diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10493 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5011305585796835,
+  "eval_steps": 500,
+  "global_step": 1496,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3170796398874254,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.7671,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.29680852396718344,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.6668,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2873156967491437,
+      "learning_rate": 3e-06,
+      "loss": 2.6609,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.32279830604138887,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.6938,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3053378475916063,
+      "learning_rate": 5e-06,
+      "loss": 2.7157,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.31356199532358553,
+      "learning_rate": 6e-06,
+      "loss": 2.6557,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.31222297787622016,
+      "learning_rate": 7e-06,
+      "loss": 2.7756,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.31394414996683867,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.6484,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.30526319548839004,
+      "learning_rate": 9e-06,
+      "loss": 2.6737,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2841622783139327,
+      "learning_rate": 1e-05,
+      "loss": 2.6651,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3046154890794536,
+      "learning_rate": 9.999997212173297e-06,
+      "loss": 2.6241,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3035503343057213,
+      "learning_rate": 9.9999888486963e-06,
+      "loss": 2.6725,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.32373433520739703,
+      "learning_rate": 9.999974909578329e-06,
+      "loss": 2.6226,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.31263516602351327,
+      "learning_rate": 9.999955394834935e-06,
+      "loss": 2.6546,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3200414604098111,
+      "learning_rate": 9.999930304487874e-06,
+      "loss": 2.703,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2733658015391629,
+      "learning_rate": 9.99989963856513e-06,
+      "loss": 2.6281,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.29070953298516844,
+      "learning_rate": 9.999863397100895e-06,
+      "loss": 2.6899,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2688747648684131,
+      "learning_rate": 9.999821580135584e-06,
+      "loss": 2.6487,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.30758393123030825,
+      "learning_rate": 9.999774187715831e-06,
+      "loss": 2.7497,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2828363345308586,
+      "learning_rate": 9.999721219894482e-06,
+      "loss": 2.5763,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2663942725881501,
+      "learning_rate": 9.999662676730606e-06,
+      "loss": 2.6937,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2689492534938777,
+      "learning_rate": 9.999598558289483e-06,
+      "loss": 2.6607,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2721982530512938,
+      "learning_rate": 9.999528864642613e-06,
+      "loss": 2.6965,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.30924763996084126,
+      "learning_rate": 9.999453595867717e-06,
+      "loss": 2.7285,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.31069570849542394,
+      "learning_rate": 9.999372752048729e-06,
+      "loss": 2.7327,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28105572772546,
+      "learning_rate": 9.999286333275797e-06,
+      "loss": 2.7391,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.29124281802756785,
+      "learning_rate": 9.999194339645292e-06,
+      "loss": 2.736,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28911610699083024,
+      "learning_rate": 9.999096771259797e-06,
+      "loss": 2.7244,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3218998422572818,
+      "learning_rate": 9.998993628228117e-06,
+      "loss": 2.6511,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.29829283998012357,
+      "learning_rate": 9.998884910665267e-06,
+      "loss": 2.6505,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2871946091545111,
+      "learning_rate": 9.998770618692483e-06,
+      "loss": 2.672,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2921996511900815,
+      "learning_rate": 9.998650752437214e-06,
+      "loss": 2.7574,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2882670075948252,
+      "learning_rate": 9.998525312033127e-06,
+      "loss": 2.6553,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.29239646966364624,
+      "learning_rate": 9.998394297620104e-06,
+      "loss": 2.7093,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.27417884873271275,
+      "learning_rate": 9.998257709344246e-06,
+      "loss": 2.7521,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2841800938763941,
+      "learning_rate": 9.998115547357862e-06,
+      "loss": 2.6787,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2856884783984889,
+      "learning_rate": 9.997967811819485e-06,
+      "loss": 2.6779,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2683793206277986,
+      "learning_rate": 9.997814502893857e-06,
+      "loss": 2.6535,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26118210674650083,
+      "learning_rate": 9.997655620751939e-06,
+      "loss": 2.543,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.284127574743679,
+      "learning_rate": 9.997491165570907e-06,
+      "loss": 2.6393,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.28201563255626655,
+      "learning_rate": 9.997321137534146e-06,
+      "loss": 2.6337,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.26033949026291303,
+      "learning_rate": 9.997145536831262e-06,
+      "loss": 2.6164,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2988779986281759,
+      "learning_rate": 9.996964363658072e-06,
+      "loss": 2.5932,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.2759490337531855,
+      "learning_rate": 9.996777618216608e-06,
+      "loss": 2.684,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28293313936126274,
+      "learning_rate": 9.996585300715117e-06,
+      "loss": 2.6431,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2847363633557033,
+      "learning_rate": 9.996387411368053e-06,
+      "loss": 2.6603,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28381392284071094,
+      "learning_rate": 9.996183950396094e-06,
+      "loss": 2.6464,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2916519709131871,
+      "learning_rate": 9.995974918026125e-06,
+      "loss": 2.6885,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2700571651342325,
+      "learning_rate": 9.995760314491243e-06,
+      "loss": 2.6409,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.277172183309257,
+      "learning_rate": 9.995540140030759e-06,
+      "loss": 2.6321,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2721933123803388,
+      "learning_rate": 9.995314394890196e-06,
+      "loss": 2.6912,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2859308096097758,
+      "learning_rate": 9.99508307932129e-06,
+      "loss": 2.692,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2988143743206093,
+      "learning_rate": 9.99484619358199e-06,
+      "loss": 2.6676,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.287084630302098,
+      "learning_rate": 9.994603737936448e-06,
+      "loss": 2.6356,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.28178162133982554,
+      "learning_rate": 9.99435571265504e-06,
+      "loss": 2.5701,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26758913727535316,
+      "learning_rate": 9.994102118014346e-06,
+      "loss": 2.5164,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27606618820293444,
+      "learning_rate": 9.993842954297155e-06,
+      "loss": 2.5834,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.27242713636520005,
+      "learning_rate": 9.993578221792469e-06,
+      "loss": 2.7215,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2839211405332648,
+      "learning_rate": 9.9933079207955e-06,
+      "loss": 2.635,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2685892394986992,
+      "learning_rate": 9.99303205160767e-06,
+      "loss": 2.7163,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2697296518070164,
+      "learning_rate": 9.992750614536606e-06,
+      "loss": 2.5452,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2854401338053572,
+      "learning_rate": 9.99246360989615e-06,
+      "loss": 2.5883,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29011457596920015,
+      "learning_rate": 9.992171038006346e-06,
+      "loss": 2.6617,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2712048800249517,
+      "learning_rate": 9.991872899193455e-06,
+      "loss": 2.6344,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29054941045481053,
+      "learning_rate": 9.991569193789938e-06,
+      "loss": 2.7087,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26950023306799387,
+      "learning_rate": 9.991259922134466e-06,
+      "loss": 2.6186,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2612400944442634,
+      "learning_rate": 9.99094508457192e-06,
+      "loss": 2.6211,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2714769376106506,
+      "learning_rate": 9.99062468145338e-06,
+      "loss": 2.708,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2896534832484691,
+      "learning_rate": 9.99029871313614e-06,
+      "loss": 2.6488,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2734636975408946,
+      "learning_rate": 9.989967179983699e-06,
+      "loss": 2.617,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.29347749407064544,
+      "learning_rate": 9.989630082365757e-06,
+      "loss": 2.5147,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.26482740031050994,
+      "learning_rate": 9.989287420658224e-06,
+      "loss": 2.5581,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.2795723576861998,
+      "learning_rate": 9.98893919524321e-06,
+      "loss": 2.5844,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3024349947315377,
+      "learning_rate": 9.988585406509035e-06,
+      "loss": 2.6604,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3063626335259863,
+      "learning_rate": 9.988226054850218e-06,
+      "loss": 2.5631,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29619182230702906,
+      "learning_rate": 9.987861140667483e-06,
+      "loss": 2.7107,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2863355575783403,
+      "learning_rate": 9.987490664367758e-06,
+      "loss": 2.7283,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29317077029265626,
+      "learning_rate": 9.987114626364172e-06,
+      "loss": 2.6591,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27596698806471515,
+      "learning_rate": 9.986733027076054e-06,
+      "loss": 2.6256,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.27656717781378126,
+      "learning_rate": 9.98634586692894e-06,
+      "loss": 2.656,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28757181359132244,
+      "learning_rate": 9.985953146354566e-06,
+      "loss": 2.7137,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2853580962365858,
+      "learning_rate": 9.985554865790862e-06,
+      "loss": 2.7061,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29503022029015374,
+      "learning_rate": 9.985151025681964e-06,
+      "loss": 2.6384,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28368753243458983,
+      "learning_rate": 9.98474162647821e-06,
+      "loss": 2.6624,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28280698760335077,
+      "learning_rate": 9.984326668636131e-06,
+      "loss": 2.6416,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2860258768386144,
+      "learning_rate": 9.983906152618458e-06,
+      "loss": 2.5637,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.284560220784243,
+      "learning_rate": 9.983480078894125e-06,
+      "loss": 2.6423,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30396814279935847,
+      "learning_rate": 9.983048447938255e-06,
+      "loss": 2.8193,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2767383522832132,
+      "learning_rate": 9.982611260232177e-06,
+      "loss": 2.6299,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2930929175146555,
+      "learning_rate": 9.98216851626341e-06,
+      "loss": 2.5475,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30608721318647636,
+      "learning_rate": 9.981720216525675e-06,
+      "loss": 2.5872,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3003738914561364,
+      "learning_rate": 9.98126636151888e-06,
+      "loss": 2.5907,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3013039295708162,
+      "learning_rate": 9.980806951749134e-06,
+      "loss": 2.7558,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3142726332807136,
+      "learning_rate": 9.98034198772874e-06,
+      "loss": 2.5396,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29862717497789243,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 2.6472,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28702408957648073,
+      "learning_rate": 9.979395399016186e-06,
+      "loss": 2.6004,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3240509173310664,
+      "learning_rate": 9.978913775379595e-06,
+      "loss": 2.6406,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28422751794327694,
+      "learning_rate": 9.978426599603492e-06,
+      "loss": 2.6116,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29107866308539426,
+      "learning_rate": 9.977933872231148e-06,
+      "loss": 2.7328,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.2952731989387591,
+      "learning_rate": 9.977435593812013e-06,
+      "loss": 2.6816,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.29516315064408405,
+      "learning_rate": 9.976931764901733e-06,
+      "loss": 2.6957,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.30826242985921637,
+      "learning_rate": 9.976422386062147e-06,
+      "loss": 2.6688,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.31006008358094334,
+      "learning_rate": 9.975907457861276e-06,
+      "loss": 2.6719,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3066420585453031,
+      "learning_rate": 9.975386980873332e-06,
+      "loss": 2.7008,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2927576957883402,
+      "learning_rate": 9.974860955678715e-06,
+      "loss": 2.6419,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29451095298628577,
+      "learning_rate": 9.974329382864013e-06,
+      "loss": 2.5173,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2950560536107638,
+      "learning_rate": 9.973792263021998e-06,
+      "loss": 2.5954,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30334538972041447,
+      "learning_rate": 9.97324959675163e-06,
+      "loss": 2.698,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.295569951999041,
+      "learning_rate": 9.972701384658052e-06,
+      "loss": 2.6575,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2779283004801193,
+      "learning_rate": 9.972147627352593e-06,
+      "loss": 2.5847,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29050023569534117,
+      "learning_rate": 9.971588325452764e-06,
+      "loss": 2.632,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28944011458219093,
+      "learning_rate": 9.971023479582258e-06,
+      "loss": 2.6059,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.30361043855717296,
+      "learning_rate": 9.970453090370955e-06,
+      "loss": 2.6696,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2903070720744162,
+      "learning_rate": 9.969877158454912e-06,
+      "loss": 2.6237,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2947711559457182,
+      "learning_rate": 9.96929568447637e-06,
+      "loss": 2.5382,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3135860050045197,
+      "learning_rate": 9.968708669083744e-06,
+      "loss": 2.4986,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28678359148213384,
+      "learning_rate": 9.96811611293164e-06,
+      "loss": 2.4735,
+      "step": 117
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2916898789138187,
+      "learning_rate": 9.967518016680828e-06,
+      "loss": 2.6435,
+      "step": 118
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2972152291286803,
+      "learning_rate": 9.966914380998267e-06,
+      "loss": 2.6284,
+      "step": 119
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3188049030703794,
+      "learning_rate": 9.966305206557092e-06,
+      "loss": 2.6791,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2735280976751751,
+      "learning_rate": 9.965690494036608e-06,
+      "loss": 2.5355,
+      "step": 121
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3085365242350297,
+      "learning_rate": 9.965070244122302e-06,
+      "loss": 2.6272,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28189250618902306,
+      "learning_rate": 9.964444457505833e-06,
+      "loss": 2.6136,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2909831086496679,
+      "learning_rate": 9.963813134885035e-06,
+      "loss": 2.5946,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.32772962273481576,
+      "learning_rate": 9.963176276963916e-06,
+      "loss": 2.5512,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29674052656232275,
+      "learning_rate": 9.962533884452654e-06,
+      "loss": 2.6141,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.32953053258652276,
+      "learning_rate": 9.961885958067604e-06,
+      "loss": 2.5961,
+      "step": 127
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.296659441540278,
+      "learning_rate": 9.961232498531283e-06,
+      "loss": 2.5345,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.29525949719018313,
+      "learning_rate": 9.960573506572391e-06,
+      "loss": 2.6226,
+      "step": 129
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3157983285097481,
+      "learning_rate": 9.959908982925783e-06,
+      "loss": 2.7319,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.28286217521963397,
+      "learning_rate": 9.959238928332494e-06,
+      "loss": 2.7099,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2989054935005055,
+      "learning_rate": 9.958563343539723e-06,
+      "loss": 2.5254,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31605549196587157,
+      "learning_rate": 9.95788222930083e-06,
+      "loss": 2.63,
+      "step": 133
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.2942778521800686,
+      "learning_rate": 9.957195586375352e-06,
+      "loss": 2.6654,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2841294889913612,
+      "learning_rate": 9.956503415528984e-06,
+      "loss": 2.5773,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33126003232194934,
+      "learning_rate": 9.955805717533586e-06,
+      "loss": 2.4466,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29995460365464693,
+      "learning_rate": 9.955102493167184e-06,
+      "loss": 2.6908,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3097272168432938,
+      "learning_rate": 9.954393743213965e-06,
+      "loss": 2.6364,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29472319118829043,
+      "learning_rate": 9.953679468464274e-06,
+      "loss": 2.6355,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30533560268177895,
+      "learning_rate": 9.952959669714627e-06,
+      "loss": 2.642,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30490767031205196,
+      "learning_rate": 9.952234347767688e-06,
+      "loss": 2.625,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30108111160509854,
+      "learning_rate": 9.95150350343229e-06,
+      "loss": 2.5336,
+      "step": 142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35556322456241785,
+      "learning_rate": 9.950767137523417e-06,
+      "loss": 2.532,
+      "step": 143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2878834805061441,
+      "learning_rate": 9.950025250862214e-06,
+      "loss": 2.5405,
+      "step": 144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3164131493463764,
+      "learning_rate": 9.94927784427598e-06,
+      "loss": 2.6325,
+      "step": 145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3003951728387244,
+      "learning_rate": 9.948524918598175e-06,
+      "loss": 2.6077,
+      "step": 146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3006003690530792,
+      "learning_rate": 9.947766474668407e-06,
+      "loss": 2.6045,
+      "step": 147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3033727253017958,
+      "learning_rate": 9.947002513332441e-06,
+      "loss": 2.6125,
+      "step": 148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.28736676466741445,
+      "learning_rate": 9.946233035442191e-06,
+      "loss": 2.598,
+      "step": 149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29270217211382965,
+      "learning_rate": 9.945458041855732e-06,
+      "loss": 2.5943,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3048173187700944,
+      "learning_rate": 9.944677533437274e-06,
+      "loss": 2.6109,
+      "step": 151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31512121937325066,
+      "learning_rate": 9.943891511057193e-06,
+      "loss": 2.5965,
+      "step": 152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32931439227554016,
+      "learning_rate": 9.943099975592003e-06,
+      "loss": 2.7177,
+      "step": 153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3125427693663409,
+      "learning_rate": 9.942302927924372e-06,
+      "loss": 2.7429,
+      "step": 154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31511491830731564,
+      "learning_rate": 9.941500368943111e-06,
+      "loss": 2.6877,
+      "step": 155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3110005885008245,
+      "learning_rate": 9.940692299543177e-06,
+      "loss": 2.6596,
+      "step": 156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33338215837921104,
+      "learning_rate": 9.939878720625675e-06,
+      "loss": 2.574,
+      "step": 157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3057233369045081,
+      "learning_rate": 9.939059633097852e-06,
+      "loss": 2.6993,
+      "step": 158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3180182331769005,
+      "learning_rate": 9.938235037873094e-06,
+      "loss": 2.6828,
+      "step": 159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29025477598221844,
+      "learning_rate": 9.937404935870938e-06,
+      "loss": 2.6985,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2984082132718923,
+      "learning_rate": 9.936569328017052e-06,
+      "loss": 2.6061,
+      "step": 161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3062414255693518,
+      "learning_rate": 9.935728215243248e-06,
+      "loss": 2.7174,
+      "step": 162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3246654168330274,
+      "learning_rate": 9.934881598487478e-06,
+      "loss": 2.5767,
+      "step": 163
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3024015011695911,
+      "learning_rate": 9.934029478693832e-06,
+      "loss": 2.7215,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29285564782553547,
+      "learning_rate": 9.933171856812533e-06,
+      "loss": 2.6268,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30639638710780326,
+      "learning_rate": 9.93230873379994e-06,
+      "loss": 2.5962,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30924253228683957,
+      "learning_rate": 9.931440110618551e-06,
+      "loss": 2.5709,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3072053847437663,
+      "learning_rate": 9.930565988236992e-06,
+      "loss": 2.6555,
+      "step": 168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3092824759025904,
+      "learning_rate": 9.929686367630025e-06,
+      "loss": 2.5353,
+      "step": 169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2986363811791144,
+      "learning_rate": 9.92880124977854e-06,
+      "loss": 2.5062,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31033745188051043,
+      "learning_rate": 9.927910635669563e-06,
+      "loss": 2.4532,
+      "step": 171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31699134746635294,
+      "learning_rate": 9.92701452629624e-06,
+      "loss": 2.7133,
+      "step": 172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30793109197125706,
+      "learning_rate": 9.926112922657857e-06,
+      "loss": 2.6064,
+      "step": 173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31386250063346854,
+      "learning_rate": 9.925205825759813e-06,
+      "loss": 2.6756,
+      "step": 174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.320593710482426,
+      "learning_rate": 9.924293236613643e-06,
+      "loss": 2.702,
+      "step": 175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3156087384897636,
+      "learning_rate": 9.923375156237001e-06,
+      "loss": 2.6653,
+      "step": 176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31854517610852556,
+      "learning_rate": 9.922451585653668e-06,
+      "loss": 2.5921,
+      "step": 177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29132946424311973,
+      "learning_rate": 9.921522525893548e-06,
+      "loss": 2.5643,
+      "step": 178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3053782409475688,
+      "learning_rate": 9.92058797799266e-06,
+      "loss": 2.6313,
+      "step": 179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3212158908250996,
+      "learning_rate": 9.91964794299315e-06,
+      "loss": 2.5354,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31759359549627775,
+      "learning_rate": 9.918702421943279e-06,
+      "loss": 2.6066,
+      "step": 181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.35379468325614266,
+      "learning_rate": 9.917751415897424e-06,
+      "loss": 2.6648,
+      "step": 182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.309402095205423,
+      "learning_rate": 9.916794925916084e-06,
+      "loss": 2.6547,
+      "step": 183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3200272243437652,
+      "learning_rate": 9.915832953065872e-06,
+      "loss": 2.6266,
+      "step": 184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33823480881963486,
+      "learning_rate": 9.91486549841951e-06,
+      "loss": 2.5839,
+      "step": 185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30035503955484993,
+      "learning_rate": 9.913892563055838e-06,
+      "loss": 2.5857,
+      "step": 186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29683903426101027,
+      "learning_rate": 9.912914148059805e-06,
+      "loss": 2.5867,
+      "step": 187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3271230659325653,
+      "learning_rate": 9.91193025452247e-06,
+      "loss": 2.5397,
+      "step": 188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30470387096928925,
+      "learning_rate": 9.910940883541009e-06,
+      "loss": 2.5745,
+      "step": 189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31498884686525297,
+      "learning_rate": 9.909946036218694e-06,
+      "loss": 2.5992,
+      "step": 190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3418035670034064,
+      "learning_rate": 9.908945713664912e-06,
+      "loss": 2.6089,
+      "step": 191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34197561767453183,
+      "learning_rate": 9.907939916995154e-06,
+      "loss": 2.6431,
+      "step": 192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32498858871967207,
+      "learning_rate": 9.906928647331011e-06,
+      "loss": 2.6333,
+      "step": 193
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31437972497517663,
+      "learning_rate": 9.905911905800186e-06,
+      "loss": 2.6326,
+      "step": 194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3133510801919385,
+      "learning_rate": 9.904889693536475e-06,
+      "loss": 2.6042,
+      "step": 195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3507289329205512,
+      "learning_rate": 9.903862011679781e-06,
+      "loss": 2.6468,
+      "step": 196
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31804925054207894,
+      "learning_rate": 9.902828861376101e-06,
+      "loss": 2.5869,
+      "step": 197
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32612924811219846,
+      "learning_rate": 9.901790243777534e-06,
+      "loss": 2.5671,
+      "step": 198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3120656810058216,
+      "learning_rate": 9.900746160042273e-06,
+      "loss": 2.5271,
+      "step": 199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3010484944903533,
+      "learning_rate": 9.899696611334612e-06,
+      "loss": 2.6452,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3239853259961015,
+      "learning_rate": 9.89864159882493e-06,
+      "loss": 2.7043,
+      "step": 201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31146310204052247,
+      "learning_rate": 9.897581123689705e-06,
+      "loss": 2.6428,
+      "step": 202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32324911380964977,
+      "learning_rate": 9.896515187111508e-06,
+      "loss": 2.5365,
+      "step": 203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3035841060643951,
+      "learning_rate": 9.895443790278995e-06,
+      "loss": 2.5286,
+      "step": 204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3122895366531592,
+      "learning_rate": 9.894366934386913e-06,
+      "loss": 2.7649,
+      "step": 205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3139971157354563,
+      "learning_rate": 9.8932846206361e-06,
+      "loss": 2.5452,
+      "step": 206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32275803833132877,
+      "learning_rate": 9.892196850233474e-06,
+      "loss": 2.6,
+      "step": 207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31242104963540335,
+      "learning_rate": 9.891103624392042e-06,
+      "loss": 2.5098,
+      "step": 208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3068416325400618,
+      "learning_rate": 9.890004944330895e-06,
+      "loss": 2.5743,
+      "step": 209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4426076961648037,
+      "learning_rate": 9.888900811275205e-06,
+      "loss": 2.6019,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32038220368233306,
+      "learning_rate": 9.887791226456223e-06,
+      "loss": 2.6093,
+      "step": 211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3189093163589122,
+      "learning_rate": 9.886676191111283e-06,
+      "loss": 2.5176,
+      "step": 212
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2975276499008464,
+      "learning_rate": 9.885555706483792e-06,
+      "loss": 2.6061,
+      "step": 213
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3416785116972448,
+      "learning_rate": 9.884429773823238e-06,
+      "loss": 2.6776,
+      "step": 214
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3324805417498029,
+      "learning_rate": 9.883298394385186e-06,
+      "loss": 2.4792,
+      "step": 215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30942312887340606,
+      "learning_rate": 9.88216156943127e-06,
+      "loss": 2.636,
+      "step": 216
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3331045418589145,
+      "learning_rate": 9.881019300229196e-06,
+      "loss": 2.6341,
+      "step": 217
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.36589877238848295,
+      "learning_rate": 9.879871588052748e-06,
+      "loss": 2.4853,
+      "step": 218
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32143496183446163,
+      "learning_rate": 9.87871843418177e-06,
+      "loss": 2.6916,
+      "step": 219
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3367329207268536,
+      "learning_rate": 9.877559839902185e-06,
+      "loss": 2.6412,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32310923351581927,
+      "learning_rate": 9.876395806505972e-06,
+      "loss": 2.6126,
+      "step": 221
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3148934971054658,
+      "learning_rate": 9.875226335291181e-06,
+      "loss": 2.6381,
+      "step": 222
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3074710317085622,
+      "learning_rate": 9.874051427561929e-06,
+      "loss": 2.7002,
+      "step": 223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34735533732290175,
+      "learning_rate": 9.872871084628387e-06,
+      "loss": 2.6666,
+      "step": 224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3058697188419932,
+      "learning_rate": 9.871685307806796e-06,
+      "loss": 2.6453,
+      "step": 225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3263148148780988,
+      "learning_rate": 9.870494098419448e-06,
+      "loss": 2.5654,
+      "step": 226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32436238979164406,
+      "learning_rate": 9.869297457794698e-06,
+      "loss": 2.6129,
+      "step": 227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3428513310711877,
+      "learning_rate": 9.868095387266961e-06,
+      "loss": 2.5645,
+      "step": 228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3228143817148822,
+      "learning_rate": 9.866887888176697e-06,
+      "loss": 2.5796,
+      "step": 229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31185480509183455,
+      "learning_rate": 9.865674961870428e-06,
+      "loss": 2.5712,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33915296042223153,
+      "learning_rate": 9.864456609700726e-06,
+      "loss": 2.7518,
+      "step": 231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.30521825876252084,
+      "learning_rate": 9.86323283302621e-06,
+      "loss": 2.6188,
+      "step": 232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3312279239531816,
+      "learning_rate": 9.862003633211555e-06,
+      "loss": 2.6723,
+      "step": 233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3155802081211641,
+      "learning_rate": 9.860769011627476e-06,
+      "loss": 2.5617,
+      "step": 234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33185307827260807,
+      "learning_rate": 9.859528969650739e-06,
+      "loss": 2.5231,
+      "step": 235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32247823002252857,
+      "learning_rate": 9.858283508664153e-06,
+      "loss": 2.6013,
+      "step": 236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.36392929357978193,
+      "learning_rate": 9.857032630056569e-06,
+      "loss": 2.6207,
+      "step": 237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.35286768129497353,
+      "learning_rate": 9.85577633522288e-06,
+      "loss": 2.6654,
+      "step": 238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3451232141077575,
+      "learning_rate": 9.854514625564018e-06,
+      "loss": 2.5755,
+      "step": 239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.345284537223161,
+      "learning_rate": 9.853247502486957e-06,
+      "loss": 2.5712,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3135899846602162,
+      "learning_rate": 9.851974967404703e-06,
+      "loss": 2.6604,
+      "step": 241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.38832041890168123,
+      "learning_rate": 9.850697021736299e-06,
+      "loss": 2.6048,
+      "step": 242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4225646492630186,
+      "learning_rate": 9.849413666906824e-06,
+      "loss": 2.5395,
+      "step": 243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32642912856178463,
+      "learning_rate": 9.84812490434738e-06,
+      "loss": 2.6747,
+      "step": 244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31796660504566643,
+      "learning_rate": 9.846830735495112e-06,
+      "loss": 2.5409,
+      "step": 245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.35431307216240776,
+      "learning_rate": 9.845531161793185e-06,
+      "loss": 2.6876,
+      "step": 246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3346496670529943,
+      "learning_rate": 9.844226184690793e-06,
+      "loss": 2.6058,
+      "step": 247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.32803919215153093,
+      "learning_rate": 9.842915805643156e-06,
+      "loss": 2.6223,
+      "step": 248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3174652696747853,
+      "learning_rate": 9.84160002611152e-06,
+      "loss": 2.5318,
+      "step": 249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3464534397356932,
+      "learning_rate": 9.840278847563147e-06,
+      "loss": 2.6256,
+      "step": 250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3603315063519056,
+      "learning_rate": 9.83895227147133e-06,
+      "loss": 2.5634,
+      "step": 251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.37082730817007653,
+      "learning_rate": 9.837620299315367e-06,
+      "loss": 2.5691,
+      "step": 252
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3204057604236187,
+      "learning_rate": 9.836282932580588e-06,
+      "loss": 2.6175,
+      "step": 253
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3185505322361351,
+      "learning_rate": 9.834940172758327e-06,
+      "loss": 2.6166,
+      "step": 254
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3384771933662536,
+      "learning_rate": 9.833592021345938e-06,
+      "loss": 2.6049,
+      "step": 255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3593085061629509,
+      "learning_rate": 9.832238479846788e-06,
+      "loss": 2.7394,
+      "step": 256
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3396962602147287,
+      "learning_rate": 9.83087954977025e-06,
+      "loss": 2.5878,
+      "step": 257
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33152125965893986,
+      "learning_rate": 9.82951523263171e-06,
+      "loss": 2.6987,
+      "step": 258
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.328453399171747,
+      "learning_rate": 9.828145529952558e-06,
+      "loss": 2.6115,
+      "step": 259
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3257263132699668,
+      "learning_rate": 9.826770443260193e-06,
+      "loss": 2.5727,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3215557023897131,
+      "learning_rate": 9.825389974088019e-06,
+      "loss": 2.6047,
+      "step": 261
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33394984016557416,
+      "learning_rate": 9.824004123975434e-06,
+      "loss": 2.5551,
+      "step": 262
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3332665073444441,
+      "learning_rate": 9.822612894467847e-06,
+      "loss": 2.6769,
+      "step": 263
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31294573942582393,
+      "learning_rate": 9.821216287116658e-06,
+      "loss": 2.6919,
+      "step": 264
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.36571617782823884,
+      "learning_rate": 9.819814303479268e-06,
+      "loss": 2.6413,
+      "step": 265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3238280366499427,
+      "learning_rate": 9.81840694511907e-06,
+      "loss": 2.5427,
+      "step": 266
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3252099752700349,
+      "learning_rate": 9.816994213605453e-06,
+      "loss": 2.5705,
+      "step": 267
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32854133260432167,
+      "learning_rate": 9.8155761105138e-06,
+      "loss": 2.6458,
+      "step": 268
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.37397526762013245,
+      "learning_rate": 9.814152637425478e-06,
+      "loss": 2.6586,
+      "step": 269
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3376369075010002,
+      "learning_rate": 9.812723795927848e-06,
+      "loss": 2.5644,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3446187615353997,
+      "learning_rate": 9.811289587614252e-06,
+      "loss": 2.5545,
+      "step": 271
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3244205903665975,
+      "learning_rate": 9.809850014084022e-06,
+      "loss": 2.5556,
+      "step": 272
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.337581890379183,
+      "learning_rate": 9.808405076942472e-06,
+      "loss": 2.6887,
+      "step": 273
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33290044670235436,
+      "learning_rate": 9.806954777800891e-06,
+      "loss": 2.6148,
+      "step": 274
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3186843594113733,
+      "learning_rate": 9.805499118276555e-06,
+      "loss": 2.6586,
+      "step": 275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3361466444273861,
+      "learning_rate": 9.804038099992716e-06,
+      "loss": 2.611,
+      "step": 276
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3402845987243979,
+      "learning_rate": 9.8025717245786e-06,
+      "loss": 2.5221,
+      "step": 277
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3417382980288246,
+      "learning_rate": 9.801099993669403e-06,
+      "loss": 2.6768,
+      "step": 278
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3719689739322445,
+      "learning_rate": 9.7996229089063e-06,
+      "loss": 2.6276,
+      "step": 279
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3350098243383065,
+      "learning_rate": 9.798140471936437e-06,
+      "loss": 2.5812,
+      "step": 280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3365657498619843,
+      "learning_rate": 9.79665268441292e-06,
+      "loss": 2.7137,
+      "step": 281
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3517854831112496,
+      "learning_rate": 9.79515954799483e-06,
+      "loss": 2.7186,
+      "step": 282
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3609844532271263,
+      "learning_rate": 9.793661064347205e-06,
+      "loss": 2.619,
+      "step": 283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3440434547093879,
+      "learning_rate": 9.792157235141054e-06,
+      "loss": 2.6764,
+      "step": 284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34983063201707953,
+      "learning_rate": 9.790648062053341e-06,
+      "loss": 2.5175,
+      "step": 285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3400967412647235,
+      "learning_rate": 9.789133546766992e-06,
+      "loss": 2.7174,
+      "step": 286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34364284665946243,
+      "learning_rate": 9.787613690970889e-06,
+      "loss": 2.5042,
+      "step": 287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3193387372655362,
+      "learning_rate": 9.78608849635987e-06,
+      "loss": 2.6363,
+      "step": 288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3548903947858929,
+      "learning_rate": 9.784557964634729e-06,
+      "loss": 2.6435,
+      "step": 289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.352340261776187,
+      "learning_rate": 9.783022097502204e-06,
+      "loss": 2.5075,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3219062651106595,
+      "learning_rate": 9.78148089667499e-06,
+      "loss": 2.4898,
+      "step": 291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.447074149512361,
+      "learning_rate": 9.77993436387173e-06,
+      "loss": 2.6164,
+      "step": 292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32502987522582333,
+      "learning_rate": 9.778382500817006e-06,
+      "loss": 2.6231,
+      "step": 293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.35790774148123833,
+      "learning_rate": 9.776825309241347e-06,
+      "loss": 2.572,
+      "step": 294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33509687106667574,
+      "learning_rate": 9.77526279088123e-06,
+      "loss": 2.5331,
+      "step": 295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3425946543760854,
+      "learning_rate": 9.773694947479064e-06,
+      "loss": 2.5942,
+      "step": 296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3611654791904134,
+      "learning_rate": 9.772121780783202e-06,
+      "loss": 2.6132,
+      "step": 297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3333364079254133,
+      "learning_rate": 9.770543292547928e-06,
+      "loss": 2.6903,
+      "step": 298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3496197107588038,
+      "learning_rate": 9.768959484533461e-06,
+      "loss": 2.6104,
+      "step": 299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3482956524862094,
+      "learning_rate": 9.767370358505958e-06,
+      "loss": 2.5578,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3540980562435214,
+      "learning_rate": 9.7657759162375e-06,
+      "loss": 2.6823,
+      "step": 301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3378288565389089,
+      "learning_rate": 9.764176159506097e-06,
+      "loss": 2.6466,
+      "step": 302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3476414623758776,
+      "learning_rate": 9.762571090095692e-06,
+      "loss": 2.6817,
+      "step": 303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3468363770470217,
+      "learning_rate": 9.760960709796141e-06,
+      "loss": 2.7139,
+      "step": 304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32968773412470165,
+      "learning_rate": 9.759345020403233e-06,
+      "loss": 2.6294,
+      "step": 305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32461622302088416,
+      "learning_rate": 9.757724023718671e-06,
+      "loss": 2.5907,
+      "step": 306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32724000019534877,
+      "learning_rate": 9.756097721550078e-06,
+      "loss": 2.5887,
+      "step": 307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.31378490221572475,
+      "learning_rate": 9.754466115710992e-06,
+      "loss": 2.6506,
+      "step": 308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3389812794104657,
+      "learning_rate": 9.75282920802087e-06,
+      "loss": 2.6872,
+      "step": 309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3410772104137017,
+      "learning_rate": 9.751187000305076e-06,
+      "loss": 2.5604,
+      "step": 310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32674524313857833,
+      "learning_rate": 9.749539494394887e-06,
+      "loss": 2.538,
+      "step": 311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.33858326387385124,
+      "learning_rate": 9.747886692127486e-06,
+      "loss": 2.6238,
+      "step": 312
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.34589766292819163,
+      "learning_rate": 9.746228595345965e-06,
+      "loss": 2.6378,
+      "step": 313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.35188753149779034,
+      "learning_rate": 9.744565205899317e-06,
+      "loss": 2.5601,
+      "step": 314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31652413044373107,
+      "learning_rate": 9.742896525642442e-06,
+      "loss": 2.5816,
+      "step": 315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3409957215748307,
+      "learning_rate": 9.741222556436132e-06,
+      "loss": 2.6467,
+      "step": 316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3960835352075213,
+      "learning_rate": 9.739543300147085e-06,
+      "loss": 2.665,
+      "step": 317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3287165509275585,
+      "learning_rate": 9.737858758647889e-06,
+      "loss": 2.6474,
+      "step": 318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3798162995387399,
+      "learning_rate": 9.73616893381703e-06,
+      "loss": 2.61,
+      "step": 319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3342253836017593,
+      "learning_rate": 9.734473827538881e-06,
+      "loss": 2.6278,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3555154455346524,
+      "learning_rate": 9.73277344170371e-06,
+      "loss": 2.6153,
+      "step": 321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4386835777780206,
+      "learning_rate": 9.731067778207665e-06,
+      "loss": 2.6937,
+      "step": 322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.36322755314204486,
+      "learning_rate": 9.729356838952788e-06,
+      "loss": 2.6125,
+      "step": 323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3824261253050603,
+      "learning_rate": 9.727640625847e-06,
+      "loss": 2.6399,
+      "step": 324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3561152981527507,
+      "learning_rate": 9.7259191408041e-06,
+      "loss": 2.6995,
+      "step": 325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31872571450635934,
+      "learning_rate": 9.724192385743767e-06,
+      "loss": 2.5594,
+      "step": 326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3456621738468028,
+      "learning_rate": 9.722460362591565e-06,
+      "loss": 2.6167,
+      "step": 327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31936564875982554,
+      "learning_rate": 9.720723073278922e-06,
+      "loss": 2.6844,
+      "step": 328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34576928319275374,
+      "learning_rate": 9.718980519743142e-06,
+      "loss": 2.6024,
+      "step": 329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34636170754314316,
+      "learning_rate": 9.717232703927402e-06,
+      "loss": 2.5263,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.364738168081352,
+      "learning_rate": 9.715479627780744e-06,
+      "loss": 2.6529,
+      "step": 331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33147850027819137,
+      "learning_rate": 9.713721293258079e-06,
+      "loss": 2.5856,
+      "step": 332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.36853607567398494,
+      "learning_rate": 9.711957702320176e-06,
+      "loss": 2.6117,
+      "step": 333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.35143052909025574,
+      "learning_rate": 9.710188856933671e-06,
+      "loss": 2.5388,
+      "step": 334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.346783453227663,
+      "learning_rate": 9.70841475907106e-06,
+      "loss": 2.5964,
+      "step": 335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3547742157857605,
+      "learning_rate": 9.70663541071069e-06,
+      "loss": 2.6319,
+      "step": 336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3767870601456973,
+      "learning_rate": 9.70485081383677e-06,
+      "loss": 2.6695,
+      "step": 337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31099203580367296,
+      "learning_rate": 9.703060970439357e-06,
+      "loss": 2.5908,
+      "step": 338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3563365195434926,
+      "learning_rate": 9.701265882514362e-06,
+      "loss": 2.6798,
+      "step": 339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34726289934734256,
+      "learning_rate": 9.69946555206354e-06,
+      "loss": 2.5322,
+      "step": 340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3516286787687523,
+      "learning_rate": 9.697659981094498e-06,
+      "loss": 2.5811,
+      "step": 341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34894053451732737,
+      "learning_rate": 9.695849171620681e-06,
+      "loss": 2.6177,
+      "step": 342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33347703647479304,
+      "learning_rate": 9.694033125661382e-06,
+      "loss": 2.4915,
+      "step": 343
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3762769415705139,
+      "learning_rate": 9.692211845241724e-06,
+      "loss": 2.6065,
+      "step": 344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3542465956570787,
+      "learning_rate": 9.690385332392676e-06,
+      "loss": 2.6482,
+      "step": 345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33894973760786873,
+      "learning_rate": 9.688553589151038e-06,
+      "loss": 2.5544,
+      "step": 346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3646766361556834,
+      "learning_rate": 9.686716617559442e-06,
+      "loss": 2.6876,
+      "step": 347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33444582420815583,
+      "learning_rate": 9.684874419666355e-06,
+      "loss": 2.6771,
+      "step": 348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3313405650870024,
+      "learning_rate": 9.683026997526063e-06,
+      "loss": 2.59,
+      "step": 349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.352783646632777,
+      "learning_rate": 9.681174353198687e-06,
+      "loss": 2.628,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3651460142908725,
+      "learning_rate": 9.679316488750165e-06,
+      "loss": 2.8014,
+      "step": 351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3512770871751507,
+      "learning_rate": 9.677453406252263e-06,
+      "loss": 2.6069,
+      "step": 352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33320113133396406,
+      "learning_rate": 9.675585107782557e-06,
+      "loss": 2.5503,
+      "step": 353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3303586794567089,
+      "learning_rate": 9.673711595424445e-06,
+      "loss": 2.6388,
+      "step": 354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3487581727564022,
+      "learning_rate": 9.67183287126714e-06,
+      "loss": 2.5335,
+      "step": 355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3255407886746363,
+      "learning_rate": 9.669948937405662e-06,
+      "loss": 2.5475,
+      "step": 356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3361925664939444,
+      "learning_rate": 9.668059795940846e-06,
+      "loss": 2.666,
+      "step": 357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3897837927203283,
+      "learning_rate": 9.66616544897933e-06,
+      "loss": 2.6835,
+      "step": 358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33831534401082114,
+      "learning_rate": 9.664265898633558e-06,
+      "loss": 2.4319,
+      "step": 359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.33799887450160604,
+      "learning_rate": 9.66236114702178e-06,
+      "loss": 2.619,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.35036876189237753,
+      "learning_rate": 9.66045119626804e-06,
+      "loss": 2.6624,
+      "step": 361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36932320091198245,
+      "learning_rate": 9.658536048502183e-06,
+      "loss": 2.538,
+      "step": 362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.36526823651834606,
+      "learning_rate": 9.65661570585985e-06,
+      "loss": 2.7111,
+      "step": 363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34134738248746543,
+      "learning_rate": 9.654690170482474e-06,
+      "loss": 2.6082,
+      "step": 364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3318367107667496,
+      "learning_rate": 9.652759444517276e-06,
+      "loss": 2.5756,
+      "step": 365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3658073133775175,
+      "learning_rate": 9.650823530117273e-06,
+      "loss": 2.5643,
+      "step": 366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3764878796606095,
+      "learning_rate": 9.648882429441258e-06,
+      "loss": 2.5856,
+      "step": 367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3520585057949475,
+      "learning_rate": 9.646936144653813e-06,
+      "loss": 2.5729,
+      "step": 368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.352936370229019,
+      "learning_rate": 9.6449846779253e-06,
+      "loss": 2.618,
+      "step": 369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3505048312671855,
+      "learning_rate": 9.64302803143186e-06,
+      "loss": 2.6206,
+      "step": 370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3679090066668967,
+      "learning_rate": 9.64106620735541e-06,
+      "loss": 2.5719,
+      "step": 371
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34289501892250546,
+      "learning_rate": 9.63909920788364e-06,
+      "loss": 2.5773,
+      "step": 372
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4083631648954588,
+      "learning_rate": 9.63712703521001e-06,
+      "loss": 2.5552,
+      "step": 373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3623205217464116,
+      "learning_rate": 9.635149691533749e-06,
+      "loss": 2.5894,
+      "step": 374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38387843054900256,
+      "learning_rate": 9.633167179059859e-06,
+      "loss": 2.6483,
+      "step": 375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3531254253553681,
+      "learning_rate": 9.631179499999094e-06,
+      "loss": 2.6182,
+      "step": 376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.382022470634699,
+      "learning_rate": 9.629186656567981e-06,
+      "loss": 2.7027,
+      "step": 377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35043041424351157,
+      "learning_rate": 9.6271886509888e-06,
+      "loss": 2.5979,
+      "step": 378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.501125499455809,
+      "learning_rate": 9.625185485489586e-06,
+      "loss": 2.5574,
+      "step": 379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34841115834861225,
+      "learning_rate": 9.623177162304132e-06,
+      "loss": 2.6124,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36414263747845266,
+      "learning_rate": 9.62116368367198e-06,
+      "loss": 2.6047,
+      "step": 381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3621790789002891,
+      "learning_rate": 9.619145051838424e-06,
+      "loss": 2.6431,
+      "step": 382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3423299610454867,
+      "learning_rate": 9.617121269054497e-06,
+      "loss": 2.6453,
+      "step": 383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34946611335508165,
+      "learning_rate": 9.615092337576987e-06,
+      "loss": 2.5827,
+      "step": 384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3795298920344452,
+      "learning_rate": 9.613058259668416e-06,
+      "loss": 2.5943,
+      "step": 385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36580476743110946,
+      "learning_rate": 9.611019037597045e-06,
+      "loss": 2.6066,
+      "step": 386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.36676561183764467,
+      "learning_rate": 9.608974673636874e-06,
+      "loss": 2.4641,
+      "step": 387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3799683344660138,
+      "learning_rate": 9.606925170067637e-06,
+      "loss": 2.624,
+      "step": 388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.38874616823423863,
+      "learning_rate": 9.604870529174797e-06,
+      "loss": 2.5052,
+      "step": 389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35818572839323115,
+      "learning_rate": 9.602810753249549e-06,
+      "loss": 2.5451,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35520000569192134,
+      "learning_rate": 9.60074584458881e-06,
+      "loss": 2.5498,
+      "step": 391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3429831599551629,
+      "learning_rate": 9.598675805495224e-06,
+      "loss": 2.5865,
+      "step": 392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.34755300210268897,
+      "learning_rate": 9.596600638277157e-06,
+      "loss": 2.548,
+      "step": 393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3592352180638645,
+      "learning_rate": 9.594520345248687e-06,
+      "loss": 2.712,
+      "step": 394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3427075162354133,
+      "learning_rate": 9.592434928729617e-06,
+      "loss": 2.5966,
+      "step": 395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3588942754040752,
+      "learning_rate": 9.590344391045454e-06,
+      "loss": 2.5614,
+      "step": 396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.37073559794680655,
+      "learning_rate": 9.588248734527428e-06,
+      "loss": 2.6412,
+      "step": 397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3832373499776409,
+      "learning_rate": 9.586147961512464e-06,
+      "loss": 2.7403,
+      "step": 398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3788810996921007,
+      "learning_rate": 9.5840420743432e-06,
+      "loss": 2.6184,
+      "step": 399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3786757839938464,
+      "learning_rate": 9.581931075367979e-06,
+      "loss": 2.5637,
+      "step": 400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3707176712124623,
+      "learning_rate": 9.579814966940833e-06,
+      "loss": 2.5527,
+      "step": 401
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3872245794218741,
+      "learning_rate": 9.577693751421507e-06,
+      "loss": 2.7156,
+      "step": 402
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.35357111289889626,
+      "learning_rate": 9.575567431175427e-06,
+      "loss": 2.6885,
+      "step": 403
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3475328290649796,
+      "learning_rate": 9.573436008573724e-06,
+      "loss": 2.55,
+      "step": 404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3691908383403371,
+      "learning_rate": 9.57129948599321e-06,
+      "loss": 2.5979,
+      "step": 405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3776062362438795,
+      "learning_rate": 9.569157865816386e-06,
+      "loss": 2.7273,
+      "step": 406
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36309179883045734,
+      "learning_rate": 9.56701115043144e-06,
+      "loss": 2.6366,
+      "step": 407
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3635509269358371,
+      "learning_rate": 9.56485934223224e-06,
+      "loss": 2.599,
+      "step": 408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37084533005857817,
+      "learning_rate": 9.562702443618332e-06,
+      "loss": 2.6457,
+      "step": 409
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36848577310953384,
+      "learning_rate": 9.56054045699494e-06,
+      "loss": 2.5668,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.42518835171568375,
+      "learning_rate": 9.558373384772964e-06,
+      "loss": 2.5535,
+      "step": 411
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3452323035069904,
+      "learning_rate": 9.55620122936897e-06,
+      "loss": 2.4259,
+      "step": 412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34117746085207956,
+      "learning_rate": 9.554023993205194e-06,
+      "loss": 2.6565,
+      "step": 413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3511654412992474,
+      "learning_rate": 9.551841678709543e-06,
+      "loss": 2.6814,
+      "step": 414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3246443838619872,
+      "learning_rate": 9.54965428831558e-06,
+      "loss": 2.6047,
+      "step": 415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36258121435735235,
+      "learning_rate": 9.547461824462534e-06,
+      "loss": 2.6464,
+      "step": 416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34507450774639337,
+      "learning_rate": 9.545264289595284e-06,
+      "loss": 2.563,
+      "step": 417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.386133078004382,
+      "learning_rate": 9.543061686164374e-06,
+      "loss": 2.653,
+      "step": 418
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.33001999653572733,
+      "learning_rate": 9.54085401662599e-06,
+      "loss": 2.5639,
+      "step": 419
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.35995662311984156,
+      "learning_rate": 9.538641283441974e-06,
+      "loss": 2.6475,
+      "step": 420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.38971795630724054,
+      "learning_rate": 9.536423489079812e-06,
+      "loss": 2.727,
+      "step": 421
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36355223854382024,
+      "learning_rate": 9.534200636012636e-06,
+      "loss": 2.5898,
+      "step": 422
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3711145094034144,
+      "learning_rate": 9.531972726719216e-06,
+      "loss": 2.6373,
+      "step": 423
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.34920429225410604,
+      "learning_rate": 9.529739763683965e-06,
+      "loss": 2.5074,
+      "step": 424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3551729879198632,
+      "learning_rate": 9.527501749396924e-06,
+      "loss": 2.6316,
+      "step": 425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3490101137697429,
+      "learning_rate": 9.525258686353776e-06,
+      "loss": 2.6075,
+      "step": 426
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.32833723742847454,
+      "learning_rate": 9.523010577055824e-06,
+      "loss": 2.4231,
+      "step": 427
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3572609544386589,
+      "learning_rate": 9.52075742401001e-06,
+      "loss": 2.611,
+      "step": 428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3731441670515594,
+      "learning_rate": 9.518499229728889e-06,
+      "loss": 2.6161,
+      "step": 429
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36098957181567365,
+      "learning_rate": 9.516235996730645e-06,
+      "loss": 2.6358,
+      "step": 430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36118588096333676,
+      "learning_rate": 9.513967727539079e-06,
+      "loss": 2.6043,
+      "step": 431
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.36325302834767914,
+      "learning_rate": 9.511694424683606e-06,
+      "loss": 2.6365,
+      "step": 432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4035556097613667,
+      "learning_rate": 9.509416090699258e-06,
+      "loss": 2.6116,
+      "step": 433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3588793695436936,
+      "learning_rate": 9.50713272812667e-06,
+      "loss": 2.611,
+      "step": 434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3593171736591926,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 2.6189,
+      "step": 435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37871224067231624,
+      "learning_rate": 9.502550927407386e-06,
+      "loss": 2.7026,
+      "step": 436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36633980073258754,
+      "learning_rate": 9.500252494369992e-06,
+      "loss": 2.5955,
+      "step": 437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3713223205151261,
+      "learning_rate": 9.497949042962971e-06,
+      "loss": 2.5978,
+      "step": 438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3616355001004457,
+      "learning_rate": 9.49564057575497e-06,
+      "loss": 2.5843,
+      "step": 439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3517382394448638,
+      "learning_rate": 9.493327095320231e-06,
+      "loss": 2.5557,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3634880871716239,
+      "learning_rate": 9.49100860423859e-06,
+      "loss": 2.6457,
+      "step": 441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3706486090076816,
+      "learning_rate": 9.488685105095464e-06,
+      "loss": 2.6207,
+      "step": 442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35264587897523175,
+      "learning_rate": 9.486356600481859e-06,
+      "loss": 2.5636,
+      "step": 443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.357439426345557,
+      "learning_rate": 9.484023092994366e-06,
+      "loss": 2.6085,
+      "step": 444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3672140395942012,
+      "learning_rate": 9.481684585235145e-06,
+      "loss": 2.6436,
+      "step": 445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.35536074831654074,
+      "learning_rate": 9.479341079811938e-06,
+      "loss": 2.6217,
+      "step": 446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37164083609535903,
+      "learning_rate": 9.476992579338066e-06,
+      "loss": 2.6707,
+      "step": 447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.33771388993976,
+      "learning_rate": 9.474639086432408e-06,
+      "loss": 2.4938,
+      "step": 448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3773044076000881,
+      "learning_rate": 9.472280603719419e-06,
+      "loss": 2.588,
+      "step": 449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36731472246303526,
+      "learning_rate": 9.469917133829114e-06,
+      "loss": 2.5765,
+      "step": 450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36401889114168867,
+      "learning_rate": 9.467548679397072e-06,
+      "loss": 2.6097,
+      "step": 451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.36397924330762615,
+      "learning_rate": 9.465175243064428e-06,
+      "loss": 2.5596,
+      "step": 452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3479437871268676,
+      "learning_rate": 9.462796827477873e-06,
+      "loss": 2.5487,
+      "step": 453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.45223388523599456,
+      "learning_rate": 9.460413435289655e-06,
+      "loss": 2.6161,
+      "step": 454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39806018990511033,
+      "learning_rate": 9.458025069157563e-06,
+      "loss": 2.5991,
+      "step": 455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.37052890554662676,
+      "learning_rate": 9.45563173174494e-06,
+      "loss": 2.5443,
+      "step": 456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40424250730130273,
+      "learning_rate": 9.45323342572067e-06,
+      "loss": 2.4869,
+      "step": 457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3912357514806361,
+      "learning_rate": 9.450830153759177e-06,
+      "loss": 2.6105,
+      "step": 458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3519370732685063,
+      "learning_rate": 9.448421918540424e-06,
+      "loss": 2.6127,
+      "step": 459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.40068454653993657,
+      "learning_rate": 9.446008722749906e-06,
+      "loss": 2.4963,
+      "step": 460
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39281092331599454,
+      "learning_rate": 9.443590569078655e-06,
+      "loss": 2.5455,
+      "step": 461
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3591307763904136,
+      "learning_rate": 9.441167460223224e-06,
+      "loss": 2.5692,
+      "step": 462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3759099133359235,
+      "learning_rate": 9.4387393988857e-06,
+      "loss": 2.5914,
+      "step": 463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4197464687478331,
+      "learning_rate": 9.436306387773687e-06,
+      "loss": 2.649,
+      "step": 464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.36478768026082575,
+      "learning_rate": 9.43386842960031e-06,
+      "loss": 2.6149,
+      "step": 465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38633531770095064,
+      "learning_rate": 9.431425527084213e-06,
+      "loss": 2.636,
+      "step": 466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38502307934138025,
+      "learning_rate": 9.428977682949548e-06,
+      "loss": 2.481,
+      "step": 467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37578040616581126,
+      "learning_rate": 9.426524899925983e-06,
+      "loss": 2.6481,
+      "step": 468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.36412837615381155,
+      "learning_rate": 9.424067180748692e-06,
+      "loss": 2.6204,
+      "step": 469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3688629963265158,
+      "learning_rate": 9.421604528158355e-06,
+      "loss": 2.5546,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37999274617098006,
+      "learning_rate": 9.419136944901146e-06,
+      "loss": 2.502,
+      "step": 471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3704608338474391,
+      "learning_rate": 9.416664433728749e-06,
+      "loss": 2.5751,
+      "step": 472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.37304381923522467,
+      "learning_rate": 9.414186997398331e-06,
+      "loss": 2.6507,
+      "step": 473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4002003287303057,
+      "learning_rate": 9.411704638672562e-06,
+      "loss": 2.6855,
+      "step": 474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.35759693298816747,
+      "learning_rate": 9.409217360319594e-06,
+      "loss": 2.6198,
+      "step": 475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.39969817039198885,
+      "learning_rate": 9.406725165113069e-06,
+      "loss": 2.534,
+      "step": 476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3807147215970043,
+      "learning_rate": 9.40422805583211e-06,
+      "loss": 2.6171,
+      "step": 477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3790985877712856,
+      "learning_rate": 9.401726035261319e-06,
+      "loss": 2.6556,
+      "step": 478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.351516529892437,
+      "learning_rate": 9.399219106190776e-06,
+      "loss": 2.6053,
+      "step": 479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38937111835779886,
+      "learning_rate": 9.396707271416035e-06,
+      "loss": 2.6586,
+      "step": 480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3785571820542292,
+      "learning_rate": 9.394190533738118e-06,
+      "loss": 2.5479,
+      "step": 481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.361269992585338,
+      "learning_rate": 9.391668895963521e-06,
+      "loss": 2.677,
+      "step": 482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3488395355853058,
+      "learning_rate": 9.389142360904197e-06,
+      "loss": 2.5981,
+      "step": 483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.366717687311896,
+      "learning_rate": 9.386610931377564e-06,
+      "loss": 2.5667,
+      "step": 484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.371686802854601,
+      "learning_rate": 9.384074610206495e-06,
+      "loss": 2.6046,
+      "step": 485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.34937925597176434,
+      "learning_rate": 9.381533400219319e-06,
+      "loss": 2.5307,
+      "step": 486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38148537077296263,
+      "learning_rate": 9.378987304249819e-06,
+      "loss": 2.6458,
+      "step": 487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3526977013061218,
+      "learning_rate": 9.376436325137224e-06,
+      "loss": 2.5112,
+      "step": 488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3672226422314367,
+      "learning_rate": 9.373880465726208e-06,
+      "loss": 2.6376,
+      "step": 489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.378500691878116,
+      "learning_rate": 9.371319728866892e-06,
+      "loss": 2.499,
+      "step": 490
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3964495526135052,
+      "learning_rate": 9.368754117414828e-06,
+      "loss": 2.5133,
+      "step": 491
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.38403991576946866,
+      "learning_rate": 9.36618363423101e-06,
+      "loss": 2.5839,
+      "step": 492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39166616920006725,
+      "learning_rate": 9.363608282181862e-06,
+      "loss": 2.6802,
+      "step": 493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.367343341952105,
+      "learning_rate": 9.36102806413924e-06,
+      "loss": 2.5199,
+      "step": 494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3686087879446113,
+      "learning_rate": 9.35844298298042e-06,
+      "loss": 2.4803,
+      "step": 495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.36947777907644486,
+      "learning_rate": 9.355853041588111e-06,
+      "loss": 2.6321,
+      "step": 496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.36917330077534516,
+      "learning_rate": 9.353258242850432e-06,
+      "loss": 2.5247,
+      "step": 497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37585804049584626,
+      "learning_rate": 9.350658589660924e-06,
+      "loss": 2.5508,
+      "step": 498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3950181920956615,
+      "learning_rate": 9.34805408491854e-06,
+      "loss": 2.563,
+      "step": 499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3612877900576221,
+      "learning_rate": 9.345444731527642e-06,
+      "loss": 2.5589,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3747026735022066,
+      "learning_rate": 9.342830532398002e-06,
+      "loss": 2.6633,
+      "step": 501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3707103957658041,
+      "learning_rate": 9.340211490444793e-06,
+      "loss": 2.5683,
+      "step": 502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3881547065398902,
+      "learning_rate": 9.337587608588588e-06,
+      "loss": 2.5465,
+      "step": 503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.38485592784501677,
+      "learning_rate": 9.33495888975536e-06,
+      "loss": 2.5729,
+      "step": 504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3631651703156107,
+      "learning_rate": 9.332325336876472e-06,
+      "loss": 2.5225,
+      "step": 505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.395946419607132,
+      "learning_rate": 9.32968695288868e-06,
+      "loss": 2.708,
+      "step": 506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3639799802195103,
+      "learning_rate": 9.327043740734129e-06,
+      "loss": 2.6279,
+      "step": 507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39253570555205214,
+      "learning_rate": 9.324395703360345e-06,
+      "loss": 2.7964,
+      "step": 508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4062943434356018,
+      "learning_rate": 9.321742843720234e-06,
+      "loss": 2.5721,
+      "step": 509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3745062677217993,
+      "learning_rate": 9.319085164772082e-06,
+      "loss": 2.5667,
+      "step": 510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3816082756208765,
+      "learning_rate": 9.31642266947955e-06,
+      "loss": 2.6327,
+      "step": 511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.34810809443669893,
+      "learning_rate": 9.313755360811665e-06,
+      "loss": 2.5067,
+      "step": 512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.669410354578167,
+      "learning_rate": 9.311083241742829e-06,
+      "loss": 2.6153,
+      "step": 513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3828088604501034,
+      "learning_rate": 9.308406315252799e-06,
+      "loss": 2.6264,
+      "step": 514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3640939993595143,
+      "learning_rate": 9.305724584326702e-06,
+      "loss": 2.5744,
+      "step": 515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.36645221137260386,
+      "learning_rate": 9.303038051955017e-06,
+      "loss": 2.6558,
+      "step": 516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.36173421361179037,
+      "learning_rate": 9.300346721133577e-06,
+      "loss": 2.637,
+      "step": 517
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3479028106069891,
+      "learning_rate": 9.297650594863572e-06,
+      "loss": 2.627,
+      "step": 518
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3779033286984007,
+      "learning_rate": 9.29494967615153e-06,
+      "loss": 2.7463,
+      "step": 519
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.393068131152805,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 2.6275,
+      "step": 520
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.36336754214963424,
+      "learning_rate": 9.289533473454194e-06,
+      "loss": 2.5236,
+      "step": 521
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37280173510784137,
+      "learning_rate": 9.286818195508672e-06,
+      "loss": 2.6627,
+      "step": 522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3814394128580523,
+      "learning_rate": 9.284098137200656e-06,
+      "loss": 2.6625,
+      "step": 523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.35570304610873393,
+      "learning_rate": 9.281373301563369e-06,
+      "loss": 2.6041,
+      "step": 524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3796685602621214,
+      "learning_rate": 9.278643691635352e-06,
+      "loss": 2.5634,
+      "step": 525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37182496960471256,
+      "learning_rate": 9.275909310460484e-06,
+      "loss": 2.6144,
+      "step": 526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36597489951785306,
+      "learning_rate": 9.273170161087954e-06,
+      "loss": 2.4505,
+      "step": 527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37863369606990943,
+      "learning_rate": 9.270426246572273e-06,
+      "loss": 2.5277,
+      "step": 528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.44730683515841135,
+      "learning_rate": 9.26767756997326e-06,
+      "loss": 2.5425,
+      "step": 529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39472069299440277,
+      "learning_rate": 9.264924134356057e-06,
+      "loss": 2.7211,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3501969413485413,
+      "learning_rate": 9.262165942791098e-06,
+      "loss": 2.5592,
+      "step": 531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38360651542726115,
+      "learning_rate": 9.259402998354125e-06,
+      "loss": 2.6333,
+      "step": 532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3863578615228763,
+      "learning_rate": 9.25663530412619e-06,
+      "loss": 2.4994,
+      "step": 533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37848956999650335,
+      "learning_rate": 9.253862863193625e-06,
+      "loss": 2.5538,
+      "step": 534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3783992836553777,
+      "learning_rate": 9.251085678648072e-06,
+      "loss": 2.6139,
+      "step": 535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3972475946136305,
+      "learning_rate": 9.248303753586449e-06,
+      "loss": 2.6361,
+      "step": 536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39414074900599666,
+      "learning_rate": 9.24551709111097e-06,
+      "loss": 2.6609,
+      "step": 537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4031966345371276,
+      "learning_rate": 9.242725694329123e-06,
+      "loss": 2.5674,
+      "step": 538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37415076496357547,
+      "learning_rate": 9.239929566353685e-06,
+      "loss": 2.7205,
+      "step": 539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.39152375777763443,
+      "learning_rate": 9.2371287103027e-06,
+      "loss": 2.6226,
+      "step": 540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38096918332633267,
+      "learning_rate": 9.234323129299493e-06,
+      "loss": 2.5059,
+      "step": 541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3624078649965304,
+      "learning_rate": 9.231512826472651e-06,
+      "loss": 2.6158,
+      "step": 542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38864706909152263,
+      "learning_rate": 9.228697804956027e-06,
+      "loss": 2.7084,
+      "step": 543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3641660027457906,
+      "learning_rate": 9.225878067888741e-06,
+      "loss": 2.5334,
+      "step": 544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38568072251846647,
+      "learning_rate": 9.223053618415168e-06,
+      "loss": 2.5861,
+      "step": 545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40430584957463456,
+      "learning_rate": 9.220224459684936e-06,
+      "loss": 2.6363,
+      "step": 546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.40214767122496736,
+      "learning_rate": 9.217390594852928e-06,
+      "loss": 2.491,
+      "step": 547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37009409729286524,
+      "learning_rate": 9.214552027079274e-06,
+      "loss": 2.5962,
+      "step": 548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3701496764243851,
+      "learning_rate": 9.211708759529347e-06,
+      "loss": 2.6619,
+      "step": 549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.38085250848302366,
+      "learning_rate": 9.208860795373765e-06,
+      "loss": 2.5934,
+      "step": 550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3703456767225791,
+      "learning_rate": 9.206008137788376e-06,
+      "loss": 2.5907,
+      "step": 551
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37521457493596533,
+      "learning_rate": 9.203150789954269e-06,
+      "loss": 2.6769,
+      "step": 552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3926093625218847,
+      "learning_rate": 9.200288755057757e-06,
+      "loss": 2.6641,
+      "step": 553
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3659477610660178,
+      "learning_rate": 9.197422036290386e-06,
+      "loss": 2.4708,
+      "step": 554
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36528777685770725,
+      "learning_rate": 9.194550636848923e-06,
+      "loss": 2.495,
+      "step": 555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36595973236411306,
+      "learning_rate": 9.191674559935349e-06,
+      "loss": 2.6522,
+      "step": 556
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.40533292402750815,
+      "learning_rate": 9.18879380875687e-06,
+      "loss": 2.6301,
+      "step": 557
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3814157774351385,
+      "learning_rate": 9.185908386525897e-06,
+      "loss": 2.6509,
+      "step": 558
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.39954200691359565,
+      "learning_rate": 9.183018296460055e-06,
+      "loss": 2.58,
+      "step": 559
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3980474994268938,
+      "learning_rate": 9.180123541782172e-06,
+      "loss": 2.5013,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.362430108666147,
+      "learning_rate": 9.177224125720274e-06,
+      "loss": 2.6252,
+      "step": 561
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36535128593944177,
+      "learning_rate": 9.174320051507595e-06,
+      "loss": 2.5497,
+      "step": 562
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3838873584463476,
+      "learning_rate": 9.171411322382552e-06,
+      "loss": 2.6119,
+      "step": 563
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3698863414225833,
+      "learning_rate": 9.16849794158876e-06,
+      "loss": 2.5775,
+      "step": 564
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3914562726368379,
+      "learning_rate": 9.16557991237502e-06,
+      "loss": 2.528,
+      "step": 565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3748930540173487,
+      "learning_rate": 9.162657237995316e-06,
+      "loss": 2.5201,
+      "step": 566
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36816944553477204,
+      "learning_rate": 9.159729921708811e-06,
+      "loss": 2.5629,
+      "step": 567
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37321218852120774,
+      "learning_rate": 9.156797966779847e-06,
+      "loss": 2.5419,
+      "step": 568
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37915436005095765,
+      "learning_rate": 9.153861376477933e-06,
+      "loss": 2.7052,
+      "step": 569
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3704407417589313,
+      "learning_rate": 9.150920154077753e-06,
+      "loss": 2.6223,
+      "step": 570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38029524275808224,
+      "learning_rate": 9.147974302859158e-06,
+      "loss": 2.5667,
+      "step": 571
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.36244234004689,
+      "learning_rate": 9.14502382610715e-06,
+      "loss": 2.6414,
+      "step": 572
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3957962679658596,
+      "learning_rate": 9.1420687271119e-06,
+      "loss": 2.63,
+      "step": 573
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37690782027680103,
+      "learning_rate": 9.139109009168732e-06,
+      "loss": 2.6165,
+      "step": 574
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.38991806911964444,
+      "learning_rate": 9.136144675578114e-06,
+      "loss": 2.5532,
+      "step": 575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37241165643256596,
+      "learning_rate": 9.133175729645667e-06,
+      "loss": 2.5109,
+      "step": 576
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3881669718704454,
+      "learning_rate": 9.130202174682154e-06,
+      "loss": 2.6121,
+      "step": 577
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3746737411938074,
+      "learning_rate": 9.127224014003478e-06,
+      "loss": 2.5589,
+      "step": 578
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.35909899188362016,
+      "learning_rate": 9.124241250930675e-06,
+      "loss": 2.4821,
+      "step": 579
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37780525071199483,
+      "learning_rate": 9.121253888789916e-06,
+      "loss": 2.5867,
+      "step": 580
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3783001722529536,
+      "learning_rate": 9.118261930912505e-06,
+      "loss": 2.6057,
+      "step": 581
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37910446409582155,
+      "learning_rate": 9.11526538063486e-06,
+      "loss": 2.5794,
+      "step": 582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3910669306460941,
+      "learning_rate": 9.112264241298527e-06,
+      "loss": 2.5727,
+      "step": 583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3979995789286049,
+      "learning_rate": 9.109258516250172e-06,
+      "loss": 2.5771,
+      "step": 584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37724573509126186,
+      "learning_rate": 9.106248208841568e-06,
+      "loss": 2.6118,
+      "step": 585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37725914506211394,
+      "learning_rate": 9.103233322429603e-06,
+      "loss": 2.6005,
+      "step": 586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.380958974481754,
+      "learning_rate": 9.10021386037627e-06,
+      "loss": 2.5719,
+      "step": 587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41455883292386886,
+      "learning_rate": 9.09718982604866e-06,
+      "loss": 2.5273,
+      "step": 588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4056643886867896,
+      "learning_rate": 9.094161222818971e-06,
+      "loss": 2.5989,
+      "step": 589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4087813630452031,
+      "learning_rate": 9.091128054064487e-06,
+      "loss": 2.529,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40494733812427797,
+      "learning_rate": 9.08809032316759e-06,
+      "loss": 2.6702,
+      "step": 591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38784529621841723,
+      "learning_rate": 9.085048033515749e-06,
+      "loss": 2.5724,
+      "step": 592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.36458722067077504,
+      "learning_rate": 9.08200118850151e-06,
+      "loss": 2.5695,
+      "step": 593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3669722311573834,
+      "learning_rate": 9.078949791522509e-06,
+      "loss": 2.4487,
+      "step": 594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4492875917179086,
+      "learning_rate": 9.075893845981445e-06,
+      "loss": 2.7234,
+      "step": 595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40343553130410426,
+      "learning_rate": 9.072833355286104e-06,
+      "loss": 2.6101,
+      "step": 596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38583067832648477,
+      "learning_rate": 9.069768322849327e-06,
+      "loss": 2.5303,
+      "step": 597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3769044202262478,
+      "learning_rate": 9.06669875208903e-06,
+      "loss": 2.5304,
+      "step": 598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39753216379422124,
+      "learning_rate": 9.063624646428184e-06,
+      "loss": 2.6055,
+      "step": 599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39269184723845535,
+      "learning_rate": 9.060546009294818e-06,
+      "loss": 2.5723,
+      "step": 600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4054397241864531,
+      "learning_rate": 9.057462844122016e-06,
+      "loss": 2.6027,
+      "step": 601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38168745739432036,
+      "learning_rate": 9.054375154347908e-06,
+      "loss": 2.6156,
+      "step": 602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.39640886311164103,
+      "learning_rate": 9.051282943415673e-06,
+      "loss": 2.5587,
+      "step": 603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3905155791570832,
+      "learning_rate": 9.048186214773531e-06,
+      "loss": 2.6544,
+      "step": 604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4218796447215651,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 2.4422,
+      "step": 605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.41284861860724453,
+      "learning_rate": 9.041979218177586e-06,
+      "loss": 2.6839,
+      "step": 606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37382785075625063,
+      "learning_rate": 9.038868957145394e-06,
+      "loss": 2.6169,
+      "step": 607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37855383618111166,
+      "learning_rate": 9.035754192246513e-06,
+      "loss": 2.6089,
+      "step": 608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40886094964428643,
+      "learning_rate": 9.032634926954311e-06,
+      "loss": 2.5738,
+      "step": 609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.38056500135812904,
+      "learning_rate": 9.029511164747175e-06,
+      "loss": 2.5429,
+      "step": 610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3479316991805906,
+      "learning_rate": 9.02638290910851e-06,
+      "loss": 2.484,
+      "step": 611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3866011315472774,
+      "learning_rate": 9.023250163526731e-06,
+      "loss": 2.5471,
+      "step": 612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3823249982531166,
+      "learning_rate": 9.020112931495256e-06,
+      "loss": 2.528,
+      "step": 613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3746182564964411,
+      "learning_rate": 9.016971216512508e-06,
+      "loss": 2.4831,
+      "step": 614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3687923342439741,
+      "learning_rate": 9.013825022081915e-06,
+      "loss": 2.586,
+      "step": 615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38139049958522336,
+      "learning_rate": 9.01067435171189e-06,
+      "loss": 2.6182,
+      "step": 616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3886422764228466,
+      "learning_rate": 9.007519208915843e-06,
+      "loss": 2.6021,
+      "step": 617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38891624551252624,
+      "learning_rate": 9.004359597212173e-06,
+      "loss": 2.581,
+      "step": 618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4032441961032781,
+      "learning_rate": 9.001195520124257e-06,
+      "loss": 2.5595,
+      "step": 619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39172542068772426,
+      "learning_rate": 8.998026981180454e-06,
+      "loss": 2.5584,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.35467448945907143,
+      "learning_rate": 8.994853983914104e-06,
+      "loss": 2.5566,
+      "step": 621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39820696201163513,
+      "learning_rate": 8.991676531863507e-06,
+      "loss": 2.5934,
+      "step": 622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37666804152353134,
+      "learning_rate": 8.988494628571943e-06,
+      "loss": 2.5704,
+      "step": 623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37676360747299015,
+      "learning_rate": 8.985308277587647e-06,
+      "loss": 2.6006,
+      "step": 624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39448469661600266,
+      "learning_rate": 8.982117482463817e-06,
+      "loss": 2.5857,
+      "step": 625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3815405598299067,
+      "learning_rate": 8.978922246758607e-06,
+      "loss": 2.5791,
+      "step": 626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4020737047310836,
+      "learning_rate": 8.975722574035122e-06,
+      "loss": 2.6392,
+      "step": 627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4058345357510237,
+      "learning_rate": 8.972518467861413e-06,
+      "loss": 2.5718,
+      "step": 628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3968737339390062,
+      "learning_rate": 8.969309931810483e-06,
+      "loss": 2.6727,
+      "step": 629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4044949511659703,
+      "learning_rate": 8.966096969460263e-06,
+      "loss": 2.6641,
+      "step": 630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4050966911512252,
+      "learning_rate": 8.96287958439363e-06,
+      "loss": 2.5858,
+      "step": 631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37963019358499156,
+      "learning_rate": 8.959657780198388e-06,
+      "loss": 2.5699,
+      "step": 632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3752560536088806,
+      "learning_rate": 8.956431560467267e-06,
+      "loss": 2.5395,
+      "step": 633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41909915260278424,
+      "learning_rate": 8.953200928797926e-06,
+      "loss": 2.5608,
+      "step": 634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3757858981917666,
+      "learning_rate": 8.94996588879294e-06,
+      "loss": 2.534,
+      "step": 635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41734799073355916,
+      "learning_rate": 8.946726444059804e-06,
+      "loss": 2.5867,
+      "step": 636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40750575386267923,
+      "learning_rate": 8.94348259821092e-06,
+      "loss": 2.606,
+      "step": 637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.37301705525833967,
+      "learning_rate": 8.940234354863599e-06,
+      "loss": 2.4973,
+      "step": 638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.39286495756761686,
+      "learning_rate": 8.936981717640061e-06,
+      "loss": 2.5394,
+      "step": 639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4361679370503528,
+      "learning_rate": 8.933724690167417e-06,
+      "loss": 2.6022,
+      "step": 640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.40947083014043045,
+      "learning_rate": 8.930463276077681e-06,
+      "loss": 2.5744,
+      "step": 641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3617481574016219,
+      "learning_rate": 8.927197479007753e-06,
+      "loss": 2.4723,
+      "step": 642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.395860622954373,
+      "learning_rate": 8.923927302599429e-06,
+      "loss": 2.5622,
+      "step": 643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4065535584975831,
+      "learning_rate": 8.920652750499378e-06,
+      "loss": 2.6154,
+      "step": 644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.367418500595609,
+      "learning_rate": 8.917373826359156e-06,
+      "loss": 2.6548,
+      "step": 645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39840840252112203,
+      "learning_rate": 8.91409053383519e-06,
+      "loss": 2.687,
+      "step": 646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37079672722035706,
+      "learning_rate": 8.910802876588783e-06,
+      "loss": 2.6192,
+      "step": 647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37845913570811357,
+      "learning_rate": 8.9075108582861e-06,
+      "loss": 2.5964,
+      "step": 648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3668806538966641,
+      "learning_rate": 8.904214482598172e-06,
+      "loss": 2.5888,
+      "step": 649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.38671034986832453,
+      "learning_rate": 8.900913753200887e-06,
+      "loss": 2.5977,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3836801777296474,
+      "learning_rate": 8.897608673774995e-06,
+      "loss": 2.5607,
+      "step": 651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3916600818621598,
+      "learning_rate": 8.894299248006088e-06,
+      "loss": 2.5144,
+      "step": 652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.38926938389539,
+      "learning_rate": 8.890985479584606e-06,
+      "loss": 2.5266,
+      "step": 653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4106809801760616,
+      "learning_rate": 8.88766737220584e-06,
+      "loss": 2.5809,
+      "step": 654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39253585739717706,
+      "learning_rate": 8.884344929569905e-06,
+      "loss": 2.6197,
+      "step": 655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3907669381712913,
+      "learning_rate": 8.881018155381766e-06,
+      "loss": 2.6676,
+      "step": 656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.38831433667568693,
+      "learning_rate": 8.877687053351208e-06,
+      "loss": 2.6853,
+      "step": 657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3859399023251534,
+      "learning_rate": 8.874351627192844e-06,
+      "loss": 2.6383,
+      "step": 658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.42425009820652615,
+      "learning_rate": 8.871011880626112e-06,
+      "loss": 2.5976,
+      "step": 659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.37616130699136113,
+      "learning_rate": 8.867667817375266e-06,
+      "loss": 2.5867,
+      "step": 660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4063149730370637,
+      "learning_rate": 8.864319441169373e-06,
+      "loss": 2.6468,
+      "step": 661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41735807710575523,
+      "learning_rate": 8.860966755742308e-06,
+      "loss": 2.6377,
+      "step": 662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4229265106947187,
+      "learning_rate": 8.857609764832758e-06,
+      "loss": 2.6582,
+      "step": 663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.39930886000035337,
+      "learning_rate": 8.854248472184202e-06,
+      "loss": 2.3838,
+      "step": 664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4024205180881961,
+      "learning_rate": 8.850882881544923e-06,
+      "loss": 2.5637,
+      "step": 665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.41146195888933934,
+      "learning_rate": 8.847512996667991e-06,
+      "loss": 2.6788,
+      "step": 666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.49138344073463536,
+      "learning_rate": 8.844138821311273e-06,
+      "loss": 2.6933,
+      "step": 667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.38353973521143947,
+      "learning_rate": 8.840760359237412e-06,
+      "loss": 2.5917,
+      "step": 668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40273612065496545,
+      "learning_rate": 8.837377614213837e-06,
+      "loss": 2.6275,
+      "step": 669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3908994854710692,
+      "learning_rate": 8.833990590012749e-06,
+      "loss": 2.5214,
+      "step": 670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3935355662204077,
+      "learning_rate": 8.830599290411123e-06,
+      "loss": 2.4964,
+      "step": 671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39568823949262405,
+      "learning_rate": 8.827203719190704e-06,
+      "loss": 2.6924,
+      "step": 672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3944176236437975,
+      "learning_rate": 8.823803880137993e-06,
+      "loss": 2.617,
+      "step": 673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3897676596222373,
+      "learning_rate": 8.820399777044258e-06,
+      "loss": 2.6128,
+      "step": 674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39558984353182575,
+      "learning_rate": 8.816991413705515e-06,
+      "loss": 2.4529,
+      "step": 675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3800721664980996,
+      "learning_rate": 8.81357879392254e-06,
+      "loss": 2.603,
+      "step": 676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45424267616939173,
+      "learning_rate": 8.810161921500846e-06,
+      "loss": 2.6688,
+      "step": 677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.41749462081390515,
+      "learning_rate": 8.806740800250695e-06,
+      "loss": 2.5672,
+      "step": 678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3809152415456229,
+      "learning_rate": 8.80331543398708e-06,
+      "loss": 2.5946,
+      "step": 679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37890440163702277,
+      "learning_rate": 8.799885826529736e-06,
+      "loss": 2.5157,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4239167503127884,
+      "learning_rate": 8.796451981703124e-06,
+      "loss": 2.5184,
+      "step": 681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3758665245833528,
+      "learning_rate": 8.793013903336428e-06,
+      "loss": 2.5525,
+      "step": 682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3967420933674155,
+      "learning_rate": 8.789571595263554e-06,
+      "loss": 2.4598,
+      "step": 683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3978377671769104,
+      "learning_rate": 8.786125061323126e-06,
+      "loss": 2.5469,
+      "step": 684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4260046486600914,
+      "learning_rate": 8.782674305358481e-06,
+      "loss": 2.5178,
+      "step": 685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3813517395660569,
+      "learning_rate": 8.77921933121766e-06,
+      "loss": 2.532,
+      "step": 686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4077111889338434,
+      "learning_rate": 8.775760142753414e-06,
+      "loss": 2.5217,
+      "step": 687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.38579419908077955,
+      "learning_rate": 8.772296743823187e-06,
+      "loss": 2.4944,
+      "step": 688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37757286950559593,
+      "learning_rate": 8.768829138289124e-06,
+      "loss": 2.5247,
+      "step": 689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3938089311489784,
+      "learning_rate": 8.765357330018056e-06,
+      "loss": 2.4969,
+      "step": 690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3911060994568757,
+      "learning_rate": 8.761881322881505e-06,
+      "loss": 2.5603,
+      "step": 691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4093559631988249,
+      "learning_rate": 8.758401120755672e-06,
+      "loss": 2.5598,
+      "step": 692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4091429591481071,
+      "learning_rate": 8.754916727521437e-06,
+      "loss": 2.618,
+      "step": 693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.41610906563454025,
+      "learning_rate": 8.751428147064356e-06,
+      "loss": 2.6246,
+      "step": 694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40454770107419946,
+      "learning_rate": 8.74793538327465e-06,
+      "loss": 2.6546,
+      "step": 695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39444404998089466,
+      "learning_rate": 8.744438440047207e-06,
+      "loss": 2.5484,
+      "step": 696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3931267733450527,
+      "learning_rate": 8.740937321281576e-06,
+      "loss": 2.5972,
+      "step": 697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3900499116002855,
+      "learning_rate": 8.737432030881962e-06,
+      "loss": 2.6069,
+      "step": 698
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3931387129910314,
+      "learning_rate": 8.733922572757223e-06,
+      "loss": 2.4599,
+      "step": 699
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40456271070229993,
+      "learning_rate": 8.730408950820864e-06,
+      "loss": 2.489,
+      "step": 700
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.42428694119007687,
+      "learning_rate": 8.726891168991028e-06,
+      "loss": 2.63,
+      "step": 701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39795530349666575,
+      "learning_rate": 8.723369231190506e-06,
+      "loss": 2.6393,
+      "step": 702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41695698716832086,
+      "learning_rate": 8.719843141346717e-06,
+      "loss": 2.6571,
+      "step": 703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40011888912769766,
+      "learning_rate": 8.716312903391712e-06,
+      "loss": 2.5803,
+      "step": 704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.43064551627086994,
+      "learning_rate": 8.71277852126217e-06,
+      "loss": 2.5659,
+      "step": 705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4163134945736012,
+      "learning_rate": 8.709239998899386e-06,
+      "loss": 2.5896,
+      "step": 706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3985848434870783,
+      "learning_rate": 8.705697340249275e-06,
+      "loss": 2.6886,
+      "step": 707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42880397473573023,
+      "learning_rate": 8.702150549262365e-06,
+      "loss": 2.5263,
+      "step": 708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.422368185202293,
+      "learning_rate": 8.698599629893794e-06,
+      "loss": 2.5641,
+      "step": 709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40746276738031384,
+      "learning_rate": 8.695044586103297e-06,
+      "loss": 2.5599,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42202722487495065,
+      "learning_rate": 8.691485421855214e-06,
+      "loss": 2.623,
+      "step": 711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39927169015807223,
+      "learning_rate": 8.68792214111848e-06,
+      "loss": 2.7374,
+      "step": 712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.41192966538348047,
+      "learning_rate": 8.684354747866614e-06,
+      "loss": 2.5767,
+      "step": 713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40163184580235517,
+      "learning_rate": 8.680783246077732e-06,
+      "loss": 2.5966,
+      "step": 714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3846715821878233,
+      "learning_rate": 8.67720763973452e-06,
+      "loss": 2.6508,
+      "step": 715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4441529117116035,
+      "learning_rate": 8.67362793282425e-06,
+      "loss": 2.6694,
+      "step": 716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3865231876844743,
+      "learning_rate": 8.67004412933876e-06,
+      "loss": 2.588,
+      "step": 717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4247407185414399,
+      "learning_rate": 8.666456233274461e-06,
+      "loss": 2.5641,
+      "step": 718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42994671718800553,
+      "learning_rate": 8.662864248632325e-06,
+      "loss": 2.6443,
+      "step": 719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3973979478265679,
+      "learning_rate": 8.659268179417886e-06,
+      "loss": 2.5991,
+      "step": 720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.387493032731209,
+      "learning_rate": 8.65566802964123e-06,
+      "loss": 2.6397,
+      "step": 721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.42717867074031024,
+      "learning_rate": 8.652063803316994e-06,
+      "loss": 2.5512,
+      "step": 722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.47574539174721103,
+      "learning_rate": 8.648455504464362e-06,
+      "loss": 2.5937,
+      "step": 723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4204495330694543,
+      "learning_rate": 8.644843137107058e-06,
+      "loss": 2.6502,
+      "step": 724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38454967577088195,
+      "learning_rate": 8.641226705273344e-06,
+      "loss": 2.5139,
+      "step": 725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4056767306969294,
+      "learning_rate": 8.637606212996016e-06,
+      "loss": 2.5353,
+      "step": 726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39146533223192914,
+      "learning_rate": 8.633981664312393e-06,
+      "loss": 2.57,
+      "step": 727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.44199713258762546,
+      "learning_rate": 8.630353063264321e-06,
+      "loss": 2.629,
+      "step": 728
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4568822242802304,
+      "learning_rate": 8.626720413898167e-06,
+      "loss": 2.6111,
+      "step": 729
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.39149382291137347,
+      "learning_rate": 8.623083720264806e-06,
+      "loss": 2.5724,
+      "step": 730
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4047779409319991,
+      "learning_rate": 8.619442986419631e-06,
+      "loss": 2.4815,
+      "step": 731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.39009103519491567,
+      "learning_rate": 8.615798216422531e-06,
+      "loss": 2.4815,
+      "step": 732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4010843175857095,
+      "learning_rate": 8.612149414337905e-06,
+      "loss": 2.6492,
+      "step": 733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.40851694118151305,
+      "learning_rate": 8.608496584234641e-06,
+      "loss": 2.5597,
+      "step": 734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3714028727338163,
+      "learning_rate": 8.604839730186125e-06,
+      "loss": 2.4603,
+      "step": 735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.37566221893800655,
+      "learning_rate": 8.601178856270224e-06,
+      "loss": 2.6894,
+      "step": 736
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4014233552191909,
+      "learning_rate": 8.597513966569295e-06,
+      "loss": 2.5056,
+      "step": 737
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3837979730634555,
+      "learning_rate": 8.593845065170165e-06,
+      "loss": 2.556,
+      "step": 738
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.45198136738439715,
+      "learning_rate": 8.590172156164139e-06,
+      "loss": 2.6372,
+      "step": 739
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3875310439318507,
+      "learning_rate": 8.586495243646992e-06,
+      "loss": 2.5632,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4249496037391712,
+      "learning_rate": 8.582814331718961e-06,
+      "loss": 2.6393,
+      "step": 741
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44396635342172164,
+      "learning_rate": 8.579129424484746e-06,
+      "loss": 2.6965,
+      "step": 742
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.44773976211089295,
+      "learning_rate": 8.575440526053497e-06,
+      "loss": 2.6293,
+      "step": 743
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.42289162813804465,
+      "learning_rate": 8.571747640538818e-06,
+      "loss": 2.5892,
+      "step": 744
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.38147902332421474,
+      "learning_rate": 8.568050772058763e-06,
+      "loss": 2.4829,
+      "step": 745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4032218757560865,
+      "learning_rate": 8.564349924735817e-06,
+      "loss": 2.555,
+      "step": 746
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4141529092512198,
+      "learning_rate": 8.560645102696914e-06,
+      "loss": 2.6312,
+      "step": 747
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.431306666645557,
+      "learning_rate": 8.556936310073413e-06,
+      "loss": 2.4882,
+      "step": 748
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.429178526013904,
+      "learning_rate": 8.553223551001101e-06,
+      "loss": 2.5607,
+      "step": 749
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4243555435023986,
+      "learning_rate": 8.549506829620193e-06,
+      "loss": 2.6239,
+      "step": 750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.414717893313139,
+      "learning_rate": 8.545786150075314e-06,
+      "loss": 2.5657,
+      "step": 751
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.38781113900215547,
+      "learning_rate": 8.542061516515512e-06,
+      "loss": 2.5971,
+      "step": 752
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3928555699106409,
+      "learning_rate": 8.538332933094237e-06,
+      "loss": 2.6138,
+      "step": 753
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.39710846064235467,
+      "learning_rate": 8.534600403969352e-06,
+      "loss": 2.5739,
+      "step": 754
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41230975156535604,
+      "learning_rate": 8.530863933303108e-06,
+      "loss": 2.6418,
+      "step": 755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4041817911339788,
+      "learning_rate": 8.527123525262162e-06,
+      "loss": 2.4362,
+      "step": 756
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3975027751225941,
+      "learning_rate": 8.52337918401756e-06,
+      "loss": 2.5741,
+      "step": 757
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.40773708260012337,
+      "learning_rate": 8.519630913744726e-06,
+      "loss": 2.67,
+      "step": 758
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4205382680251358,
+      "learning_rate": 8.515878718623473e-06,
+      "loss": 2.6794,
+      "step": 759
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4328560266745644,
+      "learning_rate": 8.512122602837993e-06,
+      "loss": 2.6785,
+      "step": 760
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3886682903181997,
+      "learning_rate": 8.508362570576842e-06,
+      "loss": 2.5902,
+      "step": 761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39657155849490777,
+      "learning_rate": 8.504598626032947e-06,
+      "loss": 2.6572,
+      "step": 762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40633343793420695,
+      "learning_rate": 8.5008307734036e-06,
+      "loss": 2.5907,
+      "step": 763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.42596452600683693,
+      "learning_rate": 8.49705901689045e-06,
+      "loss": 2.598,
+      "step": 764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.415658573716301,
+      "learning_rate": 8.493283360699496e-06,
+      "loss": 2.6445,
+      "step": 765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3817716773710336,
+      "learning_rate": 8.489503809041089e-06,
+      "loss": 2.5449,
+      "step": 766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4174933180588395,
+      "learning_rate": 8.485720366129922e-06,
+      "loss": 2.5976,
+      "step": 767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41151998959306785,
+      "learning_rate": 8.48193303618503e-06,
+      "loss": 2.5543,
+      "step": 768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39560745302793604,
+      "learning_rate": 8.47814182342978e-06,
+      "loss": 2.5391,
+      "step": 769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39459379097045677,
+      "learning_rate": 8.47434673209187e-06,
+      "loss": 2.5435,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4146931721601953,
+      "learning_rate": 8.47054776640332e-06,
+      "loss": 2.661,
+      "step": 771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4311836896585072,
+      "learning_rate": 8.46674493060048e-06,
+      "loss": 2.5523,
+      "step": 772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40199742643044956,
+      "learning_rate": 8.462938228924e-06,
+      "loss": 2.5918,
+      "step": 773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.37538274464984434,
+      "learning_rate": 8.459127665618858e-06,
+      "loss": 2.5758,
+      "step": 774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.43428204866077696,
+      "learning_rate": 8.455313244934324e-06,
+      "loss": 2.5581,
+      "step": 775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4037835921789478,
+      "learning_rate": 8.45149497112398e-06,
+      "loss": 2.6926,
+      "step": 776
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4091287731459371,
+      "learning_rate": 8.447672848445695e-06,
+      "loss": 2.5811,
+      "step": 777
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4217107594408988,
+      "learning_rate": 8.44384688116164e-06,
+      "loss": 2.5743,
+      "step": 778
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39746033764543187,
+      "learning_rate": 8.44001707353827e-06,
+      "loss": 2.5957,
+      "step": 779
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4441830214745251,
+      "learning_rate": 8.436183429846314e-06,
+      "loss": 2.6142,
+      "step": 780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40324050076689993,
+      "learning_rate": 8.43234595436079e-06,
+      "loss": 2.5527,
+      "step": 781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4211610121078822,
+      "learning_rate": 8.428504651360988e-06,
+      "loss": 2.6577,
+      "step": 782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40083303433021816,
+      "learning_rate": 8.424659525130456e-06,
+      "loss": 2.6756,
+      "step": 783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40395875512045837,
+      "learning_rate": 8.420810579957016e-06,
+      "loss": 2.5976,
+      "step": 784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3813669588134605,
+      "learning_rate": 8.416957820132743e-06,
+      "loss": 2.5569,
+      "step": 785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4134031840620254,
+      "learning_rate": 8.413101249953971e-06,
+      "loss": 2.4946,
+      "step": 786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.40916067730209765,
+      "learning_rate": 8.409240873721277e-06,
+      "loss": 2.5588,
+      "step": 787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.420463496600291,
+      "learning_rate": 8.405376695739485e-06,
+      "loss": 2.6316,
+      "step": 788
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.38726844637275737,
+      "learning_rate": 8.40150872031766e-06,
+      "loss": 2.7317,
+      "step": 789
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.46619616046782564,
+      "learning_rate": 8.397636951769099e-06,
+      "loss": 2.6267,
+      "step": 790
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.41938914728535004,
+      "learning_rate": 8.39376139441133e-06,
+      "loss": 2.5999,
+      "step": 791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4128360578700934,
+      "learning_rate": 8.389882052566106e-06,
+      "loss": 2.6008,
+      "step": 792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4236692842899877,
+      "learning_rate": 8.3859989305594e-06,
+      "loss": 2.5726,
+      "step": 793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4078628362817052,
+      "learning_rate": 8.3821120327214e-06,
+      "loss": 2.5205,
+      "step": 794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39618573898162357,
+      "learning_rate": 8.378221363386506e-06,
+      "loss": 2.5956,
+      "step": 795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4034964151601362,
+      "learning_rate": 8.37432692689332e-06,
+      "loss": 2.4988,
+      "step": 796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43923027322011543,
+      "learning_rate": 8.370428727584654e-06,
+      "loss": 2.5139,
+      "step": 797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4248305389606218,
+      "learning_rate": 8.366526769807503e-06,
+      "loss": 2.5911,
+      "step": 798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4110205308990833,
+      "learning_rate": 8.362621057913063e-06,
+      "loss": 2.5665,
+      "step": 799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4144144901313341,
+      "learning_rate": 8.358711596256712e-06,
+      "loss": 2.5092,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4197156000281137,
+      "learning_rate": 8.354798389198013e-06,
+      "loss": 2.6169,
+      "step": 801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40972816288766706,
+      "learning_rate": 8.3508814411007e-06,
+      "loss": 2.6123,
+      "step": 802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41927212859085966,
+      "learning_rate": 8.346960756332683e-06,
+      "loss": 2.4813,
+      "step": 803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4272037158112424,
+      "learning_rate": 8.34303633926604e-06,
+      "loss": 2.5397,
+      "step": 804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41655980368998585,
+      "learning_rate": 8.339108194277006e-06,
+      "loss": 2.5427,
+      "step": 805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.41470497597592165,
+      "learning_rate": 8.33517632574598e-06,
+      "loss": 2.6004,
+      "step": 806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43266402910209373,
+      "learning_rate": 8.331240738057502e-06,
+      "loss": 2.5065,
+      "step": 807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4241724505370535,
+      "learning_rate": 8.327301435600273e-06,
+      "loss": 2.6018,
+      "step": 808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4055328825578019,
+      "learning_rate": 8.32335842276713e-06,
+      "loss": 2.5533,
+      "step": 809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4080727473950192,
+      "learning_rate": 8.319411703955042e-06,
+      "loss": 2.474,
+      "step": 810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3937313923147714,
+      "learning_rate": 8.315461283565123e-06,
+      "loss": 2.5147,
+      "step": 811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4165975592841607,
+      "learning_rate": 8.311507166002604e-06,
+      "loss": 2.5524,
+      "step": 812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.40304916540901103,
+      "learning_rate": 8.307549355676843e-06,
+      "loss": 2.6405,
+      "step": 813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4089370226477569,
+      "learning_rate": 8.303587857001316e-06,
+      "loss": 2.6325,
+      "step": 814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4013332900237908,
+      "learning_rate": 8.299622674393615e-06,
+      "loss": 2.5728,
+      "step": 815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3970217140730988,
+      "learning_rate": 8.29565381227543e-06,
+      "loss": 2.578,
+      "step": 816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4450471154194791,
+      "learning_rate": 8.291681275072568e-06,
+      "loss": 2.6582,
+      "step": 817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4022231800937019,
+      "learning_rate": 8.287705067214921e-06,
+      "loss": 2.5899,
+      "step": 818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.39869647865692215,
+      "learning_rate": 8.283725193136487e-06,
+      "loss": 2.4949,
+      "step": 819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38416378762137965,
+      "learning_rate": 8.27974165727534e-06,
+      "loss": 2.5464,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4233181605786404,
+      "learning_rate": 8.275754464073645e-06,
+      "loss": 2.6254,
+      "step": 821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4102451954629436,
+      "learning_rate": 8.271763617977642e-06,
+      "loss": 2.5013,
+      "step": 822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4252367825626563,
+      "learning_rate": 8.26776912343765e-06,
+      "loss": 2.6307,
+      "step": 823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4420150339625009,
+      "learning_rate": 8.263770984908049e-06,
+      "loss": 2.5286,
+      "step": 824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43126175076604883,
+      "learning_rate": 8.259769206847286e-06,
+      "loss": 2.6439,
+      "step": 825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4014097316497402,
+      "learning_rate": 8.255763793717868e-06,
+      "loss": 2.5087,
+      "step": 826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40908127650566717,
+      "learning_rate": 8.251754749986353e-06,
+      "loss": 2.5259,
+      "step": 827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44382902301414784,
+      "learning_rate": 8.247742080123351e-06,
+      "loss": 2.5059,
+      "step": 828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5690583326008019,
+      "learning_rate": 8.243725788603509e-06,
+      "loss": 2.5809,
+      "step": 829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4005616171832035,
+      "learning_rate": 8.239705879905519e-06,
+      "loss": 2.5534,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4180142030333501,
+      "learning_rate": 8.235682358512106e-06,
+      "loss": 2.6144,
+      "step": 831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4335665651339525,
+      "learning_rate": 8.231655228910022e-06,
+      "loss": 2.5939,
+      "step": 832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.44795603542160145,
+      "learning_rate": 8.227624495590041e-06,
+      "loss": 2.6328,
+      "step": 833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.39990249279797285,
+      "learning_rate": 8.223590163046956e-06,
+      "loss": 2.4888,
+      "step": 834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.422567364553395,
+      "learning_rate": 8.219552235779578e-06,
+      "loss": 2.5965,
+      "step": 835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4314727111378184,
+      "learning_rate": 8.215510718290724e-06,
+      "loss": 2.6093,
+      "step": 836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40041522438961896,
+      "learning_rate": 8.211465615087213e-06,
+      "loss": 2.5153,
+      "step": 837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41522560863984265,
+      "learning_rate": 8.207416930679862e-06,
+      "loss": 2.6018,
+      "step": 838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42312993496111684,
+      "learning_rate": 8.203364669583485e-06,
+      "loss": 2.5895,
+      "step": 839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42288979584341124,
+      "learning_rate": 8.199308836316883e-06,
+      "loss": 2.5965,
+      "step": 840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42051194007786136,
+      "learning_rate": 8.19524943540284e-06,
+      "loss": 2.4559,
+      "step": 841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42267226014337067,
+      "learning_rate": 8.191186471368115e-06,
+      "loss": 2.5894,
+      "step": 842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4204184855877659,
+      "learning_rate": 8.18711994874345e-06,
+      "loss": 2.5612,
+      "step": 843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41720484508218253,
+      "learning_rate": 8.183049872063543e-06,
+      "loss": 2.5136,
+      "step": 844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4464853468623226,
+      "learning_rate": 8.178976245867068e-06,
+      "loss": 2.635,
+      "step": 845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4286598401042708,
+      "learning_rate": 8.174899074696645e-06,
+      "loss": 2.6825,
+      "step": 846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43245122063916924,
+      "learning_rate": 8.170818363098854e-06,
+      "loss": 2.5948,
+      "step": 847
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41369230888120184,
+      "learning_rate": 8.166734115624223e-06,
+      "loss": 2.6412,
+      "step": 848
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4293671193850312,
+      "learning_rate": 8.162646336827222e-06,
+      "loss": 2.5823,
+      "step": 849
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4369219810371937,
+      "learning_rate": 8.158555031266255e-06,
+      "loss": 2.5921,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41473690029057125,
+      "learning_rate": 8.154460203503667e-06,
+      "loss": 2.547,
+      "step": 851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4312208386941552,
+      "learning_rate": 8.150361858105724e-06,
+      "loss": 2.5732,
+      "step": 852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4390124675983267,
+      "learning_rate": 8.146259999642614e-06,
+      "loss": 2.5189,
+      "step": 853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4137874265379334,
+      "learning_rate": 8.14215463268845e-06,
+      "loss": 2.5845,
+      "step": 854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39993164923934327,
+      "learning_rate": 8.138045761821252e-06,
+      "loss": 2.6329,
+      "step": 855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4437132444069032,
+      "learning_rate": 8.133933391622944e-06,
+      "loss": 2.6292,
+      "step": 856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4235821198963829,
+      "learning_rate": 8.129817526679358e-06,
+      "loss": 2.6303,
+      "step": 857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39602619625874363,
+      "learning_rate": 8.125698171580224e-06,
+      "loss": 2.6121,
+      "step": 858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4132650361368975,
+      "learning_rate": 8.121575330919159e-06,
+      "loss": 2.6254,
+      "step": 859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4030628075050604,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 2.7212,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40489013194172396,
+      "learning_rate": 8.113319211305141e-06,
+      "loss": 2.6366,
+      "step": 861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.44092957744334427,
+      "learning_rate": 8.109185941558842e-06,
+      "loss": 2.6076,
+      "step": 862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42045429985990546,
+      "learning_rate": 8.105049204663906e-06,
+      "loss": 2.5336,
+      "step": 863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42044467762139054,
+      "learning_rate": 8.100909005233336e-06,
+      "loss": 2.6656,
+      "step": 864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4176046443329716,
+      "learning_rate": 8.096765347883995e-06,
+      "loss": 2.5584,
+      "step": 865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.40906033106637607,
+      "learning_rate": 8.092618237236602e-06,
+      "loss": 2.4977,
+      "step": 866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41217420167561647,
+      "learning_rate": 8.088467677915728e-06,
+      "loss": 2.5494,
+      "step": 867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4164262634391922,
+      "learning_rate": 8.084313674549789e-06,
+      "loss": 2.7228,
+      "step": 868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4237722192976071,
+      "learning_rate": 8.080156231771043e-06,
+      "loss": 2.5715,
+      "step": 869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.414586509360156,
+      "learning_rate": 8.075995354215578e-06,
+      "loss": 2.6481,
+      "step": 870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3930741587756925,
+      "learning_rate": 8.07183104652332e-06,
+      "loss": 2.5999,
+      "step": 871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3800161956171746,
+      "learning_rate": 8.067663313338014e-06,
+      "loss": 2.565,
+      "step": 872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42395921077528126,
+      "learning_rate": 8.06349215930723e-06,
+      "loss": 2.5922,
+      "step": 873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.41039686865262626,
+      "learning_rate": 8.059317589082349e-06,
+      "loss": 2.5915,
+      "step": 874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4214562704564119,
+      "learning_rate": 8.055139607318558e-06,
+      "loss": 2.5322,
+      "step": 875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.42154062229902023,
+      "learning_rate": 8.050958218674859e-06,
+      "loss": 2.5986,
+      "step": 876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4310299450125749,
+      "learning_rate": 8.046773427814043e-06,
+      "loss": 2.6266,
+      "step": 877
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4395628220019876,
+      "learning_rate": 8.042585239402697e-06,
+      "loss": 2.5421,
+      "step": 878
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.43732728273627564,
+      "learning_rate": 8.038393658111203e-06,
+      "loss": 2.4944,
+      "step": 879
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4099976369161659,
+      "learning_rate": 8.03419868861372e-06,
+      "loss": 2.5705,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40913216034905114,
+      "learning_rate": 8.030000335588187e-06,
+      "loss": 2.5889,
+      "step": 881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3980903983537651,
+      "learning_rate": 8.025798603716314e-06,
+      "loss": 2.5876,
+      "step": 882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4298468554380081,
+      "learning_rate": 8.021593497683587e-06,
+      "loss": 2.5664,
+      "step": 883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.46062635167931587,
+      "learning_rate": 8.01738502217924e-06,
+      "loss": 2.644,
+      "step": 884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4229239386448534,
+      "learning_rate": 8.013173181896283e-06,
+      "loss": 2.5189,
+      "step": 885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4081552650831732,
+      "learning_rate": 8.008957981531462e-06,
+      "loss": 2.4752,
+      "step": 886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4331028195271111,
+      "learning_rate": 8.004739425785277e-06,
+      "loss": 2.5987,
+      "step": 887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40920188315560446,
+      "learning_rate": 8.00051751936197e-06,
+      "loss": 2.5385,
+      "step": 888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4316135933123064,
+      "learning_rate": 7.996292266969518e-06,
+      "loss": 2.5597,
+      "step": 889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4462340506740775,
+      "learning_rate": 7.992063673319632e-06,
+      "loss": 2.4633,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42580447002405547,
+      "learning_rate": 7.987831743127742e-06,
+      "loss": 2.4647,
+      "step": 891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41925259860489633,
+      "learning_rate": 7.983596481113005e-06,
+      "loss": 2.5792,
+      "step": 892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42194763194144164,
+      "learning_rate": 7.979357891998294e-06,
+      "loss": 2.5388,
+      "step": 893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4321751284259449,
+      "learning_rate": 7.975115980510187e-06,
+      "loss": 2.6989,
+      "step": 894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.412295945588522,
+      "learning_rate": 7.97087075137897e-06,
+      "loss": 2.5899,
+      "step": 895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39834515118459113,
+      "learning_rate": 7.966622209338629e-06,
+      "loss": 2.589,
+      "step": 896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.413233484922437,
+      "learning_rate": 7.962370359126843e-06,
+      "loss": 2.6658,
+      "step": 897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4605783421036789,
+      "learning_rate": 7.95811520548498e-06,
+      "loss": 2.6009,
+      "step": 898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.44280924675391853,
+      "learning_rate": 7.953856753158096e-06,
+      "loss": 2.6904,
+      "step": 899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.399962256558219,
+      "learning_rate": 7.949595006894917e-06,
+      "loss": 2.5138,
+      "step": 900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4065872956264197,
+      "learning_rate": 7.945329971447851e-06,
+      "loss": 2.5331,
+      "step": 901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43139524087257114,
+      "learning_rate": 7.941061651572965e-06,
+      "loss": 2.654,
+      "step": 902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.43610262802638144,
+      "learning_rate": 7.93679005203e-06,
+      "loss": 2.4393,
+      "step": 903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4305492602763158,
+      "learning_rate": 7.932515177582341e-06,
+      "loss": 2.6138,
+      "step": 904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4221443270245523,
+      "learning_rate": 7.928237032997037e-06,
+      "loss": 2.618,
+      "step": 905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4422278732901728,
+      "learning_rate": 7.923955623044775e-06,
+      "loss": 2.6176,
+      "step": 906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4266337688002018,
+      "learning_rate": 7.919670952499889e-06,
+      "loss": 2.4747,
+      "step": 907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4477483651287364,
+      "learning_rate": 7.915383026140344e-06,
+      "loss": 2.5751,
+      "step": 908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4177444069840092,
+      "learning_rate": 7.91109184874774e-06,
+      "loss": 2.6049,
+      "step": 909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.42079897809566313,
+      "learning_rate": 7.9067974251073e-06,
+      "loss": 2.5587,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42684865722248433,
+      "learning_rate": 7.902499760007867e-06,
+      "loss": 2.5912,
+      "step": 911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4071847359013808,
+      "learning_rate": 7.898198858241902e-06,
+      "loss": 2.5533,
+      "step": 912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4272959827178456,
+      "learning_rate": 7.89389472460547e-06,
+      "loss": 2.6196,
+      "step": 913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4567896212371643,
+      "learning_rate": 7.889587363898242e-06,
+      "loss": 2.6187,
+      "step": 914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4213243352124343,
+      "learning_rate": 7.885276780923488e-06,
+      "loss": 2.6833,
+      "step": 915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41023311802953555,
+      "learning_rate": 7.880962980488073e-06,
+      "loss": 2.6689,
+      "step": 916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.421825264718791,
+      "learning_rate": 7.876645967402448e-06,
+      "loss": 2.608,
+      "step": 917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43157690967476947,
+      "learning_rate": 7.872325746480645e-06,
+      "loss": 2.534,
+      "step": 918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4513483239147945,
+      "learning_rate": 7.868002322540279e-06,
+      "loss": 2.6496,
+      "step": 919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.44068006685442734,
+      "learning_rate": 7.863675700402527e-06,
+      "loss": 2.6127,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4340762297818696,
+      "learning_rate": 7.859345884892143e-06,
+      "loss": 2.583,
+      "step": 921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5535594322513243,
+      "learning_rate": 7.855012880837432e-06,
+      "loss": 2.4857,
+      "step": 922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4725111746601944,
+      "learning_rate": 7.850676693070266e-06,
+      "loss": 2.5878,
+      "step": 923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41724059592129575,
+      "learning_rate": 7.846337326426057e-06,
+      "loss": 2.5041,
+      "step": 924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42325495272452096,
+      "learning_rate": 7.841994785743765e-06,
+      "loss": 2.4939,
+      "step": 925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4448404822113957,
+      "learning_rate": 7.837649075865893e-06,
+      "loss": 2.7193,
+      "step": 926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.411834046291556,
+      "learning_rate": 7.833300201638475e-06,
+      "loss": 2.5689,
+      "step": 927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.43112137576355314,
+      "learning_rate": 7.828948167911073e-06,
+      "loss": 2.4763,
+      "step": 928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41680385794684577,
+      "learning_rate": 7.824592979536774e-06,
+      "loss": 2.6338,
+      "step": 929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4122113287041295,
+      "learning_rate": 7.820234641372182e-06,
+      "loss": 2.4718,
+      "step": 930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4246695467962534,
+      "learning_rate": 7.815873158277414e-06,
+      "loss": 2.5562,
+      "step": 931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.42277124360103996,
+      "learning_rate": 7.811508535116093e-06,
+      "loss": 2.4938,
+      "step": 932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4513815520578765,
+      "learning_rate": 7.807140776755345e-06,
+      "loss": 2.4926,
+      "step": 933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45996391797568054,
+      "learning_rate": 7.80276988806579e-06,
+      "loss": 2.7031,
+      "step": 934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4831176558870303,
+      "learning_rate": 7.798395873921542e-06,
+      "loss": 2.5952,
+      "step": 935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45636962210820387,
+      "learning_rate": 7.794018739200196e-06,
+      "loss": 2.4938,
+      "step": 936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4582229304281336,
+      "learning_rate": 7.789638488782832e-06,
+      "loss": 2.748,
+      "step": 937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4515473425938092,
+      "learning_rate": 7.785255127553999e-06,
+      "loss": 2.6467,
+      "step": 938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4435663622348958,
+      "learning_rate": 7.780868660401719e-06,
+      "loss": 2.5819,
+      "step": 939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.41412840624582015,
+      "learning_rate": 7.776479092217475e-06,
+      "loss": 2.5298,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43024657897653096,
+      "learning_rate": 7.772086427896211e-06,
+      "loss": 2.571,
+      "step": 941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4374848771887548,
+      "learning_rate": 7.767690672336321e-06,
+      "loss": 2.5744,
+      "step": 942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42377155119745097,
+      "learning_rate": 7.763291830439646e-06,
+      "loss": 2.4888,
+      "step": 943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4385016759124883,
+      "learning_rate": 7.75888990711147e-06,
+      "loss": 2.6311,
+      "step": 944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4447177294698723,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 2.6521,
+      "step": 945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4077239806470596,
+      "learning_rate": 7.750076835798927e-06,
+      "loss": 2.5797,
+      "step": 946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.428657823891645,
+      "learning_rate": 7.745665697642284e-06,
+      "loss": 2.5617,
+      "step": 947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4354502823126484,
+      "learning_rate": 7.741251497709583e-06,
+      "loss": 2.6247,
+      "step": 948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4633385145231587,
+      "learning_rate": 7.736834240923233e-06,
+      "loss": 2.6145,
+      "step": 949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.398561346216612,
+      "learning_rate": 7.73241393220905e-06,
+      "loss": 2.4895,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4329616819731975,
+      "learning_rate": 7.727990576496259e-06,
+      "loss": 2.6225,
+      "step": 951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4285001823264661,
+      "learning_rate": 7.723564178717477e-06,
+      "loss": 2.4792,
+      "step": 952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4799081586852101,
+      "learning_rate": 7.71913474380872e-06,
+      "loss": 2.645,
+      "step": 953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.440591651249703,
+      "learning_rate": 7.714702276709382e-06,
+      "loss": 2.6201,
+      "step": 954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42045495551086975,
+      "learning_rate": 7.710266782362248e-06,
+      "loss": 2.4546,
+      "step": 955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44996984963607556,
+      "learning_rate": 7.705828265713468e-06,
+      "loss": 2.5964,
+      "step": 956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44860901294069305,
+      "learning_rate": 7.701386731712575e-06,
+      "loss": 2.579,
+      "step": 957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4465100265169577,
+      "learning_rate": 7.696942185312453e-06,
+      "loss": 2.5449,
+      "step": 958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44784574908323493,
+      "learning_rate": 7.692494631469356e-06,
+      "loss": 2.4899,
+      "step": 959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43842323845570746,
+      "learning_rate": 7.688044075142888e-06,
+      "loss": 2.5747,
+      "step": 960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43228122591332846,
+      "learning_rate": 7.683590521295998e-06,
+      "loss": 2.5739,
+      "step": 961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.434804720290425,
+      "learning_rate": 7.679133974894984e-06,
+      "loss": 2.6809,
+      "step": 962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42894342327405344,
+      "learning_rate": 7.674674440909475e-06,
+      "loss": 2.5423,
+      "step": 963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.41596006640315525,
+      "learning_rate": 7.670211924312437e-06,
+      "loss": 2.6565,
+      "step": 964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4399642006984142,
+      "learning_rate": 7.665746430080155e-06,
+      "loss": 2.5905,
+      "step": 965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43095602575237124,
+      "learning_rate": 7.661277963192241e-06,
+      "loss": 2.5439,
+      "step": 966
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42881278390211974,
+      "learning_rate": 7.65680652863162e-06,
+      "loss": 2.5704,
+      "step": 967
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4403334552659803,
+      "learning_rate": 7.652332131384526e-06,
+      "loss": 2.5899,
+      "step": 968
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43952778499904116,
+      "learning_rate": 7.647854776440496e-06,
+      "loss": 2.6278,
+      "step": 969
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.44350619132871466,
+      "learning_rate": 7.643374468792364e-06,
+      "loss": 2.5577,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43480932970578706,
+      "learning_rate": 7.638891213436261e-06,
+      "loss": 2.5002,
+      "step": 971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4326045635095116,
+      "learning_rate": 7.634405015371603e-06,
+      "loss": 2.5841,
+      "step": 972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45248147057167837,
+      "learning_rate": 7.629915879601087e-06,
+      "loss": 2.6901,
+      "step": 973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.47546982179960307,
+      "learning_rate": 7.625423811130681e-06,
+      "loss": 2.533,
+      "step": 974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42614586992931125,
+      "learning_rate": 7.620928814969636e-06,
+      "loss": 2.6475,
+      "step": 975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4395102230701426,
+      "learning_rate": 7.616430896130456e-06,
+      "loss": 2.5916,
+      "step": 976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43456558053515715,
+      "learning_rate": 7.611930059628909e-06,
+      "loss": 2.5627,
+      "step": 977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.459423844990916,
+      "learning_rate": 7.607426310484014e-06,
+      "loss": 2.522,
+      "step": 978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43233604854428664,
+      "learning_rate": 7.602919653718044e-06,
+      "loss": 2.5904,
+      "step": 979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4697293541204728,
+      "learning_rate": 7.5984100943565055e-06,
+      "loss": 2.5513,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4225011347016215,
+      "learning_rate": 7.5938976374281515e-06,
+      "loss": 2.4909,
+      "step": 981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43926955733620127,
+      "learning_rate": 7.589382287964957e-06,
+      "loss": 2.6104,
+      "step": 982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45591129115222123,
+      "learning_rate": 7.584864051002128e-06,
+      "loss": 2.7106,
+      "step": 983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4349453283077686,
+      "learning_rate": 7.58034293157809e-06,
+      "loss": 2.5365,
+      "step": 984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42835368432163357,
+      "learning_rate": 7.57581893473448e-06,
+      "loss": 2.5712,
+      "step": 985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.425073700713382,
+      "learning_rate": 7.571292065516149e-06,
+      "loss": 2.5438,
+      "step": 986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43117639770458216,
+      "learning_rate": 7.566762328971146e-06,
+      "loss": 2.5966,
+      "step": 987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4350355422638024,
+      "learning_rate": 7.562229730150716e-06,
+      "loss": 2.5597,
+      "step": 988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.45651613860416895,
+      "learning_rate": 7.557694274109305e-06,
+      "loss": 2.6819,
+      "step": 989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4178926630375163,
+      "learning_rate": 7.553155965904535e-06,
+      "loss": 2.6029,
+      "step": 990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42667679705117045,
+      "learning_rate": 7.548614810597214e-06,
+      "loss": 2.513,
+      "step": 991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42468254701197683,
+      "learning_rate": 7.544070813251324e-06,
+      "loss": 2.5881,
+      "step": 992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43467675094971464,
+      "learning_rate": 7.5395239789340155e-06,
+      "loss": 2.5797,
+      "step": 993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4389732792842581,
+      "learning_rate": 7.534974312715603e-06,
+      "loss": 2.5709,
+      "step": 994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4411521788061547,
+      "learning_rate": 7.530421819669558e-06,
+      "loss": 2.6326,
+      "step": 995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4557196879536489,
+      "learning_rate": 7.5258665048725065e-06,
+      "loss": 2.596,
+      "step": 996
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43871902185286676,
+      "learning_rate": 7.521308373404218e-06,
+      "loss": 2.5556,
+      "step": 997
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.42755630130788425,
+      "learning_rate": 7.516747430347607e-06,
+      "loss": 2.7271,
+      "step": 998
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4318108374082864,
+      "learning_rate": 7.51218368078872e-06,
+      "loss": 2.5897,
+      "step": 999
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.43372744999074364,
+      "learning_rate": 7.507617129816733e-06,
+      "loss": 2.6207,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4349061846690901,
+      "learning_rate": 7.503047782523949e-06,
+      "loss": 2.5824,
+      "step": 1001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44955498095402496,
+      "learning_rate": 7.498475644005787e-06,
+      "loss": 2.5644,
+      "step": 1002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4421726426503066,
+      "learning_rate": 7.4939007193607786e-06,
+      "loss": 2.6945,
+      "step": 1003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.41894787340781975,
+      "learning_rate": 7.489323013690561e-06,
+      "loss": 2.5864,
+      "step": 1004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43641982246269445,
+      "learning_rate": 7.484742532099878e-06,
+      "loss": 2.5838,
+      "step": 1005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.433542970312903,
+      "learning_rate": 7.480159279696562e-06,
+      "loss": 2.5402,
+      "step": 1006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.47002219178179716,
+      "learning_rate": 7.4755732615915385e-06,
+      "loss": 2.6275,
+      "step": 1007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42655592623385635,
+      "learning_rate": 7.470984482898818e-06,
+      "loss": 2.5379,
+      "step": 1008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4490796246910226,
+      "learning_rate": 7.46639294873549e-06,
+      "loss": 2.5878,
+      "step": 1009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4353762578998085,
+      "learning_rate": 7.461798664221711e-06,
+      "loss": 2.5528,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4359763726298193,
+      "learning_rate": 7.457201634480713e-06,
+      "loss": 2.4963,
+      "step": 1011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4293488989036144,
+      "learning_rate": 7.452601864638782e-06,
+      "loss": 2.507,
+      "step": 1012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4632695896167269,
+      "learning_rate": 7.447999359825263e-06,
+      "loss": 2.5688,
+      "step": 1013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4717146582108519,
+      "learning_rate": 7.443394125172552e-06,
+      "loss": 2.562,
+      "step": 1014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.45926730586833125,
+      "learning_rate": 7.438786165816084e-06,
+      "loss": 2.5078,
+      "step": 1015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42035212985829007,
+      "learning_rate": 7.434175486894339e-06,
+      "loss": 2.6623,
+      "step": 1016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.442070486539419,
+      "learning_rate": 7.429562093548825e-06,
+      "loss": 2.5572,
+      "step": 1017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4346339147079722,
+      "learning_rate": 7.42494599092408e-06,
+      "loss": 2.5031,
+      "step": 1018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4486843745453151,
+      "learning_rate": 7.42032718416766e-06,
+      "loss": 2.6138,
+      "step": 1019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.454243627496171,
+      "learning_rate": 7.415705678430138e-06,
+      "loss": 2.569,
+      "step": 1020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4378035377930223,
+      "learning_rate": 7.4110814788651e-06,
+      "loss": 2.5737,
+      "step": 1021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4408324571659681,
+      "learning_rate": 7.406454590629129e-06,
+      "loss": 2.5312,
+      "step": 1022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43626638605774093,
+      "learning_rate": 7.401825018881812e-06,
+      "loss": 2.6114,
+      "step": 1023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.45581278624186283,
+      "learning_rate": 7.397192768785724e-06,
+      "loss": 2.6438,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43570018858293746,
+      "learning_rate": 7.392557845506433e-06,
+      "loss": 2.5748,
+      "step": 1025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4209876795639996,
+      "learning_rate": 7.387920254212482e-06,
+      "loss": 2.5272,
+      "step": 1026
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.44083438389417995,
+      "learning_rate": 7.383280000075392e-06,
+      "loss": 2.5001,
+      "step": 1027
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.453200235369117,
+      "learning_rate": 7.378637088269652e-06,
+      "loss": 2.6287,
+      "step": 1028
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.43011889909731615,
+      "learning_rate": 7.373991523972715e-06,
+      "loss": 2.5724,
+      "step": 1029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4442709068247595,
+      "learning_rate": 7.369343312364994e-06,
+      "loss": 2.681,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43930051060921954,
+      "learning_rate": 7.3646924586298515e-06,
+      "loss": 2.4956,
+      "step": 1031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.44486461669517774,
+      "learning_rate": 7.360038967953598e-06,
+      "loss": 2.617,
+      "step": 1032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4337851986633554,
+      "learning_rate": 7.355382845525481e-06,
+      "loss": 2.5871,
+      "step": 1033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4768810458318751,
+      "learning_rate": 7.350724096537687e-06,
+      "loss": 2.6812,
+      "step": 1034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4294759402946892,
+      "learning_rate": 7.346062726185332e-06,
+      "loss": 2.4668,
+      "step": 1035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4340582069750426,
+      "learning_rate": 7.341398739666451e-06,
+      "loss": 2.4101,
+      "step": 1036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.40832331589977605,
+      "learning_rate": 7.336732142181998e-06,
+      "loss": 2.512,
+      "step": 1037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4447890099201175,
+      "learning_rate": 7.33206293893584e-06,
+      "loss": 2.6111,
+      "step": 1038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4559691714502572,
+      "learning_rate": 7.3273911351347495e-06,
+      "loss": 2.6645,
+      "step": 1039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45753809546198526,
+      "learning_rate": 7.3227167359883964e-06,
+      "loss": 2.6484,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4663086257055781,
+      "learning_rate": 7.318039746709349e-06,
+      "loss": 2.6783,
+      "step": 1041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41836926175215,
+      "learning_rate": 7.313360172513059e-06,
+      "loss": 2.6437,
+      "step": 1042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4192792544101712,
+      "learning_rate": 7.3086780186178645e-06,
+      "loss": 2.4873,
+      "step": 1043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4517224408067883,
+      "learning_rate": 7.30399329024498e-06,
+      "loss": 2.6459,
+      "step": 1044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42751398897427867,
+      "learning_rate": 7.299305992618488e-06,
+      "loss": 2.5837,
+      "step": 1045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4265308955295411,
+      "learning_rate": 7.2946161309653375e-06,
+      "loss": 2.5892,
+      "step": 1046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4238407467136127,
+      "learning_rate": 7.289923710515338e-06,
+      "loss": 2.5839,
+      "step": 1047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4614155278757042,
+      "learning_rate": 7.285228736501153e-06,
+      "loss": 2.6102,
+      "step": 1048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4422958659102835,
+      "learning_rate": 7.280531214158291e-06,
+      "loss": 2.5856,
+      "step": 1049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42237143095235286,
+      "learning_rate": 7.275831148725101e-06,
+      "loss": 2.5692,
+      "step": 1050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.45361904637478645,
+      "learning_rate": 7.271128545442774e-06,
+      "loss": 2.5652,
+      "step": 1051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4337472729775839,
+      "learning_rate": 7.266423409555324e-06,
+      "loss": 2.6164,
+      "step": 1052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42362391039093894,
+      "learning_rate": 7.261715746309594e-06,
+      "loss": 2.5273,
+      "step": 1053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.431012140550024,
+      "learning_rate": 7.257005560955245e-06,
+      "loss": 2.5411,
+      "step": 1054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.452142900799659,
+      "learning_rate": 7.252292858744747e-06,
+      "loss": 2.5771,
+      "step": 1055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.43273717452006466,
+      "learning_rate": 7.247577644933379e-06,
+      "loss": 2.6122,
+      "step": 1056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.462962261724824,
+      "learning_rate": 7.242859924779221e-06,
+      "loss": 2.6279,
+      "step": 1057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4258143210910348,
+      "learning_rate": 7.238139703543148e-06,
+      "loss": 2.5806,
+      "step": 1058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4535192221696845,
+      "learning_rate": 7.233416986488822e-06,
+      "loss": 2.609,
+      "step": 1059
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4302664584348388,
+      "learning_rate": 7.2286917788826926e-06,
+      "loss": 2.5662,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46195490134362327,
+      "learning_rate": 7.2239640859939795e-06,
+      "loss": 2.6167,
+      "step": 1061
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.44182897260584386,
+      "learning_rate": 7.219233913094683e-06,
+      "loss": 2.5813,
+      "step": 1062
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4434406021256981,
+      "learning_rate": 7.214501265459561e-06,
+      "loss": 2.6143,
+      "step": 1063
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4315144971686448,
+      "learning_rate": 7.2097661483661355e-06,
+      "loss": 2.6934,
+      "step": 1064
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.441969335136358,
+      "learning_rate": 7.2050285670946776e-06,
+      "loss": 2.4584,
+      "step": 1065
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4295633483472062,
+      "learning_rate": 7.200288526928214e-06,
+      "loss": 2.5537,
+      "step": 1066
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4590776064719756,
+      "learning_rate": 7.195546033152506e-06,
+      "loss": 2.6182,
+      "step": 1067
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5628999506795336,
+      "learning_rate": 7.1908010910560555e-06,
+      "loss": 2.5655,
+      "step": 1068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43083319078636145,
+      "learning_rate": 7.18605370593009e-06,
+      "loss": 2.5838,
+      "step": 1069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43705338433970153,
+      "learning_rate": 7.181303883068569e-06,
+      "loss": 2.556,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.41626277262476946,
+      "learning_rate": 7.176551627768163e-06,
+      "loss": 2.4507,
+      "step": 1071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8154045786389643,
+      "learning_rate": 7.171796945328257e-06,
+      "loss": 2.5791,
+      "step": 1072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6811278417412138,
+      "learning_rate": 7.167039841050947e-06,
+      "loss": 2.6046,
+      "step": 1073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4469455756630236,
+      "learning_rate": 7.16228032024102e-06,
+      "loss": 2.5503,
+      "step": 1074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.49527095536212423,
+      "learning_rate": 7.157518388205969e-06,
+      "loss": 2.446,
+      "step": 1075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5650785736288945,
+      "learning_rate": 7.152754050255967e-06,
+      "loss": 2.5739,
+      "step": 1076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4688134150524742,
+      "learning_rate": 7.147987311703875e-06,
+      "loss": 2.53,
+      "step": 1077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4343253484657035,
+      "learning_rate": 7.143218177865229e-06,
+      "loss": 2.6128,
+      "step": 1078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4380533261518903,
+      "learning_rate": 7.138446654058238e-06,
+      "loss": 2.5129,
+      "step": 1079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43821048947336705,
+      "learning_rate": 7.1336727456037716e-06,
+      "loss": 2.6101,
+      "step": 1080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42895471337291513,
+      "learning_rate": 7.128896457825364e-06,
+      "loss": 2.5302,
+      "step": 1081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46383576939242727,
+      "learning_rate": 7.124117796049199e-06,
+      "loss": 2.5467,
+      "step": 1082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46391080952373015,
+      "learning_rate": 7.11933676560411e-06,
+      "loss": 2.5796,
+      "step": 1083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43071291553448376,
+      "learning_rate": 7.114553371821571e-06,
+      "loss": 2.6143,
+      "step": 1084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.42639088514580703,
+      "learning_rate": 7.109767620035689e-06,
+      "loss": 2.5842,
+      "step": 1085
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4565996104308099,
+      "learning_rate": 7.104979515583206e-06,
+      "loss": 2.6329,
+      "step": 1086
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43550191741328553,
+      "learning_rate": 7.10018906380348e-06,
+      "loss": 2.5939,
+      "step": 1087
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.43752917124361557,
+      "learning_rate": 7.0953962700384935e-06,
+      "loss": 2.4761,
+      "step": 1088
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.45127428221951654,
+      "learning_rate": 7.090601139632838e-06,
+      "loss": 2.6407,
+      "step": 1089
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44095063114958083,
+      "learning_rate": 7.0858036779337095e-06,
+      "loss": 2.4693,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4569369650853151,
+      "learning_rate": 7.081003890290905e-06,
+      "loss": 2.6409,
+      "step": 1091
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44391721342003115,
+      "learning_rate": 7.076201782056816e-06,
+      "loss": 2.5961,
+      "step": 1092
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4809099048433727,
+      "learning_rate": 7.071397358586418e-06,
+      "loss": 2.7127,
+      "step": 1093
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44804630695230513,
+      "learning_rate": 7.066590625237274e-06,
+      "loss": 2.5516,
+      "step": 1094
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4659208433166112,
+      "learning_rate": 7.061781587369518e-06,
+      "loss": 2.6022,
+      "step": 1095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44289040654813444,
+      "learning_rate": 7.056970250345857e-06,
+      "loss": 2.5322,
+      "step": 1096
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.43869869323441935,
+      "learning_rate": 7.052156619531559e-06,
+      "loss": 2.5477,
+      "step": 1097
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.46781991237967846,
+      "learning_rate": 7.047340700294454e-06,
+      "loss": 2.5365,
+      "step": 1098
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5023220205872594,
+      "learning_rate": 7.042522498004918e-06,
+      "loss": 2.6717,
+      "step": 1099
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.46706191323763,
+      "learning_rate": 7.037702018035878e-06,
+      "loss": 2.5515,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4628447819796725,
+      "learning_rate": 7.032879265762799e-06,
+      "loss": 2.655,
+      "step": 1101
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4363362126296687,
+      "learning_rate": 7.028054246563679e-06,
+      "loss": 2.4911,
+      "step": 1102
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.45657932728300715,
+      "learning_rate": 7.023226965819046e-06,
+      "loss": 2.5302,
+      "step": 1103
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4607815559518134,
+      "learning_rate": 7.01839742891195e-06,
+      "loss": 2.526,
+      "step": 1104
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4160020494708598,
+      "learning_rate": 7.013565641227954e-06,
+      "loss": 2.4879,
+      "step": 1105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4302714974199038,
+      "learning_rate": 7.008731608155133e-06,
+      "loss": 2.6262,
+      "step": 1106
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.41895900606126363,
+      "learning_rate": 7.003895335084065e-06,
+      "loss": 2.4959,
+      "step": 1107
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.47937350542713325,
+      "learning_rate": 6.999056827407828e-06,
+      "loss": 2.5052,
+      "step": 1108
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4615463825839077,
+      "learning_rate": 6.994216090521991e-06,
+      "loss": 2.6507,
+      "step": 1109
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.43451698919033627,
+      "learning_rate": 6.989373129824605e-06,
+      "loss": 2.4712,
+      "step": 1110
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.46028446554987273,
+      "learning_rate": 6.984527950716207e-06,
+      "loss": 2.7052,
+      "step": 1111
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4737731411095733,
+      "learning_rate": 6.979680558599803e-06,
+      "loss": 2.5551,
+      "step": 1112
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.41582506140366343,
+      "learning_rate": 6.97483095888087e-06,
+      "loss": 2.5221,
+      "step": 1113
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44955640625002374,
+      "learning_rate": 6.969979156967343e-06,
+      "loss": 2.5921,
+      "step": 1114
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.41587551426031605,
+      "learning_rate": 6.965125158269619e-06,
+      "loss": 2.4685,
+      "step": 1115
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.506174379752642,
+      "learning_rate": 6.960268968200539e-06,
+      "loss": 2.642,
+      "step": 1116
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5083226209303883,
+      "learning_rate": 6.955410592175389e-06,
+      "loss": 2.4504,
+      "step": 1117
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5037711206272876,
+      "learning_rate": 6.950550035611894e-06,
+      "loss": 2.6059,
+      "step": 1118
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.489354807362807,
+      "learning_rate": 6.945687303930208e-06,
+      "loss": 2.5988,
+      "step": 1119
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43832044647163076,
+      "learning_rate": 6.940822402552914e-06,
+      "loss": 2.632,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.41684219914751336,
+      "learning_rate": 6.935955336905012e-06,
+      "loss": 2.4968,
+      "step": 1121
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4551566387768076,
+      "learning_rate": 6.931086112413919e-06,
+      "loss": 2.5133,
+      "step": 1122
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44312495281396785,
+      "learning_rate": 6.926214734509451e-06,
+      "loss": 2.6446,
+      "step": 1123
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4666828989997829,
+      "learning_rate": 6.921341208623838e-06,
+      "loss": 2.7543,
+      "step": 1124
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.46668743303052485,
+      "learning_rate": 6.916465540191692e-06,
+      "loss": 2.6665,
+      "step": 1125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4498249441961465,
+      "learning_rate": 6.911587734650024e-06,
+      "loss": 2.6079,
+      "step": 1126
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4510315846233333,
+      "learning_rate": 6.906707797438223e-06,
+      "loss": 2.611,
+      "step": 1127
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4573631385730997,
+      "learning_rate": 6.901825733998056e-06,
+      "loss": 2.4184,
+      "step": 1128
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4715636369250365,
+      "learning_rate": 6.896941549773666e-06,
+      "loss": 2.6183,
+      "step": 1129
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4484713099690168,
+      "learning_rate": 6.892055250211552e-06,
+      "loss": 2.5424,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4316676204749411,
+      "learning_rate": 6.8871668407605796e-06,
+      "loss": 2.535,
+      "step": 1131
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43208885222760374,
+      "learning_rate": 6.88227632687196e-06,
+      "loss": 2.4664,
+      "step": 1132
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4406342295768975,
+      "learning_rate": 6.87738371399926e-06,
+      "loss": 2.6311,
+      "step": 1133
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4577854161502282,
+      "learning_rate": 6.872489007598381e-06,
+      "loss": 2.5439,
+      "step": 1134
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4600202194206776,
+      "learning_rate": 6.867592213127559e-06,
+      "loss": 2.6258,
+      "step": 1135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.44800496524801076,
+      "learning_rate": 6.862693336047361e-06,
+      "loss": 2.521,
+      "step": 1136
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.46499483874748343,
+      "learning_rate": 6.8577923818206724e-06,
+      "loss": 2.5714,
+      "step": 1137
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43889677646580477,
+      "learning_rate": 6.852889355912702e-06,
+      "loss": 2.5829,
+      "step": 1138
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4844385997487572,
+      "learning_rate": 6.8479842637909624e-06,
+      "loss": 2.5358,
+      "step": 1139
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4568251772849157,
+      "learning_rate": 6.84307711092527e-06,
+      "loss": 2.459,
+      "step": 1140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.43900774957164473,
+      "learning_rate": 6.8381679027877456e-06,
+      "loss": 2.5391,
+      "step": 1141
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4354751938478969,
+      "learning_rate": 6.833256644852795e-06,
+      "loss": 2.5212,
+      "step": 1142
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45457266704424526,
+      "learning_rate": 6.828343342597114e-06,
+      "loss": 2.6235,
+      "step": 1143
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4757255960119849,
+      "learning_rate": 6.823428001499677e-06,
+      "loss": 2.5648,
+      "step": 1144
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.47773191150395056,
+      "learning_rate": 6.81851062704173e-06,
+      "loss": 2.7091,
+      "step": 1145
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4584731156153877,
+      "learning_rate": 6.813591224706788e-06,
+      "loss": 2.6848,
+      "step": 1146
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4394450886309457,
+      "learning_rate": 6.8086697999806295e-06,
+      "loss": 2.5986,
+      "step": 1147
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4544126859932802,
+      "learning_rate": 6.803746358351285e-06,
+      "loss": 2.6201,
+      "step": 1148
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45893198891226583,
+      "learning_rate": 6.798820905309036e-06,
+      "loss": 2.6446,
+      "step": 1149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46888579944630593,
+      "learning_rate": 6.793893446346405e-06,
+      "loss": 2.601,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5002077386363459,
+      "learning_rate": 6.788963986958153e-06,
+      "loss": 2.5332,
+      "step": 1151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46513972756972205,
+      "learning_rate": 6.784032532641273e-06,
+      "loss": 2.6863,
+      "step": 1152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4698499489753665,
+      "learning_rate": 6.779099088894979e-06,
+      "loss": 2.6624,
+      "step": 1153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5120231363761293,
+      "learning_rate": 6.774163661220708e-06,
+      "loss": 2.5264,
+      "step": 1154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43200849159270543,
+      "learning_rate": 6.769226255122104e-06,
+      "loss": 2.4693,
+      "step": 1155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.43094610204264894,
+      "learning_rate": 6.764286876105019e-06,
+      "loss": 2.5116,
+      "step": 1156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4605835347553326,
+      "learning_rate": 6.759345529677512e-06,
+      "loss": 2.5829,
+      "step": 1157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4710670425663677,
+      "learning_rate": 6.754402221349825e-06,
+      "loss": 2.4705,
+      "step": 1158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.439500763748165,
+      "learning_rate": 6.749456956634393e-06,
+      "loss": 2.6056,
+      "step": 1159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4895156211562015,
+      "learning_rate": 6.744509741045835e-06,
+      "loss": 2.5483,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42577663034155283,
+      "learning_rate": 6.739560580100941e-06,
+      "loss": 2.6488,
+      "step": 1161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4707206627835328,
+      "learning_rate": 6.734609479318671e-06,
+      "loss": 2.5462,
+      "step": 1162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4601301137240529,
+      "learning_rate": 6.7296564442201524e-06,
+      "loss": 2.5521,
+      "step": 1163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4450602235118456,
+      "learning_rate": 6.7247014803286645e-06,
+      "loss": 2.5476,
+      "step": 1164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4349271530782206,
+      "learning_rate": 6.719744593169642e-06,
+      "loss": 2.5598,
+      "step": 1165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46199999407153103,
+      "learning_rate": 6.714785788270658e-06,
+      "loss": 2.5668,
+      "step": 1166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4779666476345309,
+      "learning_rate": 6.709825071161431e-06,
+      "loss": 2.5069,
+      "step": 1167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46497301500221716,
+      "learning_rate": 6.704862447373808e-06,
+      "loss": 2.5101,
+      "step": 1168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.47375077813721217,
+      "learning_rate": 6.699897922441762e-06,
+      "loss": 2.638,
+      "step": 1169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4643578754928264,
+      "learning_rate": 6.6949315019013895e-06,
+      "loss": 2.6091,
+      "step": 1170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44590990249792184,
+      "learning_rate": 6.689963191290897e-06,
+      "loss": 2.5877,
+      "step": 1171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.46897866711565833,
+      "learning_rate": 6.684992996150599e-06,
+      "loss": 2.6591,
+      "step": 1172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.45022271790738155,
+      "learning_rate": 6.680020922022914e-06,
+      "loss": 2.5427,
+      "step": 1173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4587583269304219,
+      "learning_rate": 6.675046974452354e-06,
+      "loss": 2.5613,
+      "step": 1174
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49630174256587056,
+      "learning_rate": 6.670071158985521e-06,
+      "loss": 2.7151,
+      "step": 1175
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44245111809691745,
+      "learning_rate": 6.6650934811710984e-06,
+      "loss": 2.5714,
+      "step": 1176
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4787174499890503,
+      "learning_rate": 6.6601139465598485e-06,
+      "loss": 2.5309,
+      "step": 1177
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4512958769051424,
+      "learning_rate": 6.655132560704602e-06,
+      "loss": 2.54,
+      "step": 1178
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.443250600772457,
+      "learning_rate": 6.650149329160258e-06,
+      "loss": 2.6532,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4509528813478273,
+      "learning_rate": 6.645164257483767e-06,
+      "loss": 2.634,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4363624395334654,
+      "learning_rate": 6.640177351234137e-06,
+      "loss": 2.5797,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.472388215163046,
+      "learning_rate": 6.635188615972421e-06,
+      "loss": 2.493,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43214345788970226,
+      "learning_rate": 6.63019805726171e-06,
+      "loss": 2.5832,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47246380505868424,
+      "learning_rate": 6.625205680667128e-06,
+      "loss": 2.6045,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45820506063769545,
+      "learning_rate": 6.62021149175583e-06,
+      "loss": 2.6442,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4635029537308452,
+      "learning_rate": 6.615215496096987e-06,
+      "loss": 2.6954,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.46448871934946745,
+      "learning_rate": 6.610217699261788e-06,
+      "loss": 2.4956,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4633216299950541,
+      "learning_rate": 6.60521810682343e-06,
+      "loss": 2.4558,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47388358801825714,
+      "learning_rate": 6.600216724357112e-06,
+      "loss": 2.6062,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47421097222726205,
+      "learning_rate": 6.595213557440026e-06,
+      "loss": 2.7121,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44792592977955425,
+      "learning_rate": 6.59020861165136e-06,
+      "loss": 2.6372,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4331420401321665,
+      "learning_rate": 6.5852018925722815e-06,
+      "loss": 2.5859,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44548810875988937,
+      "learning_rate": 6.580193405785939e-06,
+      "loss": 2.5591,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4460985206843281,
+      "learning_rate": 6.575183156877446e-06,
+      "loss": 2.5174,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4597105895933579,
+      "learning_rate": 6.570171151433887e-06,
+      "loss": 2.5861,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4564653136768479,
+      "learning_rate": 6.5651573950443015e-06,
+      "loss": 2.6458,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4934156625854828,
+      "learning_rate": 6.5601418932996865e-06,
+      "loss": 2.634,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4940027941844852,
+      "learning_rate": 6.555124651792978e-06,
+      "loss": 2.5543,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47859111891025474,
+      "learning_rate": 6.550105676119057e-06,
+      "loss": 2.593,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45188261584964545,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 2.5211,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47426024103041675,
+      "learning_rate": 6.540062544658763e-06,
+      "loss": 2.572,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48733240633546737,
+      "learning_rate": 6.535038400071792e-06,
+      "loss": 2.575,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.522723835907445,
+      "learning_rate": 6.530012543716406e-06,
+      "loss": 2.4252,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4467541631302905,
+      "learning_rate": 6.524984981197089e-06,
+      "loss": 2.4924,
+      "step": 1204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4843494500835704,
+      "learning_rate": 6.519955718120231e-06,
+      "loss": 2.6027,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45544520295980023,
+      "learning_rate": 6.514924760094119e-06,
+      "loss": 2.5955,
+      "step": 1206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4352749030971699,
+      "learning_rate": 6.509892112728928e-06,
+      "loss": 2.4625,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4650983994465635,
+      "learning_rate": 6.504857781636717e-06,
+      "loss": 2.5992,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.44426386322003814,
+      "learning_rate": 6.499821772431421e-06,
+      "loss": 2.6038,
+      "step": 1209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44912655755103936,
+      "learning_rate": 6.494784090728852e-06,
+      "loss": 2.5598,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44476956179303945,
+      "learning_rate": 6.489744742146682e-06,
+      "loss": 2.5912,
+      "step": 1211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4724154528911073,
+      "learning_rate": 6.4847037323044435e-06,
+      "loss": 2.6931,
+      "step": 1212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46437449772269807,
+      "learning_rate": 6.47966106682352e-06,
+      "loss": 2.5635,
+      "step": 1213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4749528246592278,
+      "learning_rate": 6.474616751327143e-06,
+      "loss": 2.6371,
+      "step": 1214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4471321073249305,
+      "learning_rate": 6.469570791440385e-06,
+      "loss": 2.5681,
+      "step": 1215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46911193861592976,
+      "learning_rate": 6.464523192790149e-06,
+      "loss": 2.5962,
+      "step": 1216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4611190987744763,
+      "learning_rate": 6.459473961005168e-06,
+      "loss": 2.5888,
+      "step": 1217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4508996282471076,
+      "learning_rate": 6.454423101715994e-06,
+      "loss": 2.6209,
+      "step": 1218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4605768700329021,
+      "learning_rate": 6.4493706205549965e-06,
+      "loss": 2.5022,
+      "step": 1219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45844750355088554,
+      "learning_rate": 6.444316523156352e-06,
+      "loss": 2.5334,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4604012554345468,
+      "learning_rate": 6.439260815156039e-06,
+      "loss": 2.5601,
+      "step": 1221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4762823813736404,
+      "learning_rate": 6.434203502191832e-06,
+      "loss": 2.6728,
+      "step": 1222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.45712664008113646,
+      "learning_rate": 6.429144589903299e-06,
+      "loss": 2.5942,
+      "step": 1223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.47406668699524035,
+      "learning_rate": 6.4240840839317845e-06,
+      "loss": 2.6325,
+      "step": 1224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44825053155317557,
+      "learning_rate": 6.419021989920416e-06,
+      "loss": 2.5028,
+      "step": 1225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4775190914791269,
+      "learning_rate": 6.41395831351409e-06,
+      "loss": 2.6418,
+      "step": 1226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4921926921994469,
+      "learning_rate": 6.408893060359466e-06,
+      "loss": 2.6952,
+      "step": 1227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4780549615162635,
+      "learning_rate": 6.4038262361049655e-06,
+      "loss": 2.5216,
+      "step": 1228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4768563912136851,
+      "learning_rate": 6.398757846400757e-06,
+      "loss": 2.4883,
+      "step": 1229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.46607741771988215,
+      "learning_rate": 6.393687896898759e-06,
+      "loss": 2.5663,
+      "step": 1230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4746852547226443,
+      "learning_rate": 6.388616393252628e-06,
+      "loss": 2.675,
+      "step": 1231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.48522883193253036,
+      "learning_rate": 6.3835433411177515e-06,
+      "loss": 2.5872,
+      "step": 1232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4667006517138037,
+      "learning_rate": 6.378468746151247e-06,
+      "loss": 2.5696,
+      "step": 1233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4873283090007363,
+      "learning_rate": 6.373392614011952e-06,
+      "loss": 2.5399,
+      "step": 1234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.47787788088501654,
+      "learning_rate": 6.368314950360416e-06,
+      "loss": 2.6245,
+      "step": 1235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4765652515769694,
+      "learning_rate": 6.363235760858896e-06,
+      "loss": 2.5061,
+      "step": 1236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4800782826782663,
+      "learning_rate": 6.358155051171354e-06,
+      "loss": 2.5323,
+      "step": 1237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4552714710692263,
+      "learning_rate": 6.3530728269634454e-06,
+      "loss": 2.5443,
+      "step": 1238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46310046626499785,
+      "learning_rate": 6.347989093902514e-06,
+      "loss": 2.6454,
+      "step": 1239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44389307777978326,
+      "learning_rate": 6.342903857657585e-06,
+      "loss": 2.6258,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4930022625784188,
+      "learning_rate": 6.3378171238993636e-06,
+      "loss": 2.555,
+      "step": 1241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4613177137629376,
+      "learning_rate": 6.3327288983002215e-06,
+      "loss": 2.6009,
+      "step": 1242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4990186896350244,
+      "learning_rate": 6.327639186534195e-06,
+      "loss": 2.5925,
+      "step": 1243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4721129423818182,
+      "learning_rate": 6.322547994276978e-06,
+      "loss": 2.6107,
+      "step": 1244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4378291833055506,
+      "learning_rate": 6.317455327205916e-06,
+      "loss": 2.5039,
+      "step": 1245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5307764860202739,
+      "learning_rate": 6.312361190999997e-06,
+      "loss": 2.5824,
+      "step": 1246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4465787185489355,
+      "learning_rate": 6.307265591339851e-06,
+      "loss": 2.6157,
+      "step": 1247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44275859846600296,
+      "learning_rate": 6.302168533907735e-06,
+      "loss": 2.6128,
+      "step": 1248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5096919565573148,
+      "learning_rate": 6.297070024387535e-06,
+      "loss": 2.5575,
+      "step": 1249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46041151521427404,
+      "learning_rate": 6.291970068464755e-06,
+      "loss": 2.4884,
+      "step": 1250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45423030873995046,
+      "learning_rate": 6.286868671826513e-06,
+      "loss": 2.5099,
+      "step": 1251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45244274077454344,
+      "learning_rate": 6.281765840161532e-06,
+      "loss": 2.5465,
+      "step": 1252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4463492254861338,
+      "learning_rate": 6.276661579160137e-06,
+      "loss": 2.6482,
+      "step": 1253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45096646211265967,
+      "learning_rate": 6.271555894514245e-06,
+      "loss": 2.5977,
+      "step": 1254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.44575906414202116,
+      "learning_rate": 6.266448791917364e-06,
+      "loss": 2.6061,
+      "step": 1255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.606917389509902,
+      "learning_rate": 6.261340277064578e-06,
+      "loss": 2.6464,
+      "step": 1256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47613827005523224,
+      "learning_rate": 6.25623035565255e-06,
+      "loss": 2.5408,
+      "step": 1257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4532061701453479,
+      "learning_rate": 6.2511190333795115e-06,
+      "loss": 2.5762,
+      "step": 1258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46327531498778945,
+      "learning_rate": 6.246006315945251e-06,
+      "loss": 2.6817,
+      "step": 1259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4753389755188805,
+      "learning_rate": 6.24089220905112e-06,
+      "loss": 2.5167,
+      "step": 1260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45448269180456724,
+      "learning_rate": 6.235776718400015e-06,
+      "loss": 2.6865,
+      "step": 1261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5156535805381762,
+      "learning_rate": 6.230659849696375e-06,
+      "loss": 2.6058,
+      "step": 1262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4707312990992745,
+      "learning_rate": 6.2255416086461795e-06,
+      "loss": 2.5054,
+      "step": 1263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4975422116204126,
+      "learning_rate": 6.220422000956936e-06,
+      "loss": 2.6425,
+      "step": 1264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4860390106348795,
+      "learning_rate": 6.215301032337674e-06,
+      "loss": 2.6584,
+      "step": 1265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.47906318010056725,
+      "learning_rate": 6.210178708498945e-06,
+      "loss": 2.5822,
+      "step": 1266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4582887613413841,
+      "learning_rate": 6.2050550351528096e-06,
+      "loss": 2.5963,
+      "step": 1267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4872350241609007,
+      "learning_rate": 6.19993001801283e-06,
+      "loss": 2.5957,
+      "step": 1268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45391871555039404,
+      "learning_rate": 6.194803662794073e-06,
+      "loss": 2.5665,
+      "step": 1269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47199458318568166,
+      "learning_rate": 6.189675975213094e-06,
+      "loss": 2.4332,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4751959421676559,
+      "learning_rate": 6.184546960987936e-06,
+      "loss": 2.6663,
+      "step": 1271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5003458853262622,
+      "learning_rate": 6.179416625838116e-06,
+      "loss": 2.5985,
+      "step": 1272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.509987735226826,
+      "learning_rate": 6.174284975484632e-06,
+      "loss": 2.6042,
+      "step": 1273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5006315593789818,
+      "learning_rate": 6.169152015649944e-06,
+      "loss": 2.6218,
+      "step": 1274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.51435661366489,
+      "learning_rate": 6.164017752057972e-06,
+      "loss": 2.5504,
+      "step": 1275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4925983470181212,
+      "learning_rate": 6.158882190434092e-06,
+      "loss": 2.6199,
+      "step": 1276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4758692999734635,
+      "learning_rate": 6.153745336505125e-06,
+      "loss": 2.642,
+      "step": 1277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45794451180555695,
+      "learning_rate": 6.148607195999335e-06,
+      "loss": 2.5597,
+      "step": 1278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4991582102475872,
+      "learning_rate": 6.143467774646421e-06,
+      "loss": 2.589,
+      "step": 1279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4626964693067074,
+      "learning_rate": 6.13832707817751e-06,
+      "loss": 2.5227,
+      "step": 1280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4451044500117788,
+      "learning_rate": 6.1331851123251465e-06,
+      "loss": 2.5149,
+      "step": 1281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4473710914434683,
+      "learning_rate": 6.128041882823298e-06,
+      "loss": 2.5892,
+      "step": 1282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4675862170597165,
+      "learning_rate": 6.122897395407337e-06,
+      "loss": 2.5835,
+      "step": 1283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5179120408166407,
+      "learning_rate": 6.117751655814037e-06,
+      "loss": 2.5477,
+      "step": 1284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48641058594229025,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 2.5264,
+      "step": 1285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4798993093459797,
+      "learning_rate": 6.107456443049504e-06,
+      "loss": 2.5979,
+      "step": 1286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.45983074961829723,
+      "learning_rate": 6.10230698135878e-06,
+      "loss": 2.7149,
+      "step": 1287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4516532443780673,
+      "learning_rate": 6.097156290451721e-06,
+      "loss": 2.5407,
+      "step": 1288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4529765971287995,
+      "learning_rate": 6.092004376072021e-06,
+      "loss": 2.5561,
+      "step": 1289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4649889102249175,
+      "learning_rate": 6.0868512439647345e-06,
+      "loss": 2.5492,
+      "step": 1290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4721703987887475,
+      "learning_rate": 6.081696899876282e-06,
+      "loss": 2.3582,
+      "step": 1291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4714606272003474,
+      "learning_rate": 6.076541349554429e-06,
+      "loss": 2.5557,
+      "step": 1292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47315572719495963,
+      "learning_rate": 6.071384598748288e-06,
+      "loss": 2.5301,
+      "step": 1293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4743054990906478,
+      "learning_rate": 6.066226653208309e-06,
+      "loss": 2.54,
+      "step": 1294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.47572471896668495,
+      "learning_rate": 6.061067518686277e-06,
+      "loss": 2.5529,
+      "step": 1295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4942032544399013,
+      "learning_rate": 6.0559072009352995e-06,
+      "loss": 2.4468,
+      "step": 1296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49834056497619217,
+      "learning_rate": 6.0507457057098064e-06,
+      "loss": 2.4629,
+      "step": 1297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.46575690583675217,
+      "learning_rate": 6.045583038765538e-06,
+      "loss": 2.6405,
+      "step": 1298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.46617524002776023,
+      "learning_rate": 6.0404192058595435e-06,
+      "loss": 2.4729,
+      "step": 1299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44875109005638913,
+      "learning_rate": 6.035254212750172e-06,
+      "loss": 2.5976,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4865843928807325,
+      "learning_rate": 6.030088065197065e-06,
+      "loss": 2.4187,
+      "step": 1301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4611861641188956,
+      "learning_rate": 6.024920768961153e-06,
+      "loss": 2.7054,
+      "step": 1302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4868159318039705,
+      "learning_rate": 6.019752329804645e-06,
+      "loss": 2.5518,
+      "step": 1303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45295034528425004,
+      "learning_rate": 6.014582753491027e-06,
+      "loss": 2.5824,
+      "step": 1304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4982243539571222,
+      "learning_rate": 6.009412045785051e-06,
+      "loss": 2.5801,
+      "step": 1305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.47081479849103025,
+      "learning_rate": 6.004240212452735e-06,
+      "loss": 2.5029,
+      "step": 1306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4442053634445994,
+      "learning_rate": 5.999067259261347e-06,
+      "loss": 2.4948,
+      "step": 1307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4553893011822898,
+      "learning_rate": 5.993893191979403e-06,
+      "loss": 2.5039,
+      "step": 1308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4885722704522011,
+      "learning_rate": 5.98871801637667e-06,
+      "loss": 2.5194,
+      "step": 1309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45898969931253597,
+      "learning_rate": 5.983541738224141e-06,
+      "loss": 2.5069,
+      "step": 1310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43885936042518603,
+      "learning_rate": 5.978364363294044e-06,
+      "loss": 2.5667,
+      "step": 1311
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4696773574657853,
+      "learning_rate": 5.973185897359828e-06,
+      "loss": 2.5707,
+      "step": 1312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.46140705050441955,
+      "learning_rate": 5.968006346196158e-06,
+      "loss": 2.5789,
+      "step": 1313
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.46456435967376847,
+      "learning_rate": 5.962825715578914e-06,
+      "loss": 2.5773,
+      "step": 1314
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4994753230493636,
+      "learning_rate": 5.957644011285173e-06,
+      "loss": 2.531,
+      "step": 1315
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4488393093272709,
+      "learning_rate": 5.952461239093215e-06,
+      "loss": 2.5406,
+      "step": 1316
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5037959369889324,
+      "learning_rate": 5.947277404782505e-06,
+      "loss": 2.5275,
+      "step": 1317
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5038023849018948,
+      "learning_rate": 5.942092514133697e-06,
+      "loss": 2.5516,
+      "step": 1318
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4980262752087501,
+      "learning_rate": 5.936906572928625e-06,
+      "loss": 2.6257,
+      "step": 1319
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.47339554969076136,
+      "learning_rate": 5.931719586950286e-06,
+      "loss": 2.5523,
+      "step": 1320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5050212441593233,
+      "learning_rate": 5.92653156198285e-06,
+      "loss": 2.6781,
+      "step": 1321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.46148450366771243,
+      "learning_rate": 5.9213425038116415e-06,
+      "loss": 2.5813,
+      "step": 1322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4878858262155164,
+      "learning_rate": 5.91615241822314e-06,
+      "loss": 2.611,
+      "step": 1323
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.45313769355621,
+      "learning_rate": 5.910961311004968e-06,
+      "loss": 2.562,
+      "step": 1324
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4792309455327243,
+      "learning_rate": 5.905769187945889e-06,
+      "loss": 2.4339,
+      "step": 1325
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5114020321139143,
+      "learning_rate": 5.900576054835798e-06,
+      "loss": 2.6092,
+      "step": 1326
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4735465531601242,
+      "learning_rate": 5.895381917465718e-06,
+      "loss": 2.5793,
+      "step": 1327
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5487065027763396,
+      "learning_rate": 5.890186781627791e-06,
+      "loss": 2.5962,
+      "step": 1328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5024300176369647,
+      "learning_rate": 5.884990653115272e-06,
+      "loss": 2.5927,
+      "step": 1329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4864913176382235,
+      "learning_rate": 5.879793537722525e-06,
+      "loss": 2.6285,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4650853595013966,
+      "learning_rate": 5.874595441245008e-06,
+      "loss": 2.5393,
+      "step": 1331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4410323863734167,
+      "learning_rate": 5.8693963694792825e-06,
+      "loss": 2.5834,
+      "step": 1332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.47980881936799635,
+      "learning_rate": 5.864196328222991e-06,
+      "loss": 2.6105,
+      "step": 1333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.47461687874847147,
+      "learning_rate": 5.85899532327486e-06,
+      "loss": 2.4965,
+      "step": 1334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48759088158174807,
+      "learning_rate": 5.853793360434687e-06,
+      "loss": 2.5663,
+      "step": 1335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46464502270002156,
+      "learning_rate": 5.848590445503345e-06,
+      "loss": 2.4455,
+      "step": 1336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4564720384300828,
+      "learning_rate": 5.843386584282762e-06,
+      "loss": 2.5382,
+      "step": 1337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4571946984830581,
+      "learning_rate": 5.838181782575921e-06,
+      "loss": 2.5204,
+      "step": 1338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4640765229749111,
+      "learning_rate": 5.832976046186859e-06,
+      "loss": 2.5873,
+      "step": 1339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0440249803438744,
+      "learning_rate": 5.82776938092065e-06,
+      "loss": 2.5299,
+      "step": 1340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46834957185719356,
+      "learning_rate": 5.822561792583409e-06,
+      "loss": 2.5228,
+      "step": 1341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4903612010783117,
+      "learning_rate": 5.817353286982276e-06,
+      "loss": 2.5942,
+      "step": 1342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.512427094554068,
+      "learning_rate": 5.812143869925415e-06,
+      "loss": 2.6052,
+      "step": 1343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46534205422909963,
+      "learning_rate": 5.8069335472220066e-06,
+      "loss": 2.4895,
+      "step": 1344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4722720547247072,
+      "learning_rate": 5.801722324682243e-06,
+      "loss": 2.6285,
+      "step": 1345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46411280503858243,
+      "learning_rate": 5.796510208117317e-06,
+      "loss": 2.6635,
+      "step": 1346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.47830408831270155,
+      "learning_rate": 5.79129720333942e-06,
+      "loss": 2.4386,
+      "step": 1347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46511528354984577,
+      "learning_rate": 5.786083316161735e-06,
+      "loss": 2.5507,
+      "step": 1348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.46147315372523634,
+      "learning_rate": 5.780868552398424e-06,
+      "loss": 2.6121,
+      "step": 1349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45998273988145016,
+      "learning_rate": 5.775652917864633e-06,
+      "loss": 2.5347,
+      "step": 1350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4731335555024925,
+      "learning_rate": 5.770436418376477e-06,
+      "loss": 2.6014,
+      "step": 1351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.479270433164431,
+      "learning_rate": 5.765219059751033e-06,
+      "loss": 2.5795,
+      "step": 1352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4797835232182591,
+      "learning_rate": 5.760000847806337e-06,
+      "loss": 2.6403,
+      "step": 1353
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5111352636958504,
+      "learning_rate": 5.754781788361376e-06,
+      "loss": 2.5856,
+      "step": 1354
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4926179638330841,
+      "learning_rate": 5.749561887236088e-06,
+      "loss": 2.5253,
+      "step": 1355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5019845321863531,
+      "learning_rate": 5.744341150251342e-06,
+      "loss": 2.6454,
+      "step": 1356
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4989025320361983,
+      "learning_rate": 5.739119583228942e-06,
+      "loss": 2.6333,
+      "step": 1357
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4723994921199449,
+      "learning_rate": 5.733897191991616e-06,
+      "loss": 2.5501,
+      "step": 1358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47938048388193044,
+      "learning_rate": 5.728673982363017e-06,
+      "loss": 2.6093,
+      "step": 1359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48023812256085124,
+      "learning_rate": 5.723449960167703e-06,
+      "loss": 2.5757,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4621477241758881,
+      "learning_rate": 5.718225131231142e-06,
+      "loss": 2.574,
+      "step": 1361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4799222862394652,
+      "learning_rate": 5.712999501379701e-06,
+      "loss": 2.7045,
+      "step": 1362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47353058331138714,
+      "learning_rate": 5.70777307644064e-06,
+      "loss": 2.6185,
+      "step": 1363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47427909253482975,
+      "learning_rate": 5.702545862242107e-06,
+      "loss": 2.6429,
+      "step": 1364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.546562050785951,
+      "learning_rate": 5.697317864613127e-06,
+      "loss": 2.6603,
+      "step": 1365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.460793440338349,
+      "learning_rate": 5.692089089383603e-06,
+      "loss": 2.5108,
+      "step": 1366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4673929279538334,
+      "learning_rate": 5.686859542384301e-06,
+      "loss": 2.5553,
+      "step": 1367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.46737622177890714,
+      "learning_rate": 5.68162922944685e-06,
+      "loss": 2.6226,
+      "step": 1368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.47916144561687396,
+      "learning_rate": 5.676398156403733e-06,
+      "loss": 2.5185,
+      "step": 1369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.45965290486581706,
+      "learning_rate": 5.671166329088278e-06,
+      "loss": 2.6296,
+      "step": 1370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5029248579960017,
+      "learning_rate": 5.665933753334658e-06,
+      "loss": 2.6806,
+      "step": 1371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5146165001820266,
+      "learning_rate": 5.660700434977877e-06,
+      "loss": 2.5208,
+      "step": 1372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5018285752673409,
+      "learning_rate": 5.6554663798537715e-06,
+      "loss": 2.4773,
+      "step": 1373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.46568363511761657,
+      "learning_rate": 5.650231593798994e-06,
+      "loss": 2.5809,
+      "step": 1374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.462076719005681,
+      "learning_rate": 5.644996082651018e-06,
+      "loss": 2.5424,
+      "step": 1375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5042890233404459,
+      "learning_rate": 5.639759852248119e-06,
+      "loss": 2.6211,
+      "step": 1376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5198241247262678,
+      "learning_rate": 5.634522908429382e-06,
+      "loss": 2.579,
+      "step": 1377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4421819437169893,
+      "learning_rate": 5.629285257034682e-06,
+      "loss": 2.6423,
+      "step": 1378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.509712274909368,
+      "learning_rate": 5.624046903904684e-06,
+      "loss": 2.6235,
+      "step": 1379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4499040435103113,
+      "learning_rate": 5.6188078548808366e-06,
+      "loss": 2.5385,
+      "step": 1380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4845657588278896,
+      "learning_rate": 5.613568115805364e-06,
+      "loss": 2.59,
+      "step": 1381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4577609374918333,
+      "learning_rate": 5.608327692521263e-06,
+      "loss": 2.4767,
+      "step": 1382
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4596689030135519,
+      "learning_rate": 5.603086590872287e-06,
+      "loss": 2.451,
+      "step": 1383
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5070629456756863,
+      "learning_rate": 5.597844816702948e-06,
+      "loss": 2.6694,
+      "step": 1384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5091104791838418,
+      "learning_rate": 5.592602375858515e-06,
+      "loss": 2.6015,
+      "step": 1385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.48373983638643475,
+      "learning_rate": 5.587359274184989e-06,
+      "loss": 2.6154,
+      "step": 1386
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4664897466965815,
+      "learning_rate": 5.582115517529114e-06,
+      "loss": 2.56,
+      "step": 1387
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4886115063550699,
+      "learning_rate": 5.576871111738368e-06,
+      "loss": 2.5047,
+      "step": 1388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5054712168708521,
+      "learning_rate": 5.571626062660947e-06,
+      "loss": 2.5823,
+      "step": 1389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4811021769752877,
+      "learning_rate": 5.566380376145762e-06,
+      "loss": 2.6173,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4886745089940987,
+      "learning_rate": 5.5611340580424445e-06,
+      "loss": 2.541,
+      "step": 1391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5146651147677359,
+      "learning_rate": 5.555887114201324e-06,
+      "loss": 2.6584,
+      "step": 1392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.49482881172356813,
+      "learning_rate": 5.550639550473426e-06,
+      "loss": 2.6346,
+      "step": 1393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4750767984545818,
+      "learning_rate": 5.545391372710469e-06,
+      "loss": 2.6069,
+      "step": 1394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4901318689909373,
+      "learning_rate": 5.540142586764862e-06,
+      "loss": 2.517,
+      "step": 1395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4862073878663385,
+      "learning_rate": 5.5348931984896836e-06,
+      "loss": 2.5365,
+      "step": 1396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4954697897715803,
+      "learning_rate": 5.529643213738689e-06,
+      "loss": 2.5685,
+      "step": 1397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48910530743925806,
+      "learning_rate": 5.524392638366296e-06,
+      "loss": 2.5707,
+      "step": 1398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.46529677298505107,
+      "learning_rate": 5.519141478227582e-06,
+      "loss": 2.6055,
+      "step": 1399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4724731721275606,
+      "learning_rate": 5.51388973917828e-06,
+      "loss": 2.4231,
+      "step": 1400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48248346221332467,
+      "learning_rate": 5.508637427074762e-06,
+      "loss": 2.5628,
+      "step": 1401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4940948766707954,
+      "learning_rate": 5.503384547774043e-06,
+      "loss": 2.6035,
+      "step": 1402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4926925520221123,
+      "learning_rate": 5.4981311071337704e-06,
+      "loss": 2.5257,
+      "step": 1403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47903165435525097,
+      "learning_rate": 5.4928771110122185e-06,
+      "loss": 2.5344,
+      "step": 1404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.45944567285937526,
+      "learning_rate": 5.4876225652682776e-06,
+      "loss": 2.5259,
+      "step": 1405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4865373521405356,
+      "learning_rate": 5.482367475761452e-06,
+      "loss": 2.5353,
+      "step": 1406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4873319935458423,
+      "learning_rate": 5.477111848351856e-06,
+      "loss": 2.4752,
+      "step": 1407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48027422308809087,
+      "learning_rate": 5.471855688900198e-06,
+      "loss": 2.4997,
+      "step": 1408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4550806790390709,
+      "learning_rate": 5.466599003267787e-06,
+      "loss": 2.5089,
+      "step": 1409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47542219848601747,
+      "learning_rate": 5.46134179731651e-06,
+      "loss": 2.4786,
+      "step": 1410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48007773949504423,
+      "learning_rate": 5.456084076908842e-06,
+      "loss": 2.5718,
+      "step": 1411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4803538301316188,
+      "learning_rate": 5.450825847907824e-06,
+      "loss": 2.6203,
+      "step": 1412
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47772256956696457,
+      "learning_rate": 5.445567116177073e-06,
+      "loss": 2.4979,
+      "step": 1413
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.489455284119979,
+      "learning_rate": 5.44030788758076e-06,
+      "loss": 2.6273,
+      "step": 1414
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4568761578713499,
+      "learning_rate": 5.435048167983613e-06,
+      "loss": 2.6025,
+      "step": 1415
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47313892530460566,
+      "learning_rate": 5.429787963250905e-06,
+      "loss": 2.6403,
+      "step": 1416
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4793583514171144,
+      "learning_rate": 5.424527279248452e-06,
+      "loss": 2.5721,
+      "step": 1417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4890874996437913,
+      "learning_rate": 5.419266121842608e-06,
+      "loss": 2.6014,
+      "step": 1418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4872584502581997,
+      "learning_rate": 5.414004496900246e-06,
+      "loss": 2.5391,
+      "step": 1419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4875954656782858,
+      "learning_rate": 5.408742410288769e-06,
+      "loss": 2.5849,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4768472039957424,
+      "learning_rate": 5.403479867876087e-06,
+      "loss": 2.6094,
+      "step": 1421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5100692780040483,
+      "learning_rate": 5.398216875530626e-06,
+      "loss": 2.5434,
+      "step": 1422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.481302514864662,
+      "learning_rate": 5.392953439121311e-06,
+      "loss": 2.5323,
+      "step": 1423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48384092481853075,
+      "learning_rate": 5.387689564517558e-06,
+      "loss": 2.4878,
+      "step": 1424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47298535010674064,
+      "learning_rate": 5.382425257589277e-06,
+      "loss": 2.5716,
+      "step": 1425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5007152210333363,
+      "learning_rate": 5.377160524206859e-06,
+      "loss": 2.6258,
+      "step": 1426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.485070559786248,
+      "learning_rate": 5.371895370241168e-06,
+      "loss": 2.4971,
+      "step": 1427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4997944111157027,
+      "learning_rate": 5.366629801563541e-06,
+      "loss": 2.5767,
+      "step": 1428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4777561622573537,
+      "learning_rate": 5.361363824045772e-06,
+      "loss": 2.5328,
+      "step": 1429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.44479602845827954,
+      "learning_rate": 5.356097443560116e-06,
+      "loss": 2.4995,
+      "step": 1430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4797680405091772,
+      "learning_rate": 5.350830665979276e-06,
+      "loss": 2.5405,
+      "step": 1431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5033448810968487,
+      "learning_rate": 5.345563497176397e-06,
+      "loss": 2.566,
+      "step": 1432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5028626332441936,
+      "learning_rate": 5.340295943025059e-06,
+      "loss": 2.5561,
+      "step": 1433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.49284364470070213,
+      "learning_rate": 5.335028009399274e-06,
+      "loss": 2.4635,
+      "step": 1434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5017514250316445,
+      "learning_rate": 5.329759702173477e-06,
+      "loss": 2.6255,
+      "step": 1435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4739334253606109,
+      "learning_rate": 5.324491027222519e-06,
+      "loss": 2.595,
+      "step": 1436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4716542554992752,
+      "learning_rate": 5.319221990421662e-06,
+      "loss": 2.4898,
+      "step": 1437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47736644160252045,
+      "learning_rate": 5.3139525976465675e-06,
+      "loss": 2.5573,
+      "step": 1438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5015361256067403,
+      "learning_rate": 5.308682854773299e-06,
+      "loss": 2.4593,
+      "step": 1439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5172313752248289,
+      "learning_rate": 5.30341276767831e-06,
+      "loss": 2.4978,
+      "step": 1440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5025222461198174,
+      "learning_rate": 5.298142342238434e-06,
+      "loss": 2.5192,
+      "step": 1441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4934382818794458,
+      "learning_rate": 5.292871584330885e-06,
+      "loss": 2.5415,
+      "step": 1442
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.46466981214095376,
+      "learning_rate": 5.287600499833246e-06,
+      "loss": 2.6038,
+      "step": 1443
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4514102883849857,
+      "learning_rate": 5.282329094623466e-06,
+      "loss": 2.5755,
+      "step": 1444
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4651572668841609,
+      "learning_rate": 5.27705737457985e-06,
+      "loss": 2.6002,
+      "step": 1445
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.47035403781097285,
+      "learning_rate": 5.271785345581057e-06,
+      "loss": 2.6076,
+      "step": 1446
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.558092385820383,
+      "learning_rate": 5.266513013506085e-06,
+      "loss": 2.5004,
+      "step": 1447
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4995290356092678,
+      "learning_rate": 5.261240384234274e-06,
+      "loss": 2.4649,
+      "step": 1448
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5200915076362117,
+      "learning_rate": 5.255967463645296e-06,
+      "loss": 2.6154,
+      "step": 1449
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4691765751591343,
+      "learning_rate": 5.2506942576191466e-06,
+      "loss": 2.5766,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5212524698505389,
+      "learning_rate": 5.24542077203614e-06,
+      "loss": 2.6492,
+      "step": 1451
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4885974928051037,
+      "learning_rate": 5.240147012776899e-06,
+      "loss": 2.6256,
+      "step": 1452
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.46886963906982426,
+      "learning_rate": 5.234872985722356e-06,
+      "loss": 2.5161,
+      "step": 1453
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5131916078183119,
+      "learning_rate": 5.229598696753742e-06,
+      "loss": 2.5721,
+      "step": 1454
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48631933120069404,
+      "learning_rate": 5.224324151752575e-06,
+      "loss": 2.6087,
+      "step": 1455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5230734398885549,
+      "learning_rate": 5.219049356600665e-06,
+      "loss": 2.6447,
+      "step": 1456
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.46952691861660395,
+      "learning_rate": 5.213774317180096e-06,
+      "loss": 2.6004,
+      "step": 1457
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5122081076553424,
+      "learning_rate": 5.208499039373228e-06,
+      "loss": 2.5246,
+      "step": 1458
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48279313230710447,
+      "learning_rate": 5.203223529062684e-06,
+      "loss": 2.5698,
+      "step": 1459
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4798270181057229,
+      "learning_rate": 5.197947792131348e-06,
+      "loss": 2.5335,
+      "step": 1460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5080796932652959,
+      "learning_rate": 5.192671834462356e-06,
+      "loss": 2.4676,
+      "step": 1461
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4755061589874045,
+      "learning_rate": 5.187395661939088e-06,
+      "loss": 2.6214,
+      "step": 1462
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49052447971887797,
+      "learning_rate": 5.18211928044517e-06,
+      "loss": 2.6149,
+      "step": 1463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4854295080830313,
+      "learning_rate": 5.176842695864454e-06,
+      "loss": 2.5338,
+      "step": 1464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4953140529523987,
+      "learning_rate": 5.1715659140810225e-06,
+      "loss": 2.6496,
+      "step": 1465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5113368757065954,
+      "learning_rate": 5.166288940979175e-06,
+      "loss": 2.5616,
+      "step": 1466
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49523068242668017,
+      "learning_rate": 5.161011782443429e-06,
+      "loss": 2.6646,
+      "step": 1467
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5023566379347381,
+      "learning_rate": 5.155734444358503e-06,
+      "loss": 2.571,
+      "step": 1468
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49808651635476875,
+      "learning_rate": 5.150456932609321e-06,
+      "loss": 2.6516,
+      "step": 1469
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4959209621823317,
+      "learning_rate": 5.145179253080997e-06,
+      "loss": 2.6119,
+      "step": 1470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49478993331873267,
+      "learning_rate": 5.139901411658834e-06,
+      "loss": 2.5604,
+      "step": 1471
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49521748807814564,
+      "learning_rate": 5.134623414228315e-06,
+      "loss": 2.6045,
+      "step": 1472
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.505958046734499,
+      "learning_rate": 5.129345266675095e-06,
+      "loss": 2.5482,
+      "step": 1473
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5680064255760643,
+      "learning_rate": 5.124066974885002e-06,
+      "loss": 2.449,
+      "step": 1474
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.46019852125115945,
+      "learning_rate": 5.118788544744016e-06,
+      "loss": 2.6183,
+      "step": 1475
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5029274357124954,
+      "learning_rate": 5.113509982138281e-06,
+      "loss": 2.5581,
+      "step": 1476
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4825843197766323,
+      "learning_rate": 5.108231292954081e-06,
+      "loss": 2.5896,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4802353454848128,
+      "learning_rate": 5.102952483077846e-06,
+      "loss": 2.5429,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4951458533300915,
+      "learning_rate": 5.09767355839614e-06,
+      "loss": 2.5251,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5068067420973925,
+      "learning_rate": 5.09239452479565e-06,
+      "loss": 2.7136,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5389926975778253,
+      "learning_rate": 5.087115388163192e-06,
+      "loss": 2.5274,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4781809512293107,
+      "learning_rate": 5.081836154385692e-06,
+      "loss": 2.5492,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49952186134161425,
+      "learning_rate": 5.076556829350185e-06,
+      "loss": 2.5255,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5260028037895784,
+      "learning_rate": 5.071277418943807e-06,
+      "loss": 2.5031,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.490261031562141,
+      "learning_rate": 5.065997929053795e-06,
+      "loss": 2.6002,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.48382156811125154,
+      "learning_rate": 5.060718365567467e-06,
+      "loss": 2.536,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5075473973733333,
+      "learning_rate": 5.055438734372225e-06,
+      "loss": 2.6029,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5325905771186163,
+      "learning_rate": 5.050159041355551e-06,
+      "loss": 2.6337,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5030898644635945,
+      "learning_rate": 5.04487929240499e-06,
+      "loss": 2.5812,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49584254715744086,
+      "learning_rate": 5.039599493408154e-06,
+      "loss": 2.6302,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.508878791496978,
+      "learning_rate": 5.034319650252707e-06,
+      "loss": 2.4956,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4751330183145929,
+      "learning_rate": 5.029039768826366e-06,
+      "loss": 2.4993,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5074354679439569,
+      "learning_rate": 5.023759855016886e-06,
+      "loss": 2.5129,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5178318193661343,
+      "learning_rate": 5.018479914712065e-06,
+      "loss": 2.4461,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4931500423440274,
+      "learning_rate": 5.0131999537997235e-06,
+      "loss": 2.4936,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5045471846685965,
+      "learning_rate": 5.007919978167709e-06,
+      "loss": 2.64,
+      "step": 1496
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2985,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 374,
+  "total_flos": 2.2320632728765071e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}