diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6557 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 500,
+  "global_step": 932,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004308023694130318,
+      "grad_norm": 46.39875030517578,
+      "learning_rate": 9.9999954157983e-06,
+      "loss": 2.2558,
+      "step": 1
+    },
+    {
+      "epoch": 0.008616047388260635,
+      "grad_norm": 15.635394096374512,
+      "learning_rate": 9.999981663201606e-06,
+      "loss": 1.9294,
+      "step": 2
+    },
+    {
+      "epoch": 0.012924071082390954,
+      "grad_norm": 11.906808853149414,
+      "learning_rate": 9.999958742235133e-06,
+      "loss": 1.6914,
+      "step": 3
+    },
+    {
+      "epoch": 0.01723209477652127,
+      "grad_norm": 8.960531234741211,
+      "learning_rate": 9.999926652940914e-06,
+      "loss": 1.5586,
+      "step": 4
+    },
+    {
+      "epoch": 0.021540118470651588,
+      "grad_norm": 6.32525634765625,
+      "learning_rate": 9.999885395377788e-06,
+      "loss": 1.3009,
+      "step": 5
+    },
+    {
+      "epoch": 0.025848142164781908,
+      "grad_norm": 4.644639015197754,
+      "learning_rate": 9.999834969621408e-06,
+      "loss": 1.1522,
+      "step": 6
+    },
+    {
+      "epoch": 0.030156165858912225,
+      "grad_norm": 3.9122426509857178,
+      "learning_rate": 9.999775375764244e-06,
+      "loss": 1.061,
+      "step": 7
+    },
+    {
+      "epoch": 0.03446418955304254,
+      "grad_norm": 4.616422653198242,
+      "learning_rate": 9.999706613915567e-06,
+      "loss": 0.9941,
+      "step": 8
+    },
+    {
+      "epoch": 0.03877221324717286,
+      "grad_norm": 3.8503878116607666,
+      "learning_rate": 9.999628684201464e-06,
+      "loss": 0.8931,
+      "step": 9
+    },
+    {
+      "epoch": 0.043080236941303175,
+      "grad_norm": 3.5663602352142334,
+      "learning_rate": 9.999541586764836e-06,
+      "loss": 0.943,
+      "step": 10
+    },
+    {
+      "epoch": 0.04738826063543349,
+      "grad_norm": 3.536245584487915,
+      "learning_rate": 9.999445321765392e-06,
+      "loss": 0.8607,
+      "step": 11
+    },
+    {
+      "epoch": 0.051696284329563816,
+      "grad_norm": 3.5765976905822754,
+      "learning_rate": 9.999339889379647e-06,
+      "loss": 0.8233,
+      "step": 12
+    },
+    {
+      "epoch": 0.05600430802369413,
+      "grad_norm": 3.3080978393554688,
+      "learning_rate": 9.999225289800935e-06,
+      "loss": 0.7818,
+      "step": 13
+    },
+    {
+      "epoch": 0.06031233171782445,
+      "grad_norm": 3.949313163757324,
+      "learning_rate": 9.999101523239392e-06,
+      "loss": 0.8687,
+      "step": 14
+    },
+    {
+      "epoch": 0.06462035541195477,
+      "grad_norm": 3.9806888103485107,
+      "learning_rate": 9.998968589921969e-06,
+      "loss": 0.7793,
+      "step": 15
+    },
+    {
+      "epoch": 0.06892837910608508,
+      "grad_norm": 3.6018688678741455,
+      "learning_rate": 9.99882649009242e-06,
+      "loss": 0.7317,
+      "step": 16
+    },
+    {
+      "epoch": 0.0732364028002154,
+      "grad_norm": 3.872166872024536,
+      "learning_rate": 9.998675224011317e-06,
+      "loss": 0.7262,
+      "step": 17
+    },
+    {
+      "epoch": 0.07754442649434572,
+      "grad_norm": 5.003402233123779,
+      "learning_rate": 9.998514791956025e-06,
+      "loss": 0.7164,
+      "step": 18
+    },
+    {
+      "epoch": 0.08185245018847603,
+      "grad_norm": 3.848304033279419,
+      "learning_rate": 9.998345194220732e-06,
+      "loss": 0.6768,
+      "step": 19
+    },
+    {
+      "epoch": 0.08616047388260635,
+      "grad_norm": 3.976318359375,
+      "learning_rate": 9.998166431116421e-06,
+      "loss": 0.6572,
+      "step": 20
+    },
+    {
+      "epoch": 0.09046849757673667,
+      "grad_norm": 4.318434238433838,
+      "learning_rate": 9.99797850297089e-06,
+      "loss": 0.6814,
+      "step": 21
+    },
+    {
+      "epoch": 0.09477652127086698,
+      "grad_norm": 4.235775947570801,
+      "learning_rate": 9.997781410128737e-06,
+      "loss": 0.6579,
+      "step": 22
+    },
+    {
+      "epoch": 0.0990845449649973,
+      "grad_norm": 4.045382022857666,
+      "learning_rate": 9.99757515295137e-06,
+      "loss": 0.6532,
+      "step": 23
+    },
+    {
+      "epoch": 0.10339256865912763,
+      "grad_norm": 3.9923672676086426,
+      "learning_rate": 9.997359731816998e-06,
+      "loss": 0.6583,
+      "step": 24
+    },
+    {
+      "epoch": 0.10770059235325795,
+      "grad_norm": 4.043659687042236,
+      "learning_rate": 9.997135147120633e-06,
+      "loss": 0.6421,
+      "step": 25
+    },
+    {
+      "epoch": 0.11200861604738827,
+      "grad_norm": 4.166319370269775,
+      "learning_rate": 9.996901399274093e-06,
+      "loss": 0.6647,
+      "step": 26
+    },
+    {
+      "epoch": 0.11631663974151858,
+      "grad_norm": 4.195486545562744,
+      "learning_rate": 9.996658488705997e-06,
+      "loss": 0.6038,
+      "step": 27
+    },
+    {
+      "epoch": 0.1206246634356489,
+      "grad_norm": 3.893251657485962,
+      "learning_rate": 9.996406415861763e-06,
+      "loss": 0.5954,
+      "step": 28
+    },
+    {
+      "epoch": 0.12493268712977922,
+      "grad_norm": 3.857978582382202,
+      "learning_rate": 9.996145181203616e-06,
+      "loss": 0.5698,
+      "step": 29
+    },
+    {
+      "epoch": 0.12924071082390953,
+      "grad_norm": 3.830559492111206,
+      "learning_rate": 9.995874785210573e-06,
+      "loss": 0.5761,
+      "step": 30
+    },
+    {
+      "epoch": 0.13354873451803984,
+      "grad_norm": 3.8703525066375732,
+      "learning_rate": 9.995595228378456e-06,
+      "loss": 0.5811,
+      "step": 31
+    },
+    {
+      "epoch": 0.13785675821217017,
+      "grad_norm": 3.8460729122161865,
+      "learning_rate": 9.995306511219885e-06,
+      "loss": 0.5873,
+      "step": 32
+    },
+    {
+      "epoch": 0.1421647819063005,
+      "grad_norm": 3.9502267837524414,
+      "learning_rate": 9.995008634264272e-06,
+      "loss": 0.5795,
+      "step": 33
+    },
+    {
+      "epoch": 0.1464728056004308,
+      "grad_norm": 3.718348979949951,
+      "learning_rate": 9.994701598057828e-06,
+      "loss": 0.5406,
+      "step": 34
+    },
+    {
+      "epoch": 0.15078082929456113,
+      "grad_norm": 3.751901865005493,
+      "learning_rate": 9.99438540316356e-06,
+      "loss": 0.5619,
+      "step": 35
+    },
+    {
+      "epoch": 0.15508885298869143,
+      "grad_norm": 3.7187840938568115,
+      "learning_rate": 9.99406005016127e-06,
+      "loss": 0.5569,
+      "step": 36
+    },
+    {
+      "epoch": 0.15939687668282176,
+      "grad_norm": 3.647937536239624,
+      "learning_rate": 9.99372553964755e-06,
+      "loss": 0.5342,
+      "step": 37
+    },
+    {
+      "epoch": 0.16370490037695207,
+      "grad_norm": 3.6027166843414307,
+      "learning_rate": 9.993381872235785e-06,
+      "loss": 0.5279,
+      "step": 38
+    },
+    {
+      "epoch": 0.1680129240710824,
+      "grad_norm": 3.61478590965271,
+      "learning_rate": 9.993029048556154e-06,
+      "loss": 0.5131,
+      "step": 39
+    },
+    {
+      "epoch": 0.1723209477652127,
+      "grad_norm": 3.574810743331909,
+      "learning_rate": 9.99266706925562e-06,
+      "loss": 0.5165,
+      "step": 40
+    },
+    {
+      "epoch": 0.17662897145934303,
+      "grad_norm": 3.5639119148254395,
+      "learning_rate": 9.99229593499794e-06,
+      "loss": 0.5093,
+      "step": 41
+    },
+    {
+      "epoch": 0.18093699515347333,
+      "grad_norm": 3.4759559631347656,
+      "learning_rate": 9.991915646463652e-06,
+      "loss": 0.4864,
+      "step": 42
+    },
+    {
+      "epoch": 0.18524501884760367,
+      "grad_norm": 3.4125938415527344,
+      "learning_rate": 9.991526204350087e-06,
+      "loss": 0.4915,
+      "step": 43
+    },
+    {
+      "epoch": 0.18955304254173397,
+      "grad_norm": 3.4541664123535156,
+      "learning_rate": 9.991127609371357e-06,
+      "loss": 0.4891,
+      "step": 44
+    },
+    {
+      "epoch": 0.1938610662358643,
+      "grad_norm": 3.4495904445648193,
+      "learning_rate": 9.990719862258357e-06,
+      "loss": 0.4676,
+      "step": 45
+    },
+    {
+      "epoch": 0.1981690899299946,
+      "grad_norm": 3.2870962619781494,
+      "learning_rate": 9.990302963758765e-06,
+      "loss": 0.4679,
+      "step": 46
+    },
+    {
+      "epoch": 0.20247711362412493,
+      "grad_norm": 3.3008880615234375,
+      "learning_rate": 9.989876914637042e-06,
+      "loss": 0.4688,
+      "step": 47
+    },
+    {
+      "epoch": 0.20678513731825526,
+      "grad_norm": 3.086277723312378,
+      "learning_rate": 9.989441715674422e-06,
+      "loss": 0.4565,
+      "step": 48
+    },
+    {
+      "epoch": 0.21109316101238557,
+      "grad_norm": 3.048494338989258,
+      "learning_rate": 9.988997367668924e-06,
+      "loss": 0.4358,
+      "step": 49
+    },
+    {
+      "epoch": 0.2154011847065159,
+      "grad_norm": 2.9959912300109863,
+      "learning_rate": 9.988543871435342e-06,
+      "loss": 0.439,
+      "step": 50
+    },
+    {
+      "epoch": 0.2197092084006462,
+      "grad_norm": 2.891278028488159,
+      "learning_rate": 9.988081227805237e-06,
+      "loss": 0.4534,
+      "step": 51
+    },
+    {
+      "epoch": 0.22401723209477653,
+      "grad_norm": 3.0530009269714355,
+      "learning_rate": 9.987609437626955e-06,
+      "loss": 0.42,
+      "step": 52
+    },
+    {
+      "epoch": 0.22832525578890683,
+      "grad_norm": 2.8397328853607178,
+      "learning_rate": 9.987128501765606e-06,
+      "loss": 0.4071,
+      "step": 53
+    },
+    {
+      "epoch": 0.23263327948303716,
+      "grad_norm": 2.5578064918518066,
+      "learning_rate": 9.986638421103074e-06,
+      "loss": 0.3947,
+      "step": 54
+    },
+    {
+      "epoch": 0.23694130317716747,
+      "grad_norm": 2.627624750137329,
+      "learning_rate": 9.986139196538011e-06,
+      "loss": 0.4039,
+      "step": 55
+    },
+    {
+      "epoch": 0.2412493268712978,
+      "grad_norm": 2.515475273132324,
+      "learning_rate": 9.985630828985835e-06,
+      "loss": 0.4188,
+      "step": 56
+    },
+    {
+      "epoch": 0.2455573505654281,
+      "grad_norm": 2.286128282546997,
+      "learning_rate": 9.98511331937873e-06,
+      "loss": 0.3922,
+      "step": 57
+    },
+    {
+      "epoch": 0.24986537425955843,
+      "grad_norm": 2.2815139293670654,
+      "learning_rate": 9.984586668665641e-06,
+      "loss": 0.4452,
+      "step": 58
+    },
+    {
+      "epoch": 0.25417339795368876,
+      "grad_norm": 2.0464518070220947,
+      "learning_rate": 9.98405087781228e-06,
+      "loss": 0.391,
+      "step": 59
+    },
+    {
+      "epoch": 0.25848142164781907,
+      "grad_norm": 2.0818228721618652,
+      "learning_rate": 9.983505947801115e-06,
+      "loss": 0.4185,
+      "step": 60
+    },
+    {
+      "epoch": 0.26278944534194937,
+      "grad_norm": 1.843110203742981,
+      "learning_rate": 9.982951879631373e-06,
+      "loss": 0.3706,
+      "step": 61
+    },
+    {
+      "epoch": 0.26709746903607967,
+      "grad_norm": 1.8725769519805908,
+      "learning_rate": 9.982388674319041e-06,
+      "loss": 0.3723,
+      "step": 62
+    },
+    {
+      "epoch": 0.27140549273021003,
+      "grad_norm": 1.7885315418243408,
+      "learning_rate": 9.981816332896854e-06,
+      "loss": 0.3903,
+      "step": 63
+    },
+    {
+      "epoch": 0.27571351642434033,
+      "grad_norm": 1.5986205339431763,
+      "learning_rate": 9.981234856414306e-06,
+      "loss": 0.3759,
+      "step": 64
+    },
+    {
+      "epoch": 0.28002154011847064,
+      "grad_norm": 1.5188963413238525,
+      "learning_rate": 9.98064424593764e-06,
+      "loss": 0.3641,
+      "step": 65
+    },
+    {
+      "epoch": 0.284329563812601,
+      "grad_norm": 1.4895678758621216,
+      "learning_rate": 9.980044502549843e-06,
+      "loss": 0.3722,
+      "step": 66
+    },
+    {
+      "epoch": 0.2886375875067313,
+      "grad_norm": 1.517785668373108,
+      "learning_rate": 9.979435627350658e-06,
+      "loss": 0.4034,
+      "step": 67
+    },
+    {
+      "epoch": 0.2929456112008616,
+      "grad_norm": 1.5465859174728394,
+      "learning_rate": 9.978817621456562e-06,
+      "loss": 0.404,
+      "step": 68
+    },
+    {
+      "epoch": 0.2972536348949919,
+      "grad_norm": 1.4948019981384277,
+      "learning_rate": 9.978190486000784e-06,
+      "loss": 0.409,
+      "step": 69
+    },
+    {
+      "epoch": 0.30156165858912226,
+      "grad_norm": 1.349805474281311,
+      "learning_rate": 9.977554222133293e-06,
+      "loss": 0.3809,
+      "step": 70
+    },
+    {
+      "epoch": 0.30586968228325256,
+      "grad_norm": 1.4803236722946167,
+      "learning_rate": 9.976908831020787e-06,
+      "loss": 0.3709,
+      "step": 71
+    },
+    {
+      "epoch": 0.31017770597738287,
+      "grad_norm": 1.500108003616333,
+      "learning_rate": 9.97625431384671e-06,
+      "loss": 0.3875,
+      "step": 72
+    },
+    {
+      "epoch": 0.31448572967151317,
+      "grad_norm": 1.431912899017334,
+      "learning_rate": 9.975590671811239e-06,
+      "loss": 0.3897,
+      "step": 73
+    },
+    {
+      "epoch": 0.31879375336564353,
+      "grad_norm": 1.3888511657714844,
+      "learning_rate": 9.974917906131283e-06,
+      "loss": 0.3558,
+      "step": 74
+    },
+    {
+      "epoch": 0.32310177705977383,
+      "grad_norm": 1.3320142030715942,
+      "learning_rate": 9.974236018040476e-06,
+      "loss": 0.3619,
+      "step": 75
+    },
+    {
+      "epoch": 0.32740980075390413,
+      "grad_norm": 1.3935425281524658,
+      "learning_rate": 9.973545008789182e-06,
+      "loss": 0.3759,
+      "step": 76
+    },
+    {
+      "epoch": 0.33171782444803444,
+      "grad_norm": 1.359113335609436,
+      "learning_rate": 9.972844879644494e-06,
+      "loss": 0.3532,
+      "step": 77
+    },
+    {
+      "epoch": 0.3360258481421648,
+      "grad_norm": 1.4055691957473755,
+      "learning_rate": 9.972135631890226e-06,
+      "loss": 0.366,
+      "step": 78
+    },
+    {
+      "epoch": 0.3403338718362951,
+      "grad_norm": 1.51666259765625,
+      "learning_rate": 9.97141726682691e-06,
+      "loss": 0.3551,
+      "step": 79
+    },
+    {
+      "epoch": 0.3446418955304254,
+      "grad_norm": 1.3641939163208008,
+      "learning_rate": 9.970689785771798e-06,
+      "loss": 0.3552,
+      "step": 80
+    },
+    {
+      "epoch": 0.34894991922455576,
+      "grad_norm": 1.476531744003296,
+      "learning_rate": 9.969953190058861e-06,
+      "loss": 0.3918,
+      "step": 81
+    },
+    {
+      "epoch": 0.35325794291868606,
+      "grad_norm": 1.4188109636306763,
+      "learning_rate": 9.969207481038776e-06,
+      "loss": 0.3829,
+      "step": 82
+    },
+    {
+      "epoch": 0.35756596661281637,
+      "grad_norm": 1.3417792320251465,
+      "learning_rate": 9.968452660078939e-06,
+      "loss": 0.355,
+      "step": 83
+    },
+    {
+      "epoch": 0.36187399030694667,
+      "grad_norm": 1.4782617092132568,
+      "learning_rate": 9.967688728563446e-06,
+      "loss": 0.3721,
+      "step": 84
+    },
+    {
+      "epoch": 0.366182014001077,
+      "grad_norm": 1.6000183820724487,
+      "learning_rate": 9.966915687893109e-06,
+      "loss": 0.3935,
+      "step": 85
+    },
+    {
+      "epoch": 0.37049003769520733,
+      "grad_norm": 1.4256422519683838,
+      "learning_rate": 9.966133539485435e-06,
+      "loss": 0.3697,
+      "step": 86
+    },
+    {
+      "epoch": 0.37479806138933763,
+      "grad_norm": 1.3764766454696655,
+      "learning_rate": 9.965342284774633e-06,
+      "loss": 0.3521,
+      "step": 87
+    },
+    {
+      "epoch": 0.37910608508346794,
+      "grad_norm": 1.4846636056900024,
+      "learning_rate": 9.964541925211613e-06,
+      "loss": 0.3394,
+      "step": 88
+    },
+    {
+      "epoch": 0.3834141087775983,
+      "grad_norm": 1.4550178050994873,
+      "learning_rate": 9.963732462263979e-06,
+      "loss": 0.3502,
+      "step": 89
+    },
+    {
+      "epoch": 0.3877221324717286,
+      "grad_norm": 1.374940037727356,
+      "learning_rate": 9.962913897416029e-06,
+      "loss": 0.3333,
+      "step": 90
+    },
+    {
+      "epoch": 0.3920301561658589,
+      "grad_norm": 1.384493350982666,
+      "learning_rate": 9.962086232168747e-06,
+      "loss": 0.3291,
+      "step": 91
+    },
+    {
+      "epoch": 0.3963381798599892,
+      "grad_norm": 1.4366718530654907,
+      "learning_rate": 9.961249468039806e-06,
+      "loss": 0.3322,
+      "step": 92
+    },
+    {
+      "epoch": 0.40064620355411956,
+      "grad_norm": 1.5821336507797241,
+      "learning_rate": 9.960403606563568e-06,
+      "loss": 0.346,
+      "step": 93
+    },
+    {
+      "epoch": 0.40495422724824987,
+      "grad_norm": 1.5977967977523804,
+      "learning_rate": 9.959548649291071e-06,
+      "loss": 0.372,
+      "step": 94
+    },
+    {
+      "epoch": 0.40926225094238017,
+      "grad_norm": 1.7642663717269897,
+      "learning_rate": 9.958684597790031e-06,
+      "loss": 0.3854,
+      "step": 95
+    },
+    {
+      "epoch": 0.4135702746365105,
+      "grad_norm": 1.5853767395019531,
+      "learning_rate": 9.957811453644848e-06,
+      "loss": 0.354,
+      "step": 96
+    },
+    {
+      "epoch": 0.41787829833064083,
+      "grad_norm": 1.445274829864502,
+      "learning_rate": 9.956929218456586e-06,
+      "loss": 0.3402,
+      "step": 97
+    },
+    {
+      "epoch": 0.42218632202477113,
+      "grad_norm": 1.5509393215179443,
+      "learning_rate": 9.956037893842982e-06,
+      "loss": 0.3361,
+      "step": 98
+    },
+    {
+      "epoch": 0.42649434571890144,
+      "grad_norm": 1.512851595878601,
+      "learning_rate": 9.955137481438442e-06,
+      "loss": 0.3451,
+      "step": 99
+    },
+    {
+      "epoch": 0.4308023694130318,
+      "grad_norm": 1.4434539079666138,
+      "learning_rate": 9.954227982894034e-06,
+      "loss": 0.3113,
+      "step": 100
+    },
+    {
+      "epoch": 0.4351103931071621,
+      "grad_norm": 1.4094306230545044,
+      "learning_rate": 9.953309399877491e-06,
+      "loss": 0.3417,
+      "step": 101
+    },
+    {
+      "epoch": 0.4394184168012924,
+      "grad_norm": 1.3972957134246826,
+      "learning_rate": 9.952381734073197e-06,
+      "loss": 0.3258,
+      "step": 102
+    },
+    {
+      "epoch": 0.4437264404954227,
+      "grad_norm": 1.4385653734207153,
+      "learning_rate": 9.951444987182195e-06,
+      "loss": 0.3299,
+      "step": 103
+    },
+    {
+      "epoch": 0.44803446418955306,
+      "grad_norm": 1.4804537296295166,
+      "learning_rate": 9.950499160922184e-06,
+      "loss": 0.3011,
+      "step": 104
+    },
+    {
+      "epoch": 0.45234248788368336,
+      "grad_norm": 1.501686453819275,
+      "learning_rate": 9.949544257027503e-06,
+      "loss": 0.3309,
+      "step": 105
+    },
+    {
+      "epoch": 0.45665051157781367,
+      "grad_norm": 1.44529128074646,
+      "learning_rate": 9.948580277249142e-06,
+      "loss": 0.3138,
+      "step": 106
+    },
+    {
+      "epoch": 0.46095853527194397,
+      "grad_norm": 1.5322788953781128,
+      "learning_rate": 9.947607223354731e-06,
+      "loss": 0.3349,
+      "step": 107
+    },
+    {
+      "epoch": 0.46526655896607433,
+      "grad_norm": 1.4645901918411255,
+      "learning_rate": 9.946625097128544e-06,
+      "loss": 0.3242,
+      "step": 108
+    },
+    {
+      "epoch": 0.46957458266020463,
+      "grad_norm": 1.4635151624679565,
+      "learning_rate": 9.945633900371483e-06,
+      "loss": 0.3153,
+      "step": 109
+    },
+    {
+      "epoch": 0.47388260635433493,
+      "grad_norm": 1.5546009540557861,
+      "learning_rate": 9.94463363490109e-06,
+      "loss": 0.3121,
+      "step": 110
+    },
+    {
+      "epoch": 0.4781906300484653,
+      "grad_norm": 1.4418449401855469,
+      "learning_rate": 9.943624302551527e-06,
+      "loss": 0.2974,
+      "step": 111
+    },
+    {
+      "epoch": 0.4824986537425956,
+      "grad_norm": 1.519238829612732,
+      "learning_rate": 9.942605905173593e-06,
+      "loss": 0.3301,
+      "step": 112
+    },
+    {
+      "epoch": 0.4868066774367259,
+      "grad_norm": 1.6555852890014648,
+      "learning_rate": 9.941578444634699e-06,
+      "loss": 0.2964,
+      "step": 113
+    },
+    {
+      "epoch": 0.4911147011308562,
+      "grad_norm": 1.690470814704895,
+      "learning_rate": 9.940541922818882e-06,
+      "loss": 0.3006,
+      "step": 114
+    },
+    {
+      "epoch": 0.49542272482498656,
+      "grad_norm": 1.4557969570159912,
+      "learning_rate": 9.939496341626791e-06,
+      "loss": 0.3061,
+      "step": 115
+    },
+    {
+      "epoch": 0.49973074851911686,
+      "grad_norm": 1.5956724882125854,
+      "learning_rate": 9.938441702975689e-06,
+      "loss": 0.3177,
+      "step": 116
+    },
+    {
+      "epoch": 0.5040387722132472,
+      "grad_norm": 1.742788314819336,
+      "learning_rate": 9.937378008799448e-06,
+      "loss": 0.2813,
+      "step": 117
+    },
+    {
+      "epoch": 0.5083467959073775,
+      "grad_norm": 1.4342483282089233,
+      "learning_rate": 9.93630526104854e-06,
+      "loss": 0.284,
+      "step": 118
+    },
+    {
+      "epoch": 0.5126548196015078,
+      "grad_norm": 1.427372932434082,
+      "learning_rate": 9.935223461690042e-06,
+      "loss": 0.3,
+      "step": 119
+    },
+    {
+      "epoch": 0.5169628432956381,
+      "grad_norm": 1.4646867513656616,
+      "learning_rate": 9.934132612707631e-06,
+      "loss": 0.2841,
+      "step": 120
+    },
+    {
+      "epoch": 0.5212708669897684,
+      "grad_norm": 1.6393767595291138,
+      "learning_rate": 9.933032716101576e-06,
+      "loss": 0.3194,
+      "step": 121
+    },
+    {
+      "epoch": 0.5255788906838987,
+      "grad_norm": 1.4437857866287231,
+      "learning_rate": 9.931923773888734e-06,
+      "loss": 0.3089,
+      "step": 122
+    },
+    {
+      "epoch": 0.529886914378029,
+      "grad_norm": 1.4526920318603516,
+      "learning_rate": 9.930805788102551e-06,
+      "loss": 0.3044,
+      "step": 123
+    },
+    {
+      "epoch": 0.5341949380721593,
+      "grad_norm": 1.5132489204406738,
+      "learning_rate": 9.929678760793057e-06,
+      "loss": 0.265,
+      "step": 124
+    },
+    {
+      "epoch": 0.5385029617662898,
+      "grad_norm": 1.6274410486221313,
+      "learning_rate": 9.928542694026862e-06,
+      "loss": 0.3252,
+      "step": 125
+    },
+    {
+      "epoch": 0.5428109854604201,
+      "grad_norm": 1.4639244079589844,
+      "learning_rate": 9.927397589887144e-06,
+      "loss": 0.288,
+      "step": 126
+    },
+    {
+      "epoch": 0.5471190091545504,
+      "grad_norm": 1.4215894937515259,
+      "learning_rate": 9.926243450473664e-06,
+      "loss": 0.2738,
+      "step": 127
+    },
+    {
+      "epoch": 0.5514270328486807,
+      "grad_norm": 1.3172849416732788,
+      "learning_rate": 9.925080277902743e-06,
+      "loss": 0.2847,
+      "step": 128
+    },
+    {
+      "epoch": 0.555735056542811,
+      "grad_norm": 1.414466381072998,
+      "learning_rate": 9.923908074307267e-06,
+      "loss": 0.3024,
+      "step": 129
+    },
+    {
+      "epoch": 0.5600430802369413,
+      "grad_norm": 1.322507619857788,
+      "learning_rate": 9.922726841836685e-06,
+      "loss": 0.2826,
+      "step": 130
+    },
+    {
+      "epoch": 0.5643511039310716,
+      "grad_norm": 1.3307005167007446,
+      "learning_rate": 9.921536582657002e-06,
+      "loss": 0.2865,
+      "step": 131
+    },
+    {
+      "epoch": 0.568659127625202,
+      "grad_norm": 1.303130865097046,
+      "learning_rate": 9.920337298950767e-06,
+      "loss": 0.275,
+      "step": 132
+    },
+    {
+      "epoch": 0.5729671513193323,
+      "grad_norm": 1.3872160911560059,
+      "learning_rate": 9.919128992917086e-06,
+      "loss": 0.3168,
+      "step": 133
+    },
+    {
+      "epoch": 0.5772751750134626,
+      "grad_norm": 1.3143068552017212,
+      "learning_rate": 9.91791166677161e-06,
+      "loss": 0.2831,
+      "step": 134
+    },
+    {
+      "epoch": 0.5815831987075929,
+      "grad_norm": 1.4942936897277832,
+      "learning_rate": 9.916685322746524e-06,
+      "loss": 0.339,
+      "step": 135
+    },
+    {
+      "epoch": 0.5858912224017232,
+      "grad_norm": 1.243410587310791,
+      "learning_rate": 9.915449963090551e-06,
+      "loss": 0.2798,
+      "step": 136
+    },
+    {
+      "epoch": 0.5901992460958535,
+      "grad_norm": 1.2295470237731934,
+      "learning_rate": 9.914205590068946e-06,
+      "loss": 0.275,
+      "step": 137
+    },
+    {
+      "epoch": 0.5945072697899838,
+      "grad_norm": 1.251297950744629,
+      "learning_rate": 9.912952205963491e-06,
+      "loss": 0.2714,
+      "step": 138
+    },
+    {
+      "epoch": 0.5988152934841141,
+      "grad_norm": 1.2663097381591797,
+      "learning_rate": 9.911689813072495e-06,
+      "loss": 0.2733,
+      "step": 139
+    },
+    {
+      "epoch": 0.6031233171782445,
+      "grad_norm": 1.2405517101287842,
+      "learning_rate": 9.91041841371078e-06,
+      "loss": 0.2659,
+      "step": 140
+    },
+    {
+      "epoch": 0.6074313408723748,
+      "grad_norm": 1.178120732307434,
+      "learning_rate": 9.90913801020969e-06,
+      "loss": 0.264,
+      "step": 141
+    },
+    {
+      "epoch": 0.6117393645665051,
+      "grad_norm": 1.292300820350647,
+      "learning_rate": 9.907848604917075e-06,
+      "loss": 0.2857,
+      "step": 142
+    },
+    {
+      "epoch": 0.6160473882606354,
+      "grad_norm": 1.1422131061553955,
+      "learning_rate": 9.906550200197288e-06,
+      "loss": 0.2547,
+      "step": 143
+    },
+    {
+      "epoch": 0.6203554119547657,
+      "grad_norm": 1.1891106367111206,
+      "learning_rate": 9.905242798431196e-06,
+      "loss": 0.2644,
+      "step": 144
+    },
+    {
+      "epoch": 0.624663435648896,
+      "grad_norm": 1.1191891431808472,
+      "learning_rate": 9.903926402016153e-06,
+      "loss": 0.2602,
+      "step": 145
+    },
+    {
+      "epoch": 0.6289714593430263,
+      "grad_norm": 1.260938286781311,
+      "learning_rate": 9.902601013366009e-06,
+      "loss": 0.2954,
+      "step": 146
+    },
+    {
+      "epoch": 0.6332794830371568,
+      "grad_norm": 1.033594012260437,
+      "learning_rate": 9.901266634911104e-06,
+      "loss": 0.2624,
+      "step": 147
+    },
+    {
+      "epoch": 0.6375875067312871,
+      "grad_norm": 1.0682276487350464,
+      "learning_rate": 9.899923269098262e-06,
+      "loss": 0.2524,
+      "step": 148
+    },
+    {
+      "epoch": 0.6418955304254174,
+      "grad_norm": 1.1801897287368774,
+      "learning_rate": 9.898570918390789e-06,
+      "loss": 0.2599,
+      "step": 149
+    },
+    {
+      "epoch": 0.6462035541195477,
+      "grad_norm": 0.9911927580833435,
+      "learning_rate": 9.897209585268459e-06,
+      "loss": 0.2388,
+      "step": 150
+    },
+    {
+      "epoch": 0.650511577813678,
+      "grad_norm": 1.055557131767273,
+      "learning_rate": 9.895839272227529e-06,
+      "loss": 0.2476,
+      "step": 151
+    },
+    {
+      "epoch": 0.6548196015078083,
+      "grad_norm": 1.1163235902786255,
+      "learning_rate": 9.894459981780711e-06,
+      "loss": 0.2506,
+      "step": 152
+    },
+    {
+      "epoch": 0.6591276252019386,
+      "grad_norm": 1.1236166954040527,
+      "learning_rate": 9.893071716457183e-06,
+      "loss": 0.2765,
+      "step": 153
+    },
+    {
+      "epoch": 0.6634356488960689,
+      "grad_norm": 0.9692741632461548,
+      "learning_rate": 9.891674478802585e-06,
+      "loss": 0.237,
+      "step": 154
+    },
+    {
+      "epoch": 0.6677436725901993,
+      "grad_norm": 1.79551362991333,
+      "learning_rate": 9.890268271379e-06,
+      "loss": 0.2653,
+      "step": 155
+    },
+    {
+      "epoch": 0.6720516962843296,
+      "grad_norm": 1.0855523347854614,
+      "learning_rate": 9.888853096764963e-06,
+      "loss": 0.2556,
+      "step": 156
+    },
+    {
+      "epoch": 0.6763597199784599,
+      "grad_norm": 1.1746084690093994,
+      "learning_rate": 9.887428957555457e-06,
+      "loss": 0.2717,
+      "step": 157
+    },
+    {
+      "epoch": 0.6806677436725902,
+      "grad_norm": 1.03372061252594,
+      "learning_rate": 9.885995856361895e-06,
+      "loss": 0.2558,
+      "step": 158
+    },
+    {
+      "epoch": 0.6849757673667205,
+      "grad_norm": 0.9783212542533875,
+      "learning_rate": 9.884553795812128e-06,
+      "loss": 0.2484,
+      "step": 159
+    },
+    {
+      "epoch": 0.6892837910608508,
+      "grad_norm": 0.8822944760322571,
+      "learning_rate": 9.883102778550434e-06,
+      "loss": 0.2471,
+      "step": 160
+    },
+    {
+      "epoch": 0.6935918147549811,
+      "grad_norm": 1.0045849084854126,
+      "learning_rate": 9.881642807237515e-06,
+      "loss": 0.2597,
+      "step": 161
+    },
+    {
+      "epoch": 0.6978998384491115,
+      "grad_norm": 0.912653923034668,
+      "learning_rate": 9.880173884550495e-06,
+      "loss": 0.2428,
+      "step": 162
+    },
+    {
+      "epoch": 0.7022078621432418,
+      "grad_norm": 1.0020110607147217,
+      "learning_rate": 9.878696013182906e-06,
+      "loss": 0.2714,
+      "step": 163
+    },
+    {
+      "epoch": 0.7065158858373721,
+      "grad_norm": 0.8890970945358276,
+      "learning_rate": 9.877209195844692e-06,
+      "loss": 0.2346,
+      "step": 164
+    },
+    {
+      "epoch": 0.7108239095315024,
+      "grad_norm": 1.1626794338226318,
+      "learning_rate": 9.875713435262205e-06,
+      "loss": 0.2839,
+      "step": 165
+    },
+    {
+      "epoch": 0.7151319332256327,
+      "grad_norm": 0.8313829898834229,
+      "learning_rate": 9.874208734178187e-06,
+      "loss": 0.2376,
+      "step": 166
+    },
+    {
+      "epoch": 0.719439956919763,
+      "grad_norm": 0.9518410563468933,
+      "learning_rate": 9.872695095351784e-06,
+      "loss": 0.2536,
+      "step": 167
+    },
+    {
+      "epoch": 0.7237479806138933,
+      "grad_norm": 0.9485108852386475,
+      "learning_rate": 9.871172521558523e-06,
+      "loss": 0.2592,
+      "step": 168
+    },
+    {
+      "epoch": 0.7280560043080236,
+      "grad_norm": 0.9191961884498596,
+      "learning_rate": 9.869641015590319e-06,
+      "loss": 0.2466,
+      "step": 169
+    },
+    {
+      "epoch": 0.732364028002154,
+      "grad_norm": 0.9200321435928345,
+      "learning_rate": 9.868100580255466e-06,
+      "loss": 0.2453,
+      "step": 170
+    },
+    {
+      "epoch": 0.7366720516962844,
+      "grad_norm": 0.9513406753540039,
+      "learning_rate": 9.86655121837863e-06,
+      "loss": 0.2628,
+      "step": 171
+    },
+    {
+      "epoch": 0.7409800753904147,
+      "grad_norm": 0.8816955089569092,
+      "learning_rate": 9.864992932800845e-06,
+      "loss": 0.2514,
+      "step": 172
+    },
+    {
+      "epoch": 0.745288099084545,
+      "grad_norm": 0.9091234803199768,
+      "learning_rate": 9.863425726379512e-06,
+      "loss": 0.2377,
+      "step": 173
+    },
+    {
+      "epoch": 0.7495961227786753,
+      "grad_norm": 1.138776183128357,
+      "learning_rate": 9.861849601988384e-06,
+      "loss": 0.2772,
+      "step": 174
+    },
+    {
+      "epoch": 0.7539041464728056,
+      "grad_norm": 0.9423293471336365,
+      "learning_rate": 9.860264562517571e-06,
+      "loss": 0.2369,
+      "step": 175
+    },
+    {
+      "epoch": 0.7582121701669359,
+      "grad_norm": 0.8946828842163086,
+      "learning_rate": 9.858670610873528e-06,
+      "loss": 0.2485,
+      "step": 176
+    },
+    {
+      "epoch": 0.7625201938610663,
+      "grad_norm": 1.0042487382888794,
+      "learning_rate": 9.857067749979057e-06,
+      "loss": 0.2707,
+      "step": 177
+    },
+    {
+      "epoch": 0.7668282175551966,
+      "grad_norm": 0.9007334113121033,
+      "learning_rate": 9.855455982773288e-06,
+      "loss": 0.2721,
+      "step": 178
+    },
+    {
+      "epoch": 0.7711362412493269,
+      "grad_norm": 0.9035285115242004,
+      "learning_rate": 9.853835312211692e-06,
+      "loss": 0.2471,
+      "step": 179
+    },
+    {
+      "epoch": 0.7754442649434572,
+      "grad_norm": 1.0299876928329468,
+      "learning_rate": 9.852205741266058e-06,
+      "loss": 0.2767,
+      "step": 180
+    },
+    {
+      "epoch": 0.7797522886375875,
+      "grad_norm": 1.1170578002929688,
+      "learning_rate": 9.8505672729245e-06,
+      "loss": 0.2659,
+      "step": 181
+    },
+    {
+      "epoch": 0.7840603123317178,
+      "grad_norm": 0.9261436462402344,
+      "learning_rate": 9.848919910191446e-06,
+      "loss": 0.2569,
+      "step": 182
+    },
+    {
+      "epoch": 0.7883683360258481,
+      "grad_norm": 0.9240891933441162,
+      "learning_rate": 9.847263656087633e-06,
+      "loss": 0.2384,
+      "step": 183
+    },
+    {
+      "epoch": 0.7926763597199784,
+      "grad_norm": 1.0352424383163452,
+      "learning_rate": 9.845598513650104e-06,
+      "loss": 0.2422,
+      "step": 184
+    },
+    {
+      "epoch": 0.7969843834141088,
+      "grad_norm": 0.9237746596336365,
+      "learning_rate": 9.843924485932195e-06,
+      "loss": 0.2417,
+      "step": 185
+    },
+    {
+      "epoch": 0.8012924071082391,
+      "grad_norm": 0.9721134901046753,
+      "learning_rate": 9.84224157600354e-06,
+      "loss": 0.2504,
+      "step": 186
+    },
+    {
+      "epoch": 0.8056004308023694,
+      "grad_norm": 1.0062624216079712,
+      "learning_rate": 9.840549786950058e-06,
+      "loss": 0.2513,
+      "step": 187
+    },
+    {
+      "epoch": 0.8099084544964997,
+      "grad_norm": 0.9284637570381165,
+      "learning_rate": 9.83884912187395e-06,
+      "loss": 0.2466,
+      "step": 188
+    },
+    {
+      "epoch": 0.81421647819063,
+      "grad_norm": 0.9498764872550964,
+      "learning_rate": 9.837139583893693e-06,
+      "loss": 0.2568,
+      "step": 189
+    },
+    {
+      "epoch": 0.8185245018847603,
+      "grad_norm": 0.9171614050865173,
+      "learning_rate": 9.835421176144035e-06,
+      "loss": 0.2595,
+      "step": 190
+    },
+    {
+      "epoch": 0.8228325255788906,
+      "grad_norm": 0.914812445640564,
+      "learning_rate": 9.833693901775985e-06,
+      "loss": 0.2518,
+      "step": 191
+    },
+    {
+      "epoch": 0.827140549273021,
+      "grad_norm": 1.0692369937896729,
+      "learning_rate": 9.831957763956814e-06,
+      "loss": 0.2395,
+      "step": 192
+    },
+    {
+      "epoch": 0.8314485729671514,
+      "grad_norm": 0.9172776341438293,
+      "learning_rate": 9.830212765870043e-06,
+      "loss": 0.2523,
+      "step": 193
+    },
+    {
+      "epoch": 0.8357565966612817,
+      "grad_norm": 1.1387518644332886,
+      "learning_rate": 9.828458910715442e-06,
+      "loss": 0.2571,
+      "step": 194
+    },
+    {
+      "epoch": 0.840064620355412,
+      "grad_norm": 0.9793438911437988,
+      "learning_rate": 9.826696201709022e-06,
+      "loss": 0.2421,
+      "step": 195
+    },
+    {
+      "epoch": 0.8443726440495423,
+      "grad_norm": 0.9396490454673767,
+      "learning_rate": 9.824924642083026e-06,
+      "loss": 0.2442,
+      "step": 196
+    },
+    {
+      "epoch": 0.8486806677436726,
+      "grad_norm": 0.889569103717804,
+      "learning_rate": 9.823144235085934e-06,
+      "loss": 0.2389,
+      "step": 197
+    },
+    {
+      "epoch": 0.8529886914378029,
+      "grad_norm": 0.931921124458313,
+      "learning_rate": 9.821354983982438e-06,
+      "loss": 0.259,
+      "step": 198
+    },
+    {
+      "epoch": 0.8572967151319332,
+      "grad_norm": 0.7875272035598755,
+      "learning_rate": 9.819556892053456e-06,
+      "loss": 0.2132,
+      "step": 199
+    },
+    {
+      "epoch": 0.8616047388260636,
+      "grad_norm": 0.9318849444389343,
+      "learning_rate": 9.817749962596115e-06,
+      "loss": 0.2516,
+      "step": 200
+    },
+    {
+      "epoch": 0.8659127625201939,
+      "grad_norm": 0.948217511177063,
+      "learning_rate": 9.815934198923746e-06,
+      "loss": 0.2577,
+      "step": 201
+    },
+    {
+      "epoch": 0.8702207862143242,
+      "grad_norm": 0.8567696213722229,
+      "learning_rate": 9.814109604365878e-06,
+      "loss": 0.2431,
+      "step": 202
+    },
+    {
+      "epoch": 0.8745288099084545,
+      "grad_norm": 0.8806264996528625,
+      "learning_rate": 9.812276182268236e-06,
+      "loss": 0.2342,
+      "step": 203
+    },
+    {
+      "epoch": 0.8788368336025848,
+      "grad_norm": 0.968190610408783,
+      "learning_rate": 9.810433935992734e-06,
+      "loss": 0.2636,
+      "step": 204
+    },
+    {
+      "epoch": 0.8831448572967151,
+      "grad_norm": 0.9591284394264221,
+      "learning_rate": 9.808582868917458e-06,
+      "loss": 0.2364,
+      "step": 205
+    },
+    {
+      "epoch": 0.8874528809908454,
+      "grad_norm": 0.930848240852356,
+      "learning_rate": 9.806722984436676e-06,
+      "loss": 0.2372,
+      "step": 206
+    },
+    {
+      "epoch": 0.8917609046849758,
+      "grad_norm": 0.9203491806983948,
+      "learning_rate": 9.804854285960823e-06,
+      "loss": 0.2441,
+      "step": 207
+    },
+    {
+      "epoch": 0.8960689283791061,
+      "grad_norm": 0.8258495926856995,
+      "learning_rate": 9.802976776916493e-06,
+      "loss": 0.2254,
+      "step": 208
+    },
+    {
+      "epoch": 0.9003769520732364,
+      "grad_norm": 0.936922550201416,
+      "learning_rate": 9.801090460746442e-06,
+      "loss": 0.2519,
+      "step": 209
+    },
+    {
+      "epoch": 0.9046849757673667,
+      "grad_norm": 0.8387218117713928,
+      "learning_rate": 9.799195340909569e-06,
+      "loss": 0.2433,
+      "step": 210
+    },
+    {
+      "epoch": 0.908992999461497,
+      "grad_norm": 0.9370262026786804,
+      "learning_rate": 9.79729142088092e-06,
+      "loss": 0.2384,
+      "step": 211
+    },
+    {
+      "epoch": 0.9133010231556273,
+      "grad_norm": 0.8370246291160583,
+      "learning_rate": 9.795378704151675e-06,
+      "loss": 0.2554,
+      "step": 212
+    },
+    {
+      "epoch": 0.9176090468497576,
+      "grad_norm": 0.8435834050178528,
+      "learning_rate": 9.793457194229145e-06,
+      "loss": 0.2372,
+      "step": 213
+    },
+    {
+      "epoch": 0.9219170705438879,
+      "grad_norm": 1.1667269468307495,
+      "learning_rate": 9.791526894636767e-06,
+      "loss": 0.2446,
+      "step": 214
+    },
+    {
+      "epoch": 0.9262250942380184,
+      "grad_norm": 0.9069737792015076,
+      "learning_rate": 9.789587808914094e-06,
+      "loss": 0.2571,
+      "step": 215
+    },
+    {
+      "epoch": 0.9305331179321487,
+      "grad_norm": 0.844599187374115,
+      "learning_rate": 9.787639940616789e-06,
+      "loss": 0.2547,
+      "step": 216
+    },
+    {
+      "epoch": 0.934841141626279,
+      "grad_norm": 0.8763352632522583,
+      "learning_rate": 9.785683293316622e-06,
+      "loss": 0.2562,
+      "step": 217
+    },
+    {
+      "epoch": 0.9391491653204093,
+      "grad_norm": 0.7414149045944214,
+      "learning_rate": 9.783717870601458e-06,
+      "loss": 0.2158,
+      "step": 218
+    },
+    {
+      "epoch": 0.9434571890145396,
+      "grad_norm": 0.8541145324707031,
+      "learning_rate": 9.781743676075257e-06,
+      "loss": 0.2425,
+      "step": 219
+    },
+    {
+      "epoch": 0.9477652127086699,
+      "grad_norm": 0.869875431060791,
+      "learning_rate": 9.77976071335806e-06,
+      "loss": 0.2241,
+      "step": 220
+    },
+    {
+      "epoch": 0.9520732364028002,
+      "grad_norm": 0.8272488713264465,
+      "learning_rate": 9.777768986085985e-06,
+      "loss": 0.2454,
+      "step": 221
+    },
+    {
+      "epoch": 0.9563812600969306,
+      "grad_norm": 0.8880266547203064,
+      "learning_rate": 9.775768497911226e-06,
+      "loss": 0.2531,
+      "step": 222
+    },
+    {
+      "epoch": 0.9606892837910609,
+      "grad_norm": 0.9570372700691223,
+      "learning_rate": 9.77375925250204e-06,
+      "loss": 0.2691,
+      "step": 223
+    },
+    {
+      "epoch": 0.9649973074851912,
+      "grad_norm": 0.8235428333282471,
+      "learning_rate": 9.771741253542742e-06,
+      "loss": 0.257,
+      "step": 224
+    },
+    {
+      "epoch": 0.9693053311793215,
+      "grad_norm": 1.0435497760772705,
+      "learning_rate": 9.769714504733695e-06,
+      "loss": 0.2724,
+      "step": 225
+    },
+    {
+      "epoch": 0.9736133548734518,
+      "grad_norm": 0.8413583040237427,
+      "learning_rate": 9.767679009791312e-06,
+      "loss": 0.2519,
+      "step": 226
+    },
+    {
+      "epoch": 0.9779213785675821,
+      "grad_norm": 0.8566842079162598,
+      "learning_rate": 9.765634772448039e-06,
+      "loss": 0.24,
+      "step": 227
+    },
+    {
+      "epoch": 0.9822294022617124,
+      "grad_norm": 0.9400589466094971,
+      "learning_rate": 9.763581796452353e-06,
+      "loss": 0.2264,
+      "step": 228
+    },
+    {
+      "epoch": 0.9865374259558427,
+      "grad_norm": 0.7903165221214294,
+      "learning_rate": 9.76152008556876e-06,
+      "loss": 0.2555,
+      "step": 229
+    },
+    {
+      "epoch": 0.9908454496499731,
+      "grad_norm": 0.9579477310180664,
+      "learning_rate": 9.759449643577779e-06,
+      "loss": 0.2428,
+      "step": 230
+    },
+    {
+      "epoch": 0.9951534733441034,
+      "grad_norm": 0.8865399360656738,
+      "learning_rate": 9.757370474275938e-06,
+      "loss": 0.2425,
+      "step": 231
+    },
+    {
+      "epoch": 0.9994614970382337,
+      "grad_norm": 0.8118222951889038,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.2327,
+      "step": 232
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8118222951889038,
+      "learning_rate": 9.753185969005802e-06,
+      "loss": 0.3039,
+      "step": 233
+    },
+    {
+      "epoch": 1.0043080236941304,
+      "grad_norm": 3.10906982421875,
+      "learning_rate": 9.751080640710554e-06,
+      "loss": 0.2138,
+      "step": 234
+    },
+    {
+      "epoch": 1.0086160473882606,
+      "grad_norm": 0.8318678140640259,
+      "learning_rate": 9.748966600450526e-06,
+      "loss": 0.2287,
+      "step": 235
+    },
+    {
+      "epoch": 1.012924071082391,
+      "grad_norm": 0.8306625485420227,
+      "learning_rate": 9.746843852102191e-06,
+      "loss": 0.2144,
+      "step": 236
+    },
+    {
+      "epoch": 1.0172320947765212,
+      "grad_norm": 0.7923266291618347,
+      "learning_rate": 9.744712399557992e-06,
+      "loss": 0.2211,
+      "step": 237
+    },
+    {
+      "epoch": 1.0215401184706516,
+      "grad_norm": 0.9020610451698303,
+      "learning_rate": 9.742572246726336e-06,
+      "loss": 0.2431,
+      "step": 238
+    },
+    {
+      "epoch": 1.0258481421647818,
+      "grad_norm": 0.6959359645843506,
+      "learning_rate": 9.740423397531573e-06,
+      "loss": 0.2142,
+      "step": 239
+    },
+    {
+      "epoch": 1.0301561658589122,
+      "grad_norm": 0.725628137588501,
+      "learning_rate": 9.738265855914014e-06,
+      "loss": 0.2194,
+      "step": 240
+    },
+    {
+      "epoch": 1.0344641895530426,
+      "grad_norm": 0.8057601451873779,
+      "learning_rate": 9.736099625829894e-06,
+      "loss": 0.2358,
+      "step": 241
+    },
+    {
+      "epoch": 1.0387722132471728,
+      "grad_norm": 0.8351283669471741,
+      "learning_rate": 9.733924711251393e-06,
+      "loss": 0.2412,
+      "step": 242
+    },
+    {
+      "epoch": 1.0430802369413033,
+      "grad_norm": 0.8563336730003357,
+      "learning_rate": 9.731741116166607e-06,
+      "loss": 0.2138,
+      "step": 243
+    },
+    {
+      "epoch": 1.0473882606354334,
+      "grad_norm": 0.8097681999206543,
+      "learning_rate": 9.729548844579552e-06,
+      "loss": 0.2256,
+      "step": 244
+    },
+    {
+      "epoch": 1.0516962843295639,
+      "grad_norm": 0.7601490616798401,
+      "learning_rate": 9.727347900510155e-06,
+      "loss": 0.208,
+      "step": 245
+    },
+    {
+      "epoch": 1.056004308023694,
+      "grad_norm": 0.7128692269325256,
+      "learning_rate": 9.725138287994246e-06,
+      "loss": 0.2121,
+      "step": 246
+    },
+    {
+      "epoch": 1.0603123317178245,
+      "grad_norm": 0.7977615594863892,
+      "learning_rate": 9.722920011083546e-06,
+      "loss": 0.2319,
+      "step": 247
+    },
+    {
+      "epoch": 1.0646203554119547,
+      "grad_norm": 0.874777615070343,
+      "learning_rate": 9.720693073845668e-06,
+      "loss": 0.2166,
+      "step": 248
+    },
+    {
+      "epoch": 1.068928379106085,
+      "grad_norm": 0.8400263786315918,
+      "learning_rate": 9.718457480364103e-06,
+      "loss": 0.2517,
+      "step": 249
+    },
+    {
+      "epoch": 1.0732364028002155,
+      "grad_norm": 0.8173924684524536,
+      "learning_rate": 9.716213234738216e-06,
+      "loss": 0.2227,
+      "step": 250
+    },
+    {
+      "epoch": 1.0775444264943457,
+      "grad_norm": 0.7754281759262085,
+      "learning_rate": 9.713960341083237e-06,
+      "loss": 0.1943,
+      "step": 251
+    },
+    {
+      "epoch": 1.081852450188476,
+      "grad_norm": 0.8374117612838745,
+      "learning_rate": 9.711698803530253e-06,
+      "loss": 0.237,
+      "step": 252
+    },
+    {
+      "epoch": 1.0861604738826063,
+      "grad_norm": 0.7473996877670288,
+      "learning_rate": 9.709428626226204e-06,
+      "loss": 0.208,
+      "step": 253
+    },
+    {
+      "epoch": 1.0904684975767367,
+      "grad_norm": 0.7273510694503784,
+      "learning_rate": 9.707149813333866e-06,
+      "loss": 0.2177,
+      "step": 254
+    },
+    {
+      "epoch": 1.0947765212708669,
+      "grad_norm": 0.8145310878753662,
+      "learning_rate": 9.704862369031857e-06,
+      "loss": 0.24,
+      "step": 255
+    },
+    {
+      "epoch": 1.0990845449649973,
+      "grad_norm": 0.8657172322273254,
+      "learning_rate": 9.70256629751462e-06,
+      "loss": 0.2252,
+      "step": 256
+    },
+    {
+      "epoch": 1.1033925686591277,
+      "grad_norm": 0.9188759922981262,
+      "learning_rate": 9.700261602992417e-06,
+      "loss": 0.241,
+      "step": 257
+    },
+    {
+      "epoch": 1.107700592353258,
+      "grad_norm": 0.7244163751602173,
+      "learning_rate": 9.69794828969132e-06,
+      "loss": 0.2139,
+      "step": 258
+    },
+    {
+      "epoch": 1.1120086160473883,
+      "grad_norm": 0.9003731608390808,
+      "learning_rate": 9.695626361853207e-06,
+      "loss": 0.2256,
+      "step": 259
+    },
+    {
+      "epoch": 1.1163166397415185,
+      "grad_norm": 0.7939245700836182,
+      "learning_rate": 9.693295823735754e-06,
+      "loss": 0.2345,
+      "step": 260
+    },
+    {
+      "epoch": 1.120624663435649,
+      "grad_norm": 0.7665999531745911,
+      "learning_rate": 9.690956679612422e-06,
+      "loss": 0.2204,
+      "step": 261
+    },
+    {
+      "epoch": 1.1249326871297791,
+      "grad_norm": 0.758368194103241,
+      "learning_rate": 9.688608933772454e-06,
+      "loss": 0.2057,
+      "step": 262
+    },
+    {
+      "epoch": 1.1292407108239095,
+      "grad_norm": 0.8348474502563477,
+      "learning_rate": 9.686252590520869e-06,
+      "loss": 0.2158,
+      "step": 263
+    },
+    {
+      "epoch": 1.13354873451804,
+      "grad_norm": 0.7352973818778992,
+      "learning_rate": 9.683887654178446e-06,
+      "loss": 0.2097,
+      "step": 264
+    },
+    {
+      "epoch": 1.1378567582121701,
+      "grad_norm": 0.8477746844291687,
+      "learning_rate": 9.681514129081725e-06,
+      "loss": 0.2288,
+      "step": 265
+    },
+    {
+      "epoch": 1.1421647819063006,
+      "grad_norm": 0.7330527305603027,
+      "learning_rate": 9.679132019582988e-06,
+      "loss": 0.2179,
+      "step": 266
+    },
+    {
+      "epoch": 1.1464728056004307,
+      "grad_norm": 0.7249947190284729,
+      "learning_rate": 9.67674133005027e-06,
+      "loss": 0.2062,
+      "step": 267
+    },
+    {
+      "epoch": 1.1507808292945612,
+      "grad_norm": 0.8508214354515076,
+      "learning_rate": 9.674342064867326e-06,
+      "loss": 0.2142,
+      "step": 268
+    },
+    {
+      "epoch": 1.1550888529886914,
+      "grad_norm": 0.9992220401763916,
+      "learning_rate": 9.671934228433647e-06,
+      "loss": 0.2201,
+      "step": 269
+    },
+    {
+      "epoch": 1.1593968766828218,
+      "grad_norm": 0.722809910774231,
+      "learning_rate": 9.669517825164435e-06,
+      "loss": 0.215,
+      "step": 270
+    },
+    {
+      "epoch": 1.1637049003769522,
+      "grad_norm": 0.770807683467865,
+      "learning_rate": 9.667092859490599e-06,
+      "loss": 0.2112,
+      "step": 271
+    },
+    {
+      "epoch": 1.1680129240710824,
+      "grad_norm": 0.7955392003059387,
+      "learning_rate": 9.664659335858755e-06,
+      "loss": 0.211,
+      "step": 272
+    },
+    {
+      "epoch": 1.1723209477652128,
+      "grad_norm": 0.7562293410301208,
+      "learning_rate": 9.662217258731208e-06,
+      "loss": 0.2166,
+      "step": 273
+    },
+    {
+      "epoch": 1.176628971459343,
+      "grad_norm": 0.7826163172721863,
+      "learning_rate": 9.659766632585946e-06,
+      "loss": 0.2333,
+      "step": 274
+    },
+    {
+      "epoch": 1.1809369951534734,
+      "grad_norm": 0.7867729663848877,
+      "learning_rate": 9.657307461916637e-06,
+      "loss": 0.2181,
+      "step": 275
+    },
+    {
+      "epoch": 1.1852450188476036,
+      "grad_norm": 0.7663466334342957,
+      "learning_rate": 9.654839751232612e-06,
+      "loss": 0.2177,
+      "step": 276
+    },
+    {
+      "epoch": 1.189553042541734,
+      "grad_norm": 0.9194732308387756,
+      "learning_rate": 9.652363505058866e-06,
+      "loss": 0.2374,
+      "step": 277
+    },
+    {
+      "epoch": 1.1938610662358644,
+      "grad_norm": 0.8086763620376587,
+      "learning_rate": 9.649878727936044e-06,
+      "loss": 0.225,
+      "step": 278
+    },
+    {
+      "epoch": 1.1981690899299946,
+      "grad_norm": 0.8902127146720886,
+      "learning_rate": 9.647385424420435e-06,
+      "loss": 0.2024,
+      "step": 279
+    },
+    {
+      "epoch": 1.202477113624125,
+      "grad_norm": 0.8926809430122375,
+      "learning_rate": 9.644883599083959e-06,
+      "loss": 0.2201,
+      "step": 280
+    },
+    {
+      "epoch": 1.2067851373182552,
+      "grad_norm": 0.8426185846328735,
+      "learning_rate": 9.642373256514164e-06,
+      "loss": 0.2371,
+      "step": 281
+    },
+    {
+      "epoch": 1.2110931610123856,
+      "grad_norm": 0.7640630006790161,
+      "learning_rate": 9.639854401314219e-06,
+      "loss": 0.2148,
+      "step": 282
+    },
+    {
+      "epoch": 1.2154011847065158,
+      "grad_norm": 0.9093323349952698,
+      "learning_rate": 9.637327038102902e-06,
+      "loss": 0.2218,
+      "step": 283
+    },
+    {
+      "epoch": 1.2197092084006462,
+      "grad_norm": 0.8968831300735474,
+      "learning_rate": 9.634791171514585e-06,
+      "loss": 0.2298,
+      "step": 284
+    },
+    {
+      "epoch": 1.2240172320947766,
+      "grad_norm": 0.8954552412033081,
+      "learning_rate": 9.632246806199242e-06,
+      "loss": 0.2173,
+      "step": 285
+    },
+    {
+      "epoch": 1.2283252557889068,
+      "grad_norm": 0.8648595213890076,
+      "learning_rate": 9.629693946822423e-06,
+      "loss": 0.2273,
+      "step": 286
+    },
+    {
+      "epoch": 1.2326332794830372,
+      "grad_norm": 0.8752464056015015,
+      "learning_rate": 9.627132598065258e-06,
+      "loss": 0.2239,
+      "step": 287
+    },
+    {
+      "epoch": 1.2369413031771674,
+      "grad_norm": 0.830769419670105,
+      "learning_rate": 9.624562764624445e-06,
+      "loss": 0.2295,
+      "step": 288
+    },
+    {
+      "epoch": 1.2412493268712979,
+      "grad_norm": 0.8199811577796936,
+      "learning_rate": 9.621984451212237e-06,
+      "loss": 0.2256,
+      "step": 289
+    },
+    {
+      "epoch": 1.245557350565428,
+      "grad_norm": 0.8059448599815369,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.2196,
+      "step": 290
+    },
+    {
+      "epoch": 1.2498653742595585,
+      "grad_norm": 0.8064429759979248,
+      "learning_rate": 9.616802403400384e-06,
+      "loss": 0.2208,
+      "step": 291
+    },
+    {
+      "epoch": 1.2541733979536889,
+      "grad_norm": 0.8114342093467712,
+      "learning_rate": 9.614198678502965e-06,
+      "loss": 0.2172,
+      "step": 292
+    },
+    {
+      "epoch": 1.258481421647819,
+      "grad_norm": 0.7933779954910278,
+      "learning_rate": 9.611586492638573e-06,
+      "loss": 0.2121,
+      "step": 293
+    },
+    {
+      "epoch": 1.2627894453419493,
+      "grad_norm": 0.7357353568077087,
+      "learning_rate": 9.608965850597125e-06,
+      "loss": 0.2237,
+      "step": 294
+    },
+    {
+      "epoch": 1.2670974690360797,
+      "grad_norm": 0.7709231972694397,
+      "learning_rate": 9.606336757184041e-06,
+      "loss": 0.22,
+      "step": 295
+    },
+    {
+      "epoch": 1.27140549273021,
+      "grad_norm": 0.7920863032341003,
+      "learning_rate": 9.603699217220239e-06,
+      "loss": 0.2214,
+      "step": 296
+    },
+    {
+      "epoch": 1.2757135164243403,
+      "grad_norm": 0.8977172374725342,
+      "learning_rate": 9.601053235542124e-06,
+      "loss": 0.214,
+      "step": 297
+    },
+    {
+      "epoch": 1.2800215401184707,
+      "grad_norm": 0.750146210193634,
+      "learning_rate": 9.598398817001585e-06,
+      "loss": 0.2127,
+      "step": 298
+    },
+    {
+      "epoch": 1.284329563812601,
+      "grad_norm": 0.8150034546852112,
+      "learning_rate": 9.595735966465973e-06,
+      "loss": 0.2209,
+      "step": 299
+    },
+    {
+      "epoch": 1.2886375875067313,
+      "grad_norm": 0.819951057434082,
+      "learning_rate": 9.59306468881811e-06,
+      "loss": 0.2291,
+      "step": 300
+    },
+    {
+      "epoch": 1.2929456112008615,
+      "grad_norm": 0.7801006436347961,
+      "learning_rate": 9.590384988956264e-06,
+      "loss": 0.215,
+      "step": 301
+    },
+    {
+      "epoch": 1.297253634894992,
+      "grad_norm": 0.8777364492416382,
+      "learning_rate": 9.587696871794148e-06,
+      "loss": 0.2321,
+      "step": 302
+    },
+    {
+      "epoch": 1.3015616585891223,
+      "grad_norm": 0.8642171621322632,
+      "learning_rate": 9.585000342260914e-06,
+      "loss": 0.218,
+      "step": 303
+    },
+    {
+      "epoch": 1.3058696822832525,
+      "grad_norm": 0.9009878635406494,
+      "learning_rate": 9.582295405301131e-06,
+      "loss": 0.2283,
+      "step": 304
+    },
+    {
+      "epoch": 1.310177705977383,
+      "grad_norm": 0.9107750058174133,
+      "learning_rate": 9.579582065874794e-06,
+      "loss": 0.2297,
+      "step": 305
+    },
+    {
+      "epoch": 1.3144857296715131,
+      "grad_norm": 0.8410093784332275,
+      "learning_rate": 9.576860328957299e-06,
+      "loss": 0.2275,
+      "step": 306
+    },
+    {
+      "epoch": 1.3187937533656435,
+      "grad_norm": 0.8049638271331787,
+      "learning_rate": 9.574130199539443e-06,
+      "loss": 0.209,
+      "step": 307
+    },
+    {
+      "epoch": 1.3231017770597737,
+      "grad_norm": 0.7018632888793945,
+      "learning_rate": 9.571391682627413e-06,
+      "loss": 0.2034,
+      "step": 308
+    },
+    {
+      "epoch": 1.3274098007539041,
+      "grad_norm": 0.779147207736969,
+      "learning_rate": 9.568644783242771e-06,
+      "loss": 0.225,
+      "step": 309
+    },
+    {
+      "epoch": 1.3317178244480345,
+      "grad_norm": 0.725441575050354,
+      "learning_rate": 9.565889506422457e-06,
+      "loss": 0.2082,
+      "step": 310
+    },
+    {
+      "epoch": 1.3360258481421647,
+      "grad_norm": 0.7808791995048523,
+      "learning_rate": 9.563125857218766e-06,
+      "loss": 0.2217,
+      "step": 311
+    },
+    {
+      "epoch": 1.3403338718362952,
+      "grad_norm": 0.833756148815155,
+      "learning_rate": 9.56035384069935e-06,
+      "loss": 0.2377,
+      "step": 312
+    },
+    {
+      "epoch": 1.3446418955304253,
+      "grad_norm": 0.9143878817558289,
+      "learning_rate": 9.557573461947201e-06,
+      "loss": 0.2221,
+      "step": 313
+    },
+    {
+      "epoch": 1.3489499192245558,
+      "grad_norm": 0.8486077785491943,
+      "learning_rate": 9.554784726060647e-06,
+      "loss": 0.2403,
+      "step": 314
+    },
+    {
+      "epoch": 1.353257942918686,
+      "grad_norm": 0.7151229381561279,
+      "learning_rate": 9.551987638153339e-06,
+      "loss": 0.2175,
+      "step": 315
+    },
+    {
+      "epoch": 1.3575659666128164,
+      "grad_norm": 0.8472189903259277,
+      "learning_rate": 9.549182203354241e-06,
+      "loss": 0.232,
+      "step": 316
+    },
+    {
+      "epoch": 1.3618739903069468,
+      "grad_norm": 0.8481770753860474,
+      "learning_rate": 9.546368426807628e-06,
+      "loss": 0.2335,
+      "step": 317
+    },
+    {
+      "epoch": 1.366182014001077,
+      "grad_norm": 0.8769595623016357,
+      "learning_rate": 9.543546313673065e-06,
+      "loss": 0.226,
+      "step": 318
+    },
+    {
+      "epoch": 1.3704900376952074,
+      "grad_norm": 0.7629135251045227,
+      "learning_rate": 9.540715869125407e-06,
+      "loss": 0.2204,
+      "step": 319
+    },
+    {
+      "epoch": 1.3747980613893376,
+      "grad_norm": 0.7563349008560181,
+      "learning_rate": 9.537877098354787e-06,
+      "loss": 0.2123,
+      "step": 320
+    },
+    {
+      "epoch": 1.379106085083468,
+      "grad_norm": 0.6988772749900818,
+      "learning_rate": 9.5350300065666e-06,
+      "loss": 0.2211,
+      "step": 321
+    },
+    {
+      "epoch": 1.3834141087775982,
+      "grad_norm": 0.7229964137077332,
+      "learning_rate": 9.532174598981507e-06,
+      "loss": 0.2222,
+      "step": 322
+    },
+    {
+      "epoch": 1.3877221324717286,
+      "grad_norm": 0.7834879755973816,
+      "learning_rate": 9.529310880835414e-06,
+      "loss": 0.2038,
+      "step": 323
+    },
+    {
+      "epoch": 1.392030156165859,
+      "grad_norm": 0.7434644103050232,
+      "learning_rate": 9.526438857379463e-06,
+      "loss": 0.231,
+      "step": 324
+    },
+    {
+      "epoch": 1.3963381798599892,
+      "grad_norm": 0.9631347060203552,
+      "learning_rate": 9.52355853388003e-06,
+      "loss": 0.2262,
+      "step": 325
+    },
+    {
+      "epoch": 1.4006462035541196,
+      "grad_norm": 0.6573905348777771,
+      "learning_rate": 9.520669915618708e-06,
+      "loss": 0.2094,
+      "step": 326
+    },
+    {
+      "epoch": 1.4049542272482498,
+      "grad_norm": 0.8617786765098572,
+      "learning_rate": 9.5177730078923e-06,
+      "loss": 0.2167,
+      "step": 327
+    },
+    {
+      "epoch": 1.4092622509423802,
+      "grad_norm": 0.7743338942527771,
+      "learning_rate": 9.514867816012809e-06,
+      "loss": 0.2188,
+      "step": 328
+    },
+    {
+      "epoch": 1.4135702746365104,
+      "grad_norm": 0.8752392530441284,
+      "learning_rate": 9.511954345307432e-06,
+      "loss": 0.2245,
+      "step": 329
+    },
+    {
+      "epoch": 1.4178782983306408,
+      "grad_norm": 0.8186882734298706,
+      "learning_rate": 9.509032601118541e-06,
+      "loss": 0.2185,
+      "step": 330
+    },
+    {
+      "epoch": 1.4221863220247712,
+      "grad_norm": 0.781764805316925,
+      "learning_rate": 9.506102588803683e-06,
+      "loss": 0.2392,
+      "step": 331
+    },
+    {
+      "epoch": 1.4264943457189014,
+      "grad_norm": 0.758326530456543,
+      "learning_rate": 9.503164313735566e-06,
+      "loss": 0.2127,
+      "step": 332
+    },
+    {
+      "epoch": 1.4308023694130318,
+      "grad_norm": 0.7992167472839355,
+      "learning_rate": 9.500217781302048e-06,
+      "loss": 0.2087,
+      "step": 333
+    },
+    {
+      "epoch": 1.435110393107162,
+      "grad_norm": 0.764522910118103,
+      "learning_rate": 9.497262996906126e-06,
+      "loss": 0.209,
+      "step": 334
+    },
+    {
+      "epoch": 1.4394184168012925,
+      "grad_norm": 1.0435947179794312,
+      "learning_rate": 9.494299965965935e-06,
+      "loss": 0.2547,
+      "step": 335
+    },
+    {
+      "epoch": 1.4437264404954226,
+      "grad_norm": 0.826876163482666,
+      "learning_rate": 9.491328693914723e-06,
+      "loss": 0.2252,
+      "step": 336
+    },
+    {
+      "epoch": 1.448034464189553,
+      "grad_norm": 0.6937296986579895,
+      "learning_rate": 9.488349186200858e-06,
+      "loss": 0.2105,
+      "step": 337
+    },
+    {
+      "epoch": 1.4523424878836835,
+      "grad_norm": 0.8464234471321106,
+      "learning_rate": 9.485361448287804e-06,
+      "loss": 0.2039,
+      "step": 338
+    },
+    {
+      "epoch": 1.4566505115778137,
+      "grad_norm": 0.8548002243041992,
+      "learning_rate": 9.482365485654118e-06,
+      "loss": 0.232,
+      "step": 339
+    },
+    {
+      "epoch": 1.4609585352719439,
+      "grad_norm": 0.7955244779586792,
+      "learning_rate": 9.479361303793441e-06,
+      "loss": 0.217,
+      "step": 340
+    },
+    {
+      "epoch": 1.4652665589660743,
+      "grad_norm": 0.7540826797485352,
+      "learning_rate": 9.476348908214482e-06,
+      "loss": 0.2287,
+      "step": 341
+    },
+    {
+      "epoch": 1.4695745826602047,
+      "grad_norm": 0.8318302631378174,
+      "learning_rate": 9.47332830444101e-06,
+      "loss": 0.2193,
+      "step": 342
+    },
+    {
+      "epoch": 1.4738826063543349,
+      "grad_norm": 0.8411695957183838,
+      "learning_rate": 9.470299498011851e-06,
+      "loss": 0.2377,
+      "step": 343
+    },
+    {
+      "epoch": 1.4781906300484653,
+      "grad_norm": 0.837867021560669,
+      "learning_rate": 9.46726249448087e-06,
+      "loss": 0.2201,
+      "step": 344
+    },
+    {
+      "epoch": 1.4824986537425957,
+      "grad_norm": 0.7983565926551819,
+      "learning_rate": 9.464217299416956e-06,
+      "loss": 0.2238,
+      "step": 345
+    },
+    {
+      "epoch": 1.486806677436726,
+      "grad_norm": 0.763559103012085,
+      "learning_rate": 9.46116391840403e-06,
+      "loss": 0.2173,
+      "step": 346
+    },
+    {
+      "epoch": 1.491114701130856,
+      "grad_norm": 0.769709587097168,
+      "learning_rate": 9.458102357041017e-06,
+      "loss": 0.2347,
+      "step": 347
+    },
+    {
+      "epoch": 1.4954227248249865,
+      "grad_norm": 0.9357376098632812,
+      "learning_rate": 9.45503262094184e-06,
+      "loss": 0.2402,
+      "step": 348
+    },
+    {
+      "epoch": 1.499730748519117,
+      "grad_norm": 0.729263186454773,
+      "learning_rate": 9.451954715735416e-06,
+      "loss": 0.226,
+      "step": 349
+    },
+    {
+      "epoch": 1.504038772213247,
+      "grad_norm": 0.84844970703125,
+      "learning_rate": 9.448868647065644e-06,
+      "loss": 0.2489,
+      "step": 350
+    },
+    {
+      "epoch": 1.5083467959073775,
+      "grad_norm": 0.9604891538619995,
+      "learning_rate": 9.445774420591382e-06,
+      "loss": 0.2311,
+      "step": 351
+    },
+    {
+      "epoch": 1.512654819601508,
+      "grad_norm": 0.8172181248664856,
+      "learning_rate": 9.442672041986456e-06,
+      "loss": 0.2167,
+      "step": 352
+    },
+    {
+      "epoch": 1.5169628432956381,
+      "grad_norm": 0.9129267334938049,
+      "learning_rate": 9.43956151693964e-06,
+      "loss": 0.2422,
+      "step": 353
+    },
+    {
+      "epoch": 1.5212708669897683,
+      "grad_norm": 0.7796788811683655,
+      "learning_rate": 9.436442851154642e-06,
+      "loss": 0.2182,
+      "step": 354
+    },
+    {
+      "epoch": 1.5255788906838987,
+      "grad_norm": 0.7742928862571716,
+      "learning_rate": 9.433316050350099e-06,
+      "loss": 0.2075,
+      "step": 355
+    },
+    {
+      "epoch": 1.5298869143780292,
+      "grad_norm": 0.8352739214897156,
+      "learning_rate": 9.430181120259566e-06,
+      "loss": 0.2111,
+      "step": 356
+    },
+    {
+      "epoch": 1.5341949380721593,
+      "grad_norm": 0.7576451301574707,
+      "learning_rate": 9.427038066631502e-06,
+      "loss": 0.2066,
+      "step": 357
+    },
+    {
+      "epoch": 1.5385029617662898,
+      "grad_norm": 0.8700200319290161,
+      "learning_rate": 9.423886895229266e-06,
+      "loss": 0.2375,
+      "step": 358
+    },
+    {
+      "epoch": 1.5428109854604202,
+      "grad_norm": 0.7757928967475891,
+      "learning_rate": 9.420727611831098e-06,
+      "loss": 0.2272,
+      "step": 359
+    },
+    {
+      "epoch": 1.5471190091545504,
+      "grad_norm": 0.7936514616012573,
+      "learning_rate": 9.417560222230115e-06,
+      "loss": 0.2118,
+      "step": 360
+    },
+    {
+      "epoch": 1.5514270328486806,
+      "grad_norm": 0.9016469120979309,
+      "learning_rate": 9.414384732234301e-06,
+      "loss": 0.229,
+      "step": 361
+    },
+    {
+      "epoch": 1.555735056542811,
+      "grad_norm": 0.7308838367462158,
+      "learning_rate": 9.411201147666486e-06,
+      "loss": 0.2067,
+      "step": 362
+    },
+    {
+      "epoch": 1.5600430802369414,
+      "grad_norm": 0.7090045213699341,
+      "learning_rate": 9.408009474364353e-06,
+      "loss": 0.2198,
+      "step": 363
+    },
+    {
+      "epoch": 1.5643511039310716,
+      "grad_norm": 0.8225852251052856,
+      "learning_rate": 9.404809718180408e-06,
+      "loss": 0.2323,
+      "step": 364
+    },
+    {
+      "epoch": 1.568659127625202,
+      "grad_norm": 0.7157912254333496,
+      "learning_rate": 9.401601884981983e-06,
+      "loss": 0.2192,
+      "step": 365
+    },
+    {
+      "epoch": 1.5729671513193324,
+      "grad_norm": 0.7957601547241211,
+      "learning_rate": 9.39838598065122e-06,
+      "loss": 0.2152,
+      "step": 366
+    },
+    {
+      "epoch": 1.5772751750134626,
+      "grad_norm": 0.7559109330177307,
+      "learning_rate": 9.39516201108506e-06,
+      "loss": 0.242,
+      "step": 367
+    },
+    {
+      "epoch": 1.5815831987075928,
+      "grad_norm": 0.8238731026649475,
+      "learning_rate": 9.391929982195233e-06,
+      "loss": 0.2238,
+      "step": 368
+    },
+    {
+      "epoch": 1.5858912224017232,
+      "grad_norm": 0.8968952298164368,
+      "learning_rate": 9.38868989990825e-06,
+      "loss": 0.2228,
+      "step": 369
+    },
+    {
+      "epoch": 1.5901992460958536,
+      "grad_norm": 0.724861741065979,
+      "learning_rate": 9.385441770165385e-06,
+      "loss": 0.2201,
+      "step": 370
+    },
+    {
+      "epoch": 1.5945072697899838,
+      "grad_norm": 0.7094451189041138,
+      "learning_rate": 9.382185598922674e-06,
+      "loss": 0.2136,
+      "step": 371
+    },
+    {
+      "epoch": 1.598815293484114,
+      "grad_norm": 0.7279621958732605,
+      "learning_rate": 9.378921392150893e-06,
+      "loss": 0.2236,
+      "step": 372
+    },
+    {
+      "epoch": 1.6031233171782446,
+      "grad_norm": 0.7491332292556763,
+      "learning_rate": 9.375649155835554e-06,
+      "loss": 0.197,
+      "step": 373
+    },
+    {
+      "epoch": 1.6074313408723748,
+      "grad_norm": 0.796010434627533,
+      "learning_rate": 9.372368895976896e-06,
+      "loss": 0.2332,
+      "step": 374
+    },
+    {
+      "epoch": 1.611739364566505,
+      "grad_norm": 0.7918517589569092,
+      "learning_rate": 9.369080618589866e-06,
+      "loss": 0.2238,
+      "step": 375
+    },
+    {
+      "epoch": 1.6160473882606354,
+      "grad_norm": 0.9035218954086304,
+      "learning_rate": 9.365784329704114e-06,
+      "loss": 0.2495,
+      "step": 376
+    },
+    {
+      "epoch": 1.6203554119547658,
+      "grad_norm": 0.9317935705184937,
+      "learning_rate": 9.362480035363987e-06,
+      "loss": 0.2425,
+      "step": 377
+    },
+    {
+      "epoch": 1.624663435648896,
+      "grad_norm": 0.7318428158760071,
+      "learning_rate": 9.3591677416285e-06,
+      "loss": 0.2084,
+      "step": 378
+    },
+    {
+      "epoch": 1.6289714593430262,
+      "grad_norm": 0.798697292804718,
+      "learning_rate": 9.35584745457134e-06,
+      "loss": 0.2235,
+      "step": 379
+    },
+    {
+      "epoch": 1.6332794830371569,
+      "grad_norm": 0.7385568022727966,
+      "learning_rate": 9.352519180280862e-06,
+      "loss": 0.2124,
+      "step": 380
+    },
+    {
+      "epoch": 1.637587506731287,
+      "grad_norm": 0.7316162586212158,
+      "learning_rate": 9.34918292486005e-06,
+      "loss": 0.2357,
+      "step": 381
+    },
+    {
+      "epoch": 1.6418955304254172,
+      "grad_norm": 0.8386530876159668,
+      "learning_rate": 9.345838694426535e-06,
+      "loss": 0.2426,
+      "step": 382
+    },
+    {
+      "epoch": 1.6462035541195477,
+      "grad_norm": 0.7922302484512329,
+      "learning_rate": 9.342486495112566e-06,
+      "loss": 0.234,
+      "step": 383
+    },
+    {
+      "epoch": 1.650511577813678,
+      "grad_norm": 0.8359029293060303,
+      "learning_rate": 9.339126333065008e-06,
+      "loss": 0.2255,
+      "step": 384
+    },
+    {
+      "epoch": 1.6548196015078083,
+      "grad_norm": 0.7093521356582642,
+      "learning_rate": 9.335758214445323e-06,
+      "loss": 0.2255,
+      "step": 385
+    },
+    {
+      "epoch": 1.6591276252019385,
+      "grad_norm": 0.7251277565956116,
+      "learning_rate": 9.332382145429568e-06,
+      "loss": 0.2144,
+      "step": 386
+    },
+    {
+      "epoch": 1.6634356488960689,
+      "grad_norm": 0.7854430675506592,
+      "learning_rate": 9.328998132208373e-06,
+      "loss": 0.2191,
+      "step": 387
+    },
+    {
+      "epoch": 1.6677436725901993,
+      "grad_norm": 0.750257670879364,
+      "learning_rate": 9.325606180986938e-06,
+      "loss": 0.216,
+      "step": 388
+    },
+    {
+      "epoch": 1.6720516962843295,
+      "grad_norm": 0.7411738038063049,
+      "learning_rate": 9.32220629798502e-06,
+      "loss": 0.2179,
+      "step": 389
+    },
+    {
+      "epoch": 1.67635971997846,
+      "grad_norm": 0.7300736308097839,
+      "learning_rate": 9.318798489436917e-06,
+      "loss": 0.2101,
+      "step": 390
+    },
+    {
+      "epoch": 1.6806677436725903,
+      "grad_norm": 0.8182831406593323,
+      "learning_rate": 9.315382761591463e-06,
+      "loss": 0.214,
+      "step": 391
+    },
+    {
+      "epoch": 1.6849757673667205,
+      "grad_norm": 0.7903637290000916,
+      "learning_rate": 9.311959120712012e-06,
+      "loss": 0.2188,
+      "step": 392
+    },
+    {
+      "epoch": 1.6892837910608507,
+      "grad_norm": 0.761500358581543,
+      "learning_rate": 9.308527573076425e-06,
+      "loss": 0.2278,
+      "step": 393
+    },
+    {
+      "epoch": 1.693591814754981,
+      "grad_norm": 0.7488672733306885,
+      "learning_rate": 9.30508812497707e-06,
+      "loss": 0.2152,
+      "step": 394
+    },
+    {
+      "epoch": 1.6978998384491115,
+      "grad_norm": 0.76832115650177,
+      "learning_rate": 9.301640782720792e-06,
+      "loss": 0.2244,
+      "step": 395
+    },
+    {
+      "epoch": 1.7022078621432417,
+      "grad_norm": 0.7345932722091675,
+      "learning_rate": 9.298185552628917e-06,
+      "loss": 0.2184,
+      "step": 396
+    },
+    {
+      "epoch": 1.7065158858373721,
+      "grad_norm": 0.7528600692749023,
+      "learning_rate": 9.294722441037238e-06,
+      "loss": 0.2451,
+      "step": 397
+    },
+    {
+      "epoch": 1.7108239095315025,
+      "grad_norm": 0.7345549464225769,
+      "learning_rate": 9.291251454295989e-06,
+      "loss": 0.2043,
+      "step": 398
+    },
+    {
+      "epoch": 1.7151319332256327,
+      "grad_norm": 0.6699172258377075,
+      "learning_rate": 9.287772598769855e-06,
+      "loss": 0.2134,
+      "step": 399
+    },
+    {
+      "epoch": 1.719439956919763,
+      "grad_norm": 0.7636283040046692,
+      "learning_rate": 9.284285880837947e-06,
+      "loss": 0.2148,
+      "step": 400
+    },
+    {
+      "epoch": 1.7237479806138933,
+      "grad_norm": 0.9217869639396667,
+      "learning_rate": 9.28079130689379e-06,
+      "loss": 0.2317,
+      "step": 401
+    },
+    {
+      "epoch": 1.7280560043080238,
+      "grad_norm": 0.8372043967247009,
+      "learning_rate": 9.277288883345318e-06,
+      "loss": 0.2244,
+      "step": 402
+    },
+    {
+      "epoch": 1.732364028002154,
+      "grad_norm": 0.7365742325782776,
+      "learning_rate": 9.273778616614857e-06,
+      "loss": 0.2215,
+      "step": 403
+    },
+    {
+      "epoch": 1.7366720516962844,
+      "grad_norm": 0.712045431137085,
+      "learning_rate": 9.270260513139116e-06,
+      "loss": 0.2107,
+      "step": 404
+    },
+    {
+      "epoch": 1.7409800753904148,
+      "grad_norm": 0.6611554622650146,
+      "learning_rate": 9.266734579369172e-06,
+      "loss": 0.222,
+      "step": 405
+    },
+    {
+      "epoch": 1.745288099084545,
+      "grad_norm": 0.8067266941070557,
+      "learning_rate": 9.263200821770462e-06,
+      "loss": 0.2069,
+      "step": 406
+    },
+    {
+      "epoch": 1.7495961227786752,
+      "grad_norm": 0.7734603881835938,
+      "learning_rate": 9.25965924682277e-06,
+      "loss": 0.2177,
+      "step": 407
+    },
+    {
+      "epoch": 1.7539041464728056,
+      "grad_norm": 0.9521034359931946,
+      "learning_rate": 9.256109861020213e-06,
+      "loss": 0.2181,
+      "step": 408
+    },
+    {
+      "epoch": 1.758212170166936,
+      "grad_norm": 0.7550384402275085,
+      "learning_rate": 9.252552670871232e-06,
+      "loss": 0.2262,
+      "step": 409
+    },
+    {
+      "epoch": 1.7625201938610662,
+      "grad_norm": 0.6783749461174011,
+      "learning_rate": 9.248987682898576e-06,
+      "loss": 0.2242,
+      "step": 410
+    },
+    {
+      "epoch": 1.7668282175551966,
+      "grad_norm": 0.7524251937866211,
+      "learning_rate": 9.245414903639295e-06,
+      "loss": 0.2175,
+      "step": 411
+    },
+    {
+      "epoch": 1.771136241249327,
+      "grad_norm": 0.7920904755592346,
+      "learning_rate": 9.241834339644726e-06,
+      "loss": 0.2431,
+      "step": 412
+    },
+    {
+      "epoch": 1.7754442649434572,
+      "grad_norm": 0.7145002484321594,
+      "learning_rate": 9.23824599748048e-06,
+      "loss": 0.2123,
+      "step": 413
+    },
+    {
+      "epoch": 1.7797522886375874,
+      "grad_norm": 0.8676689863204956,
+      "learning_rate": 9.234649883726432e-06,
+      "loss": 0.245,
+      "step": 414
+    },
+    {
+      "epoch": 1.7840603123317178,
+      "grad_norm": 0.7974684834480286,
+      "learning_rate": 9.231046004976704e-06,
+      "loss": 0.2204,
+      "step": 415
+    },
+    {
+      "epoch": 1.7883683360258482,
+      "grad_norm": 0.6935186386108398,
+      "learning_rate": 9.22743436783966e-06,
+      "loss": 0.2284,
+      "step": 416
+    },
+    {
+      "epoch": 1.7926763597199784,
+      "grad_norm": 0.8282122611999512,
+      "learning_rate": 9.223814978937888e-06,
+      "loss": 0.2542,
+      "step": 417
+    },
+    {
+      "epoch": 1.7969843834141088,
+      "grad_norm": 0.7227144837379456,
+      "learning_rate": 9.220187844908194e-06,
+      "loss": 0.2142,
+      "step": 418
+    },
+    {
+      "epoch": 1.8012924071082392,
+      "grad_norm": 0.74013352394104,
+      "learning_rate": 9.216552972401582e-06,
+      "loss": 0.2266,
+      "step": 419
+    },
+    {
+      "epoch": 1.8056004308023694,
+      "grad_norm": 0.6855618953704834,
+      "learning_rate": 9.212910368083246e-06,
+      "loss": 0.2192,
+      "step": 420
+    },
+    {
+      "epoch": 1.8099084544964996,
+      "grad_norm": 0.7998465895652771,
+      "learning_rate": 9.209260038632562e-06,
+      "loss": 0.2172,
+      "step": 421
+    },
+    {
+      "epoch": 1.81421647819063,
+      "grad_norm": 0.798219621181488,
+      "learning_rate": 9.205601990743068e-06,
+      "loss": 0.236,
+      "step": 422
+    },
+    {
+      "epoch": 1.8185245018847604,
+      "grad_norm": 0.7570368051528931,
+      "learning_rate": 9.201936231122453e-06,
+      "loss": 0.2229,
+      "step": 423
+    },
+    {
+      "epoch": 1.8228325255788906,
+      "grad_norm": 0.7358865141868591,
+      "learning_rate": 9.198262766492554e-06,
+      "loss": 0.2302,
+      "step": 424
+    },
+    {
+      "epoch": 1.827140549273021,
+      "grad_norm": 0.7285174131393433,
+      "learning_rate": 9.194581603589327e-06,
+      "loss": 0.2312,
+      "step": 425
+    },
+    {
+      "epoch": 1.8314485729671515,
+      "grad_norm": 0.6818633675575256,
+      "learning_rate": 9.190892749162854e-06,
+      "loss": 0.2128,
+      "step": 426
+    },
+    {
+      "epoch": 1.8357565966612817,
+      "grad_norm": 0.7215419411659241,
+      "learning_rate": 9.187196209977314e-06,
+      "loss": 0.2142,
+      "step": 427
+    },
+    {
+      "epoch": 1.8400646203554119,
+      "grad_norm": 0.7318853139877319,
+      "learning_rate": 9.18349199281098e-06,
+      "loss": 0.2146,
+      "step": 428
+    },
+    {
+      "epoch": 1.8443726440495423,
+      "grad_norm": 0.7606166005134583,
+      "learning_rate": 9.179780104456205e-06,
+      "loss": 0.2099,
+      "step": 429
+    },
+    {
+      "epoch": 1.8486806677436727,
+      "grad_norm": 0.7998718619346619,
+      "learning_rate": 9.176060551719402e-06,
+      "loss": 0.2169,
+      "step": 430
+    },
+    {
+      "epoch": 1.8529886914378029,
+      "grad_norm": 0.788635790348053,
+      "learning_rate": 9.172333341421046e-06,
+      "loss": 0.2084,
+      "step": 431
+    },
+    {
+      "epoch": 1.857296715131933,
+      "grad_norm": 0.7982397675514221,
+      "learning_rate": 9.168598480395653e-06,
+      "loss": 0.2275,
+      "step": 432
+    },
+    {
+      "epoch": 1.8616047388260637,
+      "grad_norm": 0.783639132976532,
+      "learning_rate": 9.16485597549176e-06,
+      "loss": 0.2255,
+      "step": 433
+    },
+    {
+      "epoch": 1.865912762520194,
+      "grad_norm": 0.683904767036438,
+      "learning_rate": 9.16110583357193e-06,
+      "loss": 0.2044,
+      "step": 434
+    },
+    {
+      "epoch": 1.870220786214324,
+      "grad_norm": 0.7352596521377563,
+      "learning_rate": 9.157348061512728e-06,
+      "loss": 0.211,
+      "step": 435
+    },
+    {
+      "epoch": 1.8745288099084545,
+      "grad_norm": 0.7592405080795288,
+      "learning_rate": 9.153582666204702e-06,
+      "loss": 0.2208,
+      "step": 436
+    },
+    {
+      "epoch": 1.878836833602585,
+      "grad_norm": 0.8083028793334961,
+      "learning_rate": 9.149809654552387e-06,
+      "loss": 0.2278,
+      "step": 437
+    },
+    {
+      "epoch": 1.883144857296715,
+      "grad_norm": 0.918319046497345,
+      "learning_rate": 9.146029033474284e-06,
+      "loss": 0.2447,
+      "step": 438
+    },
+    {
+      "epoch": 1.8874528809908453,
+      "grad_norm": 0.8534688949584961,
+      "learning_rate": 9.142240809902841e-06,
+      "loss": 0.2135,
+      "step": 439
+    },
+    {
+      "epoch": 1.891760904684976,
+      "grad_norm": 0.6360190510749817,
+      "learning_rate": 9.138444990784455e-06,
+      "loss": 0.2143,
+      "step": 440
+    },
+    {
+      "epoch": 1.8960689283791061,
+      "grad_norm": 0.7315630316734314,
+      "learning_rate": 9.13464158307944e-06,
+      "loss": 0.2249,
+      "step": 441
+    },
+    {
+      "epoch": 1.9003769520732363,
+      "grad_norm": 0.7185893058776855,
+      "learning_rate": 9.130830593762037e-06,
+      "loss": 0.2219,
+      "step": 442
+    },
+    {
+      "epoch": 1.9046849757673667,
+      "grad_norm": 0.7180191874504089,
+      "learning_rate": 9.12701202982038e-06,
+      "loss": 0.2262,
+      "step": 443
+    },
+    {
+      "epoch": 1.9089929994614971,
+      "grad_norm": 0.844063937664032,
+      "learning_rate": 9.123185898256497e-06,
+      "loss": 0.2342,
+      "step": 444
+    },
+    {
+      "epoch": 1.9133010231556273,
+      "grad_norm": 0.7702510356903076,
+      "learning_rate": 9.119352206086292e-06,
+      "loss": 0.2277,
+      "step": 445
+    },
+    {
+      "epoch": 1.9176090468497575,
+      "grad_norm": 0.7082570791244507,
+      "learning_rate": 9.115510960339533e-06,
+      "loss": 0.2195,
+      "step": 446
+    },
+    {
+      "epoch": 1.921917070543888,
+      "grad_norm": 0.7621147632598877,
+      "learning_rate": 9.111662168059836e-06,
+      "loss": 0.2167,
+      "step": 447
+    },
+    {
+      "epoch": 1.9262250942380184,
+      "grad_norm": 0.7724840641021729,
+      "learning_rate": 9.107805836304658e-06,
+      "loss": 0.2276,
+      "step": 448
+    },
+    {
+      "epoch": 1.9305331179321485,
+      "grad_norm": 0.8185726404190063,
+      "learning_rate": 9.10394197214528e-06,
+      "loss": 0.227,
+      "step": 449
+    },
+    {
+      "epoch": 1.934841141626279,
+      "grad_norm": 0.8165621757507324,
+      "learning_rate": 9.100070582666796e-06,
+      "loss": 0.2159,
+      "step": 450
+    },
+    {
+      "epoch": 1.9391491653204094,
+      "grad_norm": 0.6630677580833435,
+      "learning_rate": 9.096191674968095e-06,
+      "loss": 0.2025,
+      "step": 451
+    },
+    {
+      "epoch": 1.9434571890145396,
+      "grad_norm": 0.8972369432449341,
+      "learning_rate": 9.09230525616186e-06,
+      "loss": 0.2288,
+      "step": 452
+    },
+    {
+      "epoch": 1.9477652127086698,
+      "grad_norm": 0.6295329332351685,
+      "learning_rate": 9.088411333374539e-06,
+      "loss": 0.1988,
+      "step": 453
+    },
+    {
+      "epoch": 1.9520732364028002,
+      "grad_norm": 0.8167666792869568,
+      "learning_rate": 9.084509913746342e-06,
+      "loss": 0.2316,
+      "step": 454
+    },
+    {
+      "epoch": 1.9563812600969306,
+      "grad_norm": 0.9114188551902771,
+      "learning_rate": 9.08060100443123e-06,
+      "loss": 0.2467,
+      "step": 455
+    },
+    {
+      "epoch": 1.9606892837910608,
+      "grad_norm": 0.6885513663291931,
+      "learning_rate": 9.076684612596891e-06,
+      "loss": 0.2033,
+      "step": 456
+    },
+    {
+      "epoch": 1.9649973074851912,
+      "grad_norm": 0.8051403760910034,
+      "learning_rate": 9.07276074542474e-06,
+      "loss": 0.2264,
+      "step": 457
+    },
+    {
+      "epoch": 1.9693053311793216,
+      "grad_norm": 0.7468147277832031,
+      "learning_rate": 9.068829410109893e-06,
+      "loss": 0.2166,
+      "step": 458
+    },
+    {
+      "epoch": 1.9736133548734518,
+      "grad_norm": 0.7809402346611023,
+      "learning_rate": 9.064890613861168e-06,
+      "loss": 0.2222,
+      "step": 459
+    },
+    {
+      "epoch": 1.977921378567582,
+      "grad_norm": 0.8526117205619812,
+      "learning_rate": 9.060944363901057e-06,
+      "loss": 0.2437,
+      "step": 460
+    },
+    {
+      "epoch": 1.9822294022617124,
+      "grad_norm": 0.7447945475578308,
+      "learning_rate": 9.05699066746572e-06,
+      "loss": 0.2191,
+      "step": 461
+    },
+    {
+      "epoch": 1.9865374259558428,
+      "grad_norm": 0.6959477663040161,
+      "learning_rate": 9.05302953180498e-06,
+      "loss": 0.2098,
+      "step": 462
+    },
+    {
+      "epoch": 1.990845449649973,
+      "grad_norm": 0.9427213668823242,
+      "learning_rate": 9.04906096418229e-06,
+      "loss": 0.2205,
+      "step": 463
+    },
+    {
+      "epoch": 1.9951534733441034,
+      "grad_norm": 0.8688002228736877,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.2068,
+      "step": 464
+    },
+    {
+      "epoch": 1.9994614970382338,
+      "grad_norm": 0.8308051824569702,
+      "learning_rate": 9.041101562173023e-06,
+      "loss": 0.2221,
+      "step": 465
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8308051824569702,
+      "learning_rate": 9.037110742381445e-06,
+      "loss": 0.2036,
+      "step": 466
+    },
+    {
+      "epoch": 2.00430802369413,
+      "grad_norm": 1.945624828338623,
+      "learning_rate": 9.033112519817897e-06,
+      "loss": 0.2128,
+      "step": 467
+    },
+    {
+      "epoch": 2.008616047388261,
+      "grad_norm": 0.7928284406661987,
+      "learning_rate": 9.02910690181384e-06,
+      "loss": 0.1988,
+      "step": 468
+    },
+    {
+      "epoch": 2.012924071082391,
+      "grad_norm": 0.7106955647468567,
+      "learning_rate": 9.0250938957143e-06,
+      "loss": 0.2005,
+      "step": 469
+    },
+    {
+      "epoch": 2.017232094776521,
+      "grad_norm": 0.7833051085472107,
+      "learning_rate": 9.021073508877845e-06,
+      "loss": 0.1958,
+      "step": 470
+    },
+    {
+      "epoch": 2.0215401184706514,
+      "grad_norm": 0.7723597288131714,
+      "learning_rate": 9.017045748676584e-06,
+      "loss": 0.1849,
+      "step": 471
+    },
+    {
+      "epoch": 2.025848142164782,
+      "grad_norm": 0.7105690836906433,
+      "learning_rate": 9.013010622496145e-06,
+      "loss": 0.2006,
+      "step": 472
+    },
+    {
+      "epoch": 2.0301561658589122,
+      "grad_norm": 0.7699000239372253,
+      "learning_rate": 9.008968137735655e-06,
+      "loss": 0.1946,
+      "step": 473
+    },
+    {
+      "epoch": 2.0344641895530424,
+      "grad_norm": 0.7681946158409119,
+      "learning_rate": 9.004918301807746e-06,
+      "loss": 0.1967,
+      "step": 474
+    },
+    {
+      "epoch": 2.038772213247173,
+      "grad_norm": 0.7109311819076538,
+      "learning_rate": 9.000861122138518e-06,
+      "loss": 0.1941,
+      "step": 475
+    },
+    {
+      "epoch": 2.0430802369413033,
+      "grad_norm": 0.7828920483589172,
+      "learning_rate": 8.996796606167549e-06,
+      "loss": 0.1931,
+      "step": 476
+    },
+    {
+      "epoch": 2.0473882606354334,
+      "grad_norm": 0.7319113612174988,
+      "learning_rate": 8.99272476134786e-06,
+      "loss": 0.2126,
+      "step": 477
+    },
+    {
+      "epoch": 2.0516962843295636,
+      "grad_norm": 0.7852508425712585,
+      "learning_rate": 8.988645595145913e-06,
+      "loss": 0.2024,
+      "step": 478
+    },
+    {
+      "epoch": 2.0560043080236943,
+      "grad_norm": 0.8128734827041626,
+      "learning_rate": 8.9845591150416e-06,
+      "loss": 0.2046,
+      "step": 479
+    },
+    {
+      "epoch": 2.0603123317178245,
+      "grad_norm": 0.8232097029685974,
+      "learning_rate": 8.98046532852822e-06,
+      "loss": 0.2083,
+      "step": 480
+    },
+    {
+      "epoch": 2.0646203554119547,
+      "grad_norm": 0.7391591668128967,
+      "learning_rate": 8.976364243112468e-06,
+      "loss": 0.1971,
+      "step": 481
+    },
+    {
+      "epoch": 2.0689283791060853,
+      "grad_norm": 0.7096617221832275,
+      "learning_rate": 8.972255866314425e-06,
+      "loss": 0.1974,
+      "step": 482
+    },
+    {
+      "epoch": 2.0732364028002155,
+      "grad_norm": 0.7919265627861023,
+      "learning_rate": 8.968140205667544e-06,
+      "loss": 0.1944,
+      "step": 483
+    },
+    {
+      "epoch": 2.0775444264943457,
+      "grad_norm": 0.7866784930229187,
+      "learning_rate": 8.964017268718632e-06,
+      "loss": 0.205,
+      "step": 484
+    },
+    {
+      "epoch": 2.081852450188476,
+      "grad_norm": 0.7036305069923401,
+      "learning_rate": 8.959887063027837e-06,
+      "loss": 0.1952,
+      "step": 485
+    },
+    {
+      "epoch": 2.0861604738826065,
+      "grad_norm": 0.7657865881919861,
+      "learning_rate": 8.95574959616864e-06,
+      "loss": 0.215,
+      "step": 486
+    },
+    {
+      "epoch": 2.0904684975767367,
+      "grad_norm": 0.7114456295967102,
+      "learning_rate": 8.951604875727833e-06,
+      "loss": 0.1968,
+      "step": 487
+    },
+    {
+      "epoch": 2.094776521270867,
+      "grad_norm": 0.7189726829528809,
+      "learning_rate": 8.94745290930551e-06,
+      "loss": 0.1929,
+      "step": 488
+    },
+    {
+      "epoch": 2.099084544964997,
+      "grad_norm": 0.7715457677841187,
+      "learning_rate": 8.94329370451505e-06,
+      "loss": 0.2049,
+      "step": 489
+    },
+    {
+      "epoch": 2.1033925686591277,
+      "grad_norm": 0.6534609198570251,
+      "learning_rate": 8.93912726898311e-06,
+      "loss": 0.1926,
+      "step": 490
+    },
+    {
+      "epoch": 2.107700592353258,
+      "grad_norm": 0.742912769317627,
+      "learning_rate": 8.934953610349599e-06,
+      "loss": 0.1961,
+      "step": 491
+    },
+    {
+      "epoch": 2.112008616047388,
+      "grad_norm": 0.6816502213478088,
+      "learning_rate": 8.930772736267675e-06,
+      "loss": 0.1885,
+      "step": 492
+    },
+    {
+      "epoch": 2.1163166397415187,
+      "grad_norm": 0.7203125357627869,
+      "learning_rate": 8.926584654403725e-06,
+      "loss": 0.199,
+      "step": 493
+    },
+    {
+      "epoch": 2.120624663435649,
+      "grad_norm": 0.714485228061676,
+      "learning_rate": 8.922389372437357e-06,
+      "loss": 0.2069,
+      "step": 494
+    },
+    {
+      "epoch": 2.124932687129779,
+      "grad_norm": 0.8489207029342651,
+      "learning_rate": 8.918186898061377e-06,
+      "loss": 0.1975,
+      "step": 495
+    },
+    {
+      "epoch": 2.1292407108239093,
+      "grad_norm": 0.7291126251220703,
+      "learning_rate": 8.91397723898178e-06,
+      "loss": 0.1973,
+      "step": 496
+    },
+    {
+      "epoch": 2.13354873451804,
+      "grad_norm": 0.7790775895118713,
+      "learning_rate": 8.909760402917738e-06,
+      "loss": 0.2002,
+      "step": 497
+    },
+    {
+      "epoch": 2.13785675821217,
+      "grad_norm": 0.6789630651473999,
+      "learning_rate": 8.90553639760158e-06,
+      "loss": 0.1825,
+      "step": 498
+    },
+    {
+      "epoch": 2.1421647819063003,
+      "grad_norm": 0.6676609516143799,
+      "learning_rate": 8.901305230778783e-06,
+      "loss": 0.1912,
+      "step": 499
+    },
+    {
+      "epoch": 2.146472805600431,
+      "grad_norm": 0.7658909559249878,
+      "learning_rate": 8.897066910207958e-06,
+      "loss": 0.1906,
+      "step": 500
+    },
+    {
+      "epoch": 2.150780829294561,
+      "grad_norm": 0.7476828694343567,
+      "learning_rate": 8.892821443660831e-06,
+      "loss": 0.1889,
+      "step": 501
+    },
+    {
+      "epoch": 2.1550888529886914,
+      "grad_norm": 0.7970424294471741,
+      "learning_rate": 8.888568838922231e-06,
+      "loss": 0.2097,
+      "step": 502
+    },
+    {
+      "epoch": 2.1593968766828215,
+      "grad_norm": 0.775412380695343,
+      "learning_rate": 8.884309103790078e-06,
+      "loss": 0.1901,
+      "step": 503
+    },
+    {
+      "epoch": 2.163704900376952,
+      "grad_norm": 0.822077214717865,
+      "learning_rate": 8.880042246075366e-06,
+      "loss": 0.2022,
+      "step": 504
+    },
+    {
+      "epoch": 2.1680129240710824,
+      "grad_norm": 0.7652409672737122,
+      "learning_rate": 8.875768273602148e-06,
+      "loss": 0.2038,
+      "step": 505
+    },
+    {
+      "epoch": 2.1723209477652126,
+      "grad_norm": 0.6706952452659607,
+      "learning_rate": 8.871487194207527e-06,
+      "loss": 0.1898,
+      "step": 506
+    },
+    {
+      "epoch": 2.176628971459343,
+      "grad_norm": 0.7346447706222534,
+      "learning_rate": 8.867199015741632e-06,
+      "loss": 0.1916,
+      "step": 507
+    },
+    {
+      "epoch": 2.1809369951534734,
+      "grad_norm": 0.8773221969604492,
+      "learning_rate": 8.862903746067619e-06,
+      "loss": 0.1995,
+      "step": 508
+    },
+    {
+      "epoch": 2.1852450188476036,
+      "grad_norm": 0.8744363784790039,
+      "learning_rate": 8.858601393061634e-06,
+      "loss": 0.2144,
+      "step": 509
+    },
+    {
+      "epoch": 2.1895530425417338,
+      "grad_norm": 0.7618948817253113,
+      "learning_rate": 8.854291964612824e-06,
+      "loss": 0.1927,
+      "step": 510
+    },
+    {
+      "epoch": 2.1938610662358644,
+      "grad_norm": 0.8064348697662354,
+      "learning_rate": 8.849975468623302e-06,
+      "loss": 0.2018,
+      "step": 511
+    },
+    {
+      "epoch": 2.1981690899299946,
+      "grad_norm": 0.7798653244972229,
+      "learning_rate": 8.845651913008145e-06,
+      "loss": 0.1968,
+      "step": 512
+    },
+    {
+      "epoch": 2.202477113624125,
+      "grad_norm": 0.7638295292854309,
+      "learning_rate": 8.841321305695372e-06,
+      "loss": 0.2006,
+      "step": 513
+    },
+    {
+      "epoch": 2.2067851373182554,
+      "grad_norm": 0.7909740805625916,
+      "learning_rate": 8.836983654625934e-06,
+      "loss": 0.2074,
+      "step": 514
+    },
+    {
+      "epoch": 2.2110931610123856,
+      "grad_norm": 0.7894330620765686,
+      "learning_rate": 8.832638967753699e-06,
+      "loss": 0.1924,
+      "step": 515
+    },
+    {
+      "epoch": 2.215401184706516,
+      "grad_norm": 0.7473629117012024,
+      "learning_rate": 8.828287253045436e-06,
+      "loss": 0.2004,
+      "step": 516
+    },
+    {
+      "epoch": 2.219709208400646,
+      "grad_norm": 0.666713535785675,
+      "learning_rate": 8.823928518480797e-06,
+      "loss": 0.1855,
+      "step": 517
+    },
+    {
+      "epoch": 2.2240172320947766,
+      "grad_norm": 0.7174916863441467,
+      "learning_rate": 8.819562772052312e-06,
+      "loss": 0.1951,
+      "step": 518
+    },
+    {
+      "epoch": 2.228325255788907,
+      "grad_norm": 0.8827500343322754,
+      "learning_rate": 8.815190021765365e-06,
+      "loss": 0.1914,
+      "step": 519
+    },
+    {
+      "epoch": 2.232633279483037,
+      "grad_norm": 0.7642231583595276,
+      "learning_rate": 8.810810275638183e-06,
+      "loss": 0.2155,
+      "step": 520
+    },
+    {
+      "epoch": 2.2369413031771677,
+      "grad_norm": 0.8573631644248962,
+      "learning_rate": 8.806423541701824e-06,
+      "loss": 0.2138,
+      "step": 521
+    },
+    {
+      "epoch": 2.241249326871298,
+      "grad_norm": 0.8140626549720764,
+      "learning_rate": 8.802029828000157e-06,
+      "loss": 0.2121,
+      "step": 522
+    },
+    {
+      "epoch": 2.245557350565428,
+      "grad_norm": 0.8195950388908386,
+      "learning_rate": 8.797629142589846e-06,
+      "loss": 0.2035,
+      "step": 523
+    },
+    {
+      "epoch": 2.2498653742595582,
+      "grad_norm": 0.8607699871063232,
+      "learning_rate": 8.793221493540347e-06,
+      "loss": 0.2148,
+      "step": 524
+    },
+    {
+      "epoch": 2.254173397953689,
+      "grad_norm": 0.8024702072143555,
+      "learning_rate": 8.788806888933881e-06,
+      "loss": 0.1999,
+      "step": 525
+    },
+    {
+      "epoch": 2.258481421647819,
+      "grad_norm": 0.7508124113082886,
+      "learning_rate": 8.784385336865419e-06,
+      "loss": 0.2015,
+      "step": 526
+    },
+    {
+      "epoch": 2.2627894453419493,
+      "grad_norm": 0.6819286942481995,
+      "learning_rate": 8.779956845442682e-06,
+      "loss": 0.2036,
+      "step": 527
+    },
+    {
+      "epoch": 2.26709746903608,
+      "grad_norm": 0.7056980729103088,
+      "learning_rate": 8.775521422786104e-06,
+      "loss": 0.1968,
+      "step": 528
+    },
+    {
+      "epoch": 2.27140549273021,
+      "grad_norm": 0.6826128959655762,
+      "learning_rate": 8.771079077028836e-06,
+      "loss": 0.1974,
+      "step": 529
+    },
+    {
+      "epoch": 2.2757135164243403,
+      "grad_norm": 0.7513538002967834,
+      "learning_rate": 8.766629816316722e-06,
+      "loss": 0.1985,
+      "step": 530
+    },
+    {
+      "epoch": 2.2800215401184705,
+      "grad_norm": 0.720900297164917,
+      "learning_rate": 8.762173648808283e-06,
+      "loss": 0.2054,
+      "step": 531
+    },
+    {
+      "epoch": 2.284329563812601,
+      "grad_norm": 0.90285325050354,
+      "learning_rate": 8.757710582674708e-06,
+      "loss": 0.2161,
+      "step": 532
+    },
+    {
+      "epoch": 2.2886375875067313,
+      "grad_norm": 0.877210259437561,
+      "learning_rate": 8.753240626099836e-06,
+      "loss": 0.2011,
+      "step": 533
+    },
+    {
+      "epoch": 2.2929456112008615,
+      "grad_norm": 0.7395567893981934,
+      "learning_rate": 8.748763787280142e-06,
+      "loss": 0.2003,
+      "step": 534
+    },
+    {
+      "epoch": 2.297253634894992,
+      "grad_norm": 0.7593705058097839,
+      "learning_rate": 8.744280074424713e-06,
+      "loss": 0.2002,
+      "step": 535
+    },
+    {
+      "epoch": 2.3015616585891223,
+      "grad_norm": 0.7474355697631836,
+      "learning_rate": 8.739789495755254e-06,
+      "loss": 0.1963,
+      "step": 536
+    },
+    {
+      "epoch": 2.3058696822832525,
+      "grad_norm": 0.6919746398925781,
+      "learning_rate": 8.735292059506047e-06,
+      "loss": 0.1955,
+      "step": 537
+    },
+    {
+      "epoch": 2.3101777059773827,
+      "grad_norm": 0.7640472650527954,
+      "learning_rate": 8.730787773923957e-06,
+      "loss": 0.2001,
+      "step": 538
+    },
+    {
+      "epoch": 2.3144857296715133,
+      "grad_norm": 0.6979402899742126,
+      "learning_rate": 8.726276647268403e-06,
+      "loss": 0.1985,
+      "step": 539
+    },
+    {
+      "epoch": 2.3187937533656435,
+      "grad_norm": 0.7744669914245605,
+      "learning_rate": 8.721758687811353e-06,
+      "loss": 0.206,
+      "step": 540
+    },
+    {
+      "epoch": 2.3231017770597737,
+      "grad_norm": 0.839337944984436,
+      "learning_rate": 8.717233903837298e-06,
+      "loss": 0.2062,
+      "step": 541
+    },
+    {
+      "epoch": 2.3274098007539044,
+      "grad_norm": 0.7544538974761963,
+      "learning_rate": 8.712702303643254e-06,
+      "loss": 0.2108,
+      "step": 542
+    },
+    {
+      "epoch": 2.3317178244480345,
+      "grad_norm": 0.7645125389099121,
+      "learning_rate": 8.708163895538722e-06,
+      "loss": 0.1971,
+      "step": 543
+    },
+    {
+      "epoch": 2.3360258481421647,
+      "grad_norm": 0.8445392847061157,
+      "learning_rate": 8.703618687845697e-06,
+      "loss": 0.1967,
+      "step": 544
+    },
+    {
+      "epoch": 2.340333871836295,
+      "grad_norm": 0.8427358865737915,
+      "learning_rate": 8.699066688898636e-06,
+      "loss": 0.216,
+      "step": 545
+    },
+    {
+      "epoch": 2.3446418955304256,
+      "grad_norm": 0.7722771763801575,
+      "learning_rate": 8.694507907044454e-06,
+      "loss": 0.1975,
+      "step": 546
+    },
+    {
+      "epoch": 2.3489499192245558,
+      "grad_norm": 0.689138650894165,
+      "learning_rate": 8.6899423506425e-06,
+      "loss": 0.2049,
+      "step": 547
+    },
+    {
+      "epoch": 2.353257942918686,
+      "grad_norm": 0.7982721328735352,
+      "learning_rate": 8.685370028064546e-06,
+      "loss": 0.1983,
+      "step": 548
+    },
+    {
+      "epoch": 2.3575659666128166,
+      "grad_norm": 0.8092723488807678,
+      "learning_rate": 8.680790947694772e-06,
+      "loss": 0.2092,
+      "step": 549
+    },
+    {
+      "epoch": 2.361873990306947,
+      "grad_norm": 0.7706890106201172,
+      "learning_rate": 8.676205117929752e-06,
+      "loss": 0.2121,
+      "step": 550
+    },
+    {
+      "epoch": 2.366182014001077,
+      "grad_norm": 0.8860268592834473,
+      "learning_rate": 8.671612547178428e-06,
+      "loss": 0.2071,
+      "step": 551
+    },
+    {
+      "epoch": 2.370490037695207,
+      "grad_norm": 0.8444654941558838,
+      "learning_rate": 8.667013243862113e-06,
+      "loss": 0.1887,
+      "step": 552
+    },
+    {
+      "epoch": 2.374798061389338,
+      "grad_norm": 0.6994353532791138,
+      "learning_rate": 8.66240721641446e-06,
+      "loss": 0.1928,
+      "step": 553
+    },
+    {
+      "epoch": 2.379106085083468,
+      "grad_norm": 0.7686805725097656,
+      "learning_rate": 8.657794473281447e-06,
+      "loss": 0.1974,
+      "step": 554
+    },
+    {
+      "epoch": 2.383414108777598,
+      "grad_norm": 0.6858060956001282,
+      "learning_rate": 8.65317502292138e-06,
+      "loss": 0.1914,
+      "step": 555
+    },
+    {
+      "epoch": 2.387722132471729,
+      "grad_norm": 0.7314836382865906,
+      "learning_rate": 8.64854887380485e-06,
+      "loss": 0.2014,
+      "step": 556
+    },
+    {
+      "epoch": 2.392030156165859,
+      "grad_norm": 0.7626047134399414,
+      "learning_rate": 8.643916034414741e-06,
+      "loss": 0.1925,
+      "step": 557
+    },
+    {
+      "epoch": 2.396338179859989,
+      "grad_norm": 0.6878098845481873,
+      "learning_rate": 8.639276513246199e-06,
+      "loss": 0.1925,
+      "step": 558
+    },
+    {
+      "epoch": 2.4006462035541194,
+      "grad_norm": 0.8436222076416016,
+      "learning_rate": 8.634630318806626e-06,
+      "loss": 0.217,
+      "step": 559
+    },
+    {
+      "epoch": 2.40495422724825,
+      "grad_norm": 0.7351590394973755,
+      "learning_rate": 8.629977459615655e-06,
+      "loss": 0.2068,
+      "step": 560
+    },
+    {
+      "epoch": 2.4092622509423802,
+      "grad_norm": 0.7552523612976074,
+      "learning_rate": 8.62531794420515e-06,
+      "loss": 0.2003,
+      "step": 561
+    },
+    {
+      "epoch": 2.4135702746365104,
+      "grad_norm": 0.7772282361984253,
+      "learning_rate": 8.620651781119169e-06,
+      "loss": 0.1978,
+      "step": 562
+    },
+    {
+      "epoch": 2.417878298330641,
+      "grad_norm": 0.8326259255409241,
+      "learning_rate": 8.615978978913968e-06,
+      "loss": 0.217,
+      "step": 563
+    },
+    {
+      "epoch": 2.4221863220247712,
+      "grad_norm": 0.8465754985809326,
+      "learning_rate": 8.611299546157973e-06,
+      "loss": 0.2032,
+      "step": 564
+    },
+    {
+      "epoch": 2.4264943457189014,
+      "grad_norm": 0.703048050403595,
+      "learning_rate": 8.60661349143177e-06,
+      "loss": 0.1963,
+      "step": 565
+    },
+    {
+      "epoch": 2.4308023694130316,
+      "grad_norm": 0.73744797706604,
+      "learning_rate": 8.601920823328088e-06,
+      "loss": 0.1893,
+      "step": 566
+    },
+    {
+      "epoch": 2.4351103931071623,
+      "grad_norm": 0.8171586990356445,
+      "learning_rate": 8.59722155045178e-06,
+      "loss": 0.198,
+      "step": 567
+    },
+    {
+      "epoch": 2.4394184168012925,
+      "grad_norm": 0.7568394541740417,
+      "learning_rate": 8.592515681419812e-06,
+      "loss": 0.1975,
+      "step": 568
+    },
+    {
+      "epoch": 2.4437264404954226,
+      "grad_norm": 0.874150812625885,
+      "learning_rate": 8.587803224861248e-06,
+      "loss": 0.2031,
+      "step": 569
+    },
+    {
+      "epoch": 2.4480344641895533,
+      "grad_norm": 0.7862420678138733,
+      "learning_rate": 8.583084189417225e-06,
+      "loss": 0.2001,
+      "step": 570
+    },
+    {
+      "epoch": 2.4523424878836835,
+      "grad_norm": 0.708115816116333,
+      "learning_rate": 8.578358583740947e-06,
+      "loss": 0.1989,
+      "step": 571
+    },
+    {
+      "epoch": 2.4566505115778137,
+      "grad_norm": 0.7508895397186279,
+      "learning_rate": 8.573626416497669e-06,
+      "loss": 0.1982,
+      "step": 572
+    },
+    {
+      "epoch": 2.460958535271944,
+      "grad_norm": 0.7255448698997498,
+      "learning_rate": 8.568887696364673e-06,
+      "loss": 0.1977,
+      "step": 573
+    },
+    {
+      "epoch": 2.4652665589660745,
+      "grad_norm": 0.819625198841095,
+      "learning_rate": 8.564142432031257e-06,
+      "loss": 0.1953,
+      "step": 574
+    },
+    {
+      "epoch": 2.4695745826602047,
+      "grad_norm": 0.6851021647453308,
+      "learning_rate": 8.559390632198723e-06,
+      "loss": 0.1925,
+      "step": 575
+    },
+    {
+      "epoch": 2.473882606354335,
+      "grad_norm": 0.8143554329872131,
+      "learning_rate": 8.554632305580355e-06,
+      "loss": 0.2123,
+      "step": 576
+    },
+    {
+      "epoch": 2.4781906300484655,
+      "grad_norm": 0.7244540452957153,
+      "learning_rate": 8.549867460901402e-06,
+      "loss": 0.1924,
+      "step": 577
+    },
+    {
+      "epoch": 2.4824986537425957,
+      "grad_norm": 0.7576718330383301,
+      "learning_rate": 8.545096106899068e-06,
+      "loss": 0.2173,
+      "step": 578
+    },
+    {
+      "epoch": 2.486806677436726,
+      "grad_norm": 0.7695874571800232,
+      "learning_rate": 8.540318252322493e-06,
+      "loss": 0.2073,
+      "step": 579
+    },
+    {
+      "epoch": 2.491114701130856,
+      "grad_norm": 0.6576676964759827,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.1936,
+      "step": 580
+    },
+    {
+      "epoch": 2.4954227248249867,
+      "grad_norm": 0.7426583766937256,
+      "learning_rate": 8.530743076502766e-06,
+      "loss": 0.1967,
+      "step": 581
+    },
+    {
+      "epoch": 2.499730748519117,
+      "grad_norm": 0.7785230278968811,
+      "learning_rate": 8.525945772817427e-06,
+      "loss": 0.2053,
+      "step": 582
+    },
+    {
+      "epoch": 2.504038772213247,
+      "grad_norm": 0.7582430243492126,
+      "learning_rate": 8.521142003673447e-06,
+      "loss": 0.1936,
+      "step": 583
+    },
+    {
+      "epoch": 2.5083467959073777,
+      "grad_norm": 0.7534447908401489,
+      "learning_rate": 8.5163317778794e-06,
+      "loss": 0.2121,
+      "step": 584
+    },
+    {
+      "epoch": 2.512654819601508,
+      "grad_norm": 0.7007425427436829,
+      "learning_rate": 8.51151510425571e-06,
+      "loss": 0.2067,
+      "step": 585
+    },
+    {
+      "epoch": 2.516962843295638,
+      "grad_norm": 0.7491743564605713,
+      "learning_rate": 8.506691991634612e-06,
+      "loss": 0.1993,
+      "step": 586
+    },
+    {
+      "epoch": 2.5212708669897683,
+      "grad_norm": 0.7492290139198303,
+      "learning_rate": 8.501862448860159e-06,
+      "loss": 0.2011,
+      "step": 587
+    },
+    {
+      "epoch": 2.5255788906838985,
+      "grad_norm": 0.7035954594612122,
+      "learning_rate": 8.497026484788189e-06,
+      "loss": 0.2085,
+      "step": 588
+    },
+    {
+      "epoch": 2.529886914378029,
+      "grad_norm": 0.6668585538864136,
+      "learning_rate": 8.492184108286316e-06,
+      "loss": 0.1962,
+      "step": 589
+    },
+    {
+      "epoch": 2.5341949380721593,
+      "grad_norm": 0.7750927209854126,
+      "learning_rate": 8.487335328233912e-06,
+      "loss": 0.2219,
+      "step": 590
+    },
+    {
+      "epoch": 2.53850296176629,
+      "grad_norm": 0.7335095405578613,
+      "learning_rate": 8.48248015352209e-06,
+      "loss": 0.1978,
+      "step": 591
+    },
+    {
+      "epoch": 2.54281098546042,
+      "grad_norm": 0.6468396782875061,
+      "learning_rate": 8.477618593053693e-06,
+      "loss": 0.1945,
+      "step": 592
+    },
+    {
+      "epoch": 2.5471190091545504,
+      "grad_norm": 0.7096503376960754,
+      "learning_rate": 8.47275065574327e-06,
+      "loss": 0.2109,
+      "step": 593
+    },
+    {
+      "epoch": 2.5514270328486806,
+      "grad_norm": 0.771790087223053,
+      "learning_rate": 8.46787635051706e-06,
+      "loss": 0.1988,
+      "step": 594
+    },
+    {
+      "epoch": 2.5557350565428107,
+      "grad_norm": 0.8515303134918213,
+      "learning_rate": 8.462995686312985e-06,
+      "loss": 0.1917,
+      "step": 595
+    },
+    {
+      "epoch": 2.5600430802369414,
+      "grad_norm": 0.7744242548942566,
+      "learning_rate": 8.458108672080624e-06,
+      "loss": 0.2075,
+      "step": 596
+    },
+    {
+      "epoch": 2.5643511039310716,
+      "grad_norm": 0.6713221669197083,
+      "learning_rate": 8.453215316781205e-06,
+      "loss": 0.1914,
+      "step": 597
+    },
+    {
+      "epoch": 2.568659127625202,
+      "grad_norm": 0.6832510232925415,
+      "learning_rate": 8.448315629387572e-06,
+      "loss": 0.2057,
+      "step": 598
+    },
+    {
+      "epoch": 2.5729671513193324,
+      "grad_norm": 0.8287082314491272,
+      "learning_rate": 8.44340961888419e-06,
+      "loss": 0.2012,
+      "step": 599
+    },
+    {
+      "epoch": 2.5772751750134626,
+      "grad_norm": 0.7368465662002563,
+      "learning_rate": 8.438497294267117e-06,
+      "loss": 0.2022,
+      "step": 600
+    },
+    {
+      "epoch": 2.581583198707593,
+      "grad_norm": 0.7697679996490479,
+      "learning_rate": 8.433578664543986e-06,
+      "loss": 0.2098,
+      "step": 601
+    },
+    {
+      "epoch": 2.585891222401723,
+      "grad_norm": 0.7776275277137756,
+      "learning_rate": 8.428653738733996e-06,
+      "loss": 0.1963,
+      "step": 602
+    },
+    {
+      "epoch": 2.5901992460958536,
+      "grad_norm": 0.7150554656982422,
+      "learning_rate": 8.423722525867883e-06,
+      "loss": 0.1917,
+      "step": 603
+    },
+    {
+      "epoch": 2.594507269789984,
+      "grad_norm": 0.67384934425354,
+      "learning_rate": 8.418785034987921e-06,
+      "loss": 0.1931,
+      "step": 604
+    },
+    {
+      "epoch": 2.598815293484114,
+      "grad_norm": 0.7748411297798157,
+      "learning_rate": 8.413841275147893e-06,
+      "loss": 0.1979,
+      "step": 605
+    },
+    {
+      "epoch": 2.6031233171782446,
+      "grad_norm": 0.7929915189743042,
+      "learning_rate": 8.408891255413072e-06,
+      "loss": 0.214,
+      "step": 606
+    },
+    {
+      "epoch": 2.607431340872375,
+      "grad_norm": 0.7815855145454407,
+      "learning_rate": 8.403934984860216e-06,
+      "loss": 0.2106,
+      "step": 607
+    },
+    {
+      "epoch": 2.611739364566505,
+      "grad_norm": 0.8815018534660339,
+      "learning_rate": 8.39897247257754e-06,
+      "loss": 0.2141,
+      "step": 608
+    },
+    {
+      "epoch": 2.616047388260635,
+      "grad_norm": 0.6736514568328857,
+      "learning_rate": 8.39400372766471e-06,
+      "loss": 0.204,
+      "step": 609
+    },
+    {
+      "epoch": 2.620355411954766,
+      "grad_norm": 0.6993145942687988,
+      "learning_rate": 8.389028759232816e-06,
+      "loss": 0.2135,
+      "step": 610
+    },
+    {
+      "epoch": 2.624663435648896,
+      "grad_norm": 0.8782504200935364,
+      "learning_rate": 8.38404757640436e-06,
+      "loss": 0.2129,
+      "step": 611
+    },
+    {
+      "epoch": 2.6289714593430262,
+      "grad_norm": 0.6805502772331238,
+      "learning_rate": 8.379060188313244e-06,
+      "loss": 0.2151,
+      "step": 612
+    },
+    {
+      "epoch": 2.633279483037157,
+      "grad_norm": 0.8480393886566162,
+      "learning_rate": 8.374066604104742e-06,
+      "loss": 0.2181,
+      "step": 613
+    },
+    {
+      "epoch": 2.637587506731287,
+      "grad_norm": 0.7601768970489502,
+      "learning_rate": 8.369066832935498e-06,
+      "loss": 0.2011,
+      "step": 614
+    },
+    {
+      "epoch": 2.6418955304254172,
+      "grad_norm": 0.6898226141929626,
+      "learning_rate": 8.364060883973488e-06,
+      "loss": 0.1925,
+      "step": 615
+    },
+    {
+      "epoch": 2.6462035541195474,
+      "grad_norm": 0.6958613991737366,
+      "learning_rate": 8.359048766398032e-06,
+      "loss": 0.1993,
+      "step": 616
+    },
+    {
+      "epoch": 2.650511577813678,
+      "grad_norm": 0.7246137261390686,
+      "learning_rate": 8.354030489399747e-06,
+      "loss": 0.1969,
+      "step": 617
+    },
+    {
+      "epoch": 2.6548196015078083,
+      "grad_norm": 0.7820032238960266,
+      "learning_rate": 8.349006062180552e-06,
+      "loss": 0.1914,
+      "step": 618
+    },
+    {
+      "epoch": 2.6591276252019385,
+      "grad_norm": 0.8184999823570251,
+      "learning_rate": 8.343975493953645e-06,
+      "loss": 0.1963,
+      "step": 619
+    },
+    {
+      "epoch": 2.663435648896069,
+      "grad_norm": 0.7577897906303406,
+      "learning_rate": 8.338938793943478e-06,
+      "loss": 0.2011,
+      "step": 620
+    },
+    {
+      "epoch": 2.6677436725901993,
+      "grad_norm": 0.7401598691940308,
+      "learning_rate": 8.333895971385754e-06,
+      "loss": 0.1919,
+      "step": 621
+    },
+    {
+      "epoch": 2.6720516962843295,
+      "grad_norm": 0.7432500720024109,
+      "learning_rate": 8.328847035527397e-06,
+      "loss": 0.2049,
+      "step": 622
+    },
+    {
+      "epoch": 2.6763597199784597,
+      "grad_norm": 0.7831881046295166,
+      "learning_rate": 8.323791995626543e-06,
+      "loss": 0.2062,
+      "step": 623
+    },
+    {
+      "epoch": 2.6806677436725903,
+      "grad_norm": 0.8239650130271912,
+      "learning_rate": 8.318730860952523e-06,
+      "loss": 0.2153,
+      "step": 624
+    },
+    {
+      "epoch": 2.6849757673667205,
+      "grad_norm": 0.7947770357131958,
+      "learning_rate": 8.313663640785839e-06,
+      "loss": 0.2047,
+      "step": 625
+    },
+    {
+      "epoch": 2.6892837910608507,
+      "grad_norm": 0.8108847141265869,
+      "learning_rate": 8.308590344418158e-06,
+      "loss": 0.1959,
+      "step": 626
+    },
+    {
+      "epoch": 2.6935918147549813,
+      "grad_norm": 0.7876724004745483,
+      "learning_rate": 8.303510981152283e-06,
+      "loss": 0.2138,
+      "step": 627
+    },
+    {
+      "epoch": 2.6978998384491115,
+      "grad_norm": 0.8588249683380127,
+      "learning_rate": 8.298425560302146e-06,
+      "loss": 0.2119,
+      "step": 628
+    },
+    {
+      "epoch": 2.7022078621432417,
+      "grad_norm": 0.7418671250343323,
+      "learning_rate": 8.293334091192782e-06,
+      "loss": 0.1929,
+      "step": 629
+    },
+    {
+      "epoch": 2.706515885837372,
+      "grad_norm": 0.6973020434379578,
+      "learning_rate": 8.288236583160322e-06,
+      "loss": 0.2001,
+      "step": 630
+    },
+    {
+      "epoch": 2.7108239095315025,
+      "grad_norm": 0.8735868334770203,
+      "learning_rate": 8.28313304555197e-06,
+      "loss": 0.1849,
+      "step": 631
+    },
+    {
+      "epoch": 2.7151319332256327,
+      "grad_norm": 0.700727641582489,
+      "learning_rate": 8.278023487725981e-06,
+      "loss": 0.1902,
+      "step": 632
+    },
+    {
+      "epoch": 2.719439956919763,
+      "grad_norm": 0.7866157293319702,
+      "learning_rate": 8.272907919051653e-06,
+      "loss": 0.2,
+      "step": 633
+    },
+    {
+      "epoch": 2.7237479806138936,
+      "grad_norm": 0.7389417290687561,
+      "learning_rate": 8.267786348909306e-06,
+      "loss": 0.2006,
+      "step": 634
+    },
+    {
+      "epoch": 2.7280560043080238,
+      "grad_norm": 0.7321817278862,
+      "learning_rate": 8.262658786690262e-06,
+      "loss": 0.2072,
+      "step": 635
+    },
+    {
+      "epoch": 2.732364028002154,
+      "grad_norm": 0.6979652047157288,
+      "learning_rate": 8.257525241796837e-06,
+      "loss": 0.1992,
+      "step": 636
+    },
+    {
+      "epoch": 2.736672051696284,
+      "grad_norm": 0.7699320316314697,
+      "learning_rate": 8.252385723642312e-06,
+      "loss": 0.1939,
+      "step": 637
+    },
+    {
+      "epoch": 2.7409800753904148,
+      "grad_norm": 0.7373313903808594,
+      "learning_rate": 8.247240241650918e-06,
+      "loss": 0.2051,
+      "step": 638
+    },
+    {
+      "epoch": 2.745288099084545,
+      "grad_norm": 0.6791808605194092,
+      "learning_rate": 8.242088805257832e-06,
+      "loss": 0.1934,
+      "step": 639
+    },
+    {
+      "epoch": 2.749596122778675,
+      "grad_norm": 0.6882132291793823,
+      "learning_rate": 8.23693142390914e-06,
+      "loss": 0.2032,
+      "step": 640
+    },
+    {
+      "epoch": 2.753904146472806,
+      "grad_norm": 0.7233914136886597,
+      "learning_rate": 8.231768107061831e-06,
+      "loss": 0.193,
+      "step": 641
+    },
+    {
+      "epoch": 2.758212170166936,
+      "grad_norm": 0.7256336212158203,
+      "learning_rate": 8.226598864183782e-06,
+      "loss": 0.2061,
+      "step": 642
+    },
+    {
+      "epoch": 2.762520193861066,
+      "grad_norm": 0.739739716053009,
+      "learning_rate": 8.221423704753733e-06,
+      "loss": 0.214,
+      "step": 643
+    },
+    {
+      "epoch": 2.7668282175551964,
+      "grad_norm": 0.6662392020225525,
+      "learning_rate": 8.216242638261277e-06,
+      "loss": 0.1992,
+      "step": 644
+    },
+    {
+      "epoch": 2.771136241249327,
+      "grad_norm": 0.7961679697036743,
+      "learning_rate": 8.211055674206828e-06,
+      "loss": 0.1898,
+      "step": 645
+    },
+    {
+      "epoch": 2.775444264943457,
+      "grad_norm": 0.7357142567634583,
+      "learning_rate": 8.205862822101628e-06,
+      "loss": 0.1988,
+      "step": 646
+    },
+    {
+      "epoch": 2.7797522886375874,
+      "grad_norm": 0.7799228429794312,
+      "learning_rate": 8.200664091467707e-06,
+      "loss": 0.2121,
+      "step": 647
+    },
+    {
+      "epoch": 2.784060312331718,
+      "grad_norm": 0.8995800018310547,
+      "learning_rate": 8.195459491837881e-06,
+      "loss": 0.2165,
+      "step": 648
+    },
+    {
+      "epoch": 2.788368336025848,
+      "grad_norm": 0.8075342178344727,
+      "learning_rate": 8.190249032755717e-06,
+      "loss": 0.2076,
+      "step": 649
+    },
+    {
+      "epoch": 2.7926763597199784,
+      "grad_norm": 0.8986786007881165,
+      "learning_rate": 8.18503272377554e-06,
+      "loss": 0.2093,
+      "step": 650
+    },
+    {
+      "epoch": 2.7969843834141086,
+      "grad_norm": 0.885172426700592,
+      "learning_rate": 8.179810574462388e-06,
+      "loss": 0.206,
+      "step": 651
+    },
+    {
+      "epoch": 2.8012924071082392,
+      "grad_norm": 0.9469455480575562,
+      "learning_rate": 8.17458259439202e-06,
+      "loss": 0.2077,
+      "step": 652
+    },
+    {
+      "epoch": 2.8056004308023694,
+      "grad_norm": 0.8134719729423523,
+      "learning_rate": 8.169348793150884e-06,
+      "loss": 0.2039,
+      "step": 653
+    },
+    {
+      "epoch": 2.8099084544964996,
+      "grad_norm": 0.7365888357162476,
+      "learning_rate": 8.164109180336094e-06,
+      "loss": 0.2014,
+      "step": 654
+    },
+    {
+      "epoch": 2.8142164781906303,
+      "grad_norm": 0.7723566293716431,
+      "learning_rate": 8.15886376555543e-06,
+      "loss": 0.192,
+      "step": 655
+    },
+    {
+      "epoch": 2.8185245018847604,
+      "grad_norm": 0.7469866275787354,
+      "learning_rate": 8.153612558427311e-06,
+      "loss": 0.208,
+      "step": 656
+    },
+    {
+      "epoch": 2.8228325255788906,
+      "grad_norm": 0.7777121067047119,
+      "learning_rate": 8.148355568580768e-06,
+      "loss": 0.1991,
+      "step": 657
+    },
+    {
+      "epoch": 2.827140549273021,
+      "grad_norm": 0.7026106119155884,
+      "learning_rate": 8.143092805655445e-06,
+      "loss": 0.1996,
+      "step": 658
+    },
+    {
+      "epoch": 2.8314485729671515,
+      "grad_norm": 0.778965413570404,
+      "learning_rate": 8.13782427930157e-06,
+      "loss": 0.2162,
+      "step": 659
+    },
+    {
+      "epoch": 2.8357565966612817,
+      "grad_norm": 0.8295383453369141,
+      "learning_rate": 8.132549999179934e-06,
+      "loss": 0.2013,
+      "step": 660
+    },
+    {
+      "epoch": 2.840064620355412,
+      "grad_norm": 0.7001696825027466,
+      "learning_rate": 8.127269974961886e-06,
+      "loss": 0.209,
+      "step": 661
+    },
+    {
+      "epoch": 2.8443726440495425,
+      "grad_norm": 0.6350052356719971,
+      "learning_rate": 8.121984216329303e-06,
+      "loss": 0.1946,
+      "step": 662
+    },
+    {
+      "epoch": 2.8486806677436727,
+      "grad_norm": 0.7462396621704102,
+      "learning_rate": 8.116692732974578e-06,
+      "loss": 0.2067,
+      "step": 663
+    },
+    {
+      "epoch": 2.852988691437803,
+      "grad_norm": 0.7569505572319031,
+      "learning_rate": 8.111395534600604e-06,
+      "loss": 0.1956,
+      "step": 664
+    },
+    {
+      "epoch": 2.857296715131933,
+      "grad_norm": 0.7309179902076721,
+      "learning_rate": 8.10609263092075e-06,
+      "loss": 0.2024,
+      "step": 665
+    },
+    {
+      "epoch": 2.8616047388260637,
+      "grad_norm": 0.6988533139228821,
+      "learning_rate": 8.100784031658846e-06,
+      "loss": 0.2191,
+      "step": 666
+    },
+    {
+      "epoch": 2.865912762520194,
+      "grad_norm": 0.7274155616760254,
+      "learning_rate": 8.095469746549172e-06,
+      "loss": 0.2057,
+      "step": 667
+    },
+    {
+      "epoch": 2.870220786214324,
+      "grad_norm": 0.7919628620147705,
+      "learning_rate": 8.090149785336426e-06,
+      "loss": 0.1939,
+      "step": 668
+    },
+    {
+      "epoch": 2.8745288099084547,
+      "grad_norm": 0.7325423955917358,
+      "learning_rate": 8.084824157775719e-06,
+      "loss": 0.189,
+      "step": 669
+    },
+    {
+      "epoch": 2.878836833602585,
+      "grad_norm": 0.8092797994613647,
+      "learning_rate": 8.079492873632554e-06,
+      "loss": 0.2127,
+      "step": 670
+    },
+    {
+      "epoch": 2.883144857296715,
+      "grad_norm": 0.6685104966163635,
+      "learning_rate": 8.074155942682803e-06,
+      "loss": 0.201,
+      "step": 671
+    },
+    {
+      "epoch": 2.8874528809908453,
+      "grad_norm": 0.7067998647689819,
+      "learning_rate": 8.068813374712689e-06,
+      "loss": 0.2036,
+      "step": 672
+    },
+    {
+      "epoch": 2.891760904684976,
+      "grad_norm": 0.7045204043388367,
+      "learning_rate": 8.06346517951878e-06,
+      "loss": 0.1982,
+      "step": 673
+    },
+    {
+      "epoch": 2.896068928379106,
+      "grad_norm": 0.6677000522613525,
+      "learning_rate": 8.058111366907957e-06,
+      "loss": 0.189,
+      "step": 674
+    },
+    {
+      "epoch": 2.9003769520732363,
+      "grad_norm": 0.8431633114814758,
+      "learning_rate": 8.052751946697403e-06,
+      "loss": 0.2113,
+      "step": 675
+    },
+    {
+      "epoch": 2.904684975767367,
+      "grad_norm": 0.781539797782898,
+      "learning_rate": 8.047386928714583e-06,
+      "loss": 0.2184,
+      "step": 676
+    },
+    {
+      "epoch": 2.908992999461497,
+      "grad_norm": 0.74564129114151,
+      "learning_rate": 8.042016322797227e-06,
+      "loss": 0.2001,
+      "step": 677
+    },
+    {
+      "epoch": 2.9133010231556273,
+      "grad_norm": 0.6591554880142212,
+      "learning_rate": 8.03664013879331e-06,
+      "loss": 0.1947,
+      "step": 678
+    },
+    {
+      "epoch": 2.9176090468497575,
+      "grad_norm": 0.7208926677703857,
+      "learning_rate": 8.031258386561038e-06,
+      "loss": 0.2105,
+      "step": 679
+    },
+    {
+      "epoch": 2.9219170705438877,
+      "grad_norm": 0.7640514373779297,
+      "learning_rate": 8.025871075968828e-06,
+      "loss": 0.2109,
+      "step": 680
+    },
+    {
+      "epoch": 2.9262250942380184,
+      "grad_norm": 0.7145681977272034,
+      "learning_rate": 8.020478216895282e-06,
+      "loss": 0.1874,
+      "step": 681
+    },
+    {
+      "epoch": 2.9305331179321485,
+      "grad_norm": 0.7085782885551453,
+      "learning_rate": 8.015079819229187e-06,
+      "loss": 0.2117,
+      "step": 682
+    },
+    {
+      "epoch": 2.934841141626279,
+      "grad_norm": 0.7184674739837646,
+      "learning_rate": 8.009675892869478e-06,
+      "loss": 0.2103,
+      "step": 683
+    },
+    {
+      "epoch": 2.9391491653204094,
+      "grad_norm": 0.726003885269165,
+      "learning_rate": 8.00426644772523e-06,
+      "loss": 0.2119,
+      "step": 684
+    },
+    {
+      "epoch": 2.9434571890145396,
+      "grad_norm": 0.8296722769737244,
+      "learning_rate": 7.99885149371564e-06,
+      "loss": 0.2048,
+      "step": 685
+    },
+    {
+      "epoch": 2.9477652127086698,
+      "grad_norm": 0.6646244525909424,
+      "learning_rate": 7.993431040770002e-06,
+      "loss": 0.2064,
+      "step": 686
+    },
+    {
+      "epoch": 2.9520732364028,
+      "grad_norm": 0.8237845301628113,
+      "learning_rate": 7.988005098827699e-06,
+      "loss": 0.2195,
+      "step": 687
+    },
+    {
+      "epoch": 2.9563812600969306,
+      "grad_norm": 0.7150204181671143,
+      "learning_rate": 7.982573677838172e-06,
+      "loss": 0.2053,
+      "step": 688
+    },
+    {
+      "epoch": 2.9606892837910608,
+      "grad_norm": 0.6601112484931946,
+      "learning_rate": 7.977136787760916e-06,
+      "loss": 0.1835,
+      "step": 689
+    },
+    {
+      "epoch": 2.9649973074851914,
+      "grad_norm": 0.7872753739356995,
+      "learning_rate": 7.97169443856545e-06,
+      "loss": 0.1974,
+      "step": 690
+    },
+    {
+      "epoch": 2.9693053311793216,
+      "grad_norm": 0.7161293625831604,
+      "learning_rate": 7.966246640231303e-06,
+      "loss": 0.1963,
+      "step": 691
+    },
+    {
+      "epoch": 2.973613354873452,
+      "grad_norm": 0.7552039623260498,
+      "learning_rate": 7.960793402748001e-06,
+      "loss": 0.1942,
+      "step": 692
+    },
+    {
+      "epoch": 2.977921378567582,
+      "grad_norm": 0.7136242389678955,
+      "learning_rate": 7.955334736115038e-06,
+      "loss": 0.2035,
+      "step": 693
+    },
+    {
+      "epoch": 2.982229402261712,
+      "grad_norm": 0.781005859375,
+      "learning_rate": 7.949870650341864e-06,
+      "loss": 0.194,
+      "step": 694
+    },
+    {
+      "epoch": 2.986537425955843,
+      "grad_norm": 0.6988127827644348,
+      "learning_rate": 7.944401155447872e-06,
+      "loss": 0.1936,
+      "step": 695
+    },
+    {
+      "epoch": 2.990845449649973,
+      "grad_norm": 0.7402964234352112,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.2199,
+      "step": 696
+    },
+    {
+      "epoch": 2.9951534733441036,
+      "grad_norm": 0.8665761947631836,
+      "learning_rate": 7.933445978424555e-06,
+      "loss": 0.2041,
+      "step": 697
+    },
+    {
+      "epoch": 2.999461497038234,
+      "grad_norm": 0.7297384142875671,
+      "learning_rate": 7.927960316383524e-06,
+      "loss": 0.1965,
+      "step": 698
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.7297384142875671,
+      "learning_rate": 7.92246928539823e-06,
+      "loss": 0.2041,
+      "step": 699
+    },
+    {
+      "epoch": 3.00430802369413,
+      "grad_norm": 3.1112265586853027,
+      "learning_rate": 7.916972895537471e-06,
+      "loss": 0.1905,
+      "step": 700
+    },
+    {
+      "epoch": 3.008616047388261,
+      "grad_norm": 0.7145331501960754,
+      "learning_rate": 7.911471156879866e-06,
+      "loss": 0.1807,
+      "step": 701
+    },
+    {
+      "epoch": 3.012924071082391,
+      "grad_norm": 0.6694017648696899,
+      "learning_rate": 7.905964079513851e-06,
+      "loss": 0.1801,
+      "step": 702
+    },
+    {
+      "epoch": 3.017232094776521,
+      "grad_norm": 0.6125895380973816,
+      "learning_rate": 7.900451673537646e-06,
+      "loss": 0.1847,
+      "step": 703
+    },
+    {
+      "epoch": 3.0215401184706514,
+      "grad_norm": 0.8431554436683655,
+      "learning_rate": 7.894933949059245e-06,
+      "loss": 0.1821,
+      "step": 704
+    },
+    {
+      "epoch": 3.025848142164782,
+      "grad_norm": 0.6408715844154358,
+      "learning_rate": 7.88941091619639e-06,
+      "loss": 0.1709,
+      "step": 705
+    },
+    {
+      "epoch": 3.0301561658589122,
+      "grad_norm": 0.6939838528633118,
+      "learning_rate": 7.883882585076558e-06,
+      "loss": 0.1765,
+      "step": 706
+    },
+    {
+      "epoch": 3.0344641895530424,
+      "grad_norm": 0.739832878112793,
+      "learning_rate": 7.87834896583695e-06,
+      "loss": 0.179,
+      "step": 707
+    },
+    {
+      "epoch": 3.038772213247173,
+      "grad_norm": 0.8355388045310974,
+      "learning_rate": 7.872810068624452e-06,
+      "loss": 0.1736,
+      "step": 708
+    },
+    {
+      "epoch": 3.0430802369413033,
+      "grad_norm": 0.7089900374412537,
+      "learning_rate": 7.867265903595632e-06,
+      "loss": 0.1787,
+      "step": 709
+    },
+    {
+      "epoch": 3.0473882606354334,
+      "grad_norm": 0.7276995182037354,
+      "learning_rate": 7.86171648091672e-06,
+      "loss": 0.1756,
+      "step": 710
+    },
+    {
+      "epoch": 3.0516962843295636,
+      "grad_norm": 0.6631605625152588,
+      "learning_rate": 7.856161810763584e-06,
+      "loss": 0.176,
+      "step": 711
+    },
+    {
+      "epoch": 3.0560043080236943,
+      "grad_norm": 0.7474228143692017,
+      "learning_rate": 7.850601903321717e-06,
+      "loss": 0.1817,
+      "step": 712
+    },
+    {
+      "epoch": 3.0603123317178245,
+      "grad_norm": 0.6688998937606812,
+      "learning_rate": 7.845036768786214e-06,
+      "loss": 0.1792,
+      "step": 713
+    },
+    {
+      "epoch": 3.0646203554119547,
+      "grad_norm": 0.7395163774490356,
+      "learning_rate": 7.839466417361753e-06,
+      "loss": 0.1658,
+      "step": 714
+    },
+    {
+      "epoch": 3.0689283791060853,
+      "grad_norm": 0.7095337510108948,
+      "learning_rate": 7.833890859262579e-06,
+      "loss": 0.1693,
+      "step": 715
+    },
+    {
+      "epoch": 3.0732364028002155,
+      "grad_norm": 0.8074468970298767,
+      "learning_rate": 7.828310104712488e-06,
+      "loss": 0.1751,
+      "step": 716
+    },
+    {
+      "epoch": 3.0775444264943457,
+      "grad_norm": 0.6688709259033203,
+      "learning_rate": 7.822724163944802e-06,
+      "loss": 0.1848,
+      "step": 717
+    },
+    {
+      "epoch": 3.081852450188476,
+      "grad_norm": 0.7156962752342224,
+      "learning_rate": 7.81713304720235e-06,
+      "loss": 0.1664,
+      "step": 718
+    },
+    {
+      "epoch": 3.0861604738826065,
+      "grad_norm": 0.7590762376785278,
+      "learning_rate": 7.811536764737454e-06,
+      "loss": 0.1786,
+      "step": 719
+    },
+    {
+      "epoch": 3.0904684975767367,
+      "grad_norm": 0.7737705111503601,
+      "learning_rate": 7.805935326811913e-06,
+      "loss": 0.182,
+      "step": 720
+    },
+    {
+      "epoch": 3.094776521270867,
+      "grad_norm": 0.7593074440956116,
+      "learning_rate": 7.800328743696973e-06,
+      "loss": 0.1744,
+      "step": 721
+    },
+    {
+      "epoch": 3.099084544964997,
+      "grad_norm": 0.8113828301429749,
+      "learning_rate": 7.794717025673318e-06,
+      "loss": 0.166,
+      "step": 722
+    },
+    {
+      "epoch": 3.1033925686591277,
+      "grad_norm": 0.7906281352043152,
+      "learning_rate": 7.789100183031045e-06,
+      "loss": 0.1876,
+      "step": 723
+    },
+    {
+      "epoch": 3.107700592353258,
+      "grad_norm": 0.7934765815734863,
+      "learning_rate": 7.783478226069652e-06,
+      "loss": 0.1852,
+      "step": 724
+    },
+    {
+      "epoch": 3.112008616047388,
+      "grad_norm": 0.814837634563446,
+      "learning_rate": 7.777851165098012e-06,
+      "loss": 0.1699,
+      "step": 725
+    },
+    {
+      "epoch": 3.1163166397415187,
+      "grad_norm": 0.6663200259208679,
+      "learning_rate": 7.772219010434359e-06,
+      "loss": 0.1742,
+      "step": 726
+    },
+    {
+      "epoch": 3.120624663435649,
+      "grad_norm": 0.7285510301589966,
+      "learning_rate": 7.766581772406266e-06,
+      "loss": 0.1708,
+      "step": 727
+    },
+    {
+      "epoch": 3.124932687129779,
+      "grad_norm": 0.6569226384162903,
+      "learning_rate": 7.760939461350622e-06,
+      "loss": 0.179,
+      "step": 728
+    },
+    {
+      "epoch": 3.1292407108239093,
+      "grad_norm": 0.7777225375175476,
+      "learning_rate": 7.755292087613635e-06,
+      "loss": 0.1706,
+      "step": 729
+    },
+    {
+      "epoch": 3.13354873451804,
+      "grad_norm": 0.7825784087181091,
+      "learning_rate": 7.749639661550775e-06,
+      "loss": 0.1806,
+      "step": 730
+    },
+    {
+      "epoch": 3.13785675821217,
+      "grad_norm": 0.8840234875679016,
+      "learning_rate": 7.743982193526791e-06,
+      "loss": 0.1839,
+      "step": 731
+    },
+    {
+      "epoch": 3.1421647819063003,
+      "grad_norm": 0.6787511110305786,
+      "learning_rate": 7.738319693915673e-06,
+      "loss": 0.1761,
+      "step": 732
+    },
+    {
+      "epoch": 3.146472805600431,
+      "grad_norm": 0.7972119450569153,
+      "learning_rate": 7.732652173100634e-06,
+      "loss": 0.1742,
+      "step": 733
+    },
+    {
+      "epoch": 3.150780829294561,
+      "grad_norm": 0.8184621334075928,
+      "learning_rate": 7.726979641474102e-06,
+      "loss": 0.1833,
+      "step": 734
+    },
+    {
+      "epoch": 3.1550888529886914,
+      "grad_norm": 0.75215083360672,
+      "learning_rate": 7.721302109437686e-06,
+      "loss": 0.1819,
+      "step": 735
+    },
+    {
+      "epoch": 3.1593968766828215,
+      "grad_norm": 0.8182296752929688,
+      "learning_rate": 7.715619587402165e-06,
+      "loss": 0.1819,
+      "step": 736
+    },
+    {
+      "epoch": 3.163704900376952,
+      "grad_norm": 0.8826239109039307,
+      "learning_rate": 7.709932085787473e-06,
+      "loss": 0.176,
+      "step": 737
+    },
+    {
+      "epoch": 3.1680129240710824,
+      "grad_norm": 0.7005189657211304,
+      "learning_rate": 7.704239615022671e-06,
+      "loss": 0.1758,
+      "step": 738
+    },
+    {
+      "epoch": 3.1723209477652126,
+      "grad_norm": 0.7199857831001282,
+      "learning_rate": 7.698542185545932e-06,
+      "loss": 0.1611,
+      "step": 739
+    },
+    {
+      "epoch": 3.176628971459343,
+      "grad_norm": 0.7488324046134949,
+      "learning_rate": 7.692839807804522e-06,
+      "loss": 0.1767,
+      "step": 740
+    },
+    {
+      "epoch": 3.1809369951534734,
+      "grad_norm": 0.7330511808395386,
+      "learning_rate": 7.687132492254783e-06,
+      "loss": 0.1746,
+      "step": 741
+    },
+    {
+      "epoch": 3.1852450188476036,
+      "grad_norm": 0.7547047138214111,
+      "learning_rate": 7.681420249362107e-06,
+      "loss": 0.1662,
+      "step": 742
+    },
+    {
+      "epoch": 3.1895530425417338,
+      "grad_norm": 0.6859620809555054,
+      "learning_rate": 7.675703089600926e-06,
+      "loss": 0.1768,
+      "step": 743
+    },
+    {
+      "epoch": 3.1938610662358644,
+      "grad_norm": 0.8874660730361938,
+      "learning_rate": 7.669981023454682e-06,
+      "loss": 0.1961,
+      "step": 744
+    },
+    {
+      "epoch": 3.1981690899299946,
+      "grad_norm": 0.9125708341598511,
+      "learning_rate": 7.664254061415818e-06,
+      "loss": 0.1797,
+      "step": 745
+    },
+    {
+      "epoch": 3.202477113624125,
+      "grad_norm": 0.711672842502594,
+      "learning_rate": 7.658522213985757e-06,
+      "loss": 0.1834,
+      "step": 746
+    },
+    {
+      "epoch": 3.2067851373182554,
+      "grad_norm": 0.7379571199417114,
+      "learning_rate": 7.652785491674872e-06,
+      "loss": 0.1719,
+      "step": 747
+    },
+    {
+      "epoch": 3.2110931610123856,
+      "grad_norm": 0.7924142479896545,
+      "learning_rate": 7.647043905002485e-06,
+      "loss": 0.1861,
+      "step": 748
+    },
+    {
+      "epoch": 3.215401184706516,
+      "grad_norm": 0.8076757192611694,
+      "learning_rate": 7.641297464496828e-06,
+      "loss": 0.1801,
+      "step": 749
+    },
+    {
+      "epoch": 3.219709208400646,
+      "grad_norm": 0.723019540309906,
+      "learning_rate": 7.635546180695039e-06,
+      "loss": 0.1758,
+      "step": 750
+    },
+    {
+      "epoch": 3.2240172320947766,
+      "grad_norm": 0.6700244545936584,
+      "learning_rate": 7.629790064143139e-06,
+      "loss": 0.1808,
+      "step": 751
+    },
+    {
+      "epoch": 3.228325255788907,
+      "grad_norm": 0.7725983262062073,
+      "learning_rate": 7.624029125396004e-06,
+      "loss": 0.1795,
+      "step": 752
+    },
+    {
+      "epoch": 3.232633279483037,
+      "grad_norm": 0.7336100935935974,
+      "learning_rate": 7.618263375017358e-06,
+      "loss": 0.174,
+      "step": 753
+    },
+    {
+      "epoch": 3.2369413031771677,
+      "grad_norm": 0.8215693831443787,
+      "learning_rate": 7.612492823579744e-06,
+      "loss": 0.1721,
+      "step": 754
+    },
+    {
+      "epoch": 3.241249326871298,
+      "grad_norm": 0.7242813110351562,
+      "learning_rate": 7.606717481664515e-06,
+      "loss": 0.1934,
+      "step": 755
+    },
+    {
+      "epoch": 3.245557350565428,
+      "grad_norm": 0.889627993106842,
+      "learning_rate": 7.600937359861799e-06,
+      "loss": 0.1795,
+      "step": 756
+    },
+    {
+      "epoch": 3.2498653742595582,
+      "grad_norm": 0.8864472508430481,
+      "learning_rate": 7.595152468770497e-06,
+      "loss": 0.1741,
+      "step": 757
+    },
+    {
+      "epoch": 3.254173397953689,
+      "grad_norm": 0.7779316306114197,
+      "learning_rate": 7.589362818998251e-06,
+      "loss": 0.1874,
+      "step": 758
+    },
+    {
+      "epoch": 3.258481421647819,
+      "grad_norm": 0.8183091878890991,
+      "learning_rate": 7.58356842116143e-06,
+      "loss": 0.1851,
+      "step": 759
+    },
+    {
+      "epoch": 3.2627894453419493,
+      "grad_norm": 0.8073911070823669,
+      "learning_rate": 7.57776928588511e-06,
+      "loss": 0.186,
+      "step": 760
+    },
+    {
+      "epoch": 3.26709746903608,
+      "grad_norm": 0.8156265020370483,
+      "learning_rate": 7.571965423803052e-06,
+      "loss": 0.1735,
+      "step": 761
+    },
+    {
+      "epoch": 3.27140549273021,
+      "grad_norm": 0.7891457080841064,
+      "learning_rate": 7.566156845557684e-06,
+      "loss": 0.1708,
+      "step": 762
+    },
+    {
+      "epoch": 3.2757135164243403,
+      "grad_norm": 0.7164343595504761,
+      "learning_rate": 7.560343561800087e-06,
+      "loss": 0.183,
+      "step": 763
+    },
+    {
+      "epoch": 3.2800215401184705,
+      "grad_norm": 0.7573089599609375,
+      "learning_rate": 7.554525583189969e-06,
+      "loss": 0.1883,
+      "step": 764
+    },
+    {
+      "epoch": 3.284329563812601,
+      "grad_norm": 0.7825318574905396,
+      "learning_rate": 7.548702920395639e-06,
+      "loss": 0.1787,
+      "step": 765
+    },
+    {
+      "epoch": 3.2886375875067313,
+      "grad_norm": 0.7607948184013367,
+      "learning_rate": 7.542875584094006e-06,
+      "loss": 0.1934,
+      "step": 766
+    },
+    {
+      "epoch": 3.2929456112008615,
+      "grad_norm": 0.8865452408790588,
+      "learning_rate": 7.537043584970543e-06,
+      "loss": 0.1713,
+      "step": 767
+    },
+    {
+      "epoch": 3.297253634894992,
+      "grad_norm": 0.7793084383010864,
+      "learning_rate": 7.53120693371927e-06,
+      "loss": 0.1741,
+      "step": 768
+    },
+    {
+      "epoch": 3.3015616585891223,
+      "grad_norm": 0.6900126934051514,
+      "learning_rate": 7.525365641042749e-06,
+      "loss": 0.186,
+      "step": 769
+    },
+    {
+      "epoch": 3.3058696822832525,
+      "grad_norm": 0.7191606760025024,
+      "learning_rate": 7.519519717652039e-06,
+      "loss": 0.1805,
+      "step": 770
+    },
+    {
+      "epoch": 3.3101777059773827,
+      "grad_norm": 0.8574715852737427,
+      "learning_rate": 7.5136691742667e-06,
+      "loss": 0.1846,
+      "step": 771
+    },
+    {
+      "epoch": 3.3144857296715133,
+      "grad_norm": 0.7486770153045654,
+      "learning_rate": 7.507814021614761e-06,
+      "loss": 0.1788,
+      "step": 772
+    },
+    {
+      "epoch": 3.3187937533656435,
+      "grad_norm": 0.7185223698616028,
+      "learning_rate": 7.501954270432701e-06,
+      "loss": 0.1813,
+      "step": 773
+    },
+    {
+      "epoch": 3.3231017770597737,
+      "grad_norm": 0.7054628133773804,
+      "learning_rate": 7.496089931465432e-06,
+      "loss": 0.192,
+      "step": 774
+    },
+    {
+      "epoch": 3.3274098007539044,
+      "grad_norm": 0.8348388671875,
+      "learning_rate": 7.490221015466279e-06,
+      "loss": 0.1831,
+      "step": 775
+    },
+    {
+      "epoch": 3.3317178244480345,
+      "grad_norm": 0.90952467918396,
+      "learning_rate": 7.4843475331969614e-06,
+      "loss": 0.1764,
+      "step": 776
+    },
+    {
+      "epoch": 3.3360258481421647,
+      "grad_norm": 0.7084717750549316,
+      "learning_rate": 7.478469495427569e-06,
+      "loss": 0.1865,
+      "step": 777
+    },
+    {
+      "epoch": 3.340333871836295,
+      "grad_norm": 0.771369457244873,
+      "learning_rate": 7.4725869129365484e-06,
+      "loss": 0.1846,
+      "step": 778
+    },
+    {
+      "epoch": 3.3446418955304256,
+      "grad_norm": 0.75652676820755,
+      "learning_rate": 7.4666997965106725e-06,
+      "loss": 0.1875,
+      "step": 779
+    },
+    {
+      "epoch": 3.3489499192245558,
+      "grad_norm": 0.7481139302253723,
+      "learning_rate": 7.4608081569450365e-06,
+      "loss": 0.1752,
+      "step": 780
+    },
+    {
+      "epoch": 3.353257942918686,
+      "grad_norm": 0.7712534070014954,
+      "learning_rate": 7.4549120050430265e-06,
+      "loss": 0.181,
+      "step": 781
+    },
+    {
+      "epoch": 3.3575659666128166,
+      "grad_norm": 0.8637974858283997,
+      "learning_rate": 7.449011351616302e-06,
+      "loss": 0.1774,
+      "step": 782
+    },
+    {
+      "epoch": 3.361873990306947,
+      "grad_norm": 0.712335467338562,
+      "learning_rate": 7.443106207484776e-06,
+      "loss": 0.1878,
+      "step": 783
+    },
+    {
+      "epoch": 3.366182014001077,
+      "grad_norm": 0.779151201248169,
+      "learning_rate": 7.437196583476597e-06,
+      "loss": 0.1728,
+      "step": 784
+    },
+    {
+      "epoch": 3.370490037695207,
+      "grad_norm": 0.7468299865722656,
+      "learning_rate": 7.43128249042813e-06,
+      "loss": 0.1911,
+      "step": 785
+    },
+    {
+      "epoch": 3.374798061389338,
+      "grad_norm": 0.8930696249008179,
+      "learning_rate": 7.425363939183931e-06,
+      "loss": 0.1798,
+      "step": 786
+    },
+    {
+      "epoch": 3.379106085083468,
+      "grad_norm": 0.9994642734527588,
+      "learning_rate": 7.419440940596735e-06,
+      "loss": 0.1881,
+      "step": 787
+    },
+    {
+      "epoch": 3.383414108777598,
+      "grad_norm": 0.8835253715515137,
+      "learning_rate": 7.41351350552743e-06,
+      "loss": 0.1827,
+      "step": 788
+    },
+    {
+      "epoch": 3.387722132471729,
+      "grad_norm": 0.7508931159973145,
+      "learning_rate": 7.407581644845038e-06,
+      "loss": 0.1867,
+      "step": 789
+    },
+    {
+      "epoch": 3.392030156165859,
+      "grad_norm": 0.8232507109642029,
+      "learning_rate": 7.401645369426697e-06,
+      "loss": 0.1807,
+      "step": 790
+    },
+    {
+      "epoch": 3.396338179859989,
+      "grad_norm": 0.755211353302002,
+      "learning_rate": 7.395704690157644e-06,
+      "loss": 0.1841,
+      "step": 791
+    },
+    {
+      "epoch": 3.4006462035541194,
+      "grad_norm": 0.8512404561042786,
+      "learning_rate": 7.389759617931183e-06,
+      "loss": 0.1932,
+      "step": 792
+    },
+    {
+      "epoch": 3.40495422724825,
+      "grad_norm": 0.7709099054336548,
+      "learning_rate": 7.383810163648682e-06,
+      "loss": 0.166,
+      "step": 793
+    },
+    {
+      "epoch": 3.4092622509423802,
+      "grad_norm": 0.7232650518417358,
+      "learning_rate": 7.3778563382195365e-06,
+      "loss": 0.18,
+      "step": 794
+    },
+    {
+      "epoch": 3.4135702746365104,
+      "grad_norm": 0.8082464337348938,
+      "learning_rate": 7.371898152561166e-06,
+      "loss": 0.186,
+      "step": 795
+    },
+    {
+      "epoch": 3.417878298330641,
+      "grad_norm": 0.7701036334037781,
+      "learning_rate": 7.365935617598975e-06,
+      "loss": 0.1739,
+      "step": 796
+    },
+    {
+      "epoch": 3.4221863220247712,
+      "grad_norm": 0.7623946666717529,
+      "learning_rate": 7.359968744266353e-06,
+      "loss": 0.1933,
+      "step": 797
+    },
+    {
+      "epoch": 3.4264943457189014,
+      "grad_norm": 0.81331866979599,
+      "learning_rate": 7.35399754350464e-06,
+      "loss": 0.1903,
+      "step": 798
+    },
+    {
+      "epoch": 3.4308023694130316,
+      "grad_norm": 0.7704458236694336,
+      "learning_rate": 7.3480220262631095e-06,
+      "loss": 0.1756,
+      "step": 799
+    },
+    {
+      "epoch": 3.4351103931071623,
+      "grad_norm": 0.7425993084907532,
+      "learning_rate": 7.342042203498952e-06,
+      "loss": 0.1786,
+      "step": 800
+    },
+    {
+      "epoch": 3.4394184168012925,
+      "grad_norm": 0.8658850789070129,
+      "learning_rate": 7.336058086177253e-06,
+      "loss": 0.1834,
+      "step": 801
+    },
+    {
+      "epoch": 3.4437264404954226,
+      "grad_norm": 0.8086578249931335,
+      "learning_rate": 7.330069685270976e-06,
+      "loss": 0.1808,
+      "step": 802
+    },
+    {
+      "epoch": 3.4480344641895533,
+      "grad_norm": 0.7482308745384216,
+      "learning_rate": 7.3240770117609325e-06,
+      "loss": 0.1887,
+      "step": 803
+    },
+    {
+      "epoch": 3.4523424878836835,
+      "grad_norm": 0.8038567900657654,
+      "learning_rate": 7.318080076635773e-06,
+      "loss": 0.1831,
+      "step": 804
+    },
+    {
+      "epoch": 3.4566505115778137,
+      "grad_norm": 0.7037034034729004,
+      "learning_rate": 7.312078890891962e-06,
+      "loss": 0.1766,
+      "step": 805
+    },
+    {
+      "epoch": 3.460958535271944,
+      "grad_norm": 0.7575218081474304,
+      "learning_rate": 7.306073465533759e-06,
+      "loss": 0.1753,
+      "step": 806
+    },
+    {
+      "epoch": 3.4652665589660745,
+      "grad_norm": 0.831322193145752,
+      "learning_rate": 7.300063811573194e-06,
+      "loss": 0.1988,
+      "step": 807
+    },
+    {
+      "epoch": 3.4695745826602047,
+      "grad_norm": 0.737129271030426,
+      "learning_rate": 7.294049940030055e-06,
+      "loss": 0.1904,
+      "step": 808
+    },
+    {
+      "epoch": 3.473882606354335,
+      "grad_norm": 0.7896698713302612,
+      "learning_rate": 7.2880318619318605e-06,
+      "loss": 0.1883,
+      "step": 809
+    },
+    {
+      "epoch": 3.4781906300484655,
+      "grad_norm": 0.7189210653305054,
+      "learning_rate": 7.2820095883138456e-06,
+      "loss": 0.1743,
+      "step": 810
+    },
+    {
+      "epoch": 3.4824986537425957,
+      "grad_norm": 0.7875738739967346,
+      "learning_rate": 7.2759831302189376e-06,
+      "loss": 0.1917,
+      "step": 811
+    },
+    {
+      "epoch": 3.486806677436726,
+      "grad_norm": 0.8502741456031799,
+      "learning_rate": 7.269952498697734e-06,
+      "loss": 0.193,
+      "step": 812
+    },
+    {
+      "epoch": 3.491114701130856,
+      "grad_norm": 0.7916224002838135,
+      "learning_rate": 7.2639177048084894e-06,
+      "loss": 0.1877,
+      "step": 813
+    },
+    {
+      "epoch": 3.4954227248249867,
+      "grad_norm": 0.7110468745231628,
+      "learning_rate": 7.25787875961709e-06,
+      "loss": 0.1751,
+      "step": 814
+    },
+    {
+      "epoch": 3.499730748519117,
+      "grad_norm": 0.7333479523658752,
+      "learning_rate": 7.2518356741970285e-06,
+      "loss": 0.1736,
+      "step": 815
+    },
+    {
+      "epoch": 3.504038772213247,
+      "grad_norm": 0.7988889813423157,
+      "learning_rate": 7.245788459629397e-06,
+      "loss": 0.1756,
+      "step": 816
+    },
+    {
+      "epoch": 3.5083467959073777,
+      "grad_norm": 0.7231525778770447,
+      "learning_rate": 7.239737127002854e-06,
+      "loss": 0.1748,
+      "step": 817
+    },
+    {
+      "epoch": 3.512654819601508,
+      "grad_norm": 0.7418997287750244,
+      "learning_rate": 7.233681687413614e-06,
+      "loss": 0.1809,
+      "step": 818
+    },
+    {
+      "epoch": 3.516962843295638,
+      "grad_norm": 0.7931709885597229,
+      "learning_rate": 7.227622151965418e-06,
+      "loss": 0.19,
+      "step": 819
+    },
+    {
+      "epoch": 3.5212708669897683,
+      "grad_norm": 0.8055832386016846,
+      "learning_rate": 7.221558531769519e-06,
+      "loss": 0.1841,
+      "step": 820
+    },
+    {
+      "epoch": 3.5255788906838985,
+      "grad_norm": 0.8355058431625366,
+      "learning_rate": 7.21549083794466e-06,
+      "loss": 0.1834,
+      "step": 821
+    },
+    {
+      "epoch": 3.529886914378029,
+      "grad_norm": 0.7648366093635559,
+      "learning_rate": 7.209419081617055e-06,
+      "loss": 0.1836,
+      "step": 822
+    },
+    {
+      "epoch": 3.5341949380721593,
+      "grad_norm": 0.8649141192436218,
+      "learning_rate": 7.203343273920365e-06,
+      "loss": 0.1928,
+      "step": 823
+    },
+    {
+      "epoch": 3.53850296176629,
+      "grad_norm": 0.7186275720596313,
+      "learning_rate": 7.197263425995682e-06,
+      "loss": 0.1767,
+      "step": 824
+    },
+    {
+      "epoch": 3.54281098546042,
+      "grad_norm": 0.7774980068206787,
+      "learning_rate": 7.191179548991507e-06,
+      "loss": 0.1797,
+      "step": 825
+    },
+    {
+      "epoch": 3.5471190091545504,
+      "grad_norm": 0.7177664041519165,
+      "learning_rate": 7.185091654063724e-06,
+      "loss": 0.1845,
+      "step": 826
+    },
+    {
+      "epoch": 3.5514270328486806,
+      "grad_norm": 0.7631592750549316,
+      "learning_rate": 7.1789997523755915e-06,
+      "loss": 0.1785,
+      "step": 827
+    },
+    {
+      "epoch": 3.5557350565428107,
+      "grad_norm": 0.7096567749977112,
+      "learning_rate": 7.172903855097712e-06,
+      "loss": 0.1893,
+      "step": 828
+    },
+    {
+      "epoch": 3.5600430802369414,
+      "grad_norm": 0.800830066204071,
+      "learning_rate": 7.166803973408012e-06,
+      "loss": 0.1864,
+      "step": 829
+    },
+    {
+      "epoch": 3.5643511039310716,
+      "grad_norm": 0.7556864619255066,
+      "learning_rate": 7.160700118491729e-06,
+      "loss": 0.1876,
+      "step": 830
+    },
+    {
+      "epoch": 3.568659127625202,
+      "grad_norm": 0.7628764510154724,
+      "learning_rate": 7.154592301541383e-06,
+      "loss": 0.1873,
+      "step": 831
+    },
+    {
+      "epoch": 3.5729671513193324,
+      "grad_norm": 0.7047873139381409,
+      "learning_rate": 7.148480533756759e-06,
+      "loss": 0.1781,
+      "step": 832
+    },
+    {
+      "epoch": 3.5772751750134626,
+      "grad_norm": 0.7988886833190918,
+      "learning_rate": 7.142364826344891e-06,
+      "loss": 0.1856,
+      "step": 833
+    },
+    {
+      "epoch": 3.581583198707593,
+      "grad_norm": 0.7696588635444641,
+      "learning_rate": 7.1362451905200285e-06,
+      "loss": 0.1739,
+      "step": 834
+    },
+    {
+      "epoch": 3.585891222401723,
+      "grad_norm": 0.8719116449356079,
+      "learning_rate": 7.130121637503633e-06,
+      "loss": 0.1937,
+      "step": 835
+    },
+    {
+      "epoch": 3.5901992460958536,
+      "grad_norm": 0.8098923563957214,
+      "learning_rate": 7.123994178524345e-06,
+      "loss": 0.1859,
+      "step": 836
+    },
+    {
+      "epoch": 3.594507269789984,
+      "grad_norm": 0.8798801898956299,
+      "learning_rate": 7.117862824817966e-06,
+      "loss": 0.1802,
+      "step": 837
+    },
+    {
+      "epoch": 3.598815293484114,
+      "grad_norm": 0.8650680184364319,
+      "learning_rate": 7.1117275876274425e-06,
+      "loss": 0.1806,
+      "step": 838
+    },
+    {
+      "epoch": 3.6031233171782446,
+      "grad_norm": 0.8523943424224854,
+      "learning_rate": 7.105588478202838e-06,
+      "loss": 0.1752,
+      "step": 839
+    },
+    {
+      "epoch": 3.607431340872375,
+      "grad_norm": 0.7416144609451294,
+      "learning_rate": 7.099445507801324e-06,
+      "loss": 0.1747,
+      "step": 840
+    },
+    {
+      "epoch": 3.611739364566505,
+      "grad_norm": 0.7997528910636902,
+      "learning_rate": 7.093298687687141e-06,
+      "loss": 0.1785,
+      "step": 841
+    },
+    {
+      "epoch": 3.616047388260635,
+      "grad_norm": 0.8117310404777527,
+      "learning_rate": 7.0871480291315975e-06,
+      "loss": 0.179,
+      "step": 842
+    },
+    {
+      "epoch": 3.620355411954766,
+      "grad_norm": 0.8026857972145081,
+      "learning_rate": 7.080993543413035e-06,
+      "loss": 0.1905,
+      "step": 843
+    },
+    {
+      "epoch": 3.624663435648896,
+      "grad_norm": 0.7391301393508911,
+      "learning_rate": 7.0748352418168174e-06,
+      "loss": 0.1751,
+      "step": 844
+    },
+    {
+      "epoch": 3.6289714593430262,
+      "grad_norm": 0.7324339747428894,
+      "learning_rate": 7.068673135635302e-06,
+      "loss": 0.177,
+      "step": 845
+    },
+    {
+      "epoch": 3.633279483037157,
+      "grad_norm": 0.752116858959198,
+      "learning_rate": 7.062507236167826e-06,
+      "loss": 0.1919,
+      "step": 846
+    },
+    {
+      "epoch": 3.637587506731287,
+      "grad_norm": 0.8739787340164185,
+      "learning_rate": 7.056337554720676e-06,
+      "loss": 0.18,
+      "step": 847
+    },
+    {
+      "epoch": 3.6418955304254172,
+      "grad_norm": 0.9470150470733643,
+      "learning_rate": 7.050164102607081e-06,
+      "loss": 0.1889,
+      "step": 848
+    },
+    {
+      "epoch": 3.6462035541195474,
+      "grad_norm": 0.8388432264328003,
+      "learning_rate": 7.043986891147179e-06,
+      "loss": 0.1911,
+      "step": 849
+    },
+    {
+      "epoch": 3.650511577813678,
+      "grad_norm": 0.7548778057098389,
+      "learning_rate": 7.037805931668006e-06,
+      "loss": 0.1833,
+      "step": 850
+    },
+    {
+      "epoch": 3.6548196015078083,
+      "grad_norm": 0.7865473628044128,
+      "learning_rate": 7.031621235503464e-06,
+      "loss": 0.1824,
+      "step": 851
+    },
+    {
+      "epoch": 3.6591276252019385,
+      "grad_norm": 0.7438720464706421,
+      "learning_rate": 7.025432813994315e-06,
+      "loss": 0.1812,
+      "step": 852
+    },
+    {
+      "epoch": 3.663435648896069,
+      "grad_norm": 0.7457842230796814,
+      "learning_rate": 7.019240678488145e-06,
+      "loss": 0.1948,
+      "step": 853
+    },
+    {
+      "epoch": 3.6677436725901993,
+      "grad_norm": 0.7958986759185791,
+      "learning_rate": 7.013044840339353e-06,
+      "loss": 0.1845,
+      "step": 854
+    },
+    {
+      "epoch": 3.6720516962843295,
+      "grad_norm": 0.7546960711479187,
+      "learning_rate": 7.006845310909131e-06,
+      "loss": 0.1807,
+      "step": 855
+    },
+    {
+      "epoch": 3.6763597199784597,
+      "grad_norm": 0.7642547488212585,
+      "learning_rate": 7.000642101565434e-06,
+      "loss": 0.175,
+      "step": 856
+    },
+    {
+      "epoch": 3.6806677436725903,
+      "grad_norm": 0.6758170127868652,
+      "learning_rate": 6.994435223682966e-06,
+      "loss": 0.1836,
+      "step": 857
+    },
+    {
+      "epoch": 3.6849757673667205,
+      "grad_norm": 0.7619655728340149,
+      "learning_rate": 6.9882246886431615e-06,
+      "loss": 0.1881,
+      "step": 858
+    },
+    {
+      "epoch": 3.6892837910608507,
+      "grad_norm": 0.7106654644012451,
+      "learning_rate": 6.982010507834158e-06,
+      "loss": 0.182,
+      "step": 859
+    },
+    {
+      "epoch": 3.6935918147549813,
+      "grad_norm": 0.7118704319000244,
+      "learning_rate": 6.975792692650778e-06,
+      "loss": 0.1894,
+      "step": 860
+    },
+    {
+      "epoch": 3.6978998384491115,
+      "grad_norm": 0.8190689086914062,
+      "learning_rate": 6.969571254494509e-06,
+      "loss": 0.1806,
+      "step": 861
+    },
+    {
+      "epoch": 3.7022078621432417,
+      "grad_norm": 0.8226962089538574,
+      "learning_rate": 6.963346204773483e-06,
+      "loss": 0.1784,
+      "step": 862
+    },
+    {
+      "epoch": 3.706515885837372,
+      "grad_norm": 0.6322476863861084,
+      "learning_rate": 6.957117554902452e-06,
+      "loss": 0.1706,
+      "step": 863
+    },
+    {
+      "epoch": 3.7108239095315025,
+      "grad_norm": 0.8125499486923218,
+      "learning_rate": 6.950885316302773e-06,
+      "loss": 0.1934,
+      "step": 864
+    },
+    {
+      "epoch": 3.7151319332256327,
+      "grad_norm": 0.8503420948982239,
+      "learning_rate": 6.94464950040238e-06,
+      "loss": 0.1833,
+      "step": 865
+    },
+    {
+      "epoch": 3.719439956919763,
+      "grad_norm": 0.7436283826828003,
+      "learning_rate": 6.938410118635768e-06,
+      "loss": 0.1784,
+      "step": 866
+    },
+    {
+      "epoch": 3.7237479806138936,
+      "grad_norm": 0.7340200543403625,
+      "learning_rate": 6.9321671824439715e-06,
+      "loss": 0.1855,
+      "step": 867
+    },
+    {
+      "epoch": 3.7280560043080238,
+      "grad_norm": 0.7100676894187927,
+      "learning_rate": 6.9259207032745415e-06,
+      "loss": 0.1699,
+      "step": 868
+    },
+    {
+      "epoch": 3.732364028002154,
+      "grad_norm": 0.8664551973342896,
+      "learning_rate": 6.919670692581526e-06,
+      "loss": 0.1914,
+      "step": 869
+    },
+    {
+      "epoch": 3.736672051696284,
+      "grad_norm": 0.7995779514312744,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 0.1768,
+      "step": 870
+    },
+    {
+      "epoch": 3.7409800753904148,
+      "grad_norm": 0.7231603264808655,
+      "learning_rate": 6.907160122473291e-06,
+      "loss": 0.1843,
+      "step": 871
+    },
+    {
+      "epoch": 3.745288099084545,
+      "grad_norm": 0.7902730107307434,
+      "learning_rate": 6.90089958599846e-06,
+      "loss": 0.1811,
+      "step": 872
+    },
+    {
+      "epoch": 3.749596122778675,
+      "grad_norm": 0.7823545336723328,
+      "learning_rate": 6.894635563880785e-06,
+      "loss": 0.1795,
+      "step": 873
+    },
+    {
+      "epoch": 3.753904146472806,
+      "grad_norm": 0.7500615119934082,
+      "learning_rate": 6.88836806760648e-06,
+      "loss": 0.1758,
+      "step": 874
+    },
+    {
+      "epoch": 3.758212170166936,
+      "grad_norm": 0.7018880844116211,
+      "learning_rate": 6.882097108668132e-06,
+      "loss": 0.1733,
+      "step": 875
+    },
+    {
+      "epoch": 3.762520193861066,
+      "grad_norm": 0.7450665235519409,
+      "learning_rate": 6.875822698564678e-06,
+      "loss": 0.1676,
+      "step": 876
+    },
+    {
+      "epoch": 3.7668282175551964,
+      "grad_norm": 0.7359514832496643,
+      "learning_rate": 6.869544848801383e-06,
+      "loss": 0.19,
+      "step": 877
+    },
+    {
+      "epoch": 3.771136241249327,
+      "grad_norm": 0.7743359208106995,
+      "learning_rate": 6.863263570889818e-06,
+      "loss": 0.1688,
+      "step": 878
+    },
+    {
+      "epoch": 3.775444264943457,
+      "grad_norm": 0.6148492097854614,
+      "learning_rate": 6.85697887634784e-06,
+      "loss": 0.1873,
+      "step": 879
+    },
+    {
+      "epoch": 3.7797522886375874,
+      "grad_norm": 0.7440102696418762,
+      "learning_rate": 6.850690776699574e-06,
+      "loss": 0.1903,
+      "step": 880
+    },
+    {
+      "epoch": 3.784060312331718,
+      "grad_norm": 0.7168632745742798,
+      "learning_rate": 6.844399283475384e-06,
+      "loss": 0.191,
+      "step": 881
+    },
+    {
+      "epoch": 3.788368336025848,
+      "grad_norm": 0.6857194900512695,
+      "learning_rate": 6.838104408211862e-06,
+      "loss": 0.1831,
+      "step": 882
+    },
+    {
+      "epoch": 3.7926763597199784,
+      "grad_norm": 0.6851407885551453,
+      "learning_rate": 6.831806162451799e-06,
+      "loss": 0.1842,
+      "step": 883
+    },
+    {
+      "epoch": 3.7969843834141086,
+      "grad_norm": 0.8678072094917297,
+      "learning_rate": 6.825504557744167e-06,
+      "loss": 0.1892,
+      "step": 884
+    },
+    {
+      "epoch": 3.8012924071082392,
+      "grad_norm": 0.7824573516845703,
+      "learning_rate": 6.819199605644093e-06,
+      "loss": 0.184,
+      "step": 885
+    },
+    {
+      "epoch": 3.8056004308023694,
+      "grad_norm": 0.8031753301620483,
+      "learning_rate": 6.812891317712851e-06,
+      "loss": 0.1851,
+      "step": 886
+    },
+    {
+      "epoch": 3.8099084544964996,
+      "grad_norm": 0.747918963432312,
+      "learning_rate": 6.806579705517824e-06,
+      "loss": 0.1791,
+      "step": 887
+    },
+    {
+      "epoch": 3.8142164781906303,
+      "grad_norm": 0.6846821904182434,
+      "learning_rate": 6.800264780632495e-06,
+      "loss": 0.1867,
+      "step": 888
+    },
+    {
+      "epoch": 3.8185245018847604,
+      "grad_norm": 0.6762527227401733,
+      "learning_rate": 6.793946554636417e-06,
+      "loss": 0.1946,
+      "step": 889
+    },
+    {
+      "epoch": 3.8228325255788906,
+      "grad_norm": 0.8078494071960449,
+      "learning_rate": 6.7876250391152e-06,
+      "loss": 0.1811,
+      "step": 890
+    },
+    {
+      "epoch": 3.827140549273021,
+      "grad_norm": 0.709995687007904,
+      "learning_rate": 6.781300245660487e-06,
+      "loss": 0.177,
+      "step": 891
+    },
+    {
+      "epoch": 3.8314485729671515,
+      "grad_norm": 0.7268922328948975,
+      "learning_rate": 6.774972185869928e-06,
+      "loss": 0.1909,
+      "step": 892
+    },
+    {
+      "epoch": 3.8357565966612817,
+      "grad_norm": 0.7240892052650452,
+      "learning_rate": 6.768640871347163e-06,
+      "loss": 0.1906,
+      "step": 893
+    },
+    {
+      "epoch": 3.840064620355412,
+      "grad_norm": 0.7050197720527649,
+      "learning_rate": 6.762306313701803e-06,
+      "loss": 0.1792,
+      "step": 894
+    },
+    {
+      "epoch": 3.8443726440495425,
+      "grad_norm": 0.7163313031196594,
+      "learning_rate": 6.7559685245494025e-06,
+      "loss": 0.1844,
+      "step": 895
+    },
+    {
+      "epoch": 3.8486806677436727,
+      "grad_norm": 0.7890435457229614,
+      "learning_rate": 6.749627515511443e-06,
+      "loss": 0.1886,
+      "step": 896
+    },
+    {
+      "epoch": 3.852988691437803,
+      "grad_norm": 0.7992168068885803,
+      "learning_rate": 6.743283298215312e-06,
+      "loss": 0.1907,
+      "step": 897
+    },
+    {
+      "epoch": 3.857296715131933,
+      "grad_norm": 0.904237687587738,
+      "learning_rate": 6.736935884294275e-06,
+      "loss": 0.1774,
+      "step": 898
+    },
+    {
+      "epoch": 3.8616047388260637,
+      "grad_norm": 0.7321799397468567,
+      "learning_rate": 6.730585285387465e-06,
+      "loss": 0.18,
+      "step": 899
+    },
+    {
+      "epoch": 3.865912762520194,
+      "grad_norm": 0.7467935681343079,
+      "learning_rate": 6.724231513139853e-06,
+      "loss": 0.1685,
+      "step": 900
+    },
+    {
+      "epoch": 3.870220786214324,
+      "grad_norm": 0.7189427614212036,
+      "learning_rate": 6.717874579202227e-06,
+      "loss": 0.1807,
+      "step": 901
+    },
+    {
+      "epoch": 3.8745288099084547,
+      "grad_norm": 0.810021162033081,
+      "learning_rate": 6.711514495231173e-06,
+      "loss": 0.1857,
+      "step": 902
+    },
+    {
+      "epoch": 3.878836833602585,
+      "grad_norm": 0.7699490785598755,
+      "learning_rate": 6.705151272889055e-06,
+      "loss": 0.1804,
+      "step": 903
+    },
+    {
+      "epoch": 3.883144857296715,
+      "grad_norm": 0.723810076713562,
+      "learning_rate": 6.698784923843993e-06,
+      "loss": 0.1767,
+      "step": 904
+    },
+    {
+      "epoch": 3.8874528809908453,
+      "grad_norm": 0.7115669250488281,
+      "learning_rate": 6.692415459769835e-06,
+      "loss": 0.1681,
+      "step": 905
+    },
+    {
+      "epoch": 3.891760904684976,
+      "grad_norm": 0.7950160503387451,
+      "learning_rate": 6.686042892346147e-06,
+      "loss": 0.1868,
+      "step": 906
+    },
+    {
+      "epoch": 3.896068928379106,
+      "grad_norm": 0.7404954433441162,
+      "learning_rate": 6.679667233258179e-06,
+      "loss": 0.1793,
+      "step": 907
+    },
+    {
+      "epoch": 3.9003769520732363,
+      "grad_norm": 0.7232417464256287,
+      "learning_rate": 6.673288494196858e-06,
+      "loss": 0.1698,
+      "step": 908
+    },
+    {
+      "epoch": 3.904684975767367,
+      "grad_norm": 0.6922205090522766,
+      "learning_rate": 6.666906686858753e-06,
+      "loss": 0.1878,
+      "step": 909
+    },
+    {
+      "epoch": 3.908992999461497,
+      "grad_norm": 0.8065887093544006,
+      "learning_rate": 6.66052182294606e-06,
+      "loss": 0.172,
+      "step": 910
+    },
+    {
+      "epoch": 3.9133010231556273,
+      "grad_norm": 0.6786320805549622,
+      "learning_rate": 6.654133914166582e-06,
+      "loss": 0.1818,
+      "step": 911
+    },
+    {
+      "epoch": 3.9176090468497575,
+      "grad_norm": 0.9004634022712708,
+      "learning_rate": 6.647742972233703e-06,
+      "loss": 0.1814,
+      "step": 912
+    },
+    {
+      "epoch": 3.9219170705438877,
+      "grad_norm": 0.928763747215271,
+      "learning_rate": 6.641349008866369e-06,
+      "loss": 0.1819,
+      "step": 913
+    },
+    {
+      "epoch": 3.9262250942380184,
+      "grad_norm": 0.8380340337753296,
+      "learning_rate": 6.634952035789069e-06,
+      "loss": 0.1799,
+      "step": 914
+    },
+    {
+      "epoch": 3.9305331179321485,
+      "grad_norm": 0.7506704330444336,
+      "learning_rate": 6.628552064731807e-06,
+      "loss": 0.1777,
+      "step": 915
+    },
+    {
+      "epoch": 3.934841141626279,
+      "grad_norm": 0.7512242794036865,
+      "learning_rate": 6.622149107430088e-06,
+      "loss": 0.1745,
+      "step": 916
+    },
+    {
+      "epoch": 3.9391491653204094,
+      "grad_norm": 0.817297101020813,
+      "learning_rate": 6.6157431756248906e-06,
+      "loss": 0.1851,
+      "step": 917
+    },
+    {
+      "epoch": 3.9434571890145396,
+      "grad_norm": 0.817671537399292,
+      "learning_rate": 6.609334281062647e-06,
+      "loss": 0.1791,
+      "step": 918
+    },
+    {
+      "epoch": 3.9477652127086698,
+      "grad_norm": 0.8131271004676819,
+      "learning_rate": 6.602922435495225e-06,
+      "loss": 0.18,
+      "step": 919
+    },
+    {
+      "epoch": 3.9520732364028,
+      "grad_norm": 0.7726566791534424,
+      "learning_rate": 6.5965076506799e-06,
+      "loss": 0.1738,
+      "step": 920
+    },
+    {
+      "epoch": 3.9563812600969306,
+      "grad_norm": 0.7950050830841064,
+      "learning_rate": 6.5900899383793415e-06,
+      "loss": 0.1756,
+      "step": 921
+    },
+    {
+      "epoch": 3.9606892837910608,
+      "grad_norm": 0.7116392850875854,
+      "learning_rate": 6.583669310361583e-06,
+      "loss": 0.1766,
+      "step": 922
+    },
+    {
+      "epoch": 3.9649973074851914,
+      "grad_norm": 0.7608798742294312,
+      "learning_rate": 6.577245778400006e-06,
+      "loss": 0.1897,
+      "step": 923
+    },
+    {
+      "epoch": 3.9693053311793216,
+      "grad_norm": 0.8389754891395569,
+      "learning_rate": 6.570819354273317e-06,
+      "loss": 0.1857,
+      "step": 924
+    },
+    {
+      "epoch": 3.973613354873452,
+      "grad_norm": 0.7660545110702515,
+      "learning_rate": 6.564390049765528e-06,
+      "loss": 0.1807,
+      "step": 925
+    },
+    {
+      "epoch": 3.977921378567582,
+      "grad_norm": 0.736478865146637,
+      "learning_rate": 6.557957876665926e-06,
+      "loss": 0.1777,
+      "step": 926
+    },
+    {
+      "epoch": 3.982229402261712,
+      "grad_norm": 0.7079669833183289,
+      "learning_rate": 6.551522846769067e-06,
+      "loss": 0.1826,
+      "step": 927
+    },
+    {
+      "epoch": 3.986537425955843,
+      "grad_norm": 0.7876244187355042,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.1795,
+      "step": 928
+    },
+    {
+      "epoch": 3.990845449649973,
+      "grad_norm": 0.7773596048355103,
+      "learning_rate": 6.538644263787948e-06,
+      "loss": 0.1932,
+      "step": 929
+    },
+    {
+      "epoch": 3.9951534733441036,
+      "grad_norm": 0.744175374507904,
+      "learning_rate": 6.532200734318896e-06,
+      "loss": 0.1785,
+      "step": 930
+    },
+    {
+      "epoch": 3.999461497038234,
+      "grad_norm": 0.6581542491912842,
+      "learning_rate": 6.525754395282961e-06,
+      "loss": 0.1776,
+      "step": 931
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.6581542491912842,
+      "learning_rate": 6.5193052585006666e-06,
+      "loss": 0.1884,
+      "step": 932
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2320,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 10751167365120.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}