
+      "epoch": 0.3374710433148302,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.447472299514644e-05,
+      "loss": 1.911,
+      "step": 7721
+    },
+    {
+      "epoch": 0.33751475151886007,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.44687332392779e-05,
+      "loss": 1.8569,
+      "step": 7722
+    },
+    {
+      "epoch": 0.33755845972289,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.446274302164873e-05,
+      "loss": 1.5678,
+      "step": 7723
+    },
+    {
+      "epoch": 0.3376021679269199,
+      "grad_norm": 2.0,
+      "learning_rate": 7.445675234237202e-05,
+      "loss": 1.7527,
+      "step": 7724
+    },
+    {
+      "epoch": 0.33764587613094976,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.445076120156078e-05,
+      "loss": 2.0955,
+      "step": 7725
+    },
+    {
+      "epoch": 0.3376895843349797,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.44447695993281e-05,
+      "loss": 1.7765,
+      "step": 7726
+    },
+    {
+      "epoch": 0.3377332925390096,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.443877753578702e-05,
+      "loss": 1.4968,
+      "step": 7727
+    },
+    {
+      "epoch": 0.33777700074303946,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.443278501105065e-05,
+      "loss": 2.7519,
+      "step": 7728
+    },
+    {
+      "epoch": 0.3378207089470694,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.442679202523208e-05,
+      "loss": 2.2485,
+      "step": 7729
+    },
+    {
+      "epoch": 0.33786441715109927,
+      "grad_norm": 2.25,
+      "learning_rate": 7.442079857844438e-05,
+      "loss": 2.299,
+      "step": 7730
+    },
+    {
+      "epoch": 0.33790812535512915,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.441480467080066e-05,
+      "loss": 1.8709,
+      "step": 7731
+    },
+    {
+      "epoch": 0.33795183355915903,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.440881030241407e-05,
+      "loss": 2.7728,
+      "step": 7732
+    },
+    {
+      "epoch": 0.33799554176318897,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.44028154733977e-05,
+      "loss": 1.9273,
+      "step": 7733
+    },
+    {
+      "epoch": 0.33803924996721885,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.439682018386467e-05,
+      "loss": 1.672,
+      "step": 7734
+    },
+    {
+      "epoch": 0.3380829581712487,
+      "grad_norm": 2.5,
+      "learning_rate": 7.439082443392813e-05,
+      "loss": 2.3186,
+      "step": 7735
+    },
+    {
+      "epoch": 0.33812666637527866,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.438482822370124e-05,
+      "loss": 1.9351,
+      "step": 7736
+    },
+    {
+      "epoch": 0.33817037457930854,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.437883155329715e-05,
+      "loss": 1.9687,
+      "step": 7737
+    },
+    {
+      "epoch": 0.3382140827833384,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.437283442282904e-05,
+      "loss": 1.5673,
+      "step": 7738
+    },
+    {
+      "epoch": 0.33825779098736836,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.436683683241006e-05,
+      "loss": 2.2424,
+      "step": 7739
+    },
+    {
+      "epoch": 0.33830149919139824,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.43608387821534e-05,
+      "loss": 2.1129,
+      "step": 7740
+    },
+    {
+      "epoch": 0.3383452073954281,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.435484027217225e-05,
+      "loss": 1.741,
+      "step": 7741
+    },
+    {
+      "epoch": 0.338388915599458,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.434884130257985e-05,
+      "loss": 1.89,
+      "step": 7742
+    },
+    {
+      "epoch": 0.33843262380348793,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.434284187348935e-05,
+      "loss": 1.9559,
+      "step": 7743
+    },
+    {
+      "epoch": 0.3384763320075178,
+      "grad_norm": 2.0,
+      "learning_rate": 7.4336841985014e-05,
+      "loss": 1.6128,
+      "step": 7744
+    },
+    {
+      "epoch": 0.3385200402115477,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.433084163726703e-05,
+      "loss": 1.7827,
+      "step": 7745
+    },
+    {
+      "epoch": 0.3385637484155776,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.432484083036165e-05,
+      "loss": 1.7955,
+      "step": 7746
+    },
+    {
+      "epoch": 0.3386074566196075,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.431883956441112e-05,
+      "loss": 2.5427,
+      "step": 7747
+    },
+    {
+      "epoch": 0.3386511648236374,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.431283783952872e-05,
+      "loss": 1.9095,
+      "step": 7748
+    },
+    {
+      "epoch": 0.3386948730276673,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.430683565582766e-05,
+      "loss": 1.592,
+      "step": 7749
+    },
+    {
+      "epoch": 0.3387385812316972,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.430083301342124e-05,
+      "loss": 2.0556,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3387822894357271,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.429482991242274e-05,
+      "loss": 1.5385,
+      "step": 7751
+    },
+    {
+      "epoch": 0.33882599763975696,
+      "grad_norm": 1.5390625,
+      "learning_rate": 7.428882635294543e-05,
+      "loss": 1.3219,
+      "step": 7752
+    },
+    {
+      "epoch": 0.3388697058437869,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.428282233510262e-05,
+      "loss": 1.8182,
+      "step": 7753
+    },
+    {
+      "epoch": 0.3389134140478168,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.427681785900761e-05,
+      "loss": 2.379,
+      "step": 7754
+    },
+    {
+      "epoch": 0.33895712225184665,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.427081292477371e-05,
+      "loss": 2.1487,
+      "step": 7755
+    },
+    {
+      "epoch": 0.3390008304558766,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.426480753251425e-05,
+      "loss": 2.2427,
+      "step": 7756
+    },
+    {
+      "epoch": 0.33904453865990647,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.425880168234256e-05,
+      "loss": 2.2215,
+      "step": 7757
+    },
+    {
+      "epoch": 0.33908824686393635,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.425279537437198e-05,
+      "loss": 1.9807,
+      "step": 7758
+    },
+    {
+      "epoch": 0.3391319550679663,
+      "grad_norm": 6.96875,
+      "learning_rate": 7.424678860871584e-05,
+      "loss": 3.9342,
+      "step": 7759
+    },
+    {
+      "epoch": 0.33917566327199616,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.42407813854875e-05,
+      "loss": 2.0916,
+      "step": 7760
+    },
+    {
+      "epoch": 0.33921937147602604,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.423477370480035e-05,
+      "loss": 1.8062,
+      "step": 7761
+    },
+    {
+      "epoch": 0.3392630796800559,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.422876556676776e-05,
+      "loss": 1.7046,
+      "step": 7762
+    },
+    {
+      "epoch": 0.33930678788408586,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.422275697150308e-05,
+      "loss": 1.4491,
+      "step": 7763
+    },
+    {
+      "epoch": 0.33935049608811574,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.421674791911973e-05,
+      "loss": 1.9007,
+      "step": 7764
+    },
+    {
+      "epoch": 0.3393942042921456,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.42107384097311e-05,
+      "loss": 1.5789,
+      "step": 7765
+    },
+    {
+      "epoch": 0.33943791249617555,
+      "grad_norm": 1.875,
+      "learning_rate": 7.420472844345059e-05,
+      "loss": 1.7002,
+      "step": 7766
+    },
+    {
+      "epoch": 0.33948162070020543,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.419871802039163e-05,
+      "loss": 1.5454,
+      "step": 7767
+    },
+    {
+      "epoch": 0.3395253289042353,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.419270714066765e-05,
+      "loss": 1.8283,
+      "step": 7768
+    },
+    {
+      "epoch": 0.33956903710826525,
+      "grad_norm": 1.859375,
+      "learning_rate": 7.418669580439209e-05,
+      "loss": 1.5417,
+      "step": 7769
+    },
+    {
+      "epoch": 0.33961274531229513,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.418068401167834e-05,
+      "loss": 1.8136,
+      "step": 7770
+    },
+    {
+      "epoch": 0.339656453516325,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.41746717626399e-05,
+      "loss": 1.7942,
+      "step": 7771
+    },
+    {
+      "epoch": 0.3397001617203549,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.416865905739024e-05,
+      "loss": 2.0847,
+      "step": 7772
+    },
+    {
+      "epoch": 0.3397438699243848,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.41626458960428e-05,
+      "loss": 1.6278,
+      "step": 7773
+    },
+    {
+      "epoch": 0.3397875781284147,
+      "grad_norm": 2.375,
+      "learning_rate": 7.415663227871106e-05,
+      "loss": 2.0389,
+      "step": 7774
+    },
+    {
+      "epoch": 0.3398312863324446,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.41506182055085e-05,
+      "loss": 1.7392,
+      "step": 7775
+    },
+    {
+      "epoch": 0.3398749945364745,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.414460367654864e-05,
+      "loss": 2.5617,
+      "step": 7776
+    },
+    {
+      "epoch": 0.3399187027405044,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.413858869194496e-05,
+      "loss": 1.6048,
+      "step": 7777
+    },
+    {
+      "epoch": 0.3399624109445343,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.413257325181098e-05,
+      "loss": 1.5458,
+      "step": 7778
+    },
+    {
+      "epoch": 0.3400061191485642,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.412655735626024e-05,
+      "loss": 2.3651,
+      "step": 7779
+    },
+    {
+      "epoch": 0.3400498273525941,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.412054100540623e-05,
+      "loss": 2.3673,
+      "step": 7780
+    },
+    {
+      "epoch": 0.34009353555662397,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.41145241993625e-05,
+      "loss": 2.4358,
+      "step": 7781
+    },
+    {
+      "epoch": 0.34013724376065385,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.410850693824261e-05,
+      "loss": 1.8383,
+      "step": 7782
+    },
+    {
+      "epoch": 0.3401809519646838,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.41024892221601e-05,
+      "loss": 2.3828,
+      "step": 7783
+    },
+    {
+      "epoch": 0.34022466016871367,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.409647105122854e-05,
+      "loss": 1.704,
+      "step": 7784
+    },
+    {
+      "epoch": 0.34026836837274355,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.409045242556151e-05,
+      "loss": 1.6124,
+      "step": 7785
+    },
+    {
+      "epoch": 0.3403120765767735,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.408443334527257e-05,
+      "loss": 1.0824,
+      "step": 7786
+    },
+    {
+      "epoch": 0.34035578478080336,
+      "grad_norm": 28.25,
+      "learning_rate": 7.407841381047532e-05,
+      "loss": 2.9776,
+      "step": 7787
+    },
+    {
+      "epoch": 0.34039949298483324,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.407239382128336e-05,
+      "loss": 1.7465,
+      "step": 7788
+    },
+    {
+      "epoch": 0.3404432011888632,
+      "grad_norm": 2.25,
+      "learning_rate": 7.406637337781031e-05,
+      "loss": 2.5876,
+      "step": 7789
+    },
+    {
+      "epoch": 0.34048690939289306,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.406035248016973e-05,
+      "loss": 1.9449,
+      "step": 7790
+    },
+    {
+      "epoch": 0.34053061759692294,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.40543311284753e-05,
+      "loss": 1.52,
+      "step": 7791
+    },
+    {
+      "epoch": 0.3405743258009528,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.404830932284064e-05,
+      "loss": 1.8827,
+      "step": 7792
+    },
+    {
+      "epoch": 0.34061803400498275,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.404228706337937e-05,
+      "loss": 2.3619,
+      "step": 7793
+    },
+    {
+      "epoch": 0.34066174220901263,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.403626435020516e-05,
+      "loss": 1.6467,
+      "step": 7794
+    },
+    {
+      "epoch": 0.3407054504130425,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.403024118343167e-05,
+      "loss": 2.2609,
+      "step": 7795
+    },
+    {
+      "epoch": 0.34074915861707245,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.402421756317252e-05,
+      "loss": 1.1447,
+      "step": 7796
+    },
+    {
+      "epoch": 0.3407928668211023,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.401819348954144e-05,
+      "loss": 1.6363,
+      "step": 7797
+    },
+    {
+      "epoch": 0.3408365750251322,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.401216896265208e-05,
+      "loss": 1.7098,
+      "step": 7798
+    },
+    {
+      "epoch": 0.34088028322916214,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.400614398261817e-05,
+      "loss": 2.1047,
+      "step": 7799
+    },
+    {
+      "epoch": 0.340923991433192,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.400011854955336e-05,
+      "loss": 1.68,
+      "step": 7800
+    },
+    {
+      "epoch": 0.3409676996372219,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.399409266357139e-05,
+      "loss": 1.8467,
+      "step": 7801
+    },
+    {
+      "epoch": 0.3410114078412518,
+      "grad_norm": 3.0,
+      "learning_rate": 7.398806632478598e-05,
+      "loss": 2.5555,
+      "step": 7802
+    },
+    {
+      "epoch": 0.3410551160452817,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.398203953331083e-05,
+      "loss": 1.5894,
+      "step": 7803
+    },
+    {
+      "epoch": 0.3410988242493116,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.39760122892597e-05,
+      "loss": 2.3018,
+      "step": 7804
+    },
+    {
+      "epoch": 0.3411425324533415,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.396998459274632e-05,
+      "loss": 1.5333,
+      "step": 7805
+    },
+    {
+      "epoch": 0.3411862406573714,
+      "grad_norm": 2.625,
+      "learning_rate": 7.396395644388443e-05,
+      "loss": 2.7594,
+      "step": 7806
+    },
+    {
+      "epoch": 0.3412299488614013,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.395792784278783e-05,
+      "loss": 1.6819,
+      "step": 7807
+    },
+    {
+      "epoch": 0.34127365706543117,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.395189878957025e-05,
+      "loss": 2.0002,
+      "step": 7808
+    },
+    {
+      "epoch": 0.3413173652694611,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.394586928434549e-05,
+      "loss": 1.7576,
+      "step": 7809
+    },
+    {
+      "epoch": 0.341361073473491,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.39398393272273e-05,
+      "loss": 1.5846,
+      "step": 7810
+    },
+    {
+      "epoch": 0.34140478167752086,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.393380891832951e-05,
+      "loss": 1.8085,
+      "step": 7811
+    },
+    {
+      "epoch": 0.34144848988155074,
+      "grad_norm": 3.328125,
+      "learning_rate": 7.392777805776592e-05,
+      "loss": 2.6784,
+      "step": 7812
+    },
+    {
+      "epoch": 0.3414921980855807,
+      "grad_norm": 1.8828125,
+      "learning_rate": 7.392174674565031e-05,
+      "loss": 1.4711,
+      "step": 7813
+    },
+    {
+      "epoch": 0.34153590628961056,
+      "grad_norm": 1.984375,
+      "learning_rate": 7.391571498209654e-05,
+      "loss": 1.9854,
+      "step": 7814
+    },
+    {
+      "epoch": 0.34157961449364044,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.390968276721844e-05,
+      "loss": 1.5118,
+      "step": 7815
+    },
+    {
+      "epoch": 0.3416233226976704,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.390365010112979e-05,
+      "loss": 2.5905,
+      "step": 7816
+    },
+    {
+      "epoch": 0.34166703090170025,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.389761698394449e-05,
+      "loss": 1.9055,
+      "step": 7817
+    },
+    {
+      "epoch": 0.34171073910573013,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.389158341577638e-05,
+      "loss": 1.9238,
+      "step": 7818
+    },
+    {
+      "epoch": 0.34175444730976007,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.388554939673931e-05,
+      "loss": 1.9666,
+      "step": 7819
+    },
+    {
+      "epoch": 0.34179815551378995,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.387951492694717e-05,
+      "loss": 1.648,
+      "step": 7820
+    },
+    {
+      "epoch": 0.3418418637178198,
+      "grad_norm": 2.5,
+      "learning_rate": 7.387348000651381e-05,
+      "loss": 2.3185,
+      "step": 7821
+    },
+    {
+      "epoch": 0.3418855719218497,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.386744463555316e-05,
+      "loss": 1.9617,
+      "step": 7822
+    },
+    {
+      "epoch": 0.34192928012587964,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.386140881417907e-05,
+      "loss": 1.8481,
+      "step": 7823
+    },
+    {
+      "epoch": 0.3419729883299095,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.385537254250549e-05,
+      "loss": 1.8291,
+      "step": 7824
+    },
+    {
+      "epoch": 0.3420166965339394,
+      "grad_norm": 3.734375,
+      "learning_rate": 7.38493358206463e-05,
+      "loss": 1.8099,
+      "step": 7825
+    },
+    {
+      "epoch": 0.34206040473796934,
+      "grad_norm": 2.75,
+      "learning_rate": 7.384329864871542e-05,
+      "loss": 2.5644,
+      "step": 7826
+    },
+    {
+      "epoch": 0.3421041129419992,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.38372610268268e-05,
+      "loss": 1.8474,
+      "step": 7827
+    },
+    {
+      "epoch": 0.3421478211460291,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.383122295509437e-05,
+      "loss": 2.8546,
+      "step": 7828
+    },
+    {
+      "epoch": 0.34219152935005903,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.382518443363208e-05,
+      "loss": 2.1391,
+      "step": 7829
+    },
+    {
+      "epoch": 0.3422352375540889,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.38191454625539e-05,
+      "loss": 2.2937,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3422789457581188,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.381310604197375e-05,
+      "loss": 1.79,
+      "step": 7831
+    },
+    {
+      "epoch": 0.34232265396214867,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.380706617200564e-05,
+      "loss": 2.197,
+      "step": 7832
+    },
+    {
+      "epoch": 0.3423663621661786,
+      "grad_norm": 2.0,
+      "learning_rate": 7.380102585276355e-05,
+      "loss": 2.2665,
+      "step": 7833
+    },
+    {
+      "epoch": 0.3424100703702085,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.379498508436146e-05,
+      "loss": 1.7406,
+      "step": 7834
+    },
+    {
+      "epoch": 0.34245377857423837,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.378894386691337e-05,
+      "loss": 2.1328,
+      "step": 7835
+    },
+    {
+      "epoch": 0.3424974867782683,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.378290220053328e-05,
+      "loss": 1.7703,
+      "step": 7836
+    },
+    {
+      "epoch": 0.3425411949822982,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.377686008533521e-05,
+      "loss": 2.2732,
+      "step": 7837
+    },
+    {
+      "epoch": 0.34258490318632806,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.377081752143319e-05,
+      "loss": 1.8219,
+      "step": 7838
+    },
+    {
+      "epoch": 0.342628611390358,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.376477450894124e-05,
+      "loss": 1.7202,
+      "step": 7839
+    },
+    {
+      "epoch": 0.3426723195943879,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.375873104797341e-05,
+      "loss": 1.6177,
+      "step": 7840
+    },
+    {
+      "epoch": 0.34271602779841775,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.375268713864374e-05,
+      "loss": 1.2808,
+      "step": 7841
+    },
+    {
+      "epoch": 0.34275973600244763,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.374664278106631e-05,
+      "loss": 2.2372,
+      "step": 7842
+    },
+    {
+      "epoch": 0.34280344420647757,
+      "grad_norm": 1.8359375,
+      "learning_rate": 7.374059797535517e-05,
+      "loss": 2.0804,
+      "step": 7843
+    },
+    {
+      "epoch": 0.34284715241050745,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.373455272162438e-05,
+      "loss": 1.7002,
+      "step": 7844
+    },
+    {
+      "epoch": 0.34289086061453733,
+      "grad_norm": 2.75,
+      "learning_rate": 7.372850701998803e-05,
+      "loss": 1.9006,
+      "step": 7845
+    },
+    {
+      "epoch": 0.34293456881856726,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.372246087056023e-05,
+      "loss": 2.5089,
+      "step": 7846
+    },
+    {
+      "epoch": 0.34297827702259714,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.371641427345506e-05,
+      "loss": 1.567,
+      "step": 7847
+    },
+    {
+      "epoch": 0.343021985226627,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.371036722878664e-05,
+      "loss": 2.1754,
+      "step": 7848
+    },
+    {
+      "epoch": 0.34306569343065696,
+      "grad_norm": 3.265625,
+      "learning_rate": 7.370431973666909e-05,
+      "loss": 2.6514,
+      "step": 7849
+    },
+    {
+      "epoch": 0.34310940163468684,
+      "grad_norm": 2.25,
+      "learning_rate": 7.369827179721651e-05,
+      "loss": 1.7452,
+      "step": 7850
+    },
+    {
+      "epoch": 0.3431531098387167,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.369222341054305e-05,
+      "loss": 1.6974,
+      "step": 7851
+    },
+    {
+      "epoch": 0.3431968180427466,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.368617457676286e-05,
+      "loss": 2.2866,
+      "step": 7852
+    },
+    {
+      "epoch": 0.34324052624677653,
+      "grad_norm": 3.578125,
+      "learning_rate": 7.36801252959901e-05,
+      "loss": 2.6568,
+      "step": 7853
+    },
+    {
+      "epoch": 0.3432842344508064,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.367407556833887e-05,
+      "loss": 1.8624,
+      "step": 7854
+    },
+    {
+      "epoch": 0.3433279426548363,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.366802539392341e-05,
+      "loss": 2.464,
+      "step": 7855
+    },
+    {
+      "epoch": 0.34337165085886623,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.366197477285785e-05,
+      "loss": 2.198,
+      "step": 7856
+    },
+    {
+      "epoch": 0.3434153590628961,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.365592370525639e-05,
+      "loss": 2.274,
+      "step": 7857
+    },
+    {
+      "epoch": 0.343459067266926,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.364987219123323e-05,
+      "loss": 1.6779,
+      "step": 7858
+    },
+    {
+      "epoch": 0.3435027754709559,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.364382023090255e-05,
+      "loss": 2.1957,
+      "step": 7859
+    },
+    {
+      "epoch": 0.3435464836749858,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.363776782437857e-05,
+      "loss": 1.9011,
+      "step": 7860
+    },
+    {
+      "epoch": 0.3435901918790157,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.36317149717755e-05,
+      "loss": 2.4219,
+      "step": 7861
+    },
+    {
+      "epoch": 0.34363390008304556,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.362566167320759e-05,
+      "loss": 1.9962,
+      "step": 7862
+    },
+    {
+      "epoch": 0.3436776082870755,
+      "grad_norm": 1.8359375,
+      "learning_rate": 7.361960792878906e-05,
+      "loss": 1.5491,
+      "step": 7863
+    },
+    {
+      "epoch": 0.3437213164911054,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.361355373863414e-05,
+      "loss": 1.6535,
+      "step": 7864
+    },
+    {
+      "epoch": 0.34376502469513526,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.360749910285711e-05,
+      "loss": 1.7978,
+      "step": 7865
+    },
+    {
+      "epoch": 0.3438087328991652,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.360144402157218e-05,
+      "loss": 1.3264,
+      "step": 7866
+    },
+    {
+      "epoch": 0.3438524411031951,
+      "grad_norm": 1.96875,
+      "learning_rate": 7.359538849489367e-05,
+      "loss": 1.3655,
+      "step": 7867
+    },
+    {
+      "epoch": 0.34389614930722495,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.358933252293585e-05,
+      "loss": 1.5282,
+      "step": 7868
+    },
+    {
+      "epoch": 0.3439398575112549,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.3583276105813e-05,
+      "loss": 2.3046,
+      "step": 7869
+    },
+    {
+      "epoch": 0.34398356571528477,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.357721924363937e-05,
+      "loss": 1.8875,
+      "step": 7870
+    },
+    {
+      "epoch": 0.34402727391931465,
+      "grad_norm": 1.859375,
+      "learning_rate": 7.357116193652931e-05,
+      "loss": 1.6185,
+      "step": 7871
+    },
+    {
+      "epoch": 0.3440709821233445,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.356510418459714e-05,
+      "loss": 1.548,
+      "step": 7872
+    },
+    {
+      "epoch": 0.34411469032737446,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.355904598795713e-05,
+      "loss": 1.9057,
+      "step": 7873
+    },
+    {
+      "epoch": 0.34415839853140434,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.355298734672364e-05,
+      "loss": 1.7803,
+      "step": 7874
+    },
+    {
+      "epoch": 0.3442021067354342,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.354692826101102e-05,
+      "loss": 2.1805,
+      "step": 7875
+    },
+    {
+      "epoch": 0.34424581493946416,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.354086873093356e-05,
+      "loss": 1.8715,
+      "step": 7876
+    },
+    {
+      "epoch": 0.34428952314349404,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.353480875660566e-05,
+      "loss": 1.8297,
+      "step": 7877
+    },
+    {
+      "epoch": 0.3443332313475239,
+      "grad_norm": 2.75,
+      "learning_rate": 7.352874833814168e-05,
+      "loss": 2.1321,
+      "step": 7878
+    },
+    {
+      "epoch": 0.34437693955155385,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.352268747565596e-05,
+      "loss": 2.0992,
+      "step": 7879
+    },
+    {
+      "epoch": 0.34442064775558373,
+      "grad_norm": 1.890625,
+      "learning_rate": 7.351662616926289e-05,
+      "loss": 1.477,
+      "step": 7880
+    },
+    {
+      "epoch": 0.3444643559596136,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.351056441907687e-05,
+      "loss": 1.5112,
+      "step": 7881
+    },
+    {
+      "epoch": 0.3445080641636435,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.350450222521226e-05,
+      "loss": 1.8943,
+      "step": 7882
+    },
+    {
+      "epoch": 0.3445517723676734,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.34984395877835e-05,
+      "loss": 1.7028,
+      "step": 7883
+    },
+    {
+      "epoch": 0.3445954805717033,
+      "grad_norm": 2.0,
+      "learning_rate": 7.349237650690497e-05,
+      "loss": 1.7619,
+      "step": 7884
+    },
+    {
+      "epoch": 0.3446391887757332,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.348631298269114e-05,
+      "loss": 1.6977,
+      "step": 7885
+    },
+    {
+      "epoch": 0.3446828969797631,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.348024901525635e-05,
+      "loss": 2.221,
+      "step": 7886
+    },
+    {
+      "epoch": 0.344726605183793,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.347418460471511e-05,
+      "loss": 1.3263,
+      "step": 7887
+    },
+    {
+      "epoch": 0.3447703133878229,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.346811975118185e-05,
+      "loss": 1.9857,
+      "step": 7888
+    },
+    {
+      "epoch": 0.3448140215918528,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.346205445477101e-05,
+      "loss": 1.9187,
+      "step": 7889
+    },
+    {
+      "epoch": 0.3448577297958827,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.345598871559706e-05,
+      "loss": 2.4341,
+      "step": 7890
+    },
+    {
+      "epoch": 0.3449014379999126,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.344992253377445e-05,
+      "loss": 1.9575,
+      "step": 7891
+    },
+    {
+      "epoch": 0.34494514620394245,
+      "grad_norm": 2.125,
+      "learning_rate": 7.344385590941768e-05,
+      "loss": 1.8408,
+      "step": 7892
+    },
+    {
+      "epoch": 0.3449888544079724,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.343778884264123e-05,
+      "loss": 1.9551,
+      "step": 7893
+    },
+    {
+      "epoch": 0.34503256261200227,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.343172133355958e-05,
+      "loss": 1.682,
+      "step": 7894
+    },
+    {
+      "epoch": 0.34507627081603215,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.342565338228726e-05,
+      "loss": 1.4362,
+      "step": 7895
+    },
+    {
+      "epoch": 0.3451199790200621,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.341958498893876e-05,
+      "loss": 1.7903,
+      "step": 7896
+    },
+    {
+      "epoch": 0.34516368722409196,
+      "grad_norm": 2.25,
+      "learning_rate": 7.34135161536286e-05,
+      "loss": 1.8678,
+      "step": 7897
+    },
+    {
+      "epoch": 0.34520739542812184,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.340744687647133e-05,
+      "loss": 1.4317,
+      "step": 7898
+    },
+    {
+      "epoch": 0.3452511036321518,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.340137715758146e-05,
+      "loss": 1.769,
+      "step": 7899
+    },
+    {
+      "epoch": 0.34529481183618166,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.339530699707354e-05,
+      "loss": 1.5794,
+      "step": 7900
+    },
+    {
+      "epoch": 0.34533852004021154,
+      "grad_norm": 3.25,
+      "learning_rate": 7.338923639506213e-05,
+      "loss": 2.1146,
+      "step": 7901
+    },
+    {
+      "epoch": 0.3453822282442414,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.338316535166179e-05,
+      "loss": 2.3914,
+      "step": 7902
+    },
+    {
+      "epoch": 0.34542593644827135,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.337709386698709e-05,
+      "loss": 1.5035,
+      "step": 7903
+    },
+    {
+      "epoch": 0.34546964465230123,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.33710219411526e-05,
+      "loss": 2.1352,
+      "step": 7904
+    },
+    {
+      "epoch": 0.3455133528563311,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.336494957427292e-05,
+      "loss": 1.7889,
+      "step": 7905
+    },
+    {
+      "epoch": 0.34555706106036105,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.335887676646263e-05,
+      "loss": 2.0182,
+      "step": 7906
+    },
+    {
+      "epoch": 0.34560076926439093,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.335280351783632e-05,
+      "loss": 1.5878,
+      "step": 7907
+    },
+    {
+      "epoch": 0.3456444774684208,
+      "grad_norm": 2.375,
+      "learning_rate": 7.334672982850865e-05,
+      "loss": 2.4043,
+      "step": 7908
+    },
+    {
+      "epoch": 0.34568818567245074,
+      "grad_norm": 2.5,
+      "learning_rate": 7.334065569859419e-05,
+      "loss": 2.0702,
+      "step": 7909
+    },
+    {
+      "epoch": 0.3457318938764806,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.333458112820758e-05,
+      "loss": 1.916,
+      "step": 7910
+    },
+    {
+      "epoch": 0.3457756020805105,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.332850611746346e-05,
+      "loss": 2.2221,
+      "step": 7911
+    },
+    {
+      "epoch": 0.3458193102845404,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.332243066647651e-05,
+      "loss": 2.2696,
+      "step": 7912
+    },
+    {
+      "epoch": 0.3458630184885703,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.331635477536131e-05,
+      "loss": 1.8497,
+      "step": 7913
+    },
+    {
+      "epoch": 0.3459067266926002,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.331027844423258e-05,
+      "loss": 1.73,
+      "step": 7914
+    },
+    {
+      "epoch": 0.3459504348966301,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.330420167320498e-05,
+      "loss": 2.7669,
+      "step": 7915
+    },
+    {
+      "epoch": 0.34599414310066,
+      "grad_norm": 6.78125,
+      "learning_rate": 7.329812446239315e-05,
+      "loss": 2.766,
+      "step": 7916
+    },
+    {
+      "epoch": 0.3460378513046899,
+      "grad_norm": 1.90625,
+      "learning_rate": 7.329204681191183e-05,
+      "loss": 1.4046,
+      "step": 7917
+    },
+    {
+      "epoch": 0.34608155950871977,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.328596872187567e-05,
+      "loss": 2.738,
+      "step": 7918
+    },
+    {
+      "epoch": 0.3461252677127497,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.327989019239938e-05,
+      "loss": 2.2535,
+      "step": 7919
+    },
+    {
+      "epoch": 0.3461689759167796,
+      "grad_norm": 2.75,
+      "learning_rate": 7.32738112235977e-05,
+      "loss": 2.1258,
+      "step": 7920
+    },
+    {
+      "epoch": 0.34621268412080947,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.326773181558532e-05,
+      "loss": 1.9796,
+      "step": 7921
+    },
+    {
+      "epoch": 0.34625639232483935,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.326165196847697e-05,
+      "loss": 1.7916,
+      "step": 7922
+    },
+    {
+      "epoch": 0.3463001005288693,
+      "grad_norm": 1.84375,
+      "learning_rate": 7.32555716823874e-05,
+      "loss": 1.8068,
+      "step": 7923
+    },
+    {
+      "epoch": 0.34634380873289916,
+      "grad_norm": 3.53125,
+      "learning_rate": 7.324949095743134e-05,
+      "loss": 2.206,
+      "step": 7924
+    },
+    {
+      "epoch": 0.34638751693692904,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.324340979372356e-05,
+      "loss": 2.2156,
+      "step": 7925
+    },
+    {
+      "epoch": 0.346431225140959,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.32373281913788e-05,
+      "loss": 1.6678,
+      "step": 7926
+    },
+    {
+      "epoch": 0.34647493334498886,
+      "grad_norm": 2.25,
+      "learning_rate": 7.323124615051183e-05,
+      "loss": 1.9071,
+      "step": 7927
+    },
+    {
+      "epoch": 0.34651864154901874,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.322516367123744e-05,
+      "loss": 2.493,
+      "step": 7928
+    },
+    {
+      "epoch": 0.34656234975304867,
+      "grad_norm": 3.515625,
+      "learning_rate": 7.321908075367041e-05,
+      "loss": 1.8614,
+      "step": 7929
+    },
+    {
+      "epoch": 0.34660605795707855,
+      "grad_norm": 2.046875,
+      "learning_rate": 7.321299739792552e-05,
+      "loss": 1.7816,
+      "step": 7930
+    },
+    {
+      "epoch": 0.34664976616110843,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.32069136041176e-05,
+      "loss": 1.5655,
+      "step": 7931
+    },
+    {
+      "epoch": 0.3466934743651383,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.320082937236144e-05,
+      "loss": 1.8418,
+      "step": 7932
+    },
+    {
+      "epoch": 0.34673718256916825,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.319474470277187e-05,
+      "loss": 2.4229,
+      "step": 7933
+    },
+    {
+      "epoch": 0.3467808907731981,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.318865959546369e-05,
+      "loss": 1.6238,
+      "step": 7934
+    },
+    {
+      "epoch": 0.346824598977228,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.318257405055178e-05,
+      "loss": 1.6683,
+      "step": 7935
+    },
+    {
+      "epoch": 0.34686830718125794,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.317648806815094e-05,
+      "loss": 2.9313,
+      "step": 7936
+    },
+    {
+      "epoch": 0.3469120153852878,
+      "grad_norm": 1.9296875,
+      "learning_rate": 7.317040164837604e-05,
+      "loss": 1.7694,
+      "step": 7937
+    },
+    {
+      "epoch": 0.3469557235893177,
+      "grad_norm": 1.9375,
+      "learning_rate": 7.316431479134194e-05,
+      "loss": 1.7356,
+      "step": 7938
+    },
+    {
+      "epoch": 0.34699943179334763,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.31582274971635e-05,
+      "loss": 1.583,
+      "step": 7939
+    },
+    {
+      "epoch": 0.3470431399973775,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.315213976595561e-05,
+      "loss": 1.7361,
+      "step": 7940
+    },
+    {
+      "epoch": 0.3470868482014074,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.314605159783314e-05,
+      "loss": 1.5803,
+      "step": 7941
+    },
+    {
+      "epoch": 0.3471305564054373,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.313996299291098e-05,
+      "loss": 1.6847,
+      "step": 7942
+    },
+    {
+      "epoch": 0.3471742646094672,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.313387395130406e-05,
+      "loss": 1.7097,
+      "step": 7943
+    },
+    {
+      "epoch": 0.3472179728134971,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.312778447312725e-05,
+      "loss": 2.1191,
+      "step": 7944
+    },
+    {
+      "epoch": 0.34726168101752697,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.312169455849551e-05,
+      "loss": 2.0175,
+      "step": 7945
+    },
+    {
+      "epoch": 0.3473053892215569,
+      "grad_norm": 4.1875,
+      "learning_rate": 7.311560420752373e-05,
+      "loss": 2.3162,
+      "step": 7946
+    },
+    {
+      "epoch": 0.3473490974255868,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.310951342032684e-05,
+      "loss": 1.7003,
+      "step": 7947
+    },
+    {
+      "epoch": 0.34739280562961666,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.310342219701981e-05,
+      "loss": 2.1897,
+      "step": 7948
+    },
+    {
+      "epoch": 0.3474365138336466,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.309733053771758e-05,
+      "loss": 2.1538,
+      "step": 7949
+    },
+    {
+      "epoch": 0.3474802220376765,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.309123844253511e-05,
+      "loss": 1.8306,
+      "step": 7950
+    },
+    {
+      "epoch": 0.34752393024170636,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.308514591158735e-05,
+      "loss": 1.3273,
+      "step": 7951
+    },
+    {
+      "epoch": 0.34756763844573624,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.307905294498929e-05,
+      "loss": 1.599,
+      "step": 7952
+    },
+    {
+      "epoch": 0.3476113466497662,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.30729595428559e-05,
+      "loss": 2.3818,
+      "step": 7953
+    },
+    {
+      "epoch": 0.34765505485379605,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.306686570530221e-05,
+      "loss": 2.0856,
+      "step": 7954
+    },
+    {
+      "epoch": 0.34769876305782593,
+      "grad_norm": 2.140625,
+      "learning_rate": 7.30607714324432e-05,
+      "loss": 2.4284,
+      "step": 7955
+    },
+    {
+      "epoch": 0.34774247126185587,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.305467672439384e-05,
+      "loss": 1.9622,
+      "step": 7956
+    },
+    {
+      "epoch": 0.34778617946588575,
+      "grad_norm": 1.9609375,
+      "learning_rate": 7.304858158126917e-05,
+      "loss": 1.7486,
+      "step": 7957
+    },
+    {
+      "epoch": 0.3478298876699156,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.304248600318425e-05,
+      "loss": 1.6015,
+      "step": 7958
+    },
+    {
+      "epoch": 0.34787359587394556,
+      "grad_norm": 2.171875,
+      "learning_rate": 7.303638999025406e-05,
+      "loss": 1.9826,
+      "step": 7959
+    },
+    {
+      "epoch": 0.34791730407797544,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.303029354259367e-05,
+      "loss": 1.6595,
+      "step": 7960
+    },
+    {
+      "epoch": 0.3479610122820053,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.302419666031813e-05,
+      "loss": 1.8291,
+      "step": 7961
+    },
+    {
+      "epoch": 0.3480047204860352,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.301809934354248e-05,
+      "loss": 1.7561,
+      "step": 7962
+    },
+    {
+      "epoch": 0.34804842869006514,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.30120015923818e-05,
+      "loss": 2.1202,
+      "step": 7963
+    },
+    {
+      "epoch": 0.348092136894095,
+      "grad_norm": 1.859375,
+      "learning_rate": 7.300590340695115e-05,
+      "loss": 1.761,
+      "step": 7964
+    },
+    {
+      "epoch": 0.3481358450981249,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.299980478736564e-05,
+      "loss": 1.82,
+      "step": 7965
+    },
+    {
+      "epoch": 0.34817955330215483,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.299370573374031e-05,
+      "loss": 1.8937,
+      "step": 7966
+    },
+    {
+      "epoch": 0.3482232615061847,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.298760624619029e-05,
+      "loss": 2.0898,
+      "step": 7967
+    },
+    {
+      "epoch": 0.3482669697102146,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.29815063248307e-05,
+      "loss": 2.0204,
+      "step": 7968
+    },
+    {
+      "epoch": 0.3483106779142445,
+      "grad_norm": 1.8359375,
+      "learning_rate": 7.297540596977662e-05,
+      "loss": 1.4455,
+      "step": 7969
+    },
+    {
+      "epoch": 0.3483543861182744,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.29693051811432e-05,
+      "loss": 1.4206,
+      "step": 7970
+    },
+    {
+      "epoch": 0.3483980943223043,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.296320395904556e-05,
+      "loss": 2.2885,
+      "step": 7971
+    },
+    {
+      "epoch": 0.34844180252633417,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.295710230359885e-05,
+      "loss": 2.0993,
+      "step": 7972
+    },
+    {
+      "epoch": 0.3484855107303641,
+      "grad_norm": 2.078125,
+      "learning_rate": 7.295100021491818e-05,
+      "loss": 1.7229,
+      "step": 7973
+    },
+    {
+      "epoch": 0.348529218934394,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.294489769311876e-05,
+      "loss": 1.7387,
+      "step": 7974
+    },
+    {
+      "epoch": 0.34857292713842386,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.293879473831572e-05,
+      "loss": 1.8183,
+      "step": 7975
+    },
+    {
+      "epoch": 0.3486166353424538,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.293269135062424e-05,
+      "loss": 1.9519,
+      "step": 7976
+    },
+    {
+      "epoch": 0.3486603435464837,
+      "grad_norm": 1.953125,
+      "learning_rate": 7.292658753015948e-05,
+      "loss": 1.6782,
+      "step": 7977
+    },
+    {
+      "epoch": 0.34870405175051356,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.292048327703666e-05,
+      "loss": 2.2993,
+      "step": 7978
+    },
+    {
+      "epoch": 0.3487477599545435,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.291437859137095e-05,
+      "loss": 1.7695,
+      "step": 7979
+    },
+    {
+      "epoch": 0.34879146815857337,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.290827347327758e-05,
+      "loss": 1.7113,
+      "step": 7980
+    },
+    {
+      "epoch": 0.34883517636260325,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.290216792287175e-05,
+      "loss": 1.5739,
+      "step": 7981
+    },
+    {
+      "epoch": 0.34887888456663313,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.289606194026866e-05,
+      "loss": 1.8195,
+      "step": 7982
+    },
+    {
+      "epoch": 0.34892259277066306,
+      "grad_norm": 1.9765625,
+      "learning_rate": 7.288995552558357e-05,
+      "loss": 1.7554,
+      "step": 7983
+    },
+    {
+      "epoch": 0.34896630097469294,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.28838486789317e-05,
+      "loss": 1.7297,
+      "step": 7984
+    },
+    {
+      "epoch": 0.3490100091787228,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.28777414004283e-05,
+      "loss": 2.0312,
+      "step": 7985
+    },
+    {
+      "epoch": 0.34905371738275276,
+      "grad_norm": 2.125,
+      "learning_rate": 7.287163369018863e-05,
+      "loss": 2.0484,
+      "step": 7986
+    },
+    {
+      "epoch": 0.34909742558678264,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.286552554832793e-05,
+      "loss": 2.0505,
+      "step": 7987
+    },
+    {
+      "epoch": 0.3491411337908125,
+      "grad_norm": 2.125,
+      "learning_rate": 7.28594169749615e-05,
+      "loss": 1.859,
+      "step": 7988
+    },
+    {
+      "epoch": 0.34918484199484245,
+      "grad_norm": 2.03125,
+      "learning_rate": 7.285330797020458e-05,
+      "loss": 2.0386,
+      "step": 7989
+    },
+    {
+      "epoch": 0.34922855019887233,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.28471985341725e-05,
+      "loss": 1.5225,
+      "step": 7990
+    },
+    {
+      "epoch": 0.3492722584029022,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.284108866698051e-05,
+      "loss": 1.7554,
+      "step": 7991
+    },
+    {
+      "epoch": 0.3493159666069321,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.283497836874396e-05,
+      "loss": 1.5883,
+      "step": 7992
+    },
+    {
+      "epoch": 0.34935967481096203,
+      "grad_norm": 1.9375,
+      "learning_rate": 7.282886763957812e-05,
+      "loss": 1.3351,
+      "step": 7993
+    },
+    {
+      "epoch": 0.3494033830149919,
+      "grad_norm": 1.9140625,
+      "learning_rate": 7.282275647959831e-05,
+      "loss": 1.6072,
+      "step": 7994
+    },
+    {
+      "epoch": 0.3494470912190218,
+      "grad_norm": 3.0,
+      "learning_rate": 7.281664488891988e-05,
+      "loss": 2.1401,
+      "step": 7995
+    },
+    {
+      "epoch": 0.3494907994230517,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.281053286765815e-05,
+      "loss": 1.2633,
+      "step": 7996
+    },
+    {
+      "epoch": 0.3495345076270816,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.280442041592846e-05,
+      "loss": 2.2913,
+      "step": 7997
+    },
+    {
+      "epoch": 0.3495782158311115,
+      "grad_norm": 2.375,
+      "learning_rate": 7.279830753384618e-05,
+      "loss": 2.0087,
+      "step": 7998
+    },
+    {
+      "epoch": 0.3496219240351414,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.279219422152666e-05,
+      "loss": 2.0633,
+      "step": 7999
+    },
+    {
+      "epoch": 0.3496656322391713,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.278608047908523e-05,
+      "loss": 1.8298,
+      "step": 8000
+    },
+    {
+      "epoch": 0.3497093404432012,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.277996630663734e-05,
+      "loss": 1.0935,
+      "step": 8001
+    },
+    {
+      "epoch": 0.3497530486472311,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.27738517042983e-05,
+      "loss": 1.3579,
+      "step": 8002
+    },
+    {
+      "epoch": 0.349796756851261,
+      "grad_norm": 2.015625,
+      "learning_rate": 7.276773667218354e-05,
+      "loss": 1.889,
+      "step": 8003
+    },
+    {
+      "epoch": 0.3498404650552909,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.276162121040846e-05,
+      "loss": 1.8152,
+      "step": 8004
+    },
+    {
+      "epoch": 0.34988417325932075,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.275550531908846e-05,
+      "loss": 1.5692,
+      "step": 8005
+    },
+    {
+      "epoch": 0.3499278814633507,
+      "grad_norm": 3.8125,
+      "learning_rate": 7.274938899833896e-05,
+      "loss": 2.0927,
+      "step": 8006
+    },
+    {
+      "epoch": 0.34997158966738057,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.274327224827535e-05,
+      "loss": 1.6025,
+      "step": 8007
+    },
+    {
+      "epoch": 0.35001529787141045,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.273715506901312e-05,
+      "loss": 2.1043,
+      "step": 8008
+    },
+    {
+      "epoch": 0.3500590060754404,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.273103746066767e-05,
+      "loss": 2.2297,
+      "step": 8009
+    },
+    {
+      "epoch": 0.35010271427947026,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.272491942335447e-05,
+      "loss": 1.5105,
+      "step": 8010
+    },
+    {
+      "epoch": 0.35014642248350014,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.271880095718895e-05,
+      "loss": 1.8723,
+      "step": 8011
+    },
+    {
+      "epoch": 0.3501901306875301,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.27126820622866e-05,
+      "loss": 1.9394,
+      "step": 8012
+    },
+    {
+      "epoch": 0.35023383889155996,
+      "grad_norm": 1.9921875,
+      "learning_rate": 7.270656273876289e-05,
+      "loss": 1.9664,
+      "step": 8013
+    },
+    {
+      "epoch": 0.35027754709558984,
+      "grad_norm": 1.921875,
+      "learning_rate": 7.270044298673328e-05,
+      "loss": 1.7432,
+      "step": 8014
+    },
+    {
+      "epoch": 0.3503212552996197,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.269432280631327e-05,
+      "loss": 1.5492,
+      "step": 8015
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 22879,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 229,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.2123448025219072e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}