diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4157 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 58914,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.005092297899427116,
+      "grad_norm": 3.7509827613830566,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 3.2642,
+      "step": 100
+    },
+    {
+      "epoch": 0.010184595798854232,
+      "grad_norm": 3.509660005569458,
+      "learning_rate": 4.991583636549121e-05,
+      "loss": 3.014,
+      "step": 200
+    },
+    {
+      "epoch": 0.015276893698281349,
+      "grad_norm": 3.0620908737182617,
+      "learning_rate": 4.983082259326011e-05,
+      "loss": 2.901,
+      "step": 300
+    },
+    {
+      "epoch": 0.020369191597708464,
+      "grad_norm": 2.637498617172241,
+      "learning_rate": 4.974580882102901e-05,
+      "loss": 2.8888,
+      "step": 400
+    },
+    {
+      "epoch": 0.025461489497135583,
+      "grad_norm": 2.582336902618408,
+      "learning_rate": 4.966079504879791e-05,
+      "loss": 2.8137,
+      "step": 500
+    },
+    {
+      "epoch": 0.030553787396562698,
+      "grad_norm": 2.575382709503174,
+      "learning_rate": 4.957578127656681e-05,
+      "loss": 2.8131,
+      "step": 600
+    },
+    {
+      "epoch": 0.03564608529598982,
+      "grad_norm": 2.2707173824310303,
+      "learning_rate": 4.94907675043357e-05,
+      "loss": 2.7753,
+      "step": 700
+    },
+    {
+      "epoch": 0.04073838319541693,
+      "grad_norm": 2.1095917224884033,
+      "learning_rate": 4.94057537321046e-05,
+      "loss": 2.7512,
+      "step": 800
+    },
+    {
+      "epoch": 0.04583068109484405,
+      "grad_norm": 1.7593672275543213,
+      "learning_rate": 4.93207399598735e-05,
+      "loss": 2.7556,
+      "step": 900
+    },
+    {
+      "epoch": 0.050922978994271166,
+      "grad_norm": 2.2801873683929443,
+      "learning_rate": 4.92357261876424e-05,
+      "loss": 2.7417,
+      "step": 1000
+    },
+    {
+      "epoch": 0.056015276893698285,
+      "grad_norm": 1.9631321430206299,
+      "learning_rate": 4.91507124154113e-05,
+      "loss": 2.7122,
+      "step": 1100
+    },
+    {
+      "epoch": 0.061107574793125397,
+      "grad_norm": 1.6080312728881836,
+      "learning_rate": 4.90656986431802e-05,
+      "loss": 2.687,
+      "step": 1200
+    },
+    {
+      "epoch": 0.06619987269255251,
+      "grad_norm": 2.1147282123565674,
+      "learning_rate": 4.89806848709491e-05,
+      "loss": 2.6617,
+      "step": 1300
+    },
+    {
+      "epoch": 0.07129217059197963,
+      "grad_norm": 1.905120849609375,
+      "learning_rate": 4.889567109871799e-05,
+      "loss": 2.6655,
+      "step": 1400
+    },
+    {
+      "epoch": 0.07638446849140675,
+      "grad_norm": 1.6756385564804077,
+      "learning_rate": 4.881065732648689e-05,
+      "loss": 2.6286,
+      "step": 1500
+    },
+    {
+      "epoch": 0.08147676639083386,
+      "grad_norm": 1.8816139698028564,
+      "learning_rate": 4.872564355425579e-05,
+      "loss": 2.6414,
+      "step": 1600
+    },
+    {
+      "epoch": 0.08656906429026098,
+      "grad_norm": 1.611456036567688,
+      "learning_rate": 4.864062978202469e-05,
+      "loss": 2.635,
+      "step": 1700
+    },
+    {
+      "epoch": 0.0916613621896881,
+      "grad_norm": 1.8698660135269165,
+      "learning_rate": 4.855561600979359e-05,
+      "loss": 2.6683,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0967536600891152,
+      "grad_norm": 1.6007249355316162,
+      "learning_rate": 4.847060223756249e-05,
+      "loss": 2.633,
+      "step": 1900
+    },
+    {
+      "epoch": 0.10184595798854233,
+      "grad_norm": 1.5520641803741455,
+      "learning_rate": 4.838558846533139e-05,
+      "loss": 2.5912,
+      "step": 2000
+    },
+    {
+      "epoch": 0.10693825588796944,
+      "grad_norm": 1.522303819656372,
+      "learning_rate": 4.8300574693100286e-05,
+      "loss": 2.614,
+      "step": 2100
+    },
+    {
+      "epoch": 0.11203055378739657,
+      "grad_norm": 1.752119541168213,
+      "learning_rate": 4.8215560920869186e-05,
+      "loss": 2.5984,
+      "step": 2200
+    },
+    {
+      "epoch": 0.11712285168682368,
+      "grad_norm": 1.5005803108215332,
+      "learning_rate": 4.8130547148638085e-05,
+      "loss": 2.6231,
+      "step": 2300
+    },
+    {
+      "epoch": 0.12221514958625079,
+      "grad_norm": 1.3557181358337402,
+      "learning_rate": 4.8045533376406984e-05,
+      "loss": 2.6189,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1273074474856779,
+      "grad_norm": 1.5560193061828613,
+      "learning_rate": 4.796051960417588e-05,
+      "loss": 2.5609,
+      "step": 2500
+    },
+    {
+      "epoch": 0.13239974538510502,
+      "grad_norm": 1.4254344701766968,
+      "learning_rate": 4.7875505831944776e-05,
+      "loss": 2.5632,
+      "step": 2600
+    },
+    {
+      "epoch": 0.13749204328453216,
+      "grad_norm": 1.366593599319458,
+      "learning_rate": 4.7790492059713675e-05,
+      "loss": 2.54,
+      "step": 2700
+    },
+    {
+      "epoch": 0.14258434118395927,
+      "grad_norm": 1.2629475593566895,
+      "learning_rate": 4.7705478287482575e-05,
+      "loss": 2.5679,
+      "step": 2800
+    },
+    {
+      "epoch": 0.14767663908338638,
+      "grad_norm": 1.234580159187317,
+      "learning_rate": 4.7620464515251474e-05,
+      "loss": 2.5671,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1527689369828135,
+      "grad_norm": 1.4017528295516968,
+      "learning_rate": 4.7535450743020373e-05,
+      "loss": 2.5859,
+      "step": 3000
+    },
+    {
+      "epoch": 0.1578612348822406,
+      "grad_norm": 1.3464558124542236,
+      "learning_rate": 4.7450436970789266e-05,
+      "loss": 2.5289,
+      "step": 3100
+    },
+    {
+      "epoch": 0.16295353278166771,
+      "grad_norm": 1.3121877908706665,
+      "learning_rate": 4.7365423198558165e-05,
+      "loss": 2.548,
+      "step": 3200
+    },
+    {
+      "epoch": 0.16804583068109485,
+      "grad_norm": 1.2319351434707642,
+      "learning_rate": 4.7280409426327065e-05,
+      "loss": 2.5425,
+      "step": 3300
+    },
+    {
+      "epoch": 0.17313812858052197,
+      "grad_norm": 1.243325114250183,
+      "learning_rate": 4.7195395654095964e-05,
+      "loss": 2.5798,
+      "step": 3400
+    },
+    {
+      "epoch": 0.17823042647994908,
+      "grad_norm": 1.2152389287948608,
+      "learning_rate": 4.711038188186486e-05,
+      "loss": 2.5235,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1833227243793762,
+      "grad_norm": 1.2546372413635254,
+      "learning_rate": 4.702536810963376e-05,
+      "loss": 2.5451,
+      "step": 3600
+    },
+    {
+      "epoch": 0.1884150222788033,
+      "grad_norm": 1.2566453218460083,
+      "learning_rate": 4.694035433740266e-05,
+      "loss": 2.5031,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1935073201782304,
+      "grad_norm": 1.4164502620697021,
+      "learning_rate": 4.685534056517156e-05,
+      "loss": 2.5002,
+      "step": 3800
+    },
+    {
+      "epoch": 0.19859961807765755,
+      "grad_norm": 1.2575647830963135,
+      "learning_rate": 4.677032679294046e-05,
+      "loss": 2.5175,
+      "step": 3900
+    },
+    {
+      "epoch": 0.20369191597708466,
+      "grad_norm": 1.2546263933181763,
+      "learning_rate": 4.668531302070936e-05,
+      "loss": 2.5374,
+      "step": 4000
+    },
+    {
+      "epoch": 0.20878421387651178,
+      "grad_norm": 1.4746454954147339,
+      "learning_rate": 4.660029924847826e-05,
+      "loss": 2.5077,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2138765117759389,
+      "grad_norm": 1.3161815404891968,
+      "learning_rate": 4.651528547624716e-05,
+      "loss": 2.4939,
+      "step": 4200
+    },
+    {
+      "epoch": 0.218968809675366,
+      "grad_norm": 1.2247682809829712,
+      "learning_rate": 4.643027170401605e-05,
+      "loss": 2.5047,
+      "step": 4300
+    },
+    {
+      "epoch": 0.22406110757479314,
+      "grad_norm": 1.024702787399292,
+      "learning_rate": 4.634525793178495e-05,
+      "loss": 2.4986,
+      "step": 4400
+    },
+    {
+      "epoch": 0.22915340547422025,
+      "grad_norm": 1.2271933555603027,
+      "learning_rate": 4.626024415955385e-05,
+      "loss": 2.4815,
+      "step": 4500
+    },
+    {
+      "epoch": 0.23424570337364736,
+      "grad_norm": 1.1049838066101074,
+      "learning_rate": 4.617523038732275e-05,
+      "loss": 2.5055,
+      "step": 4600
+    },
+    {
+      "epoch": 0.23933800127307447,
+      "grad_norm": 1.1865185499191284,
+      "learning_rate": 4.609021661509165e-05,
+      "loss": 2.4932,
+      "step": 4700
+    },
+    {
+      "epoch": 0.24443029917250159,
+      "grad_norm": 1.2031099796295166,
+      "learning_rate": 4.600520284286055e-05,
+      "loss": 2.4857,
+      "step": 4800
+    },
+    {
+      "epoch": 0.2495225970719287,
+      "grad_norm": 1.2100847959518433,
+      "learning_rate": 4.592018907062944e-05,
+      "loss": 2.4704,
+      "step": 4900
+    },
+    {
+      "epoch": 0.2546148949713558,
+      "grad_norm": 1.306518793106079,
+      "learning_rate": 4.583517529839834e-05,
+      "loss": 2.4679,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2597071928707829,
+      "grad_norm": 1.3596395254135132,
+      "learning_rate": 4.575016152616724e-05,
+      "loss": 2.5029,
+      "step": 5100
+    },
+    {
+      "epoch": 0.26479949077021003,
+      "grad_norm": 1.1463990211486816,
+      "learning_rate": 4.566514775393614e-05,
+      "loss": 2.4678,
+      "step": 5200
+    },
+    {
+      "epoch": 0.2698917886696372,
+      "grad_norm": 1.4843939542770386,
+      "learning_rate": 4.558013398170504e-05,
+      "loss": 2.4549,
+      "step": 5300
+    },
+    {
+      "epoch": 0.2749840865690643,
+      "grad_norm": 1.4119912385940552,
+      "learning_rate": 4.549512020947394e-05,
+      "loss": 2.4449,
+      "step": 5400
+    },
+    {
+      "epoch": 0.2800763844684914,
+      "grad_norm": 1.1640745401382446,
+      "learning_rate": 4.5410106437242836e-05,
+      "loss": 2.4133,
+      "step": 5500
+    },
+    {
+      "epoch": 0.28516868236791854,
+      "grad_norm": 1.2901395559310913,
+      "learning_rate": 4.532509266501173e-05,
+      "loss": 2.4493,
+      "step": 5600
+    },
+    {
+      "epoch": 0.29026098026734565,
+      "grad_norm": 1.3150924444198608,
+      "learning_rate": 4.5240078892780635e-05,
+      "loss": 2.4616,
+      "step": 5700
+    },
+    {
+      "epoch": 0.29535327816677276,
+      "grad_norm": 1.1391271352767944,
+      "learning_rate": 4.5155065120549534e-05,
+      "loss": 2.4491,
+      "step": 5800
+    },
+    {
+      "epoch": 0.30044557606619987,
+      "grad_norm": 1.047142505645752,
+      "learning_rate": 4.5070051348318434e-05,
+      "loss": 2.4664,
+      "step": 5900
+    },
+    {
+      "epoch": 0.305537873965627,
+      "grad_norm": 1.2513772249221802,
+      "learning_rate": 4.498503757608733e-05,
+      "loss": 2.4356,
+      "step": 6000
+    },
+    {
+      "epoch": 0.3106301718650541,
+      "grad_norm": 1.2248339653015137,
+      "learning_rate": 4.4900023803856225e-05,
+      "loss": 2.458,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3157224697644812,
+      "grad_norm": 0.9861664772033691,
+      "learning_rate": 4.4815010031625125e-05,
+      "loss": 2.4494,
+      "step": 6200
+    },
+    {
+      "epoch": 0.3208147676639083,
+      "grad_norm": 1.087272047996521,
+      "learning_rate": 4.4729996259394024e-05,
+      "loss": 2.4459,
+      "step": 6300
+    },
+    {
+      "epoch": 0.32590706556333543,
+      "grad_norm": 1.0361382961273193,
+      "learning_rate": 4.464498248716292e-05,
+      "loss": 2.451,
+      "step": 6400
+    },
+    {
+      "epoch": 0.3309993634627626,
+      "grad_norm": 1.0861406326293945,
+      "learning_rate": 4.455996871493182e-05,
+      "loss": 2.4426,
+      "step": 6500
+    },
+    {
+      "epoch": 0.3360916613621897,
+      "grad_norm": 0.9402614235877991,
+      "learning_rate": 4.447495494270072e-05,
+      "loss": 2.4189,
+      "step": 6600
+    },
+    {
+      "epoch": 0.3411839592616168,
+      "grad_norm": 0.9866734743118286,
+      "learning_rate": 4.4389941170469615e-05,
+      "loss": 2.4521,
+      "step": 6700
+    },
+    {
+      "epoch": 0.34627625716104393,
+      "grad_norm": 1.0977962017059326,
+      "learning_rate": 4.4304927398238514e-05,
+      "loss": 2.4505,
+      "step": 6800
+    },
+    {
+      "epoch": 0.35136855506047104,
+      "grad_norm": 1.1266326904296875,
+      "learning_rate": 4.421991362600741e-05,
+      "loss": 2.3999,
+      "step": 6900
+    },
+    {
+      "epoch": 0.35646085295989816,
+      "grad_norm": 1.1100637912750244,
+      "learning_rate": 4.413489985377631e-05,
+      "loss": 2.4226,
+      "step": 7000
+    },
+    {
+      "epoch": 0.36155315085932527,
+      "grad_norm": 1.1532678604125977,
+      "learning_rate": 4.404988608154521e-05,
+      "loss": 2.4048,
+      "step": 7100
+    },
+    {
+      "epoch": 0.3666454487587524,
+      "grad_norm": 1.02146315574646,
+      "learning_rate": 4.396487230931411e-05,
+      "loss": 2.4177,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3717377466581795,
+      "grad_norm": 1.1943087577819824,
+      "learning_rate": 4.387985853708301e-05,
+      "loss": 2.4276,
+      "step": 7300
+    },
+    {
+      "epoch": 0.3768300445576066,
+      "grad_norm": 1.118034839630127,
+      "learning_rate": 4.37948447648519e-05,
+      "loss": 2.3933,
+      "step": 7400
+    },
+    {
+      "epoch": 0.3819223424570337,
+      "grad_norm": 1.0506726503372192,
+      "learning_rate": 4.370983099262081e-05,
+      "loss": 2.4162,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3870146403564608,
+      "grad_norm": 1.1072652339935303,
+      "learning_rate": 4.362481722038971e-05,
+      "loss": 2.4166,
+      "step": 7600
+    },
+    {
+      "epoch": 0.392106938255888,
+      "grad_norm": 0.9805678129196167,
+      "learning_rate": 4.353980344815861e-05,
+      "loss": 2.3771,
+      "step": 7700
+    },
+    {
+      "epoch": 0.3971992361553151,
+      "grad_norm": 1.0781447887420654,
+      "learning_rate": 4.345478967592751e-05,
+      "loss": 2.3971,
+      "step": 7800
+    },
+    {
+      "epoch": 0.4022915340547422,
+      "grad_norm": 1.1752007007598877,
+      "learning_rate": 4.33697759036964e-05,
+      "loss": 2.3837,
+      "step": 7900
+    },
+    {
+      "epoch": 0.40738383195416933,
+      "grad_norm": 1.0886644124984741,
+      "learning_rate": 4.32847621314653e-05,
+      "loss": 2.4372,
+      "step": 8000
+    },
+    {
+      "epoch": 0.41247612985359644,
+      "grad_norm": 1.01775062084198,
+      "learning_rate": 4.31997483592342e-05,
+      "loss": 2.4051,
+      "step": 8100
+    },
+    {
+      "epoch": 0.41756842775302355,
+      "grad_norm": 1.0455646514892578,
+      "learning_rate": 4.31147345870031e-05,
+      "loss": 2.366,
+      "step": 8200
+    },
+    {
+      "epoch": 0.42266072565245066,
+      "grad_norm": 0.9850195646286011,
+      "learning_rate": 4.3029720814772e-05,
+      "loss": 2.3816,
+      "step": 8300
+    },
+    {
+      "epoch": 0.4277530235518778,
+      "grad_norm": 1.092155933380127,
+      "learning_rate": 4.2944707042540896e-05,
+      "loss": 2.396,
+      "step": 8400
+    },
+    {
+      "epoch": 0.4328453214513049,
+      "grad_norm": 1.008317232131958,
+      "learning_rate": 4.285969327030979e-05,
+      "loss": 2.3976,
+      "step": 8500
+    },
+    {
+      "epoch": 0.437937619350732,
+      "grad_norm": 1.1001275777816772,
+      "learning_rate": 4.277467949807869e-05,
+      "loss": 2.4009,
+      "step": 8600
+    },
+    {
+      "epoch": 0.4430299172501591,
+      "grad_norm": 0.9589524865150452,
+      "learning_rate": 4.268966572584759e-05,
+      "loss": 2.3755,
+      "step": 8700
+    },
+    {
+      "epoch": 0.4481222151495863,
+      "grad_norm": 0.9529566168785095,
+      "learning_rate": 4.260465195361649e-05,
+      "loss": 2.3961,
+      "step": 8800
+    },
+    {
+      "epoch": 0.4532145130490134,
+      "grad_norm": 1.0157649517059326,
+      "learning_rate": 4.2519638181385386e-05,
+      "loss": 2.3743,
+      "step": 8900
+    },
+    {
+      "epoch": 0.4583068109484405,
+      "grad_norm": 1.0096311569213867,
+      "learning_rate": 4.2434624409154286e-05,
+      "loss": 2.3702,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4633991088478676,
+      "grad_norm": 1.0700254440307617,
+      "learning_rate": 4.2349610636923185e-05,
+      "loss": 2.3486,
+      "step": 9100
+    },
+    {
+      "epoch": 0.4684914067472947,
+      "grad_norm": 0.9580355286598206,
+      "learning_rate": 4.226459686469208e-05,
+      "loss": 2.3686,
+      "step": 9200
+    },
+    {
+      "epoch": 0.47358370464672184,
+      "grad_norm": 1.0027587413787842,
+      "learning_rate": 4.217958309246098e-05,
+      "loss": 2.4074,
+      "step": 9300
+    },
+    {
+      "epoch": 0.47867600254614895,
+      "grad_norm": 0.9647036194801331,
+      "learning_rate": 4.209456932022988e-05,
+      "loss": 2.3631,
+      "step": 9400
+    },
+    {
+      "epoch": 0.48376830044557606,
+      "grad_norm": 1.0718977451324463,
+      "learning_rate": 4.200955554799878e-05,
+      "loss": 2.3613,
+      "step": 9500
+    },
+    {
+      "epoch": 0.48886059834500317,
+      "grad_norm": 1.1674007177352905,
+      "learning_rate": 4.192454177576768e-05,
+      "loss": 2.3604,
+      "step": 9600
+    },
+    {
+      "epoch": 0.4939528962444303,
+      "grad_norm": 0.8964582681655884,
+      "learning_rate": 4.1839528003536574e-05,
+      "loss": 2.3517,
+      "step": 9700
+    },
+    {
+      "epoch": 0.4990451941438574,
+      "grad_norm": 0.9950689673423767,
+      "learning_rate": 4.175451423130547e-05,
+      "loss": 2.3609,
+      "step": 9800
+    },
+    {
+      "epoch": 0.5041374920432845,
+      "grad_norm": 1.0391299724578857,
+      "learning_rate": 4.166950045907437e-05,
+      "loss": 2.3764,
+      "step": 9900
+    },
+    {
+      "epoch": 0.5092297899427116,
+      "grad_norm": 0.9937861561775208,
+      "learning_rate": 4.158448668684327e-05,
+      "loss": 2.3439,
+      "step": 10000
+    },
+    {
+      "epoch": 0.5143220878421387,
+      "grad_norm": 0.9637438654899597,
+      "learning_rate": 4.149947291461217e-05,
+      "loss": 2.3599,
+      "step": 10100
+    },
+    {
+      "epoch": 0.5194143857415658,
+      "grad_norm": 0.991791844367981,
+      "learning_rate": 4.141445914238107e-05,
+      "loss": 2.3688,
+      "step": 10200
+    },
+    {
+      "epoch": 0.524506683640993,
+      "grad_norm": 1.1475801467895508,
+      "learning_rate": 4.132944537014996e-05,
+      "loss": 2.351,
+      "step": 10300
+    },
+    {
+      "epoch": 0.5295989815404201,
+      "grad_norm": 1.018678069114685,
+      "learning_rate": 4.124443159791886e-05,
+      "loss": 2.3381,
+      "step": 10400
+    },
+    {
+      "epoch": 0.5346912794398472,
+      "grad_norm": 1.0166884660720825,
+      "learning_rate": 4.115941782568776e-05,
+      "loss": 2.3393,
+      "step": 10500
+    },
+    {
+      "epoch": 0.5397835773392744,
+      "grad_norm": 0.9590491652488708,
+      "learning_rate": 4.107440405345666e-05,
+      "loss": 2.3428,
+      "step": 10600
+    },
+    {
+      "epoch": 0.5448758752387015,
+      "grad_norm": 1.0007227659225464,
+      "learning_rate": 4.098939028122556e-05,
+      "loss": 2.3388,
+      "step": 10700
+    },
+    {
+      "epoch": 0.5499681731381286,
+      "grad_norm": 0.8273807764053345,
+      "learning_rate": 4.090437650899446e-05,
+      "loss": 2.3238,
+      "step": 10800
+    },
+    {
+      "epoch": 0.5550604710375557,
+      "grad_norm": 0.9188222885131836,
+      "learning_rate": 4.081936273676335e-05,
+      "loss": 2.3171,
+      "step": 10900
+    },
+    {
+      "epoch": 0.5601527689369828,
+      "grad_norm": 1.2066142559051514,
+      "learning_rate": 4.073434896453225e-05,
+      "loss": 2.385,
+      "step": 11000
+    },
+    {
+      "epoch": 0.56524506683641,
+      "grad_norm": 1.0904101133346558,
+      "learning_rate": 4.064933519230115e-05,
+      "loss": 2.341,
+      "step": 11100
+    },
+    {
+      "epoch": 0.5703373647358371,
+      "grad_norm": 1.0374412536621094,
+      "learning_rate": 4.056432142007005e-05,
+      "loss": 2.3398,
+      "step": 11200
+    },
+    {
+      "epoch": 0.5754296626352642,
+      "grad_norm": 0.9854114055633545,
+      "learning_rate": 4.0479307647838956e-05,
+      "loss": 2.3512,
+      "step": 11300
+    },
+    {
+      "epoch": 0.5805219605346913,
+      "grad_norm": 1.071382999420166,
+      "learning_rate": 4.0394293875607856e-05,
+      "loss": 2.3145,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5856142584341184,
+      "grad_norm": 0.9923407435417175,
+      "learning_rate": 4.030928010337675e-05,
+      "loss": 2.3475,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5907065563335455,
+      "grad_norm": 1.034600019454956,
+      "learning_rate": 4.022426633114565e-05,
+      "loss": 2.3196,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5957988542329726,
+      "grad_norm": 1.4072537422180176,
+      "learning_rate": 4.013925255891455e-05,
+      "loss": 2.3435,
+      "step": 11700
+    },
+    {
+      "epoch": 0.6008911521323997,
+      "grad_norm": 1.0498465299606323,
+      "learning_rate": 4.0054238786683446e-05,
+      "loss": 2.3488,
+      "step": 11800
+    },
+    {
+      "epoch": 0.6059834500318269,
+      "grad_norm": 0.9911717176437378,
+      "learning_rate": 3.9969225014452346e-05,
+      "loss": 2.3286,
+      "step": 11900
+    },
+    {
+      "epoch": 0.611075747931254,
+      "grad_norm": 0.9431672692298889,
+      "learning_rate": 3.9884211242221245e-05,
+      "loss": 2.3502,
+      "step": 12000
+    },
+    {
+      "epoch": 0.6161680458306811,
+      "grad_norm": 1.0439810752868652,
+      "learning_rate": 3.979919746999014e-05,
+      "loss": 2.3516,
+      "step": 12100
+    },
+    {
+      "epoch": 0.6212603437301082,
+      "grad_norm": 0.8762308955192566,
+      "learning_rate": 3.971418369775904e-05,
+      "loss": 2.2836,
+      "step": 12200
+    },
+    {
+      "epoch": 0.6263526416295353,
+      "grad_norm": 0.8706735372543335,
+      "learning_rate": 3.9629169925527936e-05,
+      "loss": 2.349,
+      "step": 12300
+    },
+    {
+      "epoch": 0.6314449395289624,
+      "grad_norm": 0.9823511838912964,
+      "learning_rate": 3.9544156153296836e-05,
+      "loss": 2.3356,
+      "step": 12400
+    },
+    {
+      "epoch": 0.6365372374283895,
+      "grad_norm": 0.939285933971405,
+      "learning_rate": 3.9459142381065735e-05,
+      "loss": 2.3435,
+      "step": 12500
+    },
+    {
+      "epoch": 0.6416295353278166,
+      "grad_norm": 1.033011555671692,
+      "learning_rate": 3.9374128608834634e-05,
+      "loss": 2.3208,
+      "step": 12600
+    },
+    {
+      "epoch": 0.6467218332272437,
+      "grad_norm": 0.9835578799247742,
+      "learning_rate": 3.928911483660353e-05,
+      "loss": 2.3332,
+      "step": 12700
+    },
+    {
+      "epoch": 0.6518141311266709,
+      "grad_norm": 0.9082310795783997,
+      "learning_rate": 3.9204101064372426e-05,
+      "loss": 2.3216,
+      "step": 12800
+    },
+    {
+      "epoch": 0.6569064290260981,
+      "grad_norm": 0.8588578701019287,
+      "learning_rate": 3.9119087292141325e-05,
+      "loss": 2.3114,
+      "step": 12900
+    },
+    {
+      "epoch": 0.6619987269255252,
+      "grad_norm": 1.040531873703003,
+      "learning_rate": 3.9034073519910225e-05,
+      "loss": 2.3328,
+      "step": 13000
+    },
+    {
+      "epoch": 0.6670910248249523,
+      "grad_norm": 1.0225043296813965,
+      "learning_rate": 3.894905974767913e-05,
+      "loss": 2.3245,
+      "step": 13100
+    },
+    {
+      "epoch": 0.6721833227243794,
+      "grad_norm": 1.0172550678253174,
+      "learning_rate": 3.886404597544803e-05,
+      "loss": 2.3056,
+      "step": 13200
+    },
+    {
+      "epoch": 0.6772756206238065,
+      "grad_norm": 0.9119499921798706,
+      "learning_rate": 3.877903220321692e-05,
+      "loss": 2.317,
+      "step": 13300
+    },
+    {
+      "epoch": 0.6823679185232336,
+      "grad_norm": 0.8971495032310486,
+      "learning_rate": 3.869401843098582e-05,
+      "loss": 2.3292,
+      "step": 13400
+    },
+    {
+      "epoch": 0.6874602164226608,
+      "grad_norm": 0.9643430709838867,
+      "learning_rate": 3.860900465875472e-05,
+      "loss": 2.3779,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6925525143220879,
+      "grad_norm": 0.919440507888794,
+      "learning_rate": 3.852399088652362e-05,
+      "loss": 2.2993,
+      "step": 13600
+    },
+    {
+      "epoch": 0.697644812221515,
+      "grad_norm": 0.9949972033500671,
+      "learning_rate": 3.843897711429252e-05,
+      "loss": 2.3255,
+      "step": 13700
+    },
+    {
+      "epoch": 0.7027371101209421,
+      "grad_norm": 0.9251271486282349,
+      "learning_rate": 3.835396334206142e-05,
+      "loss": 2.2997,
+      "step": 13800
+    },
+    {
+      "epoch": 0.7078294080203692,
+      "grad_norm": 0.9567040205001831,
+      "learning_rate": 3.826894956983031e-05,
+      "loss": 2.3198,
+      "step": 13900
+    },
+    {
+      "epoch": 0.7129217059197963,
+      "grad_norm": 1.1165566444396973,
+      "learning_rate": 3.818393579759921e-05,
+      "loss": 2.3074,
+      "step": 14000
+    },
+    {
+      "epoch": 0.7180140038192234,
+      "grad_norm": 0.9649367928504944,
+      "learning_rate": 3.809892202536811e-05,
+      "loss": 2.2916,
+      "step": 14100
+    },
+    {
+      "epoch": 0.7231063017186505,
+      "grad_norm": 0.8595756888389587,
+      "learning_rate": 3.801390825313701e-05,
+      "loss": 2.3386,
+      "step": 14200
+    },
+    {
+      "epoch": 0.7281985996180776,
+      "grad_norm": 0.7877846360206604,
+      "learning_rate": 3.792889448090591e-05,
+      "loss": 2.2741,
+      "step": 14300
+    },
+    {
+      "epoch": 0.7332908975175048,
+      "grad_norm": 0.9086227416992188,
+      "learning_rate": 3.784388070867481e-05,
+      "loss": 2.3186,
+      "step": 14400
+    },
+    {
+      "epoch": 0.7383831954169319,
+      "grad_norm": 0.9466003179550171,
+      "learning_rate": 3.77588669364437e-05,
+      "loss": 2.2916,
+      "step": 14500
+    },
+    {
+      "epoch": 0.743475493316359,
+      "grad_norm": 0.8069922924041748,
+      "learning_rate": 3.76738531642126e-05,
+      "loss": 2.3108,
+      "step": 14600
+    },
+    {
+      "epoch": 0.7485677912157861,
+      "grad_norm": 1.0324113368988037,
+      "learning_rate": 3.75888393919815e-05,
+      "loss": 2.3066,
+      "step": 14700
+    },
+    {
+      "epoch": 0.7536600891152132,
+      "grad_norm": 0.892573893070221,
+      "learning_rate": 3.75038256197504e-05,
+      "loss": 2.2738,
+      "step": 14800
+    },
+    {
+      "epoch": 0.7587523870146403,
+      "grad_norm": 0.7999922037124634,
+      "learning_rate": 3.74188118475193e-05,
+      "loss": 2.3195,
+      "step": 14900
+    },
+    {
+      "epoch": 0.7638446849140674,
+      "grad_norm": 1.004957914352417,
+      "learning_rate": 3.73337980752882e-05,
+      "loss": 2.2935,
+      "step": 15000
+    },
+    {
+      "epoch": 0.7689369828134945,
+      "grad_norm": 1.046640157699585,
+      "learning_rate": 3.72487843030571e-05,
+      "loss": 2.3109,
+      "step": 15100
+    },
+    {
+      "epoch": 0.7740292807129217,
+      "grad_norm": 0.9236047863960266,
+      "learning_rate": 3.7163770530825996e-05,
+      "loss": 2.3128,
+      "step": 15200
+    },
+    {
+      "epoch": 0.7791215786123489,
+      "grad_norm": 1.0190492868423462,
+      "learning_rate": 3.7078756758594896e-05,
+      "loss": 2.3018,
+      "step": 15300
+    },
+    {
+      "epoch": 0.784213876511776,
+      "grad_norm": 0.8099306225776672,
+      "learning_rate": 3.6993742986363795e-05,
+      "loss": 2.313,
+      "step": 15400
+    },
+    {
+      "epoch": 0.7893061744112031,
+      "grad_norm": 0.9618342518806458,
+      "learning_rate": 3.6908729214132694e-05,
+      "loss": 2.2864,
+      "step": 15500
+    },
+    {
+      "epoch": 0.7943984723106302,
+      "grad_norm": 1.046680212020874,
+      "learning_rate": 3.6823715441901594e-05,
+      "loss": 2.2853,
+      "step": 15600
+    },
+    {
+      "epoch": 0.7994907702100573,
+      "grad_norm": 0.8486195206642151,
+      "learning_rate": 3.6738701669670486e-05,
+      "loss": 2.2854,
+      "step": 15700
+    },
+    {
+      "epoch": 0.8045830681094844,
+      "grad_norm": 0.9708773493766785,
+      "learning_rate": 3.6653687897439386e-05,
+      "loss": 2.2928,
+      "step": 15800
+    },
+    {
+      "epoch": 0.8096753660089115,
+      "grad_norm": 0.8969681262969971,
+      "learning_rate": 3.6568674125208285e-05,
+      "loss": 2.2976,
+      "step": 15900
+    },
+    {
+      "epoch": 0.8147676639083387,
+      "grad_norm": 0.9385348558425903,
+      "learning_rate": 3.6483660352977184e-05,
+      "loss": 2.2847,
+      "step": 16000
+    },
+    {
+      "epoch": 0.8198599618077658,
+      "grad_norm": 0.8899937272071838,
+      "learning_rate": 3.6398646580746083e-05,
+      "loss": 2.2972,
+      "step": 16100
+    },
+    {
+      "epoch": 0.8249522597071929,
+      "grad_norm": 0.8900747299194336,
+      "learning_rate": 3.631363280851498e-05,
+      "loss": 2.2952,
+      "step": 16200
+    },
+    {
+      "epoch": 0.83004455760662,
+      "grad_norm": 1.026571273803711,
+      "learning_rate": 3.6228619036283875e-05,
+      "loss": 2.2842,
+      "step": 16300
+    },
+    {
+      "epoch": 0.8351368555060471,
+      "grad_norm": 0.9016963839530945,
+      "learning_rate": 3.6143605264052775e-05,
+      "loss": 2.288,
+      "step": 16400
+    },
+    {
+      "epoch": 0.8402291534054742,
+      "grad_norm": 0.8101049065589905,
+      "learning_rate": 3.6058591491821674e-05,
+      "loss": 2.2486,
+      "step": 16500
+    },
+    {
+      "epoch": 0.8453214513049013,
+      "grad_norm": 0.860748827457428,
+      "learning_rate": 3.597357771959057e-05,
+      "loss": 2.2911,
+      "step": 16600
+    },
+    {
+      "epoch": 0.8504137492043284,
+      "grad_norm": 0.9295821189880371,
+      "learning_rate": 3.588856394735947e-05,
+      "loss": 2.2477,
+      "step": 16700
+    },
+    {
+      "epoch": 0.8555060471037556,
+      "grad_norm": 0.9582170844078064,
+      "learning_rate": 3.580355017512837e-05,
+      "loss": 2.307,
+      "step": 16800
+    },
+    {
+      "epoch": 0.8605983450031827,
+      "grad_norm": 0.9199303984642029,
+      "learning_rate": 3.571853640289727e-05,
+      "loss": 2.2692,
+      "step": 16900
+    },
+    {
+      "epoch": 0.8656906429026098,
+      "grad_norm": 0.8835098743438721,
+      "learning_rate": 3.563352263066617e-05,
+      "loss": 2.2681,
+      "step": 17000
+    },
+    {
+      "epoch": 0.8707829408020369,
+      "grad_norm": 0.9898850917816162,
+      "learning_rate": 3.554850885843507e-05,
+      "loss": 2.2718,
+      "step": 17100
+    },
+    {
+      "epoch": 0.875875238701464,
+      "grad_norm": 1.0997586250305176,
+      "learning_rate": 3.546349508620397e-05,
+      "loss": 2.2577,
+      "step": 17200
+    },
+    {
+      "epoch": 0.8809675366008911,
+      "grad_norm": 0.8374606370925903,
+      "learning_rate": 3.537848131397287e-05,
+      "loss": 2.2731,
+      "step": 17300
+    },
+    {
+      "epoch": 0.8860598345003182,
+      "grad_norm": 0.9752559065818787,
+      "learning_rate": 3.529346754174177e-05,
+      "loss": 2.2776,
+      "step": 17400
+    },
+    {
+      "epoch": 0.8911521323997453,
+      "grad_norm": 0.8918510675430298,
+      "learning_rate": 3.520845376951066e-05,
+      "loss": 2.2838,
+      "step": 17500
+    },
+    {
+      "epoch": 0.8962444302991726,
+      "grad_norm": 0.9751953482627869,
+      "learning_rate": 3.512343999727956e-05,
+      "loss": 2.268,
+      "step": 17600
+    },
+    {
+      "epoch": 0.9013367281985997,
+      "grad_norm": 0.9787586331367493,
+      "learning_rate": 3.503842622504846e-05,
+      "loss": 2.2927,
+      "step": 17700
+    },
+    {
+      "epoch": 0.9064290260980268,
+      "grad_norm": 0.9199690222740173,
+      "learning_rate": 3.495341245281736e-05,
+      "loss": 2.2785,
+      "step": 17800
+    },
+    {
+      "epoch": 0.9115213239974539,
+      "grad_norm": 0.8526634573936462,
+      "learning_rate": 3.486839868058626e-05,
+      "loss": 2.2818,
+      "step": 17900
+    },
+    {
+      "epoch": 0.916613621896881,
+      "grad_norm": 0.9445266127586365,
+      "learning_rate": 3.478338490835516e-05,
+      "loss": 2.3147,
+      "step": 18000
+    },
+    {
+      "epoch": 0.9217059197963081,
+      "grad_norm": 0.9607738256454468,
+      "learning_rate": 3.469837113612405e-05,
+      "loss": 2.2663,
+      "step": 18100
+    },
+    {
+      "epoch": 0.9267982176957352,
+      "grad_norm": 0.8561920523643494,
+      "learning_rate": 3.461335736389295e-05,
+      "loss": 2.2355,
+      "step": 18200
+    },
+    {
+      "epoch": 0.9318905155951623,
+      "grad_norm": 0.8668131828308105,
+      "learning_rate": 3.452834359166185e-05,
+      "loss": 2.2801,
+      "step": 18300
+    },
+    {
+      "epoch": 0.9369828134945895,
+      "grad_norm": 0.9161975979804993,
+      "learning_rate": 3.444332981943075e-05,
+      "loss": 2.2668,
+      "step": 18400
+    },
+    {
+      "epoch": 0.9420751113940166,
+      "grad_norm": 0.9021576046943665,
+      "learning_rate": 3.435831604719965e-05,
+      "loss": 2.2887,
+      "step": 18500
+    },
+    {
+      "epoch": 0.9471674092934437,
+      "grad_norm": 0.8754701018333435,
+      "learning_rate": 3.4273302274968546e-05,
+      "loss": 2.2567,
+      "step": 18600
+    },
+    {
+      "epoch": 0.9522597071928708,
+      "grad_norm": 0.9762224555015564,
+      "learning_rate": 3.4188288502737446e-05,
+      "loss": 2.2574,
+      "step": 18700
+    },
+    {
+      "epoch": 0.9573520050922979,
+      "grad_norm": 0.8961549401283264,
+      "learning_rate": 3.4103274730506345e-05,
+      "loss": 2.252,
+      "step": 18800
+    },
+    {
+      "epoch": 0.962444302991725,
+      "grad_norm": 0.8942741751670837,
+      "learning_rate": 3.4018260958275244e-05,
+      "loss": 2.3098,
+      "step": 18900
+    },
+    {
+      "epoch": 0.9675366008911521,
+      "grad_norm": 0.8678953051567078,
+      "learning_rate": 3.3933247186044144e-05,
+      "loss": 2.2751,
+      "step": 19000
+    },
+    {
+      "epoch": 0.9726288987905792,
+      "grad_norm": 0.9803009629249573,
+      "learning_rate": 3.384823341381304e-05,
+      "loss": 2.2329,
+      "step": 19100
+    },
+    {
+      "epoch": 0.9777211966900063,
+      "grad_norm": 0.8548142313957214,
+      "learning_rate": 3.376321964158194e-05,
+      "loss": 2.2577,
+      "step": 19200
+    },
+    {
+      "epoch": 0.9828134945894335,
+      "grad_norm": 0.8247301578521729,
+      "learning_rate": 3.3678205869350835e-05,
+      "loss": 2.2776,
+      "step": 19300
+    },
+    {
+      "epoch": 0.9879057924888606,
+      "grad_norm": 0.8970145583152771,
+      "learning_rate": 3.3593192097119734e-05,
+      "loss": 2.2436,
+      "step": 19400
+    },
+    {
+      "epoch": 0.9929980903882877,
+      "grad_norm": 0.9450452923774719,
+      "learning_rate": 3.3508178324888633e-05,
+      "loss": 2.274,
+      "step": 19500
+    },
+    {
+      "epoch": 0.9980903882877148,
+      "grad_norm": 0.9455347061157227,
+      "learning_rate": 3.342316455265753e-05,
+      "loss": 2.2618,
+      "step": 19600
+    },
+    {
+      "epoch": 1.0031572246976448,
+      "grad_norm": 0.9727960228919983,
+      "learning_rate": 3.333815078042643e-05,
+      "loss": 2.2148,
+      "step": 19700
+    },
+    {
+      "epoch": 1.008249522597072,
+      "grad_norm": 1.0244638919830322,
+      "learning_rate": 3.325313700819533e-05,
+      "loss": 2.2209,
+      "step": 19800
+    },
+    {
+      "epoch": 1.013341820496499,
+      "grad_norm": 1.002837061882019,
+      "learning_rate": 3.3168123235964224e-05,
+      "loss": 2.2011,
+      "step": 19900
+    },
+    {
+      "epoch": 1.0184341183959262,
+      "grad_norm": 0.8974801898002625,
+      "learning_rate": 3.308310946373312e-05,
+      "loss": 2.2186,
+      "step": 20000
+    },
+    {
+      "epoch": 1.0235264162953532,
+      "grad_norm": 1.0660030841827393,
+      "learning_rate": 3.299809569150202e-05,
+      "loss": 2.2368,
+      "step": 20100
+    },
+    {
+      "epoch": 1.0286187141947805,
+      "grad_norm": 0.8874944448471069,
+      "learning_rate": 3.291308191927092e-05,
+      "loss": 2.2552,
+      "step": 20200
+    },
+    {
+      "epoch": 1.0337110120942075,
+      "grad_norm": 0.9332163333892822,
+      "learning_rate": 3.282806814703982e-05,
+      "loss": 2.231,
+      "step": 20300
+    },
+    {
+      "epoch": 1.0388033099936347,
+      "grad_norm": 0.8272064328193665,
+      "learning_rate": 3.274305437480872e-05,
+      "loss": 2.2287,
+      "step": 20400
+    },
+    {
+      "epoch": 1.0438956078930617,
+      "grad_norm": 0.8333924412727356,
+      "learning_rate": 3.265804060257761e-05,
+      "loss": 2.2217,
+      "step": 20500
+    },
+    {
+      "epoch": 1.048987905792489,
+      "grad_norm": 0.9589939117431641,
+      "learning_rate": 3.257302683034652e-05,
+      "loss": 2.2328,
+      "step": 20600
+    },
+    {
+      "epoch": 1.054080203691916,
+      "grad_norm": 0.8918903470039368,
+      "learning_rate": 3.248801305811542e-05,
+      "loss": 2.2169,
+      "step": 20700
+    },
+    {
+      "epoch": 1.0591725015913431,
+      "grad_norm": 0.9166114926338196,
+      "learning_rate": 3.240299928588432e-05,
+      "loss": 2.2605,
+      "step": 20800
+    },
+    {
+      "epoch": 1.0642647994907701,
+      "grad_norm": 0.8604680895805359,
+      "learning_rate": 3.231798551365322e-05,
+      "loss": 2.2591,
+      "step": 20900
+    },
+    {
+      "epoch": 1.0693570973901974,
+      "grad_norm": 0.82822185754776,
+      "learning_rate": 3.2232971741422117e-05,
+      "loss": 2.2075,
+      "step": 21000
+    },
+    {
+      "epoch": 1.0744493952896244,
+      "grad_norm": 0.8195912837982178,
+      "learning_rate": 3.214795796919101e-05,
+      "loss": 2.2054,
+      "step": 21100
+    },
+    {
+      "epoch": 1.0795416931890516,
+      "grad_norm": 0.9587050080299377,
+      "learning_rate": 3.206294419695991e-05,
+      "loss": 2.2558,
+      "step": 21200
+    },
+    {
+      "epoch": 1.0846339910884786,
+      "grad_norm": 0.9604052901268005,
+      "learning_rate": 3.197793042472881e-05,
+      "loss": 2.2023,
+      "step": 21300
+    },
+    {
+      "epoch": 1.0897262889879058,
+      "grad_norm": 0.9480250477790833,
+      "learning_rate": 3.189291665249771e-05,
+      "loss": 2.2168,
+      "step": 21400
+    },
+    {
+      "epoch": 1.094818586887333,
+      "grad_norm": 0.8999929428100586,
+      "learning_rate": 3.1807902880266606e-05,
+      "loss": 2.2089,
+      "step": 21500
+    },
+    {
+      "epoch": 1.09991088478676,
+      "grad_norm": 0.9180619716644287,
+      "learning_rate": 3.1722889108035506e-05,
+      "loss": 2.2092,
+      "step": 21600
+    },
+    {
+      "epoch": 1.105003182686187,
+      "grad_norm": 0.8434627056121826,
+      "learning_rate": 3.16378753358044e-05,
+      "loss": 2.2179,
+      "step": 21700
+    },
+    {
+      "epoch": 1.1100954805856142,
+      "grad_norm": 0.8810749053955078,
+      "learning_rate": 3.15528615635733e-05,
+      "loss": 2.1857,
+      "step": 21800
+    },
+    {
+      "epoch": 1.1151877784850415,
+      "grad_norm": 0.9257334470748901,
+      "learning_rate": 3.14678477913422e-05,
+      "loss": 2.2205,
+      "step": 21900
+    },
+    {
+      "epoch": 1.1202800763844685,
+      "grad_norm": 0.8661274313926697,
+      "learning_rate": 3.1382834019111096e-05,
+      "loss": 2.1995,
+      "step": 22000
+    },
+    {
+      "epoch": 1.1253723742838957,
+      "grad_norm": 0.8728938698768616,
+      "learning_rate": 3.1297820246879996e-05,
+      "loss": 2.2125,
+      "step": 22100
+    },
+    {
+      "epoch": 1.1304646721833227,
+      "grad_norm": 0.9176629185676575,
+      "learning_rate": 3.1212806474648895e-05,
+      "loss": 2.1908,
+      "step": 22200
+    },
+    {
+      "epoch": 1.13555697008275,
+      "grad_norm": 0.9520237445831299,
+      "learning_rate": 3.112779270241779e-05,
+      "loss": 2.2345,
+      "step": 22300
+    },
+    {
+      "epoch": 1.140649267982177,
+      "grad_norm": 0.8356249928474426,
+      "learning_rate": 3.1042778930186694e-05,
+      "loss": 2.2452,
+      "step": 22400
+    },
+    {
+      "epoch": 1.1457415658816041,
+      "grad_norm": 1.0978131294250488,
+      "learning_rate": 3.095776515795559e-05,
+      "loss": 2.1776,
+      "step": 22500
+    },
+    {
+      "epoch": 1.1508338637810311,
+      "grad_norm": 1.1184298992156982,
+      "learning_rate": 3.087275138572449e-05,
+      "loss": 2.2174,
+      "step": 22600
+    },
+    {
+      "epoch": 1.1559261616804584,
+      "grad_norm": 0.9109058380126953,
+      "learning_rate": 3.078773761349339e-05,
+      "loss": 2.2168,
+      "step": 22700
+    },
+    {
+      "epoch": 1.1610184595798854,
+      "grad_norm": 0.8274030089378357,
+      "learning_rate": 3.0702723841262284e-05,
+      "loss": 2.224,
+      "step": 22800
+    },
+    {
+      "epoch": 1.1661107574793126,
+      "grad_norm": 0.8593317270278931,
+      "learning_rate": 3.0617710069031183e-05,
+      "loss": 2.2653,
+      "step": 22900
+    },
+    {
+      "epoch": 1.1712030553787396,
+      "grad_norm": 1.1305369138717651,
+      "learning_rate": 3.053269629680008e-05,
+      "loss": 2.241,
+      "step": 23000
+    },
+    {
+      "epoch": 1.1762953532781668,
+      "grad_norm": 1.0249735116958618,
+      "learning_rate": 3.0447682524568982e-05,
+      "loss": 2.2044,
+      "step": 23100
+    },
+    {
+      "epoch": 1.1813876511775938,
+      "grad_norm": 0.762690007686615,
+      "learning_rate": 3.036266875233788e-05,
+      "loss": 2.2057,
+      "step": 23200
+    },
+    {
+      "epoch": 1.186479949077021,
+      "grad_norm": 0.7995686531066895,
+      "learning_rate": 3.0277654980106777e-05,
+      "loss": 2.2435,
+      "step": 23300
+    },
+    {
+      "epoch": 1.191572246976448,
+      "grad_norm": 1.0537996292114258,
+      "learning_rate": 3.0192641207875677e-05,
+      "loss": 2.2155,
+      "step": 23400
+    },
+    {
+      "epoch": 1.1966645448758753,
+      "grad_norm": 0.8992569446563721,
+      "learning_rate": 3.0107627435644576e-05,
+      "loss": 2.217,
+      "step": 23500
+    },
+    {
+      "epoch": 1.2017568427753023,
+      "grad_norm": 0.9041591286659241,
+      "learning_rate": 3.0022613663413472e-05,
+      "loss": 2.2277,
+      "step": 23600
+    },
+    {
+      "epoch": 1.2068491406747295,
+      "grad_norm": 0.9437869787216187,
+      "learning_rate": 2.993759989118237e-05,
+      "loss": 2.2151,
+      "step": 23700
+    },
+    {
+      "epoch": 1.2119414385741565,
+      "grad_norm": 0.7999377846717834,
+      "learning_rate": 2.985258611895127e-05,
+      "loss": 2.2103,
+      "step": 23800
+    },
+    {
+      "epoch": 1.2170337364735837,
+      "grad_norm": 0.932995080947876,
+      "learning_rate": 2.976757234672017e-05,
+      "loss": 2.1964,
+      "step": 23900
+    },
+    {
+      "epoch": 1.222126034373011,
+      "grad_norm": 0.846868097782135,
+      "learning_rate": 2.9682558574489066e-05,
+      "loss": 2.1821,
+      "step": 24000
+    },
+    {
+      "epoch": 1.227218332272438,
+      "grad_norm": 0.889284610748291,
+      "learning_rate": 2.9597544802257965e-05,
+      "loss": 2.2227,
+      "step": 24100
+    },
+    {
+      "epoch": 1.2323106301718652,
+      "grad_norm": 0.9376260042190552,
+      "learning_rate": 2.9512531030026865e-05,
+      "loss": 2.226,
+      "step": 24200
+    },
+    {
+      "epoch": 1.2374029280712922,
+      "grad_norm": 0.8779696226119995,
+      "learning_rate": 2.9427517257795767e-05,
+      "loss": 2.2086,
+      "step": 24300
+    },
+    {
+      "epoch": 1.2424952259707194,
+      "grad_norm": 0.9524549841880798,
+      "learning_rate": 2.9342503485564667e-05,
+      "loss": 2.2026,
+      "step": 24400
+    },
+    {
+      "epoch": 1.2475875238701464,
+      "grad_norm": 0.919808030128479,
+      "learning_rate": 2.9257489713333563e-05,
+      "loss": 2.192,
+      "step": 24500
+    },
+    {
+      "epoch": 1.2526798217695736,
+      "grad_norm": 1.0228092670440674,
+      "learning_rate": 2.9172475941102462e-05,
+      "loss": 2.2241,
+      "step": 24600
+    },
+    {
+      "epoch": 1.2577721196690006,
+      "grad_norm": 0.8363624811172485,
+      "learning_rate": 2.908746216887136e-05,
+      "loss": 2.1808,
+      "step": 24700
+    },
+    {
+      "epoch": 1.2628644175684278,
+      "grad_norm": 0.8711551427841187,
+      "learning_rate": 2.9002448396640257e-05,
+      "loss": 2.2093,
+      "step": 24800
+    },
+    {
+      "epoch": 1.2679567154678548,
+      "grad_norm": 0.9497014284133911,
+      "learning_rate": 2.8917434624409156e-05,
+      "loss": 2.1856,
+      "step": 24900
+    },
+    {
+      "epoch": 1.273049013367282,
+      "grad_norm": 0.9282352924346924,
+      "learning_rate": 2.8832420852178056e-05,
+      "loss": 2.1787,
+      "step": 25000
+    },
+    {
+      "epoch": 1.278141311266709,
+      "grad_norm": 0.9017792344093323,
+      "learning_rate": 2.8747407079946952e-05,
+      "loss": 2.2054,
+      "step": 25100
+    },
+    {
+      "epoch": 1.2832336091661363,
+      "grad_norm": 0.9470519423484802,
+      "learning_rate": 2.866239330771585e-05,
+      "loss": 2.1885,
+      "step": 25200
+    },
+    {
+      "epoch": 1.2883259070655633,
+      "grad_norm": 0.991397500038147,
+      "learning_rate": 2.857737953548475e-05,
+      "loss": 2.1875,
+      "step": 25300
+    },
+    {
+      "epoch": 1.2934182049649905,
+      "grad_norm": 0.920644223690033,
+      "learning_rate": 2.8492365763253646e-05,
+      "loss": 2.2418,
+      "step": 25400
+    },
+    {
+      "epoch": 1.2985105028644175,
+      "grad_norm": 0.8312422037124634,
+      "learning_rate": 2.8407351991022546e-05,
+      "loss": 2.1635,
+      "step": 25500
+    },
+    {
+      "epoch": 1.3036028007638447,
+      "grad_norm": 0.9457144737243652,
+      "learning_rate": 2.8322338218791445e-05,
+      "loss": 2.1945,
+      "step": 25600
+    },
+    {
+      "epoch": 1.308695098663272,
+      "grad_norm": 0.8914629220962524,
+      "learning_rate": 2.8237324446560344e-05,
+      "loss": 2.2092,
+      "step": 25700
+    },
+    {
+      "epoch": 1.313787396562699,
+      "grad_norm": 0.9140703082084656,
+      "learning_rate": 2.815231067432924e-05,
+      "loss": 2.2162,
+      "step": 25800
+    },
+    {
+      "epoch": 1.318879694462126,
+      "grad_norm": 0.926543116569519,
+      "learning_rate": 2.806729690209814e-05,
+      "loss": 2.1906,
+      "step": 25900
+    },
+    {
+      "epoch": 1.3239719923615532,
+      "grad_norm": 0.888692319393158,
+      "learning_rate": 2.798228312986704e-05,
+      "loss": 2.1866,
+      "step": 26000
+    },
+    {
+      "epoch": 1.3290642902609804,
+      "grad_norm": 0.7925876379013062,
+      "learning_rate": 2.7897269357635935e-05,
+      "loss": 2.1988,
+      "step": 26100
+    },
+    {
+      "epoch": 1.3341565881604074,
+      "grad_norm": 0.8814985752105713,
+      "learning_rate": 2.781225558540484e-05,
+      "loss": 2.2072,
+      "step": 26200
+    },
+    {
+      "epoch": 1.3392488860598344,
+      "grad_norm": 0.8415858745574951,
+      "learning_rate": 2.7727241813173737e-05,
+      "loss": 2.2227,
+      "step": 26300
+    },
+    {
+      "epoch": 1.3443411839592616,
+      "grad_norm": 0.9423860907554626,
+      "learning_rate": 2.7642228040942636e-05,
+      "loss": 2.2426,
+      "step": 26400
+    },
+    {
+      "epoch": 1.3494334818586888,
+      "grad_norm": 0.8816553950309753,
+      "learning_rate": 2.7557214268711535e-05,
+      "loss": 2.206,
+      "step": 26500
+    },
+    {
+      "epoch": 1.3545257797581158,
+      "grad_norm": 0.8283177018165588,
+      "learning_rate": 2.747220049648043e-05,
+      "loss": 2.1859,
+      "step": 26600
+    },
+    {
+      "epoch": 1.3596180776575428,
+      "grad_norm": 0.8860555291175842,
+      "learning_rate": 2.738718672424933e-05,
+      "loss": 2.178,
+      "step": 26700
+    },
+    {
+      "epoch": 1.36471037555697,
+      "grad_norm": 0.8853309154510498,
+      "learning_rate": 2.730217295201823e-05,
+      "loss": 2.1844,
+      "step": 26800
+    },
+    {
+      "epoch": 1.3698026734563973,
+      "grad_norm": 0.9043028950691223,
+      "learning_rate": 2.7217159179787126e-05,
+      "loss": 2.2105,
+      "step": 26900
+    },
+    {
+      "epoch": 1.3748949713558243,
+      "grad_norm": 0.8943936824798584,
+      "learning_rate": 2.7132145407556025e-05,
+      "loss": 2.1814,
+      "step": 27000
+    },
+    {
+      "epoch": 1.3799872692552515,
+      "grad_norm": 0.7901210188865662,
+      "learning_rate": 2.7047131635324925e-05,
+      "loss": 2.1819,
+      "step": 27100
+    },
+    {
+      "epoch": 1.3850795671546785,
+      "grad_norm": 0.9602735638618469,
+      "learning_rate": 2.696211786309382e-05,
+      "loss": 2.2121,
+      "step": 27200
+    },
+    {
+      "epoch": 1.3901718650541057,
+      "grad_norm": 0.8327048420906067,
+      "learning_rate": 2.687710409086272e-05,
+      "loss": 2.2128,
+      "step": 27300
+    },
+    {
+      "epoch": 1.3952641629535327,
+      "grad_norm": 0.8546739220619202,
+      "learning_rate": 2.679209031863162e-05,
+      "loss": 2.2035,
+      "step": 27400
+    },
+    {
+      "epoch": 1.40035646085296,
+      "grad_norm": 1.585236668586731,
+      "learning_rate": 2.6707076546400515e-05,
+      "loss": 2.1845,
+      "step": 27500
+    },
+    {
+      "epoch": 1.405448758752387,
+      "grad_norm": 0.9497547745704651,
+      "learning_rate": 2.6622062774169415e-05,
+      "loss": 2.1886,
+      "step": 27600
+    },
+    {
+      "epoch": 1.4105410566518142,
+      "grad_norm": 0.8747720718383789,
+      "learning_rate": 2.6537049001938314e-05,
+      "loss": 2.1735,
+      "step": 27700
+    },
+    {
+      "epoch": 1.4156333545512412,
+      "grad_norm": 0.9204273223876953,
+      "learning_rate": 2.6452035229707213e-05,
+      "loss": 2.2153,
+      "step": 27800
+    },
+    {
+      "epoch": 1.4207256524506684,
+      "grad_norm": 0.868325412273407,
+      "learning_rate": 2.636702145747611e-05,
+      "loss": 2.209,
+      "step": 27900
+    },
+    {
+      "epoch": 1.4258179503500954,
+      "grad_norm": 0.9367715716362,
+      "learning_rate": 2.6282007685245015e-05,
+      "loss": 2.1868,
+      "step": 28000
+    },
+    {
+      "epoch": 1.4309102482495226,
+      "grad_norm": 0.9658358693122864,
+      "learning_rate": 2.619699391301391e-05,
+      "loss": 2.1757,
+      "step": 28100
+    },
+    {
+      "epoch": 1.4360025461489498,
+      "grad_norm": 0.8091734051704407,
+      "learning_rate": 2.611198014078281e-05,
+      "loss": 2.1878,
+      "step": 28200
+    },
+    {
+      "epoch": 1.4410948440483768,
+      "grad_norm": 0.8200072050094604,
+      "learning_rate": 2.602696636855171e-05,
+      "loss": 2.192,
+      "step": 28300
+    },
+    {
+      "epoch": 1.4461871419478038,
+      "grad_norm": 0.9280868768692017,
+      "learning_rate": 2.5941952596320606e-05,
+      "loss": 2.1829,
+      "step": 28400
+    },
+    {
+      "epoch": 1.451279439847231,
+      "grad_norm": 0.9731032252311707,
+      "learning_rate": 2.5856938824089505e-05,
+      "loss": 2.156,
+      "step": 28500
+    },
+    {
+      "epoch": 1.4563717377466583,
+      "grad_norm": 0.8023040294647217,
+      "learning_rate": 2.5771925051858404e-05,
+      "loss": 2.1913,
+      "step": 28600
+    },
+    {
+      "epoch": 1.4614640356460853,
+      "grad_norm": 1.003476619720459,
+      "learning_rate": 2.56869112796273e-05,
+      "loss": 2.1537,
+      "step": 28700
+    },
+    {
+      "epoch": 1.4665563335455123,
+      "grad_norm": 1.0280425548553467,
+      "learning_rate": 2.56018975073962e-05,
+      "loss": 2.2106,
+      "step": 28800
+    },
+    {
+      "epoch": 1.4716486314449395,
+      "grad_norm": 0.9685016870498657,
+      "learning_rate": 2.55168837351651e-05,
+      "loss": 2.1758,
+      "step": 28900
+    },
+    {
+      "epoch": 1.4767409293443667,
+      "grad_norm": 0.8572561144828796,
+      "learning_rate": 2.5431869962933995e-05,
+      "loss": 2.1647,
+      "step": 29000
+    },
+    {
+      "epoch": 1.4818332272437937,
+      "grad_norm": 0.8688543438911438,
+      "learning_rate": 2.5346856190702894e-05,
+      "loss": 2.1973,
+      "step": 29100
+    },
+    {
+      "epoch": 1.486925525143221,
+      "grad_norm": 1.0197324752807617,
+      "learning_rate": 2.5261842418471794e-05,
+      "loss": 2.1649,
+      "step": 29200
+    },
+    {
+      "epoch": 1.492017823042648,
+      "grad_norm": 0.8760496377944946,
+      "learning_rate": 2.517682864624069e-05,
+      "loss": 2.2024,
+      "step": 29300
+    },
+    {
+      "epoch": 1.4971101209420752,
+      "grad_norm": 0.9327671527862549,
+      "learning_rate": 2.509181487400959e-05,
+      "loss": 2.2006,
+      "step": 29400
+    },
+    {
+      "epoch": 1.5022024188415022,
+      "grad_norm": 0.9184695482254028,
+      "learning_rate": 2.5006801101778488e-05,
+      "loss": 2.1616,
+      "step": 29500
+    },
+    {
+      "epoch": 1.5072947167409292,
+      "grad_norm": 0.8531858325004578,
+      "learning_rate": 2.4921787329547387e-05,
+      "loss": 2.1688,
+      "step": 29600
+    },
+    {
+      "epoch": 1.5123870146403564,
+      "grad_norm": 0.8902334570884705,
+      "learning_rate": 2.4836773557316287e-05,
+      "loss": 2.1692,
+      "step": 29700
+    },
+    {
+      "epoch": 1.5174793125397836,
+      "grad_norm": 0.8231461644172668,
+      "learning_rate": 2.4751759785085186e-05,
+      "loss": 2.1855,
+      "step": 29800
+    },
+    {
+      "epoch": 1.5225716104392109,
+      "grad_norm": 0.9362125396728516,
+      "learning_rate": 2.4666746012854082e-05,
+      "loss": 2.1798,
+      "step": 29900
+    },
+    {
+      "epoch": 1.5276639083386379,
+      "grad_norm": 0.8145864009857178,
+      "learning_rate": 2.458173224062298e-05,
+      "loss": 2.1655,
+      "step": 30000
+    },
+    {
+      "epoch": 1.5327562062380649,
+      "grad_norm": 0.9912553429603577,
+      "learning_rate": 2.449671846839188e-05,
+      "loss": 2.2025,
+      "step": 30100
+    },
+    {
+      "epoch": 1.537848504137492,
+      "grad_norm": 0.818953275680542,
+      "learning_rate": 2.4411704696160777e-05,
+      "loss": 2.1845,
+      "step": 30200
+    },
+    {
+      "epoch": 1.5429408020369193,
+      "grad_norm": 0.845649778842926,
+      "learning_rate": 2.4326690923929676e-05,
+      "loss": 2.199,
+      "step": 30300
+    },
+    {
+      "epoch": 1.5480330999363463,
+      "grad_norm": 1.0135074853897095,
+      "learning_rate": 2.424167715169858e-05,
+      "loss": 2.1912,
+      "step": 30400
+    },
+    {
+      "epoch": 1.5531253978357733,
+      "grad_norm": 0.9612752199172974,
+      "learning_rate": 2.4156663379467475e-05,
+      "loss": 2.159,
+      "step": 30500
+    },
+    {
+      "epoch": 1.5582176957352005,
+      "grad_norm": 0.8450791239738464,
+      "learning_rate": 2.4071649607236374e-05,
+      "loss": 2.1615,
+      "step": 30600
+    },
+    {
+      "epoch": 1.5633099936346277,
+      "grad_norm": 0.9979317784309387,
+      "learning_rate": 2.3986635835005273e-05,
+      "loss": 2.1713,
+      "step": 30700
+    },
+    {
+      "epoch": 1.5684022915340547,
+      "grad_norm": 0.904403567314148,
+      "learning_rate": 2.390162206277417e-05,
+      "loss": 2.2114,
+      "step": 30800
+    },
+    {
+      "epoch": 1.5734945894334817,
+      "grad_norm": 0.8977887630462646,
+      "learning_rate": 2.381660829054307e-05,
+      "loss": 2.1867,
+      "step": 30900
+    },
+    {
+      "epoch": 1.578586887332909,
+      "grad_norm": 0.9076321125030518,
+      "learning_rate": 2.3731594518311968e-05,
+      "loss": 2.167,
+      "step": 31000
+    },
+    {
+      "epoch": 1.5836791852323362,
+      "grad_norm": 0.9048725962638855,
+      "learning_rate": 2.3646580746080864e-05,
+      "loss": 2.1645,
+      "step": 31100
+    },
+    {
+      "epoch": 1.5887714831317632,
+      "grad_norm": 0.9547775387763977,
+      "learning_rate": 2.3561566973849763e-05,
+      "loss": 2.1849,
+      "step": 31200
+    },
+    {
+      "epoch": 1.5938637810311902,
+      "grad_norm": 0.7886509299278259,
+      "learning_rate": 2.3476553201618666e-05,
+      "loss": 2.187,
+      "step": 31300
+    },
+    {
+      "epoch": 1.5989560789306174,
+      "grad_norm": 0.8473970293998718,
+      "learning_rate": 2.3391539429387562e-05,
+      "loss": 2.1722,
+      "step": 31400
+    },
+    {
+      "epoch": 1.6040483768300446,
+      "grad_norm": 0.8617937564849854,
+      "learning_rate": 2.330652565715646e-05,
+      "loss": 2.2002,
+      "step": 31500
+    },
+    {
+      "epoch": 1.6091406747294716,
+      "grad_norm": 0.9672524333000183,
+      "learning_rate": 2.322151188492536e-05,
+      "loss": 2.1623,
+      "step": 31600
+    },
+    {
+      "epoch": 1.6142329726288986,
+      "grad_norm": 0.8769922852516174,
+      "learning_rate": 2.3136498112694256e-05,
+      "loss": 2.1695,
+      "step": 31700
+    },
+    {
+      "epoch": 1.6193252705283259,
+      "grad_norm": 0.8249488472938538,
+      "learning_rate": 2.3051484340463156e-05,
+      "loss": 2.1647,
+      "step": 31800
+    },
+    {
+      "epoch": 1.624417568427753,
+      "grad_norm": 0.9503587484359741,
+      "learning_rate": 2.2966470568232055e-05,
+      "loss": 2.2024,
+      "step": 31900
+    },
+    {
+      "epoch": 1.62950986632718,
+      "grad_norm": 0.9500870108604431,
+      "learning_rate": 2.288145679600095e-05,
+      "loss": 2.1467,
+      "step": 32000
+    },
+    {
+      "epoch": 1.634602164226607,
+      "grad_norm": 0.888297975063324,
+      "learning_rate": 2.279644302376985e-05,
+      "loss": 2.1586,
+      "step": 32100
+    },
+    {
+      "epoch": 1.6396944621260343,
+      "grad_norm": 0.8958535194396973,
+      "learning_rate": 2.2711429251538753e-05,
+      "loss": 2.1923,
+      "step": 32200
+    },
+    {
+      "epoch": 1.6447867600254615,
+      "grad_norm": 0.7949930429458618,
+      "learning_rate": 2.262641547930765e-05,
+      "loss": 2.1925,
+      "step": 32300
+    },
+    {
+      "epoch": 1.6498790579248888,
+      "grad_norm": 0.8516358733177185,
+      "learning_rate": 2.2541401707076548e-05,
+      "loss": 2.1818,
+      "step": 32400
+    },
+    {
+      "epoch": 1.6549713558243158,
+      "grad_norm": 0.9597014784812927,
+      "learning_rate": 2.2456387934845448e-05,
+      "loss": 2.1412,
+      "step": 32500
+    },
+    {
+      "epoch": 1.6600636537237428,
+      "grad_norm": 0.8643897771835327,
+      "learning_rate": 2.2371374162614344e-05,
+      "loss": 2.1645,
+      "step": 32600
+    },
+    {
+      "epoch": 1.66515595162317,
+      "grad_norm": 1.069393515586853,
+      "learning_rate": 2.2286360390383243e-05,
+      "loss": 2.1468,
+      "step": 32700
+    },
+    {
+      "epoch": 1.6702482495225972,
+      "grad_norm": 0.8896872401237488,
+      "learning_rate": 2.2201346618152142e-05,
+      "loss": 2.1732,
+      "step": 32800
+    },
+    {
+      "epoch": 1.6753405474220242,
+      "grad_norm": 0.8662711381912231,
+      "learning_rate": 2.2116332845921038e-05,
+      "loss": 2.1901,
+      "step": 32900
+    },
+    {
+      "epoch": 1.6804328453214512,
+      "grad_norm": 0.7606475353240967,
+      "learning_rate": 2.2031319073689937e-05,
+      "loss": 2.2045,
+      "step": 33000
+    },
+    {
+      "epoch": 1.6855251432208784,
+      "grad_norm": 0.9675360918045044,
+      "learning_rate": 2.1946305301458837e-05,
+      "loss": 2.1782,
+      "step": 33100
+    },
+    {
+      "epoch": 1.6906174411203057,
+      "grad_norm": 0.8184406757354736,
+      "learning_rate": 2.1861291529227736e-05,
+      "loss": 2.1827,
+      "step": 33200
+    },
+    {
+      "epoch": 1.6957097390197327,
+      "grad_norm": 0.8774561882019043,
+      "learning_rate": 2.1776277756996635e-05,
+      "loss": 2.1592,
+      "step": 33300
+    },
+    {
+      "epoch": 1.7008020369191597,
+      "grad_norm": 0.8667624592781067,
+      "learning_rate": 2.1691263984765535e-05,
+      "loss": 2.1779,
+      "step": 33400
+    },
+    {
+      "epoch": 1.7058943348185869,
+      "grad_norm": 0.9804625511169434,
+      "learning_rate": 2.160625021253443e-05,
+      "loss": 2.1985,
+      "step": 33500
+    },
+    {
+      "epoch": 1.710986632718014,
+      "grad_norm": 0.9614706039428711,
+      "learning_rate": 2.152123644030333e-05,
+      "loss": 2.1687,
+      "step": 33600
+    },
+    {
+      "epoch": 1.716078930617441,
+      "grad_norm": 0.8270972967147827,
+      "learning_rate": 2.143622266807223e-05,
+      "loss": 2.1539,
+      "step": 33700
+    },
+    {
+      "epoch": 1.721171228516868,
+      "grad_norm": 0.9252774119377136,
+      "learning_rate": 2.1351208895841125e-05,
+      "loss": 2.16,
+      "step": 33800
+    },
+    {
+      "epoch": 1.7262635264162953,
+      "grad_norm": 0.855818510055542,
+      "learning_rate": 2.1266195123610025e-05,
+      "loss": 2.1928,
+      "step": 33900
+    },
+    {
+      "epoch": 1.7313558243157225,
+      "grad_norm": 0.8505380153656006,
+      "learning_rate": 2.1181181351378924e-05,
+      "loss": 2.1748,
+      "step": 34000
+    },
+    {
+      "epoch": 1.7364481222151495,
+      "grad_norm": 0.8876926898956299,
+      "learning_rate": 2.1096167579147823e-05,
+      "loss": 2.2102,
+      "step": 34100
+    },
+    {
+      "epoch": 1.7415404201145765,
+      "grad_norm": 0.8772891163825989,
+      "learning_rate": 2.1011153806916723e-05,
+      "loss": 2.1691,
+      "step": 34200
+    },
+    {
+      "epoch": 1.7466327180140038,
+      "grad_norm": 0.9799501299858093,
+      "learning_rate": 2.0926140034685622e-05,
+      "loss": 2.1858,
+      "step": 34300
+    },
+    {
+      "epoch": 1.751725015913431,
+      "grad_norm": 0.8863718509674072,
+      "learning_rate": 2.0841126262454518e-05,
+      "loss": 2.1745,
+      "step": 34400
+    },
+    {
+      "epoch": 1.7568173138128582,
+      "grad_norm": 0.8394114375114441,
+      "learning_rate": 2.0756112490223417e-05,
+      "loss": 2.1629,
+      "step": 34500
+    },
+    {
+      "epoch": 1.7619096117122852,
+      "grad_norm": 0.8472095727920532,
+      "learning_rate": 2.0671098717992317e-05,
+      "loss": 2.1665,
+      "step": 34600
+    },
+    {
+      "epoch": 1.7670019096117122,
+      "grad_norm": 0.9460027813911438,
+      "learning_rate": 2.0586084945761212e-05,
+      "loss": 2.1458,
+      "step": 34700
+    },
+    {
+      "epoch": 1.7720942075111394,
+      "grad_norm": 0.9211781620979309,
+      "learning_rate": 2.0501071173530112e-05,
+      "loss": 2.1922,
+      "step": 34800
+    },
+    {
+      "epoch": 1.7771865054105667,
+      "grad_norm": 0.9996361136436462,
+      "learning_rate": 2.041605740129901e-05,
+      "loss": 2.1447,
+      "step": 34900
+    },
+    {
+      "epoch": 1.7822788033099937,
+      "grad_norm": 0.8266726136207581,
+      "learning_rate": 2.033104362906791e-05,
+      "loss": 2.1881,
+      "step": 35000
+    },
+    {
+      "epoch": 1.7873711012094207,
+      "grad_norm": 0.8855674862861633,
+      "learning_rate": 2.024602985683681e-05,
+      "loss": 2.198,
+      "step": 35100
+    },
+    {
+      "epoch": 1.7924633991088479,
+      "grad_norm": 0.9789201021194458,
+      "learning_rate": 2.016101608460571e-05,
+      "loss": 2.1685,
+      "step": 35200
+    },
+    {
+      "epoch": 1.797555697008275,
+      "grad_norm": 0.8354413509368896,
+      "learning_rate": 2.0076002312374605e-05,
+      "loss": 2.1535,
+      "step": 35300
+    },
+    {
+      "epoch": 1.8026479949077021,
+      "grad_norm": 0.9418453574180603,
+      "learning_rate": 1.9990988540143504e-05,
+      "loss": 2.1671,
+      "step": 35400
+    },
+    {
+      "epoch": 1.8077402928071291,
+      "grad_norm": 0.9462503790855408,
+      "learning_rate": 1.9905974767912404e-05,
+      "loss": 2.1339,
+      "step": 35500
+    },
+    {
+      "epoch": 1.8128325907065563,
+      "grad_norm": 0.8490837216377258,
+      "learning_rate": 1.98209609956813e-05,
+      "loss": 2.1528,
+      "step": 35600
+    },
+    {
+      "epoch": 1.8179248886059836,
+      "grad_norm": 0.9105218052864075,
+      "learning_rate": 1.97359472234502e-05,
+      "loss": 2.1717,
+      "step": 35700
+    },
+    {
+      "epoch": 1.8230171865054106,
+      "grad_norm": 0.9058020710945129,
+      "learning_rate": 1.9650933451219098e-05,
+      "loss": 2.1535,
+      "step": 35800
+    },
+    {
+      "epoch": 1.8281094844048376,
+      "grad_norm": 0.9724037647247314,
+      "learning_rate": 1.9565919678987994e-05,
+      "loss": 2.166,
+      "step": 35900
+    },
+    {
+      "epoch": 1.8332017823042648,
+      "grad_norm": 0.9018999338150024,
+      "learning_rate": 1.9480905906756897e-05,
+      "loss": 2.1528,
+      "step": 36000
+    },
+    {
+      "epoch": 1.838294080203692,
+      "grad_norm": 0.9223784804344177,
+      "learning_rate": 1.9395892134525796e-05,
+      "loss": 2.1982,
+      "step": 36100
+    },
+    {
+      "epoch": 1.843386378103119,
+      "grad_norm": 0.8883550763130188,
+      "learning_rate": 1.9310878362294692e-05,
+      "loss": 2.1701,
+      "step": 36200
+    },
+    {
+      "epoch": 1.848478676002546,
+      "grad_norm": 0.8294488787651062,
+      "learning_rate": 1.922586459006359e-05,
+      "loss": 2.2064,
+      "step": 36300
+    },
+    {
+      "epoch": 1.8535709739019732,
+      "grad_norm": 0.8737560510635376,
+      "learning_rate": 1.914085081783249e-05,
+      "loss": 2.1529,
+      "step": 36400
+    },
+    {
+      "epoch": 1.8586632718014005,
+      "grad_norm": 0.8156319260597229,
+      "learning_rate": 1.9055837045601387e-05,
+      "loss": 2.1628,
+      "step": 36500
+    },
+    {
+      "epoch": 1.8637555697008275,
+      "grad_norm": 0.8669657111167908,
+      "learning_rate": 1.8970823273370286e-05,
+      "loss": 2.2155,
+      "step": 36600
+    },
+    {
+      "epoch": 1.8688478676002545,
+      "grad_norm": 0.8657876253128052,
+      "learning_rate": 1.8885809501139185e-05,
+      "loss": 2.1506,
+      "step": 36700
+    },
+    {
+      "epoch": 1.8739401654996817,
+      "grad_norm": 0.8771129250526428,
+      "learning_rate": 1.880079572890808e-05,
+      "loss": 2.1797,
+      "step": 36800
+    },
+    {
+      "epoch": 1.879032463399109,
+      "grad_norm": 0.8845404982566833,
+      "learning_rate": 1.8715781956676984e-05,
+      "loss": 2.1434,
+      "step": 36900
+    },
+    {
+      "epoch": 1.8841247612985361,
+      "grad_norm": 0.9354609251022339,
+      "learning_rate": 1.8630768184445883e-05,
+      "loss": 2.1701,
+      "step": 37000
+    },
+    {
+      "epoch": 1.8892170591979631,
+      "grad_norm": 0.7781304717063904,
+      "learning_rate": 1.854575441221478e-05,
+      "loss": 2.2095,
+      "step": 37100
+    },
+    {
+      "epoch": 1.8943093570973901,
+      "grad_norm": 0.9069561958312988,
+      "learning_rate": 1.846074063998368e-05,
+      "loss": 2.169,
+      "step": 37200
+    },
+    {
+      "epoch": 1.8994016549968173,
+      "grad_norm": 0.9173194766044617,
+      "learning_rate": 1.8375726867752578e-05,
+      "loss": 2.121,
+      "step": 37300
+    },
+    {
+      "epoch": 1.9044939528962446,
+      "grad_norm": 0.864583432674408,
+      "learning_rate": 1.8290713095521474e-05,
+      "loss": 2.1711,
+      "step": 37400
+    },
+    {
+      "epoch": 1.9095862507956716,
+      "grad_norm": 0.7620731592178345,
+      "learning_rate": 1.8205699323290373e-05,
+      "loss": 2.1967,
+      "step": 37500
+    },
+    {
+      "epoch": 1.9146785486950986,
+      "grad_norm": 0.7830232977867126,
+      "learning_rate": 1.8120685551059273e-05,
+      "loss": 2.1574,
+      "step": 37600
+    },
+    {
+      "epoch": 1.9197708465945258,
+      "grad_norm": 0.8825329542160034,
+      "learning_rate": 1.803567177882817e-05,
+      "loss": 2.1432,
+      "step": 37700
+    },
+    {
+      "epoch": 1.924863144493953,
+      "grad_norm": 0.9680500030517578,
+      "learning_rate": 1.795065800659707e-05,
+      "loss": 2.1808,
+      "step": 37800
+    },
+    {
+      "epoch": 1.92995544239338,
+      "grad_norm": 0.9914782047271729,
+      "learning_rate": 1.786564423436597e-05,
+      "loss": 2.1945,
+      "step": 37900
+    },
+    {
+      "epoch": 1.935047740292807,
+      "grad_norm": 0.882604718208313,
+      "learning_rate": 1.7780630462134867e-05,
+      "loss": 2.15,
+      "step": 38000
+    },
+    {
+      "epoch": 1.9401400381922342,
+      "grad_norm": 0.8211714625358582,
+      "learning_rate": 1.7695616689903766e-05,
+      "loss": 2.178,
+      "step": 38100
+    },
+    {
+      "epoch": 1.9452323360916615,
+      "grad_norm": 0.9662156701087952,
+      "learning_rate": 1.7610602917672665e-05,
+      "loss": 2.164,
+      "step": 38200
+    },
+    {
+      "epoch": 1.9503246339910885,
+      "grad_norm": 0.8627343773841858,
+      "learning_rate": 1.752558914544156e-05,
+      "loss": 2.1977,
+      "step": 38300
+    },
+    {
+      "epoch": 1.9554169318905155,
+      "grad_norm": 0.8883799910545349,
+      "learning_rate": 1.744057537321046e-05,
+      "loss": 2.1813,
+      "step": 38400
+    },
+    {
+      "epoch": 1.9605092297899427,
+      "grad_norm": 0.9309747219085693,
+      "learning_rate": 1.735556160097936e-05,
+      "loss": 2.1686,
+      "step": 38500
+    },
+    {
+      "epoch": 1.96560152768937,
+      "grad_norm": 0.9126595854759216,
+      "learning_rate": 1.7270547828748256e-05,
+      "loss": 2.1571,
+      "step": 38600
+    },
+    {
+      "epoch": 1.970693825588797,
+      "grad_norm": 0.9490466117858887,
+      "learning_rate": 1.718553405651716e-05,
+      "loss": 2.1589,
+      "step": 38700
+    },
+    {
+      "epoch": 1.975786123488224,
+      "grad_norm": 0.8641236424446106,
+      "learning_rate": 1.7100520284286058e-05,
+      "loss": 2.1728,
+      "step": 38800
+    },
+    {
+      "epoch": 1.9808784213876511,
+      "grad_norm": 1.040710210800171,
+      "learning_rate": 1.7015506512054954e-05,
+      "loss": 2.1888,
+      "step": 38900
+    },
+    {
+      "epoch": 1.9859707192870784,
+      "grad_norm": 0.8207067251205444,
+      "learning_rate": 1.6930492739823853e-05,
+      "loss": 2.1504,
+      "step": 39000
+    },
+    {
+      "epoch": 1.9910630171865054,
+      "grad_norm": 0.8451477289199829,
+      "learning_rate": 1.6845478967592752e-05,
+      "loss": 2.1791,
+      "step": 39100
+    },
+    {
+      "epoch": 1.9961553150859326,
+      "grad_norm": 0.9080301523208618,
+      "learning_rate": 1.6760465195361648e-05,
+      "loss": 2.181,
+      "step": 39200
+    },
+    {
+      "epoch": 2.0012221514958624,
+      "grad_norm": 0.9207277297973633,
+      "learning_rate": 1.6675451423130548e-05,
+      "loss": 2.1537,
+      "step": 39300
+    },
+    {
+      "epoch": 2.0063144493952896,
+      "grad_norm": 0.8472919464111328,
+      "learning_rate": 1.6590437650899447e-05,
+      "loss": 2.1691,
+      "step": 39400
+    },
+    {
+      "epoch": 2.011406747294717,
+      "grad_norm": 0.9604014754295349,
+      "learning_rate": 1.6505423878668343e-05,
+      "loss": 2.1409,
+      "step": 39500
+    },
+    {
+      "epoch": 2.016499045194144,
+      "grad_norm": 0.957785964012146,
+      "learning_rate": 1.6420410106437242e-05,
+      "loss": 2.126,
+      "step": 39600
+    },
+    {
+      "epoch": 2.021591343093571,
+      "grad_norm": 0.8542806506156921,
+      "learning_rate": 1.6335396334206145e-05,
+      "loss": 2.1532,
+      "step": 39700
+    },
+    {
+      "epoch": 2.026683640992998,
+      "grad_norm": 0.9949219822883606,
+      "learning_rate": 1.625038256197504e-05,
+      "loss": 2.1437,
+      "step": 39800
+    },
+    {
+      "epoch": 2.0317759388924252,
+      "grad_norm": 0.8735845685005188,
+      "learning_rate": 1.616536878974394e-05,
+      "loss": 2.133,
+      "step": 39900
+    },
+    {
+      "epoch": 2.0368682367918525,
+      "grad_norm": 0.9472355842590332,
+      "learning_rate": 1.608035501751284e-05,
+      "loss": 2.146,
+      "step": 40000
+    },
+    {
+      "epoch": 2.0419605346912792,
+      "grad_norm": 0.9042348861694336,
+      "learning_rate": 1.5995341245281735e-05,
+      "loss": 2.1288,
+      "step": 40100
+    },
+    {
+      "epoch": 2.0470528325907065,
+      "grad_norm": 0.8667154908180237,
+      "learning_rate": 1.5910327473050635e-05,
+      "loss": 2.1182,
+      "step": 40200
+    },
+    {
+      "epoch": 2.0521451304901337,
+      "grad_norm": 0.9168582558631897,
+      "learning_rate": 1.5825313700819534e-05,
+      "loss": 2.1227,
+      "step": 40300
+    },
+    {
+      "epoch": 2.057237428389561,
+      "grad_norm": 0.8843423128128052,
+      "learning_rate": 1.574029992858843e-05,
+      "loss": 2.1564,
+      "step": 40400
+    },
+    {
+      "epoch": 2.0623297262889877,
+      "grad_norm": 0.8709278106689453,
+      "learning_rate": 1.565528615635733e-05,
+      "loss": 2.129,
+      "step": 40500
+    },
+    {
+      "epoch": 2.067422024188415,
+      "grad_norm": 1.0448068380355835,
+      "learning_rate": 1.5570272384126232e-05,
+      "loss": 2.1259,
+      "step": 40600
+    },
+    {
+      "epoch": 2.072514322087842,
+      "grad_norm": 1.014841914176941,
+      "learning_rate": 1.5485258611895128e-05,
+      "loss": 2.1526,
+      "step": 40700
+    },
+    {
+      "epoch": 2.0776066199872694,
+      "grad_norm": 0.9346544146537781,
+      "learning_rate": 1.5400244839664027e-05,
+      "loss": 2.1349,
+      "step": 40800
+    },
+    {
+      "epoch": 2.082698917886696,
+      "grad_norm": 1.029351830482483,
+      "learning_rate": 1.5315231067432927e-05,
+      "loss": 2.1224,
+      "step": 40900
+    },
+    {
+      "epoch": 2.0877912157861234,
+      "grad_norm": 0.8560373783111572,
+      "learning_rate": 1.5230217295201824e-05,
+      "loss": 2.0945,
+      "step": 41000
+    },
+    {
+      "epoch": 2.0928835136855506,
+      "grad_norm": 0.8771845698356628,
+      "learning_rate": 1.5145203522970722e-05,
+      "loss": 2.1215,
+      "step": 41100
+    },
+    {
+      "epoch": 2.097975811584978,
+      "grad_norm": 0.7786750197410583,
+      "learning_rate": 1.506018975073962e-05,
+      "loss": 2.1119,
+      "step": 41200
+    },
+    {
+      "epoch": 2.103068109484405,
+      "grad_norm": 0.8961013555526733,
+      "learning_rate": 1.4975175978508519e-05,
+      "loss": 2.1284,
+      "step": 41300
+    },
+    {
+      "epoch": 2.108160407383832,
+      "grad_norm": 0.7917054295539856,
+      "learning_rate": 1.4890162206277417e-05,
+      "loss": 2.1663,
+      "step": 41400
+    },
+    {
+      "epoch": 2.113252705283259,
+      "grad_norm": 0.9229695200920105,
+      "learning_rate": 1.4805148434046318e-05,
+      "loss": 2.1255,
+      "step": 41500
+    },
+    {
+      "epoch": 2.1183450031826863,
+      "grad_norm": 0.8761498332023621,
+      "learning_rate": 1.4720134661815215e-05,
+      "loss": 2.1271,
+      "step": 41600
+    },
+    {
+      "epoch": 2.1234373010821135,
+      "grad_norm": 0.8369442820549011,
+      "learning_rate": 1.4635120889584114e-05,
+      "loss": 2.1381,
+      "step": 41700
+    },
+    {
+      "epoch": 2.1285295989815403,
+      "grad_norm": 1.058815836906433,
+      "learning_rate": 1.4550107117353012e-05,
+      "loss": 2.1253,
+      "step": 41800
+    },
+    {
+      "epoch": 2.1336218968809675,
+      "grad_norm": 0.8793694972991943,
+      "learning_rate": 1.4465093345121911e-05,
+      "loss": 2.1327,
+      "step": 41900
+    },
+    {
+      "epoch": 2.1387141947803947,
+      "grad_norm": 0.9903535842895508,
+      "learning_rate": 1.4380079572890809e-05,
+      "loss": 2.1585,
+      "step": 42000
+    },
+    {
+      "epoch": 2.143806492679822,
+      "grad_norm": 0.8910212516784668,
+      "learning_rate": 1.4295065800659707e-05,
+      "loss": 2.1482,
+      "step": 42100
+    },
+    {
+      "epoch": 2.1488987905792487,
+      "grad_norm": 0.9088174700737,
+      "learning_rate": 1.4210052028428606e-05,
+      "loss": 2.1391,
+      "step": 42200
+    },
+    {
+      "epoch": 2.153991088478676,
+      "grad_norm": 0.9213513731956482,
+      "learning_rate": 1.4125038256197504e-05,
+      "loss": 2.1447,
+      "step": 42300
+    },
+    {
+      "epoch": 2.159083386378103,
+      "grad_norm": 0.9317104816436768,
+      "learning_rate": 1.4040024483966401e-05,
+      "loss": 2.1115,
+      "step": 42400
+    },
+    {
+      "epoch": 2.1641756842775304,
+      "grad_norm": 0.7989690899848938,
+      "learning_rate": 1.3955010711735302e-05,
+      "loss": 2.1385,
+      "step": 42500
+    },
+    {
+      "epoch": 2.169267982176957,
+      "grad_norm": 0.8436581492424011,
+      "learning_rate": 1.3869996939504202e-05,
+      "loss": 2.1487,
+      "step": 42600
+    },
+    {
+      "epoch": 2.1743602800763844,
+      "grad_norm": 0.9113427400588989,
+      "learning_rate": 1.37849831672731e-05,
+      "loss": 2.0851,
+      "step": 42700
+    },
+    {
+      "epoch": 2.1794525779758116,
+      "grad_norm": 0.8313522338867188,
+      "learning_rate": 1.3699969395041997e-05,
+      "loss": 2.1502,
+      "step": 42800
+    },
+    {
+      "epoch": 2.184544875875239,
+      "grad_norm": 0.9525701999664307,
+      "learning_rate": 1.3614955622810896e-05,
+      "loss": 2.1206,
+      "step": 42900
+    },
+    {
+      "epoch": 2.189637173774666,
+      "grad_norm": 0.9474479556083679,
+      "learning_rate": 1.3529941850579794e-05,
+      "loss": 2.1117,
+      "step": 43000
+    },
+    {
+      "epoch": 2.194729471674093,
+      "grad_norm": 0.8311910629272461,
+      "learning_rate": 1.3444928078348693e-05,
+      "loss": 2.1268,
+      "step": 43100
+    },
+    {
+      "epoch": 2.19982176957352,
+      "grad_norm": 0.879364013671875,
+      "learning_rate": 1.335991430611759e-05,
+      "loss": 2.1426,
+      "step": 43200
+    },
+    {
+      "epoch": 2.2049140674729473,
+      "grad_norm": 0.8633144497871399,
+      "learning_rate": 1.3274900533886488e-05,
+      "loss": 2.1324,
+      "step": 43300
+    },
+    {
+      "epoch": 2.210006365372374,
+      "grad_norm": 0.8333730697631836,
+      "learning_rate": 1.318988676165539e-05,
+      "loss": 2.1246,
+      "step": 43400
+    },
+    {
+      "epoch": 2.2150986632718013,
+      "grad_norm": 0.8649702072143555,
+      "learning_rate": 1.3104872989424289e-05,
+      "loss": 2.122,
+      "step": 43500
+    },
+    {
+      "epoch": 2.2201909611712285,
+      "grad_norm": 0.8680943846702576,
+      "learning_rate": 1.3019859217193186e-05,
+      "loss": 2.1295,
+      "step": 43600
+    },
+    {
+      "epoch": 2.2252832590706557,
+      "grad_norm": 0.9396230578422546,
+      "learning_rate": 1.2934845444962084e-05,
+      "loss": 2.1458,
+      "step": 43700
+    },
+    {
+      "epoch": 2.230375556970083,
+      "grad_norm": 0.9014144539833069,
+      "learning_rate": 1.2849831672730983e-05,
+      "loss": 2.1573,
+      "step": 43800
+    },
+    {
+      "epoch": 2.2354678548695097,
+      "grad_norm": 0.9344182014465332,
+      "learning_rate": 1.2764817900499881e-05,
+      "loss": 2.1516,
+      "step": 43900
+    },
+    {
+      "epoch": 2.240560152768937,
+      "grad_norm": 0.979686439037323,
+      "learning_rate": 1.267980412826878e-05,
+      "loss": 2.1307,
+      "step": 44000
+    },
+    {
+      "epoch": 2.245652450668364,
+      "grad_norm": 0.8325761556625366,
+      "learning_rate": 1.2594790356037678e-05,
+      "loss": 2.1498,
+      "step": 44100
+    },
+    {
+      "epoch": 2.2507447485677914,
+      "grad_norm": 0.8997836709022522,
+      "learning_rate": 1.2509776583806576e-05,
+      "loss": 2.1494,
+      "step": 44200
+    },
+    {
+      "epoch": 2.255837046467218,
+      "grad_norm": 0.8690670132637024,
+      "learning_rate": 1.2424762811575475e-05,
+      "loss": 2.1393,
+      "step": 44300
+    },
+    {
+      "epoch": 2.2609293443666454,
+      "grad_norm": 0.7817577719688416,
+      "learning_rate": 1.2339749039344374e-05,
+      "loss": 2.1341,
+      "step": 44400
+    },
+    {
+      "epoch": 2.2660216422660726,
+      "grad_norm": 0.8697742223739624,
+      "learning_rate": 1.2254735267113272e-05,
+      "loss": 2.1469,
+      "step": 44500
+    },
+    {
+      "epoch": 2.2711139401655,
+      "grad_norm": 0.8965489268302917,
+      "learning_rate": 1.2169721494882171e-05,
+      "loss": 2.1257,
+      "step": 44600
+    },
+    {
+      "epoch": 2.2762062380649266,
+      "grad_norm": 1.0732325315475464,
+      "learning_rate": 1.208470772265107e-05,
+      "loss": 2.1131,
+      "step": 44700
+    },
+    {
+      "epoch": 2.281298535964354,
+      "grad_norm": 0.7745924592018127,
+      "learning_rate": 1.1999693950419968e-05,
+      "loss": 2.1153,
+      "step": 44800
+    },
+    {
+      "epoch": 2.286390833863781,
+      "grad_norm": 0.8988758325576782,
+      "learning_rate": 1.1914680178188868e-05,
+      "loss": 2.1545,
+      "step": 44900
+    },
+    {
+      "epoch": 2.2914831317632083,
+      "grad_norm": 0.9772248268127441,
+      "learning_rate": 1.1829666405957767e-05,
+      "loss": 2.1333,
+      "step": 45000
+    },
+    {
+      "epoch": 2.296575429662635,
+      "grad_norm": 0.8579228520393372,
+      "learning_rate": 1.1744652633726664e-05,
+      "loss": 2.1122,
+      "step": 45100
+    },
+    {
+      "epoch": 2.3016677275620623,
+      "grad_norm": 0.8738901019096375,
+      "learning_rate": 1.1659638861495562e-05,
+      "loss": 2.0938,
+      "step": 45200
+    },
+    {
+      "epoch": 2.3067600254614895,
+      "grad_norm": 0.8962051868438721,
+      "learning_rate": 1.1574625089264461e-05,
+      "loss": 2.1216,
+      "step": 45300
+    },
+    {
+      "epoch": 2.3118523233609167,
+      "grad_norm": 0.8730968236923218,
+      "learning_rate": 1.1489611317033359e-05,
+      "loss": 2.1067,
+      "step": 45400
+    },
+    {
+      "epoch": 2.316944621260344,
+      "grad_norm": 0.9516613483428955,
+      "learning_rate": 1.1404597544802258e-05,
+      "loss": 2.1092,
+      "step": 45500
+    },
+    {
+      "epoch": 2.3220369191597707,
+      "grad_norm": 1.0411871671676636,
+      "learning_rate": 1.1319583772571158e-05,
+      "loss": 2.1199,
+      "step": 45600
+    },
+    {
+      "epoch": 2.327129217059198,
+      "grad_norm": 0.9724430441856384,
+      "learning_rate": 1.1234570000340055e-05,
+      "loss": 2.1335,
+      "step": 45700
+    },
+    {
+      "epoch": 2.332221514958625,
+      "grad_norm": 0.8349046111106873,
+      "learning_rate": 1.1149556228108955e-05,
+      "loss": 2.1249,
+      "step": 45800
+    },
+    {
+      "epoch": 2.337313812858052,
+      "grad_norm": 0.8713769316673279,
+      "learning_rate": 1.1064542455877852e-05,
+      "loss": 2.1054,
+      "step": 45900
+    },
+    {
+      "epoch": 2.342406110757479,
+      "grad_norm": 0.8659300208091736,
+      "learning_rate": 1.0979528683646752e-05,
+      "loss": 2.1095,
+      "step": 46000
+    },
+    {
+      "epoch": 2.3474984086569064,
+      "grad_norm": 1.0436406135559082,
+      "learning_rate": 1.089451491141565e-05,
+      "loss": 2.1337,
+      "step": 46100
+    },
+    {
+      "epoch": 2.3525907065563336,
+      "grad_norm": 0.8275535106658936,
+      "learning_rate": 1.0809501139184549e-05,
+      "loss": 2.1209,
+      "step": 46200
+    },
+    {
+      "epoch": 2.357683004455761,
+      "grad_norm": 0.9503908157348633,
+      "learning_rate": 1.0724487366953446e-05,
+      "loss": 2.1262,
+      "step": 46300
+    },
+    {
+      "epoch": 2.3627753023551876,
+      "grad_norm": 0.8849694728851318,
+      "learning_rate": 1.0639473594722346e-05,
+      "loss": 2.121,
+      "step": 46400
+    },
+    {
+      "epoch": 2.367867600254615,
+      "grad_norm": 0.8742644786834717,
+      "learning_rate": 1.0554459822491245e-05,
+      "loss": 2.1421,
+      "step": 46500
+    },
+    {
+      "epoch": 2.372959898154042,
+      "grad_norm": 0.8519076704978943,
+      "learning_rate": 1.0469446050260143e-05,
+      "loss": 2.1046,
+      "step": 46600
+    },
+    {
+      "epoch": 2.3780521960534693,
+      "grad_norm": 0.8561546206474304,
+      "learning_rate": 1.038443227802904e-05,
+      "loss": 2.1262,
+      "step": 46700
+    },
+    {
+      "epoch": 2.383144493952896,
+      "grad_norm": 0.8309553265571594,
+      "learning_rate": 1.029941850579794e-05,
+      "loss": 2.138,
+      "step": 46800
+    },
+    {
+      "epoch": 2.3882367918523233,
+      "grad_norm": 1.0880669355392456,
+      "learning_rate": 1.0214404733566839e-05,
+      "loss": 2.0881,
+      "step": 46900
+    },
+    {
+      "epoch": 2.3933290897517505,
+      "grad_norm": 0.9982330799102783,
+      "learning_rate": 1.0129390961335736e-05,
+      "loss": 2.1086,
+      "step": 47000
+    },
+    {
+      "epoch": 2.3984213876511777,
+      "grad_norm": 0.9612807035446167,
+      "learning_rate": 1.0044377189104636e-05,
+      "loss": 2.1207,
+      "step": 47100
+    },
+    {
+      "epoch": 2.4035136855506045,
+      "grad_norm": 0.848710298538208,
+      "learning_rate": 9.959363416873533e-06,
+      "loss": 2.1301,
+      "step": 47200
+    },
+    {
+      "epoch": 2.4086059834500317,
+      "grad_norm": 0.8840051889419556,
+      "learning_rate": 9.874349644642433e-06,
+      "loss": 2.1118,
+      "step": 47300
+    },
+    {
+      "epoch": 2.413698281349459,
+      "grad_norm": 0.916346549987793,
+      "learning_rate": 9.789335872411332e-06,
+      "loss": 2.128,
+      "step": 47400
+    },
+    {
+      "epoch": 2.418790579248886,
+      "grad_norm": 0.8974706530570984,
+      "learning_rate": 9.70432210018023e-06,
+      "loss": 2.1452,
+      "step": 47500
+    },
+    {
+      "epoch": 2.423882877148313,
+      "grad_norm": 1.0237131118774414,
+      "learning_rate": 9.619308327949127e-06,
+      "loss": 2.121,
+      "step": 47600
+    },
+    {
+      "epoch": 2.42897517504774,
+      "grad_norm": 0.9156752228736877,
+      "learning_rate": 9.534294555718027e-06,
+      "loss": 2.0985,
+      "step": 47700
+    },
+    {
+      "epoch": 2.4340674729471674,
+      "grad_norm": 0.9210427403450012,
+      "learning_rate": 9.449280783486926e-06,
+      "loss": 2.0653,
+      "step": 47800
+    },
+    {
+      "epoch": 2.4391597708465946,
+      "grad_norm": 0.8185928463935852,
+      "learning_rate": 9.364267011255824e-06,
+      "loss": 2.0994,
+      "step": 47900
+    },
+    {
+      "epoch": 2.444252068746022,
+      "grad_norm": 0.923605740070343,
+      "learning_rate": 9.279253239024723e-06,
+      "loss": 2.1402,
+      "step": 48000
+    },
+    {
+      "epoch": 2.4493443666454486,
+      "grad_norm": 0.8515633344650269,
+      "learning_rate": 9.19423946679362e-06,
+      "loss": 2.1273,
+      "step": 48100
+    },
+    {
+      "epoch": 2.454436664544876,
+      "grad_norm": 0.8325629830360413,
+      "learning_rate": 9.109225694562518e-06,
+      "loss": 2.0974,
+      "step": 48200
+    },
+    {
+      "epoch": 2.459528962444303,
+      "grad_norm": 0.8125095963478088,
+      "learning_rate": 9.02421192233142e-06,
+      "loss": 2.1157,
+      "step": 48300
+    },
+    {
+      "epoch": 2.4646212603437303,
+      "grad_norm": 0.8951058387756348,
+      "learning_rate": 8.939198150100317e-06,
+      "loss": 2.1111,
+      "step": 48400
+    },
+    {
+      "epoch": 2.469713558243157,
+      "grad_norm": 0.8785336017608643,
+      "learning_rate": 8.854184377869214e-06,
+      "loss": 2.1412,
+      "step": 48500
+    },
+    {
+      "epoch": 2.4748058561425843,
+      "grad_norm": 0.9884998202323914,
+      "learning_rate": 8.769170605638114e-06,
+      "loss": 2.1403,
+      "step": 48600
+    },
+    {
+      "epoch": 2.4798981540420115,
+      "grad_norm": 0.9092361330986023,
+      "learning_rate": 8.684156833407011e-06,
+      "loss": 2.1341,
+      "step": 48700
+    },
+    {
+      "epoch": 2.4849904519414387,
+      "grad_norm": 0.9467695951461792,
+      "learning_rate": 8.59914306117591e-06,
+      "loss": 2.1098,
+      "step": 48800
+    },
+    {
+      "epoch": 2.4900827498408655,
+      "grad_norm": 0.8339031338691711,
+      "learning_rate": 8.51412928894481e-06,
+      "loss": 2.1146,
+      "step": 48900
+    },
+    {
+      "epoch": 2.4951750477402928,
+      "grad_norm": 0.8132495284080505,
+      "learning_rate": 8.429115516713708e-06,
+      "loss": 2.1721,
+      "step": 49000
+    },
+    {
+      "epoch": 2.50026734563972,
+      "grad_norm": 0.9209297895431519,
+      "learning_rate": 8.344101744482605e-06,
+      "loss": 2.0942,
+      "step": 49100
+    },
+    {
+      "epoch": 2.505359643539147,
+      "grad_norm": 0.9470928311347961,
+      "learning_rate": 8.259087972251506e-06,
+      "loss": 2.0926,
+      "step": 49200
+    },
+    {
+      "epoch": 2.510451941438574,
+      "grad_norm": 0.9337894320487976,
+      "learning_rate": 8.174074200020404e-06,
+      "loss": 2.1189,
+      "step": 49300
+    },
+    {
+      "epoch": 2.515544239338001,
+      "grad_norm": 0.9764918088912964,
+      "learning_rate": 8.089060427789302e-06,
+      "loss": 2.1185,
+      "step": 49400
+    },
+    {
+      "epoch": 2.5206365372374284,
+      "grad_norm": 0.894453763961792,
+      "learning_rate": 8.004046655558201e-06,
+      "loss": 2.1289,
+      "step": 49500
+    },
+    {
+      "epoch": 2.5257288351368556,
+      "grad_norm": 0.8645434379577637,
+      "learning_rate": 7.919032883327099e-06,
+      "loss": 2.1025,
+      "step": 49600
+    },
+    {
+      "epoch": 2.530821133036283,
+      "grad_norm": 0.8322845101356506,
+      "learning_rate": 7.834019111095998e-06,
+      "loss": 2.1128,
+      "step": 49700
+    },
+    {
+      "epoch": 2.5359134309357096,
+      "grad_norm": 1.0294426679611206,
+      "learning_rate": 7.749005338864897e-06,
+      "loss": 2.1348,
+      "step": 49800
+    },
+    {
+      "epoch": 2.541005728835137,
+      "grad_norm": 0.9489388465881348,
+      "learning_rate": 7.663991566633795e-06,
+      "loss": 2.1089,
+      "step": 49900
+    },
+    {
+      "epoch": 2.546098026734564,
+      "grad_norm": 0.9332979917526245,
+      "learning_rate": 7.578977794402693e-06,
+      "loss": 2.1677,
+      "step": 50000
+    },
+    {
+      "epoch": 2.551190324633991,
+      "grad_norm": 0.8114882111549377,
+      "learning_rate": 7.493964022171592e-06,
+      "loss": 2.1265,
+      "step": 50100
+    },
+    {
+      "epoch": 2.556282622533418,
+      "grad_norm": 0.8496439456939697,
+      "learning_rate": 7.408950249940491e-06,
+      "loss": 2.1713,
+      "step": 50200
+    },
+    {
+      "epoch": 2.5613749204328453,
+      "grad_norm": 1.149905800819397,
+      "learning_rate": 7.32393647770939e-06,
+      "loss": 2.1234,
+      "step": 50300
+    },
+    {
+      "epoch": 2.5664672183322725,
+      "grad_norm": 1.0552695989608765,
+      "learning_rate": 7.238922705478287e-06,
+      "loss": 2.1398,
+      "step": 50400
+    },
+    {
+      "epoch": 2.5715595162316998,
+      "grad_norm": 0.9433385133743286,
+      "learning_rate": 7.153908933247186e-06,
+      "loss": 2.0986,
+      "step": 50500
+    },
+    {
+      "epoch": 2.5766518141311265,
+      "grad_norm": 0.889086127281189,
+      "learning_rate": 7.068895161016086e-06,
+      "loss": 2.1338,
+      "step": 50600
+    },
+    {
+      "epoch": 2.5817441120305538,
+      "grad_norm": 0.8793154358863831,
+      "learning_rate": 6.9838813887849835e-06,
+      "loss": 2.1095,
+      "step": 50700
+    },
+    {
+      "epoch": 2.586836409929981,
+      "grad_norm": 0.7565730214118958,
+      "learning_rate": 6.898867616553882e-06,
+      "loss": 2.1219,
+      "step": 50800
+    },
+    {
+      "epoch": 2.5919287078294078,
+      "grad_norm": 0.8305276036262512,
+      "learning_rate": 6.8138538443227805e-06,
+      "loss": 2.099,
+      "step": 50900
+    },
+    {
+      "epoch": 2.597021005728835,
+      "grad_norm": 0.9467841386795044,
+      "learning_rate": 6.728840072091679e-06,
+      "loss": 2.123,
+      "step": 51000
+    },
+    {
+      "epoch": 2.602113303628262,
+      "grad_norm": 0.9913722276687622,
+      "learning_rate": 6.643826299860578e-06,
+      "loss": 2.1189,
+      "step": 51100
+    },
+    {
+      "epoch": 2.6072056015276894,
+      "grad_norm": 0.9008012413978577,
+      "learning_rate": 6.558812527629477e-06,
+      "loss": 2.155,
+      "step": 51200
+    },
+    {
+      "epoch": 2.6122978994271167,
+      "grad_norm": 0.9230712056159973,
+      "learning_rate": 6.473798755398374e-06,
+      "loss": 2.1333,
+      "step": 51300
+    },
+    {
+      "epoch": 2.617390197326544,
+      "grad_norm": 1.0198971033096313,
+      "learning_rate": 6.388784983167273e-06,
+      "loss": 2.1374,
+      "step": 51400
+    },
+    {
+      "epoch": 2.6224824952259707,
+      "grad_norm": 0.9199273586273193,
+      "learning_rate": 6.303771210936171e-06,
+      "loss": 2.1332,
+      "step": 51500
+    },
+    {
+      "epoch": 2.627574793125398,
+      "grad_norm": 0.8723760843276978,
+      "learning_rate": 6.21875743870507e-06,
+      "loss": 2.1547,
+      "step": 51600
+    },
+    {
+      "epoch": 2.632667091024825,
+      "grad_norm": 0.9192347526550293,
+      "learning_rate": 6.133743666473969e-06,
+      "loss": 2.1192,
+      "step": 51700
+    },
+    {
+      "epoch": 2.637759388924252,
+      "grad_norm": 0.9517456889152527,
+      "learning_rate": 6.048729894242868e-06,
+      "loss": 2.1143,
+      "step": 51800
+    },
+    {
+      "epoch": 2.642851686823679,
+      "grad_norm": 0.9906876683235168,
+      "learning_rate": 5.963716122011766e-06,
+      "loss": 2.1171,
+      "step": 51900
+    },
+    {
+      "epoch": 2.6479439847231063,
+      "grad_norm": 0.9755644202232361,
+      "learning_rate": 5.878702349780665e-06,
+      "loss": 2.163,
+      "step": 52000
+    },
+    {
+      "epoch": 2.6530362826225335,
+      "grad_norm": 0.9300287961959839,
+      "learning_rate": 5.793688577549564e-06,
+      "loss": 2.1218,
+      "step": 52100
+    },
+    {
+      "epoch": 2.6581285805219608,
+      "grad_norm": 0.8865501284599304,
+      "learning_rate": 5.7086748053184616e-06,
+      "loss": 2.1356,
+      "step": 52200
+    },
+    {
+      "epoch": 2.6632208784213875,
+      "grad_norm": 0.8156447410583496,
+      "learning_rate": 5.62366103308736e-06,
+      "loss": 2.1171,
+      "step": 52300
+    },
+    {
+      "epoch": 2.6683131763208148,
+      "grad_norm": 0.8186530470848083,
+      "learning_rate": 5.538647260856259e-06,
+      "loss": 2.1052,
+      "step": 52400
+    },
+    {
+      "epoch": 2.673405474220242,
+      "grad_norm": 0.790550708770752,
+      "learning_rate": 5.453633488625157e-06,
+      "loss": 2.1071,
+      "step": 52500
+    },
+    {
+      "epoch": 2.6784977721196688,
+      "grad_norm": 0.8866438865661621,
+      "learning_rate": 5.368619716394056e-06,
+      "loss": 2.1354,
+      "step": 52600
+    },
+    {
+      "epoch": 2.683590070019096,
+      "grad_norm": 0.9953215718269348,
+      "learning_rate": 5.283605944162955e-06,
+      "loss": 2.1383,
+      "step": 52700
+    },
+    {
+      "epoch": 2.688682367918523,
+      "grad_norm": 0.9829987287521362,
+      "learning_rate": 5.198592171931853e-06,
+      "loss": 2.0919,
+      "step": 52800
+    },
+    {
+      "epoch": 2.6937746658179504,
+      "grad_norm": 0.9085790514945984,
+      "learning_rate": 5.113578399700752e-06,
+      "loss": 2.1178,
+      "step": 52900
+    },
+    {
+      "epoch": 2.6988669637173777,
+      "grad_norm": 0.8004271388053894,
+      "learning_rate": 5.02856462746965e-06,
+      "loss": 2.1239,
+      "step": 53000
+    },
+    {
+      "epoch": 2.7039592616168044,
+      "grad_norm": 0.9412344098091125,
+      "learning_rate": 4.943550855238549e-06,
+      "loss": 2.108,
+      "step": 53100
+    },
+    {
+      "epoch": 2.7090515595162317,
+      "grad_norm": 0.9245398640632629,
+      "learning_rate": 4.858537083007447e-06,
+      "loss": 2.1241,
+      "step": 53200
+    },
+    {
+      "epoch": 2.714143857415659,
+      "grad_norm": 0.9695274233818054,
+      "learning_rate": 4.7735233107763465e-06,
+      "loss": 2.1106,
+      "step": 53300
+    },
+    {
+      "epoch": 2.7192361553150857,
+      "grad_norm": 0.9269813895225525,
+      "learning_rate": 4.688509538545244e-06,
+      "loss": 2.1075,
+      "step": 53400
+    },
+    {
+      "epoch": 2.724328453214513,
+      "grad_norm": 0.9783353805541992,
+      "learning_rate": 4.6034957663141435e-06,
+      "loss": 2.1127,
+      "step": 53500
+    },
+    {
+      "epoch": 2.72942075111394,
+      "grad_norm": 0.9476038813591003,
+      "learning_rate": 4.518481994083042e-06,
+      "loss": 2.1284,
+      "step": 53600
+    },
+    {
+      "epoch": 2.7345130490133673,
+      "grad_norm": 0.93116295337677,
+      "learning_rate": 4.43346822185194e-06,
+      "loss": 2.1004,
+      "step": 53700
+    },
+    {
+      "epoch": 2.7396053469127946,
+      "grad_norm": 0.9898892641067505,
+      "learning_rate": 4.348454449620839e-06,
+      "loss": 2.1129,
+      "step": 53800
+    },
+    {
+      "epoch": 2.744697644812222,
+      "grad_norm": 0.9059526920318604,
+      "learning_rate": 4.263440677389737e-06,
+      "loss": 2.1189,
+      "step": 53900
+    },
+    {
+      "epoch": 2.7497899427116486,
+      "grad_norm": 0.8806390762329102,
+      "learning_rate": 4.178426905158636e-06,
+      "loss": 2.1416,
+      "step": 54000
+    },
+    {
+      "epoch": 2.754882240611076,
+      "grad_norm": 0.9231753945350647,
+      "learning_rate": 4.093413132927534e-06,
+      "loss": 2.1373,
+      "step": 54100
+    },
+    {
+      "epoch": 2.759974538510503,
+      "grad_norm": 0.7574446201324463,
+      "learning_rate": 4.008399360696434e-06,
+      "loss": 2.1355,
+      "step": 54200
+    },
+    {
+      "epoch": 2.76506683640993,
+      "grad_norm": 0.8553287982940674,
+      "learning_rate": 3.923385588465331e-06,
+      "loss": 2.0786,
+      "step": 54300
+    },
+    {
+      "epoch": 2.770159134309357,
+      "grad_norm": 0.7898595333099365,
+      "learning_rate": 3.83837181623423e-06,
+      "loss": 2.0941,
+      "step": 54400
+    },
+    {
+      "epoch": 2.7752514322087842,
+      "grad_norm": 0.8895372748374939,
+      "learning_rate": 3.7533580440031287e-06,
+      "loss": 2.1311,
+      "step": 54500
+    },
+    {
+      "epoch": 2.7803437301082115,
+      "grad_norm": 0.9352322816848755,
+      "learning_rate": 3.668344271772027e-06,
+      "loss": 2.102,
+      "step": 54600
+    },
+    {
+      "epoch": 2.7854360280076387,
+      "grad_norm": 1.003927230834961,
+      "learning_rate": 3.583330499540926e-06,
+      "loss": 2.1119,
+      "step": 54700
+    },
+    {
+      "epoch": 2.7905283259070655,
+      "grad_norm": 0.9228959083557129,
+      "learning_rate": 3.498316727309824e-06,
+      "loss": 2.142,
+      "step": 54800
+    },
+    {
+      "epoch": 2.7956206238064927,
+      "grad_norm": 0.9431111812591553,
+      "learning_rate": 3.413302955078723e-06,
+      "loss": 2.12,
+      "step": 54900
+    },
+    {
+      "epoch": 2.80071292170592,
+      "grad_norm": 0.9116231799125671,
+      "learning_rate": 3.3282891828476215e-06,
+      "loss": 2.1261,
+      "step": 55000
+    },
+    {
+      "epoch": 2.8058052196053467,
+      "grad_norm": 0.9542424082756042,
+      "learning_rate": 3.2432754106165196e-06,
+      "loss": 2.1151,
+      "step": 55100
+    },
+    {
+      "epoch": 2.810897517504774,
+      "grad_norm": 0.8199505805969238,
+      "learning_rate": 3.1582616383854185e-06,
+      "loss": 2.0883,
+      "step": 55200
+    },
+    {
+      "epoch": 2.815989815404201,
+      "grad_norm": 0.8526725769042969,
+      "learning_rate": 3.0732478661543174e-06,
+      "loss": 2.1094,
+      "step": 55300
+    },
+    {
+      "epoch": 2.8210821133036283,
+      "grad_norm": 0.9284189343452454,
+      "learning_rate": 2.9882340939232155e-06,
+      "loss": 2.1072,
+      "step": 55400
+    },
+    {
+      "epoch": 2.8261744112030556,
+      "grad_norm": 0.9289183616638184,
+      "learning_rate": 2.9032203216921144e-06,
+      "loss": 2.1227,
+      "step": 55500
+    },
+    {
+      "epoch": 2.8312667091024823,
+      "grad_norm": 1.0548968315124512,
+      "learning_rate": 2.818206549461013e-06,
+      "loss": 2.138,
+      "step": 55600
+    },
+    {
+      "epoch": 2.8363590070019096,
+      "grad_norm": 0.8402355313301086,
+      "learning_rate": 2.7331927772299113e-06,
+      "loss": 2.1394,
+      "step": 55700
+    },
+    {
+      "epoch": 2.841451304901337,
+      "grad_norm": 0.9172413349151611,
+      "learning_rate": 2.64817900499881e-06,
+      "loss": 2.114,
+      "step": 55800
+    },
+    {
+      "epoch": 2.8465436028007636,
+      "grad_norm": 0.8457333445549011,
+      "learning_rate": 2.5631652327677087e-06,
+      "loss": 2.1268,
+      "step": 55900
+    },
+    {
+      "epoch": 2.851635900700191,
+      "grad_norm": 0.8858858942985535,
+      "learning_rate": 2.478151460536607e-06,
+      "loss": 2.0901,
+      "step": 56000
+    },
+    {
+      "epoch": 2.856728198599618,
+      "grad_norm": 0.8789589405059814,
+      "learning_rate": 2.3931376883055057e-06,
+      "loss": 2.1154,
+      "step": 56100
+    },
+    {
+      "epoch": 2.8618204964990452,
+      "grad_norm": 0.9234612584114075,
+      "learning_rate": 2.308123916074404e-06,
+      "loss": 2.1106,
+      "step": 56200
+    },
+    {
+      "epoch": 2.8669127943984725,
+      "grad_norm": 0.8070857524871826,
+      "learning_rate": 2.2231101438433026e-06,
+      "loss": 2.1181,
+      "step": 56300
+    },
+    {
+      "epoch": 2.8720050922978997,
+      "grad_norm": 0.9172016978263855,
+      "learning_rate": 2.138096371612201e-06,
+      "loss": 2.0832,
+      "step": 56400
+    },
+    {
+      "epoch": 2.8770973901973265,
+      "grad_norm": 0.9449873566627502,
+      "learning_rate": 2.0530825993811e-06,
+      "loss": 2.126,
+      "step": 56500
+    },
+    {
+      "epoch": 2.8821896880967537,
+      "grad_norm": 1.0262093544006348,
+      "learning_rate": 1.9680688271499985e-06,
+      "loss": 2.1117,
+      "step": 56600
+    },
+    {
+      "epoch": 2.887281985996181,
+      "grad_norm": 0.7934767007827759,
+      "learning_rate": 1.8830550549188972e-06,
+      "loss": 2.1256,
+      "step": 56700
+    },
+    {
+      "epoch": 2.8923742838956077,
+      "grad_norm": 0.9590465426445007,
+      "learning_rate": 1.7980412826877954e-06,
+      "loss": 2.1335,
+      "step": 56800
+    },
+    {
+      "epoch": 2.897466581795035,
+      "grad_norm": 1.006219744682312,
+      "learning_rate": 1.713027510456694e-06,
+      "loss": 2.0888,
+      "step": 56900
+    },
+    {
+      "epoch": 2.902558879694462,
+      "grad_norm": 0.9063106179237366,
+      "learning_rate": 1.6280137382255926e-06,
+      "loss": 2.1506,
+      "step": 57000
+    },
+    {
+      "epoch": 2.9076511775938894,
+      "grad_norm": 0.8653075695037842,
+      "learning_rate": 1.542999965994491e-06,
+      "loss": 2.0845,
+      "step": 57100
+    },
+    {
+      "epoch": 2.9127434754933166,
+      "grad_norm": 0.9707706570625305,
+      "learning_rate": 1.4579861937633898e-06,
+      "loss": 2.0865,
+      "step": 57200
+    },
+    {
+      "epoch": 2.9178357733927434,
+      "grad_norm": 0.9578688740730286,
+      "learning_rate": 1.3729724215322882e-06,
+      "loss": 2.1098,
+      "step": 57300
+    },
+    {
+      "epoch": 2.9229280712921706,
+      "grad_norm": 0.8037517070770264,
+      "learning_rate": 1.2879586493011867e-06,
+      "loss": 2.085,
+      "step": 57400
+    },
+    {
+      "epoch": 2.928020369191598,
+      "grad_norm": 0.9694920182228088,
+      "learning_rate": 1.2029448770700854e-06,
+      "loss": 2.0926,
+      "step": 57500
+    },
+    {
+      "epoch": 2.9331126670910246,
+      "grad_norm": 0.8718476891517639,
+      "learning_rate": 1.1179311048389841e-06,
+      "loss": 2.1112,
+      "step": 57600
+    },
+    {
+      "epoch": 2.938204964990452,
+      "grad_norm": 0.8940988779067993,
+      "learning_rate": 1.0329173326078824e-06,
+      "loss": 2.12,
+      "step": 57700
+    },
+    {
+      "epoch": 2.943297262889879,
+      "grad_norm": 0.9514064192771912,
+      "learning_rate": 9.479035603767811e-07,
+      "loss": 2.1416,
+      "step": 57800
+    },
+    {
+      "epoch": 2.9483895607893063,
+      "grad_norm": 0.9789698719978333,
+      "learning_rate": 8.628897881456797e-07,
+      "loss": 2.0913,
+      "step": 57900
+    },
+    {
+      "epoch": 2.9534818586887335,
+      "grad_norm": 1.028600811958313,
+      "learning_rate": 7.778760159145782e-07,
+      "loss": 2.1142,
+      "step": 58000
+    },
+    {
+      "epoch": 2.9585741565881603,
+      "grad_norm": 0.850046694278717,
+      "learning_rate": 6.928622436834767e-07,
+      "loss": 2.0929,
+      "step": 58100
+    },
+    {
+      "epoch": 2.9636664544875875,
+      "grad_norm": 0.8758450150489807,
+      "learning_rate": 6.078484714523753e-07,
+      "loss": 2.0991,
+      "step": 58200
+    },
+    {
+      "epoch": 2.9687587523870147,
+      "grad_norm": 0.9652713537216187,
+      "learning_rate": 5.228346992212739e-07,
+      "loss": 2.1095,
+      "step": 58300
+    },
+    {
+      "epoch": 2.973851050286442,
+      "grad_norm": 1.0260512828826904,
+      "learning_rate": 4.3782092699017247e-07,
+      "loss": 2.1069,
+      "step": 58400
+    },
+    {
+      "epoch": 2.9789433481858687,
+      "grad_norm": 0.7857241034507751,
+      "learning_rate": 3.5280715475907095e-07,
+      "loss": 2.1014,
+      "step": 58500
+    },
+    {
+      "epoch": 2.984035646085296,
+      "grad_norm": 0.964096188545227,
+      "learning_rate": 2.6779338252796954e-07,
+      "loss": 2.0981,
+      "step": 58600
+    },
+    {
+      "epoch": 2.989127943984723,
+      "grad_norm": 0.8568851351737976,
+      "learning_rate": 1.827796102968681e-07,
+      "loss": 2.1283,
+      "step": 58700
+    },
+    {
+      "epoch": 2.9942202418841504,
+      "grad_norm": 0.9048463702201843,
+      "learning_rate": 9.776583806576667e-08,
+      "loss": 2.1011,
+      "step": 58800
+    },
+    {
+      "epoch": 2.9993125397835776,
+      "grad_norm": 0.8119781613349915,
+      "learning_rate": 1.2752065834665216e-08,
+      "loss": 2.0672,
+      "step": 58900
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 58914,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 5000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.462938693632e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}