diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4766 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 0,
+  "global_step": 676,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014792899408284023,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1e-05,
+      "loss": 2.4231,
+      "step": 1
+    },
+    {
+      "epoch": 0.0029585798816568047,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.985207100591717e-06,
+      "loss": 2.4086,
+      "step": 2
+    },
+    {
+      "epoch": 0.004437869822485207,
+      "grad_norm": 0.5,
+      "learning_rate": 9.970414201183432e-06,
+      "loss": 2.4096,
+      "step": 3
+    },
+    {
+      "epoch": 0.005917159763313609,
+      "grad_norm": 0.490234375,
+      "learning_rate": 9.95562130177515e-06,
+      "loss": 2.4641,
+      "step": 4
+    },
+    {
+      "epoch": 0.0073964497041420114,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.940828402366864e-06,
+      "loss": 2.3734,
+      "step": 5
+    },
+    {
+      "epoch": 0.008875739644970414,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.926035502958581e-06,
+      "loss": 2.3715,
+      "step": 6
+    },
+    {
+      "epoch": 0.010355029585798817,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.911242603550297e-06,
+      "loss": 2.4129,
+      "step": 7
+    },
+    {
+      "epoch": 0.011834319526627219,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.896449704142012e-06,
+      "loss": 2.3551,
+      "step": 8
+    },
+    {
+      "epoch": 0.013313609467455622,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.88165680473373e-06,
+      "loss": 2.3445,
+      "step": 9
+    },
+    {
+      "epoch": 0.014792899408284023,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.866863905325444e-06,
+      "loss": 2.2953,
+      "step": 10
+    },
+    {
+      "epoch": 0.016272189349112426,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.852071005917161e-06,
+      "loss": 2.3048,
+      "step": 11
+    },
+    {
+      "epoch": 0.01775147928994083,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.837278106508877e-06,
+      "loss": 2.2406,
+      "step": 12
+    },
+    {
+      "epoch": 0.019230769230769232,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.822485207100593e-06,
+      "loss": 2.2238,
+      "step": 13
+    },
+    {
+      "epoch": 0.020710059171597635,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.807692307692308e-06,
+      "loss": 2.1735,
+      "step": 14
+    },
+    {
+      "epoch": 0.022189349112426034,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.792899408284024e-06,
+      "loss": 2.163,
+      "step": 15
+    },
+    {
+      "epoch": 0.023668639053254437,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.778106508875741e-06,
+      "loss": 2.1651,
+      "step": 16
+    },
+    {
+      "epoch": 0.02514792899408284,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.763313609467457e-06,
+      "loss": 2.135,
+      "step": 17
+    },
+    {
+      "epoch": 0.026627218934911243,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.748520710059173e-06,
+      "loss": 2.102,
+      "step": 18
+    },
+    {
+      "epoch": 0.028106508875739646,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.733727810650888e-06,
+      "loss": 2.0616,
+      "step": 19
+    },
+    {
+      "epoch": 0.029585798816568046,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.718934911242604e-06,
+      "loss": 2.0527,
+      "step": 20
+    },
+    {
+      "epoch": 0.03106508875739645,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.70414201183432e-06,
+      "loss": 1.9993,
+      "step": 21
+    },
+    {
+      "epoch": 0.03254437869822485,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.689349112426036e-06,
+      "loss": 2.021,
+      "step": 22
+    },
+    {
+      "epoch": 0.034023668639053255,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.674556213017751e-06,
+      "loss": 1.9456,
+      "step": 23
+    },
+    {
+      "epoch": 0.03550295857988166,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.659763313609469e-06,
+      "loss": 1.9524,
+      "step": 24
+    },
+    {
+      "epoch": 0.03698224852071006,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.644970414201184e-06,
+      "loss": 1.9964,
+      "step": 25
+    },
+    {
+      "epoch": 0.038461538461538464,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.6301775147929e-06,
+      "loss": 1.9319,
+      "step": 26
+    },
+    {
+      "epoch": 0.03994082840236687,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 1.9021,
+      "step": 27
+    },
+    {
+      "epoch": 0.04142011834319527,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.600591715976331e-06,
+      "loss": 1.8616,
+      "step": 28
+    },
+    {
+      "epoch": 0.042899408284023666,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.585798816568049e-06,
+      "loss": 1.8125,
+      "step": 29
+    },
+    {
+      "epoch": 0.04437869822485207,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.571005917159763e-06,
+      "loss": 1.8167,
+      "step": 30
+    },
+    {
+      "epoch": 0.04585798816568047,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.55621301775148e-06,
+      "loss": 1.7553,
+      "step": 31
+    },
+    {
+      "epoch": 0.047337278106508875,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.541420118343196e-06,
+      "loss": 1.7764,
+      "step": 32
+    },
+    {
+      "epoch": 0.04881656804733728,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.526627218934912e-06,
+      "loss": 1.7717,
+      "step": 33
+    },
+    {
+      "epoch": 0.05029585798816568,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.511834319526629e-06,
+      "loss": 1.7104,
+      "step": 34
+    },
+    {
+      "epoch": 0.051775147928994084,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.497041420118343e-06,
+      "loss": 1.6314,
+      "step": 35
+    },
+    {
+      "epoch": 0.05325443786982249,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.48224852071006e-06,
+      "loss": 1.668,
+      "step": 36
+    },
+    {
+      "epoch": 0.05473372781065089,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.467455621301776e-06,
+      "loss": 1.6516,
+      "step": 37
+    },
+    {
+      "epoch": 0.05621301775147929,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.452662721893492e-06,
+      "loss": 1.6512,
+      "step": 38
+    },
+    {
+      "epoch": 0.057692307692307696,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.43786982248521e-06,
+      "loss": 1.6004,
+      "step": 39
+    },
+    {
+      "epoch": 0.05917159763313609,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.423076923076923e-06,
+      "loss": 1.5672,
+      "step": 40
+    },
+    {
+      "epoch": 0.060650887573964495,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.40828402366864e-06,
+      "loss": 1.5295,
+      "step": 41
+    },
+    {
+      "epoch": 0.0621301775147929,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.393491124260356e-06,
+      "loss": 1.5748,
+      "step": 42
+    },
+    {
+      "epoch": 0.06360946745562131,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.378698224852072e-06,
+      "loss": 1.55,
+      "step": 43
+    },
+    {
+      "epoch": 0.0650887573964497,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.363905325443788e-06,
+      "loss": 1.4825,
+      "step": 44
+    },
+    {
+      "epoch": 0.06656804733727811,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.349112426035503e-06,
+      "loss": 1.489,
+      "step": 45
+    },
+    {
+      "epoch": 0.06804733727810651,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.33431952662722e-06,
+      "loss": 1.4201,
+      "step": 46
+    },
+    {
+      "epoch": 0.0695266272189349,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.319526627218935e-06,
+      "loss": 1.4269,
+      "step": 47
+    },
+    {
+      "epoch": 0.07100591715976332,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.304733727810652e-06,
+      "loss": 1.4084,
+      "step": 48
+    },
+    {
+      "epoch": 0.07248520710059171,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.289940828402368e-06,
+      "loss": 1.3724,
+      "step": 49
+    },
+    {
+      "epoch": 0.07396449704142012,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.275147928994084e-06,
+      "loss": 1.4076,
+      "step": 50
+    },
+    {
+      "epoch": 0.07544378698224852,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.2603550295858e-06,
+      "loss": 1.3456,
+      "step": 51
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.245562130177515e-06,
+      "loss": 1.3309,
+      "step": 52
+    },
+    {
+      "epoch": 0.07840236686390532,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.3331,
+      "step": 53
+    },
+    {
+      "epoch": 0.07988165680473373,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.215976331360948e-06,
+      "loss": 1.3032,
+      "step": 54
+    },
+    {
+      "epoch": 0.08136094674556213,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.201183431952664e-06,
+      "loss": 1.2559,
+      "step": 55
+    },
+    {
+      "epoch": 0.08284023668639054,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.18639053254438e-06,
+      "loss": 1.289,
+      "step": 56
+    },
+    {
+      "epoch": 0.08431952662721894,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.171597633136095e-06,
+      "loss": 1.2803,
+      "step": 57
+    },
+    {
+      "epoch": 0.08579881656804733,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.15680473372781e-06,
+      "loss": 1.2438,
+      "step": 58
+    },
+    {
+      "epoch": 0.08727810650887574,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.142011834319528e-06,
+      "loss": 1.1926,
+      "step": 59
+    },
+    {
+      "epoch": 0.08875739644970414,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.127218934911244e-06,
+      "loss": 1.2031,
+      "step": 60
+    },
+    {
+      "epoch": 0.09023668639053255,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.11242603550296e-06,
+      "loss": 1.1792,
+      "step": 61
+    },
+    {
+      "epoch": 0.09171597633136094,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.097633136094675e-06,
+      "loss": 1.1981,
+      "step": 62
+    },
+    {
+      "epoch": 0.09319526627218935,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.082840236686391e-06,
+      "loss": 1.1847,
+      "step": 63
+    },
+    {
+      "epoch": 0.09467455621301775,
+      "grad_norm": 0.26953125,
+      "learning_rate": 9.068047337278107e-06,
+      "loss": 1.1543,
+      "step": 64
+    },
+    {
+      "epoch": 0.09615384615384616,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.053254437869822e-06,
+      "loss": 1.1336,
+      "step": 65
+    },
+    {
+      "epoch": 0.09763313609467456,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.03846153846154e-06,
+      "loss": 1.1526,
+      "step": 66
+    },
+    {
+      "epoch": 0.09911242603550297,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.023668639053255e-06,
+      "loss": 1.1002,
+      "step": 67
+    },
+    {
+      "epoch": 0.10059171597633136,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.008875739644971e-06,
+      "loss": 1.1317,
+      "step": 68
+    },
+    {
+      "epoch": 0.10207100591715976,
+      "grad_norm": 0.2890625,
+      "learning_rate": 8.994082840236687e-06,
+      "loss": 1.0843,
+      "step": 69
+    },
+    {
+      "epoch": 0.10355029585798817,
+      "grad_norm": 0.30078125,
+      "learning_rate": 8.979289940828403e-06,
+      "loss": 1.1229,
+      "step": 70
+    },
+    {
+      "epoch": 0.10502958579881656,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.96449704142012e-06,
+      "loss": 1.0765,
+      "step": 71
+    },
+    {
+      "epoch": 0.10650887573964497,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.949704142011834e-06,
+      "loss": 1.0386,
+      "step": 72
+    },
+    {
+      "epoch": 0.10798816568047337,
+      "grad_norm": 0.25,
+      "learning_rate": 8.934911242603551e-06,
+      "loss": 1.0286,
+      "step": 73
+    },
+    {
+      "epoch": 0.10946745562130178,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.920118343195267e-06,
+      "loss": 1.0265,
+      "step": 74
+    },
+    {
+      "epoch": 0.11094674556213018,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.905325443786983e-06,
+      "loss": 1.034,
+      "step": 75
+    },
+    {
+      "epoch": 0.11242603550295859,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.8905325443787e-06,
+      "loss": 1.0409,
+      "step": 76
+    },
+    {
+      "epoch": 0.11390532544378698,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.875739644970414e-06,
+      "loss": 0.9947,
+      "step": 77
+    },
+    {
+      "epoch": 0.11538461538461539,
+      "grad_norm": 0.2890625,
+      "learning_rate": 8.860946745562132e-06,
+      "loss": 1.0118,
+      "step": 78
+    },
+    {
+      "epoch": 0.11686390532544379,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 1.0052,
+      "step": 79
+    },
+    {
+      "epoch": 0.11834319526627218,
+      "grad_norm": 0.279296875,
+      "learning_rate": 8.831360946745563e-06,
+      "loss": 0.9843,
+      "step": 80
+    },
+    {
+      "epoch": 0.11982248520710059,
+      "grad_norm": 0.244140625,
+      "learning_rate": 8.816568047337279e-06,
+      "loss": 0.9775,
+      "step": 81
+    },
+    {
+      "epoch": 0.12130177514792899,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 8.801775147928994e-06,
+      "loss": 0.9824,
+      "step": 82
+    },
+    {
+      "epoch": 0.1227810650887574,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.786982248520712e-06,
+      "loss": 0.9429,
+      "step": 83
+    },
+    {
+      "epoch": 0.1242603550295858,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.772189349112427e-06,
+      "loss": 0.9227,
+      "step": 84
+    },
+    {
+      "epoch": 0.1257396449704142,
+      "grad_norm": 0.27734375,
+      "learning_rate": 8.757396449704143e-06,
+      "loss": 0.9233,
+      "step": 85
+    },
+    {
+      "epoch": 0.12721893491124261,
+      "grad_norm": 0.283203125,
+      "learning_rate": 8.742603550295859e-06,
+      "loss": 0.9505,
+      "step": 86
+    },
+    {
+      "epoch": 0.128698224852071,
+      "grad_norm": 0.314453125,
+      "learning_rate": 8.727810650887574e-06,
+      "loss": 0.9382,
+      "step": 87
+    },
+    {
+      "epoch": 0.1301775147928994,
+      "grad_norm": 0.318359375,
+      "learning_rate": 8.71301775147929e-06,
+      "loss": 0.9702,
+      "step": 88
+    },
+    {
+      "epoch": 0.13165680473372782,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.698224852071006e-06,
+      "loss": 0.9179,
+      "step": 89
+    },
+    {
+      "epoch": 0.13313609467455623,
+      "grad_norm": 0.3125,
+      "learning_rate": 8.683431952662723e-06,
+      "loss": 0.9399,
+      "step": 90
+    },
+    {
+      "epoch": 0.1346153846153846,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.668639053254439e-06,
+      "loss": 0.907,
+      "step": 91
+    },
+    {
+      "epoch": 0.13609467455621302,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.653846153846155e-06,
+      "loss": 0.9184,
+      "step": 92
+    },
+    {
+      "epoch": 0.13757396449704143,
+      "grad_norm": 0.24609375,
+      "learning_rate": 8.63905325443787e-06,
+      "loss": 0.9028,
+      "step": 93
+    },
+    {
+      "epoch": 0.1390532544378698,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.624260355029586e-06,
+      "loss": 0.9114,
+      "step": 94
+    },
+    {
+      "epoch": 0.14053254437869822,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.609467455621302e-06,
+      "loss": 0.8842,
+      "step": 95
+    },
+    {
+      "epoch": 0.14201183431952663,
+      "grad_norm": 0.287109375,
+      "learning_rate": 8.594674556213019e-06,
+      "loss": 0.9058,
+      "step": 96
+    },
+    {
+      "epoch": 0.14349112426035504,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.579881656804735e-06,
+      "loss": 0.8886,
+      "step": 97
+    },
+    {
+      "epoch": 0.14497041420118342,
+      "grad_norm": 0.25,
+      "learning_rate": 8.56508875739645e-06,
+      "loss": 0.8919,
+      "step": 98
+    },
+    {
+      "epoch": 0.14644970414201183,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.550295857988166e-06,
+      "loss": 0.9077,
+      "step": 99
+    },
+    {
+      "epoch": 0.14792899408284024,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.535502958579882e-06,
+      "loss": 0.864,
+      "step": 100
+    },
+    {
+      "epoch": 0.14940828402366865,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.5207100591716e-06,
+      "loss": 0.852,
+      "step": 101
+    },
+    {
+      "epoch": 0.15088757396449703,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.505917159763313e-06,
+      "loss": 0.8532,
+      "step": 102
+    },
+    {
+      "epoch": 0.15236686390532544,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.49112426035503e-06,
+      "loss": 0.8782,
+      "step": 103
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.476331360946746e-06,
+      "loss": 0.8341,
+      "step": 104
+    },
+    {
+      "epoch": 0.15532544378698224,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.8387,
+      "step": 105
+    },
+    {
+      "epoch": 0.15680473372781065,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.446745562130178e-06,
+      "loss": 0.8476,
+      "step": 106
+    },
+    {
+      "epoch": 0.15828402366863906,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.431952662721893e-06,
+      "loss": 0.8373,
+      "step": 107
+    },
+    {
+      "epoch": 0.15976331360946747,
+      "grad_norm": 0.287109375,
+      "learning_rate": 8.417159763313611e-06,
+      "loss": 0.85,
+      "step": 108
+    },
+    {
+      "epoch": 0.16124260355029585,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.402366863905327e-06,
+      "loss": 0.8027,
+      "step": 109
+    },
+    {
+      "epoch": 0.16272189349112426,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.387573964497042e-06,
+      "loss": 0.8286,
+      "step": 110
+    },
+    {
+      "epoch": 0.16420118343195267,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.372781065088758e-06,
+      "loss": 0.8317,
+      "step": 111
+    },
+    {
+      "epoch": 0.16568047337278108,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.357988165680474e-06,
+      "loss": 0.8442,
+      "step": 112
+    },
+    {
+      "epoch": 0.16715976331360946,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 8.343195266272191e-06,
+      "loss": 0.8062,
+      "step": 113
+    },
+    {
+      "epoch": 0.16863905325443787,
+      "grad_norm": 0.259765625,
+      "learning_rate": 8.328402366863905e-06,
+      "loss": 0.7936,
+      "step": 114
+    },
+    {
+      "epoch": 0.17011834319526628,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.313609467455622e-06,
+      "loss": 0.7832,
+      "step": 115
+    },
+    {
+      "epoch": 0.17159763313609466,
+      "grad_norm": 0.294921875,
+      "learning_rate": 8.298816568047338e-06,
+      "loss": 0.8205,
+      "step": 116
+    },
+    {
+      "epoch": 0.17307692307692307,
+      "grad_norm": 0.287109375,
+      "learning_rate": 8.284023668639054e-06,
+      "loss": 0.8161,
+      "step": 117
+    },
+    {
+      "epoch": 0.17455621301775148,
+      "grad_norm": 0.26171875,
+      "learning_rate": 8.26923076923077e-06,
+      "loss": 0.7978,
+      "step": 118
+    },
+    {
+      "epoch": 0.1760355029585799,
+      "grad_norm": 0.365234375,
+      "learning_rate": 8.254437869822485e-06,
+      "loss": 0.8196,
+      "step": 119
+    },
+    {
+      "epoch": 0.17751479289940827,
+      "grad_norm": 0.298828125,
+      "learning_rate": 8.239644970414203e-06,
+      "loss": 0.8038,
+      "step": 120
+    },
+    {
+      "epoch": 0.17899408284023668,
+      "grad_norm": 0.361328125,
+      "learning_rate": 8.224852071005918e-06,
+      "loss": 0.8437,
+      "step": 121
+    },
+    {
+      "epoch": 0.1804733727810651,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.210059171597634e-06,
+      "loss": 0.8056,
+      "step": 122
+    },
+    {
+      "epoch": 0.1819526627218935,
+      "grad_norm": 0.357421875,
+      "learning_rate": 8.19526627218935e-06,
+      "loss": 0.7825,
+      "step": 123
+    },
+    {
+      "epoch": 0.1834319526627219,
+      "grad_norm": 0.31640625,
+      "learning_rate": 8.180473372781065e-06,
+      "loss": 0.804,
+      "step": 124
+    },
+    {
+      "epoch": 0.1849112426035503,
+      "grad_norm": 0.31640625,
+      "learning_rate": 8.165680473372781e-06,
+      "loss": 0.8087,
+      "step": 125
+    },
+    {
+      "epoch": 0.1863905325443787,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.150887573964499e-06,
+      "loss": 0.7488,
+      "step": 126
+    },
+    {
+      "epoch": 0.1878698224852071,
+      "grad_norm": 0.36328125,
+      "learning_rate": 8.136094674556214e-06,
+      "loss": 0.7905,
+      "step": 127
+    },
+    {
+      "epoch": 0.1893491124260355,
+      "grad_norm": 0.30078125,
+      "learning_rate": 8.12130177514793e-06,
+      "loss": 0.7792,
+      "step": 128
+    },
+    {
+      "epoch": 0.1908284023668639,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.106508875739646e-06,
+      "loss": 0.7933,
+      "step": 129
+    },
+    {
+      "epoch": 0.19230769230769232,
+      "grad_norm": 0.2890625,
+      "learning_rate": 8.091715976331361e-06,
+      "loss": 0.74,
+      "step": 130
+    },
+    {
+      "epoch": 0.1937869822485207,
+      "grad_norm": 0.349609375,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 0.771,
+      "step": 131
+    },
+    {
+      "epoch": 0.1952662721893491,
+      "grad_norm": 0.302734375,
+      "learning_rate": 8.062130177514793e-06,
+      "loss": 0.7907,
+      "step": 132
+    },
+    {
+      "epoch": 0.19674556213017752,
+      "grad_norm": 0.296875,
+      "learning_rate": 8.04733727810651e-06,
+      "loss": 0.758,
+      "step": 133
+    },
+    {
+      "epoch": 0.19822485207100593,
+      "grad_norm": 0.298828125,
+      "learning_rate": 8.032544378698226e-06,
+      "loss": 0.7607,
+      "step": 134
+    },
+    {
+      "epoch": 0.1997041420118343,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.017751479289941e-06,
+      "loss": 0.7552,
+      "step": 135
+    },
+    {
+      "epoch": 0.20118343195266272,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.002958579881657e-06,
+      "loss": 0.73,
+      "step": 136
+    },
+    {
+      "epoch": 0.20266272189349113,
+      "grad_norm": 0.26953125,
+      "learning_rate": 7.988165680473373e-06,
+      "loss": 0.7354,
+      "step": 137
+    },
+    {
+      "epoch": 0.20414201183431951,
+      "grad_norm": 0.2890625,
+      "learning_rate": 7.97337278106509e-06,
+      "loss": 0.7569,
+      "step": 138
+    },
+    {
+      "epoch": 0.20562130177514792,
+      "grad_norm": 0.2890625,
+      "learning_rate": 7.958579881656804e-06,
+      "loss": 0.736,
+      "step": 139
+    },
+    {
+      "epoch": 0.20710059171597633,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.943786982248522e-06,
+      "loss": 0.7334,
+      "step": 140
+    },
+    {
+      "epoch": 0.20857988165680474,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.928994082840237e-06,
+      "loss": 0.7484,
+      "step": 141
+    },
+    {
+      "epoch": 0.21005917159763313,
+      "grad_norm": 0.279296875,
+      "learning_rate": 7.914201183431953e-06,
+      "loss": 0.7334,
+      "step": 142
+    },
+    {
+      "epoch": 0.21153846153846154,
+      "grad_norm": 0.279296875,
+      "learning_rate": 7.89940828402367e-06,
+      "loss": 0.7489,
+      "step": 143
+    },
+    {
+      "epoch": 0.21301775147928995,
+      "grad_norm": 0.26953125,
+      "learning_rate": 7.884615384615384e-06,
+      "loss": 0.7366,
+      "step": 144
+    },
+    {
+      "epoch": 0.21449704142011836,
+      "grad_norm": 0.29296875,
+      "learning_rate": 7.869822485207102e-06,
+      "loss": 0.738,
+      "step": 145
+    },
+    {
+      "epoch": 0.21597633136094674,
+      "grad_norm": 0.26953125,
+      "learning_rate": 7.855029585798818e-06,
+      "loss": 0.7493,
+      "step": 146
+    },
+    {
+      "epoch": 0.21745562130177515,
+      "grad_norm": 0.306640625,
+      "learning_rate": 7.840236686390533e-06,
+      "loss": 0.728,
+      "step": 147
+    },
+    {
+      "epoch": 0.21893491124260356,
+      "grad_norm": 0.283203125,
+      "learning_rate": 7.825443786982249e-06,
+      "loss": 0.7144,
+      "step": 148
+    },
+    {
+      "epoch": 0.22041420118343194,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.810650887573965e-06,
+      "loss": 0.7199,
+      "step": 149
+    },
+    {
+      "epoch": 0.22189349112426035,
+      "grad_norm": 0.279296875,
+      "learning_rate": 7.795857988165682e-06,
+      "loss": 0.7147,
+      "step": 150
+    },
+    {
+      "epoch": 0.22337278106508876,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.781065088757396e-06,
+      "loss": 0.7019,
+      "step": 151
+    },
+    {
+      "epoch": 0.22485207100591717,
+      "grad_norm": 0.3203125,
+      "learning_rate": 7.766272189349113e-06,
+      "loss": 0.7455,
+      "step": 152
+    },
+    {
+      "epoch": 0.22633136094674555,
+      "grad_norm": 0.298828125,
+      "learning_rate": 7.751479289940829e-06,
+      "loss": 0.7111,
+      "step": 153
+    },
+    {
+      "epoch": 0.22781065088757396,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.736686390532545e-06,
+      "loss": 0.7537,
+      "step": 154
+    },
+    {
+      "epoch": 0.22928994082840237,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.72189349112426e-06,
+      "loss": 0.7252,
+      "step": 155
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.3515625,
+      "learning_rate": 7.707100591715976e-06,
+      "loss": 0.7351,
+      "step": 156
+    },
+    {
+      "epoch": 0.23224852071005916,
+      "grad_norm": 0.30859375,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 0.7266,
+      "step": 157
+    },
+    {
+      "epoch": 0.23372781065088757,
+      "grad_norm": 0.28125,
+      "learning_rate": 7.67751479289941e-06,
+      "loss": 0.7077,
+      "step": 158
+    },
+    {
+      "epoch": 0.23520710059171598,
+      "grad_norm": 0.30859375,
+      "learning_rate": 7.662721893491125e-06,
+      "loss": 0.7025,
+      "step": 159
+    },
+    {
+      "epoch": 0.23668639053254437,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.64792899408284e-06,
+      "loss": 0.7028,
+      "step": 160
+    },
+    {
+      "epoch": 0.23816568047337278,
+      "grad_norm": 0.3046875,
+      "learning_rate": 7.633136094674556e-06,
+      "loss": 0.7053,
+      "step": 161
+    },
+    {
+      "epoch": 0.23964497041420119,
+      "grad_norm": 0.318359375,
+      "learning_rate": 7.618343195266272e-06,
+      "loss": 0.6992,
+      "step": 162
+    },
+    {
+      "epoch": 0.2411242603550296,
+      "grad_norm": 0.318359375,
+      "learning_rate": 7.603550295857989e-06,
+      "loss": 0.6892,
+      "step": 163
+    },
+    {
+      "epoch": 0.24260355029585798,
+      "grad_norm": 0.384765625,
+      "learning_rate": 7.588757396449705e-06,
+      "loss": 0.7409,
+      "step": 164
+    },
+    {
+      "epoch": 0.2440828402366864,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.573964497041421e-06,
+      "loss": 0.6857,
+      "step": 165
+    },
+    {
+      "epoch": 0.2455621301775148,
+      "grad_norm": 0.294921875,
+      "learning_rate": 7.559171597633137e-06,
+      "loss": 0.6914,
+      "step": 166
+    },
+    {
+      "epoch": 0.2470414201183432,
+      "grad_norm": 0.283203125,
+      "learning_rate": 7.544378698224852e-06,
+      "loss": 0.7089,
+      "step": 167
+    },
+    {
+      "epoch": 0.2485207100591716,
+      "grad_norm": 0.33984375,
+      "learning_rate": 7.529585798816569e-06,
+      "loss": 0.6887,
+      "step": 168
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4296875,
+      "learning_rate": 7.5147928994082845e-06,
+      "loss": 0.6972,
+      "step": 169
+    },
+    {
+      "epoch": 0.2514792899408284,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.6865,
+      "step": 170
+    },
+    {
+      "epoch": 0.2529585798816568,
+      "grad_norm": 0.306640625,
+      "learning_rate": 7.485207100591717e-06,
+      "loss": 0.6829,
+      "step": 171
+    },
+    {
+      "epoch": 0.25443786982248523,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.4704142011834324e-06,
+      "loss": 0.7128,
+      "step": 172
+    },
+    {
+      "epoch": 0.2559171597633136,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.455621301775149e-06,
+      "loss": 0.7017,
+      "step": 173
+    },
+    {
+      "epoch": 0.257396449704142,
+      "grad_norm": 0.310546875,
+      "learning_rate": 7.440828402366864e-06,
+      "loss": 0.6775,
+      "step": 174
+    },
+    {
+      "epoch": 0.2588757396449704,
+      "grad_norm": 0.29296875,
+      "learning_rate": 7.42603550295858e-06,
+      "loss": 0.6938,
+      "step": 175
+    },
+    {
+      "epoch": 0.2603550295857988,
+      "grad_norm": 0.30859375,
+      "learning_rate": 7.411242603550296e-06,
+      "loss": 0.6828,
+      "step": 176
+    },
+    {
+      "epoch": 0.2618343195266272,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.396449704142013e-06,
+      "loss": 0.65,
+      "step": 177
+    },
+    {
+      "epoch": 0.26331360946745563,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.381656804733729e-06,
+      "loss": 0.6895,
+      "step": 178
+    },
+    {
+      "epoch": 0.26479289940828404,
+      "grad_norm": 0.3046875,
+      "learning_rate": 7.366863905325444e-06,
+      "loss": 0.6829,
+      "step": 179
+    },
+    {
+      "epoch": 0.26627218934911245,
+      "grad_norm": 0.30078125,
+      "learning_rate": 7.3520710059171605e-06,
+      "loss": 0.7053,
+      "step": 180
+    },
+    {
+      "epoch": 0.2677514792899408,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.337278106508876e-06,
+      "loss": 0.6982,
+      "step": 181
+    },
+    {
+      "epoch": 0.2692307692307692,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.322485207100593e-06,
+      "loss": 0.6965,
+      "step": 182
+    },
+    {
+      "epoch": 0.27071005917159763,
+      "grad_norm": 0.318359375,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 0.6846,
+      "step": 183
+    },
+    {
+      "epoch": 0.27218934911242604,
+      "grad_norm": 0.30078125,
+      "learning_rate": 7.292899408284024e-06,
+      "loss": 0.6661,
+      "step": 184
+    },
+    {
+      "epoch": 0.27366863905325445,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.278106508875741e-06,
+      "loss": 0.6783,
+      "step": 185
+    },
+    {
+      "epoch": 0.27514792899408286,
+      "grad_norm": 0.322265625,
+      "learning_rate": 7.263313609467456e-06,
+      "loss": 0.6838,
+      "step": 186
+    },
+    {
+      "epoch": 0.27662721893491127,
+      "grad_norm": 0.310546875,
+      "learning_rate": 7.248520710059173e-06,
+      "loss": 0.6933,
+      "step": 187
+    },
+    {
+      "epoch": 0.2781065088757396,
+      "grad_norm": 0.30078125,
+      "learning_rate": 7.233727810650888e-06,
+      "loss": 0.6857,
+      "step": 188
+    },
+    {
+      "epoch": 0.27958579881656803,
+      "grad_norm": 0.291015625,
+      "learning_rate": 7.218934911242604e-06,
+      "loss": 0.6602,
+      "step": 189
+    },
+    {
+      "epoch": 0.28106508875739644,
+      "grad_norm": 0.298828125,
+      "learning_rate": 7.20414201183432e-06,
+      "loss": 0.6678,
+      "step": 190
+    },
+    {
+      "epoch": 0.28254437869822485,
+      "grad_norm": 0.3125,
+      "learning_rate": 7.189349112426036e-06,
+      "loss": 0.6593,
+      "step": 191
+    },
+    {
+      "epoch": 0.28402366863905326,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.1745562130177515e-06,
+      "loss": 0.6704,
+      "step": 192
+    },
+    {
+      "epoch": 0.28550295857988167,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.159763313609468e-06,
+      "loss": 0.6192,
+      "step": 193
+    },
+    {
+      "epoch": 0.2869822485207101,
+      "grad_norm": 0.3046875,
+      "learning_rate": 7.1449704142011845e-06,
+      "loss": 0.6726,
+      "step": 194
+    },
+    {
+      "epoch": 0.28846153846153844,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.130177514792899e-06,
+      "loss": 0.6707,
+      "step": 195
+    },
+    {
+      "epoch": 0.28994082840236685,
+      "grad_norm": 0.357421875,
+      "learning_rate": 7.115384615384616e-06,
+      "loss": 0.6881,
+      "step": 196
+    },
+    {
+      "epoch": 0.29142011834319526,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.100591715976332e-06,
+      "loss": 0.6747,
+      "step": 197
+    },
+    {
+      "epoch": 0.29289940828402367,
+      "grad_norm": 0.322265625,
+      "learning_rate": 7.085798816568048e-06,
+      "loss": 0.6782,
+      "step": 198
+    },
+    {
+      "epoch": 0.2943786982248521,
+      "grad_norm": 0.29296875,
+      "learning_rate": 7.071005917159763e-06,
+      "loss": 0.6684,
+      "step": 199
+    },
+    {
+      "epoch": 0.2958579881656805,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.0562130177514796e-06,
+      "loss": 0.6706,
+      "step": 200
+    },
+    {
+      "epoch": 0.2973372781065089,
+      "grad_norm": 0.328125,
+      "learning_rate": 7.041420118343196e-06,
+      "loss": 0.6579,
+      "step": 201
+    },
+    {
+      "epoch": 0.2988165680473373,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.026627218934912e-06,
+      "loss": 0.6951,
+      "step": 202
+    },
+    {
+      "epoch": 0.30029585798816566,
+      "grad_norm": 0.32421875,
+      "learning_rate": 7.011834319526628e-06,
+      "loss": 0.6768,
+      "step": 203
+    },
+    {
+      "epoch": 0.30177514792899407,
+      "grad_norm": 0.37109375,
+      "learning_rate": 6.997041420118343e-06,
+      "loss": 0.6589,
+      "step": 204
+    },
+    {
+      "epoch": 0.3032544378698225,
+      "grad_norm": 0.37890625,
+      "learning_rate": 6.98224852071006e-06,
+      "loss": 0.6396,
+      "step": 205
+    },
+    {
+      "epoch": 0.3047337278106509,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.9674556213017754e-06,
+      "loss": 0.6342,
+      "step": 206
+    },
+    {
+      "epoch": 0.3062130177514793,
+      "grad_norm": 0.322265625,
+      "learning_rate": 6.952662721893492e-06,
+      "loss": 0.6468,
+      "step": 207
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.9378698224852085e-06,
+      "loss": 0.6714,
+      "step": 208
+    },
+    {
+      "epoch": 0.3091715976331361,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 0.6485,
+      "step": 209
+    },
+    {
+      "epoch": 0.3106508875739645,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.90828402366864e-06,
+      "loss": 0.6593,
+      "step": 210
+    },
+    {
+      "epoch": 0.3121301775147929,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.893491124260356e-06,
+      "loss": 0.6596,
+      "step": 211
+    },
+    {
+      "epoch": 0.3136094674556213,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.878698224852071e-06,
+      "loss": 0.6367,
+      "step": 212
+    },
+    {
+      "epoch": 0.3150887573964497,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.863905325443787e-06,
+      "loss": 0.6649,
+      "step": 213
+    },
+    {
+      "epoch": 0.3165680473372781,
+      "grad_norm": 0.32421875,
+      "learning_rate": 6.8491124260355036e-06,
+      "loss": 0.673,
+      "step": 214
+    },
+    {
+      "epoch": 0.3180473372781065,
+      "grad_norm": 0.310546875,
+      "learning_rate": 6.83431952662722e-06,
+      "loss": 0.6336,
+      "step": 215
+    },
+    {
+      "epoch": 0.31952662721893493,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.819526627218935e-06,
+      "loss": 0.642,
+      "step": 216
+    },
+    {
+      "epoch": 0.3210059171597633,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.8047337278106515e-06,
+      "loss": 0.6383,
+      "step": 217
+    },
+    {
+      "epoch": 0.3224852071005917,
+      "grad_norm": 0.3203125,
+      "learning_rate": 6.789940828402367e-06,
+      "loss": 0.6429,
+      "step": 218
+    },
+    {
+      "epoch": 0.3239644970414201,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.775147928994084e-06,
+      "loss": 0.6526,
+      "step": 219
+    },
+    {
+      "epoch": 0.3254437869822485,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.760355029585799e-06,
+      "loss": 0.6615,
+      "step": 220
+    },
+    {
+      "epoch": 0.3269230769230769,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.745562130177515e-06,
+      "loss": 0.6427,
+      "step": 221
+    },
+    {
+      "epoch": 0.32840236686390534,
+      "grad_norm": 0.3203125,
+      "learning_rate": 6.730769230769232e-06,
+      "loss": 0.6501,
+      "step": 222
+    },
+    {
+      "epoch": 0.32988165680473375,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.715976331360947e-06,
+      "loss": 0.6404,
+      "step": 223
+    },
+    {
+      "epoch": 0.33136094674556216,
+      "grad_norm": 0.318359375,
+      "learning_rate": 6.701183431952664e-06,
+      "loss": 0.6448,
+      "step": 224
+    },
+    {
+      "epoch": 0.3328402366863905,
+      "grad_norm": 0.30859375,
+      "learning_rate": 6.686390532544379e-06,
+      "loss": 0.6687,
+      "step": 225
+    },
+    {
+      "epoch": 0.3343195266272189,
+      "grad_norm": 0.302734375,
+      "learning_rate": 6.671597633136095e-06,
+      "loss": 0.6425,
+      "step": 226
+    },
+    {
+      "epoch": 0.33579881656804733,
+      "grad_norm": 0.296875,
+      "learning_rate": 6.656804733727811e-06,
+      "loss": 0.6159,
+      "step": 227
+    },
+    {
+      "epoch": 0.33727810650887574,
+      "grad_norm": 0.310546875,
+      "learning_rate": 6.6420118343195276e-06,
+      "loss": 0.6343,
+      "step": 228
+    },
+    {
+      "epoch": 0.33875739644970415,
+      "grad_norm": 0.3125,
+      "learning_rate": 6.627218934911244e-06,
+      "loss": 0.6225,
+      "step": 229
+    },
+    {
+      "epoch": 0.34023668639053256,
+      "grad_norm": 0.3125,
+      "learning_rate": 6.612426035502959e-06,
+      "loss": 0.6291,
+      "step": 230
+    },
+    {
+      "epoch": 0.34171597633136097,
+      "grad_norm": 0.3125,
+      "learning_rate": 6.5976331360946755e-06,
+      "loss": 0.6427,
+      "step": 231
+    },
+    {
+      "epoch": 0.3431952662721893,
+      "grad_norm": 0.298828125,
+      "learning_rate": 6.582840236686391e-06,
+      "loss": 0.648,
+      "step": 232
+    },
+    {
+      "epoch": 0.34467455621301774,
+      "grad_norm": 0.29296875,
+      "learning_rate": 6.568047337278107e-06,
+      "loss": 0.6437,
+      "step": 233
+    },
+    {
+      "epoch": 0.34615384615384615,
+      "grad_norm": 0.322265625,
+      "learning_rate": 6.553254437869823e-06,
+      "loss": 0.6264,
+      "step": 234
+    },
+    {
+      "epoch": 0.34763313609467456,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 0.637,
+      "step": 235
+    },
+    {
+      "epoch": 0.34911242603550297,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.523668639053255e-06,
+      "loss": 0.6304,
+      "step": 236
+    },
+    {
+      "epoch": 0.3505917159763314,
+      "grad_norm": 0.310546875,
+      "learning_rate": 6.5088757396449705e-06,
+      "loss": 0.6199,
+      "step": 237
+    },
+    {
+      "epoch": 0.3520710059171598,
+      "grad_norm": 0.30078125,
+      "learning_rate": 6.494082840236687e-06,
+      "loss": 0.6256,
+      "step": 238
+    },
+    {
+      "epoch": 0.35355029585798814,
+      "grad_norm": 0.30078125,
+      "learning_rate": 6.479289940828403e-06,
+      "loss": 0.6254,
+      "step": 239
+    },
+    {
+      "epoch": 0.35502958579881655,
+      "grad_norm": 0.37109375,
+      "learning_rate": 6.464497041420119e-06,
+      "loss": 0.6292,
+      "step": 240
+    },
+    {
+      "epoch": 0.35650887573964496,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.449704142011834e-06,
+      "loss": 0.637,
+      "step": 241
+    },
+    {
+      "epoch": 0.35798816568047337,
+      "grad_norm": 0.3046875,
+      "learning_rate": 6.434911242603551e-06,
+      "loss": 0.6202,
+      "step": 242
+    },
+    {
+      "epoch": 0.3594674556213018,
+      "grad_norm": 0.30859375,
+      "learning_rate": 6.420118343195266e-06,
+      "loss": 0.6164,
+      "step": 243
+    },
+    {
+      "epoch": 0.3609467455621302,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.405325443786983e-06,
+      "loss": 0.6334,
+      "step": 244
+    },
+    {
+      "epoch": 0.3624260355029586,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.3905325443786995e-06,
+      "loss": 0.6492,
+      "step": 245
+    },
+    {
+      "epoch": 0.363905325443787,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.375739644970414e-06,
+      "loss": 0.643,
+      "step": 246
+    },
+    {
+      "epoch": 0.36538461538461536,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.360946745562131e-06,
+      "loss": 0.6303,
+      "step": 247
+    },
+    {
+      "epoch": 0.3668639053254438,
+      "grad_norm": 0.345703125,
+      "learning_rate": 6.3461538461538466e-06,
+      "loss": 0.6305,
+      "step": 248
+    },
+    {
+      "epoch": 0.3683431952662722,
+      "grad_norm": 0.32421875,
+      "learning_rate": 6.331360946745563e-06,
+      "loss": 0.6165,
+      "step": 249
+    },
+    {
+      "epoch": 0.3698224852071006,
+      "grad_norm": 0.31640625,
+      "learning_rate": 6.316568047337278e-06,
+      "loss": 0.6329,
+      "step": 250
+    },
+    {
+      "epoch": 0.371301775147929,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.3017751479289945e-06,
+      "loss": 0.6441,
+      "step": 251
+    },
+    {
+      "epoch": 0.3727810650887574,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.286982248520711e-06,
+      "loss": 0.6525,
+      "step": 252
+    },
+    {
+      "epoch": 0.3742603550295858,
+      "grad_norm": 0.294921875,
+      "learning_rate": 6.272189349112427e-06,
+      "loss": 0.6411,
+      "step": 253
+    },
+    {
+      "epoch": 0.3757396449704142,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.2573964497041425e-06,
+      "loss": 0.6141,
+      "step": 254
+    },
+    {
+      "epoch": 0.3772189349112426,
+      "grad_norm": 0.318359375,
+      "learning_rate": 6.242603550295858e-06,
+      "loss": 0.6489,
+      "step": 255
+    },
+    {
+      "epoch": 0.378698224852071,
+      "grad_norm": 0.3359375,
+      "learning_rate": 6.227810650887575e-06,
+      "loss": 0.5926,
+      "step": 256
+    },
+    {
+      "epoch": 0.3801775147928994,
+      "grad_norm": 0.302734375,
+      "learning_rate": 6.21301775147929e-06,
+      "loss": 0.6306,
+      "step": 257
+    },
+    {
+      "epoch": 0.3816568047337278,
+      "grad_norm": 0.3125,
+      "learning_rate": 6.198224852071006e-06,
+      "loss": 0.6222,
+      "step": 258
+    },
+    {
+      "epoch": 0.3831360946745562,
+      "grad_norm": 0.30859375,
+      "learning_rate": 6.183431952662723e-06,
+      "loss": 0.6073,
+      "step": 259
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.168639053254438e-06,
+      "loss": 0.6484,
+      "step": 260
+    },
+    {
+      "epoch": 0.386094674556213,
+      "grad_norm": 0.31640625,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 0.6061,
+      "step": 261
+    },
+    {
+      "epoch": 0.3875739644970414,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.13905325443787e-06,
+      "loss": 0.6175,
+      "step": 262
+    },
+    {
+      "epoch": 0.3890532544378698,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.124260355029586e-06,
+      "loss": 0.6283,
+      "step": 263
+    },
+    {
+      "epoch": 0.3905325443786982,
+      "grad_norm": 0.34375,
+      "learning_rate": 6.109467455621302e-06,
+      "loss": 0.6029,
+      "step": 264
+    },
+    {
+      "epoch": 0.39201183431952663,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.0946745562130185e-06,
+      "loss": 0.624,
+      "step": 265
+    },
+    {
+      "epoch": 0.39349112426035504,
+      "grad_norm": 0.373046875,
+      "learning_rate": 6.079881656804735e-06,
+      "loss": 0.6114,
+      "step": 266
+    },
+    {
+      "epoch": 0.39497041420118345,
+      "grad_norm": 0.365234375,
+      "learning_rate": 6.06508875739645e-06,
+      "loss": 0.6175,
+      "step": 267
+    },
+    {
+      "epoch": 0.39644970414201186,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.0502958579881664e-06,
+      "loss": 0.617,
+      "step": 268
+    },
+    {
+      "epoch": 0.3979289940828402,
+      "grad_norm": 0.3046875,
+      "learning_rate": 6.035502958579882e-06,
+      "loss": 0.6179,
+      "step": 269
+    },
+    {
+      "epoch": 0.3994082840236686,
+      "grad_norm": 0.341796875,
+      "learning_rate": 6.020710059171599e-06,
+      "loss": 0.6236,
+      "step": 270
+    },
+    {
+      "epoch": 0.40088757396449703,
+      "grad_norm": 0.310546875,
+      "learning_rate": 6.0059171597633135e-06,
+      "loss": 0.6317,
+      "step": 271
+    },
+    {
+      "epoch": 0.40236686390532544,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.99112426035503e-06,
+      "loss": 0.6417,
+      "step": 272
+    },
+    {
+      "epoch": 0.40384615384615385,
+      "grad_norm": 0.322265625,
+      "learning_rate": 5.976331360946747e-06,
+      "loss": 0.6208,
+      "step": 273
+    },
+    {
+      "epoch": 0.40532544378698226,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.961538461538462e-06,
+      "loss": 0.6333,
+      "step": 274
+    },
+    {
+      "epoch": 0.4068047337278107,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.946745562130178e-06,
+      "loss": 0.6247,
+      "step": 275
+    },
+    {
+      "epoch": 0.40828402366863903,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.931952662721894e-06,
+      "loss": 0.6073,
+      "step": 276
+    },
+    {
+      "epoch": 0.40976331360946744,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.91715976331361e-06,
+      "loss": 0.5878,
+      "step": 277
+    },
+    {
+      "epoch": 0.41124260355029585,
+      "grad_norm": 0.361328125,
+      "learning_rate": 5.902366863905326e-06,
+      "loss": 0.6207,
+      "step": 278
+    },
+    {
+      "epoch": 0.41272189349112426,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.887573964497042e-06,
+      "loss": 0.607,
+      "step": 279
+    },
+    {
+      "epoch": 0.41420118343195267,
+      "grad_norm": 0.322265625,
+      "learning_rate": 5.872781065088757e-06,
+      "loss": 0.6165,
+      "step": 280
+    },
+    {
+      "epoch": 0.4156804733727811,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.857988165680474e-06,
+      "loss": 0.6268,
+      "step": 281
+    },
+    {
+      "epoch": 0.4171597633136095,
+      "grad_norm": 0.37109375,
+      "learning_rate": 5.8431952662721904e-06,
+      "loss": 0.6327,
+      "step": 282
+    },
+    {
+      "epoch": 0.41863905325443784,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.828402366863905e-06,
+      "loss": 0.6148,
+      "step": 283
+    },
+    {
+      "epoch": 0.42011834319526625,
+      "grad_norm": 0.369140625,
+      "learning_rate": 5.813609467455622e-06,
+      "loss": 0.6173,
+      "step": 284
+    },
+    {
+      "epoch": 0.42159763313609466,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.7988165680473375e-06,
+      "loss": 0.6027,
+      "step": 285
+    },
+    {
+      "epoch": 0.4230769230769231,
+      "grad_norm": 0.369140625,
+      "learning_rate": 5.784023668639054e-06,
+      "loss": 0.6129,
+      "step": 286
+    },
+    {
+      "epoch": 0.4245562130177515,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 0.6279,
+      "step": 287
+    },
+    {
+      "epoch": 0.4260355029585799,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.7544378698224855e-06,
+      "loss": 0.5993,
+      "step": 288
+    },
+    {
+      "epoch": 0.4275147928994083,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.739644970414202e-06,
+      "loss": 0.5982,
+      "step": 289
+    },
+    {
+      "epoch": 0.4289940828402367,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.724852071005918e-06,
+      "loss": 0.6081,
+      "step": 290
+    },
+    {
+      "epoch": 0.43047337278106507,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.710059171597634e-06,
+      "loss": 0.5879,
+      "step": 291
+    },
+    {
+      "epoch": 0.4319526627218935,
+      "grad_norm": 0.376953125,
+      "learning_rate": 5.695266272189349e-06,
+      "loss": 0.5918,
+      "step": 292
+    },
+    {
+      "epoch": 0.4334319526627219,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.680473372781066e-06,
+      "loss": 0.608,
+      "step": 293
+    },
+    {
+      "epoch": 0.4349112426035503,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.665680473372781e-06,
+      "loss": 0.6122,
+      "step": 294
+    },
+    {
+      "epoch": 0.4363905325443787,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.650887573964498e-06,
+      "loss": 0.6209,
+      "step": 295
+    },
+    {
+      "epoch": 0.4378698224852071,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.636094674556214e-06,
+      "loss": 0.6056,
+      "step": 296
+    },
+    {
+      "epoch": 0.4393491124260355,
+      "grad_norm": 0.3125,
+      "learning_rate": 5.621301775147929e-06,
+      "loss": 0.6166,
+      "step": 297
+    },
+    {
+      "epoch": 0.4408284023668639,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.606508875739646e-06,
+      "loss": 0.6296,
+      "step": 298
+    },
+    {
+      "epoch": 0.4423076923076923,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.591715976331361e-06,
+      "loss": 0.618,
+      "step": 299
+    },
+    {
+      "epoch": 0.4437869822485207,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.576923076923077e-06,
+      "loss": 0.598,
+      "step": 300
+    },
+    {
+      "epoch": 0.4452662721893491,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.562130177514793e-06,
+      "loss": 0.5973,
+      "step": 301
+    },
+    {
+      "epoch": 0.4467455621301775,
+      "grad_norm": 0.3125,
+      "learning_rate": 5.5473372781065095e-06,
+      "loss": 0.6228,
+      "step": 302
+    },
+    {
+      "epoch": 0.44822485207100593,
+      "grad_norm": 0.30859375,
+      "learning_rate": 5.532544378698226e-06,
+      "loss": 0.611,
+      "step": 303
+    },
+    {
+      "epoch": 0.44970414201183434,
+      "grad_norm": 0.30859375,
+      "learning_rate": 5.517751479289941e-06,
+      "loss": 0.6094,
+      "step": 304
+    },
+    {
+      "epoch": 0.4511834319526627,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.502958579881657e-06,
+      "loss": 0.6156,
+      "step": 305
+    },
+    {
+      "epoch": 0.4526627218934911,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.488165680473373e-06,
+      "loss": 0.6077,
+      "step": 306
+    },
+    {
+      "epoch": 0.4541420118343195,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.47337278106509e-06,
+      "loss": 0.5891,
+      "step": 307
+    },
+    {
+      "epoch": 0.4556213017751479,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.4585798816568045e-06,
+      "loss": 0.6089,
+      "step": 308
+    },
+    {
+      "epoch": 0.45710059171597633,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.443786982248521e-06,
+      "loss": 0.5967,
+      "step": 309
+    },
+    {
+      "epoch": 0.45857988165680474,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.4289940828402376e-06,
+      "loss": 0.5958,
+      "step": 310
+    },
+    {
+      "epoch": 0.46005917159763315,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.414201183431953e-06,
+      "loss": 0.5989,
+      "step": 311
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.39940828402367e-06,
+      "loss": 0.588,
+      "step": 312
+    },
+    {
+      "epoch": 0.4630177514792899,
+      "grad_norm": 0.32421875,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 0.6028,
+      "step": 313
+    },
+    {
+      "epoch": 0.46449704142011833,
+      "grad_norm": 0.34375,
+      "learning_rate": 5.369822485207101e-06,
+      "loss": 0.5966,
+      "step": 314
+    },
+    {
+      "epoch": 0.46597633136094674,
+      "grad_norm": 0.3359375,
+      "learning_rate": 5.355029585798817e-06,
+      "loss": 0.5906,
+      "step": 315
+    },
+    {
+      "epoch": 0.46745562130177515,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.3402366863905334e-06,
+      "loss": 0.6084,
+      "step": 316
+    },
+    {
+      "epoch": 0.46893491124260356,
+      "grad_norm": 0.32421875,
+      "learning_rate": 5.325443786982249e-06,
+      "loss": 0.6294,
+      "step": 317
+    },
+    {
+      "epoch": 0.47041420118343197,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.310650887573965e-06,
+      "loss": 0.5992,
+      "step": 318
+    },
+    {
+      "epoch": 0.4718934911242604,
+      "grad_norm": 0.328125,
+      "learning_rate": 5.295857988165681e-06,
+      "loss": 0.6162,
+      "step": 319
+    },
+    {
+      "epoch": 0.47337278106508873,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.281065088757396e-06,
+      "loss": 0.5896,
+      "step": 320
+    },
+    {
+      "epoch": 0.47485207100591714,
+      "grad_norm": 0.33203125,
+      "learning_rate": 5.266272189349113e-06,
+      "loss": 0.6053,
+      "step": 321
+    },
+    {
+      "epoch": 0.47633136094674555,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.2514792899408285e-06,
+      "loss": 0.5949,
+      "step": 322
+    },
+    {
+      "epoch": 0.47781065088757396,
+      "grad_norm": 0.349609375,
+      "learning_rate": 5.236686390532545e-06,
+      "loss": 0.6179,
+      "step": 323
+    },
+    {
+      "epoch": 0.47928994082840237,
+      "grad_norm": 0.3515625,
+      "learning_rate": 5.22189349112426e-06,
+      "loss": 0.5907,
+      "step": 324
+    },
+    {
+      "epoch": 0.4807692307692308,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.207100591715976e-06,
+      "loss": 0.5961,
+      "step": 325
+    },
+    {
+      "epoch": 0.4822485207100592,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.192307692307693e-06,
+      "loss": 0.6169,
+      "step": 326
+    },
+    {
+      "epoch": 0.48372781065088755,
+      "grad_norm": 0.318359375,
+      "learning_rate": 5.177514792899409e-06,
+      "loss": 0.6008,
+      "step": 327
+    },
+    {
+      "epoch": 0.48520710059171596,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.162721893491125e-06,
+      "loss": 0.5942,
+      "step": 328
+    },
+    {
+      "epoch": 0.48668639053254437,
+      "grad_norm": 0.359375,
+      "learning_rate": 5.14792899408284e-06,
+      "loss": 0.5903,
+      "step": 329
+    },
+    {
+      "epoch": 0.4881656804733728,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.133136094674557e-06,
+      "loss": 0.5754,
+      "step": 330
+    },
+    {
+      "epoch": 0.4896449704142012,
+      "grad_norm": 0.3125,
+      "learning_rate": 5.118343195266272e-06,
+      "loss": 0.5818,
+      "step": 331
+    },
+    {
+      "epoch": 0.4911242603550296,
+      "grad_norm": 0.326171875,
+      "learning_rate": 5.103550295857989e-06,
+      "loss": 0.5918,
+      "step": 332
+    },
+    {
+      "epoch": 0.492603550295858,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.088757396449705e-06,
+      "loss": 0.6153,
+      "step": 333
+    },
+    {
+      "epoch": 0.4940828402366864,
+      "grad_norm": 0.3125,
+      "learning_rate": 5.07396449704142e-06,
+      "loss": 0.6038,
+      "step": 334
+    },
+    {
+      "epoch": 0.49556213017751477,
+      "grad_norm": 0.3046875,
+      "learning_rate": 5.059171597633137e-06,
+      "loss": 0.5753,
+      "step": 335
+    },
+    {
+      "epoch": 0.4970414201183432,
+      "grad_norm": 0.345703125,
+      "learning_rate": 5.0443786982248525e-06,
+      "loss": 0.5963,
+      "step": 336
+    },
+    {
+      "epoch": 0.4985207100591716,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.029585798816569e-06,
+      "loss": 0.585,
+      "step": 337
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 5.014792899408284e-06,
+      "loss": 0.5888,
+      "step": 338
+    },
+    {
+      "epoch": 0.5014792899408284,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5e-06,
+      "loss": 0.5802,
+      "step": 339
+    },
+    {
+      "epoch": 0.5029585798816568,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.985207100591716e-06,
+      "loss": 0.578,
+      "step": 340
+    },
+    {
+      "epoch": 0.5044378698224852,
+      "grad_norm": 0.31640625,
+      "learning_rate": 4.970414201183432e-06,
+      "loss": 0.5865,
+      "step": 341
+    },
+    {
+      "epoch": 0.5059171597633136,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.955621301775148e-06,
+      "loss": 0.5892,
+      "step": 342
+    },
+    {
+      "epoch": 0.507396449704142,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.940828402366865e-06,
+      "loss": 0.5974,
+      "step": 343
+    },
+    {
+      "epoch": 0.5088757396449705,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.926035502958581e-06,
+      "loss": 0.5886,
+      "step": 344
+    },
+    {
+      "epoch": 0.5103550295857988,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.911242603550296e-06,
+      "loss": 0.5863,
+      "step": 345
+    },
+    {
+      "epoch": 0.5118343195266272,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.896449704142012e-06,
+      "loss": 0.6069,
+      "step": 346
+    },
+    {
+      "epoch": 0.5133136094674556,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.8816568047337285e-06,
+      "loss": 0.6,
+      "step": 347
+    },
+    {
+      "epoch": 0.514792899408284,
+      "grad_norm": 0.322265625,
+      "learning_rate": 4.866863905325444e-06,
+      "loss": 0.5811,
+      "step": 348
+    },
+    {
+      "epoch": 0.5162721893491125,
+      "grad_norm": 0.380859375,
+      "learning_rate": 4.85207100591716e-06,
+      "loss": 0.597,
+      "step": 349
+    },
+    {
+      "epoch": 0.5177514792899408,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.837278106508876e-06,
+      "loss": 0.5923,
+      "step": 350
+    },
+    {
+      "epoch": 0.5192307692307693,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.822485207100592e-06,
+      "loss": 0.5851,
+      "step": 351
+    },
+    {
+      "epoch": 0.5207100591715976,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.807692307692308e-06,
+      "loss": 0.5761,
+      "step": 352
+    },
+    {
+      "epoch": 0.522189349112426,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.792899408284024e-06,
+      "loss": 0.5885,
+      "step": 353
+    },
+    {
+      "epoch": 0.5236686390532544,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.77810650887574e-06,
+      "loss": 0.5982,
+      "step": 354
+    },
+    {
+      "epoch": 0.5251479289940828,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.763313609467456e-06,
+      "loss": 0.6099,
+      "step": 355
+    },
+    {
+      "epoch": 0.5266272189349113,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.7485207100591715e-06,
+      "loss": 0.606,
+      "step": 356
+    },
+    {
+      "epoch": 0.5281065088757396,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.733727810650888e-06,
+      "loss": 0.6068,
+      "step": 357
+    },
+    {
+      "epoch": 0.5295857988165681,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.718934911242605e-06,
+      "loss": 0.5901,
+      "step": 358
+    },
+    {
+      "epoch": 0.5310650887573964,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.70414201183432e-06,
+      "loss": 0.6075,
+      "step": 359
+    },
+    {
+      "epoch": 0.5325443786982249,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.689349112426036e-06,
+      "loss": 0.5928,
+      "step": 360
+    },
+    {
+      "epoch": 0.5340236686390533,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.674556213017752e-06,
+      "loss": 0.6098,
+      "step": 361
+    },
+    {
+      "epoch": 0.5355029585798816,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.659763313609467e-06,
+      "loss": 0.6056,
+      "step": 362
+    },
+    {
+      "epoch": 0.5369822485207101,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.644970414201184e-06,
+      "loss": 0.6067,
+      "step": 363
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.6301775147929e-06,
+      "loss": 0.6019,
+      "step": 364
+    },
+    {
+      "epoch": 0.5399408284023669,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.5756,
+      "step": 365
+    },
+    {
+      "epoch": 0.5414201183431953,
+      "grad_norm": 0.375,
+      "learning_rate": 4.600591715976332e-06,
+      "loss": 0.5908,
+      "step": 366
+    },
+    {
+      "epoch": 0.5428994082840237,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.5857988165680475e-06,
+      "loss": 0.5841,
+      "step": 367
+    },
+    {
+      "epoch": 0.5443786982248521,
+      "grad_norm": 0.375,
+      "learning_rate": 4.571005917159764e-06,
+      "loss": 0.5753,
+      "step": 368
+    },
+    {
+      "epoch": 0.5458579881656804,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.55621301775148e-06,
+      "loss": 0.5663,
+      "step": 369
+    },
+    {
+      "epoch": 0.5473372781065089,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.5414201183431955e-06,
+      "loss": 0.6038,
+      "step": 370
+    },
+    {
+      "epoch": 0.5488165680473372,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.526627218934911e-06,
+      "loss": 0.5855,
+      "step": 371
+    },
+    {
+      "epoch": 0.5502958579881657,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.511834319526628e-06,
+      "loss": 0.6029,
+      "step": 372
+    },
+    {
+      "epoch": 0.5517751479289941,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.497041420118343e-06,
+      "loss": 0.5934,
+      "step": 373
+    },
+    {
+      "epoch": 0.5532544378698225,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.48224852071006e-06,
+      "loss": 0.5672,
+      "step": 374
+    },
+    {
+      "epoch": 0.5547337278106509,
+      "grad_norm": 0.341796875,
+      "learning_rate": 4.467455621301776e-06,
+      "loss": 0.5848,
+      "step": 375
+    },
+    {
+      "epoch": 0.5562130177514792,
+      "grad_norm": 0.328125,
+      "learning_rate": 4.452662721893491e-06,
+      "loss": 0.6083,
+      "step": 376
+    },
+    {
+      "epoch": 0.5576923076923077,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.437869822485207e-06,
+      "loss": 0.579,
+      "step": 377
+    },
+    {
+      "epoch": 0.5591715976331361,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.423076923076924e-06,
+      "loss": 0.5701,
+      "step": 378
+    },
+    {
+      "epoch": 0.5606508875739645,
+      "grad_norm": 0.3515625,
+      "learning_rate": 4.408284023668639e-06,
+      "loss": 0.5797,
+      "step": 379
+    },
+    {
+      "epoch": 0.5621301775147929,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.393491124260356e-06,
+      "loss": 0.5945,
+      "step": 380
+    },
+    {
+      "epoch": 0.5636094674556213,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.3786982248520715e-06,
+      "loss": 0.5777,
+      "step": 381
+    },
+    {
+      "epoch": 0.5650887573964497,
+      "grad_norm": 0.35546875,
+      "learning_rate": 4.363905325443787e-06,
+      "loss": 0.5934,
+      "step": 382
+    },
+    {
+      "epoch": 0.5665680473372781,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.349112426035503e-06,
+      "loss": 0.5951,
+      "step": 383
+    },
+    {
+      "epoch": 0.5680473372781065,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.3343195266272195e-06,
+      "loss": 0.5683,
+      "step": 384
+    },
+    {
+      "epoch": 0.5695266272189349,
+      "grad_norm": 0.314453125,
+      "learning_rate": 4.319526627218935e-06,
+      "loss": 0.5823,
+      "step": 385
+    },
+    {
+      "epoch": 0.5710059171597633,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.304733727810651e-06,
+      "loss": 0.5814,
+      "step": 386
+    },
+    {
+      "epoch": 0.5724852071005917,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.289940828402367e-06,
+      "loss": 0.5872,
+      "step": 387
+    },
+    {
+      "epoch": 0.5739644970414202,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.275147928994083e-06,
+      "loss": 0.5858,
+      "step": 388
+    },
+    {
+      "epoch": 0.5754437869822485,
+      "grad_norm": 0.3203125,
+      "learning_rate": 4.2603550295858e-06,
+      "loss": 0.5755,
+      "step": 389
+    },
+    {
+      "epoch": 0.5769230769230769,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.245562130177515e-06,
+      "loss": 0.5867,
+      "step": 390
+    },
+    {
+      "epoch": 0.5784023668639053,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 0.5669,
+      "step": 391
+    },
+    {
+      "epoch": 0.5798816568047337,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.215976331360947e-06,
+      "loss": 0.5628,
+      "step": 392
+    },
+    {
+      "epoch": 0.5813609467455622,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.201183431952663e-06,
+      "loss": 0.5684,
+      "step": 393
+    },
+    {
+      "epoch": 0.5828402366863905,
+      "grad_norm": 0.3359375,
+      "learning_rate": 4.186390532544379e-06,
+      "loss": 0.56,
+      "step": 394
+    },
+    {
+      "epoch": 0.584319526627219,
+      "grad_norm": 0.34765625,
+      "learning_rate": 4.1715976331360955e-06,
+      "loss": 0.5612,
+      "step": 395
+    },
+    {
+      "epoch": 0.5857988165680473,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.156804733727811e-06,
+      "loss": 0.5838,
+      "step": 396
+    },
+    {
+      "epoch": 0.5872781065088757,
+      "grad_norm": 0.330078125,
+      "learning_rate": 4.142011834319527e-06,
+      "loss": 0.5791,
+      "step": 397
+    },
+    {
+      "epoch": 0.5887573964497042,
+      "grad_norm": 0.37890625,
+      "learning_rate": 4.127218934911243e-06,
+      "loss": 0.5863,
+      "step": 398
+    },
+    {
+      "epoch": 0.5902366863905325,
+      "grad_norm": 0.30859375,
+      "learning_rate": 4.112426035502959e-06,
+      "loss": 0.5734,
+      "step": 399
+    },
+    {
+      "epoch": 0.591715976331361,
+      "grad_norm": 0.30859375,
+      "learning_rate": 4.097633136094675e-06,
+      "loss": 0.563,
+      "step": 400
+    },
+    {
+      "epoch": 0.5931952662721893,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.0828402366863906e-06,
+      "loss": 0.5916,
+      "step": 401
+    },
+    {
+      "epoch": 0.5946745562130178,
+      "grad_norm": 0.3203125,
+      "learning_rate": 4.068047337278107e-06,
+      "loss": 0.5911,
+      "step": 402
+    },
+    {
+      "epoch": 0.5961538461538461,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.053254437869823e-06,
+      "loss": 0.5944,
+      "step": 403
+    },
+    {
+      "epoch": 0.5976331360946746,
+      "grad_norm": 0.31640625,
+      "learning_rate": 4.0384615384615385e-06,
+      "loss": 0.5864,
+      "step": 404
+    },
+    {
+      "epoch": 0.599112426035503,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.023668639053255e-06,
+      "loss": 0.5917,
+      "step": 405
+    },
+    {
+      "epoch": 0.6005917159763313,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.008875739644971e-06,
+      "loss": 0.5621,
+      "step": 406
+    },
+    {
+      "epoch": 0.6020710059171598,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.9940828402366864e-06,
+      "loss": 0.5806,
+      "step": 407
+    },
+    {
+      "epoch": 0.6035502958579881,
+      "grad_norm": 0.349609375,
+      "learning_rate": 3.979289940828402e-06,
+      "loss": 0.5882,
+      "step": 408
+    },
+    {
+      "epoch": 0.6050295857988166,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.964497041420119e-06,
+      "loss": 0.5766,
+      "step": 409
+    },
+    {
+      "epoch": 0.606508875739645,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.949704142011835e-06,
+      "loss": 0.5719,
+      "step": 410
+    },
+    {
+      "epoch": 0.6079881656804734,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.934911242603551e-06,
+      "loss": 0.5925,
+      "step": 411
+    },
+    {
+      "epoch": 0.6094674556213018,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.920118343195267e-06,
+      "loss": 0.5731,
+      "step": 412
+    },
+    {
+      "epoch": 0.6109467455621301,
+      "grad_norm": 0.3046875,
+      "learning_rate": 3.905325443786982e-06,
+      "loss": 0.5922,
+      "step": 413
+    },
+    {
+      "epoch": 0.6124260355029586,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.890532544378698e-06,
+      "loss": 0.5672,
+      "step": 414
+    },
+    {
+      "epoch": 0.613905325443787,
+      "grad_norm": 0.302734375,
+      "learning_rate": 3.8757396449704146e-06,
+      "loss": 0.5772,
+      "step": 415
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.86094674556213e-06,
+      "loss": 0.5818,
+      "step": 416
+    },
+    {
+      "epoch": 0.6168639053254438,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.5888,
+      "step": 417
+    },
+    {
+      "epoch": 0.6183431952662722,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.8313609467455625e-06,
+      "loss": 0.5716,
+      "step": 418
+    },
+    {
+      "epoch": 0.6198224852071006,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.816568047337278e-06,
+      "loss": 0.5679,
+      "step": 419
+    },
+    {
+      "epoch": 0.621301775147929,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.8017751479289943e-06,
+      "loss": 0.5971,
+      "step": 420
+    },
+    {
+      "epoch": 0.6227810650887574,
+      "grad_norm": 0.314453125,
+      "learning_rate": 3.7869822485207104e-06,
+      "loss": 0.562,
+      "step": 421
+    },
+    {
+      "epoch": 0.6242603550295858,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.772189349112426e-06,
+      "loss": 0.5492,
+      "step": 422
+    },
+    {
+      "epoch": 0.6257396449704142,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.7573964497041422e-06,
+      "loss": 0.5806,
+      "step": 423
+    },
+    {
+      "epoch": 0.6272189349112426,
+      "grad_norm": 0.306640625,
+      "learning_rate": 3.7426035502958584e-06,
+      "loss": 0.5723,
+      "step": 424
+    },
+    {
+      "epoch": 0.628698224852071,
+      "grad_norm": 0.314453125,
+      "learning_rate": 3.7278106508875745e-06,
+      "loss": 0.5735,
+      "step": 425
+    },
+    {
+      "epoch": 0.6301775147928994,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.71301775147929e-06,
+      "loss": 0.5726,
+      "step": 426
+    },
+    {
+      "epoch": 0.6316568047337278,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.6982248520710063e-06,
+      "loss": 0.5876,
+      "step": 427
+    },
+    {
+      "epoch": 0.6331360946745562,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.683431952662722e-06,
+      "loss": 0.5697,
+      "step": 428
+    },
+    {
+      "epoch": 0.6346153846153846,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.668639053254438e-06,
+      "loss": 0.572,
+      "step": 429
+    },
+    {
+      "epoch": 0.636094674556213,
+      "grad_norm": 0.3515625,
+      "learning_rate": 3.653846153846154e-06,
+      "loss": 0.5775,
+      "step": 430
+    },
+    {
+      "epoch": 0.6375739644970414,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.6390532544378704e-06,
+      "loss": 0.5707,
+      "step": 431
+    },
+    {
+      "epoch": 0.6390532544378699,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.6242603550295865e-06,
+      "loss": 0.5729,
+      "step": 432
+    },
+    {
+      "epoch": 0.6405325443786982,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.609467455621302e-06,
+      "loss": 0.5782,
+      "step": 433
+    },
+    {
+      "epoch": 0.6420118343195266,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.594674556213018e-06,
+      "loss": 0.5713,
+      "step": 434
+    },
+    {
+      "epoch": 0.643491124260355,
+      "grad_norm": 0.322265625,
+      "learning_rate": 3.579881656804734e-06,
+      "loss": 0.5565,
+      "step": 435
+    },
+    {
+      "epoch": 0.6449704142011834,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.5650887573964497e-06,
+      "loss": 0.5801,
+      "step": 436
+    },
+    {
+      "epoch": 0.6464497041420119,
+      "grad_norm": 0.31640625,
+      "learning_rate": 3.550295857988166e-06,
+      "loss": 0.5766,
+      "step": 437
+    },
+    {
+      "epoch": 0.6479289940828402,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.5355029585798815e-06,
+      "loss": 0.5859,
+      "step": 438
+    },
+    {
+      "epoch": 0.6494082840236687,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.520710059171598e-06,
+      "loss": 0.5631,
+      "step": 439
+    },
+    {
+      "epoch": 0.650887573964497,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.505917159763314e-06,
+      "loss": 0.5751,
+      "step": 440
+    },
+    {
+      "epoch": 0.6523668639053254,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.49112426035503e-06,
+      "loss": 0.5752,
+      "step": 441
+    },
+    {
+      "epoch": 0.6538461538461539,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.476331360946746e-06,
+      "loss": 0.5655,
+      "step": 442
+    },
+    {
+      "epoch": 0.6553254437869822,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 0.5449,
+      "step": 443
+    },
+    {
+      "epoch": 0.6568047337278107,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.446745562130178e-06,
+      "loss": 0.5768,
+      "step": 444
+    },
+    {
+      "epoch": 0.658284023668639,
+      "grad_norm": 0.33203125,
+      "learning_rate": 3.4319526627218935e-06,
+      "loss": 0.5549,
+      "step": 445
+    },
+    {
+      "epoch": 0.6597633136094675,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.41715976331361e-06,
+      "loss": 0.5796,
+      "step": 446
+    },
+    {
+      "epoch": 0.6612426035502958,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.4023668639053257e-06,
+      "loss": 0.5644,
+      "step": 447
+    },
+    {
+      "epoch": 0.6627218934911243,
+      "grad_norm": 0.29296875,
+      "learning_rate": 3.387573964497042e-06,
+      "loss": 0.5836,
+      "step": 448
+    },
+    {
+      "epoch": 0.6642011834319527,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.3727810650887576e-06,
+      "loss": 0.5699,
+      "step": 449
+    },
+    {
+      "epoch": 0.665680473372781,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.3579881656804737e-06,
+      "loss": 0.5646,
+      "step": 450
+    },
+    {
+      "epoch": 0.6671597633136095,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.3431952662721894e-06,
+      "loss": 0.5852,
+      "step": 451
+    },
+    {
+      "epoch": 0.6686390532544378,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.3284023668639055e-06,
+      "loss": 0.5762,
+      "step": 452
+    },
+    {
+      "epoch": 0.6701183431952663,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.313609467455622e-06,
+      "loss": 0.5895,
+      "step": 453
+    },
+    {
+      "epoch": 0.6715976331360947,
+      "grad_norm": 0.330078125,
+      "learning_rate": 3.2988165680473377e-06,
+      "loss": 0.5716,
+      "step": 454
+    },
+    {
+      "epoch": 0.6730769230769231,
+      "grad_norm": 0.359375,
+      "learning_rate": 3.2840236686390534e-06,
+      "loss": 0.5847,
+      "step": 455
+    },
+    {
+      "epoch": 0.6745562130177515,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.2692307692307696e-06,
+      "loss": 0.5824,
+      "step": 456
+    },
+    {
+      "epoch": 0.6760355029585798,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.2544378698224853e-06,
+      "loss": 0.5666,
+      "step": 457
+    },
+    {
+      "epoch": 0.6775147928994083,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.2396449704142014e-06,
+      "loss": 0.5785,
+      "step": 458
+    },
+    {
+      "epoch": 0.6789940828402367,
+      "grad_norm": 0.314453125,
+      "learning_rate": 3.224852071005917e-06,
+      "loss": 0.5636,
+      "step": 459
+    },
+    {
+      "epoch": 0.6804733727810651,
+      "grad_norm": 0.328125,
+      "learning_rate": 3.210059171597633e-06,
+      "loss": 0.5583,
+      "step": 460
+    },
+    {
+      "epoch": 0.6819526627218935,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.1952662721893497e-06,
+      "loss": 0.5774,
+      "step": 461
+    },
+    {
+      "epoch": 0.6834319526627219,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.1804733727810654e-06,
+      "loss": 0.5655,
+      "step": 462
+    },
+    {
+      "epoch": 0.6849112426035503,
+      "grad_norm": 0.30859375,
+      "learning_rate": 3.1656804733727816e-06,
+      "loss": 0.571,
+      "step": 463
+    },
+    {
+      "epoch": 0.6863905325443787,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.1508875739644973e-06,
+      "loss": 0.587,
+      "step": 464
+    },
+    {
+      "epoch": 0.6878698224852071,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.1360946745562134e-06,
+      "loss": 0.5672,
+      "step": 465
+    },
+    {
+      "epoch": 0.6893491124260355,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.121301775147929e-06,
+      "loss": 0.5727,
+      "step": 466
+    },
+    {
+      "epoch": 0.6908284023668639,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.106508875739645e-06,
+      "loss": 0.5452,
+      "step": 467
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.0917159763313613e-06,
+      "loss": 0.5746,
+      "step": 468
+    },
+    {
+      "epoch": 0.6937869822485208,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.5719,
+      "step": 469
+    },
+    {
+      "epoch": 0.6952662721893491,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.062130177514793e-06,
+      "loss": 0.5875,
+      "step": 470
+    },
+    {
+      "epoch": 0.6967455621301775,
+      "grad_norm": 0.369140625,
+      "learning_rate": 3.0473372781065093e-06,
+      "loss": 0.5983,
+      "step": 471
+    },
+    {
+      "epoch": 0.6982248520710059,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.032544378698225e-06,
+      "loss": 0.5947,
+      "step": 472
+    },
+    {
+      "epoch": 0.6997041420118343,
+      "grad_norm": 0.375,
+      "learning_rate": 3.017751479289941e-06,
+      "loss": 0.5494,
+      "step": 473
+    },
+    {
+      "epoch": 0.7011834319526628,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.0029585798816568e-06,
+      "loss": 0.561,
+      "step": 474
+    },
+    {
+      "epoch": 0.7026627218934911,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.9881656804733733e-06,
+      "loss": 0.5795,
+      "step": 475
+    },
+    {
+      "epoch": 0.7041420118343196,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.973372781065089e-06,
+      "loss": 0.5694,
+      "step": 476
+    },
+    {
+      "epoch": 0.7056213017751479,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.958579881656805e-06,
+      "loss": 0.5601,
+      "step": 477
+    },
+    {
+      "epoch": 0.7071005917159763,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.943786982248521e-06,
+      "loss": 0.5641,
+      "step": 478
+    },
+    {
+      "epoch": 0.7085798816568047,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.928994082840237e-06,
+      "loss": 0.5816,
+      "step": 479
+    },
+    {
+      "epoch": 0.7100591715976331,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.9142011834319526e-06,
+      "loss": 0.5707,
+      "step": 480
+    },
+    {
+      "epoch": 0.7115384615384616,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.8994082840236688e-06,
+      "loss": 0.5626,
+      "step": 481
+    },
+    {
+      "epoch": 0.7130177514792899,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.8846153846153845e-06,
+      "loss": 0.5636,
+      "step": 482
+    },
+    {
+      "epoch": 0.7144970414201184,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.869822485207101e-06,
+      "loss": 0.5834,
+      "step": 483
+    },
+    {
+      "epoch": 0.7159763313609467,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.855029585798817e-06,
+      "loss": 0.5864,
+      "step": 484
+    },
+    {
+      "epoch": 0.7174556213017751,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.840236686390533e-06,
+      "loss": 0.5701,
+      "step": 485
+    },
+    {
+      "epoch": 0.7189349112426036,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.825443786982249e-06,
+      "loss": 0.559,
+      "step": 486
+    },
+    {
+      "epoch": 0.7204142011834319,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.8106508875739646e-06,
+      "loss": 0.5605,
+      "step": 487
+    },
+    {
+      "epoch": 0.7218934911242604,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.7958579881656803e-06,
+      "loss": 0.5703,
+      "step": 488
+    },
+    {
+      "epoch": 0.7233727810650887,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.7810650887573965e-06,
+      "loss": 0.5692,
+      "step": 489
+    },
+    {
+      "epoch": 0.7248520710059172,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.766272189349113e-06,
+      "loss": 0.5771,
+      "step": 490
+    },
+    {
+      "epoch": 0.7263313609467456,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.7514792899408287e-06,
+      "loss": 0.5505,
+      "step": 491
+    },
+    {
+      "epoch": 0.727810650887574,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.736686390532545e-06,
+      "loss": 0.57,
+      "step": 492
+    },
+    {
+      "epoch": 0.7292899408284024,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.7218934911242605e-06,
+      "loss": 0.5648,
+      "step": 493
+    },
+    {
+      "epoch": 0.7307692307692307,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.7071005917159766e-06,
+      "loss": 0.5493,
+      "step": 494
+    },
+    {
+      "epoch": 0.7322485207100592,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 0.5865,
+      "step": 495
+    },
+    {
+      "epoch": 0.7337278106508875,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.6775147928994085e-06,
+      "loss": 0.5645,
+      "step": 496
+    },
+    {
+      "epoch": 0.735207100591716,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.6627218934911246e-06,
+      "loss": 0.5788,
+      "step": 497
+    },
+    {
+      "epoch": 0.7366863905325444,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.6479289940828407e-06,
+      "loss": 0.5419,
+      "step": 498
+    },
+    {
+      "epoch": 0.7381656804733728,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.6331360946745564e-06,
+      "loss": 0.5738,
+      "step": 499
+    },
+    {
+      "epoch": 0.7396449704142012,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.6183431952662725e-06,
+      "loss": 0.5762,
+      "step": 500
+    },
+    {
+      "epoch": 0.7411242603550295,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.603550295857988e-06,
+      "loss": 0.565,
+      "step": 501
+    },
+    {
+      "epoch": 0.742603550295858,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.5887573964497043e-06,
+      "loss": 0.5741,
+      "step": 502
+    },
+    {
+      "epoch": 0.7440828402366864,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.57396449704142e-06,
+      "loss": 0.5627,
+      "step": 503
+    },
+    {
+      "epoch": 0.7455621301775148,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.559171597633136e-06,
+      "loss": 0.5762,
+      "step": 504
+    },
+    {
+      "epoch": 0.7470414201183432,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.5443786982248527e-06,
+      "loss": 0.5589,
+      "step": 505
+    },
+    {
+      "epoch": 0.7485207100591716,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.5295857988165684e-06,
+      "loss": 0.5686,
+      "step": 506
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.5147928994082845e-06,
+      "loss": 0.5663,
+      "step": 507
+    },
+    {
+      "epoch": 0.7514792899408284,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.5e-06,
+      "loss": 0.5551,
+      "step": 508
+    },
+    {
+      "epoch": 0.7529585798816568,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.485207100591716e-06,
+      "loss": 0.5739,
+      "step": 509
+    },
+    {
+      "epoch": 0.7544378698224852,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.4704142011834324e-06,
+      "loss": 0.5589,
+      "step": 510
+    },
+    {
+      "epoch": 0.7559171597633136,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.455621301775148e-06,
+      "loss": 0.5828,
+      "step": 511
+    },
+    {
+      "epoch": 0.757396449704142,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.4408284023668643e-06,
+      "loss": 0.5721,
+      "step": 512
+    },
+    {
+      "epoch": 0.7588757396449705,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.42603550295858e-06,
+      "loss": 0.569,
+      "step": 513
+    },
+    {
+      "epoch": 0.7603550295857988,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.411242603550296e-06,
+      "loss": 0.5795,
+      "step": 514
+    },
+    {
+      "epoch": 0.7618343195266272,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.396449704142012e-06,
+      "loss": 0.5573,
+      "step": 515
+    },
+    {
+      "epoch": 0.7633136094674556,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.381656804733728e-06,
+      "loss": 0.5668,
+      "step": 516
+    },
+    {
+      "epoch": 0.764792899408284,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.366863905325444e-06,
+      "loss": 0.5755,
+      "step": 517
+    },
+    {
+      "epoch": 0.7662721893491125,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.35207100591716e-06,
+      "loss": 0.5439,
+      "step": 518
+    },
+    {
+      "epoch": 0.7677514792899408,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.337278106508876e-06,
+      "loss": 0.6044,
+      "step": 519
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.322485207100592e-06,
+      "loss": 0.5703,
+      "step": 520
+    },
+    {
+      "epoch": 0.7707100591715976,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.5855,
+      "step": 521
+    },
+    {
+      "epoch": 0.772189349112426,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.2928994082840238e-06,
+      "loss": 0.5794,
+      "step": 522
+    },
+    {
+      "epoch": 0.7736686390532544,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.27810650887574e-06,
+      "loss": 0.5678,
+      "step": 523
+    },
+    {
+      "epoch": 0.7751479289940828,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.2633136094674556e-06,
+      "loss": 0.571,
+      "step": 524
+    },
+    {
+      "epoch": 0.7766272189349113,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.2485207100591717e-06,
+      "loss": 0.5697,
+      "step": 525
+    },
+    {
+      "epoch": 0.7781065088757396,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.233727810650888e-06,
+      "loss": 0.5643,
+      "step": 526
+    },
+    {
+      "epoch": 0.7795857988165681,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.2189349112426035e-06,
+      "loss": 0.5924,
+      "step": 527
+    },
+    {
+      "epoch": 0.7810650887573964,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.2041420118343196e-06,
+      "loss": 0.5483,
+      "step": 528
+    },
+    {
+      "epoch": 0.7825443786982249,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.1893491124260358e-06,
+      "loss": 0.5689,
+      "step": 529
+    },
+    {
+      "epoch": 0.7840236686390533,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.1745562130177515e-06,
+      "loss": 0.5588,
+      "step": 530
+    },
+    {
+      "epoch": 0.7855029585798816,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.1597633136094676e-06,
+      "loss": 0.5674,
+      "step": 531
+    },
+    {
+      "epoch": 0.7869822485207101,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.1449704142011837e-06,
+      "loss": 0.5715,
+      "step": 532
+    },
+    {
+      "epoch": 0.7884615384615384,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.1301775147929e-06,
+      "loss": 0.5608,
+      "step": 533
+    },
+    {
+      "epoch": 0.7899408284023669,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.1153846153846155e-06,
+      "loss": 0.5712,
+      "step": 534
+    },
+    {
+      "epoch": 0.7914201183431953,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.1005917159763316e-06,
+      "loss": 0.5642,
+      "step": 535
+    },
+    {
+      "epoch": 0.7928994082840237,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.0857988165680478e-06,
+      "loss": 0.5829,
+      "step": 536
+    },
+    {
+      "epoch": 0.7943786982248521,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.0710059171597635e-06,
+      "loss": 0.5663,
+      "step": 537
+    },
+    {
+      "epoch": 0.7958579881656804,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.0562130177514796e-06,
+      "loss": 0.5441,
+      "step": 538
+    },
+    {
+      "epoch": 0.7973372781065089,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.0414201183431953e-06,
+      "loss": 0.5885,
+      "step": 539
+    },
+    {
+      "epoch": 0.7988165680473372,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.0266272189349114e-06,
+      "loss": 0.5578,
+      "step": 540
+    },
+    {
+      "epoch": 0.8002958579881657,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.0118343195266275e-06,
+      "loss": 0.5711,
+      "step": 541
+    },
+    {
+      "epoch": 0.8017751479289941,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9970414201183432e-06,
+      "loss": 0.5577,
+      "step": 542
+    },
+    {
+      "epoch": 0.8032544378698225,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.9822485207100593e-06,
+      "loss": 0.5473,
+      "step": 543
+    },
+    {
+      "epoch": 0.8047337278106509,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.9674556213017755e-06,
+      "loss": 0.5505,
+      "step": 544
+    },
+    {
+      "epoch": 0.8062130177514792,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.952662721893491e-06,
+      "loss": 0.5484,
+      "step": 545
+    },
+    {
+      "epoch": 0.8076923076923077,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.9378698224852073e-06,
+      "loss": 0.5598,
+      "step": 546
+    },
+    {
+      "epoch": 0.8091715976331361,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.5665,
+      "step": 547
+    },
+    {
+      "epoch": 0.8106508875739645,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.908284023668639e-06,
+      "loss": 0.5909,
+      "step": 548
+    },
+    {
+      "epoch": 0.8121301775147929,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.8934911242603552e-06,
+      "loss": 0.5637,
+      "step": 549
+    },
+    {
+      "epoch": 0.8136094674556213,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8786982248520711e-06,
+      "loss": 0.5815,
+      "step": 550
+    },
+    {
+      "epoch": 0.8150887573964497,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8639053254437872e-06,
+      "loss": 0.5591,
+      "step": 551
+    },
+    {
+      "epoch": 0.8165680473372781,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.8491124260355032e-06,
+      "loss": 0.5759,
+      "step": 552
+    },
+    {
+      "epoch": 0.8180473372781065,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.834319526627219e-06,
+      "loss": 0.5888,
+      "step": 553
+    },
+    {
+      "epoch": 0.8195266272189349,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8195266272189352e-06,
+      "loss": 0.5647,
+      "step": 554
+    },
+    {
+      "epoch": 0.8210059171597633,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.804733727810651e-06,
+      "loss": 0.5686,
+      "step": 555
+    },
+    {
+      "epoch": 0.8224852071005917,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.789940828402367e-06,
+      "loss": 0.5556,
+      "step": 556
+    },
+    {
+      "epoch": 0.8239644970414202,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.775147928994083e-06,
+      "loss": 0.5613,
+      "step": 557
+    },
+    {
+      "epoch": 0.8254437869822485,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.760355029585799e-06,
+      "loss": 0.5642,
+      "step": 558
+    },
+    {
+      "epoch": 0.8269230769230769,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.745562130177515e-06,
+      "loss": 0.6028,
+      "step": 559
+    },
+    {
+      "epoch": 0.8284023668639053,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7307692307692308e-06,
+      "loss": 0.5737,
+      "step": 560
+    },
+    {
+      "epoch": 0.8298816568047337,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.7159763313609468e-06,
+      "loss": 0.5734,
+      "step": 561
+    },
+    {
+      "epoch": 0.8313609467455622,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.7011834319526629e-06,
+      "loss": 0.5527,
+      "step": 562
+    },
+    {
+      "epoch": 0.8328402366863905,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.6863905325443788e-06,
+      "loss": 0.5748,
+      "step": 563
+    },
+    {
+      "epoch": 0.834319526627219,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.6715976331360947e-06,
+      "loss": 0.5798,
+      "step": 564
+    },
+    {
+      "epoch": 0.8357988165680473,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.656804733727811e-06,
+      "loss": 0.5765,
+      "step": 565
+    },
+    {
+      "epoch": 0.8372781065088757,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.6420118343195267e-06,
+      "loss": 0.573,
+      "step": 566
+    },
+    {
+      "epoch": 0.8387573964497042,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.6272189349112426e-06,
+      "loss": 0.5562,
+      "step": 567
+    },
+    {
+      "epoch": 0.8402366863905325,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.6124260355029585e-06,
+      "loss": 0.5624,
+      "step": 568
+    },
+    {
+      "epoch": 0.841715976331361,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.5976331360946749e-06,
+      "loss": 0.5724,
+      "step": 569
+    },
+    {
+      "epoch": 0.8431952662721893,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.5828402366863908e-06,
+      "loss": 0.5572,
+      "step": 570
+    },
+    {
+      "epoch": 0.8446745562130178,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.5680473372781067e-06,
+      "loss": 0.5413,
+      "step": 571
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.5532544378698226e-06,
+      "loss": 0.5648,
+      "step": 572
+    },
+    {
+      "epoch": 0.8476331360946746,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.5582,
+      "step": 573
+    },
+    {
+      "epoch": 0.849112426035503,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.5236686390532546e-06,
+      "loss": 0.5686,
+      "step": 574
+    },
+    {
+      "epoch": 0.8505917159763313,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.5088757396449705e-06,
+      "loss": 0.5632,
+      "step": 575
+    },
+    {
+      "epoch": 0.8520710059171598,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.4940828402366867e-06,
+      "loss": 0.5714,
+      "step": 576
+    },
+    {
+      "epoch": 0.8535502958579881,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.4792899408284026e-06,
+      "loss": 0.5606,
+      "step": 577
+    },
+    {
+      "epoch": 0.8550295857988166,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.4644970414201185e-06,
+      "loss": 0.556,
+      "step": 578
+    },
+    {
+      "epoch": 0.856508875739645,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.4497041420118344e-06,
+      "loss": 0.5748,
+      "step": 579
+    },
+    {
+      "epoch": 0.8579881656804734,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.4349112426035505e-06,
+      "loss": 0.5854,
+      "step": 580
+    },
+    {
+      "epoch": 0.8594674556213018,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.4201183431952664e-06,
+      "loss": 0.5535,
+      "step": 581
+    },
+    {
+      "epoch": 0.8609467455621301,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.4053254437869823e-06,
+      "loss": 0.5738,
+      "step": 582
+    },
+    {
+      "epoch": 0.8624260355029586,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.3905325443786982e-06,
+      "loss": 0.5597,
+      "step": 583
+    },
+    {
+      "epoch": 0.863905325443787,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.3757396449704143e-06,
+      "loss": 0.5788,
+      "step": 584
+    },
+    {
+      "epoch": 0.8653846153846154,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3609467455621303e-06,
+      "loss": 0.5593,
+      "step": 585
+    },
+    {
+      "epoch": 0.8668639053254438,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.3461538461538462e-06,
+      "loss": 0.5647,
+      "step": 586
+    },
+    {
+      "epoch": 0.8683431952662722,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.3313609467455623e-06,
+      "loss": 0.5654,
+      "step": 587
+    },
+    {
+      "epoch": 0.8698224852071006,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.3165680473372782e-06,
+      "loss": 0.5731,
+      "step": 588
+    },
+    {
+      "epoch": 0.871301775147929,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.301775147928994e-06,
+      "loss": 0.5657,
+      "step": 589
+    },
+    {
+      "epoch": 0.8727810650887574,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.28698224852071e-06,
+      "loss": 0.5705,
+      "step": 590
+    },
+    {
+      "epoch": 0.8742603550295858,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.2721893491124263e-06,
+      "loss": 0.5743,
+      "step": 591
+    },
+    {
+      "epoch": 0.8757396449704142,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.2573964497041423e-06,
+      "loss": 0.5618,
+      "step": 592
+    },
+    {
+      "epoch": 0.8772189349112426,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.242603550295858e-06,
+      "loss": 0.5785,
+      "step": 593
+    },
+    {
+      "epoch": 0.878698224852071,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.227810650887574e-06,
+      "loss": 0.5398,
+      "step": 594
+    },
+    {
+      "epoch": 0.8801775147928994,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.21301775147929e-06,
+      "loss": 0.5583,
+      "step": 595
+    },
+    {
+      "epoch": 0.8816568047337278,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.198224852071006e-06,
+      "loss": 0.5545,
+      "step": 596
+    },
+    {
+      "epoch": 0.8831360946745562,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.183431952662722e-06,
+      "loss": 0.5672,
+      "step": 597
+    },
+    {
+      "epoch": 0.8846153846153846,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.168639053254438e-06,
+      "loss": 0.5593,
+      "step": 598
+    },
+    {
+      "epoch": 0.886094674556213,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 0.5657,
+      "step": 599
+    },
+    {
+      "epoch": 0.8875739644970414,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.13905325443787e-06,
+      "loss": 0.5733,
+      "step": 600
+    },
+    {
+      "epoch": 0.8890532544378699,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.1242603550295859e-06,
+      "loss": 0.5593,
+      "step": 601
+    },
+    {
+      "epoch": 0.8905325443786982,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.1094674556213018e-06,
+      "loss": 0.5633,
+      "step": 602
+    },
+    {
+      "epoch": 0.8920118343195266,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.0946745562130179e-06,
+      "loss": 0.5903,
+      "step": 603
+    },
+    {
+      "epoch": 0.893491124260355,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.0798816568047338e-06,
+      "loss": 0.5375,
+      "step": 604
+    },
+    {
+      "epoch": 0.8949704142011834,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.06508875739645e-06,
+      "loss": 0.5813,
+      "step": 605
+    },
+    {
+      "epoch": 0.8964497041420119,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.0502958579881658e-06,
+      "loss": 0.5783,
+      "step": 606
+    },
+    {
+      "epoch": 0.8979289940828402,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.0355029585798817e-06,
+      "loss": 0.5687,
+      "step": 607
+    },
+    {
+      "epoch": 0.8994082840236687,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.0207100591715976e-06,
+      "loss": 0.5787,
+      "step": 608
+    },
+    {
+      "epoch": 0.900887573964497,
+      "grad_norm": 0.375,
+      "learning_rate": 1.0059171597633138e-06,
+      "loss": 0.5585,
+      "step": 609
+    },
+    {
+      "epoch": 0.9023668639053254,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.911242603550297e-07,
+      "loss": 0.5476,
+      "step": 610
+    },
+    {
+      "epoch": 0.9038461538461539,
+      "grad_norm": 0.48046875,
+      "learning_rate": 9.763313609467456e-07,
+      "loss": 0.5506,
+      "step": 611
+    },
+    {
+      "epoch": 0.9053254437869822,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.615384615384617e-07,
+      "loss": 0.5637,
+      "step": 612
+    },
+    {
+      "epoch": 0.9068047337278107,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.467455621301776e-07,
+      "loss": 0.5886,
+      "step": 613
+    },
+    {
+      "epoch": 0.908284023668639,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.319526627218936e-07,
+      "loss": 0.5442,
+      "step": 614
+    },
+    {
+      "epoch": 0.9097633136094675,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.171597633136095e-07,
+      "loss": 0.5558,
+      "step": 615
+    },
+    {
+      "epoch": 0.9112426035502958,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.023668639053255e-07,
+      "loss": 0.5677,
+      "step": 616
+    },
+    {
+      "epoch": 0.9127218934911243,
+      "grad_norm": 0.30859375,
+      "learning_rate": 8.875739644970415e-07,
+      "loss": 0.5703,
+      "step": 617
+    },
+    {
+      "epoch": 0.9142011834319527,
+      "grad_norm": 0.30859375,
+      "learning_rate": 8.727810650887575e-07,
+      "loss": 0.5571,
+      "step": 618
+    },
+    {
+      "epoch": 0.915680473372781,
+      "grad_norm": 0.326171875,
+      "learning_rate": 8.579881656804734e-07,
+      "loss": 0.5645,
+      "step": 619
+    },
+    {
+      "epoch": 0.9171597633136095,
+      "grad_norm": 0.30078125,
+      "learning_rate": 8.431952662721894e-07,
+      "loss": 0.5711,
+      "step": 620
+    },
+    {
+      "epoch": 0.9186390532544378,
+      "grad_norm": 0.3125,
+      "learning_rate": 8.284023668639055e-07,
+      "loss": 0.5501,
+      "step": 621
+    },
+    {
+      "epoch": 0.9201183431952663,
+      "grad_norm": 0.3046875,
+      "learning_rate": 8.136094674556213e-07,
+      "loss": 0.5633,
+      "step": 622
+    },
+    {
+      "epoch": 0.9215976331360947,
+      "grad_norm": 0.310546875,
+      "learning_rate": 7.988165680473374e-07,
+      "loss": 0.5796,
+      "step": 623
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.30078125,
+      "learning_rate": 7.840236686390533e-07,
+      "loss": 0.5626,
+      "step": 624
+    },
+    {
+      "epoch": 0.9245562130177515,
+      "grad_norm": 0.296875,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.5577,
+      "step": 625
+    },
+    {
+      "epoch": 0.9260355029585798,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.544378698224853e-07,
+      "loss": 0.5671,
+      "step": 626
+    },
+    {
+      "epoch": 0.9275147928994083,
+      "grad_norm": 0.310546875,
+      "learning_rate": 7.396449704142013e-07,
+      "loss": 0.5583,
+      "step": 627
+    },
+    {
+      "epoch": 0.9289940828402367,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.248520710059172e-07,
+      "loss": 0.5706,
+      "step": 628
+    },
+    {
+      "epoch": 0.9304733727810651,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.100591715976332e-07,
+      "loss": 0.5668,
+      "step": 629
+    },
+    {
+      "epoch": 0.9319526627218935,
+      "grad_norm": 0.333984375,
+      "learning_rate": 6.952662721893491e-07,
+      "loss": 0.5597,
+      "step": 630
+    },
+    {
+      "epoch": 0.9334319526627219,
+      "grad_norm": 0.30859375,
+      "learning_rate": 6.804733727810651e-07,
+      "loss": 0.5494,
+      "step": 631
+    },
+    {
+      "epoch": 0.9349112426035503,
+      "grad_norm": 0.31640625,
+      "learning_rate": 6.656804733727811e-07,
+      "loss": 0.5948,
+      "step": 632
+    },
+    {
+      "epoch": 0.9363905325443787,
+      "grad_norm": 0.375,
+      "learning_rate": 6.50887573964497e-07,
+      "loss": 0.5771,
+      "step": 633
+    },
+    {
+      "epoch": 0.9378698224852071,
+      "grad_norm": 0.30859375,
+      "learning_rate": 6.360946745562132e-07,
+      "loss": 0.5567,
+      "step": 634
+    },
+    {
+      "epoch": 0.9393491124260355,
+      "grad_norm": 0.298828125,
+      "learning_rate": 6.21301775147929e-07,
+      "loss": 0.5566,
+      "step": 635
+    },
+    {
+      "epoch": 0.9408284023668639,
+      "grad_norm": 0.359375,
+      "learning_rate": 6.06508875739645e-07,
+      "loss": 0.565,
+      "step": 636
+    },
+    {
+      "epoch": 0.9423076923076923,
+      "grad_norm": 0.3046875,
+      "learning_rate": 5.91715976331361e-07,
+      "loss": 0.5601,
+      "step": 637
+    },
+    {
+      "epoch": 0.9437869822485208,
+      "grad_norm": 0.3125,
+      "learning_rate": 5.76923076923077e-07,
+      "loss": 0.5559,
+      "step": 638
+    },
+    {
+      "epoch": 0.9452662721893491,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.621301775147929e-07,
+      "loss": 0.571,
+      "step": 639
+    },
+    {
+      "epoch": 0.9467455621301775,
+      "grad_norm": 0.302734375,
+      "learning_rate": 5.473372781065089e-07,
+      "loss": 0.5667,
+      "step": 640
+    },
+    {
+      "epoch": 0.9482248520710059,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.32544378698225e-07,
+      "loss": 0.5524,
+      "step": 641
+    },
+    {
+      "epoch": 0.9497041420118343,
+      "grad_norm": 0.3125,
+      "learning_rate": 5.177514792899409e-07,
+      "loss": 0.5736,
+      "step": 642
+    },
+    {
+      "epoch": 0.9511834319526628,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.029585798816569e-07,
+      "loss": 0.5692,
+      "step": 643
+    },
+    {
+      "epoch": 0.9526627218934911,
+      "grad_norm": 0.3671875,
+      "learning_rate": 4.881656804733728e-07,
+      "loss": 0.5535,
+      "step": 644
+    },
+    {
+      "epoch": 0.9541420118343196,
+      "grad_norm": 0.30078125,
+      "learning_rate": 4.733727810650888e-07,
+      "loss": 0.5674,
+      "step": 645
+    },
+    {
+      "epoch": 0.9556213017751479,
+      "grad_norm": 0.326171875,
+      "learning_rate": 4.5857988165680477e-07,
+      "loss": 0.5621,
+      "step": 646
+    },
+    {
+      "epoch": 0.9571005917159763,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.4378698224852073e-07,
+      "loss": 0.5684,
+      "step": 647
+    },
+    {
+      "epoch": 0.9585798816568047,
+      "grad_norm": 0.314453125,
+      "learning_rate": 4.289940828402367e-07,
+      "loss": 0.5605,
+      "step": 648
+    },
+    {
+      "epoch": 0.9600591715976331,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.1420118343195276e-07,
+      "loss": 0.55,
+      "step": 649
+    },
+    {
+      "epoch": 0.9615384615384616,
+      "grad_norm": 0.302734375,
+      "learning_rate": 3.994082840236687e-07,
+      "loss": 0.5747,
+      "step": 650
+    },
+    {
+      "epoch": 0.9630177514792899,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 0.5614,
+      "step": 651
+    },
+    {
+      "epoch": 0.9644970414201184,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.6982248520710064e-07,
+      "loss": 0.5627,
+      "step": 652
+    },
+    {
+      "epoch": 0.9659763313609467,
+      "grad_norm": 0.322265625,
+      "learning_rate": 3.550295857988166e-07,
+      "loss": 0.5524,
+      "step": 653
+    },
+    {
+      "epoch": 0.9674556213017751,
+      "grad_norm": 0.30859375,
+      "learning_rate": 3.4023668639053256e-07,
+      "loss": 0.5469,
+      "step": 654
+    },
+    {
+      "epoch": 0.9689349112426036,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.254437869822485e-07,
+      "loss": 0.5692,
+      "step": 655
+    },
+    {
+      "epoch": 0.9704142011834319,
+      "grad_norm": 0.314453125,
+      "learning_rate": 3.106508875739645e-07,
+      "loss": 0.5611,
+      "step": 656
+    },
+    {
+      "epoch": 0.9718934911242604,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.958579881656805e-07,
+      "loss": 0.5744,
+      "step": 657
+    },
+    {
+      "epoch": 0.9733727810650887,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.8106508875739646e-07,
+      "loss": 0.5563,
+      "step": 658
+    },
+    {
+      "epoch": 0.9748520710059172,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.662721893491125e-07,
+      "loss": 0.5503,
+      "step": 659
+    },
+    {
+      "epoch": 0.9763313609467456,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.5147928994082844e-07,
+      "loss": 0.5669,
+      "step": 660
+    },
+    {
+      "epoch": 0.977810650887574,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.366863905325444e-07,
+      "loss": 0.5935,
+      "step": 661
+    },
+    {
+      "epoch": 0.9792899408284024,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.2189349112426036e-07,
+      "loss": 0.5609,
+      "step": 662
+    },
+    {
+      "epoch": 0.9807692307692307,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.0710059171597638e-07,
+      "loss": 0.5684,
+      "step": 663
+    },
+    {
+      "epoch": 0.9822485207100592,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.9230769230769234e-07,
+      "loss": 0.5626,
+      "step": 664
+    },
+    {
+      "epoch": 0.9837278106508875,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.775147928994083e-07,
+      "loss": 0.5485,
+      "step": 665
+    },
+    {
+      "epoch": 0.985207100591716,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.6272189349112426e-07,
+      "loss": 0.5812,
+      "step": 666
+    },
+    {
+      "epoch": 0.9866863905325444,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.4792899408284025e-07,
+      "loss": 0.5705,
+      "step": 667
+    },
+    {
+      "epoch": 0.9881656804733728,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3313609467455624e-07,
+      "loss": 0.5608,
+      "step": 668
+    },
+    {
+      "epoch": 0.9896449704142012,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.183431952662722e-07,
+      "loss": 0.5619,
+      "step": 669
+    },
+    {
+      "epoch": 0.9911242603550295,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.0355029585798819e-07,
+      "loss": 0.5583,
+      "step": 670
+    },
+    {
+      "epoch": 0.992603550295858,
+      "grad_norm": 0.30078125,
+      "learning_rate": 8.875739644970415e-08,
+      "loss": 0.5623,
+      "step": 671
+    },
+    {
+      "epoch": 0.9940828402366864,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.396449704142013e-08,
+      "loss": 0.5511,
+      "step": 672
+    },
+    {
+      "epoch": 0.9955621301775148,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.91715976331361e-08,
+      "loss": 0.5555,
+      "step": 673
+    },
+    {
+      "epoch": 0.9970414201183432,
+      "grad_norm": 0.32421875,
+      "learning_rate": 4.4378698224852075e-08,
+      "loss": 0.5655,
+      "step": 674
+    },
+    {
+      "epoch": 0.9985207100591716,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.958579881656805e-08,
+      "loss": 0.5678,
+      "step": 675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4792899408284025e-08,
+      "loss": 0.5862,
+      "step": 676
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 676,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.2142270255839314e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}