diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15407 @@
+{
+  "best_metric": 5.182580947875977,
+  "best_model_checkpoint": "./results/models/mistral-prot/checkpoint-937410",
+  "epoch": 7.0,
+  "eval_steps": 500,
+  "global_step": 1093645,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0032003072294940313,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00039997439754216405,
+      "loss": 5.83,
+      "step": 500
+    },
+    {
+      "epoch": 0.006400614458988063,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0003999487950843281,
+      "loss": 5.6885,
+      "step": 1000
+    },
+    {
+      "epoch": 0.009600921688482094,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0003999231926264922,
+      "loss": 5.6553,
+      "step": 1500
+    },
+    {
+      "epoch": 0.012801228917976125,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0003998975901686562,
+      "loss": 5.6407,
+      "step": 2000
+    },
+    {
+      "epoch": 0.016001536147470158,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00039987198771082024,
+      "loss": 5.6315,
+      "step": 2500
+    },
+    {
+      "epoch": 0.01920184337696419,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 0.0003998463852529843,
+      "loss": 5.6252,
+      "step": 3000
+    },
+    {
+      "epoch": 0.02240215060645822,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00039982078279514837,
+      "loss": 5.6176,
+      "step": 3500
+    },
+    {
+      "epoch": 0.02560245783595225,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0003997951803373124,
+      "loss": 5.6118,
+      "step": 4000
+    },
+    {
+      "epoch": 0.02880276506544628,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0003997695778794765,
+      "loss": 5.6065,
+      "step": 4500
+    },
+    {
+      "epoch": 0.032003072294940316,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0003997439754216405,
+      "loss": 5.6034,
+      "step": 5000
+    },
+    {
+      "epoch": 0.03520337952443434,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00039971837296380456,
+      "loss": 5.5962,
+      "step": 5500
+    },
+    {
+      "epoch": 0.03840368675392838,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0003996927705059686,
+      "loss": 5.5968,
+      "step": 6000
+    },
+    {
+      "epoch": 0.04160399398342241,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0003996671680481326,
+      "loss": 5.5895,
+      "step": 6500
+    },
+    {
+      "epoch": 0.04480430121291644,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0003996415655902967,
+      "loss": 5.5867,
+      "step": 7000
+    },
+    {
+      "epoch": 0.048004608442410474,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00039961596313246075,
+      "loss": 5.5837,
+      "step": 7500
+    },
+    {
+      "epoch": 0.0512049156719045,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0003995903606746248,
+      "loss": 5.5813,
+      "step": 8000
+    },
+    {
+      "epoch": 0.054405222901398535,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0003995647582167888,
+      "loss": 5.577,
+      "step": 8500
+    },
+    {
+      "epoch": 0.05760553013089256,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00039953915575895285,
+      "loss": 5.571,
+      "step": 9000
+    },
+    {
+      "epoch": 0.0608058373603866,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0003995135533011169,
+      "loss": 5.5684,
+      "step": 9500
+    },
+    {
+      "epoch": 0.06400614458988063,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.000399487950843281,
+      "loss": 5.566,
+      "step": 10000
+    },
+    {
+      "epoch": 0.06720645181937467,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.000399462348385445,
+      "loss": 5.5603,
+      "step": 10500
+    },
+    {
+      "epoch": 0.07040675904886869,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00039943674592760904,
+      "loss": 5.5588,
+      "step": 11000
+    },
+    {
+      "epoch": 0.07360706627836272,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 0.00039941114346977313,
+      "loss": 5.5549,
+      "step": 11500
+    },
+    {
+      "epoch": 0.07680737350785675,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00039938554101193717,
+      "loss": 5.5491,
+      "step": 12000
+    },
+    {
+      "epoch": 0.08000768073735079,
+      "grad_norm": 0.244140625,
+      "learning_rate": 0.00039935993855410125,
+      "loss": 5.5479,
+      "step": 12500
+    },
+    {
+      "epoch": 0.08320798796684482,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0003993343360962653,
+      "loss": 5.5413,
+      "step": 13000
+    },
+    {
+      "epoch": 0.08640829519633884,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0003993087336384293,
+      "loss": 5.5361,
+      "step": 13500
+    },
+    {
+      "epoch": 0.08960860242583288,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00039928313118059336,
+      "loss": 5.5309,
+      "step": 14000
+    },
+    {
+      "epoch": 0.09280890965532691,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0003992575287227574,
+      "loss": 5.5225,
+      "step": 14500
+    },
+    {
+      "epoch": 0.09600921688482095,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0003992319262649214,
+      "loss": 5.5158,
+      "step": 15000
+    },
+    {
+      "epoch": 0.09920952411431497,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0003992063238070855,
+      "loss": 5.5118,
+      "step": 15500
+    },
+    {
+      "epoch": 0.102409831343809,
+      "grad_norm": 0.2421875,
+      "learning_rate": 0.00039918072134924955,
+      "loss": 5.5003,
+      "step": 16000
+    },
+    {
+      "epoch": 0.10561013857330304,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0003991551188914136,
+      "loss": 5.4983,
+      "step": 16500
+    },
+    {
+      "epoch": 0.10881044580279707,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0003991295164335776,
+      "loss": 5.4975,
+      "step": 17000
+    },
+    {
+      "epoch": 0.1120107530322911,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00039910391397574165,
+      "loss": 5.4921,
+      "step": 17500
+    },
+    {
+      "epoch": 0.11521106026178513,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.00039907831151790574,
+      "loss": 5.4822,
+      "step": 18000
+    },
+    {
+      "epoch": 0.11841136749127916,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0003990527090600698,
+      "loss": 5.476,
+      "step": 18500
+    },
+    {
+      "epoch": 0.1216116747207732,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00039902710660223386,
+      "loss": 5.4742,
+      "step": 19000
+    },
+    {
+      "epoch": 0.12481198195026723,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.0003990015041443979,
+      "loss": 5.4732,
+      "step": 19500
+    },
+    {
+      "epoch": 0.12801228917976126,
+      "grad_norm": 0.248046875,
+      "learning_rate": 0.00039897590168656193,
+      "loss": 5.4715,
+      "step": 20000
+    },
+    {
+      "epoch": 0.13121259640925528,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.000398950299228726,
+      "loss": 5.4682,
+      "step": 20500
+    },
+    {
+      "epoch": 0.13441290363874933,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00039892469677089005,
+      "loss": 5.4649,
+      "step": 21000
+    },
+    {
+      "epoch": 0.13761321086824335,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0003988990943130541,
+      "loss": 5.4609,
+      "step": 21500
+    },
+    {
+      "epoch": 0.14081351809773737,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0003988734918552181,
+      "loss": 5.4567,
+      "step": 22000
+    },
+    {
+      "epoch": 0.14401382532723142,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00039884788939738216,
+      "loss": 5.4546,
+      "step": 22500
+    },
+    {
+      "epoch": 0.14721413255672544,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0003988222869395462,
+      "loss": 5.4503,
+      "step": 23000
+    },
+    {
+      "epoch": 0.1504144397862195,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0003987966844817103,
+      "loss": 5.4432,
+      "step": 23500
+    },
+    {
+      "epoch": 0.1536147470157135,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.0003987710820238743,
+      "loss": 5.4463,
+      "step": 24000
+    },
+    {
+      "epoch": 0.15681505424520753,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00039874547956603835,
+      "loss": 5.4436,
+      "step": 24500
+    },
+    {
+      "epoch": 0.16001536147470158,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0003987198771082024,
+      "loss": 5.4418,
+      "step": 25000
+    },
+    {
+      "epoch": 0.1632156687041956,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0003986942746503664,
+      "loss": 5.4418,
+      "step": 25500
+    },
+    {
+      "epoch": 0.16641597593368965,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0003986686721925305,
+      "loss": 5.4369,
+      "step": 26000
+    },
+    {
+      "epoch": 0.16961628316318367,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00039864306973469454,
+      "loss": 5.4352,
+      "step": 26500
+    },
+    {
+      "epoch": 0.1728165903926777,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00039861746727685863,
+      "loss": 5.4318,
+      "step": 27000
+    },
+    {
+      "epoch": 0.17601689762217174,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00039859186481902266,
+      "loss": 5.4271,
+      "step": 27500
+    },
+    {
+      "epoch": 0.17921720485166576,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0003985662623611867,
+      "loss": 5.4238,
+      "step": 28000
+    },
+    {
+      "epoch": 0.1824175120811598,
+      "grad_norm": 0.251953125,
+      "learning_rate": 0.00039854065990335073,
+      "loss": 5.4235,
+      "step": 28500
+    },
+    {
+      "epoch": 0.18561781931065383,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0003985150574455148,
+      "loss": 5.4201,
+      "step": 29000
+    },
+    {
+      "epoch": 0.18881812654014785,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00039848945498767885,
+      "loss": 5.4201,
+      "step": 29500
+    },
+    {
+      "epoch": 0.1920184337696419,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0003984638525298429,
+      "loss": 5.4195,
+      "step": 30000
+    },
+    {
+      "epoch": 0.19521874099913591,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.0003984382500720069,
+      "loss": 5.4198,
+      "step": 30500
+    },
+    {
+      "epoch": 0.19841904822862994,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.00039841264761417096,
+      "loss": 5.4172,
+      "step": 31000
+    },
+    {
+      "epoch": 0.20161935545812398,
+      "grad_norm": 0.255859375,
+      "learning_rate": 0.00039838704515633504,
+      "loss": 5.4132,
+      "step": 31500
+    },
+    {
+      "epoch": 0.204819662687618,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0003983614426984991,
+      "loss": 5.4128,
+      "step": 32000
+    },
+    {
+      "epoch": 0.20801996991711205,
+      "grad_norm": 0.25,
+      "learning_rate": 0.0003983358402406631,
+      "loss": 5.4085,
+      "step": 32500
+    },
+    {
+      "epoch": 0.21122027714660607,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00039831023778282715,
+      "loss": 5.4076,
+      "step": 33000
+    },
+    {
+      "epoch": 0.2144205843761001,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0003982846353249912,
+      "loss": 5.4045,
+      "step": 33500
+    },
+    {
+      "epoch": 0.21762089160559414,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.00039825903286715527,
+      "loss": 5.4033,
+      "step": 34000
+    },
+    {
+      "epoch": 0.22082119883508816,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.0003982334304093193,
+      "loss": 5.4007,
+      "step": 34500
+    },
+    {
+      "epoch": 0.2240215060645822,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0003982078279514834,
+      "loss": 5.3985,
+      "step": 35000
+    },
+    {
+      "epoch": 0.22722181329407623,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0003981822254936474,
+      "loss": 5.3999,
+      "step": 35500
+    },
+    {
+      "epoch": 0.23042212052357025,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 0.00039815662303581146,
+      "loss": 5.3982,
+      "step": 36000
+    },
+    {
+      "epoch": 0.2336224277530643,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0003981310205779755,
+      "loss": 5.3966,
+      "step": 36500
+    },
+    {
+      "epoch": 0.23682273498255832,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0003981054181201396,
+      "loss": 5.3965,
+      "step": 37000
+    },
+    {
+      "epoch": 0.24002304221205237,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0003980798156623036,
+      "loss": 5.3941,
+      "step": 37500
+    },
+    {
+      "epoch": 0.2432233494415464,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00039805421320446765,
+      "loss": 5.3914,
+      "step": 38000
+    },
+    {
+      "epoch": 0.2464236566710404,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0003980286107466317,
+      "loss": 5.387,
+      "step": 38500
+    },
+    {
+      "epoch": 0.24962396390053446,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0003980030082887957,
+      "loss": 5.3889,
+      "step": 39000
+    },
+    {
+      "epoch": 0.2528242711300285,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00039797740583095975,
+      "loss": 5.3829,
+      "step": 39500
+    },
+    {
+      "epoch": 0.2560245783595225,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00039795180337312384,
+      "loss": 5.3887,
+      "step": 40000
+    },
+    {
+      "epoch": 0.2592248855890165,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0003979262009152879,
+      "loss": 5.3815,
+      "step": 40500
+    },
+    {
+      "epoch": 0.26242519281851057,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0003979005984574519,
+      "loss": 5.3823,
+      "step": 41000
+    },
+    {
+      "epoch": 0.2656255000480046,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.000397874995999616,
+      "loss": 5.3804,
+      "step": 41500
+    },
+    {
+      "epoch": 0.26882580727749866,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00039784939354178003,
+      "loss": 5.3816,
+      "step": 42000
+    },
+    {
+      "epoch": 0.27202611450699266,
+      "grad_norm": 0.26171875,
+      "learning_rate": 0.0003978237910839441,
+      "loss": 5.3798,
+      "step": 42500
+    },
+    {
+      "epoch": 0.2752264217364867,
+      "grad_norm": 0.25390625,
+      "learning_rate": 0.00039779818862610816,
+      "loss": 5.3706,
+      "step": 43000
+    },
+    {
+      "epoch": 0.27842672896598075,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0003977725861682722,
+      "loss": 5.3747,
+      "step": 43500
+    },
+    {
+      "epoch": 0.28162703619547474,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0003977469837104362,
+      "loss": 5.3684,
+      "step": 44000
+    },
+    {
+      "epoch": 0.2848273434249688,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.00039772138125260026,
+      "loss": 5.3669,
+      "step": 44500
+    },
+    {
+      "epoch": 0.28802765065446284,
+      "grad_norm": 0.259765625,
+      "learning_rate": 0.0003976957787947643,
+      "loss": 5.3725,
+      "step": 45000
+    },
+    {
+      "epoch": 0.29122795788395683,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0003976701763369284,
+      "loss": 5.3737,
+      "step": 45500
+    },
+    {
+      "epoch": 0.2944282651134509,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0003976445738790924,
+      "loss": 5.3694,
+      "step": 46000
+    },
+    {
+      "epoch": 0.29762857234294493,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.00039761897142125645,
+      "loss": 5.3707,
+      "step": 46500
+    },
+    {
+      "epoch": 0.300828879572439,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0003975933689634205,
+      "loss": 5.3625,
+      "step": 47000
+    },
+    {
+      "epoch": 0.30402918680193297,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0003975677665055845,
+      "loss": 5.3647,
+      "step": 47500
+    },
+    {
+      "epoch": 0.307229494031427,
+      "grad_norm": 0.2578125,
+      "learning_rate": 0.0003975421640477486,
+      "loss": 5.3624,
+      "step": 48000
+    },
+    {
+      "epoch": 0.31042980126092107,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00039751656158991264,
+      "loss": 5.3623,
+      "step": 48500
+    },
+    {
+      "epoch": 0.31363010849041506,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0003974909591320767,
+      "loss": 5.3642,
+      "step": 49000
+    },
+    {
+      "epoch": 0.3168304157199091,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00039746535667424076,
+      "loss": 5.3574,
+      "step": 49500
+    },
+    {
+      "epoch": 0.32003072294940316,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.0003974397542164048,
+      "loss": 5.361,
+      "step": 50000
+    },
+    {
+      "epoch": 0.32323103017889715,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00039741415175856883,
+      "loss": 5.3607,
+      "step": 50500
+    },
+    {
+      "epoch": 0.3264313374083912,
+      "grad_norm": 0.265625,
+      "learning_rate": 0.0003973885493007329,
+      "loss": 5.3574,
+      "step": 51000
+    },
+    {
+      "epoch": 0.32963164463788525,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00039736294684289696,
+      "loss": 5.353,
+      "step": 51500
+    },
+    {
+      "epoch": 0.3328319518673793,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.000397337344385061,
+      "loss": 5.3515,
+      "step": 52000
+    },
+    {
+      "epoch": 0.3360322590968733,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.000397311741927225,
+      "loss": 5.3555,
+      "step": 52500
+    },
+    {
+      "epoch": 0.33923256632636734,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00039728613946938906,
+      "loss": 5.3451,
+      "step": 53000
+    },
+    {
+      "epoch": 0.3424328735558614,
+      "grad_norm": 0.267578125,
+      "learning_rate": 0.00039726053701155315,
+      "loss": 5.3454,
+      "step": 53500
+    },
+    {
+      "epoch": 0.3456331807853554,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0003972349345537172,
+      "loss": 5.3493,
+      "step": 54000
+    },
+    {
+      "epoch": 0.3488334880148494,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0003972093320958812,
+      "loss": 5.3508,
+      "step": 54500
+    },
+    {
+      "epoch": 0.3520337952443435,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00039718372963804525,
+      "loss": 5.3413,
+      "step": 55000
+    },
+    {
+      "epoch": 0.35523410247383747,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0003971581271802093,
+      "loss": 5.3471,
+      "step": 55500
+    },
+    {
+      "epoch": 0.3584344097033315,
+      "grad_norm": 0.26953125,
+      "learning_rate": 0.0003971325247223733,
+      "loss": 5.3464,
+      "step": 56000
+    },
+    {
+      "epoch": 0.36163471693282556,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0003971069222645374,
+      "loss": 5.3434,
+      "step": 56500
+    },
+    {
+      "epoch": 0.3648350241623196,
+      "grad_norm": 0.263671875,
+      "learning_rate": 0.0003970813198067015,
+      "loss": 5.3393,
+      "step": 57000
+    },
+    {
+      "epoch": 0.3680353313918136,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00039705571734886553,
+      "loss": 5.3449,
+      "step": 57500
+    },
+    {
+      "epoch": 0.37123563862130765,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00039703011489102956,
+      "loss": 5.3393,
+      "step": 58000
+    },
+    {
+      "epoch": 0.3744359458508017,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0003970045124331936,
+      "loss": 5.3411,
+      "step": 58500
+    },
+    {
+      "epoch": 0.3776362530802957,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0003969789099753577,
+      "loss": 5.3293,
+      "step": 59000
+    },
+    {
+      "epoch": 0.38083656030978974,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0003969533075175217,
+      "loss": 5.3352,
+      "step": 59500
+    },
+    {
+      "epoch": 0.3840368675392838,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.00039692770505968575,
+      "loss": 5.334,
+      "step": 60000
+    },
+    {
+      "epoch": 0.3872371747687778,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0003969021026018498,
+      "loss": 5.3365,
+      "step": 60500
+    },
+    {
+      "epoch": 0.39043748199827183,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0003968765001440138,
+      "loss": 5.3376,
+      "step": 61000
+    },
+    {
+      "epoch": 0.3936377892277659,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0003968508976861779,
+      "loss": 5.332,
+      "step": 61500
+    },
+    {
+      "epoch": 0.39683809645725987,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.00039682529522834195,
+      "loss": 5.3327,
+      "step": 62000
+    },
+    {
+      "epoch": 0.4000384036867539,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.000396799692770506,
+      "loss": 5.3324,
+      "step": 62500
+    },
+    {
+      "epoch": 0.40323871091624797,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00039677409031267,
+      "loss": 5.333,
+      "step": 63000
+    },
+    {
+      "epoch": 0.406439018145742,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00039674848785483405,
+      "loss": 5.3365,
+      "step": 63500
+    },
+    {
+      "epoch": 0.409639325375236,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.00039672288539699814,
+      "loss": 5.3299,
+      "step": 64000
+    },
+    {
+      "epoch": 0.41283963260473006,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00039669728293916217,
+      "loss": 5.3274,
+      "step": 64500
+    },
+    {
+      "epoch": 0.4160399398342241,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00039667168048132626,
+      "loss": 5.3297,
+      "step": 65000
+    },
+    {
+      "epoch": 0.4192402470637181,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0003966460780234903,
+      "loss": 5.3297,
+      "step": 65500
+    },
+    {
+      "epoch": 0.42244055429321214,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00039662047556565433,
+      "loss": 5.3258,
+      "step": 66000
+    },
+    {
+      "epoch": 0.4256408615227062,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00039659487310781836,
+      "loss": 5.3202,
+      "step": 66500
+    },
+    {
+      "epoch": 0.4288411687522002,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00039656927064998245,
+      "loss": 5.3167,
+      "step": 67000
+    },
+    {
+      "epoch": 0.43204147598169423,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0003965436681921465,
+      "loss": 5.3217,
+      "step": 67500
+    },
+    {
+      "epoch": 0.4352417832111883,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0003965180657343105,
+      "loss": 5.3199,
+      "step": 68000
+    },
+    {
+      "epoch": 0.43844209044068233,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00039649246327647455,
+      "loss": 5.3201,
+      "step": 68500
+    },
+    {
+      "epoch": 0.4416423976701763,
+      "grad_norm": 0.2734375,
+      "learning_rate": 0.0003964668608186386,
+      "loss": 5.3223,
+      "step": 69000
+    },
+    {
+      "epoch": 0.44484270489967037,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0003964412583608026,
+      "loss": 5.3276,
+      "step": 69500
+    },
+    {
+      "epoch": 0.4480430121291644,
+      "grad_norm": 0.271484375,
+      "learning_rate": 0.0003964156559029667,
+      "loss": 5.3137,
+      "step": 70000
+    },
+    {
+      "epoch": 0.4512433193586584,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00039639005344513074,
+      "loss": 5.3179,
+      "step": 70500
+    },
+    {
+      "epoch": 0.45444362658815246,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0003963644509872948,
+      "loss": 5.3192,
+      "step": 71000
+    },
+    {
+      "epoch": 0.4576439338176465,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0003963388485294588,
+      "loss": 5.317,
+      "step": 71500
+    },
+    {
+      "epoch": 0.4608442410471405,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0003963132460716229,
+      "loss": 5.3145,
+      "step": 72000
+    },
+    {
+      "epoch": 0.46404454827663455,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00039628764361378694,
+      "loss": 5.3123,
+      "step": 72500
+    },
+    {
+      "epoch": 0.4672448555061286,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.000396262041155951,
+      "loss": 5.3127,
+      "step": 73000
+    },
+    {
+      "epoch": 0.47044516273562265,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00039623643869811506,
+      "loss": 5.3195,
+      "step": 73500
+    },
+    {
+      "epoch": 0.47364546996511664,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0003962108362402791,
+      "loss": 5.3111,
+      "step": 74000
+    },
+    {
+      "epoch": 0.4768457771946107,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00039618523378244313,
+      "loss": 5.3141,
+      "step": 74500
+    },
+    {
+      "epoch": 0.48004608442410474,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00039615963132460716,
+      "loss": 5.3062,
+      "step": 75000
+    },
+    {
+      "epoch": 0.48324639165359873,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00039613402886677125,
+      "loss": 5.3127,
+      "step": 75500
+    },
+    {
+      "epoch": 0.4864466988830928,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0003961084264089353,
+      "loss": 5.31,
+      "step": 76000
+    },
+    {
+      "epoch": 0.4896470061125868,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0003960828239510993,
+      "loss": 5.3139,
+      "step": 76500
+    },
+    {
+      "epoch": 0.4928473133420808,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00039605722149326335,
+      "loss": 5.3109,
+      "step": 77000
+    },
+    {
+      "epoch": 0.49604762057157487,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0003960316190354274,
+      "loss": 5.3037,
+      "step": 77500
+    },
+    {
+      "epoch": 0.4992479278010689,
+      "grad_norm": 0.275390625,
+      "learning_rate": 0.0003960060165775915,
+      "loss": 5.3057,
+      "step": 78000
+    },
+    {
+      "epoch": 0.5024482350305629,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0003959804141197555,
+      "loss": 5.3097,
+      "step": 78500
+    },
+    {
+      "epoch": 0.505648542260057,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00039595481166191954,
+      "loss": 5.3118,
+      "step": 79000
+    },
+    {
+      "epoch": 0.508848849489551,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00039592920920408363,
+      "loss": 5.3031,
+      "step": 79500
+    },
+    {
+      "epoch": 0.512049156719045,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00039590360674624767,
+      "loss": 5.3047,
+      "step": 80000
+    },
+    {
+      "epoch": 0.5152494639485391,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0003958780042884117,
+      "loss": 5.307,
+      "step": 80500
+    },
+    {
+      "epoch": 0.518449771178033,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0003958524018305758,
+      "loss": 5.3015,
+      "step": 81000
+    },
+    {
+      "epoch": 0.5216500784075271,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0003958267993727398,
+      "loss": 5.2987,
+      "step": 81500
+    },
+    {
+      "epoch": 0.5248503856370211,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.00039580119691490386,
+      "loss": 5.2999,
+      "step": 82000
+    },
+    {
+      "epoch": 0.5280506928665152,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.0003957755944570679,
+      "loss": 5.2987,
+      "step": 82500
+    },
+    {
+      "epoch": 0.5312510000960092,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0003957499919992319,
+      "loss": 5.301,
+      "step": 83000
+    },
+    {
+      "epoch": 0.5344513073255033,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.000395724389541396,
+      "loss": 5.3026,
+      "step": 83500
+    },
+    {
+      "epoch": 0.5376516145549973,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00039569878708356005,
+      "loss": 5.2979,
+      "step": 84000
+    },
+    {
+      "epoch": 0.5408519217844913,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0003956731846257241,
+      "loss": 5.3044,
+      "step": 84500
+    },
+    {
+      "epoch": 0.5440522290139853,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0003956475821678881,
+      "loss": 5.3011,
+      "step": 85000
+    },
+    {
+      "epoch": 0.5472525362434794,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.00039562197971005215,
+      "loss": 5.3003,
+      "step": 85500
+    },
+    {
+      "epoch": 0.5504528434729734,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0003955963772522162,
+      "loss": 5.3009,
+      "step": 86000
+    },
+    {
+      "epoch": 0.5536531507024675,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.0003955707747943803,
+      "loss": 5.299,
+      "step": 86500
+    },
+    {
+      "epoch": 0.5568534579319615,
+      "grad_norm": 0.27734375,
+      "learning_rate": 0.0003955451723365443,
+      "loss": 5.295,
+      "step": 87000
+    },
+    {
+      "epoch": 0.5600537651614556,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0003955195698787084,
+      "loss": 5.2971,
+      "step": 87500
+    },
+    {
+      "epoch": 0.5632540723909495,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00039549396742087243,
+      "loss": 5.2982,
+      "step": 88000
+    },
+    {
+      "epoch": 0.5664543796204435,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00039546836496303647,
+      "loss": 5.2923,
+      "step": 88500
+    },
+    {
+      "epoch": 0.5696546868499376,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00039544276250520055,
+      "loss": 5.2898,
+      "step": 89000
+    },
+    {
+      "epoch": 0.5728549940794316,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0003954171600473646,
+      "loss": 5.298,
+      "step": 89500
+    },
+    {
+      "epoch": 0.5760553013089257,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0003953915575895286,
+      "loss": 5.2947,
+      "step": 90000
+    },
+    {
+      "epoch": 0.5792556085384197,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.00039536595513169266,
+      "loss": 5.2937,
+      "step": 90500
+    },
+    {
+      "epoch": 0.5824559157679137,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0003953403526738567,
+      "loss": 5.287,
+      "step": 91000
+    },
+    {
+      "epoch": 0.5856562229974077,
+      "grad_norm": 0.28515625,
+      "learning_rate": 0.0003953147502160207,
+      "loss": 5.2907,
+      "step": 91500
+    },
+    {
+      "epoch": 0.5888565302269018,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0003952891477581848,
+      "loss": 5.2879,
+      "step": 92000
+    },
+    {
+      "epoch": 0.5920568374563958,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00039526354530034885,
+      "loss": 5.2897,
+      "step": 92500
+    },
+    {
+      "epoch": 0.5952571446858899,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0003952379428425129,
+      "loss": 5.2899,
+      "step": 93000
+    },
+    {
+      "epoch": 0.5984574519153839,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0003952123403846769,
+      "loss": 5.2895,
+      "step": 93500
+    },
+    {
+      "epoch": 0.601657759144878,
+      "grad_norm": 0.279296875,
+      "learning_rate": 0.00039518673792684095,
+      "loss": 5.293,
+      "step": 94000
+    },
+    {
+      "epoch": 0.6048580663743719,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00039516113546900504,
+      "loss": 5.2785,
+      "step": 94500
+    },
+    {
+      "epoch": 0.6080583736038659,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00039513553301116913,
+      "loss": 5.2833,
+      "step": 95000
+    },
+    {
+      "epoch": 0.61125868083336,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.00039510993055333316,
+      "loss": 5.2893,
+      "step": 95500
+    },
+    {
+      "epoch": 0.614458988062854,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.0003950843280954972,
+      "loss": 5.2882,
+      "step": 96000
+    },
+    {
+      "epoch": 0.6176592952923481,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00039505872563766123,
+      "loss": 5.2862,
+      "step": 96500
+    },
+    {
+      "epoch": 0.6208596025218421,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0003950331231798253,
+      "loss": 5.2826,
+      "step": 97000
+    },
+    {
+      "epoch": 0.6240599097513362,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00039500752072198935,
+      "loss": 5.2812,
+      "step": 97500
+    },
+    {
+      "epoch": 0.6272602169808301,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.0003949819182641534,
+      "loss": 5.2875,
+      "step": 98000
+    },
+    {
+      "epoch": 0.6304605242103242,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0003949563158063174,
+      "loss": 5.2858,
+      "step": 98500
+    },
+    {
+      "epoch": 0.6336608314398182,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.00039493071334848146,
+      "loss": 5.2816,
+      "step": 99000
+    },
+    {
+      "epoch": 0.6368611386693123,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0003949051108906455,
+      "loss": 5.2786,
+      "step": 99500
+    },
+    {
+      "epoch": 0.6400614458988063,
+      "grad_norm": 0.28125,
+      "learning_rate": 0.0003948795084328096,
+      "loss": 5.2865,
+      "step": 100000
+    },
+    {
+      "epoch": 0.6432617531283004,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0003948539059749736,
+      "loss": 5.2846,
+      "step": 100500
+    },
+    {
+      "epoch": 0.6464620603577943,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00039482830351713765,
+      "loss": 5.2812,
+      "step": 101000
+    },
+    {
+      "epoch": 0.6496623675872883,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0003948027010593017,
+      "loss": 5.2776,
+      "step": 101500
+    },
+    {
+      "epoch": 0.6528626748167824,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00039477709860146577,
+      "loss": 5.2828,
+      "step": 102000
+    },
+    {
+      "epoch": 0.6560629820462764,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0003947514961436298,
+      "loss": 5.28,
+      "step": 102500
+    },
+    {
+      "epoch": 0.6592632892757705,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0003947258936857939,
+      "loss": 5.2795,
+      "step": 103000
+    },
+    {
+      "epoch": 0.6624635965052645,
+      "grad_norm": 0.287109375,
+      "learning_rate": 0.0003947002912279579,
+      "loss": 5.2777,
+      "step": 103500
+    },
+    {
+      "epoch": 0.6656639037347586,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00039467468877012196,
+      "loss": 5.2839,
+      "step": 104000
+    },
+    {
+      "epoch": 0.6688642109642525,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.000394649086312286,
+      "loss": 5.2749,
+      "step": 104500
+    },
+    {
+      "epoch": 0.6720645181937466,
+      "grad_norm": 0.283203125,
+      "learning_rate": 0.00039462348385445003,
+      "loss": 5.2791,
+      "step": 105000
+    },
+    {
+      "epoch": 0.6752648254232406,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0003945978813966141,
+      "loss": 5.2742,
+      "step": 105500
+    },
+    {
+      "epoch": 0.6784651326527347,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.00039457227893877815,
+      "loss": 5.2746,
+      "step": 106000
+    },
+    {
+      "epoch": 0.6816654398822287,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0003945466764809422,
+      "loss": 5.2777,
+      "step": 106500
+    },
+    {
+      "epoch": 0.6848657471117228,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0003945210740231062,
+      "loss": 5.2717,
+      "step": 107000
+    },
+    {
+      "epoch": 0.6880660543412167,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00039449547156527025,
+      "loss": 5.2749,
+      "step": 107500
+    },
+    {
+      "epoch": 0.6912663615707108,
+      "grad_norm": 0.2890625,
+      "learning_rate": 0.00039446986910743434,
+      "loss": 5.2783,
+      "step": 108000
+    },
+    {
+      "epoch": 0.6944666688002048,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0003944442666495984,
+      "loss": 5.2739,
+      "step": 108500
+    },
+    {
+      "epoch": 0.6976669760296988,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0003944186641917624,
+      "loss": 5.2767,
+      "step": 109000
+    },
+    {
+      "epoch": 0.7008672832591929,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00039439306173392645,
+      "loss": 5.2801,
+      "step": 109500
+    },
+    {
+      "epoch": 0.704067590488687,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00039436745927609053,
+      "loss": 5.2789,
+      "step": 110000
+    },
+    {
+      "epoch": 0.707267897718181,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.00039434185681825457,
+      "loss": 5.2754,
+      "step": 110500
+    },
+    {
+      "epoch": 0.7104682049476749,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00039431625436041866,
+      "loss": 5.2722,
+      "step": 111000
+    },
+    {
+      "epoch": 0.713668512177169,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.0003942906519025827,
+      "loss": 5.2671,
+      "step": 111500
+    },
+    {
+      "epoch": 0.716868819406663,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0003942650494447467,
+      "loss": 5.27,
+      "step": 112000
+    },
+    {
+      "epoch": 0.7200691266361571,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.00039423944698691076,
+      "loss": 5.2712,
+      "step": 112500
+    },
+    {
+      "epoch": 0.7232694338656511,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.0003942138445290748,
+      "loss": 5.272,
+      "step": 113000
+    },
+    {
+      "epoch": 0.7264697410951452,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0003941882420712389,
+      "loss": 5.2714,
+      "step": 113500
+    },
+    {
+      "epoch": 0.7296700483246392,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0003941626396134029,
+      "loss": 5.266,
+      "step": 114000
+    },
+    {
+      "epoch": 0.7328703555541332,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00039413703715556695,
+      "loss": 5.2666,
+      "step": 114500
+    },
+    {
+      "epoch": 0.7360706627836272,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.000394111434697731,
+      "loss": 5.2648,
+      "step": 115000
+    },
+    {
+      "epoch": 0.7392709700131213,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.000394085832239895,
+      "loss": 5.266,
+      "step": 115500
+    },
+    {
+      "epoch": 0.7424712772426153,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00039406022978205905,
+      "loss": 5.2665,
+      "step": 116000
+    },
+    {
+      "epoch": 0.7456715844721094,
+      "grad_norm": 0.296875,
+      "learning_rate": 0.00039403462732422314,
+      "loss": 5.269,
+      "step": 116500
+    },
+    {
+      "epoch": 0.7488718917016034,
+      "grad_norm": 0.302734375,
+      "learning_rate": 0.0003940090248663872,
+      "loss": 5.2671,
+      "step": 117000
+    },
+    {
+      "epoch": 0.7520721989310973,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00039398342240855126,
+      "loss": 5.2694,
+      "step": 117500
+    },
+    {
+      "epoch": 0.7552725061605914,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0003939578199507153,
+      "loss": 5.2679,
+      "step": 118000
+    },
+    {
+      "epoch": 0.7584728133900854,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00039393221749287933,
+      "loss": 5.2682,
+      "step": 118500
+    },
+    {
+      "epoch": 0.7616731206195795,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0003939066150350434,
+      "loss": 5.2669,
+      "step": 119000
+    },
+    {
+      "epoch": 0.7648734278490735,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00039388101257720746,
+      "loss": 5.2708,
+      "step": 119500
+    },
+    {
+      "epoch": 0.7680737350785676,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0003938554101193715,
+      "loss": 5.2702,
+      "step": 120000
+    },
+    {
+      "epoch": 0.7712740423080616,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0003938298076615355,
+      "loss": 5.2592,
+      "step": 120500
+    },
+    {
+      "epoch": 0.7744743495375556,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.00039380420520369956,
+      "loss": 5.261,
+      "step": 121000
+    },
+    {
+      "epoch": 0.7776746567670496,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0003937786027458636,
+      "loss": 5.2653,
+      "step": 121500
+    },
+    {
+      "epoch": 0.7808749639965437,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0003937530002880277,
+      "loss": 5.2716,
+      "step": 122000
+    },
+    {
+      "epoch": 0.7840752712260377,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0003937273978301917,
+      "loss": 5.2643,
+      "step": 122500
+    },
+    {
+      "epoch": 0.7872755784555318,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00039370179537235575,
+      "loss": 5.2618,
+      "step": 123000
+    },
+    {
+      "epoch": 0.7904758856850258,
+      "grad_norm": 0.306640625,
+      "learning_rate": 0.0003936761929145198,
+      "loss": 5.2592,
+      "step": 123500
+    },
+    {
+      "epoch": 0.7936761929145197,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0003936505904566838,
+      "loss": 5.262,
+      "step": 124000
+    },
+    {
+      "epoch": 0.7968765001440138,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0003936249879988479,
+      "loss": 5.2561,
+      "step": 124500
+    },
+    {
+      "epoch": 0.8000768073735078,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00039359938554101194,
+      "loss": 5.2625,
+      "step": 125000
+    },
+    {
+      "epoch": 0.8032771146030019,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00039357378308317603,
+      "loss": 5.2566,
+      "step": 125500
+    },
+    {
+      "epoch": 0.8064774218324959,
+      "grad_norm": 0.298828125,
+      "learning_rate": 0.00039354818062534006,
+      "loss": 5.2632,
+      "step": 126000
+    },
+    {
+      "epoch": 0.80967772906199,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0003935225781675041,
+      "loss": 5.2591,
+      "step": 126500
+    },
+    {
+      "epoch": 0.812878036291484,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.00039349697570966813,
+      "loss": 5.26,
+      "step": 127000
+    },
+    {
+      "epoch": 0.816078343520978,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0003934713732518322,
+      "loss": 5.261,
+      "step": 127500
+    },
+    {
+      "epoch": 0.819278650750472,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00039344577079399625,
+      "loss": 5.2639,
+      "step": 128000
+    },
+    {
+      "epoch": 0.8224789579799661,
+      "grad_norm": 0.294921875,
+      "learning_rate": 0.0003934201683361603,
+      "loss": 5.258,
+      "step": 128500
+    },
+    {
+      "epoch": 0.8256792652094601,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0003933945658783243,
+      "loss": 5.2634,
+      "step": 129000
+    },
+    {
+      "epoch": 0.8288795724389542,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00039336896342048836,
+      "loss": 5.2536,
+      "step": 129500
+    },
+    {
+      "epoch": 0.8320798796684482,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.00039334336096265245,
+      "loss": 5.259,
+      "step": 130000
+    },
+    {
+      "epoch": 0.8352801868979423,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0003933177585048165,
+      "loss": 5.256,
+      "step": 130500
+    },
+    {
+      "epoch": 0.8384804941274362,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0003932921560469805,
+      "loss": 5.263,
+      "step": 131000
+    },
+    {
+      "epoch": 0.8416808013569302,
+      "grad_norm": 0.291015625,
+      "learning_rate": 0.00039326655358914455,
+      "loss": 5.2594,
+      "step": 131500
+    },
+    {
+      "epoch": 0.8448811085864243,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00039324095113130864,
+      "loss": 5.254,
+      "step": 132000
+    },
+    {
+      "epoch": 0.8480814158159183,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00039321534867347267,
+      "loss": 5.2611,
+      "step": 132500
+    },
+    {
+      "epoch": 0.8512817230454124,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00039318974621563676,
+      "loss": 5.2558,
+      "step": 133000
+    },
+    {
+      "epoch": 0.8544820302749064,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0003931641437578008,
+      "loss": 5.2568,
+      "step": 133500
+    },
+    {
+      "epoch": 0.8576823375044004,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00039313854129996483,
+      "loss": 5.2554,
+      "step": 134000
+    },
+    {
+      "epoch": 0.8608826447338944,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00039311293884212886,
+      "loss": 5.2527,
+      "step": 134500
+    },
+    {
+      "epoch": 0.8640829519633885,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0003930873363842929,
+      "loss": 5.2552,
+      "step": 135000
+    },
+    {
+      "epoch": 0.8672832591928825,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.000393061733926457,
+      "loss": 5.2521,
+      "step": 135500
+    },
+    {
+      "epoch": 0.8704835664223766,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.000393036131468621,
+      "loss": 5.2537,
+      "step": 136000
+    },
+    {
+      "epoch": 0.8736838736518706,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00039301052901078505,
+      "loss": 5.2509,
+      "step": 136500
+    },
+    {
+      "epoch": 0.8768841808813647,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0003929849265529491,
+      "loss": 5.2577,
+      "step": 137000
+    },
+    {
+      "epoch": 0.8800844881108586,
+      "grad_norm": 0.3125,
+      "learning_rate": 0.0003929593240951131,
+      "loss": 5.2591,
+      "step": 137500
+    },
+    {
+      "epoch": 0.8832847953403526,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0003929337216372772,
+      "loss": 5.2543,
+      "step": 138000
+    },
+    {
+      "epoch": 0.8864851025698467,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.00039290811917944124,
+      "loss": 5.2522,
+      "step": 138500
+    },
+    {
+      "epoch": 0.8896854097993407,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0003928825167216053,
+      "loss": 5.2445,
+      "step": 139000
+    },
+    {
+      "epoch": 0.8928857170288348,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0003928569142637693,
+      "loss": 5.2539,
+      "step": 139500
+    },
+    {
+      "epoch": 0.8960860242583288,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0003928313118059334,
+      "loss": 5.2501,
+      "step": 140000
+    },
+    {
+      "epoch": 0.8992863314878228,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00039280570934809744,
+      "loss": 5.256,
+      "step": 140500
+    },
+    {
+      "epoch": 0.9024866387173168,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0003927801068902615,
+      "loss": 5.256,
+      "step": 141000
+    },
+    {
+      "epoch": 0.9056869459468109,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00039275450443242556,
+      "loss": 5.253,
+      "step": 141500
+    },
+    {
+      "epoch": 0.9088872531763049,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.0003927289019745896,
+      "loss": 5.25,
+      "step": 142000
+    },
+    {
+      "epoch": 0.912087560405799,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.00039270329951675363,
+      "loss": 5.2534,
+      "step": 142500
+    },
+    {
+      "epoch": 0.915287867635293,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00039267769705891766,
+      "loss": 5.253,
+      "step": 143000
+    },
+    {
+      "epoch": 0.9184881748647871,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00039265209460108175,
+      "loss": 5.2497,
+      "step": 143500
+    },
+    {
+      "epoch": 0.921688482094281,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.0003926264921432458,
+      "loss": 5.2563,
+      "step": 144000
+    },
+    {
+      "epoch": 0.924888789323775,
+      "grad_norm": 0.30078125,
+      "learning_rate": 0.0003926008896854098,
+      "loss": 5.2449,
+      "step": 144500
+    },
+    {
+      "epoch": 0.9280890965532691,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00039257528722757385,
+      "loss": 5.2514,
+      "step": 145000
+    },
+    {
+      "epoch": 0.9312894037827631,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0003925496847697379,
+      "loss": 5.2508,
+      "step": 145500
+    },
+    {
+      "epoch": 0.9344897110122572,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0003925240823119019,
+      "loss": 5.2492,
+      "step": 146000
+    },
+    {
+      "epoch": 0.9376900182417512,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.000392498479854066,
+      "loss": 5.2414,
+      "step": 146500
+    },
+    {
+      "epoch": 0.9408903254712453,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00039247287739623004,
+      "loss": 5.2449,
+      "step": 147000
+    },
+    {
+      "epoch": 0.9440906327007392,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0003924472749383941,
+      "loss": 5.2465,
+      "step": 147500
+    },
+    {
+      "epoch": 0.9472909399302333,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00039242167248055817,
+      "loss": 5.246,
+      "step": 148000
+    },
+    {
+      "epoch": 0.9504912471597273,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0003923960700227222,
+      "loss": 5.2489,
+      "step": 148500
+    },
+    {
+      "epoch": 0.9536915543892214,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0003923704675648863,
+      "loss": 5.2438,
+      "step": 149000
+    },
+    {
+      "epoch": 0.9568918616187154,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0003923448651070503,
+      "loss": 5.2472,
+      "step": 149500
+    },
+    {
+      "epoch": 0.9600921688482095,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00039231926264921436,
+      "loss": 5.25,
+      "step": 150000
+    },
+    {
+      "epoch": 0.9632924760777034,
+      "grad_norm": 0.29296875,
+      "learning_rate": 0.0003922936601913784,
+      "loss": 5.2486,
+      "step": 150500
+    },
+    {
+      "epoch": 0.9664927833071975,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0003922680577335424,
+      "loss": 5.2468,
+      "step": 151000
+    },
+    {
+      "epoch": 0.9696930905366915,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00039224245527570646,
+      "loss": 5.2491,
+      "step": 151500
+    },
+    {
+      "epoch": 0.9728933977661856,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00039221685281787055,
+      "loss": 5.2458,
+      "step": 152000
+    },
+    {
+      "epoch": 0.9760937049956796,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0003921912503600346,
+      "loss": 5.2511,
+      "step": 152500
+    },
+    {
+      "epoch": 0.9792940122251736,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0003921656479021986,
+      "loss": 5.2354,
+      "step": 153000
+    },
+    {
+      "epoch": 0.9824943194546677,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00039214004544436265,
+      "loss": 5.2433,
+      "step": 153500
+    },
+    {
+      "epoch": 0.9856946266841616,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0003921144429865267,
+      "loss": 5.2477,
+      "step": 154000
+    },
+    {
+      "epoch": 0.9888949339136557,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0003920888405286908,
+      "loss": 5.2436,
+      "step": 154500
+    },
+    {
+      "epoch": 0.9920952411431497,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0003920632380708548,
+      "loss": 5.2468,
+      "step": 155000
+    },
+    {
+      "epoch": 0.9952955483726438,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0003920376356130189,
+      "loss": 5.2409,
+      "step": 155500
+    },
+    {
+      "epoch": 0.9984958556021378,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00039201203315518293,
+      "loss": 5.2425,
+      "step": 156000
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 5.233697414398193,
+      "eval_runtime": 1.6051,
+      "eval_samples_per_second": 623.012,
+      "eval_steps_per_second": 9.968,
+      "step": 156235
+    },
+    {
+      "epoch": 1.0016961628316319,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00039198643069734697,
+      "loss": 5.2388,
+      "step": 156500
+    },
+    {
+      "epoch": 1.0048964700611258,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.000391960828239511,
+      "loss": 5.2386,
+      "step": 157000
+    },
+    {
+      "epoch": 1.00809677729062,
+      "grad_norm": 0.3046875,
+      "learning_rate": 0.0003919352257816751,
+      "loss": 5.2406,
+      "step": 157500
+    },
+    {
+      "epoch": 1.011297084520114,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0003919096233238391,
+      "loss": 5.2425,
+      "step": 158000
+    },
+    {
+      "epoch": 1.014497391749608,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00039188402086600316,
+      "loss": 5.2406,
+      "step": 158500
+    },
+    {
+      "epoch": 1.017697698979102,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0003918584184081672,
+      "loss": 5.2416,
+      "step": 159000
+    },
+    {
+      "epoch": 1.020898006208596,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003918328159503312,
+      "loss": 5.242,
+      "step": 159500
+    },
+    {
+      "epoch": 1.02409831343809,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0003918072134924953,
+      "loss": 5.2358,
+      "step": 160000
+    },
+    {
+      "epoch": 1.027298620667584,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00039178161103465935,
+      "loss": 5.2389,
+      "step": 160500
+    },
+    {
+      "epoch": 1.0304989278970782,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0003917560085768234,
+      "loss": 5.2403,
+      "step": 161000
+    },
+    {
+      "epoch": 1.0336992351265721,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0003917304061189874,
+      "loss": 5.238,
+      "step": 161500
+    },
+    {
+      "epoch": 1.0368995423560663,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00039170480366115145,
+      "loss": 5.2392,
+      "step": 162000
+    },
+    {
+      "epoch": 1.0400998495855602,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00039167920120331554,
+      "loss": 5.2381,
+      "step": 162500
+    },
+    {
+      "epoch": 1.0433001568150542,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.0003916535987454796,
+      "loss": 5.2354,
+      "step": 163000
+    },
+    {
+      "epoch": 1.0465004640445483,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00039162799628764366,
+      "loss": 5.2342,
+      "step": 163500
+    },
+    {
+      "epoch": 1.0497007712740423,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0003916023938298077,
+      "loss": 5.2378,
+      "step": 164000
+    },
+    {
+      "epoch": 1.0529010785035364,
+      "grad_norm": 0.314453125,
+      "learning_rate": 0.00039157679137197173,
+      "loss": 5.2414,
+      "step": 164500
+    },
+    {
+      "epoch": 1.0561013857330304,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00039155118891413576,
+      "loss": 5.2382,
+      "step": 165000
+    },
+    {
+      "epoch": 1.0593016929625243,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00039152558645629985,
+      "loss": 5.2386,
+      "step": 165500
+    },
+    {
+      "epoch": 1.0625020001920185,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0003914999839984639,
+      "loss": 5.2348,
+      "step": 166000
+    },
+    {
+      "epoch": 1.0657023074215124,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0003914743815406279,
+      "loss": 5.2314,
+      "step": 166500
+    },
+    {
+      "epoch": 1.0689026146510066,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00039144877908279196,
+      "loss": 5.23,
+      "step": 167000
+    },
+    {
+      "epoch": 1.0721029218805005,
+      "grad_norm": 0.31640625,
+      "learning_rate": 0.000391423176624956,
+      "loss": 5.2312,
+      "step": 167500
+    },
+    {
+      "epoch": 1.0753032291099947,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00039139757416712,
+      "loss": 5.2448,
+      "step": 168000
+    },
+    {
+      "epoch": 1.0785035363394886,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.0003913719717092841,
+      "loss": 5.2313,
+      "step": 168500
+    },
+    {
+      "epoch": 1.0817038435689825,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.00039134636925144815,
+      "loss": 5.2329,
+      "step": 169000
+    },
+    {
+      "epoch": 1.0849041507984767,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0003913207667936122,
+      "loss": 5.2307,
+      "step": 169500
+    },
+    {
+      "epoch": 1.0881044580279706,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00039129516433577627,
+      "loss": 5.2377,
+      "step": 170000
+    },
+    {
+      "epoch": 1.0913047652574648,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003912695618779403,
+      "loss": 5.2423,
+      "step": 170500
+    },
+    {
+      "epoch": 1.0945050724869587,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0003912439594201044,
+      "loss": 5.2413,
+      "step": 171000
+    },
+    {
+      "epoch": 1.0977053797164529,
+      "grad_norm": 0.318359375,
+      "learning_rate": 0.0003912183569622684,
+      "loss": 5.2345,
+      "step": 171500
+    },
+    {
+      "epoch": 1.1009056869459468,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00039119275450443246,
+      "loss": 5.2354,
+      "step": 172000
+    },
+    {
+      "epoch": 1.1041059941754408,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0003911671520465965,
+      "loss": 5.2357,
+      "step": 172500
+    },
+    {
+      "epoch": 1.107306301404935,
+      "grad_norm": 0.310546875,
+      "learning_rate": 0.00039114154958876053,
+      "loss": 5.2332,
+      "step": 173000
+    },
+    {
+      "epoch": 1.1105066086344288,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0003911159471309246,
+      "loss": 5.2334,
+      "step": 173500
+    },
+    {
+      "epoch": 1.113706915863923,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00039109034467308865,
+      "loss": 5.2318,
+      "step": 174000
+    },
+    {
+      "epoch": 1.116907223093417,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0003910647422152527,
+      "loss": 5.2356,
+      "step": 174500
+    },
+    {
+      "epoch": 1.1201075303229109,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0003910391397574167,
+      "loss": 5.232,
+      "step": 175000
+    },
+    {
+      "epoch": 1.123307837552405,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.00039101353729958075,
+      "loss": 5.233,
+      "step": 175500
+    },
+    {
+      "epoch": 1.126508144781899,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0003909879348417448,
+      "loss": 5.2262,
+      "step": 176000
+    },
+    {
+      "epoch": 1.1297084520113931,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0003909623323839089,
+      "loss": 5.2293,
+      "step": 176500
+    },
+    {
+      "epoch": 1.132908759240887,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0003909367299260729,
+      "loss": 5.2283,
+      "step": 177000
+    },
+    {
+      "epoch": 1.1361090664703812,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00039091112746823695,
+      "loss": 5.2333,
+      "step": 177500
+    },
+    {
+      "epoch": 1.1393093736998752,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00039088552501040103,
+      "loss": 5.2351,
+      "step": 178000
+    },
+    {
+      "epoch": 1.1425096809293693,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00039085992255256507,
+      "loss": 5.2342,
+      "step": 178500
+    },
+    {
+      "epoch": 1.1457099881588633,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00039083432009472916,
+      "loss": 5.2338,
+      "step": 179000
+    },
+    {
+      "epoch": 1.1489102953883572,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0003908087176368932,
+      "loss": 5.2308,
+      "step": 179500
+    },
+    {
+      "epoch": 1.1521106026178514,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0003907831151790572,
+      "loss": 5.2279,
+      "step": 180000
+    },
+    {
+      "epoch": 1.1553109098473453,
+      "grad_norm": 0.30859375,
+      "learning_rate": 0.00039075751272122126,
+      "loss": 5.2357,
+      "step": 180500
+    },
+    {
+      "epoch": 1.1585112170768395,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0003907319102633853,
+      "loss": 5.2233,
+      "step": 181000
+    },
+    {
+      "epoch": 1.1617115243063334,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00039070630780554933,
+      "loss": 5.2309,
+      "step": 181500
+    },
+    {
+      "epoch": 1.1649118315358273,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0003906807053477134,
+      "loss": 5.2303,
+      "step": 182000
+    },
+    {
+      "epoch": 1.1681121387653215,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.00039065510288987745,
+      "loss": 5.2274,
+      "step": 182500
+    },
+    {
+      "epoch": 1.1713124459948154,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0003906295004320415,
+      "loss": 5.2295,
+      "step": 183000
+    },
+    {
+      "epoch": 1.1745127532243096,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0003906038979742055,
+      "loss": 5.2258,
+      "step": 183500
+    },
+    {
+      "epoch": 1.1777130604538035,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00039057829551636955,
+      "loss": 5.2288,
+      "step": 184000
+    },
+    {
+      "epoch": 1.1809133676832977,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00039055269305853364,
+      "loss": 5.2293,
+      "step": 184500
+    },
+    {
+      "epoch": 1.1841136749127916,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003905270906006977,
+      "loss": 5.2275,
+      "step": 185000
+    },
+    {
+      "epoch": 1.1873139821422858,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00039050148814286176,
+      "loss": 5.2306,
+      "step": 185500
+    },
+    {
+      "epoch": 1.1905142893717797,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0003904758856850258,
+      "loss": 5.2261,
+      "step": 186000
+    },
+    {
+      "epoch": 1.1937145966012737,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00039045028322718983,
+      "loss": 5.2227,
+      "step": 186500
+    },
+    {
+      "epoch": 1.1969149038307678,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00039042468076935387,
+      "loss": 5.2282,
+      "step": 187000
+    },
+    {
+      "epoch": 1.2001152110602618,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00039039907831151796,
+      "loss": 5.2293,
+      "step": 187500
+    },
+    {
+      "epoch": 1.203315518289756,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.000390373475853682,
+      "loss": 5.2249,
+      "step": 188000
+    },
+    {
+      "epoch": 1.2065158255192499,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.000390347873395846,
+      "loss": 5.2263,
+      "step": 188500
+    },
+    {
+      "epoch": 1.2097161327487438,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00039032227093801006,
+      "loss": 5.227,
+      "step": 189000
+    },
+    {
+      "epoch": 1.212916439978238,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0003902966684801741,
+      "loss": 5.2288,
+      "step": 189500
+    },
+    {
+      "epoch": 1.2161167472077319,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0003902710660223382,
+      "loss": 5.2248,
+      "step": 190000
+    },
+    {
+      "epoch": 1.219317054437226,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0003902454635645022,
+      "loss": 5.2225,
+      "step": 190500
+    },
+    {
+      "epoch": 1.22251736166672,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00039021986110666625,
+      "loss": 5.2203,
+      "step": 191000
+    },
+    {
+      "epoch": 1.225717668896214,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0003901942586488303,
+      "loss": 5.2242,
+      "step": 191500
+    },
+    {
+      "epoch": 1.228917976125708,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0003901686561909943,
+      "loss": 5.2195,
+      "step": 192000
+    },
+    {
+      "epoch": 1.232118283355202,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0003901430537331584,
+      "loss": 5.2254,
+      "step": 192500
+    },
+    {
+      "epoch": 1.2353185905846962,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00039011745127532244,
+      "loss": 5.2238,
+      "step": 193000
+    },
+    {
+      "epoch": 1.2385188978141901,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00039009184881748653,
+      "loss": 5.23,
+      "step": 193500
+    },
+    {
+      "epoch": 1.2417192050436843,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00039006624635965056,
+      "loss": 5.2221,
+      "step": 194000
+    },
+    {
+      "epoch": 1.2449195122731782,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0003900406439018146,
+      "loss": 5.2264,
+      "step": 194500
+    },
+    {
+      "epoch": 1.2481198195026724,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00039001504144397863,
+      "loss": 5.2214,
+      "step": 195000
+    },
+    {
+      "epoch": 1.2513201267321663,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0003899894389861427,
+      "loss": 5.2222,
+      "step": 195500
+    },
+    {
+      "epoch": 1.2545204339616602,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00038996383652830676,
+      "loss": 5.2248,
+      "step": 196000
+    },
+    {
+      "epoch": 1.2577207411911544,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003899382340704708,
+      "loss": 5.2203,
+      "step": 196500
+    },
+    {
+      "epoch": 1.2609210484206483,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0003899126316126348,
+      "loss": 5.222,
+      "step": 197000
+    },
+    {
+      "epoch": 1.2641213556501425,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00038988702915479886,
+      "loss": 5.2234,
+      "step": 197500
+    },
+    {
+      "epoch": 1.2673216628796364,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003898614266969629,
+      "loss": 5.2289,
+      "step": 198000
+    },
+    {
+      "epoch": 1.2705219701091304,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.000389835824239127,
+      "loss": 5.2214,
+      "step": 198500
+    },
+    {
+      "epoch": 1.2737222773386245,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.000389810221781291,
+      "loss": 5.2235,
+      "step": 199000
+    },
+    {
+      "epoch": 1.2769225845681185,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.00038978461932345505,
+      "loss": 5.2256,
+      "step": 199500
+    },
+    {
+      "epoch": 1.2801228917976126,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0003897590168656191,
+      "loss": 5.2266,
+      "step": 200000
+    },
+    {
+      "epoch": 1.2833231990271066,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00038973341440778317,
+      "loss": 5.2288,
+      "step": 200500
+    },
+    {
+      "epoch": 1.2865235062566005,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0003897078119499472,
+      "loss": 5.2253,
+      "step": 201000
+    },
+    {
+      "epoch": 1.2897238134860947,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0003896822094921113,
+      "loss": 5.2216,
+      "step": 201500
+    },
+    {
+      "epoch": 1.2929241207155888,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00038965660703427533,
+      "loss": 5.2229,
+      "step": 202000
+    },
+    {
+      "epoch": 1.2961244279450828,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00038963100457643936,
+      "loss": 5.2239,
+      "step": 202500
+    },
+    {
+      "epoch": 1.2993247351745767,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003896054021186034,
+      "loss": 5.217,
+      "step": 203000
+    },
+    {
+      "epoch": 1.3025250424040709,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00038957979966076743,
+      "loss": 5.2295,
+      "step": 203500
+    },
+    {
+      "epoch": 1.3057253496335648,
+      "grad_norm": 0.32421875,
+      "learning_rate": 0.0003895541972029315,
+      "loss": 5.2226,
+      "step": 204000
+    },
+    {
+      "epoch": 1.308925656863059,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.00038952859474509555,
+      "loss": 5.224,
+      "step": 204500
+    },
+    {
+      "epoch": 1.3121259640925529,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0003895029922872596,
+      "loss": 5.2216,
+      "step": 205000
+    },
+    {
+      "epoch": 1.3153262713220468,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0003894773898294236,
+      "loss": 5.2219,
+      "step": 205500
+    },
+    {
+      "epoch": 1.318526578551541,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00038945178737158766,
+      "loss": 5.2175,
+      "step": 206000
+    },
+    {
+      "epoch": 1.321726885781035,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00038942618491375175,
+      "loss": 5.2167,
+      "step": 206500
+    },
+    {
+      "epoch": 1.324927193010529,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.0003894005824559158,
+      "loss": 5.2227,
+      "step": 207000
+    },
+    {
+      "epoch": 1.328127500240023,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0003893749799980798,
+      "loss": 5.2199,
+      "step": 207500
+    },
+    {
+      "epoch": 1.331327807469517,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003893493775402439,
+      "loss": 5.2198,
+      "step": 208000
+    },
+    {
+      "epoch": 1.3345281146990111,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00038932377508240794,
+      "loss": 5.2206,
+      "step": 208500
+    },
+    {
+      "epoch": 1.3377284219285053,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.000389298172624572,
+      "loss": 5.2217,
+      "step": 209000
+    },
+    {
+      "epoch": 1.3409287291579992,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00038927257016673606,
+      "loss": 5.2231,
+      "step": 209500
+    },
+    {
+      "epoch": 1.3441290363874931,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003892469677089001,
+      "loss": 5.2211,
+      "step": 210000
+    },
+    {
+      "epoch": 1.3473293436169873,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00038922136525106413,
+      "loss": 5.2185,
+      "step": 210500
+    },
+    {
+      "epoch": 1.3505296508464812,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00038919576279322816,
+      "loss": 5.2207,
+      "step": 211000
+    },
+    {
+      "epoch": 1.3537299580759754,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.0003891701603353922,
+      "loss": 5.23,
+      "step": 211500
+    },
+    {
+      "epoch": 1.3569302653054693,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0003891445578775563,
+      "loss": 5.2224,
+      "step": 212000
+    },
+    {
+      "epoch": 1.3601305725349633,
+      "grad_norm": 0.3203125,
+      "learning_rate": 0.0003891189554197203,
+      "loss": 5.2179,
+      "step": 212500
+    },
+    {
+      "epoch": 1.3633308797644574,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00038909335296188435,
+      "loss": 5.2214,
+      "step": 213000
+    },
+    {
+      "epoch": 1.3665311869939514,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0003890677505040484,
+      "loss": 5.2253,
+      "step": 213500
+    },
+    {
+      "epoch": 1.3697314942234455,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0003890421480462124,
+      "loss": 5.2184,
+      "step": 214000
+    },
+    {
+      "epoch": 1.3729318014529395,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0003890165455883765,
+      "loss": 5.2235,
+      "step": 214500
+    },
+    {
+      "epoch": 1.3761321086824334,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00038899094313054054,
+      "loss": 5.2209,
+      "step": 215000
+    },
+    {
+      "epoch": 1.3793324159119276,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003889653406727046,
+      "loss": 5.2172,
+      "step": 215500
+    },
+    {
+      "epoch": 1.3825327231414215,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00038893973821486867,
+      "loss": 5.2177,
+      "step": 216000
+    },
+    {
+      "epoch": 1.3857330303709157,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0003889141357570327,
+      "loss": 5.2171,
+      "step": 216500
+    },
+    {
+      "epoch": 1.3889333376004096,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00038888853329919674,
+      "loss": 5.2189,
+      "step": 217000
+    },
+    {
+      "epoch": 1.3921336448299035,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0003888629308413608,
+      "loss": 5.2184,
+      "step": 217500
+    },
+    {
+      "epoch": 1.3953339520593977,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00038883732838352486,
+      "loss": 5.2201,
+      "step": 218000
+    },
+    {
+      "epoch": 1.3985342592888919,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0003888117259256889,
+      "loss": 5.2189,
+      "step": 218500
+    },
+    {
+      "epoch": 1.4017345665183858,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0003887861234678529,
+      "loss": 5.219,
+      "step": 219000
+    },
+    {
+      "epoch": 1.4049348737478797,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00038876052101001696,
+      "loss": 5.2243,
+      "step": 219500
+    },
+    {
+      "epoch": 1.408135180977374,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.00038873491855218105,
+      "loss": 5.218,
+      "step": 220000
+    },
+    {
+      "epoch": 1.4113354882068678,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0003887093160943451,
+      "loss": 5.2208,
+      "step": 220500
+    },
+    {
+      "epoch": 1.414535795436362,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0003886837136365091,
+      "loss": 5.2175,
+      "step": 221000
+    },
+    {
+      "epoch": 1.417736102665856,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00038865811117867315,
+      "loss": 5.2186,
+      "step": 221500
+    },
+    {
+      "epoch": 1.4209364098953499,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0003886325087208372,
+      "loss": 5.215,
+      "step": 222000
+    },
+    {
+      "epoch": 1.424136717124844,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0003886069062630012,
+      "loss": 5.2184,
+      "step": 222500
+    },
+    {
+      "epoch": 1.427337024354338,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.0003885813038051653,
+      "loss": 5.2215,
+      "step": 223000
+    },
+    {
+      "epoch": 1.4305373315838321,
+      "grad_norm": 0.326171875,
+      "learning_rate": 0.0003885557013473294,
+      "loss": 5.2166,
+      "step": 223500
+    },
+    {
+      "epoch": 1.433737638813326,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00038853009888949343,
+      "loss": 5.2194,
+      "step": 224000
+    },
+    {
+      "epoch": 1.43693794604282,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00038850449643165747,
+      "loss": 5.2188,
+      "step": 224500
+    },
+    {
+      "epoch": 1.4401382532723142,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0003884788939738215,
+      "loss": 5.216,
+      "step": 225000
+    },
+    {
+      "epoch": 1.4433385605018083,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0003884532915159856,
+      "loss": 5.216,
+      "step": 225500
+    },
+    {
+      "epoch": 1.4465388677313022,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0003884276890581496,
+      "loss": 5.2194,
+      "step": 226000
+    },
+    {
+      "epoch": 1.4497391749607962,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00038840208660031366,
+      "loss": 5.2202,
+      "step": 226500
+    },
+    {
+      "epoch": 1.4529394821902903,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.0003883764841424777,
+      "loss": 5.2207,
+      "step": 227000
+    },
+    {
+      "epoch": 1.4561397894197843,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0003883508816846417,
+      "loss": 5.2209,
+      "step": 227500
+    },
+    {
+      "epoch": 1.4593400966492784,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00038832527922680576,
+      "loss": 5.2265,
+      "step": 228000
+    },
+    {
+      "epoch": 1.4625404038787724,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00038829967676896985,
+      "loss": 5.2122,
+      "step": 228500
+    },
+    {
+      "epoch": 1.4657407111082663,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0003882740743111339,
+      "loss": 5.211,
+      "step": 229000
+    },
+    {
+      "epoch": 1.4689410183377605,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0003882484718532979,
+      "loss": 5.2213,
+      "step": 229500
+    },
+    {
+      "epoch": 1.4721413255672544,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00038822286939546195,
+      "loss": 5.2215,
+      "step": 230000
+    },
+    {
+      "epoch": 1.4753416327967486,
+      "grad_norm": 0.33203125,
+      "learning_rate": 0.00038819726693762604,
+      "loss": 5.2211,
+      "step": 230500
+    },
+    {
+      "epoch": 1.4785419400262425,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003881716644797901,
+      "loss": 5.2163,
+      "step": 231000
+    },
+    {
+      "epoch": 1.4817422472557364,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00038814606202195416,
+      "loss": 5.2206,
+      "step": 231500
+    },
+    {
+      "epoch": 1.4849425544852306,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0003881204595641182,
+      "loss": 5.2235,
+      "step": 232000
+    },
+    {
+      "epoch": 1.4881428617147245,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00038809485710628223,
+      "loss": 5.2167,
+      "step": 232500
+    },
+    {
+      "epoch": 1.4913431689442187,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.00038806925464844626,
+      "loss": 5.2211,
+      "step": 233000
+    },
+    {
+      "epoch": 1.4945434761737126,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003880436521906103,
+      "loss": 5.2137,
+      "step": 233500
+    },
+    {
+      "epoch": 1.4977437834032066,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0003880180497327744,
+      "loss": 5.2173,
+      "step": 234000
+    },
+    {
+      "epoch": 1.5009440906327007,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0003879924472749384,
+      "loss": 5.2144,
+      "step": 234500
+    },
+    {
+      "epoch": 1.504144397862195,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00038796684481710246,
+      "loss": 5.2201,
+      "step": 235000
+    },
+    {
+      "epoch": 1.5073447050916888,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0003879412423592665,
+      "loss": 5.2115,
+      "step": 235500
+    },
+    {
+      "epoch": 1.5105450123211828,
+      "grad_norm": 0.333984375,
+      "learning_rate": 0.0003879156399014305,
+      "loss": 5.2141,
+      "step": 236000
+    },
+    {
+      "epoch": 1.5137453195506767,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0003878900374435946,
+      "loss": 5.2155,
+      "step": 236500
+    },
+    {
+      "epoch": 1.5169456267801709,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00038786443498575865,
+      "loss": 5.2129,
+      "step": 237000
+    },
+    {
+      "epoch": 1.520145934009665,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0003878388325279227,
+      "loss": 5.2213,
+      "step": 237500
+    },
+    {
+      "epoch": 1.523346241239159,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0003878132300700867,
+      "loss": 5.2209,
+      "step": 238000
+    },
+    {
+      "epoch": 1.526546548468653,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0003877876276122508,
+      "loss": 5.2169,
+      "step": 238500
+    },
+    {
+      "epoch": 1.529746855698147,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00038776202515441484,
+      "loss": 5.2133,
+      "step": 239000
+    },
+    {
+      "epoch": 1.5329471629276412,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0003877364226965789,
+      "loss": 5.2157,
+      "step": 239500
+    },
+    {
+      "epoch": 1.5361474701571352,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00038771082023874296,
+      "loss": 5.2146,
+      "step": 240000
+    },
+    {
+      "epoch": 1.539347777386629,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.000387685217780907,
+      "loss": 5.2147,
+      "step": 240500
+    },
+    {
+      "epoch": 1.542548084616123,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00038765961532307103,
+      "loss": 5.2205,
+      "step": 241000
+    },
+    {
+      "epoch": 1.5457483918456172,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00038763401286523506,
+      "loss": 5.215,
+      "step": 241500
+    },
+    {
+      "epoch": 1.5489486990751113,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00038760841040739915,
+      "loss": 5.2172,
+      "step": 242000
+    },
+    {
+      "epoch": 1.5521490063046053,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0003875828079495632,
+      "loss": 5.2126,
+      "step": 242500
+    },
+    {
+      "epoch": 1.5553493135340992,
+      "grad_norm": 0.328125,
+      "learning_rate": 0.0003875572054917272,
+      "loss": 5.2145,
+      "step": 243000
+    },
+    {
+      "epoch": 1.5585496207635932,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.00038753160303389125,
+      "loss": 5.2084,
+      "step": 243500
+    },
+    {
+      "epoch": 1.5617499279930873,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0003875060005760553,
+      "loss": 5.2203,
+      "step": 244000
+    },
+    {
+      "epoch": 1.5649502352225815,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0003874803981182193,
+      "loss": 5.2093,
+      "step": 244500
+    },
+    {
+      "epoch": 1.5681505424520754,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0003874547956603834,
+      "loss": 5.2172,
+      "step": 245000
+    },
+    {
+      "epoch": 1.5713508496815694,
+      "grad_norm": 0.322265625,
+      "learning_rate": 0.00038742919320254745,
+      "loss": 5.2117,
+      "step": 245500
+    },
+    {
+      "epoch": 1.5745511569110635,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00038740359074471153,
+      "loss": 5.2167,
+      "step": 246000
+    },
+    {
+      "epoch": 1.5777514641405574,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00038737798828687557,
+      "loss": 5.2152,
+      "step": 246500
+    },
+    {
+      "epoch": 1.5809517713700516,
+      "grad_norm": 0.337890625,
+      "learning_rate": 0.0003873523858290396,
+      "loss": 5.2127,
+      "step": 247000
+    },
+    {
+      "epoch": 1.5841520785995455,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0003873267833712037,
+      "loss": 5.2125,
+      "step": 247500
+    },
+    {
+      "epoch": 1.5873523858290395,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0003873011809133677,
+      "loss": 5.215,
+      "step": 248000
+    },
+    {
+      "epoch": 1.5905526930585336,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00038727557845553176,
+      "loss": 5.2106,
+      "step": 248500
+    },
+    {
+      "epoch": 1.5937530002880278,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0003872499759976958,
+      "loss": 5.2123,
+      "step": 249000
+    },
+    {
+      "epoch": 1.5969533075175217,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00038722437353985983,
+      "loss": 5.2162,
+      "step": 249500
+    },
+    {
+      "epoch": 1.6001536147470157,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0003871987710820239,
+      "loss": 5.2224,
+      "step": 250000
+    },
+    {
+      "epoch": 1.6033539219765096,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00038717316862418795,
+      "loss": 5.2105,
+      "step": 250500
+    },
+    {
+      "epoch": 1.6065542292060038,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.000387147566166352,
+      "loss": 5.2144,
+      "step": 251000
+    },
+    {
+      "epoch": 1.609754536435498,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.000387121963708516,
+      "loss": 5.2169,
+      "step": 251500
+    },
+    {
+      "epoch": 1.6129548436649919,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00038709636125068005,
+      "loss": 5.2141,
+      "step": 252000
+    },
+    {
+      "epoch": 1.6161551508944858,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0003870707587928441,
+      "loss": 5.2145,
+      "step": 252500
+    },
+    {
+      "epoch": 1.6193554581239797,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0003870451563350082,
+      "loss": 5.2192,
+      "step": 253000
+    },
+    {
+      "epoch": 1.622555765353474,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003870195538771722,
+      "loss": 5.2116,
+      "step": 253500
+    },
+    {
+      "epoch": 1.625756072582968,
+      "grad_norm": 0.330078125,
+      "learning_rate": 0.0003869939514193363,
+      "loss": 5.2121,
+      "step": 254000
+    },
+    {
+      "epoch": 1.628956379812462,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00038696834896150033,
+      "loss": 5.2086,
+      "step": 254500
+    },
+    {
+      "epoch": 1.632156687041956,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00038694274650366437,
+      "loss": 5.2137,
+      "step": 255000
+    },
+    {
+      "epoch": 1.63535699427145,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00038691714404582846,
+      "loss": 5.2148,
+      "step": 255500
+    },
+    {
+      "epoch": 1.6385573015009443,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003868915415879925,
+      "loss": 5.2191,
+      "step": 256000
+    },
+    {
+      "epoch": 1.6417576087304382,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0003868659391301565,
+      "loss": 5.2119,
+      "step": 256500
+    },
+    {
+      "epoch": 1.6449579159599321,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038684033667232056,
+      "loss": 5.2129,
+      "step": 257000
+    },
+    {
+      "epoch": 1.648158223189426,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0003868147342144846,
+      "loss": 5.2107,
+      "step": 257500
+    },
+    {
+      "epoch": 1.6513585304189202,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00038678913175664863,
+      "loss": 5.2092,
+      "step": 258000
+    },
+    {
+      "epoch": 1.6545588376484144,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003867635292988127,
+      "loss": 5.2142,
+      "step": 258500
+    },
+    {
+      "epoch": 1.6577591448779083,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00038673792684097675,
+      "loss": 5.2097,
+      "step": 259000
+    },
+    {
+      "epoch": 1.6609594521074023,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0003867123243831408,
+      "loss": 5.2178,
+      "step": 259500
+    },
+    {
+      "epoch": 1.6641597593368962,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0003866867219253048,
+      "loss": 5.2202,
+      "step": 260000
+    },
+    {
+      "epoch": 1.6673600665663904,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00038666111946746885,
+      "loss": 5.2108,
+      "step": 260500
+    },
+    {
+      "epoch": 1.6705603737958845,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00038663551700963294,
+      "loss": 5.2111,
+      "step": 261000
+    },
+    {
+      "epoch": 1.6737606810253784,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00038660991455179703,
+      "loss": 5.2115,
+      "step": 261500
+    },
+    {
+      "epoch": 1.6769609882548724,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.00038658431209396106,
+      "loss": 5.2137,
+      "step": 262000
+    },
+    {
+      "epoch": 1.6801612954843665,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0003865587096361251,
+      "loss": 5.2147,
+      "step": 262500
+    },
+    {
+      "epoch": 1.6833616027138605,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00038653310717828913,
+      "loss": 5.2157,
+      "step": 263000
+    },
+    {
+      "epoch": 1.6865619099433546,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00038650750472045317,
+      "loss": 5.2153,
+      "step": 263500
+    },
+    {
+      "epoch": 1.6897622171728486,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00038648190226261726,
+      "loss": 5.2073,
+      "step": 264000
+    },
+    {
+      "epoch": 1.6929625244023425,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0003864562998047813,
+      "loss": 5.2185,
+      "step": 264500
+    },
+    {
+      "epoch": 1.6961628316318367,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0003864306973469453,
+      "loss": 5.2131,
+      "step": 265000
+    },
+    {
+      "epoch": 1.6993631388613308,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00038640509488910936,
+      "loss": 5.2177,
+      "step": 265500
+    },
+    {
+      "epoch": 1.7025634460908248,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003863794924312734,
+      "loss": 5.2098,
+      "step": 266000
+    },
+    {
+      "epoch": 1.7057637533203187,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0003863538899734375,
+      "loss": 5.2088,
+      "step": 266500
+    },
+    {
+      "epoch": 1.7089640605498126,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.0003863282875156015,
+      "loss": 5.2095,
+      "step": 267000
+    },
+    {
+      "epoch": 1.7121643677793068,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00038630268505776555,
+      "loss": 5.2109,
+      "step": 267500
+    },
+    {
+      "epoch": 1.715364675008801,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0003862770825999296,
+      "loss": 5.2124,
+      "step": 268000
+    },
+    {
+      "epoch": 1.718564982238295,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00038625148014209367,
+      "loss": 5.2093,
+      "step": 268500
+    },
+    {
+      "epoch": 1.7217652894677888,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0003862258776842577,
+      "loss": 5.2065,
+      "step": 269000
+    },
+    {
+      "epoch": 1.7249655966972828,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003862002752264218,
+      "loss": 5.2177,
+      "step": 269500
+    },
+    {
+      "epoch": 1.728165903926777,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00038617467276858583,
+      "loss": 5.2129,
+      "step": 270000
+    },
+    {
+      "epoch": 1.731366211156271,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00038614907031074986,
+      "loss": 5.2104,
+      "step": 270500
+    },
+    {
+      "epoch": 1.734566518385765,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0003861234678529139,
+      "loss": 5.2146,
+      "step": 271000
+    },
+    {
+      "epoch": 1.737766825615259,
+      "grad_norm": 0.341796875,
+      "learning_rate": 0.00038609786539507793,
+      "loss": 5.21,
+      "step": 271500
+    },
+    {
+      "epoch": 1.7409671328447531,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.000386072262937242,
+      "loss": 5.2105,
+      "step": 272000
+    },
+    {
+      "epoch": 1.7441674400742473,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00038604666047940605,
+      "loss": 5.2113,
+      "step": 272500
+    },
+    {
+      "epoch": 1.7473677473037412,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003860210580215701,
+      "loss": 5.2127,
+      "step": 273000
+    },
+    {
+      "epoch": 1.7505680545332352,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0003859954555637341,
+      "loss": 5.2113,
+      "step": 273500
+    },
+    {
+      "epoch": 1.753768361762729,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.00038596985310589816,
+      "loss": 5.2137,
+      "step": 274000
+    },
+    {
+      "epoch": 1.7569686689922233,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0003859442506480622,
+      "loss": 5.2108,
+      "step": 274500
+    },
+    {
+      "epoch": 1.7601689762217174,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0003859186481902263,
+      "loss": 5.2108,
+      "step": 275000
+    },
+    {
+      "epoch": 1.7633692834512114,
+      "grad_norm": 0.345703125,
+      "learning_rate": 0.0003858930457323903,
+      "loss": 5.2107,
+      "step": 275500
+    },
+    {
+      "epoch": 1.7665695906807053,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.00038586744327455435,
+      "loss": 5.2126,
+      "step": 276000
+    },
+    {
+      "epoch": 1.7697698979101992,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00038584184081671844,
+      "loss": 5.2125,
+      "step": 276500
+    },
+    {
+      "epoch": 1.7729702051396934,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00038581623835888247,
+      "loss": 5.2136,
+      "step": 277000
+    },
+    {
+      "epoch": 1.7761705123691875,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00038579063590104656,
+      "loss": 5.2136,
+      "step": 277500
+    },
+    {
+      "epoch": 1.7793708195986815,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0003857650334432106,
+      "loss": 5.2113,
+      "step": 278000
+    },
+    {
+      "epoch": 1.7825711268281754,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.00038573943098537463,
+      "loss": 5.2126,
+      "step": 278500
+    },
+    {
+      "epoch": 1.7857714340576696,
+      "grad_norm": 0.3359375,
+      "learning_rate": 0.00038571382852753866,
+      "loss": 5.2121,
+      "step": 279000
+    },
+    {
+      "epoch": 1.7889717412871635,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0003856882260697027,
+      "loss": 5.2061,
+      "step": 279500
+    },
+    {
+      "epoch": 1.7921720485166577,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00038566262361186673,
+      "loss": 5.2132,
+      "step": 280000
+    },
+    {
+      "epoch": 1.7953723557461516,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0003856370211540308,
+      "loss": 5.2158,
+      "step": 280500
+    },
+    {
+      "epoch": 1.7985726629756456,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00038561141869619485,
+      "loss": 5.2113,
+      "step": 281000
+    },
+    {
+      "epoch": 1.8017729702051397,
+      "grad_norm": 0.34375,
+      "learning_rate": 0.0003855858162383589,
+      "loss": 5.2072,
+      "step": 281500
+    },
+    {
+      "epoch": 1.8049732774346339,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0003855602137805229,
+      "loss": 5.2132,
+      "step": 282000
+    },
+    {
+      "epoch": 1.8081735846641278,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00038553461132268696,
+      "loss": 5.2134,
+      "step": 282500
+    },
+    {
+      "epoch": 1.8113738918936217,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00038550900886485104,
+      "loss": 5.2122,
+      "step": 283000
+    },
+    {
+      "epoch": 1.8145741991231157,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003854834064070151,
+      "loss": 5.2105,
+      "step": 283500
+    },
+    {
+      "epoch": 1.8177745063526098,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00038545780394917917,
+      "loss": 5.2108,
+      "step": 284000
+    },
+    {
+      "epoch": 1.820974813582104,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003854322014913432,
+      "loss": 5.2122,
+      "step": 284500
+    },
+    {
+      "epoch": 1.824175120811598,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00038540659903350724,
+      "loss": 5.2076,
+      "step": 285000
+    },
+    {
+      "epoch": 1.8273754280410919,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.0003853809965756713,
+      "loss": 5.2107,
+      "step": 285500
+    },
+    {
+      "epoch": 1.8305757352705858,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00038535539411783536,
+      "loss": 5.2128,
+      "step": 286000
+    },
+    {
+      "epoch": 1.83377604250008,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0003853297916599994,
+      "loss": 5.2076,
+      "step": 286500
+    },
+    {
+      "epoch": 1.8369763497295741,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0003853041892021634,
+      "loss": 5.2113,
+      "step": 287000
+    },
+    {
+      "epoch": 1.840176656959068,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00038527858674432746,
+      "loss": 5.2084,
+      "step": 287500
+    },
+    {
+      "epoch": 1.843376964188562,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003852529842864915,
+      "loss": 5.2152,
+      "step": 288000
+    },
+    {
+      "epoch": 1.8465772714180562,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.0003852273818286556,
+      "loss": 5.2082,
+      "step": 288500
+    },
+    {
+      "epoch": 1.8497775786475503,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0003852017793708196,
+      "loss": 5.2145,
+      "step": 289000
+    },
+    {
+      "epoch": 1.8529778858770443,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00038517617691298365,
+      "loss": 5.2083,
+      "step": 289500
+    },
+    {
+      "epoch": 1.8561781931065382,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0003851505744551477,
+      "loss": 5.2103,
+      "step": 290000
+    },
+    {
+      "epoch": 1.8593785003360321,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003851249719973117,
+      "loss": 5.2066,
+      "step": 290500
+    },
+    {
+      "epoch": 1.8625788075655263,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0003850993695394758,
+      "loss": 5.2152,
+      "step": 291000
+    },
+    {
+      "epoch": 1.8657791147950205,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00038507376708163984,
+      "loss": 5.2144,
+      "step": 291500
+    },
+    {
+      "epoch": 1.8689794220245144,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00038504816462380393,
+      "loss": 5.2091,
+      "step": 292000
+    },
+    {
+      "epoch": 1.8721797292540083,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00038502256216596797,
+      "loss": 5.2137,
+      "step": 292500
+    },
+    {
+      "epoch": 1.8753800364835023,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.000384996959708132,
+      "loss": 5.2023,
+      "step": 293000
+    },
+    {
+      "epoch": 1.8785803437129964,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00038497135725029603,
+      "loss": 5.2073,
+      "step": 293500
+    },
+    {
+      "epoch": 1.8817806509424906,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0003849457547924601,
+      "loss": 5.2089,
+      "step": 294000
+    },
+    {
+      "epoch": 1.8849809581719845,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00038492015233462416,
+      "loss": 5.2069,
+      "step": 294500
+    },
+    {
+      "epoch": 1.8881812654014785,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003848945498767882,
+      "loss": 5.2084,
+      "step": 295000
+    },
+    {
+      "epoch": 1.8913815726309726,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0003848689474189522,
+      "loss": 5.2065,
+      "step": 295500
+    },
+    {
+      "epoch": 1.8945818798604666,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00038484334496111626,
+      "loss": 5.2123,
+      "step": 296000
+    },
+    {
+      "epoch": 1.8977821870899607,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00038481774250328035,
+      "loss": 5.2105,
+      "step": 296500
+    },
+    {
+      "epoch": 1.9009824943194547,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003847921400454444,
+      "loss": 5.2092,
+      "step": 297000
+    },
+    {
+      "epoch": 1.9041828015489486,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0003847665375876084,
+      "loss": 5.2042,
+      "step": 297500
+    },
+    {
+      "epoch": 1.9073831087784427,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00038474093512977245,
+      "loss": 5.2083,
+      "step": 298000
+    },
+    {
+      "epoch": 1.910583416007937,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038471533267193654,
+      "loss": 5.2129,
+      "step": 298500
+    },
+    {
+      "epoch": 1.9137837232374308,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0003846897302141006,
+      "loss": 5.2136,
+      "step": 299000
+    },
+    {
+      "epoch": 1.9169840304669248,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.00038466412775626466,
+      "loss": 5.2083,
+      "step": 299500
+    },
+    {
+      "epoch": 1.9201843376964187,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0003846385252984287,
+      "loss": 5.2071,
+      "step": 300000
+    },
+    {
+      "epoch": 1.9233846449259129,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00038461292284059273,
+      "loss": 5.2018,
+      "step": 300500
+    },
+    {
+      "epoch": 1.926584952155407,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00038458732038275676,
+      "loss": 5.2091,
+      "step": 301000
+    },
+    {
+      "epoch": 1.929785259384901,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003845617179249208,
+      "loss": 5.2078,
+      "step": 301500
+    },
+    {
+      "epoch": 1.932985566614395,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003845361154670849,
+      "loss": 5.2102,
+      "step": 302000
+    },
+    {
+      "epoch": 1.9361858738438888,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0003845105130092489,
+      "loss": 5.2069,
+      "step": 302500
+    },
+    {
+      "epoch": 1.939386181073383,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00038448491055141296,
+      "loss": 5.2084,
+      "step": 303000
+    },
+    {
+      "epoch": 1.9425864883028772,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.000384459308093577,
+      "loss": 5.208,
+      "step": 303500
+    },
+    {
+      "epoch": 1.945786795532371,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.000384433705635741,
+      "loss": 5.2096,
+      "step": 304000
+    },
+    {
+      "epoch": 1.948987102761865,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00038440810317790506,
+      "loss": 5.208,
+      "step": 304500
+    },
+    {
+      "epoch": 1.9521874099913592,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.00038438250072006915,
+      "loss": 5.2065,
+      "step": 305000
+    },
+    {
+      "epoch": 1.9553877172208534,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0003843568982622332,
+      "loss": 5.2092,
+      "step": 305500
+    },
+    {
+      "epoch": 1.9585880244503473,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.0003843312958043972,
+      "loss": 5.2149,
+      "step": 306000
+    },
+    {
+      "epoch": 1.9617883316798412,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003843056933465613,
+      "loss": 5.2069,
+      "step": 306500
+    },
+    {
+      "epoch": 1.9649886389093352,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00038428009088872534,
+      "loss": 5.2079,
+      "step": 307000
+    },
+    {
+      "epoch": 1.9681889461388293,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0003842544884308894,
+      "loss": 5.2044,
+      "step": 307500
+    },
+    {
+      "epoch": 1.9713892533683235,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00038422888597305346,
+      "loss": 5.2109,
+      "step": 308000
+    },
+    {
+      "epoch": 1.9745895605978174,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003842032835152175,
+      "loss": 5.2061,
+      "step": 308500
+    },
+    {
+      "epoch": 1.9777898678273114,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00038417768105738153,
+      "loss": 5.2095,
+      "step": 309000
+    },
+    {
+      "epoch": 1.9809901750568053,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00038415207859954556,
+      "loss": 5.2049,
+      "step": 309500
+    },
+    {
+      "epoch": 1.9841904822862995,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0003841264761417096,
+      "loss": 5.2041,
+      "step": 310000
+    },
+    {
+      "epoch": 1.9873907895157936,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0003841008736838737,
+      "loss": 5.1995,
+      "step": 310500
+    },
+    {
+      "epoch": 1.9905910967452876,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0003840752712260377,
+      "loss": 5.2057,
+      "step": 311000
+    },
+    {
+      "epoch": 1.9937914039747815,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00038404966876820176,
+      "loss": 5.212,
+      "step": 311500
+    },
+    {
+      "epoch": 1.9969917112042757,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0003840240663103658,
+      "loss": 5.2029,
+      "step": 312000
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 5.198747158050537,
+      "eval_runtime": 1.6976,
+      "eval_samples_per_second": 589.07,
+      "eval_steps_per_second": 9.425,
+      "step": 312470
+    },
+    {
+      "epoch": 2.00019201843377,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0003839984638525298,
+      "loss": 5.2133,
+      "step": 312500
+    },
+    {
+      "epoch": 2.0033923256632638,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003839728613946939,
+      "loss": 5.2076,
+      "step": 313000
+    },
+    {
+      "epoch": 2.0065926328927577,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00038394725893685795,
+      "loss": 5.202,
+      "step": 313500
+    },
+    {
+      "epoch": 2.0097929401222516,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000383921656479022,
+      "loss": 5.203,
+      "step": 314000
+    },
+    {
+      "epoch": 2.0129932473517456,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00038389605402118607,
+      "loss": 5.2092,
+      "step": 314500
+    },
+    {
+      "epoch": 2.01619355458124,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003838704515633501,
+      "loss": 5.1966,
+      "step": 315000
+    },
+    {
+      "epoch": 2.019393861810734,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00038384484910551414,
+      "loss": 5.2033,
+      "step": 315500
+    },
+    {
+      "epoch": 2.022594169040228,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0003838192466476782,
+      "loss": 5.2001,
+      "step": 316000
+    },
+    {
+      "epoch": 2.0257944762697218,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.00038379364418984226,
+      "loss": 5.2038,
+      "step": 316500
+    },
+    {
+      "epoch": 2.028994783499216,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0003837680417320063,
+      "loss": 5.2041,
+      "step": 317000
+    },
+    {
+      "epoch": 2.03219509072871,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00038374243927417033,
+      "loss": 5.199,
+      "step": 317500
+    },
+    {
+      "epoch": 2.035395397958204,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00038371683681633436,
+      "loss": 5.2118,
+      "step": 318000
+    },
+    {
+      "epoch": 2.038595705187698,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00038369123435849845,
+      "loss": 5.1993,
+      "step": 318500
+    },
+    {
+      "epoch": 2.041796012417192,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0003836656319006625,
+      "loss": 5.2074,
+      "step": 319000
+    },
+    {
+      "epoch": 2.0449963196466863,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003836400294428265,
+      "loss": 5.2091,
+      "step": 319500
+    },
+    {
+      "epoch": 2.04819662687618,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038361442698499055,
+      "loss": 5.2067,
+      "step": 320000
+    },
+    {
+      "epoch": 2.051396934105674,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003835888245271546,
+      "loss": 5.2012,
+      "step": 320500
+    },
+    {
+      "epoch": 2.054597241335168,
+      "grad_norm": 0.349609375,
+      "learning_rate": 0.0003835632220693187,
+      "loss": 5.2013,
+      "step": 321000
+    },
+    {
+      "epoch": 2.057797548564662,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0003835376196114827,
+      "loss": 5.2057,
+      "step": 321500
+    },
+    {
+      "epoch": 2.0609978557941564,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0003835120171536468,
+      "loss": 5.2053,
+      "step": 322000
+    },
+    {
+      "epoch": 2.0641981630236503,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00038348641469581083,
+      "loss": 5.2096,
+      "step": 322500
+    },
+    {
+      "epoch": 2.0673984702531443,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.00038346081223797487,
+      "loss": 5.2038,
+      "step": 323000
+    },
+    {
+      "epoch": 2.070598777482638,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003834352097801389,
+      "loss": 5.2057,
+      "step": 323500
+    },
+    {
+      "epoch": 2.0737990847121326,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.000383409607322303,
+      "loss": 5.2013,
+      "step": 324000
+    },
+    {
+      "epoch": 2.0769993919416265,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.000383384004864467,
+      "loss": 5.2081,
+      "step": 324500
+    },
+    {
+      "epoch": 2.0801996991711205,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038335840240663106,
+      "loss": 5.2038,
+      "step": 325000
+    },
+    {
+      "epoch": 2.0834000064006144,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003833327999487951,
+      "loss": 5.2032,
+      "step": 325500
+    },
+    {
+      "epoch": 2.0866003136301083,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00038330719749095913,
+      "loss": 5.2018,
+      "step": 326000
+    },
+    {
+      "epoch": 2.0898006208596027,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0003832815950331232,
+      "loss": 5.2015,
+      "step": 326500
+    },
+    {
+      "epoch": 2.0930009280890967,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00038325599257528725,
+      "loss": 5.1985,
+      "step": 327000
+    },
+    {
+      "epoch": 2.0962012353185906,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0003832303901174513,
+      "loss": 5.2005,
+      "step": 327500
+    },
+    {
+      "epoch": 2.0994015425480845,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003832047876596153,
+      "loss": 5.2005,
+      "step": 328000
+    },
+    {
+      "epoch": 2.1026018497775785,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00038317918520177935,
+      "loss": 5.2106,
+      "step": 328500
+    },
+    {
+      "epoch": 2.105802157007073,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00038315358274394344,
+      "loss": 5.2059,
+      "step": 329000
+    },
+    {
+      "epoch": 2.109002464236567,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0003831279802861075,
+      "loss": 5.2012,
+      "step": 329500
+    },
+    {
+      "epoch": 2.1122027714660607,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00038310237782827156,
+      "loss": 5.1964,
+      "step": 330000
+    },
+    {
+      "epoch": 2.1154030786955547,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.0003830767753704356,
+      "loss": 5.2087,
+      "step": 330500
+    },
+    {
+      "epoch": 2.1186033859250486,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00038305117291259963,
+      "loss": 5.203,
+      "step": 331000
+    },
+    {
+      "epoch": 2.121803693154543,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00038302557045476367,
+      "loss": 5.2044,
+      "step": 331500
+    },
+    {
+      "epoch": 2.125004000384037,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00038299996799692776,
+      "loss": 5.2065,
+      "step": 332000
+    },
+    {
+      "epoch": 2.128204307613531,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0003829743655390918,
+      "loss": 5.2018,
+      "step": 332500
+    },
+    {
+      "epoch": 2.131404614843025,
+      "grad_norm": 0.353515625,
+      "learning_rate": 0.0003829487630812558,
+      "loss": 5.205,
+      "step": 333000
+    },
+    {
+      "epoch": 2.134604922072519,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038292316062341986,
+      "loss": 5.2031,
+      "step": 333500
+    },
+    {
+      "epoch": 2.137805229302013,
+      "grad_norm": 0.33984375,
+      "learning_rate": 0.0003828975581655839,
+      "loss": 5.2081,
+      "step": 334000
+    },
+    {
+      "epoch": 2.141005536531507,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0003828719557077479,
+      "loss": 5.2036,
+      "step": 334500
+    },
+    {
+      "epoch": 2.144205843761001,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.000382846353249912,
+      "loss": 5.2067,
+      "step": 335000
+    },
+    {
+      "epoch": 2.147406150990495,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00038282075079207605,
+      "loss": 5.2016,
+      "step": 335500
+    },
+    {
+      "epoch": 2.1506064582199893,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0003827951483342401,
+      "loss": 5.2033,
+      "step": 336000
+    },
+    {
+      "epoch": 2.1538067654494832,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00038276954587640417,
+      "loss": 5.2013,
+      "step": 336500
+    },
+    {
+      "epoch": 2.157007072678977,
+      "grad_norm": 0.361328125,
+      "learning_rate": 0.0003827439434185682,
+      "loss": 5.1996,
+      "step": 337000
+    },
+    {
+      "epoch": 2.160207379908471,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.0003827183409607323,
+      "loss": 5.2038,
+      "step": 337500
+    },
+    {
+      "epoch": 2.163407687137965,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038269273850289633,
+      "loss": 5.2004,
+      "step": 338000
+    },
+    {
+      "epoch": 2.1666079943674594,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00038266713604506036,
+      "loss": 5.202,
+      "step": 338500
+    },
+    {
+      "epoch": 2.1698083015969534,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0003826415335872244,
+      "loss": 5.1946,
+      "step": 339000
+    },
+    {
+      "epoch": 2.1730086088264473,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00038261593112938843,
+      "loss": 5.205,
+      "step": 339500
+    },
+    {
+      "epoch": 2.1762089160559412,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00038259032867155247,
+      "loss": 5.2002,
+      "step": 340000
+    },
+    {
+      "epoch": 2.179409223285435,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038256472621371655,
+      "loss": 5.2024,
+      "step": 340500
+    },
+    {
+      "epoch": 2.1826095305149296,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003825391237558806,
+      "loss": 5.2081,
+      "step": 341000
+    },
+    {
+      "epoch": 2.1858098377444235,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003825135212980446,
+      "loss": 5.2062,
+      "step": 341500
+    },
+    {
+      "epoch": 2.1890101449739174,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00038248791884020866,
+      "loss": 5.202,
+      "step": 342000
+    },
+    {
+      "epoch": 2.1922104522034114,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003824623163823727,
+      "loss": 5.2013,
+      "step": 342500
+    },
+    {
+      "epoch": 2.1954107594329058,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003824367139245368,
+      "loss": 5.199,
+      "step": 343000
+    },
+    {
+      "epoch": 2.1986110666623997,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0003824111114667008,
+      "loss": 5.2029,
+      "step": 343500
+    },
+    {
+      "epoch": 2.2018113738918936,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038238550900886485,
+      "loss": 5.208,
+      "step": 344000
+    },
+    {
+      "epoch": 2.2050116811213876,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00038235990655102894,
+      "loss": 5.2012,
+      "step": 344500
+    },
+    {
+      "epoch": 2.2082119883508815,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00038233430409319297,
+      "loss": 5.1991,
+      "step": 345000
+    },
+    {
+      "epoch": 2.211412295580376,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.000382308701635357,
+      "loss": 5.2054,
+      "step": 345500
+    },
+    {
+      "epoch": 2.21461260280987,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003822830991775211,
+      "loss": 5.2062,
+      "step": 346000
+    },
+    {
+      "epoch": 2.2178129100393638,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.00038225749671968513,
+      "loss": 5.202,
+      "step": 346500
+    },
+    {
+      "epoch": 2.2210132172688577,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00038223189426184916,
+      "loss": 5.1993,
+      "step": 347000
+    },
+    {
+      "epoch": 2.2242135244983516,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0003822062918040132,
+      "loss": 5.1995,
+      "step": 347500
+    },
+    {
+      "epoch": 2.227413831727846,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00038218068934617723,
+      "loss": 5.1964,
+      "step": 348000
+    },
+    {
+      "epoch": 2.23061413895734,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003821550868883413,
+      "loss": 5.2061,
+      "step": 348500
+    },
+    {
+      "epoch": 2.233814446186834,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00038212948443050535,
+      "loss": 5.2081,
+      "step": 349000
+    },
+    {
+      "epoch": 2.237014753416328,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003821038819726694,
+      "loss": 5.1967,
+      "step": 349500
+    },
+    {
+      "epoch": 2.2402150606458218,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0003820782795148334,
+      "loss": 5.2061,
+      "step": 350000
+    },
+    {
+      "epoch": 2.243415367875316,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00038205267705699746,
+      "loss": 5.2018,
+      "step": 350500
+    },
+    {
+      "epoch": 2.24661567510481,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0003820270745991615,
+      "loss": 5.2016,
+      "step": 351000
+    },
+    {
+      "epoch": 2.249815982334304,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003820014721413256,
+      "loss": 5.2083,
+      "step": 351500
+    },
+    {
+      "epoch": 2.253016289563798,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003819758696834896,
+      "loss": 5.2076,
+      "step": 352000
+    },
+    {
+      "epoch": 2.2562165967932923,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003819502672256537,
+      "loss": 5.2068,
+      "step": 352500
+    },
+    {
+      "epoch": 2.2594169040227863,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038192466476781774,
+      "loss": 5.2036,
+      "step": 353000
+    },
+    {
+      "epoch": 2.26261721125228,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00038189906230998177,
+      "loss": 5.198,
+      "step": 353500
+    },
+    {
+      "epoch": 2.265817518481774,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038187345985214586,
+      "loss": 5.2034,
+      "step": 354000
+    },
+    {
+      "epoch": 2.269017825711268,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003818478573943099,
+      "loss": 5.2037,
+      "step": 354500
+    },
+    {
+      "epoch": 2.2722181329407625,
+      "grad_norm": 0.357421875,
+      "learning_rate": 0.0003818222549364739,
+      "loss": 5.2039,
+      "step": 355000
+    },
+    {
+      "epoch": 2.2754184401702564,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00038179665247863796,
+      "loss": 5.2,
+      "step": 355500
+    },
+    {
+      "epoch": 2.2786187473997503,
+      "grad_norm": 0.369140625,
+      "learning_rate": 0.000381771050020802,
+      "loss": 5.2056,
+      "step": 356000
+    },
+    {
+      "epoch": 2.2818190546292443,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.00038174544756296603,
+      "loss": 5.2023,
+      "step": 356500
+    },
+    {
+      "epoch": 2.2850193618587387,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0003817198451051301,
+      "loss": 5.1957,
+      "step": 357000
+    },
+    {
+      "epoch": 2.2882196690882326,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.00038169424264729415,
+      "loss": 5.2005,
+      "step": 357500
+    },
+    {
+      "epoch": 2.2914199763177265,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003816686401894582,
+      "loss": 5.2037,
+      "step": 358000
+    },
+    {
+      "epoch": 2.2946202835472205,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003816430377316222,
+      "loss": 5.2036,
+      "step": 358500
+    },
+    {
+      "epoch": 2.2978205907767144,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003816174352737863,
+      "loss": 5.2034,
+      "step": 359000
+    },
+    {
+      "epoch": 2.3010208980062083,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00038159183281595034,
+      "loss": 5.1988,
+      "step": 359500
+    },
+    {
+      "epoch": 2.3042212052357027,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00038156623035811443,
+      "loss": 5.2031,
+      "step": 360000
+    },
+    {
+      "epoch": 2.3074215124651967,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00038154062790027847,
+      "loss": 5.1989,
+      "step": 360500
+    },
+    {
+      "epoch": 2.3106218196946906,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003815150254424425,
+      "loss": 5.2058,
+      "step": 361000
+    },
+    {
+      "epoch": 2.3138221269241845,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.00038148942298460653,
+      "loss": 5.2032,
+      "step": 361500
+    },
+    {
+      "epoch": 2.317022434153679,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003814638205267706,
+      "loss": 5.1991,
+      "step": 362000
+    },
+    {
+      "epoch": 2.320222741383173,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.00038143821806893466,
+      "loss": 5.2017,
+      "step": 362500
+    },
+    {
+      "epoch": 2.323423048612667,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003814126156110987,
+      "loss": 5.2021,
+      "step": 363000
+    },
+    {
+      "epoch": 2.3266233558421607,
+      "grad_norm": 0.34765625,
+      "learning_rate": 0.0003813870131532627,
+      "loss": 5.2023,
+      "step": 363500
+    },
+    {
+      "epoch": 2.3298236630716547,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00038136141069542676,
+      "loss": 5.2085,
+      "step": 364000
+    },
+    {
+      "epoch": 2.333023970301149,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003813358082375908,
+      "loss": 5.2003,
+      "step": 364500
+    },
+    {
+      "epoch": 2.336224277530643,
+      "grad_norm": 0.375,
+      "learning_rate": 0.0003813102057797549,
+      "loss": 5.2016,
+      "step": 365000
+    },
+    {
+      "epoch": 2.339424584760137,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003812846033219189,
+      "loss": 5.2058,
+      "step": 365500
+    },
+    {
+      "epoch": 2.342624891989631,
+      "grad_norm": 0.359375,
+      "learning_rate": 0.00038125900086408295,
+      "loss": 5.2025,
+      "step": 366000
+    },
+    {
+      "epoch": 2.3458251992191252,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.000381233398406247,
+      "loss": 5.1953,
+      "step": 366500
+    },
+    {
+      "epoch": 2.349025506448619,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003812077959484111,
+      "loss": 5.2013,
+      "step": 367000
+    },
+    {
+      "epoch": 2.352225813678113,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003811821934905751,
+      "loss": 5.2023,
+      "step": 367500
+    },
+    {
+      "epoch": 2.355426120907607,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003811565910327392,
+      "loss": 5.2069,
+      "step": 368000
+    },
+    {
+      "epoch": 2.358626428137101,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00038113098857490323,
+      "loss": 5.2004,
+      "step": 368500
+    },
+    {
+      "epoch": 2.3618267353665954,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00038110538611706727,
+      "loss": 5.2035,
+      "step": 369000
+    },
+    {
+      "epoch": 2.3650270425960893,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003810797836592313,
+      "loss": 5.2042,
+      "step": 369500
+    },
+    {
+      "epoch": 2.3682273498255833,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00038105418120139533,
+      "loss": 5.1985,
+      "step": 370000
+    },
+    {
+      "epoch": 2.371427657055077,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003810285787435594,
+      "loss": 5.2041,
+      "step": 370500
+    },
+    {
+      "epoch": 2.3746279642845716,
+      "grad_norm": 0.35546875,
+      "learning_rate": 0.00038100297628572346,
+      "loss": 5.1956,
+      "step": 371000
+    },
+    {
+      "epoch": 2.3778282715140655,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0003809773738278875,
+      "loss": 5.2046,
+      "step": 371500
+    },
+    {
+      "epoch": 2.3810285787435594,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003809517713700515,
+      "loss": 5.1996,
+      "step": 372000
+    },
+    {
+      "epoch": 2.3842288859730534,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00038092616891221556,
+      "loss": 5.2018,
+      "step": 372500
+    },
+    {
+      "epoch": 2.3874291932025473,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.00038090056645437965,
+      "loss": 5.2014,
+      "step": 373000
+    },
+    {
+      "epoch": 2.3906295004320413,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003808749639965437,
+      "loss": 5.1966,
+      "step": 373500
+    },
+    {
+      "epoch": 2.3938298076615356,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003808493615387077,
+      "loss": 5.2013,
+      "step": 374000
+    },
+    {
+      "epoch": 2.3970301148910296,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003808237590808718,
+      "loss": 5.1956,
+      "step": 374500
+    },
+    {
+      "epoch": 2.4002304221205235,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00038079815662303584,
+      "loss": 5.1954,
+      "step": 375000
+    },
+    {
+      "epoch": 2.4034307293500174,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003807725541651999,
+      "loss": 5.1928,
+      "step": 375500
+    },
+    {
+      "epoch": 2.406631036579512,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00038074695170736396,
+      "loss": 5.2002,
+      "step": 376000
+    },
+    {
+      "epoch": 2.4098313438090058,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.000380721349249528,
+      "loss": 5.1998,
+      "step": 376500
+    },
+    {
+      "epoch": 2.4130316510384997,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00038069574679169203,
+      "loss": 5.1989,
+      "step": 377000
+    },
+    {
+      "epoch": 2.4162319582679936,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00038067014433385606,
+      "loss": 5.205,
+      "step": 377500
+    },
+    {
+      "epoch": 2.4194322654974876,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003806445418760201,
+      "loss": 5.2056,
+      "step": 378000
+    },
+    {
+      "epoch": 2.422632572726982,
+      "grad_norm": 0.376953125,
+      "learning_rate": 0.0003806189394181842,
+      "loss": 5.2047,
+      "step": 378500
+    },
+    {
+      "epoch": 2.425832879956476,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003805933369603482,
+      "loss": 5.1933,
+      "step": 379000
+    },
+    {
+      "epoch": 2.42903318718597,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00038056773450251226,
+      "loss": 5.1982,
+      "step": 379500
+    },
+    {
+      "epoch": 2.4322334944154638,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003805421320446763,
+      "loss": 5.2007,
+      "step": 380000
+    },
+    {
+      "epoch": 2.435433801644958,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003805165295868403,
+      "loss": 5.2001,
+      "step": 380500
+    },
+    {
+      "epoch": 2.438634108874452,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00038049092712900436,
+      "loss": 5.2003,
+      "step": 381000
+    },
+    {
+      "epoch": 2.441834416103946,
+      "grad_norm": 0.3515625,
+      "learning_rate": 0.00038046532467116845,
+      "loss": 5.203,
+      "step": 381500
+    },
+    {
+      "epoch": 2.44503472333344,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003804397222133325,
+      "loss": 5.1956,
+      "step": 382000
+    },
+    {
+      "epoch": 2.448235030562934,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00038041411975549657,
+      "loss": 5.1983,
+      "step": 382500
+    },
+    {
+      "epoch": 2.451435337792428,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003803885172976606,
+      "loss": 5.199,
+      "step": 383000
+    },
+    {
+      "epoch": 2.454635645021922,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00038036291483982464,
+      "loss": 5.203,
+      "step": 383500
+    },
+    {
+      "epoch": 2.457835952251416,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003803373123819887,
+      "loss": 5.2004,
+      "step": 384000
+    },
+    {
+      "epoch": 2.46103625948091,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00038031170992415276,
+      "loss": 5.2037,
+      "step": 384500
+    },
+    {
+      "epoch": 2.464236566710404,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003802861074663168,
+      "loss": 5.1961,
+      "step": 385000
+    },
+    {
+      "epoch": 2.4674368739398984,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00038026050500848083,
+      "loss": 5.2003,
+      "step": 385500
+    },
+    {
+      "epoch": 2.4706371811693923,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00038023490255064486,
+      "loss": 5.2007,
+      "step": 386000
+    },
+    {
+      "epoch": 2.4738374883988863,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003802093000928089,
+      "loss": 5.1987,
+      "step": 386500
+    },
+    {
+      "epoch": 2.4770377956283802,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.000380183697634973,
+      "loss": 5.1996,
+      "step": 387000
+    },
+    {
+      "epoch": 2.480238102857874,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.000380158095177137,
+      "loss": 5.1935,
+      "step": 387500
+    },
+    {
+      "epoch": 2.4834384100873685,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00038013249271930105,
+      "loss": 5.1977,
+      "step": 388000
+    },
+    {
+      "epoch": 2.4866387173168625,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003801068902614651,
+      "loss": 5.2041,
+      "step": 388500
+    },
+    {
+      "epoch": 2.4898390245463564,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003800812878036291,
+      "loss": 5.2037,
+      "step": 389000
+    },
+    {
+      "epoch": 2.4930393317758504,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003800556853457932,
+      "loss": 5.2075,
+      "step": 389500
+    },
+    {
+      "epoch": 2.4962396390053447,
+      "grad_norm": 0.36328125,
+      "learning_rate": 0.0003800300828879573,
+      "loss": 5.2056,
+      "step": 390000
+    },
+    {
+      "epoch": 2.4994399462348387,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00038000448043012133,
+      "loss": 5.1986,
+      "step": 390500
+    },
+    {
+      "epoch": 2.5026402534643326,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037997887797228537,
+      "loss": 5.2022,
+      "step": 391000
+    },
+    {
+      "epoch": 2.5058405606938265,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0003799532755144494,
+      "loss": 5.2009,
+      "step": 391500
+    },
+    {
+      "epoch": 2.5090408679233205,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037992767305661344,
+      "loss": 5.197,
+      "step": 392000
+    },
+    {
+      "epoch": 2.5122411751528144,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003799020705987775,
+      "loss": 5.2044,
+      "step": 392500
+    },
+    {
+      "epoch": 2.515441482382309,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00037987646814094156,
+      "loss": 5.2045,
+      "step": 393000
+    },
+    {
+      "epoch": 2.5186417896118027,
+      "grad_norm": 0.37109375,
+      "learning_rate": 0.0003798508656831056,
+      "loss": 5.2081,
+      "step": 393500
+    },
+    {
+      "epoch": 2.5218420968412967,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00037982526322526963,
+      "loss": 5.2002,
+      "step": 394000
+    },
+    {
+      "epoch": 2.525042404070791,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037979966076743366,
+      "loss": 5.1958,
+      "step": 394500
+    },
+    {
+      "epoch": 2.528242711300285,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.00037977405830959775,
+      "loss": 5.1946,
+      "step": 395000
+    },
+    {
+      "epoch": 2.531443018529779,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003797484558517618,
+      "loss": 5.2012,
+      "step": 395500
+    },
+    {
+      "epoch": 2.534643325759273,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003797228533939258,
+      "loss": 5.1959,
+      "step": 396000
+    },
+    {
+      "epoch": 2.537843632988767,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00037969725093608985,
+      "loss": 5.2028,
+      "step": 396500
+    },
+    {
+      "epoch": 2.5410439402182607,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00037967164847825394,
+      "loss": 5.1976,
+      "step": 397000
+    },
+    {
+      "epoch": 2.544244247447755,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.000379646046020418,
+      "loss": 5.1984,
+      "step": 397500
+    },
+    {
+      "epoch": 2.547444554677249,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00037962044356258206,
+      "loss": 5.199,
+      "step": 398000
+    },
+    {
+      "epoch": 2.550644861906743,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003795948411047461,
+      "loss": 5.1985,
+      "step": 398500
+    },
+    {
+      "epoch": 2.553845169136237,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00037956923864691013,
+      "loss": 5.2035,
+      "step": 399000
+    },
+    {
+      "epoch": 2.5570454763657313,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00037954363618907417,
+      "loss": 5.1987,
+      "step": 399500
+    },
+    {
+      "epoch": 2.5602457835952253,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003795180337312382,
+      "loss": 5.2027,
+      "step": 400000
+    },
+    {
+      "epoch": 2.563446090824719,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003794924312734023,
+      "loss": 5.2047,
+      "step": 400500
+    },
+    {
+      "epoch": 2.566646398054213,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003794668288155663,
+      "loss": 5.2031,
+      "step": 401000
+    },
+    {
+      "epoch": 2.569846705283707,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00037944122635773036,
+      "loss": 5.2054,
+      "step": 401500
+    },
+    {
+      "epoch": 2.573047012513201,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003794156238998944,
+      "loss": 5.203,
+      "step": 402000
+    },
+    {
+      "epoch": 2.5762473197426954,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003793900214420584,
+      "loss": 5.205,
+      "step": 402500
+    },
+    {
+      "epoch": 2.5794476269721893,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003793644189842225,
+      "loss": 5.1995,
+      "step": 403000
+    },
+    {
+      "epoch": 2.5826479342016833,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00037933881652638655,
+      "loss": 5.2005,
+      "step": 403500
+    },
+    {
+      "epoch": 2.5858482414311776,
+      "grad_norm": 0.365234375,
+      "learning_rate": 0.0003793132140685506,
+      "loss": 5.2015,
+      "step": 404000
+    },
+    {
+      "epoch": 2.5890485486606716,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003792876116107146,
+      "loss": 5.201,
+      "step": 404500
+    },
+    {
+      "epoch": 2.5922488558901655,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003792620091528787,
+      "loss": 5.2021,
+      "step": 405000
+    },
+    {
+      "epoch": 2.5954491631196595,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00037923640669504274,
+      "loss": 5.209,
+      "step": 405500
+    },
+    {
+      "epoch": 2.5986494703491534,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00037921080423720683,
+      "loss": 5.1946,
+      "step": 406000
+    },
+    {
+      "epoch": 2.6018497775786473,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00037918520177937086,
+      "loss": 5.1998,
+      "step": 406500
+    },
+    {
+      "epoch": 2.6050500848081417,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003791595993215349,
+      "loss": 5.1969,
+      "step": 407000
+    },
+    {
+      "epoch": 2.6082503920376356,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037913399686369893,
+      "loss": 5.1967,
+      "step": 407500
+    },
+    {
+      "epoch": 2.6114506992671296,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037910839440586297,
+      "loss": 5.2025,
+      "step": 408000
+    },
+    {
+      "epoch": 2.6146510064966235,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00037908279194802705,
+      "loss": 5.199,
+      "step": 408500
+    },
+    {
+      "epoch": 2.617851313726118,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003790571894901911,
+      "loss": 5.2005,
+      "step": 409000
+    },
+    {
+      "epoch": 2.621051620955612,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003790315870323551,
+      "loss": 5.1969,
+      "step": 409500
+    },
+    {
+      "epoch": 2.6242519281851058,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.00037900598457451916,
+      "loss": 5.205,
+      "step": 410000
+    },
+    {
+      "epoch": 2.6274522354145997,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0003789803821166832,
+      "loss": 5.2016,
+      "step": 410500
+    },
+    {
+      "epoch": 2.6306525426440936,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003789547796588472,
+      "loss": 5.2007,
+      "step": 411000
+    },
+    {
+      "epoch": 2.6338528498735876,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003789291772010113,
+      "loss": 5.2053,
+      "step": 411500
+    },
+    {
+      "epoch": 2.637053157103082,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.00037890357474317535,
+      "loss": 5.1981,
+      "step": 412000
+    },
+    {
+      "epoch": 2.640253464332576,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00037887797228533944,
+      "loss": 5.2052,
+      "step": 412500
+    },
+    {
+      "epoch": 2.64345377156207,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00037885236982750347,
+      "loss": 5.1987,
+      "step": 413000
+    },
+    {
+      "epoch": 2.6466540787915642,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003788267673696675,
+      "loss": 5.1951,
+      "step": 413500
+    },
+    {
+      "epoch": 2.649854386021058,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003788011649118316,
+      "loss": 5.2001,
+      "step": 414000
+    },
+    {
+      "epoch": 2.653054693250552,
+      "grad_norm": 0.375,
+      "learning_rate": 0.00037877556245399563,
+      "loss": 5.2009,
+      "step": 414500
+    },
+    {
+      "epoch": 2.656255000480046,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037874995999615966,
+      "loss": 5.199,
+      "step": 415000
+    },
+    {
+      "epoch": 2.65945530770954,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003787243575383237,
+      "loss": 5.2029,
+      "step": 415500
+    },
+    {
+      "epoch": 2.662655614939034,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00037869875508048773,
+      "loss": 5.1956,
+      "step": 416000
+    },
+    {
+      "epoch": 2.6658559221685283,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00037867315262265177,
+      "loss": 5.2,
+      "step": 416500
+    },
+    {
+      "epoch": 2.6690562293980222,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00037864755016481585,
+      "loss": 5.1974,
+      "step": 417000
+    },
+    {
+      "epoch": 2.672256536627516,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003786219477069799,
+      "loss": 5.1977,
+      "step": 417500
+    },
+    {
+      "epoch": 2.6754568438570105,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003785963452491439,
+      "loss": 5.1954,
+      "step": 418000
+    },
+    {
+      "epoch": 2.6786571510865045,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00037857074279130796,
+      "loss": 5.2027,
+      "step": 418500
+    },
+    {
+      "epoch": 2.6818574583159984,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000378545140333472,
+      "loss": 5.1943,
+      "step": 419000
+    },
+    {
+      "epoch": 2.6850577655454924,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003785195378756361,
+      "loss": 5.1955,
+      "step": 419500
+    },
+    {
+      "epoch": 2.6882580727749863,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003784939354178001,
+      "loss": 5.206,
+      "step": 420000
+    },
+    {
+      "epoch": 2.6914583800044802,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.0003784683329599642,
+      "loss": 5.2077,
+      "step": 420500
+    },
+    {
+      "epoch": 2.6946586872339746,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.00037844273050212824,
+      "loss": 5.2,
+      "step": 421000
+    },
+    {
+      "epoch": 2.6978589944634686,
+      "grad_norm": 0.384765625,
+      "learning_rate": 0.00037841712804429227,
+      "loss": 5.1981,
+      "step": 421500
+    },
+    {
+      "epoch": 2.7010593016929625,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003783915255864563,
+      "loss": 5.2035,
+      "step": 422000
+    },
+    {
+      "epoch": 2.7042596089224564,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003783659231286204,
+      "loss": 5.2,
+      "step": 422500
+    },
+    {
+      "epoch": 2.707459916151951,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003783403206707844,
+      "loss": 5.2004,
+      "step": 423000
+    },
+    {
+      "epoch": 2.7106602233814447,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037831471821294846,
+      "loss": 5.1948,
+      "step": 423500
+    },
+    {
+      "epoch": 2.7138605306109387,
+      "grad_norm": 0.373046875,
+      "learning_rate": 0.0003782891157551125,
+      "loss": 5.1975,
+      "step": 424000
+    },
+    {
+      "epoch": 2.7170608378404326,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037826351329727653,
+      "loss": 5.2018,
+      "step": 424500
+    },
+    {
+      "epoch": 2.7202611450699266,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003782379108394406,
+      "loss": 5.2005,
+      "step": 425000
+    },
+    {
+      "epoch": 2.7234614522994205,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00037821230838160465,
+      "loss": 5.2011,
+      "step": 425500
+    },
+    {
+      "epoch": 2.726661759528915,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003781867059237687,
+      "loss": 5.2041,
+      "step": 426000
+    },
+    {
+      "epoch": 2.729862066758409,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003781611034659327,
+      "loss": 5.1967,
+      "step": 426500
+    },
+    {
+      "epoch": 2.7330623739879027,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.00037813550100809676,
+      "loss": 5.2039,
+      "step": 427000
+    },
+    {
+      "epoch": 2.736262681217397,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037810989855026084,
+      "loss": 5.1936,
+      "step": 427500
+    },
+    {
+      "epoch": 2.739462988446891,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037808429609242493,
+      "loss": 5.1995,
+      "step": 428000
+    },
+    {
+      "epoch": 2.742663295676385,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037805869363458897,
+      "loss": 5.2016,
+      "step": 428500
+    },
+    {
+      "epoch": 2.745863602905879,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.000378033091176753,
+      "loss": 5.1997,
+      "step": 429000
+    },
+    {
+      "epoch": 2.749063910135373,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037800748871891703,
+      "loss": 5.2064,
+      "step": 429500
+    },
+    {
+      "epoch": 2.752264217364867,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037798188626108107,
+      "loss": 5.1999,
+      "step": 430000
+    },
+    {
+      "epoch": 2.755464524594361,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00037795628380324516,
+      "loss": 5.1978,
+      "step": 430500
+    },
+    {
+      "epoch": 2.758664831823855,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003779306813454092,
+      "loss": 5.1986,
+      "step": 431000
+    },
+    {
+      "epoch": 2.761865139053349,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.0003779050788875732,
+      "loss": 5.202,
+      "step": 431500
+    },
+    {
+      "epoch": 2.765065446282843,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00037787947642973726,
+      "loss": 5.1967,
+      "step": 432000
+    },
+    {
+      "epoch": 2.7682657535123374,
+      "grad_norm": 0.392578125,
+      "learning_rate": 0.0003778538739719013,
+      "loss": 5.2028,
+      "step": 432500
+    },
+    {
+      "epoch": 2.7714660607418313,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003778282715140654,
+      "loss": 5.1978,
+      "step": 433000
+    },
+    {
+      "epoch": 2.7746663679713253,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003778026690562294,
+      "loss": 5.197,
+      "step": 433500
+    },
+    {
+      "epoch": 2.777866675200819,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00037777706659839345,
+      "loss": 5.1927,
+      "step": 434000
+    },
+    {
+      "epoch": 2.781066982430313,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003777514641405575,
+      "loss": 5.1988,
+      "step": 434500
+    },
+    {
+      "epoch": 2.784267289659807,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003777258616827216,
+      "loss": 5.1968,
+      "step": 435000
+    },
+    {
+      "epoch": 2.7874675968893015,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003777002592248856,
+      "loss": 5.1956,
+      "step": 435500
+    },
+    {
+      "epoch": 2.7906679041187954,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003776746567670497,
+      "loss": 5.1986,
+      "step": 436000
+    },
+    {
+      "epoch": 2.7938682113482893,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00037764905430921373,
+      "loss": 5.1952,
+      "step": 436500
+    },
+    {
+      "epoch": 2.7970685185777837,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.00037762345185137777,
+      "loss": 5.2054,
+      "step": 437000
+    },
+    {
+      "epoch": 2.8002688258072777,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003775978493935418,
+      "loss": 5.1935,
+      "step": 437500
+    },
+    {
+      "epoch": 2.8034691330367716,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00037757224693570583,
+      "loss": 5.1976,
+      "step": 438000
+    },
+    {
+      "epoch": 2.8066694402662655,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003775466444778699,
+      "loss": 5.2001,
+      "step": 438500
+    },
+    {
+      "epoch": 2.8098697474957595,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00037752104202003396,
+      "loss": 5.1961,
+      "step": 439000
+    },
+    {
+      "epoch": 2.8130700547252534,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.000377495439562198,
+      "loss": 5.2053,
+      "step": 439500
+    },
+    {
+      "epoch": 2.816270361954748,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.000377469837104362,
+      "loss": 5.2005,
+      "step": 440000
+    },
+    {
+      "epoch": 2.8194706691842417,
+      "grad_norm": 0.390625,
+      "learning_rate": 0.00037744423464652606,
+      "loss": 5.2043,
+      "step": 440500
+    },
+    {
+      "epoch": 2.8226709764137357,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003774186321886901,
+      "loss": 5.1988,
+      "step": 441000
+    },
+    {
+      "epoch": 2.8258712836432296,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003773930297308542,
+      "loss": 5.2003,
+      "step": 441500
+    },
+    {
+      "epoch": 2.829071590872724,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003773674272730182,
+      "loss": 5.1983,
+      "step": 442000
+    },
+    {
+      "epoch": 2.832271898102218,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00037734182481518225,
+      "loss": 5.1978,
+      "step": 442500
+    },
+    {
+      "epoch": 2.835472205331712,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037731622235734634,
+      "loss": 5.1958,
+      "step": 443000
+    },
+    {
+      "epoch": 2.838672512561206,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003772906198995104,
+      "loss": 5.2009,
+      "step": 443500
+    },
+    {
+      "epoch": 2.8418728197906997,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.00037726501744167446,
+      "loss": 5.1955,
+      "step": 444000
+    },
+    {
+      "epoch": 2.845073127020194,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003772394149838385,
+      "loss": 5.2014,
+      "step": 444500
+    },
+    {
+      "epoch": 2.848273434249688,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037721381252600253,
+      "loss": 5.2015,
+      "step": 445000
+    },
+    {
+      "epoch": 2.851473741479182,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037718821006816656,
+      "loss": 5.1953,
+      "step": 445500
+    },
+    {
+      "epoch": 2.854674048708676,
+      "grad_norm": 0.3828125,
+      "learning_rate": 0.0003771626076103306,
+      "loss": 5.1958,
+      "step": 446000
+    },
+    {
+      "epoch": 2.8578743559381703,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037713700515249463,
+      "loss": 5.2021,
+      "step": 446500
+    },
+    {
+      "epoch": 2.8610746631676642,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.0003771114026946587,
+      "loss": 5.1982,
+      "step": 447000
+    },
+    {
+      "epoch": 2.864274970397158,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00037708580023682276,
+      "loss": 5.1954,
+      "step": 447500
+    },
+    {
+      "epoch": 2.867475277626652,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003770601977789868,
+      "loss": 5.1985,
+      "step": 448000
+    },
+    {
+      "epoch": 2.870675584856146,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003770345953211508,
+      "loss": 5.2033,
+      "step": 448500
+    },
+    {
+      "epoch": 2.87387589208564,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037700899286331486,
+      "loss": 5.1984,
+      "step": 449000
+    },
+    {
+      "epoch": 2.8770761993151344,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037698339040547895,
+      "loss": 5.201,
+      "step": 449500
+    },
+    {
+      "epoch": 2.8802765065446283,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000376957787947643,
+      "loss": 5.1991,
+      "step": 450000
+    },
+    {
+      "epoch": 2.8834768137741222,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00037693218548980707,
+      "loss": 5.1997,
+      "step": 450500
+    },
+    {
+      "epoch": 2.8866771210036166,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003769065830319711,
+      "loss": 5.1972,
+      "step": 451000
+    },
+    {
+      "epoch": 2.8898774282331106,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037688098057413514,
+      "loss": 5.2039,
+      "step": 451500
+    },
+    {
+      "epoch": 2.8930777354626045,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00037685537811629917,
+      "loss": 5.1969,
+      "step": 452000
+    },
+    {
+      "epoch": 2.8962780426920984,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00037682977565846326,
+      "loss": 5.1992,
+      "step": 452500
+    },
+    {
+      "epoch": 2.8994783499215924,
+      "grad_norm": 0.388671875,
+      "learning_rate": 0.0003768041732006273,
+      "loss": 5.1965,
+      "step": 453000
+    },
+    {
+      "epoch": 2.9026786571510863,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037677857074279133,
+      "loss": 5.1972,
+      "step": 453500
+    },
+    {
+      "epoch": 2.9058789643805807,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00037675296828495536,
+      "loss": 5.1968,
+      "step": 454000
+    },
+    {
+      "epoch": 2.9090792716100746,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003767273658271194,
+      "loss": 5.1998,
+      "step": 454500
+    },
+    {
+      "epoch": 2.9122795788395686,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003767017633692835,
+      "loss": 5.1934,
+      "step": 455000
+    },
+    {
+      "epoch": 2.9154798860690625,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003766761609114475,
+      "loss": 5.1931,
+      "step": 455500
+    },
+    {
+      "epoch": 2.918680193298557,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037665055845361155,
+      "loss": 5.2004,
+      "step": 456000
+    },
+    {
+      "epoch": 2.921880500528051,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003766249559957756,
+      "loss": 5.2008,
+      "step": 456500
+    },
+    {
+      "epoch": 2.9250808077575448,
+      "grad_norm": 0.396484375,
+      "learning_rate": 0.0003765993535379396,
+      "loss": 5.1973,
+      "step": 457000
+    },
+    {
+      "epoch": 2.9282811149870387,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003765737510801037,
+      "loss": 5.1917,
+      "step": 457500
+    },
+    {
+      "epoch": 2.9314814222165326,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037654814862226775,
+      "loss": 5.1964,
+      "step": 458000
+    },
+    {
+      "epoch": 2.9346817294460266,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037652254616443183,
+      "loss": 5.196,
+      "step": 458500
+    },
+    {
+      "epoch": 2.937882036675521,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037649694370659587,
+      "loss": 5.1973,
+      "step": 459000
+    },
+    {
+      "epoch": 2.941082343905015,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003764713412487599,
+      "loss": 5.203,
+      "step": 459500
+    },
+    {
+      "epoch": 2.944282651134509,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00037644573879092394,
+      "loss": 5.1963,
+      "step": 460000
+    },
+    {
+      "epoch": 2.947482958364003,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000376420136333088,
+      "loss": 5.1969,
+      "step": 460500
+    },
+    {
+      "epoch": 2.950683265593497,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00037639453387525206,
+      "loss": 5.1959,
+      "step": 461000
+    },
+    {
+      "epoch": 2.953883572822991,
+      "grad_norm": 0.39453125,
+      "learning_rate": 0.0003763689314174161,
+      "loss": 5.197,
+      "step": 461500
+    },
+    {
+      "epoch": 2.957083880052485,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037634332895958013,
+      "loss": 5.1998,
+      "step": 462000
+    },
+    {
+      "epoch": 2.960284187281979,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037631772650174416,
+      "loss": 5.2028,
+      "step": 462500
+    },
+    {
+      "epoch": 2.963484494511473,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003762921240439082,
+      "loss": 5.1969,
+      "step": 463000
+    },
+    {
+      "epoch": 2.9666848017409673,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003762665215860723,
+      "loss": 5.202,
+      "step": 463500
+    },
+    {
+      "epoch": 2.969885108970461,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003762409191282363,
+      "loss": 5.1935,
+      "step": 464000
+    },
+    {
+      "epoch": 2.973085416199955,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037621531667040035,
+      "loss": 5.1952,
+      "step": 464500
+    },
+    {
+      "epoch": 2.976285723429449,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003761897142125644,
+      "loss": 5.1999,
+      "step": 465000
+    },
+    {
+      "epoch": 2.9794860306589435,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003761641117547285,
+      "loss": 5.1928,
+      "step": 465500
+    },
+    {
+      "epoch": 2.9826863378884374,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.00037613850929689256,
+      "loss": 5.1958,
+      "step": 466000
+    },
+    {
+      "epoch": 2.9858866451179313,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003761129068390566,
+      "loss": 5.1954,
+      "step": 466500
+    },
+    {
+      "epoch": 2.9890869523474253,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037608730438122063,
+      "loss": 5.1946,
+      "step": 467000
+    },
+    {
+      "epoch": 2.992287259576919,
+      "grad_norm": 0.380859375,
+      "learning_rate": 0.00037606170192338467,
+      "loss": 5.2003,
+      "step": 467500
+    },
+    {
+      "epoch": 2.995487566806413,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003760360994655487,
+      "loss": 5.1928,
+      "step": 468000
+    },
+    {
+      "epoch": 2.9986878740359075,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00037601049700771274,
+      "loss": 5.2001,
+      "step": 468500
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 5.188437461853027,
+      "eval_runtime": 1.1511,
+      "eval_samples_per_second": 868.755,
+      "eval_steps_per_second": 13.9,
+      "step": 468705
+    },
+    {
+      "epoch": 3.0018881812654015,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003759848945498768,
+      "loss": 5.2012,
+      "step": 469000
+    },
+    {
+      "epoch": 3.0050884884948954,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00037595929209204086,
+      "loss": 5.195,
+      "step": 469500
+    },
+    {
+      "epoch": 3.0082887957243893,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003759336896342049,
+      "loss": 5.1946,
+      "step": 470000
+    },
+    {
+      "epoch": 3.0114891029538837,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.0003759080871763689,
+      "loss": 5.1945,
+      "step": 470500
+    },
+    {
+      "epoch": 3.0146894101833777,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00037588248471853296,
+      "loss": 5.1927,
+      "step": 471000
+    },
+    {
+      "epoch": 3.0178897174128716,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037585688226069705,
+      "loss": 5.194,
+      "step": 471500
+    },
+    {
+      "epoch": 3.0210900246423655,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003758312798028611,
+      "loss": 5.187,
+      "step": 472000
+    },
+    {
+      "epoch": 3.02429033187186,
+      "grad_norm": 0.37890625,
+      "learning_rate": 0.0003758056773450251,
+      "loss": 5.195,
+      "step": 472500
+    },
+    {
+      "epoch": 3.027490639101354,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003757800748871892,
+      "loss": 5.1955,
+      "step": 473000
+    },
+    {
+      "epoch": 3.030690946330848,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00037575447242935324,
+      "loss": 5.1973,
+      "step": 473500
+    },
+    {
+      "epoch": 3.0338912535603417,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00037572886997151733,
+      "loss": 5.1982,
+      "step": 474000
+    },
+    {
+      "epoch": 3.0370915607898357,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00037570326751368136,
+      "loss": 5.1939,
+      "step": 474500
+    },
+    {
+      "epoch": 3.04029186801933,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003756776650558454,
+      "loss": 5.202,
+      "step": 475000
+    },
+    {
+      "epoch": 3.043492175248824,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00037565206259800943,
+      "loss": 5.1921,
+      "step": 475500
+    },
+    {
+      "epoch": 3.046692482478318,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00037562646014017347,
+      "loss": 5.1992,
+      "step": 476000
+    },
+    {
+      "epoch": 3.049892789707812,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003756008576823375,
+      "loss": 5.1936,
+      "step": 476500
+    },
+    {
+      "epoch": 3.053093096937306,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003755752552245016,
+      "loss": 5.1884,
+      "step": 477000
+    },
+    {
+      "epoch": 3.0562934041668,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.0003755496527666656,
+      "loss": 5.1873,
+      "step": 477500
+    },
+    {
+      "epoch": 3.059493711396294,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00037552405030882966,
+      "loss": 5.1863,
+      "step": 478000
+    },
+    {
+      "epoch": 3.062694018625788,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003754984478509937,
+      "loss": 5.197,
+      "step": 478500
+    },
+    {
+      "epoch": 3.065894325855282,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003754728453931577,
+      "loss": 5.1983,
+      "step": 479000
+    },
+    {
+      "epoch": 3.069094633084776,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003754472429353218,
+      "loss": 5.1934,
+      "step": 479500
+    },
+    {
+      "epoch": 3.0722949403142703,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00037542164047748585,
+      "loss": 5.1938,
+      "step": 480000
+    },
+    {
+      "epoch": 3.0754952475437642,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003753960380196499,
+      "loss": 5.1926,
+      "step": 480500
+    },
+    {
+      "epoch": 3.078695554773258,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037537043556181397,
+      "loss": 5.1942,
+      "step": 481000
+    },
+    {
+      "epoch": 3.081895862002752,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.000375344833103978,
+      "loss": 5.1923,
+      "step": 481500
+    },
+    {
+      "epoch": 3.0850961692322465,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037531923064614204,
+      "loss": 5.1809,
+      "step": 482000
+    },
+    {
+      "epoch": 3.0882964764617404,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037529362818830613,
+      "loss": 5.1921,
+      "step": 482500
+    },
+    {
+      "epoch": 3.0914967836912344,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037526802573047016,
+      "loss": 5.1906,
+      "step": 483000
+    },
+    {
+      "epoch": 3.0946970909207283,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003752424232726342,
+      "loss": 5.1914,
+      "step": 483500
+    },
+    {
+      "epoch": 3.0978973981502222,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037521682081479823,
+      "loss": 5.1966,
+      "step": 484000
+    },
+    {
+      "epoch": 3.1010977053797166,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00037519121835696227,
+      "loss": 5.1933,
+      "step": 484500
+    },
+    {
+      "epoch": 3.1042980126092106,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037516561589912635,
+      "loss": 5.1917,
+      "step": 485000
+    },
+    {
+      "epoch": 3.1074983198387045,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003751400134412904,
+      "loss": 5.1991,
+      "step": 485500
+    },
+    {
+      "epoch": 3.1106986270681984,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003751144109834544,
+      "loss": 5.1917,
+      "step": 486000
+    },
+    {
+      "epoch": 3.1138989342976924,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00037508880852561846,
+      "loss": 5.192,
+      "step": 486500
+    },
+    {
+      "epoch": 3.1170992415271868,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003750632060677825,
+      "loss": 5.1985,
+      "step": 487000
+    },
+    {
+      "epoch": 3.1202995487566807,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003750376036099466,
+      "loss": 5.1917,
+      "step": 487500
+    },
+    {
+      "epoch": 3.1234998559861746,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003750120011521106,
+      "loss": 5.1907,
+      "step": 488000
+    },
+    {
+      "epoch": 3.1267001632156686,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003749863986942747,
+      "loss": 5.1917,
+      "step": 488500
+    },
+    {
+      "epoch": 3.129900470445163,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037496079623643874,
+      "loss": 5.1896,
+      "step": 489000
+    },
+    {
+      "epoch": 3.133100777674657,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00037493519377860277,
+      "loss": 5.1971,
+      "step": 489500
+    },
+    {
+      "epoch": 3.136301084904151,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003749095913207668,
+      "loss": 5.1931,
+      "step": 490000
+    },
+    {
+      "epoch": 3.1395013921336448,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003748839888629309,
+      "loss": 5.1998,
+      "step": 490500
+    },
+    {
+      "epoch": 3.1427016993631387,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00037485838640509493,
+      "loss": 5.1847,
+      "step": 491000
+    },
+    {
+      "epoch": 3.145902006592633,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00037483278394725896,
+      "loss": 5.1932,
+      "step": 491500
+    },
+    {
+      "epoch": 3.149102313822127,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.000374807181489423,
+      "loss": 5.1958,
+      "step": 492000
+    },
+    {
+      "epoch": 3.152302621051621,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037478157903158703,
+      "loss": 5.191,
+      "step": 492500
+    },
+    {
+      "epoch": 3.155502928281115,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037475597657375106,
+      "loss": 5.1945,
+      "step": 493000
+    },
+    {
+      "epoch": 3.158703235510609,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00037473037411591515,
+      "loss": 5.1945,
+      "step": 493500
+    },
+    {
+      "epoch": 3.161903542740103,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003747047716580792,
+      "loss": 5.1986,
+      "step": 494000
+    },
+    {
+      "epoch": 3.165103849969597,
+      "grad_norm": 0.400390625,
+      "learning_rate": 0.0003746791692002432,
+      "loss": 5.1928,
+      "step": 494500
+    },
+    {
+      "epoch": 3.168304157199091,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00037465356674240726,
+      "loss": 5.195,
+      "step": 495000
+    },
+    {
+      "epoch": 3.171504464428585,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037462796428457134,
+      "loss": 5.1943,
+      "step": 495500
+    },
+    {
+      "epoch": 3.174704771658079,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003746023618267354,
+      "loss": 5.1968,
+      "step": 496000
+    },
+    {
+      "epoch": 3.1779050788875733,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00037457675936889947,
+      "loss": 5.1964,
+      "step": 496500
+    },
+    {
+      "epoch": 3.1811053861170673,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003745511569110635,
+      "loss": 5.1988,
+      "step": 497000
+    },
+    {
+      "epoch": 3.184305693346561,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00037452555445322754,
+      "loss": 5.195,
+      "step": 497500
+    },
+    {
+      "epoch": 3.187506000576055,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00037449995199539157,
+      "loss": 5.1972,
+      "step": 498000
+    },
+    {
+      "epoch": 3.1907063078055495,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003744743495375556,
+      "loss": 5.1964,
+      "step": 498500
+    },
+    {
+      "epoch": 3.1939066150350435,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003744487470797197,
+      "loss": 5.1935,
+      "step": 499000
+    },
+    {
+      "epoch": 3.1971069222645374,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.0003744231446218837,
+      "loss": 5.1911,
+      "step": 499500
+    },
+    {
+      "epoch": 3.2003072294940313,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037439754216404776,
+      "loss": 5.1889,
+      "step": 500000
+    },
+    {
+      "epoch": 3.2035075367235253,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003743719397062118,
+      "loss": 5.1899,
+      "step": 500500
+    },
+    {
+      "epoch": 3.2067078439530197,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00037434633724837583,
+      "loss": 5.1958,
+      "step": 501000
+    },
+    {
+      "epoch": 3.2099081511825136,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003743207347905399,
+      "loss": 5.1986,
+      "step": 501500
+    },
+    {
+      "epoch": 3.2131084584120075,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00037429513233270395,
+      "loss": 5.1947,
+      "step": 502000
+    },
+    {
+      "epoch": 3.2163087656415015,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000374269529874868,
+      "loss": 5.188,
+      "step": 502500
+    },
+    {
+      "epoch": 3.2195090728709954,
+      "grad_norm": 0.3671875,
+      "learning_rate": 0.0003742439274170321,
+      "loss": 5.1988,
+      "step": 503000
+    },
+    {
+      "epoch": 3.22270938010049,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003742183249591961,
+      "loss": 5.1933,
+      "step": 503500
+    },
+    {
+      "epoch": 3.2259096873299837,
+      "grad_norm": 0.38671875,
+      "learning_rate": 0.00037419272250136014,
+      "loss": 5.1946,
+      "step": 504000
+    },
+    {
+      "epoch": 3.2291099945594777,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00037416712004352423,
+      "loss": 5.1928,
+      "step": 504500
+    },
+    {
+      "epoch": 3.2323103017889716,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037414151758568827,
+      "loss": 5.1958,
+      "step": 505000
+    },
+    {
+      "epoch": 3.2355106090184655,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003741159151278523,
+      "loss": 5.1934,
+      "step": 505500
+    },
+    {
+      "epoch": 3.23871091624796,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00037409031267001633,
+      "loss": 5.197,
+      "step": 506000
+    },
+    {
+      "epoch": 3.241911223477454,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037406471021218037,
+      "loss": 5.1905,
+      "step": 506500
+    },
+    {
+      "epoch": 3.245111530706948,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00037403910775434446,
+      "loss": 5.1953,
+      "step": 507000
+    },
+    {
+      "epoch": 3.2483118379364417,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003740135052965085,
+      "loss": 5.1952,
+      "step": 507500
+    },
+    {
+      "epoch": 3.251512145165936,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003739879028386725,
+      "loss": 5.1929,
+      "step": 508000
+    },
+    {
+      "epoch": 3.25471245239543,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00037396230038083656,
+      "loss": 5.1908,
+      "step": 508500
+    },
+    {
+      "epoch": 3.257912759624924,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003739366979230006,
+      "loss": 5.1931,
+      "step": 509000
+    },
+    {
+      "epoch": 3.261113066854418,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003739110954651647,
+      "loss": 5.1924,
+      "step": 509500
+    },
+    {
+      "epoch": 3.264313374083912,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003738854930073287,
+      "loss": 5.1907,
+      "step": 510000
+    },
+    {
+      "epoch": 3.2675136813134062,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037385989054949275,
+      "loss": 5.1911,
+      "step": 510500
+    },
+    {
+      "epoch": 3.2707139885429,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037383428809165684,
+      "loss": 5.1866,
+      "step": 511000
+    },
+    {
+      "epoch": 3.273914295772394,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003738086856338209,
+      "loss": 5.1897,
+      "step": 511500
+    },
+    {
+      "epoch": 3.277114603001888,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003737830831759849,
+      "loss": 5.1928,
+      "step": 512000
+    },
+    {
+      "epoch": 3.2803149102313824,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000373757480718149,
+      "loss": 5.1908,
+      "step": 512500
+    },
+    {
+      "epoch": 3.2835152174608764,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00037373187826031303,
+      "loss": 5.1901,
+      "step": 513000
+    },
+    {
+      "epoch": 3.2867155246903703,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00037370627580247706,
+      "loss": 5.1933,
+      "step": 513500
+    },
+    {
+      "epoch": 3.2899158319198643,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003736806733446411,
+      "loss": 5.1916,
+      "step": 514000
+    },
+    {
+      "epoch": 3.293116139149358,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00037365507088680513,
+      "loss": 5.1879,
+      "step": 514500
+    },
+    {
+      "epoch": 3.296316446378852,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003736294684289692,
+      "loss": 5.1881,
+      "step": 515000
+    },
+    {
+      "epoch": 3.2995167536083465,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00037360386597113326,
+      "loss": 5.1928,
+      "step": 515500
+    },
+    {
+      "epoch": 3.3027170608378404,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003735782635132973,
+      "loss": 5.1954,
+      "step": 516000
+    },
+    {
+      "epoch": 3.3059173680673344,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003735526610554613,
+      "loss": 5.1907,
+      "step": 516500
+    },
+    {
+      "epoch": 3.3091176752968283,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00037352705859762536,
+      "loss": 5.1956,
+      "step": 517000
+    },
+    {
+      "epoch": 3.3123179825263227,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003735014561397894,
+      "loss": 5.1939,
+      "step": 517500
+    },
+    {
+      "epoch": 3.3155182897558166,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003734758536819535,
+      "loss": 5.1891,
+      "step": 518000
+    },
+    {
+      "epoch": 3.3187185969853106,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0003734502512241175,
+      "loss": 5.1844,
+      "step": 518500
+    },
+    {
+      "epoch": 3.3219189042148045,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003734246487662816,
+      "loss": 5.1886,
+      "step": 519000
+    },
+    {
+      "epoch": 3.3251192114442985,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00037339904630844564,
+      "loss": 5.1915,
+      "step": 519500
+    },
+    {
+      "epoch": 3.328319518673793,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037337344385060967,
+      "loss": 5.1936,
+      "step": 520000
+    },
+    {
+      "epoch": 3.3315198259032868,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00037334784139277376,
+      "loss": 5.1928,
+      "step": 520500
+    },
+    {
+      "epoch": 3.3347201331327807,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003733222389349378,
+      "loss": 5.1946,
+      "step": 521000
+    },
+    {
+      "epoch": 3.3379204403622746,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037329663647710183,
+      "loss": 5.1969,
+      "step": 521500
+    },
+    {
+      "epoch": 3.341120747591769,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00037327103401926586,
+      "loss": 5.1986,
+      "step": 522000
+    },
+    {
+      "epoch": 3.344321054821263,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003732454315614299,
+      "loss": 5.1924,
+      "step": 522500
+    },
+    {
+      "epoch": 3.347521362050757,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00037321982910359393,
+      "loss": 5.1939,
+      "step": 523000
+    },
+    {
+      "epoch": 3.350721669280251,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.000373194226645758,
+      "loss": 5.1984,
+      "step": 523500
+    },
+    {
+      "epoch": 3.3539219765097448,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037316862418792205,
+      "loss": 5.1929,
+      "step": 524000
+    },
+    {
+      "epoch": 3.357122283739239,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003731430217300861,
+      "loss": 5.1921,
+      "step": 524500
+    },
+    {
+      "epoch": 3.360322590968733,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003731174192722501,
+      "loss": 5.1959,
+      "step": 525000
+    },
+    {
+      "epoch": 3.363522898198227,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003730918168144142,
+      "loss": 5.1947,
+      "step": 525500
+    },
+    {
+      "epoch": 3.366723205427721,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00037306621435657825,
+      "loss": 5.1939,
+      "step": 526000
+    },
+    {
+      "epoch": 3.369923512657215,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037304061189874233,
+      "loss": 5.2001,
+      "step": 526500
+    },
+    {
+      "epoch": 3.3731238198867093,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00037301500944090637,
+      "loss": 5.1944,
+      "step": 527000
+    },
+    {
+      "epoch": 3.376324127116203,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003729894069830704,
+      "loss": 5.1885,
+      "step": 527500
+    },
+    {
+      "epoch": 3.379524434345697,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.00037296380452523444,
+      "loss": 5.1954,
+      "step": 528000
+    },
+    {
+      "epoch": 3.382724741575191,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00037293820206739847,
+      "loss": 5.1898,
+      "step": 528500
+    },
+    {
+      "epoch": 3.385925048804685,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037291259960956256,
+      "loss": 5.1952,
+      "step": 529000
+    },
+    {
+      "epoch": 3.3891253560341794,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003728869971517266,
+      "loss": 5.1911,
+      "step": 529500
+    },
+    {
+      "epoch": 3.3923256632636734,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00037286139469389063,
+      "loss": 5.1901,
+      "step": 530000
+    },
+    {
+      "epoch": 3.3955259704931673,
+      "grad_norm": 0.40625,
+      "learning_rate": 0.00037283579223605466,
+      "loss": 5.1941,
+      "step": 530500
+    },
+    {
+      "epoch": 3.3987262777226612,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003728101897782187,
+      "loss": 5.1962,
+      "step": 531000
+    },
+    {
+      "epoch": 3.4019265849521556,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003727845873203828,
+      "loss": 5.1943,
+      "step": 531500
+    },
+    {
+      "epoch": 3.4051268921816495,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003727589848625468,
+      "loss": 5.1894,
+      "step": 532000
+    },
+    {
+      "epoch": 3.4083271994111435,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037273338240471085,
+      "loss": 5.1891,
+      "step": 532500
+    },
+    {
+      "epoch": 3.4115275066406374,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003727077799468749,
+      "loss": 5.1877,
+      "step": 533000
+    },
+    {
+      "epoch": 3.4147278138701314,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.000372682177489039,
+      "loss": 5.1948,
+      "step": 533500
+    },
+    {
+      "epoch": 3.4179281210996257,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000372656575031203,
+      "loss": 5.1937,
+      "step": 534000
+    },
+    {
+      "epoch": 3.4211284283291197,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003726309725733671,
+      "loss": 5.1932,
+      "step": 534500
+    },
+    {
+      "epoch": 3.4243287355586136,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00037260537011553113,
+      "loss": 5.1979,
+      "step": 535000
+    },
+    {
+      "epoch": 3.4275290427881075,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00037257976765769517,
+      "loss": 5.1899,
+      "step": 535500
+    },
+    {
+      "epoch": 3.430729350017602,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003725541651998592,
+      "loss": 5.1883,
+      "step": 536000
+    },
+    {
+      "epoch": 3.433929657247096,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00037252856274202324,
+      "loss": 5.1951,
+      "step": 536500
+    },
+    {
+      "epoch": 3.43712996447659,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003725029602841873,
+      "loss": 5.1938,
+      "step": 537000
+    },
+    {
+      "epoch": 3.4403302717060837,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037247735782635136,
+      "loss": 5.1935,
+      "step": 537500
+    },
+    {
+      "epoch": 3.4435305789355777,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003724517553685154,
+      "loss": 5.1865,
+      "step": 538000
+    },
+    {
+      "epoch": 3.4467308861650716,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00037242615291067943,
+      "loss": 5.1933,
+      "step": 538500
+    },
+    {
+      "epoch": 3.449931193394566,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00037240055045284346,
+      "loss": 5.1946,
+      "step": 539000
+    },
+    {
+      "epoch": 3.45313150062406,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003723749479950075,
+      "loss": 5.1926,
+      "step": 539500
+    },
+    {
+      "epoch": 3.456331807853554,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003723493455371716,
+      "loss": 5.1921,
+      "step": 540000
+    },
+    {
+      "epoch": 3.459532115083048,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003723237430793356,
+      "loss": 5.1952,
+      "step": 540500
+    },
+    {
+      "epoch": 3.462732422312542,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003722981406214997,
+      "loss": 5.1905,
+      "step": 541000
+    },
+    {
+      "epoch": 3.465932729542036,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.00037227253816366374,
+      "loss": 5.1907,
+      "step": 541500
+    },
+    {
+      "epoch": 3.46913303677153,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003722469357058278,
+      "loss": 5.1925,
+      "step": 542000
+    },
+    {
+      "epoch": 3.472333344001024,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037222133324799186,
+      "loss": 5.1903,
+      "step": 542500
+    },
+    {
+      "epoch": 3.475533651230518,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003721957307901559,
+      "loss": 5.1954,
+      "step": 543000
+    },
+    {
+      "epoch": 3.4787339584600123,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00037217012833231993,
+      "loss": 5.1947,
+      "step": 543500
+    },
+    {
+      "epoch": 3.4819342656895063,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00037214452587448397,
+      "loss": 5.1955,
+      "step": 544000
+    },
+    {
+      "epoch": 3.485134572919,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.000372118923416648,
+      "loss": 5.195,
+      "step": 544500
+    },
+    {
+      "epoch": 3.488334880148494,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.00037209332095881203,
+      "loss": 5.1956,
+      "step": 545000
+    },
+    {
+      "epoch": 3.4915351873779885,
+      "grad_norm": 0.3984375,
+      "learning_rate": 0.0003720677185009761,
+      "loss": 5.1975,
+      "step": 545500
+    },
+    {
+      "epoch": 3.4947354946074825,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00037204211604314016,
+      "loss": 5.1947,
+      "step": 546000
+    },
+    {
+      "epoch": 3.4979358018369764,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003720165135853042,
+      "loss": 5.1972,
+      "step": 546500
+    },
+    {
+      "epoch": 3.5011361090664703,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003719909111274682,
+      "loss": 5.1907,
+      "step": 547000
+    },
+    {
+      "epoch": 3.5043364162959643,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00037196530866963226,
+      "loss": 5.1957,
+      "step": 547500
+    },
+    {
+      "epoch": 3.507536723525458,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037193970621179635,
+      "loss": 5.1908,
+      "step": 548000
+    },
+    {
+      "epoch": 3.5107370307549526,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003719141037539604,
+      "loss": 5.1912,
+      "step": 548500
+    },
+    {
+      "epoch": 3.5139373379844465,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00037188850129612447,
+      "loss": 5.1996,
+      "step": 549000
+    },
+    {
+      "epoch": 3.5171376452139405,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003718628988382885,
+      "loss": 5.1944,
+      "step": 549500
+    },
+    {
+      "epoch": 3.520337952443435,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00037183729638045254,
+      "loss": 5.1949,
+      "step": 550000
+    },
+    {
+      "epoch": 3.5235382596729288,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00037181169392261663,
+      "loss": 5.1878,
+      "step": 550500
+    },
+    {
+      "epoch": 3.5267385669024227,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00037178609146478066,
+      "loss": 5.1897,
+      "step": 551000
+    },
+    {
+      "epoch": 3.5299388741319166,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003717604890069447,
+      "loss": 5.1953,
+      "step": 551500
+    },
+    {
+      "epoch": 3.5331391813614106,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00037173488654910873,
+      "loss": 5.1928,
+      "step": 552000
+    },
+    {
+      "epoch": 3.5363394885909045,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00037170928409127277,
+      "loss": 5.1918,
+      "step": 552500
+    },
+    {
+      "epoch": 3.539539795820399,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003716836816334368,
+      "loss": 5.1947,
+      "step": 553000
+    },
+    {
+      "epoch": 3.542740103049893,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003716580791756009,
+      "loss": 5.1898,
+      "step": 553500
+    },
+    {
+      "epoch": 3.545940410279387,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003716324767177649,
+      "loss": 5.194,
+      "step": 554000
+    },
+    {
+      "epoch": 3.5491407175088807,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00037160687425992896,
+      "loss": 5.1985,
+      "step": 554500
+    },
+    {
+      "epoch": 3.552341024738375,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.000371581271802093,
+      "loss": 5.1973,
+      "step": 555000
+    },
+    {
+      "epoch": 3.555541331967869,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.000371555669344257,
+      "loss": 5.1947,
+      "step": 555500
+    },
+    {
+      "epoch": 3.558741639197363,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003715300668864211,
+      "loss": 5.1974,
+      "step": 556000
+    },
+    {
+      "epoch": 3.561941946426857,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.0003715044644285852,
+      "loss": 5.1936,
+      "step": 556500
+    },
+    {
+      "epoch": 3.565142253656351,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00037147886197074924,
+      "loss": 5.197,
+      "step": 557000
+    },
+    {
+      "epoch": 3.568342560885845,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00037145325951291327,
+      "loss": 5.1962,
+      "step": 557500
+    },
+    {
+      "epoch": 3.571542868115339,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003714276570550773,
+      "loss": 5.1893,
+      "step": 558000
+    },
+    {
+      "epoch": 3.574743175344833,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00037140205459724134,
+      "loss": 5.1963,
+      "step": 558500
+    },
+    {
+      "epoch": 3.577943482574327,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00037137645213940543,
+      "loss": 5.1885,
+      "step": 559000
+    },
+    {
+      "epoch": 3.5811437898038214,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00037135084968156946,
+      "loss": 5.192,
+      "step": 559500
+    },
+    {
+      "epoch": 3.5843440970333154,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003713252472237335,
+      "loss": 5.1898,
+      "step": 560000
+    },
+    {
+      "epoch": 3.5875444042628093,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00037129964476589753,
+      "loss": 5.1964,
+      "step": 560500
+    },
+    {
+      "epoch": 3.5907447114923032,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00037127404230806156,
+      "loss": 5.1908,
+      "step": 561000
+    },
+    {
+      "epoch": 3.593945018721797,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00037124843985022565,
+      "loss": 5.1914,
+      "step": 561500
+    },
+    {
+      "epoch": 3.597145325951291,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.0003712228373923897,
+      "loss": 5.1917,
+      "step": 562000
+    },
+    {
+      "epoch": 3.6003456331807855,
+      "grad_norm": 0.416015625,
+      "learning_rate": 0.0003711972349345537,
+      "loss": 5.1943,
+      "step": 562500
+    },
+    {
+      "epoch": 3.6035459404102794,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00037117163247671776,
+      "loss": 5.1978,
+      "step": 563000
+    },
+    {
+      "epoch": 3.6067462476397734,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.00037114603001888184,
+      "loss": 5.1876,
+      "step": 563500
+    },
+    {
+      "epoch": 3.6099465548692673,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003711204275610459,
+      "loss": 5.191,
+      "step": 564000
+    },
+    {
+      "epoch": 3.6131468620987617,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00037109482510320997,
+      "loss": 5.1959,
+      "step": 564500
+    },
+    {
+      "epoch": 3.6163471693282556,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000371069222645374,
+      "loss": 5.1885,
+      "step": 565000
+    },
+    {
+      "epoch": 3.6195474765577496,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00037104362018753804,
+      "loss": 5.1891,
+      "step": 565500
+    },
+    {
+      "epoch": 3.6227477837872435,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00037101801772970207,
+      "loss": 5.1944,
+      "step": 566000
+    },
+    {
+      "epoch": 3.6259480910167374,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003709924152718661,
+      "loss": 5.1902,
+      "step": 566500
+    },
+    {
+      "epoch": 3.6291483982462314,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003709668128140302,
+      "loss": 5.1885,
+      "step": 567000
+    },
+    {
+      "epoch": 3.6323487054757257,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003709412103561942,
+      "loss": 5.189,
+      "step": 567500
+    },
+    {
+      "epoch": 3.6355490127052197,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00037091560789835826,
+      "loss": 5.1871,
+      "step": 568000
+    },
+    {
+      "epoch": 3.6387493199347136,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003708900054405223,
+      "loss": 5.1998,
+      "step": 568500
+    },
+    {
+      "epoch": 3.641949627164208,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00037086440298268633,
+      "loss": 5.1883,
+      "step": 569000
+    },
+    {
+      "epoch": 3.645149934393702,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00037083880052485036,
+      "loss": 5.1916,
+      "step": 569500
+    },
+    {
+      "epoch": 3.648350241623196,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00037081319806701445,
+      "loss": 5.1919,
+      "step": 570000
+    },
+    {
+      "epoch": 3.65155054885269,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0003707875956091785,
+      "loss": 5.1965,
+      "step": 570500
+    },
+    {
+      "epoch": 3.6547508560821838,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003707619931513425,
+      "loss": 5.1954,
+      "step": 571000
+    },
+    {
+      "epoch": 3.6579511633116777,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003707363906935066,
+      "loss": 5.1963,
+      "step": 571500
+    },
+    {
+      "epoch": 3.661151470541172,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00037071078823567064,
+      "loss": 5.1893,
+      "step": 572000
+    },
+    {
+      "epoch": 3.664351777770666,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00037068518577783473,
+      "loss": 5.1916,
+      "step": 572500
+    },
+    {
+      "epoch": 3.66755208500016,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00037065958331999877,
+      "loss": 5.1885,
+      "step": 573000
+    },
+    {
+      "epoch": 3.670752392229654,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003706339808621628,
+      "loss": 5.1929,
+      "step": 573500
+    },
+    {
+      "epoch": 3.6739526994591483,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00037060837840432683,
+      "loss": 5.1891,
+      "step": 574000
+    },
+    {
+      "epoch": 3.677153006688642,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00037058277594649087,
+      "loss": 5.1971,
+      "step": 574500
+    },
+    {
+      "epoch": 3.680353313918136,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0003705571734886549,
+      "loss": 5.201,
+      "step": 575000
+    },
+    {
+      "epoch": 3.68355362114763,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.000370531571030819,
+      "loss": 5.1911,
+      "step": 575500
+    },
+    {
+      "epoch": 3.686753928377124,
+      "grad_norm": 0.408203125,
+      "learning_rate": 0.000370505968572983,
+      "loss": 5.18,
+      "step": 576000
+    },
+    {
+      "epoch": 3.6899542356066184,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00037048036611514706,
+      "loss": 5.1987,
+      "step": 576500
+    },
+    {
+      "epoch": 3.6931545428361123,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003704547636573111,
+      "loss": 5.1972,
+      "step": 577000
+    },
+    {
+      "epoch": 3.6963548500656063,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00037042916119947513,
+      "loss": 5.1944,
+      "step": 577500
+    },
+    {
+      "epoch": 3.6995551572951,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003704035587416392,
+      "loss": 5.1943,
+      "step": 578000
+    },
+    {
+      "epoch": 3.7027554645245946,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.00037037795628380325,
+      "loss": 5.1932,
+      "step": 578500
+    },
+    {
+      "epoch": 3.7059557717540885,
+      "grad_norm": 0.419921875,
+      "learning_rate": 0.00037035235382596734,
+      "loss": 5.198,
+      "step": 579000
+    },
+    {
+      "epoch": 3.7091560789835825,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003703267513681314,
+      "loss": 5.201,
+      "step": 579500
+    },
+    {
+      "epoch": 3.7123563862130764,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003703011489102954,
+      "loss": 5.191,
+      "step": 580000
+    },
+    {
+      "epoch": 3.7155566934425703,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00037027554645245944,
+      "loss": 5.1979,
+      "step": 580500
+    },
+    {
+      "epoch": 3.7187570006720643,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00037024994399462353,
+      "loss": 5.1952,
+      "step": 581000
+    },
+    {
+      "epoch": 3.7219573079015587,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00037022434153678756,
+      "loss": 5.1955,
+      "step": 581500
+    },
+    {
+      "epoch": 3.7251576151310526,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0003701987390789516,
+      "loss": 5.1907,
+      "step": 582000
+    },
+    {
+      "epoch": 3.7283579223605465,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00037017313662111563,
+      "loss": 5.189,
+      "step": 582500
+    },
+    {
+      "epoch": 3.731558229590041,
+      "grad_norm": 0.4140625,
+      "learning_rate": 0.00037014753416327967,
+      "loss": 5.1898,
+      "step": 583000
+    },
+    {
+      "epoch": 3.734758536819535,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00037012193170544376,
+      "loss": 5.1901,
+      "step": 583500
+    },
+    {
+      "epoch": 3.737958844049029,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003700963292476078,
+      "loss": 5.1962,
+      "step": 584000
+    },
+    {
+      "epoch": 3.7411591512785227,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.0003700707267897718,
+      "loss": 5.1929,
+      "step": 584500
+    },
+    {
+      "epoch": 3.7443594585080167,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00037004512433193586,
+      "loss": 5.1978,
+      "step": 585000
+    },
+    {
+      "epoch": 3.7475597657375106,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003700195218740999,
+      "loss": 5.1906,
+      "step": 585500
+    },
+    {
+      "epoch": 3.750760072967005,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.000369993919416264,
+      "loss": 5.1892,
+      "step": 586000
+    },
+    {
+      "epoch": 3.753960380196499,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.000369968316958428,
+      "loss": 5.191,
+      "step": 586500
+    },
+    {
+      "epoch": 3.757160687425993,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0003699427145005921,
+      "loss": 5.1967,
+      "step": 587000
+    },
+    {
+      "epoch": 3.760360994655487,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00036991711204275614,
+      "loss": 5.1956,
+      "step": 587500
+    },
+    {
+      "epoch": 3.763561301884981,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00036989150958492017,
+      "loss": 5.1934,
+      "step": 588000
+    },
+    {
+      "epoch": 3.766761609114475,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003698659071270842,
+      "loss": 5.1938,
+      "step": 588500
+    },
+    {
+      "epoch": 3.769961916343969,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003698403046692483,
+      "loss": 5.1923,
+      "step": 589000
+    },
+    {
+      "epoch": 3.773162223573463,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00036981470221141233,
+      "loss": 5.1926,
+      "step": 589500
+    },
+    {
+      "epoch": 3.776362530802957,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00036978909975357636,
+      "loss": 5.1917,
+      "step": 590000
+    },
+    {
+      "epoch": 3.779562838032451,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003697634972957404,
+      "loss": 5.1902,
+      "step": 590500
+    },
+    {
+      "epoch": 3.7827631452619452,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00036973789483790443,
+      "loss": 5.1998,
+      "step": 591000
+    },
+    {
+      "epoch": 3.785963452491439,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003697122923800685,
+      "loss": 5.1894,
+      "step": 591500
+    },
+    {
+      "epoch": 3.789163759720933,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00036968668992223255,
+      "loss": 5.1918,
+      "step": 592000
+    },
+    {
+      "epoch": 3.7923640669504275,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003696610874643966,
+      "loss": 5.1876,
+      "step": 592500
+    },
+    {
+      "epoch": 3.7955643741799214,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003696354850065606,
+      "loss": 5.2001,
+      "step": 593000
+    },
+    {
+      "epoch": 3.7987646814094154,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00036960988254872466,
+      "loss": 5.1952,
+      "step": 593500
+    },
+    {
+      "epoch": 3.8019649886389093,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00036958428009088875,
+      "loss": 5.192,
+      "step": 594000
+    },
+    {
+      "epoch": 3.8051652958684032,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00036955867763305283,
+      "loss": 5.1968,
+      "step": 594500
+    },
+    {
+      "epoch": 3.808365603097897,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00036953307517521687,
+      "loss": 5.1993,
+      "step": 595000
+    },
+    {
+      "epoch": 3.8115659103273916,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003695074727173809,
+      "loss": 5.1949,
+      "step": 595500
+    },
+    {
+      "epoch": 3.8147662175568855,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00036948187025954494,
+      "loss": 5.1955,
+      "step": 596000
+    },
+    {
+      "epoch": 3.8179665247863794,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.00036945626780170897,
+      "loss": 5.1966,
+      "step": 596500
+    },
+    {
+      "epoch": 3.8211668320158734,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00036943066534387306,
+      "loss": 5.1922,
+      "step": 597000
+    },
+    {
+      "epoch": 3.8243671392453678,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003694050628860371,
+      "loss": 5.188,
+      "step": 597500
+    },
+    {
+      "epoch": 3.8275674464748617,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00036937946042820113,
+      "loss": 5.1951,
+      "step": 598000
+    },
+    {
+      "epoch": 3.8307677537043556,
+      "grad_norm": 0.404296875,
+      "learning_rate": 0.00036935385797036516,
+      "loss": 5.1962,
+      "step": 598500
+    },
+    {
+      "epoch": 3.8339680609338496,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003693282555125292,
+      "loss": 5.1961,
+      "step": 599000
+    },
+    {
+      "epoch": 3.8371683681633435,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.00036930265305469323,
+      "loss": 5.1996,
+      "step": 599500
+    },
+    {
+      "epoch": 3.8403686753928374,
+      "grad_norm": 0.427734375,
+      "learning_rate": 0.0003692770505968573,
+      "loss": 5.1948,
+      "step": 600000
+    },
+    {
+      "epoch": 3.843568982622332,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00036925144813902135,
+      "loss": 5.1891,
+      "step": 600500
+    },
+    {
+      "epoch": 3.8467692898518258,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0003692258456811854,
+      "loss": 5.1954,
+      "step": 601000
+    },
+    {
+      "epoch": 3.8499695970813197,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003692002432233495,
+      "loss": 5.1881,
+      "step": 601500
+    },
+    {
+      "epoch": 3.853169904310814,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003691746407655135,
+      "loss": 5.1938,
+      "step": 602000
+    },
+    {
+      "epoch": 3.856370211540308,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003691490383076776,
+      "loss": 5.1946,
+      "step": 602500
+    },
+    {
+      "epoch": 3.859570518769802,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00036912343584984163,
+      "loss": 5.1926,
+      "step": 603000
+    },
+    {
+      "epoch": 3.862770825999296,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036909783339200567,
+      "loss": 5.1925,
+      "step": 603500
+    },
+    {
+      "epoch": 3.86597113322879,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003690722309341697,
+      "loss": 5.1956,
+      "step": 604000
+    },
+    {
+      "epoch": 3.8691714404582838,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00036904662847633374,
+      "loss": 5.1917,
+      "step": 604500
+    },
+    {
+      "epoch": 3.872371747687778,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00036902102601849777,
+      "loss": 5.1941,
+      "step": 605000
+    },
+    {
+      "epoch": 3.875572054917272,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00036899542356066186,
+      "loss": 5.196,
+      "step": 605500
+    },
+    {
+      "epoch": 3.878772362146766,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003689698211028259,
+      "loss": 5.1926,
+      "step": 606000
+    },
+    {
+      "epoch": 3.88197266937626,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00036894421864498993,
+      "loss": 5.1966,
+      "step": 606500
+    },
+    {
+      "epoch": 3.8851729766057543,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.00036891861618715396,
+      "loss": 5.1931,
+      "step": 607000
+    },
+    {
+      "epoch": 3.8883732838352483,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.000368893013729318,
+      "loss": 5.1937,
+      "step": 607500
+    },
+    {
+      "epoch": 3.891573591064742,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003688674112714821,
+      "loss": 5.195,
+      "step": 608000
+    },
+    {
+      "epoch": 3.894773898294236,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0003688418088136461,
+      "loss": 5.1918,
+      "step": 608500
+    },
+    {
+      "epoch": 3.89797420552373,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036881620635581015,
+      "loss": 5.1942,
+      "step": 609000
+    },
+    {
+      "epoch": 3.9011745127532245,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00036879060389797424,
+      "loss": 5.1918,
+      "step": 609500
+    },
+    {
+      "epoch": 3.9043748199827184,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003687650014401383,
+      "loss": 5.1955,
+      "step": 610000
+    },
+    {
+      "epoch": 3.9075751272122123,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.0003687393989823023,
+      "loss": 5.19,
+      "step": 610500
+    },
+    {
+      "epoch": 3.9107754344417063,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003687137965244664,
+      "loss": 5.1947,
+      "step": 611000
+    },
+    {
+      "epoch": 3.9139757416712007,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00036868819406663043,
+      "loss": 5.1975,
+      "step": 611500
+    },
+    {
+      "epoch": 3.9171760489006946,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00036866259160879447,
+      "loss": 5.1929,
+      "step": 612000
+    },
+    {
+      "epoch": 3.9203763561301885,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003686369891509585,
+      "loss": 5.1986,
+      "step": 612500
+    },
+    {
+      "epoch": 3.9235766633596825,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00036861138669312254,
+      "loss": 5.1918,
+      "step": 613000
+    },
+    {
+      "epoch": 3.9267769705891764,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.0003685857842352866,
+      "loss": 5.1931,
+      "step": 613500
+    },
+    {
+      "epoch": 3.9299772778186703,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00036856018177745066,
+      "loss": 5.1948,
+      "step": 614000
+    },
+    {
+      "epoch": 3.9331775850481647,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003685345793196147,
+      "loss": 5.1881,
+      "step": 614500
+    },
+    {
+      "epoch": 3.9363778922776587,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003685089768617787,
+      "loss": 5.1954,
+      "step": 615000
+    },
+    {
+      "epoch": 3.9395781995071526,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00036848337440394276,
+      "loss": 5.1983,
+      "step": 615500
+    },
+    {
+      "epoch": 3.942778506736647,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00036845777194610685,
+      "loss": 5.192,
+      "step": 616000
+    },
+    {
+      "epoch": 3.945978813966141,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003684321694882709,
+      "loss": 5.1921,
+      "step": 616500
+    },
+    {
+      "epoch": 3.949179121195635,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036840656703043497,
+      "loss": 5.193,
+      "step": 617000
+    },
+    {
+      "epoch": 3.952379428425129,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000368380964572599,
+      "loss": 5.1931,
+      "step": 617500
+    },
+    {
+      "epoch": 3.9555797356546227,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00036835536211476304,
+      "loss": 5.1898,
+      "step": 618000
+    },
+    {
+      "epoch": 3.9587800428841167,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0003683297596569271,
+      "loss": 5.1993,
+      "step": 618500
+    },
+    {
+      "epoch": 3.961980350113611,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00036830415719909116,
+      "loss": 5.1947,
+      "step": 619000
+    },
+    {
+      "epoch": 3.965180657343105,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003682785547412552,
+      "loss": 5.1963,
+      "step": 619500
+    },
+    {
+      "epoch": 3.968380964572599,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00036825295228341923,
+      "loss": 5.1975,
+      "step": 620000
+    },
+    {
+      "epoch": 3.971581271802093,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00036822734982558327,
+      "loss": 5.195,
+      "step": 620500
+    },
+    {
+      "epoch": 3.9747815790315872,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003682017473677473,
+      "loss": 5.1945,
+      "step": 621000
+    },
+    {
+      "epoch": 3.977981886261081,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003681761449099114,
+      "loss": 5.1949,
+      "step": 621500
+    },
+    {
+      "epoch": 3.981182193490575,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003681505424520754,
+      "loss": 5.1965,
+      "step": 622000
+    },
+    {
+      "epoch": 3.984382500720069,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00036812493999423946,
+      "loss": 5.1927,
+      "step": 622500
+    },
+    {
+      "epoch": 3.987582807949563,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003680993375364035,
+      "loss": 5.1926,
+      "step": 623000
+    },
+    {
+      "epoch": 3.990783115179057,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003680737350785675,
+      "loss": 5.1971,
+      "step": 623500
+    },
+    {
+      "epoch": 3.9939834224085513,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003680481326207316,
+      "loss": 5.1945,
+      "step": 624000
+    },
+    {
+      "epoch": 3.9971837296380452,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00036802253016289565,
+      "loss": 5.1938,
+      "step": 624500
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 5.188127040863037,
+      "eval_runtime": 1.1918,
+      "eval_samples_per_second": 839.075,
+      "eval_steps_per_second": 13.425,
+      "step": 624940
+    },
+    {
+      "epoch": 4.00038403686754,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036799692770505974,
+      "loss": 5.1936,
+      "step": 625000
+    },
+    {
+      "epoch": 4.003584344097034,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00036797132524722377,
+      "loss": 5.1842,
+      "step": 625500
+    },
+    {
+      "epoch": 4.0067846513265275,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003679457227893878,
+      "loss": 5.1866,
+      "step": 626000
+    },
+    {
+      "epoch": 4.009984958556021,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00036792012033155184,
+      "loss": 5.1875,
+      "step": 626500
+    },
+    {
+      "epoch": 4.013185265785515,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00036789451787371593,
+      "loss": 5.1882,
+      "step": 627000
+    },
+    {
+      "epoch": 4.016385573015009,
+      "grad_norm": 0.431640625,
+      "learning_rate": 0.00036786891541587996,
+      "loss": 5.1898,
+      "step": 627500
+    },
+    {
+      "epoch": 4.019585880244503,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.000367843312958044,
+      "loss": 5.1909,
+      "step": 628000
+    },
+    {
+      "epoch": 4.022786187473997,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00036781771050020803,
+      "loss": 5.1898,
+      "step": 628500
+    },
+    {
+      "epoch": 4.025986494703491,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00036779210804237206,
+      "loss": 5.1957,
+      "step": 629000
+    },
+    {
+      "epoch": 4.029186801932986,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0003677665055845361,
+      "loss": 5.1845,
+      "step": 629500
+    },
+    {
+      "epoch": 4.03238710916248,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0003677409031267002,
+      "loss": 5.1906,
+      "step": 630000
+    },
+    {
+      "epoch": 4.035587416391974,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.0003677153006688642,
+      "loss": 5.1994,
+      "step": 630500
+    },
+    {
+      "epoch": 4.038787723621468,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036768969821102826,
+      "loss": 5.1873,
+      "step": 631000
+    },
+    {
+      "epoch": 4.041988030850962,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003676640957531923,
+      "loss": 5.1878,
+      "step": 631500
+    },
+    {
+      "epoch": 4.045188338080456,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003676384932953564,
+      "loss": 5.1948,
+      "step": 632000
+    },
+    {
+      "epoch": 4.04838864530995,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00036761289083752047,
+      "loss": 5.1922,
+      "step": 632500
+    },
+    {
+      "epoch": 4.0515889525394435,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003675872883796845,
+      "loss": 5.1802,
+      "step": 633000
+    },
+    {
+      "epoch": 4.0547892597689374,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00036756168592184854,
+      "loss": 5.1917,
+      "step": 633500
+    },
+    {
+      "epoch": 4.057989566998432,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00036753608346401257,
+      "loss": 5.1942,
+      "step": 634000
+    },
+    {
+      "epoch": 4.061189874227926,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003675104810061766,
+      "loss": 5.1912,
+      "step": 634500
+    },
+    {
+      "epoch": 4.06439018145742,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00036748487854834064,
+      "loss": 5.186,
+      "step": 635000
+    },
+    {
+      "epoch": 4.067590488686914,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003674592760905047,
+      "loss": 5.1892,
+      "step": 635500
+    },
+    {
+      "epoch": 4.070790795916408,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00036743367363266876,
+      "loss": 5.1912,
+      "step": 636000
+    },
+    {
+      "epoch": 4.073991103145902,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003674080711748328,
+      "loss": 5.1944,
+      "step": 636500
+    },
+    {
+      "epoch": 4.077191410375396,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00036738246871699683,
+      "loss": 5.1918,
+      "step": 637000
+    },
+    {
+      "epoch": 4.08039171760489,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00036735686625916086,
+      "loss": 5.1935,
+      "step": 637500
+    },
+    {
+      "epoch": 4.083592024834384,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00036733126380132495,
+      "loss": 5.1961,
+      "step": 638000
+    },
+    {
+      "epoch": 4.086792332063878,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.000367305661343489,
+      "loss": 5.1916,
+      "step": 638500
+    },
+    {
+      "epoch": 4.0899926392933725,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.000367280058885653,
+      "loss": 5.1949,
+      "step": 639000
+    },
+    {
+      "epoch": 4.0931929465228665,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003672544564278171,
+      "loss": 5.1934,
+      "step": 639500
+    },
+    {
+      "epoch": 4.09639325375236,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00036722885396998114,
+      "loss": 5.1895,
+      "step": 640000
+    },
+    {
+      "epoch": 4.099593560981854,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0003672032515121452,
+      "loss": 5.193,
+      "step": 640500
+    },
+    {
+      "epoch": 4.102793868211348,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00036717764905430927,
+      "loss": 5.1871,
+      "step": 641000
+    },
+    {
+      "epoch": 4.105994175440842,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003671520465964733,
+      "loss": 5.1882,
+      "step": 641500
+    },
+    {
+      "epoch": 4.109194482670336,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00036712644413863733,
+      "loss": 5.1871,
+      "step": 642000
+    },
+    {
+      "epoch": 4.11239478989983,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036710084168080137,
+      "loss": 5.1922,
+      "step": 642500
+    },
+    {
+      "epoch": 4.115595097129324,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003670752392229654,
+      "loss": 5.186,
+      "step": 643000
+    },
+    {
+      "epoch": 4.118795404358819,
+      "grad_norm": 0.412109375,
+      "learning_rate": 0.0003670496367651295,
+      "loss": 5.1933,
+      "step": 643500
+    },
+    {
+      "epoch": 4.121995711588313,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003670240343072935,
+      "loss": 5.1924,
+      "step": 644000
+    },
+    {
+      "epoch": 4.125196018817807,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00036699843184945756,
+      "loss": 5.1947,
+      "step": 644500
+    },
+    {
+      "epoch": 4.128396326047301,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.0003669728293916216,
+      "loss": 5.2002,
+      "step": 645000
+    },
+    {
+      "epoch": 4.131596633276795,
+      "grad_norm": 0.40234375,
+      "learning_rate": 0.00036694722693378563,
+      "loss": 5.1989,
+      "step": 645500
+    },
+    {
+      "epoch": 4.1347969405062885,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00036692162447594966,
+      "loss": 5.1922,
+      "step": 646000
+    },
+    {
+      "epoch": 4.1379972477357825,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00036689602201811375,
+      "loss": 5.1884,
+      "step": 646500
+    },
+    {
+      "epoch": 4.141197554965276,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003668704195602778,
+      "loss": 5.1934,
+      "step": 647000
+    },
+    {
+      "epoch": 4.14439786219477,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003668448171024419,
+      "loss": 5.1914,
+      "step": 647500
+    },
+    {
+      "epoch": 4.147598169424265,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003668192146446059,
+      "loss": 5.1949,
+      "step": 648000
+    },
+    {
+      "epoch": 4.150798476653759,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00036679361218676994,
+      "loss": 5.1976,
+      "step": 648500
+    },
+    {
+      "epoch": 4.153998783883253,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036676800972893403,
+      "loss": 5.1909,
+      "step": 649000
+    },
+    {
+      "epoch": 4.157199091112747,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00036674240727109806,
+      "loss": 5.1887,
+      "step": 649500
+    },
+    {
+      "epoch": 4.160399398342241,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003667168048132621,
+      "loss": 5.1877,
+      "step": 650000
+    },
+    {
+      "epoch": 4.163599705571735,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00036669120235542613,
+      "loss": 5.1908,
+      "step": 650500
+    },
+    {
+      "epoch": 4.166800012801229,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00036666559989759017,
+      "loss": 5.1921,
+      "step": 651000
+    },
+    {
+      "epoch": 4.170000320030723,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003666399974397542,
+      "loss": 5.1892,
+      "step": 651500
+    },
+    {
+      "epoch": 4.173200627260217,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003666143949819183,
+      "loss": 5.1924,
+      "step": 652000
+    },
+    {
+      "epoch": 4.176400934489711,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003665887925240823,
+      "loss": 5.195,
+      "step": 652500
+    },
+    {
+      "epoch": 4.179601241719205,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00036656319006624636,
+      "loss": 5.1932,
+      "step": 653000
+    },
+    {
+      "epoch": 4.182801548948699,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003665375876084104,
+      "loss": 5.1947,
+      "step": 653500
+    },
+    {
+      "epoch": 4.186001856178193,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0003665119851505745,
+      "loss": 5.1889,
+      "step": 654000
+    },
+    {
+      "epoch": 4.189202163407687,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.0003664863826927385,
+      "loss": 5.1934,
+      "step": 654500
+    },
+    {
+      "epoch": 4.192402470637181,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003664607802349026,
+      "loss": 5.1974,
+      "step": 655000
+    },
+    {
+      "epoch": 4.195602777866675,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00036643517777706664,
+      "loss": 5.1908,
+      "step": 655500
+    },
+    {
+      "epoch": 4.198803085096169,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.00036640957531923067,
+      "loss": 5.1891,
+      "step": 656000
+    },
+    {
+      "epoch": 4.202003392325663,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003663839728613947,
+      "loss": 5.1873,
+      "step": 656500
+    },
+    {
+      "epoch": 4.205203699555157,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036635837040355874,
+      "loss": 5.1883,
+      "step": 657000
+    },
+    {
+      "epoch": 4.208404006784651,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00036633276794572283,
+      "loss": 5.1928,
+      "step": 657500
+    },
+    {
+      "epoch": 4.211604314014146,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00036630716548788686,
+      "loss": 5.1884,
+      "step": 658000
+    },
+    {
+      "epoch": 4.21480462124364,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003662815630300509,
+      "loss": 5.1933,
+      "step": 658500
+    },
+    {
+      "epoch": 4.218004928473134,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00036625596057221493,
+      "loss": 5.1924,
+      "step": 659000
+    },
+    {
+      "epoch": 4.2212052357026275,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036623035811437897,
+      "loss": 5.1907,
+      "step": 659500
+    },
+    {
+      "epoch": 4.2244055429321214,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00036620475565654306,
+      "loss": 5.1954,
+      "step": 660000
+    },
+    {
+      "epoch": 4.227605850161615,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.0003661791531987071,
+      "loss": 5.1918,
+      "step": 660500
+    },
+    {
+      "epoch": 4.230806157391109,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003661535507408711,
+      "loss": 5.1888,
+      "step": 661000
+    },
+    {
+      "epoch": 4.234006464620603,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00036612794828303516,
+      "loss": 5.1945,
+      "step": 661500
+    },
+    {
+      "epoch": 4.237206771850097,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00036610234582519925,
+      "loss": 5.1909,
+      "step": 662000
+    },
+    {
+      "epoch": 4.240407079079592,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003660767433673633,
+      "loss": 5.1898,
+      "step": 662500
+    },
+    {
+      "epoch": 4.243607386309086,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00036605114090952737,
+      "loss": 5.1976,
+      "step": 663000
+    },
+    {
+      "epoch": 4.24680769353858,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003660255384516914,
+      "loss": 5.1939,
+      "step": 663500
+    },
+    {
+      "epoch": 4.250008000768074,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00036599993599385544,
+      "loss": 5.1878,
+      "step": 664000
+    },
+    {
+      "epoch": 4.253208307997568,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00036597433353601947,
+      "loss": 5.1858,
+      "step": 664500
+    },
+    {
+      "epoch": 4.256408615227062,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003659487310781835,
+      "loss": 5.1895,
+      "step": 665000
+    },
+    {
+      "epoch": 4.259608922456556,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003659231286203476,
+      "loss": 5.1937,
+      "step": 665500
+    },
+    {
+      "epoch": 4.26280922968605,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00036589752616251163,
+      "loss": 5.1913,
+      "step": 666000
+    },
+    {
+      "epoch": 4.2660095369155435,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00036587192370467566,
+      "loss": 5.191,
+      "step": 666500
+    },
+    {
+      "epoch": 4.269209844145038,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003658463212468397,
+      "loss": 5.1996,
+      "step": 667000
+    },
+    {
+      "epoch": 4.272410151374532,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00036582071878900373,
+      "loss": 5.1936,
+      "step": 667500
+    },
+    {
+      "epoch": 4.275610458604026,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003657951163311678,
+      "loss": 5.1897,
+      "step": 668000
+    },
+    {
+      "epoch": 4.27881076583352,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00036576951387333185,
+      "loss": 5.1917,
+      "step": 668500
+    },
+    {
+      "epoch": 4.282011073063014,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003657439114154959,
+      "loss": 5.1918,
+      "step": 669000
+    },
+    {
+      "epoch": 4.285211380292508,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00036571830895766,
+      "loss": 5.1907,
+      "step": 669500
+    },
+    {
+      "epoch": 4.288411687522002,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.000365692706499824,
+      "loss": 5.1899,
+      "step": 670000
+    },
+    {
+      "epoch": 4.291611994751496,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00036566710404198805,
+      "loss": 5.1958,
+      "step": 670500
+    },
+    {
+      "epoch": 4.29481230198099,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00036564150158415213,
+      "loss": 5.1896,
+      "step": 671000
+    },
+    {
+      "epoch": 4.298012609210485,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00036561589912631617,
+      "loss": 5.197,
+      "step": 671500
+    },
+    {
+      "epoch": 4.301212916439979,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003655902966684802,
+      "loss": 5.1875,
+      "step": 672000
+    },
+    {
+      "epoch": 4.3044132236694725,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00036556469421064424,
+      "loss": 5.1866,
+      "step": 672500
+    },
+    {
+      "epoch": 4.3076135308989665,
+      "grad_norm": 0.41015625,
+      "learning_rate": 0.00036553909175280827,
+      "loss": 5.1938,
+      "step": 673000
+    },
+    {
+      "epoch": 4.31081383812846,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00036551348929497236,
+      "loss": 5.1941,
+      "step": 673500
+    },
+    {
+      "epoch": 4.314014145357954,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003654878868371364,
+      "loss": 5.1871,
+      "step": 674000
+    },
+    {
+      "epoch": 4.317214452587448,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00036546228437930043,
+      "loss": 5.194,
+      "step": 674500
+    },
+    {
+      "epoch": 4.320414759816942,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00036543668192146446,
+      "loss": 5.1944,
+      "step": 675000
+    },
+    {
+      "epoch": 4.323615067046436,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0003654110794636285,
+      "loss": 5.1897,
+      "step": 675500
+    },
+    {
+      "epoch": 4.32681537427593,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00036538547700579253,
+      "loss": 5.195,
+      "step": 676000
+    },
+    {
+      "epoch": 4.330015681505425,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0003653598745479566,
+      "loss": 5.1907,
+      "step": 676500
+    },
+    {
+      "epoch": 4.333215988734919,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036533427209012065,
+      "loss": 5.1874,
+      "step": 677000
+    },
+    {
+      "epoch": 4.336416295964413,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00036530866963228474,
+      "loss": 5.1906,
+      "step": 677500
+    },
+    {
+      "epoch": 4.339616603193907,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003652830671744488,
+      "loss": 5.1959,
+      "step": 678000
+    },
+    {
+      "epoch": 4.342816910423401,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003652574647166128,
+      "loss": 5.1941,
+      "step": 678500
+    },
+    {
+      "epoch": 4.346017217652895,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003652318622587769,
+      "loss": 5.1884,
+      "step": 679000
+    },
+    {
+      "epoch": 4.3492175248823886,
+      "grad_norm": 0.43359375,
+      "learning_rate": 0.00036520625980094093,
+      "loss": 5.19,
+      "step": 679500
+    },
+    {
+      "epoch": 4.3524178321118825,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00036518065734310497,
+      "loss": 5.1904,
+      "step": 680000
+    },
+    {
+      "epoch": 4.355618139341376,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.000365155054885269,
+      "loss": 5.1916,
+      "step": 680500
+    },
+    {
+      "epoch": 4.35881844657087,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00036512945242743304,
+      "loss": 5.1923,
+      "step": 681000
+    },
+    {
+      "epoch": 4.362018753800365,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00036510384996959707,
+      "loss": 5.1895,
+      "step": 681500
+    },
+    {
+      "epoch": 4.365219061029859,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00036507824751176116,
+      "loss": 5.1908,
+      "step": 682000
+    },
+    {
+      "epoch": 4.368419368259353,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.0003650526450539252,
+      "loss": 5.1916,
+      "step": 682500
+    },
+    {
+      "epoch": 4.371619675488847,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003650270425960892,
+      "loss": 5.1936,
+      "step": 683000
+    },
+    {
+      "epoch": 4.374819982718341,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00036500144013825326,
+      "loss": 5.1916,
+      "step": 683500
+    },
+    {
+      "epoch": 4.378020289947835,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003649758376804173,
+      "loss": 5.1896,
+      "step": 684000
+    },
+    {
+      "epoch": 4.381220597177329,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003649502352225814,
+      "loss": 5.1947,
+      "step": 684500
+    },
+    {
+      "epoch": 4.384420904406823,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003649246327647454,
+      "loss": 5.1908,
+      "step": 685000
+    },
+    {
+      "epoch": 4.387621211636317,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003648990303069095,
+      "loss": 5.1974,
+      "step": 685500
+    },
+    {
+      "epoch": 4.3908215188658115,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00036487342784907354,
+      "loss": 5.1908,
+      "step": 686000
+    },
+    {
+      "epoch": 4.3940218260953054,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003648478253912376,
+      "loss": 5.1929,
+      "step": 686500
+    },
+    {
+      "epoch": 4.397222133324799,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003648222229334016,
+      "loss": 5.1903,
+      "step": 687000
+    },
+    {
+      "epoch": 4.400422440554293,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0003647966204755657,
+      "loss": 5.1931,
+      "step": 687500
+    },
+    {
+      "epoch": 4.403622747783787,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00036477101801772973,
+      "loss": 5.192,
+      "step": 688000
+    },
+    {
+      "epoch": 4.406823055013281,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00036474541555989377,
+      "loss": 5.1902,
+      "step": 688500
+    },
+    {
+      "epoch": 4.410023362242775,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003647198131020578,
+      "loss": 5.1876,
+      "step": 689000
+    },
+    {
+      "epoch": 4.413223669472269,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00036469421064422183,
+      "loss": 5.1888,
+      "step": 689500
+    },
+    {
+      "epoch": 4.416423976701763,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003646686081863859,
+      "loss": 5.1934,
+      "step": 690000
+    },
+    {
+      "epoch": 4.419624283931258,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00036464300572854996,
+      "loss": 5.1933,
+      "step": 690500
+    },
+    {
+      "epoch": 4.422824591160752,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.000364617403270714,
+      "loss": 5.1894,
+      "step": 691000
+    },
+    {
+      "epoch": 4.426024898390246,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.000364591800812878,
+      "loss": 5.1912,
+      "step": 691500
+    },
+    {
+      "epoch": 4.42922520561974,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.0003645661983550421,
+      "loss": 5.1926,
+      "step": 692000
+    },
+    {
+      "epoch": 4.432425512849234,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00036454059589720615,
+      "loss": 5.1901,
+      "step": 692500
+    },
+    {
+      "epoch": 4.4356258200787275,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00036451499343937024,
+      "loss": 5.1939,
+      "step": 693000
+    },
+    {
+      "epoch": 4.4388261273082215,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00036448939098153427,
+      "loss": 5.1953,
+      "step": 693500
+    },
+    {
+      "epoch": 4.442026434537715,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003644637885236983,
+      "loss": 5.1922,
+      "step": 694000
+    },
+    {
+      "epoch": 4.445226741767209,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00036443818606586234,
+      "loss": 5.1988,
+      "step": 694500
+    },
+    {
+      "epoch": 4.448427048996703,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003644125836080264,
+      "loss": 5.1935,
+      "step": 695000
+    },
+    {
+      "epoch": 4.451627356226198,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.00036438698115019046,
+      "loss": 5.1963,
+      "step": 695500
+    },
+    {
+      "epoch": 4.454827663455692,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003643613786923545,
+      "loss": 5.1938,
+      "step": 696000
+    },
+    {
+      "epoch": 4.458027970685186,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00036433577623451853,
+      "loss": 5.1942,
+      "step": 696500
+    },
+    {
+      "epoch": 4.46122827791468,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00036431017377668256,
+      "loss": 5.1892,
+      "step": 697000
+    },
+    {
+      "epoch": 4.464428585144174,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003642845713188466,
+      "loss": 5.1923,
+      "step": 697500
+    },
+    {
+      "epoch": 4.467628892373668,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0003642589688610107,
+      "loss": 5.1957,
+      "step": 698000
+    },
+    {
+      "epoch": 4.470829199603162,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003642333664031747,
+      "loss": 5.1982,
+      "step": 698500
+    },
+    {
+      "epoch": 4.474029506832656,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00036420776394533876,
+      "loss": 5.194,
+      "step": 699000
+    },
+    {
+      "epoch": 4.47722981406215,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003641821614875028,
+      "loss": 5.1964,
+      "step": 699500
+    },
+    {
+      "epoch": 4.4804301212916435,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003641565590296669,
+      "loss": 5.1911,
+      "step": 700000
+    },
+    {
+      "epoch": 4.483630428521138,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.0003641309565718309,
+      "loss": 5.195,
+      "step": 700500
+    },
+    {
+      "epoch": 4.486830735750632,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.000364105354113995,
+      "loss": 5.1903,
+      "step": 701000
+    },
+    {
+      "epoch": 4.490031042980126,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00036407975165615904,
+      "loss": 5.2005,
+      "step": 701500
+    },
+    {
+      "epoch": 4.49323135020962,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00036405414919832307,
+      "loss": 5.1917,
+      "step": 702000
+    },
+    {
+      "epoch": 4.496431657439114,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003640285467404871,
+      "loss": 5.1933,
+      "step": 702500
+    },
+    {
+      "epoch": 4.499631964668608,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00036400294428265114,
+      "loss": 5.1886,
+      "step": 703000
+    },
+    {
+      "epoch": 4.502832271898102,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003639773418248152,
+      "loss": 5.1913,
+      "step": 703500
+    },
+    {
+      "epoch": 4.506032579127596,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00036395173936697926,
+      "loss": 5.2004,
+      "step": 704000
+    },
+    {
+      "epoch": 4.50923288635709,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003639261369091433,
+      "loss": 5.1908,
+      "step": 704500
+    },
+    {
+      "epoch": 4.512433193586585,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00036390053445130733,
+      "loss": 5.1889,
+      "step": 705000
+    },
+    {
+      "epoch": 4.515633500816079,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00036387493199347136,
+      "loss": 5.1943,
+      "step": 705500
+    },
+    {
+      "epoch": 4.5188338080455726,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003638493295356354,
+      "loss": 5.1924,
+      "step": 706000
+    },
+    {
+      "epoch": 4.5220341152750665,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003638237270777995,
+      "loss": 5.1945,
+      "step": 706500
+    },
+    {
+      "epoch": 4.52523442250456,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003637981246199635,
+      "loss": 5.1936,
+      "step": 707000
+    },
+    {
+      "epoch": 4.528434729734054,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.0003637725221621276,
+      "loss": 5.1932,
+      "step": 707500
+    },
+    {
+      "epoch": 4.531635036963548,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00036374691970429164,
+      "loss": 5.1967,
+      "step": 708000
+    },
+    {
+      "epoch": 4.534835344193042,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003637213172464557,
+      "loss": 5.1921,
+      "step": 708500
+    },
+    {
+      "epoch": 4.538035651422536,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00036369571478861977,
+      "loss": 5.1905,
+      "step": 709000
+    },
+    {
+      "epoch": 4.541235958652031,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003636701123307838,
+      "loss": 5.1932,
+      "step": 709500
+    },
+    {
+      "epoch": 4.544436265881525,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00036364450987294783,
+      "loss": 5.1955,
+      "step": 710000
+    },
+    {
+      "epoch": 4.547636573111019,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00036361890741511187,
+      "loss": 5.1927,
+      "step": 710500
+    },
+    {
+      "epoch": 4.550836880340513,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003635933049572759,
+      "loss": 5.1945,
+      "step": 711000
+    },
+    {
+      "epoch": 4.554037187570007,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00036356770249943994,
+      "loss": 5.1902,
+      "step": 711500
+    },
+    {
+      "epoch": 4.557237494799501,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.000363542100041604,
+      "loss": 5.1927,
+      "step": 712000
+    },
+    {
+      "epoch": 4.560437802028995,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00036351649758376806,
+      "loss": 5.1942,
+      "step": 712500
+    },
+    {
+      "epoch": 4.563638109258489,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.0003634908951259321,
+      "loss": 5.1962,
+      "step": 713000
+    },
+    {
+      "epoch": 4.5668384164879825,
+      "grad_norm": 0.65625,
+      "learning_rate": 0.00036346529266809613,
+      "loss": 5.1954,
+      "step": 713500
+    },
+    {
+      "epoch": 4.570038723717477,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00036343969021026016,
+      "loss": 5.1929,
+      "step": 714000
+    },
+    {
+      "epoch": 4.573239030946971,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00036341408775242425,
+      "loss": 5.1943,
+      "step": 714500
+    },
+    {
+      "epoch": 4.576439338176465,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003633884852945883,
+      "loss": 5.194,
+      "step": 715000
+    },
+    {
+      "epoch": 4.579639645405959,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003633628828367524,
+      "loss": 5.194,
+      "step": 715500
+    },
+    {
+      "epoch": 4.582839952635453,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003633372803789164,
+      "loss": 5.1945,
+      "step": 716000
+    },
+    {
+      "epoch": 4.586040259864947,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00036331167792108044,
+      "loss": 5.1895,
+      "step": 716500
+    },
+    {
+      "epoch": 4.589240567094441,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003632860754632445,
+      "loss": 5.1929,
+      "step": 717000
+    },
+    {
+      "epoch": 4.592440874323935,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00036326047300540857,
+      "loss": 5.1933,
+      "step": 717500
+    },
+    {
+      "epoch": 4.595641181553429,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003632348705475726,
+      "loss": 5.1961,
+      "step": 718000
+    },
+    {
+      "epoch": 4.598841488782924,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00036320926808973663,
+      "loss": 5.1955,
+      "step": 718500
+    },
+    {
+      "epoch": 4.602041796012417,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00036318366563190067,
+      "loss": 5.1968,
+      "step": 719000
+    },
+    {
+      "epoch": 4.6052421032419115,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003631580631740647,
+      "loss": 5.195,
+      "step": 719500
+    },
+    {
+      "epoch": 4.6084424104714055,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.0003631324607162288,
+      "loss": 5.1944,
+      "step": 720000
+    },
+    {
+      "epoch": 4.611642717700899,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003631068582583928,
+      "loss": 5.1937,
+      "step": 720500
+    },
+    {
+      "epoch": 4.614843024930393,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00036308125580055686,
+      "loss": 5.196,
+      "step": 721000
+    },
+    {
+      "epoch": 4.618043332159887,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003630556533427209,
+      "loss": 5.1996,
+      "step": 721500
+    },
+    {
+      "epoch": 4.621243639389381,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00036303005088488493,
+      "loss": 5.194,
+      "step": 722000
+    },
+    {
+      "epoch": 4.624443946618875,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000363004448427049,
+      "loss": 5.1924,
+      "step": 722500
+    },
+    {
+      "epoch": 4.627644253848369,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00036297884596921305,
+      "loss": 5.2027,
+      "step": 723000
+    },
+    {
+      "epoch": 4.630844561077863,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00036295324351137714,
+      "loss": 5.199,
+      "step": 723500
+    },
+    {
+      "epoch": 4.634044868307358,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00036292764105354117,
+      "loss": 5.1987,
+      "step": 724000
+    },
+    {
+      "epoch": 4.637245175536852,
+      "grad_norm": 0.41796875,
+      "learning_rate": 0.0003629020385957052,
+      "loss": 5.193,
+      "step": 724500
+    },
+    {
+      "epoch": 4.640445482766346,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00036287643613786924,
+      "loss": 5.1978,
+      "step": 725000
+    },
+    {
+      "epoch": 4.64364578999584,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00036285083368003333,
+      "loss": 5.1938,
+      "step": 725500
+    },
+    {
+      "epoch": 4.646846097225334,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00036282523122219736,
+      "loss": 5.1967,
+      "step": 726000
+    },
+    {
+      "epoch": 4.6500464044548275,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003627996287643614,
+      "loss": 5.1909,
+      "step": 726500
+    },
+    {
+      "epoch": 4.6532467116843215,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00036277402630652543,
+      "loss": 5.1929,
+      "step": 727000
+    },
+    {
+      "epoch": 4.656447018913815,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00036274842384868947,
+      "loss": 5.1954,
+      "step": 727500
+    },
+    {
+      "epoch": 4.659647326143309,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003627228213908535,
+      "loss": 5.1988,
+      "step": 728000
+    },
+    {
+      "epoch": 4.662847633372804,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003626972189330176,
+      "loss": 5.1884,
+      "step": 728500
+    },
+    {
+      "epoch": 4.666047940602298,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003626716164751816,
+      "loss": 5.1952,
+      "step": 729000
+    },
+    {
+      "epoch": 4.669248247831792,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.00036264601401734566,
+      "loss": 5.1953,
+      "step": 729500
+    },
+    {
+      "epoch": 4.672448555061286,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00036262041155950975,
+      "loss": 5.1932,
+      "step": 730000
+    },
+    {
+      "epoch": 4.67564886229078,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0003625948091016738,
+      "loss": 5.1955,
+      "step": 730500
+    },
+    {
+      "epoch": 4.678849169520274,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00036256920664383787,
+      "loss": 5.1975,
+      "step": 731000
+    },
+    {
+      "epoch": 4.682049476749768,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003625436041860019,
+      "loss": 5.1946,
+      "step": 731500
+    },
+    {
+      "epoch": 4.685249783979262,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00036251800172816594,
+      "loss": 5.1981,
+      "step": 732000
+    },
+    {
+      "epoch": 4.688450091208756,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00036249239927032997,
+      "loss": 5.1932,
+      "step": 732500
+    },
+    {
+      "epoch": 4.6916503984382505,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.000362466796812494,
+      "loss": 5.1963,
+      "step": 733000
+    },
+    {
+      "epoch": 4.694850705667744,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00036244119435465804,
+      "loss": 5.2029,
+      "step": 733500
+    },
+    {
+      "epoch": 4.698051012897238,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00036241559189682213,
+      "loss": 5.1952,
+      "step": 734000
+    },
+    {
+      "epoch": 4.701251320126732,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00036238998943898616,
+      "loss": 5.1905,
+      "step": 734500
+    },
+    {
+      "epoch": 4.704451627356226,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0003623643869811502,
+      "loss": 5.1941,
+      "step": 735000
+    },
+    {
+      "epoch": 4.70765193458572,
+      "grad_norm": 0.421875,
+      "learning_rate": 0.00036233878452331423,
+      "loss": 5.1893,
+      "step": 735500
+    },
+    {
+      "epoch": 4.710852241815214,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00036231318206547827,
+      "loss": 5.1955,
+      "step": 736000
+    },
+    {
+      "epoch": 4.714052549044708,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00036228757960764235,
+      "loss": 5.1985,
+      "step": 736500
+    },
+    {
+      "epoch": 4.717252856274202,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003622619771498064,
+      "loss": 5.1987,
+      "step": 737000
+    },
+    {
+      "epoch": 4.720453163503697,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003622363746919704,
+      "loss": 5.1945,
+      "step": 737500
+    },
+    {
+      "epoch": 4.723653470733191,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.0003622107722341345,
+      "loss": 5.1946,
+      "step": 738000
+    },
+    {
+      "epoch": 4.726853777962685,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00036218516977629855,
+      "loss": 5.1921,
+      "step": 738500
+    },
+    {
+      "epoch": 4.730054085192179,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00036215956731846263,
+      "loss": 5.1962,
+      "step": 739000
+    },
+    {
+      "epoch": 4.733254392421673,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00036213396486062667,
+      "loss": 5.1977,
+      "step": 739500
+    },
+    {
+      "epoch": 4.7364546996511665,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003621083624027907,
+      "loss": 5.1945,
+      "step": 740000
+    },
+    {
+      "epoch": 4.73965500688066,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00036208275994495474,
+      "loss": 5.1903,
+      "step": 740500
+    },
+    {
+      "epoch": 4.742855314110154,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00036205715748711877,
+      "loss": 5.2021,
+      "step": 741000
+    },
+    {
+      "epoch": 4.746055621339648,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003620315550292828,
+      "loss": 5.1963,
+      "step": 741500
+    },
+    {
+      "epoch": 4.749255928569143,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003620059525714469,
+      "loss": 5.1947,
+      "step": 742000
+    },
+    {
+      "epoch": 4.752456235798636,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00036198035011361093,
+      "loss": 5.1968,
+      "step": 742500
+    },
+    {
+      "epoch": 4.755656543028131,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00036195474765577496,
+      "loss": 5.1955,
+      "step": 743000
+    },
+    {
+      "epoch": 4.758856850257625,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.000361929145197939,
+      "loss": 5.1982,
+      "step": 743500
+    },
+    {
+      "epoch": 4.762057157487119,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00036190354274010303,
+      "loss": 5.1986,
+      "step": 744000
+    },
+    {
+      "epoch": 4.765257464716613,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003618779402822671,
+      "loss": 5.1994,
+      "step": 744500
+    },
+    {
+      "epoch": 4.768457771946107,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00036185233782443115,
+      "loss": 5.1965,
+      "step": 745000
+    },
+    {
+      "epoch": 4.771658079175601,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00036182673536659524,
+      "loss": 5.195,
+      "step": 745500
+    },
+    {
+      "epoch": 4.774858386405095,
+      "grad_norm": 0.423828125,
+      "learning_rate": 0.0003618011329087593,
+      "loss": 5.1981,
+      "step": 746000
+    },
+    {
+      "epoch": 4.778058693634589,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003617755304509233,
+      "loss": 5.1937,
+      "step": 746500
+    },
+    {
+      "epoch": 4.7812590008640825,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00036174992799308734,
+      "loss": 5.1932,
+      "step": 747000
+    },
+    {
+      "epoch": 4.784459308093577,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00036172432553525143,
+      "loss": 5.1944,
+      "step": 747500
+    },
+    {
+      "epoch": 4.787659615323071,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00036169872307741547,
+      "loss": 5.1957,
+      "step": 748000
+    },
+    {
+      "epoch": 4.790859922552565,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003616731206195795,
+      "loss": 5.1977,
+      "step": 748500
+    },
+    {
+      "epoch": 4.794060229782059,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00036164751816174354,
+      "loss": 5.2021,
+      "step": 749000
+    },
+    {
+      "epoch": 4.797260537011553,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00036162191570390757,
+      "loss": 5.1928,
+      "step": 749500
+    },
+    {
+      "epoch": 4.800460844241047,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00036159631324607166,
+      "loss": 5.1915,
+      "step": 750000
+    },
+    {
+      "epoch": 4.803661151470541,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003615707107882357,
+      "loss": 5.1967,
+      "step": 750500
+    },
+    {
+      "epoch": 4.806861458700035,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003615451083303997,
+      "loss": 5.1971,
+      "step": 751000
+    },
+    {
+      "epoch": 4.810061765929529,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036151950587256376,
+      "loss": 5.1977,
+      "step": 751500
+    },
+    {
+      "epoch": 4.813262073159024,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003614939034147278,
+      "loss": 5.1976,
+      "step": 752000
+    },
+    {
+      "epoch": 4.816462380388518,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003614683009568919,
+      "loss": 5.19,
+      "step": 752500
+    },
+    {
+      "epoch": 4.8196626876180115,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003614426984990559,
+      "loss": 5.1953,
+      "step": 753000
+    },
+    {
+      "epoch": 4.8228629948475055,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00036141709604122,
+      "loss": 5.2064,
+      "step": 753500
+    },
+    {
+      "epoch": 4.826063302076999,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00036139149358338404,
+      "loss": 5.1947,
+      "step": 754000
+    },
+    {
+      "epoch": 4.829263609306493,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003613658911255481,
+      "loss": 5.191,
+      "step": 754500
+    },
+    {
+      "epoch": 4.832463916535987,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0003613402886677121,
+      "loss": 5.1949,
+      "step": 755000
+    },
+    {
+      "epoch": 4.835664223765481,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003613146862098762,
+      "loss": 5.1965,
+      "step": 755500
+    },
+    {
+      "epoch": 4.838864530994975,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036128908375204023,
+      "loss": 5.1932,
+      "step": 756000
+    },
+    {
+      "epoch": 4.84206483822447,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.00036126348129420427,
+      "loss": 5.1951,
+      "step": 756500
+    },
+    {
+      "epoch": 4.845265145453964,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003612378788363683,
+      "loss": 5.1925,
+      "step": 757000
+    },
+    {
+      "epoch": 4.848465452683458,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00036121227637853233,
+      "loss": 5.1934,
+      "step": 757500
+    },
+    {
+      "epoch": 4.851665759912952,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00036118667392069637,
+      "loss": 5.1932,
+      "step": 758000
+    },
+    {
+      "epoch": 4.854866067142446,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00036116107146286046,
+      "loss": 5.1952,
+      "step": 758500
+    },
+    {
+      "epoch": 4.85806637437194,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003611354690050245,
+      "loss": 5.1937,
+      "step": 759000
+    },
+    {
+      "epoch": 4.861266681601434,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003611098665471885,
+      "loss": 5.1867,
+      "step": 759500
+    },
+    {
+      "epoch": 4.8644669888309275,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00036108426408935256,
+      "loss": 5.1896,
+      "step": 760000
+    },
+    {
+      "epoch": 4.8676672960604215,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00036105866163151665,
+      "loss": 5.1969,
+      "step": 760500
+    },
+    {
+      "epoch": 4.870867603289916,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00036103305917368074,
+      "loss": 5.1907,
+      "step": 761000
+    },
+    {
+      "epoch": 4.87406791051941,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00036100745671584477,
+      "loss": 5.1907,
+      "step": 761500
+    },
+    {
+      "epoch": 4.877268217748904,
+      "grad_norm": 0.7734375,
+      "learning_rate": 0.0003609818542580088,
+      "loss": 5.1944,
+      "step": 762000
+    },
+    {
+      "epoch": 4.880468524978398,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00036095625180017284,
+      "loss": 5.1974,
+      "step": 762500
+    },
+    {
+      "epoch": 4.883668832207892,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003609306493423369,
+      "loss": 5.1938,
+      "step": 763000
+    },
+    {
+      "epoch": 4.886869139437386,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003609050468845009,
+      "loss": 5.1975,
+      "step": 763500
+    },
+    {
+      "epoch": 4.89006944666688,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000360879444426665,
+      "loss": 5.1969,
+      "step": 764000
+    },
+    {
+      "epoch": 4.893269753896374,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00036085384196882903,
+      "loss": 5.1967,
+      "step": 764500
+    },
+    {
+      "epoch": 4.896470061125868,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00036082823951099306,
+      "loss": 5.1969,
+      "step": 765000
+    },
+    {
+      "epoch": 4.899670368355362,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.0003608026370531571,
+      "loss": 5.1924,
+      "step": 765500
+    },
+    {
+      "epoch": 4.902870675584856,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00036077703459532113,
+      "loss": 5.1943,
+      "step": 766000
+    },
+    {
+      "epoch": 4.9060709828143505,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003607514321374852,
+      "loss": 5.1916,
+      "step": 766500
+    },
+    {
+      "epoch": 4.909271290043844,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00036072582967964926,
+      "loss": 5.199,
+      "step": 767000
+    },
+    {
+      "epoch": 4.912471597273338,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003607002272218133,
+      "loss": 5.1961,
+      "step": 767500
+    },
+    {
+      "epoch": 4.915671904502832,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0003606746247639774,
+      "loss": 5.196,
+      "step": 768000
+    },
+    {
+      "epoch": 4.918872211732326,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003606490223061414,
+      "loss": 5.1949,
+      "step": 768500
+    },
+    {
+      "epoch": 4.92207251896182,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00036062341984830545,
+      "loss": 5.19,
+      "step": 769000
+    },
+    {
+      "epoch": 4.925272826191314,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.00036059781739046954,
+      "loss": 5.1954,
+      "step": 769500
+    },
+    {
+      "epoch": 4.928473133420808,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00036057221493263357,
+      "loss": 5.201,
+      "step": 770000
+    },
+    {
+      "epoch": 4.931673440650302,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.0003605466124747976,
+      "loss": 5.1932,
+      "step": 770500
+    },
+    {
+      "epoch": 4.934873747879797,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00036052101001696164,
+      "loss": 5.1932,
+      "step": 771000
+    },
+    {
+      "epoch": 4.938074055109291,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00036049540755912567,
+      "loss": 5.1997,
+      "step": 771500
+    },
+    {
+      "epoch": 4.941274362338785,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00036046980510128976,
+      "loss": 5.1931,
+      "step": 772000
+    },
+    {
+      "epoch": 4.944474669568279,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003604442026434538,
+      "loss": 5.1964,
+      "step": 772500
+    },
+    {
+      "epoch": 4.947674976797773,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00036041860018561783,
+      "loss": 5.1998,
+      "step": 773000
+    },
+    {
+      "epoch": 4.9508752840272665,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00036039299772778186,
+      "loss": 5.1968,
+      "step": 773500
+    },
+    {
+      "epoch": 4.9540755912567604,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.0003603673952699459,
+      "loss": 5.1856,
+      "step": 774000
+    },
+    {
+      "epoch": 4.957275898486254,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00036034179281211,
+      "loss": 5.1989,
+      "step": 774500
+    },
+    {
+      "epoch": 4.960476205715748,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.000360316190354274,
+      "loss": 5.1884,
+      "step": 775000
+    },
+    {
+      "epoch": 4.963676512945243,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00036029058789643806,
+      "loss": 5.1902,
+      "step": 775500
+    },
+    {
+      "epoch": 4.966876820174737,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00036026498543860214,
+      "loss": 5.1938,
+      "step": 776000
+    },
+    {
+      "epoch": 4.970077127404231,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003602393829807662,
+      "loss": 5.1943,
+      "step": 776500
+    },
+    {
+      "epoch": 4.973277434633725,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003602137805229302,
+      "loss": 5.1974,
+      "step": 777000
+    },
+    {
+      "epoch": 4.976477741863219,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003601881780650943,
+      "loss": 5.1969,
+      "step": 777500
+    },
+    {
+      "epoch": 4.979678049092713,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00036016257560725833,
+      "loss": 5.1962,
+      "step": 778000
+    },
+    {
+      "epoch": 4.982878356322207,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00036013697314942237,
+      "loss": 5.1981,
+      "step": 778500
+    },
+    {
+      "epoch": 4.986078663551701,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0003601113706915864,
+      "loss": 5.2013,
+      "step": 779000
+    },
+    {
+      "epoch": 4.989278970781195,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00036008576823375044,
+      "loss": 5.1841,
+      "step": 779500
+    },
+    {
+      "epoch": 4.9924792780106895,
+      "grad_norm": 0.4375,
+      "learning_rate": 0.0003600601657759145,
+      "loss": 5.194,
+      "step": 780000
+    },
+    {
+      "epoch": 4.995679585240183,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00036003456331807856,
+      "loss": 5.1985,
+      "step": 780500
+    },
+    {
+      "epoch": 4.998879892469677,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003600089608602426,
+      "loss": 5.1993,
+      "step": 781000
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 5.186727523803711,
+      "eval_runtime": 1.1486,
+      "eval_samples_per_second": 870.654,
+      "eval_steps_per_second": 13.93,
+      "step": 781175
+    },
+    {
+      "epoch": 5.002080199699171,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00035998335840240663,
+      "loss": 5.196,
+      "step": 781500
+    },
+    {
+      "epoch": 5.005280506928665,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00035995775594457066,
+      "loss": 5.1908,
+      "step": 782000
+    },
+    {
+      "epoch": 5.008480814158159,
+      "grad_norm": 0.447265625,
+      "learning_rate": 0.00035993215348673475,
+      "loss": 5.1907,
+      "step": 782500
+    },
+    {
+      "epoch": 5.011681121387653,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003599065510288988,
+      "loss": 5.1877,
+      "step": 783000
+    },
+    {
+      "epoch": 5.014881428617147,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003598809485710629,
+      "loss": 5.1926,
+      "step": 783500
+    },
+    {
+      "epoch": 5.018081735846641,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003598553461132269,
+      "loss": 5.1938,
+      "step": 784000
+    },
+    {
+      "epoch": 5.021282043076135,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00035982974365539094,
+      "loss": 5.1873,
+      "step": 784500
+    },
+    {
+      "epoch": 5.02448235030563,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.000359804141197555,
+      "loss": 5.1941,
+      "step": 785000
+    },
+    {
+      "epoch": 5.027682657535124,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00035977853873971907,
+      "loss": 5.1923,
+      "step": 785500
+    },
+    {
+      "epoch": 5.030882964764618,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003597529362818831,
+      "loss": 5.1884,
+      "step": 786000
+    },
+    {
+      "epoch": 5.0340832719941115,
+      "grad_norm": 0.4453125,
+      "learning_rate": 0.00035972733382404713,
+      "loss": 5.1959,
+      "step": 786500
+    },
+    {
+      "epoch": 5.0372835792236055,
+      "grad_norm": 0.451171875,
+      "learning_rate": 0.00035970173136621117,
+      "loss": 5.1889,
+      "step": 787000
+    },
+    {
+      "epoch": 5.040483886453099,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003596761289083752,
+      "loss": 5.195,
+      "step": 787500
+    },
+    {
+      "epoch": 5.043684193682593,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00035965052645053924,
+      "loss": 5.1857,
+      "step": 788000
+    },
+    {
+      "epoch": 5.046884500912087,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003596249239927033,
+      "loss": 5.1941,
+      "step": 788500
+    },
+    {
+      "epoch": 5.050084808141581,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00035959932153486736,
+      "loss": 5.1934,
+      "step": 789000
+    },
+    {
+      "epoch": 5.053285115371076,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0003595737190770314,
+      "loss": 5.1901,
+      "step": 789500
+    },
+    {
+      "epoch": 5.05648542260057,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00035954811661919543,
+      "loss": 5.1927,
+      "step": 790000
+    },
+    {
+      "epoch": 5.059685729830064,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003595225141613595,
+      "loss": 5.1878,
+      "step": 790500
+    },
+    {
+      "epoch": 5.062886037059558,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.00035949691170352355,
+      "loss": 5.1916,
+      "step": 791000
+    },
+    {
+      "epoch": 5.066086344289052,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00035947130924568764,
+      "loss": 5.1882,
+      "step": 791500
+    },
+    {
+      "epoch": 5.069286651518546,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003594457067878517,
+      "loss": 5.1905,
+      "step": 792000
+    },
+    {
+      "epoch": 5.07248695874804,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003594201043300157,
+      "loss": 5.1944,
+      "step": 792500
+    },
+    {
+      "epoch": 5.075687265977534,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00035939450187217974,
+      "loss": 5.1941,
+      "step": 793000
+    },
+    {
+      "epoch": 5.0788875732070276,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003593688994143438,
+      "loss": 5.1883,
+      "step": 793500
+    },
+    {
+      "epoch": 5.0820878804365215,
+      "grad_norm": 0.435546875,
+      "learning_rate": 0.00035934329695650786,
+      "loss": 5.1958,
+      "step": 794000
+    },
+    {
+      "epoch": 5.085288187666016,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003593176944986719,
+      "loss": 5.1933,
+      "step": 794500
+    },
+    {
+      "epoch": 5.08848849489551,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00035929209204083593,
+      "loss": 5.1934,
+      "step": 795000
+    },
+    {
+      "epoch": 5.091688802125004,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00035926648958299997,
+      "loss": 5.187,
+      "step": 795500
+    },
+    {
+      "epoch": 5.094889109354498,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.000359240887125164,
+      "loss": 5.199,
+      "step": 796000
+    },
+    {
+      "epoch": 5.098089416583992,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003592152846673281,
+      "loss": 5.186,
+      "step": 796500
+    },
+    {
+      "epoch": 5.101289723813486,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003591896822094921,
+      "loss": 5.1892,
+      "step": 797000
+    },
+    {
+      "epoch": 5.10449003104298,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035916407975165616,
+      "loss": 5.1931,
+      "step": 797500
+    },
+    {
+      "epoch": 5.107690338272474,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003591384772938202,
+      "loss": 5.1888,
+      "step": 798000
+    },
+    {
+      "epoch": 5.110890645501968,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003591128748359843,
+      "loss": 5.1965,
+      "step": 798500
+    },
+    {
+      "epoch": 5.114090952731463,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0003590872723781483,
+      "loss": 5.1959,
+      "step": 799000
+    },
+    {
+      "epoch": 5.117291259960957,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003590616699203124,
+      "loss": 5.1928,
+      "step": 799500
+    },
+    {
+      "epoch": 5.1204915671904505,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00035903606746247644,
+      "loss": 5.1896,
+      "step": 800000
+    },
+    {
+      "epoch": 5.1236918744199444,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035901046500464047,
+      "loss": 5.1934,
+      "step": 800500
+    },
+    {
+      "epoch": 5.126892181649438,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003589848625468045,
+      "loss": 5.1969,
+      "step": 801000
+    },
+    {
+      "epoch": 5.130092488878932,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00035895926008896854,
+      "loss": 5.1986,
+      "step": 801500
+    },
+    {
+      "epoch": 5.133292796108426,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035893365763113263,
+      "loss": 5.1884,
+      "step": 802000
+    },
+    {
+      "epoch": 5.13649310333792,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00035890805517329666,
+      "loss": 5.195,
+      "step": 802500
+    },
+    {
+      "epoch": 5.139693410567414,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003588824527154607,
+      "loss": 5.1907,
+      "step": 803000
+    },
+    {
+      "epoch": 5.142893717796909,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00035885685025762473,
+      "loss": 5.1878,
+      "step": 803500
+    },
+    {
+      "epoch": 5.146094025026403,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035883124779978877,
+      "loss": 5.1876,
+      "step": 804000
+    },
+    {
+      "epoch": 5.149294332255897,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003588056453419528,
+      "loss": 5.19,
+      "step": 804500
+    },
+    {
+      "epoch": 5.152494639485391,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003587800428841169,
+      "loss": 5.196,
+      "step": 805000
+    },
+    {
+      "epoch": 5.155694946714885,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003587544404262809,
+      "loss": 5.1937,
+      "step": 805500
+    },
+    {
+      "epoch": 5.158895253944379,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000358728837968445,
+      "loss": 5.191,
+      "step": 806000
+    },
+    {
+      "epoch": 5.162095561173873,
+      "grad_norm": 0.455078125,
+      "learning_rate": 0.00035870323551060905,
+      "loss": 5.191,
+      "step": 806500
+    },
+    {
+      "epoch": 5.1652958684033665,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003586776330527731,
+      "loss": 5.1922,
+      "step": 807000
+    },
+    {
+      "epoch": 5.1684961756328605,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00035865203059493717,
+      "loss": 5.1869,
+      "step": 807500
+    },
+    {
+      "epoch": 5.171696482862354,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003586264281371012,
+      "loss": 5.1968,
+      "step": 808000
+    },
+    {
+      "epoch": 5.174896790091849,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00035860082567926524,
+      "loss": 5.189,
+      "step": 808500
+    },
+    {
+      "epoch": 5.178097097321343,
+      "grad_norm": 0.44140625,
+      "learning_rate": 0.00035857522322142927,
+      "loss": 5.1945,
+      "step": 809000
+    },
+    {
+      "epoch": 5.181297404550837,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003585496207635933,
+      "loss": 5.1938,
+      "step": 809500
+    },
+    {
+      "epoch": 5.184497711780331,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003585240183057574,
+      "loss": 5.1902,
+      "step": 810000
+    },
+    {
+      "epoch": 5.187698019009825,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00035849841584792143,
+      "loss": 5.1895,
+      "step": 810500
+    },
+    {
+      "epoch": 5.190898326239319,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00035847281339008546,
+      "loss": 5.1867,
+      "step": 811000
+    },
+    {
+      "epoch": 5.194098633468813,
+      "grad_norm": 0.45703125,
+      "learning_rate": 0.0003584472109322495,
+      "loss": 5.1907,
+      "step": 811500
+    },
+    {
+      "epoch": 5.197298940698307,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00035842160847441353,
+      "loss": 5.1907,
+      "step": 812000
+    },
+    {
+      "epoch": 5.200499247927801,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00035839600601657756,
+      "loss": 5.192,
+      "step": 812500
+    },
+    {
+      "epoch": 5.203699555157295,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00035837040355874165,
+      "loss": 5.2016,
+      "step": 813000
+    },
+    {
+      "epoch": 5.2068998623867895,
+      "grad_norm": 0.443359375,
+      "learning_rate": 0.0003583448011009057,
+      "loss": 5.193,
+      "step": 813500
+    },
+    {
+      "epoch": 5.210100169616283,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003583191986430698,
+      "loss": 5.1969,
+      "step": 814000
+    },
+    {
+      "epoch": 5.213300476845777,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003582935961852338,
+      "loss": 5.193,
+      "step": 814500
+    },
+    {
+      "epoch": 5.216500784075271,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035826799372739784,
+      "loss": 5.1904,
+      "step": 815000
+    },
+    {
+      "epoch": 5.219701091304765,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00035824239126956193,
+      "loss": 5.1898,
+      "step": 815500
+    },
+    {
+      "epoch": 5.222901398534259,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00035821678881172597,
+      "loss": 5.1947,
+      "step": 816000
+    },
+    {
+      "epoch": 5.226101705763753,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00035819118635389,
+      "loss": 5.201,
+      "step": 816500
+    },
+    {
+      "epoch": 5.229302012993247,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00035816558389605404,
+      "loss": 5.195,
+      "step": 817000
+    },
+    {
+      "epoch": 5.232502320222741,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00035813998143821807,
+      "loss": 5.1934,
+      "step": 817500
+    },
+    {
+      "epoch": 5.235702627452236,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003581143789803821,
+      "loss": 5.1882,
+      "step": 818000
+    },
+    {
+      "epoch": 5.23890293468173,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003580887765225462,
+      "loss": 5.1984,
+      "step": 818500
+    },
+    {
+      "epoch": 5.242103241911224,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003580631740647102,
+      "loss": 5.1947,
+      "step": 819000
+    },
+    {
+      "epoch": 5.245303549140718,
+      "grad_norm": 0.44921875,
+      "learning_rate": 0.00035803757160687426,
+      "loss": 5.1932,
+      "step": 819500
+    },
+    {
+      "epoch": 5.2485038563702116,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003580119691490383,
+      "loss": 5.1865,
+      "step": 820000
+    },
+    {
+      "epoch": 5.2517041635997055,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0003579863666912024,
+      "loss": 5.1919,
+      "step": 820500
+    },
+    {
+      "epoch": 5.254904470829199,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003579607642333664,
+      "loss": 5.196,
+      "step": 821000
+    },
+    {
+      "epoch": 5.258104778058693,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003579351617755305,
+      "loss": 5.1858,
+      "step": 821500
+    },
+    {
+      "epoch": 5.261305085288187,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035790955931769454,
+      "loss": 5.1927,
+      "step": 822000
+    },
+    {
+      "epoch": 5.264505392517682,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003578839568598586,
+      "loss": 5.1908,
+      "step": 822500
+    },
+    {
+      "epoch": 5.267705699747176,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003578583544020226,
+      "loss": 5.1927,
+      "step": 823000
+    },
+    {
+      "epoch": 5.27090600697667,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00035783275194418664,
+      "loss": 5.1974,
+      "step": 823500
+    },
+    {
+      "epoch": 5.274106314206164,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00035780714948635073,
+      "loss": 5.1912,
+      "step": 824000
+    },
+    {
+      "epoch": 5.277306621435658,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00035778154702851477,
+      "loss": 5.1906,
+      "step": 824500
+    },
+    {
+      "epoch": 5.280506928665152,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003577559445706788,
+      "loss": 5.1898,
+      "step": 825000
+    },
+    {
+      "epoch": 5.283707235894646,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00035773034211284283,
+      "loss": 5.1904,
+      "step": 825500
+    },
+    {
+      "epoch": 5.28690754312414,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00035770473965500687,
+      "loss": 5.1934,
+      "step": 826000
+    },
+    {
+      "epoch": 5.290107850353634,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00035767913719717096,
+      "loss": 5.1942,
+      "step": 826500
+    },
+    {
+      "epoch": 5.2933081575831284,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000357653534739335,
+      "loss": 5.1926,
+      "step": 827000
+    },
+    {
+      "epoch": 5.296508464812622,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.000357627932281499,
+      "loss": 5.1936,
+      "step": 827500
+    },
+    {
+      "epoch": 5.299708772042116,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.00035760232982366306,
+      "loss": 5.1895,
+      "step": 828000
+    },
+    {
+      "epoch": 5.30290907927161,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00035757672736582715,
+      "loss": 5.1925,
+      "step": 828500
+    },
+    {
+      "epoch": 5.306109386501104,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003575511249079912,
+      "loss": 5.196,
+      "step": 829000
+    },
+    {
+      "epoch": 5.309309693730598,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00035752552245015527,
+      "loss": 5.1964,
+      "step": 829500
+    },
+    {
+      "epoch": 5.312510000960092,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003574999199923193,
+      "loss": 5.1984,
+      "step": 830000
+    },
+    {
+      "epoch": 5.315710308189586,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00035747431753448334,
+      "loss": 5.1861,
+      "step": 830500
+    },
+    {
+      "epoch": 5.31891061541908,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0003574487150766474,
+      "loss": 5.1971,
+      "step": 831000
+    },
+    {
+      "epoch": 5.322110922648574,
+      "grad_norm": 0.453125,
+      "learning_rate": 0.0003574231126188114,
+      "loss": 5.2024,
+      "step": 831500
+    },
+    {
+      "epoch": 5.325311229878069,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003573975101609755,
+      "loss": 5.194,
+      "step": 832000
+    },
+    {
+      "epoch": 5.328511537107563,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00035737190770313953,
+      "loss": 5.1953,
+      "step": 832500
+    },
+    {
+      "epoch": 5.331711844337057,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00035734630524530357,
+      "loss": 5.192,
+      "step": 833000
+    },
+    {
+      "epoch": 5.3349121515665505,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003573207027874676,
+      "loss": 5.1998,
+      "step": 833500
+    },
+    {
+      "epoch": 5.3381124587960445,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00035729510032963163,
+      "loss": 5.1896,
+      "step": 834000
+    },
+    {
+      "epoch": 5.341312766025538,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00035726949787179567,
+      "loss": 5.1879,
+      "step": 834500
+    },
+    {
+      "epoch": 5.344513073255032,
+      "grad_norm": 0.42578125,
+      "learning_rate": 0.00035724389541395976,
+      "loss": 5.194,
+      "step": 835000
+    },
+    {
+      "epoch": 5.347713380484526,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003572182929561238,
+      "loss": 5.193,
+      "step": 835500
+    },
+    {
+      "epoch": 5.35091368771402,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003571926904982878,
+      "loss": 5.1903,
+      "step": 836000
+    },
+    {
+      "epoch": 5.354113994943514,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003571670880404519,
+      "loss": 5.1892,
+      "step": 836500
+    },
+    {
+      "epoch": 5.357314302173009,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00035714148558261595,
+      "loss": 5.2005,
+      "step": 837000
+    },
+    {
+      "epoch": 5.360514609402503,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035711588312478004,
+      "loss": 5.1901,
+      "step": 837500
+    },
+    {
+      "epoch": 5.363714916631997,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00035709028066694407,
+      "loss": 5.1951,
+      "step": 838000
+    },
+    {
+      "epoch": 5.366915223861491,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0003570646782091081,
+      "loss": 5.189,
+      "step": 838500
+    },
+    {
+      "epoch": 5.370115531090985,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00035703907575127214,
+      "loss": 5.1957,
+      "step": 839000
+    },
+    {
+      "epoch": 5.373315838320479,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003570134732934362,
+      "loss": 5.1892,
+      "step": 839500
+    },
+    {
+      "epoch": 5.376516145549973,
+      "grad_norm": 0.4296875,
+      "learning_rate": 0.0003569878708356002,
+      "loss": 5.192,
+      "step": 840000
+    },
+    {
+      "epoch": 5.3797164527794665,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003569622683777643,
+      "loss": 5.1936,
+      "step": 840500
+    },
+    {
+      "epoch": 5.3829167600089605,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00035693666591992833,
+      "loss": 5.1927,
+      "step": 841000
+    },
+    {
+      "epoch": 5.386117067238455,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035691106346209236,
+      "loss": 5.1942,
+      "step": 841500
+    },
+    {
+      "epoch": 5.389317374467949,
+      "grad_norm": 0.439453125,
+      "learning_rate": 0.0003568854610042564,
+      "loss": 5.1916,
+      "step": 842000
+    },
+    {
+      "epoch": 5.392517681697443,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00035685985854642043,
+      "loss": 5.1909,
+      "step": 842500
+    },
+    {
+      "epoch": 5.395717988926937,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003568342560885845,
+      "loss": 5.1895,
+      "step": 843000
+    },
+    {
+      "epoch": 5.398918296156431,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00035680865363074856,
+      "loss": 5.196,
+      "step": 843500
+    },
+    {
+      "epoch": 5.402118603385925,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00035678305117291264,
+      "loss": 5.1934,
+      "step": 844000
+    },
+    {
+      "epoch": 5.405318910615419,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003567574487150767,
+      "loss": 5.193,
+      "step": 844500
+    },
+    {
+      "epoch": 5.408519217844913,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003567318462572407,
+      "loss": 5.1885,
+      "step": 845000
+    },
+    {
+      "epoch": 5.411719525074407,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.00035670624379940475,
+      "loss": 5.191,
+      "step": 845500
+    },
+    {
+      "epoch": 5.414919832303902,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035668064134156883,
+      "loss": 5.1924,
+      "step": 846000
+    },
+    {
+      "epoch": 5.4181201395333956,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00035665503888373287,
+      "loss": 5.1941,
+      "step": 846500
+    },
+    {
+      "epoch": 5.4213204467628895,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003566294364258969,
+      "loss": 5.1947,
+      "step": 847000
+    },
+    {
+      "epoch": 5.424520753992383,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00035660383396806094,
+      "loss": 5.1892,
+      "step": 847500
+    },
+    {
+      "epoch": 5.427721061221877,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00035657823151022497,
+      "loss": 5.1924,
+      "step": 848000
+    },
+    {
+      "epoch": 5.430921368451371,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00035655262905238906,
+      "loss": 5.1932,
+      "step": 848500
+    },
+    {
+      "epoch": 5.434121675680865,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003565270265945531,
+      "loss": 5.1959,
+      "step": 849000
+    },
+    {
+      "epoch": 5.437321982910359,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00035650142413671713,
+      "loss": 5.1937,
+      "step": 849500
+    },
+    {
+      "epoch": 5.440522290139853,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00035647582167888116,
+      "loss": 5.1952,
+      "step": 850000
+    },
+    {
+      "epoch": 5.443722597369347,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003564502192210452,
+      "loss": 5.1862,
+      "step": 850500
+    },
+    {
+      "epoch": 5.446922904598842,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003564246167632093,
+      "loss": 5.1865,
+      "step": 851000
+    },
+    {
+      "epoch": 5.450123211828336,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003563990143053733,
+      "loss": 5.1919,
+      "step": 851500
+    },
+    {
+      "epoch": 5.45332351905783,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0003563734118475374,
+      "loss": 5.1846,
+      "step": 852000
+    },
+    {
+      "epoch": 5.456523826287324,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035634780938970144,
+      "loss": 5.1946,
+      "step": 852500
+    },
+    {
+      "epoch": 5.459724133516818,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0003563222069318655,
+      "loss": 5.188,
+      "step": 853000
+    },
+    {
+      "epoch": 5.462924440746312,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003562966044740295,
+      "loss": 5.189,
+      "step": 853500
+    },
+    {
+      "epoch": 5.4661247479758055,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003562710020161936,
+      "loss": 5.1924,
+      "step": 854000
+    },
+    {
+      "epoch": 5.469325055205299,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00035624539955835763,
+      "loss": 5.1974,
+      "step": 854500
+    },
+    {
+      "epoch": 5.472525362434793,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00035621979710052167,
+      "loss": 5.19,
+      "step": 855000
+    },
+    {
+      "epoch": 5.475725669664287,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003561941946426857,
+      "loss": 5.1927,
+      "step": 855500
+    },
+    {
+      "epoch": 5.478925976893782,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035616859218484974,
+      "loss": 5.1937,
+      "step": 856000
+    },
+    {
+      "epoch": 5.482126284123276,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003561429897270138,
+      "loss": 5.1928,
+      "step": 856500
+    },
+    {
+      "epoch": 5.48532659135277,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035611738726917786,
+      "loss": 5.1935,
+      "step": 857000
+    },
+    {
+      "epoch": 5.488526898582264,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003560917848113419,
+      "loss": 5.1932,
+      "step": 857500
+    },
+    {
+      "epoch": 5.491727205811758,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035606618235350593,
+      "loss": 5.1861,
+      "step": 858000
+    },
+    {
+      "epoch": 5.494927513041252,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00035604057989567,
+      "loss": 5.1931,
+      "step": 858500
+    },
+    {
+      "epoch": 5.498127820270746,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00035601497743783405,
+      "loss": 5.1891,
+      "step": 859000
+    },
+    {
+      "epoch": 5.50132812750024,
+      "grad_norm": 0.470703125,
+      "learning_rate": 0.00035598937497999814,
+      "loss": 5.1924,
+      "step": 859500
+    },
+    {
+      "epoch": 5.504528434729734,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003559637725221622,
+      "loss": 5.1958,
+      "step": 860000
+    },
+    {
+      "epoch": 5.5077287419592285,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003559381700643262,
+      "loss": 5.1897,
+      "step": 860500
+    },
+    {
+      "epoch": 5.510929049188722,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035591256760649024,
+      "loss": 5.1907,
+      "step": 861000
+    },
+    {
+      "epoch": 5.514129356418216,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0003558869651486543,
+      "loss": 5.1888,
+      "step": 861500
+    },
+    {
+      "epoch": 5.51732966364771,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035586136269081836,
+      "loss": 5.1928,
+      "step": 862000
+    },
+    {
+      "epoch": 5.520529970877204,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0003558357602329824,
+      "loss": 5.1948,
+      "step": 862500
+    },
+    {
+      "epoch": 5.523730278106698,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035581015777514643,
+      "loss": 5.1949,
+      "step": 863000
+    },
+    {
+      "epoch": 5.526930585336192,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00035578455531731047,
+      "loss": 5.1971,
+      "step": 863500
+    },
+    {
+      "epoch": 5.530130892565686,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003557589528594745,
+      "loss": 5.189,
+      "step": 864000
+    },
+    {
+      "epoch": 5.53333119979518,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00035573335040163854,
+      "loss": 5.1916,
+      "step": 864500
+    },
+    {
+      "epoch": 5.536531507024675,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003557077479438026,
+      "loss": 5.1895,
+      "step": 865000
+    },
+    {
+      "epoch": 5.539731814254169,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00035568214548596666,
+      "loss": 5.1928,
+      "step": 865500
+    },
+    {
+      "epoch": 5.542932121483663,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003556565430281307,
+      "loss": 5.1955,
+      "step": 866000
+    },
+    {
+      "epoch": 5.546132428713157,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003556309405702948,
+      "loss": 5.1908,
+      "step": 866500
+    },
+    {
+      "epoch": 5.5493327359426505,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003556053381124588,
+      "loss": 5.1921,
+      "step": 867000
+    },
+    {
+      "epoch": 5.5525330431721445,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0003555797356546229,
+      "loss": 5.187,
+      "step": 867500
+    },
+    {
+      "epoch": 5.555733350401638,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00035555413319678694,
+      "loss": 5.1901,
+      "step": 868000
+    },
+    {
+      "epoch": 5.558933657631132,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00035552853073895097,
+      "loss": 5.1981,
+      "step": 868500
+    },
+    {
+      "epoch": 5.562133964860626,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.000355502928281115,
+      "loss": 5.1823,
+      "step": 869000
+    },
+    {
+      "epoch": 5.565334272090121,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035547732582327904,
+      "loss": 5.1946,
+      "step": 869500
+    },
+    {
+      "epoch": 5.568534579319615,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003554517233654431,
+      "loss": 5.1921,
+      "step": 870000
+    },
+    {
+      "epoch": 5.571734886549109,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00035542612090760716,
+      "loss": 5.1861,
+      "step": 870500
+    },
+    {
+      "epoch": 5.574935193778603,
+      "grad_norm": 0.47265625,
+      "learning_rate": 0.0003554005184497712,
+      "loss": 5.1876,
+      "step": 871000
+    },
+    {
+      "epoch": 5.578135501008097,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00035537491599193523,
+      "loss": 5.1953,
+      "step": 871500
+    },
+    {
+      "epoch": 5.581335808237591,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00035534931353409927,
+      "loss": 5.1937,
+      "step": 872000
+    },
+    {
+      "epoch": 5.584536115467085,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0003553237110762633,
+      "loss": 5.1894,
+      "step": 872500
+    },
+    {
+      "epoch": 5.587736422696579,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.0003552981086184274,
+      "loss": 5.1917,
+      "step": 873000
+    },
+    {
+      "epoch": 5.590936729926073,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0003552725061605914,
+      "loss": 5.1891,
+      "step": 873500
+    },
+    {
+      "epoch": 5.594137037155567,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0003552469037027555,
+      "loss": 5.1961,
+      "step": 874000
+    },
+    {
+      "epoch": 5.5973373443850605,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00035522130124491955,
+      "loss": 5.1954,
+      "step": 874500
+    },
+    {
+      "epoch": 5.600537651614555,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003551956987870836,
+      "loss": 5.1918,
+      "step": 875000
+    },
+    {
+      "epoch": 5.603737958844049,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003551700963292476,
+      "loss": 5.1932,
+      "step": 875500
+    },
+    {
+      "epoch": 5.606938266073543,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003551444938714117,
+      "loss": 5.2006,
+      "step": 876000
+    },
+    {
+      "epoch": 5.610138573303037,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00035511889141357574,
+      "loss": 5.1887,
+      "step": 876500
+    },
+    {
+      "epoch": 5.613338880532531,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00035509328895573977,
+      "loss": 5.1883,
+      "step": 877000
+    },
+    {
+      "epoch": 5.616539187762025,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0003550676864979038,
+      "loss": 5.1887,
+      "step": 877500
+    },
+    {
+      "epoch": 5.619739494991519,
+      "grad_norm": 0.466796875,
+      "learning_rate": 0.00035504208404006784,
+      "loss": 5.1867,
+      "step": 878000
+    },
+    {
+      "epoch": 5.622939802221013,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00035501648158223193,
+      "loss": 5.196,
+      "step": 878500
+    },
+    {
+      "epoch": 5.626140109450507,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00035499087912439596,
+      "loss": 5.1891,
+      "step": 879000
+    },
+    {
+      "epoch": 5.629340416680002,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00035496527666656,
+      "loss": 5.1935,
+      "step": 879500
+    },
+    {
+      "epoch": 5.632540723909496,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035493967420872403,
+      "loss": 5.1882,
+      "step": 880000
+    },
+    {
+      "epoch": 5.6357410311389895,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035491407175088807,
+      "loss": 5.1948,
+      "step": 880500
+    },
+    {
+      "epoch": 5.638941338368483,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.00035488846929305215,
+      "loss": 5.1945,
+      "step": 881000
+    },
+    {
+      "epoch": 5.642141645597977,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003548628668352162,
+      "loss": 5.1885,
+      "step": 881500
+    },
+    {
+      "epoch": 5.645341952827471,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003548372643773803,
+      "loss": 5.1935,
+      "step": 882000
+    },
+    {
+      "epoch": 5.648542260056965,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003548116619195443,
+      "loss": 5.1941,
+      "step": 882500
+    },
+    {
+      "epoch": 5.651742567286459,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00035478605946170834,
+      "loss": 5.1936,
+      "step": 883000
+    },
+    {
+      "epoch": 5.654942874515953,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003547604570038724,
+      "loss": 5.1873,
+      "step": 883500
+    },
+    {
+      "epoch": 5.658143181745448,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00035473485454603647,
+      "loss": 5.1939,
+      "step": 884000
+    },
+    {
+      "epoch": 5.661343488974942,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003547092520882005,
+      "loss": 5.1914,
+      "step": 884500
+    },
+    {
+      "epoch": 5.664543796204436,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00035468364963036454,
+      "loss": 5.1913,
+      "step": 885000
+    },
+    {
+      "epoch": 5.66774410343393,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035465804717252857,
+      "loss": 5.1865,
+      "step": 885500
+    },
+    {
+      "epoch": 5.670944410663424,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003546324447146926,
+      "loss": 5.191,
+      "step": 886000
+    },
+    {
+      "epoch": 5.674144717892918,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003546068422568567,
+      "loss": 5.1936,
+      "step": 886500
+    },
+    {
+      "epoch": 5.677345025122412,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.0003545812397990207,
+      "loss": 5.1917,
+      "step": 887000
+    },
+    {
+      "epoch": 5.6805453323519055,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00035455563734118476,
+      "loss": 5.1941,
+      "step": 887500
+    },
+    {
+      "epoch": 5.6837456395813994,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0003545300348833488,
+      "loss": 5.1957,
+      "step": 888000
+    },
+    {
+      "epoch": 5.686945946810894,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035450443242551283,
+      "loss": 5.1946,
+      "step": 888500
+    },
+    {
+      "epoch": 5.690146254040388,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003544788299676769,
+      "loss": 5.1872,
+      "step": 889000
+    },
+    {
+      "epoch": 5.693346561269882,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00035445322750984095,
+      "loss": 5.1903,
+      "step": 889500
+    },
+    {
+      "epoch": 5.696546868499376,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00035442762505200504,
+      "loss": 5.1979,
+      "step": 890000
+    },
+    {
+      "epoch": 5.69974717572887,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003544020225941691,
+      "loss": 5.1933,
+      "step": 890500
+    },
+    {
+      "epoch": 5.702947482958364,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003543764201363331,
+      "loss": 5.1966,
+      "step": 891000
+    },
+    {
+      "epoch": 5.706147790187858,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00035435081767849714,
+      "loss": 5.1889,
+      "step": 891500
+    },
+    {
+      "epoch": 5.709348097417352,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00035432521522066123,
+      "loss": 5.1904,
+      "step": 892000
+    },
+    {
+      "epoch": 5.712548404646846,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.00035429961276282527,
+      "loss": 5.1933,
+      "step": 892500
+    },
+    {
+      "epoch": 5.715748711876341,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003542740103049893,
+      "loss": 5.1877,
+      "step": 893000
+    },
+    {
+      "epoch": 5.7189490191058345,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00035424840784715333,
+      "loss": 5.1965,
+      "step": 893500
+    },
+    {
+      "epoch": 5.7221493263353285,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035422280538931737,
+      "loss": 5.1922,
+      "step": 894000
+    },
+    {
+      "epoch": 5.725349633564822,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003541972029314814,
+      "loss": 5.1867,
+      "step": 894500
+    },
+    {
+      "epoch": 5.728549940794316,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0003541716004736455,
+      "loss": 5.1915,
+      "step": 895000
+    },
+    {
+      "epoch": 5.73175024802381,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003541459980158095,
+      "loss": 5.1982,
+      "step": 895500
+    },
+    {
+      "epoch": 5.734950555253304,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035412039555797356,
+      "loss": 5.1919,
+      "step": 896000
+    },
+    {
+      "epoch": 5.738150862482798,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00035409479310013765,
+      "loss": 5.1929,
+      "step": 896500
+    },
+    {
+      "epoch": 5.741351169712292,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003540691906423017,
+      "loss": 5.1914,
+      "step": 897000
+    },
+    {
+      "epoch": 5.744551476941786,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00035404358818446577,
+      "loss": 5.1876,
+      "step": 897500
+    },
+    {
+      "epoch": 5.74775178417128,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003540179857266298,
+      "loss": 5.1937,
+      "step": 898000
+    },
+    {
+      "epoch": 5.750952091400775,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035399238326879384,
+      "loss": 5.194,
+      "step": 898500
+    },
+    {
+      "epoch": 5.754152398630269,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003539667808109579,
+      "loss": 5.1912,
+      "step": 899000
+    },
+    {
+      "epoch": 5.757352705859763,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003539411783531219,
+      "loss": 5.1917,
+      "step": 899500
+    },
+    {
+      "epoch": 5.760553013089257,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00035391557589528594,
+      "loss": 5.1971,
+      "step": 900000
+    },
+    {
+      "epoch": 5.7637533203187505,
+      "grad_norm": 0.458984375,
+      "learning_rate": 0.00035388997343745003,
+      "loss": 5.1915,
+      "step": 900500
+    },
+    {
+      "epoch": 5.7669536275482445,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00035386437097961407,
+      "loss": 5.1934,
+      "step": 901000
+    },
+    {
+      "epoch": 5.770153934777738,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003538387685217781,
+      "loss": 5.1906,
+      "step": 901500
+    },
+    {
+      "epoch": 5.773354242007232,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00035381316606394213,
+      "loss": 5.195,
+      "step": 902000
+    },
+    {
+      "epoch": 5.776554549236726,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00035378756360610617,
+      "loss": 5.1931,
+      "step": 902500
+    },
+    {
+      "epoch": 5.779754856466221,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035376196114827026,
+      "loss": 5.1899,
+      "step": 903000
+    },
+    {
+      "epoch": 5.782955163695715,
+      "grad_norm": 0.609375,
+      "learning_rate": 0.0003537363586904343,
+      "loss": 5.1867,
+      "step": 903500
+    },
+    {
+      "epoch": 5.786155470925209,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003537107562325983,
+      "loss": 5.1938,
+      "step": 904000
+    },
+    {
+      "epoch": 5.789355778154703,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003536851537747624,
+      "loss": 5.1991,
+      "step": 904500
+    },
+    {
+      "epoch": 5.792556085384197,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035365955131692645,
+      "loss": 5.1889,
+      "step": 905000
+    },
+    {
+      "epoch": 5.795756392613691,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003536339488590905,
+      "loss": 5.1888,
+      "step": 905500
+    },
+    {
+      "epoch": 5.798956699843185,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035360834640125457,
+      "loss": 5.1912,
+      "step": 906000
+    },
+    {
+      "epoch": 5.802157007072679,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0003535827439434186,
+      "loss": 5.1871,
+      "step": 906500
+    },
+    {
+      "epoch": 5.805357314302173,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00035355714148558264,
+      "loss": 5.198,
+      "step": 907000
+    },
+    {
+      "epoch": 5.808557621531667,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0003535315390277467,
+      "loss": 5.1929,
+      "step": 907500
+    },
+    {
+      "epoch": 5.811757928761161,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003535059365699107,
+      "loss": 5.1934,
+      "step": 908000
+    },
+    {
+      "epoch": 5.814958235990655,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003534803341120748,
+      "loss": 5.1958,
+      "step": 908500
+    },
+    {
+      "epoch": 5.818158543220149,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00035345473165423883,
+      "loss": 5.19,
+      "step": 909000
+    },
+    {
+      "epoch": 5.821358850449643,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00035342912919640286,
+      "loss": 5.1926,
+      "step": 909500
+    },
+    {
+      "epoch": 5.824559157679137,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003534035267385669,
+      "loss": 5.196,
+      "step": 910000
+    },
+    {
+      "epoch": 5.827759464908631,
+      "grad_norm": 0.46484375,
+      "learning_rate": 0.00035337792428073093,
+      "loss": 5.1921,
+      "step": 910500
+    },
+    {
+      "epoch": 5.830959772138125,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00035335232182289497,
+      "loss": 5.1961,
+      "step": 911000
+    },
+    {
+      "epoch": 5.834160079367619,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00035332671936505906,
+      "loss": 5.1909,
+      "step": 911500
+    },
+    {
+      "epoch": 5.837360386597114,
+      "grad_norm": 0.6328125,
+      "learning_rate": 0.00035330111690722314,
+      "loss": 5.1908,
+      "step": 912000
+    },
+    {
+      "epoch": 5.840560693826608,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003532755144493872,
+      "loss": 5.199,
+      "step": 912500
+    },
+    {
+      "epoch": 5.843761001056102,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003532499119915512,
+      "loss": 5.1974,
+      "step": 913000
+    },
+    {
+      "epoch": 5.846961308285596,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035322430953371525,
+      "loss": 5.1941,
+      "step": 913500
+    },
+    {
+      "epoch": 5.8501616155150895,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035319870707587934,
+      "loss": 5.1948,
+      "step": 914000
+    },
+    {
+      "epoch": 5.8533619227445834,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00035317310461804337,
+      "loss": 5.1903,
+      "step": 914500
+    },
+    {
+      "epoch": 5.856562229974077,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003531475021602074,
+      "loss": 5.1971,
+      "step": 915000
+    },
+    {
+      "epoch": 5.859762537203571,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00035312189970237144,
+      "loss": 5.1977,
+      "step": 915500
+    },
+    {
+      "epoch": 5.862962844433065,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00035309629724453547,
+      "loss": 5.1835,
+      "step": 916000
+    },
+    {
+      "epoch": 5.86616315166256,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003530706947866995,
+      "loss": 5.1871,
+      "step": 916500
+    },
+    {
+      "epoch": 5.869363458892054,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003530450923288636,
+      "loss": 5.1921,
+      "step": 917000
+    },
+    {
+      "epoch": 5.872563766121548,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.00035301948987102763,
+      "loss": 5.1888,
+      "step": 917500
+    },
+    {
+      "epoch": 5.875764073351042,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00035299388741319166,
+      "loss": 5.1879,
+      "step": 918000
+    },
+    {
+      "epoch": 5.878964380580536,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0003529682849553557,
+      "loss": 5.1944,
+      "step": 918500
+    },
+    {
+      "epoch": 5.88216468781003,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003529426824975198,
+      "loss": 5.1927,
+      "step": 919000
+    },
+    {
+      "epoch": 5.885364995039524,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003529170800396838,
+      "loss": 5.1947,
+      "step": 919500
+    },
+    {
+      "epoch": 5.888565302269018,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003528914775818479,
+      "loss": 5.1938,
+      "step": 920000
+    },
+    {
+      "epoch": 5.891765609498512,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035286587512401194,
+      "loss": 5.1885,
+      "step": 920500
+    },
+    {
+      "epoch": 5.8949659167280055,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.000352840272666176,
+      "loss": 5.193,
+      "step": 921000
+    },
+    {
+      "epoch": 5.8981662239574995,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00035281467020834,
+      "loss": 5.1907,
+      "step": 921500
+    },
+    {
+      "epoch": 5.901366531186994,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00035278906775050405,
+      "loss": 5.1884,
+      "step": 922000
+    },
+    {
+      "epoch": 5.904566838416488,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00035276346529266813,
+      "loss": 5.1923,
+      "step": 922500
+    },
+    {
+      "epoch": 5.907767145645982,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00035273786283483217,
+      "loss": 5.1928,
+      "step": 923000
+    },
+    {
+      "epoch": 5.910967452875476,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003527122603769962,
+      "loss": 5.1888,
+      "step": 923500
+    },
+    {
+      "epoch": 5.91416776010497,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00035268665791916024,
+      "loss": 5.1952,
+      "step": 924000
+    },
+    {
+      "epoch": 5.917368067334464,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00035266105546132427,
+      "loss": 5.1968,
+      "step": 924500
+    },
+    {
+      "epoch": 5.920568374563958,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035263545300348836,
+      "loss": 5.1912,
+      "step": 925000
+    },
+    {
+      "epoch": 5.923768681793452,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0003526098505456524,
+      "loss": 5.1918,
+      "step": 925500
+    },
+    {
+      "epoch": 5.926968989022946,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00035258424808781643,
+      "loss": 5.1855,
+      "step": 926000
+    },
+    {
+      "epoch": 5.930169296252441,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00035255864562998046,
+      "loss": 5.1912,
+      "step": 926500
+    },
+    {
+      "epoch": 5.9333696034819345,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.00035253304317214455,
+      "loss": 5.1881,
+      "step": 927000
+    },
+    {
+      "epoch": 5.9365699107114285,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00035250744071430864,
+      "loss": 5.1856,
+      "step": 927500
+    },
+    {
+      "epoch": 5.939770217940922,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003524818382564727,
+      "loss": 5.1917,
+      "step": 928000
+    },
+    {
+      "epoch": 5.942970525170416,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003524562357986367,
+      "loss": 5.1948,
+      "step": 928500
+    },
+    {
+      "epoch": 5.94617083239991,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00035243063334080074,
+      "loss": 5.1927,
+      "step": 929000
+    },
+    {
+      "epoch": 5.949371139629404,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0003524050308829648,
+      "loss": 5.1937,
+      "step": 929500
+    },
+    {
+      "epoch": 5.952571446858898,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003523794284251288,
+      "loss": 5.189,
+      "step": 930000
+    },
+    {
+      "epoch": 5.955771754088392,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003523538259672929,
+      "loss": 5.1947,
+      "step": 930500
+    },
+    {
+      "epoch": 5.958972061317887,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.00035232822350945693,
+      "loss": 5.1924,
+      "step": 931000
+    },
+    {
+      "epoch": 5.962172368547381,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00035230262105162097,
+      "loss": 5.193,
+      "step": 931500
+    },
+    {
+      "epoch": 5.965372675776875,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.000352277018593785,
+      "loss": 5.1917,
+      "step": 932000
+    },
+    {
+      "epoch": 5.968572983006369,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00035225141613594904,
+      "loss": 5.1942,
+      "step": 932500
+    },
+    {
+      "epoch": 5.971773290235863,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003522258136781131,
+      "loss": 5.1945,
+      "step": 933000
+    },
+    {
+      "epoch": 5.974973597465357,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035220021122027716,
+      "loss": 5.1909,
+      "step": 933500
+    },
+    {
+      "epoch": 5.9781739046948505,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0003521746087624412,
+      "loss": 5.1906,
+      "step": 934000
+    },
+    {
+      "epoch": 5.9813742119243445,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003521490063046053,
+      "loss": 5.1929,
+      "step": 934500
+    },
+    {
+      "epoch": 5.984574519153838,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0003521234038467693,
+      "loss": 5.1918,
+      "step": 935000
+    },
+    {
+      "epoch": 5.987774826383333,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00035209780138893335,
+      "loss": 5.1913,
+      "step": 935500
+    },
+    {
+      "epoch": 5.990975133612827,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00035207219893109744,
+      "loss": 5.1875,
+      "step": 936000
+    },
+    {
+      "epoch": 5.994175440842321,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00035204659647326147,
+      "loss": 5.1937,
+      "step": 936500
+    },
+    {
+      "epoch": 5.997375748071815,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003520209940154255,
+      "loss": 5.1945,
+      "step": 937000
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 5.182580947875977,
+      "eval_runtime": 1.1238,
+      "eval_samples_per_second": 889.838,
+      "eval_steps_per_second": 14.237,
+      "step": 937410
+    },
+    {
+      "epoch": 6.000576055301309,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.00035199539155758954,
+      "loss": 5.1917,
+      "step": 937500
+    },
+    {
+      "epoch": 6.003776362530803,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003519697890997536,
+      "loss": 5.1882,
+      "step": 938000
+    },
+    {
+      "epoch": 6.006976669760297,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00035194418664191766,
+      "loss": 5.1892,
+      "step": 938500
+    },
+    {
+      "epoch": 6.010176976989791,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0003519185841840817,
+      "loss": 5.193,
+      "step": 939000
+    },
+    {
+      "epoch": 6.013377284219285,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00035189298172624573,
+      "loss": 5.1884,
+      "step": 939500
+    },
+    {
+      "epoch": 6.016577591448779,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00035186737926840977,
+      "loss": 5.1862,
+      "step": 940000
+    },
+    {
+      "epoch": 6.0197778986782735,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003518417768105738,
+      "loss": 5.1902,
+      "step": 940500
+    },
+    {
+      "epoch": 6.0229782059077674,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00035181617435273783,
+      "loss": 5.1893,
+      "step": 941000
+    },
+    {
+      "epoch": 6.026178513137261,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003517905718949019,
+      "loss": 5.1898,
+      "step": 941500
+    },
+    {
+      "epoch": 6.029378820366755,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035176496943706596,
+      "loss": 5.1912,
+      "step": 942000
+    },
+    {
+      "epoch": 6.032579127596249,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035173936697923005,
+      "loss": 5.1923,
+      "step": 942500
+    },
+    {
+      "epoch": 6.035779434825743,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0003517137645213941,
+      "loss": 5.1908,
+      "step": 943000
+    },
+    {
+      "epoch": 6.038979742055237,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003516881620635581,
+      "loss": 5.1907,
+      "step": 943500
+    },
+    {
+      "epoch": 6.042180049284731,
+      "grad_norm": 0.462890625,
+      "learning_rate": 0.0003516625596057222,
+      "loss": 5.1906,
+      "step": 944000
+    },
+    {
+      "epoch": 6.045380356514225,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035163695714788624,
+      "loss": 5.1924,
+      "step": 944500
+    },
+    {
+      "epoch": 6.04858066374372,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035161135469005027,
+      "loss": 5.1851,
+      "step": 945000
+    },
+    {
+      "epoch": 6.051780970973214,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003515857522322143,
+      "loss": 5.1892,
+      "step": 945500
+    },
+    {
+      "epoch": 6.054981278202708,
+      "grad_norm": 0.62109375,
+      "learning_rate": 0.00035156014977437834,
+      "loss": 5.1903,
+      "step": 946000
+    },
+    {
+      "epoch": 6.058181585432202,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0003515345473165424,
+      "loss": 5.1901,
+      "step": 946500
+    },
+    {
+      "epoch": 6.061381892661696,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00035150894485870646,
+      "loss": 5.1916,
+      "step": 947000
+    },
+    {
+      "epoch": 6.0645821998911895,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003514833424008705,
+      "loss": 5.1928,
+      "step": 947500
+    },
+    {
+      "epoch": 6.0677825071206835,
+      "grad_norm": 0.486328125,
+      "learning_rate": 0.00035145773994303453,
+      "loss": 5.194,
+      "step": 948000
+    },
+    {
+      "epoch": 6.070982814350177,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035143213748519857,
+      "loss": 5.1887,
+      "step": 948500
+    },
+    {
+      "epoch": 6.074183121579671,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00035140653502736265,
+      "loss": 5.1883,
+      "step": 949000
+    },
+    {
+      "epoch": 6.077383428809165,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003513809325695267,
+      "loss": 5.1863,
+      "step": 949500
+    },
+    {
+      "epoch": 6.08058373603866,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003513553301116908,
+      "loss": 5.1892,
+      "step": 950000
+    },
+    {
+      "epoch": 6.083784043268154,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0003513297276538548,
+      "loss": 5.1887,
+      "step": 950500
+    },
+    {
+      "epoch": 6.086984350497648,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035130412519601884,
+      "loss": 5.1915,
+      "step": 951000
+    },
+    {
+      "epoch": 6.090184657727142,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0003512785227381829,
+      "loss": 5.1833,
+      "step": 951500
+    },
+    {
+      "epoch": 6.093384964956636,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0003512529202803469,
+      "loss": 5.1895,
+      "step": 952000
+    },
+    {
+      "epoch": 6.09658527218613,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.000351227317822511,
+      "loss": 5.1865,
+      "step": 952500
+    },
+    {
+      "epoch": 6.099785579415624,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035120171536467504,
+      "loss": 5.1873,
+      "step": 953000
+    },
+    {
+      "epoch": 6.102985886645118,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00035117611290683907,
+      "loss": 5.1876,
+      "step": 953500
+    },
+    {
+      "epoch": 6.106186193874612,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003511505104490031,
+      "loss": 5.1868,
+      "step": 954000
+    },
+    {
+      "epoch": 6.109386501104106,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00035112490799116714,
+      "loss": 5.1895,
+      "step": 954500
+    },
+    {
+      "epoch": 6.1125868083336,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00035109930553333123,
+      "loss": 5.1849,
+      "step": 955000
+    },
+    {
+      "epoch": 6.115787115563094,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035107370307549526,
+      "loss": 5.1908,
+      "step": 955500
+    },
+    {
+      "epoch": 6.118987422792588,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003510481006176593,
+      "loss": 5.1902,
+      "step": 956000
+    },
+    {
+      "epoch": 6.122187730022082,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00035102249815982333,
+      "loss": 5.1848,
+      "step": 956500
+    },
+    {
+      "epoch": 6.125388037251576,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003509968957019874,
+      "loss": 5.1872,
+      "step": 957000
+    },
+    {
+      "epoch": 6.12858834448107,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00035097129324415145,
+      "loss": 5.1885,
+      "step": 957500
+    },
+    {
+      "epoch": 6.131788651710564,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035094569078631554,
+      "loss": 5.1939,
+      "step": 958000
+    },
+    {
+      "epoch": 6.134988958940058,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.0003509200883284796,
+      "loss": 5.1867,
+      "step": 958500
+    },
+    {
+      "epoch": 6.138189266169552,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003508944858706436,
+      "loss": 5.1911,
+      "step": 959000
+    },
+    {
+      "epoch": 6.141389573399047,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00035086888341280764,
+      "loss": 5.1907,
+      "step": 959500
+    },
+    {
+      "epoch": 6.144589880628541,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0003508432809549717,
+      "loss": 5.1826,
+      "step": 960000
+    },
+    {
+      "epoch": 6.1477901878580345,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00035081767849713577,
+      "loss": 5.1895,
+      "step": 960500
+    },
+    {
+      "epoch": 6.1509904950875285,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003507920760392998,
+      "loss": 5.1838,
+      "step": 961000
+    },
+    {
+      "epoch": 6.154190802317022,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00035076647358146384,
+      "loss": 5.189,
+      "step": 961500
+    },
+    {
+      "epoch": 6.157391109546516,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00035074087112362787,
+      "loss": 5.197,
+      "step": 962000
+    },
+    {
+      "epoch": 6.16059141677601,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003507152686657919,
+      "loss": 5.1883,
+      "step": 962500
+    },
+    {
+      "epoch": 6.163791724005504,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.000350689666207956,
+      "loss": 5.1886,
+      "step": 963000
+    },
+    {
+      "epoch": 6.166992031234998,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00035066406375012,
+      "loss": 5.1899,
+      "step": 963500
+    },
+    {
+      "epoch": 6.170192338464493,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00035063846129228406,
+      "loss": 5.188,
+      "step": 964000
+    },
+    {
+      "epoch": 6.173392645693987,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0003506128588344481,
+      "loss": 5.1878,
+      "step": 964500
+    },
+    {
+      "epoch": 6.176592952923481,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0003505872563766122,
+      "loss": 5.1893,
+      "step": 965000
+    },
+    {
+      "epoch": 6.179793260152975,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003505616539187762,
+      "loss": 5.1954,
+      "step": 965500
+    },
+    {
+      "epoch": 6.182993567382469,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003505360514609403,
+      "loss": 5.1895,
+      "step": 966000
+    },
+    {
+      "epoch": 6.186193874611963,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00035051044900310434,
+      "loss": 5.1938,
+      "step": 966500
+    },
+    {
+      "epoch": 6.189394181841457,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003504848465452684,
+      "loss": 5.1846,
+      "step": 967000
+    },
+    {
+      "epoch": 6.192594489070951,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003504592440874324,
+      "loss": 5.1904,
+      "step": 967500
+    },
+    {
+      "epoch": 6.1957947963004445,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00035043364162959644,
+      "loss": 5.1925,
+      "step": 968000
+    },
+    {
+      "epoch": 6.198995103529938,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00035040803917176053,
+      "loss": 5.1875,
+      "step": 968500
+    },
+    {
+      "epoch": 6.202195410759433,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00035038243671392457,
+      "loss": 5.1934,
+      "step": 969000
+    },
+    {
+      "epoch": 6.205395717988927,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003503568342560886,
+      "loss": 5.1878,
+      "step": 969500
+    },
+    {
+      "epoch": 6.208596025218421,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00035033123179825263,
+      "loss": 5.1826,
+      "step": 970000
+    },
+    {
+      "epoch": 6.211796332447915,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00035030562934041667,
+      "loss": 5.1873,
+      "step": 970500
+    },
+    {
+      "epoch": 6.214996639677409,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003502800268825807,
+      "loss": 5.1948,
+      "step": 971000
+    },
+    {
+      "epoch": 6.218196946906903,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003502544244247448,
+      "loss": 5.1897,
+      "step": 971500
+    },
+    {
+      "epoch": 6.221397254136397,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003502288219669088,
+      "loss": 5.19,
+      "step": 972000
+    },
+    {
+      "epoch": 6.224597561365891,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0003502032195090729,
+      "loss": 5.1866,
+      "step": 972500
+    },
+    {
+      "epoch": 6.227797868595385,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00035017761705123695,
+      "loss": 5.1908,
+      "step": 973000
+    },
+    {
+      "epoch": 6.23099817582488,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000350152014593401,
+      "loss": 5.1909,
+      "step": 973500
+    },
+    {
+      "epoch": 6.2341984830543735,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00035012641213556507,
+      "loss": 5.1882,
+      "step": 974000
+    },
+    {
+      "epoch": 6.2373987902838675,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0003501008096777291,
+      "loss": 5.1924,
+      "step": 974500
+    },
+    {
+      "epoch": 6.240599097513361,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00035007520721989314,
+      "loss": 5.1889,
+      "step": 975000
+    },
+    {
+      "epoch": 6.243799404742855,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003500496047620572,
+      "loss": 5.1881,
+      "step": 975500
+    },
+    {
+      "epoch": 6.246999711972349,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0003500240023042212,
+      "loss": 5.188,
+      "step": 976000
+    },
+    {
+      "epoch": 6.250200019201843,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00034999839984638524,
+      "loss": 5.1906,
+      "step": 976500
+    },
+    {
+      "epoch": 6.253400326431337,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00034997279738854933,
+      "loss": 5.184,
+      "step": 977000
+    },
+    {
+      "epoch": 6.256600633660831,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00034994719493071336,
+      "loss": 5.1946,
+      "step": 977500
+    },
+    {
+      "epoch": 6.259800940890326,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003499215924728774,
+      "loss": 5.1867,
+      "step": 978000
+    },
+    {
+      "epoch": 6.26300124811982,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00034989599001504143,
+      "loss": 5.1909,
+      "step": 978500
+    },
+    {
+      "epoch": 6.266201555349314,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00034987038755720547,
+      "loss": 5.192,
+      "step": 979000
+    },
+    {
+      "epoch": 6.269401862578808,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00034984478509936956,
+      "loss": 5.1862,
+      "step": 979500
+    },
+    {
+      "epoch": 6.272602169808302,
+      "grad_norm": 0.478515625,
+      "learning_rate": 0.0003498191826415336,
+      "loss": 5.1925,
+      "step": 980000
+    },
+    {
+      "epoch": 6.275802477037796,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0003497935801836977,
+      "loss": 5.1927,
+      "step": 980500
+    },
+    {
+      "epoch": 6.2790027842672895,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003497679777258617,
+      "loss": 5.1869,
+      "step": 981000
+    },
+    {
+      "epoch": 6.2822030914967835,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00034974237526802575,
+      "loss": 5.1929,
+      "step": 981500
+    },
+    {
+      "epoch": 6.285403398726277,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003497167728101898,
+      "loss": 5.1874,
+      "step": 982000
+    },
+    {
+      "epoch": 6.288603705955771,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00034969117035235387,
+      "loss": 5.1839,
+      "step": 982500
+    },
+    {
+      "epoch": 6.291804013185266,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0003496655678945179,
+      "loss": 5.1902,
+      "step": 983000
+    },
+    {
+      "epoch": 6.29500432041476,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00034963996543668194,
+      "loss": 5.1861,
+      "step": 983500
+    },
+    {
+      "epoch": 6.298204627644254,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00034961436297884597,
+      "loss": 5.1926,
+      "step": 984000
+    },
+    {
+      "epoch": 6.301404934873748,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.00034958876052101,
+      "loss": 5.1846,
+      "step": 984500
+    },
+    {
+      "epoch": 6.304605242103242,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0003495631580631741,
+      "loss": 5.1901,
+      "step": 985000
+    },
+    {
+      "epoch": 6.307805549332736,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00034953755560533813,
+      "loss": 5.1898,
+      "step": 985500
+    },
+    {
+      "epoch": 6.31100585656223,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00034951195314750216,
+      "loss": 5.1932,
+      "step": 986000
+    },
+    {
+      "epoch": 6.314206163791724,
+      "grad_norm": 0.66015625,
+      "learning_rate": 0.0003494863506896662,
+      "loss": 5.186,
+      "step": 986500
+    },
+    {
+      "epoch": 6.317406471021218,
+      "grad_norm": 0.482421875,
+      "learning_rate": 0.0003494607482318303,
+      "loss": 5.192,
+      "step": 987000
+    },
+    {
+      "epoch": 6.320606778250712,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0003494351457739943,
+      "loss": 5.193,
+      "step": 987500
+    },
+    {
+      "epoch": 6.323807085480206,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003494095433161584,
+      "loss": 5.1917,
+      "step": 988000
+    },
+    {
+      "epoch": 6.3270073927097,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00034938394085832244,
+      "loss": 5.188,
+      "step": 988500
+    },
+    {
+      "epoch": 6.330207699939194,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0003493583384004865,
+      "loss": 5.19,
+      "step": 989000
+    },
+    {
+      "epoch": 6.333408007168688,
+      "grad_norm": 0.484375,
+      "learning_rate": 0.0003493327359426505,
+      "loss": 5.1942,
+      "step": 989500
+    },
+    {
+      "epoch": 6.336608314398182,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00034930713348481455,
+      "loss": 5.1932,
+      "step": 990000
+    },
+    {
+      "epoch": 6.339808621627676,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00034928153102697863,
+      "loss": 5.189,
+      "step": 990500
+    },
+    {
+      "epoch": 6.34300892885717,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00034925592856914267,
+      "loss": 5.1888,
+      "step": 991000
+    },
+    {
+      "epoch": 6.346209236086664,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003492303261113067,
+      "loss": 5.1849,
+      "step": 991500
+    },
+    {
+      "epoch": 6.349409543316158,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00034920472365347074,
+      "loss": 5.1969,
+      "step": 992000
+    },
+    {
+      "epoch": 6.352609850545653,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00034917912119563477,
+      "loss": 5.1877,
+      "step": 992500
+    },
+    {
+      "epoch": 6.355810157775147,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003491535187377988,
+      "loss": 5.1914,
+      "step": 993000
+    },
+    {
+      "epoch": 6.359010465004641,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003491279162799629,
+      "loss": 5.1909,
+      "step": 993500
+    },
+    {
+      "epoch": 6.362210772234135,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00034910231382212693,
+      "loss": 5.1861,
+      "step": 994000
+    },
+    {
+      "epoch": 6.3654110794636285,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00034907671136429096,
+      "loss": 5.1896,
+      "step": 994500
+    },
+    {
+      "epoch": 6.368611386693122,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00034905110890645505,
+      "loss": 5.1852,
+      "step": 995000
+    },
+    {
+      "epoch": 6.371811693922616,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003490255064486191,
+      "loss": 5.1914,
+      "step": 995500
+    },
+    {
+      "epoch": 6.37501200115211,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003489999039907832,
+      "loss": 5.1831,
+      "step": 996000
+    },
+    {
+      "epoch": 6.378212308381604,
+      "grad_norm": 0.490234375,
+      "learning_rate": 0.0003489743015329472,
+      "loss": 5.1936,
+      "step": 996500
+    },
+    {
+      "epoch": 6.381412615611099,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00034894869907511124,
+      "loss": 5.1896,
+      "step": 997000
+    },
+    {
+      "epoch": 6.384612922840593,
+      "grad_norm": 0.4609375,
+      "learning_rate": 0.0003489230966172753,
+      "loss": 5.1879,
+      "step": 997500
+    },
+    {
+      "epoch": 6.387813230070087,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.0003488974941594393,
+      "loss": 5.1874,
+      "step": 998000
+    },
+    {
+      "epoch": 6.391013537299581,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0003488718917016034,
+      "loss": 5.1846,
+      "step": 998500
+    },
+    {
+      "epoch": 6.394213844529075,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00034884628924376743,
+      "loss": 5.1941,
+      "step": 999000
+    },
+    {
+      "epoch": 6.397414151758569,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00034882068678593147,
+      "loss": 5.1869,
+      "step": 999500
+    },
+    {
+      "epoch": 6.400614458988063,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003487950843280955,
+      "loss": 5.1922,
+      "step": 1000000
+    },
+    {
+      "epoch": 6.403814766217557,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00034876948187025954,
+      "loss": 5.1877,
+      "step": 1000500
+    },
+    {
+      "epoch": 6.407015073447051,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00034874387941242357,
+      "loss": 5.1912,
+      "step": 1001000
+    },
+    {
+      "epoch": 6.410215380676545,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00034871827695458766,
+      "loss": 5.1865,
+      "step": 1001500
+    },
+    {
+      "epoch": 6.413415687906039,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003486926744967517,
+      "loss": 5.1884,
+      "step": 1002000
+    },
+    {
+      "epoch": 6.416615995135533,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00034866707203891573,
+      "loss": 5.1913,
+      "step": 1002500
+    },
+    {
+      "epoch": 6.419816302365027,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0003486414695810798,
+      "loss": 5.1852,
+      "step": 1003000
+    },
+    {
+      "epoch": 6.423016609594521,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00034861586712324385,
+      "loss": 5.1951,
+      "step": 1003500
+    },
+    {
+      "epoch": 6.426216916824015,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.00034859026466540794,
+      "loss": 5.1873,
+      "step": 1004000
+    },
+    {
+      "epoch": 6.429417224053509,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00034856466220757197,
+      "loss": 5.1911,
+      "step": 1004500
+    },
+    {
+      "epoch": 6.432617531283003,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.000348539059749736,
+      "loss": 5.1974,
+      "step": 1005000
+    },
+    {
+      "epoch": 6.435817838512497,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00034851345729190004,
+      "loss": 5.1876,
+      "step": 1005500
+    },
+    {
+      "epoch": 6.439018145741991,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0003484878548340641,
+      "loss": 5.19,
+      "step": 1006000
+    },
+    {
+      "epoch": 6.442218452971486,
+      "grad_norm": 0.6015625,
+      "learning_rate": 0.0003484622523762281,
+      "loss": 5.1888,
+      "step": 1006500
+    },
+    {
+      "epoch": 6.44541876020098,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003484366499183922,
+      "loss": 5.1857,
+      "step": 1007000
+    },
+    {
+      "epoch": 6.4486190674304735,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00034841104746055623,
+      "loss": 5.1881,
+      "step": 1007500
+    },
+    {
+      "epoch": 6.4518193746599675,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00034838544500272027,
+      "loss": 5.1899,
+      "step": 1008000
+    },
+    {
+      "epoch": 6.455019681889461,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003483598425448843,
+      "loss": 5.1891,
+      "step": 1008500
+    },
+    {
+      "epoch": 6.458219989118955,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00034833424008704833,
+      "loss": 5.1972,
+      "step": 1009000
+    },
+    {
+      "epoch": 6.461420296348449,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.0003483086376292124,
+      "loss": 5.1911,
+      "step": 1009500
+    },
+    {
+      "epoch": 6.464620603577943,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00034828303517137646,
+      "loss": 5.1908,
+      "step": 1010000
+    },
+    {
+      "epoch": 6.467820910807437,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00034825743271354055,
+      "loss": 5.1851,
+      "step": 1010500
+    },
+    {
+      "epoch": 6.471021218036931,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003482318302557046,
+      "loss": 5.1934,
+      "step": 1011000
+    },
+    {
+      "epoch": 6.474221525266426,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0003482062277978686,
+      "loss": 5.1904,
+      "step": 1011500
+    },
+    {
+      "epoch": 6.47742183249592,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00034818062534003265,
+      "loss": 5.1897,
+      "step": 1012000
+    },
+    {
+      "epoch": 6.480622139725414,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00034815502288219674,
+      "loss": 5.1882,
+      "step": 1012500
+    },
+    {
+      "epoch": 6.483822446954908,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00034812942042436077,
+      "loss": 5.1908,
+      "step": 1013000
+    },
+    {
+      "epoch": 6.487022754184402,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003481038179665248,
+      "loss": 5.1889,
+      "step": 1013500
+    },
+    {
+      "epoch": 6.490223061413896,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00034807821550868884,
+      "loss": 5.1961,
+      "step": 1014000
+    },
+    {
+      "epoch": 6.4934233686433895,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.0003480526130508529,
+      "loss": 5.1916,
+      "step": 1014500
+    },
+    {
+      "epoch": 6.4966236758728835,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00034802701059301696,
+      "loss": 5.1997,
+      "step": 1015000
+    },
+    {
+      "epoch": 6.499823983102377,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.000348001408135181,
+      "loss": 5.1884,
+      "step": 1015500
+    },
+    {
+      "epoch": 6.503024290331872,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00034797580567734503,
+      "loss": 5.1854,
+      "step": 1016000
+    },
+    {
+      "epoch": 6.506224597561366,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00034795020321950907,
+      "loss": 5.1893,
+      "step": 1016500
+    },
+    {
+      "epoch": 6.50942490479086,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.0003479246007616731,
+      "loss": 5.1924,
+      "step": 1017000
+    },
+    {
+      "epoch": 6.512625212020354,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003478989983038372,
+      "loss": 5.1975,
+      "step": 1017500
+    },
+    {
+      "epoch": 6.515825519249848,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003478733958460012,
+      "loss": 5.1912,
+      "step": 1018000
+    },
+    {
+      "epoch": 6.519025826479342,
+      "grad_norm": 0.671875,
+      "learning_rate": 0.0003478477933881653,
+      "loss": 5.1807,
+      "step": 1018500
+    },
+    {
+      "epoch": 6.522226133708836,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00034782219093032935,
+      "loss": 5.1958,
+      "step": 1019000
+    },
+    {
+      "epoch": 6.52542644093833,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003477965884724934,
+      "loss": 5.1963,
+      "step": 1019500
+    },
+    {
+      "epoch": 6.528626748167824,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.0003477709860146574,
+      "loss": 5.1893,
+      "step": 1020000
+    },
+    {
+      "epoch": 6.531827055397319,
+      "grad_norm": 0.5,
+      "learning_rate": 0.0003477453835568215,
+      "loss": 5.1881,
+      "step": 1020500
+    },
+    {
+      "epoch": 6.5350273626268125,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00034771978109898554,
+      "loss": 5.1903,
+      "step": 1021000
+    },
+    {
+      "epoch": 6.538227669856306,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.00034769417864114957,
+      "loss": 5.193,
+      "step": 1021500
+    },
+    {
+      "epoch": 6.5414279770858,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0003476685761833136,
+      "loss": 5.195,
+      "step": 1022000
+    },
+    {
+      "epoch": 6.544628284315294,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.00034764297372547764,
+      "loss": 5.1893,
+      "step": 1022500
+    },
+    {
+      "epoch": 6.547828591544788,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0003476173712676417,
+      "loss": 5.1921,
+      "step": 1023000
+    },
+    {
+      "epoch": 6.551028898774282,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00034759176880980576,
+      "loss": 5.1918,
+      "step": 1023500
+    },
+    {
+      "epoch": 6.554229206003776,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003475661663519698,
+      "loss": 5.1978,
+      "step": 1024000
+    },
+    {
+      "epoch": 6.55742951323327,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00034754056389413383,
+      "loss": 5.189,
+      "step": 1024500
+    },
+    {
+      "epoch": 6.560629820462765,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.0003475149614362979,
+      "loss": 5.1892,
+      "step": 1025000
+    },
+    {
+      "epoch": 6.563830127692259,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00034748935897846195,
+      "loss": 5.1897,
+      "step": 1025500
+    },
+    {
+      "epoch": 6.567030434921753,
+      "grad_norm": 0.4921875,
+      "learning_rate": 0.00034746375652062604,
+      "loss": 5.1944,
+      "step": 1026000
+    },
+    {
+      "epoch": 6.570230742151247,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.0003474381540627901,
+      "loss": 5.192,
+      "step": 1026500
+    },
+    {
+      "epoch": 6.573431049380741,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0003474125516049541,
+      "loss": 5.1942,
+      "step": 1027000
+    },
+    {
+      "epoch": 6.576631356610235,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00034738694914711814,
+      "loss": 5.185,
+      "step": 1027500
+    },
+    {
+      "epoch": 6.5798316638397285,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003473613466892822,
+      "loss": 5.1918,
+      "step": 1028000
+    },
+    {
+      "epoch": 6.583031971069222,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003473357442314462,
+      "loss": 5.1896,
+      "step": 1028500
+    },
+    {
+      "epoch": 6.586232278298716,
+      "grad_norm": 0.46875,
+      "learning_rate": 0.0003473101417736103,
+      "loss": 5.1934,
+      "step": 1029000
+    },
+    {
+      "epoch": 6.589432585528211,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00034728453931577434,
+      "loss": 5.1935,
+      "step": 1029500
+    },
+    {
+      "epoch": 6.592632892757704,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00034725893685793837,
+      "loss": 5.1943,
+      "step": 1030000
+    },
+    {
+      "epoch": 6.595833199987199,
+      "grad_norm": 0.51171875,
+      "learning_rate": 0.0003472333344001024,
+      "loss": 5.1917,
+      "step": 1030500
+    },
+    {
+      "epoch": 6.599033507216693,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00034720773194226644,
+      "loss": 5.1918,
+      "step": 1031000
+    },
+    {
+      "epoch": 6.602233814446187,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0003471821294844305,
+      "loss": 5.1897,
+      "step": 1031500
+    },
+    {
+      "epoch": 6.605434121675681,
+      "grad_norm": 0.474609375,
+      "learning_rate": 0.00034715652702659456,
+      "loss": 5.1899,
+      "step": 1032000
+    },
+    {
+      "epoch": 6.608634428905175,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003471309245687586,
+      "loss": 5.1846,
+      "step": 1032500
+    },
+    {
+      "epoch": 6.611834736134669,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003471053221109227,
+      "loss": 5.1905,
+      "step": 1033000
+    },
+    {
+      "epoch": 6.615035043364163,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0003470797196530867,
+      "loss": 5.1933,
+      "step": 1033500
+    },
+    {
+      "epoch": 6.618235350593657,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00034705411719525075,
+      "loss": 5.197,
+      "step": 1034000
+    },
+    {
+      "epoch": 6.621435657823151,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00034702851473741484,
+      "loss": 5.1931,
+      "step": 1034500
+    },
+    {
+      "epoch": 6.624635965052645,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003470029122795789,
+      "loss": 5.1942,
+      "step": 1035000
+    },
+    {
+      "epoch": 6.627836272282139,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003469773098217429,
+      "loss": 5.1872,
+      "step": 1035500
+    },
+    {
+      "epoch": 6.631036579511633,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00034695170736390694,
+      "loss": 5.1943,
+      "step": 1036000
+    },
+    {
+      "epoch": 6.634236886741127,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.000346926104906071,
+      "loss": 5.1812,
+      "step": 1036500
+    },
+    {
+      "epoch": 6.637437193970621,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00034690050244823507,
+      "loss": 5.1939,
+      "step": 1037000
+    },
+    {
+      "epoch": 6.640637501200115,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0003468748999903991,
+      "loss": 5.1886,
+      "step": 1037500
+    },
+    {
+      "epoch": 6.643837808429609,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00034684929753256313,
+      "loss": 5.1897,
+      "step": 1038000
+    },
+    {
+      "epoch": 6.647038115659103,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00034682369507472717,
+      "loss": 5.1931,
+      "step": 1038500
+    },
+    {
+      "epoch": 6.650238422888597,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0003467980926168912,
+      "loss": 5.1965,
+      "step": 1039000
+    },
+    {
+      "epoch": 6.653438730118092,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0003467724901590553,
+      "loss": 5.1879,
+      "step": 1039500
+    },
+    {
+      "epoch": 6.656639037347586,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0003467468877012193,
+      "loss": 5.1892,
+      "step": 1040000
+    },
+    {
+      "epoch": 6.65983934457708,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.0003467212852433834,
+      "loss": 5.1931,
+      "step": 1040500
+    },
+    {
+      "epoch": 6.6630396518065735,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.00034669568278554745,
+      "loss": 5.1952,
+      "step": 1041000
+    },
+    {
+      "epoch": 6.6662399590360675,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0003466700803277115,
+      "loss": 5.1862,
+      "step": 1041500
+    },
+    {
+      "epoch": 6.669440266265561,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003466444778698755,
+      "loss": 5.1895,
+      "step": 1042000
+    },
+    {
+      "epoch": 6.672640573495055,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0003466188754120396,
+      "loss": 5.1934,
+      "step": 1042500
+    },
+    {
+      "epoch": 6.675840880724549,
+      "grad_norm": 0.63671875,
+      "learning_rate": 0.00034659327295420364,
+      "loss": 5.1865,
+      "step": 1043000
+    },
+    {
+      "epoch": 6.679041187954043,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003465676704963677,
+      "loss": 5.1901,
+      "step": 1043500
+    },
+    {
+      "epoch": 6.682241495183538,
+      "grad_norm": 0.48046875,
+      "learning_rate": 0.0003465420680385317,
+      "loss": 5.1898,
+      "step": 1044000
+    },
+    {
+      "epoch": 6.685441802413032,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00034651646558069574,
+      "loss": 5.1869,
+      "step": 1044500
+    },
+    {
+      "epoch": 6.688642109642526,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00034649086312285983,
+      "loss": 5.1884,
+      "step": 1045000
+    },
+    {
+      "epoch": 6.69184241687202,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00034646526066502386,
+      "loss": 5.1853,
+      "step": 1045500
+    },
+    {
+      "epoch": 6.695042724101514,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003464396582071879,
+      "loss": 5.1859,
+      "step": 1046000
+    },
+    {
+      "epoch": 6.698243031331008,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.00034641405574935193,
+      "loss": 5.1969,
+      "step": 1046500
+    },
+    {
+      "epoch": 6.701443338560502,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00034638845329151597,
+      "loss": 5.1975,
+      "step": 1047000
+    },
+    {
+      "epoch": 6.704643645789996,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00034636285083368006,
+      "loss": 5.1999,
+      "step": 1047500
+    },
+    {
+      "epoch": 6.7078439530194895,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003463372483758441,
+      "loss": 5.1876,
+      "step": 1048000
+    },
+    {
+      "epoch": 6.711044260248984,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003463116459180082,
+      "loss": 5.1959,
+      "step": 1048500
+    },
+    {
+      "epoch": 6.714244567478478,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.0003462860434601722,
+      "loss": 5.1904,
+      "step": 1049000
+    },
+    {
+      "epoch": 6.717444874707972,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00034626044100233625,
+      "loss": 5.1928,
+      "step": 1049500
+    },
+    {
+      "epoch": 6.720645181937466,
+      "grad_norm": 0.6640625,
+      "learning_rate": 0.0003462348385445003,
+      "loss": 5.1965,
+      "step": 1050000
+    },
+    {
+      "epoch": 6.72384548916696,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00034620923608666437,
+      "loss": 5.1949,
+      "step": 1050500
+    },
+    {
+      "epoch": 6.727045796396454,
+      "grad_norm": 0.61328125,
+      "learning_rate": 0.0003461836336288284,
+      "loss": 5.187,
+      "step": 1051000
+    },
+    {
+      "epoch": 6.730246103625948,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00034615803117099244,
+      "loss": 5.1885,
+      "step": 1051500
+    },
+    {
+      "epoch": 6.733446410855442,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00034613242871315647,
+      "loss": 5.1862,
+      "step": 1052000
+    },
+    {
+      "epoch": 6.736646718084936,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.0003461068262553205,
+      "loss": 5.1924,
+      "step": 1052500
+    },
+    {
+      "epoch": 6.73984702531443,
+      "grad_norm": 0.48828125,
+      "learning_rate": 0.00034608122379748454,
+      "loss": 5.1912,
+      "step": 1053000
+    },
+    {
+      "epoch": 6.743047332543924,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00034605562133964863,
+      "loss": 5.1894,
+      "step": 1053500
+    },
+    {
+      "epoch": 6.746247639773419,
+      "grad_norm": 0.59375,
+      "learning_rate": 0.00034603001888181266,
+      "loss": 5.1912,
+      "step": 1054000
+    },
+    {
+      "epoch": 6.7494479470029125,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003460044164239767,
+      "loss": 5.1874,
+      "step": 1054500
+    },
+    {
+      "epoch": 6.752648254232406,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00034597881396614073,
+      "loss": 5.1947,
+      "step": 1055000
+    },
+    {
+      "epoch": 6.7558485614619,
+      "grad_norm": 0.498046875,
+      "learning_rate": 0.0003459532115083048,
+      "loss": 5.1855,
+      "step": 1055500
+    },
+    {
+      "epoch": 6.759048868691394,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00034592760905046885,
+      "loss": 5.1935,
+      "step": 1056000
+    },
+    {
+      "epoch": 6.762249175920888,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.00034590200659263294,
+      "loss": 5.1905,
+      "step": 1056500
+    },
+    {
+      "epoch": 6.765449483150382,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.000345876404134797,
+      "loss": 5.1933,
+      "step": 1057000
+    },
+    {
+      "epoch": 6.768649790379876,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.000345850801676961,
+      "loss": 5.1927,
+      "step": 1057500
+    },
+    {
+      "epoch": 6.77185009760937,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00034582519921912505,
+      "loss": 5.1847,
+      "step": 1058000
+    },
+    {
+      "epoch": 6.775050404838865,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.0003457995967612891,
+      "loss": 5.1947,
+      "step": 1058500
+    },
+    {
+      "epoch": 6.778250712068359,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00034577399430345317,
+      "loss": 5.1914,
+      "step": 1059000
+    },
+    {
+      "epoch": 6.781451019297853,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0003457483918456172,
+      "loss": 5.1856,
+      "step": 1059500
+    },
+    {
+      "epoch": 6.784651326527347,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.00034572278938778124,
+      "loss": 5.1888,
+      "step": 1060000
+    },
+    {
+      "epoch": 6.787851633756841,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00034569718692994527,
+      "loss": 5.1896,
+      "step": 1060500
+    },
+    {
+      "epoch": 6.791051940986335,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.0003456715844721093,
+      "loss": 5.1867,
+      "step": 1061000
+    },
+    {
+      "epoch": 6.7942522482158285,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.0003456459820142734,
+      "loss": 5.1903,
+      "step": 1061500
+    },
+    {
+      "epoch": 6.7974525554453225,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00034562037955643743,
+      "loss": 5.1917,
+      "step": 1062000
+    },
+    {
+      "epoch": 6.800652862674816,
+      "grad_norm": 0.56640625,
+      "learning_rate": 0.00034559477709860146,
+      "loss": 5.1901,
+      "step": 1062500
+    },
+    {
+      "epoch": 6.803853169904311,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.00034556917464076555,
+      "loss": 5.1923,
+      "step": 1063000
+    },
+    {
+      "epoch": 6.807053477133805,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.0003455435721829296,
+      "loss": 5.1872,
+      "step": 1063500
+    },
+    {
+      "epoch": 6.810253784363299,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0003455179697250936,
+      "loss": 5.1963,
+      "step": 1064000
+    },
+    {
+      "epoch": 6.813454091592793,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.0003454923672672577,
+      "loss": 5.1925,
+      "step": 1064500
+    },
+    {
+      "epoch": 6.816654398822287,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00034546676480942174,
+      "loss": 5.188,
+      "step": 1065000
+    },
+    {
+      "epoch": 6.819854706051781,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0003454411623515858,
+      "loss": 5.1882,
+      "step": 1065500
+    },
+    {
+      "epoch": 6.823055013281275,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0003454155598937498,
+      "loss": 5.1874,
+      "step": 1066000
+    },
+    {
+      "epoch": 6.826255320510769,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00034538995743591384,
+      "loss": 5.1914,
+      "step": 1066500
+    },
+    {
+      "epoch": 6.829455627740263,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.00034536435497807793,
+      "loss": 5.1978,
+      "step": 1067000
+    },
+    {
+      "epoch": 6.8326559349697575,
+      "grad_norm": 0.60546875,
+      "learning_rate": 0.00034533875252024197,
+      "loss": 5.1938,
+      "step": 1067500
+    },
+    {
+      "epoch": 6.8358562421992515,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.000345313150062406,
+      "loss": 5.1847,
+      "step": 1068000
+    },
+    {
+      "epoch": 6.839056549428745,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.00034528754760457004,
+      "loss": 5.1907,
+      "step": 1068500
+    },
+    {
+      "epoch": 6.842256856658239,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00034526194514673407,
+      "loss": 5.1946,
+      "step": 1069000
+    },
+    {
+      "epoch": 6.845457163887733,
+      "grad_norm": 0.4765625,
+      "learning_rate": 0.0003452363426888981,
+      "loss": 5.1881,
+      "step": 1069500
+    },
+    {
+      "epoch": 6.848657471117227,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0003452107402310622,
+      "loss": 5.1939,
+      "step": 1070000
+    },
+    {
+      "epoch": 6.851857778346721,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00034518513777322623,
+      "loss": 5.1925,
+      "step": 1070500
+    },
+    {
+      "epoch": 6.855058085576215,
+      "grad_norm": 0.5546875,
+      "learning_rate": 0.0003451595353153903,
+      "loss": 5.1939,
+      "step": 1071000
+    },
+    {
+      "epoch": 6.858258392805709,
+      "grad_norm": 0.5078125,
+      "learning_rate": 0.00034513393285755435,
+      "loss": 5.1893,
+      "step": 1071500
+    },
+    {
+      "epoch": 6.861458700035204,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.0003451083303997184,
+      "loss": 5.1903,
+      "step": 1072000
+    },
+    {
+      "epoch": 6.864659007264698,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.00034508272794188247,
+      "loss": 5.1848,
+      "step": 1072500
+    },
+    {
+      "epoch": 6.867859314494192,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003450571254840465,
+      "loss": 5.1885,
+      "step": 1073000
+    },
+    {
+      "epoch": 6.871059621723686,
+      "grad_norm": 0.5703125,
+      "learning_rate": 0.00034503152302621054,
+      "loss": 5.1888,
+      "step": 1073500
+    },
+    {
+      "epoch": 6.87425992895318,
+      "grad_norm": 0.6171875,
+      "learning_rate": 0.0003450059205683746,
+      "loss": 5.1897,
+      "step": 1074000
+    },
+    {
+      "epoch": 6.8774602361826735,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0003449803181105386,
+      "loss": 5.1896,
+      "step": 1074500
+    },
+    {
+      "epoch": 6.8806605434121675,
+      "grad_norm": 0.640625,
+      "learning_rate": 0.0003449547156527027,
+      "loss": 5.1949,
+      "step": 1075000
+    },
+    {
+      "epoch": 6.883860850641661,
+      "grad_norm": 0.58984375,
+      "learning_rate": 0.00034492911319486673,
+      "loss": 5.1969,
+      "step": 1075500
+    },
+    {
+      "epoch": 6.887061157871155,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00034490351073703077,
+      "loss": 5.1871,
+      "step": 1076000
+    },
+    {
+      "epoch": 6.890261465100649,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003448779082791948,
+      "loss": 5.1961,
+      "step": 1076500
+    },
+    {
+      "epoch": 6.893461772330143,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00034485230582135884,
+      "loss": 5.1902,
+      "step": 1077000
+    },
+    {
+      "epoch": 6.896662079559638,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00034482670336352287,
+      "loss": 5.1938,
+      "step": 1077500
+    },
+    {
+      "epoch": 6.899862386789132,
+      "grad_norm": 0.54296875,
+      "learning_rate": 0.00034480110090568696,
+      "loss": 5.1915,
+      "step": 1078000
+    },
+    {
+      "epoch": 6.903062694018626,
+      "grad_norm": 0.5,
+      "learning_rate": 0.00034477549844785105,
+      "loss": 5.1905,
+      "step": 1078500
+    },
+    {
+      "epoch": 6.90626300124812,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003447498959900151,
+      "loss": 5.1938,
+      "step": 1079000
+    },
+    {
+      "epoch": 6.909463308477614,
+      "grad_norm": 0.53515625,
+      "learning_rate": 0.0003447242935321791,
+      "loss": 5.1896,
+      "step": 1079500
+    },
+    {
+      "epoch": 6.912663615707108,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.00034469869107434315,
+      "loss": 5.1903,
+      "step": 1080000
+    },
+    {
+      "epoch": 6.915863922936602,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00034467308861650724,
+      "loss": 5.195,
+      "step": 1080500
+    },
+    {
+      "epoch": 6.919064230166096,
+      "grad_norm": 0.49609375,
+      "learning_rate": 0.00034464748615867127,
+      "loss": 5.1908,
+      "step": 1081000
+    },
+    {
+      "epoch": 6.92226453739559,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0003446218837008353,
+      "loss": 5.1953,
+      "step": 1081500
+    },
+    {
+      "epoch": 6.925464844625084,
+      "grad_norm": 0.5234375,
+      "learning_rate": 0.00034459628124299934,
+      "loss": 5.1934,
+      "step": 1082000
+    },
+    {
+      "epoch": 6.928665151854578,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003445706787851634,
+      "loss": 5.1893,
+      "step": 1082500
+    },
+    {
+      "epoch": 6.931865459084072,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.0003445450763273274,
+      "loss": 5.1902,
+      "step": 1083000
+    },
+    {
+      "epoch": 6.935065766313566,
+      "grad_norm": 0.55859375,
+      "learning_rate": 0.0003445194738694915,
+      "loss": 5.1884,
+      "step": 1083500
+    },
+    {
+      "epoch": 6.93826607354306,
+      "grad_norm": 0.51953125,
+      "learning_rate": 0.00034449387141165553,
+      "loss": 5.1871,
+      "step": 1084000
+    },
+    {
+      "epoch": 6.941466380772554,
+      "grad_norm": 0.62890625,
+      "learning_rate": 0.00034446826895381957,
+      "loss": 5.1905,
+      "step": 1084500
+    },
+    {
+      "epoch": 6.944666688002048,
+      "grad_norm": 0.53125,
+      "learning_rate": 0.0003444426664959836,
+      "loss": 5.1862,
+      "step": 1085000
+    },
+    {
+      "epoch": 6.947866995231542,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003444170640381477,
+      "loss": 5.1877,
+      "step": 1085500
+    },
+    {
+      "epoch": 6.951067302461036,
+      "grad_norm": 0.625,
+      "learning_rate": 0.0003443914615803117,
+      "loss": 5.1917,
+      "step": 1086000
+    },
+    {
+      "epoch": 6.954267609690531,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.0003443658591224758,
+      "loss": 5.1928,
+      "step": 1086500
+    },
+    {
+      "epoch": 6.957467916920025,
+      "grad_norm": 0.5390625,
+      "learning_rate": 0.00034434025666463985,
+      "loss": 5.1903,
+      "step": 1087000
+    },
+    {
+      "epoch": 6.960668224149519,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003443146542068039,
+      "loss": 5.1885,
+      "step": 1087500
+    },
+    {
+      "epoch": 6.9638685313790125,
+      "grad_norm": 0.546875,
+      "learning_rate": 0.0003442890517489679,
+      "loss": 5.1883,
+      "step": 1088000
+    },
+    {
+      "epoch": 6.9670688386085065,
+      "grad_norm": 0.5859375,
+      "learning_rate": 0.00034426344929113195,
+      "loss": 5.184,
+      "step": 1088500
+    },
+    {
+      "epoch": 6.970269145838,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00034423784683329604,
+      "loss": 5.1899,
+      "step": 1089000
+    },
+    {
+      "epoch": 6.973469453067494,
+      "grad_norm": 0.59765625,
+      "learning_rate": 0.00034421224437546007,
+      "loss": 5.1946,
+      "step": 1089500
+    },
+    {
+      "epoch": 6.976669760296988,
+      "grad_norm": 0.5625,
+      "learning_rate": 0.0003441866419176241,
+      "loss": 5.1946,
+      "step": 1090000
+    },
+    {
+      "epoch": 6.979870067526482,
+      "grad_norm": 0.494140625,
+      "learning_rate": 0.00034416103945978814,
+      "loss": 5.1905,
+      "step": 1090500
+    },
+    {
+      "epoch": 6.983070374755977,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0003441354370019522,
+      "loss": 5.1917,
+      "step": 1091000
+    },
+    {
+      "epoch": 6.986270681985471,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.00034410983454411626,
+      "loss": 5.1895,
+      "step": 1091500
+    },
+    {
+      "epoch": 6.989470989214965,
+      "grad_norm": 0.52734375,
+      "learning_rate": 0.0003440842320862803,
+      "loss": 5.1894,
+      "step": 1092000
+    },
+    {
+      "epoch": 6.992671296444459,
+      "grad_norm": 0.58203125,
+      "learning_rate": 0.00034405862962844433,
+      "loss": 5.1908,
+      "step": 1092500
+    },
+    {
+      "epoch": 6.995871603673953,
+      "grad_norm": 0.578125,
+      "learning_rate": 0.00034403302717060836,
+      "loss": 5.1952,
+      "step": 1093000
+    },
+    {
+      "epoch": 6.999071910903447,
+      "grad_norm": 0.57421875,
+      "learning_rate": 0.00034400742471277245,
+      "loss": 5.1905,
+      "step": 1093500
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 5.1858015060424805,
+      "eval_runtime": 1.1382,
+      "eval_samples_per_second": 878.603,
+      "eval_steps_per_second": 14.058,
+      "step": 1093645
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 7811750,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 50,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 3,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.473295707163853e+19,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}