diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,29526 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 21063,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00023738308882875184,
+      "grad_norm": 6.03125,
+      "learning_rate": 2.3719165085388994e-07,
+      "loss": 0.993,
+      "step": 5
+    },
+    {
+      "epoch": 0.00047476617765750367,
+      "grad_norm": 15.9375,
+      "learning_rate": 4.743833017077799e-07,
+      "loss": 0.9493,
+      "step": 10
+    },
+    {
+      "epoch": 0.0007121492664862555,
+      "grad_norm": 19.875,
+      "learning_rate": 7.115749525616699e-07,
+      "loss": 0.9772,
+      "step": 15
+    },
+    {
+      "epoch": 0.0009495323553150073,
+      "grad_norm": 12.1875,
+      "learning_rate": 9.487666034155598e-07,
+      "loss": 0.867,
+      "step": 20
+    },
+    {
+      "epoch": 0.0011869154441437593,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1859582542694498e-06,
+      "loss": 0.9284,
+      "step": 25
+    },
+    {
+      "epoch": 0.001424298532972511,
+      "grad_norm": 16.0,
+      "learning_rate": 1.4231499051233397e-06,
+      "loss": 0.8649,
+      "step": 30
+    },
+    {
+      "epoch": 0.001661681621801263,
+      "grad_norm": 10.875,
+      "learning_rate": 1.6603415559772294e-06,
+      "loss": 0.9028,
+      "step": 35
+    },
+    {
+      "epoch": 0.0018990647106300147,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8975332068311195e-06,
+      "loss": 0.8898,
+      "step": 40
+    },
+    {
+      "epoch": 0.0021364477994587666,
+      "grad_norm": 9.75,
+      "learning_rate": 2.1347248576850095e-06,
+      "loss": 0.892,
+      "step": 45
+    },
+    {
+      "epoch": 0.0023738308882875186,
+      "grad_norm": 9.8125,
+      "learning_rate": 2.3719165085388996e-06,
+      "loss": 0.8617,
+      "step": 50
+    },
+    {
+      "epoch": 0.00261121397711627,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.6091081593927897e-06,
+      "loss": 0.8693,
+      "step": 55
+    },
+    {
+      "epoch": 0.002848597065945022,
+      "grad_norm": 8.5,
+      "learning_rate": 2.8462998102466794e-06,
+      "loss": 0.7502,
+      "step": 60
+    },
+    {
+      "epoch": 0.003085980154773774,
+      "grad_norm": 4.125,
+      "learning_rate": 3.0834914611005695e-06,
+      "loss": 0.7925,
+      "step": 65
+    },
+    {
+      "epoch": 0.003323363243602526,
+      "grad_norm": 4.40625,
+      "learning_rate": 3.320683111954459e-06,
+      "loss": 0.7839,
+      "step": 70
+    },
+    {
+      "epoch": 0.0035607463324312774,
+      "grad_norm": 8.125,
+      "learning_rate": 3.557874762808349e-06,
+      "loss": 0.7424,
+      "step": 75
+    },
+    {
+      "epoch": 0.0037981294212600294,
+      "grad_norm": 3.625,
+      "learning_rate": 3.795066413662239e-06,
+      "loss": 0.706,
+      "step": 80
+    },
+    {
+      "epoch": 0.004035512510088781,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.032258064516129e-06,
+      "loss": 0.7192,
+      "step": 85
+    },
+    {
+      "epoch": 0.004272895598917533,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.269449715370019e-06,
+      "loss": 0.6576,
+      "step": 90
+    },
+    {
+      "epoch": 0.004510278687746285,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.5066413662239095e-06,
+      "loss": 0.6379,
+      "step": 95
+    },
+    {
+      "epoch": 0.004747661776575037,
+      "grad_norm": 3.703125,
+      "learning_rate": 4.743833017077799e-06,
+      "loss": 0.627,
+      "step": 100
+    },
+    {
+      "epoch": 0.004985044865403789,
+      "grad_norm": 1.890625,
+      "learning_rate": 4.981024667931689e-06,
+      "loss": 0.5905,
+      "step": 105
+    },
+    {
+      "epoch": 0.00522242795423254,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.218216318785579e-06,
+      "loss": 0.5738,
+      "step": 110
+    },
+    {
+      "epoch": 0.0054598110430612926,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.455407969639469e-06,
+      "loss": 0.5512,
+      "step": 115
+    },
+    {
+      "epoch": 0.005697194131890044,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.692599620493359e-06,
+      "loss": 0.5384,
+      "step": 120
+    },
+    {
+      "epoch": 0.005934577220718796,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.9297912713472485e-06,
+      "loss": 0.518,
+      "step": 125
+    },
+    {
+      "epoch": 0.006171960309547548,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.166982922201139e-06,
+      "loss": 0.5024,
+      "step": 130
+    },
+    {
+      "epoch": 0.0064093433983762995,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.404174573055029e-06,
+      "loss": 0.4808,
+      "step": 135
+    },
+    {
+      "epoch": 0.006646726487205052,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.641366223908918e-06,
+      "loss": 0.4742,
+      "step": 140
+    },
+    {
+      "epoch": 0.006884109576033803,
+      "grad_norm": 3.421875,
+      "learning_rate": 6.878557874762809e-06,
+      "loss": 0.474,
+      "step": 145
+    },
+    {
+      "epoch": 0.007121492664862555,
+      "grad_norm": 1.390625,
+      "learning_rate": 7.115749525616698e-06,
+      "loss": 0.4557,
+      "step": 150
+    },
+    {
+      "epoch": 0.007358875753691307,
+      "grad_norm": 3.796875,
+      "learning_rate": 7.3529411764705884e-06,
+      "loss": 0.4491,
+      "step": 155
+    },
+    {
+      "epoch": 0.007596258842520059,
+      "grad_norm": 1.515625,
+      "learning_rate": 7.590132827324478e-06,
+      "loss": 0.4419,
+      "step": 160
+    },
+    {
+      "epoch": 0.007833641931348811,
+      "grad_norm": 2.109375,
+      "learning_rate": 7.827324478178369e-06,
+      "loss": 0.4351,
+      "step": 165
+    },
+    {
+      "epoch": 0.008071025020177562,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.4319,
+      "step": 170
+    },
+    {
+      "epoch": 0.008308408109006314,
+      "grad_norm": 8.25,
+      "learning_rate": 8.30170777988615e-06,
+      "loss": 0.4297,
+      "step": 175
+    },
+    {
+      "epoch": 0.008545791197835067,
+      "grad_norm": 1.2890625,
+      "learning_rate": 8.538899430740038e-06,
+      "loss": 0.4251,
+      "step": 180
+    },
+    {
+      "epoch": 0.008783174286663819,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.776091081593928e-06,
+      "loss": 0.418,
+      "step": 185
+    },
+    {
+      "epoch": 0.00902055737549257,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.013282732447819e-06,
+      "loss": 0.415,
+      "step": 190
+    },
+    {
+      "epoch": 0.009257940464321322,
+      "grad_norm": 5.8125,
+      "learning_rate": 9.250474383301707e-06,
+      "loss": 0.4173,
+      "step": 195
+    },
+    {
+      "epoch": 0.009495323553150074,
+      "grad_norm": 57.75,
+      "learning_rate": 9.487666034155598e-06,
+      "loss": 0.4148,
+      "step": 200
+    },
+    {
+      "epoch": 0.009732706641978825,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.724857685009488e-06,
+      "loss": 0.4059,
+      "step": 205
+    },
+    {
+      "epoch": 0.009970089730807577,
+      "grad_norm": 4.34375,
+      "learning_rate": 9.962049335863378e-06,
+      "loss": 0.4042,
+      "step": 210
+    },
+    {
+      "epoch": 0.01020747281963633,
+      "grad_norm": 3.75,
+      "learning_rate": 1.0199240986717267e-05,
+      "loss": 0.4064,
+      "step": 215
+    },
+    {
+      "epoch": 0.01044485590846508,
+      "grad_norm": 6.15625,
+      "learning_rate": 1.0436432637571159e-05,
+      "loss": 0.402,
+      "step": 220
+    },
+    {
+      "epoch": 0.010682238997293833,
+      "grad_norm": 10.5,
+      "learning_rate": 1.0673624288425049e-05,
+      "loss": 0.399,
+      "step": 225
+    },
+    {
+      "epoch": 0.010919622086122585,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.0910815939278938e-05,
+      "loss": 0.3985,
+      "step": 230
+    },
+    {
+      "epoch": 0.011157005174951336,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1148007590132828e-05,
+      "loss": 0.3927,
+      "step": 235
+    },
+    {
+      "epoch": 0.011394388263780088,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.1385199240986718e-05,
+      "loss": 0.3848,
+      "step": 240
+    },
+    {
+      "epoch": 0.01163177135260884,
+      "grad_norm": 5.0,
+      "learning_rate": 1.1622390891840607e-05,
+      "loss": 0.3867,
+      "step": 245
+    },
+    {
+      "epoch": 0.011869154441437591,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1859582542694497e-05,
+      "loss": 0.388,
+      "step": 250
+    },
+    {
+      "epoch": 0.012106537530266344,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.2096774193548388e-05,
+      "loss": 0.3836,
+      "step": 255
+    },
+    {
+      "epoch": 0.012343920619095096,
+      "grad_norm": 15.5,
+      "learning_rate": 1.2333965844402278e-05,
+      "loss": 0.3892,
+      "step": 260
+    },
+    {
+      "epoch": 0.012581303707923848,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.2571157495256166e-05,
+      "loss": 0.3915,
+      "step": 265
+    },
+    {
+      "epoch": 0.012818686796752599,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.2808349146110058e-05,
+      "loss": 0.3791,
+      "step": 270
+    },
+    {
+      "epoch": 0.013056069885581351,
+      "grad_norm": 4.375,
+      "learning_rate": 1.3045540796963949e-05,
+      "loss": 0.378,
+      "step": 275
+    },
+    {
+      "epoch": 0.013293452974410104,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3282732447817835e-05,
+      "loss": 0.3747,
+      "step": 280
+    },
+    {
+      "epoch": 0.013530836063238854,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.3519924098671727e-05,
+      "loss": 0.3721,
+      "step": 285
+    },
+    {
+      "epoch": 0.013768219152067607,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.3757115749525618e-05,
+      "loss": 0.3664,
+      "step": 290
+    },
+    {
+      "epoch": 0.014005602240896359,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.3994307400379508e-05,
+      "loss": 0.3654,
+      "step": 295
+    },
+    {
+      "epoch": 0.01424298532972511,
+      "grad_norm": 3.359375,
+      "learning_rate": 1.4231499051233396e-05,
+      "loss": 0.3587,
+      "step": 300
+    },
+    {
+      "epoch": 0.014480368418553862,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.4468690702087287e-05,
+      "loss": 0.3565,
+      "step": 305
+    },
+    {
+      "epoch": 0.014717751507382614,
+      "grad_norm": 2.25,
+      "learning_rate": 1.4705882352941177e-05,
+      "loss": 0.3563,
+      "step": 310
+    },
+    {
+      "epoch": 0.014955134596211365,
+      "grad_norm": 4.96875,
+      "learning_rate": 1.4943074003795068e-05,
+      "loss": 0.3539,
+      "step": 315
+    },
+    {
+      "epoch": 0.015192517685040118,
+      "grad_norm": 2.5,
+      "learning_rate": 1.5180265654648956e-05,
+      "loss": 0.3512,
+      "step": 320
+    },
+    {
+      "epoch": 0.01542990077386887,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.5417457305502848e-05,
+      "loss": 0.3454,
+      "step": 325
+    },
+    {
+      "epoch": 0.015667283862697622,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.5654648956356737e-05,
+      "loss": 0.3439,
+      "step": 330
+    },
+    {
+      "epoch": 0.015904666951526373,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5891840607210627e-05,
+      "loss": 0.3427,
+      "step": 335
+    },
+    {
+      "epoch": 0.016142050040355124,
+      "grad_norm": 2.0,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.3417,
+      "step": 340
+    },
+    {
+      "epoch": 0.016379433129183878,
+      "grad_norm": 4.375,
+      "learning_rate": 1.6366223908918406e-05,
+      "loss": 0.3331,
+      "step": 345
+    },
+    {
+      "epoch": 0.01661681621801263,
+      "grad_norm": 5.25,
+      "learning_rate": 1.66034155597723e-05,
+      "loss": 0.338,
+      "step": 350
+    },
+    {
+      "epoch": 0.01685419930684138,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.6840607210626186e-05,
+      "loss": 0.3295,
+      "step": 355
+    },
+    {
+      "epoch": 0.017091582395670133,
+      "grad_norm": 14.875,
+      "learning_rate": 1.7077798861480076e-05,
+      "loss": 0.3363,
+      "step": 360
+    },
+    {
+      "epoch": 0.017328965484498884,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.731499051233397e-05,
+      "loss": 0.3254,
+      "step": 365
+    },
+    {
+      "epoch": 0.017566348573327638,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.7552182163187855e-05,
+      "loss": 0.3201,
+      "step": 370
+    },
+    {
+      "epoch": 0.01780373166215639,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.7789373814041745e-05,
+      "loss": 0.3217,
+      "step": 375
+    },
+    {
+      "epoch": 0.01804111475098514,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.8026565464895638e-05,
+      "loss": 0.3203,
+      "step": 380
+    },
+    {
+      "epoch": 0.018278497839813893,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.8263757115749528e-05,
+      "loss": 0.3119,
+      "step": 385
+    },
+    {
+      "epoch": 0.018515880928642644,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8500948766603414e-05,
+      "loss": 0.3157,
+      "step": 390
+    },
+    {
+      "epoch": 0.018753264017471395,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8738140417457307e-05,
+      "loss": 0.3172,
+      "step": 395
+    },
+    {
+      "epoch": 0.01899064710630015,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.8975332068311197e-05,
+      "loss": 0.3157,
+      "step": 400
+    },
+    {
+      "epoch": 0.0192280301951289,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9212523719165086e-05,
+      "loss": 0.3091,
+      "step": 405
+    },
+    {
+      "epoch": 0.01946541328395765,
+      "grad_norm": 27.0,
+      "learning_rate": 1.9449715370018976e-05,
+      "loss": 0.3054,
+      "step": 410
+    },
+    {
+      "epoch": 0.019702796372786404,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.9686907020872866e-05,
+      "loss": 0.3147,
+      "step": 415
+    },
+    {
+      "epoch": 0.019940179461615155,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.9924098671726755e-05,
+      "loss": 0.3029,
+      "step": 420
+    },
+    {
+      "epoch": 0.020177562550443905,
+      "grad_norm": 1.625,
+      "learning_rate": 2.0161290322580645e-05,
+      "loss": 0.3023,
+      "step": 425
+    },
+    {
+      "epoch": 0.02041494563927266,
+      "grad_norm": 1.6171875,
+      "learning_rate": 2.0398481973434535e-05,
+      "loss": 0.3032,
+      "step": 430
+    },
+    {
+      "epoch": 0.02065232872810141,
+      "grad_norm": 1.4921875,
+      "learning_rate": 2.0635673624288428e-05,
+      "loss": 0.2973,
+      "step": 435
+    },
+    {
+      "epoch": 0.02088971181693016,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.0872865275142318e-05,
+      "loss": 0.2932,
+      "step": 440
+    },
+    {
+      "epoch": 0.021127094905758915,
+      "grad_norm": 1.34375,
+      "learning_rate": 2.1110056925996204e-05,
+      "loss": 0.2915,
+      "step": 445
+    },
+    {
+      "epoch": 0.021364477994587665,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.1347248576850097e-05,
+      "loss": 0.2929,
+      "step": 450
+    },
+    {
+      "epoch": 0.021601861083416416,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.1584440227703987e-05,
+      "loss": 0.2883,
+      "step": 455
+    },
+    {
+      "epoch": 0.02183924417224517,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.1821631878557876e-05,
+      "loss": 0.2882,
+      "step": 460
+    },
+    {
+      "epoch": 0.02207662726107392,
+      "grad_norm": 1.7890625,
+      "learning_rate": 2.2058823529411766e-05,
+      "loss": 0.2876,
+      "step": 465
+    },
+    {
+      "epoch": 0.02231401034990267,
+      "grad_norm": 1.65625,
+      "learning_rate": 2.2296015180265656e-05,
+      "loss": 0.2888,
+      "step": 470
+    },
+    {
+      "epoch": 0.022551393438731426,
+      "grad_norm": 2.28125,
+      "learning_rate": 2.2533206831119546e-05,
+      "loss": 0.2841,
+      "step": 475
+    },
+    {
+      "epoch": 0.022788776527560176,
+      "grad_norm": 1.3984375,
+      "learning_rate": 2.2770398481973435e-05,
+      "loss": 0.284,
+      "step": 480
+    },
+    {
+      "epoch": 0.023026159616388927,
+      "grad_norm": 2.1875,
+      "learning_rate": 2.3007590132827325e-05,
+      "loss": 0.2825,
+      "step": 485
+    },
+    {
+      "epoch": 0.02326354270521768,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.3244781783681215e-05,
+      "loss": 0.2815,
+      "step": 490
+    },
+    {
+      "epoch": 0.02350092579404643,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.3481973434535108e-05,
+      "loss": 0.2851,
+      "step": 495
+    },
+    {
+      "epoch": 0.023738308882875182,
+      "grad_norm": 1.6328125,
+      "learning_rate": 2.3719165085388994e-05,
+      "loss": 0.2821,
+      "step": 500
+    },
+    {
+      "epoch": 0.023975691971703936,
+      "grad_norm": 1.3046875,
+      "learning_rate": 2.3956356736242884e-05,
+      "loss": 0.2736,
+      "step": 505
+    },
+    {
+      "epoch": 0.024213075060532687,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.2786,
+      "step": 510
+    },
+    {
+      "epoch": 0.02445045814936144,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.4430740037950663e-05,
+      "loss": 0.2786,
+      "step": 515
+    },
+    {
+      "epoch": 0.024687841238190192,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.4667931688804556e-05,
+      "loss": 0.2805,
+      "step": 520
+    },
+    {
+      "epoch": 0.024925224327018942,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.4905123339658446e-05,
+      "loss": 0.2755,
+      "step": 525
+    },
+    {
+      "epoch": 0.025162607415847697,
+      "grad_norm": 1.25,
+      "learning_rate": 2.5142314990512332e-05,
+      "loss": 0.2753,
+      "step": 530
+    },
+    {
+      "epoch": 0.025399990504676447,
+      "grad_norm": 1.640625,
+      "learning_rate": 2.5379506641366225e-05,
+      "loss": 0.2734,
+      "step": 535
+    },
+    {
+      "epoch": 0.025637373593505198,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.5616698292220115e-05,
+      "loss": 0.2695,
+      "step": 540
+    },
+    {
+      "epoch": 0.025874756682333952,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.5853889943074005e-05,
+      "loss": 0.2759,
+      "step": 545
+    },
+    {
+      "epoch": 0.026112139771162703,
+      "grad_norm": 3.0,
+      "learning_rate": 2.6091081593927898e-05,
+      "loss": 0.2725,
+      "step": 550
+    },
+    {
+      "epoch": 0.026349522859991453,
+      "grad_norm": 2.0625,
+      "learning_rate": 2.6328273244781788e-05,
+      "loss": 0.2766,
+      "step": 555
+    },
+    {
+      "epoch": 0.026586905948820207,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.656546489563567e-05,
+      "loss": 0.2722,
+      "step": 560
+    },
+    {
+      "epoch": 0.026824289037648958,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.6802656546489564e-05,
+      "loss": 0.2738,
+      "step": 565
+    },
+    {
+      "epoch": 0.02706167212647771,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.7039848197343453e-05,
+      "loss": 0.2722,
+      "step": 570
+    },
+    {
+      "epoch": 0.027299055215306463,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.7277039848197343e-05,
+      "loss": 0.2638,
+      "step": 575
+    },
+    {
+      "epoch": 0.027536438304135213,
+      "grad_norm": 1.328125,
+      "learning_rate": 2.7514231499051236e-05,
+      "loss": 0.2732,
+      "step": 580
+    },
+    {
+      "epoch": 0.027773821392963964,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.7751423149905126e-05,
+      "loss": 0.2713,
+      "step": 585
+    },
+    {
+      "epoch": 0.028011204481792718,
+      "grad_norm": 1.859375,
+      "learning_rate": 2.7988614800759016e-05,
+      "loss": 0.2721,
+      "step": 590
+    },
+    {
+      "epoch": 0.02824858757062147,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.822580645161291e-05,
+      "loss": 0.2693,
+      "step": 595
+    },
+    {
+      "epoch": 0.02848597065945022,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.846299810246679e-05,
+      "loss": 0.264,
+      "step": 600
+    },
+    {
+      "epoch": 0.028723353748278974,
+      "grad_norm": 1.234375,
+      "learning_rate": 2.8700189753320685e-05,
+      "loss": 0.2698,
+      "step": 605
+    },
+    {
+      "epoch": 0.028960736837107724,
+      "grad_norm": 2.75,
+      "learning_rate": 2.8937381404174574e-05,
+      "loss": 0.2677,
+      "step": 610
+    },
+    {
+      "epoch": 0.029198119925936475,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.9174573055028464e-05,
+      "loss": 0.2684,
+      "step": 615
+    },
+    {
+      "epoch": 0.02943550301476523,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.9411764705882354e-05,
+      "loss": 0.2652,
+      "step": 620
+    },
+    {
+      "epoch": 0.02967288610359398,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.9648956356736247e-05,
+      "loss": 0.2622,
+      "step": 625
+    },
+    {
+      "epoch": 0.02991026919242273,
+      "grad_norm": 1.4453125,
+      "learning_rate": 2.9886148007590137e-05,
+      "loss": 0.2676,
+      "step": 630
+    },
+    {
+      "epoch": 0.030147652281251484,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.0123339658444023e-05,
+      "loss": 0.2665,
+      "step": 635
+    },
+    {
+      "epoch": 0.030385035370080235,
+      "grad_norm": 1.375,
+      "learning_rate": 3.0360531309297913e-05,
+      "loss": 0.2595,
+      "step": 640
+    },
+    {
+      "epoch": 0.030622418458908986,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.0597722960151806e-05,
+      "loss": 0.2637,
+      "step": 645
+    },
+    {
+      "epoch": 0.03085980154773774,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.0834914611005695e-05,
+      "loss": 0.2647,
+      "step": 650
+    },
+    {
+      "epoch": 0.03109718463656649,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.1072106261859585e-05,
+      "loss": 0.2639,
+      "step": 655
+    },
+    {
+      "epoch": 0.031334567725395245,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.1309297912713475e-05,
+      "loss": 0.2638,
+      "step": 660
+    },
+    {
+      "epoch": 0.031571950814223995,
+      "grad_norm": 1.5,
+      "learning_rate": 3.1546489563567364e-05,
+      "loss": 0.2617,
+      "step": 665
+    },
+    {
+      "epoch": 0.031809333903052746,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.1783681214421254e-05,
+      "loss": 0.2634,
+      "step": 670
+    },
+    {
+      "epoch": 0.032046716991881496,
+      "grad_norm": 1.46875,
+      "learning_rate": 3.2020872865275144e-05,
+      "loss": 0.26,
+      "step": 675
+    },
+    {
+      "epoch": 0.03228410008071025,
+      "grad_norm": 2.0,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.2587,
+      "step": 680
+    },
+    {
+      "epoch": 0.032521483169539005,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.249525616698292e-05,
+      "loss": 0.2663,
+      "step": 685
+    },
+    {
+      "epoch": 0.032758866258367755,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.273244781783681e-05,
+      "loss": 0.2622,
+      "step": 690
+    },
+    {
+      "epoch": 0.032996249347196506,
+      "grad_norm": 1.3671875,
+      "learning_rate": 3.29696394686907e-05,
+      "loss": 0.2576,
+      "step": 695
+    },
+    {
+      "epoch": 0.03323363243602526,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.32068311195446e-05,
+      "loss": 0.2614,
+      "step": 700
+    },
+    {
+      "epoch": 0.03347101552485401,
+      "grad_norm": 1.609375,
+      "learning_rate": 3.344402277039848e-05,
+      "loss": 0.2586,
+      "step": 705
+    },
+    {
+      "epoch": 0.03370839861368276,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.368121442125237e-05,
+      "loss": 0.2617,
+      "step": 710
+    },
+    {
+      "epoch": 0.033945781702511515,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.391840607210626e-05,
+      "loss": 0.2608,
+      "step": 715
+    },
+    {
+      "epoch": 0.034183164791340266,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.415559772296015e-05,
+      "loss": 0.2566,
+      "step": 720
+    },
+    {
+      "epoch": 0.03442054788016902,
+      "grad_norm": 1.7109375,
+      "learning_rate": 3.439278937381404e-05,
+      "loss": 0.256,
+      "step": 725
+    },
+    {
+      "epoch": 0.03465793096899777,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.462998102466794e-05,
+      "loss": 0.2636,
+      "step": 730
+    },
+    {
+      "epoch": 0.03489531405782652,
+      "grad_norm": 2.0,
+      "learning_rate": 3.486717267552183e-05,
+      "loss": 0.2573,
+      "step": 735
+    },
+    {
+      "epoch": 0.035132697146655276,
+      "grad_norm": 1.6796875,
+      "learning_rate": 3.510436432637571e-05,
+      "loss": 0.2541,
+      "step": 740
+    },
+    {
+      "epoch": 0.035370080235484026,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.53415559772296e-05,
+      "loss": 0.2561,
+      "step": 745
+    },
+    {
+      "epoch": 0.03560746332431278,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.557874762808349e-05,
+      "loss": 0.258,
+      "step": 750
+    },
+    {
+      "epoch": 0.03584484641314153,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.5815939278937386e-05,
+      "loss": 0.2513,
+      "step": 755
+    },
+    {
+      "epoch": 0.03608222950197028,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.6053130929791276e-05,
+      "loss": 0.2604,
+      "step": 760
+    },
+    {
+      "epoch": 0.03631961259079903,
+      "grad_norm": 1.8125,
+      "learning_rate": 3.6290322580645165e-05,
+      "loss": 0.2602,
+      "step": 765
+    },
+    {
+      "epoch": 0.036556995679627786,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.6527514231499055e-05,
+      "loss": 0.2526,
+      "step": 770
+    },
+    {
+      "epoch": 0.03679437876845654,
+      "grad_norm": 1.3125,
+      "learning_rate": 3.6764705882352945e-05,
+      "loss": 0.2548,
+      "step": 775
+    },
+    {
+      "epoch": 0.03703176185728529,
+      "grad_norm": 1.2265625,
+      "learning_rate": 3.700189753320683e-05,
+      "loss": 0.2536,
+      "step": 780
+    },
+    {
+      "epoch": 0.03726914494611404,
+      "grad_norm": 1.7265625,
+      "learning_rate": 3.7239089184060724e-05,
+      "loss": 0.2559,
+      "step": 785
+    },
+    {
+      "epoch": 0.03750652803494279,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.7476280834914614e-05,
+      "loss": 0.2551,
+      "step": 790
+    },
+    {
+      "epoch": 0.03774391112377154,
+      "grad_norm": 1.6640625,
+      "learning_rate": 3.7713472485768504e-05,
+      "loss": 0.2518,
+      "step": 795
+    },
+    {
+      "epoch": 0.0379812942126003,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.795066413662239e-05,
+      "loss": 0.2523,
+      "step": 800
+    },
+    {
+      "epoch": 0.03821867730142905,
+      "grad_norm": 1.2734375,
+      "learning_rate": 3.818785578747628e-05,
+      "loss": 0.2541,
+      "step": 805
+    },
+    {
+      "epoch": 0.0384560603902578,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.842504743833017e-05,
+      "loss": 0.2558,
+      "step": 810
+    },
+    {
+      "epoch": 0.03869344347908655,
+      "grad_norm": 1.421875,
+      "learning_rate": 3.866223908918406e-05,
+      "loss": 0.2529,
+      "step": 815
+    },
+    {
+      "epoch": 0.0389308265679153,
+      "grad_norm": 1.0,
+      "learning_rate": 3.889943074003795e-05,
+      "loss": 0.2549,
+      "step": 820
+    },
+    {
+      "epoch": 0.03916820965674405,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.913662239089184e-05,
+      "loss": 0.2501,
+      "step": 825
+    },
+    {
+      "epoch": 0.03940559274557281,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.937381404174573e-05,
+      "loss": 0.2542,
+      "step": 830
+    },
+    {
+      "epoch": 0.03964297583440156,
+      "grad_norm": 3.515625,
+      "learning_rate": 3.961100569259962e-05,
+      "loss": 0.2523,
+      "step": 835
+    },
+    {
+      "epoch": 0.03988035892323031,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.984819734345351e-05,
+      "loss": 0.2507,
+      "step": 840
+    },
+    {
+      "epoch": 0.04011774201205906,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.008538899430741e-05,
+      "loss": 0.25,
+      "step": 845
+    },
+    {
+      "epoch": 0.04035512510088781,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.032258064516129e-05,
+      "loss": 0.2536,
+      "step": 850
+    },
+    {
+      "epoch": 0.04059250818971656,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.055977229601518e-05,
+      "loss": 0.2566,
+      "step": 855
+    },
+    {
+      "epoch": 0.04082989127854532,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.079696394686907e-05,
+      "loss": 0.2488,
+      "step": 860
+    },
+    {
+      "epoch": 0.04106727436737407,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.103415559772296e-05,
+      "loss": 0.253,
+      "step": 865
+    },
+    {
+      "epoch": 0.04130465745620282,
+      "grad_norm": 1.3515625,
+      "learning_rate": 4.1271347248576856e-05,
+      "loss": 0.2486,
+      "step": 870
+    },
+    {
+      "epoch": 0.04154204054503157,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.1508538899430746e-05,
+      "loss": 0.2519,
+      "step": 875
+    },
+    {
+      "epoch": 0.04177942363386032,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.1745730550284635e-05,
+      "loss": 0.2524,
+      "step": 880
+    },
+    {
+      "epoch": 0.04201680672268908,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.198292220113852e-05,
+      "loss": 0.2548,
+      "step": 885
+    },
+    {
+      "epoch": 0.04225418981151783,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.222011385199241e-05,
+      "loss": 0.252,
+      "step": 890
+    },
+    {
+      "epoch": 0.04249157290034658,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.24573055028463e-05,
+      "loss": 0.2532,
+      "step": 895
+    },
+    {
+      "epoch": 0.04272895598917533,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.2694497153700194e-05,
+      "loss": 0.2519,
+      "step": 900
+    },
+    {
+      "epoch": 0.04296633907800408,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.2931688804554084e-05,
+      "loss": 0.25,
+      "step": 905
+    },
+    {
+      "epoch": 0.04320372216683283,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.3168880455407974e-05,
+      "loss": 0.2497,
+      "step": 910
+    },
+    {
+      "epoch": 0.04344110525566159,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.340607210626186e-05,
+      "loss": 0.2502,
+      "step": 915
+    },
+    {
+      "epoch": 0.04367848834449034,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.364326375711575e-05,
+      "loss": 0.2532,
+      "step": 920
+    },
+    {
+      "epoch": 0.04391587143331909,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.388045540796964e-05,
+      "loss": 0.2477,
+      "step": 925
+    },
+    {
+      "epoch": 0.04415325452214784,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.411764705882353e-05,
+      "loss": 0.2499,
+      "step": 930
+    },
+    {
+      "epoch": 0.04439063761097659,
+      "grad_norm": 1.25,
+      "learning_rate": 4.435483870967742e-05,
+      "loss": 0.2453,
+      "step": 935
+    },
+    {
+      "epoch": 0.04462802069980534,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.459203036053131e-05,
+      "loss": 0.2523,
+      "step": 940
+    },
+    {
+      "epoch": 0.0448654037886341,
+      "grad_norm": 1.5,
+      "learning_rate": 4.48292220113852e-05,
+      "loss": 0.249,
+      "step": 945
+    },
+    {
+      "epoch": 0.04510278687746285,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.506641366223909e-05,
+      "loss": 0.2482,
+      "step": 950
+    },
+    {
+      "epoch": 0.0453401699662916,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.530360531309298e-05,
+      "loss": 0.248,
+      "step": 955
+    },
+    {
+      "epoch": 0.04557755305512035,
+      "grad_norm": 1.3203125,
+      "learning_rate": 4.554079696394687e-05,
+      "loss": 0.2487,
+      "step": 960
+    },
+    {
+      "epoch": 0.0458149361439491,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.577798861480076e-05,
+      "loss": 0.2479,
+      "step": 965
+    },
+    {
+      "epoch": 0.046052319232777854,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.601518026565465e-05,
+      "loss": 0.2479,
+      "step": 970
+    },
+    {
+      "epoch": 0.04628970232160661,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.625237191650854e-05,
+      "loss": 0.2423,
+      "step": 975
+    },
+    {
+      "epoch": 0.04652708541043536,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.648956356736243e-05,
+      "loss": 0.2474,
+      "step": 980
+    },
+    {
+      "epoch": 0.04676446849926411,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.6726755218216326e-05,
+      "loss": 0.2494,
+      "step": 985
+    },
+    {
+      "epoch": 0.04700185158809286,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.6963946869070216e-05,
+      "loss": 0.2464,
+      "step": 990
+    },
+    {
+      "epoch": 0.047239234676921614,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.72011385199241e-05,
+      "loss": 0.2495,
+      "step": 995
+    },
+    {
+      "epoch": 0.047476617765750365,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.743833017077799e-05,
+      "loss": 0.251,
+      "step": 1000
+    },
+    {
+      "epoch": 0.04771400085457912,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.767552182163188e-05,
+      "loss": 0.2509,
+      "step": 1005
+    },
+    {
+      "epoch": 0.04795138394340787,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.791271347248577e-05,
+      "loss": 0.2444,
+      "step": 1010
+    },
+    {
+      "epoch": 0.048188767032236623,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.8149905123339664e-05,
+      "loss": 0.2512,
+      "step": 1015
+    },
+    {
+      "epoch": 0.048426150121065374,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.8387096774193554e-05,
+      "loss": 0.2437,
+      "step": 1020
+    },
+    {
+      "epoch": 0.048663533209894125,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.8624288425047443e-05,
+      "loss": 0.2486,
+      "step": 1025
+    },
+    {
+      "epoch": 0.04890091629872288,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.8861480075901326e-05,
+      "loss": 0.2476,
+      "step": 1030
+    },
+    {
+      "epoch": 0.04913829938755163,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.9098671726755216e-05,
+      "loss": 0.2467,
+      "step": 1035
+    },
+    {
+      "epoch": 0.049375682476380384,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.933586337760911e-05,
+      "loss": 0.2447,
+      "step": 1040
+    },
+    {
+      "epoch": 0.049613065565209134,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.9573055028463e-05,
+      "loss": 0.2482,
+      "step": 1045
+    },
+    {
+      "epoch": 0.049850448654037885,
+      "grad_norm": 1.0,
+      "learning_rate": 4.981024667931689e-05,
+      "loss": 0.2468,
+      "step": 1050
+    },
+    {
+      "epoch": 0.050087831742866636,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.9999999722667036e-05,
+      "loss": 0.245,
+      "step": 1055
+    },
+    {
+      "epoch": 0.05032521483169539,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.999999001601389e-05,
+      "loss": 0.2473,
+      "step": 1060
+    },
+    {
+      "epoch": 0.050562597920524144,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.999996644271922e-05,
+      "loss": 0.2435,
+      "step": 1065
+    },
+    {
+      "epoch": 0.050799981009352894,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.999992900279755e-05,
+      "loss": 0.2468,
+      "step": 1070
+    },
+    {
+      "epoch": 0.051037364098181645,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.999987769627194e-05,
+      "loss": 0.247,
+      "step": 1075
+    },
+    {
+      "epoch": 0.051274747187010396,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.9999812523174015e-05,
+      "loss": 0.2403,
+      "step": 1080
+    },
+    {
+      "epoch": 0.051512130275839146,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.9999733483543956e-05,
+      "loss": 0.2423,
+      "step": 1085
+    },
+    {
+      "epoch": 0.051749513364667904,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.999964057743047e-05,
+      "loss": 0.2454,
+      "step": 1090
+    },
+    {
+      "epoch": 0.051986896453496655,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.99995338048908e-05,
+      "loss": 0.2477,
+      "step": 1095
+    },
+    {
+      "epoch": 0.052224279542325405,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.999941316599075e-05,
+      "loss": 0.2407,
+      "step": 1100
+    },
+    {
+      "epoch": 0.052461662631154156,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.99992786608047e-05,
+      "loss": 0.2487,
+      "step": 1105
+    },
+    {
+      "epoch": 0.05269904571998291,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.999913028941551e-05,
+      "loss": 0.248,
+      "step": 1110
+    },
+    {
+      "epoch": 0.05293642880881166,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.999896805191463e-05,
+      "loss": 0.25,
+      "step": 1115
+    },
+    {
+      "epoch": 0.053173811897640415,
+      "grad_norm": 0.75,
+      "learning_rate": 4.999879194840206e-05,
+      "loss": 0.2412,
+      "step": 1120
+    },
+    {
+      "epoch": 0.053411194986469165,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.999860197898631e-05,
+      "loss": 0.2481,
+      "step": 1125
+    },
+    {
+      "epoch": 0.053648578075297916,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.9998398143784466e-05,
+      "loss": 0.2418,
+      "step": 1130
+    },
+    {
+      "epoch": 0.05388596116412667,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.9998180442922165e-05,
+      "loss": 0.2485,
+      "step": 1135
+    },
+    {
+      "epoch": 0.05412334425295542,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.999794887653356e-05,
+      "loss": 0.2451,
+      "step": 1140
+    },
+    {
+      "epoch": 0.05436072734178417,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.9997703444761373e-05,
+      "loss": 0.2468,
+      "step": 1145
+    },
+    {
+      "epoch": 0.054598110430612926,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.999744414775685e-05,
+      "loss": 0.2445,
+      "step": 1150
+    },
+    {
+      "epoch": 0.054835493519441676,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.99971709856798e-05,
+      "loss": 0.2433,
+      "step": 1155
+    },
+    {
+      "epoch": 0.05507287660827043,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.9996883958698584e-05,
+      "loss": 0.2463,
+      "step": 1160
+    },
+    {
+      "epoch": 0.05531025969709918,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.9996583066990076e-05,
+      "loss": 0.2443,
+      "step": 1165
+    },
+    {
+      "epoch": 0.05554764278592793,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.999626831073972e-05,
+      "loss": 0.2365,
+      "step": 1170
+    },
+    {
+      "epoch": 0.055785025874756686,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.999593969014151e-05,
+      "loss": 0.2413,
+      "step": 1175
+    },
+    {
+      "epoch": 0.056022408963585436,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.9995597205397966e-05,
+      "loss": 0.2437,
+      "step": 1180
+    },
+    {
+      "epoch": 0.05625979205241419,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.9995240856720146e-05,
+      "loss": 0.2431,
+      "step": 1185
+    },
+    {
+      "epoch": 0.05649717514124294,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.999487064432769e-05,
+      "loss": 0.2423,
+      "step": 1190
+    },
+    {
+      "epoch": 0.05673455823007169,
+      "grad_norm": 0.625,
+      "learning_rate": 4.999448656844874e-05,
+      "loss": 0.2386,
+      "step": 1195
+    },
+    {
+      "epoch": 0.05697194131890044,
+      "grad_norm": 1.125,
+      "learning_rate": 4.999408862932002e-05,
+      "loss": 0.2423,
+      "step": 1200
+    },
+    {
+      "epoch": 0.057209324407729197,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.999367682718675e-05,
+      "loss": 0.2455,
+      "step": 1205
+    },
+    {
+      "epoch": 0.05744670749655795,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.999325116230275e-05,
+      "loss": 0.2453,
+      "step": 1210
+    },
+    {
+      "epoch": 0.0576840905853867,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.999281163493034e-05,
+      "loss": 0.2472,
+      "step": 1215
+    },
+    {
+      "epoch": 0.05792147367421545,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.99923582453404e-05,
+      "loss": 0.2456,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0581588567630442,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.999189099381235e-05,
+      "loss": 0.2411,
+      "step": 1225
+    },
+    {
+      "epoch": 0.05839623985187295,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.999140988063416e-05,
+      "loss": 0.2395,
+      "step": 1230
+    },
+    {
+      "epoch": 0.05863362294070171,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.9990914906102336e-05,
+      "loss": 0.2377,
+      "step": 1235
+    },
+    {
+      "epoch": 0.05887100602953046,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.9990406070521935e-05,
+      "loss": 0.2433,
+      "step": 1240
+    },
+    {
+      "epoch": 0.05910838911835921,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.998988337420654e-05,
+      "loss": 0.2397,
+      "step": 1245
+    },
+    {
+      "epoch": 0.05934577220718796,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.9989346817478286e-05,
+      "loss": 0.2427,
+      "step": 1250
+    },
+    {
+      "epoch": 0.05958315529601671,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.998879640066786e-05,
+      "loss": 0.2375,
+      "step": 1255
+    },
+    {
+      "epoch": 0.05982053838484546,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.998823212411447e-05,
+      "loss": 0.2416,
+      "step": 1260
+    },
+    {
+      "epoch": 0.06005792147367422,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.998765398816589e-05,
+      "loss": 0.2423,
+      "step": 1265
+    },
+    {
+      "epoch": 0.06029530456250297,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.998706199317841e-05,
+      "loss": 0.2411,
+      "step": 1270
+    },
+    {
+      "epoch": 0.06053268765133172,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.9986456139516877e-05,
+      "loss": 0.2403,
+      "step": 1275
+    },
+    {
+      "epoch": 0.06077007074016047,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.998583642755468e-05,
+      "loss": 0.2406,
+      "step": 1280
+    },
+    {
+      "epoch": 0.06100745382898922,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.998520285767375e-05,
+      "loss": 0.2385,
+      "step": 1285
+    },
+    {
+      "epoch": 0.06124483691781797,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.998455543026454e-05,
+      "loss": 0.2385,
+      "step": 1290
+    },
+    {
+      "epoch": 0.06148222000664673,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.9983894145726074e-05,
+      "loss": 0.2369,
+      "step": 1295
+    },
+    {
+      "epoch": 0.06171960309547548,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.9983219004465874e-05,
+      "loss": 0.2384,
+      "step": 1300
+    },
+    {
+      "epoch": 0.06195698618430423,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.998253000690004e-05,
+      "loss": 0.2425,
+      "step": 1305
+    },
+    {
+      "epoch": 0.06219436927313298,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.99818271534532e-05,
+      "loss": 0.2413,
+      "step": 1310
+    },
+    {
+      "epoch": 0.06243175236196173,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.998111044455853e-05,
+      "loss": 0.2405,
+      "step": 1315
+    },
+    {
+      "epoch": 0.06266913545079049,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.9980379880657705e-05,
+      "loss": 0.2368,
+      "step": 1320
+    },
+    {
+      "epoch": 0.06290651853961923,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.9979635462201e-05,
+      "loss": 0.2421,
+      "step": 1325
+    },
+    {
+      "epoch": 0.06314390162844799,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.997887718964717e-05,
+      "loss": 0.2386,
+      "step": 1330
+    },
+    {
+      "epoch": 0.06338128471727675,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.9978105063463546e-05,
+      "loss": 0.2427,
+      "step": 1335
+    },
+    {
+      "epoch": 0.06361866780610549,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.997731908412599e-05,
+      "loss": 0.2418,
+      "step": 1340
+    },
+    {
+      "epoch": 0.06385605089493425,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.997651925211889e-05,
+      "loss": 0.2404,
+      "step": 1345
+    },
+    {
+      "epoch": 0.06409343398376299,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.997570556793519e-05,
+      "loss": 0.2398,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06433081707259175,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.997487803207635e-05,
+      "loss": 0.2388,
+      "step": 1355
+    },
+    {
+      "epoch": 0.0645682001614205,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.997403664505238e-05,
+      "loss": 0.2342,
+      "step": 1360
+    },
+    {
+      "epoch": 0.06480558325024925,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.9973181407381817e-05,
+      "loss": 0.236,
+      "step": 1365
+    },
+    {
+      "epoch": 0.06504296633907801,
+      "grad_norm": 0.75,
+      "learning_rate": 4.9972312319591754e-05,
+      "loss": 0.2405,
+      "step": 1370
+    },
+    {
+      "epoch": 0.06528034942790675,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.99714293822178e-05,
+      "loss": 0.2393,
+      "step": 1375
+    },
+    {
+      "epoch": 0.06551773251673551,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.9970532595804086e-05,
+      "loss": 0.2377,
+      "step": 1380
+    },
+    {
+      "epoch": 0.06575511560556425,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.996962196090332e-05,
+      "loss": 0.2401,
+      "step": 1385
+    },
+    {
+      "epoch": 0.06599249869439301,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.9968697478076725e-05,
+      "loss": 0.2417,
+      "step": 1390
+    },
+    {
+      "epoch": 0.06622988178322177,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.996775914789405e-05,
+      "loss": 0.2416,
+      "step": 1395
+    },
+    {
+      "epoch": 0.06646726487205051,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.9966806970933566e-05,
+      "loss": 0.2352,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06670464796087927,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.9965840947782114e-05,
+      "loss": 0.2374,
+      "step": 1405
+    },
+    {
+      "epoch": 0.06694203104970801,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.996486107903504e-05,
+      "loss": 0.2355,
+      "step": 1410
+    },
+    {
+      "epoch": 0.06717941413853677,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.996386736529625e-05,
+      "loss": 0.2406,
+      "step": 1415
+    },
+    {
+      "epoch": 0.06741679722736552,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.996285980717814e-05,
+      "loss": 0.2421,
+      "step": 1420
+    },
+    {
+      "epoch": 0.06765418031619427,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.9961838405301685e-05,
+      "loss": 0.2412,
+      "step": 1425
+    },
+    {
+      "epoch": 0.06789156340502303,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.996080316029636e-05,
+      "loss": 0.2392,
+      "step": 1430
+    },
+    {
+      "epoch": 0.06812894649385177,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.995975407280019e-05,
+      "loss": 0.2403,
+      "step": 1435
+    },
+    {
+      "epoch": 0.06836632958268053,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.9958691143459716e-05,
+      "loss": 0.2396,
+      "step": 1440
+    },
+    {
+      "epoch": 0.06860371267150928,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.995761437293e-05,
+      "loss": 0.2396,
+      "step": 1445
+    },
+    {
+      "epoch": 0.06884109576033803,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.9956523761874685e-05,
+      "loss": 0.2388,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06907847884916679,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.9955419310965885e-05,
+      "loss": 0.235,
+      "step": 1455
+    },
+    {
+      "epoch": 0.06931586193799553,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.995430102088428e-05,
+      "loss": 0.2371,
+      "step": 1460
+    },
+    {
+      "epoch": 0.06955324502682429,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.995316889231906e-05,
+      "loss": 0.24,
+      "step": 1465
+    },
+    {
+      "epoch": 0.06979062811565304,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.995202292596795e-05,
+      "loss": 0.2384,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0700280112044818,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.9950863122537216e-05,
+      "loss": 0.236,
+      "step": 1475
+    },
+    {
+      "epoch": 0.07026539429331055,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.994968948274163e-05,
+      "loss": 0.236,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0705027773821393,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.99485020073045e-05,
+      "loss": 0.2395,
+      "step": 1485
+    },
+    {
+      "epoch": 0.07074016047096805,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.994730069695766e-05,
+      "loss": 0.2363,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0709775435597968,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.994608555244147e-05,
+      "loss": 0.2332,
+      "step": 1495
+    },
+    {
+      "epoch": 0.07121492664862555,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.9944856574504836e-05,
+      "loss": 0.2363,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0714523097374543,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.994361376390515e-05,
+      "loss": 0.2384,
+      "step": 1505
+    },
+    {
+      "epoch": 0.07168969282628306,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.994235712140837e-05,
+      "loss": 0.2338,
+      "step": 1510
+    },
+    {
+      "epoch": 0.07192707591511181,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.9941086647788935e-05,
+      "loss": 0.2403,
+      "step": 1515
+    },
+    {
+      "epoch": 0.07216445900394056,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.993980234382985e-05,
+      "loss": 0.2378,
+      "step": 1520
+    },
+    {
+      "epoch": 0.07240184209276931,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.993850421032263e-05,
+      "loss": 0.2377,
+      "step": 1525
+    },
+    {
+      "epoch": 0.07263922518159806,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.993719224806729e-05,
+      "loss": 0.2382,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07287660827042682,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.9935866457872405e-05,
+      "loss": 0.2384,
+      "step": 1535
+    },
+    {
+      "epoch": 0.07311399135925557,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.993452684055504e-05,
+      "loss": 0.2382,
+      "step": 1540
+    },
+    {
+      "epoch": 0.07335137444808432,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.99331733969408e-05,
+      "loss": 0.2373,
+      "step": 1545
+    },
+    {
+      "epoch": 0.07358875753691307,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.993180612786381e-05,
+      "loss": 0.2409,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07382614062574182,
+      "grad_norm": 1.0,
+      "learning_rate": 4.993042503416671e-05,
+      "loss": 0.236,
+      "step": 1555
+    },
+    {
+      "epoch": 0.07406352371457058,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.9929030116700647e-05,
+      "loss": 0.2368,
+      "step": 1560
+    },
+    {
+      "epoch": 0.07430090680339932,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.9927621376325324e-05,
+      "loss": 0.2358,
+      "step": 1565
+    },
+    {
+      "epoch": 0.07453828989222808,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.992619881390893e-05,
+      "loss": 0.2389,
+      "step": 1570
+    },
+    {
+      "epoch": 0.07477567298105683,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.992476243032819e-05,
+      "loss": 0.2388,
+      "step": 1575
+    },
+    {
+      "epoch": 0.07501305606988558,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.9923312226468336e-05,
+      "loss": 0.2367,
+      "step": 1580
+    },
+    {
+      "epoch": 0.07525043915871434,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.992184820322312e-05,
+      "loss": 0.2392,
+      "step": 1585
+    },
+    {
+      "epoch": 0.07548782224754308,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.992037036149481e-05,
+      "loss": 0.2342,
+      "step": 1590
+    },
+    {
+      "epoch": 0.07572520533637184,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.991887870219421e-05,
+      "loss": 0.238,
+      "step": 1595
+    },
+    {
+      "epoch": 0.0759625884252006,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.9917373226240607e-05,
+      "loss": 0.238,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07619997151402934,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.991585393456182e-05,
+      "loss": 0.2381,
+      "step": 1605
+    },
+    {
+      "epoch": 0.0764373546028581,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.991432082809419e-05,
+      "loss": 0.2389,
+      "step": 1610
+    },
+    {
+      "epoch": 0.07667473769168684,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.991277390778256e-05,
+      "loss": 0.2316,
+      "step": 1615
+    },
+    {
+      "epoch": 0.0769121207805156,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.991121317458029e-05,
+      "loss": 0.2383,
+      "step": 1620
+    },
+    {
+      "epoch": 0.07714950386934435,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.990963862944924e-05,
+      "loss": 0.2355,
+      "step": 1625
+    },
+    {
+      "epoch": 0.0773868869581731,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.990805027335981e-05,
+      "loss": 0.2365,
+      "step": 1630
+    },
+    {
+      "epoch": 0.07762427004700186,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.9906448107290904e-05,
+      "loss": 0.2381,
+      "step": 1635
+    },
+    {
+      "epoch": 0.0778616531358306,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.990483213222992e-05,
+      "loss": 0.234,
+      "step": 1640
+    },
+    {
+      "epoch": 0.07809903622465936,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.990320234917276e-05,
+      "loss": 0.2314,
+      "step": 1645
+    },
+    {
+      "epoch": 0.0783364193134881,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.990155875912387e-05,
+      "loss": 0.2371,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07857380240231686,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.989990136309619e-05,
+      "loss": 0.2376,
+      "step": 1655
+    },
+    {
+      "epoch": 0.07881118549114562,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.9898230162111154e-05,
+      "loss": 0.2372,
+      "step": 1660
+    },
+    {
+      "epoch": 0.07904856857997436,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.9896545157198726e-05,
+      "loss": 0.2404,
+      "step": 1665
+    },
+    {
+      "epoch": 0.07928595166880312,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.9894846349397365e-05,
+      "loss": 0.2357,
+      "step": 1670
+    },
+    {
+      "epoch": 0.07952333475763186,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.9893133739754024e-05,
+      "loss": 0.24,
+      "step": 1675
+    },
+    {
+      "epoch": 0.07976071784646062,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.9891407329324195e-05,
+      "loss": 0.2319,
+      "step": 1680
+    },
+    {
+      "epoch": 0.07999810093528938,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.988966711917185e-05,
+      "loss": 0.2391,
+      "step": 1685
+    },
+    {
+      "epoch": 0.08023548402411812,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.9887913110369466e-05,
+      "loss": 0.235,
+      "step": 1690
+    },
+    {
+      "epoch": 0.08047286711294688,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.988614530399805e-05,
+      "loss": 0.2316,
+      "step": 1695
+    },
+    {
+      "epoch": 0.08071025020177562,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.988436370114707e-05,
+      "loss": 0.2359,
+      "step": 1700
+    },
+    {
+      "epoch": 0.08094763329060438,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.988256830291453e-05,
+      "loss": 0.236,
+      "step": 1705
+    },
+    {
+      "epoch": 0.08118501637943312,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.988075911040693e-05,
+      "loss": 0.2357,
+      "step": 1710
+    },
+    {
+      "epoch": 0.08142239946826188,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.9878936124739255e-05,
+      "loss": 0.2328,
+      "step": 1715
+    },
+    {
+      "epoch": 0.08165978255709064,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.987709934703502e-05,
+      "loss": 0.2269,
+      "step": 1720
+    },
+    {
+      "epoch": 0.08189716564591938,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.987524877842622e-05,
+      "loss": 0.2338,
+      "step": 1725
+    },
+    {
+      "epoch": 0.08213454873474814,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.987338442005333e-05,
+      "loss": 0.2351,
+      "step": 1730
+    },
+    {
+      "epoch": 0.08237193182357688,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.987150627306535e-05,
+      "loss": 0.2349,
+      "step": 1735
+    },
+    {
+      "epoch": 0.08260931491240564,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.986961433861981e-05,
+      "loss": 0.2348,
+      "step": 1740
+    },
+    {
+      "epoch": 0.0828466980012344,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.986770861788266e-05,
+      "loss": 0.2371,
+      "step": 1745
+    },
+    {
+      "epoch": 0.08308408109006314,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.986578911202841e-05,
+      "loss": 0.237,
+      "step": 1750
+    },
+    {
+      "epoch": 0.0833214641788919,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.986385582224003e-05,
+      "loss": 0.2386,
+      "step": 1755
+    },
+    {
+      "epoch": 0.08355884726772064,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.9861908749709004e-05,
+      "loss": 0.2349,
+      "step": 1760
+    },
+    {
+      "epoch": 0.0837962303565494,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.98599478956353e-05,
+      "loss": 0.2375,
+      "step": 1765
+    },
+    {
+      "epoch": 0.08403361344537816,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.985797326122739e-05,
+      "loss": 0.2333,
+      "step": 1770
+    },
+    {
+      "epoch": 0.0842709965342069,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.985598484770222e-05,
+      "loss": 0.2365,
+      "step": 1775
+    },
+    {
+      "epoch": 0.08450837962303566,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.985398265628526e-05,
+      "loss": 0.2406,
+      "step": 1780
+    },
+    {
+      "epoch": 0.0847457627118644,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.985196668821043e-05,
+      "loss": 0.2353,
+      "step": 1785
+    },
+    {
+      "epoch": 0.08498314580069316,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.984993694472018e-05,
+      "loss": 0.2314,
+      "step": 1790
+    },
+    {
+      "epoch": 0.0852205288895219,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9847893427065415e-05,
+      "loss": 0.2329,
+      "step": 1795
+    },
+    {
+      "epoch": 0.08545791197835066,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.984583613650556e-05,
+      "loss": 0.2371,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08569529506717942,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.98437650743085e-05,
+      "loss": 0.2379,
+      "step": 1805
+    },
+    {
+      "epoch": 0.08593267815600816,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.9841680241750636e-05,
+      "loss": 0.2365,
+      "step": 1810
+    },
+    {
+      "epoch": 0.08617006124483692,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.983958164011683e-05,
+      "loss": 0.2317,
+      "step": 1815
+    },
+    {
+      "epoch": 0.08640744433366566,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.983746927070044e-05,
+      "loss": 0.2383,
+      "step": 1820
+    },
+    {
+      "epoch": 0.08664482742249442,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.9835343134803326e-05,
+      "loss": 0.2335,
+      "step": 1825
+    },
+    {
+      "epoch": 0.08688221051132318,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.98332032337358e-05,
+      "loss": 0.2386,
+      "step": 1830
+    },
+    {
+      "epoch": 0.08711959360015192,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.983104956881667e-05,
+      "loss": 0.2368,
+      "step": 1835
+    },
+    {
+      "epoch": 0.08735697668898068,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.9828882141373245e-05,
+      "loss": 0.2358,
+      "step": 1840
+    },
+    {
+      "epoch": 0.08759435977780942,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.982670095274129e-05,
+      "loss": 0.2373,
+      "step": 1845
+    },
+    {
+      "epoch": 0.08783174286663818,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.982450600426506e-05,
+      "loss": 0.2317,
+      "step": 1850
+    },
+    {
+      "epoch": 0.08806912595546693,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.9822297297297307e-05,
+      "loss": 0.2366,
+      "step": 1855
+    },
+    {
+      "epoch": 0.08830650904429568,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.982007483319923e-05,
+      "loss": 0.2354,
+      "step": 1860
+    },
+    {
+      "epoch": 0.08854389213312444,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.981783861334053e-05,
+      "loss": 0.233,
+      "step": 1865
+    },
+    {
+      "epoch": 0.08878127522195318,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.9815588639099374e-05,
+      "loss": 0.2359,
+      "step": 1870
+    },
+    {
+      "epoch": 0.08901865831078194,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.981332491186242e-05,
+      "loss": 0.2352,
+      "step": 1875
+    },
+    {
+      "epoch": 0.08925604139961069,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.9811047433024794e-05,
+      "loss": 0.2336,
+      "step": 1880
+    },
+    {
+      "epoch": 0.08949342448843944,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.980875620399009e-05,
+      "loss": 0.2335,
+      "step": 1885
+    },
+    {
+      "epoch": 0.0897308075772682,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.980645122617037e-05,
+      "loss": 0.2343,
+      "step": 1890
+    },
+    {
+      "epoch": 0.08996819066609694,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.98041325009862e-05,
+      "loss": 0.235,
+      "step": 1895
+    },
+    {
+      "epoch": 0.0902055737549257,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.98018000298666e-05,
+      "loss": 0.2378,
+      "step": 1900
+    },
+    {
+      "epoch": 0.09044295684375445,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.979945381424905e-05,
+      "loss": 0.2325,
+      "step": 1905
+    },
+    {
+      "epoch": 0.0906803399325832,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.9797093855579515e-05,
+      "loss": 0.2332,
+      "step": 1910
+    },
+    {
+      "epoch": 0.09091772302141196,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.9794720155312436e-05,
+      "loss": 0.2337,
+      "step": 1915
+    },
+    {
+      "epoch": 0.0911551061102407,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.97923327149107e-05,
+      "loss": 0.235,
+      "step": 1920
+    },
+    {
+      "epoch": 0.09139248919906946,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.978993153584569e-05,
+      "loss": 0.2366,
+      "step": 1925
+    },
+    {
+      "epoch": 0.0916298722878982,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.978751661959722e-05,
+      "loss": 0.2351,
+      "step": 1930
+    },
+    {
+      "epoch": 0.09186725537672696,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.9785087967653613e-05,
+      "loss": 0.2364,
+      "step": 1935
+    },
+    {
+      "epoch": 0.09210463846555571,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.9782645581511635e-05,
+      "loss": 0.2382,
+      "step": 1940
+    },
+    {
+      "epoch": 0.09234202155438447,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.9780189462676506e-05,
+      "loss": 0.2396,
+      "step": 1945
+    },
+    {
+      "epoch": 0.09257940464321322,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.977771961266193e-05,
+      "loss": 0.2327,
+      "step": 1950
+    },
+    {
+      "epoch": 0.09281678773204197,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.977523603299005e-05,
+      "loss": 0.2312,
+      "step": 1955
+    },
+    {
+      "epoch": 0.09305417082087072,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.97727387251915e-05,
+      "loss": 0.2326,
+      "step": 1960
+    },
+    {
+      "epoch": 0.09329155390969947,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.9770227690805364e-05,
+      "loss": 0.2337,
+      "step": 1965
+    },
+    {
+      "epoch": 0.09352893699852823,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.976770293137917e-05,
+      "loss": 0.2345,
+      "step": 1970
+    },
+    {
+      "epoch": 0.09376632008735698,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.976516444846891e-05,
+      "loss": 0.2331,
+      "step": 1975
+    },
+    {
+      "epoch": 0.09400370317618573,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.976261224363905e-05,
+      "loss": 0.237,
+      "step": 1980
+    },
+    {
+      "epoch": 0.09424108626501448,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.97600463184625e-05,
+      "loss": 0.2344,
+      "step": 1985
+    },
+    {
+      "epoch": 0.09447846935384323,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.9757466674520626e-05,
+      "loss": 0.236,
+      "step": 1990
+    },
+    {
+      "epoch": 0.09471585244267199,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.9754873313403254e-05,
+      "loss": 0.2347,
+      "step": 1995
+    },
+    {
+      "epoch": 0.09495323553150073,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.975226623670866e-05,
+      "loss": 0.2331,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09519061862032949,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.974964544604357e-05,
+      "loss": 0.2389,
+      "step": 2005
+    },
+    {
+      "epoch": 0.09542800170915824,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.974701094302317e-05,
+      "loss": 0.2327,
+      "step": 2010
+    },
+    {
+      "epoch": 0.09566538479798699,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.974436272927109e-05,
+      "loss": 0.2325,
+      "step": 2015
+    },
+    {
+      "epoch": 0.09590276788681575,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.974170080641941e-05,
+      "loss": 0.2348,
+      "step": 2020
+    },
+    {
+      "epoch": 0.09614015097564449,
+      "grad_norm": 0.875,
+      "learning_rate": 4.973902517610866e-05,
+      "loss": 0.2353,
+      "step": 2025
+    },
+    {
+      "epoch": 0.09637753406447325,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.973633583998783e-05,
+      "loss": 0.2335,
+      "step": 2030
+    },
+    {
+      "epoch": 0.096614917153302,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.9733632799714334e-05,
+      "loss": 0.2322,
+      "step": 2035
+    },
+    {
+      "epoch": 0.09685230024213075,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.973091605695405e-05,
+      "loss": 0.2335,
+      "step": 2040
+    },
+    {
+      "epoch": 0.0970896833309595,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.972818561338128e-05,
+      "loss": 0.2335,
+      "step": 2045
+    },
+    {
+      "epoch": 0.09732706641978825,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.972544147067881e-05,
+      "loss": 0.2344,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09756444950861701,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.972268363053782e-05,
+      "loss": 0.2316,
+      "step": 2055
+    },
+    {
+      "epoch": 0.09780183259744576,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.971991209465796e-05,
+      "loss": 0.2344,
+      "step": 2060
+    },
+    {
+      "epoch": 0.09803921568627451,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.971712686474732e-05,
+      "loss": 0.2305,
+      "step": 2065
+    },
+    {
+      "epoch": 0.09827659877510327,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.971432794252243e-05,
+      "loss": 0.237,
+      "step": 2070
+    },
+    {
+      "epoch": 0.09851398186393201,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.971151532970823e-05,
+      "loss": 0.2315,
+      "step": 2075
+    },
+    {
+      "epoch": 0.09875136495276077,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.970868902803814e-05,
+      "loss": 0.2368,
+      "step": 2080
+    },
+    {
+      "epoch": 0.09898874804158951,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.9705849039253995e-05,
+      "loss": 0.2341,
+      "step": 2085
+    },
+    {
+      "epoch": 0.09922613113041827,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.9702995365106065e-05,
+      "loss": 0.2337,
+      "step": 2090
+    },
+    {
+      "epoch": 0.09946351421924703,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.970012800735305e-05,
+      "loss": 0.2371,
+      "step": 2095
+    },
+    {
+      "epoch": 0.09970089730807577,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.96972469677621e-05,
+      "loss": 0.2295,
+      "step": 2100
+    },
+    {
+      "epoch": 0.09993828039690453,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.969435224810878e-05,
+      "loss": 0.2346,
+      "step": 2105
+    },
+    {
+      "epoch": 0.10017566348573327,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.96914438501771e-05,
+      "loss": 0.2358,
+      "step": 2110
+    },
+    {
+      "epoch": 0.10041304657456203,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.9688521775759476e-05,
+      "loss": 0.2363,
+      "step": 2115
+    },
+    {
+      "epoch": 0.10065042966339079,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.968558602665679e-05,
+      "loss": 0.235,
+      "step": 2120
+    },
+    {
+      "epoch": 0.10088781275221953,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.968263660467831e-05,
+      "loss": 0.2315,
+      "step": 2125
+    },
+    {
+      "epoch": 0.10112519584104829,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.967967351164177e-05,
+      "loss": 0.2293,
+      "step": 2130
+    },
+    {
+      "epoch": 0.10136257892987703,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.96766967493733e-05,
+      "loss": 0.2311,
+      "step": 2135
+    },
+    {
+      "epoch": 0.10159996201870579,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.967370631970747e-05,
+      "loss": 0.2353,
+      "step": 2140
+    },
+    {
+      "epoch": 0.10183734510753453,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.967070222448727e-05,
+      "loss": 0.2336,
+      "step": 2145
+    },
+    {
+      "epoch": 0.10207472819636329,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.96676844655641e-05,
+      "loss": 0.2341,
+      "step": 2150
+    },
+    {
+      "epoch": 0.10231211128519205,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.96646530447978e-05,
+      "loss": 0.2339,
+      "step": 2155
+    },
+    {
+      "epoch": 0.10254949437402079,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.9661607964056605e-05,
+      "loss": 0.2365,
+      "step": 2160
+    },
+    {
+      "epoch": 0.10278687746284955,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.965854922521722e-05,
+      "loss": 0.2308,
+      "step": 2165
+    },
+    {
+      "epoch": 0.10302426055167829,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.965547683016469e-05,
+      "loss": 0.2316,
+      "step": 2170
+    },
+    {
+      "epoch": 0.10326164364050705,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.965239078079255e-05,
+      "loss": 0.2342,
+      "step": 2175
+    },
+    {
+      "epoch": 0.10349902672933581,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.96492910790027e-05,
+      "loss": 0.2346,
+      "step": 2180
+    },
+    {
+      "epoch": 0.10373640981816455,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.964617772670548e-05,
+      "loss": 0.2318,
+      "step": 2185
+    },
+    {
+      "epoch": 0.10397379290699331,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.9643050725819635e-05,
+      "loss": 0.2356,
+      "step": 2190
+    },
+    {
+      "epoch": 0.10421117599582205,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.9639910078272315e-05,
+      "loss": 0.2377,
+      "step": 2195
+    },
+    {
+      "epoch": 0.10444855908465081,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.96367557859991e-05,
+      "loss": 0.2306,
+      "step": 2200
+    },
+    {
+      "epoch": 0.10468594217347957,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.9633587850943945e-05,
+      "loss": 0.2284,
+      "step": 2205
+    },
+    {
+      "epoch": 0.10492332526230831,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.9630406275059247e-05,
+      "loss": 0.2346,
+      "step": 2210
+    },
+    {
+      "epoch": 0.10516070835113707,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.962721106030581e-05,
+      "loss": 0.2345,
+      "step": 2215
+    },
+    {
+      "epoch": 0.10539809143996581,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.96240022086528e-05,
+      "loss": 0.235,
+      "step": 2220
+    },
+    {
+      "epoch": 0.10563547452879457,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.962077972207784e-05,
+      "loss": 0.2324,
+      "step": 2225
+    },
+    {
+      "epoch": 0.10587285761762331,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.961754360256693e-05,
+      "loss": 0.2281,
+      "step": 2230
+    },
+    {
+      "epoch": 0.10611024070645207,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.961429385211447e-05,
+      "loss": 0.2321,
+      "step": 2235
+    },
+    {
+      "epoch": 0.10634762379528083,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.9611030472723266e-05,
+      "loss": 0.2307,
+      "step": 2240
+    },
+    {
+      "epoch": 0.10658500688410957,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.960775346640453e-05,
+      "loss": 0.2292,
+      "step": 2245
+    },
+    {
+      "epoch": 0.10682238997293833,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.9604462835177866e-05,
+      "loss": 0.2313,
+      "step": 2250
+    },
+    {
+      "epoch": 0.10705977306176707,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.960115858107128e-05,
+      "loss": 0.2342,
+      "step": 2255
+    },
+    {
+      "epoch": 0.10729715615059583,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.9597840706121153e-05,
+      "loss": 0.2345,
+      "step": 2260
+    },
+    {
+      "epoch": 0.10753453923942459,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.959450921237228e-05,
+      "loss": 0.2344,
+      "step": 2265
+    },
+    {
+      "epoch": 0.10777192232825333,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.959116410187786e-05,
+      "loss": 0.2304,
+      "step": 2270
+    },
+    {
+      "epoch": 0.10800930541708209,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.958780537669945e-05,
+      "loss": 0.2309,
+      "step": 2275
+    },
+    {
+      "epoch": 0.10824668850591083,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.958443303890704e-05,
+      "loss": 0.2338,
+      "step": 2280
+    },
+    {
+      "epoch": 0.10848407159473959,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.9581047090578966e-05,
+      "loss": 0.2313,
+      "step": 2285
+    },
+    {
+      "epoch": 0.10872145468356834,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.957764753380197e-05,
+      "loss": 0.2334,
+      "step": 2290
+    },
+    {
+      "epoch": 0.1089588377723971,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.95742343706712e-05,
+      "loss": 0.2316,
+      "step": 2295
+    },
+    {
+      "epoch": 0.10919622086122585,
+      "grad_norm": 0.75,
+      "learning_rate": 4.957080760329017e-05,
+      "loss": 0.2295,
+      "step": 2300
+    },
+    {
+      "epoch": 0.1094336039500546,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.956736723377078e-05,
+      "loss": 0.2343,
+      "step": 2305
+    },
+    {
+      "epoch": 0.10967098703888335,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.95639132642333e-05,
+      "loss": 0.2302,
+      "step": 2310
+    },
+    {
+      "epoch": 0.1099083701277121,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.956044569680641e-05,
+      "loss": 0.2349,
+      "step": 2315
+    },
+    {
+      "epoch": 0.11014575321654085,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.955696453362716e-05,
+      "loss": 0.2324,
+      "step": 2320
+    },
+    {
+      "epoch": 0.11038313630536961,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.955346977684097e-05,
+      "loss": 0.2327,
+      "step": 2325
+    },
+    {
+      "epoch": 0.11062051939419836,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.9549961428601624e-05,
+      "loss": 0.233,
+      "step": 2330
+    },
+    {
+      "epoch": 0.11085790248302711,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.954643949107133e-05,
+      "loss": 0.2318,
+      "step": 2335
+    },
+    {
+      "epoch": 0.11109528557185586,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.9542903966420626e-05,
+      "loss": 0.2321,
+      "step": 2340
+    },
+    {
+      "epoch": 0.11133266866068461,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.953935485682845e-05,
+      "loss": 0.2361,
+      "step": 2345
+    },
+    {
+      "epoch": 0.11157005174951337,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.9535792164482095e-05,
+      "loss": 0.2348,
+      "step": 2350
+    },
+    {
+      "epoch": 0.11180743483834212,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.9532215891577225e-05,
+      "loss": 0.231,
+      "step": 2355
+    },
+    {
+      "epoch": 0.11204481792717087,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.952862604031789e-05,
+      "loss": 0.2279,
+      "step": 2360
+    },
+    {
+      "epoch": 0.11228220101599962,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.952502261291651e-05,
+      "loss": 0.2297,
+      "step": 2365
+    },
+    {
+      "epoch": 0.11251958410482837,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.9521405611593834e-05,
+      "loss": 0.234,
+      "step": 2370
+    },
+    {
+      "epoch": 0.11275696719365712,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.9517775038579024e-05,
+      "loss": 0.233,
+      "step": 2375
+    },
+    {
+      "epoch": 0.11299435028248588,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.951413089610958e-05,
+      "loss": 0.2319,
+      "step": 2380
+    },
+    {
+      "epoch": 0.11323173337131463,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.951047318643136e-05,
+      "loss": 0.233,
+      "step": 2385
+    },
+    {
+      "epoch": 0.11346911646014338,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.950680191179862e-05,
+      "loss": 0.2351,
+      "step": 2390
+    },
+    {
+      "epoch": 0.11370649954897213,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.950311707447392e-05,
+      "loss": 0.2301,
+      "step": 2395
+    },
+    {
+      "epoch": 0.11394388263780088,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.949941867672822e-05,
+      "loss": 0.2334,
+      "step": 2400
+    },
+    {
+      "epoch": 0.11418126572662964,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.9495706720840845e-05,
+      "loss": 0.2311,
+      "step": 2405
+    },
+    {
+      "epoch": 0.11441864881545839,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.949198120909943e-05,
+      "loss": 0.23,
+      "step": 2410
+    },
+    {
+      "epoch": 0.11465603190428714,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.9488242143800004e-05,
+      "loss": 0.2299,
+      "step": 2415
+    },
+    {
+      "epoch": 0.1148934149931159,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.948448952724693e-05,
+      "loss": 0.235,
+      "step": 2420
+    },
+    {
+      "epoch": 0.11513079808194464,
+      "grad_norm": 0.5,
+      "learning_rate": 4.948072336175293e-05,
+      "loss": 0.2318,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1153681811707734,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.9476943649639066e-05,
+      "loss": 0.2272,
+      "step": 2430
+    },
+    {
+      "epoch": 0.11560556425960214,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.9473150393234784e-05,
+      "loss": 0.2286,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1158429473484309,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.946934359487782e-05,
+      "loss": 0.2354,
+      "step": 2440
+    },
+    {
+      "epoch": 0.11608033043725965,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.9465523256914314e-05,
+      "loss": 0.2289,
+      "step": 2445
+    },
+    {
+      "epoch": 0.1163177135260884,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.946168938169869e-05,
+      "loss": 0.2293,
+      "step": 2450
+    },
+    {
+      "epoch": 0.11655509661491716,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.945784197159379e-05,
+      "loss": 0.2313,
+      "step": 2455
+    },
+    {
+      "epoch": 0.1167924797037459,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.9453981028970715e-05,
+      "loss": 0.2368,
+      "step": 2460
+    },
+    {
+      "epoch": 0.11702986279257466,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.9450106556208975e-05,
+      "loss": 0.2315,
+      "step": 2465
+    },
+    {
+      "epoch": 0.11726724588140341,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.944621855569637e-05,
+      "loss": 0.2329,
+      "step": 2470
+    },
+    {
+      "epoch": 0.11750462897023216,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.9442317029829064e-05,
+      "loss": 0.2287,
+      "step": 2475
+    },
+    {
+      "epoch": 0.11774201205906092,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.943840198101156e-05,
+      "loss": 0.234,
+      "step": 2480
+    },
+    {
+      "epoch": 0.11797939514788966,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.9434473411656666e-05,
+      "loss": 0.2294,
+      "step": 2485
+    },
+    {
+      "epoch": 0.11821677823671842,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.9430531324185555e-05,
+      "loss": 0.2308,
+      "step": 2490
+    },
+    {
+      "epoch": 0.11845416132554717,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.942657572102772e-05,
+      "loss": 0.2347,
+      "step": 2495
+    },
+    {
+      "epoch": 0.11869154441437592,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.942260660462097e-05,
+      "loss": 0.23,
+      "step": 2500
+    },
+    {
+      "epoch": 0.11892892750320468,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.9418623977411455e-05,
+      "loss": 0.2308,
+      "step": 2505
+    },
+    {
+      "epoch": 0.11916631059203342,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.941462784185366e-05,
+      "loss": 0.2304,
+      "step": 2510
+    },
+    {
+      "epoch": 0.11940369368086218,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.9410618200410374e-05,
+      "loss": 0.235,
+      "step": 2515
+    },
+    {
+      "epoch": 0.11964107676969092,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.940659505555273e-05,
+      "loss": 0.2389,
+      "step": 2520
+    },
+    {
+      "epoch": 0.11987845985851968,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.9402558409760176e-05,
+      "loss": 0.2329,
+      "step": 2525
+    },
+    {
+      "epoch": 0.12011584294734844,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.9398508265520455e-05,
+      "loss": 0.2325,
+      "step": 2530
+    },
+    {
+      "epoch": 0.12035322603617718,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.9394444625329686e-05,
+      "loss": 0.2307,
+      "step": 2535
+    },
+    {
+      "epoch": 0.12059060912500594,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.9390367491692255e-05,
+      "loss": 0.2294,
+      "step": 2540
+    },
+    {
+      "epoch": 0.12082799221383468,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.938627686712087e-05,
+      "loss": 0.2339,
+      "step": 2545
+    },
+    {
+      "epoch": 0.12106537530266344,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.9382172754136574e-05,
+      "loss": 0.2292,
+      "step": 2550
+    },
+    {
+      "epoch": 0.1213027583914922,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.9378055155268726e-05,
+      "loss": 0.2302,
+      "step": 2555
+    },
+    {
+      "epoch": 0.12154014148032094,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.937392407305497e-05,
+      "loss": 0.2301,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1217775245691497,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.936977951004127e-05,
+      "loss": 0.2314,
+      "step": 2565
+    },
+    {
+      "epoch": 0.12201490765797844,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9365621468781905e-05,
+      "loss": 0.2317,
+      "step": 2570
+    },
+    {
+      "epoch": 0.1222522907468072,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.936144995183947e-05,
+      "loss": 0.2319,
+      "step": 2575
+    },
+    {
+      "epoch": 0.12248967383563594,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.935726496178483e-05,
+      "loss": 0.2309,
+      "step": 2580
+    },
+    {
+      "epoch": 0.1227270569244647,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.935306650119719e-05,
+      "loss": 0.2333,
+      "step": 2585
+    },
+    {
+      "epoch": 0.12296444001329346,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.934885457266404e-05,
+      "loss": 0.2308,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1232018231021222,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.9344629178781165e-05,
+      "loss": 0.2287,
+      "step": 2595
+    },
+    {
+      "epoch": 0.12343920619095096,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.934039032215267e-05,
+      "loss": 0.2306,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1236765892797797,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.933613800539093e-05,
+      "loss": 0.2297,
+      "step": 2605
+    },
+    {
+      "epoch": 0.12391397236860846,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.9331872231116624e-05,
+      "loss": 0.2284,
+      "step": 2610
+    },
+    {
+      "epoch": 0.12415135545743722,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.932759300195875e-05,
+      "loss": 0.2291,
+      "step": 2615
+    },
+    {
+      "epoch": 0.12438873854626596,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.9323300320554566e-05,
+      "loss": 0.2289,
+      "step": 2620
+    },
+    {
+      "epoch": 0.12462612163509472,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.931899418954964e-05,
+      "loss": 0.2321,
+      "step": 2625
+    },
+    {
+      "epoch": 0.12486350472392346,
+      "grad_norm": 0.5,
+      "learning_rate": 4.93146746115978e-05,
+      "loss": 0.2335,
+      "step": 2630
+    },
+    {
+      "epoch": 0.12510088781275222,
+      "grad_norm": 0.5,
+      "learning_rate": 4.9310341589361195e-05,
+      "loss": 0.2329,
+      "step": 2635
+    },
+    {
+      "epoch": 0.12533827090158098,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.930599512551025e-05,
+      "loss": 0.2311,
+      "step": 2640
+    },
+    {
+      "epoch": 0.12557565399040974,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.9301635222723666e-05,
+      "loss": 0.2332,
+      "step": 2645
+    },
+    {
+      "epoch": 0.12581303707923847,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.929726188368844e-05,
+      "loss": 0.2271,
+      "step": 2650
+    },
+    {
+      "epoch": 0.12605042016806722,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.9292875111099825e-05,
+      "loss": 0.2275,
+      "step": 2655
+    },
+    {
+      "epoch": 0.12628780325689598,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.928847490766138e-05,
+      "loss": 0.2327,
+      "step": 2660
+    },
+    {
+      "epoch": 0.12652518634572474,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.928406127608494e-05,
+      "loss": 0.2322,
+      "step": 2665
+    },
+    {
+      "epoch": 0.1267625694345535,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.927963421909058e-05,
+      "loss": 0.2277,
+      "step": 2670
+    },
+    {
+      "epoch": 0.12699995252338223,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.927519373940669e-05,
+      "loss": 0.2282,
+      "step": 2675
+    },
+    {
+      "epoch": 0.12723733561221098,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.9270739839769934e-05,
+      "loss": 0.2301,
+      "step": 2680
+    },
+    {
+      "epoch": 0.12747471870103974,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.926627252292519e-05,
+      "loss": 0.232,
+      "step": 2685
+    },
+    {
+      "epoch": 0.1277121017898685,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.926179179162569e-05,
+      "loss": 0.237,
+      "step": 2690
+    },
+    {
+      "epoch": 0.12794948487869723,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.925729764863285e-05,
+      "loss": 0.234,
+      "step": 2695
+    },
+    {
+      "epoch": 0.12818686796752599,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.925279009671642e-05,
+      "loss": 0.2348,
+      "step": 2700
+    },
+    {
+      "epoch": 0.12842425105635474,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.9248269138654356e-05,
+      "loss": 0.2312,
+      "step": 2705
+    },
+    {
+      "epoch": 0.1286616341451835,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.924373477723292e-05,
+      "loss": 0.2289,
+      "step": 2710
+    },
+    {
+      "epoch": 0.12889901723401226,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.923918701524663e-05,
+      "loss": 0.2326,
+      "step": 2715
+    },
+    {
+      "epoch": 0.129136400322841,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.9234625855498235e-05,
+      "loss": 0.2337,
+      "step": 2720
+    },
+    {
+      "epoch": 0.12937378341166975,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.9230051300798754e-05,
+      "loss": 0.231,
+      "step": 2725
+    },
+    {
+      "epoch": 0.1296111665004985,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.922546335396746e-05,
+      "loss": 0.2294,
+      "step": 2730
+    },
+    {
+      "epoch": 0.12984854958932726,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.922086201783191e-05,
+      "loss": 0.2311,
+      "step": 2735
+    },
+    {
+      "epoch": 0.13008593267815602,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.9216247295227865e-05,
+      "loss": 0.2312,
+      "step": 2740
+    },
+    {
+      "epoch": 0.13032331576698475,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.921161918899936e-05,
+      "loss": 0.2315,
+      "step": 2745
+    },
+    {
+      "epoch": 0.1305606988558135,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.920697770199868e-05,
+      "loss": 0.2327,
+      "step": 2750
+    },
+    {
+      "epoch": 0.13079808194464226,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.9202322837086356e-05,
+      "loss": 0.2332,
+      "step": 2755
+    },
+    {
+      "epoch": 0.13103546503347102,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.919765459713115e-05,
+      "loss": 0.2274,
+      "step": 2760
+    },
+    {
+      "epoch": 0.13127284812229978,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.919297298501008e-05,
+      "loss": 0.2299,
+      "step": 2765
+    },
+    {
+      "epoch": 0.1315102312111285,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.91882780036084e-05,
+      "loss": 0.2341,
+      "step": 2770
+    },
+    {
+      "epoch": 0.13174761429995727,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.918356965581962e-05,
+      "loss": 0.2272,
+      "step": 2775
+    },
+    {
+      "epoch": 0.13198499738878602,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.917884794454547e-05,
+      "loss": 0.2309,
+      "step": 2780
+    },
+    {
+      "epoch": 0.13222238047761478,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.917411287269589e-05,
+      "loss": 0.2282,
+      "step": 2785
+    },
+    {
+      "epoch": 0.13245976356644354,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.916936444318912e-05,
+      "loss": 0.2262,
+      "step": 2790
+    },
+    {
+      "epoch": 0.13269714665527227,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.916460265895158e-05,
+      "loss": 0.2315,
+      "step": 2795
+    },
+    {
+      "epoch": 0.13293452974410103,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.915982752291793e-05,
+      "loss": 0.2339,
+      "step": 2800
+    },
+    {
+      "epoch": 0.13317191283292978,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.915503903803108e-05,
+      "loss": 0.2342,
+      "step": 2805
+    },
+    {
+      "epoch": 0.13340929592175854,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.915023720724214e-05,
+      "loss": 0.2349,
+      "step": 2810
+    },
+    {
+      "epoch": 0.1336466790105873,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.914542203351046e-05,
+      "loss": 0.2264,
+      "step": 2815
+    },
+    {
+      "epoch": 0.13388406209941603,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.9140593519803604e-05,
+      "loss": 0.2296,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1341214451882448,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.913575166909738e-05,
+      "loss": 0.2306,
+      "step": 2825
+    },
+    {
+      "epoch": 0.13435882827707354,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.913089648437578e-05,
+      "loss": 0.2291,
+      "step": 2830
+    },
+    {
+      "epoch": 0.1345962113659023,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.912602796863104e-05,
+      "loss": 0.2312,
+      "step": 2835
+    },
+    {
+      "epoch": 0.13483359445473103,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.9121146124863606e-05,
+      "loss": 0.2306,
+      "step": 2840
+    },
+    {
+      "epoch": 0.1350709775435598,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.911625095608213e-05,
+      "loss": 0.2367,
+      "step": 2845
+    },
+    {
+      "epoch": 0.13530836063238855,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.911134246530349e-05,
+      "loss": 0.2321,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1355457437212173,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.9106420655552756e-05,
+      "loss": 0.2336,
+      "step": 2855
+    },
+    {
+      "epoch": 0.13578312681004606,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.9101485529863225e-05,
+      "loss": 0.2319,
+      "step": 2860
+    },
+    {
+      "epoch": 0.1360205098988748,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.9096537091276394e-05,
+      "loss": 0.2283,
+      "step": 2865
+    },
+    {
+      "epoch": 0.13625789298770355,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.909157534284196e-05,
+      "loss": 0.2305,
+      "step": 2870
+    },
+    {
+      "epoch": 0.1364952760765323,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.908660028761782e-05,
+      "loss": 0.2322,
+      "step": 2875
+    },
+    {
+      "epoch": 0.13673265916536106,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.908161192867009e-05,
+      "loss": 0.2305,
+      "step": 2880
+    },
+    {
+      "epoch": 0.13697004225418982,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.9076610269073064e-05,
+      "loss": 0.2309,
+      "step": 2885
+    },
+    {
+      "epoch": 0.13720742534301855,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.907159531190925e-05,
+      "loss": 0.2275,
+      "step": 2890
+    },
+    {
+      "epoch": 0.1374448084318473,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.9066567060269334e-05,
+      "loss": 0.229,
+      "step": 2895
+    },
+    {
+      "epoch": 0.13768219152067607,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.906152551725221e-05,
+      "loss": 0.2304,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13791957460950482,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.905647068596495e-05,
+      "loss": 0.2271,
+      "step": 2905
+    },
+    {
+      "epoch": 0.13815695769833358,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.905140256952284e-05,
+      "loss": 0.2253,
+      "step": 2910
+    },
+    {
+      "epoch": 0.1383943407871623,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.904632117104933e-05,
+      "loss": 0.2286,
+      "step": 2915
+    },
+    {
+      "epoch": 0.13863172387599107,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.904122649367605e-05,
+      "loss": 0.2294,
+      "step": 2920
+    },
+    {
+      "epoch": 0.13886910696481983,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.9036118540542836e-05,
+      "loss": 0.2277,
+      "step": 2925
+    },
+    {
+      "epoch": 0.13910649005364858,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.9030997314797705e-05,
+      "loss": 0.2304,
+      "step": 2930
+    },
+    {
+      "epoch": 0.13934387314247734,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.902586281959683e-05,
+      "loss": 0.2279,
+      "step": 2935
+    },
+    {
+      "epoch": 0.13958125623130607,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.902071505810459e-05,
+      "loss": 0.2333,
+      "step": 2940
+    },
+    {
+      "epoch": 0.13981863932013483,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.901555403349352e-05,
+      "loss": 0.2289,
+      "step": 2945
+    },
+    {
+      "epoch": 0.1400560224089636,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.901037974894433e-05,
+      "loss": 0.2347,
+      "step": 2950
+    },
+    {
+      "epoch": 0.14029340549779235,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.900519220764593e-05,
+      "loss": 0.2324,
+      "step": 2955
+    },
+    {
+      "epoch": 0.1405307885866211,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.899999141279536e-05,
+      "loss": 0.2261,
+      "step": 2960
+    },
+    {
+      "epoch": 0.14076817167544983,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.8994777367597836e-05,
+      "loss": 0.2315,
+      "step": 2965
+    },
+    {
+      "epoch": 0.1410055547642786,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.8989550075266784e-05,
+      "loss": 0.2299,
+      "step": 2970
+    },
+    {
+      "epoch": 0.14124293785310735,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.898430953902372e-05,
+      "loss": 0.2332,
+      "step": 2975
+    },
+    {
+      "epoch": 0.1414803209419361,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.8979055762098396e-05,
+      "loss": 0.2314,
+      "step": 2980
+    },
+    {
+      "epoch": 0.14171770403076484,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.897378874772867e-05,
+      "loss": 0.2314,
+      "step": 2985
+    },
+    {
+      "epoch": 0.1419550871195936,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.8968508499160594e-05,
+      "loss": 0.2294,
+      "step": 2990
+    },
+    {
+      "epoch": 0.14219247020842235,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.896321501964836e-05,
+      "loss": 0.2284,
+      "step": 2995
+    },
+    {
+      "epoch": 0.1424298532972511,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.8957908312454296e-05,
+      "loss": 0.2325,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14266723638607987,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.895258838084892e-05,
+      "loss": 0.2277,
+      "step": 3005
+    },
+    {
+      "epoch": 0.1429046194749086,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.8947255228110874e-05,
+      "loss": 0.228,
+      "step": 3010
+    },
+    {
+      "epoch": 0.14314200256373735,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.894190885752697e-05,
+      "loss": 0.2319,
+      "step": 3015
+    },
+    {
+      "epoch": 0.1433793856525661,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.8936549272392124e-05,
+      "loss": 0.2279,
+      "step": 3020
+    },
+    {
+      "epoch": 0.14361676874139487,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.893117647600945e-05,
+      "loss": 0.2312,
+      "step": 3025
+    },
+    {
+      "epoch": 0.14385415183022363,
+      "grad_norm": 0.5,
+      "learning_rate": 4.892579047169017e-05,
+      "loss": 0.2297,
+      "step": 3030
+    },
+    {
+      "epoch": 0.14409153491905236,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.892039126275365e-05,
+      "loss": 0.2306,
+      "step": 3035
+    },
+    {
+      "epoch": 0.1443289180078811,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.8914978852527394e-05,
+      "loss": 0.2304,
+      "step": 3040
+    },
+    {
+      "epoch": 0.14456630109670987,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.890955324434706e-05,
+      "loss": 0.2311,
+      "step": 3045
+    },
+    {
+      "epoch": 0.14480368418553863,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.890411444155641e-05,
+      "loss": 0.2294,
+      "step": 3050
+    },
+    {
+      "epoch": 0.14504106727436739,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.889866244750736e-05,
+      "loss": 0.2272,
+      "step": 3055
+    },
+    {
+      "epoch": 0.14527845036319612,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.889319726555996e-05,
+      "loss": 0.2332,
+      "step": 3060
+    },
+    {
+      "epoch": 0.14551583345202487,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.8887718899082355e-05,
+      "loss": 0.2331,
+      "step": 3065
+    },
+    {
+      "epoch": 0.14575321654085363,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.888222735145086e-05,
+      "loss": 0.2261,
+      "step": 3070
+    },
+    {
+      "epoch": 0.1459905996296824,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.8876722626049884e-05,
+      "loss": 0.2279,
+      "step": 3075
+    },
+    {
+      "epoch": 0.14622798271851115,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.8871204726271946e-05,
+      "loss": 0.2258,
+      "step": 3080
+    },
+    {
+      "epoch": 0.14646536580733988,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.886567365551772e-05,
+      "loss": 0.2305,
+      "step": 3085
+    },
+    {
+      "epoch": 0.14670274889616863,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.886012941719599e-05,
+      "loss": 0.2314,
+      "step": 3090
+    },
+    {
+      "epoch": 0.1469401319849974,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.8854572014723616e-05,
+      "loss": 0.2293,
+      "step": 3095
+    },
+    {
+      "epoch": 0.14717751507382615,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.8849001451525626e-05,
+      "loss": 0.231,
+      "step": 3100
+    },
+    {
+      "epoch": 0.1474148981626549,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.884341773103511e-05,
+      "loss": 0.2295,
+      "step": 3105
+    },
+    {
+      "epoch": 0.14765228125148364,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.8837820856693314e-05,
+      "loss": 0.2254,
+      "step": 3110
+    },
+    {
+      "epoch": 0.1478896643403124,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.883221083194955e-05,
+      "loss": 0.2297,
+      "step": 3115
+    },
+    {
+      "epoch": 0.14812704742914115,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.882658766026125e-05,
+      "loss": 0.2293,
+      "step": 3120
+    },
+    {
+      "epoch": 0.1483644305179699,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.882095134509396e-05,
+      "loss": 0.2291,
+      "step": 3125
+    },
+    {
+      "epoch": 0.14860181360679864,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.8815301889921305e-05,
+      "loss": 0.2279,
+      "step": 3130
+    },
+    {
+      "epoch": 0.1488391966956274,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.880963929822502e-05,
+      "loss": 0.2261,
+      "step": 3135
+    },
+    {
+      "epoch": 0.14907657978445615,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.880396357349495e-05,
+      "loss": 0.2287,
+      "step": 3140
+    },
+    {
+      "epoch": 0.1493139628732849,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.8798274719228996e-05,
+      "loss": 0.2329,
+      "step": 3145
+    },
+    {
+      "epoch": 0.14955134596211367,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.879257273893319e-05,
+      "loss": 0.2319,
+      "step": 3150
+    },
+    {
+      "epoch": 0.1497887290509424,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.8786857636121616e-05,
+      "loss": 0.2303,
+      "step": 3155
+    },
+    {
+      "epoch": 0.15002611213977116,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.878112941431648e-05,
+      "loss": 0.2315,
+      "step": 3160
+    },
+    {
+      "epoch": 0.1502634952285999,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.877538807704807e-05,
+      "loss": 0.2299,
+      "step": 3165
+    },
+    {
+      "epoch": 0.15050087831742867,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.876963362785473e-05,
+      "loss": 0.2343,
+      "step": 3170
+    },
+    {
+      "epoch": 0.15073826140625743,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.876386607028291e-05,
+      "loss": 0.2295,
+      "step": 3175
+    },
+    {
+      "epoch": 0.15097564449508616,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.8758085407887105e-05,
+      "loss": 0.2286,
+      "step": 3180
+    },
+    {
+      "epoch": 0.15121302758391492,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.875229164422995e-05,
+      "loss": 0.2305,
+      "step": 3185
+    },
+    {
+      "epoch": 0.15145041067274367,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.874648478288209e-05,
+      "loss": 0.2273,
+      "step": 3190
+    },
+    {
+      "epoch": 0.15168779376157243,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.8740664827422265e-05,
+      "loss": 0.2295,
+      "step": 3195
+    },
+    {
+      "epoch": 0.1519251768504012,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.8734831781437304e-05,
+      "loss": 0.2296,
+      "step": 3200
+    },
+    {
+      "epoch": 0.15216255993922992,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.8728985648522075e-05,
+      "loss": 0.2304,
+      "step": 3205
+    },
+    {
+      "epoch": 0.15239994302805868,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.872312643227952e-05,
+      "loss": 0.2354,
+      "step": 3210
+    },
+    {
+      "epoch": 0.15263732611688743,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.871725413632066e-05,
+      "loss": 0.2296,
+      "step": 3215
+    },
+    {
+      "epoch": 0.1528747092057162,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.8711368764264555e-05,
+      "loss": 0.2297,
+      "step": 3220
+    },
+    {
+      "epoch": 0.15311209229454495,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.870547031973834e-05,
+      "loss": 0.232,
+      "step": 3225
+    },
+    {
+      "epoch": 0.15334947538337368,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.869955880637719e-05,
+      "loss": 0.2322,
+      "step": 3230
+    },
+    {
+      "epoch": 0.15358685847220244,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.8693634227824355e-05,
+      "loss": 0.2311,
+      "step": 3235
+    },
+    {
+      "epoch": 0.1538242415610312,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.868769658773111e-05,
+      "loss": 0.2308,
+      "step": 3240
+    },
+    {
+      "epoch": 0.15406162464985995,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.8681745889756816e-05,
+      "loss": 0.2286,
+      "step": 3245
+    },
+    {
+      "epoch": 0.1542990077386887,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.867578213756885e-05,
+      "loss": 0.2297,
+      "step": 3250
+    },
+    {
+      "epoch": 0.15453639082751744,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.8669805334842634e-05,
+      "loss": 0.2302,
+      "step": 3255
+    },
+    {
+      "epoch": 0.1547737739163462,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.8663815485261666e-05,
+      "loss": 0.2311,
+      "step": 3260
+    },
+    {
+      "epoch": 0.15501115700517495,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.865781259251744e-05,
+      "loss": 0.229,
+      "step": 3265
+    },
+    {
+      "epoch": 0.1552485400940037,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.865179666030954e-05,
+      "loss": 0.2346,
+      "step": 3270
+    },
+    {
+      "epoch": 0.15548592318283244,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.864576769234553e-05,
+      "loss": 0.2287,
+      "step": 3275
+    },
+    {
+      "epoch": 0.1557233062716612,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.863972569234106e-05,
+      "loss": 0.2326,
+      "step": 3280
+    },
+    {
+      "epoch": 0.15596068936048996,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.863367066401977e-05,
+      "loss": 0.2242,
+      "step": 3285
+    },
+    {
+      "epoch": 0.15619807244931871,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.862760261111335e-05,
+      "loss": 0.2321,
+      "step": 3290
+    },
+    {
+      "epoch": 0.15643545553814747,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.8621521537361516e-05,
+      "loss": 0.2286,
+      "step": 3295
+    },
+    {
+      "epoch": 0.1566728386269762,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.8615427446512e-05,
+      "loss": 0.2288,
+      "step": 3300
+    },
+    {
+      "epoch": 0.15691022171580496,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.860932034232058e-05,
+      "loss": 0.2331,
+      "step": 3305
+    },
+    {
+      "epoch": 0.15714760480463372,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.8603200228551006e-05,
+      "loss": 0.2311,
+      "step": 3310
+    },
+    {
+      "epoch": 0.15738498789346247,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.8597067108975104e-05,
+      "loss": 0.2313,
+      "step": 3315
+    },
+    {
+      "epoch": 0.15762237098229123,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.859092098737267e-05,
+      "loss": 0.2275,
+      "step": 3320
+    },
+    {
+      "epoch": 0.15785975407111996,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.8584761867531554e-05,
+      "loss": 0.2316,
+      "step": 3325
+    },
+    {
+      "epoch": 0.15809713715994872,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.857858975324757e-05,
+      "loss": 0.2297,
+      "step": 3330
+    },
+    {
+      "epoch": 0.15833452024877748,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.857240464832457e-05,
+      "loss": 0.2249,
+      "step": 3335
+    },
+    {
+      "epoch": 0.15857190333760623,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.856620655657441e-05,
+      "loss": 0.2292,
+      "step": 3340
+    },
+    {
+      "epoch": 0.158809286426435,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.855999548181695e-05,
+      "loss": 0.2261,
+      "step": 3345
+    },
+    {
+      "epoch": 0.15904666951526372,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.855377142788003e-05,
+      "loss": 0.229,
+      "step": 3350
+    },
+    {
+      "epoch": 0.15928405260409248,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.854753439859952e-05,
+      "loss": 0.2349,
+      "step": 3355
+    },
+    {
+      "epoch": 0.15952143569292124,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.854128439781927e-05,
+      "loss": 0.2287,
+      "step": 3360
+    },
+    {
+      "epoch": 0.15975881878175,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.853502142939113e-05,
+      "loss": 0.2279,
+      "step": 3365
+    },
+    {
+      "epoch": 0.15999620187057875,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.852874549717494e-05,
+      "loss": 0.226,
+      "step": 3370
+    },
+    {
+      "epoch": 0.16023358495940748,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.8522456605038515e-05,
+      "loss": 0.2279,
+      "step": 3375
+    },
+    {
+      "epoch": 0.16047096804823624,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.8516154756857686e-05,
+      "loss": 0.2297,
+      "step": 3380
+    },
+    {
+      "epoch": 0.160708351137065,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.8509839956516246e-05,
+      "loss": 0.2289,
+      "step": 3385
+    },
+    {
+      "epoch": 0.16094573422589376,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.8503512207905985e-05,
+      "loss": 0.2269,
+      "step": 3390
+    },
+    {
+      "epoch": 0.1611831173147225,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.849717151492665e-05,
+      "loss": 0.2311,
+      "step": 3395
+    },
+    {
+      "epoch": 0.16142050040355124,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.8490817881486e-05,
+      "loss": 0.2317,
+      "step": 3400
+    },
+    {
+      "epoch": 0.16165788349238,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.848445131149973e-05,
+      "loss": 0.2312,
+      "step": 3405
+    },
+    {
+      "epoch": 0.16189526658120876,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.847807180889155e-05,
+      "loss": 0.2328,
+      "step": 3410
+    },
+    {
+      "epoch": 0.16213264967003752,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.847167937759311e-05,
+      "loss": 0.2326,
+      "step": 3415
+    },
+    {
+      "epoch": 0.16237003275886625,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.8465274021544026e-05,
+      "loss": 0.2327,
+      "step": 3420
+    },
+    {
+      "epoch": 0.162607415847695,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.84588557446919e-05,
+      "loss": 0.2307,
+      "step": 3425
+    },
+    {
+      "epoch": 0.16284479893652376,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.8452424550992286e-05,
+      "loss": 0.2272,
+      "step": 3430
+    },
+    {
+      "epoch": 0.16308218202535252,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.8445980444408706e-05,
+      "loss": 0.2288,
+      "step": 3435
+    },
+    {
+      "epoch": 0.16331956511418128,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.843952342891262e-05,
+      "loss": 0.2317,
+      "step": 3440
+    },
+    {
+      "epoch": 0.16355694820301,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.843305350848346e-05,
+      "loss": 0.2296,
+      "step": 3445
+    },
+    {
+      "epoch": 0.16379433129183876,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.842657068710862e-05,
+      "loss": 0.2304,
+      "step": 3450
+    },
+    {
+      "epoch": 0.16403171438066752,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.842007496878342e-05,
+      "loss": 0.2284,
+      "step": 3455
+    },
+    {
+      "epoch": 0.16426909746949628,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.841356635751115e-05,
+      "loss": 0.2316,
+      "step": 3460
+    },
+    {
+      "epoch": 0.16450648055832504,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.840704485730303e-05,
+      "loss": 0.2329,
+      "step": 3465
+    },
+    {
+      "epoch": 0.16474386364715377,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.840051047217824e-05,
+      "loss": 0.2334,
+      "step": 3470
+    },
+    {
+      "epoch": 0.16498124673598252,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.8393963206163886e-05,
+      "loss": 0.2285,
+      "step": 3475
+    },
+    {
+      "epoch": 0.16521862982481128,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.8387403063295017e-05,
+      "loss": 0.2304,
+      "step": 3480
+    },
+    {
+      "epoch": 0.16545601291364004,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.838083004761462e-05,
+      "loss": 0.2279,
+      "step": 3485
+    },
+    {
+      "epoch": 0.1656933960024688,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.837424416317362e-05,
+      "loss": 0.2295,
+      "step": 3490
+    },
+    {
+      "epoch": 0.16593077909129753,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.836764541403087e-05,
+      "loss": 0.2302,
+      "step": 3495
+    },
+    {
+      "epoch": 0.16616816218012628,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.836103380425313e-05,
+      "loss": 0.2275,
+      "step": 3500
+    },
+    {
+      "epoch": 0.16640554526895504,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.835440933791513e-05,
+      "loss": 0.2283,
+      "step": 3505
+    },
+    {
+      "epoch": 0.1666429283577838,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.834777201909948e-05,
+      "loss": 0.2318,
+      "step": 3510
+    },
+    {
+      "epoch": 0.16688031144661256,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.8341121851896734e-05,
+      "loss": 0.2257,
+      "step": 3515
+    },
+    {
+      "epoch": 0.16711769453544129,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.833445884040537e-05,
+      "loss": 0.2303,
+      "step": 3520
+    },
+    {
+      "epoch": 0.16735507762427004,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.832778298873176e-05,
+      "loss": 0.2315,
+      "step": 3525
+    },
+    {
+      "epoch": 0.1675924607130988,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.832109430099021e-05,
+      "loss": 0.228,
+      "step": 3530
+    },
+    {
+      "epoch": 0.16782984380192756,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.831439278130291e-05,
+      "loss": 0.2241,
+      "step": 3535
+    },
+    {
+      "epoch": 0.16806722689075632,
+      "grad_norm": 0.625,
+      "learning_rate": 4.83076784338e-05,
+      "loss": 0.2315,
+      "step": 3540
+    },
+    {
+      "epoch": 0.16830460997958505,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.830095126261948e-05,
+      "loss": 0.2279,
+      "step": 3545
+    },
+    {
+      "epoch": 0.1685419930684138,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.82942112719073e-05,
+      "loss": 0.2359,
+      "step": 3550
+    },
+    {
+      "epoch": 0.16877937615724256,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.828745846581726e-05,
+      "loss": 0.2282,
+      "step": 3555
+    },
+    {
+      "epoch": 0.16901675924607132,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.82806928485111e-05,
+      "loss": 0.231,
+      "step": 3560
+    },
+    {
+      "epoch": 0.16925414233490005,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.827391442415843e-05,
+      "loss": 0.2284,
+      "step": 3565
+    },
+    {
+      "epoch": 0.1694915254237288,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.8267123196936767e-05,
+      "loss": 0.2263,
+      "step": 3570
+    },
+    {
+      "epoch": 0.16972890851255756,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.826031917103151e-05,
+      "loss": 0.2312,
+      "step": 3575
+    },
+    {
+      "epoch": 0.16996629160138632,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.8253502350635957e-05,
+      "loss": 0.2319,
+      "step": 3580
+    },
+    {
+      "epoch": 0.17020367469021508,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.824667273995127e-05,
+      "loss": 0.2323,
+      "step": 3585
+    },
+    {
+      "epoch": 0.1704410577790438,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.8239830343186526e-05,
+      "loss": 0.2315,
+      "step": 3590
+    },
+    {
+      "epoch": 0.17067844086787257,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.823297516455864e-05,
+      "loss": 0.2255,
+      "step": 3595
+    },
+    {
+      "epoch": 0.17091582395670132,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.8226107208292444e-05,
+      "loss": 0.2277,
+      "step": 3600
+    },
+    {
+      "epoch": 0.17115320704553008,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.8219226478620634e-05,
+      "loss": 0.2318,
+      "step": 3605
+    },
+    {
+      "epoch": 0.17139059013435884,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.8212332979783746e-05,
+      "loss": 0.2255,
+      "step": 3610
+    },
+    {
+      "epoch": 0.17162797322318757,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.820542671603024e-05,
+      "loss": 0.2276,
+      "step": 3615
+    },
+    {
+      "epoch": 0.17186535631201633,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.81985076916164e-05,
+      "loss": 0.2251,
+      "step": 3620
+    },
+    {
+      "epoch": 0.17210273940084508,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.819157591080639e-05,
+      "loss": 0.2312,
+      "step": 3625
+    },
+    {
+      "epoch": 0.17234012248967384,
+      "grad_norm": 0.5,
+      "learning_rate": 4.8184631377872244e-05,
+      "loss": 0.2251,
+      "step": 3630
+    },
+    {
+      "epoch": 0.1725775055785026,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.8177674097093835e-05,
+      "loss": 0.2303,
+      "step": 3635
+    },
+    {
+      "epoch": 0.17281488866733133,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.8170704072758904e-05,
+      "loss": 0.2273,
+      "step": 3640
+    },
+    {
+      "epoch": 0.1730522717561601,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.816372130916305e-05,
+      "loss": 0.2294,
+      "step": 3645
+    },
+    {
+      "epoch": 0.17328965484498884,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.815672581060972e-05,
+      "loss": 0.228,
+      "step": 3650
+    },
+    {
+      "epoch": 0.1735270379338176,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.814971758141021e-05,
+      "loss": 0.2305,
+      "step": 3655
+    },
+    {
+      "epoch": 0.17376442102264636,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.814269662588365e-05,
+      "loss": 0.233,
+      "step": 3660
+    },
+    {
+      "epoch": 0.1740018041114751,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.813566294835703e-05,
+      "loss": 0.2284,
+      "step": 3665
+    },
+    {
+      "epoch": 0.17423918720030385,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.8128616553165175e-05,
+      "loss": 0.2284,
+      "step": 3670
+    },
+    {
+      "epoch": 0.1744765702891326,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.812155744465073e-05,
+      "loss": 0.2327,
+      "step": 3675
+    },
+    {
+      "epoch": 0.17471395337796136,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.811448562716422e-05,
+      "loss": 0.2279,
+      "step": 3680
+    },
+    {
+      "epoch": 0.17495133646679012,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.8107401105063957e-05,
+      "loss": 0.2299,
+      "step": 3685
+    },
+    {
+      "epoch": 0.17518871955561885,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.81003038827161e-05,
+      "loss": 0.2288,
+      "step": 3690
+    },
+    {
+      "epoch": 0.1754261026444476,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.809319396449463e-05,
+      "loss": 0.2301,
+      "step": 3695
+    },
+    {
+      "epoch": 0.17566348573327636,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.8086071354781364e-05,
+      "loss": 0.2263,
+      "step": 3700
+    },
+    {
+      "epoch": 0.17590086882210512,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.807893605796594e-05,
+      "loss": 0.2279,
+      "step": 3705
+    },
+    {
+      "epoch": 0.17613825191093385,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.80717880784458e-05,
+      "loss": 0.2282,
+      "step": 3710
+    },
+    {
+      "epoch": 0.1763756349997626,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.8064627420626215e-05,
+      "loss": 0.2256,
+      "step": 3715
+    },
+    {
+      "epoch": 0.17661301808859137,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.805745408892026e-05,
+      "loss": 0.2282,
+      "step": 3720
+    },
+    {
+      "epoch": 0.17685040117742012,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.805026808774883e-05,
+      "loss": 0.2284,
+      "step": 3725
+    },
+    {
+      "epoch": 0.17708778426624888,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.804306942154063e-05,
+      "loss": 0.2287,
+      "step": 3730
+    },
+    {
+      "epoch": 0.1773251673550776,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.803585809473216e-05,
+      "loss": 0.2282,
+      "step": 3735
+    },
+    {
+      "epoch": 0.17756255044390637,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.802863411176774e-05,
+      "loss": 0.2269,
+      "step": 3740
+    },
+    {
+      "epoch": 0.17779993353273513,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.802139747709945e-05,
+      "loss": 0.2254,
+      "step": 3745
+    },
+    {
+      "epoch": 0.17803731662156388,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.8014148195187226e-05,
+      "loss": 0.2304,
+      "step": 3750
+    },
+    {
+      "epoch": 0.17827469971039264,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.800688627049874e-05,
+      "loss": 0.2265,
+      "step": 3755
+    },
+    {
+      "epoch": 0.17851208279922137,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.799961170750951e-05,
+      "loss": 0.2326,
+      "step": 3760
+    },
+    {
+      "epoch": 0.17874946588805013,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.79923245107028e-05,
+      "loss": 0.2303,
+      "step": 3765
+    },
+    {
+      "epoch": 0.1789868489768789,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.798502468456967e-05,
+      "loss": 0.2338,
+      "step": 3770
+    },
+    {
+      "epoch": 0.17922423206570764,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.797771223360899e-05,
+      "loss": 0.2258,
+      "step": 3775
+    },
+    {
+      "epoch": 0.1794616151545364,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.797038716232737e-05,
+      "loss": 0.2305,
+      "step": 3780
+    },
+    {
+      "epoch": 0.17969899824336513,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.796304947523923e-05,
+      "loss": 0.2272,
+      "step": 3785
+    },
+    {
+      "epoch": 0.1799363813321939,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.795569917686676e-05,
+      "loss": 0.2268,
+      "step": 3790
+    },
+    {
+      "epoch": 0.18017376442102265,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.7948336271739894e-05,
+      "loss": 0.2324,
+      "step": 3795
+    },
+    {
+      "epoch": 0.1804111475098514,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.794096076439638e-05,
+      "loss": 0.2293,
+      "step": 3800
+    },
+    {
+      "epoch": 0.18064853059868016,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.793357265938169e-05,
+      "loss": 0.2272,
+      "step": 3805
+    },
+    {
+      "epoch": 0.1808859136875089,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.792617196124908e-05,
+      "loss": 0.2255,
+      "step": 3810
+    },
+    {
+      "epoch": 0.18112329677633765,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.7918758674559595e-05,
+      "loss": 0.2314,
+      "step": 3815
+    },
+    {
+      "epoch": 0.1813606798651664,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.791133280388198e-05,
+      "loss": 0.2269,
+      "step": 3820
+    },
+    {
+      "epoch": 0.18159806295399517,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.7903894353792764e-05,
+      "loss": 0.2316,
+      "step": 3825
+    },
+    {
+      "epoch": 0.18183544604282392,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.789644332887626e-05,
+      "loss": 0.2293,
+      "step": 3830
+    },
+    {
+      "epoch": 0.18207282913165265,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.788897973372447e-05,
+      "loss": 0.2305,
+      "step": 3835
+    },
+    {
+      "epoch": 0.1823102122204814,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.7881503572937186e-05,
+      "loss": 0.2316,
+      "step": 3840
+    },
+    {
+      "epoch": 0.18254759530931017,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.787401485112193e-05,
+      "loss": 0.2306,
+      "step": 3845
+    },
+    {
+      "epoch": 0.18278497839813893,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.7866513572893975e-05,
+      "loss": 0.2281,
+      "step": 3850
+    },
+    {
+      "epoch": 0.18302236148696766,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.7858999742876315e-05,
+      "loss": 0.2322,
+      "step": 3855
+    },
+    {
+      "epoch": 0.1832597445757964,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.785147336569969e-05,
+      "loss": 0.2298,
+      "step": 3860
+    },
+    {
+      "epoch": 0.18349712766462517,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.784393444600257e-05,
+      "loss": 0.2296,
+      "step": 3865
+    },
+    {
+      "epoch": 0.18373451075345393,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.783638298843117e-05,
+      "loss": 0.2273,
+      "step": 3870
+    },
+    {
+      "epoch": 0.18397189384228269,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.7828818997639404e-05,
+      "loss": 0.2319,
+      "step": 3875
+    },
+    {
+      "epoch": 0.18420927693111142,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.7821242478288935e-05,
+      "loss": 0.2291,
+      "step": 3880
+    },
+    {
+      "epoch": 0.18444666001994017,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.781365343504913e-05,
+      "loss": 0.2254,
+      "step": 3885
+    },
+    {
+      "epoch": 0.18468404310876893,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.7806051872597095e-05,
+      "loss": 0.2258,
+      "step": 3890
+    },
+    {
+      "epoch": 0.1849214261975977,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.779843779561762e-05,
+      "loss": 0.2317,
+      "step": 3895
+    },
+    {
+      "epoch": 0.18515880928642645,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.7790811208803244e-05,
+      "loss": 0.2298,
+      "step": 3900
+    },
+    {
+      "epoch": 0.18539619237525518,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.778317211685418e-05,
+      "loss": 0.2306,
+      "step": 3905
+    },
+    {
+      "epoch": 0.18563357546408393,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.7775520524478374e-05,
+      "loss": 0.2299,
+      "step": 3910
+    },
+    {
+      "epoch": 0.1858709585529127,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.7767856436391476e-05,
+      "loss": 0.23,
+      "step": 3915
+    },
+    {
+      "epoch": 0.18610834164174145,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.7760179857316815e-05,
+      "loss": 0.2305,
+      "step": 3920
+    },
+    {
+      "epoch": 0.1863457247305702,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.775249079198544e-05,
+      "loss": 0.2336,
+      "step": 3925
+    },
+    {
+      "epoch": 0.18658310781939894,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.774478924513608e-05,
+      "loss": 0.2272,
+      "step": 3930
+    },
+    {
+      "epoch": 0.1868204909082277,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.773707522151517e-05,
+      "loss": 0.2272,
+      "step": 3935
+    },
+    {
+      "epoch": 0.18705787399705645,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.7729348725876826e-05,
+      "loss": 0.2286,
+      "step": 3940
+    },
+    {
+      "epoch": 0.1872952570858852,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.7721609762982856e-05,
+      "loss": 0.2294,
+      "step": 3945
+    },
+    {
+      "epoch": 0.18753264017471397,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.771385833760274e-05,
+      "loss": 0.2244,
+      "step": 3950
+    },
+    {
+      "epoch": 0.1877700232635427,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.770609445451364e-05,
+      "loss": 0.2311,
+      "step": 3955
+    },
+    {
+      "epoch": 0.18800740635237145,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.7698318118500424e-05,
+      "loss": 0.2287,
+      "step": 3960
+    },
+    {
+      "epoch": 0.1882447894412002,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.7690529334355596e-05,
+      "loss": 0.2299,
+      "step": 3965
+    },
+    {
+      "epoch": 0.18848217253002897,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.7682728106879356e-05,
+      "loss": 0.2289,
+      "step": 3970
+    },
+    {
+      "epoch": 0.18871955561885773,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.767491444087955e-05,
+      "loss": 0.2291,
+      "step": 3975
+    },
+    {
+      "epoch": 0.18895693870768646,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.766708834117172e-05,
+      "loss": 0.2285,
+      "step": 3980
+    },
+    {
+      "epoch": 0.1891943217965152,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.7659249812579055e-05,
+      "loss": 0.2304,
+      "step": 3985
+    },
+    {
+      "epoch": 0.18943170488534397,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.765139885993241e-05,
+      "loss": 0.228,
+      "step": 3990
+    },
+    {
+      "epoch": 0.18966908797417273,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.7643535488070265e-05,
+      "loss": 0.2315,
+      "step": 3995
+    },
+    {
+      "epoch": 0.18990647106300146,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.7635659701838805e-05,
+      "loss": 0.2295,
+      "step": 4000
+    },
+    {
+      "epoch": 0.19014385415183022,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.762777150609183e-05,
+      "loss": 0.2306,
+      "step": 4005
+    },
+    {
+      "epoch": 0.19038123724065897,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.761987090569081e-05,
+      "loss": 0.2298,
+      "step": 4010
+    },
+    {
+      "epoch": 0.19061862032948773,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.761195790550484e-05,
+      "loss": 0.2285,
+      "step": 4015
+    },
+    {
+      "epoch": 0.1908560034183165,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.760403251041067e-05,
+      "loss": 0.2255,
+      "step": 4020
+    },
+    {
+      "epoch": 0.19109338650714522,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.7596094725292676e-05,
+      "loss": 0.2316,
+      "step": 4025
+    },
+    {
+      "epoch": 0.19133076959597398,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.7588144555042896e-05,
+      "loss": 0.2316,
+      "step": 4030
+    },
+    {
+      "epoch": 0.19156815268480273,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.758018200456097e-05,
+      "loss": 0.2295,
+      "step": 4035
+    },
+    {
+      "epoch": 0.1918055357736315,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.757220707875418e-05,
+      "loss": 0.2289,
+      "step": 4040
+    },
+    {
+      "epoch": 0.19204291886246025,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.756421978253745e-05,
+      "loss": 0.2285,
+      "step": 4045
+    },
+    {
+      "epoch": 0.19228030195128898,
+      "grad_norm": 0.5,
+      "learning_rate": 4.755622012083331e-05,
+      "loss": 0.2279,
+      "step": 4050
+    },
+    {
+      "epoch": 0.19251768504011774,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.75482080985719e-05,
+      "loss": 0.2301,
+      "step": 4055
+    },
+    {
+      "epoch": 0.1927550681289465,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.7540183720691006e-05,
+      "loss": 0.2295,
+      "step": 4060
+    },
+    {
+      "epoch": 0.19299245121777525,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.753214699213602e-05,
+      "loss": 0.2277,
+      "step": 4065
+    },
+    {
+      "epoch": 0.193229834306604,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.752409791785993e-05,
+      "loss": 0.2294,
+      "step": 4070
+    },
+    {
+      "epoch": 0.19346721739543274,
+      "grad_norm": 0.625,
+      "learning_rate": 4.751603650282335e-05,
+      "loss": 0.229,
+      "step": 4075
+    },
+    {
+      "epoch": 0.1937046004842615,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.75079627519945e-05,
+      "loss": 0.2295,
+      "step": 4080
+    },
+    {
+      "epoch": 0.19394198357309025,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.749987667034918e-05,
+      "loss": 0.229,
+      "step": 4085
+    },
+    {
+      "epoch": 0.194179366661919,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.7491778262870825e-05,
+      "loss": 0.2263,
+      "step": 4090
+    },
+    {
+      "epoch": 0.19441674975074777,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.748366753455042e-05,
+      "loss": 0.2257,
+      "step": 4095
+    },
+    {
+      "epoch": 0.1946541328395765,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.74755444903866e-05,
+      "loss": 0.2286,
+      "step": 4100
+    },
+    {
+      "epoch": 0.19489151592840526,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.7467409135385535e-05,
+      "loss": 0.2297,
+      "step": 4105
+    },
+    {
+      "epoch": 0.19512889901723401,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.745926147456102e-05,
+      "loss": 0.2269,
+      "step": 4110
+    },
+    {
+      "epoch": 0.19536628210606277,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.745110151293442e-05,
+      "loss": 0.2268,
+      "step": 4115
+    },
+    {
+      "epoch": 0.19560366519489153,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.744292925553468e-05,
+      "loss": 0.227,
+      "step": 4120
+    },
+    {
+      "epoch": 0.19584104828372026,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.7434744707398335e-05,
+      "loss": 0.2252,
+      "step": 4125
+    },
+    {
+      "epoch": 0.19607843137254902,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.7426547873569466e-05,
+      "loss": 0.2294,
+      "step": 4130
+    },
+    {
+      "epoch": 0.19631581446137777,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.741833875909977e-05,
+      "loss": 0.2268,
+      "step": 4135
+    },
+    {
+      "epoch": 0.19655319755020653,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.7410117369048455e-05,
+      "loss": 0.2282,
+      "step": 4140
+    },
+    {
+      "epoch": 0.19679058063903526,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.7401883708482356e-05,
+      "loss": 0.23,
+      "step": 4145
+    },
+    {
+      "epoch": 0.19702796372786402,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.739363778247582e-05,
+      "loss": 0.2262,
+      "step": 4150
+    },
+    {
+      "epoch": 0.19726534681669278,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.738537959611078e-05,
+      "loss": 0.2208,
+      "step": 4155
+    },
+    {
+      "epoch": 0.19750272990552153,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.737710915447672e-05,
+      "loss": 0.2284,
+      "step": 4160
+    },
+    {
+      "epoch": 0.1977401129943503,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.736882646267068e-05,
+      "loss": 0.2269,
+      "step": 4165
+    },
+    {
+      "epoch": 0.19797749608317902,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.736053152579724e-05,
+      "loss": 0.2254,
+      "step": 4170
+    },
+    {
+      "epoch": 0.19821487917200778,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.735222434896853e-05,
+      "loss": 0.2257,
+      "step": 4175
+    },
+    {
+      "epoch": 0.19845226226083654,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.7343904937304233e-05,
+      "loss": 0.2272,
+      "step": 4180
+    },
+    {
+      "epoch": 0.1986896453496653,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.7335573295931554e-05,
+      "loss": 0.2289,
+      "step": 4185
+    },
+    {
+      "epoch": 0.19892702843849405,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.732722942998526e-05,
+      "loss": 0.2271,
+      "step": 4190
+    },
+    {
+      "epoch": 0.19916441152732278,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.731887334460762e-05,
+      "loss": 0.2315,
+      "step": 4195
+    },
+    {
+      "epoch": 0.19940179461615154,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.731050504494847e-05,
+      "loss": 0.2297,
+      "step": 4200
+    },
+    {
+      "epoch": 0.1996391777049803,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.7302124536165135e-05,
+      "loss": 0.2288,
+      "step": 4205
+    },
+    {
+      "epoch": 0.19987656079380905,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.72937318234225e-05,
+      "loss": 0.2285,
+      "step": 4210
+    },
+    {
+      "epoch": 0.2001139438826378,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.7285326911892954e-05,
+      "loss": 0.2306,
+      "step": 4215
+    },
+    {
+      "epoch": 0.20035132697146654,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.7276909806756394e-05,
+      "loss": 0.2276,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2005887100602953,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.7268480513200256e-05,
+      "loss": 0.2269,
+      "step": 4225
+    },
+    {
+      "epoch": 0.20082609314912406,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.7260039036419474e-05,
+      "loss": 0.229,
+      "step": 4230
+    },
+    {
+      "epoch": 0.20106347623795282,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.7251585381616484e-05,
+      "loss": 0.2288,
+      "step": 4235
+    },
+    {
+      "epoch": 0.20130085932678157,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.724311955400125e-05,
+      "loss": 0.2269,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2015382424156103,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.723464155879121e-05,
+      "loss": 0.2296,
+      "step": 4245
+    },
+    {
+      "epoch": 0.20177562550443906,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.722615140121131e-05,
+      "loss": 0.2261,
+      "step": 4250
+    },
+    {
+      "epoch": 0.20201300859326782,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.721764908649401e-05,
+      "loss": 0.2286,
+      "step": 4255
+    },
+    {
+      "epoch": 0.20225039168209658,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.720913461987923e-05,
+      "loss": 0.2282,
+      "step": 4260
+    },
+    {
+      "epoch": 0.20248777477092533,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.720060800661442e-05,
+      "loss": 0.2294,
+      "step": 4265
+    },
+    {
+      "epoch": 0.20272515785975406,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.719206925195448e-05,
+      "loss": 0.2297,
+      "step": 4270
+    },
+    {
+      "epoch": 0.20296254094858282,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.718351836116179e-05,
+      "loss": 0.2333,
+      "step": 4275
+    },
+    {
+      "epoch": 0.20319992403741158,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.717495533950625e-05,
+      "loss": 0.226,
+      "step": 4280
+    },
+    {
+      "epoch": 0.20343730712624034,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.7166380192265205e-05,
+      "loss": 0.2273,
+      "step": 4285
+    },
+    {
+      "epoch": 0.20367469021506907,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.7157792924723476e-05,
+      "loss": 0.2269,
+      "step": 4290
+    },
+    {
+      "epoch": 0.20391207330389782,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.714919354217336e-05,
+      "loss": 0.2253,
+      "step": 4295
+    },
+    {
+      "epoch": 0.20414945639272658,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.714058204991462e-05,
+      "loss": 0.2273,
+      "step": 4300
+    },
+    {
+      "epoch": 0.20438683948155534,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.7131958453254466e-05,
+      "loss": 0.2282,
+      "step": 4305
+    },
+    {
+      "epoch": 0.2046242225703841,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.7123322757507586e-05,
+      "loss": 0.2282,
+      "step": 4310
+    },
+    {
+      "epoch": 0.20486160565921283,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.711467496799613e-05,
+      "loss": 0.231,
+      "step": 4315
+    },
+    {
+      "epoch": 0.20509898874804158,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.7106015090049686e-05,
+      "loss": 0.2252,
+      "step": 4320
+    },
+    {
+      "epoch": 0.20533637183687034,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.70973431290053e-05,
+      "loss": 0.2255,
+      "step": 4325
+    },
+    {
+      "epoch": 0.2055737549256991,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.708865909020745e-05,
+      "loss": 0.2289,
+      "step": 4330
+    },
+    {
+      "epoch": 0.20581113801452786,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.7079962979008086e-05,
+      "loss": 0.2311,
+      "step": 4335
+    },
+    {
+      "epoch": 0.20604852110335659,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.707125480076657e-05,
+      "loss": 0.2235,
+      "step": 4340
+    },
+    {
+      "epoch": 0.20628590419218534,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.706253456084972e-05,
+      "loss": 0.2303,
+      "step": 4345
+    },
+    {
+      "epoch": 0.2065232872810141,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.705380226463178e-05,
+      "loss": 0.2319,
+      "step": 4350
+    },
+    {
+      "epoch": 0.20676067036984286,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.704505791749441e-05,
+      "loss": 0.2297,
+      "step": 4355
+    },
+    {
+      "epoch": 0.20699805345867162,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.7036301524826735e-05,
+      "loss": 0.223,
+      "step": 4360
+    },
+    {
+      "epoch": 0.20723543654750035,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.702753309202526e-05,
+      "loss": 0.2294,
+      "step": 4365
+    },
+    {
+      "epoch": 0.2074728196363291,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.701875262449395e-05,
+      "loss": 0.227,
+      "step": 4370
+    },
+    {
+      "epoch": 0.20771020272515786,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.700996012764414e-05,
+      "loss": 0.2267,
+      "step": 4375
+    },
+    {
+      "epoch": 0.20794758581398662,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.700115560689463e-05,
+      "loss": 0.2278,
+      "step": 4380
+    },
+    {
+      "epoch": 0.20818496890281538,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.69923390676716e-05,
+      "loss": 0.225,
+      "step": 4385
+    },
+    {
+      "epoch": 0.2084223519916441,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.698351051540864e-05,
+      "loss": 0.2291,
+      "step": 4390
+    },
+    {
+      "epoch": 0.20865973508047286,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.697466995554674e-05,
+      "loss": 0.2262,
+      "step": 4395
+    },
+    {
+      "epoch": 0.20889711816930162,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.696581739353431e-05,
+      "loss": 0.2292,
+      "step": 4400
+    },
+    {
+      "epoch": 0.20913450125813038,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.6956952834827135e-05,
+      "loss": 0.2275,
+      "step": 4405
+    },
+    {
+      "epoch": 0.20937188434695914,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.694807628488841e-05,
+      "loss": 0.2303,
+      "step": 4410
+    },
+    {
+      "epoch": 0.20960926743578787,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.69391877491887e-05,
+      "loss": 0.2284,
+      "step": 4415
+    },
+    {
+      "epoch": 0.20984665052461662,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.6930287233205976e-05,
+      "loss": 0.2243,
+      "step": 4420
+    },
+    {
+      "epoch": 0.21008403361344538,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.692137474242559e-05,
+      "loss": 0.2245,
+      "step": 4425
+    },
+    {
+      "epoch": 0.21032141670227414,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.6912450282340264e-05,
+      "loss": 0.2262,
+      "step": 4430
+    },
+    {
+      "epoch": 0.21055879979110287,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.690351385845011e-05,
+      "loss": 0.2301,
+      "step": 4435
+    },
+    {
+      "epoch": 0.21079618287993163,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.689456547626259e-05,
+      "loss": 0.2308,
+      "step": 4440
+    },
+    {
+      "epoch": 0.21103356596876038,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.688560514129257e-05,
+      "loss": 0.2298,
+      "step": 4445
+    },
+    {
+      "epoch": 0.21127094905758914,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.687663285906225e-05,
+      "loss": 0.2281,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2115083321464179,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.68676486351012e-05,
+      "loss": 0.2244,
+      "step": 4455
+    },
+    {
+      "epoch": 0.21174571523524663,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.685865247494638e-05,
+      "loss": 0.2328,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2119830983240754,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.684964438414206e-05,
+      "loss": 0.2292,
+      "step": 4465
+    },
+    {
+      "epoch": 0.21222048141290414,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.68406243682399e-05,
+      "loss": 0.2289,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2124578645017329,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.6831592432798896e-05,
+      "loss": 0.2285,
+      "step": 4475
+    },
+    {
+      "epoch": 0.21269524759056166,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.682254858338537e-05,
+      "loss": 0.2246,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2129326306793904,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.681349282557303e-05,
+      "loss": 0.2295,
+      "step": 4485
+    },
+    {
+      "epoch": 0.21317001376821915,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.680442516494287e-05,
+      "loss": 0.2287,
+      "step": 4490
+    },
+    {
+      "epoch": 0.2134073968570479,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.679534560708328e-05,
+      "loss": 0.2267,
+      "step": 4495
+    },
+    {
+      "epoch": 0.21364477994587666,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.678625415758993e-05,
+      "loss": 0.2313,
+      "step": 4500
+    },
+    {
+      "epoch": 0.21388216303470542,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.677715082206584e-05,
+      "loss": 0.2301,
+      "step": 4505
+    },
+    {
+      "epoch": 0.21411954612353415,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.676803560612136e-05,
+      "loss": 0.2326,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2143569292123629,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.675890851537415e-05,
+      "loss": 0.2284,
+      "step": 4515
+    },
+    {
+      "epoch": 0.21459431230119166,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.6749769555449214e-05,
+      "loss": 0.2237,
+      "step": 4520
+    },
+    {
+      "epoch": 0.21483169539002042,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.674061873197882e-05,
+      "loss": 0.2293,
+      "step": 4525
+    },
+    {
+      "epoch": 0.21506907847884918,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.67314560506026e-05,
+      "loss": 0.2303,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2153064615676779,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.672228151696747e-05,
+      "loss": 0.2267,
+      "step": 4535
+    },
+    {
+      "epoch": 0.21554384465650667,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.671309513672765e-05,
+      "loss": 0.2324,
+      "step": 4540
+    },
+    {
+      "epoch": 0.21578122774533542,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.670389691554466e-05,
+      "loss": 0.2294,
+      "step": 4545
+    },
+    {
+      "epoch": 0.21601861083416418,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.6694686859087324e-05,
+      "loss": 0.2281,
+      "step": 4550
+    },
+    {
+      "epoch": 0.21625599392299294,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.668546497303177e-05,
+      "loss": 0.2282,
+      "step": 4555
+    },
+    {
+      "epoch": 0.21649337701182167,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.6676231263061386e-05,
+      "loss": 0.2269,
+      "step": 4560
+    },
+    {
+      "epoch": 0.21673076010065043,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.6666985734866884e-05,
+      "loss": 0.2301,
+      "step": 4565
+    },
+    {
+      "epoch": 0.21696814318947918,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.665772839414623e-05,
+      "loss": 0.2258,
+      "step": 4570
+    },
+    {
+      "epoch": 0.21720552627830794,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.664845924660469e-05,
+      "loss": 0.225,
+      "step": 4575
+    },
+    {
+      "epoch": 0.21744290936713667,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.663917829795478e-05,
+      "loss": 0.2323,
+      "step": 4580
+    },
+    {
+      "epoch": 0.21768029245596543,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.662988555391632e-05,
+      "loss": 0.2253,
+      "step": 4585
+    },
+    {
+      "epoch": 0.2179176755447942,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.662058102021638e-05,
+      "loss": 0.2253,
+      "step": 4590
+    },
+    {
+      "epoch": 0.21815505863362294,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.661126470258931e-05,
+      "loss": 0.229,
+      "step": 4595
+    },
+    {
+      "epoch": 0.2183924417224517,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.660193660677671e-05,
+      "loss": 0.2258,
+      "step": 4600
+    },
+    {
+      "epoch": 0.21862982481128043,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.659259673852743e-05,
+      "loss": 0.2262,
+      "step": 4605
+    },
+    {
+      "epoch": 0.2188672079001092,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.658324510359762e-05,
+      "loss": 0.2272,
+      "step": 4610
+    },
+    {
+      "epoch": 0.21910459098893795,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.6573881707750616e-05,
+      "loss": 0.2266,
+      "step": 4615
+    },
+    {
+      "epoch": 0.2193419740777667,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.6564506556757054e-05,
+      "loss": 0.2307,
+      "step": 4620
+    },
+    {
+      "epoch": 0.21957935716659546,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.655511965639479e-05,
+      "loss": 0.229,
+      "step": 4625
+    },
+    {
+      "epoch": 0.2198167402554242,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.654572101244893e-05,
+      "loss": 0.2278,
+      "step": 4630
+    },
+    {
+      "epoch": 0.22005412334425295,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.653631063071182e-05,
+      "loss": 0.2274,
+      "step": 4635
+    },
+    {
+      "epoch": 0.2202915064330817,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.652688851698304e-05,
+      "loss": 0.2275,
+      "step": 4640
+    },
+    {
+      "epoch": 0.22052888952191047,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.651745467706938e-05,
+      "loss": 0.2302,
+      "step": 4645
+    },
+    {
+      "epoch": 0.22076627261073922,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.6508009116784885e-05,
+      "loss": 0.2227,
+      "step": 4650
+    },
+    {
+      "epoch": 0.22100365569956795,
+      "grad_norm": 0.5,
+      "learning_rate": 4.6498551841950805e-05,
+      "loss": 0.2262,
+      "step": 4655
+    },
+    {
+      "epoch": 0.2212410387883967,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.6489082858395616e-05,
+      "loss": 0.2295,
+      "step": 4660
+    },
+    {
+      "epoch": 0.22147842187722547,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.6479602171955014e-05,
+      "loss": 0.2285,
+      "step": 4665
+    },
+    {
+      "epoch": 0.22171580496605423,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.6470109788471893e-05,
+      "loss": 0.2283,
+      "step": 4670
+    },
+    {
+      "epoch": 0.22195318805488298,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.6460605713796387e-05,
+      "loss": 0.232,
+      "step": 4675
+    },
+    {
+      "epoch": 0.2221905711437117,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.645108995378578e-05,
+      "loss": 0.2285,
+      "step": 4680
+    },
+    {
+      "epoch": 0.22242795423254047,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.6441562514304625e-05,
+      "loss": 0.2317,
+      "step": 4685
+    },
+    {
+      "epoch": 0.22266533732136923,
+      "grad_norm": 0.5,
+      "learning_rate": 4.643202340122462e-05,
+      "loss": 0.2293,
+      "step": 4690
+    },
+    {
+      "epoch": 0.22290272041019799,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.6422472620424686e-05,
+      "loss": 0.2281,
+      "step": 4695
+    },
+    {
+      "epoch": 0.22314010349902674,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.6412910177790926e-05,
+      "loss": 0.2294,
+      "step": 4700
+    },
+    {
+      "epoch": 0.22337748658785547,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.640333607921662e-05,
+      "loss": 0.2274,
+      "step": 4705
+    },
+    {
+      "epoch": 0.22361486967668423,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.6393750330602244e-05,
+      "loss": 0.2251,
+      "step": 4710
+    },
+    {
+      "epoch": 0.223852252765513,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.638415293785546e-05,
+      "loss": 0.2318,
+      "step": 4715
+    },
+    {
+      "epoch": 0.22408963585434175,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.637454390689109e-05,
+      "loss": 0.2274,
+      "step": 4720
+    },
+    {
+      "epoch": 0.22432701894317048,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.636492324363115e-05,
+      "loss": 0.2278,
+      "step": 4725
+    },
+    {
+      "epoch": 0.22456440203199923,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.635529095400479e-05,
+      "loss": 0.2314,
+      "step": 4730
+    },
+    {
+      "epoch": 0.224801785120828,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.634564704394836e-05,
+      "loss": 0.226,
+      "step": 4735
+    },
+    {
+      "epoch": 0.22503916820965675,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.6335991519405354e-05,
+      "loss": 0.2237,
+      "step": 4740
+    },
+    {
+      "epoch": 0.2252765512984855,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.6326324386326424e-05,
+      "loss": 0.2302,
+      "step": 4745
+    },
+    {
+      "epoch": 0.22551393438731424,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.6316645650669385e-05,
+      "loss": 0.2322,
+      "step": 4750
+    },
+    {
+      "epoch": 0.225751317476143,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.630695531839919e-05,
+      "loss": 0.2277,
+      "step": 4755
+    },
+    {
+      "epoch": 0.22598870056497175,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.629725339548796e-05,
+      "loss": 0.2301,
+      "step": 4760
+    },
+    {
+      "epoch": 0.2262260836538005,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.628753988791495e-05,
+      "loss": 0.2273,
+      "step": 4765
+    },
+    {
+      "epoch": 0.22646346674262927,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.6277814801666524e-05,
+      "loss": 0.2294,
+      "step": 4770
+    },
+    {
+      "epoch": 0.226700849831458,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.626807814273624e-05,
+      "loss": 0.229,
+      "step": 4775
+    },
+    {
+      "epoch": 0.22693823292028675,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.6258329917124724e-05,
+      "loss": 0.2263,
+      "step": 4780
+    },
+    {
+      "epoch": 0.2271756160091155,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.624857013083979e-05,
+      "loss": 0.2241,
+      "step": 4785
+    },
+    {
+      "epoch": 0.22741299909794427,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.6238798789896336e-05,
+      "loss": 0.2261,
+      "step": 4790
+    },
+    {
+      "epoch": 0.22765038218677303,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.6229015900316404e-05,
+      "loss": 0.2291,
+      "step": 4795
+    },
+    {
+      "epoch": 0.22788776527560176,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.621922146812914e-05,
+      "loss": 0.2269,
+      "step": 4800
+    },
+    {
+      "epoch": 0.2281251483644305,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.62094154993708e-05,
+      "loss": 0.2291,
+      "step": 4805
+    },
+    {
+      "epoch": 0.22836253145325927,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.619959800008477e-05,
+      "loss": 0.2288,
+      "step": 4810
+    },
+    {
+      "epoch": 0.22859991454208803,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.6189768976321524e-05,
+      "loss": 0.2298,
+      "step": 4815
+    },
+    {
+      "epoch": 0.22883729763091679,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.6179928434138644e-05,
+      "loss": 0.2257,
+      "step": 4820
+    },
+    {
+      "epoch": 0.22907468071974552,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.6170076379600825e-05,
+      "loss": 0.2261,
+      "step": 4825
+    },
+    {
+      "epoch": 0.22931206380857427,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.616021281877982e-05,
+      "loss": 0.2284,
+      "step": 4830
+    },
+    {
+      "epoch": 0.22954944689740303,
+      "grad_norm": 0.5,
+      "learning_rate": 4.6150337757754515e-05,
+      "loss": 0.2277,
+      "step": 4835
+    },
+    {
+      "epoch": 0.2297868299862318,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.614045120261086e-05,
+      "loss": 0.2285,
+      "step": 4840
+    },
+    {
+      "epoch": 0.23002421307506055,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.6130553159441884e-05,
+      "loss": 0.2249,
+      "step": 4845
+    },
+    {
+      "epoch": 0.23026159616388928,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.612064363434772e-05,
+      "loss": 0.2274,
+      "step": 4850
+    },
+    {
+      "epoch": 0.23049897925271803,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.611072263343556e-05,
+      "loss": 0.2306,
+      "step": 4855
+    },
+    {
+      "epoch": 0.2307363623415468,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.610079016281967e-05,
+      "loss": 0.2259,
+      "step": 4860
+    },
+    {
+      "epoch": 0.23097374543037555,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.6090846228621384e-05,
+      "loss": 0.2279,
+      "step": 4865
+    },
+    {
+      "epoch": 0.23121112851920428,
+      "grad_norm": 0.400390625,
+      "learning_rate": 4.608089083696911e-05,
+      "loss": 0.2265,
+      "step": 4870
+    },
+    {
+      "epoch": 0.23144851160803304,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.60709239939983e-05,
+      "loss": 0.2239,
+      "step": 4875
+    },
+    {
+      "epoch": 0.2316858946968618,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.606094570585149e-05,
+      "loss": 0.2281,
+      "step": 4880
+    },
+    {
+      "epoch": 0.23192327778569055,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.605095597867823e-05,
+      "loss": 0.2284,
+      "step": 4885
+    },
+    {
+      "epoch": 0.2321606608745193,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.604095481863517e-05,
+      "loss": 0.2241,
+      "step": 4890
+    },
+    {
+      "epoch": 0.23239804396334804,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.603094223188596e-05,
+      "loss": 0.2299,
+      "step": 4895
+    },
+    {
+      "epoch": 0.2326354270521768,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.6020918224601324e-05,
+      "loss": 0.2255,
+      "step": 4900
+    },
+    {
+      "epoch": 0.23287281014100555,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.6010882802959e-05,
+      "loss": 0.2247,
+      "step": 4905
+    },
+    {
+      "epoch": 0.2331101932298343,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.6000835973143784e-05,
+      "loss": 0.2229,
+      "step": 4910
+    },
+    {
+      "epoch": 0.23334757631866307,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.5990777741347485e-05,
+      "loss": 0.229,
+      "step": 4915
+    },
+    {
+      "epoch": 0.2335849594074918,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.598070811376895e-05,
+      "loss": 0.2275,
+      "step": 4920
+    },
+    {
+      "epoch": 0.23382234249632056,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.597062709661404e-05,
+      "loss": 0.227,
+      "step": 4925
+    },
+    {
+      "epoch": 0.23405972558514931,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.596053469609564e-05,
+      "loss": 0.2277,
+      "step": 4930
+    },
+    {
+      "epoch": 0.23429710867397807,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.595043091843366e-05,
+      "loss": 0.2272,
+      "step": 4935
+    },
+    {
+      "epoch": 0.23453449176280683,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.5940315769855004e-05,
+      "loss": 0.2284,
+      "step": 4940
+    },
+    {
+      "epoch": 0.23477187485163556,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.593018925659359e-05,
+      "loss": 0.2277,
+      "step": 4945
+    },
+    {
+      "epoch": 0.23500925794046432,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.592005138489034e-05,
+      "loss": 0.2299,
+      "step": 4950
+    },
+    {
+      "epoch": 0.23524664102929307,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.590990216099319e-05,
+      "loss": 0.2252,
+      "step": 4955
+    },
+    {
+      "epoch": 0.23548402411812183,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.589974159115705e-05,
+      "loss": 0.2275,
+      "step": 4960
+    },
+    {
+      "epoch": 0.2357214072069506,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.588956968164383e-05,
+      "loss": 0.227,
+      "step": 4965
+    },
+    {
+      "epoch": 0.23595879029577932,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.587938643872246e-05,
+      "loss": 0.2301,
+      "step": 4970
+    },
+    {
+      "epoch": 0.23619617338460808,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.586919186866879e-05,
+      "loss": 0.2333,
+      "step": 4975
+    },
+    {
+      "epoch": 0.23643355647343683,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.58589859777657e-05,
+      "loss": 0.2302,
+      "step": 4980
+    },
+    {
+      "epoch": 0.2366709395622656,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.584876877230304e-05,
+      "loss": 0.2266,
+      "step": 4985
+    },
+    {
+      "epoch": 0.23690832265109435,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.5838540258577625e-05,
+      "loss": 0.2286,
+      "step": 4990
+    },
+    {
+      "epoch": 0.23714570573992308,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.5828300442893236e-05,
+      "loss": 0.2284,
+      "step": 4995
+    },
+    {
+      "epoch": 0.23738308882875184,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.5818049331560634e-05,
+      "loss": 0.2307,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2376204719175806,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.580778693089753e-05,
+      "loss": 0.2271,
+      "step": 5005
+    },
+    {
+      "epoch": 0.23785785500640935,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.579751324722859e-05,
+      "loss": 0.2276,
+      "step": 5010
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.578722828688543e-05,
+      "loss": 0.2251,
+      "step": 5015
+    },
+    {
+      "epoch": 0.23833262118406684,
+      "grad_norm": 0.5,
+      "learning_rate": 4.577693205620663e-05,
+      "loss": 0.2299,
+      "step": 5020
+    },
+    {
+      "epoch": 0.2385700042728956,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.576662456153773e-05,
+      "loss": 0.2277,
+      "step": 5025
+    },
+    {
+      "epoch": 0.23880738736172435,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.5756305809231176e-05,
+      "loss": 0.227,
+      "step": 5030
+    },
+    {
+      "epoch": 0.2390447704505531,
+      "grad_norm": 0.5,
+      "learning_rate": 4.5745975805646366e-05,
+      "loss": 0.2311,
+      "step": 5035
+    },
+    {
+      "epoch": 0.23928215353938184,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.573563455714964e-05,
+      "loss": 0.2275,
+      "step": 5040
+    },
+    {
+      "epoch": 0.2395195366282106,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.572528207011425e-05,
+      "loss": 0.2287,
+      "step": 5045
+    },
+    {
+      "epoch": 0.23975691971703936,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.57149183509204e-05,
+      "loss": 0.2253,
+      "step": 5050
+    },
+    {
+      "epoch": 0.23999430280586811,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.570454340595521e-05,
+      "loss": 0.2316,
+      "step": 5055
+    },
+    {
+      "epoch": 0.24023168589469687,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.569415724161269e-05,
+      "loss": 0.2296,
+      "step": 5060
+    },
+    {
+      "epoch": 0.2404690689835256,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.56837598642938e-05,
+      "loss": 0.2238,
+      "step": 5065
+    },
+    {
+      "epoch": 0.24070645207235436,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.567335128040639e-05,
+      "loss": 0.2296,
+      "step": 5070
+    },
+    {
+      "epoch": 0.24094383516118312,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.566293149636523e-05,
+      "loss": 0.2239,
+      "step": 5075
+    },
+    {
+      "epoch": 0.24118121825001188,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.5652500518591975e-05,
+      "loss": 0.2252,
+      "step": 5080
+    },
+    {
+      "epoch": 0.24141860133884063,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.5642058353515204e-05,
+      "loss": 0.231,
+      "step": 5085
+    },
+    {
+      "epoch": 0.24165598442766936,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.5631605007570366e-05,
+      "loss": 0.2297,
+      "step": 5090
+    },
+    {
+      "epoch": 0.24189336751649812,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.5621140487199804e-05,
+      "loss": 0.2278,
+      "step": 5095
+    },
+    {
+      "epoch": 0.24213075060532688,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.561066479885277e-05,
+      "loss": 0.2292,
+      "step": 5100
+    },
+    {
+      "epoch": 0.24236813369415564,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.560017794898538e-05,
+      "loss": 0.2265,
+      "step": 5105
+    },
+    {
+      "epoch": 0.2426055167829844,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.5589679944060616e-05,
+      "loss": 0.2244,
+      "step": 5110
+    },
+    {
+      "epoch": 0.24284289987181312,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.557917079054837e-05,
+      "loss": 0.231,
+      "step": 5115
+    },
+    {
+      "epoch": 0.24308028296064188,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.5568650494925376e-05,
+      "loss": 0.2248,
+      "step": 5120
+    },
+    {
+      "epoch": 0.24331766604947064,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.555811906367524e-05,
+      "loss": 0.2281,
+      "step": 5125
+    },
+    {
+      "epoch": 0.2435550491382994,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.554757650328845e-05,
+      "loss": 0.2311,
+      "step": 5130
+    },
+    {
+      "epoch": 0.24379243222712815,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.553702282026232e-05,
+      "loss": 0.2271,
+      "step": 5135
+    },
+    {
+      "epoch": 0.24402981531595688,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.552645802110105e-05,
+      "loss": 0.2273,
+      "step": 5140
+    },
+    {
+      "epoch": 0.24426719840478564,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.551588211231568e-05,
+      "loss": 0.2277,
+      "step": 5145
+    },
+    {
+      "epoch": 0.2445045814936144,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.5505295100424076e-05,
+      "loss": 0.2287,
+      "step": 5150
+    },
+    {
+      "epoch": 0.24474196458244316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.5494696991950987e-05,
+      "loss": 0.2265,
+      "step": 5155
+    },
+    {
+      "epoch": 0.24497934767127189,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.548408779342797e-05,
+      "loss": 0.2288,
+      "step": 5160
+    },
+    {
+      "epoch": 0.24521673076010064,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.547346751139342e-05,
+      "loss": 0.2246,
+      "step": 5165
+    },
+    {
+      "epoch": 0.2454541138489294,
+      "grad_norm": 0.5,
+      "learning_rate": 4.546283615239258e-05,
+      "loss": 0.2268,
+      "step": 5170
+    },
+    {
+      "epoch": 0.24569149693775816,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.545219372297749e-05,
+      "loss": 0.2308,
+      "step": 5175
+    },
+    {
+      "epoch": 0.24592888002658692,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.544154022970705e-05,
+      "loss": 0.2282,
+      "step": 5180
+    },
+    {
+      "epoch": 0.24616626311541565,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.5430875679146955e-05,
+      "loss": 0.2288,
+      "step": 5185
+    },
+    {
+      "epoch": 0.2464036462042444,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.542020007786972e-05,
+      "loss": 0.2258,
+      "step": 5190
+    },
+    {
+      "epoch": 0.24664102929307316,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.540951343245465e-05,
+      "loss": 0.2248,
+      "step": 5195
+    },
+    {
+      "epoch": 0.24687841238190192,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.539881574948791e-05,
+      "loss": 0.2254,
+      "step": 5200
+    },
+    {
+      "epoch": 0.24711579547073068,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.5388107035562406e-05,
+      "loss": 0.2229,
+      "step": 5205
+    },
+    {
+      "epoch": 0.2473531785595594,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.5377387297277876e-05,
+      "loss": 0.2268,
+      "step": 5210
+    },
+    {
+      "epoch": 0.24759056164838816,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.536665654124085e-05,
+      "loss": 0.2315,
+      "step": 5215
+    },
+    {
+      "epoch": 0.24782794473721692,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.535591477406466e-05,
+      "loss": 0.2258,
+      "step": 5220
+    },
+    {
+      "epoch": 0.24806532782604568,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.534516200236937e-05,
+      "loss": 0.2288,
+      "step": 5225
+    },
+    {
+      "epoch": 0.24830271091487444,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.533439823278189e-05,
+      "loss": 0.2278,
+      "step": 5230
+    },
+    {
+      "epoch": 0.24854009400370317,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.532362347193589e-05,
+      "loss": 0.2234,
+      "step": 5235
+    },
+    {
+      "epoch": 0.24877747709253192,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.531283772647178e-05,
+      "loss": 0.2267,
+      "step": 5240
+    },
+    {
+      "epoch": 0.24901486018136068,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.530204100303678e-05,
+      "loss": 0.2249,
+      "step": 5245
+    },
+    {
+      "epoch": 0.24925224327018944,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.529123330828487e-05,
+      "loss": 0.2257,
+      "step": 5250
+    },
+    {
+      "epoch": 0.2494896263590182,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.5280414648876765e-05,
+      "loss": 0.2289,
+      "step": 5255
+    },
+    {
+      "epoch": 0.24972700944784693,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.526958503147997e-05,
+      "loss": 0.225,
+      "step": 5260
+    },
+    {
+      "epoch": 0.24996439253667568,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.525874446276873e-05,
+      "loss": 0.2296,
+      "step": 5265
+    },
+    {
+      "epoch": 0.25020177562550444,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.524789294942402e-05,
+      "loss": 0.2267,
+      "step": 5270
+    },
+    {
+      "epoch": 0.2504391587143332,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.52370304981336e-05,
+      "loss": 0.2239,
+      "step": 5275
+    },
+    {
+      "epoch": 0.25067654180316196,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.5226157115591937e-05,
+      "loss": 0.2232,
+      "step": 5280
+    },
+    {
+      "epoch": 0.2509139248919907,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.521527280850025e-05,
+      "loss": 0.2255,
+      "step": 5285
+    },
+    {
+      "epoch": 0.25115130798081947,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.5204377583566494e-05,
+      "loss": 0.228,
+      "step": 5290
+    },
+    {
+      "epoch": 0.2513886910696482,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.5193471447505335e-05,
+      "loss": 0.2265,
+      "step": 5295
+    },
+    {
+      "epoch": 0.25162607415847693,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.518255440703818e-05,
+      "loss": 0.2276,
+      "step": 5300
+    },
+    {
+      "epoch": 0.2518634572473057,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.517162646889315e-05,
+      "loss": 0.226,
+      "step": 5305
+    },
+    {
+      "epoch": 0.25210084033613445,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.516068763980509e-05,
+      "loss": 0.2297,
+      "step": 5310
+    },
+    {
+      "epoch": 0.2523382234249632,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.514973792651555e-05,
+      "loss": 0.2279,
+      "step": 5315
+    },
+    {
+      "epoch": 0.25257560651379196,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.5138777335772784e-05,
+      "loss": 0.2283,
+      "step": 5320
+    },
+    {
+      "epoch": 0.2528129896026207,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.512780587433175e-05,
+      "loss": 0.2253,
+      "step": 5325
+    },
+    {
+      "epoch": 0.2530503726914495,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.5116823548954116e-05,
+      "loss": 0.2262,
+      "step": 5330
+    },
+    {
+      "epoch": 0.25328775578027823,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.510583036640824e-05,
+      "loss": 0.2233,
+      "step": 5335
+    },
+    {
+      "epoch": 0.253525138869107,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.5094826333469165e-05,
+      "loss": 0.2256,
+      "step": 5340
+    },
+    {
+      "epoch": 0.2537625219579357,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.5083811456918635e-05,
+      "loss": 0.2276,
+      "step": 5345
+    },
+    {
+      "epoch": 0.25399990504676445,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.5072785743545056e-05,
+      "loss": 0.2266,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2542372881355932,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.5061749200143535e-05,
+      "loss": 0.2227,
+      "step": 5355
+    },
+    {
+      "epoch": 0.25447467122442197,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.505070183351584e-05,
+      "loss": 0.2227,
+      "step": 5360
+    },
+    {
+      "epoch": 0.2547120543132507,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.503964365047042e-05,
+      "loss": 0.2265,
+      "step": 5365
+    },
+    {
+      "epoch": 0.2549494374020795,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.502857465782237e-05,
+      "loss": 0.2265,
+      "step": 5370
+    },
+    {
+      "epoch": 0.25518682049090824,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.5017494862393474e-05,
+      "loss": 0.2262,
+      "step": 5375
+    },
+    {
+      "epoch": 0.255424203579737,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.500640427101215e-05,
+      "loss": 0.2241,
+      "step": 5380
+    },
+    {
+      "epoch": 0.25566158666856575,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.499530289051348e-05,
+      "loss": 0.2267,
+      "step": 5385
+    },
+    {
+      "epoch": 0.25589896975739446,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.4984190727739186e-05,
+      "loss": 0.2299,
+      "step": 5390
+    },
+    {
+      "epoch": 0.2561363528462232,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.497306778953765e-05,
+      "loss": 0.2324,
+      "step": 5395
+    },
+    {
+      "epoch": 0.25637373593505197,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.4961934082763915e-05,
+      "loss": 0.2256,
+      "step": 5400
+    },
+    {
+      "epoch": 0.25661111902388073,
+      "grad_norm": 0.4296875,
+      "learning_rate": 4.495078961427959e-05,
+      "loss": 0.2269,
+      "step": 5405
+    },
+    {
+      "epoch": 0.2568485021127095,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.493963439095299e-05,
+      "loss": 0.229,
+      "step": 5410
+    },
+    {
+      "epoch": 0.25708588520153824,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.492846841965901e-05,
+      "loss": 0.2248,
+      "step": 5415
+    },
+    {
+      "epoch": 0.257323268290367,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.49172917072792e-05,
+      "loss": 0.2239,
+      "step": 5420
+    },
+    {
+      "epoch": 0.25756065137919576,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.4906104260701716e-05,
+      "loss": 0.2291,
+      "step": 5425
+    },
+    {
+      "epoch": 0.2577980344680245,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.489490608682132e-05,
+      "loss": 0.2239,
+      "step": 5430
+    },
+    {
+      "epoch": 0.2580354175568533,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.488369719253941e-05,
+      "loss": 0.2284,
+      "step": 5435
+    },
+    {
+      "epoch": 0.258272800645682,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.4872477584763966e-05,
+      "loss": 0.2245,
+      "step": 5440
+    },
+    {
+      "epoch": 0.25851018373451073,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.486124727040958e-05,
+      "loss": 0.2304,
+      "step": 5445
+    },
+    {
+      "epoch": 0.2587475668233395,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.485000625639746e-05,
+      "loss": 0.2265,
+      "step": 5450
+    },
+    {
+      "epoch": 0.25898494991216825,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.4838754549655376e-05,
+      "loss": 0.223,
+      "step": 5455
+    },
+    {
+      "epoch": 0.259222333000997,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.482749215711771e-05,
+      "loss": 0.2259,
+      "step": 5460
+    },
+    {
+      "epoch": 0.25945971608982576,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.4816219085725404e-05,
+      "loss": 0.227,
+      "step": 5465
+    },
+    {
+      "epoch": 0.2596970991786545,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.4804935342426036e-05,
+      "loss": 0.2301,
+      "step": 5470
+    },
+    {
+      "epoch": 0.2599344822674833,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.4793640934173705e-05,
+      "loss": 0.2292,
+      "step": 5475
+    },
+    {
+      "epoch": 0.26017186535631204,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.4782335867929096e-05,
+      "loss": 0.2212,
+      "step": 5480
+    },
+    {
+      "epoch": 0.2604092484451408,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.477102015065949e-05,
+      "loss": 0.2268,
+      "step": 5485
+    },
+    {
+      "epoch": 0.2606466315339695,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.47596937893387e-05,
+      "loss": 0.2267,
+      "step": 5490
+    },
+    {
+      "epoch": 0.26088401462279825,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.4748356790947114e-05,
+      "loss": 0.2307,
+      "step": 5495
+    },
+    {
+      "epoch": 0.261121397711627,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.4737009162471666e-05,
+      "loss": 0.2288,
+      "step": 5500
+    },
+    {
+      "epoch": 0.26135878080045577,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.4725650910905866e-05,
+      "loss": 0.2252,
+      "step": 5505
+    },
+    {
+      "epoch": 0.2615961638892845,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.471428204324972e-05,
+      "loss": 0.2289,
+      "step": 5510
+    },
+    {
+      "epoch": 0.2618335469781133,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.4702902566509846e-05,
+      "loss": 0.2272,
+      "step": 5515
+    },
+    {
+      "epoch": 0.26207093006694204,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.469151248769934e-05,
+      "loss": 0.2294,
+      "step": 5520
+    },
+    {
+      "epoch": 0.2623083131557708,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.4680111813837864e-05,
+      "loss": 0.2288,
+      "step": 5525
+    },
+    {
+      "epoch": 0.26254569624459956,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.4668700551951607e-05,
+      "loss": 0.2255,
+      "step": 5530
+    },
+    {
+      "epoch": 0.26278307933342826,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.4657278709073266e-05,
+      "loss": 0.2273,
+      "step": 5535
+    },
+    {
+      "epoch": 0.263020462422257,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.464584629224207e-05,
+      "loss": 0.2259,
+      "step": 5540
+    },
+    {
+      "epoch": 0.2632578455110858,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.463440330850378e-05,
+      "loss": 0.2277,
+      "step": 5545
+    },
+    {
+      "epoch": 0.26349522859991453,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.462294976491064e-05,
+      "loss": 0.2275,
+      "step": 5550
+    },
+    {
+      "epoch": 0.2637326116887433,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.461148566852143e-05,
+      "loss": 0.2254,
+      "step": 5555
+    },
+    {
+      "epoch": 0.26396999477757205,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.460001102640141e-05,
+      "loss": 0.2233,
+      "step": 5560
+    },
+    {
+      "epoch": 0.2642073778664008,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.4588525845622356e-05,
+      "loss": 0.2276,
+      "step": 5565
+    },
+    {
+      "epoch": 0.26444476095522956,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.4577030133262534e-05,
+      "loss": 0.2292,
+      "step": 5570
+    },
+    {
+      "epoch": 0.2646821440440583,
+      "grad_norm": 0.5,
+      "learning_rate": 4.4565523896406696e-05,
+      "loss": 0.2266,
+      "step": 5575
+    },
+    {
+      "epoch": 0.2649195271328871,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.455400714214607e-05,
+      "loss": 0.2275,
+      "step": 5580
+    },
+    {
+      "epoch": 0.2651569102217158,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.4542479877578404e-05,
+      "loss": 0.2295,
+      "step": 5585
+    },
+    {
+      "epoch": 0.26539429331054454,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.453094210980788e-05,
+      "loss": 0.2287,
+      "step": 5590
+    },
+    {
+      "epoch": 0.2656316763993733,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.451939384594519e-05,
+      "loss": 0.2249,
+      "step": 5595
+    },
+    {
+      "epoch": 0.26586905948820205,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.450783509310745e-05,
+      "loss": 0.2263,
+      "step": 5600
+    },
+    {
+      "epoch": 0.2661064425770308,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.44962658584183e-05,
+      "loss": 0.2255,
+      "step": 5605
+    },
+    {
+      "epoch": 0.26634382566585957,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.448468614900777e-05,
+      "loss": 0.2258,
+      "step": 5610
+    },
+    {
+      "epoch": 0.2665812087546883,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.4473095972012414e-05,
+      "loss": 0.2264,
+      "step": 5615
+    },
+    {
+      "epoch": 0.2668185918435171,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.4461495334575196e-05,
+      "loss": 0.2255,
+      "step": 5620
+    },
+    {
+      "epoch": 0.26705597493234584,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.4449884243845524e-05,
+      "loss": 0.2286,
+      "step": 5625
+    },
+    {
+      "epoch": 0.2672933580211746,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.443826270697927e-05,
+      "loss": 0.223,
+      "step": 5630
+    },
+    {
+      "epoch": 0.2675307411100033,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.442663073113873e-05,
+      "loss": 0.229,
+      "step": 5635
+    },
+    {
+      "epoch": 0.26776812419883206,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.4414988323492645e-05,
+      "loss": 0.2265,
+      "step": 5640
+    },
+    {
+      "epoch": 0.2680055072876608,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.4403335491216176e-05,
+      "loss": 0.2261,
+      "step": 5645
+    },
+    {
+      "epoch": 0.2682428903764896,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.43916722414909e-05,
+      "loss": 0.2282,
+      "step": 5650
+    },
+    {
+      "epoch": 0.26848027346531833,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.437999858150484e-05,
+      "loss": 0.2318,
+      "step": 5655
+    },
+    {
+      "epoch": 0.2687176565541471,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.436831451845242e-05,
+      "loss": 0.2305,
+      "step": 5660
+    },
+    {
+      "epoch": 0.26895503964297585,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.435662005953446e-05,
+      "loss": 0.2288,
+      "step": 5665
+    },
+    {
+      "epoch": 0.2691924227318046,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.434491521195821e-05,
+      "loss": 0.2265,
+      "step": 5670
+    },
+    {
+      "epoch": 0.26942980582063336,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.433319998293732e-05,
+      "loss": 0.2244,
+      "step": 5675
+    },
+    {
+      "epoch": 0.26966718890946206,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.432147437969182e-05,
+      "loss": 0.2235,
+      "step": 5680
+    },
+    {
+      "epoch": 0.2699045719982908,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.430973840944815e-05,
+      "loss": 0.2283,
+      "step": 5685
+    },
+    {
+      "epoch": 0.2701419550871196,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.429799207943914e-05,
+      "loss": 0.2252,
+      "step": 5690
+    },
+    {
+      "epoch": 0.27037933817594834,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.4286235396904e-05,
+      "loss": 0.2297,
+      "step": 5695
+    },
+    {
+      "epoch": 0.2706167212647771,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.4274468369088315e-05,
+      "loss": 0.2279,
+      "step": 5700
+    },
+    {
+      "epoch": 0.27085410435360585,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.4262691003244056e-05,
+      "loss": 0.2267,
+      "step": 5705
+    },
+    {
+      "epoch": 0.2710914874424346,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.4250903306629554e-05,
+      "loss": 0.2265,
+      "step": 5710
+    },
+    {
+      "epoch": 0.27132887053126337,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.423910528650951e-05,
+      "loss": 0.2234,
+      "step": 5715
+    },
+    {
+      "epoch": 0.2715662536200921,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.4227296950155006e-05,
+      "loss": 0.2248,
+      "step": 5720
+    },
+    {
+      "epoch": 0.2718036367089209,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.4215478304843455e-05,
+      "loss": 0.232,
+      "step": 5725
+    },
+    {
+      "epoch": 0.2720410197977496,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.4203649357858625e-05,
+      "loss": 0.2236,
+      "step": 5730
+    },
+    {
+      "epoch": 0.27227840288657834,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.4191810116490666e-05,
+      "loss": 0.2248,
+      "step": 5735
+    },
+    {
+      "epoch": 0.2725157859754071,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.4179960588036026e-05,
+      "loss": 0.2287,
+      "step": 5740
+    },
+    {
+      "epoch": 0.27275316906423586,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.4168100779797533e-05,
+      "loss": 0.2257,
+      "step": 5745
+    },
+    {
+      "epoch": 0.2729905521530646,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.415623069908432e-05,
+      "loss": 0.2304,
+      "step": 5750
+    },
+    {
+      "epoch": 0.27322793524189337,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.414435035321187e-05,
+      "loss": 0.2256,
+      "step": 5755
+    },
+    {
+      "epoch": 0.27346531833072213,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.4132459749501975e-05,
+      "loss": 0.2248,
+      "step": 5760
+    },
+    {
+      "epoch": 0.2737027014195509,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.412055889528278e-05,
+      "loss": 0.2286,
+      "step": 5765
+    },
+    {
+      "epoch": 0.27394008450837964,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.410864779788872e-05,
+      "loss": 0.223,
+      "step": 5770
+    },
+    {
+      "epoch": 0.2741774675972084,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.409672646466054e-05,
+      "loss": 0.2262,
+      "step": 5775
+    },
+    {
+      "epoch": 0.2744148506860371,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.4084794902945306e-05,
+      "loss": 0.2306,
+      "step": 5780
+    },
+    {
+      "epoch": 0.27465223377486586,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.4072853120096405e-05,
+      "loss": 0.2273,
+      "step": 5785
+    },
+    {
+      "epoch": 0.2748896168636946,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.4060901123473476e-05,
+      "loss": 0.2279,
+      "step": 5790
+    },
+    {
+      "epoch": 0.2751269999525234,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.4048938920442496e-05,
+      "loss": 0.2237,
+      "step": 5795
+    },
+    {
+      "epoch": 0.27536438304135213,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.4036966518375706e-05,
+      "loss": 0.2261,
+      "step": 5800
+    },
+    {
+      "epoch": 0.2756017661301809,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.402498392465165e-05,
+      "loss": 0.2298,
+      "step": 5805
+    },
+    {
+      "epoch": 0.27583914921900965,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.401299114665516e-05,
+      "loss": 0.2268,
+      "step": 5810
+    },
+    {
+      "epoch": 0.2760765323078384,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.40009881917773e-05,
+      "loss": 0.229,
+      "step": 5815
+    },
+    {
+      "epoch": 0.27631391539666716,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.398897506741547e-05,
+      "loss": 0.2236,
+      "step": 5820
+    },
+    {
+      "epoch": 0.27655129848549587,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.397695178097327e-05,
+      "loss": 0.2302,
+      "step": 5825
+    },
+    {
+      "epoch": 0.2767886815743246,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.396491833986062e-05,
+      "loss": 0.2287,
+      "step": 5830
+    },
+    {
+      "epoch": 0.2770260646631534,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.395287475149367e-05,
+      "loss": 0.2259,
+      "step": 5835
+    },
+    {
+      "epoch": 0.27726344775198214,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.394082102329482e-05,
+      "loss": 0.2267,
+      "step": 5840
+    },
+    {
+      "epoch": 0.2775008308408109,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.392875716269274e-05,
+      "loss": 0.2297,
+      "step": 5845
+    },
+    {
+      "epoch": 0.27773821392963965,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.391668317712233e-05,
+      "loss": 0.227,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2779755970184684,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.390459907402472e-05,
+      "loss": 0.2216,
+      "step": 5855
+    },
+    {
+      "epoch": 0.27821298010729717,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.389250486084731e-05,
+      "loss": 0.227,
+      "step": 5860
+    },
+    {
+      "epoch": 0.2784503631961259,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.3880400545043686e-05,
+      "loss": 0.2252,
+      "step": 5865
+    },
+    {
+      "epoch": 0.2786877462849547,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.3868286134073704e-05,
+      "loss": 0.226,
+      "step": 5870
+    },
+    {
+      "epoch": 0.2789251293737834,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.38561616354034e-05,
+      "loss": 0.2236,
+      "step": 5875
+    },
+    {
+      "epoch": 0.27916251246261214,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.384402705650506e-05,
+      "loss": 0.2246,
+      "step": 5880
+    },
+    {
+      "epoch": 0.2793998955514409,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.3831882404857165e-05,
+      "loss": 0.2254,
+      "step": 5885
+    },
+    {
+      "epoch": 0.27963727864026966,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.3819727687944416e-05,
+      "loss": 0.2248,
+      "step": 5890
+    },
+    {
+      "epoch": 0.2798746617290984,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.3807562913257696e-05,
+      "loss": 0.2273,
+      "step": 5895
+    },
+    {
+      "epoch": 0.2801120448179272,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.3795388088294116e-05,
+      "loss": 0.2283,
+      "step": 5900
+    },
+    {
+      "epoch": 0.28034942790675593,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.378320322055695e-05,
+      "loss": 0.225,
+      "step": 5905
+    },
+    {
+      "epoch": 0.2805868109955847,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.377100831755569e-05,
+      "loss": 0.2281,
+      "step": 5910
+    },
+    {
+      "epoch": 0.28082419408441345,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.3758803386805984e-05,
+      "loss": 0.2252,
+      "step": 5915
+    },
+    {
+      "epoch": 0.2810615771732422,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.37465884358297e-05,
+      "loss": 0.2265,
+      "step": 5920
+    },
+    {
+      "epoch": 0.2812989602620709,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.373436347215483e-05,
+      "loss": 0.2289,
+      "step": 5925
+    },
+    {
+      "epoch": 0.28153634335089966,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.372212850331558e-05,
+      "loss": 0.2295,
+      "step": 5930
+    },
+    {
+      "epoch": 0.2817737264397284,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.37098835368523e-05,
+      "loss": 0.2282,
+      "step": 5935
+    },
+    {
+      "epoch": 0.2820111095285572,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.3697628580311505e-05,
+      "loss": 0.2275,
+      "step": 5940
+    },
+    {
+      "epoch": 0.28224849261738594,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.368536364124587e-05,
+      "loss": 0.2247,
+      "step": 5945
+    },
+    {
+      "epoch": 0.2824858757062147,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.367308872721423e-05,
+      "loss": 0.2271,
+      "step": 5950
+    },
+    {
+      "epoch": 0.28272325879504345,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.366080384578154e-05,
+      "loss": 0.2313,
+      "step": 5955
+    },
+    {
+      "epoch": 0.2829606418838722,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.3648509004518946e-05,
+      "loss": 0.224,
+      "step": 5960
+    },
+    {
+      "epoch": 0.28319802497270097,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.363620421100367e-05,
+      "loss": 0.2288,
+      "step": 5965
+    },
+    {
+      "epoch": 0.28343540806152967,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.362388947281912e-05,
+      "loss": 0.2289,
+      "step": 5970
+    },
+    {
+      "epoch": 0.2836727911503584,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.361156479755481e-05,
+      "loss": 0.2263,
+      "step": 5975
+    },
+    {
+      "epoch": 0.2839101742391872,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.359923019280638e-05,
+      "loss": 0.232,
+      "step": 5980
+    },
+    {
+      "epoch": 0.28414755732801594,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.3586885666175584e-05,
+      "loss": 0.2226,
+      "step": 5985
+    },
+    {
+      "epoch": 0.2843849404168447,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.3574531225270316e-05,
+      "loss": 0.2255,
+      "step": 5990
+    },
+    {
+      "epoch": 0.28462232350567346,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.3562166877704533e-05,
+      "loss": 0.2296,
+      "step": 5995
+    },
+    {
+      "epoch": 0.2848597065945022,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.3549792631098356e-05,
+      "loss": 0.2322,
+      "step": 6000
+    },
+    {
+      "epoch": 0.285097089683331,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.3537408493077966e-05,
+      "loss": 0.2257,
+      "step": 6005
+    },
+    {
+      "epoch": 0.28533447277215973,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.3525014471275646e-05,
+      "loss": 0.2257,
+      "step": 6010
+    },
+    {
+      "epoch": 0.2855718558609885,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.351261057332977e-05,
+      "loss": 0.2279,
+      "step": 6015
+    },
+    {
+      "epoch": 0.2858092389498172,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.350019680688482e-05,
+      "loss": 0.2278,
+      "step": 6020
+    },
+    {
+      "epoch": 0.28604662203864595,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.348777317959133e-05,
+      "loss": 0.2274,
+      "step": 6025
+    },
+    {
+      "epoch": 0.2862840051274747,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.347533969910594e-05,
+      "loss": 0.2236,
+      "step": 6030
+    },
+    {
+      "epoch": 0.28652138821630346,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.346289637309133e-05,
+      "loss": 0.2303,
+      "step": 6035
+    },
+    {
+      "epoch": 0.2867587713051322,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.345044320921628e-05,
+      "loss": 0.2259,
+      "step": 6040
+    },
+    {
+      "epoch": 0.286996154393961,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.34379802151556e-05,
+      "loss": 0.2246,
+      "step": 6045
+    },
+    {
+      "epoch": 0.28723353748278974,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.342550739859021e-05,
+      "loss": 0.2253,
+      "step": 6050
+    },
+    {
+      "epoch": 0.2874709205716185,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.341302476720701e-05,
+      "loss": 0.2289,
+      "step": 6055
+    },
+    {
+      "epoch": 0.28770830366044725,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.3400532328699016e-05,
+      "loss": 0.2269,
+      "step": 6060
+    },
+    {
+      "epoch": 0.287945686749276,
+      "grad_norm": 0.5,
+      "learning_rate": 4.338803009076525e-05,
+      "loss": 0.2227,
+      "step": 6065
+    },
+    {
+      "epoch": 0.2881830698381047,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.3375518061110785e-05,
+      "loss": 0.2205,
+      "step": 6070
+    },
+    {
+      "epoch": 0.28842045292693347,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.336299624744674e-05,
+      "loss": 0.2258,
+      "step": 6075
+    },
+    {
+      "epoch": 0.2886578360157622,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.335046465749022e-05,
+      "loss": 0.2271,
+      "step": 6080
+    },
+    {
+      "epoch": 0.288895219104591,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.3337923298964426e-05,
+      "loss": 0.2268,
+      "step": 6085
+    },
+    {
+      "epoch": 0.28913260219341974,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.3325372179598524e-05,
+      "loss": 0.2276,
+      "step": 6090
+    },
+    {
+      "epoch": 0.2893699852822485,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.33128113071277e-05,
+      "loss": 0.2235,
+      "step": 6095
+    },
+    {
+      "epoch": 0.28960736837107726,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.3300240689293185e-05,
+      "loss": 0.2288,
+      "step": 6100
+    },
+    {
+      "epoch": 0.289844751459906,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.328766033384217e-05,
+      "loss": 0.2284,
+      "step": 6105
+    },
+    {
+      "epoch": 0.29008213454873477,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.3275070248527896e-05,
+      "loss": 0.228,
+      "step": 6110
+    },
+    {
+      "epoch": 0.2903195176375635,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.326247044110956e-05,
+      "loss": 0.2318,
+      "step": 6115
+    },
+    {
+      "epoch": 0.29055690072639223,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.324986091935236e-05,
+      "loss": 0.2255,
+      "step": 6120
+    },
+    {
+      "epoch": 0.290794283815221,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.3237241691027495e-05,
+      "loss": 0.2262,
+      "step": 6125
+    },
+    {
+      "epoch": 0.29103166690404975,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.3224612763912146e-05,
+      "loss": 0.2256,
+      "step": 6130
+    },
+    {
+      "epoch": 0.2912690499928785,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.321197414578945e-05,
+      "loss": 0.2253,
+      "step": 6135
+    },
+    {
+      "epoch": 0.29150643308170726,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.3199325844448534e-05,
+      "loss": 0.2282,
+      "step": 6140
+    },
+    {
+      "epoch": 0.291743816170536,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.318666786768449e-05,
+      "loss": 0.2309,
+      "step": 6145
+    },
+    {
+      "epoch": 0.2919811992593648,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.317400022329838e-05,
+      "loss": 0.2287,
+      "step": 6150
+    },
+    {
+      "epoch": 0.29221858234819353,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.31613229190972e-05,
+      "loss": 0.2257,
+      "step": 6155
+    },
+    {
+      "epoch": 0.2924559654370223,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.314863596289393e-05,
+      "loss": 0.2278,
+      "step": 6160
+    },
+    {
+      "epoch": 0.292693348525851,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.3135939362507463e-05,
+      "loss": 0.2247,
+      "step": 6165
+    },
+    {
+      "epoch": 0.29293073161467975,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.3123233125762666e-05,
+      "loss": 0.2226,
+      "step": 6170
+    },
+    {
+      "epoch": 0.2931681147035085,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.3110517260490346e-05,
+      "loss": 0.2261,
+      "step": 6175
+    },
+    {
+      "epoch": 0.29340549779233727,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.3097791774527226e-05,
+      "loss": 0.2262,
+      "step": 6180
+    },
+    {
+      "epoch": 0.293642880881166,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.3085056675715965e-05,
+      "loss": 0.2285,
+      "step": 6185
+    },
+    {
+      "epoch": 0.2938802639699948,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.307231197190514e-05,
+      "loss": 0.2244,
+      "step": 6190
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.3059557670949264e-05,
+      "loss": 0.2244,
+      "step": 6195
+    },
+    {
+      "epoch": 0.2943550301476523,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.3046793780708746e-05,
+      "loss": 0.2301,
+      "step": 6200
+    },
+    {
+      "epoch": 0.29459241323648105,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.303402030904991e-05,
+      "loss": 0.2285,
+      "step": 6205
+    },
+    {
+      "epoch": 0.2948297963253098,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.3021237263845005e-05,
+      "loss": 0.2218,
+      "step": 6210
+    },
+    {
+      "epoch": 0.2950671794141385,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.300844465297215e-05,
+      "loss": 0.2232,
+      "step": 6215
+    },
+    {
+      "epoch": 0.29530456250296727,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.299564248431539e-05,
+      "loss": 0.2269,
+      "step": 6220
+    },
+    {
+      "epoch": 0.29554194559179603,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.298283076576461e-05,
+      "loss": 0.2269,
+      "step": 6225
+    },
+    {
+      "epoch": 0.2957793286806248,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.297000950521565e-05,
+      "loss": 0.2296,
+      "step": 6230
+    },
+    {
+      "epoch": 0.29601671176945354,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.295717871057017e-05,
+      "loss": 0.2268,
+      "step": 6235
+    },
+    {
+      "epoch": 0.2962540948582823,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.294433838973574e-05,
+      "loss": 0.2261,
+      "step": 6240
+    },
+    {
+      "epoch": 0.29649147794711106,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.2931488550625796e-05,
+      "loss": 0.2284,
+      "step": 6245
+    },
+    {
+      "epoch": 0.2967288610359398,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.2918629201159624e-05,
+      "loss": 0.2314,
+      "step": 6250
+    },
+    {
+      "epoch": 0.2969662441247686,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.2905760349262397e-05,
+      "loss": 0.2293,
+      "step": 6255
+    },
+    {
+      "epoch": 0.2972036272135973,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.2892882002865114e-05,
+      "loss": 0.2248,
+      "step": 6260
+    },
+    {
+      "epoch": 0.29744101030242603,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.287999416990465e-05,
+      "loss": 0.2276,
+      "step": 6265
+    },
+    {
+      "epoch": 0.2976783933912548,
+      "grad_norm": 0.494140625,
+      "learning_rate": 4.286709685832372e-05,
+      "loss": 0.2288,
+      "step": 6270
+    },
+    {
+      "epoch": 0.29791577648008355,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.285419007607087e-05,
+      "loss": 0.2269,
+      "step": 6275
+    },
+    {
+      "epoch": 0.2981531595689123,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.28412738311005e-05,
+      "loss": 0.2282,
+      "step": 6280
+    },
+    {
+      "epoch": 0.29839054265774106,
+      "grad_norm": 0.404296875,
+      "learning_rate": 4.2828348131372834e-05,
+      "loss": 0.226,
+      "step": 6285
+    },
+    {
+      "epoch": 0.2986279257465698,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.281541298485392e-05,
+      "loss": 0.2263,
+      "step": 6290
+    },
+    {
+      "epoch": 0.2988653088353986,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.2802468399515613e-05,
+      "loss": 0.2265,
+      "step": 6295
+    },
+    {
+      "epoch": 0.29910269192422734,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.278951438333563e-05,
+      "loss": 0.2281,
+      "step": 6300
+    },
+    {
+      "epoch": 0.2993400750130561,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.277655094429745e-05,
+      "loss": 0.222,
+      "step": 6305
+    },
+    {
+      "epoch": 0.2995774581018848,
+      "grad_norm": 0.3984375,
+      "learning_rate": 4.2763578090390394e-05,
+      "loss": 0.2257,
+      "step": 6310
+    },
+    {
+      "epoch": 0.29981484119071355,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.275059582960957e-05,
+      "loss": 0.2246,
+      "step": 6315
+    },
+    {
+      "epoch": 0.3000522242795423,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.273760416995589e-05,
+      "loss": 0.2256,
+      "step": 6320
+    },
+    {
+      "epoch": 0.30028960736837107,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.272460311943604e-05,
+      "loss": 0.2249,
+      "step": 6325
+    },
+    {
+      "epoch": 0.3005269904571998,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.271159268606252e-05,
+      "loss": 0.2251,
+      "step": 6330
+    },
+    {
+      "epoch": 0.3007643735460286,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.2698572877853605e-05,
+      "loss": 0.2266,
+      "step": 6335
+    },
+    {
+      "epoch": 0.30100175663485734,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.2685543702833334e-05,
+      "loss": 0.2291,
+      "step": 6340
+    },
+    {
+      "epoch": 0.3012391397236861,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.267250516903154e-05,
+      "loss": 0.2269,
+      "step": 6345
+    },
+    {
+      "epoch": 0.30147652281251486,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.26594572844838e-05,
+      "loss": 0.2271,
+      "step": 6350
+    },
+    {
+      "epoch": 0.3017139059013436,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.264640005723147e-05,
+      "loss": 0.2289,
+      "step": 6355
+    },
+    {
+      "epoch": 0.3019512889901723,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.263333349532166e-05,
+      "loss": 0.2293,
+      "step": 6360
+    },
+    {
+      "epoch": 0.3021886720790011,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.262025760680723e-05,
+      "loss": 0.2295,
+      "step": 6365
+    },
+    {
+      "epoch": 0.30242605516782983,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.2607172399746796e-05,
+      "loss": 0.2273,
+      "step": 6370
+    },
+    {
+      "epoch": 0.3026634382566586,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.2594077882204706e-05,
+      "loss": 0.2255,
+      "step": 6375
+    },
+    {
+      "epoch": 0.30290082134548735,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.258097406225106e-05,
+      "loss": 0.2235,
+      "step": 6380
+    },
+    {
+      "epoch": 0.3031382044343161,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.256786094796168e-05,
+      "loss": 0.2246,
+      "step": 6385
+    },
+    {
+      "epoch": 0.30337558752314486,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.2554738547418114e-05,
+      "loss": 0.2289,
+      "step": 6390
+    },
+    {
+      "epoch": 0.3036129706119736,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.254160686870764e-05,
+      "loss": 0.2268,
+      "step": 6395
+    },
+    {
+      "epoch": 0.3038503537008024,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.2528465919923245e-05,
+      "loss": 0.2244,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3040877367896311,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.251531570916365e-05,
+      "loss": 0.2243,
+      "step": 6405
+    },
+    {
+      "epoch": 0.30432511987845984,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.250215624453326e-05,
+      "loss": 0.2198,
+      "step": 6410
+    },
+    {
+      "epoch": 0.3045625029672886,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.248898753414219e-05,
+      "loss": 0.2218,
+      "step": 6415
+    },
+    {
+      "epoch": 0.30479988605611735,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.247580958610628e-05,
+      "loss": 0.2213,
+      "step": 6420
+    },
+    {
+      "epoch": 0.3050372691449461,
+      "grad_norm": 0.384765625,
+      "learning_rate": 4.2462622408547005e-05,
+      "loss": 0.2255,
+      "step": 6425
+    },
+    {
+      "epoch": 0.30527465223377487,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.244942600959159e-05,
+      "loss": 0.2277,
+      "step": 6430
+    },
+    {
+      "epoch": 0.3055120353226036,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.24362203973729e-05,
+      "loss": 0.2259,
+      "step": 6435
+    },
+    {
+      "epoch": 0.3057494184114324,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.2423005580029504e-05,
+      "loss": 0.2291,
+      "step": 6440
+    },
+    {
+      "epoch": 0.30598680150026114,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.2409781565705614e-05,
+      "loss": 0.2221,
+      "step": 6445
+    },
+    {
+      "epoch": 0.3062241845890899,
+      "grad_norm": 0.439453125,
+      "learning_rate": 4.239654836255116e-05,
+      "loss": 0.2252,
+      "step": 6450
+    },
+    {
+      "epoch": 0.3064615676779186,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.238330597872168e-05,
+      "loss": 0.2253,
+      "step": 6455
+    },
+    {
+      "epoch": 0.30669895076674736,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.237005442237841e-05,
+      "loss": 0.2264,
+      "step": 6460
+    },
+    {
+      "epoch": 0.3069363338555761,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.235679370168822e-05,
+      "loss": 0.2258,
+      "step": 6465
+    },
+    {
+      "epoch": 0.3071737169444049,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.234352382482363e-05,
+      "loss": 0.2219,
+      "step": 6470
+    },
+    {
+      "epoch": 0.30741110003323363,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.2330244799962795e-05,
+      "loss": 0.2253,
+      "step": 6475
+    },
+    {
+      "epoch": 0.3076484831220624,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.231695663528953e-05,
+      "loss": 0.2269,
+      "step": 6480
+    },
+    {
+      "epoch": 0.30788586621089115,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.230365933899328e-05,
+      "loss": 0.2263,
+      "step": 6485
+    },
+    {
+      "epoch": 0.3081232492997199,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.229035291926909e-05,
+      "loss": 0.2268,
+      "step": 6490
+    },
+    {
+      "epoch": 0.30836063238854866,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.227703738431765e-05,
+      "loss": 0.2305,
+      "step": 6495
+    },
+    {
+      "epoch": 0.3085980154773774,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.226371274234527e-05,
+      "loss": 0.2231,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3088353985662061,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.2250379001563854e-05,
+      "loss": 0.2249,
+      "step": 6505
+    },
+    {
+      "epoch": 0.3090727816550349,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.223703617019094e-05,
+      "loss": 0.2238,
+      "step": 6510
+    },
+    {
+      "epoch": 0.30931016474386364,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.222368425644965e-05,
+      "loss": 0.2258,
+      "step": 6515
+    },
+    {
+      "epoch": 0.3095475478326924,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.2210323268568704e-05,
+      "loss": 0.2239,
+      "step": 6520
+    },
+    {
+      "epoch": 0.30978493092152115,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.219695321478242e-05,
+      "loss": 0.2261,
+      "step": 6525
+    },
+    {
+      "epoch": 0.3100223140103499,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.21835741033307e-05,
+      "loss": 0.2258,
+      "step": 6530
+    },
+    {
+      "epoch": 0.31025969709917867,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.217018594245903e-05,
+      "loss": 0.2249,
+      "step": 6535
+    },
+    {
+      "epoch": 0.3104970801880074,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.215678874041848e-05,
+      "loss": 0.2287,
+      "step": 6540
+    },
+    {
+      "epoch": 0.3107344632768362,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.2143382505465676e-05,
+      "loss": 0.2256,
+      "step": 6545
+    },
+    {
+      "epoch": 0.3109718463656649,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.212996724586282e-05,
+      "loss": 0.2269,
+      "step": 6550
+    },
+    {
+      "epoch": 0.31120922945449364,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.211654296987768e-05,
+      "loss": 0.2288,
+      "step": 6555
+    },
+    {
+      "epoch": 0.3114466125433224,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.210310968578357e-05,
+      "loss": 0.224,
+      "step": 6560
+    },
+    {
+      "epoch": 0.31168399563215116,
+      "grad_norm": 0.43359375,
+      "learning_rate": 4.2089667401859374e-05,
+      "loss": 0.2304,
+      "step": 6565
+    },
+    {
+      "epoch": 0.3119213787209799,
+      "grad_norm": 0.47265625,
+      "learning_rate": 4.207621612638949e-05,
+      "loss": 0.2277,
+      "step": 6570
+    },
+    {
+      "epoch": 0.31215876180980867,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.2062755867663895e-05,
+      "loss": 0.2256,
+      "step": 6575
+    },
+    {
+      "epoch": 0.31239614489863743,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.2049286633978085e-05,
+      "loss": 0.2256,
+      "step": 6580
+    },
+    {
+      "epoch": 0.3126335279874662,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.203580843363308e-05,
+      "loss": 0.2271,
+      "step": 6585
+    },
+    {
+      "epoch": 0.31287091107629494,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.202232127493543e-05,
+      "loss": 0.2258,
+      "step": 6590
+    },
+    {
+      "epoch": 0.3131082941651237,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.200882516619723e-05,
+      "loss": 0.2238,
+      "step": 6595
+    },
+    {
+      "epoch": 0.3133456772539524,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.199532011573605e-05,
+      "loss": 0.2255,
+      "step": 6600
+    },
+    {
+      "epoch": 0.31358306034278116,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.198180613187499e-05,
+      "loss": 0.2265,
+      "step": 6605
+    },
+    {
+      "epoch": 0.3138204434316099,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.1968283222942666e-05,
+      "loss": 0.2266,
+      "step": 6610
+    },
+    {
+      "epoch": 0.3140578265204387,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.195475139727318e-05,
+      "loss": 0.2265,
+      "step": 6615
+    },
+    {
+      "epoch": 0.31429520960926743,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.1941210663206155e-05,
+      "loss": 0.2232,
+      "step": 6620
+    },
+    {
+      "epoch": 0.3145325926980962,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.192766102908665e-05,
+      "loss": 0.2275,
+      "step": 6625
+    },
+    {
+      "epoch": 0.31476997578692495,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.1914102503265274e-05,
+      "loss": 0.2259,
+      "step": 6630
+    },
+    {
+      "epoch": 0.3150073588757537,
+      "grad_norm": 0.5,
+      "learning_rate": 4.190053509409806e-05,
+      "loss": 0.2309,
+      "step": 6635
+    },
+    {
+      "epoch": 0.31524474196458246,
+      "grad_norm": 0.455078125,
+      "learning_rate": 4.1886958809946564e-05,
+      "loss": 0.2235,
+      "step": 6640
+    },
+    {
+      "epoch": 0.3154821250534112,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.187337365917776e-05,
+      "loss": 0.2242,
+      "step": 6645
+    },
+    {
+      "epoch": 0.3157195081422399,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.185977965016413e-05,
+      "loss": 0.2264,
+      "step": 6650
+    },
+    {
+      "epoch": 0.3159568912310687,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.184617679128362e-05,
+      "loss": 0.2269,
+      "step": 6655
+    },
+    {
+      "epoch": 0.31619427431989744,
+      "grad_norm": 0.5,
+      "learning_rate": 4.1832565090919576e-05,
+      "loss": 0.2277,
+      "step": 6660
+    },
+    {
+      "epoch": 0.3164316574087262,
+      "grad_norm": 0.5,
+      "learning_rate": 4.181894455746083e-05,
+      "loss": 0.2253,
+      "step": 6665
+    },
+    {
+      "epoch": 0.31666904049755495,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.180531519930169e-05,
+      "loss": 0.226,
+      "step": 6670
+    },
+    {
+      "epoch": 0.3169064235863837,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.179167702484183e-05,
+      "loss": 0.2259,
+      "step": 6675
+    },
+    {
+      "epoch": 0.31714380667521247,
+      "grad_norm": 0.5,
+      "learning_rate": 4.177803004248641e-05,
+      "loss": 0.2257,
+      "step": 6680
+    },
+    {
+      "epoch": 0.3173811897640412,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.1764374260646e-05,
+      "loss": 0.2229,
+      "step": 6685
+    },
+    {
+      "epoch": 0.31761857285287,
+      "grad_norm": 0.3828125,
+      "learning_rate": 4.17507096877366e-05,
+      "loss": 0.2241,
+      "step": 6690
+    },
+    {
+      "epoch": 0.3178559559416987,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.173703633217962e-05,
+      "loss": 0.226,
+      "step": 6695
+    },
+    {
+      "epoch": 0.31809333903052744,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.172335420240189e-05,
+      "loss": 0.2259,
+      "step": 6700
+    },
+    {
+      "epoch": 0.3183307221193562,
+      "grad_norm": 0.49609375,
+      "learning_rate": 4.170966330683564e-05,
+      "loss": 0.2265,
+      "step": 6705
+    },
+    {
+      "epoch": 0.31856810520818496,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.1695963653918516e-05,
+      "loss": 0.2235,
+      "step": 6710
+    },
+    {
+      "epoch": 0.3188054882970137,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.168225525209355e-05,
+      "loss": 0.2239,
+      "step": 6715
+    },
+    {
+      "epoch": 0.3190428713858425,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.1668538109809155e-05,
+      "loss": 0.2244,
+      "step": 6720
+    },
+    {
+      "epoch": 0.31928025447467123,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.165481223551915e-05,
+      "loss": 0.2261,
+      "step": 6725
+    },
+    {
+      "epoch": 0.3195176375635,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.164107763768274e-05,
+      "loss": 0.2313,
+      "step": 6730
+    },
+    {
+      "epoch": 0.31975502065232875,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.162733432476448e-05,
+      "loss": 0.2277,
+      "step": 6735
+    },
+    {
+      "epoch": 0.3199924037411575,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.1613582305234314e-05,
+      "loss": 0.2278,
+      "step": 6740
+    },
+    {
+      "epoch": 0.3202297868299862,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.159982158756756e-05,
+      "loss": 0.2237,
+      "step": 6745
+    },
+    {
+      "epoch": 0.32046716991881496,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.158605218024487e-05,
+      "loss": 0.2258,
+      "step": 6750
+    },
+    {
+      "epoch": 0.3207045530076437,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.157227409175228e-05,
+      "loss": 0.2318,
+      "step": 6755
+    },
+    {
+      "epoch": 0.3209419360964725,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.1558487330581153e-05,
+      "loss": 0.2237,
+      "step": 6760
+    },
+    {
+      "epoch": 0.32117931918530124,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.1544691905228215e-05,
+      "loss": 0.2177,
+      "step": 6765
+    },
+    {
+      "epoch": 0.32141670227413,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.1530887824195526e-05,
+      "loss": 0.226,
+      "step": 6770
+    },
+    {
+      "epoch": 0.32165408536295875,
+      "grad_norm": 0.4765625,
+      "learning_rate": 4.151707509599047e-05,
+      "loss": 0.2283,
+      "step": 6775
+    },
+    {
+      "epoch": 0.3218914684517875,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.1503253729125776e-05,
+      "loss": 0.2257,
+      "step": 6780
+    },
+    {
+      "epoch": 0.32212885154061627,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.148942373211948e-05,
+      "loss": 0.2239,
+      "step": 6785
+    },
+    {
+      "epoch": 0.322366234629445,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.147558511349496e-05,
+      "loss": 0.2284,
+      "step": 6790
+    },
+    {
+      "epoch": 0.3226036177182737,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.146173788178088e-05,
+      "loss": 0.2259,
+      "step": 6795
+    },
+    {
+      "epoch": 0.3228410008071025,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.1447882045511234e-05,
+      "loss": 0.2296,
+      "step": 6800
+    },
+    {
+      "epoch": 0.32307838389593124,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.143401761322531e-05,
+      "loss": 0.2279,
+      "step": 6805
+    },
+    {
+      "epoch": 0.32331576698476,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.1420144593467685e-05,
+      "loss": 0.2247,
+      "step": 6810
+    },
+    {
+      "epoch": 0.32355315007358876,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.1406262994788256e-05,
+      "loss": 0.2228,
+      "step": 6815
+    },
+    {
+      "epoch": 0.3237905331624175,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.139237282574217e-05,
+      "loss": 0.2236,
+      "step": 6820
+    },
+    {
+      "epoch": 0.3240279162512463,
+      "grad_norm": 0.625,
+      "learning_rate": 4.1378474094889895e-05,
+      "loss": 0.2244,
+      "step": 6825
+    },
+    {
+      "epoch": 0.32426529934007503,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.136456681079715e-05,
+      "loss": 0.2255,
+      "step": 6830
+    },
+    {
+      "epoch": 0.3245026824289038,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.1350650982034915e-05,
+      "loss": 0.2254,
+      "step": 6835
+    },
+    {
+      "epoch": 0.3247400655177325,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.133672661717946e-05,
+      "loss": 0.2284,
+      "step": 6840
+    },
+    {
+      "epoch": 0.32497744860656125,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.132279372481232e-05,
+      "loss": 0.2277,
+      "step": 6845
+    },
+    {
+      "epoch": 0.32521483169539,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.130885231352026e-05,
+      "loss": 0.2262,
+      "step": 6850
+    },
+    {
+      "epoch": 0.32545221478421876,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.129490239189531e-05,
+      "loss": 0.2276,
+      "step": 6855
+    },
+    {
+      "epoch": 0.3256895978730475,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.1280943968534745e-05,
+      "loss": 0.2268,
+      "step": 6860
+    },
+    {
+      "epoch": 0.3259269809618763,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.126697705204109e-05,
+      "loss": 0.224,
+      "step": 6865
+    },
+    {
+      "epoch": 0.32616436405070504,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.125300165102207e-05,
+      "loss": 0.2265,
+      "step": 6870
+    },
+    {
+      "epoch": 0.3264017471395338,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.1239017774090674e-05,
+      "loss": 0.2293,
+      "step": 6875
+    },
+    {
+      "epoch": 0.32663913022836255,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.122502542986511e-05,
+      "loss": 0.2283,
+      "step": 6880
+    },
+    {
+      "epoch": 0.3268765133171913,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.121102462696877e-05,
+      "loss": 0.2278,
+      "step": 6885
+    },
+    {
+      "epoch": 0.32711389640602,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.119701537403031e-05,
+      "loss": 0.2261,
+      "step": 6890
+    },
+    {
+      "epoch": 0.32735127949484877,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.118299767968356e-05,
+      "loss": 0.2281,
+      "step": 6895
+    },
+    {
+      "epoch": 0.3275886625836775,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.116897155256755e-05,
+      "loss": 0.2319,
+      "step": 6900
+    },
+    {
+      "epoch": 0.3278260456725063,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.115493700132653e-05,
+      "loss": 0.2259,
+      "step": 6905
+    },
+    {
+      "epoch": 0.32806342876133504,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.114089403460994e-05,
+      "loss": 0.2253,
+      "step": 6910
+    },
+    {
+      "epoch": 0.3283008118501638,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.112684266107238e-05,
+      "loss": 0.2268,
+      "step": 6915
+    },
+    {
+      "epoch": 0.32853819493899256,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.111278288937366e-05,
+      "loss": 0.2284,
+      "step": 6920
+    },
+    {
+      "epoch": 0.3287755780278213,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.109871472817874e-05,
+      "loss": 0.2242,
+      "step": 6925
+    },
+    {
+      "epoch": 0.32901296111665007,
+      "grad_norm": 0.474609375,
+      "learning_rate": 4.108463818615778e-05,
+      "loss": 0.223,
+      "step": 6930
+    },
+    {
+      "epoch": 0.32925034420547883,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.1070553271986065e-05,
+      "loss": 0.2277,
+      "step": 6935
+    },
+    {
+      "epoch": 0.32948772729430753,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.105645999434409e-05,
+      "loss": 0.2276,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3297251103831363,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.104235836191746e-05,
+      "loss": 0.2298,
+      "step": 6945
+    },
+    {
+      "epoch": 0.32996249347196505,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.1028248383396954e-05,
+      "loss": 0.2234,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3301998765607938,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.101413006747849e-05,
+      "loss": 0.2279,
+      "step": 6955
+    },
+    {
+      "epoch": 0.33043725964962256,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.1000003422863115e-05,
+      "loss": 0.2252,
+      "step": 6960
+    },
+    {
+      "epoch": 0.3306746427384513,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.0985868458257024e-05,
+      "loss": 0.2257,
+      "step": 6965
+    },
+    {
+      "epoch": 0.3309120258272801,
+      "grad_norm": 0.484375,
+      "learning_rate": 4.097172518237152e-05,
+      "loss": 0.2265,
+      "step": 6970
+    },
+    {
+      "epoch": 0.33114940891610883,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.095757360392305e-05,
+      "loss": 0.2261,
+      "step": 6975
+    },
+    {
+      "epoch": 0.3313867920049376,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.0943413731633166e-05,
+      "loss": 0.2273,
+      "step": 6980
+    },
+    {
+      "epoch": 0.3316241750937663,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.092924557422853e-05,
+      "loss": 0.2311,
+      "step": 6985
+    },
+    {
+      "epoch": 0.33186155818259505,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.091506914044091e-05,
+      "loss": 0.2284,
+      "step": 6990
+    },
+    {
+      "epoch": 0.3320989412714238,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.090088443900718e-05,
+      "loss": 0.2255,
+      "step": 6995
+    },
+    {
+      "epoch": 0.33233632436025257,
+      "grad_norm": 0.5,
+      "learning_rate": 4.088669147866931e-05,
+      "loss": 0.2257,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3325737074490813,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.0872490268174356e-05,
+      "loss": 0.2267,
+      "step": 7005
+    },
+    {
+      "epoch": 0.3328110905379101,
+      "grad_norm": 0.478515625,
+      "learning_rate": 4.0858280816274464e-05,
+      "loss": 0.2281,
+      "step": 7010
+    },
+    {
+      "epoch": 0.33304847362673884,
+      "grad_norm": 0.48828125,
+      "learning_rate": 4.084406313172684e-05,
+      "loss": 0.2243,
+      "step": 7015
+    },
+    {
+      "epoch": 0.3332858567155676,
+      "grad_norm": 0.470703125,
+      "learning_rate": 4.082983722329379e-05,
+      "loss": 0.2249,
+      "step": 7020
+    },
+    {
+      "epoch": 0.33352323980439635,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.081560309974268e-05,
+      "loss": 0.2265,
+      "step": 7025
+    },
+    {
+      "epoch": 0.3337606228932251,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.080136076984593e-05,
+      "loss": 0.2242,
+      "step": 7030
+    },
+    {
+      "epoch": 0.3339980059820538,
+      "grad_norm": 0.458984375,
+      "learning_rate": 4.078711024238102e-05,
+      "loss": 0.2262,
+      "step": 7035
+    },
+    {
+      "epoch": 0.33423538907088257,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.0772851526130487e-05,
+      "loss": 0.2253,
+      "step": 7040
+    },
+    {
+      "epoch": 0.33447277215971133,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.075858462988191e-05,
+      "loss": 0.2245,
+      "step": 7045
+    },
+    {
+      "epoch": 0.3347101552485401,
+      "grad_norm": 0.453125,
+      "learning_rate": 4.074430956242793e-05,
+      "loss": 0.2289,
+      "step": 7050
+    },
+    {
+      "epoch": 0.33494753833736884,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.073002633256618e-05,
+      "loss": 0.2267,
+      "step": 7055
+    },
+    {
+      "epoch": 0.3351849214261976,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.071573494909937e-05,
+      "loss": 0.2278,
+      "step": 7060
+    },
+    {
+      "epoch": 0.33542230451502636,
+      "grad_norm": 0.486328125,
+      "learning_rate": 4.07014354208352e-05,
+      "loss": 0.2226,
+      "step": 7065
+    },
+    {
+      "epoch": 0.3356596876038551,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.0687127756586416e-05,
+      "loss": 0.2266,
+      "step": 7070
+    },
+    {
+      "epoch": 0.3358970706926839,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.0672811965170765e-05,
+      "loss": 0.2267,
+      "step": 7075
+    },
+    {
+      "epoch": 0.33613445378151263,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.0658488055410996e-05,
+      "loss": 0.2285,
+      "step": 7080
+    },
+    {
+      "epoch": 0.33637183687034133,
+      "grad_norm": 0.412109375,
+      "learning_rate": 4.064415603613487e-05,
+      "loss": 0.2272,
+      "step": 7085
+    },
+    {
+      "epoch": 0.3366092199591701,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.062981591617516e-05,
+      "loss": 0.229,
+      "step": 7090
+    },
+    {
+      "epoch": 0.33684660304799885,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.0615467704369594e-05,
+      "loss": 0.2241,
+      "step": 7095
+    },
+    {
+      "epoch": 0.3370839861368276,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.060111140956093e-05,
+      "loss": 0.2228,
+      "step": 7100
+    },
+    {
+      "epoch": 0.33732136922565636,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.0586747040596884e-05,
+      "loss": 0.2227,
+      "step": 7105
+    },
+    {
+      "epoch": 0.3375587523144851,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.057237460633013e-05,
+      "loss": 0.2282,
+      "step": 7110
+    },
+    {
+      "epoch": 0.3377961354033139,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.055799411561836e-05,
+      "loss": 0.2298,
+      "step": 7115
+    },
+    {
+      "epoch": 0.33803351849214264,
+      "grad_norm": 0.451171875,
+      "learning_rate": 4.054360557732421e-05,
+      "loss": 0.2254,
+      "step": 7120
+    },
+    {
+      "epoch": 0.3382709015809714,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.052920900031524e-05,
+      "loss": 0.2263,
+      "step": 7125
+    },
+    {
+      "epoch": 0.3385082846698001,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.051480439346402e-05,
+      "loss": 0.2222,
+      "step": 7130
+    },
+    {
+      "epoch": 0.33874566775862885,
+      "grad_norm": 0.4140625,
+      "learning_rate": 4.0500391765648014e-05,
+      "loss": 0.2272,
+      "step": 7135
+    },
+    {
+      "epoch": 0.3389830508474576,
+      "grad_norm": 0.44140625,
+      "learning_rate": 4.0485971125749704e-05,
+      "loss": 0.2267,
+      "step": 7140
+    },
+    {
+      "epoch": 0.33922043393628637,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.0471542482656424e-05,
+      "loss": 0.2272,
+      "step": 7145
+    },
+    {
+      "epoch": 0.3394578170251151,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.045710584526051e-05,
+      "loss": 0.2267,
+      "step": 7150
+    },
+    {
+      "epoch": 0.3396952001139439,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.0442661222459175e-05,
+      "loss": 0.2244,
+      "step": 7155
+    },
+    {
+      "epoch": 0.33993258320277264,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.0428208623154586e-05,
+      "loss": 0.2303,
+      "step": 7160
+    },
+    {
+      "epoch": 0.3401699662916014,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.041374805625381e-05,
+      "loss": 0.2243,
+      "step": 7165
+    },
+    {
+      "epoch": 0.34040734938043016,
+      "grad_norm": 0.42578125,
+      "learning_rate": 4.039927953066884e-05,
+      "loss": 0.2262,
+      "step": 7170
+    },
+    {
+      "epoch": 0.3406447324692589,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.038480305531654e-05,
+      "loss": 0.2254,
+      "step": 7175
+    },
+    {
+      "epoch": 0.3408821155580876,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.037031863911871e-05,
+      "loss": 0.2216,
+      "step": 7180
+    },
+    {
+      "epoch": 0.3411194986469164,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.0355826291002033e-05,
+      "loss": 0.2262,
+      "step": 7185
+    },
+    {
+      "epoch": 0.34135688173574513,
+      "grad_norm": 0.51171875,
+      "learning_rate": 4.0341326019898075e-05,
+      "loss": 0.2234,
+      "step": 7190
+    },
+    {
+      "epoch": 0.3415942648245739,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.032681783474327e-05,
+      "loss": 0.2288,
+      "step": 7195
+    },
+    {
+      "epoch": 0.34183164791340265,
+      "grad_norm": 0.482421875,
+      "learning_rate": 4.031230174447897e-05,
+      "loss": 0.2239,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3420690310022314,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.0297777758051366e-05,
+      "loss": 0.2251,
+      "step": 7205
+    },
+    {
+      "epoch": 0.34230641409106016,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.028324588441153e-05,
+      "loss": 0.2235,
+      "step": 7210
+    },
+    {
+      "epoch": 0.3425437971798889,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.0268706132515377e-05,
+      "loss": 0.2255,
+      "step": 7215
+    },
+    {
+      "epoch": 0.3427811802687177,
+      "grad_norm": 0.50390625,
+      "learning_rate": 4.025415851132371e-05,
+      "loss": 0.2251,
+      "step": 7220
+    },
+    {
+      "epoch": 0.34301856335754644,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.023960302980214e-05,
+      "loss": 0.2271,
+      "step": 7225
+    },
+    {
+      "epoch": 0.34325594644637514,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.0225039696921154e-05,
+      "loss": 0.2247,
+      "step": 7230
+    },
+    {
+      "epoch": 0.3434933295352039,
+      "grad_norm": 0.4921875,
+      "learning_rate": 4.0210468521656065e-05,
+      "loss": 0.2218,
+      "step": 7235
+    },
+    {
+      "epoch": 0.34373071262403265,
+      "grad_norm": 0.46484375,
+      "learning_rate": 4.0195889512987035e-05,
+      "loss": 0.2232,
+      "step": 7240
+    },
+    {
+      "epoch": 0.3439680957128614,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.018130267989902e-05,
+      "loss": 0.2296,
+      "step": 7245
+    },
+    {
+      "epoch": 0.34420547880169017,
+      "grad_norm": 0.4609375,
+      "learning_rate": 4.016670803138183e-05,
+      "loss": 0.2233,
+      "step": 7250
+    },
+    {
+      "epoch": 0.3444428618905189,
+      "grad_norm": 0.46875,
+      "learning_rate": 4.015210557643009e-05,
+      "loss": 0.2258,
+      "step": 7255
+    },
+    {
+      "epoch": 0.3446802449793477,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.013749532404321e-05,
+      "loss": 0.2267,
+      "step": 7260
+    },
+    {
+      "epoch": 0.34491762806817644,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.012287728322542e-05,
+      "loss": 0.2236,
+      "step": 7265
+    },
+    {
+      "epoch": 0.3451550111570052,
+      "grad_norm": 0.416015625,
+      "learning_rate": 4.010825146298577e-05,
+      "loss": 0.2292,
+      "step": 7270
+    },
+    {
+      "epoch": 0.3453923942458339,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.009361787233808e-05,
+      "loss": 0.224,
+      "step": 7275
+    },
+    {
+      "epoch": 0.34562977733466266,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.007897652030096e-05,
+      "loss": 0.2309,
+      "step": 7280
+    },
+    {
+      "epoch": 0.3458671604234914,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.0064327415897815e-05,
+      "loss": 0.2247,
+      "step": 7285
+    },
+    {
+      "epoch": 0.3461045435123202,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.00496705681568e-05,
+      "loss": 0.2248,
+      "step": 7290
+    },
+    {
+      "epoch": 0.34634192660114893,
+      "grad_norm": 0.48046875,
+      "learning_rate": 4.00350059861109e-05,
+      "loss": 0.2248,
+      "step": 7295
+    },
+    {
+      "epoch": 0.3465793096899777,
+      "grad_norm": 0.490234375,
+      "learning_rate": 4.00203336787978e-05,
+      "loss": 0.2253,
+      "step": 7300
+    },
+    {
+      "epoch": 0.34681669277880645,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.000565365525999e-05,
+      "loss": 0.2235,
+      "step": 7305
+    },
+    {
+      "epoch": 0.3470540758676352,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.99909659245447e-05,
+      "loss": 0.2301,
+      "step": 7310
+    },
+    {
+      "epoch": 0.34729145895646396,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.997627049570392e-05,
+      "loss": 0.2229,
+      "step": 7315
+    },
+    {
+      "epoch": 0.3475288420452927,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.996156737779436e-05,
+      "loss": 0.2261,
+      "step": 7320
+    },
+    {
+      "epoch": 0.3477662251341214,
+      "grad_norm": 0.5,
+      "learning_rate": 3.994685657987749e-05,
+      "loss": 0.2225,
+      "step": 7325
+    },
+    {
+      "epoch": 0.3480036082229502,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.993213811101951e-05,
+      "loss": 0.2273,
+      "step": 7330
+    },
+    {
+      "epoch": 0.34824099131177894,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.991741198029135e-05,
+      "loss": 0.2209,
+      "step": 7335
+    },
+    {
+      "epoch": 0.3484783744006077,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.990267819676865e-05,
+      "loss": 0.2244,
+      "step": 7340
+    },
+    {
+      "epoch": 0.34871575748943645,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.988793676953177e-05,
+      "loss": 0.2256,
+      "step": 7345
+    },
+    {
+      "epoch": 0.3489531405782652,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.98731877076658e-05,
+      "loss": 0.2259,
+      "step": 7350
+    },
+    {
+      "epoch": 0.34919052366709397,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.985843102026051e-05,
+      "loss": 0.2262,
+      "step": 7355
+    },
+    {
+      "epoch": 0.3494279067559227,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.984366671641037e-05,
+      "loss": 0.2216,
+      "step": 7360
+    },
+    {
+      "epoch": 0.3496652898447515,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.9828894805214576e-05,
+      "loss": 0.2259,
+      "step": 7365
+    },
+    {
+      "epoch": 0.34990267293358024,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.981411529577697e-05,
+      "loss": 0.2223,
+      "step": 7370
+    },
+    {
+      "epoch": 0.35014005602240894,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.97993281972061e-05,
+      "loss": 0.2245,
+      "step": 7375
+    },
+    {
+      "epoch": 0.3503774391112377,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.97845335186152e-05,
+      "loss": 0.2295,
+      "step": 7380
+    },
+    {
+      "epoch": 0.35061482220006646,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.976973126912215e-05,
+      "loss": 0.2235,
+      "step": 7385
+    },
+    {
+      "epoch": 0.3508522052888952,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.9754921457849516e-05,
+      "loss": 0.2274,
+      "step": 7390
+    },
+    {
+      "epoch": 0.35108958837772397,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.974010409392451e-05,
+      "loss": 0.2258,
+      "step": 7395
+    },
+    {
+      "epoch": 0.35132697146655273,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.9725279186479006e-05,
+      "loss": 0.2264,
+      "step": 7400
+    },
+    {
+      "epoch": 0.3515643545553815,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.971044674464954e-05,
+      "loss": 0.222,
+      "step": 7405
+    },
+    {
+      "epoch": 0.35180173764421024,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.969560677757727e-05,
+      "loss": 0.2235,
+      "step": 7410
+    },
+    {
+      "epoch": 0.352039120733039,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.9680759294408e-05,
+      "loss": 0.2244,
+      "step": 7415
+    },
+    {
+      "epoch": 0.3522765038218677,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.966590430429216e-05,
+      "loss": 0.2301,
+      "step": 7420
+    },
+    {
+      "epoch": 0.35251388691069646,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.9651041816384824e-05,
+      "loss": 0.2288,
+      "step": 7425
+    },
+    {
+      "epoch": 0.3527512699995252,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.9636171839845675e-05,
+      "loss": 0.2247,
+      "step": 7430
+    },
+    {
+      "epoch": 0.352988653088354,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.962129438383901e-05,
+      "loss": 0.2238,
+      "step": 7435
+    },
+    {
+      "epoch": 0.35322603617718273,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.960640945753372e-05,
+      "loss": 0.2249,
+      "step": 7440
+    },
+    {
+      "epoch": 0.3534634192660115,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.959151707010335e-05,
+      "loss": 0.2252,
+      "step": 7445
+    },
+    {
+      "epoch": 0.35370080235484025,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.9576617230725984e-05,
+      "loss": 0.2259,
+      "step": 7450
+    },
+    {
+      "epoch": 0.353938185443669,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.956170994858434e-05,
+      "loss": 0.2234,
+      "step": 7455
+    },
+    {
+      "epoch": 0.35417556853249776,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.954679523286569e-05,
+      "loss": 0.2258,
+      "step": 7460
+    },
+    {
+      "epoch": 0.3544129516213265,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.953187309276194e-05,
+      "loss": 0.2248,
+      "step": 7465
+    },
+    {
+      "epoch": 0.3546503347101552,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.95169435374695e-05,
+      "loss": 0.2278,
+      "step": 7470
+    },
+    {
+      "epoch": 0.354887717798984,
+      "grad_norm": 0.5,
+      "learning_rate": 3.950200657618942e-05,
+      "loss": 0.2256,
+      "step": 7475
+    },
+    {
+      "epoch": 0.35512510088781274,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.9487062218127246e-05,
+      "loss": 0.2253,
+      "step": 7480
+    },
+    {
+      "epoch": 0.3553624839766415,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.9472110472493144e-05,
+      "loss": 0.2249,
+      "step": 7485
+    },
+    {
+      "epoch": 0.35559986706547025,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.945715134850179e-05,
+      "loss": 0.2249,
+      "step": 7490
+    },
+    {
+      "epoch": 0.355837250154299,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.944218485537245e-05,
+      "loss": 0.2224,
+      "step": 7495
+    },
+    {
+      "epoch": 0.35607463324312777,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.942721100232887e-05,
+      "loss": 0.2275,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3563120163319565,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.94122297985994e-05,
+      "loss": 0.2248,
+      "step": 7505
+    },
+    {
+      "epoch": 0.3565493994207853,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.9397241253416864e-05,
+      "loss": 0.225,
+      "step": 7510
+    },
+    {
+      "epoch": 0.35678678250961404,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.938224537601865e-05,
+      "loss": 0.2218,
+      "step": 7515
+    },
+    {
+      "epoch": 0.35702416559844274,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.936724217564665e-05,
+      "loss": 0.2282,
+      "step": 7520
+    },
+    {
+      "epoch": 0.3572615486872715,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.9352231661547244e-05,
+      "loss": 0.2222,
+      "step": 7525
+    },
+    {
+      "epoch": 0.35749893177610026,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.933721384297137e-05,
+      "loss": 0.2239,
+      "step": 7530
+    },
+    {
+      "epoch": 0.357736314864929,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.932218872917443e-05,
+      "loss": 0.2261,
+      "step": 7535
+    },
+    {
+      "epoch": 0.3579736979537578,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.930715632941633e-05,
+      "loss": 0.2312,
+      "step": 7540
+    },
+    {
+      "epoch": 0.35821108104258653,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.92921166529615e-05,
+      "loss": 0.2254,
+      "step": 7545
+    },
+    {
+      "epoch": 0.3584484641314153,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.927706970907877e-05,
+      "loss": 0.2254,
+      "step": 7550
+    },
+    {
+      "epoch": 0.35868584722024405,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.926201550704155e-05,
+      "loss": 0.2261,
+      "step": 7555
+    },
+    {
+      "epoch": 0.3589232303090728,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.924695405612765e-05,
+      "loss": 0.2243,
+      "step": 7560
+    },
+    {
+      "epoch": 0.3591606133979015,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.9231885365619385e-05,
+      "loss": 0.225,
+      "step": 7565
+    },
+    {
+      "epoch": 0.35939799648673026,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.921680944480352e-05,
+      "loss": 0.2231,
+      "step": 7570
+    },
+    {
+      "epoch": 0.359635379575559,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.920172630297127e-05,
+      "loss": 0.2266,
+      "step": 7575
+    },
+    {
+      "epoch": 0.3598727626643878,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.9186635949418317e-05,
+      "loss": 0.2261,
+      "step": 7580
+    },
+    {
+      "epoch": 0.36011014575321654,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.917153839344476e-05,
+      "loss": 0.2246,
+      "step": 7585
+    },
+    {
+      "epoch": 0.3603475288420453,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.9156433644355175e-05,
+      "loss": 0.2234,
+      "step": 7590
+    },
+    {
+      "epoch": 0.36058491193087405,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.914132171145854e-05,
+      "loss": 0.2278,
+      "step": 7595
+    },
+    {
+      "epoch": 0.3608222950197028,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.9126202604068265e-05,
+      "loss": 0.2262,
+      "step": 7600
+    },
+    {
+      "epoch": 0.36105967810853157,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.91110763315022e-05,
+      "loss": 0.2209,
+      "step": 7605
+    },
+    {
+      "epoch": 0.3612970611973603,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.909594290308259e-05,
+      "loss": 0.2219,
+      "step": 7610
+    },
+    {
+      "epoch": 0.361534444286189,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.90808023281361e-05,
+      "loss": 0.2251,
+      "step": 7615
+    },
+    {
+      "epoch": 0.3617718273750178,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.906565461599379e-05,
+      "loss": 0.2269,
+      "step": 7620
+    },
+    {
+      "epoch": 0.36200921046384654,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.905049977599114e-05,
+      "loss": 0.2271,
+      "step": 7625
+    },
+    {
+      "epoch": 0.3622465935526753,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.9035337817468e-05,
+      "loss": 0.2247,
+      "step": 7630
+    },
+    {
+      "epoch": 0.36248397664150406,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.902016874976862e-05,
+      "loss": 0.2243,
+      "step": 7635
+    },
+    {
+      "epoch": 0.3627213597303328,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.900499258224162e-05,
+      "loss": 0.2233,
+      "step": 7640
+    },
+    {
+      "epoch": 0.3629587428191616,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.8989809324240006e-05,
+      "loss": 0.2287,
+      "step": 7645
+    },
+    {
+      "epoch": 0.36319612590799033,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.897461898512116e-05,
+      "loss": 0.223,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3634335089968191,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.8959421574246816e-05,
+      "loss": 0.2251,
+      "step": 7655
+    },
+    {
+      "epoch": 0.36367089208564785,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.894421710098306e-05,
+      "loss": 0.2305,
+      "step": 7660
+    },
+    {
+      "epoch": 0.36390827517447655,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.892900557470034e-05,
+      "loss": 0.2247,
+      "step": 7665
+    },
+    {
+      "epoch": 0.3641456582633053,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.891378700477347e-05,
+      "loss": 0.2265,
+      "step": 7670
+    },
+    {
+      "epoch": 0.36438304135213406,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.889856140058156e-05,
+      "loss": 0.2273,
+      "step": 7675
+    },
+    {
+      "epoch": 0.3646204244409628,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.888332877150809e-05,
+      "loss": 0.2262,
+      "step": 7680
+    },
+    {
+      "epoch": 0.3648578075297916,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.886808912694087e-05,
+      "loss": 0.2245,
+      "step": 7685
+    },
+    {
+      "epoch": 0.36509519061862034,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.8852842476272006e-05,
+      "loss": 0.2259,
+      "step": 7690
+    },
+    {
+      "epoch": 0.3653325737074491,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.8837588828897954e-05,
+      "loss": 0.2258,
+      "step": 7695
+    },
+    {
+      "epoch": 0.36556995679627785,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.8822328194219463e-05,
+      "loss": 0.2259,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3658073398851066,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.8807060581641594e-05,
+      "loss": 0.2297,
+      "step": 7705
+    },
+    {
+      "epoch": 0.3660447229739353,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.87917860005737e-05,
+      "loss": 0.2252,
+      "step": 7710
+    },
+    {
+      "epoch": 0.36628210606276407,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.877650446042944e-05,
+      "loss": 0.2265,
+      "step": 7715
+    },
+    {
+      "epoch": 0.3665194891515928,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.876121597062677e-05,
+      "loss": 0.2262,
+      "step": 7720
+    },
+    {
+      "epoch": 0.3667568722404216,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.874592054058789e-05,
+      "loss": 0.221,
+      "step": 7725
+    },
+    {
+      "epoch": 0.36699425532925034,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.873061817973933e-05,
+      "loss": 0.2255,
+      "step": 7730
+    },
+    {
+      "epoch": 0.3672316384180791,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.8715308897511856e-05,
+      "loss": 0.223,
+      "step": 7735
+    },
+    {
+      "epoch": 0.36746902150690786,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.8699992703340506e-05,
+      "loss": 0.2222,
+      "step": 7740
+    },
+    {
+      "epoch": 0.3677064045957366,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.868466960666458e-05,
+      "loss": 0.2268,
+      "step": 7745
+    },
+    {
+      "epoch": 0.36794378768456537,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.8669339616927644e-05,
+      "loss": 0.2295,
+      "step": 7750
+    },
+    {
+      "epoch": 0.36818117077339413,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.865400274357748e-05,
+      "loss": 0.2233,
+      "step": 7755
+    },
+    {
+      "epoch": 0.36841855386222283,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.863865899606615e-05,
+      "loss": 0.2257,
+      "step": 7760
+    },
+    {
+      "epoch": 0.3686559369510516,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.8623308383849936e-05,
+      "loss": 0.2265,
+      "step": 7765
+    },
+    {
+      "epoch": 0.36889332003988035,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.8607950916389344e-05,
+      "loss": 0.22,
+      "step": 7770
+    },
+    {
+      "epoch": 0.3691307031287091,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.859258660314912e-05,
+      "loss": 0.2237,
+      "step": 7775
+    },
+    {
+      "epoch": 0.36936808621753786,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.85772154535982e-05,
+      "loss": 0.2245,
+      "step": 7780
+    },
+    {
+      "epoch": 0.3696054693063666,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.856183747720978e-05,
+      "loss": 0.2239,
+      "step": 7785
+    },
+    {
+      "epoch": 0.3698428523951954,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.854645268346121e-05,
+      "loss": 0.2254,
+      "step": 7790
+    },
+    {
+      "epoch": 0.37008023548402413,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.8531061081834104e-05,
+      "loss": 0.2259,
+      "step": 7795
+    },
+    {
+      "epoch": 0.3703176185728529,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.8515662681814194e-05,
+      "loss": 0.2264,
+      "step": 7800
+    },
+    {
+      "epoch": 0.37055500166168165,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.8500257492891484e-05,
+      "loss": 0.2266,
+      "step": 7805
+    },
+    {
+      "epoch": 0.37079238475051035,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.8484845524560095e-05,
+      "loss": 0.2253,
+      "step": 7810
+    },
+    {
+      "epoch": 0.3710297678393391,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.8469426786318357e-05,
+      "loss": 0.2244,
+      "step": 7815
+    },
+    {
+      "epoch": 0.37126715092816787,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.845400128766877e-05,
+      "loss": 0.2253,
+      "step": 7820
+    },
+    {
+      "epoch": 0.3715045340169966,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.843856903811801e-05,
+      "loss": 0.2266,
+      "step": 7825
+    },
+    {
+      "epoch": 0.3717419171058254,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.8423130047176886e-05,
+      "loss": 0.2261,
+      "step": 7830
+    },
+    {
+      "epoch": 0.37197930019465414,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.8407684324360396e-05,
+      "loss": 0.225,
+      "step": 7835
+    },
+    {
+      "epoch": 0.3722166832834829,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.839223187918764e-05,
+      "loss": 0.2287,
+      "step": 7840
+    },
+    {
+      "epoch": 0.37245406637231165,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.837677272118192e-05,
+      "loss": 0.2264,
+      "step": 7845
+    },
+    {
+      "epoch": 0.3726914494611404,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.8361306859870624e-05,
+      "loss": 0.2257,
+      "step": 7850
+    },
+    {
+      "epoch": 0.3729288325499691,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.8345834304785305e-05,
+      "loss": 0.2207,
+      "step": 7855
+    },
+    {
+      "epoch": 0.37316621563879787,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.8330355065461625e-05,
+      "loss": 0.2227,
+      "step": 7860
+    },
+    {
+      "epoch": 0.37340359872762663,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.831486915143937e-05,
+      "loss": 0.2232,
+      "step": 7865
+    },
+    {
+      "epoch": 0.3736409818164554,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.829937657226243e-05,
+      "loss": 0.2233,
+      "step": 7870
+    },
+    {
+      "epoch": 0.37387836490528414,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.8283877337478816e-05,
+      "loss": 0.2278,
+      "step": 7875
+    },
+    {
+      "epoch": 0.3741157479941129,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.826837145664064e-05,
+      "loss": 0.2248,
+      "step": 7880
+    },
+    {
+      "epoch": 0.37435313108294166,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.82528589393041e-05,
+      "loss": 0.2248,
+      "step": 7885
+    },
+    {
+      "epoch": 0.3745905141717704,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.823733979502949e-05,
+      "loss": 0.228,
+      "step": 7890
+    },
+    {
+      "epoch": 0.3748278972605992,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.82218140333812e-05,
+      "loss": 0.2237,
+      "step": 7895
+    },
+    {
+      "epoch": 0.37506528034942793,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.820628166392766e-05,
+      "loss": 0.2226,
+      "step": 7900
+    },
+    {
+      "epoch": 0.37530266343825663,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.8190742696241426e-05,
+      "loss": 0.2249,
+      "step": 7905
+    },
+    {
+      "epoch": 0.3755400465270854,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.817519713989909e-05,
+      "loss": 0.2269,
+      "step": 7910
+    },
+    {
+      "epoch": 0.37577742961591415,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.815964500448129e-05,
+      "loss": 0.2252,
+      "step": 7915
+    },
+    {
+      "epoch": 0.3760148127047429,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.8144086299572746e-05,
+      "loss": 0.2255,
+      "step": 7920
+    },
+    {
+      "epoch": 0.37625219579357166,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.812852103476222e-05,
+      "loss": 0.2204,
+      "step": 7925
+    },
+    {
+      "epoch": 0.3764895788824004,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.8112949219642514e-05,
+      "loss": 0.2241,
+      "step": 7930
+    },
+    {
+      "epoch": 0.3767269619712292,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.809737086381046e-05,
+      "loss": 0.2248,
+      "step": 7935
+    },
+    {
+      "epoch": 0.37696434506005794,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.8081785976866934e-05,
+      "loss": 0.2293,
+      "step": 7940
+    },
+    {
+      "epoch": 0.3772017281488867,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.806619456841682e-05,
+      "loss": 0.2235,
+      "step": 7945
+    },
+    {
+      "epoch": 0.37743911123771545,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.805059664806906e-05,
+      "loss": 0.2242,
+      "step": 7950
+    },
+    {
+      "epoch": 0.37767649432654415,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.8034992225436554e-05,
+      "loss": 0.2256,
+      "step": 7955
+    },
+    {
+      "epoch": 0.3779138774153729,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.801938131013625e-05,
+      "loss": 0.2303,
+      "step": 7960
+    },
+    {
+      "epoch": 0.37815126050420167,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.800376391178907e-05,
+      "loss": 0.2249,
+      "step": 7965
+    },
+    {
+      "epoch": 0.3783886435930304,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.798814004001997e-05,
+      "loss": 0.2226,
+      "step": 7970
+    },
+    {
+      "epoch": 0.3786260266818592,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.7972509704457835e-05,
+      "loss": 0.2295,
+      "step": 7975
+    },
+    {
+      "epoch": 0.37886340977068794,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.7956872914735605e-05,
+      "loss": 0.2246,
+      "step": 7980
+    },
+    {
+      "epoch": 0.3791007928595167,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.7941229680490144e-05,
+      "loss": 0.2272,
+      "step": 7985
+    },
+    {
+      "epoch": 0.37933817594834546,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.792558001136232e-05,
+      "loss": 0.2257,
+      "step": 7990
+    },
+    {
+      "epoch": 0.3795755590371742,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.7909923916996924e-05,
+      "loss": 0.2251,
+      "step": 7995
+    },
+    {
+      "epoch": 0.3798129421260029,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.789426140704277e-05,
+      "loss": 0.2246,
+      "step": 8000
+    },
+    {
+      "epoch": 0.3800503252148317,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.787859249115258e-05,
+      "loss": 0.2215,
+      "step": 8005
+    },
+    {
+      "epoch": 0.38028770830366043,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.786291717898303e-05,
+      "loss": 0.224,
+      "step": 8010
+    },
+    {
+      "epoch": 0.3805250913924892,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.7847235480194736e-05,
+      "loss": 0.2263,
+      "step": 8015
+    },
+    {
+      "epoch": 0.38076247448131795,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.783154740445227e-05,
+      "loss": 0.2182,
+      "step": 8020
+    },
+    {
+      "epoch": 0.3809998575701467,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.7815852961424124e-05,
+      "loss": 0.2242,
+      "step": 8025
+    },
+    {
+      "epoch": 0.38123724065897546,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.780015216078271e-05,
+      "loss": 0.2255,
+      "step": 8030
+    },
+    {
+      "epoch": 0.3814746237478042,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.7784445012204356e-05,
+      "loss": 0.2242,
+      "step": 8035
+    },
+    {
+      "epoch": 0.381712006836633,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.776873152536931e-05,
+      "loss": 0.2232,
+      "step": 8040
+    },
+    {
+      "epoch": 0.38194938992546174,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.7753011709961704e-05,
+      "loss": 0.226,
+      "step": 8045
+    },
+    {
+      "epoch": 0.38218677301429044,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.7737285575669626e-05,
+      "loss": 0.2211,
+      "step": 8050
+    },
+    {
+      "epoch": 0.3824241561031192,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.7721553132184987e-05,
+      "loss": 0.2255,
+      "step": 8055
+    },
+    {
+      "epoch": 0.38266153919194795,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.770581438920364e-05,
+      "loss": 0.2284,
+      "step": 8060
+    },
+    {
+      "epoch": 0.3828989222807767,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.7690069356425286e-05,
+      "loss": 0.2234,
+      "step": 8065
+    },
+    {
+      "epoch": 0.38313630536960547,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.7674318043553535e-05,
+      "loss": 0.2258,
+      "step": 8070
+    },
+    {
+      "epoch": 0.3833736884584342,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.765856046029583e-05,
+      "loss": 0.2263,
+      "step": 8075
+    },
+    {
+      "epoch": 0.383611071547263,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.764279661636352e-05,
+      "loss": 0.225,
+      "step": 8080
+    },
+    {
+      "epoch": 0.38384845463609174,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.762702652147177e-05,
+      "loss": 0.2227,
+      "step": 8085
+    },
+    {
+      "epoch": 0.3840858377249205,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.7611250185339625e-05,
+      "loss": 0.2265,
+      "step": 8090
+    },
+    {
+      "epoch": 0.38432322081374926,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.759546761768997e-05,
+      "loss": 0.2273,
+      "step": 8095
+    },
+    {
+      "epoch": 0.38456060390257796,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.7579678828249526e-05,
+      "loss": 0.225,
+      "step": 8100
+    },
+    {
+      "epoch": 0.3847979869914067,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.756388382674887e-05,
+      "loss": 0.2266,
+      "step": 8105
+    },
+    {
+      "epoch": 0.3850353700802355,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.754808262292235e-05,
+      "loss": 0.2233,
+      "step": 8110
+    },
+    {
+      "epoch": 0.38527275316906423,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.753227522650822e-05,
+      "loss": 0.2275,
+      "step": 8115
+    },
+    {
+      "epoch": 0.385510136257893,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.751646164724849e-05,
+      "loss": 0.2215,
+      "step": 8120
+    },
+    {
+      "epoch": 0.38574751934672175,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.750064189488899e-05,
+      "loss": 0.2261,
+      "step": 8125
+    },
+    {
+      "epoch": 0.3859849024355505,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.748481597917936e-05,
+      "loss": 0.2243,
+      "step": 8130
+    },
+    {
+      "epoch": 0.38622228552437926,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.746898390987305e-05,
+      "loss": 0.2212,
+      "step": 8135
+    },
+    {
+      "epoch": 0.386459668613208,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.7453145696727295e-05,
+      "loss": 0.2285,
+      "step": 8140
+    },
+    {
+      "epoch": 0.3866970517020367,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.7437301349503105e-05,
+      "loss": 0.2243,
+      "step": 8145
+    },
+    {
+      "epoch": 0.3869344347908655,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.742145087796527e-05,
+      "loss": 0.2196,
+      "step": 8150
+    },
+    {
+      "epoch": 0.38717181787969424,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.7405594291882394e-05,
+      "loss": 0.2257,
+      "step": 8155
+    },
+    {
+      "epoch": 0.387409200968523,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.738973160102679e-05,
+      "loss": 0.2253,
+      "step": 8160
+    },
+    {
+      "epoch": 0.38764658405735175,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.737386281517457e-05,
+      "loss": 0.2271,
+      "step": 8165
+    },
+    {
+      "epoch": 0.3878839671461805,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.73579879441056e-05,
+      "loss": 0.2281,
+      "step": 8170
+    },
+    {
+      "epoch": 0.38812135023500927,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.734210699760348e-05,
+      "loss": 0.2253,
+      "step": 8175
+    },
+    {
+      "epoch": 0.388358733323838,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.732621998545558e-05,
+      "loss": 0.2263,
+      "step": 8180
+    },
+    {
+      "epoch": 0.3885961164126668,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.731032691745298e-05,
+      "loss": 0.2263,
+      "step": 8185
+    },
+    {
+      "epoch": 0.38883349950149554,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.729442780339052e-05,
+      "loss": 0.2244,
+      "step": 8190
+    },
+    {
+      "epoch": 0.38907088259032424,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.727852265306674e-05,
+      "loss": 0.2275,
+      "step": 8195
+    },
+    {
+      "epoch": 0.389308265679153,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.726261147628392e-05,
+      "loss": 0.2246,
+      "step": 8200
+    },
+    {
+      "epoch": 0.38954564876798176,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.724669428284803e-05,
+      "loss": 0.2279,
+      "step": 8205
+    },
+    {
+      "epoch": 0.3897830318568105,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.723077108256878e-05,
+      "loss": 0.2289,
+      "step": 8210
+    },
+    {
+      "epoch": 0.39002041494563927,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.721484188525956e-05,
+      "loss": 0.2246,
+      "step": 8215
+    },
+    {
+      "epoch": 0.39025779803446803,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.7198906700737466e-05,
+      "loss": 0.2263,
+      "step": 8220
+    },
+    {
+      "epoch": 0.3904951811232968,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.718296553882327e-05,
+      "loss": 0.2263,
+      "step": 8225
+    },
+    {
+      "epoch": 0.39073256421212554,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.716701840934144e-05,
+      "loss": 0.2255,
+      "step": 8230
+    },
+    {
+      "epoch": 0.3909699473009543,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.715106532212014e-05,
+      "loss": 0.2301,
+      "step": 8235
+    },
+    {
+      "epoch": 0.39120733038978306,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.7135106286991154e-05,
+      "loss": 0.2297,
+      "step": 8240
+    },
+    {
+      "epoch": 0.39144471347861176,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.711914131378998e-05,
+      "loss": 0.2229,
+      "step": 8245
+    },
+    {
+      "epoch": 0.3916820965674405,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.710317041235576e-05,
+      "loss": 0.2262,
+      "step": 8250
+    },
+    {
+      "epoch": 0.3919194796562693,
+      "grad_norm": 0.5,
+      "learning_rate": 3.708719359253128e-05,
+      "loss": 0.2255,
+      "step": 8255
+    },
+    {
+      "epoch": 0.39215686274509803,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.707121086416299e-05,
+      "loss": 0.2233,
+      "step": 8260
+    },
+    {
+      "epoch": 0.3923942458339268,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.7055222237100965e-05,
+      "loss": 0.2247,
+      "step": 8265
+    },
+    {
+      "epoch": 0.39263162892275555,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.703922772119892e-05,
+      "loss": 0.2282,
+      "step": 8270
+    },
+    {
+      "epoch": 0.3928690120115843,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.702322732631422e-05,
+      "loss": 0.223,
+      "step": 8275
+    },
+    {
+      "epoch": 0.39310639510041306,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.700722106230781e-05,
+      "loss": 0.2244,
+      "step": 8280
+    },
+    {
+      "epoch": 0.3933437781892418,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.69912089390443e-05,
+      "loss": 0.2283,
+      "step": 8285
+    },
+    {
+      "epoch": 0.3935811612780705,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.697519096639188e-05,
+      "loss": 0.2239,
+      "step": 8290
+    },
+    {
+      "epoch": 0.3938185443668993,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.695916715422236e-05,
+      "loss": 0.2264,
+      "step": 8295
+    },
+    {
+      "epoch": 0.39405592745572804,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.694313751241113e-05,
+      "loss": 0.2273,
+      "step": 8300
+    },
+    {
+      "epoch": 0.3942933105445568,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.692710205083719e-05,
+      "loss": 0.2247,
+      "step": 8305
+    },
+    {
+      "epoch": 0.39453069363338555,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.6911060779383126e-05,
+      "loss": 0.2236,
+      "step": 8310
+    },
+    {
+      "epoch": 0.3947680767222143,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.6895013707935095e-05,
+      "loss": 0.2239,
+      "step": 8315
+    },
+    {
+      "epoch": 0.39500545981104307,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.687896084638285e-05,
+      "loss": 0.2261,
+      "step": 8320
+    },
+    {
+      "epoch": 0.3952428428998718,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.686290220461967e-05,
+      "loss": 0.2239,
+      "step": 8325
+    },
+    {
+      "epoch": 0.3954802259887006,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.6846837792542446e-05,
+      "loss": 0.2249,
+      "step": 8330
+    },
+    {
+      "epoch": 0.39571760907752934,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.683076762005159e-05,
+      "loss": 0.2239,
+      "step": 8335
+    },
+    {
+      "epoch": 0.39595499216635804,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.681469169705109e-05,
+      "loss": 0.223,
+      "step": 8340
+    },
+    {
+      "epoch": 0.3961923752551868,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.6798610033448434e-05,
+      "loss": 0.2253,
+      "step": 8345
+    },
+    {
+      "epoch": 0.39642975834401556,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.678252263915471e-05,
+      "loss": 0.2263,
+      "step": 8350
+    },
+    {
+      "epoch": 0.3966671414328443,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.67664295240845e-05,
+      "loss": 0.2286,
+      "step": 8355
+    },
+    {
+      "epoch": 0.3969045245216731,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.675033069815591e-05,
+      "loss": 0.2292,
+      "step": 8360
+    },
+    {
+      "epoch": 0.39714190761050183,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.6734226171290556e-05,
+      "loss": 0.2274,
+      "step": 8365
+    },
+    {
+      "epoch": 0.3973792906993306,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.671811595341362e-05,
+      "loss": 0.2252,
+      "step": 8370
+    },
+    {
+      "epoch": 0.39761667378815935,
+      "grad_norm": 0.5,
+      "learning_rate": 3.670200005445374e-05,
+      "loss": 0.2269,
+      "step": 8375
+    },
+    {
+      "epoch": 0.3978540568769881,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.6685878484343075e-05,
+      "loss": 0.2248,
+      "step": 8380
+    },
+    {
+      "epoch": 0.39809143996581686,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.666975125301726e-05,
+      "loss": 0.2249,
+      "step": 8385
+    },
+    {
+      "epoch": 0.39832882305464556,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.665361837041545e-05,
+      "loss": 0.2255,
+      "step": 8390
+    },
+    {
+      "epoch": 0.3985662061434743,
+      "grad_norm": 0.5,
+      "learning_rate": 3.663747984648026e-05,
+      "loss": 0.2284,
+      "step": 8395
+    },
+    {
+      "epoch": 0.3988035892323031,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.66213356911578e-05,
+      "loss": 0.2235,
+      "step": 8400
+    },
+    {
+      "epoch": 0.39904097232113184,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.6605185914397625e-05,
+      "loss": 0.2289,
+      "step": 8405
+    },
+    {
+      "epoch": 0.3992783554099606,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.658903052615278e-05,
+      "loss": 0.2219,
+      "step": 8410
+    },
+    {
+      "epoch": 0.39951573849878935,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.6572869536379736e-05,
+      "loss": 0.2271,
+      "step": 8415
+    },
+    {
+      "epoch": 0.3997531215876181,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.655670295503847e-05,
+      "loss": 0.2214,
+      "step": 8420
+    },
+    {
+      "epoch": 0.39999050467644687,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.654053079209235e-05,
+      "loss": 0.2273,
+      "step": 8425
+    },
+    {
+      "epoch": 0.4002278877652756,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.652435305750822e-05,
+      "loss": 0.2247,
+      "step": 8430
+    },
+    {
+      "epoch": 0.4004652708541043,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.650816976125633e-05,
+      "loss": 0.2264,
+      "step": 8435
+    },
+    {
+      "epoch": 0.4007026539429331,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.649198091331039e-05,
+      "loss": 0.227,
+      "step": 8440
+    },
+    {
+      "epoch": 0.40094003703176184,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.647578652364749e-05,
+      "loss": 0.2279,
+      "step": 8445
+    },
+    {
+      "epoch": 0.4011774201205906,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.645958660224818e-05,
+      "loss": 0.2251,
+      "step": 8450
+    },
+    {
+      "epoch": 0.40141480320941936,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.644338115909638e-05,
+      "loss": 0.225,
+      "step": 8455
+    },
+    {
+      "epoch": 0.4016521862982481,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.642717020417945e-05,
+      "loss": 0.2246,
+      "step": 8460
+    },
+    {
+      "epoch": 0.4018895693870769,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.641095374748811e-05,
+      "loss": 0.2265,
+      "step": 8465
+    },
+    {
+      "epoch": 0.40212695247590563,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.63947317990165e-05,
+      "loss": 0.2261,
+      "step": 8470
+    },
+    {
+      "epoch": 0.4023643355647344,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.6378504368762124e-05,
+      "loss": 0.2242,
+      "step": 8475
+    },
+    {
+      "epoch": 0.40260171865356315,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.636227146672589e-05,
+      "loss": 0.2224,
+      "step": 8480
+    },
+    {
+      "epoch": 0.40283910174239185,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.634603310291204e-05,
+      "loss": 0.2248,
+      "step": 8485
+    },
+    {
+      "epoch": 0.4030764848312206,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.632978928732823e-05,
+      "loss": 0.2293,
+      "step": 8490
+    },
+    {
+      "epoch": 0.40331386792004936,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.6313540029985414e-05,
+      "loss": 0.2236,
+      "step": 8495
+    },
+    {
+      "epoch": 0.4035512510088781,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.6297285340897965e-05,
+      "loss": 0.2296,
+      "step": 8500
+    },
+    {
+      "epoch": 0.4037886340977069,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.6281025230083566e-05,
+      "loss": 0.2238,
+      "step": 8505
+    },
+    {
+      "epoch": 0.40402601718653564,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.626475970756324e-05,
+      "loss": 0.2295,
+      "step": 8510
+    },
+    {
+      "epoch": 0.4042634002753644,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.624848878336137e-05,
+      "loss": 0.2246,
+      "step": 8515
+    },
+    {
+      "epoch": 0.40450078336419315,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.623221246750565e-05,
+      "loss": 0.2259,
+      "step": 8520
+    },
+    {
+      "epoch": 0.4047381664530219,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.6215930770027083e-05,
+      "loss": 0.224,
+      "step": 8525
+    },
+    {
+      "epoch": 0.40497554954185067,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.619964370096002e-05,
+      "loss": 0.2281,
+      "step": 8530
+    },
+    {
+      "epoch": 0.40521293263067937,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.61833512703421e-05,
+      "loss": 0.2224,
+      "step": 8535
+    },
+    {
+      "epoch": 0.4054503157195081,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.6167053488214286e-05,
+      "loss": 0.2222,
+      "step": 8540
+    },
+    {
+      "epoch": 0.4056876988083369,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.6150750364620825e-05,
+      "loss": 0.2248,
+      "step": 8545
+    },
+    {
+      "epoch": 0.40592508189716564,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.613444190960924e-05,
+      "loss": 0.2241,
+      "step": 8550
+    },
+    {
+      "epoch": 0.4061624649859944,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.6118128133230374e-05,
+      "loss": 0.2192,
+      "step": 8555
+    },
+    {
+      "epoch": 0.40639984807482316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.610180904553832e-05,
+      "loss": 0.2236,
+      "step": 8560
+    },
+    {
+      "epoch": 0.4066372311636519,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.608548465659048e-05,
+      "loss": 0.2273,
+      "step": 8565
+    },
+    {
+      "epoch": 0.40687461425248067,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.606915497644748e-05,
+      "loss": 0.2271,
+      "step": 8570
+    },
+    {
+      "epoch": 0.40711199734130943,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.6052820015173236e-05,
+      "loss": 0.2222,
+      "step": 8575
+    },
+    {
+      "epoch": 0.40734938043013813,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.6036479782834906e-05,
+      "loss": 0.2236,
+      "step": 8580
+    },
+    {
+      "epoch": 0.4075867635189669,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.602013428950291e-05,
+      "loss": 0.2237,
+      "step": 8585
+    },
+    {
+      "epoch": 0.40782414660779565,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.6003783545250886e-05,
+      "loss": 0.2276,
+      "step": 8590
+    },
+    {
+      "epoch": 0.4080615296966244,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.5987427560155744e-05,
+      "loss": 0.2244,
+      "step": 8595
+    },
+    {
+      "epoch": 0.40829891278545316,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.597106634429759e-05,
+      "loss": 0.2238,
+      "step": 8600
+    },
+    {
+      "epoch": 0.4085362958742819,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.595469990775976e-05,
+      "loss": 0.2254,
+      "step": 8605
+    },
+    {
+      "epoch": 0.4087736789631107,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.593832826062883e-05,
+      "loss": 0.2248,
+      "step": 8610
+    },
+    {
+      "epoch": 0.40901106205193943,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.592195141299456e-05,
+      "loss": 0.2241,
+      "step": 8615
+    },
+    {
+      "epoch": 0.4092484451407682,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.5905569374949935e-05,
+      "loss": 0.2249,
+      "step": 8620
+    },
+    {
+      "epoch": 0.40948582822959695,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.588918215659113e-05,
+      "loss": 0.22,
+      "step": 8625
+    },
+    {
+      "epoch": 0.40972321131842565,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.587278976801751e-05,
+      "loss": 0.2285,
+      "step": 8630
+    },
+    {
+      "epoch": 0.4099605944072544,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.585639221933162e-05,
+      "loss": 0.2283,
+      "step": 8635
+    },
+    {
+      "epoch": 0.41019797749608317,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.583998952063921e-05,
+      "loss": 0.2266,
+      "step": 8640
+    },
+    {
+      "epoch": 0.4104353605849119,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.582358168204919e-05,
+      "loss": 0.2227,
+      "step": 8645
+    },
+    {
+      "epoch": 0.4106727436737407,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.5807168713673624e-05,
+      "loss": 0.2275,
+      "step": 8650
+    },
+    {
+      "epoch": 0.41091012676256944,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.579075062562776e-05,
+      "loss": 0.2245,
+      "step": 8655
+    },
+    {
+      "epoch": 0.4111475098513982,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.577432742802997e-05,
+      "loss": 0.2212,
+      "step": 8660
+    },
+    {
+      "epoch": 0.41138489294022695,
+      "grad_norm": 0.498046875,
+      "learning_rate": 3.575789913100183e-05,
+      "loss": 0.2257,
+      "step": 8665
+    },
+    {
+      "epoch": 0.4116222760290557,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.574146574466799e-05,
+      "loss": 0.2251,
+      "step": 8670
+    },
+    {
+      "epoch": 0.41185965911788447,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.5725027279156304e-05,
+      "loss": 0.2289,
+      "step": 8675
+    },
+    {
+      "epoch": 0.41209704220671317,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.5708583744597694e-05,
+      "loss": 0.2261,
+      "step": 8680
+    },
+    {
+      "epoch": 0.41233442529554193,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.569213515112626e-05,
+      "loss": 0.2289,
+      "step": 8685
+    },
+    {
+      "epoch": 0.4125718083843707,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.567568150887918e-05,
+      "loss": 0.2242,
+      "step": 8690
+    },
+    {
+      "epoch": 0.41280919147319944,
+      "grad_norm": 0.37890625,
+      "learning_rate": 3.565922282799676e-05,
+      "loss": 0.2211,
+      "step": 8695
+    },
+    {
+      "epoch": 0.4130465745620282,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.564275911862243e-05,
+      "loss": 0.2266,
+      "step": 8700
+    },
+    {
+      "epoch": 0.41328395765085696,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.562629039090266e-05,
+      "loss": 0.2276,
+      "step": 8705
+    },
+    {
+      "epoch": 0.4135213407396857,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.5609816654987095e-05,
+      "loss": 0.2245,
+      "step": 8710
+    },
+    {
+      "epoch": 0.4137587238285145,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.559333792102839e-05,
+      "loss": 0.223,
+      "step": 8715
+    },
+    {
+      "epoch": 0.41399610691734323,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.557685419918233e-05,
+      "loss": 0.2285,
+      "step": 8720
+    },
+    {
+      "epoch": 0.41423349000617193,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.556036549960776e-05,
+      "loss": 0.2253,
+      "step": 8725
+    },
+    {
+      "epoch": 0.4144708730950007,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.554387183246658e-05,
+      "loss": 0.2197,
+      "step": 8730
+    },
+    {
+      "epoch": 0.41470825618382945,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.552737320792377e-05,
+      "loss": 0.2245,
+      "step": 8735
+    },
+    {
+      "epoch": 0.4149456392726582,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.551086963614736e-05,
+      "loss": 0.2263,
+      "step": 8740
+    },
+    {
+      "epoch": 0.41518302236148696,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.549436112730841e-05,
+      "loss": 0.2249,
+      "step": 8745
+    },
+    {
+      "epoch": 0.4154204054503157,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.547784769158106e-05,
+      "loss": 0.2239,
+      "step": 8750
+    },
+    {
+      "epoch": 0.4156577885391445,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.546132933914245e-05,
+      "loss": 0.2282,
+      "step": 8755
+    },
+    {
+      "epoch": 0.41589517162797324,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.544480608017278e-05,
+      "loss": 0.2229,
+      "step": 8760
+    },
+    {
+      "epoch": 0.416132554716802,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.5428277924855236e-05,
+      "loss": 0.2226,
+      "step": 8765
+    },
+    {
+      "epoch": 0.41636993780563075,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.541174488337607e-05,
+      "loss": 0.2276,
+      "step": 8770
+    },
+    {
+      "epoch": 0.41660732089445945,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.5395206965924504e-05,
+      "loss": 0.2231,
+      "step": 8775
+    },
+    {
+      "epoch": 0.4168447039832882,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.537866418269279e-05,
+      "loss": 0.2234,
+      "step": 8780
+    },
+    {
+      "epoch": 0.41708208707211697,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.5362116543876164e-05,
+      "loss": 0.2244,
+      "step": 8785
+    },
+    {
+      "epoch": 0.4173194701609457,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.534556405967287e-05,
+      "loss": 0.2263,
+      "step": 8790
+    },
+    {
+      "epoch": 0.4175568532497745,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.532900674028411e-05,
+      "loss": 0.2289,
+      "step": 8795
+    },
+    {
+      "epoch": 0.41779423633860324,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.531244459591411e-05,
+      "loss": 0.2251,
+      "step": 8800
+    },
+    {
+      "epoch": 0.418031619427432,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.529587763677003e-05,
+      "loss": 0.2273,
+      "step": 8805
+    },
+    {
+      "epoch": 0.41826900251626076,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.5279305873062005e-05,
+      "loss": 0.2265,
+      "step": 8810
+    },
+    {
+      "epoch": 0.4185063856050895,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.5262729315003154e-05,
+      "loss": 0.2244,
+      "step": 8815
+    },
+    {
+      "epoch": 0.4187437686939183,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.5246147972809517e-05,
+      "loss": 0.2261,
+      "step": 8820
+    },
+    {
+      "epoch": 0.418981151782747,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.522956185670011e-05,
+      "loss": 0.2269,
+      "step": 8825
+    },
+    {
+      "epoch": 0.41921853487157573,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.521297097689689e-05,
+      "loss": 0.2239,
+      "step": 8830
+    },
+    {
+      "epoch": 0.4194559179604045,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.519637534362472e-05,
+      "loss": 0.2222,
+      "step": 8835
+    },
+    {
+      "epoch": 0.41969330104923325,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.517977496711142e-05,
+      "loss": 0.2247,
+      "step": 8840
+    },
+    {
+      "epoch": 0.419930684138062,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.516316985758774e-05,
+      "loss": 0.2253,
+      "step": 8845
+    },
+    {
+      "epoch": 0.42016806722689076,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.5146560025287325e-05,
+      "loss": 0.2224,
+      "step": 8850
+    },
+    {
+      "epoch": 0.4204054503157195,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.512994548044673e-05,
+      "loss": 0.2284,
+      "step": 8855
+    },
+    {
+      "epoch": 0.4206428334045483,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.511332623330544e-05,
+      "loss": 0.2257,
+      "step": 8860
+    },
+    {
+      "epoch": 0.42088021649337704,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.50967022941058e-05,
+      "loss": 0.2249,
+      "step": 8865
+    },
+    {
+      "epoch": 0.42111759958220574,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.508007367309309e-05,
+      "loss": 0.2243,
+      "step": 8870
+    },
+    {
+      "epoch": 0.4213549826710345,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.506344038051544e-05,
+      "loss": 0.2215,
+      "step": 8875
+    },
+    {
+      "epoch": 0.42159236575986325,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.5046802426623875e-05,
+      "loss": 0.2199,
+      "step": 8880
+    },
+    {
+      "epoch": 0.421829748848692,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.5030159821672296e-05,
+      "loss": 0.2266,
+      "step": 8885
+    },
+    {
+      "epoch": 0.42206713193752077,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.501351257591746e-05,
+      "loss": 0.2253,
+      "step": 8890
+    },
+    {
+      "epoch": 0.4223045150263495,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.499686069961899e-05,
+      "loss": 0.2221,
+      "step": 8895
+    },
+    {
+      "epoch": 0.4225418981151783,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.4980204203039366e-05,
+      "loss": 0.2239,
+      "step": 8900
+    },
+    {
+      "epoch": 0.42277928120400704,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.49635430964439e-05,
+      "loss": 0.2258,
+      "step": 8905
+    },
+    {
+      "epoch": 0.4230166642928358,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.4946877390100766e-05,
+      "loss": 0.2254,
+      "step": 8910
+    },
+    {
+      "epoch": 0.42325404738166456,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.493020709428096e-05,
+      "loss": 0.2272,
+      "step": 8915
+    },
+    {
+      "epoch": 0.42349143047049326,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.491353221925831e-05,
+      "loss": 0.2264,
+      "step": 8920
+    },
+    {
+      "epoch": 0.423728813559322,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.489685277530946e-05,
+      "loss": 0.2226,
+      "step": 8925
+    },
+    {
+      "epoch": 0.4239661966481508,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.488016877271388e-05,
+      "loss": 0.2251,
+      "step": 8930
+    },
+    {
+      "epoch": 0.42420357973697953,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.486348022175386e-05,
+      "loss": 0.2242,
+      "step": 8935
+    },
+    {
+      "epoch": 0.4244409628258083,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.484678713271445e-05,
+      "loss": 0.2268,
+      "step": 8940
+    },
+    {
+      "epoch": 0.42467834591463705,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.483008951588354e-05,
+      "loss": 0.226,
+      "step": 8945
+    },
+    {
+      "epoch": 0.4249157290034658,
+      "grad_norm": 0.5,
+      "learning_rate": 3.481338738155179e-05,
+      "loss": 0.2252,
+      "step": 8950
+    },
+    {
+      "epoch": 0.42515311209229456,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.479668074001266e-05,
+      "loss": 0.2243,
+      "step": 8955
+    },
+    {
+      "epoch": 0.4253904951811233,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.477996960156237e-05,
+      "loss": 0.2266,
+      "step": 8960
+    },
+    {
+      "epoch": 0.4256278782699521,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.476325397649992e-05,
+      "loss": 0.2238,
+      "step": 8965
+    },
+    {
+      "epoch": 0.4258652613587808,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.4746533875127065e-05,
+      "loss": 0.2257,
+      "step": 8970
+    },
+    {
+      "epoch": 0.42610264444760954,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.472980930774834e-05,
+      "loss": 0.2309,
+      "step": 8975
+    },
+    {
+      "epoch": 0.4263400275364383,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.4713080284671006e-05,
+      "loss": 0.2237,
+      "step": 8980
+    },
+    {
+      "epoch": 0.42657741062526705,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.469634681620511e-05,
+      "loss": 0.2206,
+      "step": 8985
+    },
+    {
+      "epoch": 0.4268147937140958,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.4679608912663385e-05,
+      "loss": 0.2207,
+      "step": 8990
+    },
+    {
+      "epoch": 0.42705217680292457,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.466286658436134e-05,
+      "loss": 0.2281,
+      "step": 8995
+    },
+    {
+      "epoch": 0.4272895598917533,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.464611984161719e-05,
+      "loss": 0.2286,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4275269429805821,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.4629368694751877e-05,
+      "loss": 0.2254,
+      "step": 9005
+    },
+    {
+      "epoch": 0.42776432606941084,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.461261315408904e-05,
+      "loss": 0.2208,
+      "step": 9010
+    },
+    {
+      "epoch": 0.42800170915823954,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.4595853229955076e-05,
+      "loss": 0.2271,
+      "step": 9015
+    },
+    {
+      "epoch": 0.4282390922470683,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.457908893267904e-05,
+      "loss": 0.2245,
+      "step": 9020
+    },
+    {
+      "epoch": 0.42847647533589706,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.456232027259267e-05,
+      "loss": 0.2219,
+      "step": 9025
+    },
+    {
+      "epoch": 0.4287138584247258,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.454554726003044e-05,
+      "loss": 0.229,
+      "step": 9030
+    },
+    {
+      "epoch": 0.42895124151355457,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.452876990532947e-05,
+      "loss": 0.227,
+      "step": 9035
+    },
+    {
+      "epoch": 0.42918862460238333,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.451198821882957e-05,
+      "loss": 0.2212,
+      "step": 9040
+    },
+    {
+      "epoch": 0.4294260076912121,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.449520221087321e-05,
+      "loss": 0.2213,
+      "step": 9045
+    },
+    {
+      "epoch": 0.42966339078004084,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.4478411891805545e-05,
+      "loss": 0.2256,
+      "step": 9050
+    },
+    {
+      "epoch": 0.4299007738688696,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.446161727197437e-05,
+      "loss": 0.2277,
+      "step": 9055
+    },
+    {
+      "epoch": 0.43013815695769836,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.444481836173011e-05,
+      "loss": 0.2212,
+      "step": 9060
+    },
+    {
+      "epoch": 0.43037554004652706,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.442801517142589e-05,
+      "loss": 0.2225,
+      "step": 9065
+    },
+    {
+      "epoch": 0.4306129231353558,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.441120771141741e-05,
+      "loss": 0.2227,
+      "step": 9070
+    },
+    {
+      "epoch": 0.4308503062241846,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.439439599206305e-05,
+      "loss": 0.2252,
+      "step": 9075
+    },
+    {
+      "epoch": 0.43108768931301333,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.437758002372379e-05,
+      "loss": 0.2235,
+      "step": 9080
+    },
+    {
+      "epoch": 0.4313250724018421,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.436075981676324e-05,
+      "loss": 0.2218,
+      "step": 9085
+    },
+    {
+      "epoch": 0.43156245549067085,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.4343935381547615e-05,
+      "loss": 0.2245,
+      "step": 9090
+    },
+    {
+      "epoch": 0.4317998385794996,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.4327106728445734e-05,
+      "loss": 0.2254,
+      "step": 9095
+    },
+    {
+      "epoch": 0.43203722166832836,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.431027386782902e-05,
+      "loss": 0.2268,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4322746047571571,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.429343681007148e-05,
+      "loss": 0.224,
+      "step": 9105
+    },
+    {
+      "epoch": 0.4325119878459859,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.4276595565549746e-05,
+      "loss": 0.2277,
+      "step": 9110
+    },
+    {
+      "epoch": 0.4327493709348146,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.425975014464297e-05,
+      "loss": 0.2222,
+      "step": 9115
+    },
+    {
+      "epoch": 0.43298675402364334,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.424290055773291e-05,
+      "loss": 0.2243,
+      "step": 9120
+    },
+    {
+      "epoch": 0.4332241371124721,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.422604681520391e-05,
+      "loss": 0.2295,
+      "step": 9125
+    },
+    {
+      "epoch": 0.43346152020130085,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.420918892744284e-05,
+      "loss": 0.2189,
+      "step": 9130
+    },
+    {
+      "epoch": 0.4336989032901296,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.4192326904839143e-05,
+      "loss": 0.2251,
+      "step": 9135
+    },
+    {
+      "epoch": 0.43393628637895837,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.417546075778481e-05,
+      "loss": 0.2204,
+      "step": 9140
+    },
+    {
+      "epoch": 0.4341736694677871,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.4158590496674365e-05,
+      "loss": 0.2249,
+      "step": 9145
+    },
+    {
+      "epoch": 0.4344110525566159,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.414171613190487e-05,
+      "loss": 0.2265,
+      "step": 9150
+    },
+    {
+      "epoch": 0.43464843564544464,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.412483767387592e-05,
+      "loss": 0.2268,
+      "step": 9155
+    },
+    {
+      "epoch": 0.43488581873427334,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.410795513298966e-05,
+      "loss": 0.2264,
+      "step": 9160
+    },
+    {
+      "epoch": 0.4351232018231021,
+      "grad_norm": 0.5,
+      "learning_rate": 3.4091068519650677e-05,
+      "loss": 0.2249,
+      "step": 9165
+    },
+    {
+      "epoch": 0.43536058491193086,
+      "grad_norm": 0.474609375,
+      "learning_rate": 3.407417784426615e-05,
+      "loss": 0.224,
+      "step": 9170
+    },
+    {
+      "epoch": 0.4355979680007596,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.40572831172457e-05,
+      "loss": 0.2241,
+      "step": 9175
+    },
+    {
+      "epoch": 0.4358353510895884,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.404038434900151e-05,
+      "loss": 0.2263,
+      "step": 9180
+    },
+    {
+      "epoch": 0.43607273417841713,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.402348154994816e-05,
+      "loss": 0.2272,
+      "step": 9185
+    },
+    {
+      "epoch": 0.4363101172672459,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.4006574730502806e-05,
+      "loss": 0.2274,
+      "step": 9190
+    },
+    {
+      "epoch": 0.43654750035607465,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.398966390108503e-05,
+      "loss": 0.2258,
+      "step": 9195
+    },
+    {
+      "epoch": 0.4367848834449034,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.39727490721169e-05,
+      "loss": 0.2196,
+      "step": 9200
+    },
+    {
+      "epoch": 0.43702226653373216,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.395583025402296e-05,
+      "loss": 0.2257,
+      "step": 9205
+    },
+    {
+      "epoch": 0.43725964962256086,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.393890745723018e-05,
+      "loss": 0.225,
+      "step": 9210
+    },
+    {
+      "epoch": 0.4374970327113896,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.392198069216801e-05,
+      "loss": 0.2285,
+      "step": 9215
+    },
+    {
+      "epoch": 0.4377344158002184,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.390504996926836e-05,
+      "loss": 0.2225,
+      "step": 9220
+    },
+    {
+      "epoch": 0.43797179888904714,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.388811529896552e-05,
+      "loss": 0.2276,
+      "step": 9225
+    },
+    {
+      "epoch": 0.4382091819778759,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.387117669169629e-05,
+      "loss": 0.2178,
+      "step": 9230
+    },
+    {
+      "epoch": 0.43844656506670465,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.385423415789983e-05,
+      "loss": 0.2289,
+      "step": 9235
+    },
+    {
+      "epoch": 0.4386839481555334,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.3837287708017744e-05,
+      "loss": 0.2253,
+      "step": 9240
+    },
+    {
+      "epoch": 0.43892133124436217,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.382033735249407e-05,
+      "loss": 0.2258,
+      "step": 9245
+    },
+    {
+      "epoch": 0.4391587143331909,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.380338310177522e-05,
+      "loss": 0.222,
+      "step": 9250
+    },
+    {
+      "epoch": 0.4393960974220197,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.378642496631004e-05,
+      "loss": 0.2246,
+      "step": 9255
+    },
+    {
+      "epoch": 0.4396334805108484,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.376946295654973e-05,
+      "loss": 0.2256,
+      "step": 9260
+    },
+    {
+      "epoch": 0.43987086359967714,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.37524970829479e-05,
+      "loss": 0.2258,
+      "step": 9265
+    },
+    {
+      "epoch": 0.4401082466885059,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.373552735596057e-05,
+      "loss": 0.223,
+      "step": 9270
+    },
+    {
+      "epoch": 0.44034562977733466,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.371855378604607e-05,
+      "loss": 0.2286,
+      "step": 9275
+    },
+    {
+      "epoch": 0.4405830128661634,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.3701576383665156e-05,
+      "loss": 0.2227,
+      "step": 9280
+    },
+    {
+      "epoch": 0.4408203959549922,
+      "grad_norm": 0.51171875,
+      "learning_rate": 3.368459515928092e-05,
+      "loss": 0.222,
+      "step": 9285
+    },
+    {
+      "epoch": 0.44105777904382093,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.3667610123358805e-05,
+      "loss": 0.2254,
+      "step": 9290
+    },
+    {
+      "epoch": 0.4412951621326497,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.3650621286366623e-05,
+      "loss": 0.2234,
+      "step": 9295
+    },
+    {
+      "epoch": 0.44153254522147845,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.36336286587745e-05,
+      "loss": 0.2275,
+      "step": 9300
+    },
+    {
+      "epoch": 0.44176992831030715,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.361663225105494e-05,
+      "loss": 0.2256,
+      "step": 9305
+    },
+    {
+      "epoch": 0.4420073113991359,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.359963207368273e-05,
+      "loss": 0.2268,
+      "step": 9310
+    },
+    {
+      "epoch": 0.44224469448796466,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.3582628137135016e-05,
+      "loss": 0.2251,
+      "step": 9315
+    },
+    {
+      "epoch": 0.4424820775767934,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.3565620451891234e-05,
+      "loss": 0.2246,
+      "step": 9320
+    },
+    {
+      "epoch": 0.4427194606656222,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.354860902843315e-05,
+      "loss": 0.2281,
+      "step": 9325
+    },
+    {
+      "epoch": 0.44295684375445094,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.353159387724483e-05,
+      "loss": 0.2259,
+      "step": 9330
+    },
+    {
+      "epoch": 0.4431942268432797,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.351457500881263e-05,
+      "loss": 0.2235,
+      "step": 9335
+    },
+    {
+      "epoch": 0.44343160993210845,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.34975524336252e-05,
+      "loss": 0.2232,
+      "step": 9340
+    },
+    {
+      "epoch": 0.4436689930209372,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.348052616217348e-05,
+      "loss": 0.2245,
+      "step": 9345
+    },
+    {
+      "epoch": 0.44390637610976597,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.346349620495068e-05,
+      "loss": 0.2259,
+      "step": 9350
+    },
+    {
+      "epoch": 0.44414375919859467,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.3446462572452284e-05,
+      "loss": 0.2275,
+      "step": 9355
+    },
+    {
+      "epoch": 0.4443811422874234,
+      "grad_norm": 0.5,
+      "learning_rate": 3.342942527517605e-05,
+      "loss": 0.2199,
+      "step": 9360
+    },
+    {
+      "epoch": 0.4446185253762522,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.341238432362199e-05,
+      "loss": 0.2246,
+      "step": 9365
+    },
+    {
+      "epoch": 0.44485590846508094,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.339533972829234e-05,
+      "loss": 0.2259,
+      "step": 9370
+    },
+    {
+      "epoch": 0.4450932915539097,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.337829149969164e-05,
+      "loss": 0.2257,
+      "step": 9375
+    },
+    {
+      "epoch": 0.44533067464273846,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.3361239648326616e-05,
+      "loss": 0.2258,
+      "step": 9380
+    },
+    {
+      "epoch": 0.4455680577315672,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.334418418470624e-05,
+      "loss": 0.2239,
+      "step": 9385
+    },
+    {
+      "epoch": 0.44580544082039597,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.332712511934173e-05,
+      "loss": 0.2244,
+      "step": 9390
+    },
+    {
+      "epoch": 0.44604282390922473,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.3310062462746514e-05,
+      "loss": 0.2197,
+      "step": 9395
+    },
+    {
+      "epoch": 0.4462802069980535,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.329299622543621e-05,
+      "loss": 0.2258,
+      "step": 9400
+    },
+    {
+      "epoch": 0.4465175900868822,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.327592641792868e-05,
+      "loss": 0.2233,
+      "step": 9405
+    },
+    {
+      "epoch": 0.44675497317571095,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.325885305074396e-05,
+      "loss": 0.226,
+      "step": 9410
+    },
+    {
+      "epoch": 0.4469923562645397,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.324177613440429e-05,
+      "loss": 0.2256,
+      "step": 9415
+    },
+    {
+      "epoch": 0.44722973935336846,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.322469567943411e-05,
+      "loss": 0.2246,
+      "step": 9420
+    },
+    {
+      "epoch": 0.4474671224421972,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.3207611696360017e-05,
+      "loss": 0.2247,
+      "step": 9425
+    },
+    {
+      "epoch": 0.447704505531026,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.319052419571078e-05,
+      "loss": 0.2236,
+      "step": 9430
+    },
+    {
+      "epoch": 0.44794188861985473,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.3173433188017354e-05,
+      "loss": 0.2317,
+      "step": 9435
+    },
+    {
+      "epoch": 0.4481792717086835,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.3156338683812867e-05,
+      "loss": 0.2249,
+      "step": 9440
+    },
+    {
+      "epoch": 0.44841665479751225,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.3139240693632555e-05,
+      "loss": 0.2232,
+      "step": 9445
+    },
+    {
+      "epoch": 0.44865403788634095,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.312213922801386e-05,
+      "loss": 0.2281,
+      "step": 9450
+    },
+    {
+      "epoch": 0.4488914209751697,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.310503429749631e-05,
+      "loss": 0.222,
+      "step": 9455
+    },
+    {
+      "epoch": 0.44912880406399847,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.308792591262162e-05,
+      "loss": 0.2217,
+      "step": 9460
+    },
+    {
+      "epoch": 0.4493661871528272,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.307081408393361e-05,
+      "loss": 0.2242,
+      "step": 9465
+    },
+    {
+      "epoch": 0.449603570241656,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.305369882197819e-05,
+      "loss": 0.2258,
+      "step": 9470
+    },
+    {
+      "epoch": 0.44984095333048474,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.3036580137303465e-05,
+      "loss": 0.2207,
+      "step": 9475
+    },
+    {
+      "epoch": 0.4500783364193135,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.301945804045956e-05,
+      "loss": 0.2228,
+      "step": 9480
+    },
+    {
+      "epoch": 0.45031571950814225,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.300233254199877e-05,
+      "loss": 0.2254,
+      "step": 9485
+    },
+    {
+      "epoch": 0.450553102596971,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.2985203652475455e-05,
+      "loss": 0.2231,
+      "step": 9490
+    },
+    {
+      "epoch": 0.45079048568579977,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.296807138244606e-05,
+      "loss": 0.2278,
+      "step": 9495
+    },
+    {
+      "epoch": 0.45102786877462847,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.2950935742469155e-05,
+      "loss": 0.2243,
+      "step": 9500
+    },
+    {
+      "epoch": 0.45126525186345723,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.2933796743105334e-05,
+      "loss": 0.2274,
+      "step": 9505
+    },
+    {
+      "epoch": 0.451502634952286,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.291665439491729e-05,
+      "loss": 0.2229,
+      "step": 9510
+    },
+    {
+      "epoch": 0.45174001804111474,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.289950870846977e-05,
+      "loss": 0.2227,
+      "step": 9515
+    },
+    {
+      "epoch": 0.4519774011299435,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.2882359694329587e-05,
+      "loss": 0.2233,
+      "step": 9520
+    },
+    {
+      "epoch": 0.45221478421877226,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.28652073630656e-05,
+      "loss": 0.2227,
+      "step": 9525
+    },
+    {
+      "epoch": 0.452452167307601,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.284805172524871e-05,
+      "loss": 0.2246,
+      "step": 9530
+    },
+    {
+      "epoch": 0.4526895503964298,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.283089279145187e-05,
+      "loss": 0.2267,
+      "step": 9535
+    },
+    {
+      "epoch": 0.45292693348525853,
+      "grad_norm": 0.5,
+      "learning_rate": 3.2813730572250034e-05,
+      "loss": 0.222,
+      "step": 9540
+    },
+    {
+      "epoch": 0.4531643165740873,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.279656507822021e-05,
+      "loss": 0.2268,
+      "step": 9545
+    },
+    {
+      "epoch": 0.453401699662916,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.277939631994142e-05,
+      "loss": 0.2237,
+      "step": 9550
+    },
+    {
+      "epoch": 0.45363908275174475,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.276222430799468e-05,
+      "loss": 0.2227,
+      "step": 9555
+    },
+    {
+      "epoch": 0.4538764658405735,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.274504905296303e-05,
+      "loss": 0.2261,
+      "step": 9560
+    },
+    {
+      "epoch": 0.45411384892940226,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.27278705654315e-05,
+      "loss": 0.226,
+      "step": 9565
+    },
+    {
+      "epoch": 0.454351232018231,
+      "grad_norm": 0.3671875,
+      "learning_rate": 3.2710688855987114e-05,
+      "loss": 0.2223,
+      "step": 9570
+    },
+    {
+      "epoch": 0.4545886151070598,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.269350393521888e-05,
+      "loss": 0.2241,
+      "step": 9575
+    },
+    {
+      "epoch": 0.45482599819588854,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.2676315813717785e-05,
+      "loss": 0.2228,
+      "step": 9580
+    },
+    {
+      "epoch": 0.4550633812847173,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.2659124502076793e-05,
+      "loss": 0.2216,
+      "step": 9585
+    },
+    {
+      "epoch": 0.45530076437354605,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.264193001089084e-05,
+      "loss": 0.2262,
+      "step": 9590
+    },
+    {
+      "epoch": 0.45553814746237475,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.2624732350756784e-05,
+      "loss": 0.2227,
+      "step": 9595
+    },
+    {
+      "epoch": 0.4557755305512035,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.26075315322735e-05,
+      "loss": 0.2227,
+      "step": 9600
+    },
+    {
+      "epoch": 0.45601291364003227,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.2590327566041753e-05,
+      "loss": 0.2271,
+      "step": 9605
+    },
+    {
+      "epoch": 0.456250296728861,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.257312046266427e-05,
+      "loss": 0.2253,
+      "step": 9610
+    },
+    {
+      "epoch": 0.4564876798176898,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.255591023274572e-05,
+      "loss": 0.2248,
+      "step": 9615
+    },
+    {
+      "epoch": 0.45672506290651854,
+      "grad_norm": 0.5078125,
+      "learning_rate": 3.253869688689268e-05,
+      "loss": 0.2217,
+      "step": 9620
+    },
+    {
+      "epoch": 0.4569624459953473,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.252148043571367e-05,
+      "loss": 0.2259,
+      "step": 9625
+    },
+    {
+      "epoch": 0.45719982908417606,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.250426088981909e-05,
+      "loss": 0.2247,
+      "step": 9630
+    },
+    {
+      "epoch": 0.4574372121730048,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.248703825982129e-05,
+      "loss": 0.2281,
+      "step": 9635
+    },
+    {
+      "epoch": 0.45767459526183357,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.2469812556334484e-05,
+      "loss": 0.2289,
+      "step": 9640
+    },
+    {
+      "epoch": 0.4579119783506623,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.24525837899748e-05,
+      "loss": 0.224,
+      "step": 9645
+    },
+    {
+      "epoch": 0.45814936143949103,
+      "grad_norm": 0.490234375,
+      "learning_rate": 3.2435351971360244e-05,
+      "loss": 0.2208,
+      "step": 9650
+    },
+    {
+      "epoch": 0.4583867445283198,
+      "grad_norm": 0.48046875,
+      "learning_rate": 3.241811711111071e-05,
+      "loss": 0.2215,
+      "step": 9655
+    },
+    {
+      "epoch": 0.45862412761714855,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.240087921984797e-05,
+      "loss": 0.2257,
+      "step": 9660
+    },
+    {
+      "epoch": 0.4588615107059773,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.238363830819565e-05,
+      "loss": 0.227,
+      "step": 9665
+    },
+    {
+      "epoch": 0.45909889379480606,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.236639438677925e-05,
+      "loss": 0.2227,
+      "step": 9670
+    },
+    {
+      "epoch": 0.4593362768836348,
+      "grad_norm": 0.5234375,
+      "learning_rate": 3.2349147466226125e-05,
+      "loss": 0.2252,
+      "step": 9675
+    },
+    {
+      "epoch": 0.4595736599724636,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.233189755716545e-05,
+      "loss": 0.2285,
+      "step": 9680
+    },
+    {
+      "epoch": 0.45981104306129233,
+      "grad_norm": 0.5,
+      "learning_rate": 3.23146446702283e-05,
+      "loss": 0.2241,
+      "step": 9685
+    },
+    {
+      "epoch": 0.4600484261501211,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.229738881604752e-05,
+      "loss": 0.2248,
+      "step": 9690
+    },
+    {
+      "epoch": 0.4602858092389498,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.228013000525784e-05,
+      "loss": 0.2259,
+      "step": 9695
+    },
+    {
+      "epoch": 0.46052319232777855,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.226286824849576e-05,
+      "loss": 0.2232,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4607605754166073,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.224560355639964e-05,
+      "loss": 0.2245,
+      "step": 9705
+    },
+    {
+      "epoch": 0.46099795850543607,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.222833593960963e-05,
+      "loss": 0.2266,
+      "step": 9710
+    },
+    {
+      "epoch": 0.4612353415942648,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.221106540876769e-05,
+      "loss": 0.2262,
+      "step": 9715
+    },
+    {
+      "epoch": 0.4614727246830936,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.219379197451756e-05,
+      "loss": 0.2264,
+      "step": 9720
+    },
+    {
+      "epoch": 0.46171010777192234,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.217651564750478e-05,
+      "loss": 0.2204,
+      "step": 9725
+    },
+    {
+      "epoch": 0.4619474908607511,
+      "grad_norm": 0.375,
+      "learning_rate": 3.2159236438376664e-05,
+      "loss": 0.2239,
+      "step": 9730
+    },
+    {
+      "epoch": 0.46218487394957986,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.2141954357782325e-05,
+      "loss": 0.2274,
+      "step": 9735
+    },
+    {
+      "epoch": 0.46242225703840856,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.212466941637261e-05,
+      "loss": 0.2231,
+      "step": 9740
+    },
+    {
+      "epoch": 0.4626596401272373,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.2107381624800184e-05,
+      "loss": 0.224,
+      "step": 9745
+    },
+    {
+      "epoch": 0.4628970232160661,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.20900909937194e-05,
+      "loss": 0.2282,
+      "step": 9750
+    },
+    {
+      "epoch": 0.46313440630489483,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.2072797533786415e-05,
+      "loss": 0.2251,
+      "step": 9755
+    },
+    {
+      "epoch": 0.4633717893937236,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.20555012556591e-05,
+      "loss": 0.2262,
+      "step": 9760
+    },
+    {
+      "epoch": 0.46360917248255235,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.203820216999708e-05,
+      "loss": 0.2257,
+      "step": 9765
+    },
+    {
+      "epoch": 0.4638465555713811,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.2020900287461695e-05,
+      "loss": 0.2251,
+      "step": 9770
+    },
+    {
+      "epoch": 0.46408393866020986,
+      "grad_norm": 0.416015625,
+      "learning_rate": 3.200359561871603e-05,
+      "loss": 0.2228,
+      "step": 9775
+    },
+    {
+      "epoch": 0.4643213217490386,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.1986288174424856e-05,
+      "loss": 0.2261,
+      "step": 9780
+    },
+    {
+      "epoch": 0.4645587048378674,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.196897796525468e-05,
+      "loss": 0.2277,
+      "step": 9785
+    },
+    {
+      "epoch": 0.4647960879266961,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.19516650018737e-05,
+      "loss": 0.2267,
+      "step": 9790
+    },
+    {
+      "epoch": 0.46503347101552484,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.193434929495184e-05,
+      "loss": 0.2246,
+      "step": 9795
+    },
+    {
+      "epoch": 0.4652708541043536,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.1917030855160664e-05,
+      "loss": 0.2262,
+      "step": 9800
+    },
+    {
+      "epoch": 0.46550823719318235,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.1899709693173465e-05,
+      "loss": 0.2238,
+      "step": 9805
+    },
+    {
+      "epoch": 0.4657456202820111,
+      "grad_norm": 0.50390625,
+      "learning_rate": 3.188238581966518e-05,
+      "loss": 0.2278,
+      "step": 9810
+    },
+    {
+      "epoch": 0.46598300337083987,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.186505924531244e-05,
+      "loss": 0.2255,
+      "step": 9815
+    },
+    {
+      "epoch": 0.4662203864596686,
+      "grad_norm": 0.48828125,
+      "learning_rate": 3.184772998079355e-05,
+      "loss": 0.2265,
+      "step": 9820
+    },
+    {
+      "epoch": 0.4664577695484974,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.183039803678843e-05,
+      "loss": 0.2217,
+      "step": 9825
+    },
+    {
+      "epoch": 0.46669515263732614,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.1813063423978704e-05,
+      "loss": 0.2237,
+      "step": 9830
+    },
+    {
+      "epoch": 0.4669325357261549,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.179572615304759e-05,
+      "loss": 0.222,
+      "step": 9835
+    },
+    {
+      "epoch": 0.4671699188149836,
+      "grad_norm": 0.412109375,
+      "learning_rate": 3.177838623467999e-05,
+      "loss": 0.228,
+      "step": 9840
+    },
+    {
+      "epoch": 0.46740730190381236,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.17610436795624e-05,
+      "loss": 0.2257,
+      "step": 9845
+    },
+    {
+      "epoch": 0.4676446849926411,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.174369849838297e-05,
+      "loss": 0.2224,
+      "step": 9850
+    },
+    {
+      "epoch": 0.46788206808146987,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.1726350701831454e-05,
+      "loss": 0.227,
+      "step": 9855
+    },
+    {
+      "epoch": 0.46811945117029863,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.1709000300599224e-05,
+      "loss": 0.2269,
+      "step": 9860
+    },
+    {
+      "epoch": 0.4683568342591274,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.1691647305379245e-05,
+      "loss": 0.228,
+      "step": 9865
+    },
+    {
+      "epoch": 0.46859421734795614,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.1674291726866095e-05,
+      "loss": 0.2264,
+      "step": 9870
+    },
+    {
+      "epoch": 0.4688316004367849,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.165693357575593e-05,
+      "loss": 0.2228,
+      "step": 9875
+    },
+    {
+      "epoch": 0.46906898352561366,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.163957286274651e-05,
+      "loss": 0.2211,
+      "step": 9880
+    },
+    {
+      "epoch": 0.46930636661444236,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.162220959853716e-05,
+      "loss": 0.224,
+      "step": 9885
+    },
+    {
+      "epoch": 0.4695437497032711,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.160484379382877e-05,
+      "loss": 0.2203,
+      "step": 9890
+    },
+    {
+      "epoch": 0.4697811327920999,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.1587475459323816e-05,
+      "loss": 0.2208,
+      "step": 9895
+    },
+    {
+      "epoch": 0.47001851588092863,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.1570104605726334e-05,
+      "loss": 0.2247,
+      "step": 9900
+    },
+    {
+      "epoch": 0.4702558989697574,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.155273124374189e-05,
+      "loss": 0.2233,
+      "step": 9905
+    },
+    {
+      "epoch": 0.47049328205858615,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.153535538407761e-05,
+      "loss": 0.2213,
+      "step": 9910
+    },
+    {
+      "epoch": 0.4707306651474149,
+      "grad_norm": 0.408203125,
+      "learning_rate": 3.1517977037442145e-05,
+      "loss": 0.2245,
+      "step": 9915
+    },
+    {
+      "epoch": 0.47096804823624366,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.150059621454572e-05,
+      "loss": 0.221,
+      "step": 9920
+    },
+    {
+      "epoch": 0.4712054313250724,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.148321292610003e-05,
+      "loss": 0.2261,
+      "step": 9925
+    },
+    {
+      "epoch": 0.4714428144139012,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.146582718281833e-05,
+      "loss": 0.2221,
+      "step": 9930
+    },
+    {
+      "epoch": 0.4716801975027299,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.1448438995415365e-05,
+      "loss": 0.2269,
+      "step": 9935
+    },
+    {
+      "epoch": 0.47191758059155864,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.1431048374607405e-05,
+      "loss": 0.2252,
+      "step": 9940
+    },
+    {
+      "epoch": 0.4721549636803874,
+      "grad_norm": 0.47265625,
+      "learning_rate": 3.141365533111221e-05,
+      "loss": 0.2295,
+      "step": 9945
+    },
+    {
+      "epoch": 0.47239234676921615,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.139625987564903e-05,
+      "loss": 0.2229,
+      "step": 9950
+    },
+    {
+      "epoch": 0.4726297298580449,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.13788620189386e-05,
+      "loss": 0.2238,
+      "step": 9955
+    },
+    {
+      "epoch": 0.47286711294687367,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.136146177170316e-05,
+      "loss": 0.2239,
+      "step": 9960
+    },
+    {
+      "epoch": 0.4731044960357024,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.1344059144666375e-05,
+      "loss": 0.2221,
+      "step": 9965
+    },
+    {
+      "epoch": 0.4733418791245312,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.132665414855343e-05,
+      "loss": 0.2228,
+      "step": 9970
+    },
+    {
+      "epoch": 0.47357926221335994,
+      "grad_norm": 0.435546875,
+      "learning_rate": 3.130924679409093e-05,
+      "loss": 0.2253,
+      "step": 9975
+    },
+    {
+      "epoch": 0.4738166453021887,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.1291837092006954e-05,
+      "loss": 0.2221,
+      "step": 9980
+    },
+    {
+      "epoch": 0.4740540283910174,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.127442505303101e-05,
+      "loss": 0.2236,
+      "step": 9985
+    },
+    {
+      "epoch": 0.47429141147984616,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.125701068789408e-05,
+      "loss": 0.2213,
+      "step": 9990
+    },
+    {
+      "epoch": 0.4745287945686749,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.123959400732855e-05,
+      "loss": 0.2204,
+      "step": 9995
+    },
+    {
+      "epoch": 0.4747661776575037,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.1222175022068236e-05,
+      "loss": 0.2279,
+      "step": 10000
+    },
+    {
+      "epoch": 0.47500356074633243,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.120475374284838e-05,
+      "loss": 0.2218,
+      "step": 10005
+    },
+    {
+      "epoch": 0.4752409438351612,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.118733018040564e-05,
+      "loss": 0.2274,
+      "step": 10010
+    },
+    {
+      "epoch": 0.47547832692398995,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.116990434547807e-05,
+      "loss": 0.2254,
+      "step": 10015
+    },
+    {
+      "epoch": 0.4757157100128187,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.1152476248805145e-05,
+      "loss": 0.2259,
+      "step": 10020
+    },
+    {
+      "epoch": 0.47595309310164746,
+      "grad_norm": 0.494140625,
+      "learning_rate": 3.1135045901127726e-05,
+      "loss": 0.2232,
+      "step": 10025
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.111761331318804e-05,
+      "loss": 0.2227,
+      "step": 10030
+    },
+    {
+      "epoch": 0.4764278592793049,
+      "grad_norm": 0.5,
+      "learning_rate": 3.1100178495729716e-05,
+      "loss": 0.224,
+      "step": 10035
+    },
+    {
+      "epoch": 0.4766652423681337,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.108274145949776e-05,
+      "loss": 0.2256,
+      "step": 10040
+    },
+    {
+      "epoch": 0.47690262545696244,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.1065302215238534e-05,
+      "loss": 0.2236,
+      "step": 10045
+    },
+    {
+      "epoch": 0.4771400085457912,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.1047860773699764e-05,
+      "loss": 0.2255,
+      "step": 10050
+    },
+    {
+      "epoch": 0.47737739163461995,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.103041714563054e-05,
+      "loss": 0.2226,
+      "step": 10055
+    },
+    {
+      "epoch": 0.4776147747234487,
+      "grad_norm": 0.396484375,
+      "learning_rate": 3.101297134178128e-05,
+      "loss": 0.2217,
+      "step": 10060
+    },
+    {
+      "epoch": 0.47785215781227747,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.099552337290375e-05,
+      "loss": 0.2229,
+      "step": 10065
+    },
+    {
+      "epoch": 0.4780895409011062,
+      "grad_norm": 0.42578125,
+      "learning_rate": 3.0978073249751066e-05,
+      "loss": 0.2244,
+      "step": 10070
+    },
+    {
+      "epoch": 0.478326923989935,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.096062098307765e-05,
+      "loss": 0.2216,
+      "step": 10075
+    },
+    {
+      "epoch": 0.4785643070787637,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.094316658363926e-05,
+      "loss": 0.2262,
+      "step": 10080
+    },
+    {
+      "epoch": 0.47880169016759244,
+      "grad_norm": 0.404296875,
+      "learning_rate": 3.092571006219296e-05,
+      "loss": 0.2228,
+      "step": 10085
+    },
+    {
+      "epoch": 0.4790390732564212,
+      "grad_norm": 0.5,
+      "learning_rate": 3.090825142949713e-05,
+      "loss": 0.225,
+      "step": 10090
+    },
+    {
+      "epoch": 0.47927645634524996,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.089079069631144e-05,
+      "loss": 0.2258,
+      "step": 10095
+    },
+    {
+      "epoch": 0.4795138394340787,
+      "grad_norm": 0.515625,
+      "learning_rate": 3.087332787339687e-05,
+      "loss": 0.2238,
+      "step": 10100
+    },
+    {
+      "epoch": 0.4797512225229075,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.085586297151567e-05,
+      "loss": 0.2208,
+      "step": 10105
+    },
+    {
+      "epoch": 0.47998860561173623,
+      "grad_norm": 0.486328125,
+      "learning_rate": 3.083839600143137e-05,
+      "loss": 0.2199,
+      "step": 10110
+    },
+    {
+      "epoch": 0.480225988700565,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.0820926973908806e-05,
+      "loss": 0.2238,
+      "step": 10115
+    },
+    {
+      "epoch": 0.48046337178939375,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.080345589971404e-05,
+      "loss": 0.2211,
+      "step": 10120
+    },
+    {
+      "epoch": 0.4807007548782225,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.078598278961443e-05,
+      "loss": 0.2254,
+      "step": 10125
+    },
+    {
+      "epoch": 0.4809381379670512,
+      "grad_norm": 0.41796875,
+      "learning_rate": 3.076850765437857e-05,
+      "loss": 0.2219,
+      "step": 10130
+    },
+    {
+      "epoch": 0.48117552105587996,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.0751030504776315e-05,
+      "loss": 0.221,
+      "step": 10135
+    },
+    {
+      "epoch": 0.4814129041447087,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.0733551351578724e-05,
+      "loss": 0.2242,
+      "step": 10140
+    },
+    {
+      "epoch": 0.4816502872335375,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.071607020555815e-05,
+      "loss": 0.2239,
+      "step": 10145
+    },
+    {
+      "epoch": 0.48188767032236623,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.069858707748813e-05,
+      "loss": 0.2208,
+      "step": 10150
+    },
+    {
+      "epoch": 0.482125053411195,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.068110197814344e-05,
+      "loss": 0.2246,
+      "step": 10155
+    },
+    {
+      "epoch": 0.48236243650002375,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.066361491830007e-05,
+      "loss": 0.2243,
+      "step": 10160
+    },
+    {
+      "epoch": 0.4825998195888525,
+      "grad_norm": 0.49609375,
+      "learning_rate": 3.064612590873521e-05,
+      "loss": 0.2263,
+      "step": 10165
+    },
+    {
+      "epoch": 0.48283720267768127,
+      "grad_norm": 0.458984375,
+      "learning_rate": 3.062863496022725e-05,
+      "loss": 0.2259,
+      "step": 10170
+    },
+    {
+      "epoch": 0.48307458576650997,
+      "grad_norm": 0.43359375,
+      "learning_rate": 3.0611142083555806e-05,
+      "loss": 0.2243,
+      "step": 10175
+    },
+    {
+      "epoch": 0.4833119688553387,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.059364728950163e-05,
+      "loss": 0.226,
+      "step": 10180
+    },
+    {
+      "epoch": 0.4835493519441675,
+      "grad_norm": 0.41015625,
+      "learning_rate": 3.0576150588846704e-05,
+      "loss": 0.2261,
+      "step": 10185
+    },
+    {
+      "epoch": 0.48378673503299624,
+      "grad_norm": 0.462890625,
+      "learning_rate": 3.0558651992374154e-05,
+      "loss": 0.2251,
+      "step": 10190
+    },
+    {
+      "epoch": 0.484024118121825,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.0541151510868296e-05,
+      "loss": 0.226,
+      "step": 10195
+    },
+    {
+      "epoch": 0.48426150121065376,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.052364915511458e-05,
+      "loss": 0.223,
+      "step": 10200
+    },
+    {
+      "epoch": 0.4844988842994825,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.050614493589965e-05,
+      "loss": 0.2247,
+      "step": 10205
+    },
+    {
+      "epoch": 0.48473626738831127,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.0488638864011264e-05,
+      "loss": 0.2268,
+      "step": 10210
+    },
+    {
+      "epoch": 0.48497365047714003,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.0471130950238334e-05,
+      "loss": 0.2258,
+      "step": 10215
+    },
+    {
+      "epoch": 0.4852110335659688,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.0453621205370914e-05,
+      "loss": 0.219,
+      "step": 10220
+    },
+    {
+      "epoch": 0.4854484166547975,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.0436109640200173e-05,
+      "loss": 0.2202,
+      "step": 10225
+    },
+    {
+      "epoch": 0.48568579974362625,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.041859626551841e-05,
+      "loss": 0.2263,
+      "step": 10230
+    },
+    {
+      "epoch": 0.485923182832455,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.0401081092119045e-05,
+      "loss": 0.2268,
+      "step": 10235
+    },
+    {
+      "epoch": 0.48616056592128376,
+      "grad_norm": 0.453125,
+      "learning_rate": 3.0383564130796588e-05,
+      "loss": 0.2279,
+      "step": 10240
+    },
+    {
+      "epoch": 0.4863979490101125,
+      "grad_norm": 0.451171875,
+      "learning_rate": 3.0366045392346674e-05,
+      "loss": 0.2237,
+      "step": 10245
+    },
+    {
+      "epoch": 0.4866353320989413,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.034852488756602e-05,
+      "loss": 0.2227,
+      "step": 10250
+    },
+    {
+      "epoch": 0.48687271518777003,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.033100262725243e-05,
+      "loss": 0.2274,
+      "step": 10255
+    },
+    {
+      "epoch": 0.4871100982765988,
+      "grad_norm": 0.52734375,
+      "learning_rate": 3.03134786222048e-05,
+      "loss": 0.2257,
+      "step": 10260
+    },
+    {
+      "epoch": 0.48734748136542755,
+      "grad_norm": 0.40625,
+      "learning_rate": 3.029595288322308e-05,
+      "loss": 0.2226,
+      "step": 10265
+    },
+    {
+      "epoch": 0.4875848644542563,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.0278425421108315e-05,
+      "loss": 0.221,
+      "step": 10270
+    },
+    {
+      "epoch": 0.487822247543085,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.026089624666259e-05,
+      "loss": 0.2235,
+      "step": 10275
+    },
+    {
+      "epoch": 0.48805963063191377,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.0243365370689075e-05,
+      "loss": 0.2242,
+      "step": 10280
+    },
+    {
+      "epoch": 0.4882970137207425,
+      "grad_norm": 0.4609375,
+      "learning_rate": 3.022583280399195e-05,
+      "loss": 0.226,
+      "step": 10285
+    },
+    {
+      "epoch": 0.4885343968095713,
+      "grad_norm": 0.482421875,
+      "learning_rate": 3.020829855737647e-05,
+      "loss": 0.2221,
+      "step": 10290
+    },
+    {
+      "epoch": 0.48877177989840004,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.019076264164889e-05,
+      "loss": 0.2224,
+      "step": 10295
+    },
+    {
+      "epoch": 0.4890091629872288,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.017322506761654e-05,
+      "loss": 0.2248,
+      "step": 10300
+    },
+    {
+      "epoch": 0.48924654607605755,
+      "grad_norm": 0.484375,
+      "learning_rate": 3.015568584608774e-05,
+      "loss": 0.2243,
+      "step": 10305
+    },
+    {
+      "epoch": 0.4894839291648863,
+      "grad_norm": 0.470703125,
+      "learning_rate": 3.0138144987871825e-05,
+      "loss": 0.2256,
+      "step": 10310
+    },
+    {
+      "epoch": 0.48972131225371507,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.012060250377915e-05,
+      "loss": 0.2266,
+      "step": 10315
+    },
+    {
+      "epoch": 0.48995869534254377,
+      "grad_norm": 0.466796875,
+      "learning_rate": 3.010305840462107e-05,
+      "loss": 0.2232,
+      "step": 10320
+    },
+    {
+      "epoch": 0.49019607843137253,
+      "grad_norm": 0.4375,
+      "learning_rate": 3.0085512701209935e-05,
+      "loss": 0.2228,
+      "step": 10325
+    },
+    {
+      "epoch": 0.4904334615202013,
+      "grad_norm": 0.4921875,
+      "learning_rate": 3.0067965404359077e-05,
+      "loss": 0.224,
+      "step": 10330
+    },
+    {
+      "epoch": 0.49067084460903004,
+      "grad_norm": 0.46875,
+      "learning_rate": 3.0050416524882807e-05,
+      "loss": 0.2248,
+      "step": 10335
+    },
+    {
+      "epoch": 0.4909082276978588,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.0032866073596445e-05,
+      "loss": 0.2242,
+      "step": 10340
+    },
+    {
+      "epoch": 0.49114561078668756,
+      "grad_norm": 0.478515625,
+      "learning_rate": 3.0015314061316236e-05,
+      "loss": 0.2251,
+      "step": 10345
+    },
+    {
+      "epoch": 0.4913829938755163,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.9997760498859407e-05,
+      "loss": 0.227,
+      "step": 10350
+    },
+    {
+      "epoch": 0.4916203769643451,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.998020539704414e-05,
+      "loss": 0.2267,
+      "step": 10355
+    },
+    {
+      "epoch": 0.49185776005317383,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.9962648766689573e-05,
+      "loss": 0.2246,
+      "step": 10360
+    },
+    {
+      "epoch": 0.4920951431420026,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.9945090618615756e-05,
+      "loss": 0.2257,
+      "step": 10365
+    },
+    {
+      "epoch": 0.4923325262308313,
+      "grad_norm": 0.5,
+      "learning_rate": 2.992753096364372e-05,
+      "loss": 0.2241,
+      "step": 10370
+    },
+    {
+      "epoch": 0.49256990931966005,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.9909969812595385e-05,
+      "loss": 0.2245,
+      "step": 10375
+    },
+    {
+      "epoch": 0.4928072924084888,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.9892407176293612e-05,
+      "loss": 0.2261,
+      "step": 10380
+    },
+    {
+      "epoch": 0.49304467549731756,
+      "grad_norm": 0.5,
+      "learning_rate": 2.987484306556218e-05,
+      "loss": 0.2186,
+      "step": 10385
+    },
+    {
+      "epoch": 0.4932820585861463,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.9857277491225767e-05,
+      "loss": 0.2212,
+      "step": 10390
+    },
+    {
+      "epoch": 0.4935194416749751,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.9839710464109944e-05,
+      "loss": 0.2243,
+      "step": 10395
+    },
+    {
+      "epoch": 0.49375682476380384,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.9822141995041214e-05,
+      "loss": 0.2266,
+      "step": 10400
+    },
+    {
+      "epoch": 0.4939942078526326,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.9804572094846924e-05,
+      "loss": 0.2257,
+      "step": 10405
+    },
+    {
+      "epoch": 0.49423159094146135,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.978700077435534e-05,
+      "loss": 0.2222,
+      "step": 10410
+    },
+    {
+      "epoch": 0.4944689740302901,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.9769428044395568e-05,
+      "loss": 0.2271,
+      "step": 10415
+    },
+    {
+      "epoch": 0.4947063571191188,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.9751853915797617e-05,
+      "loss": 0.2225,
+      "step": 10420
+    },
+    {
+      "epoch": 0.49494374020794757,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.973427839939234e-05,
+      "loss": 0.2256,
+      "step": 10425
+    },
+    {
+      "epoch": 0.4951811232967763,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.971670150601144e-05,
+      "loss": 0.2247,
+      "step": 10430
+    },
+    {
+      "epoch": 0.4954185063856051,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.9699123246487487e-05,
+      "loss": 0.2238,
+      "step": 10435
+    },
+    {
+      "epoch": 0.49565588947443384,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.9681543631653873e-05,
+      "loss": 0.2215,
+      "step": 10440
+    },
+    {
+      "epoch": 0.4958932725632626,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.9663962672344842e-05,
+      "loss": 0.2245,
+      "step": 10445
+    },
+    {
+      "epoch": 0.49613065565209136,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.9646380379395456e-05,
+      "loss": 0.2213,
+      "step": 10450
+    },
+    {
+      "epoch": 0.4963680387409201,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.9628796763641607e-05,
+      "loss": 0.2276,
+      "step": 10455
+    },
+    {
+      "epoch": 0.49660542182974887,
+      "grad_norm": 0.5,
+      "learning_rate": 2.9611211835919982e-05,
+      "loss": 0.2258,
+      "step": 10460
+    },
+    {
+      "epoch": 0.4968428049185776,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.9593625607068116e-05,
+      "loss": 0.2213,
+      "step": 10465
+    },
+    {
+      "epoch": 0.49708018800740633,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.9576038087924297e-05,
+      "loss": 0.2233,
+      "step": 10470
+    },
+    {
+      "epoch": 0.4973175710962351,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.955844928932765e-05,
+      "loss": 0.2263,
+      "step": 10475
+    },
+    {
+      "epoch": 0.49755495418506385,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.9540859222118068e-05,
+      "loss": 0.2219,
+      "step": 10480
+    },
+    {
+      "epoch": 0.4977923372738926,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.9523267897136226e-05,
+      "loss": 0.2252,
+      "step": 10485
+    },
+    {
+      "epoch": 0.49802972036272136,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.9505675325223575e-05,
+      "loss": 0.2244,
+      "step": 10490
+    },
+    {
+      "epoch": 0.4982671034515501,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.9488081517222347e-05,
+      "loss": 0.2239,
+      "step": 10495
+    },
+    {
+      "epoch": 0.4985044865403789,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.9470486483975517e-05,
+      "loss": 0.2217,
+      "step": 10500
+    },
+    {
+      "epoch": 0.49874186962920763,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.9452890236326834e-05,
+      "loss": 0.2248,
+      "step": 10505
+    },
+    {
+      "epoch": 0.4989792527180364,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.9435292785120775e-05,
+      "loss": 0.2266,
+      "step": 10510
+    },
+    {
+      "epoch": 0.4992166358068651,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.9417694141202572e-05,
+      "loss": 0.222,
+      "step": 10515
+    },
+    {
+      "epoch": 0.49945401889569385,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.9400094315418193e-05,
+      "loss": 0.2276,
+      "step": 10520
+    },
+    {
+      "epoch": 0.4996914019845226,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.9382493318614323e-05,
+      "loss": 0.2228,
+      "step": 10525
+    },
+    {
+      "epoch": 0.49992878507335137,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.9364891161638382e-05,
+      "loss": 0.2272,
+      "step": 10530
+    },
+    {
+      "epoch": 0.5001661681621802,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.9347287855338502e-05,
+      "loss": 0.2258,
+      "step": 10535
+    },
+    {
+      "epoch": 0.5004035512510089,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.932968341056351e-05,
+      "loss": 0.2237,
+      "step": 10540
+    },
+    {
+      "epoch": 0.5006409343398376,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.931207783816296e-05,
+      "loss": 0.2229,
+      "step": 10545
+    },
+    {
+      "epoch": 0.5008783174286664,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.9294471148987074e-05,
+      "loss": 0.2213,
+      "step": 10550
+    },
+    {
+      "epoch": 0.5011157005174951,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.927686335388678e-05,
+      "loss": 0.2223,
+      "step": 10555
+    },
+    {
+      "epoch": 0.5013530836063239,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.925925446371368e-05,
+      "loss": 0.2272,
+      "step": 10560
+    },
+    {
+      "epoch": 0.5015904666951526,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.924164448932006e-05,
+      "loss": 0.2232,
+      "step": 10565
+    },
+    {
+      "epoch": 0.5018278497839814,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.9224033441558852e-05,
+      "loss": 0.222,
+      "step": 10570
+    },
+    {
+      "epoch": 0.5020652328728101,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.920642133128368e-05,
+      "loss": 0.2235,
+      "step": 10575
+    },
+    {
+      "epoch": 0.5023026159616389,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.9188808169348797e-05,
+      "loss": 0.2273,
+      "step": 10580
+    },
+    {
+      "epoch": 0.5025399990504676,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.9171193966609124e-05,
+      "loss": 0.2228,
+      "step": 10585
+    },
+    {
+      "epoch": 0.5027773821392963,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.91535787339202e-05,
+      "loss": 0.2204,
+      "step": 10590
+    },
+    {
+      "epoch": 0.5030147652281252,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.913596248213823e-05,
+      "loss": 0.2246,
+      "step": 10595
+    },
+    {
+      "epoch": 0.5032521483169539,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.911834522212002e-05,
+      "loss": 0.2227,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5034895314057827,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.9100726964723007e-05,
+      "loss": 0.2244,
+      "step": 10605
+    },
+    {
+      "epoch": 0.5037269144946114,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.9083107720805247e-05,
+      "loss": 0.2252,
+      "step": 10610
+    },
+    {
+      "epoch": 0.5039642975834402,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.9065487501225393e-05,
+      "loss": 0.2223,
+      "step": 10615
+    },
+    {
+      "epoch": 0.5042016806722689,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.9047866316842713e-05,
+      "loss": 0.2248,
+      "step": 10620
+    },
+    {
+      "epoch": 0.5044390637610977,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.9030244178517058e-05,
+      "loss": 0.2217,
+      "step": 10625
+    },
+    {
+      "epoch": 0.5046764468499264,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.9012621097108876e-05,
+      "loss": 0.223,
+      "step": 10630
+    },
+    {
+      "epoch": 0.5049138299387551,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.899499708347919e-05,
+      "loss": 0.2255,
+      "step": 10635
+    },
+    {
+      "epoch": 0.5051512130275839,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.8977372148489607e-05,
+      "loss": 0.2282,
+      "step": 10640
+    },
+    {
+      "epoch": 0.5053885961164126,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.895974630300228e-05,
+      "loss": 0.2251,
+      "step": 10645
+    },
+    {
+      "epoch": 0.5056259792052414,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.894211955787995e-05,
+      "loss": 0.224,
+      "step": 10650
+    },
+    {
+      "epoch": 0.5058633622940701,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.8924491923985886e-05,
+      "loss": 0.2246,
+      "step": 10655
+    },
+    {
+      "epoch": 0.506100745382899,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.8906863412183933e-05,
+      "loss": 0.2223,
+      "step": 10660
+    },
+    {
+      "epoch": 0.5063381284717277,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.888923403333845e-05,
+      "loss": 0.2242,
+      "step": 10665
+    },
+    {
+      "epoch": 0.5065755115605565,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.887160379831435e-05,
+      "loss": 0.2309,
+      "step": 10670
+    },
+    {
+      "epoch": 0.5068128946493852,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.8853972717977075e-05,
+      "loss": 0.2205,
+      "step": 10675
+    },
+    {
+      "epoch": 0.507050277738214,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.8836340803192565e-05,
+      "loss": 0.2292,
+      "step": 10680
+    },
+    {
+      "epoch": 0.5072876608270427,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.8818708064827295e-05,
+      "loss": 0.2234,
+      "step": 10685
+    },
+    {
+      "epoch": 0.5075250439158714,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.8801074513748237e-05,
+      "loss": 0.2262,
+      "step": 10690
+    },
+    {
+      "epoch": 0.5077624270047002,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.878344016082288e-05,
+      "loss": 0.224,
+      "step": 10695
+    },
+    {
+      "epoch": 0.5079998100935289,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.876580501691919e-05,
+      "loss": 0.2242,
+      "step": 10700
+    },
+    {
+      "epoch": 0.5082371931823577,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.874816909290562e-05,
+      "loss": 0.2226,
+      "step": 10705
+    },
+    {
+      "epoch": 0.5084745762711864,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.873053239965111e-05,
+      "loss": 0.2288,
+      "step": 10710
+    },
+    {
+      "epoch": 0.5087119593600152,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.8712894948025083e-05,
+      "loss": 0.2262,
+      "step": 10715
+    },
+    {
+      "epoch": 0.5089493424488439,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.8695256748897413e-05,
+      "loss": 0.2271,
+      "step": 10720
+    },
+    {
+      "epoch": 0.5091867255376727,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.867761781313844e-05,
+      "loss": 0.2217,
+      "step": 10725
+    },
+    {
+      "epoch": 0.5094241086265014,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.8659978151618965e-05,
+      "loss": 0.2254,
+      "step": 10730
+    },
+    {
+      "epoch": 0.5096614917153302,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.864233777521024e-05,
+      "loss": 0.2241,
+      "step": 10735
+    },
+    {
+      "epoch": 0.509898874804159,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.862469669478393e-05,
+      "loss": 0.2165,
+      "step": 10740
+    },
+    {
+      "epoch": 0.5101362578929877,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.8607054921212156e-05,
+      "loss": 0.2262,
+      "step": 10745
+    },
+    {
+      "epoch": 0.5103736409818165,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.8589412465367478e-05,
+      "loss": 0.2242,
+      "step": 10750
+    },
+    {
+      "epoch": 0.5106110240706452,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.857176933812285e-05,
+      "loss": 0.2234,
+      "step": 10755
+    },
+    {
+      "epoch": 0.510848407159474,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.8554125550351645e-05,
+      "loss": 0.2249,
+      "step": 10760
+    },
+    {
+      "epoch": 0.5110857902483027,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.853648111292767e-05,
+      "loss": 0.2218,
+      "step": 10765
+    },
+    {
+      "epoch": 0.5113231733371315,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.851883603672509e-05,
+      "loss": 0.225,
+      "step": 10770
+    },
+    {
+      "epoch": 0.5115605564259602,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.8501190332618495e-05,
+      "loss": 0.2232,
+      "step": 10775
+    },
+    {
+      "epoch": 0.5117979395147889,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.8483544011482856e-05,
+      "loss": 0.2224,
+      "step": 10780
+    },
+    {
+      "epoch": 0.5120353226036177,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.8465897084193506e-05,
+      "loss": 0.2221,
+      "step": 10785
+    },
+    {
+      "epoch": 0.5122727056924464,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.844824956162618e-05,
+      "loss": 0.223,
+      "step": 10790
+    },
+    {
+      "epoch": 0.5125100887812752,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.8430601454656957e-05,
+      "loss": 0.2232,
+      "step": 10795
+    },
+    {
+      "epoch": 0.5127474718701039,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.8412952774162292e-05,
+      "loss": 0.223,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5129848549589328,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.8395303531018985e-05,
+      "loss": 0.2227,
+      "step": 10805
+    },
+    {
+      "epoch": 0.5132222380477615,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.8377653736104177e-05,
+      "loss": 0.2212,
+      "step": 10810
+    },
+    {
+      "epoch": 0.5134596211365903,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.836000340029537e-05,
+      "loss": 0.2273,
+      "step": 10815
+    },
+    {
+      "epoch": 0.513697004225419,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.8342352534470373e-05,
+      "loss": 0.2229,
+      "step": 10820
+    },
+    {
+      "epoch": 0.5139343873142478,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.8324701149507342e-05,
+      "loss": 0.2218,
+      "step": 10825
+    },
+    {
+      "epoch": 0.5141717704030765,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.8307049256284746e-05,
+      "loss": 0.2239,
+      "step": 10830
+    },
+    {
+      "epoch": 0.5144091534919052,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.8289396865681356e-05,
+      "loss": 0.2246,
+      "step": 10835
+    },
+    {
+      "epoch": 0.514646536580734,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.8271743988576273e-05,
+      "loss": 0.2226,
+      "step": 10840
+    },
+    {
+      "epoch": 0.5148839196695627,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.8254090635848884e-05,
+      "loss": 0.226,
+      "step": 10845
+    },
+    {
+      "epoch": 0.5151213027583915,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.8236436818378857e-05,
+      "loss": 0.2257,
+      "step": 10850
+    },
+    {
+      "epoch": 0.5153586858472202,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.821878254704617e-05,
+      "loss": 0.2244,
+      "step": 10855
+    },
+    {
+      "epoch": 0.515596068936049,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.820112783273106e-05,
+      "loss": 0.2248,
+      "step": 10860
+    },
+    {
+      "epoch": 0.5158334520248777,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.8183472686314065e-05,
+      "loss": 0.2211,
+      "step": 10865
+    },
+    {
+      "epoch": 0.5160708351137065,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.8165817118675957e-05,
+      "loss": 0.2251,
+      "step": 10870
+    },
+    {
+      "epoch": 0.5163082182025353,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.8148161140697793e-05,
+      "loss": 0.2249,
+      "step": 10875
+    },
+    {
+      "epoch": 0.516545601291364,
+      "grad_norm": 0.5,
+      "learning_rate": 2.8130504763260856e-05,
+      "loss": 0.2307,
+      "step": 10880
+    },
+    {
+      "epoch": 0.5167829843801928,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.8112847997246707e-05,
+      "loss": 0.2208,
+      "step": 10885
+    },
+    {
+      "epoch": 0.5170203674690215,
+      "grad_norm": 0.5,
+      "learning_rate": 2.8095190853537118e-05,
+      "loss": 0.2263,
+      "step": 10890
+    },
+    {
+      "epoch": 0.5172577505578503,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.8077533343014124e-05,
+      "loss": 0.2222,
+      "step": 10895
+    },
+    {
+      "epoch": 0.517495133646679,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.8059875476559944e-05,
+      "loss": 0.2228,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5177325167355078,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.8042217265057064e-05,
+      "loss": 0.2268,
+      "step": 10905
+    },
+    {
+      "epoch": 0.5179698998243365,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.8024558719388143e-05,
+      "loss": 0.2217,
+      "step": 10910
+    },
+    {
+      "epoch": 0.5182072829131653,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.800689985043607e-05,
+      "loss": 0.2285,
+      "step": 10915
+    },
+    {
+      "epoch": 0.518444666001994,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.798924066908392e-05,
+      "loss": 0.2236,
+      "step": 10920
+    },
+    {
+      "epoch": 0.5186820490908227,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.797158118621498e-05,
+      "loss": 0.2281,
+      "step": 10925
+    },
+    {
+      "epoch": 0.5189194321796515,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.7953921412712685e-05,
+      "loss": 0.2253,
+      "step": 10930
+    },
+    {
+      "epoch": 0.5191568152684802,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.7936261359460697e-05,
+      "loss": 0.2251,
+      "step": 10935
+    },
+    {
+      "epoch": 0.519394198357309,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.7918601037342816e-05,
+      "loss": 0.2237,
+      "step": 10940
+    },
+    {
+      "epoch": 0.5196315814461377,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.790094045724302e-05,
+      "loss": 0.222,
+      "step": 10945
+    },
+    {
+      "epoch": 0.5198689645349666,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.7883279630045445e-05,
+      "loss": 0.2244,
+      "step": 10950
+    },
+    {
+      "epoch": 0.5201063476237953,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.786561856663437e-05,
+      "loss": 0.2208,
+      "step": 10955
+    },
+    {
+      "epoch": 0.5203437307126241,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.7847957277894245e-05,
+      "loss": 0.2244,
+      "step": 10960
+    },
+    {
+      "epoch": 0.5205811138014528,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.7830295774709625e-05,
+      "loss": 0.2195,
+      "step": 10965
+    },
+    {
+      "epoch": 0.5208184968902816,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.7812634067965227e-05,
+      "loss": 0.2246,
+      "step": 10970
+    },
+    {
+      "epoch": 0.5210558799791103,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.779497216854587e-05,
+      "loss": 0.2257,
+      "step": 10975
+    },
+    {
+      "epoch": 0.521293263067939,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.7777310087336507e-05,
+      "loss": 0.2247,
+      "step": 10980
+    },
+    {
+      "epoch": 0.5215306461567678,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.7759647835222197e-05,
+      "loss": 0.2257,
+      "step": 10985
+    },
+    {
+      "epoch": 0.5217680292455965,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.7741985423088114e-05,
+      "loss": 0.2199,
+      "step": 10990
+    },
+    {
+      "epoch": 0.5220054123344253,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.7724322861819508e-05,
+      "loss": 0.2217,
+      "step": 10995
+    },
+    {
+      "epoch": 0.522242795423254,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.770666016230174e-05,
+      "loss": 0.2261,
+      "step": 11000
+    },
+    {
+      "epoch": 0.5224801785120828,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.7688997335420253e-05,
+      "loss": 0.2238,
+      "step": 11005
+    },
+    {
+      "epoch": 0.5227175616009115,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.7671334392060577e-05,
+      "loss": 0.227,
+      "step": 11010
+    },
+    {
+      "epoch": 0.5229549446897404,
+      "grad_norm": 0.5,
+      "learning_rate": 2.7653671343108283e-05,
+      "loss": 0.2204,
+      "step": 11015
+    },
+    {
+      "epoch": 0.523192327778569,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.763600819944905e-05,
+      "loss": 0.2214,
+      "step": 11020
+    },
+    {
+      "epoch": 0.5234297108673978,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.7618344971968575e-05,
+      "loss": 0.2248,
+      "step": 11025
+    },
+    {
+      "epoch": 0.5236670939562266,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.7600681671552636e-05,
+      "loss": 0.2229,
+      "step": 11030
+    },
+    {
+      "epoch": 0.5239044770450553,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.7583018309087038e-05,
+      "loss": 0.2243,
+      "step": 11035
+    },
+    {
+      "epoch": 0.5241418601338841,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.756535489545764e-05,
+      "loss": 0.2261,
+      "step": 11040
+    },
+    {
+      "epoch": 0.5243792432227128,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.7547691441550306e-05,
+      "loss": 0.2226,
+      "step": 11045
+    },
+    {
+      "epoch": 0.5246166263115416,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.7530027958250965e-05,
+      "loss": 0.2217,
+      "step": 11050
+    },
+    {
+      "epoch": 0.5248540094003703,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.7512364456445516e-05,
+      "loss": 0.2275,
+      "step": 11055
+    },
+    {
+      "epoch": 0.5250913924891991,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.7494700947019924e-05,
+      "loss": 0.2218,
+      "step": 11060
+    },
+    {
+      "epoch": 0.5253287755780278,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.7477037440860103e-05,
+      "loss": 0.2252,
+      "step": 11065
+    },
+    {
+      "epoch": 0.5255661586668565,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.7459373948852006e-05,
+      "loss": 0.2243,
+      "step": 11070
+    },
+    {
+      "epoch": 0.5258035417556853,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.7441710481881566e-05,
+      "loss": 0.2227,
+      "step": 11075
+    },
+    {
+      "epoch": 0.526040924844514,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.7424047050834678e-05,
+      "loss": 0.2256,
+      "step": 11080
+    },
+    {
+      "epoch": 0.5262783079333428,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.7406383666597256e-05,
+      "loss": 0.2264,
+      "step": 11085
+    },
+    {
+      "epoch": 0.5265156910221716,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.7388720340055152e-05,
+      "loss": 0.2223,
+      "step": 11090
+    },
+    {
+      "epoch": 0.5267530741110004,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.7371057082094198e-05,
+      "loss": 0.2231,
+      "step": 11095
+    },
+    {
+      "epoch": 0.5269904571998291,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.7353393903600178e-05,
+      "loss": 0.2292,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5272278402886579,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.733573081545883e-05,
+      "loss": 0.2251,
+      "step": 11105
+    },
+    {
+      "epoch": 0.5274652233774866,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.7318067828555826e-05,
+      "loss": 0.2243,
+      "step": 11110
+    },
+    {
+      "epoch": 0.5277026064663154,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.7300404953776803e-05,
+      "loss": 0.2271,
+      "step": 11115
+    },
+    {
+      "epoch": 0.5279399895551441,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.7282742202007296e-05,
+      "loss": 0.2209,
+      "step": 11120
+    },
+    {
+      "epoch": 0.5281773726439728,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.7265079584132775e-05,
+      "loss": 0.2212,
+      "step": 11125
+    },
+    {
+      "epoch": 0.5284147557328016,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.7247417111038643e-05,
+      "loss": 0.2225,
+      "step": 11130
+    },
+    {
+      "epoch": 0.5286521388216303,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.7229754793610195e-05,
+      "loss": 0.227,
+      "step": 11135
+    },
+    {
+      "epoch": 0.5288895219104591,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.7212092642732644e-05,
+      "loss": 0.2253,
+      "step": 11140
+    },
+    {
+      "epoch": 0.5291269049992878,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.7194430669291083e-05,
+      "loss": 0.2231,
+      "step": 11145
+    },
+    {
+      "epoch": 0.5293642880881166,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.717676888417051e-05,
+      "loss": 0.2249,
+      "step": 11150
+    },
+    {
+      "epoch": 0.5296016711769453,
+      "grad_norm": 0.5,
+      "learning_rate": 2.715910729825581e-05,
+      "loss": 0.2277,
+      "step": 11155
+    },
+    {
+      "epoch": 0.5298390542657742,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.7141445922431724e-05,
+      "loss": 0.2276,
+      "step": 11160
+    },
+    {
+      "epoch": 0.5300764373546029,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.71237847675829e-05,
+      "loss": 0.2228,
+      "step": 11165
+    },
+    {
+      "epoch": 0.5303138204434316,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.7106123844593795e-05,
+      "loss": 0.2198,
+      "step": 11170
+    },
+    {
+      "epoch": 0.5305512035322604,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.708846316434879e-05,
+      "loss": 0.2239,
+      "step": 11175
+    },
+    {
+      "epoch": 0.5307885866210891,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.7070802737732053e-05,
+      "loss": 0.2256,
+      "step": 11180
+    },
+    {
+      "epoch": 0.5310259697099179,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.7053142575627638e-05,
+      "loss": 0.2271,
+      "step": 11185
+    },
+    {
+      "epoch": 0.5312633527987466,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.703548268891942e-05,
+      "loss": 0.223,
+      "step": 11190
+    },
+    {
+      "epoch": 0.5315007358875754,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.7017823088491118e-05,
+      "loss": 0.2235,
+      "step": 11195
+    },
+    {
+      "epoch": 0.5317381189764041,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.7000163785226246e-05,
+      "loss": 0.2244,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5319755020652329,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.6982504790008167e-05,
+      "loss": 0.226,
+      "step": 11205
+    },
+    {
+      "epoch": 0.5322128851540616,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.6964846113720027e-05,
+      "loss": 0.2221,
+      "step": 11210
+    },
+    {
+      "epoch": 0.5324502682428903,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.6947187767244798e-05,
+      "loss": 0.2244,
+      "step": 11215
+    },
+    {
+      "epoch": 0.5326876513317191,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.6929529761465235e-05,
+      "loss": 0.2262,
+      "step": 11220
+    },
+    {
+      "epoch": 0.5329250344205478,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.6911872107263874e-05,
+      "loss": 0.2243,
+      "step": 11225
+    },
+    {
+      "epoch": 0.5331624175093767,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.6894214815523073e-05,
+      "loss": 0.2255,
+      "step": 11230
+    },
+    {
+      "epoch": 0.5333998005982054,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.6876557897124916e-05,
+      "loss": 0.2271,
+      "step": 11235
+    },
+    {
+      "epoch": 0.5336371836870342,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.68589013629513e-05,
+      "loss": 0.2245,
+      "step": 11240
+    },
+    {
+      "epoch": 0.5338745667758629,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.6841245223883855e-05,
+      "loss": 0.2226,
+      "step": 11245
+    },
+    {
+      "epoch": 0.5341119498646917,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.6823589490803987e-05,
+      "loss": 0.2253,
+      "step": 11250
+    },
+    {
+      "epoch": 0.5343493329535204,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.680593417459283e-05,
+      "loss": 0.2267,
+      "step": 11255
+    },
+    {
+      "epoch": 0.5345867160423492,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.678827928613129e-05,
+      "loss": 0.2278,
+      "step": 11260
+    },
+    {
+      "epoch": 0.5348240991311779,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.6770624836300003e-05,
+      "loss": 0.2228,
+      "step": 11265
+    },
+    {
+      "epoch": 0.5350614822200066,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.675297083597929e-05,
+      "loss": 0.2217,
+      "step": 11270
+    },
+    {
+      "epoch": 0.5352988653088354,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.673531729604927e-05,
+      "loss": 0.2235,
+      "step": 11275
+    },
+    {
+      "epoch": 0.5355362483976641,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.6717664227389716e-05,
+      "loss": 0.2226,
+      "step": 11280
+    },
+    {
+      "epoch": 0.5357736314864929,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.6700011640880146e-05,
+      "loss": 0.2279,
+      "step": 11285
+    },
+    {
+      "epoch": 0.5360110145753216,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.6682359547399753e-05,
+      "loss": 0.2246,
+      "step": 11290
+    },
+    {
+      "epoch": 0.5362483976641504,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.6664707957827467e-05,
+      "loss": 0.2238,
+      "step": 11295
+    },
+    {
+      "epoch": 0.5364857807529791,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.6647056883041853e-05,
+      "loss": 0.2291,
+      "step": 11300
+    },
+    {
+      "epoch": 0.536723163841808,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.662940633392122e-05,
+      "loss": 0.2265,
+      "step": 11305
+    },
+    {
+      "epoch": 0.5369605469306367,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.66117563213435e-05,
+      "loss": 0.2214,
+      "step": 11310
+    },
+    {
+      "epoch": 0.5371979300194654,
+      "grad_norm": 0.625,
+      "learning_rate": 2.6594106856186317e-05,
+      "loss": 0.2252,
+      "step": 11315
+    },
+    {
+      "epoch": 0.5374353131082942,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.657645794932696e-05,
+      "loss": 0.2233,
+      "step": 11320
+    },
+    {
+      "epoch": 0.5376726961971229,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.6558809611642364e-05,
+      "loss": 0.2271,
+      "step": 11325
+    },
+    {
+      "epoch": 0.5379100792859517,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.6541161854009134e-05,
+      "loss": 0.2229,
+      "step": 11330
+    },
+    {
+      "epoch": 0.5381474623747804,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.6523514687303485e-05,
+      "loss": 0.2257,
+      "step": 11335
+    },
+    {
+      "epoch": 0.5383848454636092,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.6505868122401295e-05,
+      "loss": 0.2274,
+      "step": 11340
+    },
+    {
+      "epoch": 0.5386222285524379,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.6488222170178062e-05,
+      "loss": 0.2194,
+      "step": 11345
+    },
+    {
+      "epoch": 0.5388596116412667,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.6470576841508905e-05,
+      "loss": 0.2299,
+      "step": 11350
+    },
+    {
+      "epoch": 0.5390969947300954,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.645293214726856e-05,
+      "loss": 0.2234,
+      "step": 11355
+    },
+    {
+      "epoch": 0.5393343778189241,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.6435288098331372e-05,
+      "loss": 0.2228,
+      "step": 11360
+    },
+    {
+      "epoch": 0.5395717609077529,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.6417644705571286e-05,
+      "loss": 0.2205,
+      "step": 11365
+    },
+    {
+      "epoch": 0.5398091439965816,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.6400001979861843e-05,
+      "loss": 0.2261,
+      "step": 11370
+    },
+    {
+      "epoch": 0.5400465270854105,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.6382359932076185e-05,
+      "loss": 0.2278,
+      "step": 11375
+    },
+    {
+      "epoch": 0.5402839101742392,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.636471857308701e-05,
+      "loss": 0.224,
+      "step": 11380
+    },
+    {
+      "epoch": 0.540521293263068,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.6347077913766622e-05,
+      "loss": 0.2273,
+      "step": 11385
+    },
+    {
+      "epoch": 0.5407586763518967,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.632943796498687e-05,
+      "loss": 0.2232,
+      "step": 11390
+    },
+    {
+      "epoch": 0.5409960594407255,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.631179873761918e-05,
+      "loss": 0.2233,
+      "step": 11395
+    },
+    {
+      "epoch": 0.5412334425295542,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.6294160242534522e-05,
+      "loss": 0.2255,
+      "step": 11400
+    },
+    {
+      "epoch": 0.541470825618383,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.627652249060343e-05,
+      "loss": 0.224,
+      "step": 11405
+    },
+    {
+      "epoch": 0.5417082087072117,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.625888549269596e-05,
+      "loss": 0.2237,
+      "step": 11410
+    },
+    {
+      "epoch": 0.5419455917960404,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.624124925968171e-05,
+      "loss": 0.2257,
+      "step": 11415
+    },
+    {
+      "epoch": 0.5421829748848692,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.6223613802429832e-05,
+      "loss": 0.2239,
+      "step": 11420
+    },
+    {
+      "epoch": 0.5424203579736979,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.6205979131808944e-05,
+      "loss": 0.2224,
+      "step": 11425
+    },
+    {
+      "epoch": 0.5426577410625267,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.618834525868725e-05,
+      "loss": 0.2232,
+      "step": 11430
+    },
+    {
+      "epoch": 0.5428951241513554,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.6170712193932394e-05,
+      "loss": 0.2195,
+      "step": 11435
+    },
+    {
+      "epoch": 0.5431325072401842,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.6153079948411584e-05,
+      "loss": 0.2186,
+      "step": 11440
+    },
+    {
+      "epoch": 0.543369890329013,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.6135448532991462e-05,
+      "loss": 0.2258,
+      "step": 11445
+    },
+    {
+      "epoch": 0.5436072734178418,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.6117817958538217e-05,
+      "loss": 0.226,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5438446565066705,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.610018823591746e-05,
+      "loss": 0.2256,
+      "step": 11455
+    },
+    {
+      "epoch": 0.5440820395954992,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.6082559375994352e-05,
+      "loss": 0.2225,
+      "step": 11460
+    },
+    {
+      "epoch": 0.544319422684328,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6064931389633445e-05,
+      "loss": 0.2265,
+      "step": 11465
+    },
+    {
+      "epoch": 0.5445568057731567,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.6047304287698794e-05,
+      "loss": 0.2248,
+      "step": 11470
+    },
+    {
+      "epoch": 0.5447941888619855,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.6029678081053903e-05,
+      "loss": 0.2227,
+      "step": 11475
+    },
+    {
+      "epoch": 0.5450315719508142,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.6012052780561728e-05,
+      "loss": 0.2237,
+      "step": 11480
+    },
+    {
+      "epoch": 0.545268955039643,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.5994428397084658e-05,
+      "loss": 0.2266,
+      "step": 11485
+    },
+    {
+      "epoch": 0.5455063381284717,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.5976804941484516e-05,
+      "loss": 0.2272,
+      "step": 11490
+    },
+    {
+      "epoch": 0.5457437212173005,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.5959182424622568e-05,
+      "loss": 0.225,
+      "step": 11495
+    },
+    {
+      "epoch": 0.5459811043061292,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.5941560857359488e-05,
+      "loss": 0.2248,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5462184873949579,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.592394025055536e-05,
+      "loss": 0.2245,
+      "step": 11505
+    },
+    {
+      "epoch": 0.5464558704837867,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.590632061506969e-05,
+      "loss": 0.2217,
+      "step": 11510
+    },
+    {
+      "epoch": 0.5466932535726154,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.5888701961761376e-05,
+      "loss": 0.2229,
+      "step": 11515
+    },
+    {
+      "epoch": 0.5469306366614443,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.587108430148872e-05,
+      "loss": 0.222,
+      "step": 11520
+    },
+    {
+      "epoch": 0.547168019750273,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.5853467645109392e-05,
+      "loss": 0.2294,
+      "step": 11525
+    },
+    {
+      "epoch": 0.5474054028391018,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.5835852003480475e-05,
+      "loss": 0.2242,
+      "step": 11530
+    },
+    {
+      "epoch": 0.5476427859279305,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.58182373874584e-05,
+      "loss": 0.2264,
+      "step": 11535
+    },
+    {
+      "epoch": 0.5478801690167593,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.5800623807898976e-05,
+      "loss": 0.2222,
+      "step": 11540
+    },
+    {
+      "epoch": 0.548117552105588,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.5783011275657366e-05,
+      "loss": 0.2197,
+      "step": 11545
+    },
+    {
+      "epoch": 0.5483549351944168,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.5765399801588108e-05,
+      "loss": 0.2267,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5485923182832455,
+      "grad_norm": 0.408203125,
+      "learning_rate": 2.574778939654507e-05,
+      "loss": 0.22,
+      "step": 11555
+    },
+    {
+      "epoch": 0.5488297013720742,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.5730180071381437e-05,
+      "loss": 0.2242,
+      "step": 11560
+    },
+    {
+      "epoch": 0.549067084460903,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.5712571836949802e-05,
+      "loss": 0.2232,
+      "step": 11565
+    },
+    {
+      "epoch": 0.5493044675497317,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.5694964704101998e-05,
+      "loss": 0.2232,
+      "step": 11570
+    },
+    {
+      "epoch": 0.5495418506385605,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.567735868368925e-05,
+      "loss": 0.225,
+      "step": 11575
+    },
+    {
+      "epoch": 0.5497792337273892,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.565975378656204e-05,
+      "loss": 0.2249,
+      "step": 11580
+    },
+    {
+      "epoch": 0.550016616816218,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.5642150023570226e-05,
+      "loss": 0.2238,
+      "step": 11585
+    },
+    {
+      "epoch": 0.5502539999050468,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.5624547405562892e-05,
+      "loss": 0.2237,
+      "step": 11590
+    },
+    {
+      "epoch": 0.5504913829938756,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.5606945943388468e-05,
+      "loss": 0.2258,
+      "step": 11595
+    },
+    {
+      "epoch": 0.5507287660827043,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.5589345647894643e-05,
+      "loss": 0.2217,
+      "step": 11600
+    },
+    {
+      "epoch": 0.550966149171533,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.5571746529928413e-05,
+      "loss": 0.2199,
+      "step": 11605
+    },
+    {
+      "epoch": 0.5512035322603618,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.5554148600336032e-05,
+      "loss": 0.2254,
+      "step": 11610
+    },
+    {
+      "epoch": 0.5514409153491905,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.5536551869963015e-05,
+      "loss": 0.2245,
+      "step": 11615
+    },
+    {
+      "epoch": 0.5516782984380193,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.551895634965415e-05,
+      "loss": 0.2252,
+      "step": 11620
+    },
+    {
+      "epoch": 0.551915681526848,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.550136205025348e-05,
+      "loss": 0.2262,
+      "step": 11625
+    },
+    {
+      "epoch": 0.5521530646156768,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.5483768982604288e-05,
+      "loss": 0.2296,
+      "step": 11630
+    },
+    {
+      "epoch": 0.5523904477045055,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.5466177157549108e-05,
+      "loss": 0.2245,
+      "step": 11635
+    },
+    {
+      "epoch": 0.5526278307933343,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.5448586585929695e-05,
+      "loss": 0.2258,
+      "step": 11640
+    },
+    {
+      "epoch": 0.552865213882163,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.5430997278587042e-05,
+      "loss": 0.2261,
+      "step": 11645
+    },
+    {
+      "epoch": 0.5531025969709917,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.541340924636136e-05,
+      "loss": 0.2249,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5533399800598205,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.539582250009207e-05,
+      "loss": 0.2263,
+      "step": 11655
+    },
+    {
+      "epoch": 0.5535773631486492,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.53782370506178e-05,
+      "loss": 0.2247,
+      "step": 11660
+    },
+    {
+      "epoch": 0.5538147462374781,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.53606529087764e-05,
+      "loss": 0.2237,
+      "step": 11665
+    },
+    {
+      "epoch": 0.5540521293263068,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.5343070085404864e-05,
+      "loss": 0.2207,
+      "step": 11670
+    },
+    {
+      "epoch": 0.5542895124151356,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.532548859133944e-05,
+      "loss": 0.2232,
+      "step": 11675
+    },
+    {
+      "epoch": 0.5545268955039643,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.5307908437415502e-05,
+      "loss": 0.2257,
+      "step": 11680
+    },
+    {
+      "epoch": 0.5547642785927931,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.5290329634467634e-05,
+      "loss": 0.2231,
+      "step": 11685
+    },
+    {
+      "epoch": 0.5550016616816218,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.5272752193329553e-05,
+      "loss": 0.2261,
+      "step": 11690
+    },
+    {
+      "epoch": 0.5552390447704506,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.525517612483418e-05,
+      "loss": 0.2213,
+      "step": 11695
+    },
+    {
+      "epoch": 0.5554764278592793,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.523760143981354e-05,
+      "loss": 0.2247,
+      "step": 11700
+    },
+    {
+      "epoch": 0.555713810948108,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.5220028149098862e-05,
+      "loss": 0.2244,
+      "step": 11705
+    },
+    {
+      "epoch": 0.5559511940369368,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.520245626352046e-05,
+      "loss": 0.2207,
+      "step": 11710
+    },
+    {
+      "epoch": 0.5561885771257655,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.518488579390781e-05,
+      "loss": 0.2221,
+      "step": 11715
+    },
+    {
+      "epoch": 0.5564259602145943,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.516731675108952e-05,
+      "loss": 0.2294,
+      "step": 11720
+    },
+    {
+      "epoch": 0.556663343303423,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.514974914589331e-05,
+      "loss": 0.2204,
+      "step": 11725
+    },
+    {
+      "epoch": 0.5569007263922519,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.5132182989146015e-05,
+      "loss": 0.2252,
+      "step": 11730
+    },
+    {
+      "epoch": 0.5571381094810806,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.511461829167357e-05,
+      "loss": 0.2202,
+      "step": 11735
+    },
+    {
+      "epoch": 0.5573754925699094,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.5097055064301033e-05,
+      "loss": 0.2316,
+      "step": 11740
+    },
+    {
+      "epoch": 0.5576128756587381,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.5079493317852526e-05,
+      "loss": 0.2229,
+      "step": 11745
+    },
+    {
+      "epoch": 0.5578502587475668,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.506193306315129e-05,
+      "loss": 0.2219,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5580876418363956,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.5044374311019615e-05,
+      "loss": 0.2227,
+      "step": 11755
+    },
+    {
+      "epoch": 0.5583250249252243,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.502681707227888e-05,
+      "loss": 0.228,
+      "step": 11760
+    },
+    {
+      "epoch": 0.5585624080140531,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.5009261357749547e-05,
+      "loss": 0.2256,
+      "step": 11765
+    },
+    {
+      "epoch": 0.5587997911028818,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.4991707178251107e-05,
+      "loss": 0.2234,
+      "step": 11770
+    },
+    {
+      "epoch": 0.5590371741917106,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.497415454460213e-05,
+      "loss": 0.2231,
+      "step": 11775
+    },
+    {
+      "epoch": 0.5592745572805393,
+      "grad_norm": 0.5,
+      "learning_rate": 2.4956603467620214e-05,
+      "loss": 0.2222,
+      "step": 11780
+    },
+    {
+      "epoch": 0.5595119403693681,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.493905395812202e-05,
+      "loss": 0.2272,
+      "step": 11785
+    },
+    {
+      "epoch": 0.5597493234581968,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.4921506026923232e-05,
+      "loss": 0.2209,
+      "step": 11790
+    },
+    {
+      "epoch": 0.5599867065470255,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.4903959684838547e-05,
+      "loss": 0.222,
+      "step": 11795
+    },
+    {
+      "epoch": 0.5602240896358543,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.488641494268171e-05,
+      "loss": 0.224,
+      "step": 11800
+    },
+    {
+      "epoch": 0.560461472724683,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.486887181126545e-05,
+      "loss": 0.2275,
+      "step": 11805
+    },
+    {
+      "epoch": 0.5606988558135119,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.4851330301401542e-05,
+      "loss": 0.2256,
+      "step": 11810
+    },
+    {
+      "epoch": 0.5609362389023406,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.483379042390071e-05,
+      "loss": 0.2236,
+      "step": 11815
+    },
+    {
+      "epoch": 0.5611736219911694,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.4816252189572726e-05,
+      "loss": 0.2249,
+      "step": 11820
+    },
+    {
+      "epoch": 0.5614110050799981,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.4798715609226297e-05,
+      "loss": 0.2235,
+      "step": 11825
+    },
+    {
+      "epoch": 0.5616483881688269,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.4781180693669177e-05,
+      "loss": 0.226,
+      "step": 11830
+    },
+    {
+      "epoch": 0.5618857712576556,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.4763647453708018e-05,
+      "loss": 0.2235,
+      "step": 11835
+    },
+    {
+      "epoch": 0.5621231543464844,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.4746115900148497e-05,
+      "loss": 0.2269,
+      "step": 11840
+    },
+    {
+      "epoch": 0.5623605374353131,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.4728586043795216e-05,
+      "loss": 0.2237,
+      "step": 11845
+    },
+    {
+      "epoch": 0.5625979205241418,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.471105789545176e-05,
+      "loss": 0.2245,
+      "step": 11850
+    },
+    {
+      "epoch": 0.5628353036129706,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.4693531465920645e-05,
+      "loss": 0.2203,
+      "step": 11855
+    },
+    {
+      "epoch": 0.5630726867017993,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.4676006766003318e-05,
+      "loss": 0.2258,
+      "step": 11860
+    },
+    {
+      "epoch": 0.5633100697906281,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.4658483806500183e-05,
+      "loss": 0.2223,
+      "step": 11865
+    },
+    {
+      "epoch": 0.5635474528794568,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.4640962598210554e-05,
+      "loss": 0.2273,
+      "step": 11870
+    },
+    {
+      "epoch": 0.5637848359682857,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.4623443151932672e-05,
+      "loss": 0.2212,
+      "step": 11875
+    },
+    {
+      "epoch": 0.5640222190571144,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.4605925478463697e-05,
+      "loss": 0.2253,
+      "step": 11880
+    },
+    {
+      "epoch": 0.5642596021459432,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.4588409588599696e-05,
+      "loss": 0.2211,
+      "step": 11885
+    },
+    {
+      "epoch": 0.5644969852347719,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.457089549313562e-05,
+      "loss": 0.2225,
+      "step": 11890
+    },
+    {
+      "epoch": 0.5647343683236006,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.455338320286533e-05,
+      "loss": 0.2251,
+      "step": 11895
+    },
+    {
+      "epoch": 0.5649717514124294,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.4535872728581583e-05,
+      "loss": 0.2235,
+      "step": 11900
+    },
+    {
+      "epoch": 0.5652091345012581,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.4518364081075985e-05,
+      "loss": 0.2235,
+      "step": 11905
+    },
+    {
+      "epoch": 0.5654465175900869,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.4500857271139048e-05,
+      "loss": 0.2259,
+      "step": 11910
+    },
+    {
+      "epoch": 0.5656839006789156,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.4483352309560132e-05,
+      "loss": 0.2231,
+      "step": 11915
+    },
+    {
+      "epoch": 0.5659212837677444,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.4465849207127477e-05,
+      "loss": 0.2226,
+      "step": 11920
+    },
+    {
+      "epoch": 0.5661586668565731,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.4448347974628154e-05,
+      "loss": 0.2273,
+      "step": 11925
+    },
+    {
+      "epoch": 0.5663960499454019,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.44308486228481e-05,
+      "loss": 0.2235,
+      "step": 11930
+    },
+    {
+      "epoch": 0.5666334330342306,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.4413351162572086e-05,
+      "loss": 0.2205,
+      "step": 11935
+    },
+    {
+      "epoch": 0.5668708161230593,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.4395855604583713e-05,
+      "loss": 0.2229,
+      "step": 11940
+    },
+    {
+      "epoch": 0.5671081992118882,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.437836195966542e-05,
+      "loss": 0.2247,
+      "step": 11945
+    },
+    {
+      "epoch": 0.5673455823007169,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.4360870238598438e-05,
+      "loss": 0.2254,
+      "step": 11950
+    },
+    {
+      "epoch": 0.5675829653895457,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.4343380452162874e-05,
+      "loss": 0.2246,
+      "step": 11955
+    },
+    {
+      "epoch": 0.5678203484783744,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.4325892611137563e-05,
+      "loss": 0.2244,
+      "step": 11960
+    },
+    {
+      "epoch": 0.5680577315672032,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.4308406726300224e-05,
+      "loss": 0.2256,
+      "step": 11965
+    },
+    {
+      "epoch": 0.5682951146560319,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.4290922808427284e-05,
+      "loss": 0.225,
+      "step": 11970
+    },
+    {
+      "epoch": 0.5685324977448607,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.4273440868294047e-05,
+      "loss": 0.2218,
+      "step": 11975
+    },
+    {
+      "epoch": 0.5687698808336894,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.4255960916674512e-05,
+      "loss": 0.2252,
+      "step": 11980
+    },
+    {
+      "epoch": 0.5690072639225182,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.4238482964341524e-05,
+      "loss": 0.2229,
+      "step": 11985
+    },
+    {
+      "epoch": 0.5692446470113469,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.4221007022066648e-05,
+      "loss": 0.2229,
+      "step": 11990
+    },
+    {
+      "epoch": 0.5694820301001756,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.420353310062024e-05,
+      "loss": 0.2279,
+      "step": 11995
+    },
+    {
+      "epoch": 0.5697194131890044,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.4186061210771388e-05,
+      "loss": 0.2201,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5699567962778331,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.416859136328794e-05,
+      "loss": 0.2251,
+      "step": 12005
+    },
+    {
+      "epoch": 0.570194179366662,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.415112356893649e-05,
+      "loss": 0.2252,
+      "step": 12010
+    },
+    {
+      "epoch": 0.5704315624554906,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.4133657838482348e-05,
+      "loss": 0.2227,
+      "step": 12015
+    },
+    {
+      "epoch": 0.5706689455443195,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.4116194182689582e-05,
+      "loss": 0.2224,
+      "step": 12020
+    },
+    {
+      "epoch": 0.5709063286331482,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.4098732612320945e-05,
+      "loss": 0.2256,
+      "step": 12025
+    },
+    {
+      "epoch": 0.571143711721977,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.4081273138137943e-05,
+      "loss": 0.2253,
+      "step": 12030
+    },
+    {
+      "epoch": 0.5713810948108057,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.406381577090075e-05,
+      "loss": 0.2248,
+      "step": 12035
+    },
+    {
+      "epoch": 0.5716184778996344,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.404636052136828e-05,
+      "loss": 0.2262,
+      "step": 12040
+    },
+    {
+      "epoch": 0.5718558609884632,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.4028907400298113e-05,
+      "loss": 0.2226,
+      "step": 12045
+    },
+    {
+      "epoch": 0.5720932440772919,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.4011456418446526e-05,
+      "loss": 0.2277,
+      "step": 12050
+    },
+    {
+      "epoch": 0.5723306271661207,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.3994007586568497e-05,
+      "loss": 0.2231,
+      "step": 12055
+    },
+    {
+      "epoch": 0.5725680102549494,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.3976560915417635e-05,
+      "loss": 0.2227,
+      "step": 12060
+    },
+    {
+      "epoch": 0.5728053933437782,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.395911641574627e-05,
+      "loss": 0.2264,
+      "step": 12065
+    },
+    {
+      "epoch": 0.5730427764326069,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.394167409830535e-05,
+      "loss": 0.2237,
+      "step": 12070
+    },
+    {
+      "epoch": 0.5732801595214357,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.3924233973844507e-05,
+      "loss": 0.2289,
+      "step": 12075
+    },
+    {
+      "epoch": 0.5735175426102644,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.3906796053111997e-05,
+      "loss": 0.2226,
+      "step": 12080
+    },
+    {
+      "epoch": 0.5737549256990931,
+      "grad_norm": 0.5,
+      "learning_rate": 2.388936034685475e-05,
+      "loss": 0.2238,
+      "step": 12085
+    },
+    {
+      "epoch": 0.573992308787922,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.387192686581829e-05,
+      "loss": 0.2206,
+      "step": 12090
+    },
+    {
+      "epoch": 0.5742296918767507,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.3854495620746825e-05,
+      "loss": 0.2217,
+      "step": 12095
+    },
+    {
+      "epoch": 0.5744670749655795,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.383706662238312e-05,
+      "loss": 0.2224,
+      "step": 12100
+    },
+    {
+      "epoch": 0.5747044580544082,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.3819639881468592e-05,
+      "loss": 0.2233,
+      "step": 12105
+    },
+    {
+      "epoch": 0.574941841143237,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.3802215408743282e-05,
+      "loss": 0.2258,
+      "step": 12110
+    },
+    {
+      "epoch": 0.5751792242320657,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.378479321494579e-05,
+      "loss": 0.2279,
+      "step": 12115
+    },
+    {
+      "epoch": 0.5754166073208945,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.376737331081335e-05,
+      "loss": 0.2286,
+      "step": 12120
+    },
+    {
+      "epoch": 0.5756539904097232,
+      "grad_norm": 0.51171875,
+      "learning_rate": 2.3749955707081767e-05,
+      "loss": 0.2239,
+      "step": 12125
+    },
+    {
+      "epoch": 0.575891373498552,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.3732540414485428e-05,
+      "loss": 0.2243,
+      "step": 12130
+    },
+    {
+      "epoch": 0.5761287565873807,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.3715127443757295e-05,
+      "loss": 0.2225,
+      "step": 12135
+    },
+    {
+      "epoch": 0.5763661396762094,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.3697716805628912e-05,
+      "loss": 0.2217,
+      "step": 12140
+    },
+    {
+      "epoch": 0.5766035227650382,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.3680308510830374e-05,
+      "loss": 0.2229,
+      "step": 12145
+    },
+    {
+      "epoch": 0.5768409058538669,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.3662902570090323e-05,
+      "loss": 0.2222,
+      "step": 12150
+    },
+    {
+      "epoch": 0.5770782889426957,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.3645498994135977e-05,
+      "loss": 0.2269,
+      "step": 12155
+    },
+    {
+      "epoch": 0.5773156720315245,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.362809779369307e-05,
+      "loss": 0.2253,
+      "step": 12160
+    },
+    {
+      "epoch": 0.5775530551203533,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.361069897948589e-05,
+      "loss": 0.2254,
+      "step": 12165
+    },
+    {
+      "epoch": 0.577790438209182,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.3593302562237237e-05,
+      "loss": 0.2233,
+      "step": 12170
+    },
+    {
+      "epoch": 0.5780278212980108,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.3575908552668453e-05,
+      "loss": 0.2213,
+      "step": 12175
+    },
+    {
+      "epoch": 0.5782652043868395,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.3558516961499383e-05,
+      "loss": 0.2203,
+      "step": 12180
+    },
+    {
+      "epoch": 0.5785025874756682,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.3541127799448386e-05,
+      "loss": 0.2206,
+      "step": 12185
+    },
+    {
+      "epoch": 0.578739970564497,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.3523741077232337e-05,
+      "loss": 0.2261,
+      "step": 12190
+    },
+    {
+      "epoch": 0.5789773536533257,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.350635680556656e-05,
+      "loss": 0.2225,
+      "step": 12195
+    },
+    {
+      "epoch": 0.5792147367421545,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.3488974995164936e-05,
+      "loss": 0.2245,
+      "step": 12200
+    },
+    {
+      "epoch": 0.5794521198309832,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.3471595656739776e-05,
+      "loss": 0.2222,
+      "step": 12205
+    },
+    {
+      "epoch": 0.579689502919812,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.3454218801001903e-05,
+      "loss": 0.227,
+      "step": 12210
+    },
+    {
+      "epoch": 0.5799268860086407,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.343684443866057e-05,
+      "loss": 0.2223,
+      "step": 12215
+    },
+    {
+      "epoch": 0.5801642690974695,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.3419472580423552e-05,
+      "loss": 0.2247,
+      "step": 12220
+    },
+    {
+      "epoch": 0.5804016521862982,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.3402103236997012e-05,
+      "loss": 0.2229,
+      "step": 12225
+    },
+    {
+      "epoch": 0.580639035275127,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.338473641908563e-05,
+      "loss": 0.2237,
+      "step": 12230
+    },
+    {
+      "epoch": 0.5808764183639558,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.3367372137392457e-05,
+      "loss": 0.219,
+      "step": 12235
+    },
+    {
+      "epoch": 0.5811138014527845,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.335001040261906e-05,
+      "loss": 0.2273,
+      "step": 12240
+    },
+    {
+      "epoch": 0.5813511845416133,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.333265122546538e-05,
+      "loss": 0.2258,
+      "step": 12245
+    },
+    {
+      "epoch": 0.581588567630442,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.331529461662979e-05,
+      "loss": 0.2221,
+      "step": 12250
+    },
+    {
+      "epoch": 0.5818259507192708,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.3297940586809107e-05,
+      "loss": 0.225,
+      "step": 12255
+    },
+    {
+      "epoch": 0.5820633338080995,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.3280589146698522e-05,
+      "loss": 0.2265,
+      "step": 12260
+    },
+    {
+      "epoch": 0.5823007168969283,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.326324030699167e-05,
+      "loss": 0.2245,
+      "step": 12265
+    },
+    {
+      "epoch": 0.582538099985757,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.3245894078380547e-05,
+      "loss": 0.2234,
+      "step": 12270
+    },
+    {
+      "epoch": 0.5827754830745858,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.322855047155556e-05,
+      "loss": 0.2226,
+      "step": 12275
+    },
+    {
+      "epoch": 0.5830128661634145,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.3211209497205487e-05,
+      "loss": 0.2205,
+      "step": 12280
+    },
+    {
+      "epoch": 0.5832502492522432,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.3193871166017505e-05,
+      "loss": 0.2264,
+      "step": 12285
+    },
+    {
+      "epoch": 0.583487632341072,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.3176535488677143e-05,
+      "loss": 0.225,
+      "step": 12290
+    },
+    {
+      "epoch": 0.5837250154299007,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.3159202475868293e-05,
+      "loss": 0.2272,
+      "step": 12295
+    },
+    {
+      "epoch": 0.5839623985187296,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.3141872138273226e-05,
+      "loss": 0.2256,
+      "step": 12300
+    },
+    {
+      "epoch": 0.5841997816075583,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.312454448657253e-05,
+      "loss": 0.2263,
+      "step": 12305
+    },
+    {
+      "epoch": 0.5844371646963871,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.3107219531445185e-05,
+      "loss": 0.224,
+      "step": 12310
+    },
+    {
+      "epoch": 0.5846745477852158,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.3089897283568456e-05,
+      "loss": 0.2243,
+      "step": 12315
+    },
+    {
+      "epoch": 0.5849119308740446,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.3072577753617987e-05,
+      "loss": 0.2216,
+      "step": 12320
+    },
+    {
+      "epoch": 0.5851493139628733,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.3055260952267704e-05,
+      "loss": 0.2228,
+      "step": 12325
+    },
+    {
+      "epoch": 0.585386697051702,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.3037946890189894e-05,
+      "loss": 0.2275,
+      "step": 12330
+    },
+    {
+      "epoch": 0.5856240801405308,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.302063557805512e-05,
+      "loss": 0.2245,
+      "step": 12335
+    },
+    {
+      "epoch": 0.5858614632293595,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.3003327026532275e-05,
+      "loss": 0.2215,
+      "step": 12340
+    },
+    {
+      "epoch": 0.5860988463181883,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.2986021246288535e-05,
+      "loss": 0.224,
+      "step": 12345
+    },
+    {
+      "epoch": 0.586336229407017,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.2968718247989368e-05,
+      "loss": 0.2242,
+      "step": 12350
+    },
+    {
+      "epoch": 0.5865736124958458,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.295141804229855e-05,
+      "loss": 0.2206,
+      "step": 12355
+    },
+    {
+      "epoch": 0.5868109955846745,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.2934120639878093e-05,
+      "loss": 0.2226,
+      "step": 12360
+    },
+    {
+      "epoch": 0.5870483786735033,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.2916826051388342e-05,
+      "loss": 0.2252,
+      "step": 12365
+    },
+    {
+      "epoch": 0.587285761762332,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.2899534287487834e-05,
+      "loss": 0.2275,
+      "step": 12370
+    },
+    {
+      "epoch": 0.5875231448511608,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.2882245358833442e-05,
+      "loss": 0.2253,
+      "step": 12375
+    },
+    {
+      "epoch": 0.5877605279399896,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.2864959276080228e-05,
+      "loss": 0.2235,
+      "step": 12380
+    },
+    {
+      "epoch": 0.5879979110288183,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.2847676049881546e-05,
+      "loss": 0.2215,
+      "step": 12385
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.283039569088895e-05,
+      "loss": 0.2256,
+      "step": 12390
+    },
+    {
+      "epoch": 0.5884726772064758,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.2813118209752256e-05,
+      "loss": 0.2193,
+      "step": 12395
+    },
+    {
+      "epoch": 0.5887100602953046,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.2795843617119495e-05,
+      "loss": 0.2236,
+      "step": 12400
+    },
+    {
+      "epoch": 0.5889474433841333,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.277857192363692e-05,
+      "loss": 0.2247,
+      "step": 12405
+    },
+    {
+      "epoch": 0.5891848264729621,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.276130313994901e-05,
+      "loss": 0.2281,
+      "step": 12410
+    },
+    {
+      "epoch": 0.5894222095617908,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.274403727669841e-05,
+      "loss": 0.2284,
+      "step": 12415
+    },
+    {
+      "epoch": 0.5896595926506196,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.272677434452602e-05,
+      "loss": 0.2243,
+      "step": 12420
+    },
+    {
+      "epoch": 0.5898969757394483,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.2709514354070893e-05,
+      "loss": 0.224,
+      "step": 12425
+    },
+    {
+      "epoch": 0.590134358828277,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.2692257315970288e-05,
+      "loss": 0.2268,
+      "step": 12430
+    },
+    {
+      "epoch": 0.5903717419171058,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.2675003240859636e-05,
+      "loss": 0.2275,
+      "step": 12435
+    },
+    {
+      "epoch": 0.5906091250059345,
+      "grad_norm": 0.5,
+      "learning_rate": 2.2657752139372547e-05,
+      "loss": 0.2242,
+      "step": 12440
+    },
+    {
+      "epoch": 0.5908465080947634,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.2640504022140796e-05,
+      "loss": 0.2287,
+      "step": 12445
+    },
+    {
+      "epoch": 0.5910838911835921,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.2623258899794307e-05,
+      "loss": 0.2264,
+      "step": 12450
+    },
+    {
+      "epoch": 0.5913212742724209,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.2606016782961197e-05,
+      "loss": 0.219,
+      "step": 12455
+    },
+    {
+      "epoch": 0.5915586573612496,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.258877768226768e-05,
+      "loss": 0.2238,
+      "step": 12460
+    },
+    {
+      "epoch": 0.5917960404500784,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.257154160833815e-05,
+      "loss": 0.2244,
+      "step": 12465
+    },
+    {
+      "epoch": 0.5920334235389071,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.2554308571795107e-05,
+      "loss": 0.219,
+      "step": 12470
+    },
+    {
+      "epoch": 0.5922708066277358,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.2537078583259215e-05,
+      "loss": 0.2247,
+      "step": 12475
+    },
+    {
+      "epoch": 0.5925081897165646,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.251985165334921e-05,
+      "loss": 0.2225,
+      "step": 12480
+    },
+    {
+      "epoch": 0.5927455728053933,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.2502627792682e-05,
+      "loss": 0.2279,
+      "step": 12485
+    },
+    {
+      "epoch": 0.5929829558942221,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.248540701187255e-05,
+      "loss": 0.2252,
+      "step": 12490
+    },
+    {
+      "epoch": 0.5932203389830508,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.246818932153395e-05,
+      "loss": 0.2248,
+      "step": 12495
+    },
+    {
+      "epoch": 0.5934577220718796,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.2450974732277396e-05,
+      "loss": 0.2211,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5936951051607083,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.2433763254712152e-05,
+      "loss": 0.2258,
+      "step": 12505
+    },
+    {
+      "epoch": 0.5939324882495371,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.241655489944558e-05,
+      "loss": 0.2231,
+      "step": 12510
+    },
+    {
+      "epoch": 0.5941698713383659,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.2399349677083103e-05,
+      "loss": 0.2217,
+      "step": 12515
+    },
+    {
+      "epoch": 0.5944072544271946,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.2382147598228226e-05,
+      "loss": 0.2275,
+      "step": 12520
+    },
+    {
+      "epoch": 0.5946446375160234,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.236494867348251e-05,
+      "loss": 0.2258,
+      "step": 12525
+    },
+    {
+      "epoch": 0.5948820206048521,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.2347752913445587e-05,
+      "loss": 0.2236,
+      "step": 12530
+    },
+    {
+      "epoch": 0.5951194036936809,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.2330560328715106e-05,
+      "loss": 0.2255,
+      "step": 12535
+    },
+    {
+      "epoch": 0.5953567867825096,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.2313370929886783e-05,
+      "loss": 0.2243,
+      "step": 12540
+    },
+    {
+      "epoch": 0.5955941698713384,
+      "grad_norm": 0.376953125,
+      "learning_rate": 2.229618472755438e-05,
+      "loss": 0.2209,
+      "step": 12545
+    },
+    {
+      "epoch": 0.5958315529601671,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.2279001732309657e-05,
+      "loss": 0.2226,
+      "step": 12550
+    },
+    {
+      "epoch": 0.5960689360489959,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.2261821954742434e-05,
+      "loss": 0.224,
+      "step": 12555
+    },
+    {
+      "epoch": 0.5963063191378246,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.2244645405440513e-05,
+      "loss": 0.226,
+      "step": 12560
+    },
+    {
+      "epoch": 0.5965437022266534,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.222747209498973e-05,
+      "loss": 0.226,
+      "step": 12565
+    },
+    {
+      "epoch": 0.5967810853154821,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.2210302033973924e-05,
+      "loss": 0.2269,
+      "step": 12570
+    },
+    {
+      "epoch": 0.5970184684043108,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.219313523297492e-05,
+      "loss": 0.2256,
+      "step": 12575
+    },
+    {
+      "epoch": 0.5972558514931396,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.2175971702572535e-05,
+      "loss": 0.2272,
+      "step": 12580
+    },
+    {
+      "epoch": 0.5974932345819683,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.2158811453344585e-05,
+      "loss": 0.222,
+      "step": 12585
+    },
+    {
+      "epoch": 0.5977306176707972,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.2141654495866853e-05,
+      "loss": 0.2183,
+      "step": 12590
+    },
+    {
+      "epoch": 0.5979680007596259,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.212450084071308e-05,
+      "loss": 0.2251,
+      "step": 12595
+    },
+    {
+      "epoch": 0.5982053838484547,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.2107350498455004e-05,
+      "loss": 0.2209,
+      "step": 12600
+    },
+    {
+      "epoch": 0.5984427669372834,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.2090203479662282e-05,
+      "loss": 0.2238,
+      "step": 12605
+    },
+    {
+      "epoch": 0.5986801500261122,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.207305979490256e-05,
+      "loss": 0.2245,
+      "step": 12610
+    },
+    {
+      "epoch": 0.5989175331149409,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.2055919454741407e-05,
+      "loss": 0.2222,
+      "step": 12615
+    },
+    {
+      "epoch": 0.5991549162037696,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.203878246974235e-05,
+      "loss": 0.2214,
+      "step": 12620
+    },
+    {
+      "epoch": 0.5993922992925984,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.2021648850466803e-05,
+      "loss": 0.2203,
+      "step": 12625
+    },
+    {
+      "epoch": 0.5996296823814271,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.2004518607474173e-05,
+      "loss": 0.2256,
+      "step": 12630
+    },
+    {
+      "epoch": 0.5998670654702559,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.1987391751321735e-05,
+      "loss": 0.2251,
+      "step": 12635
+    },
+    {
+      "epoch": 0.6001044485590846,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.1970268292564683e-05,
+      "loss": 0.2244,
+      "step": 12640
+    },
+    {
+      "epoch": 0.6003418316479134,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.195314824175615e-05,
+      "loss": 0.2236,
+      "step": 12645
+    },
+    {
+      "epoch": 0.6005792147367421,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.1936031609447117e-05,
+      "loss": 0.2247,
+      "step": 12650
+    },
+    {
+      "epoch": 0.600816597825571,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.191891840618652e-05,
+      "loss": 0.2239,
+      "step": 12655
+    },
+    {
+      "epoch": 0.6010539809143997,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.190180864252111e-05,
+      "loss": 0.2238,
+      "step": 12660
+    },
+    {
+      "epoch": 0.6012913640032284,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.188470232899559e-05,
+      "loss": 0.223,
+      "step": 12665
+    },
+    {
+      "epoch": 0.6015287470920572,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.1867599476152484e-05,
+      "loss": 0.2259,
+      "step": 12670
+    },
+    {
+      "epoch": 0.6017661301808859,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.185050009453221e-05,
+      "loss": 0.2231,
+      "step": 12675
+    },
+    {
+      "epoch": 0.6020035132697147,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.1833404194673045e-05,
+      "loss": 0.2237,
+      "step": 12680
+    },
+    {
+      "epoch": 0.6022408963585434,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.18163117871111e-05,
+      "loss": 0.2238,
+      "step": 12685
+    },
+    {
+      "epoch": 0.6024782794473722,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.1799222882380354e-05,
+      "loss": 0.2239,
+      "step": 12690
+    },
+    {
+      "epoch": 0.6027156625362009,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.1782137491012623e-05,
+      "loss": 0.227,
+      "step": 12695
+    },
+    {
+      "epoch": 0.6029530456250297,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.176505562353756e-05,
+      "loss": 0.2215,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6031904287138584,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.1747977290482634e-05,
+      "loss": 0.2212,
+      "step": 12705
+    },
+    {
+      "epoch": 0.6034278118026872,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.1730902502373145e-05,
+      "loss": 0.2224,
+      "step": 12710
+    },
+    {
+      "epoch": 0.6036651948915159,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.1713831269732214e-05,
+      "loss": 0.2242,
+      "step": 12715
+    },
+    {
+      "epoch": 0.6039025779803446,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.169676360308076e-05,
+      "loss": 0.2182,
+      "step": 12720
+    },
+    {
+      "epoch": 0.6041399610691734,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.1679699512937507e-05,
+      "loss": 0.2232,
+      "step": 12725
+    },
+    {
+      "epoch": 0.6043773441580021,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.1662639009818974e-05,
+      "loss": 0.2186,
+      "step": 12730
+    },
+    {
+      "epoch": 0.604614727246831,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.1645582104239485e-05,
+      "loss": 0.2248,
+      "step": 12735
+    },
+    {
+      "epoch": 0.6048521103356597,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.1628528806711114e-05,
+      "loss": 0.2245,
+      "step": 12740
+    },
+    {
+      "epoch": 0.6050894934244885,
+      "grad_norm": 0.4921875,
+      "learning_rate": 2.161147912774375e-05,
+      "loss": 0.2251,
+      "step": 12745
+    },
+    {
+      "epoch": 0.6053268765133172,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.1594433077845006e-05,
+      "loss": 0.2241,
+      "step": 12750
+    },
+    {
+      "epoch": 0.605564259602146,
+      "grad_norm": 0.48046875,
+      "learning_rate": 2.1577390667520313e-05,
+      "loss": 0.2207,
+      "step": 12755
+    },
+    {
+      "epoch": 0.6058016426909747,
+      "grad_norm": 0.52734375,
+      "learning_rate": 2.1560351907272813e-05,
+      "loss": 0.2231,
+      "step": 12760
+    },
+    {
+      "epoch": 0.6060390257798034,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.1543316807603436e-05,
+      "loss": 0.221,
+      "step": 12765
+    },
+    {
+      "epoch": 0.6062764088686322,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.1526285379010808e-05,
+      "loss": 0.2259,
+      "step": 12770
+    },
+    {
+      "epoch": 0.6065137919574609,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.1509257631991346e-05,
+      "loss": 0.227,
+      "step": 12775
+    },
+    {
+      "epoch": 0.6067511750462897,
+      "grad_norm": 0.486328125,
+      "learning_rate": 2.1492233577039167e-05,
+      "loss": 0.2195,
+      "step": 12780
+    },
+    {
+      "epoch": 0.6069885581351184,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.1475213224646107e-05,
+      "loss": 0.2252,
+      "step": 12785
+    },
+    {
+      "epoch": 0.6072259412239472,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.145819658530174e-05,
+      "loss": 0.2267,
+      "step": 12790
+    },
+    {
+      "epoch": 0.6074633243127759,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.1441183669493348e-05,
+      "loss": 0.2241,
+      "step": 12795
+    },
+    {
+      "epoch": 0.6077007074016048,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.142417448770591e-05,
+      "loss": 0.2242,
+      "step": 12800
+    },
+    {
+      "epoch": 0.6079380904904335,
+      "grad_norm": 0.375,
+      "learning_rate": 2.14071690504221e-05,
+      "loss": 0.2202,
+      "step": 12805
+    },
+    {
+      "epoch": 0.6081754735792622,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.1390167368122305e-05,
+      "loss": 0.2232,
+      "step": 12810
+    },
+    {
+      "epoch": 0.608412856668091,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.137316945128457e-05,
+      "loss": 0.2265,
+      "step": 12815
+    },
+    {
+      "epoch": 0.6086502397569197,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.1356175310384645e-05,
+      "loss": 0.2209,
+      "step": 12820
+    },
+    {
+      "epoch": 0.6088876228457485,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.1339184955895936e-05,
+      "loss": 0.2208,
+      "step": 12825
+    },
+    {
+      "epoch": 0.6091250059345772,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.1322198398289527e-05,
+      "loss": 0.2247,
+      "step": 12830
+    },
+    {
+      "epoch": 0.609362389023406,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.130521564803416e-05,
+      "loss": 0.2213,
+      "step": 12835
+    },
+    {
+      "epoch": 0.6095997721122347,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.128823671559621e-05,
+      "loss": 0.222,
+      "step": 12840
+    },
+    {
+      "epoch": 0.6098371552010635,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.1271261611439726e-05,
+      "loss": 0.2195,
+      "step": 12845
+    },
+    {
+      "epoch": 0.6100745382898922,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.1254290346026383e-05,
+      "loss": 0.2234,
+      "step": 12850
+    },
+    {
+      "epoch": 0.610311921378721,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.1237322929815508e-05,
+      "loss": 0.2258,
+      "step": 12855
+    },
+    {
+      "epoch": 0.6105493044675497,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.1220359373264027e-05,
+      "loss": 0.2186,
+      "step": 12860
+    },
+    {
+      "epoch": 0.6107866875563784,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.1203399686826513e-05,
+      "loss": 0.2226,
+      "step": 12865
+    },
+    {
+      "epoch": 0.6110240706452073,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.118644388095513e-05,
+      "loss": 0.2252,
+      "step": 12870
+    },
+    {
+      "epoch": 0.611261453734036,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.116949196609968e-05,
+      "loss": 0.224,
+      "step": 12875
+    },
+    {
+      "epoch": 0.6114988368228648,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.1152543952707553e-05,
+      "loss": 0.2252,
+      "step": 12880
+    },
+    {
+      "epoch": 0.6117362199116935,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.1135599851223713e-05,
+      "loss": 0.2207,
+      "step": 12885
+    },
+    {
+      "epoch": 0.6119736030005223,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.1118659672090746e-05,
+      "loss": 0.219,
+      "step": 12890
+    },
+    {
+      "epoch": 0.612210986089351,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.1101723425748803e-05,
+      "loss": 0.222,
+      "step": 12895
+    },
+    {
+      "epoch": 0.6124483691781798,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.1084791122635623e-05,
+      "loss": 0.2204,
+      "step": 12900
+    },
+    {
+      "epoch": 0.6126857522670085,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.10678627731865e-05,
+      "loss": 0.2242,
+      "step": 12905
+    },
+    {
+      "epoch": 0.6129231353558372,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.1050938387834302e-05,
+      "loss": 0.2191,
+      "step": 12910
+    },
+    {
+      "epoch": 0.613160518444666,
+      "grad_norm": 0.40625,
+      "learning_rate": 2.1034017977009447e-05,
+      "loss": 0.2213,
+      "step": 12915
+    },
+    {
+      "epoch": 0.6133979015334947,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.1017101551139905e-05,
+      "loss": 0.2257,
+      "step": 12920
+    },
+    {
+      "epoch": 0.6136352846223235,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.100018912065121e-05,
+      "loss": 0.2255,
+      "step": 12925
+    },
+    {
+      "epoch": 0.6138726677111522,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.098328069596639e-05,
+      "loss": 0.2218,
+      "step": 12930
+    },
+    {
+      "epoch": 0.614110050799981,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.096637628750605e-05,
+      "loss": 0.2215,
+      "step": 12935
+    },
+    {
+      "epoch": 0.6143474338888097,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.0949475905688293e-05,
+      "loss": 0.2215,
+      "step": 12940
+    },
+    {
+      "epoch": 0.6145848169776386,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.093257956092875e-05,
+      "loss": 0.2262,
+      "step": 12945
+    },
+    {
+      "epoch": 0.6148222000664673,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.0915687263640555e-05,
+      "loss": 0.2283,
+      "step": 12950
+    },
+    {
+      "epoch": 0.615059583155296,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.0898799024234363e-05,
+      "loss": 0.2244,
+      "step": 12955
+    },
+    {
+      "epoch": 0.6152969662441248,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.0881914853118305e-05,
+      "loss": 0.2257,
+      "step": 12960
+    },
+    {
+      "epoch": 0.6155343493329535,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.0865034760698037e-05,
+      "loss": 0.2231,
+      "step": 12965
+    },
+    {
+      "epoch": 0.6157717324217823,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.0848158757376675e-05,
+      "loss": 0.2236,
+      "step": 12970
+    },
+    {
+      "epoch": 0.616009115510611,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.083128685355482e-05,
+      "loss": 0.2262,
+      "step": 12975
+    },
+    {
+      "epoch": 0.6162464985994398,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.0814419059630564e-05,
+      "loss": 0.2216,
+      "step": 12980
+    },
+    {
+      "epoch": 0.6164838816882685,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.0797555385999422e-05,
+      "loss": 0.2213,
+      "step": 12985
+    },
+    {
+      "epoch": 0.6167212647770973,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.0780695843054432e-05,
+      "loss": 0.2272,
+      "step": 12990
+    },
+    {
+      "epoch": 0.616958647865926,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.0763840441186023e-05,
+      "loss": 0.2228,
+      "step": 12995
+    },
+    {
+      "epoch": 0.6171960309547548,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.0746989190782133e-05,
+      "loss": 0.2226,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6174334140435835,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.0730142102228083e-05,
+      "loss": 0.2218,
+      "step": 13005
+    },
+    {
+      "epoch": 0.6176707971324122,
+      "grad_norm": 0.46484375,
+      "learning_rate": 2.0713299185906687e-05,
+      "loss": 0.2246,
+      "step": 13010
+    },
+    {
+      "epoch": 0.617908180221241,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.069646045219813e-05,
+      "loss": 0.2226,
+      "step": 13015
+    },
+    {
+      "epoch": 0.6181455633100698,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.0679625911480065e-05,
+      "loss": 0.2213,
+      "step": 13020
+    },
+    {
+      "epoch": 0.6183829463988986,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.0662795574127535e-05,
+      "loss": 0.2241,
+      "step": 13025
+    },
+    {
+      "epoch": 0.6186203294877273,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.0645969450513003e-05,
+      "loss": 0.2265,
+      "step": 13030
+    },
+    {
+      "epoch": 0.6188577125765561,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.0629147551006327e-05,
+      "loss": 0.2206,
+      "step": 13035
+    },
+    {
+      "epoch": 0.6190950956653848,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.0612329885974774e-05,
+      "loss": 0.2217,
+      "step": 13040
+    },
+    {
+      "epoch": 0.6193324787542136,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.0595516465782987e-05,
+      "loss": 0.2245,
+      "step": 13045
+    },
+    {
+      "epoch": 0.6195698618430423,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.057870730079301e-05,
+      "loss": 0.2207,
+      "step": 13050
+    },
+    {
+      "epoch": 0.619807244931871,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.0561902401364254e-05,
+      "loss": 0.2226,
+      "step": 13055
+    },
+    {
+      "epoch": 0.6200446280206998,
+      "grad_norm": 0.48828125,
+      "learning_rate": 2.0545101777853487e-05,
+      "loss": 0.2251,
+      "step": 13060
+    },
+    {
+      "epoch": 0.6202820111095285,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.0528305440614875e-05,
+      "loss": 0.2217,
+      "step": 13065
+    },
+    {
+      "epoch": 0.6205193941983573,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.0511513399999916e-05,
+      "loss": 0.222,
+      "step": 13070
+    },
+    {
+      "epoch": 0.620756777287186,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.0494725666357473e-05,
+      "loss": 0.2221,
+      "step": 13075
+    },
+    {
+      "epoch": 0.6209941603760148,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.0477942250033742e-05,
+      "loss": 0.2203,
+      "step": 13080
+    },
+    {
+      "epoch": 0.6212315434648435,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.0461163161372277e-05,
+      "loss": 0.2258,
+      "step": 13085
+    },
+    {
+      "epoch": 0.6214689265536724,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.0444388410713944e-05,
+      "loss": 0.225,
+      "step": 13090
+    },
+    {
+      "epoch": 0.6217063096425011,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.042761800839695e-05,
+      "loss": 0.2241,
+      "step": 13095
+    },
+    {
+      "epoch": 0.6219436927313298,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.0410851964756816e-05,
+      "loss": 0.2265,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6221810758201586,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.0394090290126373e-05,
+      "loss": 0.2232,
+      "step": 13105
+    },
+    {
+      "epoch": 0.6224184589089873,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.0377332994835778e-05,
+      "loss": 0.224,
+      "step": 13110
+    },
+    {
+      "epoch": 0.6226558419978161,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.0360580089212466e-05,
+      "loss": 0.2207,
+      "step": 13115
+    },
+    {
+      "epoch": 0.6228932250866448,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.0343831583581186e-05,
+      "loss": 0.2265,
+      "step": 13120
+    },
+    {
+      "epoch": 0.6231306081754736,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.032708748826397e-05,
+      "loss": 0.2279,
+      "step": 13125
+    },
+    {
+      "epoch": 0.6233679912643023,
+      "grad_norm": 0.4609375,
+      "learning_rate": 2.0310347813580095e-05,
+      "loss": 0.2244,
+      "step": 13130
+    },
+    {
+      "epoch": 0.6236053743531311,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.0293612569846184e-05,
+      "loss": 0.2271,
+      "step": 13135
+    },
+    {
+      "epoch": 0.6238427574419598,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.027688176737606e-05,
+      "loss": 0.2231,
+      "step": 13140
+    },
+    {
+      "epoch": 0.6240801405307886,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.0260155416480875e-05,
+      "loss": 0.2215,
+      "step": 13145
+    },
+    {
+      "epoch": 0.6243175236196173,
+      "grad_norm": 0.41796875,
+      "learning_rate": 2.024343352746897e-05,
+      "loss": 0.2229,
+      "step": 13150
+    },
+    {
+      "epoch": 0.624554906708446,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.0226716110646e-05,
+      "loss": 0.222,
+      "step": 13155
+    },
+    {
+      "epoch": 0.6247922897972749,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.02100031763148e-05,
+      "loss": 0.2232,
+      "step": 13160
+    },
+    {
+      "epoch": 0.6250296728861036,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.0193294734775507e-05,
+      "loss": 0.2226,
+      "step": 13165
+    },
+    {
+      "epoch": 0.6252670559749324,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.017659079632544e-05,
+      "loss": 0.2245,
+      "step": 13170
+    },
+    {
+      "epoch": 0.6255044390637611,
+      "grad_norm": 0.423828125,
+      "learning_rate": 2.015989137125916e-05,
+      "loss": 0.2264,
+      "step": 13175
+    },
+    {
+      "epoch": 0.6257418221525899,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.0143196469868454e-05,
+      "loss": 0.2229,
+      "step": 13180
+    },
+    {
+      "epoch": 0.6259792052414186,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.01265061024423e-05,
+      "loss": 0.2194,
+      "step": 13185
+    },
+    {
+      "epoch": 0.6262165883302474,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.0109820279266915e-05,
+      "loss": 0.2247,
+      "step": 13190
+    },
+    {
+      "epoch": 0.6264539714190761,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.009313901062568e-05,
+      "loss": 0.2227,
+      "step": 13195
+    },
+    {
+      "epoch": 0.6266913545079048,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.0076462306799192e-05,
+      "loss": 0.2255,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6269287375967336,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.005979017806523e-05,
+      "loss": 0.2237,
+      "step": 13205
+    },
+    {
+      "epoch": 0.6271661206855623,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.0043122634698753e-05,
+      "loss": 0.223,
+      "step": 13210
+    },
+    {
+      "epoch": 0.6274035037743911,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.0026459686971876e-05,
+      "loss": 0.2226,
+      "step": 13215
+    },
+    {
+      "epoch": 0.6276408868632198,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.0009801345153917e-05,
+      "loss": 0.2236,
+      "step": 13220
+    },
+    {
+      "epoch": 0.6278782699520487,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.999314761951133e-05,
+      "loss": 0.2244,
+      "step": 13225
+    },
+    {
+      "epoch": 0.6281156530408774,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.9976498520307723e-05,
+      "loss": 0.2216,
+      "step": 13230
+    },
+    {
+      "epoch": 0.6283530361297062,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9959854057803872e-05,
+      "loss": 0.2238,
+      "step": 13235
+    },
+    {
+      "epoch": 0.6285904192185349,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.9943214242257675e-05,
+      "loss": 0.2235,
+      "step": 13240
+    },
+    {
+      "epoch": 0.6288278023073636,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.992657908392418e-05,
+      "loss": 0.226,
+      "step": 13245
+    },
+    {
+      "epoch": 0.6290651853961924,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.990994859305555e-05,
+      "loss": 0.2254,
+      "step": 13250
+    },
+    {
+      "epoch": 0.6293025684850211,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.9893322779901095e-05,
+      "loss": 0.2219,
+      "step": 13255
+    },
+    {
+      "epoch": 0.6295399515738499,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9876701654707213e-05,
+      "loss": 0.2228,
+      "step": 13260
+    },
+    {
+      "epoch": 0.6297773346626786,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.9860085227717444e-05,
+      "loss": 0.225,
+      "step": 13265
+    },
+    {
+      "epoch": 0.6300147177515074,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.984347350917241e-05,
+      "loss": 0.2221,
+      "step": 13270
+    },
+    {
+      "epoch": 0.6302521008403361,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.982686650930982e-05,
+      "loss": 0.2245,
+      "step": 13275
+    },
+    {
+      "epoch": 0.6304894839291649,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.9810264238364528e-05,
+      "loss": 0.2253,
+      "step": 13280
+    },
+    {
+      "epoch": 0.6307268670179936,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9793666706568403e-05,
+      "loss": 0.2251,
+      "step": 13285
+    },
+    {
+      "epoch": 0.6309642501068224,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9777073924150452e-05,
+      "loss": 0.2261,
+      "step": 13290
+    },
+    {
+      "epoch": 0.6312016331956511,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.976048590133672e-05,
+      "loss": 0.2239,
+      "step": 13295
+    },
+    {
+      "epoch": 0.6314390162844798,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.974390264835034e-05,
+      "loss": 0.2215,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6316763993733087,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.9727324175411473e-05,
+      "loss": 0.2214,
+      "step": 13305
+    },
+    {
+      "epoch": 0.6319137824621374,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9710750492737383e-05,
+      "loss": 0.2286,
+      "step": 13310
+    },
+    {
+      "epoch": 0.6321511655509662,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.9694181610542338e-05,
+      "loss": 0.2224,
+      "step": 13315
+    },
+    {
+      "epoch": 0.6323885486397949,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.967761753903768e-05,
+      "loss": 0.2214,
+      "step": 13320
+    },
+    {
+      "epoch": 0.6326259317286237,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.9661058288431756e-05,
+      "loss": 0.222,
+      "step": 13325
+    },
+    {
+      "epoch": 0.6328633148174524,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.9644503868929958e-05,
+      "loss": 0.2204,
+      "step": 13330
+    },
+    {
+      "epoch": 0.6331006979062812,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.962795429073471e-05,
+      "loss": 0.2235,
+      "step": 13335
+    },
+    {
+      "epoch": 0.6333380809951099,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.9611409564045434e-05,
+      "loss": 0.2244,
+      "step": 13340
+    },
+    {
+      "epoch": 0.6335754640839386,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9594869699058578e-05,
+      "loss": 0.2209,
+      "step": 13345
+    },
+    {
+      "epoch": 0.6338128471727674,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.9578334705967567e-05,
+      "loss": 0.225,
+      "step": 13350
+    },
+    {
+      "epoch": 0.6340502302615961,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.9561804594962868e-05,
+      "loss": 0.2252,
+      "step": 13355
+    },
+    {
+      "epoch": 0.6342876133504249,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.954527937623189e-05,
+      "loss": 0.2247,
+      "step": 13360
+    },
+    {
+      "epoch": 0.6345249964392536,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.9528759059959063e-05,
+      "loss": 0.2209,
+      "step": 13365
+    },
+    {
+      "epoch": 0.6347623795280825,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.951224365632578e-05,
+      "loss": 0.2245,
+      "step": 13370
+    },
+    {
+      "epoch": 0.6349997626169112,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.9495733175510404e-05,
+      "loss": 0.2246,
+      "step": 13375
+    },
+    {
+      "epoch": 0.63523714570574,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.9479227627688286e-05,
+      "loss": 0.2215,
+      "step": 13380
+    },
+    {
+      "epoch": 0.6354745287945687,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.946272702303169e-05,
+      "loss": 0.2177,
+      "step": 13385
+    },
+    {
+      "epoch": 0.6357119118833974,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.94462313717099e-05,
+      "loss": 0.2237,
+      "step": 13390
+    },
+    {
+      "epoch": 0.6359492949722262,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.9429740683889076e-05,
+      "loss": 0.2221,
+      "step": 13395
+    },
+    {
+      "epoch": 0.6361866780610549,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.9413254969732385e-05,
+      "loss": 0.225,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6364240611498837,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.9396774239399872e-05,
+      "loss": 0.2293,
+      "step": 13405
+    },
+    {
+      "epoch": 0.6366614442387124,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.9380298503048554e-05,
+      "loss": 0.2183,
+      "step": 13410
+    },
+    {
+      "epoch": 0.6368988273275412,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.936382777083235e-05,
+      "loss": 0.2243,
+      "step": 13415
+    },
+    {
+      "epoch": 0.6371362104163699,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9347362052902086e-05,
+      "loss": 0.224,
+      "step": 13420
+    },
+    {
+      "epoch": 0.6373735935051987,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.9330901359405516e-05,
+      "loss": 0.2248,
+      "step": 13425
+    },
+    {
+      "epoch": 0.6376109765940274,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.9314445700487288e-05,
+      "loss": 0.2217,
+      "step": 13430
+    },
+    {
+      "epoch": 0.6378483596828562,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.929799508628896e-05,
+      "loss": 0.2237,
+      "step": 13435
+    },
+    {
+      "epoch": 0.638085742771685,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.9281549526948957e-05,
+      "loss": 0.2223,
+      "step": 13440
+    },
+    {
+      "epoch": 0.6383231258605137,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.9265109032602614e-05,
+      "loss": 0.2273,
+      "step": 13445
+    },
+    {
+      "epoch": 0.6385605089493425,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9248673613382124e-05,
+      "loss": 0.2243,
+      "step": 13450
+    },
+    {
+      "epoch": 0.6387978920381712,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.923224327941657e-05,
+      "loss": 0.2257,
+      "step": 13455
+    },
+    {
+      "epoch": 0.639035275127,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.9215818040831873e-05,
+      "loss": 0.2261,
+      "step": 13460
+    },
+    {
+      "epoch": 0.6392726582158287,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.9199397907750854e-05,
+      "loss": 0.2252,
+      "step": 13465
+    },
+    {
+      "epoch": 0.6395100413046575,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.9182982890293166e-05,
+      "loss": 0.2249,
+      "step": 13470
+    },
+    {
+      "epoch": 0.6397474243934862,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.9166572998575294e-05,
+      "loss": 0.2232,
+      "step": 13475
+    },
+    {
+      "epoch": 0.639984807482315,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.9150168242710594e-05,
+      "loss": 0.228,
+      "step": 13480
+    },
+    {
+      "epoch": 0.6402221905711437,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.913376863280923e-05,
+      "loss": 0.2242,
+      "step": 13485
+    },
+    {
+      "epoch": 0.6404595736599724,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.911737417897822e-05,
+      "loss": 0.2242,
+      "step": 13490
+    },
+    {
+      "epoch": 0.6406969567488012,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.910098489132138e-05,
+      "loss": 0.2242,
+      "step": 13495
+    },
+    {
+      "epoch": 0.6409343398376299,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9084600779939366e-05,
+      "loss": 0.2218,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6411717229264587,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.9068221854929614e-05,
+      "loss": 0.2255,
+      "step": 13505
+    },
+    {
+      "epoch": 0.6414091060152874,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.9051848126386397e-05,
+      "loss": 0.2263,
+      "step": 13510
+    },
+    {
+      "epoch": 0.6416464891041163,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.903547960440077e-05,
+      "loss": 0.2249,
+      "step": 13515
+    },
+    {
+      "epoch": 0.641883872192945,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9019116299060557e-05,
+      "loss": 0.2203,
+      "step": 13520
+    },
+    {
+      "epoch": 0.6421212552817738,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.9002758220450417e-05,
+      "loss": 0.2258,
+      "step": 13525
+    },
+    {
+      "epoch": 0.6423586383706025,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.8986405378651732e-05,
+      "loss": 0.2209,
+      "step": 13530
+    },
+    {
+      "epoch": 0.6425960214594312,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.8970057783742712e-05,
+      "loss": 0.2217,
+      "step": 13535
+    },
+    {
+      "epoch": 0.64283340454826,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.895371544579828e-05,
+      "loss": 0.2243,
+      "step": 13540
+    },
+    {
+      "epoch": 0.6430707876370887,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.8937378374890165e-05,
+      "loss": 0.2238,
+      "step": 13545
+    },
+    {
+      "epoch": 0.6433081707259175,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.892104658108681e-05,
+      "loss": 0.2246,
+      "step": 13550
+    },
+    {
+      "epoch": 0.6435455538147462,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.8904720074453447e-05,
+      "loss": 0.2232,
+      "step": 13555
+    },
+    {
+      "epoch": 0.643782936903575,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.888839886505201e-05,
+      "loss": 0.2226,
+      "step": 13560
+    },
+    {
+      "epoch": 0.6440203199924037,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.8872082962941203e-05,
+      "loss": 0.2218,
+      "step": 13565
+    },
+    {
+      "epoch": 0.6442577030812325,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.8855772378176423e-05,
+      "loss": 0.2269,
+      "step": 13570
+    },
+    {
+      "epoch": 0.6444950861700612,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.883946712080981e-05,
+      "loss": 0.2262,
+      "step": 13575
+    },
+    {
+      "epoch": 0.64473246925889,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8823167200890244e-05,
+      "loss": 0.2194,
+      "step": 13580
+    },
+    {
+      "epoch": 0.6449698523477188,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.880687262846327e-05,
+      "loss": 0.2189,
+      "step": 13585
+    },
+    {
+      "epoch": 0.6452072354365475,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.879058341357116e-05,
+      "loss": 0.2216,
+      "step": 13590
+    },
+    {
+      "epoch": 0.6454446185253763,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8774299566252894e-05,
+      "loss": 0.2196,
+      "step": 13595
+    },
+    {
+      "epoch": 0.645682001614205,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8758021096544127e-05,
+      "loss": 0.2238,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6459193847030338,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.8741748014477194e-05,
+      "loss": 0.2219,
+      "step": 13605
+    },
+    {
+      "epoch": 0.6461567677918625,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8725480330081148e-05,
+      "loss": 0.2269,
+      "step": 13610
+    },
+    {
+      "epoch": 0.6463941508806913,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.8709218053381673e-05,
+      "loss": 0.2223,
+      "step": 13615
+    },
+    {
+      "epoch": 0.64663153396952,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.869296119440113e-05,
+      "loss": 0.224,
+      "step": 13620
+    },
+    {
+      "epoch": 0.6468689170583488,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.8676709763158557e-05,
+      "loss": 0.2206,
+      "step": 13625
+    },
+    {
+      "epoch": 0.6471063001471775,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.8660463769669634e-05,
+      "loss": 0.2259,
+      "step": 13630
+    },
+    {
+      "epoch": 0.6473436832360062,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8644223223946695e-05,
+      "loss": 0.2294,
+      "step": 13635
+    },
+    {
+      "epoch": 0.647581066324835,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8627988135998708e-05,
+      "loss": 0.2246,
+      "step": 13640
+    },
+    {
+      "epoch": 0.6478184494136637,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.8611758515831297e-05,
+      "loss": 0.2275,
+      "step": 13645
+    },
+    {
+      "epoch": 0.6480558325024925,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.8595534373446687e-05,
+      "loss": 0.2183,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6482932155913212,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.8579315718843758e-05,
+      "loss": 0.2236,
+      "step": 13655
+    },
+    {
+      "epoch": 0.6485305986801501,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.8563102562017993e-05,
+      "loss": 0.2279,
+      "step": 13660
+    },
+    {
+      "epoch": 0.6487679817689788,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.854689491296146e-05,
+      "loss": 0.2225,
+      "step": 13665
+    },
+    {
+      "epoch": 0.6490053648578076,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.8530692781662895e-05,
+      "loss": 0.2231,
+      "step": 13670
+    },
+    {
+      "epoch": 0.6492427479466363,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.8514496178107572e-05,
+      "loss": 0.2231,
+      "step": 13675
+    },
+    {
+      "epoch": 0.649480131035465,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.8498305112277408e-05,
+      "loss": 0.2283,
+      "step": 13680
+    },
+    {
+      "epoch": 0.6497175141242938,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.8482119594150853e-05,
+      "loss": 0.2197,
+      "step": 13685
+    },
+    {
+      "epoch": 0.6499548972131225,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.8465939633703e-05,
+      "loss": 0.2224,
+      "step": 13690
+    },
+    {
+      "epoch": 0.6501922803019513,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.844976524090546e-05,
+      "loss": 0.2231,
+      "step": 13695
+    },
+    {
+      "epoch": 0.65042966339078,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.843359642572646e-05,
+      "loss": 0.2214,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6506670464796088,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.841743319813074e-05,
+      "loss": 0.2224,
+      "step": 13705
+    },
+    {
+      "epoch": 0.6509044295684375,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.8401275568079645e-05,
+      "loss": 0.2231,
+      "step": 13710
+    },
+    {
+      "epoch": 0.6511418126572663,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.8385123545531045e-05,
+      "loss": 0.2242,
+      "step": 13715
+    },
+    {
+      "epoch": 0.651379195746095,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.8368977140439354e-05,
+      "loss": 0.2205,
+      "step": 13720
+    },
+    {
+      "epoch": 0.6516165788349239,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.835283636275553e-05,
+      "loss": 0.2225,
+      "step": 13725
+    },
+    {
+      "epoch": 0.6518539619237526,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.8336701222427057e-05,
+      "loss": 0.2248,
+      "step": 13730
+    },
+    {
+      "epoch": 0.6520913450125813,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.8320571729397956e-05,
+      "loss": 0.225,
+      "step": 13735
+    },
+    {
+      "epoch": 0.6523287281014101,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.8304447893608757e-05,
+      "loss": 0.2213,
+      "step": 13740
+    },
+    {
+      "epoch": 0.6525661111902388,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.82883297249965e-05,
+      "loss": 0.2268,
+      "step": 13745
+    },
+    {
+      "epoch": 0.6528034942790676,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.827221723349475e-05,
+      "loss": 0.2212,
+      "step": 13750
+    },
+    {
+      "epoch": 0.6530408773678963,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.825611042903356e-05,
+      "loss": 0.2233,
+      "step": 13755
+    },
+    {
+      "epoch": 0.6532782604567251,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8240009321539476e-05,
+      "loss": 0.2268,
+      "step": 13760
+    },
+    {
+      "epoch": 0.6535156435455538,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.822391392093553e-05,
+      "loss": 0.2243,
+      "step": 13765
+    },
+    {
+      "epoch": 0.6537530266343826,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.8207824237141267e-05,
+      "loss": 0.2222,
+      "step": 13770
+    },
+    {
+      "epoch": 0.6539904097232113,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.819174028007266e-05,
+      "loss": 0.2219,
+      "step": 13775
+    },
+    {
+      "epoch": 0.65422779281204,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.81756620596422e-05,
+      "loss": 0.2231,
+      "step": 13780
+    },
+    {
+      "epoch": 0.6544651759008688,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.8159589585758806e-05,
+      "loss": 0.2217,
+      "step": 13785
+    },
+    {
+      "epoch": 0.6547025589896975,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.8143522868327884e-05,
+      "loss": 0.2264,
+      "step": 13790
+    },
+    {
+      "epoch": 0.6549399420785263,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.8127461917251266e-05,
+      "loss": 0.2214,
+      "step": 13795
+    },
+    {
+      "epoch": 0.655177325167355,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.8111406742427245e-05,
+      "loss": 0.223,
+      "step": 13800
+    },
+    {
+      "epoch": 0.6554147082561839,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.8095357353750553e-05,
+      "loss": 0.2197,
+      "step": 13805
+    },
+    {
+      "epoch": 0.6556520913450126,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.8079313761112372e-05,
+      "loss": 0.2234,
+      "step": 13810
+    },
+    {
+      "epoch": 0.6558894744338414,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.8063275974400274e-05,
+      "loss": 0.2235,
+      "step": 13815
+    },
+    {
+      "epoch": 0.6561268575226701,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.8047244003498283e-05,
+      "loss": 0.2242,
+      "step": 13820
+    },
+    {
+      "epoch": 0.6563642406114988,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.8031217858286824e-05,
+      "loss": 0.223,
+      "step": 13825
+    },
+    {
+      "epoch": 0.6566016237003276,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.8015197548642737e-05,
+      "loss": 0.2242,
+      "step": 13830
+    },
+    {
+      "epoch": 0.6568390067891563,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.799918308443928e-05,
+      "loss": 0.2225,
+      "step": 13835
+    },
+    {
+      "epoch": 0.6570763898779851,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.798317447554608e-05,
+      "loss": 0.2237,
+      "step": 13840
+    },
+    {
+      "epoch": 0.6573137729668138,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.796717173182918e-05,
+      "loss": 0.2261,
+      "step": 13845
+    },
+    {
+      "epoch": 0.6575511560556426,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.7951174863150982e-05,
+      "loss": 0.2273,
+      "step": 13850
+    },
+    {
+      "epoch": 0.6577885391444713,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.7935183879370298e-05,
+      "loss": 0.2255,
+      "step": 13855
+    },
+    {
+      "epoch": 0.6580259222333001,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.7919198790342313e-05,
+      "loss": 0.2261,
+      "step": 13860
+    },
+    {
+      "epoch": 0.6582633053221288,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7903219605918524e-05,
+      "loss": 0.2247,
+      "step": 13865
+    },
+    {
+      "epoch": 0.6585006884109577,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.788724633594686e-05,
+      "loss": 0.2254,
+      "step": 13870
+    },
+    {
+      "epoch": 0.6587380714997864,
+      "grad_norm": 0.5,
+      "learning_rate": 1.787127899027156e-05,
+      "loss": 0.2252,
+      "step": 13875
+    },
+    {
+      "epoch": 0.6589754545886151,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.785531757873325e-05,
+      "loss": 0.2297,
+      "step": 13880
+    },
+    {
+      "epoch": 0.6592128376774439,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.7839362111168846e-05,
+      "loss": 0.225,
+      "step": 13885
+    },
+    {
+      "epoch": 0.6594502207662726,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7823412597411642e-05,
+      "loss": 0.2207,
+      "step": 13890
+    },
+    {
+      "epoch": 0.6596876038551014,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.780746904729125e-05,
+      "loss": 0.225,
+      "step": 13895
+    },
+    {
+      "epoch": 0.6599249869439301,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.7791531470633625e-05,
+      "loss": 0.2226,
+      "step": 13900
+    },
+    {
+      "epoch": 0.6601623700327589,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.7775599877261002e-05,
+      "loss": 0.2294,
+      "step": 13905
+    },
+    {
+      "epoch": 0.6603997531215876,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.7759674276991938e-05,
+      "loss": 0.2244,
+      "step": 13910
+    },
+    {
+      "epoch": 0.6606371362104164,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.774375467964134e-05,
+      "loss": 0.2227,
+      "step": 13915
+    },
+    {
+      "epoch": 0.6608745192992451,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.7727841095020365e-05,
+      "loss": 0.2255,
+      "step": 13920
+    },
+    {
+      "epoch": 0.6611119023880738,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.7711933532936487e-05,
+      "loss": 0.2235,
+      "step": 13925
+    },
+    {
+      "epoch": 0.6613492854769026,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.7696032003193452e-05,
+      "loss": 0.2235,
+      "step": 13930
+    },
+    {
+      "epoch": 0.6615866685657313,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.7680136515591322e-05,
+      "loss": 0.22,
+      "step": 13935
+    },
+    {
+      "epoch": 0.6618240516545602,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7664247079926406e-05,
+      "loss": 0.2236,
+      "step": 13940
+    },
+    {
+      "epoch": 0.6620614347433889,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.764836370599129e-05,
+      "loss": 0.2237,
+      "step": 13945
+    },
+    {
+      "epoch": 0.6622988178322177,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.763248640357481e-05,
+      "loss": 0.2194,
+      "step": 13950
+    },
+    {
+      "epoch": 0.6625362009210464,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.76166151824621e-05,
+      "loss": 0.22,
+      "step": 13955
+    },
+    {
+      "epoch": 0.6627735840098752,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.7600750052434506e-05,
+      "loss": 0.2266,
+      "step": 13960
+    },
+    {
+      "epoch": 0.6630109670987039,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.7584891023269645e-05,
+      "loss": 0.2215,
+      "step": 13965
+    },
+    {
+      "epoch": 0.6632483501875326,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.7569038104741355e-05,
+      "loss": 0.2239,
+      "step": 13970
+    },
+    {
+      "epoch": 0.6634857332763614,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.7553191306619715e-05,
+      "loss": 0.2243,
+      "step": 13975
+    },
+    {
+      "epoch": 0.6637231163651901,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.7537350638671047e-05,
+      "loss": 0.2234,
+      "step": 13980
+    },
+    {
+      "epoch": 0.6639604994540189,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.752151611065788e-05,
+      "loss": 0.2222,
+      "step": 13985
+    },
+    {
+      "epoch": 0.6641978825428476,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.750568773233896e-05,
+      "loss": 0.2226,
+      "step": 13990
+    },
+    {
+      "epoch": 0.6644352656316764,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.7489865513469232e-05,
+      "loss": 0.2248,
+      "step": 13995
+    },
+    {
+      "epoch": 0.6646726487205051,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.7474049463799887e-05,
+      "loss": 0.2252,
+      "step": 14000
+    },
+    {
+      "epoch": 0.664910031809334,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.7458239593078258e-05,
+      "loss": 0.2208,
+      "step": 14005
+    },
+    {
+      "epoch": 0.6651474148981626,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.7442435911047896e-05,
+      "loss": 0.2261,
+      "step": 14010
+    },
+    {
+      "epoch": 0.6653847979869915,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.742663842744856e-05,
+      "loss": 0.2232,
+      "step": 14015
+    },
+    {
+      "epoch": 0.6656221810758202,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.741084715201615e-05,
+      "loss": 0.2255,
+      "step": 14020
+    },
+    {
+      "epoch": 0.6658595641646489,
+      "grad_norm": 0.5,
+      "learning_rate": 1.739506209448276e-05,
+      "loss": 0.2261,
+      "step": 14025
+    },
+    {
+      "epoch": 0.6660969472534777,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.737928326457664e-05,
+      "loss": 0.2201,
+      "step": 14030
+    },
+    {
+      "epoch": 0.6663343303423064,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.7363510672022237e-05,
+      "loss": 0.2205,
+      "step": 14035
+    },
+    {
+      "epoch": 0.6665717134311352,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.7347744326540112e-05,
+      "loss": 0.2256,
+      "step": 14040
+    },
+    {
+      "epoch": 0.6668090965199639,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.7331984237846986e-05,
+      "loss": 0.2244,
+      "step": 14045
+    },
+    {
+      "epoch": 0.6670464796087927,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.7316230415655738e-05,
+      "loss": 0.2215,
+      "step": 14050
+    },
+    {
+      "epoch": 0.6672838626976214,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.730048286967538e-05,
+      "loss": 0.2234,
+      "step": 14055
+    },
+    {
+      "epoch": 0.6675212457864502,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.7284741609611045e-05,
+      "loss": 0.2249,
+      "step": 14060
+    },
+    {
+      "epoch": 0.6677586288752789,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.726900664516401e-05,
+      "loss": 0.226,
+      "step": 14065
+    },
+    {
+      "epoch": 0.6679960119641076,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.7253277986031657e-05,
+      "loss": 0.2262,
+      "step": 14070
+    },
+    {
+      "epoch": 0.6682333950529364,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.7237555641907472e-05,
+      "loss": 0.2225,
+      "step": 14075
+    },
+    {
+      "epoch": 0.6684707781417651,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.722183962248109e-05,
+      "loss": 0.2231,
+      "step": 14080
+    },
+    {
+      "epoch": 0.668708161230594,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.7206129937438207e-05,
+      "loss": 0.2234,
+      "step": 14085
+    },
+    {
+      "epoch": 0.6689455443194227,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.719042659646064e-05,
+      "loss": 0.2241,
+      "step": 14090
+    },
+    {
+      "epoch": 0.6691829274082515,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.717472960922627e-05,
+      "loss": 0.2193,
+      "step": 14095
+    },
+    {
+      "epoch": 0.6694203104970802,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.7159038985409098e-05,
+      "loss": 0.2203,
+      "step": 14100
+    },
+    {
+      "epoch": 0.669657693585909,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.714335473467918e-05,
+      "loss": 0.2232,
+      "step": 14105
+    },
+    {
+      "epoch": 0.6698950766747377,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7127676866702634e-05,
+      "loss": 0.2246,
+      "step": 14110
+    },
+    {
+      "epoch": 0.6701324597635664,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.7112005391141673e-05,
+      "loss": 0.2232,
+      "step": 14115
+    },
+    {
+      "epoch": 0.6703698428523952,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7096340317654547e-05,
+      "loss": 0.2238,
+      "step": 14120
+    },
+    {
+      "epoch": 0.6706072259412239,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.7080681655895587e-05,
+      "loss": 0.2252,
+      "step": 14125
+    },
+    {
+      "epoch": 0.6708446090300527,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.7065029415515125e-05,
+      "loss": 0.2243,
+      "step": 14130
+    },
+    {
+      "epoch": 0.6710819921188814,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.7049383606159596e-05,
+      "loss": 0.2201,
+      "step": 14135
+    },
+    {
+      "epoch": 0.6713193752077102,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.7033744237471416e-05,
+      "loss": 0.2286,
+      "step": 14140
+    },
+    {
+      "epoch": 0.6715567582965389,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.7018111319089087e-05,
+      "loss": 0.2239,
+      "step": 14145
+    },
+    {
+      "epoch": 0.6717941413853677,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.7002484860647082e-05,
+      "loss": 0.2215,
+      "step": 14150
+    },
+    {
+      "epoch": 0.6720315244741965,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.6986864871775908e-05,
+      "loss": 0.2254,
+      "step": 14155
+    },
+    {
+      "epoch": 0.6722689075630253,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.697125136210212e-05,
+      "loss": 0.2235,
+      "step": 14160
+    },
+    {
+      "epoch": 0.672506290651854,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.695564434124824e-05,
+      "loss": 0.2252,
+      "step": 14165
+    },
+    {
+      "epoch": 0.6727436737406827,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.6940043818832803e-05,
+      "loss": 0.2249,
+      "step": 14170
+    },
+    {
+      "epoch": 0.6729810568295115,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6924449804470332e-05,
+      "loss": 0.2255,
+      "step": 14175
+    },
+    {
+      "epoch": 0.6732184399183402,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.6908862307771362e-05,
+      "loss": 0.2254,
+      "step": 14180
+    },
+    {
+      "epoch": 0.673455823007169,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.6893281338342392e-05,
+      "loss": 0.2302,
+      "step": 14185
+    },
+    {
+      "epoch": 0.6736932060959977,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.68777069057859e-05,
+      "loss": 0.2268,
+      "step": 14190
+    },
+    {
+      "epoch": 0.6739305891848265,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.6862139019700335e-05,
+      "loss": 0.224,
+      "step": 14195
+    },
+    {
+      "epoch": 0.6741679722736552,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.6846577689680124e-05,
+      "loss": 0.2204,
+      "step": 14200
+    },
+    {
+      "epoch": 0.674405355362484,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6831022925315636e-05,
+      "loss": 0.2238,
+      "step": 14205
+    },
+    {
+      "epoch": 0.6746427384513127,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.6815474736193208e-05,
+      "loss": 0.2266,
+      "step": 14210
+    },
+    {
+      "epoch": 0.6748801215401414,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.6799933131895114e-05,
+      "loss": 0.2271,
+      "step": 14215
+    },
+    {
+      "epoch": 0.6751175046289702,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.6784398121999564e-05,
+      "loss": 0.2191,
+      "step": 14220
+    },
+    {
+      "epoch": 0.675354887717799,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.6768869716080732e-05,
+      "loss": 0.2252,
+      "step": 14225
+    },
+    {
+      "epoch": 0.6755922708066278,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.6753347923708697e-05,
+      "loss": 0.2281,
+      "step": 14230
+    },
+    {
+      "epoch": 0.6758296538954565,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.6737832754449466e-05,
+      "loss": 0.2225,
+      "step": 14235
+    },
+    {
+      "epoch": 0.6760670369842853,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.6722324217864965e-05,
+      "loss": 0.2213,
+      "step": 14240
+    },
+    {
+      "epoch": 0.676304420073114,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.6706822323513044e-05,
+      "loss": 0.2228,
+      "step": 14245
+    },
+    {
+      "epoch": 0.6765418031619428,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.669132708094746e-05,
+      "loss": 0.226,
+      "step": 14250
+    },
+    {
+      "epoch": 0.6767791862507715,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.6675838499717822e-05,
+      "loss": 0.2229,
+      "step": 14255
+    },
+    {
+      "epoch": 0.6770165693396002,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.6660356589369703e-05,
+      "loss": 0.2192,
+      "step": 14260
+    },
+    {
+      "epoch": 0.677253952428429,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.6644881359444518e-05,
+      "loss": 0.2233,
+      "step": 14265
+    },
+    {
+      "epoch": 0.6774913355172577,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.6629412819479616e-05,
+      "loss": 0.2261,
+      "step": 14270
+    },
+    {
+      "epoch": 0.6777287186060865,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.6613950979008135e-05,
+      "loss": 0.226,
+      "step": 14275
+    },
+    {
+      "epoch": 0.6779661016949152,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.6598495847559175e-05,
+      "loss": 0.225,
+      "step": 14280
+    },
+    {
+      "epoch": 0.678203484783744,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.6583047434657643e-05,
+      "loss": 0.2241,
+      "step": 14285
+    },
+    {
+      "epoch": 0.6784408678725727,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.656760574982435e-05,
+      "loss": 0.2249,
+      "step": 14290
+    },
+    {
+      "epoch": 0.6786782509614016,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.6552170802575906e-05,
+      "loss": 0.224,
+      "step": 14295
+    },
+    {
+      "epoch": 0.6789156340502303,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.653674260242482e-05,
+      "loss": 0.2219,
+      "step": 14300
+    },
+    {
+      "epoch": 0.6791530171390591,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.6521321158879415e-05,
+      "loss": 0.2238,
+      "step": 14305
+    },
+    {
+      "epoch": 0.6793904002278878,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.6505906481443854e-05,
+      "loss": 0.2232,
+      "step": 14310
+    },
+    {
+      "epoch": 0.6796277833167165,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.6490498579618136e-05,
+      "loss": 0.2241,
+      "step": 14315
+    },
+    {
+      "epoch": 0.6798651664055453,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.647509746289807e-05,
+      "loss": 0.2235,
+      "step": 14320
+    },
+    {
+      "epoch": 0.680102549494374,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.6459703140775316e-05,
+      "loss": 0.2301,
+      "step": 14325
+    },
+    {
+      "epoch": 0.6803399325832028,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.644431562273731e-05,
+      "loss": 0.2239,
+      "step": 14330
+    },
+    {
+      "epoch": 0.6805773156720315,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.6428934918267314e-05,
+      "loss": 0.2255,
+      "step": 14335
+    },
+    {
+      "epoch": 0.6808146987608603,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.6413561036844373e-05,
+      "loss": 0.2223,
+      "step": 14340
+    },
+    {
+      "epoch": 0.681052081849689,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.6398193987943362e-05,
+      "loss": 0.2248,
+      "step": 14345
+    },
+    {
+      "epoch": 0.6812894649385178,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.6382833781034914e-05,
+      "loss": 0.2236,
+      "step": 14350
+    },
+    {
+      "epoch": 0.6815268480273465,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.6367480425585447e-05,
+      "loss": 0.2245,
+      "step": 14355
+    },
+    {
+      "epoch": 0.6817642311161752,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.635213393105717e-05,
+      "loss": 0.222,
+      "step": 14360
+    },
+    {
+      "epoch": 0.682001614205004,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.633679430690806e-05,
+      "loss": 0.223,
+      "step": 14365
+    },
+    {
+      "epoch": 0.6822389972938327,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.632146156259186e-05,
+      "loss": 0.2269,
+      "step": 14370
+    },
+    {
+      "epoch": 0.6824763803826616,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.6306135707558065e-05,
+      "loss": 0.2254,
+      "step": 14375
+    },
+    {
+      "epoch": 0.6827137634714903,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.629081675125193e-05,
+      "loss": 0.2256,
+      "step": 14380
+    },
+    {
+      "epoch": 0.6829511465603191,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.6275504703114458e-05,
+      "loss": 0.2261,
+      "step": 14385
+    },
+    {
+      "epoch": 0.6831885296491478,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.6260199572582395e-05,
+      "loss": 0.2246,
+      "step": 14390
+    },
+    {
+      "epoch": 0.6834259127379766,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.6244901369088237e-05,
+      "loss": 0.2254,
+      "step": 14395
+    },
+    {
+      "epoch": 0.6836632958268053,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.622961010206017e-05,
+      "loss": 0.223,
+      "step": 14400
+    },
+    {
+      "epoch": 0.683900678915634,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.6214325780922154e-05,
+      "loss": 0.227,
+      "step": 14405
+    },
+    {
+      "epoch": 0.6841380620044628,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.6199048415093842e-05,
+      "loss": 0.2262,
+      "step": 14410
+    },
+    {
+      "epoch": 0.6843754450932915,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.61837780139906e-05,
+      "loss": 0.2241,
+      "step": 14415
+    },
+    {
+      "epoch": 0.6846128281821203,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.61685145870235e-05,
+      "loss": 0.2251,
+      "step": 14420
+    },
+    {
+      "epoch": 0.684850211270949,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.6153258143599344e-05,
+      "loss": 0.2202,
+      "step": 14425
+    },
+    {
+      "epoch": 0.6850875943597778,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.613800869312059e-05,
+      "loss": 0.2228,
+      "step": 14430
+    },
+    {
+      "epoch": 0.6853249774486065,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.612276624498542e-05,
+      "loss": 0.2227,
+      "step": 14435
+    },
+    {
+      "epoch": 0.6855623605374354,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.6107530808587667e-05,
+      "loss": 0.2215,
+      "step": 14440
+    },
+    {
+      "epoch": 0.6857997436262641,
+      "grad_norm": 0.5,
+      "learning_rate": 1.6092302393316876e-05,
+      "loss": 0.2234,
+      "step": 14445
+    },
+    {
+      "epoch": 0.6860371267150929,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.607708100855825e-05,
+      "loss": 0.2269,
+      "step": 14450
+    },
+    {
+      "epoch": 0.6862745098039216,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.6061866663692652e-05,
+      "loss": 0.2227,
+      "step": 14455
+    },
+    {
+      "epoch": 0.6865118928927503,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.6046659368096617e-05,
+      "loss": 0.2177,
+      "step": 14460
+    },
+    {
+      "epoch": 0.6867492759815791,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.6031459131142325e-05,
+      "loss": 0.2248,
+      "step": 14465
+    },
+    {
+      "epoch": 0.6869866590704078,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.6016265962197626e-05,
+      "loss": 0.2207,
+      "step": 14470
+    },
+    {
+      "epoch": 0.6872240421592366,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.600107987062599e-05,
+      "loss": 0.2259,
+      "step": 14475
+    },
+    {
+      "epoch": 0.6874614252480653,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.5985900865786547e-05,
+      "loss": 0.2265,
+      "step": 14480
+    },
+    {
+      "epoch": 0.6876988083368941,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.5970728957034032e-05,
+      "loss": 0.2217,
+      "step": 14485
+    },
+    {
+      "epoch": 0.6879361914257228,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.595556415371884e-05,
+      "loss": 0.222,
+      "step": 14490
+    },
+    {
+      "epoch": 0.6881735745145516,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5940406465186975e-05,
+      "loss": 0.2258,
+      "step": 14495
+    },
+    {
+      "epoch": 0.6884109576033803,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.5925255900780022e-05,
+      "loss": 0.2241,
+      "step": 14500
+    },
+    {
+      "epoch": 0.688648340692209,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.5910112469835232e-05,
+      "loss": 0.2215,
+      "step": 14505
+    },
+    {
+      "epoch": 0.6888857237810379,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.589497618168541e-05,
+      "loss": 0.2225,
+      "step": 14510
+    },
+    {
+      "epoch": 0.6891231068698666,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.5879847045659025e-05,
+      "loss": 0.2219,
+      "step": 14515
+    },
+    {
+      "epoch": 0.6893604899586954,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.586472507108004e-05,
+      "loss": 0.2263,
+      "step": 14520
+    },
+    {
+      "epoch": 0.6895978730475241,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5849610267268093e-05,
+      "loss": 0.2236,
+      "step": 14525
+    },
+    {
+      "epoch": 0.6898352561363529,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.583450264353835e-05,
+      "loss": 0.2275,
+      "step": 14530
+    },
+    {
+      "epoch": 0.6900726392251816,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.5819402209201602e-05,
+      "loss": 0.2229,
+      "step": 14535
+    },
+    {
+      "epoch": 0.6903100223140104,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.5804308973564137e-05,
+      "loss": 0.2205,
+      "step": 14540
+    },
+    {
+      "epoch": 0.6905474054028391,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.5789222945927877e-05,
+      "loss": 0.2245,
+      "step": 14545
+    },
+    {
+      "epoch": 0.6907847884916678,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5774144135590265e-05,
+      "loss": 0.2235,
+      "step": 14550
+    },
+    {
+      "epoch": 0.6910221715804966,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.57590725518443e-05,
+      "loss": 0.2292,
+      "step": 14555
+    },
+    {
+      "epoch": 0.6912595546693253,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.574400820397853e-05,
+      "loss": 0.2206,
+      "step": 14560
+    },
+    {
+      "epoch": 0.6914969377581541,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.572895110127704e-05,
+      "loss": 0.2258,
+      "step": 14565
+    },
+    {
+      "epoch": 0.6917343208469828,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.5713901253019463e-05,
+      "loss": 0.2245,
+      "step": 14570
+    },
+    {
+      "epoch": 0.6919717039358116,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.5698858668480955e-05,
+      "loss": 0.2228,
+      "step": 14575
+    },
+    {
+      "epoch": 0.6922090870246403,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.5683823356932183e-05,
+      "loss": 0.2236,
+      "step": 14580
+    },
+    {
+      "epoch": 0.6924464701134692,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.5668795327639335e-05,
+      "loss": 0.2242,
+      "step": 14585
+    },
+    {
+      "epoch": 0.6926838532022979,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.5653774589864134e-05,
+      "loss": 0.2249,
+      "step": 14590
+    },
+    {
+      "epoch": 0.6929212362911267,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5638761152863785e-05,
+      "loss": 0.2202,
+      "step": 14595
+    },
+    {
+      "epoch": 0.6931586193799554,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.5623755025891003e-05,
+      "loss": 0.2214,
+      "step": 14600
+    },
+    {
+      "epoch": 0.6933960024687841,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.560875621819399e-05,
+      "loss": 0.2279,
+      "step": 14605
+    },
+    {
+      "epoch": 0.6936333855576129,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.559376473901644e-05,
+      "loss": 0.218,
+      "step": 14610
+    },
+    {
+      "epoch": 0.6938707686464416,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.557878059759755e-05,
+      "loss": 0.2237,
+      "step": 14615
+    },
+    {
+      "epoch": 0.6941081517352704,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.556380380317197e-05,
+      "loss": 0.2212,
+      "step": 14620
+    },
+    {
+      "epoch": 0.6943455348240991,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.5548834364969823e-05,
+      "loss": 0.2219,
+      "step": 14625
+    },
+    {
+      "epoch": 0.6945829179129279,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.5533872292216704e-05,
+      "loss": 0.2256,
+      "step": 14630
+    },
+    {
+      "epoch": 0.6948203010017566,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.5518917594133685e-05,
+      "loss": 0.2236,
+      "step": 14635
+    },
+    {
+      "epoch": 0.6950576840905854,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.5503970279937285e-05,
+      "loss": 0.2237,
+      "step": 14640
+    },
+    {
+      "epoch": 0.6952950671794141,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.548903035883943e-05,
+      "loss": 0.2265,
+      "step": 14645
+    },
+    {
+      "epoch": 0.6955324502682428,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.5474097840047555e-05,
+      "loss": 0.225,
+      "step": 14650
+    },
+    {
+      "epoch": 0.6957698333570717,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.545917273276449e-05,
+      "loss": 0.2203,
+      "step": 14655
+    },
+    {
+      "epoch": 0.6960072164459004,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.544425504618853e-05,
+      "loss": 0.2232,
+      "step": 14660
+    },
+    {
+      "epoch": 0.6962445995347292,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.542934478951335e-05,
+      "loss": 0.2191,
+      "step": 14665
+    },
+    {
+      "epoch": 0.6964819826235579,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.5414441971928096e-05,
+      "loss": 0.2193,
+      "step": 14670
+    },
+    {
+      "epoch": 0.6967193657123867,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5399546602617284e-05,
+      "loss": 0.2242,
+      "step": 14675
+    },
+    {
+      "epoch": 0.6969567488012154,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.53846586907609e-05,
+      "loss": 0.2264,
+      "step": 14680
+    },
+    {
+      "epoch": 0.6971941318900442,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.536977824553425e-05,
+      "loss": 0.2213,
+      "step": 14685
+    },
+    {
+      "epoch": 0.6974315149788729,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.5354905276108117e-05,
+      "loss": 0.2265,
+      "step": 14690
+    },
+    {
+      "epoch": 0.6976688980677016,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.5340039791648636e-05,
+      "loss": 0.2214,
+      "step": 14695
+    },
+    {
+      "epoch": 0.6979062811565304,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.5325181801317337e-05,
+      "loss": 0.2194,
+      "step": 14700
+    },
+    {
+      "epoch": 0.6981436642453591,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.5310331314271127e-05,
+      "loss": 0.2258,
+      "step": 14705
+    },
+    {
+      "epoch": 0.6983810473341879,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.5295488339662296e-05,
+      "loss": 0.2242,
+      "step": 14710
+    },
+    {
+      "epoch": 0.6986184304230166,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.5280652886638507e-05,
+      "loss": 0.2206,
+      "step": 14715
+    },
+    {
+      "epoch": 0.6988558135118454,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.526582496434278e-05,
+      "loss": 0.2261,
+      "step": 14720
+    },
+    {
+      "epoch": 0.6990931966006741,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.5251004581913491e-05,
+      "loss": 0.2214,
+      "step": 14725
+    },
+    {
+      "epoch": 0.699330579689503,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.5236191748484374e-05,
+      "loss": 0.221,
+      "step": 14730
+    },
+    {
+      "epoch": 0.6995679627783317,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.5221386473184524e-05,
+      "loss": 0.2198,
+      "step": 14735
+    },
+    {
+      "epoch": 0.6998053458671605,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5206588765138354e-05,
+      "loss": 0.2212,
+      "step": 14740
+    },
+    {
+      "epoch": 0.7000427289559892,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.5191798633465626e-05,
+      "loss": 0.2256,
+      "step": 14745
+    },
+    {
+      "epoch": 0.7002801120448179,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.5177016087281432e-05,
+      "loss": 0.2225,
+      "step": 14750
+    },
+    {
+      "epoch": 0.7005174951336467,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.5162241135696178e-05,
+      "loss": 0.2241,
+      "step": 14755
+    },
+    {
+      "epoch": 0.7007548782224754,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.5147473787815613e-05,
+      "loss": 0.2233,
+      "step": 14760
+    },
+    {
+      "epoch": 0.7009922613113042,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.5132714052740788e-05,
+      "loss": 0.2209,
+      "step": 14765
+    },
+    {
+      "epoch": 0.7012296444001329,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.511796193956805e-05,
+      "loss": 0.2245,
+      "step": 14770
+    },
+    {
+      "epoch": 0.7014670274889617,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.5103217457389056e-05,
+      "loss": 0.227,
+      "step": 14775
+    },
+    {
+      "epoch": 0.7017044105777904,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5088480615290779e-05,
+      "loss": 0.2238,
+      "step": 14780
+    },
+    {
+      "epoch": 0.7019417936666192,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.5073751422355465e-05,
+      "loss": 0.2212,
+      "step": 14785
+    },
+    {
+      "epoch": 0.7021791767554479,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.5059029887660636e-05,
+      "loss": 0.2235,
+      "step": 14790
+    },
+    {
+      "epoch": 0.7024165598442766,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.5044316020279117e-05,
+      "loss": 0.2234,
+      "step": 14795
+    },
+    {
+      "epoch": 0.7026539429331055,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.502960982927899e-05,
+      "loss": 0.2213,
+      "step": 14800
+    },
+    {
+      "epoch": 0.7028913260219342,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.5014911323723627e-05,
+      "loss": 0.2217,
+      "step": 14805
+    },
+    {
+      "epoch": 0.703128709110763,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.5000220512671632e-05,
+      "loss": 0.2215,
+      "step": 14810
+    },
+    {
+      "epoch": 0.7033660921995917,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.4985537405176898e-05,
+      "loss": 0.2265,
+      "step": 14815
+    },
+    {
+      "epoch": 0.7036034752884205,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.4970862010288555e-05,
+      "loss": 0.2179,
+      "step": 14820
+    },
+    {
+      "epoch": 0.7038408583772492,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.4956194337050982e-05,
+      "loss": 0.2196,
+      "step": 14825
+    },
+    {
+      "epoch": 0.704078241466078,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.4941534394503792e-05,
+      "loss": 0.2224,
+      "step": 14830
+    },
+    {
+      "epoch": 0.7043156245549067,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4926882191681854e-05,
+      "loss": 0.2243,
+      "step": 14835
+    },
+    {
+      "epoch": 0.7045530076437354,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.4912237737615248e-05,
+      "loss": 0.2257,
+      "step": 14840
+    },
+    {
+      "epoch": 0.7047903907325642,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.489760104132928e-05,
+      "loss": 0.2279,
+      "step": 14845
+    },
+    {
+      "epoch": 0.7050277738213929,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4882972111844485e-05,
+      "loss": 0.2256,
+      "step": 14850
+    },
+    {
+      "epoch": 0.7052651569102217,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.4868350958176597e-05,
+      "loss": 0.2218,
+      "step": 14855
+    },
+    {
+      "epoch": 0.7055025399990504,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.4853737589336581e-05,
+      "loss": 0.2197,
+      "step": 14860
+    },
+    {
+      "epoch": 0.7057399230878793,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.4839132014330587e-05,
+      "loss": 0.2265,
+      "step": 14865
+    },
+    {
+      "epoch": 0.705977306176708,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4824534242159955e-05,
+      "loss": 0.2211,
+      "step": 14870
+    },
+    {
+      "epoch": 0.7062146892655368,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.480994428182122e-05,
+      "loss": 0.2234,
+      "step": 14875
+    },
+    {
+      "epoch": 0.7064520723543655,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.4795362142306126e-05,
+      "loss": 0.2242,
+      "step": 14880
+    },
+    {
+      "epoch": 0.7066894554431943,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.4780787832601583e-05,
+      "loss": 0.2238,
+      "step": 14885
+    },
+    {
+      "epoch": 0.706926838532023,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.4766221361689642e-05,
+      "loss": 0.2226,
+      "step": 14890
+    },
+    {
+      "epoch": 0.7071642216208517,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4751662738547584e-05,
+      "loss": 0.2233,
+      "step": 14895
+    },
+    {
+      "epoch": 0.7074016047096805,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.4737111972147794e-05,
+      "loss": 0.2228,
+      "step": 14900
+    },
+    {
+      "epoch": 0.7076389877985092,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.4722569071457882e-05,
+      "loss": 0.2242,
+      "step": 14905
+    },
+    {
+      "epoch": 0.707876370887338,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4708034045440528e-05,
+      "loss": 0.219,
+      "step": 14910
+    },
+    {
+      "epoch": 0.7081137539761667,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.469350690305363e-05,
+      "loss": 0.2192,
+      "step": 14915
+    },
+    {
+      "epoch": 0.7083511370649955,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.4678987653250187e-05,
+      "loss": 0.2258,
+      "step": 14920
+    },
+    {
+      "epoch": 0.7085885201538242,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.4664476304978367e-05,
+      "loss": 0.2225,
+      "step": 14925
+    },
+    {
+      "epoch": 0.708825903242653,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4649972867181414e-05,
+      "loss": 0.2201,
+      "step": 14930
+    },
+    {
+      "epoch": 0.7090632863314817,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.463547734879776e-05,
+      "loss": 0.2246,
+      "step": 14935
+    },
+    {
+      "epoch": 0.7093006694203104,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.4620989758760919e-05,
+      "loss": 0.2237,
+      "step": 14940
+    },
+    {
+      "epoch": 0.7095380525091393,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.4606510105999521e-05,
+      "loss": 0.2272,
+      "step": 14945
+    },
+    {
+      "epoch": 0.709775435597968,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.4592038399437318e-05,
+      "loss": 0.222,
+      "step": 14950
+    },
+    {
+      "epoch": 0.7100128186867968,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.4577574647993145e-05,
+      "loss": 0.2208,
+      "step": 14955
+    },
+    {
+      "epoch": 0.7102502017756255,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.456311886058096e-05,
+      "loss": 0.2228,
+      "step": 14960
+    },
+    {
+      "epoch": 0.7104875848644543,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.4548671046109791e-05,
+      "loss": 0.2249,
+      "step": 14965
+    },
+    {
+      "epoch": 0.710724967953283,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4534231213483767e-05,
+      "loss": 0.2204,
+      "step": 14970
+    },
+    {
+      "epoch": 0.7109623510421118,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.4519799371602072e-05,
+      "loss": 0.2225,
+      "step": 14975
+    },
+    {
+      "epoch": 0.7111997341309405,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4505375529359014e-05,
+      "loss": 0.2226,
+      "step": 14980
+    },
+    {
+      "epoch": 0.7114371172197692,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4490959695643924e-05,
+      "loss": 0.2231,
+      "step": 14985
+    },
+    {
+      "epoch": 0.711674500308598,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.4476551879341215e-05,
+      "loss": 0.223,
+      "step": 14990
+    },
+    {
+      "epoch": 0.7119118833974267,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4462152089330367e-05,
+      "loss": 0.2252,
+      "step": 14995
+    },
+    {
+      "epoch": 0.7121492664862555,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.4447760334485882e-05,
+      "loss": 0.2218,
+      "step": 15000
+    },
+    {
+      "epoch": 0.7123866495750842,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.443337662367737e-05,
+      "loss": 0.2244,
+      "step": 15005
+    },
+    {
+      "epoch": 0.712624032663913,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.441900096576942e-05,
+      "loss": 0.2252,
+      "step": 15010
+    },
+    {
+      "epoch": 0.7128614157527418,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.440463336962169e-05,
+      "loss": 0.2246,
+      "step": 15015
+    },
+    {
+      "epoch": 0.7130987988415706,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.4390273844088859e-05,
+      "loss": 0.2258,
+      "step": 15020
+    },
+    {
+      "epoch": 0.7133361819303993,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.4375922398020649e-05,
+      "loss": 0.2232,
+      "step": 15025
+    },
+    {
+      "epoch": 0.7135735650192281,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.4361579040261796e-05,
+      "loss": 0.2232,
+      "step": 15030
+    },
+    {
+      "epoch": 0.7138109481080568,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.434724377965202e-05,
+      "loss": 0.2257,
+      "step": 15035
+    },
+    {
+      "epoch": 0.7140483311968855,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.4332916625026101e-05,
+      "loss": 0.2179,
+      "step": 15040
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.4318597585213783e-05,
+      "loss": 0.2249,
+      "step": 15045
+    },
+    {
+      "epoch": 0.714523097374543,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.430428666903985e-05,
+      "loss": 0.2278,
+      "step": 15050
+    },
+    {
+      "epoch": 0.7147604804633718,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.4289983885324021e-05,
+      "loss": 0.2195,
+      "step": 15055
+    },
+    {
+      "epoch": 0.7149978635522005,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4275689242881063e-05,
+      "loss": 0.2236,
+      "step": 15060
+    },
+    {
+      "epoch": 0.7152352466410293,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.4261402750520681e-05,
+      "loss": 0.2224,
+      "step": 15065
+    },
+    {
+      "epoch": 0.715472629729858,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.4247124417047597e-05,
+      "loss": 0.2231,
+      "step": 15070
+    },
+    {
+      "epoch": 0.7157100128186868,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.4232854251261458e-05,
+      "loss": 0.2256,
+      "step": 15075
+    },
+    {
+      "epoch": 0.7159473959075155,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.421859226195692e-05,
+      "loss": 0.2233,
+      "step": 15080
+    },
+    {
+      "epoch": 0.7161847789963443,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.4204338457923575e-05,
+      "loss": 0.2172,
+      "step": 15085
+    },
+    {
+      "epoch": 0.7164221620851731,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.4190092847945979e-05,
+      "loss": 0.2259,
+      "step": 15090
+    },
+    {
+      "epoch": 0.7166595451740018,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.4175855440803635e-05,
+      "loss": 0.2211,
+      "step": 15095
+    },
+    {
+      "epoch": 0.7168969282628306,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.4161626245270986e-05,
+      "loss": 0.2233,
+      "step": 15100
+    },
+    {
+      "epoch": 0.7171343113516593,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.4147405270117441e-05,
+      "loss": 0.224,
+      "step": 15105
+    },
+    {
+      "epoch": 0.7173716944404881,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.4133192524107305e-05,
+      "loss": 0.2221,
+      "step": 15110
+    },
+    {
+      "epoch": 0.7176090775293168,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.4118988015999834e-05,
+      "loss": 0.2225,
+      "step": 15115
+    },
+    {
+      "epoch": 0.7178464606181456,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.4104791754549195e-05,
+      "loss": 0.2239,
+      "step": 15120
+    },
+    {
+      "epoch": 0.7180838437069743,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4090603748504488e-05,
+      "loss": 0.2213,
+      "step": 15125
+    },
+    {
+      "epoch": 0.718321226795803,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.407642400660972e-05,
+      "loss": 0.2209,
+      "step": 15130
+    },
+    {
+      "epoch": 0.7185586098846318,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.4062252537603793e-05,
+      "loss": 0.2221,
+      "step": 15135
+    },
+    {
+      "epoch": 0.7187959929734605,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.4048089350220522e-05,
+      "loss": 0.2222,
+      "step": 15140
+    },
+    {
+      "epoch": 0.7190333760622893,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.4033934453188608e-05,
+      "loss": 0.2205,
+      "step": 15145
+    },
+    {
+      "epoch": 0.719270759151118,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.4019787855231659e-05,
+      "loss": 0.2217,
+      "step": 15150
+    },
+    {
+      "epoch": 0.7195081422399469,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.4005649565068157e-05,
+      "loss": 0.2239,
+      "step": 15155
+    },
+    {
+      "epoch": 0.7197455253287756,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.3991519591411467e-05,
+      "loss": 0.2217,
+      "step": 15160
+    },
+    {
+      "epoch": 0.7199829084176044,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.397739794296982e-05,
+      "loss": 0.2244,
+      "step": 15165
+    },
+    {
+      "epoch": 0.7202202915064331,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.3963284628446333e-05,
+      "loss": 0.2249,
+      "step": 15170
+    },
+    {
+      "epoch": 0.7204576745952619,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.3949179656538974e-05,
+      "loss": 0.2208,
+      "step": 15175
+    },
+    {
+      "epoch": 0.7206950576840906,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3935083035940577e-05,
+      "loss": 0.2243,
+      "step": 15180
+    },
+    {
+      "epoch": 0.7209324407729193,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.3920994775338814e-05,
+      "loss": 0.2258,
+      "step": 15185
+    },
+    {
+      "epoch": 0.7211698238617481,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.3906914883416221e-05,
+      "loss": 0.2277,
+      "step": 15190
+    },
+    {
+      "epoch": 0.7214072069505768,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.389284336885019e-05,
+      "loss": 0.2246,
+      "step": 15195
+    },
+    {
+      "epoch": 0.7216445900394056,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3878780240312893e-05,
+      "loss": 0.2234,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7218819731282343,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.3864725506471404e-05,
+      "loss": 0.224,
+      "step": 15205
+    },
+    {
+      "epoch": 0.7221193562170631,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.385067917598758e-05,
+      "loss": 0.2248,
+      "step": 15210
+    },
+    {
+      "epoch": 0.7223567393058918,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3836641257518116e-05,
+      "loss": 0.2226,
+      "step": 15215
+    },
+    {
+      "epoch": 0.7225941223947207,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.3822611759714505e-05,
+      "loss": 0.2207,
+      "step": 15220
+    },
+    {
+      "epoch": 0.7228315054835494,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3808590691223074e-05,
+      "loss": 0.2171,
+      "step": 15225
+    },
+    {
+      "epoch": 0.723068888572378,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.3794578060684945e-05,
+      "loss": 0.222,
+      "step": 15230
+    },
+    {
+      "epoch": 0.7233062716612069,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3780573876736034e-05,
+      "loss": 0.2238,
+      "step": 15235
+    },
+    {
+      "epoch": 0.7235436547500356,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.3766578148007054e-05,
+      "loss": 0.2251,
+      "step": 15240
+    },
+    {
+      "epoch": 0.7237810378388644,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.3752590883123505e-05,
+      "loss": 0.2246,
+      "step": 15245
+    },
+    {
+      "epoch": 0.7240184209276931,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.3738612090705687e-05,
+      "loss": 0.2237,
+      "step": 15250
+    },
+    {
+      "epoch": 0.7242558040165219,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3724641779368663e-05,
+      "loss": 0.2228,
+      "step": 15255
+    },
+    {
+      "epoch": 0.7244931871053506,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.3710679957722267e-05,
+      "loss": 0.2243,
+      "step": 15260
+    },
+    {
+      "epoch": 0.7247305701941794,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.3696726634371104e-05,
+      "loss": 0.227,
+      "step": 15265
+    },
+    {
+      "epoch": 0.7249679532830081,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3682781817914553e-05,
+      "loss": 0.2235,
+      "step": 15270
+    },
+    {
+      "epoch": 0.7252053363718368,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.3668845516946737e-05,
+      "loss": 0.2263,
+      "step": 15275
+    },
+    {
+      "epoch": 0.7254427194606656,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.3654917740056535e-05,
+      "loss": 0.2214,
+      "step": 15280
+    },
+    {
+      "epoch": 0.7256801025494943,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3640998495827573e-05,
+      "loss": 0.2247,
+      "step": 15285
+    },
+    {
+      "epoch": 0.7259174856383231,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.3627087792838205e-05,
+      "loss": 0.225,
+      "step": 15290
+    },
+    {
+      "epoch": 0.7261548687271518,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.3613185639661563e-05,
+      "loss": 0.2236,
+      "step": 15295
+    },
+    {
+      "epoch": 0.7263922518159807,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.3599292044865447e-05,
+      "loss": 0.2229,
+      "step": 15300
+    },
+    {
+      "epoch": 0.7266296349048094,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.3585407017012445e-05,
+      "loss": 0.2249,
+      "step": 15305
+    },
+    {
+      "epoch": 0.7268670179936382,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3571530564659812e-05,
+      "loss": 0.2194,
+      "step": 15310
+    },
+    {
+      "epoch": 0.7271044010824669,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.3557662696359574e-05,
+      "loss": 0.2244,
+      "step": 15315
+    },
+    {
+      "epoch": 0.7273417841712957,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.3543803420658397e-05,
+      "loss": 0.2196,
+      "step": 15320
+    },
+    {
+      "epoch": 0.7275791672601244,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.3529952746097717e-05,
+      "loss": 0.2247,
+      "step": 15325
+    },
+    {
+      "epoch": 0.7278165503489531,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3516110681213634e-05,
+      "loss": 0.2237,
+      "step": 15330
+    },
+    {
+      "epoch": 0.7280539334377819,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.3502277234536948e-05,
+      "loss": 0.223,
+      "step": 15335
+    },
+    {
+      "epoch": 0.7282913165266106,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.3488452414593153e-05,
+      "loss": 0.2251,
+      "step": 15340
+    },
+    {
+      "epoch": 0.7285286996154394,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3474636229902409e-05,
+      "loss": 0.2271,
+      "step": 15345
+    },
+    {
+      "epoch": 0.7287660827042681,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.3460828688979587e-05,
+      "loss": 0.2248,
+      "step": 15350
+    },
+    {
+      "epoch": 0.7290034657930969,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.3447029800334204e-05,
+      "loss": 0.2248,
+      "step": 15355
+    },
+    {
+      "epoch": 0.7292408488819256,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.3433239572470452e-05,
+      "loss": 0.2279,
+      "step": 15360
+    },
+    {
+      "epoch": 0.7294782319707545,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3419458013887179e-05,
+      "loss": 0.2237,
+      "step": 15365
+    },
+    {
+      "epoch": 0.7297156150595832,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.340568513307791e-05,
+      "loss": 0.2268,
+      "step": 15370
+    },
+    {
+      "epoch": 0.7299529981484119,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.3391920938530799e-05,
+      "loss": 0.2227,
+      "step": 15375
+    },
+    {
+      "epoch": 0.7301903812372407,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.3378165438728665e-05,
+      "loss": 0.2277,
+      "step": 15380
+    },
+    {
+      "epoch": 0.7304277643260694,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.3364418642148952e-05,
+      "loss": 0.2243,
+      "step": 15385
+    },
+    {
+      "epoch": 0.7306651474148982,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.3350680557263744e-05,
+      "loss": 0.2234,
+      "step": 15390
+    },
+    {
+      "epoch": 0.7309025305037269,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.3336951192539771e-05,
+      "loss": 0.2222,
+      "step": 15395
+    },
+    {
+      "epoch": 0.7311399135925557,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.332323055643838e-05,
+      "loss": 0.2242,
+      "step": 15400
+    },
+    {
+      "epoch": 0.7313772966813844,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.3309518657415526e-05,
+      "loss": 0.2274,
+      "step": 15405
+    },
+    {
+      "epoch": 0.7316146797702132,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.3295815503921788e-05,
+      "loss": 0.2256,
+      "step": 15410
+    },
+    {
+      "epoch": 0.7318520628590419,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.3282121104402373e-05,
+      "loss": 0.2227,
+      "step": 15415
+    },
+    {
+      "epoch": 0.7320894459478706,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.326843546729707e-05,
+      "loss": 0.2239,
+      "step": 15420
+    },
+    {
+      "epoch": 0.7323268290366994,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.325475860104027e-05,
+      "loss": 0.2219,
+      "step": 15425
+    },
+    {
+      "epoch": 0.7325642121255281,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3241090514060967e-05,
+      "loss": 0.221,
+      "step": 15430
+    },
+    {
+      "epoch": 0.732801595214357,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.3227431214782732e-05,
+      "loss": 0.2235,
+      "step": 15435
+    },
+    {
+      "epoch": 0.7330389783031857,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.3213780711623752e-05,
+      "loss": 0.2197,
+      "step": 15440
+    },
+    {
+      "epoch": 0.7332763613920145,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.3200139012996742e-05,
+      "loss": 0.2251,
+      "step": 15445
+    },
+    {
+      "epoch": 0.7335137444808432,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.318650612730904e-05,
+      "loss": 0.2221,
+      "step": 15450
+    },
+    {
+      "epoch": 0.733751127569672,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.3172882062962517e-05,
+      "loss": 0.2201,
+      "step": 15455
+    },
+    {
+      "epoch": 0.7339885106585007,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.3159266828353641e-05,
+      "loss": 0.2261,
+      "step": 15460
+    },
+    {
+      "epoch": 0.7342258937473295,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.3145660431873396e-05,
+      "loss": 0.2224,
+      "step": 15465
+    },
+    {
+      "epoch": 0.7344632768361582,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.3132062881907359e-05,
+      "loss": 0.2199,
+      "step": 15470
+    },
+    {
+      "epoch": 0.7347006599249869,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.3118474186835628e-05,
+      "loss": 0.2256,
+      "step": 15475
+    },
+    {
+      "epoch": 0.7349380430138157,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.3104894355032863e-05,
+      "loss": 0.2206,
+      "step": 15480
+    },
+    {
+      "epoch": 0.7351754261026444,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3091323394868246e-05,
+      "loss": 0.2199,
+      "step": 15485
+    },
+    {
+      "epoch": 0.7354128091914732,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.3077761314705495e-05,
+      "loss": 0.2227,
+      "step": 15490
+    },
+    {
+      "epoch": 0.7356501922803019,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.3064208122902866e-05,
+      "loss": 0.2212,
+      "step": 15495
+    },
+    {
+      "epoch": 0.7358875753691307,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.3050663827813132e-05,
+      "loss": 0.2213,
+      "step": 15500
+    },
+    {
+      "epoch": 0.7361249584579594,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3037128437783569e-05,
+      "loss": 0.2205,
+      "step": 15505
+    },
+    {
+      "epoch": 0.7363623415467883,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.302360196115598e-05,
+      "loss": 0.2212,
+      "step": 15510
+    },
+    {
+      "epoch": 0.736599724635617,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.3010084406266675e-05,
+      "loss": 0.2253,
+      "step": 15515
+    },
+    {
+      "epoch": 0.7368371077244457,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.2996575781446468e-05,
+      "loss": 0.2236,
+      "step": 15520
+    },
+    {
+      "epoch": 0.7370744908132745,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.2983076095020653e-05,
+      "loss": 0.2209,
+      "step": 15525
+    },
+    {
+      "epoch": 0.7373118739021032,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2969585355309021e-05,
+      "loss": 0.2227,
+      "step": 15530
+    },
+    {
+      "epoch": 0.737549256990932,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.295610357062586e-05,
+      "loss": 0.2257,
+      "step": 15535
+    },
+    {
+      "epoch": 0.7377866400797607,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.2942630749279938e-05,
+      "loss": 0.2212,
+      "step": 15540
+    },
+    {
+      "epoch": 0.7380240231685895,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.2929166899574488e-05,
+      "loss": 0.2249,
+      "step": 15545
+    },
+    {
+      "epoch": 0.7382614062574182,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.291571202980722e-05,
+      "loss": 0.227,
+      "step": 15550
+    },
+    {
+      "epoch": 0.738498789346247,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.29022661482703e-05,
+      "loss": 0.2253,
+      "step": 15555
+    },
+    {
+      "epoch": 0.7387361724350757,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.2888829263250381e-05,
+      "loss": 0.2239,
+      "step": 15560
+    },
+    {
+      "epoch": 0.7389735555239044,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.2875401383028546e-05,
+      "loss": 0.2219,
+      "step": 15565
+    },
+    {
+      "epoch": 0.7392109386127332,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.2861982515880333e-05,
+      "loss": 0.2234,
+      "step": 15570
+    },
+    {
+      "epoch": 0.7394483217015619,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.2848572670075734e-05,
+      "loss": 0.2218,
+      "step": 15575
+    },
+    {
+      "epoch": 0.7396857047903908,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.2835171853879167e-05,
+      "loss": 0.2219,
+      "step": 15580
+    },
+    {
+      "epoch": 0.7399230878792195,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.2821780075549516e-05,
+      "loss": 0.2264,
+      "step": 15585
+    },
+    {
+      "epoch": 0.7401604709680483,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.2808397343340039e-05,
+      "loss": 0.2246,
+      "step": 15590
+    },
+    {
+      "epoch": 0.740397854056877,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.2795023665498482e-05,
+      "loss": 0.224,
+      "step": 15595
+    },
+    {
+      "epoch": 0.7406352371457058,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2781659050266965e-05,
+      "loss": 0.2223,
+      "step": 15600
+    },
+    {
+      "epoch": 0.7408726202345345,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.2768303505882062e-05,
+      "loss": 0.2162,
+      "step": 15605
+    },
+    {
+      "epoch": 0.7411100033233633,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.27549570405747e-05,
+      "loss": 0.2233,
+      "step": 15610
+    },
+    {
+      "epoch": 0.741347386412192,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.2741619662570273e-05,
+      "loss": 0.2261,
+      "step": 15615
+    },
+    {
+      "epoch": 0.7415847695010207,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.2728291380088536e-05,
+      "loss": 0.2217,
+      "step": 15620
+    },
+    {
+      "epoch": 0.7418221525898495,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.2714972201343653e-05,
+      "loss": 0.2281,
+      "step": 15625
+    },
+    {
+      "epoch": 0.7420595356786782,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.2701662134544167e-05,
+      "loss": 0.2242,
+      "step": 15630
+    },
+    {
+      "epoch": 0.742296918767507,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2688361187893006e-05,
+      "loss": 0.2233,
+      "step": 15635
+    },
+    {
+      "epoch": 0.7425343018563357,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.2675069369587506e-05,
+      "loss": 0.2232,
+      "step": 15640
+    },
+    {
+      "epoch": 0.7427716849451645,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.2661786687819332e-05,
+      "loss": 0.2242,
+      "step": 15645
+    },
+    {
+      "epoch": 0.7430090680339932,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.264851315077455e-05,
+      "loss": 0.2215,
+      "step": 15650
+    },
+    {
+      "epoch": 0.7432464511228221,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2635248766633573e-05,
+      "loss": 0.22,
+      "step": 15655
+    },
+    {
+      "epoch": 0.7434838342116508,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.2621993543571193e-05,
+      "loss": 0.2235,
+      "step": 15660
+    },
+    {
+      "epoch": 0.7437212173004795,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.2608747489756536e-05,
+      "loss": 0.2234,
+      "step": 15665
+    },
+    {
+      "epoch": 0.7439586003893083,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.259551061335309e-05,
+      "loss": 0.2246,
+      "step": 15670
+    },
+    {
+      "epoch": 0.744195983478137,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.2582282922518674e-05,
+      "loss": 0.2228,
+      "step": 15675
+    },
+    {
+      "epoch": 0.7444333665669658,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.256906442540545e-05,
+      "loss": 0.221,
+      "step": 15680
+    },
+    {
+      "epoch": 0.7446707496557945,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2555855130159933e-05,
+      "loss": 0.2249,
+      "step": 15685
+    },
+    {
+      "epoch": 0.7449081327446233,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.2542655044922944e-05,
+      "loss": 0.2247,
+      "step": 15690
+    },
+    {
+      "epoch": 0.745145515833452,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2529464177829633e-05,
+      "loss": 0.2206,
+      "step": 15695
+    },
+    {
+      "epoch": 0.7453828989222808,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.2516282537009466e-05,
+      "loss": 0.2249,
+      "step": 15700
+    },
+    {
+      "epoch": 0.7456202820111095,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.2503110130586254e-05,
+      "loss": 0.2268,
+      "step": 15705
+    },
+    {
+      "epoch": 0.7458576650999382,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.2489946966678055e-05,
+      "loss": 0.2233,
+      "step": 15710
+    },
+    {
+      "epoch": 0.746095048188767,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.247679305339729e-05,
+      "loss": 0.2218,
+      "step": 15715
+    },
+    {
+      "epoch": 0.7463324312775957,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.2463648398850656e-05,
+      "loss": 0.2243,
+      "step": 15720
+    },
+    {
+      "epoch": 0.7465698143664246,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.2450513011139137e-05,
+      "loss": 0.2209,
+      "step": 15725
+    },
+    {
+      "epoch": 0.7468071974552533,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.2437386898358013e-05,
+      "loss": 0.2239,
+      "step": 15730
+    },
+    {
+      "epoch": 0.7470445805440821,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.2424270068596843e-05,
+      "loss": 0.2228,
+      "step": 15735
+    },
+    {
+      "epoch": 0.7472819636329108,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.2411162529939482e-05,
+      "loss": 0.2274,
+      "step": 15740
+    },
+    {
+      "epoch": 0.7475193467217396,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.2398064290464041e-05,
+      "loss": 0.2195,
+      "step": 15745
+    },
+    {
+      "epoch": 0.7477567298105683,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.2384975358242906e-05,
+      "loss": 0.2252,
+      "step": 15750
+    },
+    {
+      "epoch": 0.7479941128993971,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.2371895741342714e-05,
+      "loss": 0.2203,
+      "step": 15755
+    },
+    {
+      "epoch": 0.7482314959882258,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.2358825447824394e-05,
+      "loss": 0.2262,
+      "step": 15760
+    },
+    {
+      "epoch": 0.7484688790770545,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.2345764485743098e-05,
+      "loss": 0.224,
+      "step": 15765
+    },
+    {
+      "epoch": 0.7487062621658833,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.2332712863148236e-05,
+      "loss": 0.2206,
+      "step": 15770
+    },
+    {
+      "epoch": 0.748943645254712,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.231967058808347e-05,
+      "loss": 0.2224,
+      "step": 15775
+    },
+    {
+      "epoch": 0.7491810283435408,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2306637668586681e-05,
+      "loss": 0.2194,
+      "step": 15780
+    },
+    {
+      "epoch": 0.7494184114323695,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.2293614112690014e-05,
+      "loss": 0.227,
+      "step": 15785
+    },
+    {
+      "epoch": 0.7496557945211983,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.2280599928419825e-05,
+      "loss": 0.2256,
+      "step": 15790
+    },
+    {
+      "epoch": 0.749893177610027,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.2267595123796692e-05,
+      "loss": 0.2207,
+      "step": 15795
+    },
+    {
+      "epoch": 0.7501305606988559,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.2254599706835408e-05,
+      "loss": 0.2223,
+      "step": 15800
+    },
+    {
+      "epoch": 0.7503679437876846,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.2241613685545007e-05,
+      "loss": 0.2254,
+      "step": 15805
+    },
+    {
+      "epoch": 0.7506053268765133,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2228637067928706e-05,
+      "loss": 0.2248,
+      "step": 15810
+    },
+    {
+      "epoch": 0.7508427099653421,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.2215669861983937e-05,
+      "loss": 0.224,
+      "step": 15815
+    },
+    {
+      "epoch": 0.7510800930541708,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.2202712075702332e-05,
+      "loss": 0.2271,
+      "step": 15820
+    },
+    {
+      "epoch": 0.7513174761429996,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.21897637170697e-05,
+      "loss": 0.2235,
+      "step": 15825
+    },
+    {
+      "epoch": 0.7515548592318283,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.2176824794066089e-05,
+      "loss": 0.2241,
+      "step": 15830
+    },
+    {
+      "epoch": 0.7517922423206571,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.216389531466566e-05,
+      "loss": 0.2199,
+      "step": 15835
+    },
+    {
+      "epoch": 0.7520296254094858,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.2150975286836817e-05,
+      "loss": 0.2222,
+      "step": 15840
+    },
+    {
+      "epoch": 0.7522670084983146,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.2138064718542104e-05,
+      "loss": 0.2235,
+      "step": 15845
+    },
+    {
+      "epoch": 0.7525043915871433,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.2125163617738261e-05,
+      "loss": 0.2254,
+      "step": 15850
+    },
+    {
+      "epoch": 0.752741774675972,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.2112271992376152e-05,
+      "loss": 0.222,
+      "step": 15855
+    },
+    {
+      "epoch": 0.7529791577648008,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.2099389850400853e-05,
+      "loss": 0.2215,
+      "step": 15860
+    },
+    {
+      "epoch": 0.7532165408536295,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.2086517199751552e-05,
+      "loss": 0.2231,
+      "step": 15865
+    },
+    {
+      "epoch": 0.7534539239424584,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2073654048361612e-05,
+      "loss": 0.2273,
+      "step": 15870
+    },
+    {
+      "epoch": 0.7536913070312871,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.2060800404158535e-05,
+      "loss": 0.2238,
+      "step": 15875
+    },
+    {
+      "epoch": 0.7539286901201159,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.2047956275063951e-05,
+      "loss": 0.223,
+      "step": 15880
+    },
+    {
+      "epoch": 0.7541660732089446,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.203512166899366e-05,
+      "loss": 0.2223,
+      "step": 15885
+    },
+    {
+      "epoch": 0.7544034562977734,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.2022296593857561e-05,
+      "loss": 0.2173,
+      "step": 15890
+    },
+    {
+      "epoch": 0.7546408393866021,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.2009481057559687e-05,
+      "loss": 0.2262,
+      "step": 15895
+    },
+    {
+      "epoch": 0.7548782224754309,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.1996675067998192e-05,
+      "loss": 0.2275,
+      "step": 15900
+    },
+    {
+      "epoch": 0.7551156055642596,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.198387863306536e-05,
+      "loss": 0.2252,
+      "step": 15905
+    },
+    {
+      "epoch": 0.7553529886530883,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.1971091760647569e-05,
+      "loss": 0.2239,
+      "step": 15910
+    },
+    {
+      "epoch": 0.7555903717419171,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.1958314458625312e-05,
+      "loss": 0.2275,
+      "step": 15915
+    },
+    {
+      "epoch": 0.7558277548307458,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.1945546734873181e-05,
+      "loss": 0.2207,
+      "step": 15920
+    },
+    {
+      "epoch": 0.7560651379195746,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.1932788597259859e-05,
+      "loss": 0.2195,
+      "step": 15925
+    },
+    {
+      "epoch": 0.7563025210084033,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.1920040053648142e-05,
+      "loss": 0.2218,
+      "step": 15930
+    },
+    {
+      "epoch": 0.7565399040972322,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.1907301111894895e-05,
+      "loss": 0.2223,
+      "step": 15935
+    },
+    {
+      "epoch": 0.7567772871860609,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.1894571779851065e-05,
+      "loss": 0.2218,
+      "step": 15940
+    },
+    {
+      "epoch": 0.7570146702748897,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.188185206536168e-05,
+      "loss": 0.2208,
+      "step": 15945
+    },
+    {
+      "epoch": 0.7572520533637184,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.1869141976265848e-05,
+      "loss": 0.2224,
+      "step": 15950
+    },
+    {
+      "epoch": 0.7574894364525471,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.1856441520396742e-05,
+      "loss": 0.2203,
+      "step": 15955
+    },
+    {
+      "epoch": 0.7577268195413759,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.1843750705581589e-05,
+      "loss": 0.2258,
+      "step": 15960
+    },
+    {
+      "epoch": 0.7579642026302046,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.1831069539641685e-05,
+      "loss": 0.2228,
+      "step": 15965
+    },
+    {
+      "epoch": 0.7582015857190334,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.1818398030392364e-05,
+      "loss": 0.2198,
+      "step": 15970
+    },
+    {
+      "epoch": 0.7584389688078621,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.1805736185643038e-05,
+      "loss": 0.2257,
+      "step": 15975
+    },
+    {
+      "epoch": 0.7586763518966909,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.1793084013197126e-05,
+      "loss": 0.2224,
+      "step": 15980
+    },
+    {
+      "epoch": 0.7589137349855196,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.1780441520852117e-05,
+      "loss": 0.2224,
+      "step": 15985
+    },
+    {
+      "epoch": 0.7591511180743484,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.1767808716399506e-05,
+      "loss": 0.2236,
+      "step": 15990
+    },
+    {
+      "epoch": 0.7593885011631771,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.1755185607624863e-05,
+      "loss": 0.2205,
+      "step": 15995
+    },
+    {
+      "epoch": 0.7596258842520058,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.1742572202307716e-05,
+      "loss": 0.2219,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7598632673408346,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.1729968508221674e-05,
+      "loss": 0.2232,
+      "step": 16005
+    },
+    {
+      "epoch": 0.7601006504296633,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.171737453313433e-05,
+      "loss": 0.2192,
+      "step": 16010
+    },
+    {
+      "epoch": 0.7603380335184922,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.1704790284807293e-05,
+      "loss": 0.2222,
+      "step": 16015
+    },
+    {
+      "epoch": 0.7605754166073209,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.1692215770996178e-05,
+      "loss": 0.2272,
+      "step": 16020
+    },
+    {
+      "epoch": 0.7608127996961497,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.1679650999450591e-05,
+      "loss": 0.2219,
+      "step": 16025
+    },
+    {
+      "epoch": 0.7610501827849784,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.1667095977914163e-05,
+      "loss": 0.2237,
+      "step": 16030
+    },
+    {
+      "epoch": 0.7612875658738072,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.1654550714124484e-05,
+      "loss": 0.2185,
+      "step": 16035
+    },
+    {
+      "epoch": 0.7615249489626359,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.1642015215813145e-05,
+      "loss": 0.2216,
+      "step": 16040
+    },
+    {
+      "epoch": 0.7617623320514647,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1629489490705713e-05,
+      "loss": 0.2251,
+      "step": 16045
+    },
+    {
+      "epoch": 0.7619997151402934,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.1616973546521743e-05,
+      "loss": 0.2232,
+      "step": 16050
+    },
+    {
+      "epoch": 0.7622370982291221,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.1604467390974754e-05,
+      "loss": 0.2244,
+      "step": 16055
+    },
+    {
+      "epoch": 0.7624744813179509,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.1591971031772227e-05,
+      "loss": 0.2269,
+      "step": 16060
+    },
+    {
+      "epoch": 0.7627118644067796,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.1579484476615617e-05,
+      "loss": 0.2215,
+      "step": 16065
+    },
+    {
+      "epoch": 0.7629492474956084,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.156700773320032e-05,
+      "loss": 0.2231,
+      "step": 16070
+    },
+    {
+      "epoch": 0.7631866305844371,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.1554540809215712e-05,
+      "loss": 0.2237,
+      "step": 16075
+    },
+    {
+      "epoch": 0.763424013673266,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.1542083712345094e-05,
+      "loss": 0.2259,
+      "step": 16080
+    },
+    {
+      "epoch": 0.7636613967620947,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.152963645026572e-05,
+      "loss": 0.2241,
+      "step": 16085
+    },
+    {
+      "epoch": 0.7638987798509235,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.1517199030648772e-05,
+      "loss": 0.2217,
+      "step": 16090
+    },
+    {
+      "epoch": 0.7641361629397522,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.1504771461159388e-05,
+      "loss": 0.2214,
+      "step": 16095
+    },
+    {
+      "epoch": 0.7643735460285809,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.1492353749456617e-05,
+      "loss": 0.2252,
+      "step": 16100
+    },
+    {
+      "epoch": 0.7646109291174097,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.147994590319344e-05,
+      "loss": 0.2183,
+      "step": 16105
+    },
+    {
+      "epoch": 0.7648483122062384,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.1467547930016753e-05,
+      "loss": 0.2248,
+      "step": 16110
+    },
+    {
+      "epoch": 0.7650856952950672,
+      "grad_norm": 0.5,
+      "learning_rate": 1.1455159837567372e-05,
+      "loss": 0.2248,
+      "step": 16115
+    },
+    {
+      "epoch": 0.7653230783838959,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.1442781633480023e-05,
+      "loss": 0.2222,
+      "step": 16120
+    },
+    {
+      "epoch": 0.7655604614727247,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.1430413325383328e-05,
+      "loss": 0.2268,
+      "step": 16125
+    },
+    {
+      "epoch": 0.7657978445615534,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.1418054920899835e-05,
+      "loss": 0.2222,
+      "step": 16130
+    },
+    {
+      "epoch": 0.7660352276503822,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.1405706427645966e-05,
+      "loss": 0.2207,
+      "step": 16135
+    },
+    {
+      "epoch": 0.7662726107392109,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.1393367853232037e-05,
+      "loss": 0.2253,
+      "step": 16140
+    },
+    {
+      "epoch": 0.7665099938280396,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.1381039205262251e-05,
+      "loss": 0.2246,
+      "step": 16145
+    },
+    {
+      "epoch": 0.7667473769168685,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.1368720491334714e-05,
+      "loss": 0.2198,
+      "step": 16150
+    },
+    {
+      "epoch": 0.7669847600056972,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.1356411719041384e-05,
+      "loss": 0.2227,
+      "step": 16155
+    },
+    {
+      "epoch": 0.767222143094526,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.1344112895968099e-05,
+      "loss": 0.2241,
+      "step": 16160
+    },
+    {
+      "epoch": 0.7674595261833547,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.1331824029694572e-05,
+      "loss": 0.2257,
+      "step": 16165
+    },
+    {
+      "epoch": 0.7676969092721835,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.1319545127794367e-05,
+      "loss": 0.2232,
+      "step": 16170
+    },
+    {
+      "epoch": 0.7679342923610122,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.1307276197834928e-05,
+      "loss": 0.2206,
+      "step": 16175
+    },
+    {
+      "epoch": 0.768171675449841,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.1295017247377535e-05,
+      "loss": 0.2246,
+      "step": 16180
+    },
+    {
+      "epoch": 0.7684090585386697,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.1282768283977322e-05,
+      "loss": 0.22,
+      "step": 16185
+    },
+    {
+      "epoch": 0.7686464416274985,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.1270529315183266e-05,
+      "loss": 0.221,
+      "step": 16190
+    },
+    {
+      "epoch": 0.7688838247163272,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.12583003485382e-05,
+      "loss": 0.2238,
+      "step": 16195
+    },
+    {
+      "epoch": 0.7691212078051559,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.124608139157877e-05,
+      "loss": 0.2247,
+      "step": 16200
+    },
+    {
+      "epoch": 0.7693585908939847,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.1233872451835468e-05,
+      "loss": 0.2269,
+      "step": 16205
+    },
+    {
+      "epoch": 0.7695959739828134,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.1221673536832605e-05,
+      "loss": 0.2242,
+      "step": 16210
+    },
+    {
+      "epoch": 0.7698333570716422,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.1209484654088315e-05,
+      "loss": 0.2251,
+      "step": 16215
+    },
+    {
+      "epoch": 0.770070740160471,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.1197305811114565e-05,
+      "loss": 0.2252,
+      "step": 16220
+    },
+    {
+      "epoch": 0.7703081232492998,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.11851370154171e-05,
+      "loss": 0.2215,
+      "step": 16225
+    },
+    {
+      "epoch": 0.7705455063381285,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.1172978274495507e-05,
+      "loss": 0.2233,
+      "step": 16230
+    },
+    {
+      "epoch": 0.7707828894269573,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.1160829595843153e-05,
+      "loss": 0.2192,
+      "step": 16235
+    },
+    {
+      "epoch": 0.771020272515786,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.1148690986947235e-05,
+      "loss": 0.2229,
+      "step": 16240
+    },
+    {
+      "epoch": 0.7712576556046147,
+      "grad_norm": 0.484375,
+      "learning_rate": 1.1136562455288693e-05,
+      "loss": 0.2242,
+      "step": 16245
+    },
+    {
+      "epoch": 0.7714950386934435,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.1124444008342306e-05,
+      "loss": 0.2286,
+      "step": 16250
+    },
+    {
+      "epoch": 0.7717324217822722,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.1112335653576612e-05,
+      "loss": 0.226,
+      "step": 16255
+    },
+    {
+      "epoch": 0.771969804871101,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.1100237398453933e-05,
+      "loss": 0.2248,
+      "step": 16260
+    },
+    {
+      "epoch": 0.7722071879599297,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.1088149250430374e-05,
+      "loss": 0.2199,
+      "step": 16265
+    },
+    {
+      "epoch": 0.7724445710487585,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.1076071216955793e-05,
+      "loss": 0.2216,
+      "step": 16270
+    },
+    {
+      "epoch": 0.7726819541375872,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.1064003305473845e-05,
+      "loss": 0.2228,
+      "step": 16275
+    },
+    {
+      "epoch": 0.772919337226416,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.1051945523421917e-05,
+      "loss": 0.2271,
+      "step": 16280
+    },
+    {
+      "epoch": 0.7731567203152447,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.1039897878231167e-05,
+      "loss": 0.2242,
+      "step": 16285
+    },
+    {
+      "epoch": 0.7733941034040734,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.1027860377326504e-05,
+      "loss": 0.2226,
+      "step": 16290
+    },
+    {
+      "epoch": 0.7736314864929023,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.101583302812659e-05,
+      "loss": 0.2256,
+      "step": 16295
+    },
+    {
+      "epoch": 0.773868869581731,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.1003815838043824e-05,
+      "loss": 0.2216,
+      "step": 16300
+    },
+    {
+      "epoch": 0.7741062526705598,
+      "grad_norm": 0.51171875,
+      "learning_rate": 1.0991808814484341e-05,
+      "loss": 0.2258,
+      "step": 16305
+    },
+    {
+      "epoch": 0.7743436357593885,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.0979811964848016e-05,
+      "loss": 0.223,
+      "step": 16310
+    },
+    {
+      "epoch": 0.7745810188482173,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.0967825296528453e-05,
+      "loss": 0.2191,
+      "step": 16315
+    },
+    {
+      "epoch": 0.774818401937046,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0955848816912987e-05,
+      "loss": 0.2185,
+      "step": 16320
+    },
+    {
+      "epoch": 0.7750557850258748,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.0943882533382662e-05,
+      "loss": 0.2267,
+      "step": 16325
+    },
+    {
+      "epoch": 0.7752931681147035,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.0931926453312249e-05,
+      "loss": 0.2229,
+      "step": 16330
+    },
+    {
+      "epoch": 0.7755305512035323,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.0919980584070214e-05,
+      "loss": 0.2223,
+      "step": 16335
+    },
+    {
+      "epoch": 0.775767934292361,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.0908044933018757e-05,
+      "loss": 0.2219,
+      "step": 16340
+    },
+    {
+      "epoch": 0.7760053173811897,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.0896119507513757e-05,
+      "loss": 0.2252,
+      "step": 16345
+    },
+    {
+      "epoch": 0.7762427004700185,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.0884204314904802e-05,
+      "loss": 0.2243,
+      "step": 16350
+    },
+    {
+      "epoch": 0.7764800835588472,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.0872299362535173e-05,
+      "loss": 0.2229,
+      "step": 16355
+    },
+    {
+      "epoch": 0.776717466647676,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.086040465774183e-05,
+      "loss": 0.222,
+      "step": 16360
+    },
+    {
+      "epoch": 0.7769548497365047,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.0848520207855439e-05,
+      "loss": 0.2203,
+      "step": 16365
+    },
+    {
+      "epoch": 0.7771922328253336,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.0836646020200314e-05,
+      "loss": 0.2203,
+      "step": 16370
+    },
+    {
+      "epoch": 0.7774296159141623,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.082478210209448e-05,
+      "loss": 0.224,
+      "step": 16375
+    },
+    {
+      "epoch": 0.7776669990029911,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.0812928460849603e-05,
+      "loss": 0.2233,
+      "step": 16380
+    },
+    {
+      "epoch": 0.7779043820918198,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.0801085103771044e-05,
+      "loss": 0.2208,
+      "step": 16385
+    },
+    {
+      "epoch": 0.7781417651806485,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.0789252038157791e-05,
+      "loss": 0.2198,
+      "step": 16390
+    },
+    {
+      "epoch": 0.7783791482694773,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.0777429271302522e-05,
+      "loss": 0.2263,
+      "step": 16395
+    },
+    {
+      "epoch": 0.778616531358306,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0765616810491544e-05,
+      "loss": 0.2252,
+      "step": 16400
+    },
+    {
+      "epoch": 0.7788539144471348,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.0753814663004843e-05,
+      "loss": 0.2274,
+      "step": 16405
+    },
+    {
+      "epoch": 0.7790912975359635,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.0742022836116005e-05,
+      "loss": 0.2241,
+      "step": 16410
+    },
+    {
+      "epoch": 0.7793286806247923,
+      "grad_norm": 0.490234375,
+      "learning_rate": 1.0730241337092287e-05,
+      "loss": 0.2239,
+      "step": 16415
+    },
+    {
+      "epoch": 0.779566063713621,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.071847017319458e-05,
+      "loss": 0.2274,
+      "step": 16420
+    },
+    {
+      "epoch": 0.7798034468024498,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.0706709351677397e-05,
+      "loss": 0.222,
+      "step": 16425
+    },
+    {
+      "epoch": 0.7800408298912785,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.0694958879788874e-05,
+      "loss": 0.2186,
+      "step": 16430
+    },
+    {
+      "epoch": 0.7802782129801072,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.0683218764770766e-05,
+      "loss": 0.2277,
+      "step": 16435
+    },
+    {
+      "epoch": 0.7805155960689361,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.0671489013858473e-05,
+      "loss": 0.2243,
+      "step": 16440
+    },
+    {
+      "epoch": 0.7807529791577648,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.0659769634280976e-05,
+      "loss": 0.2194,
+      "step": 16445
+    },
+    {
+      "epoch": 0.7809903622465936,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.0648060633260872e-05,
+      "loss": 0.2186,
+      "step": 16450
+    },
+    {
+      "epoch": 0.7812277453354223,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.0636362018014371e-05,
+      "loss": 0.223,
+      "step": 16455
+    },
+    {
+      "epoch": 0.7814651284242511,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.0624673795751268e-05,
+      "loss": 0.2229,
+      "step": 16460
+    },
+    {
+      "epoch": 0.7817025115130798,
+      "grad_norm": 0.388671875,
+      "learning_rate": 1.0612995973674972e-05,
+      "loss": 0.2233,
+      "step": 16465
+    },
+    {
+      "epoch": 0.7819398946019086,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.0601328558982468e-05,
+      "loss": 0.2239,
+      "step": 16470
+    },
+    {
+      "epoch": 0.7821772776907373,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.0589671558864328e-05,
+      "loss": 0.2251,
+      "step": 16475
+    },
+    {
+      "epoch": 0.7824146607795661,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.0578024980504706e-05,
+      "loss": 0.2209,
+      "step": 16480
+    },
+    {
+      "epoch": 0.7826520438683948,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.0566388831081341e-05,
+      "loss": 0.2217,
+      "step": 16485
+    },
+    {
+      "epoch": 0.7828894269572235,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.0554763117765543e-05,
+      "loss": 0.2247,
+      "step": 16490
+    },
+    {
+      "epoch": 0.7831268100460523,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.0543147847722174e-05,
+      "loss": 0.2268,
+      "step": 16495
+    },
+    {
+      "epoch": 0.783364193134881,
+      "grad_norm": 0.498046875,
+      "learning_rate": 1.0531543028109681e-05,
+      "loss": 0.2227,
+      "step": 16500
+    },
+    {
+      "epoch": 0.7836015762237099,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.0519948666080052e-05,
+      "loss": 0.2253,
+      "step": 16505
+    },
+    {
+      "epoch": 0.7838389593125386,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.0508364768778859e-05,
+      "loss": 0.2206,
+      "step": 16510
+    },
+    {
+      "epoch": 0.7840763424013674,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.049679134334518e-05,
+      "loss": 0.2266,
+      "step": 16515
+    },
+    {
+      "epoch": 0.7843137254901961,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.0485228396911685e-05,
+      "loss": 0.2216,
+      "step": 16520
+    },
+    {
+      "epoch": 0.7845511085790249,
+      "grad_norm": 0.453125,
+      "learning_rate": 1.0473675936604555e-05,
+      "loss": 0.2216,
+      "step": 16525
+    },
+    {
+      "epoch": 0.7847884916678536,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.0462133969543523e-05,
+      "loss": 0.223,
+      "step": 16530
+    },
+    {
+      "epoch": 0.7850258747566823,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.0450602502841841e-05,
+      "loss": 0.2199,
+      "step": 16535
+    },
+    {
+      "epoch": 0.7852632578455111,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.0439081543606314e-05,
+      "loss": 0.2269,
+      "step": 16540
+    },
+    {
+      "epoch": 0.7855006409343398,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.0427571098937252e-05,
+      "loss": 0.2193,
+      "step": 16545
+    },
+    {
+      "epoch": 0.7857380240231686,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.0416071175928486e-05,
+      "loss": 0.2222,
+      "step": 16550
+    },
+    {
+      "epoch": 0.7859754071119973,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.0404581781667369e-05,
+      "loss": 0.2229,
+      "step": 16555
+    },
+    {
+      "epoch": 0.7862127902008261,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.0393102923234755e-05,
+      "loss": 0.2238,
+      "step": 16560
+    },
+    {
+      "epoch": 0.7864501732896548,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.0381634607705023e-05,
+      "loss": 0.2236,
+      "step": 16565
+    },
+    {
+      "epoch": 0.7866875563784836,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.0370176842146042e-05,
+      "loss": 0.2196,
+      "step": 16570
+    },
+    {
+      "epoch": 0.7869249394673123,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.0358729633619182e-05,
+      "loss": 0.2201,
+      "step": 16575
+    },
+    {
+      "epoch": 0.787162322556141,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.034729298917929e-05,
+      "loss": 0.2246,
+      "step": 16580
+    },
+    {
+      "epoch": 0.7873997056449699,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.033586691587474e-05,
+      "loss": 0.2224,
+      "step": 16585
+    },
+    {
+      "epoch": 0.7876370887337986,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.0324451420747355e-05,
+      "loss": 0.223,
+      "step": 16590
+    },
+    {
+      "epoch": 0.7878744718226274,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.031304651083246e-05,
+      "loss": 0.2237,
+      "step": 16595
+    },
+    {
+      "epoch": 0.7881118549114561,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.0301652193158845e-05,
+      "loss": 0.2242,
+      "step": 16600
+    },
+    {
+      "epoch": 0.7883492380002849,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.029026847474877e-05,
+      "loss": 0.2235,
+      "step": 16605
+    },
+    {
+      "epoch": 0.7885866210891136,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.0278895362617988e-05,
+      "loss": 0.2222,
+      "step": 16610
+    },
+    {
+      "epoch": 0.7888240041779424,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.0267532863775672e-05,
+      "loss": 0.2271,
+      "step": 16615
+    },
+    {
+      "epoch": 0.7890613872667711,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.0256180985224498e-05,
+      "loss": 0.2222,
+      "step": 16620
+    },
+    {
+      "epoch": 0.7892987703555999,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.0244839733960563e-05,
+      "loss": 0.2247,
+      "step": 16625
+    },
+    {
+      "epoch": 0.7895361534444286,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.0233509116973447e-05,
+      "loss": 0.2238,
+      "step": 16630
+    },
+    {
+      "epoch": 0.7897735365332573,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.0222189141246132e-05,
+      "loss": 0.2215,
+      "step": 16635
+    },
+    {
+      "epoch": 0.7900109196220861,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.021087981375509e-05,
+      "loss": 0.2238,
+      "step": 16640
+    },
+    {
+      "epoch": 0.7902483027109148,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.0199581141470194e-05,
+      "loss": 0.2222,
+      "step": 16645
+    },
+    {
+      "epoch": 0.7904856857997437,
+      "grad_norm": 0.4140625,
+      "learning_rate": 1.0188293131354782e-05,
+      "loss": 0.2255,
+      "step": 16650
+    },
+    {
+      "epoch": 0.7907230688885724,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.0177015790365583e-05,
+      "loss": 0.222,
+      "step": 16655
+    },
+    {
+      "epoch": 0.7909604519774012,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.0165749125452775e-05,
+      "loss": 0.2259,
+      "step": 16660
+    },
+    {
+      "epoch": 0.7911978350662299,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.0154493143559962e-05,
+      "loss": 0.2254,
+      "step": 16665
+    },
+    {
+      "epoch": 0.7914352181550587,
+      "grad_norm": 0.439453125,
+      "learning_rate": 1.0143247851624149e-05,
+      "loss": 0.2221,
+      "step": 16670
+    },
+    {
+      "epoch": 0.7916726012438874,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.0132013256575761e-05,
+      "loss": 0.2234,
+      "step": 16675
+    },
+    {
+      "epoch": 0.7919099843327161,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.0120789365338617e-05,
+      "loss": 0.2249,
+      "step": 16680
+    },
+    {
+      "epoch": 0.7921473674215449,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.0109576184829963e-05,
+      "loss": 0.22,
+      "step": 16685
+    },
+    {
+      "epoch": 0.7923847505103736,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.0098373721960429e-05,
+      "loss": 0.2234,
+      "step": 16690
+    },
+    {
+      "epoch": 0.7926221335992024,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.0087181983634039e-05,
+      "loss": 0.2227,
+      "step": 16695
+    },
+    {
+      "epoch": 0.7928595166880311,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.007600097674821e-05,
+      "loss": 0.2213,
+      "step": 16700
+    },
+    {
+      "epoch": 0.7930968997768599,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.0064830708193738e-05,
+      "loss": 0.2221,
+      "step": 16705
+    },
+    {
+      "epoch": 0.7933342828656886,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.0053671184854824e-05,
+      "loss": 0.2214,
+      "step": 16710
+    },
+    {
+      "epoch": 0.7935716659545174,
+      "grad_norm": 0.462890625,
+      "learning_rate": 1.0042522413609028e-05,
+      "loss": 0.2214,
+      "step": 16715
+    },
+    {
+      "epoch": 0.7938090490433461,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0031384401327283e-05,
+      "loss": 0.2235,
+      "step": 16720
+    },
+    {
+      "epoch": 0.7940464321321749,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.0020257154873888e-05,
+      "loss": 0.2241,
+      "step": 16725
+    },
+    {
+      "epoch": 0.7942838152210037,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.0009140681106531e-05,
+      "loss": 0.2208,
+      "step": 16730
+    },
+    {
+      "epoch": 0.7945211983098324,
+      "grad_norm": 0.4765625,
+      "learning_rate": 9.998034986876236e-06,
+      "loss": 0.224,
+      "step": 16735
+    },
+    {
+      "epoch": 0.7947585813986612,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.986940079027388e-06,
+      "loss": 0.2213,
+      "step": 16740
+    },
+    {
+      "epoch": 0.7949959644874899,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.975855964397735e-06,
+      "loss": 0.2231,
+      "step": 16745
+    },
+    {
+      "epoch": 0.7952333475763187,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.964782649818355e-06,
+      "loss": 0.2219,
+      "step": 16750
+    },
+    {
+      "epoch": 0.7954707306651474,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.953720142113705e-06,
+      "loss": 0.2242,
+      "step": 16755
+    },
+    {
+      "epoch": 0.7957081137539762,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.942668448101528e-06,
+      "loss": 0.2245,
+      "step": 16760
+    },
+    {
+      "epoch": 0.7959454968428049,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.931627574592952e-06,
+      "loss": 0.2203,
+      "step": 16765
+    },
+    {
+      "epoch": 0.7961828799316337,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.920597528392412e-06,
+      "loss": 0.2227,
+      "step": 16770
+    },
+    {
+      "epoch": 0.7964202630204624,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.909578316297685e-06,
+      "loss": 0.2204,
+      "step": 16775
+    },
+    {
+      "epoch": 0.7966576461092911,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.898569945099839e-06,
+      "loss": 0.2258,
+      "step": 16780
+    },
+    {
+      "epoch": 0.7968950291981199,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.887572421583306e-06,
+      "loss": 0.2227,
+      "step": 16785
+    },
+    {
+      "epoch": 0.7971324122869486,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.87658575252579e-06,
+      "loss": 0.2212,
+      "step": 16790
+    },
+    {
+      "epoch": 0.7973697953757775,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.86560994469835e-06,
+      "loss": 0.2177,
+      "step": 16795
+    },
+    {
+      "epoch": 0.7976071784646062,
+      "grad_norm": 0.462890625,
+      "learning_rate": 9.854645004865307e-06,
+      "loss": 0.2267,
+      "step": 16800
+    },
+    {
+      "epoch": 0.797844561553435,
+      "grad_norm": 0.5,
+      "learning_rate": 9.8436909397843e-06,
+      "loss": 0.2258,
+      "step": 16805
+    },
+    {
+      "epoch": 0.7980819446422637,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.832747756206285e-06,
+      "loss": 0.2235,
+      "step": 16810
+    },
+    {
+      "epoch": 0.7983193277310925,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.821815460875491e-06,
+      "loss": 0.2228,
+      "step": 16815
+    },
+    {
+      "epoch": 0.7985567108199212,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.810894060529438e-06,
+      "loss": 0.23,
+      "step": 16820
+    },
+    {
+      "epoch": 0.7987940939087499,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.79998356189893e-06,
+      "loss": 0.223,
+      "step": 16825
+    },
+    {
+      "epoch": 0.7990314769975787,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.789083971708077e-06,
+      "loss": 0.2265,
+      "step": 16830
+    },
+    {
+      "epoch": 0.7992688600864074,
+      "grad_norm": 0.470703125,
+      "learning_rate": 9.77819529667423e-06,
+      "loss": 0.223,
+      "step": 16835
+    },
+    {
+      "epoch": 0.7995062431752362,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.767317543508037e-06,
+      "loss": 0.221,
+      "step": 16840
+    },
+    {
+      "epoch": 0.7997436262640649,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.756450718913404e-06,
+      "loss": 0.2215,
+      "step": 16845
+    },
+    {
+      "epoch": 0.7999810093528937,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.745594829587501e-06,
+      "loss": 0.2206,
+      "step": 16850
+    },
+    {
+      "epoch": 0.8002183924417224,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.734749882220779e-06,
+      "loss": 0.2241,
+      "step": 16855
+    },
+    {
+      "epoch": 0.8004557755305513,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.723915883496917e-06,
+      "loss": 0.2199,
+      "step": 16860
+    },
+    {
+      "epoch": 0.80069315861938,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.713092840092864e-06,
+      "loss": 0.2233,
+      "step": 16865
+    },
+    {
+      "epoch": 0.8009305417082087,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.702280758678806e-06,
+      "loss": 0.2226,
+      "step": 16870
+    },
+    {
+      "epoch": 0.8011679247970375,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.691479645918189e-06,
+      "loss": 0.2258,
+      "step": 16875
+    },
+    {
+      "epoch": 0.8014053078858662,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.680689508467685e-06,
+      "loss": 0.2273,
+      "step": 16880
+    },
+    {
+      "epoch": 0.801642690974695,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.669910352977204e-06,
+      "loss": 0.2226,
+      "step": 16885
+    },
+    {
+      "epoch": 0.8018800740635237,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.659142186089893e-06,
+      "loss": 0.2249,
+      "step": 16890
+    },
+    {
+      "epoch": 0.8021174571523525,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.648385014442127e-06,
+      "loss": 0.2236,
+      "step": 16895
+    },
+    {
+      "epoch": 0.8023548402411812,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.637638844663504e-06,
+      "loss": 0.2237,
+      "step": 16900
+    },
+    {
+      "epoch": 0.80259222333001,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.626903683376828e-06,
+      "loss": 0.2256,
+      "step": 16905
+    },
+    {
+      "epoch": 0.8028296064188387,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.616179537198141e-06,
+      "loss": 0.2197,
+      "step": 16910
+    },
+    {
+      "epoch": 0.8030669895076675,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.605466412736675e-06,
+      "loss": 0.2232,
+      "step": 16915
+    },
+    {
+      "epoch": 0.8033043725964962,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.594764316594901e-06,
+      "loss": 0.2232,
+      "step": 16920
+    },
+    {
+      "epoch": 0.8035417556853249,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.584073255368444e-06,
+      "loss": 0.2246,
+      "step": 16925
+    },
+    {
+      "epoch": 0.8037791387741537,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.573393235646177e-06,
+      "loss": 0.2245,
+      "step": 16930
+    },
+    {
+      "epoch": 0.8040165218629824,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.562724264010139e-06,
+      "loss": 0.2262,
+      "step": 16935
+    },
+    {
+      "epoch": 0.8042539049518113,
+      "grad_norm": 0.478515625,
+      "learning_rate": 9.552066347035573e-06,
+      "loss": 0.2242,
+      "step": 16940
+    },
+    {
+      "epoch": 0.80449128804064,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.541419491290902e-06,
+      "loss": 0.2255,
+      "step": 16945
+    },
+    {
+      "epoch": 0.8047286711294688,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.530783703337727e-06,
+      "loss": 0.2212,
+      "step": 16950
+    },
+    {
+      "epoch": 0.8049660542182975,
+      "grad_norm": 0.37890625,
+      "learning_rate": 9.52015898973085e-06,
+      "loss": 0.2252,
+      "step": 16955
+    },
+    {
+      "epoch": 0.8052034373071263,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.509545357018232e-06,
+      "loss": 0.2219,
+      "step": 16960
+    },
+    {
+      "epoch": 0.805440820395955,
+      "grad_norm": 0.484375,
+      "learning_rate": 9.498942811740997e-06,
+      "loss": 0.2217,
+      "step": 16965
+    },
+    {
+      "epoch": 0.8056782034847837,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.488351360433447e-06,
+      "loss": 0.2213,
+      "step": 16970
+    },
+    {
+      "epoch": 0.8059155865736125,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.477771009623055e-06,
+      "loss": 0.2195,
+      "step": 16975
+    },
+    {
+      "epoch": 0.8061529696624412,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.467201765830435e-06,
+      "loss": 0.226,
+      "step": 16980
+    },
+    {
+      "epoch": 0.80639035275127,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.456643635569367e-06,
+      "loss": 0.2231,
+      "step": 16985
+    },
+    {
+      "epoch": 0.8066277358400987,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.446096625346778e-06,
+      "loss": 0.2238,
+      "step": 16990
+    },
+    {
+      "epoch": 0.8068651189289275,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.435560741662735e-06,
+      "loss": 0.2282,
+      "step": 16995
+    },
+    {
+      "epoch": 0.8071025020177562,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.425035991010472e-06,
+      "loss": 0.223,
+      "step": 17000
+    },
+    {
+      "epoch": 0.807339885106585,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.414522379876331e-06,
+      "loss": 0.2244,
+      "step": 17005
+    },
+    {
+      "epoch": 0.8075772681954138,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.404019914739808e-06,
+      "loss": 0.2259,
+      "step": 17010
+    },
+    {
+      "epoch": 0.8078146512842425,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.393528602073517e-06,
+      "loss": 0.2215,
+      "step": 17015
+    },
+    {
+      "epoch": 0.8080520343730713,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.383048448343226e-06,
+      "loss": 0.2253,
+      "step": 17020
+    },
+    {
+      "epoch": 0.8082894174619,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.372579460007778e-06,
+      "loss": 0.2215,
+      "step": 17025
+    },
+    {
+      "epoch": 0.8085268005507288,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.362121643519186e-06,
+      "loss": 0.2223,
+      "step": 17030
+    },
+    {
+      "epoch": 0.8087641836395575,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.35167500532254e-06,
+      "loss": 0.2225,
+      "step": 17035
+    },
+    {
+      "epoch": 0.8090015667283863,
+      "grad_norm": 0.5,
+      "learning_rate": 9.341239551856071e-06,
+      "loss": 0.2242,
+      "step": 17040
+    },
+    {
+      "epoch": 0.809238949817215,
+      "grad_norm": 0.48046875,
+      "learning_rate": 9.33081528955109e-06,
+      "loss": 0.2216,
+      "step": 17045
+    },
+    {
+      "epoch": 0.8094763329060438,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.320402224832018e-06,
+      "loss": 0.2212,
+      "step": 17050
+    },
+    {
+      "epoch": 0.8097137159948725,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.310000364116395e-06,
+      "loss": 0.222,
+      "step": 17055
+    },
+    {
+      "epoch": 0.8099510990837013,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.299609713814827e-06,
+      "loss": 0.2255,
+      "step": 17060
+    },
+    {
+      "epoch": 0.81018848217253,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.289230280331029e-06,
+      "loss": 0.2233,
+      "step": 17065
+    },
+    {
+      "epoch": 0.8104258652613587,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.278862070061798e-06,
+      "loss": 0.2251,
+      "step": 17070
+    },
+    {
+      "epoch": 0.8106632483501875,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.26850508939702e-06,
+      "loss": 0.2226,
+      "step": 17075
+    },
+    {
+      "epoch": 0.8109006314390163,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.25815934471965e-06,
+      "loss": 0.2256,
+      "step": 17080
+    },
+    {
+      "epoch": 0.8111380145278451,
+      "grad_norm": 0.478515625,
+      "learning_rate": 9.247824842405724e-06,
+      "loss": 0.225,
+      "step": 17085
+    },
+    {
+      "epoch": 0.8113753976166738,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.237501588824352e-06,
+      "loss": 0.2205,
+      "step": 17090
+    },
+    {
+      "epoch": 0.8116127807055026,
+      "grad_norm": 0.447265625,
+      "learning_rate": 9.227189590337697e-06,
+      "loss": 0.2234,
+      "step": 17095
+    },
+    {
+      "epoch": 0.8118501637943313,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.216888853301012e-06,
+      "loss": 0.2226,
+      "step": 17100
+    },
+    {
+      "epoch": 0.8120875468831601,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.206599384062585e-06,
+      "loss": 0.2234,
+      "step": 17105
+    },
+    {
+      "epoch": 0.8123249299719888,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.196321188963775e-06,
+      "loss": 0.2273,
+      "step": 17110
+    },
+    {
+      "epoch": 0.8125623130608175,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.186054274338977e-06,
+      "loss": 0.226,
+      "step": 17115
+    },
+    {
+      "epoch": 0.8127996961496463,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.175798646515655e-06,
+      "loss": 0.2237,
+      "step": 17120
+    },
+    {
+      "epoch": 0.813037079238475,
+      "grad_norm": 0.462890625,
+      "learning_rate": 9.165554311814304e-06,
+      "loss": 0.2235,
+      "step": 17125
+    },
+    {
+      "epoch": 0.8132744623273038,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.155321276548454e-06,
+      "loss": 0.2244,
+      "step": 17130
+    },
+    {
+      "epoch": 0.8135118454161325,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.145099547024678e-06,
+      "loss": 0.224,
+      "step": 17135
+    },
+    {
+      "epoch": 0.8137492285049613,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.134889129542593e-06,
+      "loss": 0.2234,
+      "step": 17140
+    },
+    {
+      "epoch": 0.81398661159379,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.124690030394827e-06,
+      "loss": 0.2238,
+      "step": 17145
+    },
+    {
+      "epoch": 0.8142239946826189,
+      "grad_norm": 0.4765625,
+      "learning_rate": 9.11450225586703e-06,
+      "loss": 0.2256,
+      "step": 17150
+    },
+    {
+      "epoch": 0.8144613777714476,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.104325812237893e-06,
+      "loss": 0.2268,
+      "step": 17155
+    },
+    {
+      "epoch": 0.8146987608602763,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.094160705779101e-06,
+      "loss": 0.2211,
+      "step": 17160
+    },
+    {
+      "epoch": 0.8149361439491051,
+      "grad_norm": 0.49609375,
+      "learning_rate": 9.084006942755378e-06,
+      "loss": 0.2235,
+      "step": 17165
+    },
+    {
+      "epoch": 0.8151735270379338,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.073864529424423e-06,
+      "loss": 0.2262,
+      "step": 17170
+    },
+    {
+      "epoch": 0.8154109101267626,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.063733472036973e-06,
+      "loss": 0.2191,
+      "step": 17175
+    },
+    {
+      "epoch": 0.8156482932155913,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.053613776836745e-06,
+      "loss": 0.2225,
+      "step": 17180
+    },
+    {
+      "epoch": 0.8158856763044201,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.043505450060475e-06,
+      "loss": 0.2183,
+      "step": 17185
+    },
+    {
+      "epoch": 0.8161230593932488,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.033408497937865e-06,
+      "loss": 0.2231,
+      "step": 17190
+    },
+    {
+      "epoch": 0.8163604424820776,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.023322926691614e-06,
+      "loss": 0.225,
+      "step": 17195
+    },
+    {
+      "epoch": 0.8165978255709063,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.013248742537433e-06,
+      "loss": 0.2218,
+      "step": 17200
+    },
+    {
+      "epoch": 0.8168352086597351,
+      "grad_norm": 0.46875,
+      "learning_rate": 9.003185951683988e-06,
+      "loss": 0.2198,
+      "step": 17205
+    },
+    {
+      "epoch": 0.8170725917485638,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.993134560332928e-06,
+      "loss": 0.2183,
+      "step": 17210
+    },
+    {
+      "epoch": 0.8173099748373925,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.983094574678877e-06,
+      "loss": 0.2215,
+      "step": 17215
+    },
+    {
+      "epoch": 0.8175473579262214,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.973066000909442e-06,
+      "loss": 0.2232,
+      "step": 17220
+    },
+    {
+      "epoch": 0.81778474101505,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.963048845205182e-06,
+      "loss": 0.2233,
+      "step": 17225
+    },
+    {
+      "epoch": 0.8180221241038789,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.953043113739626e-06,
+      "loss": 0.2253,
+      "step": 17230
+    },
+    {
+      "epoch": 0.8182595071927076,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.943048812679261e-06,
+      "loss": 0.2253,
+      "step": 17235
+    },
+    {
+      "epoch": 0.8184968902815364,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.933065948183525e-06,
+      "loss": 0.2268,
+      "step": 17240
+    },
+    {
+      "epoch": 0.8187342733703651,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.923094526404815e-06,
+      "loss": 0.2227,
+      "step": 17245
+    },
+    {
+      "epoch": 0.8189716564591939,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.913134553488478e-06,
+      "loss": 0.218,
+      "step": 17250
+    },
+    {
+      "epoch": 0.8192090395480226,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.903186035572795e-06,
+      "loss": 0.2254,
+      "step": 17255
+    },
+    {
+      "epoch": 0.8194464226368513,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.893248978788984e-06,
+      "loss": 0.2223,
+      "step": 17260
+    },
+    {
+      "epoch": 0.8196838057256801,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.883323389261224e-06,
+      "loss": 0.2247,
+      "step": 17265
+    },
+    {
+      "epoch": 0.8199211888145088,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.873409273106604e-06,
+      "loss": 0.2234,
+      "step": 17270
+    },
+    {
+      "epoch": 0.8201585719033376,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.863506636435146e-06,
+      "loss": 0.2217,
+      "step": 17275
+    },
+    {
+      "epoch": 0.8203959549921663,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.8536154853498e-06,
+      "loss": 0.2232,
+      "step": 17280
+    },
+    {
+      "epoch": 0.8206333380809951,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.843735825946442e-06,
+      "loss": 0.2235,
+      "step": 17285
+    },
+    {
+      "epoch": 0.8208707211698238,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.833867664313863e-06,
+      "loss": 0.2188,
+      "step": 17290
+    },
+    {
+      "epoch": 0.8211081042586527,
+      "grad_norm": 0.4609375,
+      "learning_rate": 8.824011006533754e-06,
+      "loss": 0.2208,
+      "step": 17295
+    },
+    {
+      "epoch": 0.8213454873474814,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.814165858680737e-06,
+      "loss": 0.2215,
+      "step": 17300
+    },
+    {
+      "epoch": 0.8215828704363101,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.804332226822325e-06,
+      "loss": 0.2269,
+      "step": 17305
+    },
+    {
+      "epoch": 0.8218202535251389,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.794510117018961e-06,
+      "loss": 0.2173,
+      "step": 17310
+    },
+    {
+      "epoch": 0.8220576366139676,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.784699535323943e-06,
+      "loss": 0.2215,
+      "step": 17315
+    },
+    {
+      "epoch": 0.8222950197027964,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.774900487783499e-06,
+      "loss": 0.2223,
+      "step": 17320
+    },
+    {
+      "epoch": 0.8225324027916251,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.76511298043673e-06,
+      "loss": 0.225,
+      "step": 17325
+    },
+    {
+      "epoch": 0.8227697858804539,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.75533701931565e-06,
+      "loss": 0.2259,
+      "step": 17330
+    },
+    {
+      "epoch": 0.8230071689692826,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.745572610445118e-06,
+      "loss": 0.224,
+      "step": 17335
+    },
+    {
+      "epoch": 0.8232445520581114,
+      "grad_norm": 0.40234375,
+      "learning_rate": 8.735819759842903e-06,
+      "loss": 0.2248,
+      "step": 17340
+    },
+    {
+      "epoch": 0.8234819351469401,
+      "grad_norm": 0.484375,
+      "learning_rate": 8.726078473519647e-06,
+      "loss": 0.2299,
+      "step": 17345
+    },
+    {
+      "epoch": 0.8237193182357689,
+      "grad_norm": 0.359375,
+      "learning_rate": 8.716348757478857e-06,
+      "loss": 0.2229,
+      "step": 17350
+    },
+    {
+      "epoch": 0.8239567013245976,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.706630617716913e-06,
+      "loss": 0.2245,
+      "step": 17355
+    },
+    {
+      "epoch": 0.8241940844134263,
+      "grad_norm": 0.384765625,
+      "learning_rate": 8.69692406022305e-06,
+      "loss": 0.2213,
+      "step": 17360
+    },
+    {
+      "epoch": 0.8244314675022552,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.687229090979395e-06,
+      "loss": 0.2235,
+      "step": 17365
+    },
+    {
+      "epoch": 0.8246688505910839,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.677545715960903e-06,
+      "loss": 0.2197,
+      "step": 17370
+    },
+    {
+      "epoch": 0.8249062336799127,
+      "grad_norm": 0.38671875,
+      "learning_rate": 8.667873941135396e-06,
+      "loss": 0.2248,
+      "step": 17375
+    },
+    {
+      "epoch": 0.8251436167687414,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.658213772463536e-06,
+      "loss": 0.2233,
+      "step": 17380
+    },
+    {
+      "epoch": 0.8253809998575702,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.648565215898854e-06,
+      "loss": 0.2211,
+      "step": 17385
+    },
+    {
+      "epoch": 0.8256183829463989,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.638928277387703e-06,
+      "loss": 0.2221,
+      "step": 17390
+    },
+    {
+      "epoch": 0.8258557660352277,
+      "grad_norm": 0.458984375,
+      "learning_rate": 8.62930296286929e-06,
+      "loss": 0.225,
+      "step": 17395
+    },
+    {
+      "epoch": 0.8260931491240564,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.619689278275653e-06,
+      "loss": 0.2226,
+      "step": 17400
+    },
+    {
+      "epoch": 0.8263305322128851,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.61008722953165e-06,
+      "loss": 0.2241,
+      "step": 17405
+    },
+    {
+      "epoch": 0.8265679153017139,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.600496822554995e-06,
+      "loss": 0.2212,
+      "step": 17410
+    },
+    {
+      "epoch": 0.8268052983905426,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.590918063256206e-06,
+      "loss": 0.2256,
+      "step": 17415
+    },
+    {
+      "epoch": 0.8270426814793714,
+      "grad_norm": 0.5078125,
+      "learning_rate": 8.581350957538628e-06,
+      "loss": 0.2261,
+      "step": 17420
+    },
+    {
+      "epoch": 0.8272800645682001,
+      "grad_norm": 0.453125,
+      "learning_rate": 8.571795511298423e-06,
+      "loss": 0.2243,
+      "step": 17425
+    },
+    {
+      "epoch": 0.827517447657029,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.562251730424581e-06,
+      "loss": 0.2247,
+      "step": 17430
+    },
+    {
+      "epoch": 0.8277548307458577,
+      "grad_norm": 0.404296875,
+      "learning_rate": 8.552719620798874e-06,
+      "loss": 0.2245,
+      "step": 17435
+    },
+    {
+      "epoch": 0.8279922138346865,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.543199188295902e-06,
+      "loss": 0.2238,
+      "step": 17440
+    },
+    {
+      "epoch": 0.8282295969235152,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.533690438783074e-06,
+      "loss": 0.2214,
+      "step": 17445
+    },
+    {
+      "epoch": 0.8284669800123439,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.524193378120581e-06,
+      "loss": 0.2212,
+      "step": 17450
+    },
+    {
+      "epoch": 0.8287043631011727,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.514708012161421e-06,
+      "loss": 0.2235,
+      "step": 17455
+    },
+    {
+      "epoch": 0.8289417461900014,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.505234346751378e-06,
+      "loss": 0.2177,
+      "step": 17460
+    },
+    {
+      "epoch": 0.8291791292788302,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.495772387729034e-06,
+      "loss": 0.2218,
+      "step": 17465
+    },
+    {
+      "epoch": 0.8294165123676589,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.486322140925748e-06,
+      "loss": 0.2236,
+      "step": 17470
+    },
+    {
+      "epoch": 0.8296538954564877,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.47688361216567e-06,
+      "loss": 0.2228,
+      "step": 17475
+    },
+    {
+      "epoch": 0.8298912785453164,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.467456807265714e-06,
+      "loss": 0.2207,
+      "step": 17480
+    },
+    {
+      "epoch": 0.8301286616341452,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.458041732035575e-06,
+      "loss": 0.2238,
+      "step": 17485
+    },
+    {
+      "epoch": 0.8303660447229739,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.448638392277732e-06,
+      "loss": 0.2198,
+      "step": 17490
+    },
+    {
+      "epoch": 0.8306034278118027,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.439246793787417e-06,
+      "loss": 0.2217,
+      "step": 17495
+    },
+    {
+      "epoch": 0.8308408109006314,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.429866942352627e-06,
+      "loss": 0.2243,
+      "step": 17500
+    },
+    {
+      "epoch": 0.8310781939894601,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.420498843754118e-06,
+      "loss": 0.2222,
+      "step": 17505
+    },
+    {
+      "epoch": 0.831315577078289,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.411142503765415e-06,
+      "loss": 0.2218,
+      "step": 17510
+    },
+    {
+      "epoch": 0.8315529601671177,
+      "grad_norm": 0.416015625,
+      "learning_rate": 8.401797928152786e-06,
+      "loss": 0.2229,
+      "step": 17515
+    },
+    {
+      "epoch": 0.8317903432559465,
+      "grad_norm": 0.494140625,
+      "learning_rate": 8.392465122675244e-06,
+      "loss": 0.2236,
+      "step": 17520
+    },
+    {
+      "epoch": 0.8320277263447752,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.383144093084558e-06,
+      "loss": 0.2209,
+      "step": 17525
+    },
+    {
+      "epoch": 0.832265109433604,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.373834845125238e-06,
+      "loss": 0.2264,
+      "step": 17530
+    },
+    {
+      "epoch": 0.8325024925224327,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.364537384534537e-06,
+      "loss": 0.2211,
+      "step": 17535
+    },
+    {
+      "epoch": 0.8327398756112615,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.355251717042423e-06,
+      "loss": 0.2236,
+      "step": 17540
+    },
+    {
+      "epoch": 0.8329772587000902,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.345977848371622e-06,
+      "loss": 0.2241,
+      "step": 17545
+    },
+    {
+      "epoch": 0.8332146417889189,
+      "grad_norm": 0.46875,
+      "learning_rate": 8.336715784237567e-06,
+      "loss": 0.2268,
+      "step": 17550
+    },
+    {
+      "epoch": 0.8334520248777477,
+      "grad_norm": 0.4765625,
+      "learning_rate": 8.327465530348444e-06,
+      "loss": 0.2201,
+      "step": 17555
+    },
+    {
+      "epoch": 0.8336894079665764,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.31822709240512e-06,
+      "loss": 0.2255,
+      "step": 17560
+    },
+    {
+      "epoch": 0.8339267910554052,
+      "grad_norm": 0.482421875,
+      "learning_rate": 8.309000476101222e-06,
+      "loss": 0.2248,
+      "step": 17565
+    },
+    {
+      "epoch": 0.8341641741442339,
+      "grad_norm": 0.4296875,
+      "learning_rate": 8.299785687123054e-06,
+      "loss": 0.217,
+      "step": 17570
+    },
+    {
+      "epoch": 0.8344015572330628,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.290582731149667e-06,
+      "loss": 0.2253,
+      "step": 17575
+    },
+    {
+      "epoch": 0.8346389403218915,
+      "grad_norm": 0.462890625,
+      "learning_rate": 8.28139161385279e-06,
+      "loss": 0.2235,
+      "step": 17580
+    },
+    {
+      "epoch": 0.8348763234107203,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.272212340896863e-06,
+      "loss": 0.224,
+      "step": 17585
+    },
+    {
+      "epoch": 0.835113706499549,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.263044917939038e-06,
+      "loss": 0.2233,
+      "step": 17590
+    },
+    {
+      "epoch": 0.8353510895883777,
+      "grad_norm": 0.4375,
+      "learning_rate": 8.253889350629154e-06,
+      "loss": 0.2198,
+      "step": 17595
+    },
+    {
+      "epoch": 0.8355884726772065,
+      "grad_norm": 0.45703125,
+      "learning_rate": 8.244745644609748e-06,
+      "loss": 0.2248,
+      "step": 17600
+    },
+    {
+      "epoch": 0.8358258557660352,
+      "grad_norm": 0.408203125,
+      "learning_rate": 8.235613805516029e-06,
+      "loss": 0.2182,
+      "step": 17605
+    },
+    {
+      "epoch": 0.836063238854864,
+      "grad_norm": 0.484375,
+      "learning_rate": 8.226493838975932e-06,
+      "loss": 0.2234,
+      "step": 17610
+    },
+    {
+      "epoch": 0.8363006219436927,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.217385750610038e-06,
+      "loss": 0.2231,
+      "step": 17615
+    },
+    {
+      "epoch": 0.8365380050325215,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.20828954603162e-06,
+      "loss": 0.2249,
+      "step": 17620
+    },
+    {
+      "epoch": 0.8367753881213502,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.199205230846624e-06,
+      "loss": 0.2228,
+      "step": 17625
+    },
+    {
+      "epoch": 0.837012771210179,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.190132810653686e-06,
+      "loss": 0.224,
+      "step": 17630
+    },
+    {
+      "epoch": 0.8372501542990077,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.181072291044085e-06,
+      "loss": 0.225,
+      "step": 17635
+    },
+    {
+      "epoch": 0.8374875373878365,
+      "grad_norm": 0.412109375,
+      "learning_rate": 8.172023677601784e-06,
+      "loss": 0.2203,
+      "step": 17640
+    },
+    {
+      "epoch": 0.8377249204766652,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.162986975903397e-06,
+      "loss": 0.2231,
+      "step": 17645
+    },
+    {
+      "epoch": 0.837962303565494,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.153962191518205e-06,
+      "loss": 0.2236,
+      "step": 17650
+    },
+    {
+      "epoch": 0.8381996866543228,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.144949330008144e-06,
+      "loss": 0.224,
+      "step": 17655
+    },
+    {
+      "epoch": 0.8384370697431515,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.135948396927797e-06,
+      "loss": 0.2258,
+      "step": 17660
+    },
+    {
+      "epoch": 0.8386744528319803,
+      "grad_norm": 0.423828125,
+      "learning_rate": 8.126959397824399e-06,
+      "loss": 0.2263,
+      "step": 17665
+    },
+    {
+      "epoch": 0.838911835920809,
+      "grad_norm": 0.4921875,
+      "learning_rate": 8.117982338237822e-06,
+      "loss": 0.2263,
+      "step": 17670
+    },
+    {
+      "epoch": 0.8391492190096378,
+      "grad_norm": 0.451171875,
+      "learning_rate": 8.1090172237006e-06,
+      "loss": 0.2259,
+      "step": 17675
+    },
+    {
+      "epoch": 0.8393866020984665,
+      "grad_norm": 0.431640625,
+      "learning_rate": 8.100064059737894e-06,
+      "loss": 0.2276,
+      "step": 17680
+    },
+    {
+      "epoch": 0.8396239851872953,
+      "grad_norm": 0.455078125,
+      "learning_rate": 8.091122851867476e-06,
+      "loss": 0.2225,
+      "step": 17685
+    },
+    {
+      "epoch": 0.839861368276124,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.082193605599792e-06,
+      "loss": 0.2276,
+      "step": 17690
+    },
+    {
+      "epoch": 0.8400987513649527,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.073276326437886e-06,
+      "loss": 0.2235,
+      "step": 17695
+    },
+    {
+      "epoch": 0.8403361344537815,
+      "grad_norm": 0.4453125,
+      "learning_rate": 8.064371019877448e-06,
+      "loss": 0.2252,
+      "step": 17700
+    },
+    {
+      "epoch": 0.8405735175426102,
+      "grad_norm": 0.44921875,
+      "learning_rate": 8.055477691406764e-06,
+      "loss": 0.2277,
+      "step": 17705
+    },
+    {
+      "epoch": 0.840810900631439,
+      "grad_norm": 0.427734375,
+      "learning_rate": 8.04659634650676e-06,
+      "loss": 0.222,
+      "step": 17710
+    },
+    {
+      "epoch": 0.8410482837202677,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.037726990650963e-06,
+      "loss": 0.223,
+      "step": 17715
+    },
+    {
+      "epoch": 0.8412856668090966,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.028869629305533e-06,
+      "loss": 0.224,
+      "step": 17720
+    },
+    {
+      "epoch": 0.8415230498979253,
+      "grad_norm": 0.46484375,
+      "learning_rate": 8.0200242679292e-06,
+      "loss": 0.2201,
+      "step": 17725
+    },
+    {
+      "epoch": 0.8417604329867541,
+      "grad_norm": 0.474609375,
+      "learning_rate": 8.011190911973328e-06,
+      "loss": 0.2221,
+      "step": 17730
+    },
+    {
+      "epoch": 0.8419978160755828,
+      "grad_norm": 0.478515625,
+      "learning_rate": 8.002369566881877e-06,
+      "loss": 0.2231,
+      "step": 17735
+    },
+    {
+      "epoch": 0.8422351991644115,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.9935602380914e-06,
+      "loss": 0.2199,
+      "step": 17740
+    },
+    {
+      "epoch": 0.8424725822532403,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.984762931031048e-06,
+      "loss": 0.2222,
+      "step": 17745
+    },
+    {
+      "epoch": 0.842709965342069,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.975977651122555e-06,
+      "loss": 0.2266,
+      "step": 17750
+    },
+    {
+      "epoch": 0.8429473484308978,
+      "grad_norm": 0.376953125,
+      "learning_rate": 7.967204403780258e-06,
+      "loss": 0.2213,
+      "step": 17755
+    },
+    {
+      "epoch": 0.8431847315197265,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.958443194411062e-06,
+      "loss": 0.223,
+      "step": 17760
+    },
+    {
+      "epoch": 0.8434221146085553,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.949694028414467e-06,
+      "loss": 0.2224,
+      "step": 17765
+    },
+    {
+      "epoch": 0.843659497697384,
+      "grad_norm": 0.4609375,
+      "learning_rate": 7.940956911182537e-06,
+      "loss": 0.2257,
+      "step": 17770
+    },
+    {
+      "epoch": 0.8438968807862128,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.932231848099924e-06,
+      "loss": 0.2248,
+      "step": 17775
+    },
+    {
+      "epoch": 0.8441342638750415,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.923518844543842e-06,
+      "loss": 0.2245,
+      "step": 17780
+    },
+    {
+      "epoch": 0.8443716469638703,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.914817905884077e-06,
+      "loss": 0.2233,
+      "step": 17785
+    },
+    {
+      "epoch": 0.844609030052699,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.906129037482975e-06,
+      "loss": 0.2216,
+      "step": 17790
+    },
+    {
+      "epoch": 0.8448464131415278,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.897452244695442e-06,
+      "loss": 0.2202,
+      "step": 17795
+    },
+    {
+      "epoch": 0.8450837962303566,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.888787532868958e-06,
+      "loss": 0.2218,
+      "step": 17800
+    },
+    {
+      "epoch": 0.8453211793191853,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.880134907343539e-06,
+      "loss": 0.2208,
+      "step": 17805
+    },
+    {
+      "epoch": 0.8455585624080141,
+      "grad_norm": 0.38671875,
+      "learning_rate": 7.87149437345176e-06,
+      "loss": 0.2224,
+      "step": 17810
+    },
+    {
+      "epoch": 0.8457959454968428,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.862865936518735e-06,
+      "loss": 0.222,
+      "step": 17815
+    },
+    {
+      "epoch": 0.8460333285856716,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.854249601862147e-06,
+      "loss": 0.2184,
+      "step": 17820
+    },
+    {
+      "epoch": 0.8462707116745003,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.845645374792189e-06,
+      "loss": 0.2207,
+      "step": 17825
+    },
+    {
+      "epoch": 0.8465080947633291,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.837053260611607e-06,
+      "loss": 0.2233,
+      "step": 17830
+    },
+    {
+      "epoch": 0.8467454778521578,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.828473264615692e-06,
+      "loss": 0.2228,
+      "step": 17835
+    },
+    {
+      "epoch": 0.8469828609409865,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.819905392092252e-06,
+      "loss": 0.2232,
+      "step": 17840
+    },
+    {
+      "epoch": 0.8472202440298153,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.811349648321627e-06,
+      "loss": 0.2221,
+      "step": 17845
+    },
+    {
+      "epoch": 0.847457627118644,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.802806038576676e-06,
+      "loss": 0.2232,
+      "step": 17850
+    },
+    {
+      "epoch": 0.8476950102074728,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.7942745681228e-06,
+      "loss": 0.225,
+      "step": 17855
+    },
+    {
+      "epoch": 0.8479323932963015,
+      "grad_norm": 0.47265625,
+      "learning_rate": 7.7857552422179e-06,
+      "loss": 0.2211,
+      "step": 17860
+    },
+    {
+      "epoch": 0.8481697763851304,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.777248066112397e-06,
+      "loss": 0.2183,
+      "step": 17865
+    },
+    {
+      "epoch": 0.8484071594739591,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.768753045049223e-06,
+      "loss": 0.2261,
+      "step": 17870
+    },
+    {
+      "epoch": 0.8486445425627879,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.760270184263824e-06,
+      "loss": 0.2197,
+      "step": 17875
+    },
+    {
+      "epoch": 0.8488819256516166,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.75179948898415e-06,
+      "loss": 0.2234,
+      "step": 17880
+    },
+    {
+      "epoch": 0.8491193087404453,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.743340964430647e-06,
+      "loss": 0.218,
+      "step": 17885
+    },
+    {
+      "epoch": 0.8493566918292741,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.734894615816271e-06,
+      "loss": 0.2283,
+      "step": 17890
+    },
+    {
+      "epoch": 0.8495940749181028,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.726460448346457e-06,
+      "loss": 0.2262,
+      "step": 17895
+    },
+    {
+      "epoch": 0.8498314580069316,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.718038467219159e-06,
+      "loss": 0.2255,
+      "step": 17900
+    },
+    {
+      "epoch": 0.8500688410957603,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.709628677624799e-06,
+      "loss": 0.2228,
+      "step": 17905
+    },
+    {
+      "epoch": 0.8503062241845891,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.701231084746292e-06,
+      "loss": 0.2241,
+      "step": 17910
+    },
+    {
+      "epoch": 0.8505436072734178,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.692845693759029e-06,
+      "loss": 0.2264,
+      "step": 17915
+    },
+    {
+      "epoch": 0.8507809903622466,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.6844725098309e-06,
+      "loss": 0.2235,
+      "step": 17920
+    },
+    {
+      "epoch": 0.8510183734510753,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.676111538122262e-06,
+      "loss": 0.2205,
+      "step": 17925
+    },
+    {
+      "epoch": 0.8512557565399042,
+      "grad_norm": 0.498046875,
+      "learning_rate": 7.667762783785928e-06,
+      "loss": 0.2261,
+      "step": 17930
+    },
+    {
+      "epoch": 0.8514931396287329,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.659426251967212e-06,
+      "loss": 0.2252,
+      "step": 17935
+    },
+    {
+      "epoch": 0.8517305227175616,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.65110194780387e-06,
+      "loss": 0.2218,
+      "step": 17940
+    },
+    {
+      "epoch": 0.8519679058063904,
+      "grad_norm": 0.380859375,
+      "learning_rate": 7.642789876426148e-06,
+      "loss": 0.2216,
+      "step": 17945
+    },
+    {
+      "epoch": 0.8522052888952191,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.634490042956719e-06,
+      "loss": 0.2227,
+      "step": 17950
+    },
+    {
+      "epoch": 0.8524426719840479,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.6262024525107456e-06,
+      "loss": 0.22,
+      "step": 17955
+    },
+    {
+      "epoch": 0.8526800550728766,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.617927110195825e-06,
+      "loss": 0.2224,
+      "step": 17960
+    },
+    {
+      "epoch": 0.8529174381617054,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.609664021112024e-06,
+      "loss": 0.2201,
+      "step": 17965
+    },
+    {
+      "epoch": 0.8531548212505341,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.6014131903518375e-06,
+      "loss": 0.2239,
+      "step": 17970
+    },
+    {
+      "epoch": 0.8533922043393629,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.593174623000209e-06,
+      "loss": 0.2222,
+      "step": 17975
+    },
+    {
+      "epoch": 0.8536295874281916,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.584948324134543e-06,
+      "loss": 0.2233,
+      "step": 17980
+    },
+    {
+      "epoch": 0.8538669705170203,
+      "grad_norm": 0.39453125,
+      "learning_rate": 7.576734298824666e-06,
+      "loss": 0.2214,
+      "step": 17985
+    },
+    {
+      "epoch": 0.8541043536058491,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.56853255213284e-06,
+      "loss": 0.2234,
+      "step": 17990
+    },
+    {
+      "epoch": 0.8543417366946778,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.5603430891137635e-06,
+      "loss": 0.2206,
+      "step": 17995
+    },
+    {
+      "epoch": 0.8545791197835066,
+      "grad_norm": 0.48828125,
+      "learning_rate": 7.552165914814576e-06,
+      "loss": 0.2241,
+      "step": 18000
+    },
+    {
+      "epoch": 0.8548165028723353,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.5440010342748235e-06,
+      "loss": 0.2221,
+      "step": 18005
+    },
+    {
+      "epoch": 0.8550538859611642,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.535848452526489e-06,
+      "loss": 0.2257,
+      "step": 18010
+    },
+    {
+      "epoch": 0.8552912690499929,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.527708174593966e-06,
+      "loss": 0.2255,
+      "step": 18015
+    },
+    {
+      "epoch": 0.8555286521388217,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.519580205494079e-06,
+      "loss": 0.2227,
+      "step": 18020
+    },
+    {
+      "epoch": 0.8557660352276504,
+      "grad_norm": 0.427734375,
+      "learning_rate": 7.511464550236052e-06,
+      "loss": 0.2206,
+      "step": 18025
+    },
+    {
+      "epoch": 0.8560034183164791,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.50336121382153e-06,
+      "loss": 0.2197,
+      "step": 18030
+    },
+    {
+      "epoch": 0.8562408014053079,
+      "grad_norm": 0.4765625,
+      "learning_rate": 7.49527020124456e-06,
+      "loss": 0.2218,
+      "step": 18035
+    },
+    {
+      "epoch": 0.8564781844941366,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.487191517491593e-06,
+      "loss": 0.221,
+      "step": 18040
+    },
+    {
+      "epoch": 0.8567155675829654,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.4791251675414925e-06,
+      "loss": 0.2201,
+      "step": 18045
+    },
+    {
+      "epoch": 0.8569529506717941,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.471071156365511e-06,
+      "loss": 0.2204,
+      "step": 18050
+    },
+    {
+      "epoch": 0.8571903337606229,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.463029488927299e-06,
+      "loss": 0.2243,
+      "step": 18055
+    },
+    {
+      "epoch": 0.8574277168494516,
+      "grad_norm": 0.5,
+      "learning_rate": 7.455000170182892e-06,
+      "loss": 0.2224,
+      "step": 18060
+    },
+    {
+      "epoch": 0.8576650999382804,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.446983205080735e-06,
+      "loss": 0.2212,
+      "step": 18065
+    },
+    {
+      "epoch": 0.8579024830271091,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.4389785985616454e-06,
+      "loss": 0.2211,
+      "step": 18070
+    },
+    {
+      "epoch": 0.858139866115938,
+      "grad_norm": 0.388671875,
+      "learning_rate": 7.4309863555588124e-06,
+      "loss": 0.2256,
+      "step": 18075
+    },
+    {
+      "epoch": 0.8583772492047667,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.4230064809978325e-06,
+      "loss": 0.2253,
+      "step": 18080
+    },
+    {
+      "epoch": 0.8586146322935954,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.415038979796657e-06,
+      "loss": 0.222,
+      "step": 18085
+    },
+    {
+      "epoch": 0.8588520153824242,
+      "grad_norm": 0.439453125,
+      "learning_rate": 7.407083856865636e-06,
+      "loss": 0.2245,
+      "step": 18090
+    },
+    {
+      "epoch": 0.8590893984712529,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.399141117107454e-06,
+      "loss": 0.2201,
+      "step": 18095
+    },
+    {
+      "epoch": 0.8593267815600817,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.391210765417203e-06,
+      "loss": 0.2233,
+      "step": 18100
+    },
+    {
+      "epoch": 0.8595641646489104,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.38329280668231e-06,
+      "loss": 0.2221,
+      "step": 18105
+    },
+    {
+      "epoch": 0.8598015477377392,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.375387245782595e-06,
+      "loss": 0.2227,
+      "step": 18110
+    },
+    {
+      "epoch": 0.8600389308265679,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.367494087590196e-06,
+      "loss": 0.2218,
+      "step": 18115
+    },
+    {
+      "epoch": 0.8602763139153967,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.359613336969648e-06,
+      "loss": 0.2216,
+      "step": 18120
+    },
+    {
+      "epoch": 0.8605136970042254,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.351744998777814e-06,
+      "loss": 0.2232,
+      "step": 18125
+    },
+    {
+      "epoch": 0.8607510800930541,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.343889077863915e-06,
+      "loss": 0.223,
+      "step": 18130
+    },
+    {
+      "epoch": 0.8609884631818829,
+      "grad_norm": 0.50390625,
+      "learning_rate": 7.336045579069525e-06,
+      "loss": 0.2169,
+      "step": 18135
+    },
+    {
+      "epoch": 0.8612258462707116,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.328214507228546e-06,
+      "loss": 0.2221,
+      "step": 18140
+    },
+    {
+      "epoch": 0.8614632293595404,
+      "grad_norm": 0.431640625,
+      "learning_rate": 7.320395867167243e-06,
+      "loss": 0.2219,
+      "step": 18145
+    },
+    {
+      "epoch": 0.8617006124483692,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.312589663704206e-06,
+      "loss": 0.2231,
+      "step": 18150
+    },
+    {
+      "epoch": 0.861937995537198,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.304795901650357e-06,
+      "loss": 0.2243,
+      "step": 18155
+    },
+    {
+      "epoch": 0.8621753786260267,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.297014585808959e-06,
+      "loss": 0.225,
+      "step": 18160
+    },
+    {
+      "epoch": 0.8624127617148555,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.289245720975606e-06,
+      "loss": 0.2229,
+      "step": 18165
+    },
+    {
+      "epoch": 0.8626501448036842,
+      "grad_norm": 0.5,
+      "learning_rate": 7.281489311938208e-06,
+      "loss": 0.2268,
+      "step": 18170
+    },
+    {
+      "epoch": 0.8628875278925129,
+      "grad_norm": 0.4453125,
+      "learning_rate": 7.273745363477011e-06,
+      "loss": 0.2248,
+      "step": 18175
+    },
+    {
+      "epoch": 0.8631249109813417,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.266013880364568e-06,
+      "loss": 0.2206,
+      "step": 18180
+    },
+    {
+      "epoch": 0.8633622940701704,
+      "grad_norm": 0.486328125,
+      "learning_rate": 7.258294867365757e-06,
+      "loss": 0.2201,
+      "step": 18185
+    },
+    {
+      "epoch": 0.8635996771589992,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.250588329237777e-06,
+      "loss": 0.2224,
+      "step": 18190
+    },
+    {
+      "epoch": 0.8638370602478279,
+      "grad_norm": 0.45703125,
+      "learning_rate": 7.24289427073013e-06,
+      "loss": 0.2262,
+      "step": 18195
+    },
+    {
+      "epoch": 0.8640744433366567,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.23521269658463e-06,
+      "loss": 0.2216,
+      "step": 18200
+    },
+    {
+      "epoch": 0.8643118264254854,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.227543611535388e-06,
+      "loss": 0.2233,
+      "step": 18205
+    },
+    {
+      "epoch": 0.8645492095143142,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.219887020308838e-06,
+      "loss": 0.2226,
+      "step": 18210
+    },
+    {
+      "epoch": 0.864786592603143,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.212242927623701e-06,
+      "loss": 0.2254,
+      "step": 18215
+    },
+    {
+      "epoch": 0.8650239756919718,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.204611338190984e-06,
+      "loss": 0.2259,
+      "step": 18220
+    },
+    {
+      "epoch": 0.8652613587808005,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.196992256714015e-06,
+      "loss": 0.2208,
+      "step": 18225
+    },
+    {
+      "epoch": 0.8654987418696292,
+      "grad_norm": 0.44140625,
+      "learning_rate": 7.189385687888395e-06,
+      "loss": 0.2219,
+      "step": 18230
+    },
+    {
+      "epoch": 0.865736124958458,
+      "grad_norm": 0.3984375,
+      "learning_rate": 7.181791636402013e-06,
+      "loss": 0.2191,
+      "step": 18235
+    },
+    {
+      "epoch": 0.8659735080472867,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.174210106935051e-06,
+      "loss": 0.2225,
+      "step": 18240
+    },
+    {
+      "epoch": 0.8662108911361155,
+      "grad_norm": 0.451171875,
+      "learning_rate": 7.166641104159977e-06,
+      "loss": 0.2213,
+      "step": 18245
+    },
+    {
+      "epoch": 0.8664482742249442,
+      "grad_norm": 0.51171875,
+      "learning_rate": 7.1590846327415265e-06,
+      "loss": 0.2253,
+      "step": 18250
+    },
+    {
+      "epoch": 0.866685657313773,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.15154069733672e-06,
+      "loss": 0.2252,
+      "step": 18255
+    },
+    {
+      "epoch": 0.8669230404026017,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.144009302594848e-06,
+      "loss": 0.2231,
+      "step": 18260
+    },
+    {
+      "epoch": 0.8671604234914305,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.13649045315748e-06,
+      "loss": 0.227,
+      "step": 18265
+    },
+    {
+      "epoch": 0.8673978065802592,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.128984153658448e-06,
+      "loss": 0.2229,
+      "step": 18270
+    },
+    {
+      "epoch": 0.8676351896690879,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.121490408723847e-06,
+      "loss": 0.2244,
+      "step": 18275
+    },
+    {
+      "epoch": 0.8678725727579167,
+      "grad_norm": 0.376953125,
+      "learning_rate": 7.11400922297204e-06,
+      "loss": 0.2229,
+      "step": 18280
+    },
+    {
+      "epoch": 0.8681099558467454,
+      "grad_norm": 0.40234375,
+      "learning_rate": 7.106540601013644e-06,
+      "loss": 0.2253,
+      "step": 18285
+    },
+    {
+      "epoch": 0.8683473389355743,
+      "grad_norm": 0.466796875,
+      "learning_rate": 7.0990845474515426e-06,
+      "loss": 0.221,
+      "step": 18290
+    },
+    {
+      "epoch": 0.868584722024403,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.091641066880867e-06,
+      "loss": 0.2182,
+      "step": 18295
+    },
+    {
+      "epoch": 0.8688221051132318,
+      "grad_norm": 0.40625,
+      "learning_rate": 7.084210163888999e-06,
+      "loss": 0.2235,
+      "step": 18300
+    },
+    {
+      "epoch": 0.8690594882020605,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.076791843055562e-06,
+      "loss": 0.2222,
+      "step": 18305
+    },
+    {
+      "epoch": 0.8692968712908893,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.069386108952448e-06,
+      "loss": 0.2239,
+      "step": 18310
+    },
+    {
+      "epoch": 0.869534254379718,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.061992966143775e-06,
+      "loss": 0.2234,
+      "step": 18315
+    },
+    {
+      "epoch": 0.8697716374685467,
+      "grad_norm": 0.396484375,
+      "learning_rate": 7.054612419185891e-06,
+      "loss": 0.2275,
+      "step": 18320
+    },
+    {
+      "epoch": 0.8700090205573755,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.047244472627404e-06,
+      "loss": 0.2264,
+      "step": 18325
+    },
+    {
+      "epoch": 0.8702464036462042,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.039889131009141e-06,
+      "loss": 0.2227,
+      "step": 18330
+    },
+    {
+      "epoch": 0.870483786735033,
+      "grad_norm": 0.4921875,
+      "learning_rate": 7.032546398864174e-06,
+      "loss": 0.2241,
+      "step": 18335
+    },
+    {
+      "epoch": 0.8707211698238617,
+      "grad_norm": 0.49609375,
+      "learning_rate": 7.025216280717783e-06,
+      "loss": 0.2264,
+      "step": 18340
+    },
+    {
+      "epoch": 0.8709585529126905,
+      "grad_norm": 0.400390625,
+      "learning_rate": 7.017898781087495e-06,
+      "loss": 0.2267,
+      "step": 18345
+    },
+    {
+      "epoch": 0.8711959360015192,
+      "grad_norm": 0.447265625,
+      "learning_rate": 7.010593904483047e-06,
+      "loss": 0.2262,
+      "step": 18350
+    },
+    {
+      "epoch": 0.871433319090348,
+      "grad_norm": 0.515625,
+      "learning_rate": 7.003301655406409e-06,
+      "loss": 0.2203,
+      "step": 18355
+    },
+    {
+      "epoch": 0.8716707021791767,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.996022038351756e-06,
+      "loss": 0.2222,
+      "step": 18360
+    },
+    {
+      "epoch": 0.8719080852680056,
+      "grad_norm": 0.4453125,
+      "learning_rate": 6.9887550578054765e-06,
+      "loss": 0.2195,
+      "step": 18365
+    },
+    {
+      "epoch": 0.8721454683568343,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.981500718246187e-06,
+      "loss": 0.2211,
+      "step": 18370
+    },
+    {
+      "epoch": 0.872382851445663,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.974259024144701e-06,
+      "loss": 0.2234,
+      "step": 18375
+    },
+    {
+      "epoch": 0.8726202345344918,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.967029979964043e-06,
+      "loss": 0.2237,
+      "step": 18380
+    },
+    {
+      "epoch": 0.8728576176233205,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.959813590159434e-06,
+      "loss": 0.2225,
+      "step": 18385
+    },
+    {
+      "epoch": 0.8730950007121493,
+      "grad_norm": 0.41015625,
+      "learning_rate": 6.952609859178309e-06,
+      "loss": 0.2243,
+      "step": 18390
+    },
+    {
+      "epoch": 0.873332383800978,
+      "grad_norm": 0.40625,
+      "learning_rate": 6.9454187914602935e-06,
+      "loss": 0.2248,
+      "step": 18395
+    },
+    {
+      "epoch": 0.8735697668898068,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.938240391437209e-06,
+      "loss": 0.2258,
+      "step": 18400
+    },
+    {
+      "epoch": 0.8738071499786355,
+      "grad_norm": 0.443359375,
+      "learning_rate": 6.931074663533065e-06,
+      "loss": 0.2266,
+      "step": 18405
+    },
+    {
+      "epoch": 0.8740445330674643,
+      "grad_norm": 0.45703125,
+      "learning_rate": 6.9239216121640745e-06,
+      "loss": 0.2243,
+      "step": 18410
+    },
+    {
+      "epoch": 0.874281916156293,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.916781241738632e-06,
+      "loss": 0.2212,
+      "step": 18415
+    },
+    {
+      "epoch": 0.8745192992451217,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.90965355665731e-06,
+      "loss": 0.2263,
+      "step": 18420
+    },
+    {
+      "epoch": 0.8747566823339505,
+      "grad_norm": 0.427734375,
+      "learning_rate": 6.902538561312873e-06,
+      "loss": 0.2188,
+      "step": 18425
+    },
+    {
+      "epoch": 0.8749940654227792,
+      "grad_norm": 0.41015625,
+      "learning_rate": 6.89543626009025e-06,
+      "loss": 0.2248,
+      "step": 18430
+    },
+    {
+      "epoch": 0.8752314485116081,
+      "grad_norm": 0.4453125,
+      "learning_rate": 6.8883466573665695e-06,
+      "loss": 0.2228,
+      "step": 18435
+    },
+    {
+      "epoch": 0.8754688316004368,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.8812697575111174e-06,
+      "loss": 0.2249,
+      "step": 18440
+    },
+    {
+      "epoch": 0.8757062146892656,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.8742055648853515e-06,
+      "loss": 0.2261,
+      "step": 18445
+    },
+    {
+      "epoch": 0.8759435977780943,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.867154083842902e-06,
+      "loss": 0.223,
+      "step": 18450
+    },
+    {
+      "epoch": 0.8761809808669231,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.860115318729568e-06,
+      "loss": 0.2245,
+      "step": 18455
+    },
+    {
+      "epoch": 0.8764183639557518,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.853089273883313e-06,
+      "loss": 0.2215,
+      "step": 18460
+    },
+    {
+      "epoch": 0.8766557470445805,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.846075953634241e-06,
+      "loss": 0.2226,
+      "step": 18465
+    },
+    {
+      "epoch": 0.8768931301334093,
+      "grad_norm": 0.4453125,
+      "learning_rate": 6.839075362304645e-06,
+      "loss": 0.2261,
+      "step": 18470
+    },
+    {
+      "epoch": 0.877130513222238,
+      "grad_norm": 0.443359375,
+      "learning_rate": 6.832087504208947e-06,
+      "loss": 0.2208,
+      "step": 18475
+    },
+    {
+      "epoch": 0.8773678963110668,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.825112383653742e-06,
+      "loss": 0.2222,
+      "step": 18480
+    },
+    {
+      "epoch": 0.8776052793998955,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.818150004937754e-06,
+      "loss": 0.2219,
+      "step": 18485
+    },
+    {
+      "epoch": 0.8778426624887243,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.8112003723518755e-06,
+      "loss": 0.2274,
+      "step": 18490
+    },
+    {
+      "epoch": 0.878080045577553,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.804263490179122e-06,
+      "loss": 0.2233,
+      "step": 18495
+    },
+    {
+      "epoch": 0.8783174286663818,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.797339362694678e-06,
+      "loss": 0.2222,
+      "step": 18500
+    },
+    {
+      "epoch": 0.8785548117552106,
+      "grad_norm": 0.435546875,
+      "learning_rate": 6.790427994165838e-06,
+      "loss": 0.2249,
+      "step": 18505
+    },
+    {
+      "epoch": 0.8787921948440394,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.783529388852053e-06,
+      "loss": 0.2243,
+      "step": 18510
+    },
+    {
+      "epoch": 0.8790295779328681,
+      "grad_norm": 0.443359375,
+      "learning_rate": 6.776643551004903e-06,
+      "loss": 0.2227,
+      "step": 18515
+    },
+    {
+      "epoch": 0.8792669610216968,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.769770484868099e-06,
+      "loss": 0.2279,
+      "step": 18520
+    },
+    {
+      "epoch": 0.8795043441105256,
+      "grad_norm": 0.47265625,
+      "learning_rate": 6.762910194677474e-06,
+      "loss": 0.2223,
+      "step": 18525
+    },
+    {
+      "epoch": 0.8797417271993543,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.756062684660997e-06,
+      "loss": 0.2219,
+      "step": 18530
+    },
+    {
+      "epoch": 0.8799791102881831,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.749227959038759e-06,
+      "loss": 0.2241,
+      "step": 18535
+    },
+    {
+      "epoch": 0.8802164933770118,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.742406022022971e-06,
+      "loss": 0.2208,
+      "step": 18540
+    },
+    {
+      "epoch": 0.8804538764658406,
+      "grad_norm": 0.53515625,
+      "learning_rate": 6.735596877817959e-06,
+      "loss": 0.2217,
+      "step": 18545
+    },
+    {
+      "epoch": 0.8806912595546693,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.7288005306201694e-06,
+      "loss": 0.2231,
+      "step": 18550
+    },
+    {
+      "epoch": 0.8809286426434981,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.722016984618159e-06,
+      "loss": 0.2245,
+      "step": 18555
+    },
+    {
+      "epoch": 0.8811660257323268,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.715246243992597e-06,
+      "loss": 0.2182,
+      "step": 18560
+    },
+    {
+      "epoch": 0.8814034088211555,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.708488312916261e-06,
+      "loss": 0.2239,
+      "step": 18565
+    },
+    {
+      "epoch": 0.8816407919099843,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.701743195554032e-06,
+      "loss": 0.2233,
+      "step": 18570
+    },
+    {
+      "epoch": 0.881878174998813,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.6950108960628905e-06,
+      "loss": 0.2211,
+      "step": 18575
+    },
+    {
+      "epoch": 0.8821155580876419,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.688291418591933e-06,
+      "loss": 0.2228,
+      "step": 18580
+    },
+    {
+      "epoch": 0.8823529411764706,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.6815847672823364e-06,
+      "loss": 0.2258,
+      "step": 18585
+    },
+    {
+      "epoch": 0.8825903242652994,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.674890946267379e-06,
+      "loss": 0.2224,
+      "step": 18590
+    },
+    {
+      "epoch": 0.8828277073541281,
+      "grad_norm": 0.46875,
+      "learning_rate": 6.668209959672428e-06,
+      "loss": 0.2218,
+      "step": 18595
+    },
+    {
+      "epoch": 0.8830650904429569,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.661541811614955e-06,
+      "loss": 0.2212,
+      "step": 18600
+    },
+    {
+      "epoch": 0.8833024735317856,
+      "grad_norm": 0.4453125,
+      "learning_rate": 6.654886506204508e-06,
+      "loss": 0.2242,
+      "step": 18605
+    },
+    {
+      "epoch": 0.8835398566206143,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.6482440475427105e-06,
+      "loss": 0.2225,
+      "step": 18610
+    },
+    {
+      "epoch": 0.8837772397094431,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.641614439723291e-06,
+      "loss": 0.2249,
+      "step": 18615
+    },
+    {
+      "epoch": 0.8840146227982718,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.634997686832038e-06,
+      "loss": 0.2249,
+      "step": 18620
+    },
+    {
+      "epoch": 0.8842520058871006,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.628393792946838e-06,
+      "loss": 0.2219,
+      "step": 18625
+    },
+    {
+      "epoch": 0.8844893889759293,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.6218027621376246e-06,
+      "loss": 0.2291,
+      "step": 18630
+    },
+    {
+      "epoch": 0.8847267720647581,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.615224598466437e-06,
+      "loss": 0.2198,
+      "step": 18635
+    },
+    {
+      "epoch": 0.8849641551535868,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.608659305987355e-06,
+      "loss": 0.223,
+      "step": 18640
+    },
+    {
+      "epoch": 0.8852015382424157,
+      "grad_norm": 0.408203125,
+      "learning_rate": 6.602106888746548e-06,
+      "loss": 0.22,
+      "step": 18645
+    },
+    {
+      "epoch": 0.8854389213312444,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.595567350782231e-06,
+      "loss": 0.2219,
+      "step": 18650
+    },
+    {
+      "epoch": 0.8856763044200732,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.589040696124696e-06,
+      "loss": 0.2199,
+      "step": 18655
+    },
+    {
+      "epoch": 0.8859136875089019,
+      "grad_norm": 0.48046875,
+      "learning_rate": 6.582526928796297e-06,
+      "loss": 0.2263,
+      "step": 18660
+    },
+    {
+      "epoch": 0.8861510705977306,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.576026052811429e-06,
+      "loss": 0.2263,
+      "step": 18665
+    },
+    {
+      "epoch": 0.8863884536865594,
+      "grad_norm": 0.4453125,
+      "learning_rate": 6.569538072176558e-06,
+      "loss": 0.2227,
+      "step": 18670
+    },
+    {
+      "epoch": 0.8866258367753881,
+      "grad_norm": 0.474609375,
+      "learning_rate": 6.563062990890192e-06,
+      "loss": 0.2238,
+      "step": 18675
+    },
+    {
+      "epoch": 0.8868632198642169,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.556600812942902e-06,
+      "loss": 0.2197,
+      "step": 18680
+    },
+    {
+      "epoch": 0.8871006029530456,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.550151542317295e-06,
+      "loss": 0.2228,
+      "step": 18685
+    },
+    {
+      "epoch": 0.8873379860418744,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.543715182988027e-06,
+      "loss": 0.2246,
+      "step": 18690
+    },
+    {
+      "epoch": 0.8875753691307031,
+      "grad_norm": 0.40625,
+      "learning_rate": 6.537291738921792e-06,
+      "loss": 0.2232,
+      "step": 18695
+    },
+    {
+      "epoch": 0.8878127522195319,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.530881214077338e-06,
+      "loss": 0.2205,
+      "step": 18700
+    },
+    {
+      "epoch": 0.8880501353083606,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.524483612405439e-06,
+      "loss": 0.2218,
+      "step": 18705
+    },
+    {
+      "epoch": 0.8882875183971893,
+      "grad_norm": 0.427734375,
+      "learning_rate": 6.518098937848907e-06,
+      "loss": 0.2256,
+      "step": 18710
+    },
+    {
+      "epoch": 0.8885249014860181,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.511727194342588e-06,
+      "loss": 0.2223,
+      "step": 18715
+    },
+    {
+      "epoch": 0.8887622845748469,
+      "grad_norm": 0.41015625,
+      "learning_rate": 6.5053683858133545e-06,
+      "loss": 0.2213,
+      "step": 18720
+    },
+    {
+      "epoch": 0.8889996676636757,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.499022516180124e-06,
+      "loss": 0.2225,
+      "step": 18725
+    },
+    {
+      "epoch": 0.8892370507525044,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.492689589353809e-06,
+      "loss": 0.2186,
+      "step": 18730
+    },
+    {
+      "epoch": 0.8894744338413332,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.486369609237379e-06,
+      "loss": 0.2218,
+      "step": 18735
+    },
+    {
+      "epoch": 0.8897118169301619,
+      "grad_norm": 0.4609375,
+      "learning_rate": 6.4800625797257996e-06,
+      "loss": 0.2173,
+      "step": 18740
+    },
+    {
+      "epoch": 0.8899492000189907,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.473768504706075e-06,
+      "loss": 0.2282,
+      "step": 18745
+    },
+    {
+      "epoch": 0.8901865831078194,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.467487388057202e-06,
+      "loss": 0.224,
+      "step": 18750
+    },
+    {
+      "epoch": 0.8904239661966481,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.46121923365021e-06,
+      "loss": 0.2274,
+      "step": 18755
+    },
+    {
+      "epoch": 0.8906613492854769,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.45496404534814e-06,
+      "loss": 0.2205,
+      "step": 18760
+    },
+    {
+      "epoch": 0.8908987323743056,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.448721827006027e-06,
+      "loss": 0.2258,
+      "step": 18765
+    },
+    {
+      "epoch": 0.8911361154631344,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.4424925824709225e-06,
+      "loss": 0.2249,
+      "step": 18770
+    },
+    {
+      "epoch": 0.8913734985519631,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.436276315581884e-06,
+      "loss": 0.2256,
+      "step": 18775
+    },
+    {
+      "epoch": 0.8916108816407919,
+      "grad_norm": 0.44921875,
+      "learning_rate": 6.430073030169973e-06,
+      "loss": 0.218,
+      "step": 18780
+    },
+    {
+      "epoch": 0.8918482647296206,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.42388273005824e-06,
+      "loss": 0.2274,
+      "step": 18785
+    },
+    {
+      "epoch": 0.8920856478184495,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.41770541906174e-06,
+      "loss": 0.2234,
+      "step": 18790
+    },
+    {
+      "epoch": 0.8923230309072782,
+      "grad_norm": 0.498046875,
+      "learning_rate": 6.411541100987521e-06,
+      "loss": 0.2253,
+      "step": 18795
+    },
+    {
+      "epoch": 0.892560413996107,
+      "grad_norm": 0.435546875,
+      "learning_rate": 6.405389779634625e-06,
+      "loss": 0.222,
+      "step": 18800
+    },
+    {
+      "epoch": 0.8927977970849357,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.399251458794088e-06,
+      "loss": 0.2224,
+      "step": 18805
+    },
+    {
+      "epoch": 0.8930351801737644,
+      "grad_norm": 0.47265625,
+      "learning_rate": 6.393126142248924e-06,
+      "loss": 0.2225,
+      "step": 18810
+    },
+    {
+      "epoch": 0.8932725632625932,
+      "grad_norm": 0.5,
+      "learning_rate": 6.387013833774139e-06,
+      "loss": 0.225,
+      "step": 18815
+    },
+    {
+      "epoch": 0.8935099463514219,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.380914537136716e-06,
+      "loss": 0.2234,
+      "step": 18820
+    },
+    {
+      "epoch": 0.8937473294402507,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.374828256095633e-06,
+      "loss": 0.2244,
+      "step": 18825
+    },
+    {
+      "epoch": 0.8939847125290794,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.368754994401832e-06,
+      "loss": 0.2236,
+      "step": 18830
+    },
+    {
+      "epoch": 0.8942220956179082,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.36269475579824e-06,
+      "loss": 0.2203,
+      "step": 18835
+    },
+    {
+      "epoch": 0.8944594787067369,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.356647544019746e-06,
+      "loss": 0.2239,
+      "step": 18840
+    },
+    {
+      "epoch": 0.8946968617955657,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.3506133627932275e-06,
+      "loss": 0.2246,
+      "step": 18845
+    },
+    {
+      "epoch": 0.8949342448843944,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.344592215837526e-06,
+      "loss": 0.2246,
+      "step": 18850
+    },
+    {
+      "epoch": 0.8951716279732231,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.338584106863435e-06,
+      "loss": 0.2213,
+      "step": 18855
+    },
+    {
+      "epoch": 0.895409011062052,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.332589039573733e-06,
+      "loss": 0.2269,
+      "step": 18860
+    },
+    {
+      "epoch": 0.8956463941508807,
+      "grad_norm": 0.423828125,
+      "learning_rate": 6.326607017663148e-06,
+      "loss": 0.2252,
+      "step": 18865
+    },
+    {
+      "epoch": 0.8958837772397095,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.320638044818385e-06,
+      "loss": 0.2217,
+      "step": 18870
+    },
+    {
+      "epoch": 0.8961211603285382,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.314682124718075e-06,
+      "loss": 0.223,
+      "step": 18875
+    },
+    {
+      "epoch": 0.896358543417367,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.308739261032841e-06,
+      "loss": 0.2208,
+      "step": 18880
+    },
+    {
+      "epoch": 0.8965959265061957,
+      "grad_norm": 0.49609375,
+      "learning_rate": 6.3028094574252325e-06,
+      "loss": 0.2227,
+      "step": 18885
+    },
+    {
+      "epoch": 0.8968333095950245,
+      "grad_norm": 0.404296875,
+      "learning_rate": 6.296892717549774e-06,
+      "loss": 0.2219,
+      "step": 18890
+    },
+    {
+      "epoch": 0.8970706926838532,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.290989045052909e-06,
+      "loss": 0.2249,
+      "step": 18895
+    },
+    {
+      "epoch": 0.8973080757726819,
+      "grad_norm": 0.4609375,
+      "learning_rate": 6.2850984435730505e-06,
+      "loss": 0.2238,
+      "step": 18900
+    },
+    {
+      "epoch": 0.8975454588615107,
+      "grad_norm": 0.45703125,
+      "learning_rate": 6.279220916740555e-06,
+      "loss": 0.2247,
+      "step": 18905
+    },
+    {
+      "epoch": 0.8977828419503394,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.2733564681777075e-06,
+      "loss": 0.226,
+      "step": 18910
+    },
+    {
+      "epoch": 0.8980202250391682,
+      "grad_norm": 0.482421875,
+      "learning_rate": 6.267505101498747e-06,
+      "loss": 0.2233,
+      "step": 18915
+    },
+    {
+      "epoch": 0.8982576081279969,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.261666820309837e-06,
+      "loss": 0.2209,
+      "step": 18920
+    },
+    {
+      "epoch": 0.8984949912168257,
+      "grad_norm": 0.431640625,
+      "learning_rate": 6.255841628209092e-06,
+      "loss": 0.2252,
+      "step": 18925
+    },
+    {
+      "epoch": 0.8987323743056544,
+      "grad_norm": 0.419921875,
+      "learning_rate": 6.250029528786547e-06,
+      "loss": 0.2252,
+      "step": 18930
+    },
+    {
+      "epoch": 0.8989697573944833,
+      "grad_norm": 0.43359375,
+      "learning_rate": 6.244230525624174e-06,
+      "loss": 0.2208,
+      "step": 18935
+    },
+    {
+      "epoch": 0.899207140483312,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.238444622295868e-06,
+      "loss": 0.2273,
+      "step": 18940
+    },
+    {
+      "epoch": 0.8994445235721408,
+      "grad_norm": 0.44921875,
+      "learning_rate": 6.232671822367461e-06,
+      "loss": 0.2242,
+      "step": 18945
+    },
+    {
+      "epoch": 0.8996819066609695,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.226912129396699e-06,
+      "loss": 0.2216,
+      "step": 18950
+    },
+    {
+      "epoch": 0.8999192897497982,
+      "grad_norm": 0.443359375,
+      "learning_rate": 6.221165546933258e-06,
+      "loss": 0.2246,
+      "step": 18955
+    },
+    {
+      "epoch": 0.900156672838627,
+      "grad_norm": 0.38671875,
+      "learning_rate": 6.2154320785187276e-06,
+      "loss": 0.2207,
+      "step": 18960
+    },
+    {
+      "epoch": 0.9003940559274557,
+      "grad_norm": 0.42578125,
+      "learning_rate": 6.209711727686618e-06,
+      "loss": 0.2226,
+      "step": 18965
+    },
+    {
+      "epoch": 0.9006314390162845,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.20400449796236e-06,
+      "loss": 0.2224,
+      "step": 18970
+    },
+    {
+      "epoch": 0.9008688221051132,
+      "grad_norm": 0.419921875,
+      "learning_rate": 6.1983103928632875e-06,
+      "loss": 0.2226,
+      "step": 18975
+    },
+    {
+      "epoch": 0.901106205193942,
+      "grad_norm": 0.455078125,
+      "learning_rate": 6.192629415898657e-06,
+      "loss": 0.2216,
+      "step": 18980
+    },
+    {
+      "epoch": 0.9013435882827707,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.18696157056962e-06,
+      "loss": 0.2259,
+      "step": 18985
+    },
+    {
+      "epoch": 0.9015809713715995,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.181306860369256e-06,
+      "loss": 0.2223,
+      "step": 18990
+    },
+    {
+      "epoch": 0.9018183544604282,
+      "grad_norm": 0.44921875,
+      "learning_rate": 6.175665288782531e-06,
+      "loss": 0.2228,
+      "step": 18995
+    },
+    {
+      "epoch": 0.9020557375492569,
+      "grad_norm": 0.451171875,
+      "learning_rate": 6.170036859286317e-06,
+      "loss": 0.2227,
+      "step": 19000
+    },
+    {
+      "epoch": 0.9022931206380858,
+      "grad_norm": 0.44140625,
+      "learning_rate": 6.164421575349397e-06,
+      "loss": 0.2216,
+      "step": 19005
+    },
+    {
+      "epoch": 0.9025305037269145,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.1588194404324345e-06,
+      "loss": 0.2258,
+      "step": 19010
+    },
+    {
+      "epoch": 0.9027678868157433,
+      "grad_norm": 0.4375,
+      "learning_rate": 6.153230457988011e-06,
+      "loss": 0.2214,
+      "step": 19015
+    },
+    {
+      "epoch": 0.903005269904572,
+      "grad_norm": 0.416015625,
+      "learning_rate": 6.14765463146058e-06,
+      "loss": 0.2227,
+      "step": 19020
+    },
+    {
+      "epoch": 0.9032426529934008,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.142091964286509e-06,
+      "loss": 0.2238,
+      "step": 19025
+    },
+    {
+      "epoch": 0.9034800360822295,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.136542459894033e-06,
+      "loss": 0.2245,
+      "step": 19030
+    },
+    {
+      "epoch": 0.9037174191710583,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.131006121703301e-06,
+      "loss": 0.225,
+      "step": 19035
+    },
+    {
+      "epoch": 0.903954802259887,
+      "grad_norm": 0.43359375,
+      "learning_rate": 6.1254829531263175e-06,
+      "loss": 0.2252,
+      "step": 19040
+    },
+    {
+      "epoch": 0.9041921853487157,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.119972957566996e-06,
+      "loss": 0.2241,
+      "step": 19045
+    },
+    {
+      "epoch": 0.9044295684375445,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.11447613842112e-06,
+      "loss": 0.2229,
+      "step": 19050
+    },
+    {
+      "epoch": 0.9046669515263732,
+      "grad_norm": 0.478515625,
+      "learning_rate": 6.108992499076352e-06,
+      "loss": 0.2209,
+      "step": 19055
+    },
+    {
+      "epoch": 0.904904334615202,
+      "grad_norm": 0.392578125,
+      "learning_rate": 6.103522042912236e-06,
+      "loss": 0.2233,
+      "step": 19060
+    },
+    {
+      "epoch": 0.9051417177040307,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.098064773300183e-06,
+      "loss": 0.2221,
+      "step": 19065
+    },
+    {
+      "epoch": 0.9053791007928595,
+      "grad_norm": 0.466796875,
+      "learning_rate": 6.092620693603495e-06,
+      "loss": 0.2232,
+      "step": 19070
+    },
+    {
+      "epoch": 0.9056164838816883,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.087189807177325e-06,
+      "loss": 0.221,
+      "step": 19075
+    },
+    {
+      "epoch": 0.9058538669705171,
+      "grad_norm": 0.435546875,
+      "learning_rate": 6.081772117368707e-06,
+      "loss": 0.2198,
+      "step": 19080
+    },
+    {
+      "epoch": 0.9060912500593458,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.076367627516534e-06,
+      "loss": 0.2224,
+      "step": 19085
+    },
+    {
+      "epoch": 0.9063286331481746,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.070976340951577e-06,
+      "loss": 0.2235,
+      "step": 19090
+    },
+    {
+      "epoch": 0.9065660162370033,
+      "grad_norm": 0.447265625,
+      "learning_rate": 6.065598260996454e-06,
+      "loss": 0.2259,
+      "step": 19095
+    },
+    {
+      "epoch": 0.906803399325832,
+      "grad_norm": 0.421875,
+      "learning_rate": 6.060233390965654e-06,
+      "loss": 0.2233,
+      "step": 19100
+    },
+    {
+      "epoch": 0.9070407824146608,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.0548817341655215e-06,
+      "loss": 0.2229,
+      "step": 19105
+    },
+    {
+      "epoch": 0.9072781655034895,
+      "grad_norm": 0.41015625,
+      "learning_rate": 6.049543293894259e-06,
+      "loss": 0.2225,
+      "step": 19110
+    },
+    {
+      "epoch": 0.9075155485923183,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.044218073441922e-06,
+      "loss": 0.2268,
+      "step": 19115
+    },
+    {
+      "epoch": 0.907752931681147,
+      "grad_norm": 0.515625,
+      "learning_rate": 6.03890607609042e-06,
+      "loss": 0.2222,
+      "step": 19120
+    },
+    {
+      "epoch": 0.9079903147699758,
+      "grad_norm": 0.470703125,
+      "learning_rate": 6.033607305113514e-06,
+      "loss": 0.2247,
+      "step": 19125
+    },
+    {
+      "epoch": 0.9082276978588045,
+      "grad_norm": 0.474609375,
+      "learning_rate": 6.028321763776809e-06,
+      "loss": 0.2204,
+      "step": 19130
+    },
+    {
+      "epoch": 0.9084650809476333,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.0230494553377655e-06,
+      "loss": 0.2245,
+      "step": 19135
+    },
+    {
+      "epoch": 0.908702464036462,
+      "grad_norm": 0.3828125,
+      "learning_rate": 6.017790383045676e-06,
+      "loss": 0.2216,
+      "step": 19140
+    },
+    {
+      "epoch": 0.9089398471252907,
+      "grad_norm": 0.400390625,
+      "learning_rate": 6.01254455014169e-06,
+      "loss": 0.2266,
+      "step": 19145
+    },
+    {
+      "epoch": 0.9091772302141196,
+      "grad_norm": 0.44921875,
+      "learning_rate": 6.007311959858789e-06,
+      "loss": 0.2264,
+      "step": 19150
+    },
+    {
+      "epoch": 0.9094146133029483,
+      "grad_norm": 0.439453125,
+      "learning_rate": 6.002092615421789e-06,
+      "loss": 0.2246,
+      "step": 19155
+    },
+    {
+      "epoch": 0.9096519963917771,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.9968865200473554e-06,
+      "loss": 0.2189,
+      "step": 19160
+    },
+    {
+      "epoch": 0.9098893794806058,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.991693676943974e-06,
+      "loss": 0.2215,
+      "step": 19165
+    },
+    {
+      "epoch": 0.9101267625694346,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.986514089311978e-06,
+      "loss": 0.2234,
+      "step": 19170
+    },
+    {
+      "epoch": 0.9103641456582633,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.9813477603435196e-06,
+      "loss": 0.2205,
+      "step": 19175
+    },
+    {
+      "epoch": 0.9106015287470921,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.976194693222584e-06,
+      "loss": 0.2239,
+      "step": 19180
+    },
+    {
+      "epoch": 0.9108389118359208,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.97105489112498e-06,
+      "loss": 0.224,
+      "step": 19185
+    },
+    {
+      "epoch": 0.9110762949247495,
+      "grad_norm": 0.390625,
+      "learning_rate": 5.965928357218352e-06,
+      "loss": 0.2223,
+      "step": 19190
+    },
+    {
+      "epoch": 0.9113136780135783,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.960815094662152e-06,
+      "loss": 0.22,
+      "step": 19195
+    },
+    {
+      "epoch": 0.911551061102407,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.955715106607669e-06,
+      "loss": 0.222,
+      "step": 19200
+    },
+    {
+      "epoch": 0.9117884441912358,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.9506283961979934e-06,
+      "loss": 0.2266,
+      "step": 19205
+    },
+    {
+      "epoch": 0.9120258272800645,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.945554966568042e-06,
+      "loss": 0.222,
+      "step": 19210
+    },
+    {
+      "epoch": 0.9122632103688934,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.940494820844555e-06,
+      "loss": 0.2262,
+      "step": 19215
+    },
+    {
+      "epoch": 0.912500593457722,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.935447962146071e-06,
+      "loss": 0.2256,
+      "step": 19220
+    },
+    {
+      "epoch": 0.9127379765465509,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.930414393582946e-06,
+      "loss": 0.2229,
+      "step": 19225
+    },
+    {
+      "epoch": 0.9129753596353796,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.925394118257345e-06,
+      "loss": 0.2208,
+      "step": 19230
+    },
+    {
+      "epoch": 0.9132127427242084,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.9203871392632444e-06,
+      "loss": 0.2237,
+      "step": 19235
+    },
+    {
+      "epoch": 0.9134501258130371,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.915393459686419e-06,
+      "loss": 0.221,
+      "step": 19240
+    },
+    {
+      "epoch": 0.9136875089018658,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.9104130826044546e-06,
+      "loss": 0.2234,
+      "step": 19245
+    },
+    {
+      "epoch": 0.9139248919906946,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.905446011086732e-06,
+      "loss": 0.2237,
+      "step": 19250
+    },
+    {
+      "epoch": 0.9141622750795233,
+      "grad_norm": 0.458984375,
+      "learning_rate": 5.9004922481944325e-06,
+      "loss": 0.2231,
+      "step": 19255
+    },
+    {
+      "epoch": 0.9143996581683521,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.8955517969805465e-06,
+      "loss": 0.227,
+      "step": 19260
+    },
+    {
+      "epoch": 0.9146370412571808,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.89062466048984e-06,
+      "loss": 0.2254,
+      "step": 19265
+    },
+    {
+      "epoch": 0.9148744243460096,
+      "grad_norm": 0.478515625,
+      "learning_rate": 5.885710841758891e-06,
+      "loss": 0.2252,
+      "step": 19270
+    },
+    {
+      "epoch": 0.9151118074348383,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.880810343816064e-06,
+      "loss": 0.2219,
+      "step": 19275
+    },
+    {
+      "epoch": 0.9153491905236671,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.875923169681516e-06,
+      "loss": 0.227,
+      "step": 19280
+    },
+    {
+      "epoch": 0.9155865736124958,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.871049322367181e-06,
+      "loss": 0.226,
+      "step": 19285
+    },
+    {
+      "epoch": 0.9158239567013245,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.866188804876801e-06,
+      "loss": 0.2189,
+      "step": 19290
+    },
+    {
+      "epoch": 0.9160613397901534,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.861341620205882e-06,
+      "loss": 0.2243,
+      "step": 19295
+    },
+    {
+      "epoch": 0.9162987228789821,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.856507771341724e-06,
+      "loss": 0.2259,
+      "step": 19300
+    },
+    {
+      "epoch": 0.9165361059678109,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.851687261263409e-06,
+      "loss": 0.2214,
+      "step": 19305
+    },
+    {
+      "epoch": 0.9167734890566396,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.846880092941789e-06,
+      "loss": 0.2239,
+      "step": 19310
+    },
+    {
+      "epoch": 0.9170108721454684,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.842086269339506e-06,
+      "loss": 0.2254,
+      "step": 19315
+    },
+    {
+      "epoch": 0.9172482552342971,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.837305793410971e-06,
+      "loss": 0.2233,
+      "step": 19320
+    },
+    {
+      "epoch": 0.9174856383231259,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.832538668102369e-06,
+      "loss": 0.2253,
+      "step": 19325
+    },
+    {
+      "epoch": 0.9177230214119546,
+      "grad_norm": 0.484375,
+      "learning_rate": 5.827784896351651e-06,
+      "loss": 0.2235,
+      "step": 19330
+    },
+    {
+      "epoch": 0.9179604045007833,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.823044481088556e-06,
+      "loss": 0.2219,
+      "step": 19335
+    },
+    {
+      "epoch": 0.9181977875896121,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.818317425234576e-06,
+      "loss": 0.2227,
+      "step": 19340
+    },
+    {
+      "epoch": 0.9184351706784408,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.813603731702971e-06,
+      "loss": 0.2224,
+      "step": 19345
+    },
+    {
+      "epoch": 0.9186725537672696,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.808903403398774e-06,
+      "loss": 0.2256,
+      "step": 19350
+    },
+    {
+      "epoch": 0.9189099368560983,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.80421644321877e-06,
+      "loss": 0.2204,
+      "step": 19355
+    },
+    {
+      "epoch": 0.9191473199449272,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.799542854051515e-06,
+      "loss": 0.2218,
+      "step": 19360
+    },
+    {
+      "epoch": 0.9193847030337559,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.79488263877732e-06,
+      "loss": 0.223,
+      "step": 19365
+    },
+    {
+      "epoch": 0.9196220861225847,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.7902358002682565e-06,
+      "loss": 0.2265,
+      "step": 19370
+    },
+    {
+      "epoch": 0.9198594692114134,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.7856023413881444e-06,
+      "loss": 0.2188,
+      "step": 19375
+    },
+    {
+      "epoch": 0.9200968523002422,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.780982264992571e-06,
+      "loss": 0.2238,
+      "step": 19380
+    },
+    {
+      "epoch": 0.9203342353890709,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.776375573928863e-06,
+      "loss": 0.2213,
+      "step": 19385
+    },
+    {
+      "epoch": 0.9205716184778996,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.771782271036107e-06,
+      "loss": 0.2254,
+      "step": 19390
+    },
+    {
+      "epoch": 0.9208090015667284,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.7672023591451275e-06,
+      "loss": 0.2211,
+      "step": 19395
+    },
+    {
+      "epoch": 0.9210463846555571,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.762635841078508e-06,
+      "loss": 0.222,
+      "step": 19400
+    },
+    {
+      "epoch": 0.9212837677443859,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.758082719650579e-06,
+      "loss": 0.2256,
+      "step": 19405
+    },
+    {
+      "epoch": 0.9215211508332146,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.753542997667398e-06,
+      "loss": 0.2221,
+      "step": 19410
+    },
+    {
+      "epoch": 0.9217585339220434,
+      "grad_norm": 0.46484375,
+      "learning_rate": 5.749016677926778e-06,
+      "loss": 0.222,
+      "step": 19415
+    },
+    {
+      "epoch": 0.9219959170108721,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.744503763218272e-06,
+      "loss": 0.2294,
+      "step": 19420
+    },
+    {
+      "epoch": 0.922233300099701,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.740004256323171e-06,
+      "loss": 0.2222,
+      "step": 19425
+    },
+    {
+      "epoch": 0.9224706831885296,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.735518160014493e-06,
+      "loss": 0.2252,
+      "step": 19430
+    },
+    {
+      "epoch": 0.9227080662773584,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.731045477057006e-06,
+      "loss": 0.2276,
+      "step": 19435
+    },
+    {
+      "epoch": 0.9229454493661872,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.7265862102072e-06,
+      "loss": 0.2211,
+      "step": 19440
+    },
+    {
+      "epoch": 0.9231828324550159,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.722140362213302e-06,
+      "loss": 0.224,
+      "step": 19445
+    },
+    {
+      "epoch": 0.9234202155438447,
+      "grad_norm": 0.478515625,
+      "learning_rate": 5.717707935815266e-06,
+      "loss": 0.2237,
+      "step": 19450
+    },
+    {
+      "epoch": 0.9236575986326734,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.713288933744777e-06,
+      "loss": 0.223,
+      "step": 19455
+    },
+    {
+      "epoch": 0.9238949817215022,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.7088833587252504e-06,
+      "loss": 0.2239,
+      "step": 19460
+    },
+    {
+      "epoch": 0.9241323648103309,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.70449121347182e-06,
+      "loss": 0.2267,
+      "step": 19465
+    },
+    {
+      "epoch": 0.9243697478991597,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.7001125006913406e-06,
+      "loss": 0.2214,
+      "step": 19470
+    },
+    {
+      "epoch": 0.9246071309879884,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.695747223082394e-06,
+      "loss": 0.2243,
+      "step": 19475
+    },
+    {
+      "epoch": 0.9248445140768171,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.691395383335288e-06,
+      "loss": 0.2237,
+      "step": 19480
+    },
+    {
+      "epoch": 0.9250818971656459,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.687056984132033e-06,
+      "loss": 0.2225,
+      "step": 19485
+    },
+    {
+      "epoch": 0.9253192802544746,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.682732028146372e-06,
+      "loss": 0.2258,
+      "step": 19490
+    },
+    {
+      "epoch": 0.9255566633433034,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.678420518043747e-06,
+      "loss": 0.2193,
+      "step": 19495
+    },
+    {
+      "epoch": 0.9257940464321321,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.67412245648133e-06,
+      "loss": 0.2225,
+      "step": 19500
+    },
+    {
+      "epoch": 0.926031429520961,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.669837846107992e-06,
+      "loss": 0.223,
+      "step": 19505
+    },
+    {
+      "epoch": 0.9262688126097897,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.665566689564322e-06,
+      "loss": 0.2215,
+      "step": 19510
+    },
+    {
+      "epoch": 0.9265061956986185,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.661308989482615e-06,
+      "loss": 0.2229,
+      "step": 19515
+    },
+    {
+      "epoch": 0.9267435787874472,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.657064748486865e-06,
+      "loss": 0.2282,
+      "step": 19520
+    },
+    {
+      "epoch": 0.926980961876276,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.652833969192789e-06,
+      "loss": 0.2255,
+      "step": 19525
+    },
+    {
+      "epoch": 0.9272183449651047,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.648616654207791e-06,
+      "loss": 0.2262,
+      "step": 19530
+    },
+    {
+      "epoch": 0.9274557280539334,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.644412806130981e-06,
+      "loss": 0.2251,
+      "step": 19535
+    },
+    {
+      "epoch": 0.9276931111427622,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.640222427553178e-06,
+      "loss": 0.2249,
+      "step": 19540
+    },
+    {
+      "epoch": 0.9279304942315909,
+      "grad_norm": 0.5,
+      "learning_rate": 5.636045521056887e-06,
+      "loss": 0.2257,
+      "step": 19545
+    },
+    {
+      "epoch": 0.9281678773204197,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.6318820892163195e-06,
+      "loss": 0.2245,
+      "step": 19550
+    },
+    {
+      "epoch": 0.9284052604092484,
+      "grad_norm": 0.484375,
+      "learning_rate": 5.6277321345973795e-06,
+      "loss": 0.2241,
+      "step": 19555
+    },
+    {
+      "epoch": 0.9286426434980772,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.623595659757664e-06,
+      "loss": 0.2239,
+      "step": 19560
+    },
+    {
+      "epoch": 0.9288800265869059,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.619472667246464e-06,
+      "loss": 0.2239,
+      "step": 19565
+    },
+    {
+      "epoch": 0.9291174096757348,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.615363159604763e-06,
+      "loss": 0.2261,
+      "step": 19570
+    },
+    {
+      "epoch": 0.9293547927645635,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.611267139365227e-06,
+      "loss": 0.2239,
+      "step": 19575
+    },
+    {
+      "epoch": 0.9295921758533922,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.60718460905222e-06,
+      "loss": 0.2258,
+      "step": 19580
+    },
+    {
+      "epoch": 0.929829558942221,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.6031155711817864e-06,
+      "loss": 0.2269,
+      "step": 19585
+    },
+    {
+      "epoch": 0.9300669420310497,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.599060028261654e-06,
+      "loss": 0.2231,
+      "step": 19590
+    },
+    {
+      "epoch": 0.9303043251198785,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.5950179827912355e-06,
+      "loss": 0.227,
+      "step": 19595
+    },
+    {
+      "epoch": 0.9305417082087072,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.590989437261621e-06,
+      "loss": 0.2256,
+      "step": 19600
+    },
+    {
+      "epoch": 0.930779091297536,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.586974394155595e-06,
+      "loss": 0.2265,
+      "step": 19605
+    },
+    {
+      "epoch": 0.9310164743863647,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.5829728559475995e-06,
+      "loss": 0.2226,
+      "step": 19610
+    },
+    {
+      "epoch": 0.9312538574751935,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.578984825103772e-06,
+      "loss": 0.2264,
+      "step": 19615
+    },
+    {
+      "epoch": 0.9314912405640222,
+      "grad_norm": 0.384765625,
+      "learning_rate": 5.575010304081913e-06,
+      "loss": 0.2246,
+      "step": 19620
+    },
+    {
+      "epoch": 0.9317286236528509,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.571049295331506e-06,
+      "loss": 0.223,
+      "step": 19625
+    },
+    {
+      "epoch": 0.9319660067416797,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.567101801293697e-06,
+      "loss": 0.2217,
+      "step": 19630
+    },
+    {
+      "epoch": 0.9322033898305084,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.563167824401318e-06,
+      "loss": 0.2219,
+      "step": 19635
+    },
+    {
+      "epoch": 0.9324407729193372,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.559247367078852e-06,
+      "loss": 0.2214,
+      "step": 19640
+    },
+    {
+      "epoch": 0.932678156008166,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.555340431742457e-06,
+      "loss": 0.2223,
+      "step": 19645
+    },
+    {
+      "epoch": 0.9329155390969948,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.5514470207999724e-06,
+      "loss": 0.2214,
+      "step": 19650
+    },
+    {
+      "epoch": 0.9331529221858235,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.5475671366508765e-06,
+      "loss": 0.2238,
+      "step": 19655
+    },
+    {
+      "epoch": 0.9333903052746523,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.5437007816863315e-06,
+      "loss": 0.2235,
+      "step": 19660
+    },
+    {
+      "epoch": 0.933627688363481,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.539847958289152e-06,
+      "loss": 0.2194,
+      "step": 19665
+    },
+    {
+      "epoch": 0.9338650714523098,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.536008668833818e-06,
+      "loss": 0.2252,
+      "step": 19670
+    },
+    {
+      "epoch": 0.9341024545411385,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.532182915686461e-06,
+      "loss": 0.2261,
+      "step": 19675
+    },
+    {
+      "epoch": 0.9343398376299672,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.528370701204879e-06,
+      "loss": 0.2211,
+      "step": 19680
+    },
+    {
+      "epoch": 0.934577220718796,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.524572027738523e-06,
+      "loss": 0.2228,
+      "step": 19685
+    },
+    {
+      "epoch": 0.9348146038076247,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.5207868976284975e-06,
+      "loss": 0.2228,
+      "step": 19690
+    },
+    {
+      "epoch": 0.9350519868964535,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.517015313207559e-06,
+      "loss": 0.2226,
+      "step": 19695
+    },
+    {
+      "epoch": 0.9352893699852822,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.513257276800121e-06,
+      "loss": 0.2236,
+      "step": 19700
+    },
+    {
+      "epoch": 0.935526753074111,
+      "grad_norm": 0.5,
+      "learning_rate": 5.50951279072224e-06,
+      "loss": 0.2245,
+      "step": 19705
+    },
+    {
+      "epoch": 0.9357641361629397,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.505781857281632e-06,
+      "loss": 0.2194,
+      "step": 19710
+    },
+    {
+      "epoch": 0.9360015192517686,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.502064478777647e-06,
+      "loss": 0.2245,
+      "step": 19715
+    },
+    {
+      "epoch": 0.9362389023405973,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.4983606575012915e-06,
+      "loss": 0.2219,
+      "step": 19720
+    },
+    {
+      "epoch": 0.936476285429426,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.494670395735217e-06,
+      "loss": 0.2216,
+      "step": 19725
+    },
+    {
+      "epoch": 0.9367136685182548,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.490993695753712e-06,
+      "loss": 0.2211,
+      "step": 19730
+    },
+    {
+      "epoch": 0.9369510516070835,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.487330559822711e-06,
+      "loss": 0.2223,
+      "step": 19735
+    },
+    {
+      "epoch": 0.9371884346959123,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.483680990199788e-06,
+      "loss": 0.2299,
+      "step": 19740
+    },
+    {
+      "epoch": 0.937425817784741,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.48004498913415e-06,
+      "loss": 0.219,
+      "step": 19745
+    },
+    {
+      "epoch": 0.9376632008735698,
+      "grad_norm": 0.486328125,
+      "learning_rate": 5.476422558866659e-06,
+      "loss": 0.2225,
+      "step": 19750
+    },
+    {
+      "epoch": 0.9379005839623985,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.472813701629796e-06,
+      "loss": 0.2265,
+      "step": 19755
+    },
+    {
+      "epoch": 0.9381379670512273,
+      "grad_norm": 0.37109375,
+      "learning_rate": 5.469218419647681e-06,
+      "loss": 0.2165,
+      "step": 19760
+    },
+    {
+      "epoch": 0.938375350140056,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.465636715136074e-06,
+      "loss": 0.2237,
+      "step": 19765
+    },
+    {
+      "epoch": 0.9386127332288847,
+      "grad_norm": 0.46484375,
+      "learning_rate": 5.462068590302365e-06,
+      "loss": 0.2253,
+      "step": 19770
+    },
+    {
+      "epoch": 0.9388501163177135,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.458514047345568e-06,
+      "loss": 0.2208,
+      "step": 19775
+    },
+    {
+      "epoch": 0.9390874994065422,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.454973088456336e-06,
+      "loss": 0.223,
+      "step": 19780
+    },
+    {
+      "epoch": 0.939324882495371,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.451445715816942e-06,
+      "loss": 0.2221,
+      "step": 19785
+    },
+    {
+      "epoch": 0.9395622655841998,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.447931931601291e-06,
+      "loss": 0.2221,
+      "step": 19790
+    },
+    {
+      "epoch": 0.9397996486730286,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.444431737974915e-06,
+      "loss": 0.2201,
+      "step": 19795
+    },
+    {
+      "epoch": 0.9400370317618573,
+      "grad_norm": 0.51171875,
+      "learning_rate": 5.440945137094963e-06,
+      "loss": 0.2231,
+      "step": 19800
+    },
+    {
+      "epoch": 0.9402744148506861,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.437472131110213e-06,
+      "loss": 0.2193,
+      "step": 19805
+    },
+    {
+      "epoch": 0.9405117979395148,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.434012722161063e-06,
+      "loss": 0.223,
+      "step": 19810
+    },
+    {
+      "epoch": 0.9407491810283436,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.4305669123795355e-06,
+      "loss": 0.2259,
+      "step": 19815
+    },
+    {
+      "epoch": 0.9409865641171723,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.427134703889259e-06,
+      "loss": 0.2229,
+      "step": 19820
+    },
+    {
+      "epoch": 0.941223947206001,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.423716098805493e-06,
+      "loss": 0.2257,
+      "step": 19825
+    },
+    {
+      "epoch": 0.9414613302948298,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.420311099235107e-06,
+      "loss": 0.2217,
+      "step": 19830
+    },
+    {
+      "epoch": 0.9416987133836585,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.4169197072765895e-06,
+      "loss": 0.2232,
+      "step": 19835
+    },
+    {
+      "epoch": 0.9419360964724873,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.413541925020038e-06,
+      "loss": 0.2254,
+      "step": 19840
+    },
+    {
+      "epoch": 0.942173479561316,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.410177754547158e-06,
+      "loss": 0.2232,
+      "step": 19845
+    },
+    {
+      "epoch": 0.9424108626501448,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.406827197931282e-06,
+      "loss": 0.2248,
+      "step": 19850
+    },
+    {
+      "epoch": 0.9426482457389735,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.403490257237342e-06,
+      "loss": 0.2224,
+      "step": 19855
+    },
+    {
+      "epoch": 0.9428856288278024,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.400166934521874e-06,
+      "loss": 0.221,
+      "step": 19860
+    },
+    {
+      "epoch": 0.9431230119166311,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.396857231833028e-06,
+      "loss": 0.2256,
+      "step": 19865
+    },
+    {
+      "epoch": 0.9433603950054598,
+      "grad_norm": 0.458984375,
+      "learning_rate": 5.39356115121056e-06,
+      "loss": 0.2181,
+      "step": 19870
+    },
+    {
+      "epoch": 0.9435977780942886,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.3902786946858314e-06,
+      "loss": 0.2238,
+      "step": 19875
+    },
+    {
+      "epoch": 0.9438351611831173,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.387009864281804e-06,
+      "loss": 0.2195,
+      "step": 19880
+    },
+    {
+      "epoch": 0.9440725442719461,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.383754662013042e-06,
+      "loss": 0.2228,
+      "step": 19885
+    },
+    {
+      "epoch": 0.9443099273607748,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.380513089885711e-06,
+      "loss": 0.2244,
+      "step": 19890
+    },
+    {
+      "epoch": 0.9445473104496036,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.377285149897577e-06,
+      "loss": 0.2262,
+      "step": 19895
+    },
+    {
+      "epoch": 0.9447846935384323,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.3740708440380085e-06,
+      "loss": 0.2207,
+      "step": 19900
+    },
+    {
+      "epoch": 0.9450220766272611,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.370870174287965e-06,
+      "loss": 0.2238,
+      "step": 19905
+    },
+    {
+      "epoch": 0.9452594597160898,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.367683142620001e-06,
+      "loss": 0.2212,
+      "step": 19910
+    },
+    {
+      "epoch": 0.9454968428049185,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.364509750998279e-06,
+      "loss": 0.2274,
+      "step": 19915
+    },
+    {
+      "epoch": 0.9457342258937473,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.361350001378539e-06,
+      "loss": 0.222,
+      "step": 19920
+    },
+    {
+      "epoch": 0.945971608982576,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.35820389570812e-06,
+      "loss": 0.2242,
+      "step": 19925
+    },
+    {
+      "epoch": 0.9462089920714049,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.355071435925956e-06,
+      "loss": 0.2241,
+      "step": 19930
+    },
+    {
+      "epoch": 0.9464463751602336,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.351952623962563e-06,
+      "loss": 0.219,
+      "step": 19935
+    },
+    {
+      "epoch": 0.9466837582490624,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.348847461740058e-06,
+      "loss": 0.2251,
+      "step": 19940
+    },
+    {
+      "epoch": 0.9469211413378911,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.34575595117213e-06,
+      "loss": 0.225,
+      "step": 19945
+    },
+    {
+      "epoch": 0.9471585244267199,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.342678094164067e-06,
+      "loss": 0.2287,
+      "step": 19950
+    },
+    {
+      "epoch": 0.9473959075155486,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.3396138926127424e-06,
+      "loss": 0.2221,
+      "step": 19955
+    },
+    {
+      "epoch": 0.9476332906043774,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.336563348406602e-06,
+      "loss": 0.2194,
+      "step": 19960
+    },
+    {
+      "epoch": 0.9478706736932061,
+      "grad_norm": 0.458984375,
+      "learning_rate": 5.3335264634256855e-06,
+      "loss": 0.2244,
+      "step": 19965
+    },
+    {
+      "epoch": 0.9481080567820348,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.330503239541612e-06,
+      "loss": 0.2216,
+      "step": 19970
+    },
+    {
+      "epoch": 0.9483454398708636,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.327493678617581e-06,
+      "loss": 0.2233,
+      "step": 19975
+    },
+    {
+      "epoch": 0.9485828229596923,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.324497782508371e-06,
+      "loss": 0.2197,
+      "step": 19980
+    },
+    {
+      "epoch": 0.9488202060485211,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.321515553060338e-06,
+      "loss": 0.2217,
+      "step": 19985
+    },
+    {
+      "epoch": 0.9490575891373498,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.318546992111413e-06,
+      "loss": 0.2249,
+      "step": 19990
+    },
+    {
+      "epoch": 0.9492949722261786,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.315592101491115e-06,
+      "loss": 0.2261,
+      "step": 19995
+    },
+    {
+      "epoch": 0.9495323553150073,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.312650883020524e-06,
+      "loss": 0.2245,
+      "step": 20000
+    },
+    {
+      "epoch": 0.9497697384038362,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.309723338512304e-06,
+      "loss": 0.2205,
+      "step": 20005
+    },
+    {
+      "epoch": 0.9500071214926649,
+      "grad_norm": 0.38671875,
+      "learning_rate": 5.30680946977068e-06,
+      "loss": 0.2194,
+      "step": 20010
+    },
+    {
+      "epoch": 0.9502445045814936,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.303909278591461e-06,
+      "loss": 0.2263,
+      "step": 20015
+    },
+    {
+      "epoch": 0.9504818876703224,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.301022766762022e-06,
+      "loss": 0.222,
+      "step": 20020
+    },
+    {
+      "epoch": 0.9507192707591511,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.298149936061306e-06,
+      "loss": 0.2232,
+      "step": 20025
+    },
+    {
+      "epoch": 0.9509566538479799,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.295290788259823e-06,
+      "loss": 0.2229,
+      "step": 20030
+    },
+    {
+      "epoch": 0.9511940369368086,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.292445325119657e-06,
+      "loss": 0.2238,
+      "step": 20035
+    },
+    {
+      "epoch": 0.9514314200256374,
+      "grad_norm": 0.484375,
+      "learning_rate": 5.289613548394455e-06,
+      "loss": 0.2238,
+      "step": 20040
+    },
+    {
+      "epoch": 0.9516688031144661,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.286795459829422e-06,
+      "loss": 0.2255,
+      "step": 20045
+    },
+    {
+      "epoch": 0.9519061862032949,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.283991061161339e-06,
+      "loss": 0.2246,
+      "step": 20050
+    },
+    {
+      "epoch": 0.9521435692921236,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.281200354118539e-06,
+      "loss": 0.2224,
+      "step": 20055
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.278423340420927e-06,
+      "loss": 0.2261,
+      "step": 20060
+    },
+    {
+      "epoch": 0.9526183354697811,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.275660021779959e-06,
+      "loss": 0.225,
+      "step": 20065
+    },
+    {
+      "epoch": 0.9528557185586098,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.272910399898661e-06,
+      "loss": 0.2237,
+      "step": 20070
+    },
+    {
+      "epoch": 0.9530931016474387,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.270174476471609e-06,
+      "loss": 0.2244,
+      "step": 20075
+    },
+    {
+      "epoch": 0.9533304847362674,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.267452253184943e-06,
+      "loss": 0.2239,
+      "step": 20080
+    },
+    {
+      "epoch": 0.9535678678250962,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.264743731716352e-06,
+      "loss": 0.2218,
+      "step": 20085
+    },
+    {
+      "epoch": 0.9538052509139249,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.262048913735088e-06,
+      "loss": 0.2212,
+      "step": 20090
+    },
+    {
+      "epoch": 0.9540426340027537,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.259367800901958e-06,
+      "loss": 0.2244,
+      "step": 20095
+    },
+    {
+      "epoch": 0.9542800170915824,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.256700394869315e-06,
+      "loss": 0.2237,
+      "step": 20100
+    },
+    {
+      "epoch": 0.9545174001804112,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.254046697281073e-06,
+      "loss": 0.2245,
+      "step": 20105
+    },
+    {
+      "epoch": 0.9547547832692399,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.251406709772689e-06,
+      "loss": 0.2252,
+      "step": 20110
+    },
+    {
+      "epoch": 0.9549921663580686,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.248780433971183e-06,
+      "loss": 0.2204,
+      "step": 20115
+    },
+    {
+      "epoch": 0.9552295494468974,
+      "grad_norm": 0.375,
+      "learning_rate": 5.246167871495108e-06,
+      "loss": 0.2253,
+      "step": 20120
+    },
+    {
+      "epoch": 0.9554669325357261,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.243569023954579e-06,
+      "loss": 0.2214,
+      "step": 20125
+    },
+    {
+      "epoch": 0.9557043156245549,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.240983892951256e-06,
+      "loss": 0.2265,
+      "step": 20130
+    },
+    {
+      "epoch": 0.9559416987133836,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.238412480078337e-06,
+      "loss": 0.2273,
+      "step": 20135
+    },
+    {
+      "epoch": 0.9561790818022124,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.235854786920579e-06,
+      "loss": 0.2216,
+      "step": 20140
+    },
+    {
+      "epoch": 0.9564164648910412,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.233310815054275e-06,
+      "loss": 0.2179,
+      "step": 20145
+    },
+    {
+      "epoch": 0.95665384797987,
+      "grad_norm": 0.474609375,
+      "learning_rate": 5.230780566047259e-06,
+      "loss": 0.2236,
+      "step": 20150
+    },
+    {
+      "epoch": 0.9568912310686987,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.228264041458915e-06,
+      "loss": 0.2227,
+      "step": 20155
+    },
+    {
+      "epoch": 0.9571286141575274,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.2257612428401676e-06,
+      "loss": 0.2256,
+      "step": 20160
+    },
+    {
+      "epoch": 0.9573659972463562,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.223272171733476e-06,
+      "loss": 0.2233,
+      "step": 20165
+    },
+    {
+      "epoch": 0.9576033803351849,
+      "grad_norm": 0.5,
+      "learning_rate": 5.220796829672847e-06,
+      "loss": 0.2241,
+      "step": 20170
+    },
+    {
+      "epoch": 0.9578407634240137,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.218335218183821e-06,
+      "loss": 0.2206,
+      "step": 20175
+    },
+    {
+      "epoch": 0.9580781465128424,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.215887338783476e-06,
+      "loss": 0.2212,
+      "step": 20180
+    },
+    {
+      "epoch": 0.9583155296016712,
+      "grad_norm": 0.458984375,
+      "learning_rate": 5.213453192980435e-06,
+      "loss": 0.2239,
+      "step": 20185
+    },
+    {
+      "epoch": 0.9585529126904999,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.211032782274844e-06,
+      "loss": 0.2229,
+      "step": 20190
+    },
+    {
+      "epoch": 0.9587902957793287,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.208626108158396e-06,
+      "loss": 0.2241,
+      "step": 20195
+    },
+    {
+      "epoch": 0.9590276788681574,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.20623317211431e-06,
+      "loss": 0.2209,
+      "step": 20200
+    },
+    {
+      "epoch": 0.9592650619569861,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.203853975617345e-06,
+      "loss": 0.2175,
+      "step": 20205
+    },
+    {
+      "epoch": 0.959502445045815,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.201488520133786e-06,
+      "loss": 0.2217,
+      "step": 20210
+    },
+    {
+      "epoch": 0.9597398281346436,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.199136807121454e-06,
+      "loss": 0.2237,
+      "step": 20215
+    },
+    {
+      "epoch": 0.9599772112234725,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.1967988380296955e-06,
+      "loss": 0.2196,
+      "step": 20220
+    },
+    {
+      "epoch": 0.9602145943123012,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.1944746142993995e-06,
+      "loss": 0.2246,
+      "step": 20225
+    },
+    {
+      "epoch": 0.96045197740113,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.192164137362965e-06,
+      "loss": 0.221,
+      "step": 20230
+    },
+    {
+      "epoch": 0.9606893604899587,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.189867408644332e-06,
+      "loss": 0.2233,
+      "step": 20235
+    },
+    {
+      "epoch": 0.9609267435787875,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.187584429558964e-06,
+      "loss": 0.2222,
+      "step": 20240
+    },
+    {
+      "epoch": 0.9611641266676162,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.185315201513848e-06,
+      "loss": 0.2212,
+      "step": 20245
+    },
+    {
+      "epoch": 0.961401509756445,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.183059725907505e-06,
+      "loss": 0.2246,
+      "step": 20250
+    },
+    {
+      "epoch": 0.9616388928452737,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.180818004129968e-06,
+      "loss": 0.2247,
+      "step": 20255
+    },
+    {
+      "epoch": 0.9618762759341024,
+      "grad_norm": 0.4375,
+      "learning_rate": 5.178590037562804e-06,
+      "loss": 0.2165,
+      "step": 20260
+    },
+    {
+      "epoch": 0.9621136590229312,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.176375827579097e-06,
+      "loss": 0.2217,
+      "step": 20265
+    },
+    {
+      "epoch": 0.9623510421117599,
+      "grad_norm": 0.470703125,
+      "learning_rate": 5.1741753755434535e-06,
+      "loss": 0.2228,
+      "step": 20270
+    },
+    {
+      "epoch": 0.9625884252005887,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.1719886828120065e-06,
+      "loss": 0.2221,
+      "step": 20275
+    },
+    {
+      "epoch": 0.9628258082894174,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.1698157507323995e-06,
+      "loss": 0.2221,
+      "step": 20280
+    },
+    {
+      "epoch": 0.9630631913782463,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.167656580643803e-06,
+      "loss": 0.2207,
+      "step": 20285
+    },
+    {
+      "epoch": 0.963300574467075,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.165511173876904e-06,
+      "loss": 0.225,
+      "step": 20290
+    },
+    {
+      "epoch": 0.9635379575559038,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.163379531753907e-06,
+      "loss": 0.2237,
+      "step": 20295
+    },
+    {
+      "epoch": 0.9637753406447325,
+      "grad_norm": 0.5,
+      "learning_rate": 5.16126165558853e-06,
+      "loss": 0.2222,
+      "step": 20300
+    },
+    {
+      "epoch": 0.9640127237335612,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.159157546686015e-06,
+      "loss": 0.2219,
+      "step": 20305
+    },
+    {
+      "epoch": 0.96425010682239,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.157067206343113e-06,
+      "loss": 0.2238,
+      "step": 20310
+    },
+    {
+      "epoch": 0.9644874899112187,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.154990635848092e-06,
+      "loss": 0.2214,
+      "step": 20315
+    },
+    {
+      "epoch": 0.9647248730000475,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.1529278364807336e-06,
+      "loss": 0.223,
+      "step": 20320
+    },
+    {
+      "epoch": 0.9649622560888762,
+      "grad_norm": 0.470703125,
+      "learning_rate": 5.150878809512328e-06,
+      "loss": 0.2241,
+      "step": 20325
+    },
+    {
+      "epoch": 0.965199639177705,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.148843556205683e-06,
+      "loss": 0.2251,
+      "step": 20330
+    },
+    {
+      "epoch": 0.9654370222665337,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.146822077815119e-06,
+      "loss": 0.2238,
+      "step": 20335
+    },
+    {
+      "epoch": 0.9656744053553625,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.144814375586458e-06,
+      "loss": 0.2258,
+      "step": 20340
+    },
+    {
+      "epoch": 0.9659117884441912,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.1428204507570446e-06,
+      "loss": 0.2243,
+      "step": 20345
+    },
+    {
+      "epoch": 0.9661491715330199,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.140840304555722e-06,
+      "loss": 0.2249,
+      "step": 20350
+    },
+    {
+      "epoch": 0.9663865546218487,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.138873938202843e-06,
+      "loss": 0.2176,
+      "step": 20355
+    },
+    {
+      "epoch": 0.9666239377106774,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.1369213529102735e-06,
+      "loss": 0.2266,
+      "step": 20360
+    },
+    {
+      "epoch": 0.9668613207995063,
+      "grad_norm": 0.486328125,
+      "learning_rate": 5.134982549881383e-06,
+      "loss": 0.2245,
+      "step": 20365
+    },
+    {
+      "epoch": 0.967098703888335,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.133057530311045e-06,
+      "loss": 0.2244,
+      "step": 20370
+    },
+    {
+      "epoch": 0.9673360869771638,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.131146295385643e-06,
+      "loss": 0.2269,
+      "step": 20375
+    },
+    {
+      "epoch": 0.9675734700659925,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.129248846283058e-06,
+      "loss": 0.2253,
+      "step": 20380
+    },
+    {
+      "epoch": 0.9678108531548213,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.127365184172685e-06,
+      "loss": 0.2235,
+      "step": 20385
+    },
+    {
+      "epoch": 0.96804823624365,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.125495310215413e-06,
+      "loss": 0.2209,
+      "step": 20390
+    },
+    {
+      "epoch": 0.9682856193324788,
+      "grad_norm": 0.455078125,
+      "learning_rate": 5.123639225563635e-06,
+      "loss": 0.2236,
+      "step": 20395
+    },
+    {
+      "epoch": 0.9685230024213075,
+      "grad_norm": 0.486328125,
+      "learning_rate": 5.121796931361251e-06,
+      "loss": 0.2195,
+      "step": 20400
+    },
+    {
+      "epoch": 0.9687603855101362,
+      "grad_norm": 0.45703125,
+      "learning_rate": 5.119968428743658e-06,
+      "loss": 0.222,
+      "step": 20405
+    },
+    {
+      "epoch": 0.968997768598965,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.118153718837755e-06,
+      "loss": 0.2227,
+      "step": 20410
+    },
+    {
+      "epoch": 0.9692351516877937,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.116352802761936e-06,
+      "loss": 0.2194,
+      "step": 20415
+    },
+    {
+      "epoch": 0.9694725347766225,
+      "grad_norm": 0.388671875,
+      "learning_rate": 5.114565681626099e-06,
+      "loss": 0.2224,
+      "step": 20420
+    },
+    {
+      "epoch": 0.9697099178654512,
+      "grad_norm": 0.46484375,
+      "learning_rate": 5.112792356531637e-06,
+      "loss": 0.2254,
+      "step": 20425
+    },
+    {
+      "epoch": 0.9699473009542801,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.111032828571447e-06,
+      "loss": 0.2193,
+      "step": 20430
+    },
+    {
+      "epoch": 0.9701846840431088,
+      "grad_norm": 0.478515625,
+      "learning_rate": 5.109287098829913e-06,
+      "loss": 0.2231,
+      "step": 20435
+    },
+    {
+      "epoch": 0.9704220671319376,
+      "grad_norm": 0.46875,
+      "learning_rate": 5.1075551683829224e-06,
+      "loss": 0.2248,
+      "step": 20440
+    },
+    {
+      "epoch": 0.9706594502207663,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.105837038297855e-06,
+      "loss": 0.2206,
+      "step": 20445
+    },
+    {
+      "epoch": 0.970896833309595,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.10413270963359e-06,
+      "loss": 0.2224,
+      "step": 20450
+    },
+    {
+      "epoch": 0.9711342163984238,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.1024421834404926e-06,
+      "loss": 0.2195,
+      "step": 20455
+    },
+    {
+      "epoch": 0.9713715994872525,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.100765460760431e-06,
+      "loss": 0.2225,
+      "step": 20460
+    },
+    {
+      "epoch": 0.9716089825760813,
+      "grad_norm": 0.494140625,
+      "learning_rate": 5.099102542626757e-06,
+      "loss": 0.2258,
+      "step": 20465
+    },
+    {
+      "epoch": 0.97184636566491,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.097453430064324e-06,
+      "loss": 0.2216,
+      "step": 20470
+    },
+    {
+      "epoch": 0.9720837487537388,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.09581812408947e-06,
+      "loss": 0.2234,
+      "step": 20475
+    },
+    {
+      "epoch": 0.9723211318425675,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.094196625710026e-06,
+      "loss": 0.223,
+      "step": 20480
+    },
+    {
+      "epoch": 0.9725585149313963,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.092588935925316e-06,
+      "loss": 0.2177,
+      "step": 20485
+    },
+    {
+      "epoch": 0.972795898020225,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5.090995055726152e-06,
+      "loss": 0.2219,
+      "step": 20490
+    },
+    {
+      "epoch": 0.9730332811090537,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.0894149860948345e-06,
+      "loss": 0.2217,
+      "step": 20495
+    },
+    {
+      "epoch": 0.9732706641978826,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.0878487280051495e-06,
+      "loss": 0.2197,
+      "step": 20500
+    },
+    {
+      "epoch": 0.9735080472867113,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.08629628242238e-06,
+      "loss": 0.2217,
+      "step": 20505
+    },
+    {
+      "epoch": 0.9737454303755401,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.084757650303292e-06,
+      "loss": 0.2205,
+      "step": 20510
+    },
+    {
+      "epoch": 0.9739828134643688,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.083232832596133e-06,
+      "loss": 0.2234,
+      "step": 20515
+    },
+    {
+      "epoch": 0.9742201965531976,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.081721830240644e-06,
+      "loss": 0.2212,
+      "step": 20520
+    },
+    {
+      "epoch": 0.9744575796420263,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.0802246441680455e-06,
+      "loss": 0.2176,
+      "step": 20525
+    },
+    {
+      "epoch": 0.9746949627308551,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.078741275301051e-06,
+      "loss": 0.2215,
+      "step": 20530
+    },
+    {
+      "epoch": 0.9749323458196838,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.07727172455385e-06,
+      "loss": 0.2249,
+      "step": 20535
+    },
+    {
+      "epoch": 0.9751697289085126,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.075815992832124e-06,
+      "loss": 0.2212,
+      "step": 20540
+    },
+    {
+      "epoch": 0.9754071119973413,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.0743740810330304e-06,
+      "loss": 0.225,
+      "step": 20545
+    },
+    {
+      "epoch": 0.97564449508617,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.072945990045213e-06,
+      "loss": 0.2194,
+      "step": 20550
+    },
+    {
+      "epoch": 0.9758818781749988,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.071531720748802e-06,
+      "loss": 0.2229,
+      "step": 20555
+    },
+    {
+      "epoch": 0.9761192612638275,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.0701312740154004e-06,
+      "loss": 0.2255,
+      "step": 20560
+    },
+    {
+      "epoch": 0.9763566443526563,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.0687446507080995e-06,
+      "loss": 0.2236,
+      "step": 20565
+    },
+    {
+      "epoch": 0.976594027441485,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.067371851681466e-06,
+      "loss": 0.223,
+      "step": 20570
+    },
+    {
+      "epoch": 0.9768314105303139,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.066012877781554e-06,
+      "loss": 0.2252,
+      "step": 20575
+    },
+    {
+      "epoch": 0.9770687936191426,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.06466772984589e-06,
+      "loss": 0.2251,
+      "step": 20580
+    },
+    {
+      "epoch": 0.9773061767079714,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.063336408703486e-06,
+      "loss": 0.2205,
+      "step": 20585
+    },
+    {
+      "epoch": 0.9775435597968001,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.0620189151748235e-06,
+      "loss": 0.2251,
+      "step": 20590
+    },
+    {
+      "epoch": 0.9777809428856288,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.060715250071874e-06,
+      "loss": 0.2198,
+      "step": 20595
+    },
+    {
+      "epoch": 0.9780183259744576,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.059425414198073e-06,
+      "loss": 0.2259,
+      "step": 20600
+    },
+    {
+      "epoch": 0.9782557090632863,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.0581494083483495e-06,
+      "loss": 0.2266,
+      "step": 20605
+    },
+    {
+      "epoch": 0.9784930921521151,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.056887233309092e-06,
+      "loss": 0.2199,
+      "step": 20610
+    },
+    {
+      "epoch": 0.9787304752409438,
+      "grad_norm": 0.49609375,
+      "learning_rate": 5.0556388898581785e-06,
+      "loss": 0.219,
+      "step": 20615
+    },
+    {
+      "epoch": 0.9789678583297726,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.054404378764955e-06,
+      "loss": 0.2217,
+      "step": 20620
+    },
+    {
+      "epoch": 0.9792052414186013,
+      "grad_norm": 0.4296875,
+      "learning_rate": 5.053183700790245e-06,
+      "loss": 0.2234,
+      "step": 20625
+    },
+    {
+      "epoch": 0.9794426245074301,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.051976856686348e-06,
+      "loss": 0.2258,
+      "step": 20630
+    },
+    {
+      "epoch": 0.9796800075962588,
+      "grad_norm": 0.46875,
+      "learning_rate": 5.050783847197036e-06,
+      "loss": 0.2218,
+      "step": 20635
+    },
+    {
+      "epoch": 0.9799173906850875,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.049604673057554e-06,
+      "loss": 0.2221,
+      "step": 20640
+    },
+    {
+      "epoch": 0.9801547737739164,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.048439334994625e-06,
+      "loss": 0.225,
+      "step": 20645
+    },
+    {
+      "epoch": 0.9803921568627451,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.047287833726437e-06,
+      "loss": 0.2237,
+      "step": 20650
+    },
+    {
+      "epoch": 0.9806295399515739,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.04615016996266e-06,
+      "loss": 0.2247,
+      "step": 20655
+    },
+    {
+      "epoch": 0.9808669230404026,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.045026344404426e-06,
+      "loss": 0.2233,
+      "step": 20660
+    },
+    {
+      "epoch": 0.9811043061292314,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.043916357744347e-06,
+      "loss": 0.2226,
+      "step": 20665
+    },
+    {
+      "epoch": 0.9813416892180601,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.042820210666502e-06,
+      "loss": 0.2224,
+      "step": 20670
+    },
+    {
+      "epoch": 0.9815790723068889,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.04173790384644e-06,
+      "loss": 0.221,
+      "step": 20675
+    },
+    {
+      "epoch": 0.9818164553957176,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.040669437951186e-06,
+      "loss": 0.2229,
+      "step": 20680
+    },
+    {
+      "epoch": 0.9820538384845464,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.0396148136392255e-06,
+      "loss": 0.2242,
+      "step": 20685
+    },
+    {
+      "epoch": 0.9822912215733751,
+      "grad_norm": 0.392578125,
+      "learning_rate": 5.03857403156052e-06,
+      "loss": 0.223,
+      "step": 20690
+    },
+    {
+      "epoch": 0.9825286046622038,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.037547092356501e-06,
+      "loss": 0.2272,
+      "step": 20695
+    },
+    {
+      "epoch": 0.9827659877510326,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.036533996660064e-06,
+      "loss": 0.2218,
+      "step": 20700
+    },
+    {
+      "epoch": 0.9830033708398613,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.035534745095576e-06,
+      "loss": 0.2275,
+      "step": 20705
+    },
+    {
+      "epoch": 0.9832407539286901,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.034549338278872e-06,
+      "loss": 0.2249,
+      "step": 20710
+    },
+    {
+      "epoch": 0.9834781370175188,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.033577776817254e-06,
+      "loss": 0.2221,
+      "step": 20715
+    },
+    {
+      "epoch": 0.9837155201063477,
+      "grad_norm": 0.458984375,
+      "learning_rate": 5.03262006130949e-06,
+      "loss": 0.2222,
+      "step": 20720
+    },
+    {
+      "epoch": 0.9839529031951764,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.031676192345815e-06,
+      "loss": 0.2203,
+      "step": 20725
+    },
+    {
+      "epoch": 0.9841902862840052,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.030746170507934e-06,
+      "loss": 0.2258,
+      "step": 20730
+    },
+    {
+      "epoch": 0.9844276693728339,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.0298299963690096e-06,
+      "loss": 0.2217,
+      "step": 20735
+    },
+    {
+      "epoch": 0.9846650524616626,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.028927670493684e-06,
+      "loss": 0.2207,
+      "step": 20740
+    },
+    {
+      "epoch": 0.9849024355504914,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.028039193438049e-06,
+      "loss": 0.2225,
+      "step": 20745
+    },
+    {
+      "epoch": 0.9851398186393201,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.0271645657496725e-06,
+      "loss": 0.2264,
+      "step": 20750
+    },
+    {
+      "epoch": 0.9853772017281489,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.026303787967583e-06,
+      "loss": 0.2216,
+      "step": 20755
+    },
+    {
+      "epoch": 0.9856145848169776,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.025456860622273e-06,
+      "loss": 0.2231,
+      "step": 20760
+    },
+    {
+      "epoch": 0.9858519679058064,
+      "grad_norm": 0.375,
+      "learning_rate": 5.024623784235702e-06,
+      "loss": 0.2232,
+      "step": 20765
+    },
+    {
+      "epoch": 0.9860893509946351,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.023804559321289e-06,
+      "loss": 0.2267,
+      "step": 20770
+    },
+    {
+      "epoch": 0.9863267340834639,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.022999186383922e-06,
+      "loss": 0.2245,
+      "step": 20775
+    },
+    {
+      "epoch": 0.9865641171722926,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.022207665919942e-06,
+      "loss": 0.2219,
+      "step": 20780
+    },
+    {
+      "epoch": 0.9868015002611213,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.0214299984171685e-06,
+      "loss": 0.2204,
+      "step": 20785
+    },
+    {
+      "epoch": 0.9870388833499502,
+      "grad_norm": 0.427734375,
+      "learning_rate": 5.020666184354867e-06,
+      "loss": 0.2232,
+      "step": 20790
+    },
+    {
+      "epoch": 0.9872762664387789,
+      "grad_norm": 0.3828125,
+      "learning_rate": 5.019916224203776e-06,
+      "loss": 0.2221,
+      "step": 20795
+    },
+    {
+      "epoch": 0.9875136495276077,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.019180118426092e-06,
+      "loss": 0.2231,
+      "step": 20800
+    },
+    {
+      "epoch": 0.9877510326164364,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.018457867475475e-06,
+      "loss": 0.2247,
+      "step": 20805
+    },
+    {
+      "epoch": 0.9879884157052652,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.01774947179704e-06,
+      "loss": 0.2234,
+      "step": 20810
+    },
+    {
+      "epoch": 0.9882257987940939,
+      "grad_norm": 0.490234375,
+      "learning_rate": 5.017054931827374e-06,
+      "loss": 0.2261,
+      "step": 20815
+    },
+    {
+      "epoch": 0.9884631818829227,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.016374247994518e-06,
+      "loss": 0.223,
+      "step": 20820
+    },
+    {
+      "epoch": 0.9887005649717514,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.015707420717969e-06,
+      "loss": 0.2211,
+      "step": 20825
+    },
+    {
+      "epoch": 0.9889379480605802,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.015054450408696e-06,
+      "loss": 0.2222,
+      "step": 20830
+    },
+    {
+      "epoch": 0.9891753311494089,
+      "grad_norm": 0.3984375,
+      "learning_rate": 5.014415337469117e-06,
+      "loss": 0.2248,
+      "step": 20835
+    },
+    {
+      "epoch": 0.9894127142382376,
+      "grad_norm": 0.4609375,
+      "learning_rate": 5.013790082293118e-06,
+      "loss": 0.2252,
+      "step": 20840
+    },
+    {
+      "epoch": 0.9896500973270664,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.01317868526604e-06,
+      "loss": 0.2258,
+      "step": 20845
+    },
+    {
+      "epoch": 0.9898874804158951,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.012581146764683e-06,
+      "loss": 0.2221,
+      "step": 20850
+    },
+    {
+      "epoch": 0.990124863504724,
+      "grad_norm": 0.4921875,
+      "learning_rate": 5.011997467157306e-06,
+      "loss": 0.2233,
+      "step": 20855
+    },
+    {
+      "epoch": 0.9903622465935527,
+      "grad_norm": 0.4453125,
+      "learning_rate": 5.011427646803632e-06,
+      "loss": 0.2248,
+      "step": 20860
+    },
+    {
+      "epoch": 0.9905996296823815,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.010871686054833e-06,
+      "loss": 0.2254,
+      "step": 20865
+    },
+    {
+      "epoch": 0.9908370127712102,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.010329585253552e-06,
+      "loss": 0.2225,
+      "step": 20870
+    },
+    {
+      "epoch": 0.991074395860039,
+      "grad_norm": 0.447265625,
+      "learning_rate": 5.009801344733879e-06,
+      "loss": 0.2233,
+      "step": 20875
+    },
+    {
+      "epoch": 0.9913117789488677,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.009286964821368e-06,
+      "loss": 0.2232,
+      "step": 20880
+    },
+    {
+      "epoch": 0.9915491620376964,
+      "grad_norm": 0.41796875,
+      "learning_rate": 5.0087864458330265e-06,
+      "loss": 0.2208,
+      "step": 20885
+    },
+    {
+      "epoch": 0.9917865451265252,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.008299788077325e-06,
+      "loss": 0.2222,
+      "step": 20890
+    },
+    {
+      "epoch": 0.9920239282153539,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.007826991854186e-06,
+      "loss": 0.2227,
+      "step": 20895
+    },
+    {
+      "epoch": 0.9922613113041827,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.007368057454995e-06,
+      "loss": 0.223,
+      "step": 20900
+    },
+    {
+      "epoch": 0.9924986943930114,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.006922985162586e-06,
+      "loss": 0.2169,
+      "step": 20905
+    },
+    {
+      "epoch": 0.9927360774818402,
+      "grad_norm": 0.396484375,
+      "learning_rate": 5.006491775251261e-06,
+      "loss": 0.2231,
+      "step": 20910
+    },
+    {
+      "epoch": 0.9929734605706689,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.006074427986766e-06,
+      "loss": 0.2226,
+      "step": 20915
+    },
+    {
+      "epoch": 0.9932108436594977,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.0056709436263185e-06,
+      "loss": 0.225,
+      "step": 20920
+    },
+    {
+      "epoch": 0.9934482267483264,
+      "grad_norm": 0.482421875,
+      "learning_rate": 5.0052813224185765e-06,
+      "loss": 0.2242,
+      "step": 20925
+    },
+    {
+      "epoch": 0.9936856098371551,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.0049055646036656e-06,
+      "loss": 0.2203,
+      "step": 20930
+    },
+    {
+      "epoch": 0.993922992925984,
+      "grad_norm": 0.50390625,
+      "learning_rate": 5.004543670413163e-06,
+      "loss": 0.2219,
+      "step": 20935
+    },
+    {
+      "epoch": 0.9941603760148127,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.004195640070101e-06,
+      "loss": 0.218,
+      "step": 20940
+    },
+    {
+      "epoch": 0.9943977591036415,
+      "grad_norm": 0.443359375,
+      "learning_rate": 5.003861473788972e-06,
+      "loss": 0.226,
+      "step": 20945
+    },
+    {
+      "epoch": 0.9946351421924702,
+      "grad_norm": 0.421875,
+      "learning_rate": 5.0035411717757195e-06,
+      "loss": 0.2249,
+      "step": 20950
+    },
+    {
+      "epoch": 0.994872525281299,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.003234734227743e-06,
+      "loss": 0.2255,
+      "step": 20955
+    },
+    {
+      "epoch": 0.9951099083701277,
+      "grad_norm": 0.423828125,
+      "learning_rate": 5.002942161333901e-06,
+      "loss": 0.2257,
+      "step": 20960
+    },
+    {
+      "epoch": 0.9953472914589565,
+      "grad_norm": 0.40625,
+      "learning_rate": 5.002663453274505e-06,
+      "loss": 0.2232,
+      "step": 20965
+    },
+    {
+      "epoch": 0.9955846745477852,
+      "grad_norm": 0.384765625,
+      "learning_rate": 5.002398610221319e-06,
+      "loss": 0.2195,
+      "step": 20970
+    },
+    {
+      "epoch": 0.995822057636614,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.002147632337564e-06,
+      "loss": 0.2194,
+      "step": 20975
+    },
+    {
+      "epoch": 0.9960594407254427,
+      "grad_norm": 0.48828125,
+      "learning_rate": 5.001910519777919e-06,
+      "loss": 0.2213,
+      "step": 20980
+    },
+    {
+      "epoch": 0.9962968238142714,
+      "grad_norm": 0.408203125,
+      "learning_rate": 5.001687272688517e-06,
+      "loss": 0.2266,
+      "step": 20985
+    },
+    {
+      "epoch": 0.9965342069031002,
+      "grad_norm": 0.4140625,
+      "learning_rate": 5.001477891206939e-06,
+      "loss": 0.2211,
+      "step": 20990
+    },
+    {
+      "epoch": 0.9967715899919289,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.001282375462229e-06,
+      "loss": 0.2245,
+      "step": 20995
+    },
+    {
+      "epoch": 0.9970089730807578,
+      "grad_norm": 0.498046875,
+      "learning_rate": 5.001100725574883e-06,
+      "loss": 0.2252,
+      "step": 21000
+    },
+    {
+      "epoch": 0.9972463561695865,
+      "grad_norm": 0.486328125,
+      "learning_rate": 5.000932941656851e-06,
+      "loss": 0.2212,
+      "step": 21005
+    },
+    {
+      "epoch": 0.9974837392584153,
+      "grad_norm": 0.435546875,
+      "learning_rate": 5.0007790238115375e-06,
+      "loss": 0.2215,
+      "step": 21010
+    },
+    {
+      "epoch": 0.997721122347244,
+      "grad_norm": 0.466796875,
+      "learning_rate": 5.000638972133797e-06,
+      "loss": 0.2214,
+      "step": 21015
+    },
+    {
+      "epoch": 0.9979585054360728,
+      "grad_norm": 0.44921875,
+      "learning_rate": 5.000512786709948e-06,
+      "loss": 0.2208,
+      "step": 21020
+    },
+    {
+      "epoch": 0.9981958885249015,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.000400467617755e-06,
+      "loss": 0.2237,
+      "step": 21025
+    },
+    {
+      "epoch": 0.9984332716137302,
+      "grad_norm": 0.412109375,
+      "learning_rate": 5.000302014926443e-06,
+      "loss": 0.2256,
+      "step": 21030
+    },
+    {
+      "epoch": 0.998670654702559,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.0002174286966856e-06,
+      "loss": 0.2231,
+      "step": 21035
+    },
+    {
+      "epoch": 0.9989080377913877,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.000146708980612e-06,
+      "loss": 0.2259,
+      "step": 21040
+    },
+    {
+      "epoch": 0.9991454208802165,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.000089855821809e-06,
+      "loss": 0.2235,
+      "step": 21045
+    },
+    {
+      "epoch": 0.9993828039690452,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.000046869255312e-06,
+      "loss": 0.2198,
+      "step": 21050
+    },
+    {
+      "epoch": 0.999620187057874,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.000017749307616e-06,
+      "loss": 0.2214,
+      "step": 21055
+    },
+    {
+      "epoch": 0.9998575701467027,
+      "grad_norm": 0.4765625,
+      "learning_rate": 5.0000024959966645e-06,
+      "loss": 0.2229,
+      "step": 21060
+    },
+    {
+      "epoch": 1.0,
+      "step": 21063,
+      "total_flos": 5.928116376894964e+18,
+      "train_loss": 0.23212106142810157,
+      "train_runtime": 36092.2941,
+      "train_samples_per_second": 4.669,
+      "train_steps_per_second": 0.584
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 21063,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.928116376894964e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}