diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,21616 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.995172609630441,
+  "eval_steps": 500,
+  "global_step": 3081,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000973393900064893,
+      "grad_norm": 3.8125,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.9271,
+      "step": 1
+    },
+    {
+      "epoch": 0.001946787800129786,
+      "grad_norm": 3.8125,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.947,
+      "step": 2
+    },
+    {
+      "epoch": 0.0029201817001946787,
+      "grad_norm": 3.5625,
+      "learning_rate": 1.5e-06,
+      "loss": 1.8864,
+      "step": 3
+    },
+    {
+      "epoch": 0.003893575600259572,
+      "grad_norm": 3.546875,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.9048,
+      "step": 4
+    },
+    {
+      "epoch": 0.004866969500324465,
+      "grad_norm": 3.40625,
+      "learning_rate": 2.5e-06,
+      "loss": 1.9077,
+      "step": 5
+    },
+    {
+      "epoch": 0.005840363400389357,
+      "grad_norm": 3.0,
+      "learning_rate": 3e-06,
+      "loss": 1.9014,
+      "step": 6
+    },
+    {
+      "epoch": 0.00681375730045425,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.5e-06,
+      "loss": 1.912,
+      "step": 7
+    },
+    {
+      "epoch": 0.007787151200519144,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.8813,
+      "step": 8
+    },
+    {
+      "epoch": 0.008760545100584036,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.5e-06,
+      "loss": 1.8683,
+      "step": 9
+    },
+    {
+      "epoch": 0.00973393900064893,
+      "grad_norm": 1.8125,
+      "learning_rate": 5e-06,
+      "loss": 1.9061,
+      "step": 10
+    },
+    {
+      "epoch": 0.010707332900713823,
+      "grad_norm": 1.765625,
+      "learning_rate": 4.999998691872464e-06,
+      "loss": 1.8596,
+      "step": 11
+    },
+    {
+      "epoch": 0.011680726800778715,
+      "grad_norm": 1.7734375,
+      "learning_rate": 4.9999947674912255e-06,
+      "loss": 1.9086,
+      "step": 12
+    },
+    {
+      "epoch": 0.012654120700843608,
+      "grad_norm": 1.71875,
+      "learning_rate": 4.99998822686039e-06,
+      "loss": 1.8465,
+      "step": 13
+    },
+    {
+      "epoch": 0.0136275146009085,
+      "grad_norm": 1.609375,
+      "learning_rate": 4.999979069986803e-06,
+      "loss": 1.835,
+      "step": 14
+    },
+    {
+      "epoch": 0.014600908500973394,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.999967296880048e-06,
+      "loss": 1.8464,
+      "step": 15
+    },
+    {
+      "epoch": 0.015574302401038288,
+      "grad_norm": 1.375,
+      "learning_rate": 4.999952907552444e-06,
+      "loss": 1.8715,
+      "step": 16
+    },
+    {
+      "epoch": 0.01654769630110318,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.999935902019051e-06,
+      "loss": 1.8614,
+      "step": 17
+    },
+    {
+      "epoch": 0.01752109020116807,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.999916280297664e-06,
+      "loss": 1.8847,
+      "step": 18
+    },
+    {
+      "epoch": 0.018494484101232965,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.999894042408818e-06,
+      "loss": 1.8204,
+      "step": 19
+    },
+    {
+      "epoch": 0.01946787800129786,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.999869188375784e-06,
+      "loss": 1.8313,
+      "step": 20
+    },
+    {
+      "epoch": 0.020441271901362752,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.999841718224574e-06,
+      "loss": 1.8221,
+      "step": 21
+    },
+    {
+      "epoch": 0.021414665801427646,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.999811631983934e-06,
+      "loss": 1.8274,
+      "step": 22
+    },
+    {
+      "epoch": 0.022388059701492536,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.999778929685348e-06,
+      "loss": 1.8124,
+      "step": 23
+    },
+    {
+      "epoch": 0.02336145360155743,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.999743611363042e-06,
+      "loss": 1.8222,
+      "step": 24
+    },
+    {
+      "epoch": 0.024334847501622323,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.9997056770539745e-06,
+      "loss": 1.8039,
+      "step": 25
+    },
+    {
+      "epoch": 0.025308241401687217,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.9996651267978446e-06,
+      "loss": 1.7916,
+      "step": 26
+    },
+    {
+      "epoch": 0.02628163530175211,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.999621960637089e-06,
+      "loss": 1.7906,
+      "step": 27
+    },
+    {
+      "epoch": 0.027255029201817,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.99957617861688e-06,
+      "loss": 1.8106,
+      "step": 28
+    },
+    {
+      "epoch": 0.028228423101881894,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.999527780785127e-06,
+      "loss": 1.7894,
+      "step": 29
+    },
+    {
+      "epoch": 0.029201817001946788,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.9994767671924826e-06,
+      "loss": 1.7524,
+      "step": 30
+    },
+    {
+      "epoch": 0.03017521090201168,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.999423137892329e-06,
+      "loss": 1.809,
+      "step": 31
+    },
+    {
+      "epoch": 0.031148604802076575,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.9993668929407916e-06,
+      "loss": 1.7914,
+      "step": 32
+    },
+    {
+      "epoch": 0.03212199870214147,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.999308032396729e-06,
+      "loss": 1.7832,
+      "step": 33
+    },
+    {
+      "epoch": 0.03309539260220636,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.999246556321741e-06,
+      "loss": 1.7584,
+      "step": 34
+    },
+    {
+      "epoch": 0.03406878650227125,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.999182464780161e-06,
+      "loss": 1.7985,
+      "step": 35
+    },
+    {
+      "epoch": 0.03504218040233614,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.99911575783906e-06,
+      "loss": 1.7827,
+      "step": 36
+    },
+    {
+      "epoch": 0.036015574302401036,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.99904643556825e-06,
+      "loss": 1.7835,
+      "step": 37
+    },
+    {
+      "epoch": 0.03698896820246593,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.998974498040273e-06,
+      "loss": 1.7315,
+      "step": 38
+    },
+    {
+      "epoch": 0.037962362102530824,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.998899945330416e-06,
+      "loss": 1.7871,
+      "step": 39
+    },
+    {
+      "epoch": 0.03893575600259572,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.9988227775166954e-06,
+      "loss": 1.7366,
+      "step": 40
+    },
+    {
+      "epoch": 0.03990914990266061,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.9987429946798684e-06,
+      "loss": 1.7826,
+      "step": 41
+    },
+    {
+      "epoch": 0.040882543802725504,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.998660596903428e-06,
+      "loss": 1.772,
+      "step": 42
+    },
+    {
+      "epoch": 0.0418559377027904,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.998575584273604e-06,
+      "loss": 1.7348,
+      "step": 43
+    },
+    {
+      "epoch": 0.04282933160285529,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.998487956879361e-06,
+      "loss": 1.755,
+      "step": 44
+    },
+    {
+      "epoch": 0.04380272550292018,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.998397714812403e-06,
+      "loss": 1.7603,
+      "step": 45
+    },
+    {
+      "epoch": 0.04477611940298507,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.9983048581671676e-06,
+      "loss": 1.7151,
+      "step": 46
+    },
+    {
+      "epoch": 0.045749513303049966,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.998209387040829e-06,
+      "loss": 1.7456,
+      "step": 47
+    },
+    {
+      "epoch": 0.04672290720311486,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.998111301533299e-06,
+      "loss": 1.7458,
+      "step": 48
+    },
+    {
+      "epoch": 0.04769630110317975,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.998010601747223e-06,
+      "loss": 1.7416,
+      "step": 49
+    },
+    {
+      "epoch": 0.04866969500324465,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.997907287787985e-06,
+      "loss": 1.7257,
+      "step": 50
+    },
+    {
+      "epoch": 0.04964308890330954,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.997801359763702e-06,
+      "loss": 1.742,
+      "step": 51
+    },
+    {
+      "epoch": 0.050616482803374434,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.997692817785229e-06,
+      "loss": 1.7521,
+      "step": 52
+    },
+    {
+      "epoch": 0.05158987670343933,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.997581661966154e-06,
+      "loss": 1.7133,
+      "step": 53
+    },
+    {
+      "epoch": 0.05256327060350422,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.997467892422804e-06,
+      "loss": 1.7177,
+      "step": 54
+    },
+    {
+      "epoch": 0.05353666450356911,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.997351509274236e-06,
+      "loss": 1.7096,
+      "step": 55
+    },
+    {
+      "epoch": 0.054510058403634,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.997232512642248e-06,
+      "loss": 1.7169,
+      "step": 56
+    },
+    {
+      "epoch": 0.055483452303698895,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.997110902651368e-06,
+      "loss": 1.7097,
+      "step": 57
+    },
+    {
+      "epoch": 0.05645684620376379,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.996986679428863e-06,
+      "loss": 1.696,
+      "step": 58
+    },
+    {
+      "epoch": 0.05743024010382868,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.996859843104732e-06,
+      "loss": 1.7385,
+      "step": 59
+    },
+    {
+      "epoch": 0.058403634003893576,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.9967303938117095e-06,
+      "loss": 1.6892,
+      "step": 60
+    },
+    {
+      "epoch": 0.05937702790395847,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.9965983316852655e-06,
+      "loss": 1.7318,
+      "step": 61
+    },
+    {
+      "epoch": 0.06035042180402336,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.996463656863601e-06,
+      "loss": 1.7081,
+      "step": 62
+    },
+    {
+      "epoch": 0.06132381570408826,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.996326369487654e-06,
+      "loss": 1.7108,
+      "step": 63
+    },
+    {
+      "epoch": 0.06229720960415315,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.996186469701098e-06,
+      "loss": 1.7007,
+      "step": 64
+    },
+    {
+      "epoch": 0.06327060350421804,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.996043957650337e-06,
+      "loss": 1.7047,
+      "step": 65
+    },
+    {
+      "epoch": 0.06424399740428294,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.99589883348451e-06,
+      "loss": 1.6933,
+      "step": 66
+    },
+    {
+      "epoch": 0.06521739130434782,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.99575109735549e-06,
+      "loss": 1.7098,
+      "step": 67
+    },
+    {
+      "epoch": 0.06619078520441272,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.995600749417883e-06,
+      "loss": 1.7075,
+      "step": 68
+    },
+    {
+      "epoch": 0.06716417910447761,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.9954477898290285e-06,
+      "loss": 1.7302,
+      "step": 69
+    },
+    {
+      "epoch": 0.0681375730045425,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.9952922187490005e-06,
+      "loss": 1.6886,
+      "step": 70
+    },
+    {
+      "epoch": 0.0691109669046074,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.995134036340602e-06,
+      "loss": 1.6862,
+      "step": 71
+    },
+    {
+      "epoch": 0.07008436080467229,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.994973242769372e-06,
+      "loss": 1.698,
+      "step": 72
+    },
+    {
+      "epoch": 0.07105775470473719,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.994809838203582e-06,
+      "loss": 1.6917,
+      "step": 73
+    },
+    {
+      "epoch": 0.07203114860480207,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.9946438228142345e-06,
+      "loss": 1.6791,
+      "step": 74
+    },
+    {
+      "epoch": 0.07300454250486697,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.994475196775066e-06,
+      "loss": 1.6923,
+      "step": 75
+    },
+    {
+      "epoch": 0.07397793640493186,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.9943039602625435e-06,
+      "loss": 1.6791,
+      "step": 76
+    },
+    {
+      "epoch": 0.07495133030499676,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.994130113455865e-06,
+      "loss": 1.7237,
+      "step": 77
+    },
+    {
+      "epoch": 0.07592472420506165,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.9939536565369625e-06,
+      "loss": 1.6223,
+      "step": 78
+    },
+    {
+      "epoch": 0.07689811810512655,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.9937745896905e-06,
+      "loss": 1.6775,
+      "step": 79
+    },
+    {
+      "epoch": 0.07787151200519143,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.993592913103868e-06,
+      "loss": 1.6915,
+      "step": 80
+    },
+    {
+      "epoch": 0.07884490590525632,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.993408626967196e-06,
+      "loss": 1.6782,
+      "step": 81
+    },
+    {
+      "epoch": 0.07981829980532122,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.993221731473335e-06,
+      "loss": 1.6581,
+      "step": 82
+    },
+    {
+      "epoch": 0.08079169370538611,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.993032226817874e-06,
+      "loss": 1.6621,
+      "step": 83
+    },
+    {
+      "epoch": 0.08176508760545101,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.992840113199131e-06,
+      "loss": 1.667,
+      "step": 84
+    },
+    {
+      "epoch": 0.0827384815055159,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.992645390818151e-06,
+      "loss": 1.6677,
+      "step": 85
+    },
+    {
+      "epoch": 0.0837118754055808,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.992448059878713e-06,
+      "loss": 1.6475,
+      "step": 86
+    },
+    {
+      "epoch": 0.08468526930564568,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.992248120587323e-06,
+      "loss": 1.6859,
+      "step": 87
+    },
+    {
+      "epoch": 0.08565866320571058,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.992045573153218e-06,
+      "loss": 1.6847,
+      "step": 88
+    },
+    {
+      "epoch": 0.08663205710577547,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.9918404177883655e-06,
+      "loss": 1.6697,
+      "step": 89
+    },
+    {
+      "epoch": 0.08760545100584036,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.99163265470746e-06,
+      "loss": 1.6756,
+      "step": 90
+    },
+    {
+      "epoch": 0.08857884490590526,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.991422284127927e-06,
+      "loss": 1.6589,
+      "step": 91
+    },
+    {
+      "epoch": 0.08955223880597014,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.991209306269918e-06,
+      "loss": 1.6587,
+      "step": 92
+    },
+    {
+      "epoch": 0.09052563270603504,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.990993721356317e-06,
+      "loss": 1.6607,
+      "step": 93
+    },
+    {
+      "epoch": 0.09149902660609993,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.990775529612732e-06,
+      "loss": 1.6582,
+      "step": 94
+    },
+    {
+      "epoch": 0.09247242050616483,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.990554731267502e-06,
+      "loss": 1.669,
+      "step": 95
+    },
+    {
+      "epoch": 0.09344581440622972,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.990331326551693e-06,
+      "loss": 1.6586,
+      "step": 96
+    },
+    {
+      "epoch": 0.09441920830629462,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.9901053156990984e-06,
+      "loss": 1.6385,
+      "step": 97
+    },
+    {
+      "epoch": 0.0953926022063595,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.9898766989462385e-06,
+      "loss": 1.6393,
+      "step": 98
+    },
+    {
+      "epoch": 0.0963659961064244,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.989645476532362e-06,
+      "loss": 1.6478,
+      "step": 99
+    },
+    {
+      "epoch": 0.0973393900064893,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.9894116486994425e-06,
+      "loss": 1.648,
+      "step": 100
+    },
+    {
+      "epoch": 0.09831278390655418,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.9891752156921835e-06,
+      "loss": 1.6419,
+      "step": 101
+    },
+    {
+      "epoch": 0.09928617780661908,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.988936177758011e-06,
+      "loss": 1.6285,
+      "step": 102
+    },
+    {
+      "epoch": 0.10025957170668397,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.9886945351470775e-06,
+      "loss": 1.6654,
+      "step": 103
+    },
+    {
+      "epoch": 0.10123296560674887,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.988450288112265e-06,
+      "loss": 1.6355,
+      "step": 104
+    },
+    {
+      "epoch": 0.10220635950681375,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.988203436909177e-06,
+      "loss": 1.6376,
+      "step": 105
+    },
+    {
+      "epoch": 0.10317975340687865,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.987953981796144e-06,
+      "loss": 1.6474,
+      "step": 106
+    },
+    {
+      "epoch": 0.10415314730694354,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.987701923034222e-06,
+      "loss": 1.641,
+      "step": 107
+    },
+    {
+      "epoch": 0.10512654120700844,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.98744726088719e-06,
+      "loss": 1.6212,
+      "step": 108
+    },
+    {
+      "epoch": 0.10609993510707333,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.987189995621553e-06,
+      "loss": 1.6218,
+      "step": 109
+    },
+    {
+      "epoch": 0.10707332900713822,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.9869301275065394e-06,
+      "loss": 1.6329,
+      "step": 110
+    },
+    {
+      "epoch": 0.10804672290720312,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.986667656814102e-06,
+      "loss": 1.6328,
+      "step": 111
+    },
+    {
+      "epoch": 0.109020116807268,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.9864025838189165e-06,
+      "loss": 1.6055,
+      "step": 112
+    },
+    {
+      "epoch": 0.1099935107073329,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.986134908798383e-06,
+      "loss": 1.6214,
+      "step": 113
+    },
+    {
+      "epoch": 0.11096690460739779,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.985864632032623e-06,
+      "loss": 1.6265,
+      "step": 114
+    },
+    {
+      "epoch": 0.11194029850746269,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.985591753804483e-06,
+      "loss": 1.6191,
+      "step": 115
+    },
+    {
+      "epoch": 0.11291369240752758,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.985316274399529e-06,
+      "loss": 1.615,
+      "step": 116
+    },
+    {
+      "epoch": 0.11388708630759248,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.985038194106052e-06,
+      "loss": 1.6472,
+      "step": 117
+    },
+    {
+      "epoch": 0.11486048020765736,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.984757513215063e-06,
+      "loss": 1.6116,
+      "step": 118
+    },
+    {
+      "epoch": 0.11583387410772227,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.984474232020296e-06,
+      "loss": 1.5963,
+      "step": 119
+    },
+    {
+      "epoch": 0.11680726800778715,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.984188350818204e-06,
+      "loss": 1.6276,
+      "step": 120
+    },
+    {
+      "epoch": 0.11778066190785204,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.983899869907963e-06,
+      "loss": 1.6063,
+      "step": 121
+    },
+    {
+      "epoch": 0.11875405580791694,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.983608789591468e-06,
+      "loss": 1.6289,
+      "step": 122
+    },
+    {
+      "epoch": 0.11972744970798183,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.983315110173337e-06,
+      "loss": 1.587,
+      "step": 123
+    },
+    {
+      "epoch": 0.12070084360804673,
+      "grad_norm": 0.5,
+      "learning_rate": 4.9830188319609045e-06,
+      "loss": 1.629,
+      "step": 124
+    },
+    {
+      "epoch": 0.12167423750811161,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.982719955264227e-06,
+      "loss": 1.6104,
+      "step": 125
+    },
+    {
+      "epoch": 0.12264763140817651,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.9824184803960794e-06,
+      "loss": 1.571,
+      "step": 126
+    },
+    {
+      "epoch": 0.1236210253082414,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.982114407671955e-06,
+      "loss": 1.6102,
+      "step": 127
+    },
+    {
+      "epoch": 0.1245944192083063,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.981807737410068e-06,
+      "loss": 1.6285,
+      "step": 128
+    },
+    {
+      "epoch": 0.1255678131083712,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.981498469931348e-06,
+      "loss": 1.6085,
+      "step": 129
+    },
+    {
+      "epoch": 0.12654120700843607,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.981186605559445e-06,
+      "loss": 1.6036,
+      "step": 130
+    },
+    {
+      "epoch": 0.12751460090850097,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.980872144620726e-06,
+      "loss": 1.603,
+      "step": 131
+    },
+    {
+      "epoch": 0.12848799480856588,
+      "grad_norm": 0.5,
+      "learning_rate": 4.9805550874442735e-06,
+      "loss": 1.6262,
+      "step": 132
+    },
+    {
+      "epoch": 0.12946138870863075,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.9802354343618895e-06,
+      "loss": 1.6447,
+      "step": 133
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.979913185708093e-06,
+      "loss": 1.5935,
+      "step": 134
+    },
+    {
+      "epoch": 0.13140817650876055,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.979588341820114e-06,
+      "loss": 1.6112,
+      "step": 135
+    },
+    {
+      "epoch": 0.13238157040882545,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.979260903037906e-06,
+      "loss": 1.6473,
+      "step": 136
+    },
+    {
+      "epoch": 0.13335496430889032,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.9789308697041325e-06,
+      "loss": 1.633,
+      "step": 137
+    },
+    {
+      "epoch": 0.13432835820895522,
+      "grad_norm": 0.5,
+      "learning_rate": 4.9785982421641756e-06,
+      "loss": 1.607,
+      "step": 138
+    },
+    {
+      "epoch": 0.13530175210902012,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.978263020766129e-06,
+      "loss": 1.616,
+      "step": 139
+    },
+    {
+      "epoch": 0.136275146009085,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.977925205860803e-06,
+      "loss": 1.598,
+      "step": 140
+    },
+    {
+      "epoch": 0.1372485399091499,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.977584797801722e-06,
+      "loss": 1.6183,
+      "step": 141
+    },
+    {
+      "epoch": 0.1382219338092148,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.977241796945123e-06,
+      "loss": 1.5891,
+      "step": 142
+    },
+    {
+      "epoch": 0.1391953277092797,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.976896203649958e-06,
+      "loss": 1.6197,
+      "step": 143
+    },
+    {
+      "epoch": 0.14016872160934457,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.976548018277891e-06,
+      "loss": 1.607,
+      "step": 144
+    },
+    {
+      "epoch": 0.14114211550940947,
+      "grad_norm": 1.9140625,
+      "learning_rate": 4.976197241193298e-06,
+      "loss": 1.6206,
+      "step": 145
+    },
+    {
+      "epoch": 0.14211550940947437,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.975843872763269e-06,
+      "loss": 1.5968,
+      "step": 146
+    },
+    {
+      "epoch": 0.14308890330953927,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.975487913357603e-06,
+      "loss": 1.5844,
+      "step": 147
+    },
+    {
+      "epoch": 0.14406229720960415,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.975129363348814e-06,
+      "loss": 1.5843,
+      "step": 148
+    },
+    {
+      "epoch": 0.14503569110966905,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.974768223112125e-06,
+      "loss": 1.6291,
+      "step": 149
+    },
+    {
+      "epoch": 0.14600908500973395,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.974404493025469e-06,
+      "loss": 1.5926,
+      "step": 150
+    },
+    {
+      "epoch": 0.14698247890979882,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.974038173469491e-06,
+      "loss": 1.5854,
+      "step": 151
+    },
+    {
+      "epoch": 0.14795587280986372,
+      "grad_norm": 0.5,
+      "learning_rate": 4.973669264827545e-06,
+      "loss": 1.6067,
+      "step": 152
+    },
+    {
+      "epoch": 0.14892926670992862,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.973297767485695e-06,
+      "loss": 1.5998,
+      "step": 153
+    },
+    {
+      "epoch": 0.14990266060999352,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.972923681832714e-06,
+      "loss": 1.5724,
+      "step": 154
+    },
+    {
+      "epoch": 0.1508760545100584,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.972547008260083e-06,
+      "loss": 1.6057,
+      "step": 155
+    },
+    {
+      "epoch": 0.1518494484101233,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.97216774716199e-06,
+      "loss": 1.5807,
+      "step": 156
+    },
+    {
+      "epoch": 0.1528228423101882,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.971785898935335e-06,
+      "loss": 1.5693,
+      "step": 157
+    },
+    {
+      "epoch": 0.1537962362102531,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.971401463979722e-06,
+      "loss": 1.5948,
+      "step": 158
+    },
+    {
+      "epoch": 0.15476963011031797,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.971014442697463e-06,
+      "loss": 1.5749,
+      "step": 159
+    },
+    {
+      "epoch": 0.15574302401038287,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.970624835493576e-06,
+      "loss": 1.5615,
+      "step": 160
+    },
+    {
+      "epoch": 0.15671641791044777,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.970232642775786e-06,
+      "loss": 1.5725,
+      "step": 161
+    },
+    {
+      "epoch": 0.15768981181051264,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.969837864954524e-06,
+      "loss": 1.5923,
+      "step": 162
+    },
+    {
+      "epoch": 0.15866320571057754,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.969440502442926e-06,
+      "loss": 1.5681,
+      "step": 163
+    },
+    {
+      "epoch": 0.15963659961064244,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.969040555656831e-06,
+      "loss": 1.5895,
+      "step": 164
+    },
+    {
+      "epoch": 0.16060999351070734,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.968638025014786e-06,
+      "loss": 1.5866,
+      "step": 165
+    },
+    {
+      "epoch": 0.16158338741077222,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.96823291093804e-06,
+      "loss": 1.566,
+      "step": 166
+    },
+    {
+      "epoch": 0.16255678131083712,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.967825213850545e-06,
+      "loss": 1.5402,
+      "step": 167
+    },
+    {
+      "epoch": 0.16353017521090202,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.9674149341789554e-06,
+      "loss": 1.5918,
+      "step": 168
+    },
+    {
+      "epoch": 0.16450356911096692,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.9670020723526325e-06,
+      "loss": 1.5902,
+      "step": 169
+    },
+    {
+      "epoch": 0.1654769630110318,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.9665866288036354e-06,
+      "loss": 1.5843,
+      "step": 170
+    },
+    {
+      "epoch": 0.1664503569110967,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.966168603966727e-06,
+      "loss": 1.5759,
+      "step": 171
+    },
+    {
+      "epoch": 0.1674237508111616,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.965747998279371e-06,
+      "loss": 1.5591,
+      "step": 172
+    },
+    {
+      "epoch": 0.16839714471122647,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.965324812181733e-06,
+      "loss": 1.5646,
+      "step": 173
+    },
+    {
+      "epoch": 0.16937053861129137,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.964899046116677e-06,
+      "loss": 1.5582,
+      "step": 174
+    },
+    {
+      "epoch": 0.17034393251135627,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.964470700529769e-06,
+      "loss": 1.579,
+      "step": 175
+    },
+    {
+      "epoch": 0.17131732641142117,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.964039775869271e-06,
+      "loss": 1.5866,
+      "step": 176
+    },
+    {
+      "epoch": 0.17229072031148604,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.96360627258615e-06,
+      "loss": 1.5627,
+      "step": 177
+    },
+    {
+      "epoch": 0.17326411421155094,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.963170191134067e-06,
+      "loss": 1.5562,
+      "step": 178
+    },
+    {
+      "epoch": 0.17423750811161584,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.96273153196938e-06,
+      "loss": 1.5689,
+      "step": 179
+    },
+    {
+      "epoch": 0.1752109020116807,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.96229029555115e-06,
+      "loss": 1.5693,
+      "step": 180
+    },
+    {
+      "epoch": 0.17618429591174561,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.9618464823411285e-06,
+      "loss": 1.576,
+      "step": 181
+    },
+    {
+      "epoch": 0.17715768981181051,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.9614000928037694e-06,
+      "loss": 1.539,
+      "step": 182
+    },
+    {
+      "epoch": 0.17813108371187542,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.96095112740622e-06,
+      "loss": 1.58,
+      "step": 183
+    },
+    {
+      "epoch": 0.1791044776119403,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.960499586618322e-06,
+      "loss": 1.5663,
+      "step": 184
+    },
+    {
+      "epoch": 0.1800778715120052,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.960045470912615e-06,
+      "loss": 1.5625,
+      "step": 185
+    },
+    {
+      "epoch": 0.1810512654120701,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.959588780764333e-06,
+      "loss": 1.5449,
+      "step": 186
+    },
+    {
+      "epoch": 0.182024659312135,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.9591295166514e-06,
+      "loss": 1.5538,
+      "step": 187
+    },
+    {
+      "epoch": 0.18299805321219986,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.9586676790544395e-06,
+      "loss": 1.5541,
+      "step": 188
+    },
+    {
+      "epoch": 0.18397144711226476,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.958203268456765e-06,
+      "loss": 1.5581,
+      "step": 189
+    },
+    {
+      "epoch": 0.18494484101232966,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.957736285344383e-06,
+      "loss": 1.5518,
+      "step": 190
+    },
+    {
+      "epoch": 0.18591823491239454,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.957266730205991e-06,
+      "loss": 1.5802,
+      "step": 191
+    },
+    {
+      "epoch": 0.18689162881245944,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.956794603532981e-06,
+      "loss": 1.5601,
+      "step": 192
+    },
+    {
+      "epoch": 0.18786502271252434,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.956319905819433e-06,
+      "loss": 1.5332,
+      "step": 193
+    },
+    {
+      "epoch": 0.18883841661258924,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.955842637562121e-06,
+      "loss": 1.5651,
+      "step": 194
+    },
+    {
+      "epoch": 0.1898118105126541,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.955362799260507e-06,
+      "loss": 1.564,
+      "step": 195
+    },
+    {
+      "epoch": 0.190785204412719,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.954880391416742e-06,
+      "loss": 1.5524,
+      "step": 196
+    },
+    {
+      "epoch": 0.1917585983127839,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.954395414535666e-06,
+      "loss": 1.5427,
+      "step": 197
+    },
+    {
+      "epoch": 0.1927319922128488,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.95390786912481e-06,
+      "loss": 1.5694,
+      "step": 198
+    },
+    {
+      "epoch": 0.19370538611291369,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.95341775569439e-06,
+      "loss": 1.5655,
+      "step": 199
+    },
+    {
+      "epoch": 0.1946787800129786,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.952925074757311e-06,
+      "loss": 1.5213,
+      "step": 200
+    },
+    {
+      "epoch": 0.1956521739130435,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.9524298268291646e-06,
+      "loss": 1.552,
+      "step": 201
+    },
+    {
+      "epoch": 0.19662556781310836,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.951932012428229e-06,
+      "loss": 1.5534,
+      "step": 202
+    },
+    {
+      "epoch": 0.19759896171317326,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.951431632075468e-06,
+      "loss": 1.5501,
+      "step": 203
+    },
+    {
+      "epoch": 0.19857235561323816,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.950928686294531e-06,
+      "loss": 1.5389,
+      "step": 204
+    },
+    {
+      "epoch": 0.19954574951330306,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.950423175611751e-06,
+      "loss": 1.561,
+      "step": 205
+    },
+    {
+      "epoch": 0.20051914341336793,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.949915100556146e-06,
+      "loss": 1.5662,
+      "step": 206
+    },
+    {
+      "epoch": 0.20149253731343283,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.949404461659417e-06,
+      "loss": 1.5414,
+      "step": 207
+    },
+    {
+      "epoch": 0.20246593121349774,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.948891259455951e-06,
+      "loss": 1.5569,
+      "step": 208
+    },
+    {
+      "epoch": 0.2034393251135626,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.948375494482813e-06,
+      "loss": 1.5561,
+      "step": 209
+    },
+    {
+      "epoch": 0.2044127190136275,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.947857167279753e-06,
+      "loss": 1.5661,
+      "step": 210
+    },
+    {
+      "epoch": 0.2053861129136924,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.947336278389201e-06,
+      "loss": 1.5579,
+      "step": 211
+    },
+    {
+      "epoch": 0.2063595068137573,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.946812828356268e-06,
+      "loss": 1.5325,
+      "step": 212
+    },
+    {
+      "epoch": 0.20733290071382218,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.946286817728746e-06,
+      "loss": 1.5208,
+      "step": 213
+    },
+    {
+      "epoch": 0.20830629461388708,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.945758247057107e-06,
+      "loss": 1.5509,
+      "step": 214
+    },
+    {
+      "epoch": 0.20927968851395198,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.9452271168945e-06,
+      "loss": 1.5414,
+      "step": 215
+    },
+    {
+      "epoch": 0.21025308241401688,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.944693427796754e-06,
+      "loss": 1.5746,
+      "step": 216
+    },
+    {
+      "epoch": 0.21122647631408176,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.944157180322377e-06,
+      "loss": 1.5315,
+      "step": 217
+    },
+    {
+      "epoch": 0.21219987021414666,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.9436183750325505e-06,
+      "loss": 1.5468,
+      "step": 218
+    },
+    {
+      "epoch": 0.21317326411421156,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.943077012491138e-06,
+      "loss": 1.5213,
+      "step": 219
+    },
+    {
+      "epoch": 0.21414665801427643,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.942533093264675e-06,
+      "loss": 1.547,
+      "step": 220
+    },
+    {
+      "epoch": 0.21512005191434133,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.941986617922374e-06,
+      "loss": 1.5467,
+      "step": 221
+    },
+    {
+      "epoch": 0.21609344581440623,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.941437587036123e-06,
+      "loss": 1.5447,
+      "step": 222
+    },
+    {
+      "epoch": 0.21706683971447113,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.940886001180485e-06,
+      "loss": 1.5026,
+      "step": 223
+    },
+    {
+      "epoch": 0.218040233614536,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.940331860932694e-06,
+      "loss": 1.5049,
+      "step": 224
+    },
+    {
+      "epoch": 0.2190136275146009,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.9397751668726595e-06,
+      "loss": 1.5445,
+      "step": 225
+    },
+    {
+      "epoch": 0.2199870214146658,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.939215919582963e-06,
+      "loss": 1.5168,
+      "step": 226
+    },
+    {
+      "epoch": 0.2209604153147307,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.938654119648858e-06,
+      "loss": 1.5318,
+      "step": 227
+    },
+    {
+      "epoch": 0.22193380921479558,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.938089767658269e-06,
+      "loss": 1.4995,
+      "step": 228
+    },
+    {
+      "epoch": 0.22290720311486048,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.937522864201792e-06,
+      "loss": 1.513,
+      "step": 229
+    },
+    {
+      "epoch": 0.22388059701492538,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.936953409872692e-06,
+      "loss": 1.549,
+      "step": 230
+    },
+    {
+      "epoch": 0.22485399091499025,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.936381405266904e-06,
+      "loss": 1.5253,
+      "step": 231
+    },
+    {
+      "epoch": 0.22582738481505515,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.935806850983034e-06,
+      "loss": 1.536,
+      "step": 232
+    },
+    {
+      "epoch": 0.22680077871512005,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.935229747622352e-06,
+      "loss": 1.5603,
+      "step": 233
+    },
+    {
+      "epoch": 0.22777417261518496,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.934650095788798e-06,
+      "loss": 1.538,
+      "step": 234
+    },
+    {
+      "epoch": 0.22874756651524983,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.934067896088979e-06,
+      "loss": 1.5108,
+      "step": 235
+    },
+    {
+      "epoch": 0.22972096041531473,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.9334831491321685e-06,
+      "loss": 1.52,
+      "step": 236
+    },
+    {
+      "epoch": 0.23069435431537963,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.932895855530307e-06,
+      "loss": 1.5124,
+      "step": 237
+    },
+    {
+      "epoch": 0.23166774821544453,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.932306015897995e-06,
+      "loss": 1.5195,
+      "step": 238
+    },
+    {
+      "epoch": 0.2326411421155094,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.9317136308525025e-06,
+      "loss": 1.5301,
+      "step": 239
+    },
+    {
+      "epoch": 0.2336145360155743,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.931118701013763e-06,
+      "loss": 1.518,
+      "step": 240
+    },
+    {
+      "epoch": 0.2345879299156392,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.93052122700437e-06,
+      "loss": 1.5004,
+      "step": 241
+    },
+    {
+      "epoch": 0.23556132381570408,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.9299212094495816e-06,
+      "loss": 1.5155,
+      "step": 242
+    },
+    {
+      "epoch": 0.23653471771576898,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.929318648977318e-06,
+      "loss": 1.5499,
+      "step": 243
+    },
+    {
+      "epoch": 0.23750811161583388,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.92871354621816e-06,
+      "loss": 1.5221,
+      "step": 244
+    },
+    {
+      "epoch": 0.23848150551589878,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.9281059018053475e-06,
+      "loss": 1.535,
+      "step": 245
+    },
+    {
+      "epoch": 0.23945489941596365,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.927495716374783e-06,
+      "loss": 1.512,
+      "step": 246
+    },
+    {
+      "epoch": 0.24042829331602855,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.9268829905650274e-06,
+      "loss": 1.5274,
+      "step": 247
+    },
+    {
+      "epoch": 0.24140168721609345,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.926267725017297e-06,
+      "loss": 1.5027,
+      "step": 248
+    },
+    {
+      "epoch": 0.24237508111615833,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.925649920375471e-06,
+      "loss": 1.5368,
+      "step": 249
+    },
+    {
+      "epoch": 0.24334847501622323,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.9250295772860815e-06,
+      "loss": 1.522,
+      "step": 250
+    },
+    {
+      "epoch": 0.24432186891628813,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.924406696398319e-06,
+      "loss": 1.5478,
+      "step": 251
+    },
+    {
+      "epoch": 0.24529526281635303,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.9237812783640304e-06,
+      "loss": 1.5049,
+      "step": 252
+    },
+    {
+      "epoch": 0.2462686567164179,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.923153323837717e-06,
+      "loss": 1.5182,
+      "step": 253
+    },
+    {
+      "epoch": 0.2472420506164828,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.922522833476533e-06,
+      "loss": 1.5137,
+      "step": 254
+    },
+    {
+      "epoch": 0.2482154445165477,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.92188980794029e-06,
+      "loss": 1.5455,
+      "step": 255
+    },
+    {
+      "epoch": 0.2491888384166126,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.921254247891449e-06,
+      "loss": 1.501,
+      "step": 256
+    },
+    {
+      "epoch": 0.2501622323166775,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.920616153995126e-06,
+      "loss": 1.5195,
+      "step": 257
+    },
+    {
+      "epoch": 0.2511356262167424,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.9199755269190865e-06,
+      "loss": 1.5014,
+      "step": 258
+    },
+    {
+      "epoch": 0.25210902011680725,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.919332367333748e-06,
+      "loss": 1.4972,
+      "step": 259
+    },
+    {
+      "epoch": 0.25308241401687215,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.918686675912178e-06,
+      "loss": 1.5035,
+      "step": 260
+    },
+    {
+      "epoch": 0.25405580791693705,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.918038453330095e-06,
+      "loss": 1.5082,
+      "step": 261
+    },
+    {
+      "epoch": 0.25502920181700195,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.917387700265866e-06,
+      "loss": 1.518,
+      "step": 262
+    },
+    {
+      "epoch": 0.25600259571706685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.916734417400503e-06,
+      "loss": 1.4958,
+      "step": 263
+    },
+    {
+      "epoch": 0.25697598961713175,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.91607860541767e-06,
+      "loss": 1.5237,
+      "step": 264
+    },
+    {
+      "epoch": 0.25794938351719665,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.915420265003674e-06,
+      "loss": 1.4898,
+      "step": 265
+    },
+    {
+      "epoch": 0.2589227774172615,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.9147593968474705e-06,
+      "loss": 1.4946,
+      "step": 266
+    },
+    {
+      "epoch": 0.2598961713173264,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.914096001640659e-06,
+      "loss": 1.4958,
+      "step": 267
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.913430080077483e-06,
+      "loss": 1.5221,
+      "step": 268
+    },
+    {
+      "epoch": 0.2618429591174562,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.912761632854834e-06,
+      "loss": 1.4937,
+      "step": 269
+    },
+    {
+      "epoch": 0.2628163530175211,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.91209066067224e-06,
+      "loss": 1.5216,
+      "step": 270
+    },
+    {
+      "epoch": 0.263789746917586,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.911417164231875e-06,
+      "loss": 1.5069,
+      "step": 271
+    },
+    {
+      "epoch": 0.2647631408176509,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.910741144238556e-06,
+      "loss": 1.5145,
+      "step": 272
+    },
+    {
+      "epoch": 0.26573653471771574,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.910062601399739e-06,
+      "loss": 1.5049,
+      "step": 273
+    },
+    {
+      "epoch": 0.26670992861778064,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.9093815364255204e-06,
+      "loss": 1.5153,
+      "step": 274
+    },
+    {
+      "epoch": 0.26768332251784555,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.9086979500286345e-06,
+      "loss": 1.5106,
+      "step": 275
+    },
+    {
+      "epoch": 0.26865671641791045,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.908011842924458e-06,
+      "loss": 1.5176,
+      "step": 276
+    },
+    {
+      "epoch": 0.26963011031797535,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.9073232158310025e-06,
+      "loss": 1.4883,
+      "step": 277
+    },
+    {
+      "epoch": 0.27060350421804025,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.906632069468917e-06,
+      "loss": 1.5318,
+      "step": 278
+    },
+    {
+      "epoch": 0.27157689811810515,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.905938404561489e-06,
+      "loss": 1.518,
+      "step": 279
+    },
+    {
+      "epoch": 0.27255029201817,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.905242221834638e-06,
+      "loss": 1.4833,
+      "step": 280
+    },
+    {
+      "epoch": 0.2735236859182349,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.904543522016923e-06,
+      "loss": 1.522,
+      "step": 281
+    },
+    {
+      "epoch": 0.2744970798182998,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.903842305839534e-06,
+      "loss": 1.4828,
+      "step": 282
+    },
+    {
+      "epoch": 0.2754704737183647,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.903138574036295e-06,
+      "loss": 1.5441,
+      "step": 283
+    },
+    {
+      "epoch": 0.2764438676184296,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.902432327343662e-06,
+      "loss": 1.4891,
+      "step": 284
+    },
+    {
+      "epoch": 0.2774172615184945,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.901723566500725e-06,
+      "loss": 1.4813,
+      "step": 285
+    },
+    {
+      "epoch": 0.2783906554185594,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.901012292249203e-06,
+      "loss": 1.4753,
+      "step": 286
+    },
+    {
+      "epoch": 0.2793640493186243,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.900298505333446e-06,
+      "loss": 1.4873,
+      "step": 287
+    },
+    {
+      "epoch": 0.28033744321868914,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.899582206500433e-06,
+      "loss": 1.4863,
+      "step": 288
+    },
+    {
+      "epoch": 0.28131083711875404,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.898863396499772e-06,
+      "loss": 1.5016,
+      "step": 289
+    },
+    {
+      "epoch": 0.28228423101881894,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.898142076083701e-06,
+      "loss": 1.4733,
+      "step": 290
+    },
+    {
+      "epoch": 0.28325762491888384,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.8974182460070814e-06,
+      "loss": 1.5243,
+      "step": 291
+    },
+    {
+      "epoch": 0.28423101881894874,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.896691907027404e-06,
+      "loss": 1.489,
+      "step": 292
+    },
+    {
+      "epoch": 0.28520441271901364,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.895963059904782e-06,
+      "loss": 1.4813,
+      "step": 293
+    },
+    {
+      "epoch": 0.28617780661907855,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.895231705401958e-06,
+      "loss": 1.4781,
+      "step": 294
+    },
+    {
+      "epoch": 0.2871512005191434,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.8944978442842944e-06,
+      "loss": 1.4688,
+      "step": 295
+    },
+    {
+      "epoch": 0.2881245944192083,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.893761477319779e-06,
+      "loss": 1.4892,
+      "step": 296
+    },
+    {
+      "epoch": 0.2890979883192732,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.8930226052790204e-06,
+      "loss": 1.5027,
+      "step": 297
+    },
+    {
+      "epoch": 0.2900713822193381,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.892281228935252e-06,
+      "loss": 1.502,
+      "step": 298
+    },
+    {
+      "epoch": 0.291044776119403,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.891537349064322e-06,
+      "loss": 1.4795,
+      "step": 299
+    },
+    {
+      "epoch": 0.2920181700194679,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.890790966444705e-06,
+      "loss": 1.4875,
+      "step": 300
+    },
+    {
+      "epoch": 0.2929915639195328,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.8900420818574915e-06,
+      "loss": 1.4629,
+      "step": 301
+    },
+    {
+      "epoch": 0.29396495781959764,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.889290696086391e-06,
+      "loss": 1.4935,
+      "step": 302
+    },
+    {
+      "epoch": 0.29493835171966254,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.888536809917728e-06,
+      "loss": 1.47,
+      "step": 303
+    },
+    {
+      "epoch": 0.29591174561972744,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.887780424140448e-06,
+      "loss": 1.4882,
+      "step": 304
+    },
+    {
+      "epoch": 0.29688513951979234,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.88702153954611e-06,
+      "loss": 1.4968,
+      "step": 305
+    },
+    {
+      "epoch": 0.29785853341985724,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.8862601569288885e-06,
+      "loss": 1.4711,
+      "step": 306
+    },
+    {
+      "epoch": 0.29883192731992214,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.885496277085571e-06,
+      "loss": 1.4824,
+      "step": 307
+    },
+    {
+      "epoch": 0.29980532121998704,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.884729900815559e-06,
+      "loss": 1.4758,
+      "step": 308
+    },
+    {
+      "epoch": 0.3007787151200519,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.8839610289208695e-06,
+      "loss": 1.4932,
+      "step": 309
+    },
+    {
+      "epoch": 0.3017521090201168,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.8831896622061256e-06,
+      "loss": 1.4829,
+      "step": 310
+    },
+    {
+      "epoch": 0.3027255029201817,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.882415801478565e-06,
+      "loss": 1.4996,
+      "step": 311
+    },
+    {
+      "epoch": 0.3036988968202466,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.881639447548034e-06,
+      "loss": 1.4822,
+      "step": 312
+    },
+    {
+      "epoch": 0.3046722907203115,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.88086060122699e-06,
+      "loss": 1.485,
+      "step": 313
+    },
+    {
+      "epoch": 0.3056456846203764,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.880079263330497e-06,
+      "loss": 1.4654,
+      "step": 314
+    },
+    {
+      "epoch": 0.3066190785204413,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.8792954346762256e-06,
+      "loss": 1.4699,
+      "step": 315
+    },
+    {
+      "epoch": 0.3075924724205062,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.878509116084455e-06,
+      "loss": 1.4921,
+      "step": 316
+    },
+    {
+      "epoch": 0.30856586632057104,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.8777203083780675e-06,
+      "loss": 1.4881,
+      "step": 317
+    },
+    {
+      "epoch": 0.30953926022063594,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.876929012382555e-06,
+      "loss": 1.4741,
+      "step": 318
+    },
+    {
+      "epoch": 0.31051265412070084,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.876135228926008e-06,
+      "loss": 1.475,
+      "step": 319
+    },
+    {
+      "epoch": 0.31148604802076574,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.875338958839123e-06,
+      "loss": 1.5087,
+      "step": 320
+    },
+    {
+      "epoch": 0.31245944192083064,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.8745402029551995e-06,
+      "loss": 1.4926,
+      "step": 321
+    },
+    {
+      "epoch": 0.31343283582089554,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.873738962110135e-06,
+      "loss": 1.4801,
+      "step": 322
+    },
+    {
+      "epoch": 0.31440622972096044,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.872935237142431e-06,
+      "loss": 1.4819,
+      "step": 323
+    },
+    {
+      "epoch": 0.3153796236210253,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.872129028893186e-06,
+      "loss": 1.4921,
+      "step": 324
+    },
+    {
+      "epoch": 0.3163530175210902,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.871320338206101e-06,
+      "loss": 1.4897,
+      "step": 325
+    },
+    {
+      "epoch": 0.3173264114211551,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.870509165927471e-06,
+      "loss": 1.4856,
+      "step": 326
+    },
+    {
+      "epoch": 0.31829980532122,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.86969551290619e-06,
+      "loss": 1.4761,
+      "step": 327
+    },
+    {
+      "epoch": 0.3192731992212849,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.868879379993746e-06,
+      "loss": 1.5015,
+      "step": 328
+    },
+    {
+      "epoch": 0.3202465931213498,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.868060768044225e-06,
+      "loss": 1.4904,
+      "step": 329
+    },
+    {
+      "epoch": 0.3212199870214147,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.867239677914306e-06,
+      "loss": 1.4805,
+      "step": 330
+    },
+    {
+      "epoch": 0.32219338092147953,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.866416110463261e-06,
+      "loss": 1.4899,
+      "step": 331
+    },
+    {
+      "epoch": 0.32316677482154443,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.8655900665529565e-06,
+      "loss": 1.4891,
+      "step": 332
+    },
+    {
+      "epoch": 0.32414016872160933,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.864761547047847e-06,
+      "loss": 1.4681,
+      "step": 333
+    },
+    {
+      "epoch": 0.32511356262167423,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.863930552814981e-06,
+      "loss": 1.493,
+      "step": 334
+    },
+    {
+      "epoch": 0.32608695652173914,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.863097084723996e-06,
+      "loss": 1.4591,
+      "step": 335
+    },
+    {
+      "epoch": 0.32706035042180404,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.862261143647117e-06,
+      "loss": 1.4805,
+      "step": 336
+    },
+    {
+      "epoch": 0.32803374432186894,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.861422730459159e-06,
+      "loss": 1.4848,
+      "step": 337
+    },
+    {
+      "epoch": 0.32900713822193384,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.860581846037522e-06,
+      "loss": 1.4645,
+      "step": 338
+    },
+    {
+      "epoch": 0.3299805321219987,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.859738491262195e-06,
+      "loss": 1.4922,
+      "step": 339
+    },
+    {
+      "epoch": 0.3309539260220636,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.858892667015749e-06,
+      "loss": 1.4678,
+      "step": 340
+    },
+    {
+      "epoch": 0.3319273199221285,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.8580443741833404e-06,
+      "loss": 1.4982,
+      "step": 341
+    },
+    {
+      "epoch": 0.3329007138221934,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.857193613652711e-06,
+      "loss": 1.4497,
+      "step": 342
+    },
+    {
+      "epoch": 0.3338741077222583,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.8563403863141825e-06,
+      "loss": 1.4563,
+      "step": 343
+    },
+    {
+      "epoch": 0.3348475016223232,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.855484693060658e-06,
+      "loss": 1.4787,
+      "step": 344
+    },
+    {
+      "epoch": 0.3358208955223881,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.854626534787625e-06,
+      "loss": 1.493,
+      "step": 345
+    },
+    {
+      "epoch": 0.33679428942245293,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.853765912393146e-06,
+      "loss": 1.4569,
+      "step": 346
+    },
+    {
+      "epoch": 0.33776768332251783,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.852902826777863e-06,
+      "loss": 1.4463,
+      "step": 347
+    },
+    {
+      "epoch": 0.33874107722258273,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.852037278845e-06,
+      "loss": 1.4696,
+      "step": 348
+    },
+    {
+      "epoch": 0.33971447112264763,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.851169269500351e-06,
+      "loss": 1.4719,
+      "step": 349
+    },
+    {
+      "epoch": 0.34068786502271253,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.850298799652293e-06,
+      "loss": 1.4561,
+      "step": 350
+    },
+    {
+      "epoch": 0.34166125892277743,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.8494258702117715e-06,
+      "loss": 1.4814,
+      "step": 351
+    },
+    {
+      "epoch": 0.34263465282284233,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.8485504820923115e-06,
+      "loss": 1.4758,
+      "step": 352
+    },
+    {
+      "epoch": 0.3436080467229072,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.847672636210005e-06,
+      "loss": 1.4657,
+      "step": 353
+    },
+    {
+      "epoch": 0.3445814406229721,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.8467923334835245e-06,
+      "loss": 1.4647,
+      "step": 354
+    },
+    {
+      "epoch": 0.345554834523037,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.8459095748341045e-06,
+      "loss": 1.4916,
+      "step": 355
+    },
+    {
+      "epoch": 0.3465282284231019,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.845024361185555e-06,
+      "loss": 1.4676,
+      "step": 356
+    },
+    {
+      "epoch": 0.3475016223231668,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.8441366934642545e-06,
+      "loss": 1.4645,
+      "step": 357
+    },
+    {
+      "epoch": 0.3484750162232317,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.8432465725991475e-06,
+      "loss": 1.4519,
+      "step": 358
+    },
+    {
+      "epoch": 0.3494484101232966,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.842353999521749e-06,
+      "loss": 1.4563,
+      "step": 359
+    },
+    {
+      "epoch": 0.3504218040233614,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.841458975166137e-06,
+      "loss": 1.4553,
+      "step": 360
+    },
+    {
+      "epoch": 0.35139519792342633,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.840561500468958e-06,
+      "loss": 1.4512,
+      "step": 361
+    },
+    {
+      "epoch": 0.35236859182349123,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.839661576369419e-06,
+      "loss": 1.4644,
+      "step": 362
+    },
+    {
+      "epoch": 0.35334198572355613,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.838759203809295e-06,
+      "loss": 1.466,
+      "step": 363
+    },
+    {
+      "epoch": 0.35431537962362103,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.837854383732918e-06,
+      "loss": 1.4882,
+      "step": 364
+    },
+    {
+      "epoch": 0.35528877352368593,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.836947117087186e-06,
+      "loss": 1.4511,
+      "step": 365
+    },
+    {
+      "epoch": 0.35626216742375083,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.836037404821554e-06,
+      "loss": 1.4536,
+      "step": 366
+    },
+    {
+      "epoch": 0.35723556132381573,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.83512524788804e-06,
+      "loss": 1.4477,
+      "step": 367
+    },
+    {
+      "epoch": 0.3582089552238806,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.834210647241215e-06,
+      "loss": 1.4668,
+      "step": 368
+    },
+    {
+      "epoch": 0.3591823491239455,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.8332936038382125e-06,
+      "loss": 1.5002,
+      "step": 369
+    },
+    {
+      "epoch": 0.3601557430240104,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.8323741186387205e-06,
+      "loss": 1.4618,
+      "step": 370
+    },
+    {
+      "epoch": 0.3611291369240753,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.831452192604981e-06,
+      "loss": 1.4647,
+      "step": 371
+    },
+    {
+      "epoch": 0.3621025308241402,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.830527826701791e-06,
+      "loss": 1.4363,
+      "step": 372
+    },
+    {
+      "epoch": 0.3630759247242051,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.829601021896503e-06,
+      "loss": 1.4425,
+      "step": 373
+    },
+    {
+      "epoch": 0.36404931862427,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.828671779159019e-06,
+      "loss": 1.4738,
+      "step": 374
+    },
+    {
+      "epoch": 0.3650227125243348,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.827740099461793e-06,
+      "loss": 1.4419,
+      "step": 375
+    },
+    {
+      "epoch": 0.3659961064243997,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.826805983779831e-06,
+      "loss": 1.4411,
+      "step": 376
+    },
+    {
+      "epoch": 0.3669695003244646,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.825869433090686e-06,
+      "loss": 1.442,
+      "step": 377
+    },
+    {
+      "epoch": 0.3679428942245295,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.824930448374462e-06,
+      "loss": 1.4477,
+      "step": 378
+    },
+    {
+      "epoch": 0.3689162881245944,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.823989030613805e-06,
+      "loss": 1.4556,
+      "step": 379
+    },
+    {
+      "epoch": 0.3698896820246593,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.823045180793914e-06,
+      "loss": 1.4586,
+      "step": 380
+    },
+    {
+      "epoch": 0.37086307592472423,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.822098899902527e-06,
+      "loss": 1.463,
+      "step": 381
+    },
+    {
+      "epoch": 0.3718364698247891,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.82115018892993e-06,
+      "loss": 1.4751,
+      "step": 382
+    },
+    {
+      "epoch": 0.372809863724854,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.8201990488689524e-06,
+      "loss": 1.4391,
+      "step": 383
+    },
+    {
+      "epoch": 0.3737832576249189,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.819245480714962e-06,
+      "loss": 1.4352,
+      "step": 384
+    },
+    {
+      "epoch": 0.3747566515249838,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.8182894854658715e-06,
+      "loss": 1.4346,
+      "step": 385
+    },
+    {
+      "epoch": 0.3757300454250487,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.81733106412213e-06,
+      "loss": 1.4661,
+      "step": 386
+    },
+    {
+      "epoch": 0.3767034393251136,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.816370217686729e-06,
+      "loss": 1.4604,
+      "step": 387
+    },
+    {
+      "epoch": 0.3776768332251785,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.8154069471651956e-06,
+      "loss": 1.443,
+      "step": 388
+    },
+    {
+      "epoch": 0.3786502271252433,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.814441253565594e-06,
+      "loss": 1.4584,
+      "step": 389
+    },
+    {
+      "epoch": 0.3796236210253082,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.813473137898526e-06,
+      "loss": 1.4461,
+      "step": 390
+    },
+    {
+      "epoch": 0.3805970149253731,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.8125026011771244e-06,
+      "loss": 1.4514,
+      "step": 391
+    },
+    {
+      "epoch": 0.381570408825438,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.81152964441706e-06,
+      "loss": 1.4393,
+      "step": 392
+    },
+    {
+      "epoch": 0.3825438027255029,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.810554268636532e-06,
+      "loss": 1.4668,
+      "step": 393
+    },
+    {
+      "epoch": 0.3835171966255678,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.8095764748562754e-06,
+      "loss": 1.4354,
+      "step": 394
+    },
+    {
+      "epoch": 0.3844905905256327,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.808596264099552e-06,
+      "loss": 1.4275,
+      "step": 395
+    },
+    {
+      "epoch": 0.3854639844256976,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.807613637392153e-06,
+      "loss": 1.4668,
+      "step": 396
+    },
+    {
+      "epoch": 0.38643737832576247,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.806628595762403e-06,
+      "loss": 1.4209,
+      "step": 397
+    },
+    {
+      "epoch": 0.38741077222582737,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.805641140241146e-06,
+      "loss": 1.4311,
+      "step": 398
+    },
+    {
+      "epoch": 0.38838416612589227,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.8046512718617585e-06,
+      "loss": 1.4645,
+      "step": 399
+    },
+    {
+      "epoch": 0.3893575600259572,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.80365899166014e-06,
+      "loss": 1.4332,
+      "step": 400
+    },
+    {
+      "epoch": 0.3903309539260221,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.802664300674712e-06,
+      "loss": 1.4599,
+      "step": 401
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.801667199946422e-06,
+      "loss": 1.45,
+      "step": 402
+    },
+    {
+      "epoch": 0.3922777417261519,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.800667690518737e-06,
+      "loss": 1.4397,
+      "step": 403
+    },
+    {
+      "epoch": 0.3932511356262167,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.799665773437648e-06,
+      "loss": 1.4442,
+      "step": 404
+    },
+    {
+      "epoch": 0.3942245295262816,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.79866144975166e-06,
+      "loss": 1.4539,
+      "step": 405
+    },
+    {
+      "epoch": 0.3951979234263465,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.797654720511802e-06,
+      "loss": 1.4537,
+      "step": 406
+    },
+    {
+      "epoch": 0.3961713173264114,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.7966455867716165e-06,
+      "loss": 1.4368,
+      "step": 407
+    },
+    {
+      "epoch": 0.3971447112264763,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.795634049587165e-06,
+      "loss": 1.4435,
+      "step": 408
+    },
+    {
+      "epoch": 0.3981181051265412,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.794620110017025e-06,
+      "loss": 1.4547,
+      "step": 409
+    },
+    {
+      "epoch": 0.3990914990266061,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.793603769122283e-06,
+      "loss": 1.4532,
+      "step": 410
+    },
+    {
+      "epoch": 0.40006489292667097,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.792585027966544e-06,
+      "loss": 1.4506,
+      "step": 411
+    },
+    {
+      "epoch": 0.40103828682673587,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.791563887615921e-06,
+      "loss": 1.4642,
+      "step": 412
+    },
+    {
+      "epoch": 0.40201168072680077,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.790540349139041e-06,
+      "loss": 1.4293,
+      "step": 413
+    },
+    {
+      "epoch": 0.40298507462686567,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.789514413607039e-06,
+      "loss": 1.4423,
+      "step": 414
+    },
+    {
+      "epoch": 0.40395846852693057,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.7884860820935574e-06,
+      "loss": 1.4504,
+      "step": 415
+    },
+    {
+      "epoch": 0.40493186242699547,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.787455355674748e-06,
+      "loss": 1.4428,
+      "step": 416
+    },
+    {
+      "epoch": 0.40590525632706037,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.786422235429269e-06,
+      "loss": 1.4195,
+      "step": 417
+    },
+    {
+      "epoch": 0.4068786502271252,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.785386722438281e-06,
+      "loss": 1.4364,
+      "step": 418
+    },
+    {
+      "epoch": 0.4078520441271901,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.7843488177854516e-06,
+      "loss": 1.4277,
+      "step": 419
+    },
+    {
+      "epoch": 0.408825438027255,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.783308522556949e-06,
+      "loss": 1.452,
+      "step": 420
+    },
+    {
+      "epoch": 0.4097988319273199,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.782265837841446e-06,
+      "loss": 1.4367,
+      "step": 421
+    },
+    {
+      "epoch": 0.4107722258273848,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.781220764730113e-06,
+      "loss": 1.4328,
+      "step": 422
+    },
+    {
+      "epoch": 0.4117456197274497,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.780173304316622e-06,
+      "loss": 1.439,
+      "step": 423
+    },
+    {
+      "epoch": 0.4127190136275146,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.779123457697142e-06,
+      "loss": 1.4275,
+      "step": 424
+    },
+    {
+      "epoch": 0.4136924075275795,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.77807122597034e-06,
+      "loss": 1.3996,
+      "step": 425
+    },
+    {
+      "epoch": 0.41466580142764436,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.777016610237377e-06,
+      "loss": 1.4428,
+      "step": 426
+    },
+    {
+      "epoch": 0.41563919532770927,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.7759596116019124e-06,
+      "loss": 1.4351,
+      "step": 427
+    },
+    {
+      "epoch": 0.41661258922777417,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.774900231170096e-06,
+      "loss": 1.4532,
+      "step": 428
+    },
+    {
+      "epoch": 0.41758598312783907,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.773838470050574e-06,
+      "loss": 1.4287,
+      "step": 429
+    },
+    {
+      "epoch": 0.41855937702790397,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.772774329354479e-06,
+      "loss": 1.4431,
+      "step": 430
+    },
+    {
+      "epoch": 0.41953277092796887,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.771707810195437e-06,
+      "loss": 1.436,
+      "step": 431
+    },
+    {
+      "epoch": 0.42050616482803377,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.770638913689563e-06,
+      "loss": 1.4336,
+      "step": 432
+    },
+    {
+      "epoch": 0.4214795587280986,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.7695676409554595e-06,
+      "loss": 1.436,
+      "step": 433
+    },
+    {
+      "epoch": 0.4224529526281635,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.768493993114215e-06,
+      "loss": 1.4055,
+      "step": 434
+    },
+    {
+      "epoch": 0.4234263465282284,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.767417971289403e-06,
+      "loss": 1.4437,
+      "step": 435
+    },
+    {
+      "epoch": 0.4243997404282933,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.7663395766070854e-06,
+      "loss": 1.4219,
+      "step": 436
+    },
+    {
+      "epoch": 0.4253731343283582,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.765258810195802e-06,
+      "loss": 1.4476,
+      "step": 437
+    },
+    {
+      "epoch": 0.4263465282284231,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.764175673186579e-06,
+      "loss": 1.4434,
+      "step": 438
+    },
+    {
+      "epoch": 0.427319922128488,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.763090166712919e-06,
+      "loss": 1.4167,
+      "step": 439
+    },
+    {
+      "epoch": 0.42829331602855286,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.76200229191081e-06,
+      "loss": 1.4278,
+      "step": 440
+    },
+    {
+      "epoch": 0.42926670992861776,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.760912049918711e-06,
+      "loss": 1.4133,
+      "step": 441
+    },
+    {
+      "epoch": 0.43024010382868266,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.759819441877567e-06,
+      "loss": 1.4294,
+      "step": 442
+    },
+    {
+      "epoch": 0.43121349772874756,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.758724468930791e-06,
+      "loss": 1.4389,
+      "step": 443
+    },
+    {
+      "epoch": 0.43218689162881246,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.757627132224276e-06,
+      "loss": 1.4482,
+      "step": 444
+    },
+    {
+      "epoch": 0.43316028552887736,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.756527432906387e-06,
+      "loss": 1.4239,
+      "step": 445
+    },
+    {
+      "epoch": 0.43413367942894227,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.755425372127961e-06,
+      "loss": 1.4439,
+      "step": 446
+    },
+    {
+      "epoch": 0.43510707332900717,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.754320951042307e-06,
+      "loss": 1.4092,
+      "step": 447
+    },
+    {
+      "epoch": 0.436080467229072,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.753214170805205e-06,
+      "loss": 1.4312,
+      "step": 448
+    },
+    {
+      "epoch": 0.4370538611291369,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.752105032574902e-06,
+      "loss": 1.4296,
+      "step": 449
+    },
+    {
+      "epoch": 0.4380272550292018,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.7509935375121115e-06,
+      "loss": 1.4436,
+      "step": 450
+    },
+    {
+      "epoch": 0.4390006489292667,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.749879686780019e-06,
+      "loss": 1.4108,
+      "step": 451
+    },
+    {
+      "epoch": 0.4399740428293316,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.74876348154427e-06,
+      "loss": 1.4423,
+      "step": 452
+    },
+    {
+      "epoch": 0.4409474367293965,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.747644922972973e-06,
+      "loss": 1.4195,
+      "step": 453
+    },
+    {
+      "epoch": 0.4419208306294614,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.746524012236706e-06,
+      "loss": 1.4384,
+      "step": 454
+    },
+    {
+      "epoch": 0.44289422452952626,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.745400750508502e-06,
+      "loss": 1.4084,
+      "step": 455
+    },
+    {
+      "epoch": 0.44386761842959116,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.7442751389638564e-06,
+      "loss": 1.4368,
+      "step": 456
+    },
+    {
+      "epoch": 0.44484101232965606,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.743147178780725e-06,
+      "loss": 1.4065,
+      "step": 457
+    },
+    {
+      "epoch": 0.44581440622972096,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.7420168711395194e-06,
+      "loss": 1.4054,
+      "step": 458
+    },
+    {
+      "epoch": 0.44678780012978586,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.7408842172231115e-06,
+      "loss": 1.3868,
+      "step": 459
+    },
+    {
+      "epoch": 0.44776119402985076,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.739749218216823e-06,
+      "loss": 1.4552,
+      "step": 460
+    },
+    {
+      "epoch": 0.44873458792991566,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.7386118753084325e-06,
+      "loss": 1.4349,
+      "step": 461
+    },
+    {
+      "epoch": 0.4497079818299805,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.737472189688173e-06,
+      "loss": 1.4277,
+      "step": 462
+    },
+    {
+      "epoch": 0.4506813757300454,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.736330162548729e-06,
+      "loss": 1.4151,
+      "step": 463
+    },
+    {
+      "epoch": 0.4516547696301103,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.735185795085231e-06,
+      "loss": 1.4178,
+      "step": 464
+    },
+    {
+      "epoch": 0.4526281635301752,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.734039088495265e-06,
+      "loss": 1.4401,
+      "step": 465
+    },
+    {
+      "epoch": 0.4536015574302401,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.73289004397886e-06,
+      "loss": 1.416,
+      "step": 466
+    },
+    {
+      "epoch": 0.454574951330305,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.731738662738494e-06,
+      "loss": 1.4224,
+      "step": 467
+    },
+    {
+      "epoch": 0.4555483452303699,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.73058494597909e-06,
+      "loss": 1.4286,
+      "step": 468
+    },
+    {
+      "epoch": 0.45652173913043476,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.729428894908013e-06,
+      "loss": 1.4023,
+      "step": 469
+    },
+    {
+      "epoch": 0.45749513303049966,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.728270510735076e-06,
+      "loss": 1.4043,
+      "step": 470
+    },
+    {
+      "epoch": 0.45846852693056456,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.727109794672528e-06,
+      "loss": 1.4357,
+      "step": 471
+    },
+    {
+      "epoch": 0.45944192083062946,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.725946747935062e-06,
+      "loss": 1.4129,
+      "step": 472
+    },
+    {
+      "epoch": 0.46041531473069436,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.724781371739807e-06,
+      "loss": 1.4292,
+      "step": 473
+    },
+    {
+      "epoch": 0.46138870863075926,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.723613667306333e-06,
+      "loss": 1.4386,
+      "step": 474
+    },
+    {
+      "epoch": 0.46236210253082416,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.722443635856646e-06,
+      "loss": 1.4274,
+      "step": 475
+    },
+    {
+      "epoch": 0.46333549643088906,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.721271278615185e-06,
+      "loss": 1.4315,
+      "step": 476
+    },
+    {
+      "epoch": 0.4643088903309539,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.720096596808824e-06,
+      "loss": 1.4202,
+      "step": 477
+    },
+    {
+      "epoch": 0.4652822842310188,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.718919591666871e-06,
+      "loss": 1.42,
+      "step": 478
+    },
+    {
+      "epoch": 0.4662556781310837,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.717740264421063e-06,
+      "loss": 1.4341,
+      "step": 479
+    },
+    {
+      "epoch": 0.4672290720311486,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.716558616305568e-06,
+      "loss": 1.4016,
+      "step": 480
+    },
+    {
+      "epoch": 0.4682024659312135,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.715374648556985e-06,
+      "loss": 1.4382,
+      "step": 481
+    },
+    {
+      "epoch": 0.4691758598312784,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.714188362414337e-06,
+      "loss": 1.4271,
+      "step": 482
+    },
+    {
+      "epoch": 0.4701492537313433,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.712999759119076e-06,
+      "loss": 1.43,
+      "step": 483
+    },
+    {
+      "epoch": 0.47112264763140815,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.7118088399150776e-06,
+      "loss": 1.4367,
+      "step": 484
+    },
+    {
+      "epoch": 0.47209604153147305,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.71061560604864e-06,
+      "loss": 1.4129,
+      "step": 485
+    },
+    {
+      "epoch": 0.47306943543153795,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.709420058768487e-06,
+      "loss": 1.3979,
+      "step": 486
+    },
+    {
+      "epoch": 0.47404282933160286,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.708222199325759e-06,
+      "loss": 1.418,
+      "step": 487
+    },
+    {
+      "epoch": 0.47501622323166776,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.70702202897402e-06,
+      "loss": 1.4164,
+      "step": 488
+    },
+    {
+      "epoch": 0.47598961713173266,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.70581954896925e-06,
+      "loss": 1.4043,
+      "step": 489
+    },
+    {
+      "epoch": 0.47696301103179756,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.704614760569846e-06,
+      "loss": 1.4203,
+      "step": 490
+    },
+    {
+      "epoch": 0.4779364049318624,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.703407665036622e-06,
+      "loss": 1.4177,
+      "step": 491
+    },
+    {
+      "epoch": 0.4789097988319273,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.702198263632808e-06,
+      "loss": 1.4099,
+      "step": 492
+    },
+    {
+      "epoch": 0.4798831927319922,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.700986557624041e-06,
+      "loss": 1.4183,
+      "step": 493
+    },
+    {
+      "epoch": 0.4808565866320571,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.699772548278378e-06,
+      "loss": 1.4027,
+      "step": 494
+    },
+    {
+      "epoch": 0.481829980532122,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.6985562368662795e-06,
+      "loss": 1.3994,
+      "step": 495
+    },
+    {
+      "epoch": 0.4828033744321869,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.697337624660619e-06,
+      "loss": 1.4274,
+      "step": 496
+    },
+    {
+      "epoch": 0.4837767683322518,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.696116712936676e-06,
+      "loss": 1.4226,
+      "step": 497
+    },
+    {
+      "epoch": 0.48475016223231665,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.694893502972137e-06,
+      "loss": 1.43,
+      "step": 498
+    },
+    {
+      "epoch": 0.48572355613238155,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.693667996047094e-06,
+      "loss": 1.4295,
+      "step": 499
+    },
+    {
+      "epoch": 0.48669695003244645,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.692440193444043e-06,
+      "loss": 1.4063,
+      "step": 500
+    },
+    {
+      "epoch": 0.48767034393251135,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.6912100964478825e-06,
+      "loss": 1.4152,
+      "step": 501
+    },
+    {
+      "epoch": 0.48864373783257625,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.689977706345909e-06,
+      "loss": 1.4157,
+      "step": 502
+    },
+    {
+      "epoch": 0.48961713173264115,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.6887430244278235e-06,
+      "loss": 1.4212,
+      "step": 503
+    },
+    {
+      "epoch": 0.49059052563270605,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.6875060519857215e-06,
+      "loss": 1.4256,
+      "step": 504
+    },
+    {
+      "epoch": 0.49156391953277095,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.686266790314099e-06,
+      "loss": 1.4218,
+      "step": 505
+    },
+    {
+      "epoch": 0.4925373134328358,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.685025240709845e-06,
+      "loss": 1.4024,
+      "step": 506
+    },
+    {
+      "epoch": 0.4935107073329007,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.683781404472243e-06,
+      "loss": 1.4169,
+      "step": 507
+    },
+    {
+      "epoch": 0.4944841012329656,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.6825352829029705e-06,
+      "loss": 1.4094,
+      "step": 508
+    },
+    {
+      "epoch": 0.4954574951330305,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.6812868773060975e-06,
+      "loss": 1.4059,
+      "step": 509
+    },
+    {
+      "epoch": 0.4964308890330954,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.6800361889880805e-06,
+      "loss": 1.4261,
+      "step": 510
+    },
+    {
+      "epoch": 0.4974042829331603,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.67878321925777e-06,
+      "loss": 1.4211,
+      "step": 511
+    },
+    {
+      "epoch": 0.4983776768332252,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.6775279694264e-06,
+      "loss": 1.3971,
+      "step": 512
+    },
+    {
+      "epoch": 0.49935107073329005,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.6762704408075925e-06,
+      "loss": 1.3995,
+      "step": 513
+    },
+    {
+      "epoch": 0.500324464633355,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.675010634717353e-06,
+      "loss": 1.4011,
+      "step": 514
+    },
+    {
+      "epoch": 0.5012978585334199,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.673748552474071e-06,
+      "loss": 1.4245,
+      "step": 515
+    },
+    {
+      "epoch": 0.5022712524334848,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.672484195398519e-06,
+      "loss": 1.4161,
+      "step": 516
+    },
+    {
+      "epoch": 0.5032446463335496,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.671217564813849e-06,
+      "loss": 1.4103,
+      "step": 517
+    },
+    {
+      "epoch": 0.5042180402336145,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.669948662045593e-06,
+      "loss": 1.407,
+      "step": 518
+    },
+    {
+      "epoch": 0.5051914341336794,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.668677488421659e-06,
+      "loss": 1.4174,
+      "step": 519
+    },
+    {
+      "epoch": 0.5061648280337443,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.667404045272334e-06,
+      "loss": 1.4044,
+      "step": 520
+    },
+    {
+      "epoch": 0.5071382219338092,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.666128333930278e-06,
+      "loss": 1.3874,
+      "step": 521
+    },
+    {
+      "epoch": 0.5081116158338741,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.664850355730526e-06,
+      "loss": 1.408,
+      "step": 522
+    },
+    {
+      "epoch": 0.509085009733939,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.663570112010485e-06,
+      "loss": 1.4181,
+      "step": 523
+    },
+    {
+      "epoch": 0.5100584036340039,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.662287604109932e-06,
+      "loss": 1.4218,
+      "step": 524
+    },
+    {
+      "epoch": 0.5110317975340688,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.661002833371014e-06,
+      "loss": 1.4302,
+      "step": 525
+    },
+    {
+      "epoch": 0.5120051914341337,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.659715801138247e-06,
+      "loss": 1.4138,
+      "step": 526
+    },
+    {
+      "epoch": 0.5129785853341986,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.658426508758512e-06,
+      "loss": 1.3998,
+      "step": 527
+    },
+    {
+      "epoch": 0.5139519792342635,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.657134957581057e-06,
+      "loss": 1.3993,
+      "step": 528
+    },
+    {
+      "epoch": 0.5149253731343284,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.655841148957493e-06,
+      "loss": 1.4133,
+      "step": 529
+    },
+    {
+      "epoch": 0.5158987670343933,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.654545084241792e-06,
+      "loss": 1.4066,
+      "step": 530
+    },
+    {
+      "epoch": 0.5168721609344581,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.653246764790289e-06,
+      "loss": 1.3966,
+      "step": 531
+    },
+    {
+      "epoch": 0.517845554834523,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.6519461919616795e-06,
+      "loss": 1.3933,
+      "step": 532
+    },
+    {
+      "epoch": 0.5188189487345879,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.650643367117013e-06,
+      "loss": 1.4092,
+      "step": 533
+    },
+    {
+      "epoch": 0.5197923426346528,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.649338291619699e-06,
+      "loss": 1.3936,
+      "step": 534
+    },
+    {
+      "epoch": 0.5207657365347177,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.648030966835502e-06,
+      "loss": 1.3975,
+      "step": 535
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.646721394132541e-06,
+      "loss": 1.4235,
+      "step": 536
+    },
+    {
+      "epoch": 0.5227125243348475,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.645409574881287e-06,
+      "loss": 1.3871,
+      "step": 537
+    },
+    {
+      "epoch": 0.5236859182349124,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.6440955104545595e-06,
+      "loss": 1.4192,
+      "step": 538
+    },
+    {
+      "epoch": 0.5246593121349773,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.6427792022275296e-06,
+      "loss": 1.4187,
+      "step": 539
+    },
+    {
+      "epoch": 0.5256327060350422,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.641460651577717e-06,
+      "loss": 1.3986,
+      "step": 540
+    },
+    {
+      "epoch": 0.5266060999351071,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.640139859884989e-06,
+      "loss": 1.371,
+      "step": 541
+    },
+    {
+      "epoch": 0.527579493835172,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.638816828531555e-06,
+      "loss": 1.4026,
+      "step": 542
+    },
+    {
+      "epoch": 0.5285528877352369,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.637491558901971e-06,
+      "loss": 1.4008,
+      "step": 543
+    },
+    {
+      "epoch": 0.5295262816353018,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.636164052383134e-06,
+      "loss": 1.3944,
+      "step": 544
+    },
+    {
+      "epoch": 0.5304996755353667,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.634834310364282e-06,
+      "loss": 1.421,
+      "step": 545
+    },
+    {
+      "epoch": 0.5314730694354315,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.633502334236993e-06,
+      "loss": 1.4209,
+      "step": 546
+    },
+    {
+      "epoch": 0.5324464633354964,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.632168125395183e-06,
+      "loss": 1.3769,
+      "step": 547
+    },
+    {
+      "epoch": 0.5334198572355613,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.6308316852351036e-06,
+      "loss": 1.4062,
+      "step": 548
+    },
+    {
+      "epoch": 0.5343932511356262,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.629493015155343e-06,
+      "loss": 1.4029,
+      "step": 549
+    },
+    {
+      "epoch": 0.5353666450356911,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.628152116556821e-06,
+      "loss": 1.4101,
+      "step": 550
+    },
+    {
+      "epoch": 0.536340038935756,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.626808990842793e-06,
+      "loss": 1.4199,
+      "step": 551
+    },
+    {
+      "epoch": 0.5373134328358209,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.625463639418839e-06,
+      "loss": 1.4257,
+      "step": 552
+    },
+    {
+      "epoch": 0.5382868267358858,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.624116063692875e-06,
+      "loss": 1.3929,
+      "step": 553
+    },
+    {
+      "epoch": 0.5392602206359507,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.622766265075141e-06,
+      "loss": 1.4009,
+      "step": 554
+    },
+    {
+      "epoch": 0.5402336145360156,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.621414244978204e-06,
+      "loss": 1.3887,
+      "step": 555
+    },
+    {
+      "epoch": 0.5412070084360805,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.620060004816957e-06,
+      "loss": 1.4077,
+      "step": 556
+    },
+    {
+      "epoch": 0.5421804023361454,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.618703546008611e-06,
+      "loss": 1.4113,
+      "step": 557
+    },
+    {
+      "epoch": 0.5431537962362103,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.617344869972707e-06,
+      "loss": 1.4042,
+      "step": 558
+    },
+    {
+      "epoch": 0.5441271901362752,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.615983978131102e-06,
+      "loss": 1.4071,
+      "step": 559
+    },
+    {
+      "epoch": 0.54510058403634,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.614620871907969e-06,
+      "loss": 1.3912,
+      "step": 560
+    },
+    {
+      "epoch": 0.5460739779364049,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.613255552729805e-06,
+      "loss": 1.4088,
+      "step": 561
+    },
+    {
+      "epoch": 0.5470473718364698,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.611888022025417e-06,
+      "loss": 1.3976,
+      "step": 562
+    },
+    {
+      "epoch": 0.5480207657365347,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.610518281225929e-06,
+      "loss": 1.4268,
+      "step": 563
+    },
+    {
+      "epoch": 0.5489941596365996,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.609146331764778e-06,
+      "loss": 1.3967,
+      "step": 564
+    },
+    {
+      "epoch": 0.5499675535366645,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.607772175077712e-06,
+      "loss": 1.4042,
+      "step": 565
+    },
+    {
+      "epoch": 0.5509409474367294,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.606395812602788e-06,
+      "loss": 1.4014,
+      "step": 566
+    },
+    {
+      "epoch": 0.5519143413367943,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.605017245780372e-06,
+      "loss": 1.417,
+      "step": 567
+    },
+    {
+      "epoch": 0.5528877352368592,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.603636476053139e-06,
+      "loss": 1.387,
+      "step": 568
+    },
+    {
+      "epoch": 0.5538611291369241,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.602253504866066e-06,
+      "loss": 1.4108,
+      "step": 569
+    },
+    {
+      "epoch": 0.554834523036989,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.600868333666434e-06,
+      "loss": 1.4,
+      "step": 570
+    },
+    {
+      "epoch": 0.5558079169370539,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.5994809639038285e-06,
+      "loss": 1.398,
+      "step": 571
+    },
+    {
+      "epoch": 0.5567813108371188,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.598091397030136e-06,
+      "loss": 1.3986,
+      "step": 572
+    },
+    {
+      "epoch": 0.5577547047371837,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.596699634499538e-06,
+      "loss": 1.3918,
+      "step": 573
+    },
+    {
+      "epoch": 0.5587280986372486,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.59530567776852e-06,
+      "loss": 1.3921,
+      "step": 574
+    },
+    {
+      "epoch": 0.5597014925373134,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.593909528295859e-06,
+      "loss": 1.3768,
+      "step": 575
+    },
+    {
+      "epoch": 0.5606748864373783,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.5925111875426285e-06,
+      "loss": 1.4077,
+      "step": 576
+    },
+    {
+      "epoch": 0.5616482803374432,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.591110656972195e-06,
+      "loss": 1.411,
+      "step": 577
+    },
+    {
+      "epoch": 0.5626216742375081,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.589707938050216e-06,
+      "loss": 1.3958,
+      "step": 578
+    },
+    {
+      "epoch": 0.563595068137573,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.588303032244641e-06,
+      "loss": 1.4149,
+      "step": 579
+    },
+    {
+      "epoch": 0.5645684620376379,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.586895941025705e-06,
+      "loss": 1.3885,
+      "step": 580
+    },
+    {
+      "epoch": 0.5655418559377028,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.585486665865933e-06,
+      "loss": 1.3821,
+      "step": 581
+    },
+    {
+      "epoch": 0.5665152498377677,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.584075208240135e-06,
+      "loss": 1.3888,
+      "step": 582
+    },
+    {
+      "epoch": 0.5674886437378326,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.5826615696254026e-06,
+      "loss": 1.374,
+      "step": 583
+    },
+    {
+      "epoch": 0.5684620376378975,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.581245751501113e-06,
+      "loss": 1.3988,
+      "step": 584
+    },
+    {
+      "epoch": 0.5694354315379624,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.579827755348921e-06,
+      "loss": 1.3786,
+      "step": 585
+    },
+    {
+      "epoch": 0.5704088254380273,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.578407582652764e-06,
+      "loss": 1.3826,
+      "step": 586
+    },
+    {
+      "epoch": 0.5713822193380922,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.576985234898855e-06,
+      "loss": 1.3793,
+      "step": 587
+    },
+    {
+      "epoch": 0.5723556132381571,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.575560713575684e-06,
+      "loss": 1.4009,
+      "step": 588
+    },
+    {
+      "epoch": 0.5733290071382219,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.5741340201740146e-06,
+      "loss": 1.4198,
+      "step": 589
+    },
+    {
+      "epoch": 0.5743024010382868,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.572705156186886e-06,
+      "loss": 1.4197,
+      "step": 590
+    },
+    {
+      "epoch": 0.5752757949383517,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.571274123109606e-06,
+      "loss": 1.3729,
+      "step": 591
+    },
+    {
+      "epoch": 0.5762491888384166,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.569840922439753e-06,
+      "loss": 1.3909,
+      "step": 592
+    },
+    {
+      "epoch": 0.5772225827384815,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.568405555677177e-06,
+      "loss": 1.3993,
+      "step": 593
+    },
+    {
+      "epoch": 0.5781959766385464,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.566968024323989e-06,
+      "loss": 1.405,
+      "step": 594
+    },
+    {
+      "epoch": 0.5791693705386113,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.565528329884571e-06,
+      "loss": 1.4054,
+      "step": 595
+    },
+    {
+      "epoch": 0.5801427644386762,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.564086473865565e-06,
+      "loss": 1.3597,
+      "step": 596
+    },
+    {
+      "epoch": 0.5811161583387411,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.562642457775876e-06,
+      "loss": 1.3909,
+      "step": 597
+    },
+    {
+      "epoch": 0.582089552238806,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.56119628312667e-06,
+      "loss": 1.4039,
+      "step": 598
+    },
+    {
+      "epoch": 0.5830629461388709,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.559747951431372e-06,
+      "loss": 1.395,
+      "step": 599
+    },
+    {
+      "epoch": 0.5840363400389358,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.558297464205666e-06,
+      "loss": 1.3963,
+      "step": 600
+    },
+    {
+      "epoch": 0.5850097339390007,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.556844822967486e-06,
+      "loss": 1.4153,
+      "step": 601
+    },
+    {
+      "epoch": 0.5859831278390656,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.555390029237026e-06,
+      "loss": 1.3682,
+      "step": 602
+    },
+    {
+      "epoch": 0.5869565217391305,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.55393308453673e-06,
+      "loss": 1.4046,
+      "step": 603
+    },
+    {
+      "epoch": 0.5879299156391953,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.552473990391294e-06,
+      "loss": 1.3929,
+      "step": 604
+    },
+    {
+      "epoch": 0.5889033095392602,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.551012748327663e-06,
+      "loss": 1.3667,
+      "step": 605
+    },
+    {
+      "epoch": 0.5898767034393251,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.549549359875031e-06,
+      "loss": 1.3875,
+      "step": 606
+    },
+    {
+      "epoch": 0.59085009733939,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.548083826564834e-06,
+      "loss": 1.3889,
+      "step": 607
+    },
+    {
+      "epoch": 0.5918234912394549,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.546616149930758e-06,
+      "loss": 1.3711,
+      "step": 608
+    },
+    {
+      "epoch": 0.5927968851395198,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.54514633150873e-06,
+      "loss": 1.3763,
+      "step": 609
+    },
+    {
+      "epoch": 0.5937702790395847,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.543674372836915e-06,
+      "loss": 1.3942,
+      "step": 610
+    },
+    {
+      "epoch": 0.5947436729396496,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.542200275455724e-06,
+      "loss": 1.3893,
+      "step": 611
+    },
+    {
+      "epoch": 0.5957170668397145,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.5407240409078e-06,
+      "loss": 1.3826,
+      "step": 612
+    },
+    {
+      "epoch": 0.5966904607397794,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.539245670738029e-06,
+      "loss": 1.3885,
+      "step": 613
+    },
+    {
+      "epoch": 0.5976638546398443,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.537765166493524e-06,
+      "loss": 1.4051,
+      "step": 614
+    },
+    {
+      "epoch": 0.5986372485399092,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.5362825297236394e-06,
+      "loss": 1.3893,
+      "step": 615
+    },
+    {
+      "epoch": 0.5996106424399741,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.534797761979955e-06,
+      "loss": 1.3847,
+      "step": 616
+    },
+    {
+      "epoch": 0.600584036340039,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.533310864816286e-06,
+      "loss": 1.3936,
+      "step": 617
+    },
+    {
+      "epoch": 0.6015574302401038,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.531821839788671e-06,
+      "loss": 1.383,
+      "step": 618
+    },
+    {
+      "epoch": 0.6025308241401687,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.5303306884553785e-06,
+      "loss": 1.3955,
+      "step": 619
+    },
+    {
+      "epoch": 0.6035042180402336,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.528837412376902e-06,
+      "loss": 1.3963,
+      "step": 620
+    },
+    {
+      "epoch": 0.6044776119402985,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.527342013115956e-06,
+      "loss": 1.3767,
+      "step": 621
+    },
+    {
+      "epoch": 0.6054510058403634,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.525844492237481e-06,
+      "loss": 1.3808,
+      "step": 622
+    },
+    {
+      "epoch": 0.6064243997404283,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.524344851308635e-06,
+      "loss": 1.3882,
+      "step": 623
+    },
+    {
+      "epoch": 0.6073977936404932,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.522843091898795e-06,
+      "loss": 1.3824,
+      "step": 624
+    },
+    {
+      "epoch": 0.6083711875405581,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.521339215579555e-06,
+      "loss": 1.3764,
+      "step": 625
+    },
+    {
+      "epoch": 0.609344581440623,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.519833223924725e-06,
+      "loss": 1.3805,
+      "step": 626
+    },
+    {
+      "epoch": 0.6103179753406879,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.518325118510328e-06,
+      "loss": 1.3689,
+      "step": 627
+    },
+    {
+      "epoch": 0.6112913692407528,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.516814900914601e-06,
+      "loss": 1.386,
+      "step": 628
+    },
+    {
+      "epoch": 0.6122647631408177,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.515302572717987e-06,
+      "loss": 1.4075,
+      "step": 629
+    },
+    {
+      "epoch": 0.6132381570408826,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.513788135503142e-06,
+      "loss": 1.3808,
+      "step": 630
+    },
+    {
+      "epoch": 0.6142115509409475,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.512271590854929e-06,
+      "loss": 1.3767,
+      "step": 631
+    },
+    {
+      "epoch": 0.6151849448410124,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.5107529403604126e-06,
+      "loss": 1.3818,
+      "step": 632
+    },
+    {
+      "epoch": 0.6161583387410772,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.509232185608864e-06,
+      "loss": 1.3886,
+      "step": 633
+    },
+    {
+      "epoch": 0.6171317326411421,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.507709328191758e-06,
+      "loss": 1.3751,
+      "step": 634
+    },
+    {
+      "epoch": 0.618105126541207,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.506184369702766e-06,
+      "loss": 1.3608,
+      "step": 635
+    },
+    {
+      "epoch": 0.6190785204412719,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.5046573117377616e-06,
+      "loss": 1.3771,
+      "step": 636
+    },
+    {
+      "epoch": 0.6200519143413368,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.503128155894812e-06,
+      "loss": 1.3854,
+      "step": 637
+    },
+    {
+      "epoch": 0.6210253082414017,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.501596903774184e-06,
+      "loss": 1.3916,
+      "step": 638
+    },
+    {
+      "epoch": 0.6219987021414666,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.5000635569783365e-06,
+      "loss": 1.3768,
+      "step": 639
+    },
+    {
+      "epoch": 0.6229720960415315,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.498528117111918e-06,
+      "loss": 1.3629,
+      "step": 640
+    },
+    {
+      "epoch": 0.6239454899415964,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.49699058578177e-06,
+      "loss": 1.3698,
+      "step": 641
+    },
+    {
+      "epoch": 0.6249188838416613,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.495450964596923e-06,
+      "loss": 1.3571,
+      "step": 642
+    },
+    {
+      "epoch": 0.6258922777417262,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.493909255168592e-06,
+      "loss": 1.3787,
+      "step": 643
+    },
+    {
+      "epoch": 0.6268656716417911,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.492365459110182e-06,
+      "loss": 1.3611,
+      "step": 644
+    },
+    {
+      "epoch": 0.627839065541856,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.490819578037275e-06,
+      "loss": 1.3921,
+      "step": 645
+    },
+    {
+      "epoch": 0.6288124594419209,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.4892716135676415e-06,
+      "loss": 1.4044,
+      "step": 646
+    },
+    {
+      "epoch": 0.6297858533419857,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.487721567321229e-06,
+      "loss": 1.3714,
+      "step": 647
+    },
+    {
+      "epoch": 0.6307592472420506,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.486169440920163e-06,
+      "loss": 1.3805,
+      "step": 648
+    },
+    {
+      "epoch": 0.6317326411421155,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.484615235988747e-06,
+      "loss": 1.3888,
+      "step": 649
+    },
+    {
+      "epoch": 0.6327060350421804,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.4830589541534615e-06,
+      "loss": 1.3453,
+      "step": 650
+    },
+    {
+      "epoch": 0.6336794289422453,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.481500597042956e-06,
+      "loss": 1.4119,
+      "step": 651
+    },
+    {
+      "epoch": 0.6346528228423102,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.479940166288056e-06,
+      "loss": 1.3965,
+      "step": 652
+    },
+    {
+      "epoch": 0.6356262167423751,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.4783776635217555e-06,
+      "loss": 1.4012,
+      "step": 653
+    },
+    {
+      "epoch": 0.63659961064244,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.476813090379216e-06,
+      "loss": 1.3946,
+      "step": 654
+    },
+    {
+      "epoch": 0.6375730045425049,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.475246448497766e-06,
+      "loss": 1.3893,
+      "step": 655
+    },
+    {
+      "epoch": 0.6385463984425698,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.473677739516901e-06,
+      "loss": 1.3874,
+      "step": 656
+    },
+    {
+      "epoch": 0.6395197923426347,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.472106965078277e-06,
+      "loss": 1.3767,
+      "step": 657
+    },
+    {
+      "epoch": 0.6404931862426996,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.470534126825714e-06,
+      "loss": 1.3864,
+      "step": 658
+    },
+    {
+      "epoch": 0.6414665801427645,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.468959226405188e-06,
+      "loss": 1.3849,
+      "step": 659
+    },
+    {
+      "epoch": 0.6424399740428294,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.467382265464838e-06,
+      "loss": 1.3789,
+      "step": 660
+    },
+    {
+      "epoch": 0.6434133679428943,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.465803245654955e-06,
+      "loss": 1.3473,
+      "step": 661
+    },
+    {
+      "epoch": 0.6443867618429591,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.464222168627987e-06,
+      "loss": 1.3614,
+      "step": 662
+    },
+    {
+      "epoch": 0.645360155743024,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.462639036038536e-06,
+      "loss": 1.368,
+      "step": 663
+    },
+    {
+      "epoch": 0.6463335496430889,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.461053849543351e-06,
+      "loss": 1.3965,
+      "step": 664
+    },
+    {
+      "epoch": 0.6473069435431538,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.459466610801333e-06,
+      "loss": 1.3626,
+      "step": 665
+    },
+    {
+      "epoch": 0.6482803374432187,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.457877321473532e-06,
+      "loss": 1.3897,
+      "step": 666
+    },
+    {
+      "epoch": 0.6492537313432836,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.456285983223143e-06,
+      "loss": 1.3792,
+      "step": 667
+    },
+    {
+      "epoch": 0.6502271252433485,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.454692597715502e-06,
+      "loss": 1.3975,
+      "step": 668
+    },
+    {
+      "epoch": 0.6512005191434134,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.4530971666180925e-06,
+      "loss": 1.3743,
+      "step": 669
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.451499691600536e-06,
+      "loss": 1.3572,
+      "step": 670
+    },
+    {
+      "epoch": 0.6531473069435432,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.449900174334592e-06,
+      "loss": 1.3806,
+      "step": 671
+    },
+    {
+      "epoch": 0.6541207008436081,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.44829861649416e-06,
+      "loss": 1.3852,
+      "step": 672
+    },
+    {
+      "epoch": 0.655094094743673,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.446695019755274e-06,
+      "loss": 1.3605,
+      "step": 673
+    },
+    {
+      "epoch": 0.6560674886437379,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.445089385796099e-06,
+      "loss": 1.3732,
+      "step": 674
+    },
+    {
+      "epoch": 0.6570408825438028,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.443481716296936e-06,
+      "loss": 1.3804,
+      "step": 675
+    },
+    {
+      "epoch": 0.6580142764438677,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.4418720129402145e-06,
+      "loss": 1.3737,
+      "step": 676
+    },
+    {
+      "epoch": 0.6589876703439325,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.440260277410491e-06,
+      "loss": 1.3599,
+      "step": 677
+    },
+    {
+      "epoch": 0.6599610642439974,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.438646511394451e-06,
+      "loss": 1.3439,
+      "step": 678
+    },
+    {
+      "epoch": 0.6609344581440623,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.437030716580904e-06,
+      "loss": 1.3733,
+      "step": 679
+    },
+    {
+      "epoch": 0.6619078520441272,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.435412894660782e-06,
+      "loss": 1.365,
+      "step": 680
+    },
+    {
+      "epoch": 0.6628812459441921,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.433793047327138e-06,
+      "loss": 1.3667,
+      "step": 681
+    },
+    {
+      "epoch": 0.663854639844257,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.432171176275149e-06,
+      "loss": 1.3674,
+      "step": 682
+    },
+    {
+      "epoch": 0.6648280337443219,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.430547283202103e-06,
+      "loss": 1.3543,
+      "step": 683
+    },
+    {
+      "epoch": 0.6658014276443868,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.428921369807407e-06,
+      "loss": 1.3757,
+      "step": 684
+    },
+    {
+      "epoch": 0.6667748215444517,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.427293437792585e-06,
+      "loss": 1.371,
+      "step": 685
+    },
+    {
+      "epoch": 0.6677482154445166,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.42566348886127e-06,
+      "loss": 1.3876,
+      "step": 686
+    },
+    {
+      "epoch": 0.6687216093445815,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.424031524719208e-06,
+      "loss": 1.3651,
+      "step": 687
+    },
+    {
+      "epoch": 0.6696950032446464,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.422397547074252e-06,
+      "loss": 1.3677,
+      "step": 688
+    },
+    {
+      "epoch": 0.6706683971447113,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.420761557636362e-06,
+      "loss": 1.3676,
+      "step": 689
+    },
+    {
+      "epoch": 0.6716417910447762,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.419123558117605e-06,
+      "loss": 1.3757,
+      "step": 690
+    },
+    {
+      "epoch": 0.672615184944841,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.417483550232151e-06,
+      "loss": 1.3686,
+      "step": 691
+    },
+    {
+      "epoch": 0.6735885788449059,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.415841535696271e-06,
+      "loss": 1.3566,
+      "step": 692
+    },
+    {
+      "epoch": 0.6745619727449708,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.414197516228338e-06,
+      "loss": 1.3666,
+      "step": 693
+    },
+    {
+      "epoch": 0.6755353666450357,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.41255149354882e-06,
+      "loss": 1.3482,
+      "step": 694
+    },
+    {
+      "epoch": 0.6765087605451006,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.410903469380284e-06,
+      "loss": 1.3735,
+      "step": 695
+    },
+    {
+      "epoch": 0.6774821544451655,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.40925344544739e-06,
+      "loss": 1.3773,
+      "step": 696
+    },
+    {
+      "epoch": 0.6784555483452304,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.407601423476893e-06,
+      "loss": 1.3859,
+      "step": 697
+    },
+    {
+      "epoch": 0.6794289422452953,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.405947405197635e-06,
+      "loss": 1.3765,
+      "step": 698
+    },
+    {
+      "epoch": 0.6804023361453602,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.404291392340551e-06,
+      "loss": 1.3512,
+      "step": 699
+    },
+    {
+      "epoch": 0.6813757300454251,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.402633386638662e-06,
+      "loss": 1.3713,
+      "step": 700
+    },
+    {
+      "epoch": 0.68234912394549,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.400973389827072e-06,
+      "loss": 1.3983,
+      "step": 701
+    },
+    {
+      "epoch": 0.6833225178455549,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.399311403642975e-06,
+      "loss": 1.3858,
+      "step": 702
+    },
+    {
+      "epoch": 0.6842959117456198,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.3976474298256395e-06,
+      "loss": 1.3683,
+      "step": 703
+    },
+    {
+      "epoch": 0.6852693056456847,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.395981470116419e-06,
+      "loss": 1.3935,
+      "step": 704
+    },
+    {
+      "epoch": 0.6862426995457496,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.394313526258743e-06,
+      "loss": 1.373,
+      "step": 705
+    },
+    {
+      "epoch": 0.6872160934458144,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.3926435999981194e-06,
+      "loss": 1.3831,
+      "step": 706
+    },
+    {
+      "epoch": 0.6881894873458793,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.390971693082128e-06,
+      "loss": 1.3843,
+      "step": 707
+    },
+    {
+      "epoch": 0.6891628812459442,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.3892978072604235e-06,
+      "loss": 1.3679,
+      "step": 708
+    },
+    {
+      "epoch": 0.6901362751460091,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.38762194428473e-06,
+      "loss": 1.3644,
+      "step": 709
+    },
+    {
+      "epoch": 0.691109669046074,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.3859441059088435e-06,
+      "loss": 1.3766,
+      "step": 710
+    },
+    {
+      "epoch": 0.6920830629461389,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.384264293888624e-06,
+      "loss": 1.3654,
+      "step": 711
+    },
+    {
+      "epoch": 0.6930564568462038,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.382582509981996e-06,
+      "loss": 1.3694,
+      "step": 712
+    },
+    {
+      "epoch": 0.6940298507462687,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.3808987559489536e-06,
+      "loss": 1.3698,
+      "step": 713
+    },
+    {
+      "epoch": 0.6950032446463336,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.379213033551547e-06,
+      "loss": 1.3681,
+      "step": 714
+    },
+    {
+      "epoch": 0.6959766385463985,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.377525344553888e-06,
+      "loss": 1.3757,
+      "step": 715
+    },
+    {
+      "epoch": 0.6969500324464634,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.375835690722147e-06,
+      "loss": 1.3604,
+      "step": 716
+    },
+    {
+      "epoch": 0.6979234263465283,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.374144073824549e-06,
+      "loss": 1.3715,
+      "step": 717
+    },
+    {
+      "epoch": 0.6988968202465932,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.372450495631376e-06,
+      "loss": 1.3486,
+      "step": 718
+    },
+    {
+      "epoch": 0.6998702141466581,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.3707549579149605e-06,
+      "loss": 1.3803,
+      "step": 719
+    },
+    {
+      "epoch": 0.7008436080467229,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.369057462449686e-06,
+      "loss": 1.3779,
+      "step": 720
+    },
+    {
+      "epoch": 0.7018170019467878,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.367358011011985e-06,
+      "loss": 1.375,
+      "step": 721
+    },
+    {
+      "epoch": 0.7027903958468527,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.365656605380338e-06,
+      "loss": 1.351,
+      "step": 722
+    },
+    {
+      "epoch": 0.7037637897469176,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.363953247335267e-06,
+      "loss": 1.3634,
+      "step": 723
+    },
+    {
+      "epoch": 0.7047371836469825,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.362247938659342e-06,
+      "loss": 1.3717,
+      "step": 724
+    },
+    {
+      "epoch": 0.7057105775470474,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.36054068113717e-06,
+      "loss": 1.3568,
+      "step": 725
+    },
+    {
+      "epoch": 0.7066839714471123,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.358831476555401e-06,
+      "loss": 1.3884,
+      "step": 726
+    },
+    {
+      "epoch": 0.7076573653471772,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.357120326702721e-06,
+      "loss": 1.3781,
+      "step": 727
+    },
+    {
+      "epoch": 0.7086307592472421,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.35540723336985e-06,
+      "loss": 1.3788,
+      "step": 728
+    },
+    {
+      "epoch": 0.709604153147307,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.353692198349547e-06,
+      "loss": 1.3691,
+      "step": 729
+    },
+    {
+      "epoch": 0.7105775470473719,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.351975223436597e-06,
+      "loss": 1.3503,
+      "step": 730
+    },
+    {
+      "epoch": 0.7115509409474368,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.3502563104278175e-06,
+      "loss": 1.3927,
+      "step": 731
+    },
+    {
+      "epoch": 0.7125243348475017,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.3485354611220555e-06,
+      "loss": 1.366,
+      "step": 732
+    },
+    {
+      "epoch": 0.7134977287475666,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.346812677320183e-06,
+      "loss": 1.3639,
+      "step": 733
+    },
+    {
+      "epoch": 0.7144711226476315,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.345087960825098e-06,
+      "loss": 1.3415,
+      "step": 734
+    },
+    {
+      "epoch": 0.7154445165476963,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.343361313441717e-06,
+      "loss": 1.3652,
+      "step": 735
+    },
+    {
+      "epoch": 0.7164179104477612,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.3416327369769824e-06,
+      "loss": 1.3618,
+      "step": 736
+    },
+    {
+      "epoch": 0.717391304347826,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.339902233239853e-06,
+      "loss": 1.366,
+      "step": 737
+    },
+    {
+      "epoch": 0.718364698247891,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.338169804041303e-06,
+      "loss": 1.3752,
+      "step": 738
+    },
+    {
+      "epoch": 0.7193380921479559,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.336435451194324e-06,
+      "loss": 1.337,
+      "step": 739
+    },
+    {
+      "epoch": 0.7203114860480208,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.334699176513919e-06,
+      "loss": 1.3579,
+      "step": 740
+    },
+    {
+      "epoch": 0.7212848799480857,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.3329609818171035e-06,
+      "loss": 1.3837,
+      "step": 741
+    },
+    {
+      "epoch": 0.7222582738481506,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.3312208689229026e-06,
+      "loss": 1.3677,
+      "step": 742
+    },
+    {
+      "epoch": 0.7232316677482155,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.3294788396523465e-06,
+      "loss": 1.3692,
+      "step": 743
+    },
+    {
+      "epoch": 0.7242050616482804,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.327734895828473e-06,
+      "loss": 1.3583,
+      "step": 744
+    },
+    {
+      "epoch": 0.7251784555483453,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.325989039276323e-06,
+      "loss": 1.353,
+      "step": 745
+    },
+    {
+      "epoch": 0.7261518494484102,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.324241271822939e-06,
+      "loss": 1.3476,
+      "step": 746
+    },
+    {
+      "epoch": 0.7271252433484751,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.322491595297363e-06,
+      "loss": 1.3385,
+      "step": 747
+    },
+    {
+      "epoch": 0.72809863724854,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.320740011530634e-06,
+      "loss": 1.3517,
+      "step": 748
+    },
+    {
+      "epoch": 0.7290720311486047,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.31898652235579e-06,
+      "loss": 1.3732,
+      "step": 749
+    },
+    {
+      "epoch": 0.7300454250486696,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.317231129607859e-06,
+      "loss": 1.3647,
+      "step": 750
+    },
+    {
+      "epoch": 0.7310188189487346,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.3154738351238655e-06,
+      "loss": 1.3595,
+      "step": 751
+    },
+    {
+      "epoch": 0.7319922128487995,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.3137146407428196e-06,
+      "loss": 1.3575,
+      "step": 752
+    },
+    {
+      "epoch": 0.7329656067488644,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.311953548305722e-06,
+      "loss": 1.3698,
+      "step": 753
+    },
+    {
+      "epoch": 0.7339390006489293,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.31019055965556e-06,
+      "loss": 1.3651,
+      "step": 754
+    },
+    {
+      "epoch": 0.7349123945489942,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.3084256766373056e-06,
+      "loss": 1.3609,
+      "step": 755
+    },
+    {
+      "epoch": 0.735885788449059,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.306658901097911e-06,
+      "loss": 1.3398,
+      "step": 756
+    },
+    {
+      "epoch": 0.736859182349124,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.3048902348863116e-06,
+      "loss": 1.3856,
+      "step": 757
+    },
+    {
+      "epoch": 0.7378325762491889,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.303119679853419e-06,
+      "loss": 1.3818,
+      "step": 758
+    },
+    {
+      "epoch": 0.7388059701492538,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.3013472378521236e-06,
+      "loss": 1.3709,
+      "step": 759
+    },
+    {
+      "epoch": 0.7397793640493187,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.299572910737289e-06,
+      "loss": 1.3466,
+      "step": 760
+    },
+    {
+      "epoch": 0.7407527579493836,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.297796700365752e-06,
+      "loss": 1.3511,
+      "step": 761
+    },
+    {
+      "epoch": 0.7417261518494485,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.296018608596321e-06,
+      "loss": 1.3448,
+      "step": 762
+    },
+    {
+      "epoch": 0.7426995457495134,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.294238637289772e-06,
+      "loss": 1.3608,
+      "step": 763
+    },
+    {
+      "epoch": 0.7436729396495781,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.29245678830885e-06,
+      "loss": 1.3773,
+      "step": 764
+    },
+    {
+      "epoch": 0.744646333549643,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.290673063518261e-06,
+      "loss": 1.3641,
+      "step": 765
+    },
+    {
+      "epoch": 0.745619727449708,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.288887464784679e-06,
+      "loss": 1.3438,
+      "step": 766
+    },
+    {
+      "epoch": 0.7465931213497728,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.287099993976735e-06,
+      "loss": 1.3731,
+      "step": 767
+    },
+    {
+      "epoch": 0.7475665152498377,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.2853106529650225e-06,
+      "loss": 1.3327,
+      "step": 768
+    },
+    {
+      "epoch": 0.7485399091499026,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.28351944362209e-06,
+      "loss": 1.3578,
+      "step": 769
+    },
+    {
+      "epoch": 0.7495133030499675,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.28172636782244e-06,
+      "loss": 1.3547,
+      "step": 770
+    },
+    {
+      "epoch": 0.7504866969500325,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.279931427442532e-06,
+      "loss": 1.3565,
+      "step": 771
+    },
+    {
+      "epoch": 0.7514600908500974,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.278134624360773e-06,
+      "loss": 1.357,
+      "step": 772
+    },
+    {
+      "epoch": 0.7524334847501623,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.276335960457522e-06,
+      "loss": 1.3664,
+      "step": 773
+    },
+    {
+      "epoch": 0.7534068786502272,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.2745354376150865e-06,
+      "loss": 1.3623,
+      "step": 774
+    },
+    {
+      "epoch": 0.754380272550292,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.272733057717714e-06,
+      "loss": 1.3567,
+      "step": 775
+    },
+    {
+      "epoch": 0.755353666450357,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.2709288226516e-06,
+      "loss": 1.3633,
+      "step": 776
+    },
+    {
+      "epoch": 0.7563270603504219,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.269122734304881e-06,
+      "loss": 1.361,
+      "step": 777
+    },
+    {
+      "epoch": 0.7573004542504866,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.2673147945676305e-06,
+      "loss": 1.3854,
+      "step": 778
+    },
+    {
+      "epoch": 0.7582738481505515,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.265505005331863e-06,
+      "loss": 1.3854,
+      "step": 779
+    },
+    {
+      "epoch": 0.7592472420506164,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.263693368491524e-06,
+      "loss": 1.3353,
+      "step": 780
+    },
+    {
+      "epoch": 0.7602206359506813,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.261879885942497e-06,
+      "loss": 1.3398,
+      "step": 781
+    },
+    {
+      "epoch": 0.7611940298507462,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.260064559582596e-06,
+      "loss": 1.3785,
+      "step": 782
+    },
+    {
+      "epoch": 0.7621674237508111,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.258247391311562e-06,
+      "loss": 1.3685,
+      "step": 783
+    },
+    {
+      "epoch": 0.763140817650876,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.256428383031065e-06,
+      "loss": 1.3556,
+      "step": 784
+    },
+    {
+      "epoch": 0.764114211550941,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.254607536644702e-06,
+      "loss": 1.3519,
+      "step": 785
+    },
+    {
+      "epoch": 0.7650876054510058,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.252784854057993e-06,
+      "loss": 1.3622,
+      "step": 786
+    },
+    {
+      "epoch": 0.7660609993510707,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.2509603371783776e-06,
+      "loss": 1.3613,
+      "step": 787
+    },
+    {
+      "epoch": 0.7670343932511356,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.249133987915217e-06,
+      "loss": 1.3614,
+      "step": 788
+    },
+    {
+      "epoch": 0.7680077871512005,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.247305808179789e-06,
+      "loss": 1.3472,
+      "step": 789
+    },
+    {
+      "epoch": 0.7689811810512654,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.245475799885288e-06,
+      "loss": 1.342,
+      "step": 790
+    },
+    {
+      "epoch": 0.7699545749513304,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.243643964946821e-06,
+      "loss": 1.3809,
+      "step": 791
+    },
+    {
+      "epoch": 0.7709279688513953,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.241810305281407e-06,
+      "loss": 1.3536,
+      "step": 792
+    },
+    {
+      "epoch": 0.77190136275146,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.239974822807976e-06,
+      "loss": 1.3412,
+      "step": 793
+    },
+    {
+      "epoch": 0.7728747566515249,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.238137519447362e-06,
+      "loss": 1.3441,
+      "step": 794
+    },
+    {
+      "epoch": 0.7738481505515898,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.236298397122307e-06,
+      "loss": 1.3554,
+      "step": 795
+    },
+    {
+      "epoch": 0.7748215444516547,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.234457457757457e-06,
+      "loss": 1.3543,
+      "step": 796
+    },
+    {
+      "epoch": 0.7757949383517196,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.232614703279359e-06,
+      "loss": 1.3508,
+      "step": 797
+    },
+    {
+      "epoch": 0.7767683322517845,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.230770135616459e-06,
+      "loss": 1.3559,
+      "step": 798
+    },
+    {
+      "epoch": 0.7777417261518494,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.2289237566991e-06,
+      "loss": 1.3658,
+      "step": 799
+    },
+    {
+      "epoch": 0.7787151200519143,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.227075568459522e-06,
+      "loss": 1.367,
+      "step": 800
+    },
+    {
+      "epoch": 0.7796885139519792,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.225225572831858e-06,
+      "loss": 1.3504,
+      "step": 801
+    },
+    {
+      "epoch": 0.7806619078520441,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.223373771752131e-06,
+      "loss": 1.3512,
+      "step": 802
+    },
+    {
+      "epoch": 0.781635301752109,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.221520167158257e-06,
+      "loss": 1.3597,
+      "step": 803
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.219664760990035e-06,
+      "loss": 1.3292,
+      "step": 804
+    },
+    {
+      "epoch": 0.7835820895522388,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.217807555189151e-06,
+      "loss": 1.3392,
+      "step": 805
+    },
+    {
+      "epoch": 0.7845554834523037,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.215948551699175e-06,
+      "loss": 1.3377,
+      "step": 806
+    },
+    {
+      "epoch": 0.7855288773523685,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.21408775246556e-06,
+      "loss": 1.3513,
+      "step": 807
+    },
+    {
+      "epoch": 0.7865022712524334,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.212225159435633e-06,
+      "loss": 1.3877,
+      "step": 808
+    },
+    {
+      "epoch": 0.7874756651524983,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.210360774558604e-06,
+      "loss": 1.3519,
+      "step": 809
+    },
+    {
+      "epoch": 0.7884490590525632,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.208494599785553e-06,
+      "loss": 1.3441,
+      "step": 810
+    },
+    {
+      "epoch": 0.7894224529526281,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.206626637069438e-06,
+      "loss": 1.3656,
+      "step": 811
+    },
+    {
+      "epoch": 0.790395846852693,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.204756888365085e-06,
+      "loss": 1.3504,
+      "step": 812
+    },
+    {
+      "epoch": 0.7913692407527579,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.202885355629189e-06,
+      "loss": 1.3429,
+      "step": 813
+    },
+    {
+      "epoch": 0.7923426346528228,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.201012040820314e-06,
+      "loss": 1.3432,
+      "step": 814
+    },
+    {
+      "epoch": 0.7933160285528877,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.199136945898887e-06,
+      "loss": 1.3612,
+      "step": 815
+    },
+    {
+      "epoch": 0.7942894224529526,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.197260072827199e-06,
+      "loss": 1.3556,
+      "step": 816
+    },
+    {
+      "epoch": 0.7952628163530175,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.1953814235694015e-06,
+      "loss": 1.3423,
+      "step": 817
+    },
+    {
+      "epoch": 0.7962362102530824,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.193501000091504e-06,
+      "loss": 1.37,
+      "step": 818
+    },
+    {
+      "epoch": 0.7972096041531473,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.1916188043613754e-06,
+      "loss": 1.3564,
+      "step": 819
+    },
+    {
+      "epoch": 0.7981829980532122,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.189734838348736e-06,
+      "loss": 1.3681,
+      "step": 820
+    },
+    {
+      "epoch": 0.7991563919532771,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.187849104025159e-06,
+      "loss": 1.38,
+      "step": 821
+    },
+    {
+      "epoch": 0.8001297858533419,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.18596160336407e-06,
+      "loss": 1.3596,
+      "step": 822
+    },
+    {
+      "epoch": 0.8011031797534068,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.184072338340743e-06,
+      "loss": 1.3484,
+      "step": 823
+    },
+    {
+      "epoch": 0.8020765736534717,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.1821813109322975e-06,
+      "loss": 1.3658,
+      "step": 824
+    },
+    {
+      "epoch": 0.8030499675535366,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.180288523117697e-06,
+      "loss": 1.3523,
+      "step": 825
+    },
+    {
+      "epoch": 0.8040233614536015,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.178393976877749e-06,
+      "loss": 1.3549,
+      "step": 826
+    },
+    {
+      "epoch": 0.8049967553536664,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.176497674195098e-06,
+      "loss": 1.3393,
+      "step": 827
+    },
+    {
+      "epoch": 0.8059701492537313,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.17459961705423e-06,
+      "loss": 1.351,
+      "step": 828
+    },
+    {
+      "epoch": 0.8069435431537962,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.172699807441465e-06,
+      "loss": 1.3442,
+      "step": 829
+    },
+    {
+      "epoch": 0.8079169370538611,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.1707982473449584e-06,
+      "loss": 1.3427,
+      "step": 830
+    },
+    {
+      "epoch": 0.808890330953926,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.168894938754696e-06,
+      "loss": 1.3437,
+      "step": 831
+    },
+    {
+      "epoch": 0.8098637248539909,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.166989883662495e-06,
+      "loss": 1.3536,
+      "step": 832
+    },
+    {
+      "epoch": 0.8108371187540558,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.165083084061997e-06,
+      "loss": 1.3342,
+      "step": 833
+    },
+    {
+      "epoch": 0.8118105126541207,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.1631745419486744e-06,
+      "loss": 1.3379,
+      "step": 834
+    },
+    {
+      "epoch": 0.8127839065541856,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.16126425931982e-06,
+      "loss": 1.3536,
+      "step": 835
+    },
+    {
+      "epoch": 0.8137573004542504,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.1593522381745465e-06,
+      "loss": 1.3368,
+      "step": 836
+    },
+    {
+      "epoch": 0.8147306943543153,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.15743848051379e-06,
+      "loss": 1.3444,
+      "step": 837
+    },
+    {
+      "epoch": 0.8157040882543802,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.155522988340301e-06,
+      "loss": 1.3621,
+      "step": 838
+    },
+    {
+      "epoch": 0.8166774821544451,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.153605763658645e-06,
+      "loss": 1.3682,
+      "step": 839
+    },
+    {
+      "epoch": 0.81765087605451,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.151686808475204e-06,
+      "loss": 1.3607,
+      "step": 840
+    },
+    {
+      "epoch": 0.8186242699545749,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.149766124798166e-06,
+      "loss": 1.3299,
+      "step": 841
+    },
+    {
+      "epoch": 0.8195976638546398,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.1478437146375315e-06,
+      "loss": 1.3649,
+      "step": 842
+    },
+    {
+      "epoch": 0.8205710577547047,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.145919580005107e-06,
+      "loss": 1.372,
+      "step": 843
+    },
+    {
+      "epoch": 0.8215444516547696,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.143993722914502e-06,
+      "loss": 1.3601,
+      "step": 844
+    },
+    {
+      "epoch": 0.8225178455548345,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.142066145381131e-06,
+      "loss": 1.3618,
+      "step": 845
+    },
+    {
+      "epoch": 0.8234912394548994,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.1401368494222075e-06,
+      "loss": 1.3613,
+      "step": 846
+    },
+    {
+      "epoch": 0.8244646333549643,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.138205837056743e-06,
+      "loss": 1.3517,
+      "step": 847
+    },
+    {
+      "epoch": 0.8254380272550292,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.136273110305547e-06,
+      "loss": 1.3322,
+      "step": 848
+    },
+    {
+      "epoch": 0.8264114211550941,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.134338671191221e-06,
+      "loss": 1.359,
+      "step": 849
+    },
+    {
+      "epoch": 0.827384815055159,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.132402521738159e-06,
+      "loss": 1.3477,
+      "step": 850
+    },
+    {
+      "epoch": 0.8283582089552238,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.130464663972548e-06,
+      "loss": 1.3535,
+      "step": 851
+    },
+    {
+      "epoch": 0.8293316028552887,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.128525099922357e-06,
+      "loss": 1.3613,
+      "step": 852
+    },
+    {
+      "epoch": 0.8303049967553536,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.1265838316173455e-06,
+      "loss": 1.3557,
+      "step": 853
+    },
+    {
+      "epoch": 0.8312783906554185,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.124640861089055e-06,
+      "loss": 1.336,
+      "step": 854
+    },
+    {
+      "epoch": 0.8322517845554834,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.122696190370805e-06,
+      "loss": 1.3498,
+      "step": 855
+    },
+    {
+      "epoch": 0.8332251784555483,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.1207498214977e-06,
+      "loss": 1.3572,
+      "step": 856
+    },
+    {
+      "epoch": 0.8341985723556132,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.11880175650662e-06,
+      "loss": 1.3538,
+      "step": 857
+    },
+    {
+      "epoch": 0.8351719662556781,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.1168519974362175e-06,
+      "loss": 1.3554,
+      "step": 858
+    },
+    {
+      "epoch": 0.836145360155743,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.1149005463269186e-06,
+      "loss": 1.324,
+      "step": 859
+    },
+    {
+      "epoch": 0.8371187540558079,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.112947405220921e-06,
+      "loss": 1.3603,
+      "step": 860
+    },
+    {
+      "epoch": 0.8380921479558728,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.110992576162193e-06,
+      "loss": 1.3518,
+      "step": 861
+    },
+    {
+      "epoch": 0.8390655418559377,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.1090360611964644e-06,
+      "loss": 1.3701,
+      "step": 862
+    },
+    {
+      "epoch": 0.8400389357560026,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.107077862371235e-06,
+      "loss": 1.3422,
+      "step": 863
+    },
+    {
+      "epoch": 0.8410123296560675,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.1051179817357605e-06,
+      "loss": 1.3492,
+      "step": 864
+    },
+    {
+      "epoch": 0.8419857235561323,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.103156421341062e-06,
+      "loss": 1.3381,
+      "step": 865
+    },
+    {
+      "epoch": 0.8429591174561972,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.101193183239916e-06,
+      "loss": 1.3399,
+      "step": 866
+    },
+    {
+      "epoch": 0.8439325113562621,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.0992282694868555e-06,
+      "loss": 1.3459,
+      "step": 867
+    },
+    {
+      "epoch": 0.844905905256327,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.097261682138166e-06,
+      "loss": 1.331,
+      "step": 868
+    },
+    {
+      "epoch": 0.8458792991563919,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.095293423251886e-06,
+      "loss": 1.3371,
+      "step": 869
+    },
+    {
+      "epoch": 0.8468526930564568,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.0933234948878025e-06,
+      "loss": 1.3529,
+      "step": 870
+    },
+    {
+      "epoch": 0.8478260869565217,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.091351899107448e-06,
+      "loss": 1.3533,
+      "step": 871
+    },
+    {
+      "epoch": 0.8487994808565866,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.089378637974103e-06,
+      "loss": 1.3438,
+      "step": 872
+    },
+    {
+      "epoch": 0.8497728747566515,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.087403713552789e-06,
+      "loss": 1.3272,
+      "step": 873
+    },
+    {
+      "epoch": 0.8507462686567164,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.085427127910268e-06,
+      "loss": 1.3609,
+      "step": 874
+    },
+    {
+      "epoch": 0.8517196625567813,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.083448883115041e-06,
+      "loss": 1.371,
+      "step": 875
+    },
+    {
+      "epoch": 0.8526930564568462,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.081468981237345e-06,
+      "loss": 1.3452,
+      "step": 876
+    },
+    {
+      "epoch": 0.8536664503569111,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.079487424349152e-06,
+      "loss": 1.34,
+      "step": 877
+    },
+    {
+      "epoch": 0.854639844256976,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.077504214524166e-06,
+      "loss": 1.3307,
+      "step": 878
+    },
+    {
+      "epoch": 0.8556132381570409,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.075519353837818e-06,
+      "loss": 1.3515,
+      "step": 879
+    },
+    {
+      "epoch": 0.8565866320571057,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.0735328443672694e-06,
+      "loss": 1.3461,
+      "step": 880
+    },
+    {
+      "epoch": 0.8575600259571706,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.071544688191407e-06,
+      "loss": 1.3386,
+      "step": 881
+    },
+    {
+      "epoch": 0.8585334198572355,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.06955488739084e-06,
+      "loss": 1.3717,
+      "step": 882
+    },
+    {
+      "epoch": 0.8595068137573004,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.0675634440478995e-06,
+      "loss": 1.3352,
+      "step": 883
+    },
+    {
+      "epoch": 0.8604802076573653,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.065570360246634e-06,
+      "loss": 1.3568,
+      "step": 884
+    },
+    {
+      "epoch": 0.8614536015574302,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.0635756380728096e-06,
+      "loss": 1.3352,
+      "step": 885
+    },
+    {
+      "epoch": 0.8624269954574951,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.061579279613909e-06,
+      "loss": 1.3333,
+      "step": 886
+    },
+    {
+      "epoch": 0.86340038935756,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.0595812869591235e-06,
+      "loss": 1.3464,
+      "step": 887
+    },
+    {
+      "epoch": 0.8643737832576249,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.057581662199357e-06,
+      "loss": 1.3583,
+      "step": 888
+    },
+    {
+      "epoch": 0.8653471771576898,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.055580407427222e-06,
+      "loss": 1.363,
+      "step": 889
+    },
+    {
+      "epoch": 0.8663205710577547,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.053577524737034e-06,
+      "loss": 1.3543,
+      "step": 890
+    },
+    {
+      "epoch": 0.8672939649578196,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.051573016224813e-06,
+      "loss": 1.3434,
+      "step": 891
+    },
+    {
+      "epoch": 0.8682673588578845,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.0495668839882846e-06,
+      "loss": 1.3433,
+      "step": 892
+    },
+    {
+      "epoch": 0.8692407527579494,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.047559130126868e-06,
+      "loss": 1.3248,
+      "step": 893
+    },
+    {
+      "epoch": 0.8702141466580143,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.045549756741682e-06,
+      "loss": 1.3511,
+      "step": 894
+    },
+    {
+      "epoch": 0.8711875405580791,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.043538765935539e-06,
+      "loss": 1.3574,
+      "step": 895
+    },
+    {
+      "epoch": 0.872160934458144,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.0415261598129465e-06,
+      "loss": 1.3671,
+      "step": 896
+    },
+    {
+      "epoch": 0.8731343283582089,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.0395119404801e-06,
+      "loss": 1.3383,
+      "step": 897
+    },
+    {
+      "epoch": 0.8741077222582738,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.037496110044885e-06,
+      "loss": 1.3695,
+      "step": 898
+    },
+    {
+      "epoch": 0.8750811161583387,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.03547867061687e-06,
+      "loss": 1.3367,
+      "step": 899
+    },
+    {
+      "epoch": 0.8760545100584036,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.0334596243073125e-06,
+      "loss": 1.3357,
+      "step": 900
+    },
+    {
+      "epoch": 0.8770279039584685,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.031438973229147e-06,
+      "loss": 1.3463,
+      "step": 901
+    },
+    {
+      "epoch": 0.8780012978585334,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.029416719496988e-06,
+      "loss": 1.3597,
+      "step": 902
+    },
+    {
+      "epoch": 0.8789746917585983,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.027392865227131e-06,
+      "loss": 1.3586,
+      "step": 903
+    },
+    {
+      "epoch": 0.8799480856586632,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.025367412537539e-06,
+      "loss": 1.3201,
+      "step": 904
+    },
+    {
+      "epoch": 0.8809214795587281,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.023340363547858e-06,
+      "loss": 1.3621,
+      "step": 905
+    },
+    {
+      "epoch": 0.881894873458793,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.021311720379394e-06,
+      "loss": 1.372,
+      "step": 906
+    },
+    {
+      "epoch": 0.8828682673588579,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.0192814851551284e-06,
+      "loss": 1.3457,
+      "step": 907
+    },
+    {
+      "epoch": 0.8838416612589228,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.017249659999707e-06,
+      "loss": 1.3675,
+      "step": 908
+    },
+    {
+      "epoch": 0.8848150551589876,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.015216247039438e-06,
+      "loss": 1.3445,
+      "step": 909
+    },
+    {
+      "epoch": 0.8857884490590525,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.013181248402293e-06,
+      "loss": 1.3588,
+      "step": 910
+    },
+    {
+      "epoch": 0.8867618429591174,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.0111446662179e-06,
+      "loss": 1.3382,
+      "step": 911
+    },
+    {
+      "epoch": 0.8877352368591823,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.009106502617548e-06,
+      "loss": 1.3339,
+      "step": 912
+    },
+    {
+      "epoch": 0.8887086307592472,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.0070667597341806e-06,
+      "loss": 1.3351,
+      "step": 913
+    },
+    {
+      "epoch": 0.8896820246593121,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.005025439702391e-06,
+      "loss": 1.3449,
+      "step": 914
+    },
+    {
+      "epoch": 0.890655418559377,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.002982544658426e-06,
+      "loss": 1.3419,
+      "step": 915
+    },
+    {
+      "epoch": 0.8916288124594419,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.000938076740177e-06,
+      "loss": 1.3651,
+      "step": 916
+    },
+    {
+      "epoch": 0.8926022063595068,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.998892038087187e-06,
+      "loss": 1.3451,
+      "step": 917
+    },
+    {
+      "epoch": 0.8935756002595717,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.996844430840637e-06,
+      "loss": 1.3337,
+      "step": 918
+    },
+    {
+      "epoch": 0.8945489941596366,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.994795257143355e-06,
+      "loss": 1.3516,
+      "step": 919
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.992744519139803e-06,
+      "loss": 1.3304,
+      "step": 920
+    },
+    {
+      "epoch": 0.8964957819597664,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.990692218976082e-06,
+      "loss": 1.3498,
+      "step": 921
+    },
+    {
+      "epoch": 0.8974691758598313,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.988638358799931e-06,
+      "loss": 1.3217,
+      "step": 922
+    },
+    {
+      "epoch": 0.8984425697598962,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.986582940760717e-06,
+      "loss": 1.3579,
+      "step": 923
+    },
+    {
+      "epoch": 0.899415963659961,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.98452596700944e-06,
+      "loss": 1.3152,
+      "step": 924
+    },
+    {
+      "epoch": 0.9003893575600259,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.982467439698725e-06,
+      "loss": 1.3425,
+      "step": 925
+    },
+    {
+      "epoch": 0.9013627514600908,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.980407360982828e-06,
+      "loss": 1.3399,
+      "step": 926
+    },
+    {
+      "epoch": 0.9023361453601557,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.978345733017624e-06,
+      "loss": 1.3438,
+      "step": 927
+    },
+    {
+      "epoch": 0.9033095392602206,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.976282557960611e-06,
+      "loss": 1.3449,
+      "step": 928
+    },
+    {
+      "epoch": 0.9042829331602855,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.974217837970906e-06,
+      "loss": 1.3385,
+      "step": 929
+    },
+    {
+      "epoch": 0.9052563270603504,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.9721515752092434e-06,
+      "loss": 1.3309,
+      "step": 930
+    },
+    {
+      "epoch": 0.9062297209604153,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.970083771837969e-06,
+      "loss": 1.3521,
+      "step": 931
+    },
+    {
+      "epoch": 0.9072031148604802,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.968014430021046e-06,
+      "loss": 1.3293,
+      "step": 932
+    },
+    {
+      "epoch": 0.9081765087605451,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.965943551924043e-06,
+      "loss": 1.3597,
+      "step": 933
+    },
+    {
+      "epoch": 0.90914990266061,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.963871139714139e-06,
+      "loss": 1.3457,
+      "step": 934
+    },
+    {
+      "epoch": 0.9101232965606749,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.961797195560118e-06,
+      "loss": 1.3425,
+      "step": 935
+    },
+    {
+      "epoch": 0.9110966904607398,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.959721721632366e-06,
+      "loss": 1.3502,
+      "step": 936
+    },
+    {
+      "epoch": 0.9120700843608047,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.95764472010287e-06,
+      "loss": 1.3603,
+      "step": 937
+    },
+    {
+      "epoch": 0.9130434782608695,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.955566193145218e-06,
+      "loss": 1.3416,
+      "step": 938
+    },
+    {
+      "epoch": 0.9140168721609344,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.953486142934591e-06,
+      "loss": 1.3304,
+      "step": 939
+    },
+    {
+      "epoch": 0.9149902660609993,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.9514045716477665e-06,
+      "loss": 1.3487,
+      "step": 940
+    },
+    {
+      "epoch": 0.9159636599610642,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.949321481463114e-06,
+      "loss": 1.3587,
+      "step": 941
+    },
+    {
+      "epoch": 0.9169370538611291,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.947236874560591e-06,
+      "loss": 1.3468,
+      "step": 942
+    },
+    {
+      "epoch": 0.917910447761194,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.945150753121742e-06,
+      "loss": 1.3397,
+      "step": 943
+    },
+    {
+      "epoch": 0.9188838416612589,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.9430631193296976e-06,
+      "loss": 1.3269,
+      "step": 944
+    },
+    {
+      "epoch": 0.9198572355613238,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.940973975369171e-06,
+      "loss": 1.3579,
+      "step": 945
+    },
+    {
+      "epoch": 0.9208306294613887,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.938883323426456e-06,
+      "loss": 1.3617,
+      "step": 946
+    },
+    {
+      "epoch": 0.9218040233614536,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.936791165689424e-06,
+      "loss": 1.3583,
+      "step": 947
+    },
+    {
+      "epoch": 0.9227774172615185,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.9346975043475216e-06,
+      "loss": 1.3346,
+      "step": 948
+    },
+    {
+      "epoch": 0.9237508111615834,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.9326023415917704e-06,
+      "loss": 1.334,
+      "step": 949
+    },
+    {
+      "epoch": 0.9247242050616483,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.930505679614762e-06,
+      "loss": 1.3414,
+      "step": 950
+    },
+    {
+      "epoch": 0.9256975989617132,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.928407520610658e-06,
+      "loss": 1.3375,
+      "step": 951
+    },
+    {
+      "epoch": 0.9266709928617781,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.926307866775186e-06,
+      "loss": 1.3527,
+      "step": 952
+    },
+    {
+      "epoch": 0.9276443867618429,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.924206720305638e-06,
+      "loss": 1.3498,
+      "step": 953
+    },
+    {
+      "epoch": 0.9286177806619078,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.922104083400866e-06,
+      "loss": 1.3638,
+      "step": 954
+    },
+    {
+      "epoch": 0.9295911745619727,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.919999958261287e-06,
+      "loss": 1.334,
+      "step": 955
+    },
+    {
+      "epoch": 0.9305645684620376,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.91789434708887e-06,
+      "loss": 1.3619,
+      "step": 956
+    },
+    {
+      "epoch": 0.9315379623621025,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.915787252087143e-06,
+      "loss": 1.3265,
+      "step": 957
+    },
+    {
+      "epoch": 0.9325113562621674,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.913678675461184e-06,
+      "loss": 1.345,
+      "step": 958
+    },
+    {
+      "epoch": 0.9334847501622323,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.9115686194176225e-06,
+      "loss": 1.3243,
+      "step": 959
+    },
+    {
+      "epoch": 0.9344581440622972,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.909457086164638e-06,
+      "loss": 1.33,
+      "step": 960
+    },
+    {
+      "epoch": 0.9354315379623621,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.907344077911952e-06,
+      "loss": 1.3443,
+      "step": 961
+    },
+    {
+      "epoch": 0.936404931862427,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.905229596870833e-06,
+      "loss": 1.3452,
+      "step": 962
+    },
+    {
+      "epoch": 0.9373783257624919,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.9031136452540915e-06,
+      "loss": 1.3114,
+      "step": 963
+    },
+    {
+      "epoch": 0.9383517196625568,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.900996225276073e-06,
+      "loss": 1.3469,
+      "step": 964
+    },
+    {
+      "epoch": 0.9393251135626217,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.8988773391526626e-06,
+      "loss": 1.3281,
+      "step": 965
+    },
+    {
+      "epoch": 0.9402985074626866,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.896756989101278e-06,
+      "loss": 1.3396,
+      "step": 966
+    },
+    {
+      "epoch": 0.9412719013627514,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.894635177340871e-06,
+      "loss": 1.3279,
+      "step": 967
+    },
+    {
+      "epoch": 0.9422452952628163,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.892511906091921e-06,
+      "loss": 1.3368,
+      "step": 968
+    },
+    {
+      "epoch": 0.9432186891628812,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.890387177576437e-06,
+      "loss": 1.3565,
+      "step": 969
+    },
+    {
+      "epoch": 0.9441920830629461,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.88826099401795e-06,
+      "loss": 1.3467,
+      "step": 970
+    },
+    {
+      "epoch": 0.945165476963011,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.886133357641516e-06,
+      "loss": 1.351,
+      "step": 971
+    },
+    {
+      "epoch": 0.9461388708630759,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.884004270673711e-06,
+      "loss": 1.312,
+      "step": 972
+    },
+    {
+      "epoch": 0.9471122647631408,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.88187373534263e-06,
+      "loss": 1.334,
+      "step": 973
+    },
+    {
+      "epoch": 0.9480856586632057,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.879741753877881e-06,
+      "loss": 1.3522,
+      "step": 974
+    },
+    {
+      "epoch": 0.9490590525632706,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.877608328510587e-06,
+      "loss": 1.3301,
+      "step": 975
+    },
+    {
+      "epoch": 0.9500324464633355,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.875473461473383e-06,
+      "loss": 1.3216,
+      "step": 976
+    },
+    {
+      "epoch": 0.9510058403634004,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.873337155000409e-06,
+      "loss": 1.3288,
+      "step": 977
+    },
+    {
+      "epoch": 0.9519792342634653,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.871199411327318e-06,
+      "loss": 1.3373,
+      "step": 978
+    },
+    {
+      "epoch": 0.9529526281635302,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.86906023269126e-06,
+      "loss": 1.3399,
+      "step": 979
+    },
+    {
+      "epoch": 0.9539260220635951,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.866919621330892e-06,
+      "loss": 1.329,
+      "step": 980
+    },
+    {
+      "epoch": 0.95489941596366,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.864777579486366e-06,
+      "loss": 1.3293,
+      "step": 981
+    },
+    {
+      "epoch": 0.9558728098637248,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.8626341093993346e-06,
+      "loss": 1.3483,
+      "step": 982
+    },
+    {
+      "epoch": 0.9568462037637897,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.860489213312943e-06,
+      "loss": 1.3413,
+      "step": 983
+    },
+    {
+      "epoch": 0.9578195976638546,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.85834289347183e-06,
+      "loss": 1.3715,
+      "step": 984
+    },
+    {
+      "epoch": 0.9587929915639195,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.856195152122123e-06,
+      "loss": 1.3392,
+      "step": 985
+    },
+    {
+      "epoch": 0.9597663854639844,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.854045991511438e-06,
+      "loss": 1.3204,
+      "step": 986
+    },
+    {
+      "epoch": 0.9607397793640493,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.851895413888875e-06,
+      "loss": 1.3382,
+      "step": 987
+    },
+    {
+      "epoch": 0.9617131732641142,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.849743421505019e-06,
+      "loss": 1.347,
+      "step": 988
+    },
+    {
+      "epoch": 0.9626865671641791,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.847590016611934e-06,
+      "loss": 1.3517,
+      "step": 989
+    },
+    {
+      "epoch": 0.963659961064244,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.8454352014631624e-06,
+      "loss": 1.3339,
+      "step": 990
+    },
+    {
+      "epoch": 0.9646333549643089,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.843278978313724e-06,
+      "loss": 1.3211,
+      "step": 991
+    },
+    {
+      "epoch": 0.9656067488643738,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.841121349420109e-06,
+      "loss": 1.3338,
+      "step": 992
+    },
+    {
+      "epoch": 0.9665801427644387,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.83896231704028e-06,
+      "loss": 1.3416,
+      "step": 993
+    },
+    {
+      "epoch": 0.9675535366645036,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.8368018834336694e-06,
+      "loss": 1.3275,
+      "step": 994
+    },
+    {
+      "epoch": 0.9685269305645685,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.834640050861177e-06,
+      "loss": 1.3369,
+      "step": 995
+    },
+    {
+      "epoch": 0.9695003244646333,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.832476821585164e-06,
+      "loss": 1.333,
+      "step": 996
+    },
+    {
+      "epoch": 0.9704737183646982,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.830312197869453e-06,
+      "loss": 1.3361,
+      "step": 997
+    },
+    {
+      "epoch": 0.9714471122647631,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.828146181979327e-06,
+      "loss": 1.3312,
+      "step": 998
+    },
+    {
+      "epoch": 0.972420506164828,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.825978776181528e-06,
+      "loss": 1.3344,
+      "step": 999
+    },
+    {
+      "epoch": 0.9733939000648929,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.8238099827442494e-06,
+      "loss": 1.3319,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9743672939649578,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.821639803937138e-06,
+      "loss": 1.3529,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9753406878650227,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.819468242031291e-06,
+      "loss": 1.3247,
+      "step": 1002
+    },
+    {
+      "epoch": 0.9763140817650876,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.8172952992992515e-06,
+      "loss": 1.3332,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9772874756651525,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.815120978015008e-06,
+      "loss": 1.3292,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9782608695652174,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.8129452804539934e-06,
+      "loss": 1.3356,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9792342634652823,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.8107682088930797e-06,
+      "loss": 1.3534,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9802076573653472,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.808589765610575e-06,
+      "loss": 1.3265,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9811810512654121,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.806409952886226e-06,
+      "loss": 1.2973,
+      "step": 1008
+    },
+    {
+      "epoch": 0.982154445165477,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.8042287730012117e-06,
+      "loss": 1.3629,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9831278390655419,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.8020462282381397e-06,
+      "loss": 1.3478,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9841012329656067,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.799862320881048e-06,
+      "loss": 1.3204,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9850746268656716,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.7976770532154006e-06,
+      "loss": 1.3189,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9860480207657365,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.7954904275280844e-06,
+      "loss": 1.333,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9870214146658014,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.7933024461074075e-06,
+      "loss": 1.3276,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9879948085658663,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.7911131112430966e-06,
+      "loss": 1.3269,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9889682024659312,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.7889224252262956e-06,
+      "loss": 1.3074,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9899415963659961,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.786730390349561e-06,
+      "loss": 1.3545,
+      "step": 1017
+    },
+    {
+      "epoch": 0.990914990266061,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.7845370089068626e-06,
+      "loss": 1.318,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9918883841661259,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.7823422831935796e-06,
+      "loss": 1.3276,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9928617780661908,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.780146215506494e-06,
+      "loss": 1.3135,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9938351719662557,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.777948808143797e-06,
+      "loss": 1.3509,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9948085658663206,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.7757500634050797e-06,
+      "loss": 1.3448,
+      "step": 1022
+    },
+    {
+      "epoch": 0.9957819597663855,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.7735499835913324e-06,
+      "loss": 1.3266,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9967553536664504,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.7713485710049445e-06,
+      "loss": 1.3315,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9977287475665152,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.769145827949697e-06,
+      "loss": 1.3565,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9987021414665801,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.766941756730766e-06,
+      "loss": 1.3316,
+      "step": 1026
+    },
+    {
+      "epoch": 0.999675535366645,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.764736359654716e-06,
+      "loss": 1.3254,
+      "step": 1027
+    },
+    {
+      "epoch": 1.00064892926671,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.7625296390294996e-06,
+      "loss": 1.3144,
+      "step": 1028
+    },
+    {
+      "epoch": 1.001622323166775,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.7603215971644545e-06,
+      "loss": 1.3256,
+      "step": 1029
+    },
+    {
+      "epoch": 1.001622323166775,
+      "eval_loss": 1.3576956987380981,
+      "eval_runtime": 1522.156,
+      "eval_samples_per_second": 27.431,
+      "eval_steps_per_second": 3.429,
+      "step": 1029
+    },
+    {
+      "epoch": 1.0005680204487362,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.7581122363703016e-06,
+      "loss": 1.3287,
+      "step": 1030
+    },
+    {
+      "epoch": 1.0015417697894267,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.7559015589591397e-06,
+      "loss": 1.3425,
+      "step": 1031
+    },
+    {
+      "epoch": 1.0025155191301172,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.753689567244449e-06,
+      "loss": 1.3431,
+      "step": 1032
+    },
+    {
+      "epoch": 1.0034892684708079,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.751476263541083e-06,
+      "loss": 1.3238,
+      "step": 1033
+    },
+    {
+      "epoch": 1.0044630178114984,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.7492616501652674e-06,
+      "loss": 1.3389,
+      "step": 1034
+    },
+    {
+      "epoch": 1.0054367671521889,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.7470457294346012e-06,
+      "loss": 1.3284,
+      "step": 1035
+    },
+    {
+      "epoch": 1.0064105164928794,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.744828503668049e-06,
+      "loss": 1.3488,
+      "step": 1036
+    },
+    {
+      "epoch": 1.00738426583357,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.7426099751859413e-06,
+      "loss": 1.3519,
+      "step": 1037
+    },
+    {
+      "epoch": 1.0083580151742606,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.7403901463099745e-06,
+      "loss": 1.3461,
+      "step": 1038
+    },
+    {
+      "epoch": 1.009331764514951,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.7381690193632027e-06,
+      "loss": 1.3512,
+      "step": 1039
+    },
+    {
+      "epoch": 1.0103055138556416,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.7359465966700405e-06,
+      "loss": 1.3379,
+      "step": 1040
+    },
+    {
+      "epoch": 1.011279263196332,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.7337228805562566e-06,
+      "loss": 1.3377,
+      "step": 1041
+    },
+    {
+      "epoch": 1.0122530125370228,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.7314978733489754e-06,
+      "loss": 1.3243,
+      "step": 1042
+    },
+    {
+      "epoch": 1.0132267618777133,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.7292715773766715e-06,
+      "loss": 1.3242,
+      "step": 1043
+    },
+    {
+      "epoch": 1.0142005112184038,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.7270439949691677e-06,
+      "loss": 1.3469,
+      "step": 1044
+    },
+    {
+      "epoch": 1.0151742605590943,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.7248151284576347e-06,
+      "loss": 1.3463,
+      "step": 1045
+    },
+    {
+      "epoch": 1.016148009899785,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.7225849801745835e-06,
+      "loss": 1.3616,
+      "step": 1046
+    },
+    {
+      "epoch": 1.0171217592404755,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.7203535524538704e-06,
+      "loss": 1.317,
+      "step": 1047
+    },
+    {
+      "epoch": 1.018095508581166,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.7181208476306892e-06,
+      "loss": 1.3045,
+      "step": 1048
+    },
+    {
+      "epoch": 1.0190692579218565,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.71588686804157e-06,
+      "loss": 1.3566,
+      "step": 1049
+    },
+    {
+      "epoch": 1.0200430072625473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.713651616024376e-06,
+      "loss": 1.3379,
+      "step": 1050
+    },
+    {
+      "epoch": 1.0210167566032378,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.711415093918303e-06,
+      "loss": 1.3284,
+      "step": 1051
+    },
+    {
+      "epoch": 1.0219905059439283,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.709177304063877e-06,
+      "loss": 1.3292,
+      "step": 1052
+    },
+    {
+      "epoch": 1.0229642552846188,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.7069382488029494e-06,
+      "loss": 1.3282,
+      "step": 1053
+    },
+    {
+      "epoch": 1.0239380046253093,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.7046979304786958e-06,
+      "loss": 1.3306,
+      "step": 1054
+    },
+    {
+      "epoch": 1.024911753966,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.7024563514356132e-06,
+      "loss": 1.3347,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0258855033066905,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.7002135140195193e-06,
+      "loss": 1.3151,
+      "step": 1056
+    },
+    {
+      "epoch": 1.026859252647381,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.6979694205775486e-06,
+      "loss": 1.3277,
+      "step": 1057
+    },
+    {
+      "epoch": 1.0278330019880715,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.695724073458149e-06,
+      "loss": 1.3234,
+      "step": 1058
+    },
+    {
+      "epoch": 1.0288067513287622,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.6934774750110796e-06,
+      "loss": 1.3497,
+      "step": 1059
+    },
+    {
+      "epoch": 1.0297805006694527,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.691229627587412e-06,
+      "loss": 1.3346,
+      "step": 1060
+    },
+    {
+      "epoch": 1.0307542500101432,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.6889805335395216e-06,
+      "loss": 1.3188,
+      "step": 1061
+    },
+    {
+      "epoch": 1.0317279993508337,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.686730195221091e-06,
+      "loss": 1.3469,
+      "step": 1062
+    },
+    {
+      "epoch": 1.0327017486915244,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.6844786149871044e-06,
+      "loss": 1.347,
+      "step": 1063
+    },
+    {
+      "epoch": 1.033675498032215,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.6822257951938432e-06,
+      "loss": 1.3236,
+      "step": 1064
+    },
+    {
+      "epoch": 1.0346492473729054,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.6799717381988885e-06,
+      "loss": 1.3313,
+      "step": 1065
+    },
+    {
+      "epoch": 1.035622996713596,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.677716446361116e-06,
+      "loss": 1.3147,
+      "step": 1066
+    },
+    {
+      "epoch": 1.0365967460542864,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.6754599220406935e-06,
+      "loss": 1.3147,
+      "step": 1067
+    },
+    {
+      "epoch": 1.0375704953949771,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.673202167599078e-06,
+      "loss": 1.3392,
+      "step": 1068
+    },
+    {
+      "epoch": 1.0385442447356676,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.670943185399013e-06,
+      "loss": 1.3457,
+      "step": 1069
+    },
+    {
+      "epoch": 1.0395179940763581,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.668682977804529e-06,
+      "loss": 1.3177,
+      "step": 1070
+    },
+    {
+      "epoch": 1.0404917434170486,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.666421547180938e-06,
+      "loss": 1.3379,
+      "step": 1071
+    },
+    {
+      "epoch": 1.0414654927577394,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.6641588958948315e-06,
+      "loss": 1.3175,
+      "step": 1072
+    },
+    {
+      "epoch": 1.0424392420984299,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.6618950263140785e-06,
+      "loss": 1.3534,
+      "step": 1073
+    },
+    {
+      "epoch": 1.0434129914391204,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.6596299408078228e-06,
+      "loss": 1.349,
+      "step": 1074
+    },
+    {
+      "epoch": 1.0443867407798109,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.657363641746482e-06,
+      "loss": 1.3124,
+      "step": 1075
+    },
+    {
+      "epoch": 1.0453604901205016,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.6550961315017412e-06,
+      "loss": 1.3346,
+      "step": 1076
+    },
+    {
+      "epoch": 1.046334239461192,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.652827412446556e-06,
+      "loss": 1.3266,
+      "step": 1077
+    },
+    {
+      "epoch": 1.0473079888018826,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.650557486955145e-06,
+      "loss": 1.3395,
+      "step": 1078
+    },
+    {
+      "epoch": 1.048281738142573,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.6482863574029893e-06,
+      "loss": 1.3218,
+      "step": 1079
+    },
+    {
+      "epoch": 1.0492554874832636,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.646014026166831e-06,
+      "loss": 1.3231,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0502292368239543,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.6437404956246686e-06,
+      "loss": 1.3384,
+      "step": 1081
+    },
+    {
+      "epoch": 1.0512029861646448,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.6414657681557576e-06,
+      "loss": 1.3292,
+      "step": 1082
+    },
+    {
+      "epoch": 1.0521767355053353,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.6391898461406045e-06,
+      "loss": 1.3263,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0531504848460258,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.636912731960966e-06,
+      "loss": 1.3308,
+      "step": 1084
+    },
+    {
+      "epoch": 1.0541242341867165,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.634634427999847e-06,
+      "loss": 1.3353,
+      "step": 1085
+    },
+    {
+      "epoch": 1.055097983527407,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.632354936641497e-06,
+      "loss": 1.3165,
+      "step": 1086
+    },
+    {
+      "epoch": 1.0560717328680975,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.630074260271409e-06,
+      "loss": 1.3247,
+      "step": 1087
+    },
+    {
+      "epoch": 1.057045482208788,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.6277924012763145e-06,
+      "loss": 1.346,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0580192315494787,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.6255093620441835e-06,
+      "loss": 1.3533,
+      "step": 1089
+    },
+    {
+      "epoch": 1.0589929808901692,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.6232251449642225e-06,
+      "loss": 1.3351,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0599667302308597,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.620939752426868e-06,
+      "loss": 1.3338,
+      "step": 1091
+    },
+    {
+      "epoch": 1.0609404795715502,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.6186531868237882e-06,
+      "loss": 1.3226,
+      "step": 1092
+    },
+    {
+      "epoch": 1.0619142289122407,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.6163654505478796e-06,
+      "loss": 1.3064,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0628879782529315,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.6140765459932615e-06,
+      "loss": 1.3237,
+      "step": 1094
+    },
+    {
+      "epoch": 1.063861727593622,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.611786475555278e-06,
+      "loss": 1.3342,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0648354769343125,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.6094952416304923e-06,
+      "loss": 1.3248,
+      "step": 1096
+    },
+    {
+      "epoch": 1.065809226275003,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.607202846616685e-06,
+      "loss": 1.3255,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0667829756156937,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.6049092929128527e-06,
+      "loss": 1.322,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0677567249563842,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.6026145829192033e-06,
+      "loss": 1.3385,
+      "step": 1099
+    },
+    {
+      "epoch": 1.0687304742970747,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.600318719037156e-06,
+      "loss": 1.3149,
+      "step": 1100
+    },
+    {
+      "epoch": 1.0697042236377652,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.598021703669337e-06,
+      "loss": 1.3329,
+      "step": 1101
+    },
+    {
+      "epoch": 1.070677972978456,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.5957235392195777e-06,
+      "loss": 1.3234,
+      "step": 1102
+    },
+    {
+      "epoch": 1.0716517223191464,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.593424228092911e-06,
+      "loss": 1.3348,
+      "step": 1103
+    },
+    {
+      "epoch": 1.072625471659837,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.5911237726955717e-06,
+      "loss": 1.3505,
+      "step": 1104
+    },
+    {
+      "epoch": 1.0735992210005274,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.58882217543499e-06,
+      "loss": 1.3292,
+      "step": 1105
+    },
+    {
+      "epoch": 1.074572970341218,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.5865194387197932e-06,
+      "loss": 1.321,
+      "step": 1106
+    },
+    {
+      "epoch": 1.0755467196819086,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.5842155649597993e-06,
+      "loss": 1.2894,
+      "step": 1107
+    },
+    {
+      "epoch": 1.0765204690225991,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.581910556566017e-06,
+      "loss": 1.3462,
+      "step": 1108
+    },
+    {
+      "epoch": 1.0774942183632896,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.579604415950642e-06,
+      "loss": 1.3344,
+      "step": 1109
+    },
+    {
+      "epoch": 1.0784679677039801,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.5772971455270554e-06,
+      "loss": 1.3305,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0794417170446708,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.5749887477098213e-06,
+      "loss": 1.3326,
+      "step": 1111
+    },
+    {
+      "epoch": 1.0804154663853613,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.5726792249146814e-06,
+      "loss": 1.331,
+      "step": 1112
+    },
+    {
+      "epoch": 1.0813892157260518,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.5703685795585558e-06,
+      "loss": 1.3332,
+      "step": 1113
+    },
+    {
+      "epoch": 1.0823629650667423,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.56805681405954e-06,
+      "loss": 1.3129,
+      "step": 1114
+    },
+    {
+      "epoch": 1.0833367144074328,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.565743930836902e-06,
+      "loss": 1.323,
+      "step": 1115
+    },
+    {
+      "epoch": 1.0843104637481236,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.5634299323110777e-06,
+      "loss": 1.3322,
+      "step": 1116
+    },
+    {
+      "epoch": 1.085284213088814,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.5611148209036716e-06,
+      "loss": 1.3294,
+      "step": 1117
+    },
+    {
+      "epoch": 1.0862579624295046,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.5587985990374535e-06,
+      "loss": 1.3162,
+      "step": 1118
+    },
+    {
+      "epoch": 1.087231711770195,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.5564812691363527e-06,
+      "loss": 1.3204,
+      "step": 1119
+    },
+    {
+      "epoch": 1.0882054611108858,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.55416283362546e-06,
+      "loss": 1.3289,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0891792104515763,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.551843294931024e-06,
+      "loss": 1.3371,
+      "step": 1121
+    },
+    {
+      "epoch": 1.0901529597922668,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.5495226554804453e-06,
+      "loss": 1.3267,
+      "step": 1122
+    },
+    {
+      "epoch": 1.0911267091329573,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.547200917702279e-06,
+      "loss": 1.3404,
+      "step": 1123
+    },
+    {
+      "epoch": 1.092100458473648,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.5448780840262274e-06,
+      "loss": 1.3423,
+      "step": 1124
+    },
+    {
+      "epoch": 1.0930742078143385,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.5425541568831406e-06,
+      "loss": 1.3314,
+      "step": 1125
+    },
+    {
+      "epoch": 1.094047957155029,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.5402291387050135e-06,
+      "loss": 1.3427,
+      "step": 1126
+    },
+    {
+      "epoch": 1.0950217064957195,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.537903031924983e-06,
+      "loss": 1.3438,
+      "step": 1127
+    },
+    {
+      "epoch": 1.0959954558364102,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.535575838977323e-06,
+      "loss": 1.3431,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0969692051771007,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.5332475622974477e-06,
+      "loss": 1.3289,
+      "step": 1129
+    },
+    {
+      "epoch": 1.0979429545177912,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.5309182043219016e-06,
+      "loss": 1.3165,
+      "step": 1130
+    },
+    {
+      "epoch": 1.0989167038584817,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.528587767488363e-06,
+      "loss": 1.3249,
+      "step": 1131
+    },
+    {
+      "epoch": 1.0998904531991722,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.526256254235638e-06,
+      "loss": 1.3293,
+      "step": 1132
+    },
+    {
+      "epoch": 1.100864202539863,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.523923667003662e-06,
+      "loss": 1.3224,
+      "step": 1133
+    },
+    {
+      "epoch": 1.1018379518805534,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.521590008233491e-06,
+      "loss": 1.3258,
+      "step": 1134
+    },
+    {
+      "epoch": 1.102811701221244,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.5192552803673043e-06,
+      "loss": 1.3404,
+      "step": 1135
+    },
+    {
+      "epoch": 1.1037854505619344,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.516919485848398e-06,
+      "loss": 1.322,
+      "step": 1136
+    },
+    {
+      "epoch": 1.1047591999026252,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.5145826271211862e-06,
+      "loss": 1.3295,
+      "step": 1137
+    },
+    {
+      "epoch": 1.1057329492433157,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.5122447066311973e-06,
+      "loss": 1.3475,
+      "step": 1138
+    },
+    {
+      "epoch": 1.1067066985840062,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.5099057268250692e-06,
+      "loss": 1.3253,
+      "step": 1139
+    },
+    {
+      "epoch": 1.1076804479246967,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.5075656901505495e-06,
+      "loss": 1.32,
+      "step": 1140
+    },
+    {
+      "epoch": 1.1086541972653872,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.5052245990564902e-06,
+      "loss": 1.3076,
+      "step": 1141
+    },
+    {
+      "epoch": 1.1096279466060779,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.5028824559928488e-06,
+      "loss": 1.3443,
+      "step": 1142
+    },
+    {
+      "epoch": 1.1106016959467684,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.5005392634106827e-06,
+      "loss": 1.3314,
+      "step": 1143
+    },
+    {
+      "epoch": 1.1115754452874589,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.4981950237621476e-06,
+      "loss": 1.3521,
+      "step": 1144
+    },
+    {
+      "epoch": 1.1125491946281494,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.4958497395004946e-06,
+      "loss": 1.3114,
+      "step": 1145
+    },
+    {
+      "epoch": 1.11352294396884,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.4935034130800693e-06,
+      "loss": 1.3359,
+      "step": 1146
+    },
+    {
+      "epoch": 1.1144966933095306,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.491156046956306e-06,
+      "loss": 1.3377,
+      "step": 1147
+    },
+    {
+      "epoch": 1.115470442650221,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.4888076435857286e-06,
+      "loss": 1.3571,
+      "step": 1148
+    },
+    {
+      "epoch": 1.1164441919909116,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.4864582054259474e-06,
+      "loss": 1.3289,
+      "step": 1149
+    },
+    {
+      "epoch": 1.1174179413316023,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.4841077349356518e-06,
+      "loss": 1.3388,
+      "step": 1150
+    },
+    {
+      "epoch": 1.1183916906722928,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.4817562345746146e-06,
+      "loss": 1.318,
+      "step": 1151
+    },
+    {
+      "epoch": 1.1193654400129833,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.4794037068036867e-06,
+      "loss": 1.3307,
+      "step": 1152
+    },
+    {
+      "epoch": 1.1203391893536738,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.4770501540847935e-06,
+      "loss": 1.3155,
+      "step": 1153
+    },
+    {
+      "epoch": 1.1213129386943643,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.4746955788809306e-06,
+      "loss": 1.3097,
+      "step": 1154
+    },
+    {
+      "epoch": 1.122286688035055,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.472339983656167e-06,
+      "loss": 1.3396,
+      "step": 1155
+    },
+    {
+      "epoch": 1.1232604373757455,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.469983370875638e-06,
+      "loss": 1.3431,
+      "step": 1156
+    },
+    {
+      "epoch": 1.124234186716436,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.4676257430055438e-06,
+      "loss": 1.3206,
+      "step": 1157
+    },
+    {
+      "epoch": 1.1252079360571265,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.4652671025131457e-06,
+      "loss": 1.3194,
+      "step": 1158
+    },
+    {
+      "epoch": 1.1261816853978173,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.4629074518667666e-06,
+      "loss": 1.3156,
+      "step": 1159
+    },
+    {
+      "epoch": 1.1271554347385078,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.4605467935357856e-06,
+      "loss": 1.3325,
+      "step": 1160
+    },
+    {
+      "epoch": 1.1281291840791983,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.4581851299906367e-06,
+      "loss": 1.326,
+      "step": 1161
+    },
+    {
+      "epoch": 1.1291029334198888,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.4558224637028055e-06,
+      "loss": 1.3283,
+      "step": 1162
+    },
+    {
+      "epoch": 1.1300766827605795,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.4534587971448265e-06,
+      "loss": 1.3237,
+      "step": 1163
+    },
+    {
+      "epoch": 1.13105043210127,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.4510941327902815e-06,
+      "loss": 1.327,
+      "step": 1164
+    },
+    {
+      "epoch": 1.1320241814419605,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.448728473113797e-06,
+      "loss": 1.3273,
+      "step": 1165
+    },
+    {
+      "epoch": 1.132997930782651,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.446361820591041e-06,
+      "loss": 1.3341,
+      "step": 1166
+    },
+    {
+      "epoch": 1.1339716801233415,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.4439941776987195e-06,
+      "loss": 1.3253,
+      "step": 1167
+    },
+    {
+      "epoch": 1.1349454294640322,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.4416255469145765e-06,
+      "loss": 1.3398,
+      "step": 1168
+    },
+    {
+      "epoch": 1.1359191788047227,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.4392559307173876e-06,
+      "loss": 1.3284,
+      "step": 1169
+    },
+    {
+      "epoch": 1.1368929281454132,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.4368853315869616e-06,
+      "loss": 1.3433,
+      "step": 1170
+    },
+    {
+      "epoch": 1.1378666774861037,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.4345137520041354e-06,
+      "loss": 1.3196,
+      "step": 1171
+    },
+    {
+      "epoch": 1.1388404268267944,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.432141194450772e-06,
+      "loss": 1.3346,
+      "step": 1172
+    },
+    {
+      "epoch": 1.139814176167485,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.4297676614097573e-06,
+      "loss": 1.3232,
+      "step": 1173
+    },
+    {
+      "epoch": 1.1407879255081754,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.4273931553649986e-06,
+      "loss": 1.3498,
+      "step": 1174
+    },
+    {
+      "epoch": 1.141761674848866,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.4250176788014217e-06,
+      "loss": 1.349,
+      "step": 1175
+    },
+    {
+      "epoch": 1.1427354241895564,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.4226412342049662e-06,
+      "loss": 1.3685,
+      "step": 1176
+    },
+    {
+      "epoch": 1.1437091735302471,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.420263824062588e-06,
+      "loss": 1.3228,
+      "step": 1177
+    },
+    {
+      "epoch": 1.1446829228709376,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.4178854508622506e-06,
+      "loss": 1.3361,
+      "step": 1178
+    },
+    {
+      "epoch": 1.1456566722116281,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.415506117092927e-06,
+      "loss": 1.3417,
+      "step": 1179
+    },
+    {
+      "epoch": 1.1466304215523186,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.4131258252445947e-06,
+      "loss": 1.3278,
+      "step": 1180
+    },
+    {
+      "epoch": 1.1476041708930094,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.4107445778082334e-06,
+      "loss": 1.3197,
+      "step": 1181
+    },
+    {
+      "epoch": 1.1485779202336999,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.4083623772758236e-06,
+      "loss": 1.3394,
+      "step": 1182
+    },
+    {
+      "epoch": 1.1495516695743904,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.4059792261403422e-06,
+      "loss": 1.3305,
+      "step": 1183
+    },
+    {
+      "epoch": 1.1505254189150809,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.4035951268957647e-06,
+      "loss": 1.3281,
+      "step": 1184
+    },
+    {
+      "epoch": 1.1514991682557716,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.401210082037052e-06,
+      "loss": 1.3526,
+      "step": 1185
+    },
+    {
+      "epoch": 1.152472917596462,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.3988240940601604e-06,
+      "loss": 1.3178,
+      "step": 1186
+    },
+    {
+      "epoch": 1.1534466669371526,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.3964371654620305e-06,
+      "loss": 1.3464,
+      "step": 1187
+    },
+    {
+      "epoch": 1.154420416277843,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.3940492987405888e-06,
+      "loss": 1.3251,
+      "step": 1188
+    },
+    {
+      "epoch": 1.1553941656185338,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.391660496394742e-06,
+      "loss": 1.3338,
+      "step": 1189
+    },
+    {
+      "epoch": 1.1563679149592243,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.389270760924377e-06,
+      "loss": 1.3217,
+      "step": 1190
+    },
+    {
+      "epoch": 1.1573416642999148,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.386880094830356e-06,
+      "loss": 1.3445,
+      "step": 1191
+    },
+    {
+      "epoch": 1.1583154136406053,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.3844885006145162e-06,
+      "loss": 1.3222,
+      "step": 1192
+    },
+    {
+      "epoch": 1.1592891629812958,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.382095980779666e-06,
+      "loss": 1.3295,
+      "step": 1193
+    },
+    {
+      "epoch": 1.1602629123219865,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.3797025378295826e-06,
+      "loss": 1.3269,
+      "step": 1194
+    },
+    {
+      "epoch": 1.161236661662677,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.3773081742690097e-06,
+      "loss": 1.3061,
+      "step": 1195
+    },
+    {
+      "epoch": 1.1622104110033675,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.374912892603651e-06,
+      "loss": 1.3263,
+      "step": 1196
+    },
+    {
+      "epoch": 1.163184160344058,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.372516695340176e-06,
+      "loss": 1.3231,
+      "step": 1197
+    },
+    {
+      "epoch": 1.1641579096847487,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.370119584986209e-06,
+      "loss": 1.3239,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1651316590254392,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.36772156405033e-06,
+      "loss": 1.3341,
+      "step": 1199
+    },
+    {
+      "epoch": 1.1661054083661297,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.365322635042075e-06,
+      "loss": 1.338,
+      "step": 1200
+    },
+    {
+      "epoch": 1.1670791577068202,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.362922800471927e-06,
+      "loss": 1.3264,
+      "step": 1201
+    },
+    {
+      "epoch": 1.1680529070475107,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.360522062851317e-06,
+      "loss": 1.3335,
+      "step": 1202
+    },
+    {
+      "epoch": 1.1690266563882015,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.3581204246926223e-06,
+      "loss": 1.3147,
+      "step": 1203
+    },
+    {
+      "epoch": 1.170000405728892,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.3557178885091625e-06,
+      "loss": 1.3319,
+      "step": 1204
+    },
+    {
+      "epoch": 1.1709741550695825,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.3533144568151956e-06,
+      "loss": 1.3282,
+      "step": 1205
+    },
+    {
+      "epoch": 1.171947904410273,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.35091013212592e-06,
+      "loss": 1.325,
+      "step": 1206
+    },
+    {
+      "epoch": 1.1729216537509637,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.348504916957463e-06,
+      "loss": 1.3234,
+      "step": 1207
+    },
+    {
+      "epoch": 1.1738954030916542,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.3460988138268895e-06,
+      "loss": 1.3274,
+      "step": 1208
+    },
+    {
+      "epoch": 1.1748691524323447,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.343691825252191e-06,
+      "loss": 1.3238,
+      "step": 1209
+    },
+    {
+      "epoch": 1.1758429017730352,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.3412839537522858e-06,
+      "loss": 1.3196,
+      "step": 1210
+    },
+    {
+      "epoch": 1.1768166511137257,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.3388752018470156e-06,
+      "loss": 1.3278,
+      "step": 1211
+    },
+    {
+      "epoch": 1.1777904004544164,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.3364655720571453e-06,
+      "loss": 1.3312,
+      "step": 1212
+    },
+    {
+      "epoch": 1.178764149795107,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.334055066904357e-06,
+      "loss": 1.3238,
+      "step": 1213
+    },
+    {
+      "epoch": 1.1797378991357974,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.331643688911248e-06,
+      "loss": 1.3206,
+      "step": 1214
+    },
+    {
+      "epoch": 1.1807116484764881,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.329231440601332e-06,
+      "loss": 1.311,
+      "step": 1215
+    },
+    {
+      "epoch": 1.1816853978171786,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.3268183244990308e-06,
+      "loss": 1.3446,
+      "step": 1216
+    },
+    {
+      "epoch": 1.1826591471578691,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.324404343129676e-06,
+      "loss": 1.3251,
+      "step": 1217
+    },
+    {
+      "epoch": 1.1836328964985596,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.3219894990195036e-06,
+      "loss": 1.3382,
+      "step": 1218
+    },
+    {
+      "epoch": 1.1846066458392501,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.3195737946956525e-06,
+      "loss": 1.3156,
+      "step": 1219
+    },
+    {
+      "epoch": 1.1855803951799408,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.3171572326861624e-06,
+      "loss": 1.3352,
+      "step": 1220
+    },
+    {
+      "epoch": 1.1865541445206313,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.3147398155199706e-06,
+      "loss": 1.2905,
+      "step": 1221
+    },
+    {
+      "epoch": 1.1875278938613218,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.3123215457269086e-06,
+      "loss": 1.3017,
+      "step": 1222
+    },
+    {
+      "epoch": 1.1885016432020123,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.3099024258377017e-06,
+      "loss": 1.3375,
+      "step": 1223
+    },
+    {
+      "epoch": 1.189475392542703,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.307482458383962e-06,
+      "loss": 1.3315,
+      "step": 1224
+    },
+    {
+      "epoch": 1.1904491418833936,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.305061645898192e-06,
+      "loss": 1.3343,
+      "step": 1225
+    },
+    {
+      "epoch": 1.191422891224084,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.3026399909137757e-06,
+      "loss": 1.3336,
+      "step": 1226
+    },
+    {
+      "epoch": 1.1923966405647746,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.300217495964981e-06,
+      "loss": 1.3156,
+      "step": 1227
+    },
+    {
+      "epoch": 1.193370389905465,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.2977941635869537e-06,
+      "loss": 1.3331,
+      "step": 1228
+    },
+    {
+      "epoch": 1.1943441392461558,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.295369996315715e-06,
+      "loss": 1.3356,
+      "step": 1229
+    },
+    {
+      "epoch": 1.1953178885868463,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.292944996688161e-06,
+      "loss": 1.3177,
+      "step": 1230
+    },
+    {
+      "epoch": 1.1962916379275368,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.29051916724206e-06,
+      "loss": 1.3256,
+      "step": 1231
+    },
+    {
+      "epoch": 1.1972653872682273,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.2880925105160464e-06,
+      "loss": 1.3083,
+      "step": 1232
+    },
+    {
+      "epoch": 1.198239136608918,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.2856650290496216e-06,
+      "loss": 1.3297,
+      "step": 1233
+    },
+    {
+      "epoch": 1.1992128859496085,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.2832367253831506e-06,
+      "loss": 1.3139,
+      "step": 1234
+    },
+    {
+      "epoch": 1.200186635290299,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.280807602057857e-06,
+      "loss": 1.3298,
+      "step": 1235
+    },
+    {
+      "epoch": 1.2011603846309895,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.278377661615823e-06,
+      "loss": 1.3233,
+      "step": 1236
+    },
+    {
+      "epoch": 1.20213413397168,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.2759469065999877e-06,
+      "loss": 1.3144,
+      "step": 1237
+    },
+    {
+      "epoch": 1.2031078833123707,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.2735153395541403e-06,
+      "loss": 1.3305,
+      "step": 1238
+    },
+    {
+      "epoch": 1.2040816326530612,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.2710829630229208e-06,
+      "loss": 1.3223,
+      "step": 1239
+    },
+    {
+      "epoch": 1.2050553819937517,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.268649779551816e-06,
+      "loss": 1.3294,
+      "step": 1240
+    },
+    {
+      "epoch": 1.2060291313344425,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.266215791687157e-06,
+      "loss": 1.3111,
+      "step": 1241
+    },
+    {
+      "epoch": 1.207002880675133,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.263781001976118e-06,
+      "loss": 1.3286,
+      "step": 1242
+    },
+    {
+      "epoch": 1.2079766300158234,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.261345412966711e-06,
+      "loss": 1.3251,
+      "step": 1243
+    },
+    {
+      "epoch": 1.208950379356514,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.2589090272077845e-06,
+      "loss": 1.3358,
+      "step": 1244
+    },
+    {
+      "epoch": 1.2099241286972044,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.2564718472490215e-06,
+      "loss": 1.3325,
+      "step": 1245
+    },
+    {
+      "epoch": 1.2108978780378952,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.2540338756409353e-06,
+      "loss": 1.3274,
+      "step": 1246
+    },
+    {
+      "epoch": 1.2118716273785857,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.2515951149348683e-06,
+      "loss": 1.3338,
+      "step": 1247
+    },
+    {
+      "epoch": 1.2128453767192762,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.2491555676829894e-06,
+      "loss": 1.3228,
+      "step": 1248
+    },
+    {
+      "epoch": 1.2138191260599667,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.2467152364382883e-06,
+      "loss": 1.328,
+      "step": 1249
+    },
+    {
+      "epoch": 1.2147928754006574,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.244274123754578e-06,
+      "loss": 1.3115,
+      "step": 1250
+    },
+    {
+      "epoch": 1.215766624741348,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.2418322321864866e-06,
+      "loss": 1.3481,
+      "step": 1251
+    },
+    {
+      "epoch": 1.2167403740820384,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.2393895642894596e-06,
+      "loss": 1.3045,
+      "step": 1252
+    },
+    {
+      "epoch": 1.217714123422729,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.236946122619753e-06,
+      "loss": 1.304,
+      "step": 1253
+    },
+    {
+      "epoch": 1.2186878727634194,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.2345019097344336e-06,
+      "loss": 1.3004,
+      "step": 1254
+    },
+    {
+      "epoch": 1.21966162210411,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.232056928191376e-06,
+      "loss": 1.3315,
+      "step": 1255
+    },
+    {
+      "epoch": 1.2206353714448006,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.229611180549258e-06,
+      "loss": 1.303,
+      "step": 1256
+    },
+    {
+      "epoch": 1.221609120785491,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.2271646693675596e-06,
+      "loss": 1.3242,
+      "step": 1257
+    },
+    {
+      "epoch": 1.2225828701261816,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.2247173972065593e-06,
+      "loss": 1.2989,
+      "step": 1258
+    },
+    {
+      "epoch": 1.2235566194668723,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.2222693666273318e-06,
+      "loss": 1.336,
+      "step": 1259
+    },
+    {
+      "epoch": 1.2245303688075628,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.219820580191747e-06,
+      "loss": 1.336,
+      "step": 1260
+    },
+    {
+      "epoch": 1.2255041181482533,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.2173710404624646e-06,
+      "loss": 1.3578,
+      "step": 1261
+    },
+    {
+      "epoch": 1.2264778674889438,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.2149207500029337e-06,
+      "loss": 1.3048,
+      "step": 1262
+    },
+    {
+      "epoch": 1.2274516168296343,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.212469711377387e-06,
+      "loss": 1.3316,
+      "step": 1263
+    },
+    {
+      "epoch": 1.228425366170325,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.210017927150842e-06,
+      "loss": 1.3255,
+      "step": 1264
+    },
+    {
+      "epoch": 1.2293991155110155,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.2075653998890953e-06,
+      "loss": 1.3153,
+      "step": 1265
+    },
+    {
+      "epoch": 1.230372864851706,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.205112132158722e-06,
+      "loss": 1.3091,
+      "step": 1266
+    },
+    {
+      "epoch": 1.2313466141923965,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.202658126527073e-06,
+      "loss": 1.3172,
+      "step": 1267
+    },
+    {
+      "epoch": 1.2323203635330873,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.2002033855622683e-06,
+      "loss": 1.3196,
+      "step": 1268
+    },
+    {
+      "epoch": 1.2332941128737778,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.1977479118331994e-06,
+      "loss": 1.3345,
+      "step": 1269
+    },
+    {
+      "epoch": 1.2342678622144683,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.195291707909526e-06,
+      "loss": 1.3345,
+      "step": 1270
+    },
+    {
+      "epoch": 1.2352416115551588,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.192834776361669e-06,
+      "loss": 1.3376,
+      "step": 1271
+    },
+    {
+      "epoch": 1.2362153608958493,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.190377119760813e-06,
+      "loss": 1.332,
+      "step": 1272
+    },
+    {
+      "epoch": 1.23718911023654,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.1879187406789014e-06,
+      "loss": 1.3407,
+      "step": 1273
+    },
+    {
+      "epoch": 1.2381628595772305,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.1854596416886313e-06,
+      "loss": 1.3402,
+      "step": 1274
+    },
+    {
+      "epoch": 1.239136608917921,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.1829998253634554e-06,
+      "loss": 1.3287,
+      "step": 1275
+    },
+    {
+      "epoch": 1.2401103582586117,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.180539294277577e-06,
+      "loss": 1.3168,
+      "step": 1276
+    },
+    {
+      "epoch": 1.2410841075993022,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.178078051005946e-06,
+      "loss": 1.3151,
+      "step": 1277
+    },
+    {
+      "epoch": 1.2420578569399927,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.1756160981242596e-06,
+      "loss": 1.3262,
+      "step": 1278
+    },
+    {
+      "epoch": 1.2430316062806832,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.1731534382089552e-06,
+      "loss": 1.3139,
+      "step": 1279
+    },
+    {
+      "epoch": 1.2440053556213737,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.170690073837212e-06,
+      "loss": 1.3259,
+      "step": 1280
+    },
+    {
+      "epoch": 1.2449791049620644,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.168226007586946e-06,
+      "loss": 1.3313,
+      "step": 1281
+    },
+    {
+      "epoch": 1.245952854302755,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.1657612420368074e-06,
+      "loss": 1.3243,
+      "step": 1282
+    },
+    {
+      "epoch": 1.2469266036434454,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.163295779766178e-06,
+      "loss": 1.3218,
+      "step": 1283
+    },
+    {
+      "epoch": 1.247900352984136,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.16082962335517e-06,
+      "loss": 1.3141,
+      "step": 1284
+    },
+    {
+      "epoch": 1.2488741023248267,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.15836277538462e-06,
+      "loss": 1.3284,
+      "step": 1285
+    },
+    {
+      "epoch": 1.2498478516655172,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.15589523843609e-06,
+      "loss": 1.3417,
+      "step": 1286
+    },
+    {
+      "epoch": 1.2508216010062077,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.1534270150918616e-06,
+      "loss": 1.3158,
+      "step": 1287
+    },
+    {
+      "epoch": 1.2517953503468982,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.1509581079349373e-06,
+      "loss": 1.3368,
+      "step": 1288
+    },
+    {
+      "epoch": 1.2527690996875886,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.1484885195490323e-06,
+      "loss": 1.3687,
+      "step": 1289
+    },
+    {
+      "epoch": 1.2537428490282794,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.146018252518575e-06,
+      "loss": 1.3235,
+      "step": 1290
+    },
+    {
+      "epoch": 1.2547165983689699,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.1435473094287063e-06,
+      "loss": 1.3594,
+      "step": 1291
+    },
+    {
+      "epoch": 1.2556903477096604,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.1410756928652723e-06,
+      "loss": 1.3163,
+      "step": 1292
+    },
+    {
+      "epoch": 1.256664097050351,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.1386034054148258e-06,
+      "loss": 1.324,
+      "step": 1293
+    },
+    {
+      "epoch": 1.2576378463910416,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.1361304496646195e-06,
+      "loss": 1.3288,
+      "step": 1294
+    },
+    {
+      "epoch": 1.258611595731732,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.133656828202607e-06,
+      "loss": 1.3164,
+      "step": 1295
+    },
+    {
+      "epoch": 1.2595853450724226,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.131182543617439e-06,
+      "loss": 1.3116,
+      "step": 1296
+    },
+    {
+      "epoch": 1.260559094413113,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.1287075984984573e-06,
+      "loss": 1.3235,
+      "step": 1297
+    },
+    {
+      "epoch": 1.2615328437538036,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.126231995435699e-06,
+      "loss": 1.3236,
+      "step": 1298
+    },
+    {
+      "epoch": 1.2625065930944943,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.123755737019886e-06,
+      "loss": 1.3198,
+      "step": 1299
+    },
+    {
+      "epoch": 1.2634803424351848,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.1212788258424304e-06,
+      "loss": 1.3232,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2644540917758753,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.118801264495422e-06,
+      "loss": 1.3241,
+      "step": 1301
+    },
+    {
+      "epoch": 1.265427841116566,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.116323055571635e-06,
+      "loss": 1.3101,
+      "step": 1302
+    },
+    {
+      "epoch": 1.2664015904572565,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.1138442016645197e-06,
+      "loss": 1.3365,
+      "step": 1303
+    },
+    {
+      "epoch": 1.267375339797947,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.1113647053682024e-06,
+      "loss": 1.3186,
+      "step": 1304
+    },
+    {
+      "epoch": 1.2683490891386375,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.1088845692774798e-06,
+      "loss": 1.3203,
+      "step": 1305
+    },
+    {
+      "epoch": 1.269322838479328,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.106403795987821e-06,
+      "loss": 1.3171,
+      "step": 1306
+    },
+    {
+      "epoch": 1.2702965878200185,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.1039223880953593e-06,
+      "loss": 1.3087,
+      "step": 1307
+    },
+    {
+      "epoch": 1.2712703371607093,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.1014403481968936e-06,
+      "loss": 1.3124,
+      "step": 1308
+    },
+    {
+      "epoch": 1.2722440865013998,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.098957678889882e-06,
+      "loss": 1.3306,
+      "step": 1309
+    },
+    {
+      "epoch": 1.2732178358420903,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.0964743827724453e-06,
+      "loss": 1.3571,
+      "step": 1310
+    },
+    {
+      "epoch": 1.274191585182781,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.093990462443357e-06,
+      "loss": 1.3433,
+      "step": 1311
+    },
+    {
+      "epoch": 1.2751653345234715,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.0915059205020443e-06,
+      "loss": 1.3189,
+      "step": 1312
+    },
+    {
+      "epoch": 1.276139083864162,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.0890207595485856e-06,
+      "loss": 1.3024,
+      "step": 1313
+    },
+    {
+      "epoch": 1.2771128332048525,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.086534982183707e-06,
+      "loss": 1.3039,
+      "step": 1314
+    },
+    {
+      "epoch": 1.278086582545543,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.0840485910087797e-06,
+      "loss": 1.3412,
+      "step": 1315
+    },
+    {
+      "epoch": 1.2790603318862337,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.081561588625817e-06,
+      "loss": 1.3334,
+      "step": 1316
+    },
+    {
+      "epoch": 1.2800340812269242,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.079073977637472e-06,
+      "loss": 1.33,
+      "step": 1317
+    },
+    {
+      "epoch": 1.2810078305676147,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.0765857606470353e-06,
+      "loss": 1.329,
+      "step": 1318
+    },
+    {
+      "epoch": 1.2819815799083052,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.0740969402584287e-06,
+      "loss": 1.3007,
+      "step": 1319
+    },
+    {
+      "epoch": 1.282955329248996,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.0716075190762103e-06,
+      "loss": 1.3283,
+      "step": 1320
+    },
+    {
+      "epoch": 1.2839290785896864,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.0691174997055632e-06,
+      "loss": 1.3142,
+      "step": 1321
+    },
+    {
+      "epoch": 1.284902827930377,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.0666268847522984e-06,
+      "loss": 1.3155,
+      "step": 1322
+    },
+    {
+      "epoch": 1.2858765772710674,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.064135676822849e-06,
+      "loss": 1.3017,
+      "step": 1323
+    },
+    {
+      "epoch": 1.286850326611758,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.061643878524269e-06,
+      "loss": 1.3107,
+      "step": 1324
+    },
+    {
+      "epoch": 1.2878240759524486,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.059151492464231e-06,
+      "loss": 1.3271,
+      "step": 1325
+    },
+    {
+      "epoch": 1.2887978252931391,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.0566585212510213e-06,
+      "loss": 1.32,
+      "step": 1326
+    },
+    {
+      "epoch": 1.2897715746338296,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.05416496749354e-06,
+      "loss": 1.3155,
+      "step": 1327
+    },
+    {
+      "epoch": 1.2907453239745204,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.051670833801296e-06,
+      "loss": 1.3437,
+      "step": 1328
+    },
+    {
+      "epoch": 1.2917190733152109,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.0491761227844054e-06,
+      "loss": 1.313,
+      "step": 1329
+    },
+    {
+      "epoch": 1.2926928226559014,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.0466808370535878e-06,
+      "loss": 1.3255,
+      "step": 1330
+    },
+    {
+      "epoch": 1.2936665719965919,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.044184979220165e-06,
+      "loss": 1.3475,
+      "step": 1331
+    },
+    {
+      "epoch": 1.2946403213372824,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.041688551896057e-06,
+      "loss": 1.3305,
+      "step": 1332
+    },
+    {
+      "epoch": 1.2956140706779729,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.039191557693782e-06,
+      "loss": 1.2907,
+      "step": 1333
+    },
+    {
+      "epoch": 1.2965878200186636,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.0366939992264467e-06,
+      "loss": 1.3218,
+      "step": 1334
+    },
+    {
+      "epoch": 1.297561569359354,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.0341958791077524e-06,
+      "loss": 1.3171,
+      "step": 1335
+    },
+    {
+      "epoch": 1.2985353187000446,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.031697199951987e-06,
+      "loss": 1.3191,
+      "step": 1336
+    },
+    {
+      "epoch": 1.2995090680407353,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.0291979643740235e-06,
+      "loss": 1.3216,
+      "step": 1337
+    },
+    {
+      "epoch": 1.3004828173814258,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.026698174989316e-06,
+      "loss": 1.3069,
+      "step": 1338
+    },
+    {
+      "epoch": 1.3014565667221163,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.0241978344139e-06,
+      "loss": 1.3342,
+      "step": 1339
+    },
+    {
+      "epoch": 1.3024303160628068,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.021696945264387e-06,
+      "loss": 1.3129,
+      "step": 1340
+    },
+    {
+      "epoch": 1.3034040654034973,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.019195510157962e-06,
+      "loss": 1.3072,
+      "step": 1341
+    },
+    {
+      "epoch": 1.304377814744188,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.0166935317123824e-06,
+      "loss": 1.3239,
+      "step": 1342
+    },
+    {
+      "epoch": 1.3053515640848785,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.0141910125459744e-06,
+      "loss": 1.3281,
+      "step": 1343
+    },
+    {
+      "epoch": 1.306325313425569,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.011687955277628e-06,
+      "loss": 1.3332,
+      "step": 1344
+    },
+    {
+      "epoch": 1.3072990627662595,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.0091843625267975e-06,
+      "loss": 1.3286,
+      "step": 1345
+    },
+    {
+      "epoch": 1.3082728121069502,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.0066802369134994e-06,
+      "loss": 1.3131,
+      "step": 1346
+    },
+    {
+      "epoch": 1.3092465614476407,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.0041755810583057e-06,
+      "loss": 1.3039,
+      "step": 1347
+    },
+    {
+      "epoch": 1.3102203107883312,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.0016703975823424e-06,
+      "loss": 1.3248,
+      "step": 1348
+    },
+    {
+      "epoch": 1.3111940601290217,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.9991646891072913e-06,
+      "loss": 1.2892,
+      "step": 1349
+    },
+    {
+      "epoch": 1.3121678094697122,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.99665845825538e-06,
+      "loss": 1.3322,
+      "step": 1350
+    },
+    {
+      "epoch": 1.313141558810403,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.994151707649384e-06,
+      "loss": 1.3231,
+      "step": 1351
+    },
+    {
+      "epoch": 1.3141153081510935,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.9916444399126245e-06,
+      "loss": 1.3194,
+      "step": 1352
+    },
+    {
+      "epoch": 1.315089057491784,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.98913665766896e-06,
+      "loss": 1.3136,
+      "step": 1353
+    },
+    {
+      "epoch": 1.3160628068324747,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.986628363542792e-06,
+      "loss": 1.3007,
+      "step": 1354
+    },
+    {
+      "epoch": 1.3170365561731652,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.9841195601590526e-06,
+      "loss": 1.3282,
+      "step": 1355
+    },
+    {
+      "epoch": 1.3180103055138557,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.981610250143213e-06,
+      "loss": 1.3295,
+      "step": 1356
+    },
+    {
+      "epoch": 1.3189840548545462,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.9791004361212687e-06,
+      "loss": 1.3109,
+      "step": 1357
+    },
+    {
+      "epoch": 1.3199578041952367,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.976590120719746e-06,
+      "loss": 1.3077,
+      "step": 1358
+    },
+    {
+      "epoch": 1.3209315535359272,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.9740793065656953e-06,
+      "loss": 1.3384,
+      "step": 1359
+    },
+    {
+      "epoch": 1.321905302876618,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.9715679962866885e-06,
+      "loss": 1.2938,
+      "step": 1360
+    },
+    {
+      "epoch": 1.3228790522173084,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.9690561925108168e-06,
+      "loss": 1.3282,
+      "step": 1361
+    },
+    {
+      "epoch": 1.323852801557999,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.966543897866688e-06,
+      "loss": 1.3424,
+      "step": 1362
+    },
+    {
+      "epoch": 1.3248265508986896,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.9640311149834234e-06,
+      "loss": 1.3368,
+      "step": 1363
+    },
+    {
+      "epoch": 1.3258003002393801,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.9615178464906558e-06,
+      "loss": 1.318,
+      "step": 1364
+    },
+    {
+      "epoch": 1.3267740495800706,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.959004095018525e-06,
+      "loss": 1.3023,
+      "step": 1365
+    },
+    {
+      "epoch": 1.3277477989207611,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.9564898631976786e-06,
+      "loss": 1.306,
+      "step": 1366
+    },
+    {
+      "epoch": 1.3287215482614516,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.953975153659263e-06,
+      "loss": 1.3101,
+      "step": 1367
+    },
+    {
+      "epoch": 1.3296952976021421,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.9514599690349287e-06,
+      "loss": 1.3358,
+      "step": 1368
+    },
+    {
+      "epoch": 1.3306690469428328,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.9489443119568204e-06,
+      "loss": 1.3106,
+      "step": 1369
+    },
+    {
+      "epoch": 1.3316427962835233,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.9464281850575795e-06,
+      "loss": 1.3143,
+      "step": 1370
+    },
+    {
+      "epoch": 1.3326165456242138,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.9439115909703366e-06,
+      "loss": 1.3247,
+      "step": 1371
+    },
+    {
+      "epoch": 1.3335902949649046,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.9413945323287136e-06,
+      "loss": 1.3274,
+      "step": 1372
+    },
+    {
+      "epoch": 1.334564044305595,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.938877011766817e-06,
+      "loss": 1.3221,
+      "step": 1373
+    },
+    {
+      "epoch": 1.3355377936462856,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.936359031919237e-06,
+      "loss": 1.3119,
+      "step": 1374
+    },
+    {
+      "epoch": 1.336511542986976,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.9338405954210446e-06,
+      "loss": 1.3163,
+      "step": 1375
+    },
+    {
+      "epoch": 1.3374852923276666,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.9313217049077887e-06,
+      "loss": 1.3262,
+      "step": 1376
+    },
+    {
+      "epoch": 1.3384590416683573,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.928802363015494e-06,
+      "loss": 1.3322,
+      "step": 1377
+    },
+    {
+      "epoch": 1.3394327910090478,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.9262825723806563e-06,
+      "loss": 1.3207,
+      "step": 1378
+    },
+    {
+      "epoch": 1.3404065403497383,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.9237623356402423e-06,
+      "loss": 1.33,
+      "step": 1379
+    },
+    {
+      "epoch": 1.3413802896904288,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.9212416554316837e-06,
+      "loss": 1.3286,
+      "step": 1380
+    },
+    {
+      "epoch": 1.3423540390311195,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.9187205343928783e-06,
+      "loss": 1.3208,
+      "step": 1381
+    },
+    {
+      "epoch": 1.34332778837181,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.916198975162184e-06,
+      "loss": 1.3396,
+      "step": 1382
+    },
+    {
+      "epoch": 1.3443015377125005,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.913676980378418e-06,
+      "loss": 1.3267,
+      "step": 1383
+    },
+    {
+      "epoch": 1.345275287053191,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.9111545526808528e-06,
+      "loss": 1.3116,
+      "step": 1384
+    },
+    {
+      "epoch": 1.3462490363938815,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.9086316947092135e-06,
+      "loss": 1.3048,
+      "step": 1385
+    },
+    {
+      "epoch": 1.3472227857345722,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.9061084091036767e-06,
+      "loss": 1.3175,
+      "step": 1386
+    },
+    {
+      "epoch": 1.3481965350752627,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.9035846985048666e-06,
+      "loss": 1.3457,
+      "step": 1387
+    },
+    {
+      "epoch": 1.3491702844159532,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.90106056555385e-06,
+      "loss": 1.3214,
+      "step": 1388
+    },
+    {
+      "epoch": 1.350144033756644,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.8985360128921388e-06,
+      "loss": 1.316,
+      "step": 1389
+    },
+    {
+      "epoch": 1.3511177830973344,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.8960110431616806e-06,
+      "loss": 1.3004,
+      "step": 1390
+    },
+    {
+      "epoch": 1.352091532438025,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.8934856590048625e-06,
+      "loss": 1.3256,
+      "step": 1391
+    },
+    {
+      "epoch": 1.3530652817787154,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.8909598630645042e-06,
+      "loss": 1.3187,
+      "step": 1392
+    },
+    {
+      "epoch": 1.354039031119406,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.8884336579838556e-06,
+      "loss": 1.3232,
+      "step": 1393
+    },
+    {
+      "epoch": 1.3550127804600964,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.8859070464065963e-06,
+      "loss": 1.3082,
+      "step": 1394
+    },
+    {
+      "epoch": 1.3559865298007872,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.88338003097683e-06,
+      "loss": 1.3194,
+      "step": 1395
+    },
+    {
+      "epoch": 1.3569602791414777,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.880852614339083e-06,
+      "loss": 1.3119,
+      "step": 1396
+    },
+    {
+      "epoch": 1.3579340284821682,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.878324799138303e-06,
+      "loss": 1.2957,
+      "step": 1397
+    },
+    {
+      "epoch": 1.3589077778228589,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.8757965880198523e-06,
+      "loss": 1.3294,
+      "step": 1398
+    },
+    {
+      "epoch": 1.3598815271635494,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.8732679836295103e-06,
+      "loss": 1.3227,
+      "step": 1399
+    },
+    {
+      "epoch": 1.3608552765042399,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.870738988613467e-06,
+      "loss": 1.3249,
+      "step": 1400
+    },
+    {
+      "epoch": 1.3618290258449304,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.868209605618319e-06,
+      "loss": 1.3452,
+      "step": 1401
+    },
+    {
+      "epoch": 1.3628027751856209,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.8656798372910714e-06,
+      "loss": 1.3414,
+      "step": 1402
+    },
+    {
+      "epoch": 1.3637765245263116,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.863149686279133e-06,
+      "loss": 1.3234,
+      "step": 1403
+    },
+    {
+      "epoch": 1.364750273867002,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.8606191552303103e-06,
+      "loss": 1.3342,
+      "step": 1404
+    },
+    {
+      "epoch": 1.3657240232076926,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.858088246792811e-06,
+      "loss": 1.3379,
+      "step": 1405
+    },
+    {
+      "epoch": 1.366697772548383,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.855556963615234e-06,
+      "loss": 1.3388,
+      "step": 1406
+    },
+    {
+      "epoch": 1.3676715218890738,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.853025308346574e-06,
+      "loss": 1.331,
+      "step": 1407
+    },
+    {
+      "epoch": 1.3686452712297643,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.850493283636212e-06,
+      "loss": 1.3115,
+      "step": 1408
+    },
+    {
+      "epoch": 1.3696190205704548,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.847960892133918e-06,
+      "loss": 1.3168,
+      "step": 1409
+    },
+    {
+      "epoch": 1.3705927699111453,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.845428136489844e-06,
+      "loss": 1.319,
+      "step": 1410
+    },
+    {
+      "epoch": 1.3715665192518358,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.842895019354525e-06,
+      "loss": 1.347,
+      "step": 1411
+    },
+    {
+      "epoch": 1.3725402685925265,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.8403615433788722e-06,
+      "loss": 1.3205,
+      "step": 1412
+    },
+    {
+      "epoch": 1.373514017933217,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.837827711214173e-06,
+      "loss": 1.306,
+      "step": 1413
+    },
+    {
+      "epoch": 1.3744877672739075,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.8352935255120893e-06,
+      "loss": 1.3281,
+      "step": 1414
+    },
+    {
+      "epoch": 1.3754615166145983,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.8327589889246513e-06,
+      "loss": 1.3142,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3764352659552888,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.8302241041042564e-06,
+      "loss": 1.3356,
+      "step": 1416
+    },
+    {
+      "epoch": 1.3774090152959793,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.8276888737036657e-06,
+      "loss": 1.3273,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3783827646366698,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.8251533003760044e-06,
+      "loss": 1.3069,
+      "step": 1418
+    },
+    {
+      "epoch": 1.3793565139773603,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.822617386774754e-06,
+      "loss": 1.3231,
+      "step": 1419
+    },
+    {
+      "epoch": 1.3803302633180508,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.8200811355537543e-06,
+      "loss": 1.3291,
+      "step": 1420
+    },
+    {
+      "epoch": 1.3813040126587415,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.817544549367197e-06,
+      "loss": 1.3474,
+      "step": 1421
+    },
+    {
+      "epoch": 1.382277761999432,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.8150076308696247e-06,
+      "loss": 1.3157,
+      "step": 1422
+    },
+    {
+      "epoch": 1.3832515113401225,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.812470382715927e-06,
+      "loss": 1.3027,
+      "step": 1423
+    },
+    {
+      "epoch": 1.3842252606808132,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.8099328075613403e-06,
+      "loss": 1.3373,
+      "step": 1424
+    },
+    {
+      "epoch": 1.3851990100215037,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.807394908061441e-06,
+      "loss": 1.3396,
+      "step": 1425
+    },
+    {
+      "epoch": 1.3861727593621942,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.8048566868721473e-06,
+      "loss": 1.3278,
+      "step": 1426
+    },
+    {
+      "epoch": 1.3871465087028847,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.802318146649713e-06,
+      "loss": 1.2906,
+      "step": 1427
+    },
+    {
+      "epoch": 1.3881202580435752,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.7997792900507236e-06,
+      "loss": 1.3173,
+      "step": 1428
+    },
+    {
+      "epoch": 1.389094007384266,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.7972401197321e-06,
+      "loss": 1.3236,
+      "step": 1429
+    },
+    {
+      "epoch": 1.3900677567249564,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.7947006383510868e-06,
+      "loss": 1.3112,
+      "step": 1430
+    },
+    {
+      "epoch": 1.391041506065647,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.7921608485652585e-06,
+      "loss": 1.3085,
+      "step": 1431
+    },
+    {
+      "epoch": 1.3920152554063374,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.7896207530325093e-06,
+      "loss": 1.3015,
+      "step": 1432
+    },
+    {
+      "epoch": 1.3929890047470281,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.7870803544110546e-06,
+      "loss": 1.3219,
+      "step": 1433
+    },
+    {
+      "epoch": 1.3939627540877186,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.7845396553594267e-06,
+      "loss": 1.3122,
+      "step": 1434
+    },
+    {
+      "epoch": 1.3949365034284091,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.7819986585364715e-06,
+      "loss": 1.2978,
+      "step": 1435
+    },
+    {
+      "epoch": 1.3959102527690996,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.7794573666013485e-06,
+      "loss": 1.3255,
+      "step": 1436
+    },
+    {
+      "epoch": 1.3968840021097901,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.7769157822135246e-06,
+      "loss": 1.3289,
+      "step": 1437
+    },
+    {
+      "epoch": 1.3978577514504809,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.7743739080327724e-06,
+      "loss": 1.2938,
+      "step": 1438
+    },
+    {
+      "epoch": 1.3988315007911714,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.771831746719169e-06,
+      "loss": 1.3264,
+      "step": 1439
+    },
+    {
+      "epoch": 1.3998052501318619,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.769289300933091e-06,
+      "loss": 1.3336,
+      "step": 1440
+    },
+    {
+      "epoch": 1.4007789994725526,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.7667465733352133e-06,
+      "loss": 1.317,
+      "step": 1441
+    },
+    {
+      "epoch": 1.401752748813243,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.7642035665865065e-06,
+      "loss": 1.3164,
+      "step": 1442
+    },
+    {
+      "epoch": 1.4027264981539336,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.76166028334823e-06,
+      "loss": 1.3167,
+      "step": 1443
+    },
+    {
+      "epoch": 1.403700247494624,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.7591167262819356e-06,
+      "loss": 1.3123,
+      "step": 1444
+    },
+    {
+      "epoch": 1.4046739968353146,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.7565728980494622e-06,
+      "loss": 1.3292,
+      "step": 1445
+    },
+    {
+      "epoch": 1.405647746176005,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.754028801312931e-06,
+      "loss": 1.325,
+      "step": 1446
+    },
+    {
+      "epoch": 1.4066214955166958,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.7514844387347434e-06,
+      "loss": 1.3307,
+      "step": 1447
+    },
+    {
+      "epoch": 1.4075952448573863,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.7489398129775803e-06,
+      "loss": 1.3054,
+      "step": 1448
+    },
+    {
+      "epoch": 1.4085689941980768,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.7463949267043978e-06,
+      "loss": 1.3235,
+      "step": 1449
+    },
+    {
+      "epoch": 1.4095427435387675,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.7438497825784242e-06,
+      "loss": 1.3009,
+      "step": 1450
+    },
+    {
+      "epoch": 1.410516492879458,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.74130438326316e-06,
+      "loss": 1.3134,
+      "step": 1451
+    },
+    {
+      "epoch": 1.4114902422201485,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.7387587314223673e-06,
+      "loss": 1.3411,
+      "step": 1452
+    },
+    {
+      "epoch": 1.412463991560839,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.7362128297200784e-06,
+      "loss": 1.3121,
+      "step": 1453
+    },
+    {
+      "epoch": 1.4134377409015295,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.7336666808205844e-06,
+      "loss": 1.3225,
+      "step": 1454
+    },
+    {
+      "epoch": 1.41441149024222,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.7311202873884345e-06,
+      "loss": 1.31,
+      "step": 1455
+    },
+    {
+      "epoch": 1.4153852395829107,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.7285736520884355e-06,
+      "loss": 1.3111,
+      "step": 1456
+    },
+    {
+      "epoch": 1.4163589889236012,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.726026777585645e-06,
+      "loss": 1.3287,
+      "step": 1457
+    },
+    {
+      "epoch": 1.4173327382642917,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.723479666545373e-06,
+      "loss": 1.3063,
+      "step": 1458
+    },
+    {
+      "epoch": 1.4183064876049825,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.7209323216331774e-06,
+      "loss": 1.3414,
+      "step": 1459
+    },
+    {
+      "epoch": 1.419280236945673,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.7183847455148594e-06,
+      "loss": 1.3401,
+      "step": 1460
+    },
+    {
+      "epoch": 1.4202539862863635,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.715836940856461e-06,
+      "loss": 1.3144,
+      "step": 1461
+    },
+    {
+      "epoch": 1.421227735627054,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.7132889103242664e-06,
+      "loss": 1.322,
+      "step": 1462
+    },
+    {
+      "epoch": 1.4222014849677445,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.7107406565847943e-06,
+      "loss": 1.3164,
+      "step": 1463
+    },
+    {
+      "epoch": 1.4231752343084352,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.708192182304798e-06,
+      "loss": 1.333,
+      "step": 1464
+    },
+    {
+      "epoch": 1.4241489836491257,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.70564349015126e-06,
+      "loss": 1.3385,
+      "step": 1465
+    },
+    {
+      "epoch": 1.4251227329898162,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.7030945827913935e-06,
+      "loss": 1.312,
+      "step": 1466
+    },
+    {
+      "epoch": 1.4260964823305067,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.700545462892633e-06,
+      "loss": 1.3395,
+      "step": 1467
+    },
+    {
+      "epoch": 1.4270702316711974,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.6979961331226386e-06,
+      "loss": 1.3315,
+      "step": 1468
+    },
+    {
+      "epoch": 1.428043981011888,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.6954465961492892e-06,
+      "loss": 1.306,
+      "step": 1469
+    },
+    {
+      "epoch": 1.4290177303525784,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.692896854640681e-06,
+      "loss": 1.3272,
+      "step": 1470
+    },
+    {
+      "epoch": 1.429991479693269,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.6903469112651233e-06,
+      "loss": 1.3382,
+      "step": 1471
+    },
+    {
+      "epoch": 1.4309652290339594,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.687796768691136e-06,
+      "loss": 1.3116,
+      "step": 1472
+    },
+    {
+      "epoch": 1.4319389783746501,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.685246429587449e-06,
+      "loss": 1.3214,
+      "step": 1473
+    },
+    {
+      "epoch": 1.4329127277153406,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.6826958966229988e-06,
+      "loss": 1.3184,
+      "step": 1474
+    },
+    {
+      "epoch": 1.4338864770560311,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.680145172466921e-06,
+      "loss": 1.3083,
+      "step": 1475
+    },
+    {
+      "epoch": 1.4348602263967218,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.677594259788555e-06,
+      "loss": 1.3113,
+      "step": 1476
+    },
+    {
+      "epoch": 1.4358339757374123,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.6750431612574366e-06,
+      "loss": 1.335,
+      "step": 1477
+    },
+    {
+      "epoch": 1.4368077250781028,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.6724918795432942e-06,
+      "loss": 1.3067,
+      "step": 1478
+    },
+    {
+      "epoch": 1.4377814744187933,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.6699404173160496e-06,
+      "loss": 1.3199,
+      "step": 1479
+    },
+    {
+      "epoch": 1.4387552237594838,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.667388777245814e-06,
+      "loss": 1.2809,
+      "step": 1480
+    },
+    {
+      "epoch": 1.4397289731001743,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.664836962002883e-06,
+      "loss": 1.3153,
+      "step": 1481
+    },
+    {
+      "epoch": 1.440702722440865,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.662284974257738e-06,
+      "loss": 1.3068,
+      "step": 1482
+    },
+    {
+      "epoch": 1.4416764717815556,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.659732816681037e-06,
+      "loss": 1.3182,
+      "step": 1483
+    },
+    {
+      "epoch": 1.442650221122246,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.6571804919436195e-06,
+      "loss": 1.3285,
+      "step": 1484
+    },
+    {
+      "epoch": 1.4436239704629368,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.6546280027164985e-06,
+      "loss": 1.3073,
+      "step": 1485
+    },
+    {
+      "epoch": 1.4445977198036273,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.652075351670858e-06,
+      "loss": 1.3153,
+      "step": 1486
+    },
+    {
+      "epoch": 1.4455714691443178,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.6495225414780545e-06,
+      "loss": 1.3037,
+      "step": 1487
+    },
+    {
+      "epoch": 1.4465452184850083,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.646969574809608e-06,
+      "loss": 1.326,
+      "step": 1488
+    },
+    {
+      "epoch": 1.4475189678256988,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.644416454337203e-06,
+      "loss": 1.3135,
+      "step": 1489
+    },
+    {
+      "epoch": 1.4484927171663895,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.6418631827326857e-06,
+      "loss": 1.3028,
+      "step": 1490
+    },
+    {
+      "epoch": 1.44946646650708,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.639309762668059e-06,
+      "loss": 1.3188,
+      "step": 1491
+    },
+    {
+      "epoch": 1.4504402158477705,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.636756196815484e-06,
+      "loss": 1.3191,
+      "step": 1492
+    },
+    {
+      "epoch": 1.451413965188461,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.634202487847271e-06,
+      "loss": 1.3261,
+      "step": 1493
+    },
+    {
+      "epoch": 1.4523877145291517,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.631648638435883e-06,
+      "loss": 1.3176,
+      "step": 1494
+    },
+    {
+      "epoch": 1.4533614638698422,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.629094651253927e-06,
+      "loss": 1.3236,
+      "step": 1495
+    },
+    {
+      "epoch": 1.4543352132105327,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.6265405289741567e-06,
+      "loss": 1.3197,
+      "step": 1496
+    },
+    {
+      "epoch": 1.4553089625512232,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.6239862742694668e-06,
+      "loss": 1.3219,
+      "step": 1497
+    },
+    {
+      "epoch": 1.4562827118919137,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.621431889812888e-06,
+      "loss": 1.3157,
+      "step": 1498
+    },
+    {
+      "epoch": 1.4572564612326044,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.6188773782775917e-06,
+      "loss": 1.318,
+      "step": 1499
+    },
+    {
+      "epoch": 1.458230210573295,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.616322742336877e-06,
+      "loss": 1.3216,
+      "step": 1500
+    },
+    {
+      "epoch": 1.4592039599139854,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.6137679846641765e-06,
+      "loss": 1.3314,
+      "step": 1501
+    },
+    {
+      "epoch": 1.4601777092546762,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.6112131079330494e-06,
+      "loss": 1.3215,
+      "step": 1502
+    },
+    {
+      "epoch": 1.4611514585953667,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.6086581148171784e-06,
+      "loss": 1.3238,
+      "step": 1503
+    },
+    {
+      "epoch": 1.4621252079360572,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.606103007990371e-06,
+      "loss": 1.3354,
+      "step": 1504
+    },
+    {
+      "epoch": 1.4630989572767477,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.6035477901265506e-06,
+      "loss": 1.325,
+      "step": 1505
+    },
+    {
+      "epoch": 1.4640727066174382,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.6009924638997566e-06,
+      "loss": 1.3116,
+      "step": 1506
+    },
+    {
+      "epoch": 1.4650464559581287,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.5984370319841444e-06,
+      "loss": 1.3227,
+      "step": 1507
+    },
+    {
+      "epoch": 1.4660202052988194,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.595881497053979e-06,
+      "loss": 1.3134,
+      "step": 1508
+    },
+    {
+      "epoch": 1.4669939546395099,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.593325861783632e-06,
+      "loss": 1.3098,
+      "step": 1509
+    },
+    {
+      "epoch": 1.4679677039802004,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.590770128847582e-06,
+      "loss": 1.3104,
+      "step": 1510
+    },
+    {
+      "epoch": 1.468941453320891,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.5882143009204074e-06,
+      "loss": 1.3373,
+      "step": 1511
+    },
+    {
+      "epoch": 1.4699152026615816,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.585658380676788e-06,
+      "loss": 1.3148,
+      "step": 1512
+    },
+    {
+      "epoch": 1.470888952002272,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.5831023707915e-06,
+      "loss": 1.299,
+      "step": 1513
+    },
+    {
+      "epoch": 1.4718627013429626,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.5805462739394118e-06,
+      "loss": 1.3312,
+      "step": 1514
+    },
+    {
+      "epoch": 1.472836450683653,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.5779900927954844e-06,
+      "loss": 1.2993,
+      "step": 1515
+    },
+    {
+      "epoch": 1.4738102000243436,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.575433830034767e-06,
+      "loss": 1.3354,
+      "step": 1516
+    },
+    {
+      "epoch": 1.4747839493650343,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.572877488332393e-06,
+      "loss": 1.3343,
+      "step": 1517
+    },
+    {
+      "epoch": 1.4757576987057248,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.57032107036358e-06,
+      "loss": 1.317,
+      "step": 1518
+    },
+    {
+      "epoch": 1.4767314480464153,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.5677645788036245e-06,
+      "loss": 1.3155,
+      "step": 1519
+    },
+    {
+      "epoch": 1.477705197387106,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.5652080163278986e-06,
+      "loss": 1.3329,
+      "step": 1520
+    },
+    {
+      "epoch": 1.4786789467277965,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.5626513856118527e-06,
+      "loss": 1.3123,
+      "step": 1521
+    },
+    {
+      "epoch": 1.479652696068487,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.5600946893310043e-06,
+      "loss": 1.329,
+      "step": 1522
+    },
+    {
+      "epoch": 1.4806264454091775,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.557537930160941e-06,
+      "loss": 1.3235,
+      "step": 1523
+    },
+    {
+      "epoch": 1.481600194749868,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.5549811107773175e-06,
+      "loss": 1.314,
+      "step": 1524
+    },
+    {
+      "epoch": 1.4825739440905588,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.5524242338558498e-06,
+      "loss": 1.3204,
+      "step": 1525
+    },
+    {
+      "epoch": 1.4835476934312493,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.5498673020723147e-06,
+      "loss": 1.31,
+      "step": 1526
+    },
+    {
+      "epoch": 1.4845214427719398,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.547310318102548e-06,
+      "loss": 1.3296,
+      "step": 1527
+    },
+    {
+      "epoch": 1.4854951921126303,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.5447532846224364e-06,
+      "loss": 1.3187,
+      "step": 1528
+    },
+    {
+      "epoch": 1.486468941453321,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.5421962043079217e-06,
+      "loss": 1.3136,
+      "step": 1529
+    },
+    {
+      "epoch": 1.4874426907940115,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.539639079834994e-06,
+      "loss": 1.3299,
+      "step": 1530
+    },
+    {
+      "epoch": 1.488416440134702,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.5370819138796884e-06,
+      "loss": 1.3279,
+      "step": 1531
+    },
+    {
+      "epoch": 1.4893901894753925,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.534524709118085e-06,
+      "loss": 1.3395,
+      "step": 1532
+    },
+    {
+      "epoch": 1.490363938816083,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.5319674682263033e-06,
+      "loss": 1.3349,
+      "step": 1533
+    },
+    {
+      "epoch": 1.4913376881567737,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.5294101938805017e-06,
+      "loss": 1.3386,
+      "step": 1534
+    },
+    {
+      "epoch": 1.4923114374974642,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.5268528887568723e-06,
+      "loss": 1.3233,
+      "step": 1535
+    },
+    {
+      "epoch": 1.4932851868381547,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.5242955555316396e-06,
+      "loss": 1.3073,
+      "step": 1536
+    },
+    {
+      "epoch": 1.4942589361788454,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.52173819688106e-06,
+      "loss": 1.3273,
+      "step": 1537
+    },
+    {
+      "epoch": 1.495232685519536,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.5191808154814124e-06,
+      "loss": 1.3105,
+      "step": 1538
+    },
+    {
+      "epoch": 1.4962064348602264,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.516623414009003e-06,
+      "loss": 1.309,
+      "step": 1539
+    },
+    {
+      "epoch": 1.497180184200917,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.514065995140156e-06,
+      "loss": 1.316,
+      "step": 1540
+    },
+    {
+      "epoch": 1.4981539335416074,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.5115085615512173e-06,
+      "loss": 1.3237,
+      "step": 1541
+    },
+    {
+      "epoch": 1.499127682882298,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.5089511159185465e-06,
+      "loss": 1.309,
+      "step": 1542
+    },
+    {
+      "epoch": 1.5001014322229886,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.5063936609185133e-06,
+      "loss": 1.3211,
+      "step": 1543
+    },
+    {
+      "epoch": 1.5010751815636791,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.503836199227502e-06,
+      "loss": 1.3328,
+      "step": 1544
+    },
+    {
+      "epoch": 1.5020489309043699,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.5012787335218997e-06,
+      "loss": 1.3355,
+      "step": 1545
+    },
+    {
+      "epoch": 1.5030226802450604,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.498721266478101e-06,
+      "loss": 1.3007,
+      "step": 1546
+    },
+    {
+      "epoch": 1.5039964295857509,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.496163800772499e-06,
+      "loss": 1.308,
+      "step": 1547
+    },
+    {
+      "epoch": 1.5049701789264414,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.493606339081487e-06,
+      "loss": 1.326,
+      "step": 1548
+    },
+    {
+      "epoch": 1.5059439282671319,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.4910488840814543e-06,
+      "loss": 1.3158,
+      "step": 1549
+    },
+    {
+      "epoch": 1.5069176776078224,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.4884914384487822e-06,
+      "loss": 1.3075,
+      "step": 1550
+    },
+    {
+      "epoch": 1.5078914269485129,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.4859340048598438e-06,
+      "loss": 1.3076,
+      "step": 1551
+    },
+    {
+      "epoch": 1.5088651762892036,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.483376585990998e-06,
+      "loss": 1.317,
+      "step": 1552
+    },
+    {
+      "epoch": 1.509838925629894,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.4808191845185885e-06,
+      "loss": 1.3139,
+      "step": 1553
+    },
+    {
+      "epoch": 1.5108126749705848,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.4782618031189405e-06,
+      "loss": 1.3104,
+      "step": 1554
+    },
+    {
+      "epoch": 1.5117864243112753,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.475704444468361e-06,
+      "loss": 1.3422,
+      "step": 1555
+    },
+    {
+      "epoch": 1.5127601736519658,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.473147111243129e-06,
+      "loss": 1.3094,
+      "step": 1556
+    },
+    {
+      "epoch": 1.5137339229926563,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.470589806119499e-06,
+      "loss": 1.3046,
+      "step": 1557
+    },
+    {
+      "epoch": 1.5147076723333468,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.468032531773697e-06,
+      "loss": 1.3303,
+      "step": 1558
+    },
+    {
+      "epoch": 1.5156814216740373,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.4654752908819153e-06,
+      "loss": 1.319,
+      "step": 1559
+    },
+    {
+      "epoch": 1.5166551710147278,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.4629180861203116e-06,
+      "loss": 1.3279,
+      "step": 1560
+    },
+    {
+      "epoch": 1.5176289203554185,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.4603609201650066e-06,
+      "loss": 1.3235,
+      "step": 1561
+    },
+    {
+      "epoch": 1.518602669696109,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.4578037956920783e-06,
+      "loss": 1.3155,
+      "step": 1562
+    },
+    {
+      "epoch": 1.5195764190367997,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.455246715377564e-06,
+      "loss": 1.3035,
+      "step": 1563
+    },
+    {
+      "epoch": 1.5205501683774902,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.4526896818974534e-06,
+      "loss": 1.312,
+      "step": 1564
+    },
+    {
+      "epoch": 1.5215239177181807,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.4501326979276857e-06,
+      "loss": 1.3306,
+      "step": 1565
+    },
+    {
+      "epoch": 1.5224976670588712,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.447575766144151e-06,
+      "loss": 1.3141,
+      "step": 1566
+    },
+    {
+      "epoch": 1.5234714163995617,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.4450188892226834e-06,
+      "loss": 1.3198,
+      "step": 1567
+    },
+    {
+      "epoch": 1.5244451657402522,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.4424620698390597e-06,
+      "loss": 1.3094,
+      "step": 1568
+    },
+    {
+      "epoch": 1.525418915080943,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.4399053106689965e-06,
+      "loss": 1.3242,
+      "step": 1569
+    },
+    {
+      "epoch": 1.5263926644216335,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.4373486143881477e-06,
+      "loss": 1.3158,
+      "step": 1570
+    },
+    {
+      "epoch": 1.527366413762324,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.4347919836721014e-06,
+      "loss": 1.3152,
+      "step": 1571
+    },
+    {
+      "epoch": 1.5283401631030147,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.432235421196376e-06,
+      "loss": 1.3307,
+      "step": 1572
+    },
+    {
+      "epoch": 1.5293139124437052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.4296789296364202e-06,
+      "loss": 1.3083,
+      "step": 1573
+    },
+    {
+      "epoch": 1.5302876617843957,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.427122511667608e-06,
+      "loss": 1.3176,
+      "step": 1574
+    },
+    {
+      "epoch": 1.5312614111250862,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.4245661699652343e-06,
+      "loss": 1.3168,
+      "step": 1575
+    },
+    {
+      "epoch": 1.5322351604657767,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.4220099072045164e-06,
+      "loss": 1.327,
+      "step": 1576
+    },
+    {
+      "epoch": 1.5332089098064672,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.4194537260605895e-06,
+      "loss": 1.3198,
+      "step": 1577
+    },
+    {
+      "epoch": 1.534182659147158,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.4168976292085013e-06,
+      "loss": 1.3137,
+      "step": 1578
+    },
+    {
+      "epoch": 1.5351564084878484,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.4143416193232123e-06,
+      "loss": 1.3177,
+      "step": 1579
+    },
+    {
+      "epoch": 1.5361301578285391,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.411785699079593e-06,
+      "loss": 1.3057,
+      "step": 1580
+    },
+    {
+      "epoch": 1.5371039071692296,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.4092298711524183e-06,
+      "loss": 1.3536,
+      "step": 1581
+    },
+    {
+      "epoch": 1.5380776565099201,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.406674138216368e-06,
+      "loss": 1.303,
+      "step": 1582
+    },
+    {
+      "epoch": 1.5390514058506106,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.4041185029460214e-06,
+      "loss": 1.3182,
+      "step": 1583
+    },
+    {
+      "epoch": 1.5400251551913011,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.4015629680158555e-06,
+      "loss": 1.3098,
+      "step": 1584
+    },
+    {
+      "epoch": 1.5409989045319916,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.3990075361002447e-06,
+      "loss": 1.3131,
+      "step": 1585
+    },
+    {
+      "epoch": 1.5419726538726821,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.396452209873451e-06,
+      "loss": 1.3019,
+      "step": 1586
+    },
+    {
+      "epoch": 1.5429464032133728,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.39389699200963e-06,
+      "loss": 1.3238,
+      "step": 1587
+    },
+    {
+      "epoch": 1.5439201525540633,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.391341885182822e-06,
+      "loss": 1.3219,
+      "step": 1588
+    },
+    {
+      "epoch": 1.544893901894754,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.3887868920669514e-06,
+      "loss": 1.3393,
+      "step": 1589
+    },
+    {
+      "epoch": 1.5458676512354446,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.3862320153358243e-06,
+      "loss": 1.3254,
+      "step": 1590
+    },
+    {
+      "epoch": 1.546841400576135,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.3836772576631236e-06,
+      "loss": 1.3146,
+      "step": 1591
+    },
+    {
+      "epoch": 1.5478151499168256,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.3811226217224087e-06,
+      "loss": 1.3222,
+      "step": 1592
+    },
+    {
+      "epoch": 1.548788899257516,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.378568110187112e-06,
+      "loss": 1.3232,
+      "step": 1593
+    },
+    {
+      "epoch": 1.5497626485982066,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.3760137257305336e-06,
+      "loss": 1.3076,
+      "step": 1594
+    },
+    {
+      "epoch": 1.5507363979388973,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.373459471025843e-06,
+      "loss": 1.3447,
+      "step": 1595
+    },
+    {
+      "epoch": 1.5517101472795878,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.3709053487460738e-06,
+      "loss": 1.3181,
+      "step": 1596
+    },
+    {
+      "epoch": 1.5526838966202783,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.368351361564118e-06,
+      "loss": 1.3174,
+      "step": 1597
+    },
+    {
+      "epoch": 1.553657645960969,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.3657975121527295e-06,
+      "loss": 1.3236,
+      "step": 1598
+    },
+    {
+      "epoch": 1.5546313953016595,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.3632438031845167e-06,
+      "loss": 1.3348,
+      "step": 1599
+    },
+    {
+      "epoch": 1.55560514464235,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.3606902373319414e-06,
+      "loss": 1.3159,
+      "step": 1600
+    },
+    {
+      "epoch": 1.5565788939830405,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.358136817267315e-06,
+      "loss": 1.2833,
+      "step": 1601
+    },
+    {
+      "epoch": 1.557552643323731,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.3555835456627976e-06,
+      "loss": 1.3318,
+      "step": 1602
+    },
+    {
+      "epoch": 1.5585263926644215,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.3530304251903926e-06,
+      "loss": 1.2971,
+      "step": 1603
+    },
+    {
+      "epoch": 1.5595001420051122,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.350477458521946e-06,
+      "loss": 1.3297,
+      "step": 1604
+    },
+    {
+      "epoch": 1.5604738913458027,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.347924648329142e-06,
+      "loss": 1.3151,
+      "step": 1605
+    },
+    {
+      "epoch": 1.5614476406864934,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.345371997283502e-06,
+      "loss": 1.3315,
+      "step": 1606
+    },
+    {
+      "epoch": 1.562421390027184,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.342819508056382e-06,
+      "loss": 1.3173,
+      "step": 1607
+    },
+    {
+      "epoch": 1.5633951393678744,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.3402671833189646e-06,
+      "loss": 1.3461,
+      "step": 1608
+    },
+    {
+      "epoch": 1.564368888708565,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.3377150257422637e-06,
+      "loss": 1.312,
+      "step": 1609
+    },
+    {
+      "epoch": 1.5653426380492554,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.3351630379971176e-06,
+      "loss": 1.3165,
+      "step": 1610
+    },
+    {
+      "epoch": 1.566316387389946,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.332611222754187e-06,
+      "loss": 1.3003,
+      "step": 1611
+    },
+    {
+      "epoch": 1.5672901367306364,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.330059582683951e-06,
+      "loss": 1.315,
+      "step": 1612
+    },
+    {
+      "epoch": 1.5682638860713272,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.3275081204567066e-06,
+      "loss": 1.3279,
+      "step": 1613
+    },
+    {
+      "epoch": 1.5692376354120177,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.3249568387425642e-06,
+      "loss": 1.306,
+      "step": 1614
+    },
+    {
+      "epoch": 1.5702113847527084,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.322405740211445e-06,
+      "loss": 1.3177,
+      "step": 1615
+    },
+    {
+      "epoch": 1.5711851340933989,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.319854827533079e-06,
+      "loss": 1.339,
+      "step": 1616
+    },
+    {
+      "epoch": 1.5721588834340894,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.3173041033770016e-06,
+      "loss": 1.3263,
+      "step": 1617
+    },
+    {
+      "epoch": 1.5731326327747799,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.314753570412552e-06,
+      "loss": 1.3069,
+      "step": 1618
+    },
+    {
+      "epoch": 1.5741063821154704,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.312203231308865e-06,
+      "loss": 1.3169,
+      "step": 1619
+    },
+    {
+      "epoch": 1.5750801314561609,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.309653088734878e-06,
+      "loss": 1.2937,
+      "step": 1620
+    },
+    {
+      "epoch": 1.5760538807968514,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.30710314535932e-06,
+      "loss": 1.3169,
+      "step": 1621
+    },
+    {
+      "epoch": 1.577027630137542,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.304553403850711e-06,
+      "loss": 1.2974,
+      "step": 1622
+    },
+    {
+      "epoch": 1.5780013794782326,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.302003866877362e-06,
+      "loss": 1.3381,
+      "step": 1623
+    },
+    {
+      "epoch": 1.5789751288189233,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.2994545371073677e-06,
+      "loss": 1.3296,
+      "step": 1624
+    },
+    {
+      "epoch": 1.5799488781596138,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.2969054172086073e-06,
+      "loss": 1.3238,
+      "step": 1625
+    },
+    {
+      "epoch": 1.5809226275003043,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.29435650984874e-06,
+      "loss": 1.318,
+      "step": 1626
+    },
+    {
+      "epoch": 1.5818963768409948,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.291807817695202e-06,
+      "loss": 1.3192,
+      "step": 1627
+    },
+    {
+      "epoch": 1.5828701261816853,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.289259343415206e-06,
+      "loss": 1.3055,
+      "step": 1628
+    },
+    {
+      "epoch": 1.5838438755223758,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.286711089675735e-06,
+      "loss": 1.3111,
+      "step": 1629
+    },
+    {
+      "epoch": 1.5848176248630665,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.2841630591435403e-06,
+      "loss": 1.3275,
+      "step": 1630
+    },
+    {
+      "epoch": 1.585791374203757,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.2816152544851423e-06,
+      "loss": 1.3026,
+      "step": 1631
+    },
+    {
+      "epoch": 1.5867651235444478,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.2790676783668235e-06,
+      "loss": 1.3393,
+      "step": 1632
+    },
+    {
+      "epoch": 1.5877388728851383,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.2765203334546272e-06,
+      "loss": 1.3204,
+      "step": 1633
+    },
+    {
+      "epoch": 1.5887126222258288,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.2739732224143556e-06,
+      "loss": 1.3085,
+      "step": 1634
+    },
+    {
+      "epoch": 1.5896863715665193,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.2714263479115658e-06,
+      "loss": 1.3125,
+      "step": 1635
+    },
+    {
+      "epoch": 1.5906601209072098,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.268879712611566e-06,
+      "loss": 1.3001,
+      "step": 1636
+    },
+    {
+      "epoch": 1.5916338702479003,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.2663333191794165e-06,
+      "loss": 1.3272,
+      "step": 1637
+    },
+    {
+      "epoch": 1.5926076195885908,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.263787170279922e-06,
+      "loss": 1.3116,
+      "step": 1638
+    },
+    {
+      "epoch": 1.5935813689292815,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.2612412685776335e-06,
+      "loss": 1.308,
+      "step": 1639
+    },
+    {
+      "epoch": 1.594555118269972,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.258695616736842e-06,
+      "loss": 1.3227,
+      "step": 1640
+    },
+    {
+      "epoch": 1.5955288676106627,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.256150217421576e-06,
+      "loss": 1.3295,
+      "step": 1641
+    },
+    {
+      "epoch": 1.5965026169513532,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.2536050732956035e-06,
+      "loss": 1.3188,
+      "step": 1642
+    },
+    {
+      "epoch": 1.5974763662920437,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.2510601870224206e-06,
+      "loss": 1.3081,
+      "step": 1643
+    },
+    {
+      "epoch": 1.5984501156327342,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.2485155612652574e-06,
+      "loss": 1.3342,
+      "step": 1644
+    },
+    {
+      "epoch": 1.5994238649734247,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.2459711986870704e-06,
+      "loss": 1.3203,
+      "step": 1645
+    },
+    {
+      "epoch": 1.6003976143141152,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.243427101950538e-06,
+      "loss": 1.3271,
+      "step": 1646
+    },
+    {
+      "epoch": 1.6013713636548057,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.2408832737180648e-06,
+      "loss": 1.3244,
+      "step": 1647
+    },
+    {
+      "epoch": 1.6023451129954964,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.238339716651771e-06,
+      "loss": 1.3197,
+      "step": 1648
+    },
+    {
+      "epoch": 1.603318862336187,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.2357964334134947e-06,
+      "loss": 1.3157,
+      "step": 1649
+    },
+    {
+      "epoch": 1.6042926116768776,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.2332534266647867e-06,
+      "loss": 1.3465,
+      "step": 1650
+    },
+    {
+      "epoch": 1.6052663610175681,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.2307106990669093e-06,
+      "loss": 1.33,
+      "step": 1651
+    },
+    {
+      "epoch": 1.6062401103582586,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.2281682532808317e-06,
+      "loss": 1.3204,
+      "step": 1652
+    },
+    {
+      "epoch": 1.6072138596989491,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.225626091967228e-06,
+      "loss": 1.3131,
+      "step": 1653
+    },
+    {
+      "epoch": 1.6081876090396396,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.2230842177864766e-06,
+      "loss": 1.3227,
+      "step": 1654
+    },
+    {
+      "epoch": 1.6091613583803301,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.220542633398652e-06,
+      "loss": 1.3009,
+      "step": 1655
+    },
+    {
+      "epoch": 1.6101351077210209,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.218001341463529e-06,
+      "loss": 1.3153,
+      "step": 1656
+    },
+    {
+      "epoch": 1.6111088570617114,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.215460344640574e-06,
+      "loss": 1.3216,
+      "step": 1657
+    },
+    {
+      "epoch": 1.6120826064024019,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.212919645588946e-06,
+      "loss": 1.3238,
+      "step": 1658
+    },
+    {
+      "epoch": 1.6130563557430926,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.210379246967491e-06,
+      "loss": 1.3253,
+      "step": 1659
+    },
+    {
+      "epoch": 1.614030105083783,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.207839151434742e-06,
+      "loss": 1.3132,
+      "step": 1660
+    },
+    {
+      "epoch": 1.6150038544244736,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.205299361648913e-06,
+      "loss": 1.3274,
+      "step": 1661
+    },
+    {
+      "epoch": 1.615977603765164,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.2027598802679013e-06,
+      "loss": 1.3096,
+      "step": 1662
+    },
+    {
+      "epoch": 1.6169513531058546,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.2002207099492776e-06,
+      "loss": 1.319,
+      "step": 1663
+    },
+    {
+      "epoch": 1.617925102446545,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.1976818533502886e-06,
+      "loss": 1.3269,
+      "step": 1664
+    },
+    {
+      "epoch": 1.6188988517872358,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.1951433131278535e-06,
+      "loss": 1.3117,
+      "step": 1665
+    },
+    {
+      "epoch": 1.6198726011279263,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.1926050919385594e-06,
+      "loss": 1.3188,
+      "step": 1666
+    },
+    {
+      "epoch": 1.620846350468617,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.1900671924386606e-06,
+      "loss": 1.3052,
+      "step": 1667
+    },
+    {
+      "epoch": 1.6218200998093075,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.1875296172840737e-06,
+      "loss": 1.3015,
+      "step": 1668
+    },
+    {
+      "epoch": 1.622793849149998,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.1849923691303757e-06,
+      "loss": 1.3102,
+      "step": 1669
+    },
+    {
+      "epoch": 1.6237675984906885,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.1824554506328033e-06,
+      "loss": 1.3063,
+      "step": 1670
+    },
+    {
+      "epoch": 1.624741347831379,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.1799188644462457e-06,
+      "loss": 1.3198,
+      "step": 1671
+    },
+    {
+      "epoch": 1.6257150971720695,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.1773826132252456e-06,
+      "loss": 1.3444,
+      "step": 1672
+    },
+    {
+      "epoch": 1.62668884651276,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.174846699623997e-06,
+      "loss": 1.315,
+      "step": 1673
+    },
+    {
+      "epoch": 1.6276625958534507,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.172311126296335e-06,
+      "loss": 1.3202,
+      "step": 1674
+    },
+    {
+      "epoch": 1.6286363451941412,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.169775895895745e-06,
+      "loss": 1.3349,
+      "step": 1675
+    },
+    {
+      "epoch": 1.629610094534832,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.1672410110753495e-06,
+      "loss": 1.3099,
+      "step": 1676
+    },
+    {
+      "epoch": 1.6305838438755225,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.164706474487911e-06,
+      "loss": 1.3121,
+      "step": 1677
+    },
+    {
+      "epoch": 1.631557593216213,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.1621722887858273e-06,
+      "loss": 1.3133,
+      "step": 1678
+    },
+    {
+      "epoch": 1.6325313425569035,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.1596384566211286e-06,
+      "loss": 1.3271,
+      "step": 1679
+    },
+    {
+      "epoch": 1.633505091897594,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.157104980645476e-06,
+      "loss": 1.3363,
+      "step": 1680
+    },
+    {
+      "epoch": 1.6344788412382845,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.1545718635101563e-06,
+      "loss": 1.3166,
+      "step": 1681
+    },
+    {
+      "epoch": 1.635452590578975,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.1520391078660823e-06,
+      "loss": 1.2876,
+      "step": 1682
+    },
+    {
+      "epoch": 1.6364263399196657,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.149506716363788e-06,
+      "loss": 1.3082,
+      "step": 1683
+    },
+    {
+      "epoch": 1.6374000892603562,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.146974691653427e-06,
+      "loss": 1.318,
+      "step": 1684
+    },
+    {
+      "epoch": 1.638373838601047,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.144443036384767e-06,
+      "loss": 1.317,
+      "step": 1685
+    },
+    {
+      "epoch": 1.6393475879417374,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.1419117532071906e-06,
+      "loss": 1.3125,
+      "step": 1686
+    },
+    {
+      "epoch": 1.640321337282428,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.1393808447696906e-06,
+      "loss": 1.3225,
+      "step": 1687
+    },
+    {
+      "epoch": 1.6412950866231184,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.136850313720868e-06,
+      "loss": 1.3023,
+      "step": 1688
+    },
+    {
+      "epoch": 1.642268835963809,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.134320162708929e-06,
+      "loss": 1.3211,
+      "step": 1689
+    },
+    {
+      "epoch": 1.6432425853044994,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.131790394381682e-06,
+      "loss": 1.3269,
+      "step": 1690
+    },
+    {
+      "epoch": 1.6442163346451901,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.129261011386534e-06,
+      "loss": 1.3202,
+      "step": 1691
+    },
+    {
+      "epoch": 1.6451900839858806,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.1267320163704897e-06,
+      "loss": 1.3397,
+      "step": 1692
+    },
+    {
+      "epoch": 1.6461638333265713,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.1242034119801477e-06,
+      "loss": 1.3212,
+      "step": 1693
+    },
+    {
+      "epoch": 1.6471375826672618,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.1216752008616974e-06,
+      "loss": 1.3154,
+      "step": 1694
+    },
+    {
+      "epoch": 1.6481113320079523,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.1191473856609182e-06,
+      "loss": 1.2978,
+      "step": 1695
+    },
+    {
+      "epoch": 1.6490850813486428,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.1166199690231715e-06,
+      "loss": 1.3023,
+      "step": 1696
+    },
+    {
+      "epoch": 1.6500588306893333,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.114092953593405e-06,
+      "loss": 1.3054,
+      "step": 1697
+    },
+    {
+      "epoch": 1.6510325800300238,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.111566342016145e-06,
+      "loss": 1.3176,
+      "step": 1698
+    },
+    {
+      "epoch": 1.6520063293707143,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.1090401369354966e-06,
+      "loss": 1.3254,
+      "step": 1699
+    },
+    {
+      "epoch": 1.652980078711405,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.1065143409951384e-06,
+      "loss": 1.3353,
+      "step": 1700
+    },
+    {
+      "epoch": 1.6539538280520956,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.1039889568383202e-06,
+      "loss": 1.3259,
+      "step": 1701
+    },
+    {
+      "epoch": 1.6549275773927863,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.1014639871078625e-06,
+      "loss": 1.3283,
+      "step": 1702
+    },
+    {
+      "epoch": 1.6559013267334768,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.09893943444615e-06,
+      "loss": 1.3329,
+      "step": 1703
+    },
+    {
+      "epoch": 1.6568750760741673,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.096415301495134e-06,
+      "loss": 1.2812,
+      "step": 1704
+    },
+    {
+      "epoch": 1.6578488254148578,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.0938915908963232e-06,
+      "loss": 1.3271,
+      "step": 1705
+    },
+    {
+      "epoch": 1.6588225747555483,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.091368305290788e-06,
+      "loss": 1.3063,
+      "step": 1706
+    },
+    {
+      "epoch": 1.6597963240962388,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.088845447319149e-06,
+      "loss": 1.3102,
+      "step": 1707
+    },
+    {
+      "epoch": 1.6607700734369293,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.0863230196215833e-06,
+      "loss": 1.3243,
+      "step": 1708
+    },
+    {
+      "epoch": 1.66174382277762,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.083801024837817e-06,
+      "loss": 1.3017,
+      "step": 1709
+    },
+    {
+      "epoch": 1.6627175721183105,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.081279465607123e-06,
+      "loss": 1.3047,
+      "step": 1710
+    },
+    {
+      "epoch": 1.6636913214590012,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.078758344568317e-06,
+      "loss": 1.2974,
+      "step": 1711
+    },
+    {
+      "epoch": 1.6646650707996917,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.0762376643597586e-06,
+      "loss": 1.3087,
+      "step": 1712
+    },
+    {
+      "epoch": 1.6656388201403822,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.073717427619344e-06,
+      "loss": 1.3216,
+      "step": 1713
+    },
+    {
+      "epoch": 1.6666125694810727,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.0711976369845065e-06,
+      "loss": 1.341,
+      "step": 1714
+    },
+    {
+      "epoch": 1.6675863188217632,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.0686782950922113e-06,
+      "loss": 1.3201,
+      "step": 1715
+    },
+    {
+      "epoch": 1.6685600681624537,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.066159404578956e-06,
+      "loss": 1.3223,
+      "step": 1716
+    },
+    {
+      "epoch": 1.6695338175031444,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.0636409680807646e-06,
+      "loss": 1.3329,
+      "step": 1717
+    },
+    {
+      "epoch": 1.670507566843835,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.0611229882331844e-06,
+      "loss": 1.3013,
+      "step": 1718
+    },
+    {
+      "epoch": 1.6714813161845254,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.0586054676712873e-06,
+      "loss": 1.3448,
+      "step": 1719
+    },
+    {
+      "epoch": 1.6724550655252162,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.0560884090296642e-06,
+      "loss": 1.3304,
+      "step": 1720
+    },
+    {
+      "epoch": 1.6734288148659067,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.0535718149424213e-06,
+      "loss": 1.3149,
+      "step": 1721
+    },
+    {
+      "epoch": 1.6744025642065972,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.05105568804318e-06,
+      "loss": 1.3121,
+      "step": 1722
+    },
+    {
+      "epoch": 1.6753763135472877,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.048540030965072e-06,
+      "loss": 1.3351,
+      "step": 1723
+    },
+    {
+      "epoch": 1.6763500628879782,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.0460248463407377e-06,
+      "loss": 1.3046,
+      "step": 1724
+    },
+    {
+      "epoch": 1.6773238122286687,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.0435101368023222e-06,
+      "loss": 1.3182,
+      "step": 1725
+    },
+    {
+      "epoch": 1.6782975615693594,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.0409959049814753e-06,
+      "loss": 1.334,
+      "step": 1726
+    },
+    {
+      "epoch": 1.6792713109100499,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.038482153509345e-06,
+      "loss": 1.327,
+      "step": 1727
+    },
+    {
+      "epoch": 1.6802450602507406,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.0359688850165775e-06,
+      "loss": 1.324,
+      "step": 1728
+    },
+    {
+      "epoch": 1.681218809591431,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.033456102133313e-06,
+      "loss": 1.3314,
+      "step": 1729
+    },
+    {
+      "epoch": 1.6821925589321216,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.030943807489184e-06,
+      "loss": 1.316,
+      "step": 1730
+    },
+    {
+      "epoch": 1.683166308272812,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.0284320037133124e-06,
+      "loss": 1.3223,
+      "step": 1731
+    },
+    {
+      "epoch": 1.6841400576135026,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.025920693434305e-06,
+      "loss": 1.3131,
+      "step": 1732
+    },
+    {
+      "epoch": 1.685113806954193,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.0234098792802543e-06,
+      "loss": 1.3433,
+      "step": 1733
+    },
+    {
+      "epoch": 1.6860875562948836,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.0208995638787317e-06,
+      "loss": 1.3107,
+      "step": 1734
+    },
+    {
+      "epoch": 1.6870613056355743,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.0183897498567874e-06,
+      "loss": 1.3225,
+      "step": 1735
+    },
+    {
+      "epoch": 1.6880350549762648,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.0158804398409478e-06,
+      "loss": 1.3166,
+      "step": 1736
+    },
+    {
+      "epoch": 1.6890088043169555,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.0133716364572094e-06,
+      "loss": 1.3202,
+      "step": 1737
+    },
+    {
+      "epoch": 1.689982553657646,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.0108633423310407e-06,
+      "loss": 1.3136,
+      "step": 1738
+    },
+    {
+      "epoch": 1.6909563029983365,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.008355560087377e-06,
+      "loss": 1.3043,
+      "step": 1739
+    },
+    {
+      "epoch": 1.691930052339027,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.0058482923506168e-06,
+      "loss": 1.3084,
+      "step": 1740
+    },
+    {
+      "epoch": 1.6929038016797175,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.0033415417446213e-06,
+      "loss": 1.2842,
+      "step": 1741
+    },
+    {
+      "epoch": 1.693877551020408,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.0008353108927096e-06,
+      "loss": 1.3056,
+      "step": 1742
+    },
+    {
+      "epoch": 1.6948513003610988,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.998329602417658e-06,
+      "loss": 1.3268,
+      "step": 1743
+    },
+    {
+      "epoch": 1.6958250497017893,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.9958244189416955e-06,
+      "loss": 1.3373,
+      "step": 1744
+    },
+    {
+      "epoch": 1.6967987990424798,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9933197630865014e-06,
+      "loss": 1.3278,
+      "step": 1745
+    },
+    {
+      "epoch": 1.6977725483831705,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.990815637473203e-06,
+      "loss": 1.3251,
+      "step": 1746
+    },
+    {
+      "epoch": 1.698746297723861,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.988312044722373e-06,
+      "loss": 1.3134,
+      "step": 1747
+    },
+    {
+      "epoch": 1.6997200470645515,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9858089874540264e-06,
+      "loss": 1.3539,
+      "step": 1748
+    },
+    {
+      "epoch": 1.700693796405242,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9833064682876175e-06,
+      "loss": 1.3348,
+      "step": 1749
+    },
+    {
+      "epoch": 1.7016675457459325,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9808044898420387e-06,
+      "loss": 1.3189,
+      "step": 1750
+    },
+    {
+      "epoch": 1.702641295086623,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9783030547356134e-06,
+      "loss": 1.3063,
+      "step": 1751
+    },
+    {
+      "epoch": 1.7036150444273137,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9758021655861005e-06,
+      "loss": 1.3115,
+      "step": 1752
+    },
+    {
+      "epoch": 1.7045887937680042,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.973301825010685e-06,
+      "loss": 1.3155,
+      "step": 1753
+    },
+    {
+      "epoch": 1.705562543108695,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.970802035625978e-06,
+      "loss": 1.301,
+      "step": 1754
+    },
+    {
+      "epoch": 1.7065362924493854,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9683028000480135e-06,
+      "loss": 1.3158,
+      "step": 1755
+    },
+    {
+      "epoch": 1.707510041790076,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.965804120892248e-06,
+      "loss": 1.3212,
+      "step": 1756
+    },
+    {
+      "epoch": 1.7084837911307664,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.963306000773554e-06,
+      "loss": 1.3197,
+      "step": 1757
+    },
+    {
+      "epoch": 1.709457540471457,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.960808442306219e-06,
+      "loss": 1.336,
+      "step": 1758
+    },
+    {
+      "epoch": 1.7104312898121474,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9583114481039428e-06,
+      "loss": 1.3246,
+      "step": 1759
+    },
+    {
+      "epoch": 1.711405039152838,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.955815020779835e-06,
+      "loss": 1.3058,
+      "step": 1760
+    },
+    {
+      "epoch": 1.7123787884935286,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.953319162946413e-06,
+      "loss": 1.3121,
+      "step": 1761
+    },
+    {
+      "epoch": 1.7133525378342191,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.950823877215596e-06,
+      "loss": 1.3095,
+      "step": 1762
+    },
+    {
+      "epoch": 1.7143262871749099,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.948329166198705e-06,
+      "loss": 1.3092,
+      "step": 1763
+    },
+    {
+      "epoch": 1.7153000365156004,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9458350325064606e-06,
+      "loss": 1.3149,
+      "step": 1764
+    },
+    {
+      "epoch": 1.7162737858562909,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.943341478748979e-06,
+      "loss": 1.3227,
+      "step": 1765
+    },
+    {
+      "epoch": 1.7172475351969814,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.9408485075357695e-06,
+      "loss": 1.2979,
+      "step": 1766
+    },
+    {
+      "epoch": 1.7182212845376719,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9383561214757317e-06,
+      "loss": 1.3171,
+      "step": 1767
+    },
+    {
+      "epoch": 1.7191950338783624,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9358643231771517e-06,
+      "loss": 1.3133,
+      "step": 1768
+    },
+    {
+      "epoch": 1.7201687832190529,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.933373115247702e-06,
+      "loss": 1.309,
+      "step": 1769
+    },
+    {
+      "epoch": 1.7211425325597436,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.930882500294437e-06,
+      "loss": 1.3193,
+      "step": 1770
+    },
+    {
+      "epoch": 1.722116281900434,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.92839248092379e-06,
+      "loss": 1.3358,
+      "step": 1771
+    },
+    {
+      "epoch": 1.7230900312411248,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.9259030597415725e-06,
+      "loss": 1.308,
+      "step": 1772
+    },
+    {
+      "epoch": 1.7240637805818153,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.9234142393529664e-06,
+      "loss": 1.3055,
+      "step": 1773
+    },
+    {
+      "epoch": 1.7250375299225058,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.920926022362529e-06,
+      "loss": 1.2888,
+      "step": 1774
+    },
+    {
+      "epoch": 1.7260112792631963,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.918438411374184e-06,
+      "loss": 1.3207,
+      "step": 1775
+    },
+    {
+      "epoch": 1.7269850286038868,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.915951408991221e-06,
+      "loss": 1.3243,
+      "step": 1776
+    },
+    {
+      "epoch": 1.7279587779445773,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9134650178162937e-06,
+      "loss": 1.3249,
+      "step": 1777
+    },
+    {
+      "epoch": 1.728932527285268,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.910979240451415e-06,
+      "loss": 1.3181,
+      "step": 1778
+    },
+    {
+      "epoch": 1.7299062766259585,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.908494079497956e-06,
+      "loss": 1.3125,
+      "step": 1779
+    },
+    {
+      "epoch": 1.7308800259666492,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9060095375566434e-06,
+      "loss": 1.313,
+      "step": 1780
+    },
+    {
+      "epoch": 1.7318537753073397,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.903525617227555e-06,
+      "loss": 1.3111,
+      "step": 1781
+    },
+    {
+      "epoch": 1.7328275246480302,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9010423211101179e-06,
+      "loss": 1.291,
+      "step": 1782
+    },
+    {
+      "epoch": 1.7338012739887207,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8985596518031069e-06,
+      "loss": 1.3198,
+      "step": 1783
+    },
+    {
+      "epoch": 1.7347750233294112,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8960776119046417e-06,
+      "loss": 1.3269,
+      "step": 1784
+    },
+    {
+      "epoch": 1.7357487726701017,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8935962040121797e-06,
+      "loss": 1.3087,
+      "step": 1785
+    },
+    {
+      "epoch": 1.7367225220107922,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8911154307225204e-06,
+      "loss": 1.3191,
+      "step": 1786
+    },
+    {
+      "epoch": 1.737696271351483,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8886352946317984e-06,
+      "loss": 1.3191,
+      "step": 1787
+    },
+    {
+      "epoch": 1.7386700206921735,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8861557983354812e-06,
+      "loss": 1.3295,
+      "step": 1788
+    },
+    {
+      "epoch": 1.7396437700328642,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8836769444283658e-06,
+      "loss": 1.3094,
+      "step": 1789
+    },
+    {
+      "epoch": 1.7406175193735547,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.8811987355045789e-06,
+      "loss": 1.3324,
+      "step": 1790
+    },
+    {
+      "epoch": 1.7415912687142452,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8787211741575706e-06,
+      "loss": 1.3338,
+      "step": 1791
+    },
+    {
+      "epoch": 1.7425650180549357,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.8762442629801139e-06,
+      "loss": 1.3038,
+      "step": 1792
+    },
+    {
+      "epoch": 1.7435387673956262,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8737680045643013e-06,
+      "loss": 1.3108,
+      "step": 1793
+    },
+    {
+      "epoch": 1.7445125167363167,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.8712924015015429e-06,
+      "loss": 1.3124,
+      "step": 1794
+    },
+    {
+      "epoch": 1.7454862660770072,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8688174563825629e-06,
+      "loss": 1.343,
+      "step": 1795
+    },
+    {
+      "epoch": 1.746460015417698,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.8663431717973939e-06,
+      "loss": 1.3321,
+      "step": 1796
+    },
+    {
+      "epoch": 1.7474337647583884,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8638695503353816e-06,
+      "loss": 1.3017,
+      "step": 1797
+    },
+    {
+      "epoch": 1.7484075140990791,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8613965945851753e-06,
+      "loss": 1.3163,
+      "step": 1798
+    },
+    {
+      "epoch": 1.7493812634397696,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8589243071347279e-06,
+      "loss": 1.3126,
+      "step": 1799
+    },
+    {
+      "epoch": 1.7503550127804601,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8564526905712943e-06,
+      "loss": 1.3137,
+      "step": 1800
+    },
+    {
+      "epoch": 1.7513287621211506,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8539817474814257e-06,
+      "loss": 1.3132,
+      "step": 1801
+    },
+    {
+      "epoch": 1.7523025114618411,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8515114804509687e-06,
+      "loss": 1.3407,
+      "step": 1802
+    },
+    {
+      "epoch": 1.7532762608025316,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8490418920650633e-06,
+      "loss": 1.3373,
+      "step": 1803
+    },
+    {
+      "epoch": 1.7542500101432223,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.8465729849081382e-06,
+      "loss": 1.3004,
+      "step": 1804
+    },
+    {
+      "epoch": 1.7552237594839128,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.8441047615639103e-06,
+      "loss": 1.3301,
+      "step": 1805
+    },
+    {
+      "epoch": 1.7561975088246033,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.8416372246153813e-06,
+      "loss": 1.2893,
+      "step": 1806
+    },
+    {
+      "epoch": 1.757171258165294,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.8391703766448314e-06,
+      "loss": 1.3386,
+      "step": 1807
+    },
+    {
+      "epoch": 1.7581450075059846,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8367042202338228e-06,
+      "loss": 1.3159,
+      "step": 1808
+    },
+    {
+      "epoch": 1.759118756846675,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8342387579631938e-06,
+      "loss": 1.3156,
+      "step": 1809
+    },
+    {
+      "epoch": 1.7600925061873656,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.8317739924130548e-06,
+      "loss": 1.3203,
+      "step": 1810
+    },
+    {
+      "epoch": 1.761066255528056,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.8293099261627887e-06,
+      "loss": 1.2946,
+      "step": 1811
+    },
+    {
+      "epoch": 1.7620400048687466,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8268465617910456e-06,
+      "loss": 1.2961,
+      "step": 1812
+    },
+    {
+      "epoch": 1.7630137542094373,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.8243839018757412e-06,
+      "loss": 1.3179,
+      "step": 1813
+    },
+    {
+      "epoch": 1.7639875035501278,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8219219489940542e-06,
+      "loss": 1.3049,
+      "step": 1814
+    },
+    {
+      "epoch": 1.7649612528908185,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8194607057224234e-06,
+      "loss": 1.3293,
+      "step": 1815
+    },
+    {
+      "epoch": 1.765935002231509,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8170001746365445e-06,
+      "loss": 1.3112,
+      "step": 1816
+    },
+    {
+      "epoch": 1.7669087515721995,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.81454035831137e-06,
+      "loss": 1.3313,
+      "step": 1817
+    },
+    {
+      "epoch": 1.76788250091289,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.8120812593210999e-06,
+      "loss": 1.3131,
+      "step": 1818
+    },
+    {
+      "epoch": 1.7688562502535805,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8096228802391875e-06,
+      "loss": 1.2929,
+      "step": 1819
+    },
+    {
+      "epoch": 1.769829999594271,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.8071652236383316e-06,
+      "loss": 1.3389,
+      "step": 1820
+    },
+    {
+      "epoch": 1.7708037489349615,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8047082920904748e-06,
+      "loss": 1.3255,
+      "step": 1821
+    },
+    {
+      "epoch": 1.7717774982756522,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.802252088166801e-06,
+      "loss": 1.3184,
+      "step": 1822
+    },
+    {
+      "epoch": 1.7727512476163427,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.7997966144377328e-06,
+      "loss": 1.3293,
+      "step": 1823
+    },
+    {
+      "epoch": 1.7737249969570335,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.7973418734729278e-06,
+      "loss": 1.3226,
+      "step": 1824
+    },
+    {
+      "epoch": 1.774698746297724,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7948878678412779e-06,
+      "loss": 1.3325,
+      "step": 1825
+    },
+    {
+      "epoch": 1.7756724956384144,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.792434600110905e-06,
+      "loss": 1.2998,
+      "step": 1826
+    },
+    {
+      "epoch": 1.776646244979105,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.789982072849159e-06,
+      "loss": 1.3398,
+      "step": 1827
+    },
+    {
+      "epoch": 1.7776199943197954,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7875302886226143e-06,
+      "loss": 1.3189,
+      "step": 1828
+    },
+    {
+      "epoch": 1.778593743660486,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7850792499970673e-06,
+      "loss": 1.3301,
+      "step": 1829
+    },
+    {
+      "epoch": 1.7795674930011764,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7826289595375356e-06,
+      "loss": 1.3093,
+      "step": 1830
+    },
+    {
+      "epoch": 1.7805412423418672,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7801794198082534e-06,
+      "loss": 1.3059,
+      "step": 1831
+    },
+    {
+      "epoch": 1.7815149916825577,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7777306333726689e-06,
+      "loss": 1.3308,
+      "step": 1832
+    },
+    {
+      "epoch": 1.7824887410232484,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7752826027934418e-06,
+      "loss": 1.3162,
+      "step": 1833
+    },
+    {
+      "epoch": 1.783462490363939,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7728353306324408e-06,
+      "loss": 1.2821,
+      "step": 1834
+    },
+    {
+      "epoch": 1.7844362397046294,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7703888194507425e-06,
+      "loss": 1.3234,
+      "step": 1835
+    },
+    {
+      "epoch": 1.78540998904532,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7679430718086244e-06,
+      "loss": 1.3237,
+      "step": 1836
+    },
+    {
+      "epoch": 1.7863837383860104,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7654980902655666e-06,
+      "loss": 1.3349,
+      "step": 1837
+    },
+    {
+      "epoch": 1.7873574877267009,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7630538773802477e-06,
+      "loss": 1.3151,
+      "step": 1838
+    },
+    {
+      "epoch": 1.7883312370673916,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7606104357105418e-06,
+      "loss": 1.2914,
+      "step": 1839
+    },
+    {
+      "epoch": 1.789304986408082,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.7581677678135146e-06,
+      "loss": 1.3065,
+      "step": 1840
+    },
+    {
+      "epoch": 1.7902787357487728,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7557258762454232e-06,
+      "loss": 1.3159,
+      "step": 1841
+    },
+    {
+      "epoch": 1.7912524850894633,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.7532847635617123e-06,
+      "loss": 1.3185,
+      "step": 1842
+    },
+    {
+      "epoch": 1.7922262344301538,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7508444323170115e-06,
+      "loss": 1.3384,
+      "step": 1843
+    },
+    {
+      "epoch": 1.7931999837708443,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7484048850651325e-06,
+      "loss": 1.3023,
+      "step": 1844
+    },
+    {
+      "epoch": 1.7941737331115348,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7459661243590656e-06,
+      "loss": 1.3117,
+      "step": 1845
+    },
+    {
+      "epoch": 1.7951474824522253,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.7435281527509796e-06,
+      "loss": 1.3083,
+      "step": 1846
+    },
+    {
+      "epoch": 1.7961212317929158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.741090972792216e-06,
+      "loss": 1.3406,
+      "step": 1847
+    },
+    {
+      "epoch": 1.7970949811336066,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7386545870332893e-06,
+      "loss": 1.3224,
+      "step": 1848
+    },
+    {
+      "epoch": 1.798068730474297,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.736218998023882e-06,
+      "loss": 1.3155,
+      "step": 1849
+    },
+    {
+      "epoch": 1.7990424798149878,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.7337842083128435e-06,
+      "loss": 1.3394,
+      "step": 1850
+    },
+    {
+      "epoch": 1.8000162291556783,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.7313502204481847e-06,
+      "loss": 1.2991,
+      "step": 1851
+    },
+    {
+      "epoch": 1.8009899784963688,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7289170369770797e-06,
+      "loss": 1.3287,
+      "step": 1852
+    },
+    {
+      "epoch": 1.8019637278370593,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7264846604458607e-06,
+      "loss": 1.2975,
+      "step": 1853
+    },
+    {
+      "epoch": 1.8029374771777498,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7240530934000134e-06,
+      "loss": 1.2853,
+      "step": 1854
+    },
+    {
+      "epoch": 1.8039112265184403,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7216223383841774e-06,
+      "loss": 1.3112,
+      "step": 1855
+    },
+    {
+      "epoch": 1.8048849758591308,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.719192397942144e-06,
+      "loss": 1.3069,
+      "step": 1856
+    },
+    {
+      "epoch": 1.8058587251998215,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7167632746168503e-06,
+      "loss": 1.2939,
+      "step": 1857
+    },
+    {
+      "epoch": 1.806832474540512,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7143349709503786e-06,
+      "loss": 1.3234,
+      "step": 1858
+    },
+    {
+      "epoch": 1.8078062238812027,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7119074894839538e-06,
+      "loss": 1.33,
+      "step": 1859
+    },
+    {
+      "epoch": 1.8087799732218932,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7094808327579401e-06,
+      "loss": 1.3097,
+      "step": 1860
+    },
+    {
+      "epoch": 1.8097537225625837,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.7070550033118393e-06,
+      "loss": 1.3193,
+      "step": 1861
+    },
+    {
+      "epoch": 1.8107274719032742,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7046300036842864e-06,
+      "loss": 1.3302,
+      "step": 1862
+    },
+    {
+      "epoch": 1.8117012212439647,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.7022058364130478e-06,
+      "loss": 1.3024,
+      "step": 1863
+    },
+    {
+      "epoch": 1.8126749705846552,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6997825040350196e-06,
+      "loss": 1.299,
+      "step": 1864
+    },
+    {
+      "epoch": 1.813648719925346,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6973600090862247e-06,
+      "loss": 1.3096,
+      "step": 1865
+    },
+    {
+      "epoch": 1.8146224692660364,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.6949383541018088e-06,
+      "loss": 1.3065,
+      "step": 1866
+    },
+    {
+      "epoch": 1.815596218606727,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6925175416160387e-06,
+      "loss": 1.3274,
+      "step": 1867
+    },
+    {
+      "epoch": 1.8165699679474177,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6900975741622994e-06,
+      "loss": 1.3031,
+      "step": 1868
+    },
+    {
+      "epoch": 1.8175437172881082,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6876784542730918e-06,
+      "loss": 1.3244,
+      "step": 1869
+    },
+    {
+      "epoch": 1.8185174666287987,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6852601844800298e-06,
+      "loss": 1.3006,
+      "step": 1870
+    },
+    {
+      "epoch": 1.8194912159694892,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6828427673138378e-06,
+      "loss": 1.3155,
+      "step": 1871
+    },
+    {
+      "epoch": 1.8204649653101797,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6804262053043488e-06,
+      "loss": 1.3076,
+      "step": 1872
+    },
+    {
+      "epoch": 1.8214387146508701,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6780105009804976e-06,
+      "loss": 1.3134,
+      "step": 1873
+    },
+    {
+      "epoch": 1.8224124639915609,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6755956568703247e-06,
+      "loss": 1.3144,
+      "step": 1874
+    },
+    {
+      "epoch": 1.8233862133322514,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6731816755009696e-06,
+      "loss": 1.3084,
+      "step": 1875
+    },
+    {
+      "epoch": 1.824359962672942,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6707685593986687e-06,
+      "loss": 1.2964,
+      "step": 1876
+    },
+    {
+      "epoch": 1.8253337120136326,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6683563110887523e-06,
+      "loss": 1.308,
+      "step": 1877
+    },
+    {
+      "epoch": 1.826307461354323,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.665944933095644e-06,
+      "loss": 1.3091,
+      "step": 1878
+    },
+    {
+      "epoch": 1.8272812106950136,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6635344279428553e-06,
+      "loss": 1.3191,
+      "step": 1879
+    },
+    {
+      "epoch": 1.828254960035704,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6611247981529846e-06,
+      "loss": 1.3227,
+      "step": 1880
+    },
+    {
+      "epoch": 1.8292287093763946,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6587160462477149e-06,
+      "loss": 1.3134,
+      "step": 1881
+    },
+    {
+      "epoch": 1.830202458717085,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.6563081747478093e-06,
+      "loss": 1.3145,
+      "step": 1882
+    },
+    {
+      "epoch": 1.8311762080577758,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6539011861731115e-06,
+      "loss": 1.3177,
+      "step": 1883
+    },
+    {
+      "epoch": 1.8321499573984663,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.651495083042538e-06,
+      "loss": 1.3174,
+      "step": 1884
+    },
+    {
+      "epoch": 1.833123706739157,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6490898678740819e-06,
+      "loss": 1.3085,
+      "step": 1885
+    },
+    {
+      "epoch": 1.8340974560798475,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.646685543184805e-06,
+      "loss": 1.3278,
+      "step": 1886
+    },
+    {
+      "epoch": 1.835071205420538,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6442821114908385e-06,
+      "loss": 1.33,
+      "step": 1887
+    },
+    {
+      "epoch": 1.8360449547612285,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6418795753073785e-06,
+      "loss": 1.3125,
+      "step": 1888
+    },
+    {
+      "epoch": 1.837018704101919,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6394779371486838e-06,
+      "loss": 1.3102,
+      "step": 1889
+    },
+    {
+      "epoch": 1.8379924534426095,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6370771995280737e-06,
+      "loss": 1.3222,
+      "step": 1890
+    },
+    {
+      "epoch": 1.8389662027833003,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.634677364957925e-06,
+      "loss": 1.3038,
+      "step": 1891
+    },
+    {
+      "epoch": 1.8399399521239908,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6322784359496697e-06,
+      "loss": 1.335,
+      "step": 1892
+    },
+    {
+      "epoch": 1.8409137014646813,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.6298804150137914e-06,
+      "loss": 1.3319,
+      "step": 1893
+    },
+    {
+      "epoch": 1.841887450805372,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6274833046598254e-06,
+      "loss": 1.3062,
+      "step": 1894
+    },
+    {
+      "epoch": 1.8428612001460625,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6250871073963498e-06,
+      "loss": 1.314,
+      "step": 1895
+    },
+    {
+      "epoch": 1.843834949486753,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.6226918257309916e-06,
+      "loss": 1.3173,
+      "step": 1896
+    },
+    {
+      "epoch": 1.8448086988274435,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6202974621704176e-06,
+      "loss": 1.3172,
+      "step": 1897
+    },
+    {
+      "epoch": 1.845782448168134,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6179040192203343e-06,
+      "loss": 1.3011,
+      "step": 1898
+    },
+    {
+      "epoch": 1.8467561975088245,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6155114993854846e-06,
+      "loss": 1.3096,
+      "step": 1899
+    },
+    {
+      "epoch": 1.8477299468495152,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.613119905169645e-06,
+      "loss": 1.3185,
+      "step": 1900
+    },
+    {
+      "epoch": 1.8487036961902057,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.6107292390756241e-06,
+      "loss": 1.2993,
+      "step": 1901
+    },
+    {
+      "epoch": 1.8496774455308964,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6083395036052586e-06,
+      "loss": 1.3028,
+      "step": 1902
+    },
+    {
+      "epoch": 1.850651194871587,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6059507012594116e-06,
+      "loss": 1.29,
+      "step": 1903
+    },
+    {
+      "epoch": 1.8516249442122774,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.6035628345379695e-06,
+      "loss": 1.3278,
+      "step": 1904
+    },
+    {
+      "epoch": 1.852598693552968,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.601175905939841e-06,
+      "loss": 1.3225,
+      "step": 1905
+    },
+    {
+      "epoch": 1.8535724428936584,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5987899179629492e-06,
+      "loss": 1.3327,
+      "step": 1906
+    },
+    {
+      "epoch": 1.854546192234349,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.596404873104237e-06,
+      "loss": 1.3084,
+      "step": 1907
+    },
+    {
+      "epoch": 1.8555199415750394,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.594020773859658e-06,
+      "loss": 1.3321,
+      "step": 1908
+    },
+    {
+      "epoch": 1.8564936909157301,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5916376227241776e-06,
+      "loss": 1.3147,
+      "step": 1909
+    },
+    {
+      "epoch": 1.8574674402564206,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.5892554221917677e-06,
+      "loss": 1.3217,
+      "step": 1910
+    },
+    {
+      "epoch": 1.8584411895971114,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5868741747554061e-06,
+      "loss": 1.3225,
+      "step": 1911
+    },
+    {
+      "epoch": 1.8594149389378019,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5844938829070733e-06,
+      "loss": 1.3308,
+      "step": 1912
+    },
+    {
+      "epoch": 1.8603886882784924,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5821145491377494e-06,
+      "loss": 1.3021,
+      "step": 1913
+    },
+    {
+      "epoch": 1.8613624376191829,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.579736175937412e-06,
+      "loss": 1.3113,
+      "step": 1914
+    },
+    {
+      "epoch": 1.8623361869598734,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5773587657950338e-06,
+      "loss": 1.3239,
+      "step": 1915
+    },
+    {
+      "epoch": 1.8633099363005639,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5749823211985798e-06,
+      "loss": 1.323,
+      "step": 1916
+    },
+    {
+      "epoch": 1.8642836856412544,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5726068446350024e-06,
+      "loss": 1.3104,
+      "step": 1917
+    },
+    {
+      "epoch": 1.865257434981945,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5702323385902435e-06,
+      "loss": 1.3393,
+      "step": 1918
+    },
+    {
+      "epoch": 1.8662311843226356,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5678588055492289e-06,
+      "loss": 1.3089,
+      "step": 1919
+    },
+    {
+      "epoch": 1.8672049336633263,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5654862479958652e-06,
+      "loss": 1.306,
+      "step": 1920
+    },
+    {
+      "epoch": 1.8681786830040168,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.5631146684130389e-06,
+      "loss": 1.3229,
+      "step": 1921
+    },
+    {
+      "epoch": 1.8691524323447073,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5607440692826132e-06,
+      "loss": 1.3009,
+      "step": 1922
+    },
+    {
+      "epoch": 1.8701261816853978,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5583744530854243e-06,
+      "loss": 1.322,
+      "step": 1923
+    },
+    {
+      "epoch": 1.8710999310260883,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.5560058223012805e-06,
+      "loss": 1.3319,
+      "step": 1924
+    },
+    {
+      "epoch": 1.8720736803667788,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.553638179408959e-06,
+      "loss": 1.323,
+      "step": 1925
+    },
+    {
+      "epoch": 1.8730474297074695,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5512715268862033e-06,
+      "loss": 1.3077,
+      "step": 1926
+    },
+    {
+      "epoch": 1.87402117904816,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5489058672097195e-06,
+      "loss": 1.325,
+      "step": 1927
+    },
+    {
+      "epoch": 1.8749949283888507,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.546541202855175e-06,
+      "loss": 1.3195,
+      "step": 1928
+    },
+    {
+      "epoch": 1.8759686777295412,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5441775362971955e-06,
+      "loss": 1.3225,
+      "step": 1929
+    },
+    {
+      "epoch": 1.8769424270702317,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.541814870009364e-06,
+      "loss": 1.3365,
+      "step": 1930
+    },
+    {
+      "epoch": 1.8779161764109222,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5394532064642148e-06,
+      "loss": 1.321,
+      "step": 1931
+    },
+    {
+      "epoch": 1.8788899257516127,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5370925481332338e-06,
+      "loss": 1.3168,
+      "step": 1932
+    },
+    {
+      "epoch": 1.8798636750923032,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.534732897486855e-06,
+      "loss": 1.3231,
+      "step": 1933
+    },
+    {
+      "epoch": 1.8808374244329937,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5323742569944573e-06,
+      "loss": 1.3295,
+      "step": 1934
+    },
+    {
+      "epoch": 1.8818111737736845,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.530016629124363e-06,
+      "loss": 1.294,
+      "step": 1935
+    },
+    {
+      "epoch": 1.882784923114375,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5276600163438338e-06,
+      "loss": 1.2921,
+      "step": 1936
+    },
+    {
+      "epoch": 1.8837586724550657,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5253044211190705e-06,
+      "loss": 1.2998,
+      "step": 1937
+    },
+    {
+      "epoch": 1.8847324217957562,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.522949845915208e-06,
+      "loss": 1.3007,
+      "step": 1938
+    },
+    {
+      "epoch": 1.8857061711364467,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5205962931963135e-06,
+      "loss": 1.329,
+      "step": 1939
+    },
+    {
+      "epoch": 1.8866799204771372,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5182437654253856e-06,
+      "loss": 1.2925,
+      "step": 1940
+    },
+    {
+      "epoch": 1.8876536698178277,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.515892265064349e-06,
+      "loss": 1.2956,
+      "step": 1941
+    },
+    {
+      "epoch": 1.8886274191585182,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5135417945740533e-06,
+      "loss": 1.3301,
+      "step": 1942
+    },
+    {
+      "epoch": 1.8896011684992087,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5111923564142716e-06,
+      "loss": 1.3004,
+      "step": 1943
+    },
+    {
+      "epoch": 1.8905749178398994,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5088439530436943e-06,
+      "loss": 1.3135,
+      "step": 1944
+    },
+    {
+      "epoch": 1.89154866718059,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5064965869199316e-06,
+      "loss": 1.3066,
+      "step": 1945
+    },
+    {
+      "epoch": 1.8925224165212806,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5041502604995056e-06,
+      "loss": 1.3019,
+      "step": 1946
+    },
+    {
+      "epoch": 1.8934961658619711,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5018049762378528e-06,
+      "loss": 1.3055,
+      "step": 1947
+    },
+    {
+      "epoch": 1.8944699152026616,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4994607365893173e-06,
+      "loss": 1.3339,
+      "step": 1948
+    },
+    {
+      "epoch": 1.8954436645433521,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4971175440071516e-06,
+      "loss": 1.3305,
+      "step": 1949
+    },
+    {
+      "epoch": 1.8964174138840426,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.49477540094351e-06,
+      "loss": 1.3034,
+      "step": 1950
+    },
+    {
+      "epoch": 1.8973911632247331,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.492434309849451e-06,
+      "loss": 1.3158,
+      "step": 1951
+    },
+    {
+      "epoch": 1.8983649125654238,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4900942731749314e-06,
+      "loss": 1.3214,
+      "step": 1952
+    },
+    {
+      "epoch": 1.8993386619061143,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4877552933688033e-06,
+      "loss": 1.3366,
+      "step": 1953
+    },
+    {
+      "epoch": 1.9003124112468048,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4854173728788144e-06,
+      "loss": 1.3046,
+      "step": 1954
+    },
+    {
+      "epoch": 1.9012861605874956,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.483080514151603e-06,
+      "loss": 1.3136,
+      "step": 1955
+    },
+    {
+      "epoch": 1.902259909928186,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4807447196326967e-06,
+      "loss": 1.3015,
+      "step": 1956
+    },
+    {
+      "epoch": 1.9032336592688766,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4784099917665094e-06,
+      "loss": 1.3246,
+      "step": 1957
+    },
+    {
+      "epoch": 1.904207408609567,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4760763329963378e-06,
+      "loss": 1.3252,
+      "step": 1958
+    },
+    {
+      "epoch": 1.9051811579502576,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4737437457643616e-06,
+      "loss": 1.3049,
+      "step": 1959
+    },
+    {
+      "epoch": 1.906154907290948,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.471412232511638e-06,
+      "loss": 1.3029,
+      "step": 1960
+    },
+    {
+      "epoch": 1.9071286566316388,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4690817956781e-06,
+      "loss": 1.3111,
+      "step": 1961
+    },
+    {
+      "epoch": 1.9081024059723293,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4667524377025535e-06,
+      "loss": 1.3004,
+      "step": 1962
+    },
+    {
+      "epoch": 1.90907615531302,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4644241610226776e-06,
+      "loss": 1.3212,
+      "step": 1963
+    },
+    {
+      "epoch": 1.9100499046537105,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.462096968075018e-06,
+      "loss": 1.3024,
+      "step": 1964
+    },
+    {
+      "epoch": 1.911023653994401,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.459770861294987e-06,
+      "loss": 1.2928,
+      "step": 1965
+    },
+    {
+      "epoch": 1.9119974033350915,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.45744584311686e-06,
+      "loss": 1.3056,
+      "step": 1966
+    },
+    {
+      "epoch": 1.912971152675782,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4551219159737728e-06,
+      "loss": 1.3234,
+      "step": 1967
+    },
+    {
+      "epoch": 1.9139449020164725,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4527990822977216e-06,
+      "loss": 1.3294,
+      "step": 1968
+    },
+    {
+      "epoch": 1.914918651357163,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4504773445195544e-06,
+      "loss": 1.315,
+      "step": 1969
+    },
+    {
+      "epoch": 1.9158924006978537,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4481567050689764e-06,
+      "loss": 1.303,
+      "step": 1970
+    },
+    {
+      "epoch": 1.9168661500385442,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.44583716637454e-06,
+      "loss": 1.3114,
+      "step": 1971
+    },
+    {
+      "epoch": 1.917839899379235,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4435187308636486e-06,
+      "loss": 1.3197,
+      "step": 1972
+    },
+    {
+      "epoch": 1.9188136487199254,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4412014009625476e-06,
+      "loss": 1.3281,
+      "step": 1973
+    },
+    {
+      "epoch": 1.919787398060616,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.438885179096329e-06,
+      "loss": 1.3262,
+      "step": 1974
+    },
+    {
+      "epoch": 1.9207611474013064,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4365700676889227e-06,
+      "loss": 1.3261,
+      "step": 1975
+    },
+    {
+      "epoch": 1.921734896741997,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4342560691630991e-06,
+      "loss": 1.3037,
+      "step": 1976
+    },
+    {
+      "epoch": 1.9227086460826874,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4319431859404603e-06,
+      "loss": 1.3175,
+      "step": 1977
+    },
+    {
+      "epoch": 1.923682395423378,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4296314204414453e-06,
+      "loss": 1.3127,
+      "step": 1978
+    },
+    {
+      "epoch": 1.9246561447640687,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.427320775085319e-06,
+      "loss": 1.3142,
+      "step": 1979
+    },
+    {
+      "epoch": 1.9256298941047592,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4250112522901794e-06,
+      "loss": 1.3255,
+      "step": 1980
+    },
+    {
+      "epoch": 1.9266036434454499,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4227028544729448e-06,
+      "loss": 1.3185,
+      "step": 1981
+    },
+    {
+      "epoch": 1.9275773927861404,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4203955840493588e-06,
+      "loss": 1.3147,
+      "step": 1982
+    },
+    {
+      "epoch": 1.9285511421268309,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.4180894434339836e-06,
+      "loss": 1.3391,
+      "step": 1983
+    },
+    {
+      "epoch": 1.9295248914675214,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4157844350402017e-06,
+      "loss": 1.2939,
+      "step": 1984
+    },
+    {
+      "epoch": 1.9304986408082119,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4134805612802072e-06,
+      "loss": 1.3273,
+      "step": 1985
+    },
+    {
+      "epoch": 1.9314723901489024,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4111778245650107e-06,
+      "loss": 1.3191,
+      "step": 1986
+    },
+    {
+      "epoch": 1.932446139489593,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.4088762273044287e-06,
+      "loss": 1.3335,
+      "step": 1987
+    },
+    {
+      "epoch": 1.9334198888302836,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4065757719070896e-06,
+      "loss": 1.3201,
+      "step": 1988
+    },
+    {
+      "epoch": 1.9343936381709743,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4042764607804238e-06,
+      "loss": 1.3164,
+      "step": 1989
+    },
+    {
+      "epoch": 1.9353673875116648,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4019782963306636e-06,
+      "loss": 1.3188,
+      "step": 1990
+    },
+    {
+      "epoch": 1.9363411368523553,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.399681280962845e-06,
+      "loss": 1.3138,
+      "step": 1991
+    },
+    {
+      "epoch": 1.9373148861930458,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.397385417080797e-06,
+      "loss": 1.3319,
+      "step": 1992
+    },
+    {
+      "epoch": 1.9382886355337363,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3950907070871483e-06,
+      "loss": 1.2966,
+      "step": 1993
+    },
+    {
+      "epoch": 1.9392623848744268,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3927971533833163e-06,
+      "loss": 1.3071,
+      "step": 1994
+    },
+    {
+      "epoch": 1.9402361342151173,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3905047583695086e-06,
+      "loss": 1.3234,
+      "step": 1995
+    },
+    {
+      "epoch": 1.941209883555808,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.388213524444723e-06,
+      "loss": 1.3245,
+      "step": 1996
+    },
+    {
+      "epoch": 1.9421836328964985,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3859234540067397e-06,
+      "loss": 1.2904,
+      "step": 1997
+    },
+    {
+      "epoch": 1.9431573822371893,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3836345494521215e-06,
+      "loss": 1.3329,
+      "step": 1998
+    },
+    {
+      "epoch": 1.9441311315778798,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3813468131762126e-06,
+      "loss": 1.3248,
+      "step": 1999
+    },
+    {
+      "epoch": 1.9451048809185703,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3790602475731323e-06,
+      "loss": 1.3367,
+      "step": 2000
+    },
+    {
+      "epoch": 1.9460786302592608,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3767748550357785e-06,
+      "loss": 1.3169,
+      "step": 2001
+    },
+    {
+      "epoch": 1.9470523795999513,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.3744906379558165e-06,
+      "loss": 1.3082,
+      "step": 2002
+    },
+    {
+      "epoch": 1.9480261289406418,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3722075987236865e-06,
+      "loss": 1.3302,
+      "step": 2003
+    },
+    {
+      "epoch": 1.9489998782813323,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3699257397285926e-06,
+      "loss": 1.3229,
+      "step": 2004
+    },
+    {
+      "epoch": 1.949973627622023,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3676450633585037e-06,
+      "loss": 1.2998,
+      "step": 2005
+    },
+    {
+      "epoch": 1.9509473769627135,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.365365572000154e-06,
+      "loss": 1.3457,
+      "step": 2006
+    },
+    {
+      "epoch": 1.9519211263034042,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3630872680390356e-06,
+      "loss": 1.3216,
+      "step": 2007
+    },
+    {
+      "epoch": 1.9528948756440947,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3608101538593965e-06,
+      "loss": 1.3328,
+      "step": 2008
+    },
+    {
+      "epoch": 1.9538686249847852,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3585342318442434e-06,
+      "loss": 1.2837,
+      "step": 2009
+    },
+    {
+      "epoch": 1.9548423743254757,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3562595043753318e-06,
+      "loss": 1.32,
+      "step": 2010
+    },
+    {
+      "epoch": 1.9558161236661662,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3539859738331707e-06,
+      "loss": 1.3208,
+      "step": 2011
+    },
+    {
+      "epoch": 1.9567898730068567,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3517136425970115e-06,
+      "loss": 1.3032,
+      "step": 2012
+    },
+    {
+      "epoch": 1.9577636223475474,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3494425130448562e-06,
+      "loss": 1.3074,
+      "step": 2013
+    },
+    {
+      "epoch": 1.958737371688238,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.347172587553444e-06,
+      "loss": 1.3137,
+      "step": 2014
+    },
+    {
+      "epoch": 1.9597111210289284,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3449038684982602e-06,
+      "loss": 1.3113,
+      "step": 2015
+    },
+    {
+      "epoch": 1.9606848703696191,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3426363582535193e-06,
+      "loss": 1.3086,
+      "step": 2016
+    },
+    {
+      "epoch": 1.9616586197103096,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3403700591921787e-06,
+      "loss": 1.3039,
+      "step": 2017
+    },
+    {
+      "epoch": 1.9626323690510001,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3381049736859225e-06,
+      "loss": 1.3114,
+      "step": 2018
+    },
+    {
+      "epoch": 1.9636061183916906,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.33584110410517e-06,
+      "loss": 1.319,
+      "step": 2019
+    },
+    {
+      "epoch": 1.9645798677323811,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3335784528190627e-06,
+      "loss": 1.3085,
+      "step": 2020
+    },
+    {
+      "epoch": 1.9655536170730716,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3313170221954719e-06,
+      "loss": 1.3145,
+      "step": 2021
+    },
+    {
+      "epoch": 1.9665273664137624,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3290568146009874e-06,
+      "loss": 1.3138,
+      "step": 2022
+    },
+    {
+      "epoch": 1.9675011157544529,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3267978324009235e-06,
+      "loss": 1.3036,
+      "step": 2023
+    },
+    {
+      "epoch": 1.9684748650951436,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3245400779593067e-06,
+      "loss": 1.3135,
+      "step": 2024
+    },
+    {
+      "epoch": 1.969448614435834,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3222835536388845e-06,
+      "loss": 1.307,
+      "step": 2025
+    },
+    {
+      "epoch": 1.9704223637765246,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3200282618011115e-06,
+      "loss": 1.3144,
+      "step": 2026
+    },
+    {
+      "epoch": 1.971396113117215,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3177742048061587e-06,
+      "loss": 1.3344,
+      "step": 2027
+    },
+    {
+      "epoch": 1.9723698624579056,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3155213850128968e-06,
+      "loss": 1.2997,
+      "step": 2028
+    },
+    {
+      "epoch": 1.973343611798596,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.31326980477891e-06,
+      "loss": 1.3037,
+      "step": 2029
+    },
+    {
+      "epoch": 1.9743173611392866,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3110194664604786e-06,
+      "loss": 1.3135,
+      "step": 2030
+    },
+    {
+      "epoch": 1.9752911104799773,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3087703724125895e-06,
+      "loss": 1.3181,
+      "step": 2031
+    },
+    {
+      "epoch": 1.9762648598206678,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.3065225249889208e-06,
+      "loss": 1.3113,
+      "step": 2032
+    },
+    {
+      "epoch": 1.9772386091613585,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.3042759265418523e-06,
+      "loss": 1.3233,
+      "step": 2033
+    },
+    {
+      "epoch": 1.978212358502049,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3020305794224516e-06,
+      "loss": 1.3193,
+      "step": 2034
+    },
+    {
+      "epoch": 1.9791861078427395,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.299786485980481e-06,
+      "loss": 1.3201,
+      "step": 2035
+    },
+    {
+      "epoch": 1.98015985718343,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2975436485643865e-06,
+      "loss": 1.3062,
+      "step": 2036
+    },
+    {
+      "epoch": 1.9811336065241205,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2953020695213048e-06,
+      "loss": 1.3092,
+      "step": 2037
+    },
+    {
+      "epoch": 1.982107355864811,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2930617511970514e-06,
+      "loss": 1.3274,
+      "step": 2038
+    },
+    {
+      "epoch": 1.9830811052055017,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2908226959361241e-06,
+      "loss": 1.293,
+      "step": 2039
+    },
+    {
+      "epoch": 1.9840548545461922,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2885849060816973e-06,
+      "loss": 1.304,
+      "step": 2040
+    },
+    {
+      "epoch": 1.9850286038868827,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2863483839756254e-06,
+      "loss": 1.3259,
+      "step": 2041
+    },
+    {
+      "epoch": 1.9860023532275735,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.284113131958431e-06,
+      "loss": 1.3236,
+      "step": 2042
+    },
+    {
+      "epoch": 1.986976102568264,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2818791523693114e-06,
+      "loss": 1.3307,
+      "step": 2043
+    },
+    {
+      "epoch": 1.9879498519089545,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2796464475461296e-06,
+      "loss": 1.317,
+      "step": 2044
+    },
+    {
+      "epoch": 1.988923601249645,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.277415019825417e-06,
+      "loss": 1.3171,
+      "step": 2045
+    },
+    {
+      "epoch": 1.9898973505903355,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.275184871542366e-06,
+      "loss": 1.3081,
+      "step": 2046
+    },
+    {
+      "epoch": 1.990871099931026,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.2729560050308325e-06,
+      "loss": 1.328,
+      "step": 2047
+    },
+    {
+      "epoch": 1.9918448492717167,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2707284226233283e-06,
+      "loss": 1.3265,
+      "step": 2048
+    },
+    {
+      "epoch": 1.9928185986124072,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2685021266510256e-06,
+      "loss": 1.3072,
+      "step": 2049
+    },
+    {
+      "epoch": 1.993792347953098,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.266277119443744e-06,
+      "loss": 1.3155,
+      "step": 2050
+    },
+    {
+      "epoch": 1.9947660972937884,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.264053403329961e-06,
+      "loss": 1.3152,
+      "step": 2051
+    },
+    {
+      "epoch": 1.995739846634479,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.261830980636798e-06,
+      "loss": 1.2953,
+      "step": 2052
+    },
+    {
+      "epoch": 1.9967135959751694,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2596098536900263e-06,
+      "loss": 1.3345,
+      "step": 2053
+    },
+    {
+      "epoch": 1.99768734531586,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2573900248140586e-06,
+      "loss": 1.3066,
+      "step": 2054
+    },
+    {
+      "epoch": 1.9986610946565504,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.255171496331952e-06,
+      "loss": 1.3078,
+      "step": 2055
+    },
+    {
+      "epoch": 1.999634843997241,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2529542705653992e-06,
+      "loss": 1.3204,
+      "step": 2056
+    },
+    {
+      "epoch": 2.0006085933379314,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.2507383498347328e-06,
+      "loss": 1.3295,
+      "step": 2057
+    },
+    {
+      "epoch": 2.0015823426786223,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.248523736458917e-06,
+      "loss": 1.3092,
+      "step": 2058
+    },
+    {
+      "epoch": 2.0015823426786223,
+      "eval_loss": 1.337652564048767,
+      "eval_runtime": 1517.6913,
+      "eval_samples_per_second": 27.512,
+      "eval_steps_per_second": 3.439,
+      "step": 2058
+    },
+    {
+      "epoch": 2.000162265222506,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.246310432755551e-06,
+      "loss": 1.3348,
+      "step": 2059
+    },
+    {
+      "epoch": 2.0011358565575432,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.2440984410408607e-06,
+      "loss": 1.3299,
+      "step": 2060
+    },
+    {
+      "epoch": 2.0021094478925803,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2418877636297e-06,
+      "loss": 1.3285,
+      "step": 2061
+    },
+    {
+      "epoch": 2.0030830392276173,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.239678402835546e-06,
+      "loss": 1.2974,
+      "step": 2062
+    },
+    {
+      "epoch": 2.004056630562655,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2374703609705014e-06,
+      "loss": 1.3299,
+      "step": 2063
+    },
+    {
+      "epoch": 2.005030221897692,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2352636403452844e-06,
+      "loss": 1.3071,
+      "step": 2064
+    },
+    {
+      "epoch": 2.006003813232729,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2330582432692349e-06,
+      "loss": 1.3317,
+      "step": 2065
+    },
+    {
+      "epoch": 2.006977404567766,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2308541720503029e-06,
+      "loss": 1.3287,
+      "step": 2066
+    },
+    {
+      "epoch": 2.007950995902803,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.228651428995056e-06,
+      "loss": 1.3298,
+      "step": 2067
+    },
+    {
+      "epoch": 2.00892458723784,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2264500164086667e-06,
+      "loss": 1.325,
+      "step": 2068
+    },
+    {
+      "epoch": 2.0098981785728776,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2242499365949203e-06,
+      "loss": 1.3052,
+      "step": 2069
+    },
+    {
+      "epoch": 2.0108717699079146,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2220511918562036e-06,
+      "loss": 1.3097,
+      "step": 2070
+    },
+    {
+      "epoch": 2.0118453612429517,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.219853784493507e-06,
+      "loss": 1.3139,
+      "step": 2071
+    },
+    {
+      "epoch": 2.0128189525779887,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2176577168064213e-06,
+      "loss": 1.3104,
+      "step": 2072
+    },
+    {
+      "epoch": 2.0137925439130258,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2154629910931376e-06,
+      "loss": 1.3206,
+      "step": 2073
+    },
+    {
+      "epoch": 2.014766135248063,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.213269609650439e-06,
+      "loss": 1.3183,
+      "step": 2074
+    },
+    {
+      "epoch": 2.0157397265831003,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.2110775747737052e-06,
+      "loss": 1.3096,
+      "step": 2075
+    },
+    {
+      "epoch": 2.0167133179181373,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2088868887569036e-06,
+      "loss": 1.3128,
+      "step": 2076
+    },
+    {
+      "epoch": 2.0176869092531744,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.206697553892593e-06,
+      "loss": 1.3306,
+      "step": 2077
+    },
+    {
+      "epoch": 2.0186605005882114,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2045095724719156e-06,
+      "loss": 1.3094,
+      "step": 2078
+    },
+    {
+      "epoch": 2.0196340919232485,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2023229467845996e-06,
+      "loss": 1.342,
+      "step": 2079
+    },
+    {
+      "epoch": 2.0206076832582855,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2001376791189526e-06,
+      "loss": 1.3286,
+      "step": 2080
+    },
+    {
+      "epoch": 2.0215812745933226,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.1979537717618605e-06,
+      "loss": 1.3285,
+      "step": 2081
+    },
+    {
+      "epoch": 2.02255486592836,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.195771226998789e-06,
+      "loss": 1.3068,
+      "step": 2082
+    },
+    {
+      "epoch": 2.023528457263397,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.1935900471137742e-06,
+      "loss": 1.3012,
+      "step": 2083
+    },
+    {
+      "epoch": 2.024502048598434,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.191410234389425e-06,
+      "loss": 1.3158,
+      "step": 2084
+    },
+    {
+      "epoch": 2.0254756399334712,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.1892317911069212e-06,
+      "loss": 1.3224,
+      "step": 2085
+    },
+    {
+      "epoch": 2.0264492312685083,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.1870547195460063e-06,
+      "loss": 1.3266,
+      "step": 2086
+    },
+    {
+      "epoch": 2.0274228226035453,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1848790219849923e-06,
+      "loss": 1.321,
+      "step": 2087
+    },
+    {
+      "epoch": 2.028396413938583,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1827047007007497e-06,
+      "loss": 1.3121,
+      "step": 2088
+    },
+    {
+      "epoch": 2.02937000527362,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.1805317579687095e-06,
+      "loss": 1.3191,
+      "step": 2089
+    },
+    {
+      "epoch": 2.030343596608657,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.1783601960628624e-06,
+      "loss": 1.3041,
+      "step": 2090
+    },
+    {
+      "epoch": 2.031317187943694,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.1761900172557508e-06,
+      "loss": 1.3053,
+      "step": 2091
+    },
+    {
+      "epoch": 2.032290779278731,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1740212238184726e-06,
+      "loss": 1.3006,
+      "step": 2092
+    },
+    {
+      "epoch": 2.033264370613768,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.171853818020674e-06,
+      "loss": 1.3158,
+      "step": 2093
+    },
+    {
+      "epoch": 2.034237961948805,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.1696878021305483e-06,
+      "loss": 1.3132,
+      "step": 2094
+    },
+    {
+      "epoch": 2.0352115532838426,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.1675231784148378e-06,
+      "loss": 1.3303,
+      "step": 2095
+    },
+    {
+      "epoch": 2.0361851446188797,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.1653599491388234e-06,
+      "loss": 1.3153,
+      "step": 2096
+    },
+    {
+      "epoch": 2.0371587359539167,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1631981165663308e-06,
+      "loss": 1.3044,
+      "step": 2097
+    },
+    {
+      "epoch": 2.0381323272889538,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.1610376829597215e-06,
+      "loss": 1.3269,
+      "step": 2098
+    },
+    {
+      "epoch": 2.039105918623991,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.1588786505798924e-06,
+      "loss": 1.3155,
+      "step": 2099
+    },
+    {
+      "epoch": 2.040079509959028,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1567210216862773e-06,
+      "loss": 1.314,
+      "step": 2100
+    },
+    {
+      "epoch": 2.0410531012940654,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.1545647985368375e-06,
+      "loss": 1.3135,
+      "step": 2101
+    },
+    {
+      "epoch": 2.0420266926291024,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.1524099833880667e-06,
+      "loss": 1.3141,
+      "step": 2102
+    },
+    {
+      "epoch": 2.0430002839641395,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.150256578494981e-06,
+      "loss": 1.3114,
+      "step": 2103
+    },
+    {
+      "epoch": 2.0439738752991765,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.1481045861111256e-06,
+      "loss": 1.3159,
+      "step": 2104
+    },
+    {
+      "epoch": 2.0449474666342136,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.1459540084885635e-06,
+      "loss": 1.3034,
+      "step": 2105
+    },
+    {
+      "epoch": 2.0459210579692506,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.1438048478778786e-06,
+      "loss": 1.3174,
+      "step": 2106
+    },
+    {
+      "epoch": 2.0468946493042877,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1416571065281706e-06,
+      "loss": 1.3028,
+      "step": 2107
+    },
+    {
+      "epoch": 2.047868240639325,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.1395107866870579e-06,
+      "loss": 1.3072,
+      "step": 2108
+    },
+    {
+      "epoch": 2.048841831974362,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.1373658906006656e-06,
+      "loss": 1.2953,
+      "step": 2109
+    },
+    {
+      "epoch": 2.0498154233093993,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.1352224205136348e-06,
+      "loss": 1.2935,
+      "step": 2110
+    },
+    {
+      "epoch": 2.0507890146444363,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.1330803786691086e-06,
+      "loss": 1.3207,
+      "step": 2111
+    },
+    {
+      "epoch": 2.0517626059794734,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.1309397673087405e-06,
+      "loss": 1.3186,
+      "step": 2112
+    },
+    {
+      "epoch": 2.0527361973145104,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.1288005886726823e-06,
+      "loss": 1.2882,
+      "step": 2113
+    },
+    {
+      "epoch": 2.053709788649548,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.1266628449995913e-06,
+      "loss": 1.2869,
+      "step": 2114
+    },
+    {
+      "epoch": 2.054683379984585,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1245265385266187e-06,
+      "loss": 1.3011,
+      "step": 2115
+    },
+    {
+      "epoch": 2.055656971319622,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1223916714894146e-06,
+      "loss": 1.3219,
+      "step": 2116
+    },
+    {
+      "epoch": 2.056630562654659,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.1202582461221202e-06,
+      "loss": 1.3128,
+      "step": 2117
+    },
+    {
+      "epoch": 2.057604153989696,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1181262646573712e-06,
+      "loss": 1.3329,
+      "step": 2118
+    },
+    {
+      "epoch": 2.058577745324733,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.1159957293262888e-06,
+      "loss": 1.2994,
+      "step": 2119
+    },
+    {
+      "epoch": 2.05955133665977,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.1138666423584848e-06,
+      "loss": 1.3412,
+      "step": 2120
+    },
+    {
+      "epoch": 2.0605249279948077,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.1117390059820507e-06,
+      "loss": 1.3094,
+      "step": 2121
+    },
+    {
+      "epoch": 2.0614985193298447,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.1096128224235644e-06,
+      "loss": 1.2937,
+      "step": 2122
+    },
+    {
+      "epoch": 2.062472110664882,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.1074880939080793e-06,
+      "loss": 1.338,
+      "step": 2123
+    },
+    {
+      "epoch": 2.063445701999919,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1053648226591298e-06,
+      "loss": 1.3177,
+      "step": 2124
+    },
+    {
+      "epoch": 2.064419293334956,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.1032430108987221e-06,
+      "loss": 1.322,
+      "step": 2125
+    },
+    {
+      "epoch": 2.065392884669993,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.1011226608473391e-06,
+      "loss": 1.3191,
+      "step": 2126
+    },
+    {
+      "epoch": 2.0663664760050304,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0990037747239278e-06,
+      "loss": 1.3111,
+      "step": 2127
+    },
+    {
+      "epoch": 2.0673400673400675,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0968863547459096e-06,
+      "loss": 1.3288,
+      "step": 2128
+    },
+    {
+      "epoch": 2.0683136586751045,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0947704031291668e-06,
+      "loss": 1.3115,
+      "step": 2129
+    },
+    {
+      "epoch": 2.0692872500101416,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.092655922088049e-06,
+      "loss": 1.3068,
+      "step": 2130
+    },
+    {
+      "epoch": 2.0702608413451786,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.0905429138353628e-06,
+      "loss": 1.3019,
+      "step": 2131
+    },
+    {
+      "epoch": 2.0712344326802157,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0884313805823783e-06,
+      "loss": 1.2997,
+      "step": 2132
+    },
+    {
+      "epoch": 2.0722080240152527,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0863213245388167e-06,
+      "loss": 1.3196,
+      "step": 2133
+    },
+    {
+      "epoch": 2.07318161535029,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.0842127479128579e-06,
+      "loss": 1.3132,
+      "step": 2134
+    },
+    {
+      "epoch": 2.0741552066853273,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0821056529111296e-06,
+      "loss": 1.3221,
+      "step": 2135
+    },
+    {
+      "epoch": 2.0751287980203643,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.0800000417387132e-06,
+      "loss": 1.3214,
+      "step": 2136
+    },
+    {
+      "epoch": 2.0761023893554014,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.0778959165991342e-06,
+      "loss": 1.3201,
+      "step": 2137
+    },
+    {
+      "epoch": 2.0770759806904384,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.0757932796943637e-06,
+      "loss": 1.3115,
+      "step": 2138
+    },
+    {
+      "epoch": 2.0780495720254755,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.0736921332248145e-06,
+      "loss": 1.3333,
+      "step": 2139
+    },
+    {
+      "epoch": 2.079023163360513,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.0715924793893425e-06,
+      "loss": 1.3109,
+      "step": 2140
+    },
+    {
+      "epoch": 2.07999675469555,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.0694943203852382e-06,
+      "loss": 1.3066,
+      "step": 2141
+    },
+    {
+      "epoch": 2.080970346030587,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.0673976584082304e-06,
+      "loss": 1.3067,
+      "step": 2142
+    },
+    {
+      "epoch": 2.081943937365624,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.0653024956524788e-06,
+      "loss": 1.3261,
+      "step": 2143
+    },
+    {
+      "epoch": 2.082917528700661,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.063208834310577e-06,
+      "loss": 1.3376,
+      "step": 2144
+    },
+    {
+      "epoch": 2.083891120035698,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.0611166765735442e-06,
+      "loss": 1.3078,
+      "step": 2145
+    },
+    {
+      "epoch": 2.0848647113707353,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0590260246308296e-06,
+      "loss": 1.3155,
+      "step": 2146
+    },
+    {
+      "epoch": 2.0858383027057728,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.0569368806703029e-06,
+      "loss": 1.3193,
+      "step": 2147
+    },
+    {
+      "epoch": 2.08681189404081,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.0548492468782598e-06,
+      "loss": 1.3227,
+      "step": 2148
+    },
+    {
+      "epoch": 2.087785485375847,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.0527631254394103e-06,
+      "loss": 1.3228,
+      "step": 2149
+    },
+    {
+      "epoch": 2.088759076710884,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.050678518536887e-06,
+      "loss": 1.3149,
+      "step": 2150
+    },
+    {
+      "epoch": 2.089732668045921,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.0485954283522335e-06,
+      "loss": 1.3298,
+      "step": 2151
+    },
+    {
+      "epoch": 2.090706259380958,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0465138570654096e-06,
+      "loss": 1.3265,
+      "step": 2152
+    },
+    {
+      "epoch": 2.0916798507159955,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.0444338068547824e-06,
+      "loss": 1.3284,
+      "step": 2153
+    },
+    {
+      "epoch": 2.0926534420510325,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.0423552798971303e-06,
+      "loss": 1.3074,
+      "step": 2154
+    },
+    {
+      "epoch": 2.0936270333860696,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0402782783676343e-06,
+      "loss": 1.2978,
+      "step": 2155
+    },
+    {
+      "epoch": 2.0946006247211066,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0382028044398823e-06,
+      "loss": 1.3092,
+      "step": 2156
+    },
+    {
+      "epoch": 2.0955742160561437,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0361288602858605e-06,
+      "loss": 1.3075,
+      "step": 2157
+    },
+    {
+      "epoch": 2.0965478073911807,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.0340564480759568e-06,
+      "loss": 1.3138,
+      "step": 2158
+    },
+    {
+      "epoch": 2.097521398726218,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.0319855699789546e-06,
+      "loss": 1.3134,
+      "step": 2159
+    },
+    {
+      "epoch": 2.0984949900612553,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.0299162281620318e-06,
+      "loss": 1.3094,
+      "step": 2160
+    },
+    {
+      "epoch": 2.0994685813962923,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0278484247907578e-06,
+      "loss": 1.3168,
+      "step": 2161
+    },
+    {
+      "epoch": 2.1004421727313294,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.0257821620290948e-06,
+      "loss": 1.3154,
+      "step": 2162
+    },
+    {
+      "epoch": 2.1014157640663664,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.0237174420393894e-06,
+      "loss": 1.3323,
+      "step": 2163
+    },
+    {
+      "epoch": 2.1023893554014035,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.0216542669823768e-06,
+      "loss": 1.3194,
+      "step": 2164
+    },
+    {
+      "epoch": 2.1033629467364405,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0195926390171724e-06,
+      "loss": 1.3194,
+      "step": 2165
+    },
+    {
+      "epoch": 2.104336538071478,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.0175325603012754e-06,
+      "loss": 1.3186,
+      "step": 2166
+    },
+    {
+      "epoch": 2.105310129406515,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.0154740329905607e-06,
+      "loss": 1.32,
+      "step": 2167
+    },
+    {
+      "epoch": 2.106283720741552,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0134170592392837e-06,
+      "loss": 1.3077,
+      "step": 2168
+    },
+    {
+      "epoch": 2.107257312076589,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0113616412000688e-06,
+      "loss": 1.3067,
+      "step": 2169
+    },
+    {
+      "epoch": 2.1082309034116262,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.0093077810239185e-06,
+      "loss": 1.328,
+      "step": 2170
+    },
+    {
+      "epoch": 2.1092044947466633,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.0072554808601981e-06,
+      "loss": 1.3042,
+      "step": 2171
+    },
+    {
+      "epoch": 2.1101780860817003,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0052047428566461e-06,
+      "loss": 1.3384,
+      "step": 2172
+    },
+    {
+      "epoch": 2.111151677416738,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.0031555691593627e-06,
+      "loss": 1.3132,
+      "step": 2173
+    },
+    {
+      "epoch": 2.112125268751775,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.001107961912814e-06,
+      "loss": 1.303,
+      "step": 2174
+    },
+    {
+      "epoch": 2.113098860086812,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.990619232598228e-07,
+      "loss": 1.2988,
+      "step": 2175
+    },
+    {
+      "epoch": 2.114072451421849,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.970174553415754e-07,
+      "loss": 1.2889,
+      "step": 2176
+    },
+    {
+      "epoch": 2.115046042756886,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.94974560297609e-07,
+      "loss": 1.3143,
+      "step": 2177
+    },
+    {
+      "epoch": 2.116019634091923,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.929332402658198e-07,
+      "loss": 1.3233,
+      "step": 2178
+    },
+    {
+      "epoch": 2.1169932254269606,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.908934973824522e-07,
+      "loss": 1.3423,
+      "step": 2179
+    },
+    {
+      "epoch": 2.1179668167619976,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.888553337821005e-07,
+      "loss": 1.3449,
+      "step": 2180
+    },
+    {
+      "epoch": 2.1189404080970347,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.868187515977085e-07,
+      "loss": 1.3163,
+      "step": 2181
+    },
+    {
+      "epoch": 2.1199139994320717,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.847837529605631e-07,
+      "loss": 1.3289,
+      "step": 2182
+    },
+    {
+      "epoch": 2.1208875907671088,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.827503400002933e-07,
+      "loss": 1.3202,
+      "step": 2183
+    },
+    {
+      "epoch": 2.121861182102146,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.80718514844872e-07,
+      "loss": 1.3237,
+      "step": 2184
+    },
+    {
+      "epoch": 2.122834773437183,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.786882796206062e-07,
+      "loss": 1.3416,
+      "step": 2185
+    },
+    {
+      "epoch": 2.1238083647722203,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.76659636452143e-07,
+      "loss": 1.325,
+      "step": 2186
+    },
+    {
+      "epoch": 2.1247819561072574,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.74632587462461e-07,
+      "loss": 1.3403,
+      "step": 2187
+    },
+    {
+      "epoch": 2.1257555474422944,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.726071347728702e-07,
+      "loss": 1.2848,
+      "step": 2188
+    },
+    {
+      "epoch": 2.1267291387773315,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.705832805030124e-07,
+      "loss": 1.3335,
+      "step": 2189
+    },
+    {
+      "epoch": 2.1277027301123685,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.685610267708533e-07,
+      "loss": 1.3331,
+      "step": 2190
+    },
+    {
+      "epoch": 2.1286763214474056,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.66540375692688e-07,
+      "loss": 1.3259,
+      "step": 2191
+    },
+    {
+      "epoch": 2.129649912782443,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.645213293831304e-07,
+      "loss": 1.3321,
+      "step": 2192
+    },
+    {
+      "epoch": 2.13062350411748,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.625038899551162e-07,
+      "loss": 1.3234,
+      "step": 2193
+    },
+    {
+      "epoch": 2.131597095452517,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.604880595199011e-07,
+      "loss": 1.3221,
+      "step": 2194
+    },
+    {
+      "epoch": 2.1325706867875542,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.584738401870545e-07,
+      "loss": 1.3266,
+      "step": 2195
+    },
+    {
+      "epoch": 2.1335442781225913,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.564612340644622e-07,
+      "loss": 1.2918,
+      "step": 2196
+    },
+    {
+      "epoch": 2.1345178694576283,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.5445024325832e-07,
+      "loss": 1.2987,
+      "step": 2197
+    },
+    {
+      "epoch": 2.1354914607926654,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.52440869873133e-07,
+      "loss": 1.338,
+      "step": 2198
+    },
+    {
+      "epoch": 2.136465052127703,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.504331160117164e-07,
+      "loss": 1.3319,
+      "step": 2199
+    },
+    {
+      "epoch": 2.13743864346274,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.484269837751869e-07,
+      "loss": 1.3222,
+      "step": 2200
+    },
+    {
+      "epoch": 2.138412234797777,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.464224752629675e-07,
+      "loss": 1.2974,
+      "step": 2201
+    },
+    {
+      "epoch": 2.139385826132814,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.444195925727786e-07,
+      "loss": 1.3124,
+      "step": 2202
+    },
+    {
+      "epoch": 2.140359417467851,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.424183378006433e-07,
+      "loss": 1.3252,
+      "step": 2203
+    },
+    {
+      "epoch": 2.141333008802888,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.404187130408773e-07,
+      "loss": 1.3056,
+      "step": 2204
+    },
+    {
+      "epoch": 2.1423066001379256,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.384207203860923e-07,
+      "loss": 1.3379,
+      "step": 2205
+    },
+    {
+      "epoch": 2.1432801914729627,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.364243619271906e-07,
+      "loss": 1.3085,
+      "step": 2206
+    },
+    {
+      "epoch": 2.1442537828079997,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.344296397533672e-07,
+      "loss": 1.3085,
+      "step": 2207
+    },
+    {
+      "epoch": 2.1452273741430368,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.324365559521015e-07,
+      "loss": 1.3084,
+      "step": 2208
+    },
+    {
+      "epoch": 2.146200965478074,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.30445112609161e-07,
+      "loss": 1.2956,
+      "step": 2209
+    },
+    {
+      "epoch": 2.147174556813111,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.284553118085934e-07,
+      "loss": 1.3094,
+      "step": 2210
+    },
+    {
+      "epoch": 2.148148148148148,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.264671556327315e-07,
+      "loss": 1.3072,
+      "step": 2211
+    },
+    {
+      "epoch": 2.1491217394831854,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.244806461621828e-07,
+      "loss": 1.2836,
+      "step": 2212
+    },
+    {
+      "epoch": 2.1500953308182225,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.224957854758352e-07,
+      "loss": 1.2966,
+      "step": 2213
+    },
+    {
+      "epoch": 2.1510689221532595,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.205125756508485e-07,
+      "loss": 1.3254,
+      "step": 2214
+    },
+    {
+      "epoch": 2.1520425134882966,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.185310187626556e-07,
+      "loss": 1.3218,
+      "step": 2215
+    },
+    {
+      "epoch": 2.1530161048233336,
+      "grad_norm": 0.486328125,
+      "learning_rate": 9.165511168849594e-07,
+      "loss": 1.3148,
+      "step": 2216
+    },
+    {
+      "epoch": 2.1539896961583707,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.145728720897329e-07,
+      "loss": 1.2954,
+      "step": 2217
+    },
+    {
+      "epoch": 2.154963287493408,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.125962864472115e-07,
+      "loss": 1.329,
+      "step": 2218
+    },
+    {
+      "epoch": 2.155936878828445,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.106213620258977e-07,
+      "loss": 1.3077,
+      "step": 2219
+    },
+    {
+      "epoch": 2.1569104701634823,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.086481008925521e-07,
+      "loss": 1.308,
+      "step": 2220
+    },
+    {
+      "epoch": 2.1578840614985193,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.066765051121985e-07,
+      "loss": 1.3217,
+      "step": 2221
+    },
+    {
+      "epoch": 2.1588576528335564,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.047065767481139e-07,
+      "loss": 1.3222,
+      "step": 2222
+    },
+    {
+      "epoch": 2.1598312441685934,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.02738317861834e-07,
+      "loss": 1.3065,
+      "step": 2223
+    },
+    {
+      "epoch": 2.1608048355036304,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.007717305131444e-07,
+      "loss": 1.313,
+      "step": 2224
+    },
+    {
+      "epoch": 2.161778426838668,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.988068167600849e-07,
+      "loss": 1.3338,
+      "step": 2225
+    },
+    {
+      "epoch": 2.162752018173705,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.968435786589388e-07,
+      "loss": 1.2983,
+      "step": 2226
+    },
+    {
+      "epoch": 2.163725609508742,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.948820182642406e-07,
+      "loss": 1.3118,
+      "step": 2227
+    },
+    {
+      "epoch": 2.164699200843779,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.929221376287661e-07,
+      "loss": 1.3278,
+      "step": 2228
+    },
+    {
+      "epoch": 2.165672792178816,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.90963938803536e-07,
+      "loss": 1.3068,
+      "step": 2229
+    },
+    {
+      "epoch": 2.166646383513853,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.890074238378074e-07,
+      "loss": 1.3066,
+      "step": 2230
+    },
+    {
+      "epoch": 2.1676199748488907,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.870525947790792e-07,
+      "loss": 1.3118,
+      "step": 2231
+    },
+    {
+      "epoch": 2.1685935661839277,
+      "grad_norm": 0.390625,
+      "learning_rate": 8.85099453673082e-07,
+      "loss": 1.2933,
+      "step": 2232
+    },
+    {
+      "epoch": 2.169567157518965,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.831480025637837e-07,
+      "loss": 1.3178,
+      "step": 2233
+    },
+    {
+      "epoch": 2.170540748854002,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.8119824349338e-07,
+      "loss": 1.306,
+      "step": 2234
+    },
+    {
+      "epoch": 2.171514340189039,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.792501785022997e-07,
+      "loss": 1.3078,
+      "step": 2235
+    },
+    {
+      "epoch": 2.172487931524076,
+      "grad_norm": 0.390625,
+      "learning_rate": 8.773038096291958e-07,
+      "loss": 1.3018,
+      "step": 2236
+    },
+    {
+      "epoch": 2.1734615228591134,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.753591389109475e-07,
+      "loss": 1.344,
+      "step": 2237
+    },
+    {
+      "epoch": 2.1744351141941505,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.734161683826554e-07,
+      "loss": 1.3137,
+      "step": 2238
+    },
+    {
+      "epoch": 2.1754087055291875,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.714749000776441e-07,
+      "loss": 1.3293,
+      "step": 2239
+    },
+    {
+      "epoch": 2.1763822968642246,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.695353360274527e-07,
+      "loss": 1.3067,
+      "step": 2240
+    },
+    {
+      "epoch": 2.1773558881992616,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.67597478261841e-07,
+      "loss": 1.3205,
+      "step": 2241
+    },
+    {
+      "epoch": 2.1783294795342987,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.656613288087795e-07,
+      "loss": 1.3126,
+      "step": 2242
+    },
+    {
+      "epoch": 2.1793030708693357,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.63726889694454e-07,
+      "loss": 1.2934,
+      "step": 2243
+    },
+    {
+      "epoch": 2.180276662204373,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.617941629432572e-07,
+      "loss": 1.2984,
+      "step": 2244
+    },
+    {
+      "epoch": 2.1812502535394103,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.598631505777932e-07,
+      "loss": 1.3209,
+      "step": 2245
+    },
+    {
+      "epoch": 2.1822238448744473,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.57933854618869e-07,
+      "loss": 1.314,
+      "step": 2246
+    },
+    {
+      "epoch": 2.1831974362094844,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.56006277085499e-07,
+      "loss": 1.3065,
+      "step": 2247
+    },
+    {
+      "epoch": 2.1841710275445214,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.540804199948938e-07,
+      "loss": 1.3275,
+      "step": 2248
+    },
+    {
+      "epoch": 2.1851446188795585,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.521562853624693e-07,
+      "loss": 1.3211,
+      "step": 2249
+    },
+    {
+      "epoch": 2.186118210214596,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.502338752018344e-07,
+      "loss": 1.335,
+      "step": 2250
+    },
+    {
+      "epoch": 2.187091801549633,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.483131915247969e-07,
+      "loss": 1.3178,
+      "step": 2251
+    },
+    {
+      "epoch": 2.18806539288467,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.463942363413546e-07,
+      "loss": 1.3232,
+      "step": 2252
+    },
+    {
+      "epoch": 2.189038984219707,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.444770116596998e-07,
+      "loss": 1.3106,
+      "step": 2253
+    },
+    {
+      "epoch": 2.190012575554744,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.425615194862102e-07,
+      "loss": 1.3297,
+      "step": 2254
+    },
+    {
+      "epoch": 2.190986166889781,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.406477618254538e-07,
+      "loss": 1.3106,
+      "step": 2255
+    },
+    {
+      "epoch": 2.1919597582248187,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.387357406801804e-07,
+      "loss": 1.3031,
+      "step": 2256
+    },
+    {
+      "epoch": 2.1929333495598557,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.368254580513261e-07,
+      "loss": 1.3099,
+      "step": 2257
+    },
+    {
+      "epoch": 2.193906940894893,
+      "grad_norm": 0.388671875,
+      "learning_rate": 8.349169159380027e-07,
+      "loss": 1.3115,
+      "step": 2258
+    },
+    {
+      "epoch": 2.19488053222993,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.33010116337507e-07,
+      "loss": 1.3064,
+      "step": 2259
+    },
+    {
+      "epoch": 2.195854123564967,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.311050612453048e-07,
+      "loss": 1.3139,
+      "step": 2260
+    },
+    {
+      "epoch": 2.196827714900004,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.292017526550425e-07,
+      "loss": 1.3012,
+      "step": 2261
+    },
+    {
+      "epoch": 2.197801306235041,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.273001925585353e-07,
+      "loss": 1.3185,
+      "step": 2262
+    },
+    {
+      "epoch": 2.1987748975700785,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.254003829457707e-07,
+      "loss": 1.3078,
+      "step": 2263
+    },
+    {
+      "epoch": 2.1997484889051155,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.235023258049024e-07,
+      "loss": 1.3091,
+      "step": 2264
+    },
+    {
+      "epoch": 2.2007220802401526,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.21606023122252e-07,
+      "loss": 1.3116,
+      "step": 2265
+    },
+    {
+      "epoch": 2.2016956715751896,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.197114768823027e-07,
+      "loss": 1.342,
+      "step": 2266
+    },
+    {
+      "epoch": 2.2026692629102267,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.178186890677029e-07,
+      "loss": 1.3108,
+      "step": 2267
+    },
+    {
+      "epoch": 2.2036428542452637,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.159276616592568e-07,
+      "loss": 1.3002,
+      "step": 2268
+    },
+    {
+      "epoch": 2.2046164455803012,
+      "grad_norm": 0.390625,
+      "learning_rate": 8.140383966359302e-07,
+      "loss": 1.3349,
+      "step": 2269
+    },
+    {
+      "epoch": 2.2055900369153383,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.121508959748423e-07,
+      "loss": 1.3135,
+      "step": 2270
+    },
+    {
+      "epoch": 2.2065636282503753,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.102651616512658e-07,
+      "loss": 1.3086,
+      "step": 2271
+    },
+    {
+      "epoch": 2.2075372195854124,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.083811956386253e-07,
+      "loss": 1.324,
+      "step": 2272
+    },
+    {
+      "epoch": 2.2085108109204494,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.06498999908496e-07,
+      "loss": 1.3149,
+      "step": 2273
+    },
+    {
+      "epoch": 2.2094844022554865,
+      "grad_norm": 0.390625,
+      "learning_rate": 8.046185764305986e-07,
+      "loss": 1.3083,
+      "step": 2274
+    },
+    {
+      "epoch": 2.2104579935905235,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.027399271728012e-07,
+      "loss": 1.292,
+      "step": 2275
+    },
+    {
+      "epoch": 2.211431584925561,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.008630541011128e-07,
+      "loss": 1.3117,
+      "step": 2276
+    },
+    {
+      "epoch": 2.212405176260598,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.989879591796862e-07,
+      "loss": 1.3037,
+      "step": 2277
+    },
+    {
+      "epoch": 2.213378767595635,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.971146443708117e-07,
+      "loss": 1.3396,
+      "step": 2278
+    },
+    {
+      "epoch": 2.214352358930672,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.952431116349155e-07,
+      "loss": 1.3452,
+      "step": 2279
+    },
+    {
+      "epoch": 2.215325950265709,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.933733629305626e-07,
+      "loss": 1.3219,
+      "step": 2280
+    },
+    {
+      "epoch": 2.2162995416007463,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.915054002144478e-07,
+      "loss": 1.3486,
+      "step": 2281
+    },
+    {
+      "epoch": 2.2172731329357838,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.89639225441397e-07,
+      "loss": 1.3059,
+      "step": 2282
+    },
+    {
+      "epoch": 2.218246724270821,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.877748405643676e-07,
+      "loss": 1.3048,
+      "step": 2283
+    },
+    {
+      "epoch": 2.219220315605858,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.859122475344408e-07,
+      "loss": 1.3334,
+      "step": 2284
+    },
+    {
+      "epoch": 2.220193906940895,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.840514483008252e-07,
+      "loss": 1.3033,
+      "step": 2285
+    },
+    {
+      "epoch": 2.221167498275932,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.821924448108492e-07,
+      "loss": 1.343,
+      "step": 2286
+    },
+    {
+      "epoch": 2.222141089610969,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.803352390099661e-07,
+      "loss": 1.3136,
+      "step": 2287
+    },
+    {
+      "epoch": 2.223114680946006,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.784798328417439e-07,
+      "loss": 1.3015,
+      "step": 2288
+    },
+    {
+      "epoch": 2.2240882722810436,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.766262282478687e-07,
+      "loss": 1.3181,
+      "step": 2289
+    },
+    {
+      "epoch": 2.2250618636160806,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.747744271681429e-07,
+      "loss": 1.2962,
+      "step": 2290
+    },
+    {
+      "epoch": 2.2260354549511177,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.729244315404782e-07,
+      "loss": 1.3244,
+      "step": 2291
+    },
+    {
+      "epoch": 2.2270090462861547,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.710762433009009e-07,
+      "loss": 1.3174,
+      "step": 2292
+    },
+    {
+      "epoch": 2.2279826376211918,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.692298643835424e-07,
+      "loss": 1.312,
+      "step": 2293
+    },
+    {
+      "epoch": 2.228956228956229,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.673852967206416e-07,
+      "loss": 1.306,
+      "step": 2294
+    },
+    {
+      "epoch": 2.2299298202912663,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.655425422425436e-07,
+      "loss": 1.3128,
+      "step": 2295
+    },
+    {
+      "epoch": 2.2309034116263033,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.637016028776941e-07,
+      "loss": 1.3313,
+      "step": 2296
+    },
+    {
+      "epoch": 2.2318770029613404,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.618624805526389e-07,
+      "loss": 1.3239,
+      "step": 2297
+    },
+    {
+      "epoch": 2.2328505942963774,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.600251771920253e-07,
+      "loss": 1.3198,
+      "step": 2298
+    },
+    {
+      "epoch": 2.2338241856314145,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.58189694718593e-07,
+      "loss": 1.305,
+      "step": 2299
+    },
+    {
+      "epoch": 2.2347977769664515,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.563560350531795e-07,
+      "loss": 1.32,
+      "step": 2300
+    },
+    {
+      "epoch": 2.2357713683014886,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.545242001147123e-07,
+      "loss": 1.2973,
+      "step": 2301
+    },
+    {
+      "epoch": 2.236744959636526,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.526941918202116e-07,
+      "loss": 1.3227,
+      "step": 2302
+    },
+    {
+      "epoch": 2.237718550971563,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.508660120847841e-07,
+      "loss": 1.3226,
+      "step": 2303
+    },
+    {
+      "epoch": 2.2386921423066,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.490396628216237e-07,
+      "loss": 1.3061,
+      "step": 2304
+    },
+    {
+      "epoch": 2.2396657336416372,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.472151459420079e-07,
+      "loss": 1.3107,
+      "step": 2305
+    },
+    {
+      "epoch": 2.2406393249766743,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.453924633552986e-07,
+      "loss": 1.312,
+      "step": 2306
+    },
+    {
+      "epoch": 2.2416129163117113,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.435716169689355e-07,
+      "loss": 1.3287,
+      "step": 2307
+    },
+    {
+      "epoch": 2.242586507646749,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.417526086884394e-07,
+      "loss": 1.3028,
+      "step": 2308
+    },
+    {
+      "epoch": 2.243560098981786,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.399354404174047e-07,
+      "loss": 1.2748,
+      "step": 2309
+    },
+    {
+      "epoch": 2.244533690316823,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.381201140575031e-07,
+      "loss": 1.3154,
+      "step": 2310
+    },
+    {
+      "epoch": 2.24550728165186,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.363066315084761e-07,
+      "loss": 1.3212,
+      "step": 2311
+    },
+    {
+      "epoch": 2.246480872986897,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.344949946681382e-07,
+      "loss": 1.3299,
+      "step": 2312
+    },
+    {
+      "epoch": 2.247454464321934,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.326852054323693e-07,
+      "loss": 1.3035,
+      "step": 2313
+    },
+    {
+      "epoch": 2.248428055656971,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.308772656951202e-07,
+      "loss": 1.3132,
+      "step": 2314
+    },
+    {
+      "epoch": 2.2494016469920086,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.290711773484002e-07,
+      "loss": 1.3047,
+      "step": 2315
+    },
+    {
+      "epoch": 2.2503752383270457,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.272669422822867e-07,
+      "loss": 1.3009,
+      "step": 2316
+    },
+    {
+      "epoch": 2.2513488296620827,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.254645623849141e-07,
+      "loss": 1.3161,
+      "step": 2317
+    },
+    {
+      "epoch": 2.2523224209971198,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.23664039542478e-07,
+      "loss": 1.2989,
+      "step": 2318
+    },
+    {
+      "epoch": 2.253296012332157,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.218653756392272e-07,
+      "loss": 1.3291,
+      "step": 2319
+    },
+    {
+      "epoch": 2.254269603667194,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.200685725574694e-07,
+      "loss": 1.2851,
+      "step": 2320
+    },
+    {
+      "epoch": 2.2552431950022314,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.182736321775607e-07,
+      "loss": 1.3195,
+      "step": 2321
+    },
+    {
+      "epoch": 2.2562167863372684,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.164805563779115e-07,
+      "loss": 1.2928,
+      "step": 2322
+    },
+    {
+      "epoch": 2.2571903776723055,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.146893470349778e-07,
+      "loss": 1.3178,
+      "step": 2323
+    },
+    {
+      "epoch": 2.2581639690073425,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.129000060232649e-07,
+      "loss": 1.2987,
+      "step": 2324
+    },
+    {
+      "epoch": 2.2591375603423796,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.111125352153217e-07,
+      "loss": 1.3162,
+      "step": 2325
+    },
+    {
+      "epoch": 2.2601111516774166,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.093269364817398e-07,
+      "loss": 1.3139,
+      "step": 2326
+    },
+    {
+      "epoch": 2.2610847430124537,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.07543211691151e-07,
+      "loss": 1.3175,
+      "step": 2327
+    },
+    {
+      "epoch": 2.262058334347491,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.057613627102286e-07,
+      "loss": 1.3247,
+      "step": 2328
+    },
+    {
+      "epoch": 2.263031925682528,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.039813914036792e-07,
+      "loss": 1.3079,
+      "step": 2329
+    },
+    {
+      "epoch": 2.2640055170175652,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.022032996342485e-07,
+      "loss": 1.3464,
+      "step": 2330
+    },
+    {
+      "epoch": 2.2649791083526023,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.004270892627113e-07,
+      "loss": 1.2976,
+      "step": 2331
+    },
+    {
+      "epoch": 2.2659526996876393,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.986527621478773e-07,
+      "loss": 1.3219,
+      "step": 2332
+    },
+    {
+      "epoch": 2.2669262910226764,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.96880320146581e-07,
+      "loss": 1.2983,
+      "step": 2333
+    },
+    {
+      "epoch": 2.267899882357714,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.95109765113689e-07,
+      "loss": 1.3336,
+      "step": 2334
+    },
+    {
+      "epoch": 2.268873473692751,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.933410989020886e-07,
+      "loss": 1.297,
+      "step": 2335
+    },
+    {
+      "epoch": 2.269847065027788,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.915743233626954e-07,
+      "loss": 1.3146,
+      "step": 2336
+    },
+    {
+      "epoch": 2.270820656362825,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.8980944034444e-07,
+      "loss": 1.313,
+      "step": 2337
+    },
+    {
+      "epoch": 2.271794247697862,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.880464516942787e-07,
+      "loss": 1.3195,
+      "step": 2338
+    },
+    {
+      "epoch": 2.272767839032899,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.86285359257181e-07,
+      "loss": 1.3295,
+      "step": 2339
+    },
+    {
+      "epoch": 2.273741430367936,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.845261648761353e-07,
+      "loss": 1.3069,
+      "step": 2340
+    },
+    {
+      "epoch": 2.2747150217029737,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.827688703921407e-07,
+      "loss": 1.3146,
+      "step": 2341
+    },
+    {
+      "epoch": 2.2756886130380107,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.810134776442107e-07,
+      "loss": 1.318,
+      "step": 2342
+    },
+    {
+      "epoch": 2.276662204373048,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.79259988469366e-07,
+      "loss": 1.2957,
+      "step": 2343
+    },
+    {
+      "epoch": 2.277635795708085,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.775084047026381e-07,
+      "loss": 1.3402,
+      "step": 2344
+    },
+    {
+      "epoch": 2.278609387043122,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.757587281770614e-07,
+      "loss": 1.3118,
+      "step": 2345
+    },
+    {
+      "epoch": 2.279582978378159,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.740109607236775e-07,
+      "loss": 1.3179,
+      "step": 2346
+    },
+    {
+      "epoch": 2.2805565697131964,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.722651041715278e-07,
+      "loss": 1.313,
+      "step": 2347
+    },
+    {
+      "epoch": 2.2815301610482335,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.705211603476547e-07,
+      "loss": 1.3097,
+      "step": 2348
+    },
+    {
+      "epoch": 2.2825037523832705,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.687791310770983e-07,
+      "loss": 1.3099,
+      "step": 2349
+    },
+    {
+      "epoch": 2.2834773437183076,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.670390181828973e-07,
+      "loss": 1.3247,
+      "step": 2350
+    },
+    {
+      "epoch": 2.2844509350533446,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.653008234860814e-07,
+      "loss": 1.3361,
+      "step": 2351
+    },
+    {
+      "epoch": 2.2854245263883817,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.635645488056771e-07,
+      "loss": 1.3343,
+      "step": 2352
+    },
+    {
+      "epoch": 2.2863981177234187,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.618301959586973e-07,
+      "loss": 1.3328,
+      "step": 2353
+    },
+    {
+      "epoch": 2.287371709058456,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.600977667601477e-07,
+      "loss": 1.3181,
+      "step": 2354
+    },
+    {
+      "epoch": 2.2883453003934933,
+      "grad_norm": 0.388671875,
+      "learning_rate": 6.58367263023017e-07,
+      "loss": 1.3199,
+      "step": 2355
+    },
+    {
+      "epoch": 2.2893188917285303,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.566386865582827e-07,
+      "loss": 1.3364,
+      "step": 2356
+    },
+    {
+      "epoch": 2.2902924830635674,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.549120391749023e-07,
+      "loss": 1.2999,
+      "step": 2357
+    },
+    {
+      "epoch": 2.2912660743986044,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.531873226798174e-07,
+      "loss": 1.3147,
+      "step": 2358
+    },
+    {
+      "epoch": 2.2922396657336415,
+      "grad_norm": 0.388671875,
+      "learning_rate": 6.514645388779447e-07,
+      "loss": 1.2999,
+      "step": 2359
+    },
+    {
+      "epoch": 2.293213257068679,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.497436895721834e-07,
+      "loss": 1.3116,
+      "step": 2360
+    },
+    {
+      "epoch": 2.294186848403716,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.48024776563404e-07,
+      "loss": 1.3199,
+      "step": 2361
+    },
+    {
+      "epoch": 2.295160439738753,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.46307801650454e-07,
+      "loss": 1.3237,
+      "step": 2362
+    },
+    {
+      "epoch": 2.29613403107379,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.445927666301494e-07,
+      "loss": 1.3241,
+      "step": 2363
+    },
+    {
+      "epoch": 2.297107622408827,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.428796732972797e-07,
+      "loss": 1.3209,
+      "step": 2364
+    },
+    {
+      "epoch": 2.298081213743864,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.411685234445989e-07,
+      "loss": 1.3029,
+      "step": 2365
+    },
+    {
+      "epoch": 2.2990548050789013,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.394593188628303e-07,
+      "loss": 1.2924,
+      "step": 2366
+    },
+    {
+      "epoch": 2.3000283964139387,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.377520613406585e-07,
+      "loss": 1.328,
+      "step": 2367
+    },
+    {
+      "epoch": 2.301001987748976,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.360467526647333e-07,
+      "loss": 1.3369,
+      "step": 2368
+    },
+    {
+      "epoch": 2.301975579084013,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.343433946196631e-07,
+      "loss": 1.3079,
+      "step": 2369
+    },
+    {
+      "epoch": 2.30294917041905,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.326419889880156e-07,
+      "loss": 1.3376,
+      "step": 2370
+    },
+    {
+      "epoch": 2.303922761754087,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.309425375503144e-07,
+      "loss": 1.298,
+      "step": 2371
+    },
+    {
+      "epoch": 2.304896353089124,
+      "grad_norm": 0.390625,
+      "learning_rate": 6.292450420850402e-07,
+      "loss": 1.3031,
+      "step": 2372
+    },
+    {
+      "epoch": 2.3058699444241615,
+      "grad_norm": 0.408203125,
+      "learning_rate": 6.275495043686244e-07,
+      "loss": 1.3244,
+      "step": 2373
+    },
+    {
+      "epoch": 2.3068435357591985,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.258559261754518e-07,
+      "loss": 1.3114,
+      "step": 2374
+    },
+    {
+      "epoch": 2.3078171270942356,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.24164309277854e-07,
+      "loss": 1.3265,
+      "step": 2375
+    },
+    {
+      "epoch": 2.3087907184292726,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.224746554461128e-07,
+      "loss": 1.2972,
+      "step": 2376
+    },
+    {
+      "epoch": 2.3097643097643097,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.207869664484542e-07,
+      "loss": 1.3105,
+      "step": 2377
+    },
+    {
+      "epoch": 2.3107379010993467,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.191012440510469e-07,
+      "loss": 1.3157,
+      "step": 2378
+    },
+    {
+      "epoch": 2.311711492434384,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.174174900180044e-07,
+      "loss": 1.3024,
+      "step": 2379
+    },
+    {
+      "epoch": 2.3126850837694213,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.157357061113781e-07,
+      "loss": 1.3134,
+      "step": 2380
+    },
+    {
+      "epoch": 2.3136586751044583,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.140558940911573e-07,
+      "loss": 1.3184,
+      "step": 2381
+    },
+    {
+      "epoch": 2.3146322664394954,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.123780557152703e-07,
+      "loss": 1.3197,
+      "step": 2382
+    },
+    {
+      "epoch": 2.3156058577745324,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.10702192739577e-07,
+      "loss": 1.337,
+      "step": 2383
+    },
+    {
+      "epoch": 2.3165794491095695,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.09028306917873e-07,
+      "loss": 1.303,
+      "step": 2384
+    },
+    {
+      "epoch": 2.3175530404446065,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.073564000018811e-07,
+      "loss": 1.3265,
+      "step": 2385
+    },
+    {
+      "epoch": 2.318526631779644,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.056864737412574e-07,
+      "loss": 1.3251,
+      "step": 2386
+    },
+    {
+      "epoch": 2.319500223114681,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.040185298835821e-07,
+      "loss": 1.3304,
+      "step": 2387
+    },
+    {
+      "epoch": 2.320473814449718,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.02352570174361e-07,
+      "loss": 1.2987,
+      "step": 2388
+    },
+    {
+      "epoch": 2.321447405784755,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.006885963570261e-07,
+      "loss": 1.3073,
+      "step": 2389
+    },
+    {
+      "epoch": 2.322420997119792,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.990266101729278e-07,
+      "loss": 1.3253,
+      "step": 2390
+    },
+    {
+      "epoch": 2.3233945884548293,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.973666133613393e-07,
+      "loss": 1.2776,
+      "step": 2391
+    },
+    {
+      "epoch": 2.3243681797898663,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.957086076594502e-07,
+      "loss": 1.3026,
+      "step": 2392
+    },
+    {
+      "epoch": 2.325341771124904,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.940525948023657e-07,
+      "loss": 1.3391,
+      "step": 2393
+    },
+    {
+      "epoch": 2.326315362459941,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.923985765231082e-07,
+      "loss": 1.307,
+      "step": 2394
+    },
+    {
+      "epoch": 2.327288953794978,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.907465545526109e-07,
+      "loss": 1.3212,
+      "step": 2395
+    },
+    {
+      "epoch": 2.328262545130015,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.890965306197166e-07,
+      "loss": 1.319,
+      "step": 2396
+    },
+    {
+      "epoch": 2.329236136465052,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.874485064511809e-07,
+      "loss": 1.3012,
+      "step": 2397
+    },
+    {
+      "epoch": 2.330209727800089,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.858024837716628e-07,
+      "loss": 1.3028,
+      "step": 2398
+    },
+    {
+      "epoch": 2.3311833191351266,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.841584643037295e-07,
+      "loss": 1.3361,
+      "step": 2399
+    },
+    {
+      "epoch": 2.3321569104701636,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.825164497678495e-07,
+      "loss": 1.3167,
+      "step": 2400
+    },
+    {
+      "epoch": 2.3331305018052007,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.808764418823959e-07,
+      "loss": 1.318,
+      "step": 2401
+    },
+    {
+      "epoch": 2.3341040931402377,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.792384423636394e-07,
+      "loss": 1.3019,
+      "step": 2402
+    },
+    {
+      "epoch": 2.3350776844752748,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.77602452925749e-07,
+      "loss": 1.3256,
+      "step": 2403
+    },
+    {
+      "epoch": 2.336051275810312,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.759684752807925e-07,
+      "loss": 1.3099,
+      "step": 2404
+    },
+    {
+      "epoch": 2.337024867145349,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.743365111387303e-07,
+      "loss": 1.319,
+      "step": 2405
+    },
+    {
+      "epoch": 2.3379984584803863,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.727065622074154e-07,
+      "loss": 1.3009,
+      "step": 2406
+    },
+    {
+      "epoch": 2.3389720498154234,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.710786301925938e-07,
+      "loss": 1.3294,
+      "step": 2407
+    },
+    {
+      "epoch": 2.3399456411504604,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.694527167978986e-07,
+      "loss": 1.2959,
+      "step": 2408
+    },
+    {
+      "epoch": 2.3409192324854975,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.678288237248525e-07,
+      "loss": 1.3163,
+      "step": 2409
+    },
+    {
+      "epoch": 2.3418928238205345,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.662069526728617e-07,
+      "loss": 1.3368,
+      "step": 2410
+    },
+    {
+      "epoch": 2.3428664151555716,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.645871053392191e-07,
+      "loss": 1.3278,
+      "step": 2411
+    },
+    {
+      "epoch": 2.343840006490609,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.629692834190963e-07,
+      "loss": 1.3079,
+      "step": 2412
+    },
+    {
+      "epoch": 2.344813597825646,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.613534886055502e-07,
+      "loss": 1.2973,
+      "step": 2413
+    },
+    {
+      "epoch": 2.345787189160683,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.597397225895098e-07,
+      "loss": 1.3207,
+      "step": 2414
+    },
+    {
+      "epoch": 2.3467607804957202,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.581279870597866e-07,
+      "loss": 1.3242,
+      "step": 2415
+    },
+    {
+      "epoch": 2.3477343718307573,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.565182837030645e-07,
+      "loss": 1.3133,
+      "step": 2416
+    },
+    {
+      "epoch": 2.3487079631657943,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.549106142039018e-07,
+      "loss": 1.3171,
+      "step": 2417
+    },
+    {
+      "epoch": 2.3496815545008314,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.533049802447268e-07,
+      "loss": 1.286,
+      "step": 2418
+    },
+    {
+      "epoch": 2.350655145835869,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.517013835058404e-07,
+      "loss": 1.3036,
+      "step": 2419
+    },
+    {
+      "epoch": 2.351628737170906,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.50099825665408e-07,
+      "loss": 1.3215,
+      "step": 2420
+    },
+    {
+      "epoch": 2.352602328505943,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.485003083994649e-07,
+      "loss": 1.3113,
+      "step": 2421
+    },
+    {
+      "epoch": 2.35357591984098,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.469028333819077e-07,
+      "loss": 1.3337,
+      "step": 2422
+    },
+    {
+      "epoch": 2.354549511176017,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.453074022844984e-07,
+      "loss": 1.2914,
+      "step": 2423
+    },
+    {
+      "epoch": 2.3555231025110546,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.437140167768582e-07,
+      "loss": 1.2937,
+      "step": 2424
+    },
+    {
+      "epoch": 2.3564966938460916,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.421226785264685e-07,
+      "loss": 1.3063,
+      "step": 2425
+    },
+    {
+      "epoch": 2.3574702851811287,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.405333891986672e-07,
+      "loss": 1.3183,
+      "step": 2426
+    },
+    {
+      "epoch": 2.3584438765161657,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.389461504566503e-07,
+      "loss": 1.3231,
+      "step": 2427
+    },
+    {
+      "epoch": 2.3594174678512028,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.37360963961465e-07,
+      "loss": 1.3191,
+      "step": 2428
+    },
+    {
+      "epoch": 2.36039105918624,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.35777831372013e-07,
+      "loss": 1.3077,
+      "step": 2429
+    },
+    {
+      "epoch": 2.361364650521277,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.341967543450452e-07,
+      "loss": 1.3377,
+      "step": 2430
+    },
+    {
+      "epoch": 2.362338241856314,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.326177345351627e-07,
+      "loss": 1.3204,
+      "step": 2431
+    },
+    {
+      "epoch": 2.3633118331913514,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.31040773594812e-07,
+      "loss": 1.3141,
+      "step": 2432
+    },
+    {
+      "epoch": 2.3642854245263885,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.294658731742869e-07,
+      "loss": 1.3197,
+      "step": 2433
+    },
+    {
+      "epoch": 2.3652590158614255,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.278930349217226e-07,
+      "loss": 1.3094,
+      "step": 2434
+    },
+    {
+      "epoch": 2.3662326071964626,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.263222604831001e-07,
+      "loss": 1.3223,
+      "step": 2435
+    },
+    {
+      "epoch": 2.3672061985314996,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.247535515022345e-07,
+      "loss": 1.3183,
+      "step": 2436
+    },
+    {
+      "epoch": 2.368179789866537,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.231869096207853e-07,
+      "loss": 1.3117,
+      "step": 2437
+    },
+    {
+      "epoch": 2.369153381201574,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.216223364782455e-07,
+      "loss": 1.3197,
+      "step": 2438
+    },
+    {
+      "epoch": 2.370126972536611,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.200598337119447e-07,
+      "loss": 1.3198,
+      "step": 2439
+    },
+    {
+      "epoch": 2.3711005638716482,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.184994029570443e-07,
+      "loss": 1.3035,
+      "step": 2440
+    },
+    {
+      "epoch": 2.3720741552066853,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.169410458465393e-07,
+      "loss": 1.3379,
+      "step": 2441
+    },
+    {
+      "epoch": 2.3730477465417223,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.153847640112528e-07,
+      "loss": 1.3303,
+      "step": 2442
+    },
+    {
+      "epoch": 2.3740213378767594,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.138305590798376e-07,
+      "loss": 1.3176,
+      "step": 2443
+    },
+    {
+      "epoch": 2.3749949292117964,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.122784326787711e-07,
+      "loss": 1.3209,
+      "step": 2444
+    },
+    {
+      "epoch": 2.375968520546834,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.107283864323584e-07,
+      "loss": 1.2801,
+      "step": 2445
+    },
+    {
+      "epoch": 2.376942111881871,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.091804219627253e-07,
+      "loss": 1.3275,
+      "step": 2446
+    },
+    {
+      "epoch": 2.377915703216908,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.076345408898192e-07,
+      "loss": 1.3178,
+      "step": 2447
+    },
+    {
+      "epoch": 2.378889294551945,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.060907448314078e-07,
+      "loss": 1.3144,
+      "step": 2448
+    },
+    {
+      "epoch": 2.379862885886982,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.04549035403078e-07,
+      "loss": 1.3149,
+      "step": 2449
+    },
+    {
+      "epoch": 2.3808364772220196,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.030094142182301e-07,
+      "loss": 1.3017,
+      "step": 2450
+    },
+    {
+      "epoch": 2.3818100685570567,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.014718828880827e-07,
+      "loss": 1.3011,
+      "step": 2451
+    },
+    {
+      "epoch": 2.3827836598920937,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.999364430216639e-07,
+      "loss": 1.3088,
+      "step": 2452
+    },
+    {
+      "epoch": 2.383757251227131,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.984030962258158e-07,
+      "loss": 1.3191,
+      "step": 2453
+    },
+    {
+      "epoch": 2.384730842562168,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.968718441051876e-07,
+      "loss": 1.3256,
+      "step": 2454
+    },
+    {
+      "epoch": 2.385704433897205,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.953426882622392e-07,
+      "loss": 1.3182,
+      "step": 2455
+    },
+    {
+      "epoch": 2.386678025232242,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.938156302972338e-07,
+      "loss": 1.3034,
+      "step": 2456
+    },
+    {
+      "epoch": 2.387651616567279,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.922906718082431e-07,
+      "loss": 1.3164,
+      "step": 2457
+    },
+    {
+      "epoch": 2.3886252079023165,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.907678143911363e-07,
+      "loss": 1.324,
+      "step": 2458
+    },
+    {
+      "epoch": 2.3895987992373535,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.892470596395887e-07,
+      "loss": 1.3342,
+      "step": 2459
+    },
+    {
+      "epoch": 2.3905723905723906,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.877284091450718e-07,
+      "loss": 1.3079,
+      "step": 2460
+    },
+    {
+      "epoch": 2.3915459819074276,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.862118644968584e-07,
+      "loss": 1.3134,
+      "step": 2461
+    },
+    {
+      "epoch": 2.3925195732424647,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.846974272820131e-07,
+      "loss": 1.285,
+      "step": 2462
+    },
+    {
+      "epoch": 2.393493164577502,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.831850990854e-07,
+      "loss": 1.3168,
+      "step": 2463
+    },
+    {
+      "epoch": 2.394466755912539,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.816748814896716e-07,
+      "loss": 1.326,
+      "step": 2464
+    },
+    {
+      "epoch": 2.3954403472475763,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.801667760752754e-07,
+      "loss": 1.2944,
+      "step": 2465
+    },
+    {
+      "epoch": 2.3964139385826133,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.786607844204449e-07,
+      "loss": 1.3209,
+      "step": 2466
+    },
+    {
+      "epoch": 2.3973875299176504,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.771569081012053e-07,
+      "loss": 1.3157,
+      "step": 2467
+    },
+    {
+      "epoch": 2.3983611212526874,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.756551486913655e-07,
+      "loss": 1.3212,
+      "step": 2468
+    },
+    {
+      "epoch": 2.3993347125877245,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.741555077625193e-07,
+      "loss": 1.2997,
+      "step": 2469
+    },
+    {
+      "epoch": 2.4003083039227615,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.726579868840439e-07,
+      "loss": 1.3039,
+      "step": 2470
+    },
+    {
+      "epoch": 2.401281895257799,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.711625876230988e-07,
+      "loss": 1.3264,
+      "step": 2471
+    },
+    {
+      "epoch": 2.402255486592836,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.696693115446216e-07,
+      "loss": 1.295,
+      "step": 2472
+    },
+    {
+      "epoch": 2.403229077927873,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.6817816021132916e-07,
+      "loss": 1.3141,
+      "step": 2473
+    },
+    {
+      "epoch": 2.40420266926291,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.6668913518371413e-07,
+      "loss": 1.2946,
+      "step": 2474
+    },
+    {
+      "epoch": 2.405176260597947,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.6520223802004464e-07,
+      "loss": 1.3183,
+      "step": 2475
+    },
+    {
+      "epoch": 2.4061498519329847,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.637174702763608e-07,
+      "loss": 1.308,
+      "step": 2476
+    },
+    {
+      "epoch": 2.4071234432680217,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.622348335064761e-07,
+      "loss": 1.2958,
+      "step": 2477
+    },
+    {
+      "epoch": 2.408097034603059,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.607543292619726e-07,
+      "loss": 1.3342,
+      "step": 2478
+    },
+    {
+      "epoch": 2.409070625938096,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.592759590922005e-07,
+      "loss": 1.3052,
+      "step": 2479
+    },
+    {
+      "epoch": 2.410044217273133,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.5779972454427694e-07,
+      "loss": 1.3167,
+      "step": 2480
+    },
+    {
+      "epoch": 2.41101780860817,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.563256271630856e-07,
+      "loss": 1.2966,
+      "step": 2481
+    },
+    {
+      "epoch": 2.411991399943207,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.5485366849127105e-07,
+      "loss": 1.3211,
+      "step": 2482
+    },
+    {
+      "epoch": 2.412964991278244,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.5338385006924223e-07,
+      "loss": 1.3314,
+      "step": 2483
+    },
+    {
+      "epoch": 2.4139385826132815,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.5191617343516596e-07,
+      "loss": 1.3233,
+      "step": 2484
+    },
+    {
+      "epoch": 2.4149121739483186,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.5045064012497003e-07,
+      "loss": 1.3321,
+      "step": 2485
+    },
+    {
+      "epoch": 2.4158857652833556,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.489872516723373e-07,
+      "loss": 1.3154,
+      "step": 2486
+    },
+    {
+      "epoch": 2.4168593566183927,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.475260096087064e-07,
+      "loss": 1.3271,
+      "step": 2487
+    },
+    {
+      "epoch": 2.4178329479534297,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.46066915463271e-07,
+      "loss": 1.3055,
+      "step": 2488
+    },
+    {
+      "epoch": 2.4188065392884672,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.4460997076297504e-07,
+      "loss": 1.3123,
+      "step": 2489
+    },
+    {
+      "epoch": 2.4197801306235043,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.4315517703251514e-07,
+      "loss": 1.3276,
+      "step": 2490
+    },
+    {
+      "epoch": 2.4207537219585413,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.417025357943355e-07,
+      "loss": 1.3138,
+      "step": 2491
+    },
+    {
+      "epoch": 2.4217273132935784,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.402520485686276e-07,
+      "loss": 1.2835,
+      "step": 2492
+    },
+    {
+      "epoch": 2.4227009046286154,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.3880371687333027e-07,
+      "loss": 1.3218,
+      "step": 2493
+    },
+    {
+      "epoch": 2.4236744959636525,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.3735754222412494e-07,
+      "loss": 1.3177,
+      "step": 2494
+    },
+    {
+      "epoch": 2.4246480872986895,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.3591352613443606e-07,
+      "loss": 1.3138,
+      "step": 2495
+    },
+    {
+      "epoch": 2.425621678633727,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.3447167011543e-07,
+      "loss": 1.3114,
+      "step": 2496
+    },
+    {
+      "epoch": 2.426595269968764,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.3303197567601125e-07,
+      "loss": 1.3237,
+      "step": 2497
+    },
+    {
+      "epoch": 2.427568861303801,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.3159444432282426e-07,
+      "loss": 1.3405,
+      "step": 2498
+    },
+    {
+      "epoch": 2.428542452638838,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.3015907756024683e-07,
+      "loss": 1.3163,
+      "step": 2499
+    },
+    {
+      "epoch": 2.429516043973875,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.2872587689039486e-07,
+      "loss": 1.3104,
+      "step": 2500
+    },
+    {
+      "epoch": 2.4304896353089123,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.272948438131144e-07,
+      "loss": 1.2947,
+      "step": 2501
+    },
+    {
+      "epoch": 2.4314632266439498,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.2586597982598536e-07,
+      "loss": 1.3231,
+      "step": 2502
+    },
+    {
+      "epoch": 2.432436817978987,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.244392864243169e-07,
+      "loss": 1.3201,
+      "step": 2503
+    },
+    {
+      "epoch": 2.433410409314024,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.230147651011457e-07,
+      "loss": 1.3002,
+      "step": 2504
+    },
+    {
+      "epoch": 2.434384000649061,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.215924173472363e-07,
+      "loss": 1.3024,
+      "step": 2505
+    },
+    {
+      "epoch": 2.435357591984098,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.2017224465107947e-07,
+      "loss": 1.2984,
+      "step": 2506
+    },
+    {
+      "epoch": 2.436331183319135,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.187542484988874e-07,
+      "loss": 1.3176,
+      "step": 2507
+    },
+    {
+      "epoch": 2.437304774654172,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.1733843037459754e-07,
+      "loss": 1.3094,
+      "step": 2508
+    },
+    {
+      "epoch": 2.4382783659892096,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.1592479175986494e-07,
+      "loss": 1.2799,
+      "step": 2509
+    },
+    {
+      "epoch": 2.4392519573242466,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.145133341340665e-07,
+      "loss": 1.303,
+      "step": 2510
+    },
+    {
+      "epoch": 2.4402255486592836,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.1310405897429463e-07,
+      "loss": 1.3046,
+      "step": 2511
+    },
+    {
+      "epoch": 2.4411991399943207,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.1169696775535923e-07,
+      "loss": 1.3219,
+      "step": 2512
+    },
+    {
+      "epoch": 2.4421727313293577,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.102920619497841e-07,
+      "loss": 1.321,
+      "step": 2513
+    },
+    {
+      "epoch": 2.443146322664395,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.088893430278057e-07,
+      "loss": 1.3411,
+      "step": 2514
+    },
+    {
+      "epoch": 2.4441199139994323,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.0748881245737173e-07,
+      "loss": 1.2986,
+      "step": 2515
+    },
+    {
+      "epoch": 2.4450935053344693,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.060904717041417e-07,
+      "loss": 1.3207,
+      "step": 2516
+    },
+    {
+      "epoch": 2.4460670966695064,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.046943222314803e-07,
+      "loss": 1.3151,
+      "step": 2517
+    },
+    {
+      "epoch": 2.4470406880045434,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.033003655004622e-07,
+      "loss": 1.3021,
+      "step": 2518
+    },
+    {
+      "epoch": 2.4480142793395805,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.019086029698649e-07,
+      "loss": 1.3171,
+      "step": 2519
+    },
+    {
+      "epoch": 2.4489878706746175,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.0051903609617195e-07,
+      "loss": 1.3259,
+      "step": 2520
+    },
+    {
+      "epoch": 2.4499614620096546,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.9913166633356645e-07,
+      "loss": 1.3218,
+      "step": 2521
+    },
+    {
+      "epoch": 2.450935053344692,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.977464951339352e-07,
+      "loss": 1.3049,
+      "step": 2522
+    },
+    {
+      "epoch": 2.451908644679729,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.9636352394686134e-07,
+      "loss": 1.3084,
+      "step": 2523
+    },
+    {
+      "epoch": 2.452882236014766,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.9498275421962854e-07,
+      "loss": 1.3172,
+      "step": 2524
+    },
+    {
+      "epoch": 2.4538558273498032,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.9360418739721295e-07,
+      "loss": 1.3097,
+      "step": 2525
+    },
+    {
+      "epoch": 2.4548294186848403,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.922278249222894e-07,
+      "loss": 1.3161,
+      "step": 2526
+    },
+    {
+      "epoch": 2.4558030100198773,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.908536682352229e-07,
+      "loss": 1.3021,
+      "step": 2527
+    },
+    {
+      "epoch": 2.456776601354915,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.8948171877407207e-07,
+      "loss": 1.3062,
+      "step": 2528
+    },
+    {
+      "epoch": 2.457750192689952,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.88111977974584e-07,
+      "loss": 1.3074,
+      "step": 2529
+    },
+    {
+      "epoch": 2.458723784024989,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.867444472701959e-07,
+      "loss": 1.3241,
+      "step": 2530
+    },
+    {
+      "epoch": 2.459697375360026,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.8537912809203075e-07,
+      "loss": 1.2956,
+      "step": 2531
+    },
+    {
+      "epoch": 2.460670966695063,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.8401602186889904e-07,
+      "loss": 1.3255,
+      "step": 2532
+    },
+    {
+      "epoch": 2.4616445580301,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.826551300272924e-07,
+      "loss": 1.3229,
+      "step": 2533
+    },
+    {
+      "epoch": 2.462618149365137,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.812964539913888e-07,
+      "loss": 1.294,
+      "step": 2534
+    },
+    {
+      "epoch": 2.4635917407001746,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.7993999518304433e-07,
+      "loss": 1.3159,
+      "step": 2535
+    },
+    {
+      "epoch": 2.4645653320352117,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.7858575502179613e-07,
+      "loss": 1.3169,
+      "step": 2536
+    },
+    {
+      "epoch": 2.4655389233702487,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.772337349248589e-07,
+      "loss": 1.3146,
+      "step": 2537
+    },
+    {
+      "epoch": 2.4665125147052858,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.7588393630712513e-07,
+      "loss": 1.3161,
+      "step": 2538
+    },
+    {
+      "epoch": 2.467486106040323,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.745363605811611e-07,
+      "loss": 1.3002,
+      "step": 2539
+    },
+    {
+      "epoch": 2.46845969737536,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.731910091572083e-07,
+      "loss": 1.2935,
+      "step": 2540
+    },
+    {
+      "epoch": 2.4694332887103974,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.718478834431788e-07,
+      "loss": 1.3054,
+      "step": 2541
+    },
+    {
+      "epoch": 2.4704068800454344,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.7050698484465767e-07,
+      "loss": 1.3202,
+      "step": 2542
+    },
+    {
+      "epoch": 2.4713804713804715,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.691683147648964e-07,
+      "loss": 1.3242,
+      "step": 2543
+    },
+    {
+      "epoch": 2.4723540627155085,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.6783187460481763e-07,
+      "loss": 1.3024,
+      "step": 2544
+    },
+    {
+      "epoch": 2.4733276540505456,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.6649766576300707e-07,
+      "loss": 1.3246,
+      "step": 2545
+    },
+    {
+      "epoch": 2.4743012453855826,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.651656896357189e-07,
+      "loss": 1.3226,
+      "step": 2546
+    },
+    {
+      "epoch": 2.4752748367206197,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.638359476168668e-07,
+      "loss": 1.305,
+      "step": 2547
+    },
+    {
+      "epoch": 2.476248428055657,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.625084410980298e-07,
+      "loss": 1.3425,
+      "step": 2548
+    },
+    {
+      "epoch": 2.477222019390694,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.6118317146844524e-07,
+      "loss": 1.309,
+      "step": 2549
+    },
+    {
+      "epoch": 2.4781956107257312,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.5986014011501175e-07,
+      "loss": 1.3193,
+      "step": 2550
+    },
+    {
+      "epoch": 2.4791692020607683,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.585393484222829e-07,
+      "loss": 1.3187,
+      "step": 2551
+    },
+    {
+      "epoch": 2.4801427933958053,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.572207977724709e-07,
+      "loss": 1.3145,
+      "step": 2552
+    },
+    {
+      "epoch": 2.4811163847308424,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.559044895454411e-07,
+      "loss": 1.3195,
+      "step": 2553
+    },
+    {
+      "epoch": 2.48208997606588,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.545904251187135e-07,
+      "loss": 1.3314,
+      "step": 2554
+    },
+    {
+      "epoch": 2.483063567400917,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.532786058674581e-07,
+      "loss": 1.2886,
+      "step": 2555
+    },
+    {
+      "epoch": 2.484037158735954,
+      "grad_norm": 0.3828125,
+      "learning_rate": 3.519690331644973e-07,
+      "loss": 1.3003,
+      "step": 2556
+    },
+    {
+      "epoch": 2.485010750070991,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.5066170838030153e-07,
+      "loss": 1.3048,
+      "step": 2557
+    },
+    {
+      "epoch": 2.485984341406028,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.493566328829884e-07,
+      "loss": 1.3262,
+      "step": 2558
+    },
+    {
+      "epoch": 2.486957932741065,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.4805380803832167e-07,
+      "loss": 1.3179,
+      "step": 2559
+    },
+    {
+      "epoch": 2.487931524076102,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.4675323520971137e-07,
+      "loss": 1.3123,
+      "step": 2560
+    },
+    {
+      "epoch": 2.4889051154111397,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.454549157582082e-07,
+      "loss": 1.3129,
+      "step": 2561
+    },
+    {
+      "epoch": 2.4898787067461767,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.4415885104250787e-07,
+      "loss": 1.3066,
+      "step": 2562
+    },
+    {
+      "epoch": 2.490852298081214,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.4286504241894283e-07,
+      "loss": 1.3166,
+      "step": 2563
+    },
+    {
+      "epoch": 2.491825889416251,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.415734912414878e-07,
+      "loss": 1.3,
+      "step": 2564
+    },
+    {
+      "epoch": 2.492799480751288,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.4028419886175306e-07,
+      "loss": 1.3168,
+      "step": 2565
+    },
+    {
+      "epoch": 2.493773072086325,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.3899716662898623e-07,
+      "loss": 1.2988,
+      "step": 2566
+    },
+    {
+      "epoch": 2.4947466634213624,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.377123958900688e-07,
+      "loss": 1.314,
+      "step": 2567
+    },
+    {
+      "epoch": 2.4957202547563995,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.3642988798951587e-07,
+      "loss": 1.3127,
+      "step": 2568
+    },
+    {
+      "epoch": 2.4966938460914365,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.3514964426947433e-07,
+      "loss": 1.3219,
+      "step": 2569
+    },
+    {
+      "epoch": 2.4976674374264736,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.338716660697225e-07,
+      "loss": 1.3277,
+      "step": 2570
+    },
+    {
+      "epoch": 2.4986410287615106,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.3259595472766635e-07,
+      "loss": 1.3038,
+      "step": 2571
+    },
+    {
+      "epoch": 2.4996146200965477,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.3132251157834137e-07,
+      "loss": 1.318,
+      "step": 2572
+    },
+    {
+      "epoch": 2.5005882114315847,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.3005133795440734e-07,
+      "loss": 1.315,
+      "step": 2573
+    },
+    {
+      "epoch": 2.5015618027666218,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.2878243518615127e-07,
+      "loss": 1.3067,
+      "step": 2574
+    },
+    {
+      "epoch": 2.5025353941016593,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.2751580460148094e-07,
+      "loss": 1.311,
+      "step": 2575
+    },
+    {
+      "epoch": 2.5035089854366963,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.262514475259296e-07,
+      "loss": 1.3077,
+      "step": 2576
+    },
+    {
+      "epoch": 2.5044825767717334,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.249893652826483e-07,
+      "loss": 1.3187,
+      "step": 2577
+    },
+    {
+      "epoch": 2.5054561681067704,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.2372955919240834e-07,
+      "loss": 1.3228,
+      "step": 2578
+    },
+    {
+      "epoch": 2.506429759441808,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.2247203057360066e-07,
+      "loss": 1.3064,
+      "step": 2579
+    },
+    {
+      "epoch": 2.507403350776845,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.212167807422306e-07,
+      "loss": 1.2972,
+      "step": 2580
+    },
+    {
+      "epoch": 2.508376942111882,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.1996381101191936e-07,
+      "loss": 1.3273,
+      "step": 2581
+    },
+    {
+      "epoch": 2.509350533446919,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.187131226939036e-07,
+      "loss": 1.3228,
+      "step": 2582
+    },
+    {
+      "epoch": 2.510324124781956,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.1746471709702963e-07,
+      "loss": 1.3191,
+      "step": 2583
+    },
+    {
+      "epoch": 2.511297716116993,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.1621859552775774e-07,
+      "loss": 1.2978,
+      "step": 2584
+    },
+    {
+      "epoch": 2.51227130745203,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.1497475929015614e-07,
+      "loss": 1.307,
+      "step": 2585
+    },
+    {
+      "epoch": 2.5132448987870672,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.137332096859014e-07,
+      "loss": 1.3124,
+      "step": 2586
+    },
+    {
+      "epoch": 2.5142184901221047,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.124939480142786e-07,
+      "loss": 1.3189,
+      "step": 2587
+    },
+    {
+      "epoch": 2.515192081457142,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.11256975572177e-07,
+      "loss": 1.327,
+      "step": 2588
+    },
+    {
+      "epoch": 2.516165672792179,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.100222936540914e-07,
+      "loss": 1.322,
+      "step": 2589
+    },
+    {
+      "epoch": 2.517139264127216,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.0878990355211886e-07,
+      "loss": 1.316,
+      "step": 2590
+    },
+    {
+      "epoch": 2.518112855462253,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.075598065559571e-07,
+      "loss": 1.3309,
+      "step": 2591
+    },
+    {
+      "epoch": 2.5190864467972904,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.063320039529064e-07,
+      "loss": 1.2845,
+      "step": 2592
+    },
+    {
+      "epoch": 2.5200600381323275,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.051064970278633e-07,
+      "loss": 1.3211,
+      "step": 2593
+    },
+    {
+      "epoch": 2.5210336294673645,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.038832870633249e-07,
+      "loss": 1.3163,
+      "step": 2594
+    },
+    {
+      "epoch": 2.5220072208024016,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.0266237533938204e-07,
+      "loss": 1.3347,
+      "step": 2595
+    },
+    {
+      "epoch": 2.5229808121374386,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.014437631337211e-07,
+      "loss": 1.3153,
+      "step": 2596
+    },
+    {
+      "epoch": 2.5239544034724757,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.002274517216228e-07,
+      "loss": 1.3202,
+      "step": 2597
+    },
+    {
+      "epoch": 2.5249279948075127,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.9901344237595856e-07,
+      "loss": 1.2834,
+      "step": 2598
+    },
+    {
+      "epoch": 2.52590158614255,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.978017363671931e-07,
+      "loss": 1.3174,
+      "step": 2599
+    },
+    {
+      "epoch": 2.5268751774775873,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.965923349633779e-07,
+      "loss": 1.3101,
+      "step": 2600
+    },
+    {
+      "epoch": 2.5278487688126243,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.9538523943015455e-07,
+      "loss": 1.2955,
+      "step": 2601
+    },
+    {
+      "epoch": 2.5288223601476614,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.9418045103075137e-07,
+      "loss": 1.3198,
+      "step": 2602
+    },
+    {
+      "epoch": 2.5297959514826984,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.929779710259811e-07,
+      "loss": 1.2972,
+      "step": 2603
+    },
+    {
+      "epoch": 2.5307695428177355,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.917778006742414e-07,
+      "loss": 1.3272,
+      "step": 2604
+    },
+    {
+      "epoch": 2.531743134152773,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.905799412315141e-07,
+      "loss": 1.306,
+      "step": 2605
+    },
+    {
+      "epoch": 2.53271672548781,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.8938439395135995e-07,
+      "loss": 1.322,
+      "step": 2606
+    },
+    {
+      "epoch": 2.533690316822847,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.8819116008492327e-07,
+      "loss": 1.3134,
+      "step": 2607
+    },
+    {
+      "epoch": 2.534663908157884,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.8700024088092415e-07,
+      "loss": 1.2964,
+      "step": 2608
+    },
+    {
+      "epoch": 2.535637499492921,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.8581163758566346e-07,
+      "loss": 1.3143,
+      "step": 2609
+    },
+    {
+      "epoch": 2.536611090827958,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.8462535144301554e-07,
+      "loss": 1.3134,
+      "step": 2610
+    },
+    {
+      "epoch": 2.5375846821629953,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.834413836944325e-07,
+      "loss": 1.3109,
+      "step": 2611
+    },
+    {
+      "epoch": 2.5385582734980323,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.822597355789383e-07,
+      "loss": 1.2995,
+      "step": 2612
+    },
+    {
+      "epoch": 2.53953186483307,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.8108040833313035e-07,
+      "loss": 1.3202,
+      "step": 2613
+    },
+    {
+      "epoch": 2.540505456168107,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.799034031911765e-07,
+      "loss": 1.3064,
+      "step": 2614
+    },
+    {
+      "epoch": 2.541479047503144,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.7872872138481557e-07,
+      "loss": 1.3254,
+      "step": 2615
+    },
+    {
+      "epoch": 2.542452638838181,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.77556364143354e-07,
+      "loss": 1.3129,
+      "step": 2616
+    },
+    {
+      "epoch": 2.543426230173218,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.7638633269366666e-07,
+      "loss": 1.3286,
+      "step": 2617
+    },
+    {
+      "epoch": 2.5443998215082555,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.7521862826019317e-07,
+      "loss": 1.3266,
+      "step": 2618
+    },
+    {
+      "epoch": 2.5453734128432925,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.7405325206493914e-07,
+      "loss": 1.2944,
+      "step": 2619
+    },
+    {
+      "epoch": 2.5463470041783296,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.7289020532747263e-07,
+      "loss": 1.3001,
+      "step": 2620
+    },
+    {
+      "epoch": 2.5473205955133666,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.7172948926492497e-07,
+      "loss": 1.3071,
+      "step": 2621
+    },
+    {
+      "epoch": 2.5482941868484037,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.705711050919871e-07,
+      "loss": 1.3266,
+      "step": 2622
+    },
+    {
+      "epoch": 2.5492677781834407,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.694150540209117e-07,
+      "loss": 1.3262,
+      "step": 2623
+    },
+    {
+      "epoch": 2.550241369518478,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.68261337261507e-07,
+      "loss": 1.3199,
+      "step": 2624
+    },
+    {
+      "epoch": 2.551214960853515,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.67109956021141e-07,
+      "loss": 1.3013,
+      "step": 2625
+    },
+    {
+      "epoch": 2.5521885521885523,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.659609115047354e-07,
+      "loss": 1.3188,
+      "step": 2626
+    },
+    {
+      "epoch": 2.5531621435235894,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.648142049147692e-07,
+      "loss": 1.3065,
+      "step": 2627
+    },
+    {
+      "epoch": 2.5541357348586264,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.6366983745127197e-07,
+      "loss": 1.3006,
+      "step": 2628
+    },
+    {
+      "epoch": 2.5551093261936635,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.62527810311827e-07,
+      "loss": 1.3011,
+      "step": 2629
+    },
+    {
+      "epoch": 2.5560829175287005,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.6138812469156784e-07,
+      "loss": 1.2994,
+      "step": 2630
+    },
+    {
+      "epoch": 2.557056508863738,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.602507817831784e-07,
+      "loss": 1.3111,
+      "step": 2631
+    },
+    {
+      "epoch": 2.558030100198775,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.591157827768892e-07,
+      "loss": 1.3137,
+      "step": 2632
+    },
+    {
+      "epoch": 2.559003691533812,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.5798312886048034e-07,
+      "loss": 1.3057,
+      "step": 2633
+    },
+    {
+      "epoch": 2.559977282868849,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.568528212192756e-07,
+      "loss": 1.3158,
+      "step": 2634
+    },
+    {
+      "epoch": 2.5609508742038862,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.557248610361443e-07,
+      "loss": 1.3071,
+      "step": 2635
+    },
+    {
+      "epoch": 2.5619244655389233,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.5459924949149896e-07,
+      "loss": 1.3166,
+      "step": 2636
+    },
+    {
+      "epoch": 2.5628980568739603,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.53475987763295e-07,
+      "loss": 1.312,
+      "step": 2637
+    },
+    {
+      "epoch": 2.5638716482089974,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.523550770270269e-07,
+      "loss": 1.2926,
+      "step": 2638
+    },
+    {
+      "epoch": 2.564845239544035,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.512365184557314e-07,
+      "loss": 1.3246,
+      "step": 2639
+    },
+    {
+      "epoch": 2.565818830879072,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.5012031321998117e-07,
+      "loss": 1.3306,
+      "step": 2640
+    },
+    {
+      "epoch": 2.566792422214109,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.4900646248788823e-07,
+      "loss": 1.3108,
+      "step": 2641
+    },
+    {
+      "epoch": 2.567766013549146,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.4789496742509883e-07,
+      "loss": 1.3182,
+      "step": 2642
+    },
+    {
+      "epoch": 2.568739604884183,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.4678582919479557e-07,
+      "loss": 1.3009,
+      "step": 2643
+    },
+    {
+      "epoch": 2.5697131962192206,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.4567904895769267e-07,
+      "loss": 1.3197,
+      "step": 2644
+    },
+    {
+      "epoch": 2.5706867875542576,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.445746278720398e-07,
+      "loss": 1.3177,
+      "step": 2645
+    },
+    {
+      "epoch": 2.5716603788892947,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.434725670936139e-07,
+      "loss": 1.3048,
+      "step": 2646
+    },
+    {
+      "epoch": 2.5726339702243317,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.423728677757248e-07,
+      "loss": 1.3191,
+      "step": 2647
+    },
+    {
+      "epoch": 2.5736075615593688,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.412755310692097e-07,
+      "loss": 1.3201,
+      "step": 2648
+    },
+    {
+      "epoch": 2.574581152894406,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.4018055812243394e-07,
+      "loss": 1.3176,
+      "step": 2649
+    },
+    {
+      "epoch": 2.575554744229443,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.3908795008128873e-07,
+      "loss": 1.3146,
+      "step": 2650
+    },
+    {
+      "epoch": 2.57652833556448,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.3799770808919126e-07,
+      "loss": 1.3261,
+      "step": 2651
+    },
+    {
+      "epoch": 2.5775019268995174,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.369098332870809e-07,
+      "loss": 1.3223,
+      "step": 2652
+    },
+    {
+      "epoch": 2.5784755182345545,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.3582432681342194e-07,
+      "loss": 1.3249,
+      "step": 2653
+    },
+    {
+      "epoch": 2.5794491095695915,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.347411898041979e-07,
+      "loss": 1.3046,
+      "step": 2654
+    },
+    {
+      "epoch": 2.5804227009046286,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.3366042339291517e-07,
+      "loss": 1.3086,
+      "step": 2655
+    },
+    {
+      "epoch": 2.5813962922396656,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.325820287105973e-07,
+      "loss": 1.2968,
+      "step": 2656
+    },
+    {
+      "epoch": 2.582369883574703,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.3150600688578661e-07,
+      "loss": 1.302,
+      "step": 2657
+    },
+    {
+      "epoch": 2.58334347490974,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.3043235904454148e-07,
+      "loss": 1.3028,
+      "step": 2658
+    },
+    {
+      "epoch": 2.584317066244777,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.2936108631043785e-07,
+      "loss": 1.3239,
+      "step": 2659
+    },
+    {
+      "epoch": 2.5852906575798142,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.2829218980456342e-07,
+      "loss": 1.3075,
+      "step": 2660
+    },
+    {
+      "epoch": 2.5862642489148513,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.272256706455217e-07,
+      "loss": 1.3313,
+      "step": 2661
+    },
+    {
+      "epoch": 2.5872378402498883,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.261615299494263e-07,
+      "loss": 1.3042,
+      "step": 2662
+    },
+    {
+      "epoch": 2.5882114315849254,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.2509976882990364e-07,
+      "loss": 1.3205,
+      "step": 2663
+    },
+    {
+      "epoch": 2.5891850229199624,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.2404038839808766e-07,
+      "loss": 1.3346,
+      "step": 2664
+    },
+    {
+      "epoch": 2.590158614255,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.2298338976262324e-07,
+      "loss": 1.3188,
+      "step": 2665
+    },
+    {
+      "epoch": 2.591132205590037,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.219287740296605e-07,
+      "loss": 1.3062,
+      "step": 2666
+    },
+    {
+      "epoch": 2.592105796925074,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.208765423028586e-07,
+      "loss": 1.3295,
+      "step": 2667
+    },
+    {
+      "epoch": 2.593079388260111,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.1982669568337806e-07,
+      "loss": 1.3132,
+      "step": 2668
+    },
+    {
+      "epoch": 2.594052979595148,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.18779235269887e-07,
+      "loss": 1.3178,
+      "step": 2669
+    },
+    {
+      "epoch": 2.5950265709301856,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.1773416215855407e-07,
+      "loss": 1.3137,
+      "step": 2670
+    },
+    {
+      "epoch": 2.5960001622652227,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.166914774430512e-07,
+      "loss": 1.3098,
+      "step": 2671
+    },
+    {
+      "epoch": 2.5969737536002597,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.1565118221454905e-07,
+      "loss": 1.3295,
+      "step": 2672
+    },
+    {
+      "epoch": 2.5979473449352968,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.146132775617199e-07,
+      "loss": 1.3178,
+      "step": 2673
+    },
+    {
+      "epoch": 2.598920936270334,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.135777645707318e-07,
+      "loss": 1.3179,
+      "step": 2674
+    },
+    {
+      "epoch": 2.599894527605371,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.1254464432525214e-07,
+      "loss": 1.3161,
+      "step": 2675
+    },
+    {
+      "epoch": 2.600868118940408,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.1151391790644322e-07,
+      "loss": 1.3216,
+      "step": 2676
+    },
+    {
+      "epoch": 2.601841710275445,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.104855863929617e-07,
+      "loss": 1.3194,
+      "step": 2677
+    },
+    {
+      "epoch": 2.6028153016104825,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.094596508609595e-07,
+      "loss": 1.31,
+      "step": 2678
+    },
+    {
+      "epoch": 2.6037888929455195,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.0843611238407945e-07,
+      "loss": 1.3241,
+      "step": 2679
+    },
+    {
+      "epoch": 2.6047624842805566,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.0741497203345673e-07,
+      "loss": 1.3255,
+      "step": 2680
+    },
+    {
+      "epoch": 2.6057360756155936,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.063962308777176e-07,
+      "loss": 1.2954,
+      "step": 2681
+    },
+    {
+      "epoch": 2.6067096669506307,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.0537988998297565e-07,
+      "loss": 1.3256,
+      "step": 2682
+    },
+    {
+      "epoch": 2.607683258285668,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.0436595041283454e-07,
+      "loss": 1.3044,
+      "step": 2683
+    },
+    {
+      "epoch": 2.608656849620705,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.033544132283838e-07,
+      "loss": 1.327,
+      "step": 2684
+    },
+    {
+      "epoch": 2.6096304409557423,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.023452794881986e-07,
+      "loss": 1.3128,
+      "step": 2685
+    },
+    {
+      "epoch": 2.6106040322907793,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.013385502483406e-07,
+      "loss": 1.3102,
+      "step": 2686
+    },
+    {
+      "epoch": 2.6115776236258164,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.0033422656235258e-07,
+      "loss": 1.3093,
+      "step": 2687
+    },
+    {
+      "epoch": 2.6125512149608534,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.993323094812627e-07,
+      "loss": 1.3181,
+      "step": 2688
+    },
+    {
+      "epoch": 2.6135248062958905,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9833280005357864e-07,
+      "loss": 1.3326,
+      "step": 2689
+    },
+    {
+      "epoch": 2.6144983976309275,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.973356993252884e-07,
+      "loss": 1.3082,
+      "step": 2690
+    },
+    {
+      "epoch": 2.615471988965965,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.963410083398609e-07,
+      "loss": 1.3226,
+      "step": 2691
+    },
+    {
+      "epoch": 2.616445580301002,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9534872813824158e-07,
+      "loss": 1.3134,
+      "step": 2692
+    },
+    {
+      "epoch": 2.617419171636039,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9435885975885443e-07,
+      "loss": 1.3277,
+      "step": 2693
+    },
+    {
+      "epoch": 2.618392762971076,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.9337140423759838e-07,
+      "loss": 1.3206,
+      "step": 2694
+    },
+    {
+      "epoch": 2.619366354306113,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.9238636260784675e-07,
+      "loss": 1.3258,
+      "step": 2695
+    },
+    {
+      "epoch": 2.6203399456411507,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.914037359004489e-07,
+      "loss": 1.3277,
+      "step": 2696
+    },
+    {
+      "epoch": 2.6213135369761877,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.9042352514372504e-07,
+      "loss": 1.3244,
+      "step": 2697
+    },
+    {
+      "epoch": 2.622287128311225,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.894457313634679e-07,
+      "loss": 1.296,
+      "step": 2698
+    },
+    {
+      "epoch": 2.623260719646262,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.8847035558294037e-07,
+      "loss": 1.3252,
+      "step": 2699
+    },
+    {
+      "epoch": 2.624234310981299,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8749739882287566e-07,
+      "loss": 1.304,
+      "step": 2700
+    },
+    {
+      "epoch": 2.625207902316336,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.8652686210147485e-07,
+      "loss": 1.3261,
+      "step": 2701
+    },
+    {
+      "epoch": 2.626181493651373,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.8555874643440662e-07,
+      "loss": 1.3186,
+      "step": 2702
+    },
+    {
+      "epoch": 2.62715508498641,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8459305283480528e-07,
+      "loss": 1.3255,
+      "step": 2703
+    },
+    {
+      "epoch": 2.6281286763214475,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8362978231327184e-07,
+      "loss": 1.3142,
+      "step": 2704
+    },
+    {
+      "epoch": 2.6291022676564846,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.826689358778705e-07,
+      "loss": 1.2996,
+      "step": 2705
+    },
+    {
+      "epoch": 2.6300758589915216,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.817105145341297e-07,
+      "loss": 1.3071,
+      "step": 2706
+    },
+    {
+      "epoch": 2.6310494503265587,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.807545192850385e-07,
+      "loss": 1.3153,
+      "step": 2707
+    },
+    {
+      "epoch": 2.6320230416615957,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.7980095113104835e-07,
+      "loss": 1.2973,
+      "step": 2708
+    },
+    {
+      "epoch": 2.6329966329966332,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7884981107006981e-07,
+      "loss": 1.3153,
+      "step": 2709
+    },
+    {
+      "epoch": 2.6339702243316703,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7790110009747368e-07,
+      "loss": 1.3058,
+      "step": 2710
+    },
+    {
+      "epoch": 2.6349438156667073,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7695481920608716e-07,
+      "loss": 1.3462,
+      "step": 2711
+    },
+    {
+      "epoch": 2.6359174070017444,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7601096938619556e-07,
+      "loss": 1.2838,
+      "step": 2712
+    },
+    {
+      "epoch": 2.6368909983367814,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.7506955162553908e-07,
+      "loss": 1.2936,
+      "step": 2713
+    },
+    {
+      "epoch": 2.6378645896718185,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.7413056690931406e-07,
+      "loss": 1.3187,
+      "step": 2714
+    },
+    {
+      "epoch": 2.6388381810068555,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.7319401622016897e-07,
+      "loss": 1.3051,
+      "step": 2715
+    },
+    {
+      "epoch": 2.6398117723418926,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7225990053820724e-07,
+      "loss": 1.3195,
+      "step": 2716
+    },
+    {
+      "epoch": 2.64078536367693,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7132822084098165e-07,
+      "loss": 1.3136,
+      "step": 2717
+    },
+    {
+      "epoch": 2.641758955011967,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7039897810349786e-07,
+      "loss": 1.3162,
+      "step": 2718
+    },
+    {
+      "epoch": 2.642732546347004,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6947217329820937e-07,
+      "loss": 1.3118,
+      "step": 2719
+    },
+    {
+      "epoch": 2.643706137682041,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6854780739502003e-07,
+      "loss": 1.3109,
+      "step": 2720
+    },
+    {
+      "epoch": 2.6446797290170783,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6762588136127995e-07,
+      "loss": 1.3261,
+      "step": 2721
+    },
+    {
+      "epoch": 2.6456533203521158,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6670639616178792e-07,
+      "loss": 1.3326,
+      "step": 2722
+    },
+    {
+      "epoch": 2.646626911687153,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6578935275878533e-07,
+      "loss": 1.3104,
+      "step": 2723
+    },
+    {
+      "epoch": 2.64760050302219,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6487475211196118e-07,
+      "loss": 1.3226,
+      "step": 2724
+    },
+    {
+      "epoch": 2.648574094357227,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6396259517844598e-07,
+      "loss": 1.3035,
+      "step": 2725
+    },
+    {
+      "epoch": 2.649547685692264,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.6305288291281474e-07,
+      "loss": 1.3106,
+      "step": 2726
+    },
+    {
+      "epoch": 2.650521277027301,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6214561626708258e-07,
+      "loss": 1.3109,
+      "step": 2727
+    },
+    {
+      "epoch": 2.651494868362338,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6124079619070614e-07,
+      "loss": 1.3198,
+      "step": 2728
+    },
+    {
+      "epoch": 2.652468459697375,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6033842363058134e-07,
+      "loss": 1.3292,
+      "step": 2729
+    },
+    {
+      "epoch": 2.6534420510324126,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.594384995310433e-07,
+      "loss": 1.2973,
+      "step": 2730
+    },
+    {
+      "epoch": 2.6544156423674496,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.585410248338634e-07,
+      "loss": 1.3261,
+      "step": 2731
+    },
+    {
+      "epoch": 2.6553892337024867,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5764600047825207e-07,
+      "loss": 1.3151,
+      "step": 2732
+    },
+    {
+      "epoch": 2.6563628250375237,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.567534274008531e-07,
+      "loss": 1.3178,
+      "step": 2733
+    },
+    {
+      "epoch": 2.657336416372561,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5586330653574704e-07,
+      "loss": 1.3188,
+      "step": 2734
+    },
+    {
+      "epoch": 2.6583100077075983,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.5497563881444577e-07,
+      "loss": 1.3305,
+      "step": 2735
+    },
+    {
+      "epoch": 2.6592835990426353,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.5409042516589646e-07,
+      "loss": 1.3147,
+      "step": 2736
+    },
+    {
+      "epoch": 2.6602571903776724,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5320766651647613e-07,
+      "loss": 1.3299,
+      "step": 2737
+    },
+    {
+      "epoch": 2.6612307817127094,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.5232736378999468e-07,
+      "loss": 1.345,
+      "step": 2738
+    },
+    {
+      "epoch": 2.6622043730477465,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5144951790768942e-07,
+      "loss": 1.3179,
+      "step": 2739
+    },
+    {
+      "epoch": 2.6631779643827835,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.5057412978822906e-07,
+      "loss": 1.3291,
+      "step": 2740
+    },
+    {
+      "epoch": 2.6641515557178206,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4970120034770775e-07,
+      "loss": 1.3238,
+      "step": 2741
+    },
+    {
+      "epoch": 2.6651251470528576,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.488307304996492e-07,
+      "loss": 1.3265,
+      "step": 2742
+    },
+    {
+      "epoch": 2.666098738387895,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.4796272115500082e-07,
+      "loss": 1.3026,
+      "step": 2743
+    },
+    {
+      "epoch": 2.667072329722932,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.4709717322213712e-07,
+      "loss": 1.3354,
+      "step": 2744
+    },
+    {
+      "epoch": 2.6680459210579692,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.462340876068549e-07,
+      "loss": 1.3258,
+      "step": 2745
+    },
+    {
+      "epoch": 2.6690195123930063,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.453734652123756e-07,
+      "loss": 1.3267,
+      "step": 2746
+    },
+    {
+      "epoch": 2.6699931037280433,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.445153069393418e-07,
+      "loss": 1.3174,
+      "step": 2747
+    },
+    {
+      "epoch": 2.670966695063081,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.4365961368581844e-07,
+      "loss": 1.3072,
+      "step": 2748
+    },
+    {
+      "epoch": 2.671940286398118,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.428063863472895e-07,
+      "loss": 1.3099,
+      "step": 2749
+    },
+    {
+      "epoch": 2.672913877733155,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4195562581666017e-07,
+      "loss": 1.3071,
+      "step": 2750
+    },
+    {
+      "epoch": 2.673887469068192,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.411073329842519e-07,
+      "loss": 1.3049,
+      "step": 2751
+    },
+    {
+      "epoch": 2.674861060403229,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.4026150873780564e-07,
+      "loss": 1.3209,
+      "step": 2752
+    },
+    {
+      "epoch": 2.675834651738266,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3941815396247783e-07,
+      "loss": 1.3175,
+      "step": 2753
+    },
+    {
+      "epoch": 2.676808243073303,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3857726954084134e-07,
+      "loss": 1.3318,
+      "step": 2754
+    },
+    {
+      "epoch": 2.67778183440834,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3773885635288308e-07,
+      "loss": 1.3257,
+      "step": 2755
+    },
+    {
+      "epoch": 2.6787554257433777,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3690291527600458e-07,
+      "loss": 1.2978,
+      "step": 2756
+    },
+    {
+      "epoch": 2.6797290170784147,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3606944718501908e-07,
+      "loss": 1.3257,
+      "step": 2757
+    },
+    {
+      "epoch": 2.6807026084134518,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.3523845295215332e-07,
+      "loss": 1.3187,
+      "step": 2758
+    },
+    {
+      "epoch": 2.681676199748489,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.344099334470439e-07,
+      "loss": 1.3108,
+      "step": 2759
+    },
+    {
+      "epoch": 2.682649791083526,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.335838895367389e-07,
+      "loss": 1.3245,
+      "step": 2760
+    },
+    {
+      "epoch": 2.6836233824185634,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.3276032208569407e-07,
+      "loss": 1.309,
+      "step": 2761
+    },
+    {
+      "epoch": 2.6845969737536004,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.319392319557755e-07,
+      "loss": 1.3319,
+      "step": 2762
+    },
+    {
+      "epoch": 2.6855705650886375,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3112062000625452e-07,
+      "loss": 1.3149,
+      "step": 2763
+    },
+    {
+      "epoch": 2.6865441564236745,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3030448709381082e-07,
+      "loss": 1.3122,
+      "step": 2764
+    },
+    {
+      "epoch": 2.6875177477587116,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2949083407252898e-07,
+      "loss": 1.303,
+      "step": 2765
+    },
+    {
+      "epoch": 2.6884913390937486,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2867966179389902e-07,
+      "loss": 1.3133,
+      "step": 2766
+    },
+    {
+      "epoch": 2.6894649304287856,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.278709711068138e-07,
+      "loss": 1.3121,
+      "step": 2767
+    },
+    {
+      "epoch": 2.6904385217638227,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2706476285756997e-07,
+      "loss": 1.323,
+      "step": 2768
+    },
+    {
+      "epoch": 2.69141211309886,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.26261037889866e-07,
+      "loss": 1.3148,
+      "step": 2769
+    },
+    {
+      "epoch": 2.6923857044338972,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2545979704480181e-07,
+      "loss": 1.2991,
+      "step": 2770
+    },
+    {
+      "epoch": 2.6933592957689343,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2466104116087728e-07,
+      "loss": 1.3178,
+      "step": 2771
+    },
+    {
+      "epoch": 2.6943328871039713,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2386477107399264e-07,
+      "loss": 1.3164,
+      "step": 2772
+    },
+    {
+      "epoch": 2.695306478439009,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2307098761744547e-07,
+      "loss": 1.3033,
+      "step": 2773
+    },
+    {
+      "epoch": 2.696280069774046,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2227969162193238e-07,
+      "loss": 1.3221,
+      "step": 2774
+    },
+    {
+      "epoch": 2.697253661109083,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2149088391554599e-07,
+      "loss": 1.3038,
+      "step": 2775
+    },
+    {
+      "epoch": 2.69822725244412,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2070456532377483e-07,
+      "loss": 1.2977,
+      "step": 2776
+    },
+    {
+      "epoch": 2.699200843779157,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1992073666950376e-07,
+      "loss": 1.3398,
+      "step": 2777
+    },
+    {
+      "epoch": 2.700174435114194,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1913939877301023e-07,
+      "loss": 1.3146,
+      "step": 2778
+    },
+    {
+      "epoch": 2.701148026449231,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.1836055245196598e-07,
+      "loss": 1.3235,
+      "step": 2779
+    },
+    {
+      "epoch": 2.702121617784268,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.1758419852143599e-07,
+      "loss": 1.2857,
+      "step": 2780
+    },
+    {
+      "epoch": 2.7030952091193052,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.1681033779387507e-07,
+      "loss": 1.3169,
+      "step": 2781
+    },
+    {
+      "epoch": 2.7040688004543427,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.1603897107913126e-07,
+      "loss": 1.3133,
+      "step": 2782
+    },
+    {
+      "epoch": 2.7050423917893798,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.1527009918444076e-07,
+      "loss": 1.3187,
+      "step": 2783
+    },
+    {
+      "epoch": 2.706015983124417,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.145037229144294e-07,
+      "loss": 1.3237,
+      "step": 2784
+    },
+    {
+      "epoch": 2.706989574459454,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.137398430711123e-07,
+      "loss": 1.3193,
+      "step": 2785
+    },
+    {
+      "epoch": 2.7079631657944914,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.1297846045389e-07,
+      "loss": 1.3331,
+      "step": 2786
+    },
+    {
+      "epoch": 2.7089367571295284,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.1221957585955207e-07,
+      "loss": 1.2998,
+      "step": 2787
+    },
+    {
+      "epoch": 2.7099103484645655,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.1146319008227214e-07,
+      "loss": 1.3189,
+      "step": 2788
+    },
+    {
+      "epoch": 2.7108839397996025,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1070930391361007e-07,
+      "loss": 1.3314,
+      "step": 2789
+    },
+    {
+      "epoch": 2.7118575311346396,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.0995791814250894e-07,
+      "loss": 1.2824,
+      "step": 2790
+    },
+    {
+      "epoch": 2.7128311224696766,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.0920903355529505e-07,
+      "loss": 1.3075,
+      "step": 2791
+    },
+    {
+      "epoch": 2.7138047138047137,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0846265093567815e-07,
+      "loss": 1.2965,
+      "step": 2792
+    },
+    {
+      "epoch": 2.7147783051397507,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0771877106474927e-07,
+      "loss": 1.3235,
+      "step": 2793
+    },
+    {
+      "epoch": 2.715751896474788,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0697739472097956e-07,
+      "loss": 1.3308,
+      "step": 2794
+    },
+    {
+      "epoch": 2.7167254878098253,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.062385226802215e-07,
+      "loss": 1.338,
+      "step": 2795
+    },
+    {
+      "epoch": 2.7176990791448623,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0550215571570599e-07,
+      "loss": 1.3218,
+      "step": 2796
+    },
+    {
+      "epoch": 2.7186726704798994,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.0476829459804245e-07,
+      "loss": 1.3114,
+      "step": 2797
+    },
+    {
+      "epoch": 2.7196462618149364,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.0403694009521793e-07,
+      "loss": 1.3237,
+      "step": 2798
+    },
+    {
+      "epoch": 2.720619853149974,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.0330809297259714e-07,
+      "loss": 1.33,
+      "step": 2799
+    },
+    {
+      "epoch": 2.721593444485011,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.0258175399291914e-07,
+      "loss": 1.3131,
+      "step": 2800
+    },
+    {
+      "epoch": 2.722567035820048,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.0185792391629978e-07,
+      "loss": 1.3231,
+      "step": 2801
+    },
+    {
+      "epoch": 2.723540627155085,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.0113660350022786e-07,
+      "loss": 1.3088,
+      "step": 2802
+    },
+    {
+      "epoch": 2.724514218490122,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.0041779349956788e-07,
+      "loss": 1.3269,
+      "step": 2803
+    },
+    {
+      "epoch": 2.725487809825159,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.970149466655477e-08,
+      "loss": 1.3081,
+      "step": 2804
+    },
+    {
+      "epoch": 2.726461401160196,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.898770775079752e-08,
+      "loss": 1.3188,
+      "step": 2805
+    },
+    {
+      "epoch": 2.7274349924952332,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.82764334992753e-08,
+      "loss": 1.303,
+      "step": 2806
+    },
+    {
+      "epoch": 2.7284085838302707,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.75676726563382e-08,
+      "loss": 1.312,
+      "step": 2807
+    },
+    {
+      "epoch": 2.729382175165308,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.686142596370545e-08,
+      "loss": 1.315,
+      "step": 2808
+    },
+    {
+      "epoch": 2.730355766500345,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.615769416046639e-08,
+      "loss": 1.2908,
+      "step": 2809
+    },
+    {
+      "epoch": 2.731329357835382,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.54564779830769e-08,
+      "loss": 1.3208,
+      "step": 2810
+    },
+    {
+      "epoch": 2.732302949170419,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.475777816536219e-08,
+      "loss": 1.3289,
+      "step": 2811
+    },
+    {
+      "epoch": 2.7332765405054564,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.406159543851184e-08,
+      "loss": 1.3351,
+      "step": 2812
+    },
+    {
+      "epoch": 2.7342501318404935,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.336793053108361e-08,
+      "loss": 1.3239,
+      "step": 2813
+    },
+    {
+      "epoch": 2.7352237231755305,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.267678416899823e-08,
+      "loss": 1.3134,
+      "step": 2814
+    },
+    {
+      "epoch": 2.7361973145105676,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.198815707554237e-08,
+      "loss": 1.3196,
+      "step": 2815
+    },
+    {
+      "epoch": 2.7371709058456046,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.130204997136543e-08,
+      "loss": 1.306,
+      "step": 2816
+    },
+    {
+      "epoch": 2.7381444971806417,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.061846357448028e-08,
+      "loss": 1.2892,
+      "step": 2817
+    },
+    {
+      "epoch": 2.7391180885156787,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.993739860026108e-08,
+      "loss": 1.3408,
+      "step": 2818
+    },
+    {
+      "epoch": 2.7400916798507158,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.925885576144377e-08,
+      "loss": 1.3204,
+      "step": 2819
+    },
+    {
+      "epoch": 2.7410652711857533,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.858283576812482e-08,
+      "loss": 1.3129,
+      "step": 2820
+    },
+    {
+      "epoch": 2.7420388625207903,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.79093393277608e-08,
+      "loss": 1.3181,
+      "step": 2821
+    },
+    {
+      "epoch": 2.7430124538558274,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.723836714516681e-08,
+      "loss": 1.3159,
+      "step": 2822
+    },
+    {
+      "epoch": 2.7439860451908644,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.656991992251674e-08,
+      "loss": 1.3085,
+      "step": 2823
+    },
+    {
+      "epoch": 2.7449596365259015,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.590399835934154e-08,
+      "loss": 1.3067,
+      "step": 2824
+    },
+    {
+      "epoch": 2.745933227860939,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.524060315253019e-08,
+      "loss": 1.3227,
+      "step": 2825
+    },
+    {
+      "epoch": 2.746906819195976,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.45797349963265e-08,
+      "loss": 1.3254,
+      "step": 2826
+    },
+    {
+      "epoch": 2.747880410531013,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.392139458233056e-08,
+      "loss": 1.3098,
+      "step": 2827
+    },
+    {
+      "epoch": 2.74885400186605,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.326558259949713e-08,
+      "loss": 1.3429,
+      "step": 2828
+    },
+    {
+      "epoch": 2.749827593201087,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.26122997341347e-08,
+      "loss": 1.3122,
+      "step": 2829
+    },
+    {
+      "epoch": 2.750801184536124,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.196154666990475e-08,
+      "loss": 1.3015,
+      "step": 2830
+    },
+    {
+      "epoch": 2.7517747758711613,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.131332408782222e-08,
+      "loss": 1.3588,
+      "step": 2831
+    },
+    {
+      "epoch": 2.7527483672061983,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.066763266625283e-08,
+      "loss": 1.3171,
+      "step": 2832
+    },
+    {
+      "epoch": 2.753721958541236,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.002447308091466e-08,
+      "loss": 1.3136,
+      "step": 2833
+    },
+    {
+      "epoch": 2.754695549876273,
+      "grad_norm": 0.384765625,
+      "learning_rate": 7.938384600487486e-08,
+      "loss": 1.3255,
+      "step": 2834
+    },
+    {
+      "epoch": 2.75566914121131,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.874575210855134e-08,
+      "loss": 1.2997,
+      "step": 2835
+    },
+    {
+      "epoch": 2.756642732546347,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.811019205971021e-08,
+      "loss": 1.3134,
+      "step": 2836
+    },
+    {
+      "epoch": 2.757616323881384,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.747716652346692e-08,
+      "loss": 1.3239,
+      "step": 2837
+    },
+    {
+      "epoch": 2.7585899152164215,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.684667616228353e-08,
+      "loss": 1.3154,
+      "step": 2838
+    },
+    {
+      "epoch": 2.7595635065514585,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.621872163596999e-08,
+      "loss": 1.3054,
+      "step": 2839
+    },
+    {
+      "epoch": 2.7605370978864956,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.559330360168148e-08,
+      "loss": 1.3236,
+      "step": 2840
+    },
+    {
+      "epoch": 2.7615106892215326,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.497042271391947e-08,
+      "loss": 1.2818,
+      "step": 2841
+    },
+    {
+      "epoch": 2.7624842805565697,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.435007962452972e-08,
+      "loss": 1.3123,
+      "step": 2842
+    },
+    {
+      "epoch": 2.7634578718916067,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.373227498270353e-08,
+      "loss": 1.3257,
+      "step": 2843
+    },
+    {
+      "epoch": 2.764431463226644,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.311700943497369e-08,
+      "loss": 1.3013,
+      "step": 2844
+    },
+    {
+      "epoch": 2.765405054561681,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.250428362521711e-08,
+      "loss": 1.3261,
+      "step": 2845
+    },
+    {
+      "epoch": 2.7663786458967183,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.189409819465255e-08,
+      "loss": 1.3097,
+      "step": 2846
+    },
+    {
+      "epoch": 2.7673522372317554,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.128645378184085e-08,
+      "loss": 1.2819,
+      "step": 2847
+    },
+    {
+      "epoch": 2.7683258285667924,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.068135102268226e-08,
+      "loss": 1.3146,
+      "step": 2848
+    },
+    {
+      "epoch": 2.7692994199018295,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.007879055041855e-08,
+      "loss": 1.3027,
+      "step": 2849
+    },
+    {
+      "epoch": 2.7702730112368665,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.947877299563032e-08,
+      "loss": 1.3155,
+      "step": 2850
+    },
+    {
+      "epoch": 2.771246602571904,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.888129898623752e-08,
+      "loss": 1.3179,
+      "step": 2851
+    },
+    {
+      "epoch": 2.772220193906941,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.828636914749748e-08,
+      "loss": 1.3121,
+      "step": 2852
+    },
+    {
+      "epoch": 2.773193785241978,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.769398410200579e-08,
+      "loss": 1.3119,
+      "step": 2853
+    },
+    {
+      "epoch": 2.774167376577015,
+      "grad_norm": 0.388671875,
+      "learning_rate": 6.710414446969405e-08,
+      "loss": 1.3194,
+      "step": 2854
+    },
+    {
+      "epoch": 2.7751409679120522,
+      "grad_norm": 0.390625,
+      "learning_rate": 6.651685086783155e-08,
+      "loss": 1.3126,
+      "step": 2855
+    },
+    {
+      "epoch": 2.7761145592470893,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.593210391102139e-08,
+      "loss": 1.3027,
+      "step": 2856
+    },
+    {
+      "epoch": 2.7770881505821263,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.534990421120296e-08,
+      "loss": 1.3011,
+      "step": 2857
+    },
+    {
+      "epoch": 2.7780617419171634,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.477025237764889e-08,
+      "loss": 1.3246,
+      "step": 2858
+    },
+    {
+      "epoch": 2.779035333252201,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.419314901696671e-08,
+      "loss": 1.3045,
+      "step": 2859
+    },
+    {
+      "epoch": 2.780008924587238,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.361859473309556e-08,
+      "loss": 1.3222,
+      "step": 2860
+    },
+    {
+      "epoch": 2.780982515922275,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.304659012730835e-08,
+      "loss": 1.3308,
+      "step": 2861
+    },
+    {
+      "epoch": 2.781956107257312,
+      "grad_norm": 0.38671875,
+      "learning_rate": 6.247713579820847e-08,
+      "loss": 1.3075,
+      "step": 2862
+    },
+    {
+      "epoch": 2.782929698592349,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.191023234173143e-08,
+      "loss": 1.3197,
+      "step": 2863
+    },
+    {
+      "epoch": 2.7839032899273866,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.134588035114242e-08,
+      "loss": 1.3026,
+      "step": 2864
+    },
+    {
+      "epoch": 2.7848768812624236,
+      "grad_norm": 0.388671875,
+      "learning_rate": 6.078408041703732e-08,
+      "loss": 1.311,
+      "step": 2865
+    },
+    {
+      "epoch": 2.7858504725974607,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.022483312734112e-08,
+      "loss": 1.3102,
+      "step": 2866
+    },
+    {
+      "epoch": 2.7868240639324977,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.966813906730679e-08,
+      "loss": 1.3189,
+      "step": 2867
+    },
+    {
+      "epoch": 2.7877976552675348,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.9113998819515515e-08,
+      "loss": 1.3078,
+      "step": 2868
+    },
+    {
+      "epoch": 2.788771246602572,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.8562412963877044e-08,
+      "loss": 1.2978,
+      "step": 2869
+    },
+    {
+      "epoch": 2.789744837937609,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.80133820776263e-08,
+      "loss": 1.3239,
+      "step": 2870
+    },
+    {
+      "epoch": 2.790718429272646,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.74669067353259e-08,
+      "loss": 1.3089,
+      "step": 2871
+    },
+    {
+      "epoch": 2.7916920206076834,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.6922987508862546e-08,
+      "loss": 1.294,
+      "step": 2872
+    },
+    {
+      "epoch": 2.7926656119427204,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.638162496744981e-08,
+      "loss": 1.3267,
+      "step": 2873
+    },
+    {
+      "epoch": 2.7936392032777575,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.5842819677624225e-08,
+      "loss": 1.3228,
+      "step": 2874
+    },
+    {
+      "epoch": 2.7946127946127945,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.530657220324615e-08,
+      "loss": 1.3066,
+      "step": 2875
+    },
+    {
+      "epoch": 2.7955863859478316,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.477288310550055e-08,
+      "loss": 1.302,
+      "step": 2876
+    },
+    {
+      "epoch": 2.796559977282869,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.424175294289374e-08,
+      "loss": 1.3167,
+      "step": 2877
+    },
+    {
+      "epoch": 2.797533568617906,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.371318227125416e-08,
+      "loss": 1.2813,
+      "step": 2878
+    },
+    {
+      "epoch": 2.798507159952943,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.318717164373266e-08,
+      "loss": 1.3026,
+      "step": 2879
+    },
+    {
+      "epoch": 2.7994807512879802,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.266372161079975e-08,
+      "loss": 1.3161,
+      "step": 2880
+    },
+    {
+      "epoch": 2.8004543426230173,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.2142832720247784e-08,
+      "loss": 1.3258,
+      "step": 2881
+    },
+    {
+      "epoch": 2.8014279339580543,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.16245055171874e-08,
+      "loss": 1.3059,
+      "step": 2882
+    },
+    {
+      "epoch": 2.8024015252930914,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.110874054404941e-08,
+      "loss": 1.306,
+      "step": 2883
+    },
+    {
+      "epoch": 2.8033751166281284,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.059553834058289e-08,
+      "loss": 1.3173,
+      "step": 2884
+    },
+    {
+      "epoch": 2.804348707963166,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.008489944385464e-08,
+      "loss": 1.3198,
+      "step": 2885
+    },
+    {
+      "epoch": 2.805322299298203,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.957682438824996e-08,
+      "loss": 1.3263,
+      "step": 2886
+    },
+    {
+      "epoch": 2.80629589063324,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.9071313705469635e-08,
+      "loss": 1.3138,
+      "step": 2887
+    },
+    {
+      "epoch": 2.807269481968277,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.856836792453218e-08,
+      "loss": 1.298,
+      "step": 2888
+    },
+    {
+      "epoch": 2.808243073303314,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.806798757177128e-08,
+      "loss": 1.3065,
+      "step": 2889
+    },
+    {
+      "epoch": 2.8092166646383516,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.7570173170835846e-08,
+      "loss": 1.3114,
+      "step": 2890
+    },
+    {
+      "epoch": 2.8101902559733887,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.70749252426897e-08,
+      "loss": 1.3133,
+      "step": 2891
+    },
+    {
+      "epoch": 2.8111638473084257,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.6582244305611034e-08,
+      "loss": 1.3155,
+      "step": 2892
+    },
+    {
+      "epoch": 2.8121374386434628,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.6092130875190766e-08,
+      "loss": 1.3006,
+      "step": 2893
+    },
+    {
+      "epoch": 2.8131110299785,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.5604585464334436e-08,
+      "loss": 1.3369,
+      "step": 2894
+    },
+    {
+      "epoch": 2.814084621313537,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.5119608583258646e-08,
+      "loss": 1.3226,
+      "step": 2895
+    },
+    {
+      "epoch": 2.815058212648574,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.4637200739493514e-08,
+      "loss": 1.3146,
+      "step": 2896
+    },
+    {
+      "epoch": 2.816031803983611,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.415736243787882e-08,
+      "loss": 1.308,
+      "step": 2897
+    },
+    {
+      "epoch": 2.8170053953186485,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.368009418056707e-08,
+      "loss": 1.3103,
+      "step": 2898
+    },
+    {
+      "epoch": 2.8179789866536855,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.32053964670201e-08,
+      "loss": 1.3023,
+      "step": 2899
+    },
+    {
+      "epoch": 2.8189525779887226,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.273326979400999e-08,
+      "loss": 1.3158,
+      "step": 2900
+    },
+    {
+      "epoch": 2.8199261693237596,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.226371465561846e-08,
+      "loss": 1.3245,
+      "step": 2901
+    },
+    {
+      "epoch": 2.8208997606587967,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.179673154323605e-08,
+      "loss": 1.2983,
+      "step": 2902
+    },
+    {
+      "epoch": 2.821873351993834,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.1332320945561e-08,
+      "loss": 1.3317,
+      "step": 2903
+    },
+    {
+      "epoch": 2.822846943328871,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.0870483348600386e-08,
+      "loss": 1.3029,
+      "step": 2904
+    },
+    {
+      "epoch": 2.8238205346639083,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.0411219235667877e-08,
+      "loss": 1.3209,
+      "step": 2905
+    },
+    {
+      "epoch": 2.8247941259989453,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.9954529087384844e-08,
+      "loss": 1.3222,
+      "step": 2906
+    },
+    {
+      "epoch": 2.8257677173339824,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.950041338167787e-08,
+      "loss": 1.3242,
+      "step": 2907
+    },
+    {
+      "epoch": 2.8267413086690194,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.9048872593780426e-08,
+      "loss": 1.3201,
+      "step": 2908
+    },
+    {
+      "epoch": 2.8277149000040565,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.8599907196230636e-08,
+      "loss": 1.3172,
+      "step": 2909
+    },
+    {
+      "epoch": 2.8286884913390935,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.815351765887182e-08,
+      "loss": 1.3079,
+      "step": 2910
+    },
+    {
+      "epoch": 2.829662082674131,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.770970444885114e-08,
+      "loss": 1.3196,
+      "step": 2911
+    },
+    {
+      "epoch": 2.830635674009168,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.72684680306204e-08,
+      "loss": 1.3144,
+      "step": 2912
+    },
+    {
+      "epoch": 2.831609265344205,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.682980886593412e-08,
+      "loss": 1.3211,
+      "step": 2913
+    },
+    {
+      "epoch": 2.832582856679242,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.639372741385039e-08,
+      "loss": 1.3257,
+      "step": 2914
+    },
+    {
+      "epoch": 2.833556448014279,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.596022413072886e-08,
+      "loss": 1.3351,
+      "step": 2915
+    },
+    {
+      "epoch": 2.8345300393493167,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.5529299470232206e-08,
+      "loss": 1.3217,
+      "step": 2916
+    },
+    {
+      "epoch": 2.8355036306843537,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.51009538833233e-08,
+      "loss": 1.3064,
+      "step": 2917
+    },
+    {
+      "epoch": 2.836477222019391,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.4675187818267174e-08,
+      "loss": 1.3091,
+      "step": 2918
+    },
+    {
+      "epoch": 2.837450813354428,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.4252001720628816e-08,
+      "loss": 1.3193,
+      "step": 2919
+    },
+    {
+      "epoch": 2.838424404689465,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.383139603327312e-08,
+      "loss": 1.3046,
+      "step": 2920
+    },
+    {
+      "epoch": 2.839397996024502,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.3413371196364954e-08,
+      "loss": 1.3227,
+      "step": 2921
+    },
+    {
+      "epoch": 2.840371587359539,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.299792764736798e-08,
+      "loss": 1.2949,
+      "step": 2922
+    },
+    {
+      "epoch": 2.841345178694576,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.25850658210447e-08,
+      "loss": 1.3157,
+      "step": 2923
+    },
+    {
+      "epoch": 2.8423187700296135,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.2174786149456184e-08,
+      "loss": 1.32,
+      "step": 2924
+    },
+    {
+      "epoch": 2.8432923613646506,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.176708906196063e-08,
+      "loss": 1.3047,
+      "step": 2925
+    },
+    {
+      "epoch": 2.8442659526996876,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.136197498521398e-08,
+      "loss": 1.3176,
+      "step": 2926
+    },
+    {
+      "epoch": 2.8452395440347247,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.0959444343169055e-08,
+      "loss": 1.3141,
+      "step": 2927
+    },
+    {
+      "epoch": 2.8462131353697617,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.0559497557074715e-08,
+      "loss": 1.3156,
+      "step": 2928
+    },
+    {
+      "epoch": 2.847186726704799,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.016213504547616e-08,
+      "loss": 1.3105,
+      "step": 2929
+    },
+    {
+      "epoch": 2.8481603180398363,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.9767357224214365e-08,
+      "loss": 1.3098,
+      "step": 2930
+    },
+    {
+      "epoch": 2.8491339093748733,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.9375164506424693e-08,
+      "loss": 1.311,
+      "step": 2931
+    },
+    {
+      "epoch": 2.8501075007099104,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.8985557302537994e-08,
+      "loss": 1.3018,
+      "step": 2932
+    },
+    {
+      "epoch": 2.8510810920449474,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.8598536020278678e-08,
+      "loss": 1.3184,
+      "step": 2933
+    },
+    {
+      "epoch": 2.8520546833799845,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.8214101064665545e-08,
+      "loss": 1.3352,
+      "step": 2934
+    },
+    {
+      "epoch": 2.8530282747150215,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.7832252838010387e-08,
+      "loss": 1.3224,
+      "step": 2935
+    },
+    {
+      "epoch": 2.8540018660500586,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.7452991739918277e-08,
+      "loss": 1.3408,
+      "step": 2936
+    },
+    {
+      "epoch": 2.854975457385096,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.7076318167286452e-08,
+      "loss": 1.3088,
+      "step": 2937
+    },
+    {
+      "epoch": 2.855949048720133,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.670223251430515e-08,
+      "loss": 1.298,
+      "step": 2938
+    },
+    {
+      "epoch": 2.85692264005517,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.63307351724551e-08,
+      "loss": 1.2909,
+      "step": 2939
+    },
+    {
+      "epoch": 2.857896231390207,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.5961826530509214e-08,
+      "loss": 1.3313,
+      "step": 2940
+    },
+    {
+      "epoch": 2.8588698227252443,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.5595506974531448e-08,
+      "loss": 1.3273,
+      "step": 2941
+    },
+    {
+      "epoch": 2.8598434140602818,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.5231776887875703e-08,
+      "loss": 1.3218,
+      "step": 2942
+    },
+    {
+      "epoch": 2.860817005395319,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.4870636651186388e-08,
+      "loss": 1.3012,
+      "step": 2943
+    },
+    {
+      "epoch": 2.861790596730356,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.451208664239757e-08,
+      "loss": 1.3256,
+      "step": 2944
+    },
+    {
+      "epoch": 2.862764188065393,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.4156127236732162e-08,
+      "loss": 1.3031,
+      "step": 2945
+    },
+    {
+      "epoch": 2.86373777940043,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.3802758806702463e-08,
+      "loss": 1.3073,
+      "step": 2946
+    },
+    {
+      "epoch": 2.864711370735467,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.3451981722109608e-08,
+      "loss": 1.3062,
+      "step": 2947
+    },
+    {
+      "epoch": 2.865684962070504,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.310379635004245e-08,
+      "loss": 1.3162,
+      "step": 2948
+    },
+    {
+      "epoch": 2.866658553405541,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.2758203054877302e-08,
+      "loss": 1.338,
+      "step": 2949
+    },
+    {
+      "epoch": 2.8676321447405786,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.241520219827903e-08,
+      "loss": 1.3093,
+      "step": 2950
+    },
+    {
+      "epoch": 2.8686057360756156,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.207479413919772e-08,
+      "loss": 1.3074,
+      "step": 2951
+    },
+    {
+      "epoch": 2.8695793274106527,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.1736979233872025e-08,
+      "loss": 1.3125,
+      "step": 2952
+    },
+    {
+      "epoch": 2.8705529187456897,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.1401757835824988e-08,
+      "loss": 1.3036,
+      "step": 2953
+    },
+    {
+      "epoch": 2.871526510080727,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.1069130295867656e-08,
+      "loss": 1.3257,
+      "step": 2954
+    },
+    {
+      "epoch": 2.8725001014157643,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.073909696209436e-08,
+      "loss": 1.3061,
+      "step": 2955
+    },
+    {
+      "epoch": 2.8734736927508013,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.0411658179886053e-08,
+      "loss": 1.3274,
+      "step": 2956
+    },
+    {
+      "epoch": 2.8744472840858384,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.0086814291908064e-08,
+      "loss": 1.3454,
+      "step": 2957
+    },
+    {
+      "epoch": 2.8754208754208754,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.976456563811041e-08,
+      "loss": 1.3012,
+      "step": 2958
+    },
+    {
+      "epoch": 2.8763944667559125,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9444912555726668e-08,
+      "loss": 1.3053,
+      "step": 2959
+    },
+    {
+      "epoch": 2.8773680580909495,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9127855379274528e-08,
+      "loss": 1.2956,
+      "step": 2960
+    },
+    {
+      "epoch": 2.8783416494259866,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8813394440554966e-08,
+      "loss": 1.292,
+      "step": 2961
+    },
+    {
+      "epoch": 2.8793152407610236,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.850153006865224e-08,
+      "loss": 1.323,
+      "step": 2962
+    },
+    {
+      "epoch": 2.880288832096061,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8192262589932507e-08,
+      "loss": 1.3126,
+      "step": 2963
+    },
+    {
+      "epoch": 2.881262423431098,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.788559232804521e-08,
+      "loss": 1.3386,
+      "step": 2964
+    },
+    {
+      "epoch": 2.8822360147661352,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.7581519603921406e-08,
+      "loss": 1.3285,
+      "step": 2965
+    },
+    {
+      "epoch": 2.8832096061011723,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.7280044735773493e-08,
+      "loss": 1.3347,
+      "step": 2966
+    },
+    {
+      "epoch": 2.8841831974362093,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.6981168039095774e-08,
+      "loss": 1.3068,
+      "step": 2967
+    },
+    {
+      "epoch": 2.885156788771247,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.668488982666361e-08,
+      "loss": 1.3398,
+      "step": 2968
+    },
+    {
+      "epoch": 2.886130380106284,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.6391210408532032e-08,
+      "loss": 1.3475,
+      "step": 2969
+    },
+    {
+      "epoch": 2.887103971441321,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.6100130092037704e-08,
+      "loss": 1.3005,
+      "step": 2970
+    },
+    {
+      "epoch": 2.888077562776358,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.5811649181796673e-08,
+      "loss": 1.306,
+      "step": 2971
+    },
+    {
+      "epoch": 2.889051154111395,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.5525767979704675e-08,
+      "loss": 1.3159,
+      "step": 2972
+    },
+    {
+      "epoch": 2.890024745446432,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5242486784937115e-08,
+      "loss": 1.3084,
+      "step": 2973
+    },
+    {
+      "epoch": 2.890998336781469,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4961805893948245e-08,
+      "loss": 1.2883,
+      "step": 2974
+    },
+    {
+      "epoch": 2.891971928116506,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.4683725600471155e-08,
+      "loss": 1.3183,
+      "step": 2975
+    },
+    {
+      "epoch": 2.8929455194515437,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.4408246195517506e-08,
+      "loss": 1.3028,
+      "step": 2976
+    },
+    {
+      "epoch": 2.8939191107865807,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4135367967377244e-08,
+      "loss": 1.3117,
+      "step": 2977
+    },
+    {
+      "epoch": 2.8948927021216178,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.3865091201617498e-08,
+      "loss": 1.3144,
+      "step": 2978
+    },
+    {
+      "epoch": 2.895866293456655,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.3597416181083678e-08,
+      "loss": 1.3064,
+      "step": 2979
+    },
+    {
+      "epoch": 2.896839884791692,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.3332343185898377e-08,
+      "loss": 1.3151,
+      "step": 2980
+    },
+    {
+      "epoch": 2.8978134761267293,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.3069872493461089e-08,
+      "loss": 1.2951,
+      "step": 2981
+    },
+    {
+      "epoch": 2.8987870674617664,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.2810004378447649e-08,
+      "loss": 1.3181,
+      "step": 2982
+    },
+    {
+      "epoch": 2.8997606587968034,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.255273911281052e-08,
+      "loss": 1.3188,
+      "step": 2983
+    },
+    {
+      "epoch": 2.9007342501318405,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2298076965778782e-08,
+      "loss": 1.3106,
+      "step": 2984
+    },
+    {
+      "epoch": 2.9017078414668775,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.204601820385648e-08,
+      "loss": 1.3223,
+      "step": 2985
+    },
+    {
+      "epoch": 2.9026814328019146,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.1796563090823443e-08,
+      "loss": 1.2958,
+      "step": 2986
+    },
+    {
+      "epoch": 2.9036550241369516,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.154971188773557e-08,
+      "loss": 1.2908,
+      "step": 2987
+    },
+    {
+      "epoch": 2.9046286154719887,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.1305464852922887e-08,
+      "loss": 1.3144,
+      "step": 2988
+    },
+    {
+      "epoch": 2.905602206807026,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1063822241990097e-08,
+      "loss": 1.3207,
+      "step": 2989
+    },
+    {
+      "epoch": 2.9065757981420632,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.0824784307817138e-08,
+      "loss": 1.3336,
+      "step": 2990
+    },
+    {
+      "epoch": 2.9075493894771003,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.058835130055752e-08,
+      "loss": 1.2953,
+      "step": 2991
+    },
+    {
+      "epoch": 2.9085229808121373,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.03545234676386e-08,
+      "loss": 1.3062,
+      "step": 2992
+    },
+    {
+      "epoch": 2.909496572147175,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.0123301053762136e-08,
+      "loss": 1.3272,
+      "step": 2993
+    },
+    {
+      "epoch": 2.910470163482212,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.894684300902347e-09,
+      "loss": 1.3112,
+      "step": 2994
+    },
+    {
+      "epoch": 2.911443754817249,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.668673448307575e-09,
+      "loss": 1.3279,
+      "step": 2995
+    },
+    {
+      "epoch": 2.912417346152286,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.445268732498625e-09,
+      "loss": 1.3128,
+      "step": 2996
+    },
+    {
+      "epoch": 2.913390937487323,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.224470387268759e-09,
+      "loss": 1.3207,
+      "step": 2997
+    },
+    {
+      "epoch": 2.91436452882236,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.006278643683697e-09,
+      "loss": 1.3169,
+      "step": 2998
+    },
+    {
+      "epoch": 2.915338120157397,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.790693730082179e-09,
+      "loss": 1.3137,
+      "step": 2999
+    },
+    {
+      "epoch": 2.916311711492434,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.577715872073734e-09,
+      "loss": 1.3052,
+      "step": 3000
+    },
+    {
+      "epoch": 2.9172853028274712,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.367345292540074e-09,
+      "loss": 1.3251,
+      "step": 3001
+    },
+    {
+      "epoch": 2.9182588941625087,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.15958221163482e-09,
+      "loss": 1.3077,
+      "step": 3002
+    },
+    {
+      "epoch": 2.9192324854975458,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.95442684678238e-09,
+      "loss": 1.3223,
+      "step": 3003
+    },
+    {
+      "epoch": 2.920206076832583,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.751879412677966e-09,
+      "loss": 1.3218,
+      "step": 3004
+    },
+    {
+      "epoch": 2.92117966816762,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.551940121288126e-09,
+      "loss": 1.2997,
+      "step": 3005
+    },
+    {
+      "epoch": 2.9221532595026574,
+      "grad_norm": 0.390625,
+      "learning_rate": 7.354609181849659e-09,
+      "loss": 1.3248,
+      "step": 3006
+    },
+    {
+      "epoch": 2.9231268508376944,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.159886800869875e-09,
+      "loss": 1.3029,
+      "step": 3007
+    },
+    {
+      "epoch": 2.9241004421727315,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.967773182126048e-09,
+      "loss": 1.3156,
+      "step": 3008
+    },
+    {
+      "epoch": 2.9250740335077685,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.77826852666541e-09,
+      "loss": 1.32,
+      "step": 3009
+    },
+    {
+      "epoch": 2.9260476248428056,
+      "grad_norm": 0.390625,
+      "learning_rate": 6.591373032805437e-09,
+      "loss": 1.2957,
+      "step": 3010
+    },
+    {
+      "epoch": 2.9270212161778426,
+      "grad_norm": 0.390625,
+      "learning_rate": 6.407086896131898e-09,
+      "loss": 1.3003,
+      "step": 3011
+    },
+    {
+      "epoch": 2.9279948075128797,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.225410309501079e-09,
+      "loss": 1.3187,
+      "step": 3012
+    },
+    {
+      "epoch": 2.9289683988479167,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.046343463037563e-09,
+      "loss": 1.3073,
+      "step": 3013
+    },
+    {
+      "epoch": 2.929941990182954,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.869886544135617e-09,
+      "loss": 1.3149,
+      "step": 3014
+    },
+    {
+      "epoch": 2.9309155815179913,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.696039737457526e-09,
+      "loss": 1.3011,
+      "step": 3015
+    },
+    {
+      "epoch": 2.9318891728530283,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.524803224934427e-09,
+      "loss": 1.3196,
+      "step": 3016
+    },
+    {
+      "epoch": 2.9328627641880654,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.356177185765477e-09,
+      "loss": 1.3142,
+      "step": 3017
+    },
+    {
+      "epoch": 2.9338363555231024,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.190161796418125e-09,
+      "loss": 1.3129,
+      "step": 3018
+    },
+    {
+      "epoch": 2.93480994685814,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.026757230628121e-09,
+      "loss": 1.3309,
+      "step": 3019
+    },
+    {
+      "epoch": 2.935783538193177,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.865963659398676e-09,
+      "loss": 1.3174,
+      "step": 3020
+    },
+    {
+      "epoch": 2.936757129528214,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.7077812510001875e-09,
+      "loss": 1.317,
+      "step": 3021
+    },
+    {
+      "epoch": 2.937730720863251,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.5522101709713496e-09,
+      "loss": 1.3266,
+      "step": 3022
+    },
+    {
+      "epoch": 2.938704312198288,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.3992505821172095e-09,
+      "loss": 1.3269,
+      "step": 3023
+    },
+    {
+      "epoch": 2.939677903533325,
+      "grad_norm": 0.392578125,
+      "learning_rate": 4.248902644510555e-09,
+      "loss": 1.31,
+      "step": 3024
+    },
+    {
+      "epoch": 2.940651494868362,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.1011665154905285e-09,
+      "loss": 1.3415,
+      "step": 3025
+    },
+    {
+      "epoch": 2.9416250862033992,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.956042349663736e-09,
+      "loss": 1.3199,
+      "step": 3026
+    },
+    {
+      "epoch": 2.9425986775384367,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.813530298902579e-09,
+      "loss": 1.3236,
+      "step": 3027
+    },
+    {
+      "epoch": 2.943572268873474,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.6736305123460938e-09,
+      "loss": 1.315,
+      "step": 3028
+    },
+    {
+      "epoch": 2.944545860208511,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.5363431363999447e-09,
+      "loss": 1.3063,
+      "step": 3029
+    },
+    {
+      "epoch": 2.945519451543548,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.4016683147355954e-09,
+      "loss": 1.2863,
+      "step": 3030
+    },
+    {
+      "epoch": 2.946493042878585,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.2696061882905862e-09,
+      "loss": 1.3141,
+      "step": 3031
+    },
+    {
+      "epoch": 2.9474666342136224,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.1401568952679763e-09,
+      "loss": 1.3131,
+      "step": 3032
+    },
+    {
+      "epoch": 2.9484402255486595,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.013320571136902e-09,
+      "loss": 1.3477,
+      "step": 3033
+    },
+    {
+      "epoch": 2.9494138168836965,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.8890973486320193e-09,
+      "loss": 1.2948,
+      "step": 3034
+    },
+    {
+      "epoch": 2.9503874082187336,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.767487357752674e-09,
+      "loss": 1.319,
+      "step": 3035
+    },
+    {
+      "epoch": 2.9513609995537706,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.6484907257642856e-09,
+      "loss": 1.2956,
+      "step": 3036
+    },
+    {
+      "epoch": 2.9523345908888077,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.5321075771969627e-09,
+      "loss": 1.3126,
+      "step": 3037
+    },
+    {
+      "epoch": 2.9533081822238447,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.418338033846057e-09,
+      "loss": 1.2997,
+      "step": 3038
+    },
+    {
+      "epoch": 2.9542817735588818,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.3071822147716083e-09,
+      "loss": 1.3128,
+      "step": 3039
+    },
+    {
+      "epoch": 2.9552553648939193,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.198640236298344e-09,
+      "loss": 1.3263,
+      "step": 3040
+    },
+    {
+      "epoch": 2.9562289562289563,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.092712212015402e-09,
+      "loss": 1.2956,
+      "step": 3041
+    },
+    {
+      "epoch": 2.9572025475639934,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.989398252777164e-09,
+      "loss": 1.3155,
+      "step": 3042
+    },
+    {
+      "epoch": 2.9581761388990304,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.8886984667015883e-09,
+      "loss": 1.3278,
+      "step": 3043
+    },
+    {
+      "epoch": 2.9591497302340675,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7906129591713228e-09,
+      "loss": 1.3078,
+      "step": 3044
+    },
+    {
+      "epoch": 2.960123321569105,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.695141832833147e-09,
+      "loss": 1.3291,
+      "step": 3045
+    },
+    {
+      "epoch": 2.961096912904142,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.6022851875974189e-09,
+      "loss": 1.338,
+      "step": 3046
+    },
+    {
+      "epoch": 2.962070504239179,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5120431206391839e-09,
+      "loss": 1.3142,
+      "step": 3047
+    },
+    {
+      "epoch": 2.963044095574216,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4244157263967884e-09,
+      "loss": 1.3227,
+      "step": 3048
+    },
+    {
+      "epoch": 2.964017686909253,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.3394030965724335e-09,
+      "loss": 1.2864,
+      "step": 3049
+    },
+    {
+      "epoch": 2.96499127824429,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.2570053201318987e-09,
+      "loss": 1.3075,
+      "step": 3050
+    },
+    {
+      "epoch": 2.9659648695793273,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.177222483305096e-09,
+      "loss": 1.3226,
+      "step": 3051
+    },
+    {
+      "epoch": 2.9669384609143643,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.100054669584405e-09,
+      "loss": 1.3121,
+      "step": 3052
+    },
+    {
+      "epoch": 2.967912052249402,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0255019597266158e-09,
+      "loss": 1.3362,
+      "step": 3053
+    },
+    {
+      "epoch": 2.968885643584439,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.535644317507086e-10,
+      "loss": 1.3209,
+      "step": 3054
+    },
+    {
+      "epoch": 2.969859234919476,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.842421609397966e-10,
+      "loss": 1.3304,
+      "step": 3055
+    },
+    {
+      "epoch": 2.970832826254513,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.17535219839738e-10,
+      "loss": 1.3378,
+      "step": 3056
+    },
+    {
+      "epoch": 2.97180641758955,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.534436782594135e-10,
+      "loss": 1.3223,
+      "step": 3057
+    },
+    {
+      "epoch": 2.9727800089245875,
+      "grad_norm": 0.40625,
+      "learning_rate": 6.919676032710044e-10,
+      "loss": 1.327,
+      "step": 3058
+    },
+    {
+      "epoch": 2.9737536002596245,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.331070592088817e-10,
+      "loss": 1.3237,
+      "step": 3059
+    },
+    {
+      "epoch": 2.9747271915946616,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.768621076712721e-10,
+      "loss": 1.3338,
+      "step": 3060
+    },
+    {
+      "epoch": 2.9757007829296986,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.232328075180371e-10,
+      "loss": 1.3052,
+      "step": 3061
+    },
+    {
+      "epoch": 2.9766743742647357,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.722192148728932e-10,
+      "loss": 1.3246,
+      "step": 3062
+    },
+    {
+      "epoch": 2.9776479655997727,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.2382138312119237e-10,
+      "loss": 1.32,
+      "step": 3063
+    },
+    {
+      "epoch": 2.97862155693481,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.7803936291186395e-10,
+      "loss": 1.3072,
+      "step": 3064
+    },
+    {
+      "epoch": 2.979595148269847,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.3487320215547235e-10,
+      "loss": 1.3114,
+      "step": 3065
+    },
+    {
+      "epoch": 2.9805687396048843,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.9432294602560476e-10,
+      "loss": 1.313,
+      "step": 3066
+    },
+    {
+      "epoch": 2.9815423309399214,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.563886369583157e-10,
+      "loss": 1.3014,
+      "step": 3067
+    },
+    {
+      "epoch": 2.9825159222749584,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.2107031465185002e-10,
+      "loss": 1.3306,
+      "step": 3068
+    },
+    {
+      "epoch": 2.9834895136099955,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.8836801606664233e-10,
+      "loss": 1.3305,
+      "step": 3069
+    },
+    {
+      "epoch": 2.9844631049450325,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.5828177542642765e-10,
+      "loss": 1.318,
+      "step": 3070
+    },
+    {
+      "epoch": 2.98543669628007,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3081162421574312e-10,
+      "loss": 1.3064,
+      "step": 3071
+    },
+    {
+      "epoch": 2.986410287615107,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.059575911824262e-10,
+      "loss": 1.2982,
+      "step": 3072
+    },
+    {
+      "epoch": 2.987383878950144,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.37197023365044e-11,
+      "loss": 1.3013,
+      "step": 3073
+    },
+    {
+      "epoch": 2.988357470285181,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.40979809499176e-11,
+      "loss": 1.3008,
+      "step": 3074
+    },
+    {
+      "epoch": 2.989331061620218,
+      "grad_norm": 0.388671875,
+      "learning_rate": 4.7092447556518204e-11,
+      "loss": 1.3004,
+      "step": 3075
+    },
+    {
+      "epoch": 2.9903046529552553,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.270311995262621e-11,
+      "loss": 1.3092,
+      "step": 3076
+    },
+    {
+      "epoch": 2.9912782442902923,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.0930013197306697e-11,
+      "loss": 1.297,
+      "step": 3077
+    },
+    {
+      "epoch": 2.9922518356253294,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1773139610427031e-11,
+      "loss": 1.3032,
+      "step": 3078
+    },
+    {
+      "epoch": 2.993225426960367,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.232508775154799e-12,
+      "loss": 1.3081,
+      "step": 3079
+    },
+    {
+      "epoch": 2.994199018295404,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.308127536014947e-12,
+      "loss": 1.3196,
+      "step": 3080
+    },
+    {
+      "epoch": 2.995172609630441,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0,
+      "loss": 1.3089,
+      "step": 3081
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3081,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1027,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.829516122582994e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}