diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5649 @@
+{
+  "best_metric": 3.6955573558807373,
+  "best_model_checkpoint": "checkpoints/test_1M_1-2025-02-12-12-32/checkpoint-20000",
+  "epoch": 1.5829046299960428,
+  "eval_steps": 10000,
+  "global_step": 20000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019786307874950534,
+      "grad_norm": 254.82342529296875,
+      "learning_rate": 2.499208537244918e-07,
+      "loss": 5.7705,
+      "step": 25
+    },
+    {
+      "epoch": 0.003957261574990107,
+      "grad_norm": 153.19989013671875,
+      "learning_rate": 2.498384096875041e-07,
+      "loss": 5.6747,
+      "step": 50
+    },
+    {
+      "epoch": 0.00593589236248516,
+      "grad_norm": 224.5292510986328,
+      "learning_rate": 2.4975596565051644e-07,
+      "loss": 5.6201,
+      "step": 75
+    },
+    {
+      "epoch": 0.007914523149980214,
+      "grad_norm": 175.854248046875,
+      "learning_rate": 2.4967352161352873e-07,
+      "loss": 5.6974,
+      "step": 100
+    },
+    {
+      "epoch": 0.009893153937475268,
+      "grad_norm": 163.52769470214844,
+      "learning_rate": 2.49591077576541e-07,
+      "loss": 5.5417,
+      "step": 125
+    },
+    {
+      "epoch": 0.01187178472497032,
+      "grad_norm": 254.2264862060547,
+      "learning_rate": 2.4950863353955335e-07,
+      "loss": 5.8201,
+      "step": 150
+    },
+    {
+      "epoch": 0.013850415512465374,
+      "grad_norm": 175.30279541015625,
+      "learning_rate": 2.4942618950256564e-07,
+      "loss": 5.5302,
+      "step": 175
+    },
+    {
+      "epoch": 0.015829046299960427,
+      "grad_norm": 300.1286315917969,
+      "learning_rate": 2.4934374546557797e-07,
+      "loss": 5.6572,
+      "step": 200
+    },
+    {
+      "epoch": 0.01780767708745548,
+      "grad_norm": 201.56961059570312,
+      "learning_rate": 2.4926130142859026e-07,
+      "loss": 5.2914,
+      "step": 225
+    },
+    {
+      "epoch": 0.019786307874950535,
+      "grad_norm": 245.64854431152344,
+      "learning_rate": 2.491788573916026e-07,
+      "loss": 5.4478,
+      "step": 250
+    },
+    {
+      "epoch": 0.02176493866244559,
+      "grad_norm": 239.78257751464844,
+      "learning_rate": 2.490964133546149e-07,
+      "loss": 5.4161,
+      "step": 275
+    },
+    {
+      "epoch": 0.02374356944994064,
+      "grad_norm": 150.18310546875,
+      "learning_rate": 2.4901396931762717e-07,
+      "loss": 5.4978,
+      "step": 300
+    },
+    {
+      "epoch": 0.025722200237435693,
+      "grad_norm": 172.03607177734375,
+      "learning_rate": 2.489315252806395e-07,
+      "loss": 5.4105,
+      "step": 325
+    },
+    {
+      "epoch": 0.027700831024930747,
+      "grad_norm": 343.2570495605469,
+      "learning_rate": 2.488490812436518e-07,
+      "loss": 5.5195,
+      "step": 350
+    },
+    {
+      "epoch": 0.0296794618124258,
+      "grad_norm": 329.7228698730469,
+      "learning_rate": 2.4876663720666413e-07,
+      "loss": 5.4494,
+      "step": 375
+    },
+    {
+      "epoch": 0.031658092599920855,
+      "grad_norm": 174.63136291503906,
+      "learning_rate": 2.486841931696764e-07,
+      "loss": 5.2864,
+      "step": 400
+    },
+    {
+      "epoch": 0.033636723387415905,
+      "grad_norm": 356.6216125488281,
+      "learning_rate": 2.486017491326887e-07,
+      "loss": 5.4376,
+      "step": 425
+    },
+    {
+      "epoch": 0.03561535417491096,
+      "grad_norm": 166.16783142089844,
+      "learning_rate": 2.4851930509570104e-07,
+      "loss": 5.3141,
+      "step": 450
+    },
+    {
+      "epoch": 0.03759398496240601,
+      "grad_norm": 220.06170654296875,
+      "learning_rate": 2.484368610587133e-07,
+      "loss": 5.5457,
+      "step": 475
+    },
+    {
+      "epoch": 0.03957261574990107,
+      "grad_norm": 154.55517578125,
+      "learning_rate": 2.483544170217256e-07,
+      "loss": 5.1264,
+      "step": 500
+    },
+    {
+      "epoch": 0.04155124653739612,
+      "grad_norm": 184.18443298339844,
+      "learning_rate": 2.4827197298473794e-07,
+      "loss": 5.3702,
+      "step": 525
+    },
+    {
+      "epoch": 0.04352987732489118,
+      "grad_norm": 128.84693908691406,
+      "learning_rate": 2.4818952894775023e-07,
+      "loss": 5.0207,
+      "step": 550
+    },
+    {
+      "epoch": 0.04550850811238623,
+      "grad_norm": 196.2894287109375,
+      "learning_rate": 2.4810708491076257e-07,
+      "loss": 5.315,
+      "step": 575
+    },
+    {
+      "epoch": 0.04748713889988128,
+      "grad_norm": 200.00257873535156,
+      "learning_rate": 2.4802464087377485e-07,
+      "loss": 5.2215,
+      "step": 600
+    },
+    {
+      "epoch": 0.049465769687376336,
+      "grad_norm": 271.8963928222656,
+      "learning_rate": 2.479421968367872e-07,
+      "loss": 5.3286,
+      "step": 625
+    },
+    {
+      "epoch": 0.051444400474871387,
+      "grad_norm": 181.56686401367188,
+      "learning_rate": 2.478597527997995e-07,
+      "loss": 4.9967,
+      "step": 650
+    },
+    {
+      "epoch": 0.053423031262366444,
+      "grad_norm": 242.8925323486328,
+      "learning_rate": 2.477773087628118e-07,
+      "loss": 5.1984,
+      "step": 675
+    },
+    {
+      "epoch": 0.055401662049861494,
+      "grad_norm": 210.05746459960938,
+      "learning_rate": 2.476948647258241e-07,
+      "loss": 5.0975,
+      "step": 700
+    },
+    {
+      "epoch": 0.05738029283735655,
+      "grad_norm": 181.1220245361328,
+      "learning_rate": 2.476124206888364e-07,
+      "loss": 5.0036,
+      "step": 725
+    },
+    {
+      "epoch": 0.0593589236248516,
+      "grad_norm": 166.00709533691406,
+      "learning_rate": 2.475299766518487e-07,
+      "loss": 5.3082,
+      "step": 750
+    },
+    {
+      "epoch": 0.06133755441234666,
+      "grad_norm": 151.4649200439453,
+      "learning_rate": 2.47447532614861e-07,
+      "loss": 5.1391,
+      "step": 775
+    },
+    {
+      "epoch": 0.06331618519984171,
+      "grad_norm": 149.88165283203125,
+      "learning_rate": 2.4736508857787335e-07,
+      "loss": 5.0783,
+      "step": 800
+    },
+    {
+      "epoch": 0.06529481598733676,
+      "grad_norm": 172.47061157226562,
+      "learning_rate": 2.4728264454088563e-07,
+      "loss": 4.9624,
+      "step": 825
+    },
+    {
+      "epoch": 0.06727344677483181,
+      "grad_norm": 298.1490478515625,
+      "learning_rate": 2.4720020050389797e-07,
+      "loss": 5.1937,
+      "step": 850
+    },
+    {
+      "epoch": 0.06925207756232687,
+      "grad_norm": 164.37867736816406,
+      "learning_rate": 2.4711775646691025e-07,
+      "loss": 5.1792,
+      "step": 875
+    },
+    {
+      "epoch": 0.07123070834982193,
+      "grad_norm": 216.8033905029297,
+      "learning_rate": 2.4703531242992254e-07,
+      "loss": 5.152,
+      "step": 900
+    },
+    {
+      "epoch": 0.07320933913731698,
+      "grad_norm": 211.95762634277344,
+      "learning_rate": 2.469528683929349e-07,
+      "loss": 4.9146,
+      "step": 925
+    },
+    {
+      "epoch": 0.07518796992481203,
+      "grad_norm": 257.61968994140625,
+      "learning_rate": 2.4687042435594716e-07,
+      "loss": 5.095,
+      "step": 950
+    },
+    {
+      "epoch": 0.07716660071230709,
+      "grad_norm": 179.43719482421875,
+      "learning_rate": 2.467879803189595e-07,
+      "loss": 5.0316,
+      "step": 975
+    },
+    {
+      "epoch": 0.07914523149980214,
+      "grad_norm": 180.3157958984375,
+      "learning_rate": 2.467055362819718e-07,
+      "loss": 4.9441,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08112386228729719,
+      "grad_norm": 162.77447509765625,
+      "learning_rate": 2.4662309224498407e-07,
+      "loss": 4.9724,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08310249307479224,
+      "grad_norm": 123.65939331054688,
+      "learning_rate": 2.465406482079964e-07,
+      "loss": 5.2271,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08508112386228729,
+      "grad_norm": 163.114990234375,
+      "learning_rate": 2.464582041710087e-07,
+      "loss": 4.9724,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08705975464978236,
+      "grad_norm": 204.76400756835938,
+      "learning_rate": 2.46375760134021e-07,
+      "loss": 4.8724,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0890383854372774,
+      "grad_norm": 307.8963623046875,
+      "learning_rate": 2.462933160970333e-07,
+      "loss": 4.9256,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09101701622477246,
+      "grad_norm": 133.03707885742188,
+      "learning_rate": 2.462108720600456e-07,
+      "loss": 4.8792,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09299564701226751,
+      "grad_norm": 161.41697692871094,
+      "learning_rate": 2.4612842802305794e-07,
+      "loss": 5.054,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09497427779976256,
+      "grad_norm": 135.36228942871094,
+      "learning_rate": 2.460459839860702e-07,
+      "loss": 4.8655,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09695290858725762,
+      "grad_norm": 179.60646057128906,
+      "learning_rate": 2.4596353994908256e-07,
+      "loss": 4.7832,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09893153937475267,
+      "grad_norm": 335.71380615234375,
+      "learning_rate": 2.4588109591209485e-07,
+      "loss": 4.9979,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10091017016224772,
+      "grad_norm": 149.5147247314453,
+      "learning_rate": 2.457986518751072e-07,
+      "loss": 4.714,
+      "step": 1275
+    },
+    {
+      "epoch": 0.10288880094974277,
+      "grad_norm": 154.0236358642578,
+      "learning_rate": 2.4571620783811947e-07,
+      "loss": 4.8015,
+      "step": 1300
+    },
+    {
+      "epoch": 0.10486743173723784,
+      "grad_norm": 450.5319519042969,
+      "learning_rate": 2.456337638011318e-07,
+      "loss": 4.6914,
+      "step": 1325
+    },
+    {
+      "epoch": 0.10684606252473289,
+      "grad_norm": 195.87863159179688,
+      "learning_rate": 2.455513197641441e-07,
+      "loss": 5.0124,
+      "step": 1350
+    },
+    {
+      "epoch": 0.10882469331222794,
+      "grad_norm": 198.12225341796875,
+      "learning_rate": 2.454688757271564e-07,
+      "loss": 4.5305,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11080332409972299,
+      "grad_norm": 161.57623291015625,
+      "learning_rate": 2.453864316901687e-07,
+      "loss": 4.7806,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11278195488721804,
+      "grad_norm": 187.8081817626953,
+      "learning_rate": 2.45303987653181e-07,
+      "loss": 4.9401,
+      "step": 1425
+    },
+    {
+      "epoch": 0.1147605856747131,
+      "grad_norm": 160.1893768310547,
+      "learning_rate": 2.4522154361619334e-07,
+      "loss": 4.8119,
+      "step": 1450
+    },
+    {
+      "epoch": 0.11673921646220815,
+      "grad_norm": 181.8563995361328,
+      "learning_rate": 2.4513909957920563e-07,
+      "loss": 4.7979,
+      "step": 1475
+    },
+    {
+      "epoch": 0.1187178472497032,
+      "grad_norm": 184.80641174316406,
+      "learning_rate": 2.4505665554221796e-07,
+      "loss": 4.8448,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12069647803719825,
+      "grad_norm": 151.4502410888672,
+      "learning_rate": 2.4497421150523025e-07,
+      "loss": 4.7101,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12267510882469332,
+      "grad_norm": 163.2119598388672,
+      "learning_rate": 2.4489176746824253e-07,
+      "loss": 4.8802,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12465373961218837,
+      "grad_norm": 147.33741760253906,
+      "learning_rate": 2.4480932343125487e-07,
+      "loss": 4.6433,
+      "step": 1575
+    },
+    {
+      "epoch": 0.12663237039968342,
+      "grad_norm": 145.84716796875,
+      "learning_rate": 2.4472687939426716e-07,
+      "loss": 4.4118,
+      "step": 1600
+    },
+    {
+      "epoch": 0.12861100118717847,
+      "grad_norm": 111.55641174316406,
+      "learning_rate": 2.4464443535727944e-07,
+      "loss": 4.819,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13058963197467352,
+      "grad_norm": 145.68092346191406,
+      "learning_rate": 2.445619913202918e-07,
+      "loss": 4.7752,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13256826276216857,
+      "grad_norm": 274.0830078125,
+      "learning_rate": 2.4447954728330407e-07,
+      "loss": 4.8566,
+      "step": 1675
+    },
+    {
+      "epoch": 0.13454689354966362,
+      "grad_norm": 141.83982849121094,
+      "learning_rate": 2.4439710324631635e-07,
+      "loss": 4.6643,
+      "step": 1700
+    },
+    {
+      "epoch": 0.1365255243371587,
+      "grad_norm": 182.46160888671875,
+      "learning_rate": 2.443146592093287e-07,
+      "loss": 4.731,
+      "step": 1725
+    },
+    {
+      "epoch": 0.13850415512465375,
+      "grad_norm": 200.28773498535156,
+      "learning_rate": 2.44232215172341e-07,
+      "loss": 4.5525,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1404827859121488,
+      "grad_norm": 163.7792510986328,
+      "learning_rate": 2.441497711353533e-07,
+      "loss": 4.8076,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14246141669964385,
+      "grad_norm": 422.9642639160156,
+      "learning_rate": 2.440673270983656e-07,
+      "loss": 4.7045,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1444400474871389,
+      "grad_norm": 187.99957275390625,
+      "learning_rate": 2.4398488306137794e-07,
+      "loss": 4.6615,
+      "step": 1825
+    },
+    {
+      "epoch": 0.14641867827463395,
+      "grad_norm": 144.52732849121094,
+      "learning_rate": 2.439024390243902e-07,
+      "loss": 4.7912,
+      "step": 1850
+    },
+    {
+      "epoch": 0.148397309062129,
+      "grad_norm": 192.0771026611328,
+      "learning_rate": 2.4381999498740256e-07,
+      "loss": 4.7916,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15037593984962405,
+      "grad_norm": 148.06878662109375,
+      "learning_rate": 2.4373755095041484e-07,
+      "loss": 4.7782,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1523545706371191,
+      "grad_norm": 131.4456329345703,
+      "learning_rate": 2.436551069134272e-07,
+      "loss": 4.579,
+      "step": 1925
+    },
+    {
+      "epoch": 0.15433320142461418,
+      "grad_norm": 141.84681701660156,
+      "learning_rate": 2.4357266287643947e-07,
+      "loss": 4.5776,
+      "step": 1950
+    },
+    {
+      "epoch": 0.15631183221210923,
+      "grad_norm": 122.31990051269531,
+      "learning_rate": 2.4349021883945175e-07,
+      "loss": 4.5185,
+      "step": 1975
+    },
+    {
+      "epoch": 0.15829046299960428,
+      "grad_norm": 229.08372497558594,
+      "learning_rate": 2.434077748024641e-07,
+      "loss": 4.6352,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16026909378709933,
+      "grad_norm": 136.54153442382812,
+      "learning_rate": 2.433253307654764e-07,
+      "loss": 4.5512,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16224772457459438,
+      "grad_norm": 237.05514526367188,
+      "learning_rate": 2.432428867284887e-07,
+      "loss": 4.7146,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16422635536208943,
+      "grad_norm": 149.2750244140625,
+      "learning_rate": 2.43160442691501e-07,
+      "loss": 4.6935,
+      "step": 2075
+    },
+    {
+      "epoch": 0.16620498614958448,
+      "grad_norm": 149.77297973632812,
+      "learning_rate": 2.4307799865451334e-07,
+      "loss": 4.8223,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16818361693707953,
+      "grad_norm": 235.3883056640625,
+      "learning_rate": 2.429955546175256e-07,
+      "loss": 4.6266,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17016224772457458,
+      "grad_norm": 137.77316284179688,
+      "learning_rate": 2.429131105805379e-07,
+      "loss": 4.8543,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17214087851206966,
+      "grad_norm": 143.8935089111328,
+      "learning_rate": 2.4283066654355025e-07,
+      "loss": 4.651,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1741195092995647,
+      "grad_norm": 191.43856811523438,
+      "learning_rate": 2.4274822250656253e-07,
+      "loss": 4.4166,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17609814008705976,
+      "grad_norm": 135.82838439941406,
+      "learning_rate": 2.426657784695748e-07,
+      "loss": 4.7078,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1780767708745548,
+      "grad_norm": 114.28646087646484,
+      "learning_rate": 2.4258333443258715e-07,
+      "loss": 4.5316,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18005540166204986,
+      "grad_norm": 237.41001892089844,
+      "learning_rate": 2.4250089039559944e-07,
+      "loss": 4.4699,
+      "step": 2275
+    },
+    {
+      "epoch": 0.1820340324495449,
+      "grad_norm": 124.57892608642578,
+      "learning_rate": 2.424184463586117e-07,
+      "loss": 4.5101,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18401266323703996,
+      "grad_norm": 147.15554809570312,
+      "learning_rate": 2.4233600232162406e-07,
+      "loss": 4.5974,
+      "step": 2325
+    },
+    {
+      "epoch": 0.18599129402453501,
+      "grad_norm": 166.0609588623047,
+      "learning_rate": 2.4225355828463635e-07,
+      "loss": 4.5105,
+      "step": 2350
+    },
+    {
+      "epoch": 0.18796992481203006,
+      "grad_norm": 188.97705078125,
+      "learning_rate": 2.421711142476487e-07,
+      "loss": 4.587,
+      "step": 2375
+    },
+    {
+      "epoch": 0.18994855559952512,
+      "grad_norm": 243.09271240234375,
+      "learning_rate": 2.4208867021066097e-07,
+      "loss": 4.7686,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1919271863870202,
+      "grad_norm": 127.40078735351562,
+      "learning_rate": 2.420062261736733e-07,
+      "loss": 4.4476,
+      "step": 2425
+    },
+    {
+      "epoch": 0.19390581717451524,
+      "grad_norm": 253.8776092529297,
+      "learning_rate": 2.419237821366856e-07,
+      "loss": 4.5478,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1958844479620103,
+      "grad_norm": 123.27115631103516,
+      "learning_rate": 2.4184133809969793e-07,
+      "loss": 4.3502,
+      "step": 2475
+    },
+    {
+      "epoch": 0.19786307874950534,
+      "grad_norm": 138.00375366210938,
+      "learning_rate": 2.417588940627102e-07,
+      "loss": 4.3534,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1998417095370004,
+      "grad_norm": 115.53954315185547,
+      "learning_rate": 2.4167645002572256e-07,
+      "loss": 4.7066,
+      "step": 2525
+    },
+    {
+      "epoch": 0.20182034032449545,
+      "grad_norm": 180.38809204101562,
+      "learning_rate": 2.4159400598873484e-07,
+      "loss": 4.6605,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2037989711119905,
+      "grad_norm": 129.8457489013672,
+      "learning_rate": 2.415115619517472e-07,
+      "loss": 4.3849,
+      "step": 2575
+    },
+    {
+      "epoch": 0.20577760189948555,
+      "grad_norm": 156.64404296875,
+      "learning_rate": 2.4142911791475946e-07,
+      "loss": 4.3434,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2077562326869806,
+      "grad_norm": 162.81320190429688,
+      "learning_rate": 2.4134667387777175e-07,
+      "loss": 4.5466,
+      "step": 2625
+    },
+    {
+      "epoch": 0.20973486347447567,
+      "grad_norm": 128.7244873046875,
+      "learning_rate": 2.412642298407841e-07,
+      "loss": 4.5358,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21171349426197072,
+      "grad_norm": 217.59042358398438,
+      "learning_rate": 2.4118178580379637e-07,
+      "loss": 4.5235,
+      "step": 2675
+    },
+    {
+      "epoch": 0.21369212504946578,
+      "grad_norm": 144.84365844726562,
+      "learning_rate": 2.410993417668087e-07,
+      "loss": 4.3811,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21567075583696083,
+      "grad_norm": 146.22451782226562,
+      "learning_rate": 2.41016897729821e-07,
+      "loss": 4.3797,
+      "step": 2725
+    },
+    {
+      "epoch": 0.21764938662445588,
+      "grad_norm": 198.39772033691406,
+      "learning_rate": 2.409344536928333e-07,
+      "loss": 4.4303,
+      "step": 2750
+    },
+    {
+      "epoch": 0.21962801741195093,
+      "grad_norm": 158.10592651367188,
+      "learning_rate": 2.408520096558456e-07,
+      "loss": 4.3633,
+      "step": 2775
+    },
+    {
+      "epoch": 0.22160664819944598,
+      "grad_norm": 166.79954528808594,
+      "learning_rate": 2.407695656188579e-07,
+      "loss": 4.5392,
+      "step": 2800
+    },
+    {
+      "epoch": 0.22358527898694103,
+      "grad_norm": 207.30593872070312,
+      "learning_rate": 2.406871215818702e-07,
+      "loss": 4.5003,
+      "step": 2825
+    },
+    {
+      "epoch": 0.22556390977443608,
+      "grad_norm": 128.81883239746094,
+      "learning_rate": 2.4060467754488253e-07,
+      "loss": 4.5416,
+      "step": 2850
+    },
+    {
+      "epoch": 0.22754254056193116,
+      "grad_norm": 181.48960876464844,
+      "learning_rate": 2.405222335078948e-07,
+      "loss": 4.1725,
+      "step": 2875
+    },
+    {
+      "epoch": 0.2295211713494262,
+      "grad_norm": 179.47384643554688,
+      "learning_rate": 2.4043978947090715e-07,
+      "loss": 4.5229,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23149980213692126,
+      "grad_norm": 144.242919921875,
+      "learning_rate": 2.4035734543391943e-07,
+      "loss": 4.3295,
+      "step": 2925
+    },
+    {
+      "epoch": 0.2334784329244163,
+      "grad_norm": 177.61968994140625,
+      "learning_rate": 2.402749013969317e-07,
+      "loss": 4.4266,
+      "step": 2950
+    },
+    {
+      "epoch": 0.23545706371191136,
+      "grad_norm": 143.8682861328125,
+      "learning_rate": 2.4019245735994406e-07,
+      "loss": 4.2341,
+      "step": 2975
+    },
+    {
+      "epoch": 0.2374356944994064,
+      "grad_norm": 128.8461151123047,
+      "learning_rate": 2.4011001332295634e-07,
+      "loss": 4.3676,
+      "step": 3000
+    },
+    {
+      "epoch": 0.23941432528690146,
+      "grad_norm": 160.70687866210938,
+      "learning_rate": 2.400275692859687e-07,
+      "loss": 4.3945,
+      "step": 3025
+    },
+    {
+      "epoch": 0.2413929560743965,
+      "grad_norm": 157.65855407714844,
+      "learning_rate": 2.3994512524898097e-07,
+      "loss": 4.4967,
+      "step": 3050
+    },
+    {
+      "epoch": 0.24337158686189156,
+      "grad_norm": 125.79988861083984,
+      "learning_rate": 2.398626812119933e-07,
+      "loss": 4.279,
+      "step": 3075
+    },
+    {
+      "epoch": 0.24535021764938664,
+      "grad_norm": 168.8534698486328,
+      "learning_rate": 2.397802371750056e-07,
+      "loss": 4.4813,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2473288484368817,
+      "grad_norm": 120.4126968383789,
+      "learning_rate": 2.3969779313801793e-07,
+      "loss": 4.1997,
+      "step": 3125
+    },
+    {
+      "epoch": 0.24930747922437674,
+      "grad_norm": 115.56365203857422,
+      "learning_rate": 2.396153491010302e-07,
+      "loss": 4.4076,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2512861100118718,
+      "grad_norm": 152.89859008789062,
+      "learning_rate": 2.3953290506404255e-07,
+      "loss": 4.2893,
+      "step": 3175
+    },
+    {
+      "epoch": 0.25326474079936684,
+      "grad_norm": 177.6272735595703,
+      "learning_rate": 2.3945046102705484e-07,
+      "loss": 4.4892,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2552433715868619,
+      "grad_norm": 131.46661376953125,
+      "learning_rate": 2.393680169900671e-07,
+      "loss": 4.2702,
+      "step": 3225
+    },
+    {
+      "epoch": 0.25722200237435694,
+      "grad_norm": 101.60210418701172,
+      "learning_rate": 2.3928557295307946e-07,
+      "loss": 4.2209,
+      "step": 3250
+    },
+    {
+      "epoch": 0.259200633161852,
+      "grad_norm": 199.7799835205078,
+      "learning_rate": 2.3920312891609174e-07,
+      "loss": 4.1502,
+      "step": 3275
+    },
+    {
+      "epoch": 0.26117926394934704,
+      "grad_norm": 163.44424438476562,
+      "learning_rate": 2.391206848791041e-07,
+      "loss": 4.3423,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 148.59519958496094,
+      "learning_rate": 2.3903824084211637e-07,
+      "loss": 4.4833,
+      "step": 3325
+    },
+    {
+      "epoch": 0.26513652552433714,
+      "grad_norm": 129.75927734375,
+      "learning_rate": 2.3895579680512865e-07,
+      "loss": 4.4745,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2671151563118322,
+      "grad_norm": 126.6795654296875,
+      "learning_rate": 2.38873352768141e-07,
+      "loss": 4.3964,
+      "step": 3375
+    },
+    {
+      "epoch": 0.26909378709932724,
+      "grad_norm": 157.1032257080078,
+      "learning_rate": 2.387909087311533e-07,
+      "loss": 4.3419,
+      "step": 3400
+    },
+    {
+      "epoch": 0.2710724178868223,
+      "grad_norm": 142.79139709472656,
+      "learning_rate": 2.3870846469416556e-07,
+      "loss": 4.2243,
+      "step": 3425
+    },
+    {
+      "epoch": 0.2730510486743174,
+      "grad_norm": 137.3797607421875,
+      "learning_rate": 2.386260206571779e-07,
+      "loss": 4.1661,
+      "step": 3450
+    },
+    {
+      "epoch": 0.27502967946181245,
+      "grad_norm": 148.77401733398438,
+      "learning_rate": 2.385435766201902e-07,
+      "loss": 4.483,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2770083102493075,
+      "grad_norm": 124.54267120361328,
+      "learning_rate": 2.384611325832025e-07,
+      "loss": 4.369,
+      "step": 3500
+    },
+    {
+      "epoch": 0.27898694103680255,
+      "grad_norm": 113.43370056152344,
+      "learning_rate": 2.383786885462148e-07,
+      "loss": 4.1491,
+      "step": 3525
+    },
+    {
+      "epoch": 0.2809655718242976,
+      "grad_norm": 155.67677307128906,
+      "learning_rate": 2.3829624450922712e-07,
+      "loss": 4.3403,
+      "step": 3550
+    },
+    {
+      "epoch": 0.28294420261179265,
+      "grad_norm": 201.27784729003906,
+      "learning_rate": 2.3821380047223943e-07,
+      "loss": 4.3563,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2849228333992877,
+      "grad_norm": 104.74275970458984,
+      "learning_rate": 2.3813135643525174e-07,
+      "loss": 4.2706,
+      "step": 3600
+    },
+    {
+      "epoch": 0.28690146418678275,
+      "grad_norm": 133.6251678466797,
+      "learning_rate": 2.3804891239826405e-07,
+      "loss": 4.3637,
+      "step": 3625
+    },
+    {
+      "epoch": 0.2888800949742778,
+      "grad_norm": 102.35352325439453,
+      "learning_rate": 2.3796646836127634e-07,
+      "loss": 4.2585,
+      "step": 3650
+    },
+    {
+      "epoch": 0.29085872576177285,
+      "grad_norm": 156.72654724121094,
+      "learning_rate": 2.3788402432428868e-07,
+      "loss": 4.3448,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2928373565492679,
+      "grad_norm": 121.19142150878906,
+      "learning_rate": 2.3780158028730096e-07,
+      "loss": 4.1475,
+      "step": 3700
+    },
+    {
+      "epoch": 0.29481598733676295,
+      "grad_norm": 138.72952270507812,
+      "learning_rate": 2.3771913625031327e-07,
+      "loss": 4.2475,
+      "step": 3725
+    },
+    {
+      "epoch": 0.296794618124258,
+      "grad_norm": 314.35113525390625,
+      "learning_rate": 2.3763669221332559e-07,
+      "loss": 4.2643,
+      "step": 3750
+    },
+    {
+      "epoch": 0.29877324891175305,
+      "grad_norm": 131.71240234375,
+      "learning_rate": 2.375542481763379e-07,
+      "loss": 4.2741,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3007518796992481,
+      "grad_norm": 193.2744598388672,
+      "learning_rate": 2.374718041393502e-07,
+      "loss": 4.2314,
+      "step": 3800
+    },
+    {
+      "epoch": 0.30273051048674315,
+      "grad_norm": 146.98760986328125,
+      "learning_rate": 2.3738936010236252e-07,
+      "loss": 4.5421,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3047091412742382,
+      "grad_norm": 106.49159240722656,
+      "learning_rate": 2.373069160653748e-07,
+      "loss": 4.0922,
+      "step": 3850
+    },
+    {
+      "epoch": 0.30668777206173325,
+      "grad_norm": 128.12686157226562,
+      "learning_rate": 2.3722447202838712e-07,
+      "loss": 4.3171,
+      "step": 3875
+    },
+    {
+      "epoch": 0.30866640284922836,
+      "grad_norm": 165.8458251953125,
+      "learning_rate": 2.3714202799139943e-07,
+      "loss": 4.1937,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3106450336367234,
+      "grad_norm": 129.49652099609375,
+      "learning_rate": 2.3705958395441171e-07,
+      "loss": 4.2486,
+      "step": 3925
+    },
+    {
+      "epoch": 0.31262366442421846,
+      "grad_norm": 113.08882141113281,
+      "learning_rate": 2.3697713991742405e-07,
+      "loss": 3.9533,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3146022952117135,
+      "grad_norm": 116.51021575927734,
+      "learning_rate": 2.3689469588043634e-07,
+      "loss": 4.2525,
+      "step": 3975
+    },
+    {
+      "epoch": 0.31658092599920856,
+      "grad_norm": 95.54279327392578,
+      "learning_rate": 2.3681225184344867e-07,
+      "loss": 4.1355,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3185595567867036,
+      "grad_norm": 123.10621643066406,
+      "learning_rate": 2.3672980780646096e-07,
+      "loss": 4.3705,
+      "step": 4025
+    },
+    {
+      "epoch": 0.32053818757419866,
+      "grad_norm": 142.11273193359375,
+      "learning_rate": 2.3664736376947327e-07,
+      "loss": 4.2712,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3225168183616937,
+      "grad_norm": 162.17141723632812,
+      "learning_rate": 2.3656491973248558e-07,
+      "loss": 4.1127,
+      "step": 4075
+    },
+    {
+      "epoch": 0.32449544914918876,
+      "grad_norm": 160.26893615722656,
+      "learning_rate": 2.364824756954979e-07,
+      "loss": 4.2687,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3264740799366838,
+      "grad_norm": 134.65093994140625,
+      "learning_rate": 2.3640003165851018e-07,
+      "loss": 4.3567,
+      "step": 4125
+    },
+    {
+      "epoch": 0.32845271072417886,
+      "grad_norm": 178.23516845703125,
+      "learning_rate": 2.3631758762152252e-07,
+      "loss": 4.097,
+      "step": 4150
+    },
+    {
+      "epoch": 0.3304313415116739,
+      "grad_norm": 151.1556396484375,
+      "learning_rate": 2.362351435845348e-07,
+      "loss": 4.1602,
+      "step": 4175
+    },
+    {
+      "epoch": 0.33240997229916897,
+      "grad_norm": 154.64442443847656,
+      "learning_rate": 2.3615269954754711e-07,
+      "loss": 4.2365,
+      "step": 4200
+    },
+    {
+      "epoch": 0.334388603086664,
+      "grad_norm": 226.6827850341797,
+      "learning_rate": 2.3607025551055943e-07,
+      "loss": 4.3196,
+      "step": 4225
+    },
+    {
+      "epoch": 0.33636723387415907,
+      "grad_norm": 172.67916870117188,
+      "learning_rate": 2.359878114735717e-07,
+      "loss": 4.4476,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3383458646616541,
+      "grad_norm": 124.78984069824219,
+      "learning_rate": 2.3590536743658405e-07,
+      "loss": 4.4006,
+      "step": 4275
+    },
+    {
+      "epoch": 0.34032449544914917,
+      "grad_norm": 156.81365966796875,
+      "learning_rate": 2.3582292339959633e-07,
+      "loss": 4.3914,
+      "step": 4300
+    },
+    {
+      "epoch": 0.3423031262366442,
+      "grad_norm": 116.53181457519531,
+      "learning_rate": 2.3574047936260865e-07,
+      "loss": 4.2846,
+      "step": 4325
+    },
+    {
+      "epoch": 0.3442817570241393,
+      "grad_norm": 146.16543579101562,
+      "learning_rate": 2.3565803532562096e-07,
+      "loss": 4.1371,
+      "step": 4350
+    },
+    {
+      "epoch": 0.3462603878116344,
+      "grad_norm": 213.07974243164062,
+      "learning_rate": 2.3557559128863327e-07,
+      "loss": 4.2294,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3482390185991294,
+      "grad_norm": 99.38206481933594,
+      "learning_rate": 2.3549314725164558e-07,
+      "loss": 4.1726,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3502176493866245,
+      "grad_norm": 162.97059631347656,
+      "learning_rate": 2.354107032146579e-07,
+      "loss": 4.0507,
+      "step": 4425
+    },
+    {
+      "epoch": 0.3521962801741195,
+      "grad_norm": 132.77474975585938,
+      "learning_rate": 2.3532825917767018e-07,
+      "loss": 4.0016,
+      "step": 4450
+    },
+    {
+      "epoch": 0.3541749109616146,
+      "grad_norm": 126.9658203125,
+      "learning_rate": 2.3524581514068252e-07,
+      "loss": 4.2731,
+      "step": 4475
+    },
+    {
+      "epoch": 0.3561535417491096,
+      "grad_norm": 194.47755432128906,
+      "learning_rate": 2.351633711036948e-07,
+      "loss": 4.1119,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3581321725366047,
+      "grad_norm": 153.6606903076172,
+      "learning_rate": 2.3508092706670709e-07,
+      "loss": 4.4556,
+      "step": 4525
+    },
+    {
+      "epoch": 0.3601108033240997,
+      "grad_norm": 146.66709899902344,
+      "learning_rate": 2.3499848302971942e-07,
+      "loss": 4.3314,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3620894341115948,
+      "grad_norm": 111.01129913330078,
+      "learning_rate": 2.349160389927317e-07,
+      "loss": 4.2929,
+      "step": 4575
+    },
+    {
+      "epoch": 0.3640680648990898,
+      "grad_norm": 137.40582275390625,
+      "learning_rate": 2.3483359495574405e-07,
+      "loss": 4.2198,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3660466956865849,
+      "grad_norm": 142.0623779296875,
+      "learning_rate": 2.3475115091875633e-07,
+      "loss": 4.2013,
+      "step": 4625
+    },
+    {
+      "epoch": 0.36802532647407993,
+      "grad_norm": 135.2795867919922,
+      "learning_rate": 2.3466870688176864e-07,
+      "loss": 4.231,
+      "step": 4650
+    },
+    {
+      "epoch": 0.370003957261575,
+      "grad_norm": 127.59281158447266,
+      "learning_rate": 2.3458626284478096e-07,
+      "loss": 3.9613,
+      "step": 4675
+    },
+    {
+      "epoch": 0.37198258804907003,
+      "grad_norm": 132.48663330078125,
+      "learning_rate": 2.3450381880779327e-07,
+      "loss": 4.1925,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3739612188365651,
+      "grad_norm": 135.35409545898438,
+      "learning_rate": 2.3442137477080555e-07,
+      "loss": 4.1828,
+      "step": 4725
+    },
+    {
+      "epoch": 0.37593984962406013,
+      "grad_norm": 107.55503845214844,
+      "learning_rate": 2.343389307338179e-07,
+      "loss": 4.2578,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3779184804115552,
+      "grad_norm": 132.79620361328125,
+      "learning_rate": 2.3425648669683018e-07,
+      "loss": 4.0254,
+      "step": 4775
+    },
+    {
+      "epoch": 0.37989711119905023,
+      "grad_norm": 123.6044692993164,
+      "learning_rate": 2.341740426598425e-07,
+      "loss": 3.9981,
+      "step": 4800
+    },
+    {
+      "epoch": 0.38187574198654534,
+      "grad_norm": 149.656005859375,
+      "learning_rate": 2.340915986228548e-07,
+      "loss": 4.1067,
+      "step": 4825
+    },
+    {
+      "epoch": 0.3838543727740404,
+      "grad_norm": 122.97380065917969,
+      "learning_rate": 2.3400915458586708e-07,
+      "loss": 4.2396,
+      "step": 4850
+    },
+    {
+      "epoch": 0.38583300356153544,
+      "grad_norm": 140.10183715820312,
+      "learning_rate": 2.3392671054887942e-07,
+      "loss": 4.1309,
+      "step": 4875
+    },
+    {
+      "epoch": 0.3878116343490305,
+      "grad_norm": 137.91583251953125,
+      "learning_rate": 2.338442665118917e-07,
+      "loss": 4.0575,
+      "step": 4900
+    },
+    {
+      "epoch": 0.38979026513652554,
+      "grad_norm": 137.72152709960938,
+      "learning_rate": 2.3376182247490402e-07,
+      "loss": 4.159,
+      "step": 4925
+    },
+    {
+      "epoch": 0.3917688959240206,
+      "grad_norm": 84.3819808959961,
+      "learning_rate": 2.3367937843791633e-07,
+      "loss": 4.2611,
+      "step": 4950
+    },
+    {
+      "epoch": 0.39374752671151564,
+      "grad_norm": 200.3111114501953,
+      "learning_rate": 2.3359693440092864e-07,
+      "loss": 4.167,
+      "step": 4975
+    },
+    {
+      "epoch": 0.3957261574990107,
+      "grad_norm": 123.27460479736328,
+      "learning_rate": 2.3351449036394095e-07,
+      "loss": 4.2918,
+      "step": 5000
+    },
+    {
+      "epoch": 0.39770478828650574,
+      "grad_norm": 111.70620727539062,
+      "learning_rate": 2.3343204632695327e-07,
+      "loss": 4.2242,
+      "step": 5025
+    },
+    {
+      "epoch": 0.3996834190740008,
+      "grad_norm": 107.74165344238281,
+      "learning_rate": 2.3334960228996555e-07,
+      "loss": 4.3572,
+      "step": 5050
+    },
+    {
+      "epoch": 0.40166204986149584,
+      "grad_norm": 138.31423950195312,
+      "learning_rate": 2.332671582529779e-07,
+      "loss": 4.1759,
+      "step": 5075
+    },
+    {
+      "epoch": 0.4036406806489909,
+      "grad_norm": 104.73587799072266,
+      "learning_rate": 2.3318471421599017e-07,
+      "loss": 4.1695,
+      "step": 5100
+    },
+    {
+      "epoch": 0.40561931143648594,
+      "grad_norm": 138.1061553955078,
+      "learning_rate": 2.3310227017900246e-07,
+      "loss": 4.0986,
+      "step": 5125
+    },
+    {
+      "epoch": 0.407597942223981,
+      "grad_norm": 148.92279052734375,
+      "learning_rate": 2.330198261420148e-07,
+      "loss": 4.3455,
+      "step": 5150
+    },
+    {
+      "epoch": 0.40957657301147604,
+      "grad_norm": 321.29852294921875,
+      "learning_rate": 2.3293738210502708e-07,
+      "loss": 4.1285,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4115552037989711,
+      "grad_norm": 114.85989379882812,
+      "learning_rate": 2.3285493806803942e-07,
+      "loss": 3.9628,
+      "step": 5200
+    },
+    {
+      "epoch": 0.41353383458646614,
+      "grad_norm": 137.27610778808594,
+      "learning_rate": 2.327724940310517e-07,
+      "loss": 4.1521,
+      "step": 5225
+    },
+    {
+      "epoch": 0.4155124653739612,
+      "grad_norm": 96.02686309814453,
+      "learning_rate": 2.3269004999406402e-07,
+      "loss": 4.027,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4174910961614563,
+      "grad_norm": 213.81649780273438,
+      "learning_rate": 2.3260760595707633e-07,
+      "loss": 4.0522,
+      "step": 5275
+    },
+    {
+      "epoch": 0.41946972694895135,
+      "grad_norm": 160.4125518798828,
+      "learning_rate": 2.3252516192008864e-07,
+      "loss": 4.09,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4214483577364464,
+      "grad_norm": 167.58741760253906,
+      "learning_rate": 2.3244271788310093e-07,
+      "loss": 4.1128,
+      "step": 5325
+    },
+    {
+      "epoch": 0.42342698852394145,
+      "grad_norm": 159.55303955078125,
+      "learning_rate": 2.3236027384611326e-07,
+      "loss": 4.0867,
+      "step": 5350
+    },
+    {
+      "epoch": 0.4254056193114365,
+      "grad_norm": 122.51324462890625,
+      "learning_rate": 2.3227782980912555e-07,
+      "loss": 4.2261,
+      "step": 5375
+    },
+    {
+      "epoch": 0.42738425009893155,
+      "grad_norm": 185.9108428955078,
+      "learning_rate": 2.3219538577213789e-07,
+      "loss": 3.9684,
+      "step": 5400
+    },
+    {
+      "epoch": 0.4293628808864266,
+      "grad_norm": 195.37579345703125,
+      "learning_rate": 2.3211294173515017e-07,
+      "loss": 4.0779,
+      "step": 5425
+    },
+    {
+      "epoch": 0.43134151167392165,
+      "grad_norm": 157.84371948242188,
+      "learning_rate": 2.3203049769816246e-07,
+      "loss": 4.1991,
+      "step": 5450
+    },
+    {
+      "epoch": 0.4333201424614167,
+      "grad_norm": 111.01512908935547,
+      "learning_rate": 2.319480536611748e-07,
+      "loss": 3.9962,
+      "step": 5475
+    },
+    {
+      "epoch": 0.43529877324891175,
+      "grad_norm": 114.49053955078125,
+      "learning_rate": 2.3186560962418708e-07,
+      "loss": 3.8972,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4372774040364068,
+      "grad_norm": 168.17874145507812,
+      "learning_rate": 2.317831655871994e-07,
+      "loss": 4.1913,
+      "step": 5525
+    },
+    {
+      "epoch": 0.43925603482390185,
+      "grad_norm": 140.61912536621094,
+      "learning_rate": 2.317007215502117e-07,
+      "loss": 4.1396,
+      "step": 5550
+    },
+    {
+      "epoch": 0.4412346656113969,
+      "grad_norm": 138.01805114746094,
+      "learning_rate": 2.3161827751322401e-07,
+      "loss": 4.1399,
+      "step": 5575
+    },
+    {
+      "epoch": 0.44321329639889195,
+      "grad_norm": 188.0181427001953,
+      "learning_rate": 2.3153583347623633e-07,
+      "loss": 4.0329,
+      "step": 5600
+    },
+    {
+      "epoch": 0.445191927186387,
+      "grad_norm": 170.8402099609375,
+      "learning_rate": 2.3145338943924864e-07,
+      "loss": 4.3414,
+      "step": 5625
+    },
+    {
+      "epoch": 0.44717055797388205,
+      "grad_norm": 200.65077209472656,
+      "learning_rate": 2.3137094540226092e-07,
+      "loss": 4.2154,
+      "step": 5650
+    },
+    {
+      "epoch": 0.4491491887613771,
+      "grad_norm": 120.18091583251953,
+      "learning_rate": 2.3128850136527326e-07,
+      "loss": 4.0372,
+      "step": 5675
+    },
+    {
+      "epoch": 0.45112781954887216,
+      "grad_norm": 89.9730224609375,
+      "learning_rate": 2.3120605732828555e-07,
+      "loss": 4.1059,
+      "step": 5700
+    },
+    {
+      "epoch": 0.45310645033636726,
+      "grad_norm": 133.7999267578125,
+      "learning_rate": 2.3112361329129786e-07,
+      "loss": 4.2035,
+      "step": 5725
+    },
+    {
+      "epoch": 0.4550850811238623,
+      "grad_norm": 88.3386459350586,
+      "learning_rate": 2.3104116925431017e-07,
+      "loss": 4.0566,
+      "step": 5750
+    },
+    {
+      "epoch": 0.45706371191135736,
+      "grad_norm": 130.95127868652344,
+      "learning_rate": 2.3095872521732245e-07,
+      "loss": 4.3084,
+      "step": 5775
+    },
+    {
+      "epoch": 0.4590423426988524,
+      "grad_norm": 162.55679321289062,
+      "learning_rate": 2.308762811803348e-07,
+      "loss": 4.0288,
+      "step": 5800
+    },
+    {
+      "epoch": 0.46102097348634746,
+      "grad_norm": 104.4178695678711,
+      "learning_rate": 2.3079383714334708e-07,
+      "loss": 3.9244,
+      "step": 5825
+    },
+    {
+      "epoch": 0.4629996042738425,
+      "grad_norm": 235.28123474121094,
+      "learning_rate": 2.307113931063594e-07,
+      "loss": 4.1106,
+      "step": 5850
+    },
+    {
+      "epoch": 0.46497823506133756,
+      "grad_norm": 289.6645812988281,
+      "learning_rate": 2.306289490693717e-07,
+      "loss": 4.0457,
+      "step": 5875
+    },
+    {
+      "epoch": 0.4669568658488326,
+      "grad_norm": 99.97111511230469,
+      "learning_rate": 2.30546505032384e-07,
+      "loss": 4.2542,
+      "step": 5900
+    },
+    {
+      "epoch": 0.46893549663632766,
+      "grad_norm": 260.0950622558594,
+      "learning_rate": 2.304640609953963e-07,
+      "loss": 4.1564,
+      "step": 5925
+    },
+    {
+      "epoch": 0.4709141274238227,
+      "grad_norm": 113.74392700195312,
+      "learning_rate": 2.3038161695840864e-07,
+      "loss": 4.0403,
+      "step": 5950
+    },
+    {
+      "epoch": 0.47289275821131777,
+      "grad_norm": 79.32340240478516,
+      "learning_rate": 2.3029917292142092e-07,
+      "loss": 4.0408,
+      "step": 5975
+    },
+    {
+      "epoch": 0.4748713889988128,
+      "grad_norm": 95.92308807373047,
+      "learning_rate": 2.3021672888443326e-07,
+      "loss": 3.9811,
+      "step": 6000
+    },
+    {
+      "epoch": 0.47685001978630787,
+      "grad_norm": 94.5758285522461,
+      "learning_rate": 2.3013428484744554e-07,
+      "loss": 4.1102,
+      "step": 6025
+    },
+    {
+      "epoch": 0.4788286505738029,
+      "grad_norm": 142.32131958007812,
+      "learning_rate": 2.3005184081045786e-07,
+      "loss": 3.989,
+      "step": 6050
+    },
+    {
+      "epoch": 0.48080728136129797,
+      "grad_norm": 97.84469604492188,
+      "learning_rate": 2.2996939677347017e-07,
+      "loss": 3.9512,
+      "step": 6075
+    },
+    {
+      "epoch": 0.482785912148793,
+      "grad_norm": 94.38491821289062,
+      "learning_rate": 2.2988695273648245e-07,
+      "loss": 3.9475,
+      "step": 6100
+    },
+    {
+      "epoch": 0.48476454293628807,
+      "grad_norm": 124.32872772216797,
+      "learning_rate": 2.2980450869949476e-07,
+      "loss": 4.1352,
+      "step": 6125
+    },
+    {
+      "epoch": 0.4867431737237831,
+      "grad_norm": 196.1511993408203,
+      "learning_rate": 2.2972206466250708e-07,
+      "loss": 4.2956,
+      "step": 6150
+    },
+    {
+      "epoch": 0.48872180451127817,
+      "grad_norm": 144.1227264404297,
+      "learning_rate": 2.296396206255194e-07,
+      "loss": 3.9718,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4907004352987733,
+      "grad_norm": 115.52275085449219,
+      "learning_rate": 2.295571765885317e-07,
+      "loss": 3.9135,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4926790660862683,
+      "grad_norm": 117.71548461914062,
+      "learning_rate": 2.29474732551544e-07,
+      "loss": 3.9026,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4946576968737634,
+      "grad_norm": 135.42698669433594,
+      "learning_rate": 2.293922885145563e-07,
+      "loss": 4.0369,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4966363276612584,
+      "grad_norm": 142.4741973876953,
+      "learning_rate": 2.2930984447756863e-07,
+      "loss": 4.3588,
+      "step": 6275
+    },
+    {
+      "epoch": 0.4986149584487535,
+      "grad_norm": 128.56195068359375,
+      "learning_rate": 2.2922740044058092e-07,
+      "loss": 3.9089,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5005935892362485,
+      "grad_norm": 96.84894561767578,
+      "learning_rate": 2.2914495640359323e-07,
+      "loss": 4.1722,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5025722200237436,
+      "grad_norm": 236.92965698242188,
+      "learning_rate": 2.2906251236660554e-07,
+      "loss": 3.9729,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5045508508112386,
+      "grad_norm": 135.83609008789062,
+      "learning_rate": 2.2898006832961783e-07,
+      "loss": 4.0322,
+      "step": 6375
+    },
+    {
+      "epoch": 0.5065294815987337,
+      "grad_norm": 123.36375427246094,
+      "learning_rate": 2.2889762429263017e-07,
+      "loss": 4.0042,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5085081123862287,
+      "grad_norm": 118.30574035644531,
+      "learning_rate": 2.2881518025564245e-07,
+      "loss": 4.1079,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5104867431737238,
+      "grad_norm": 107.81358337402344,
+      "learning_rate": 2.2873273621865476e-07,
+      "loss": 4.1198,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5124653739612188,
+      "grad_norm": 146.2493438720703,
+      "learning_rate": 2.2865029218166707e-07,
+      "loss": 4.0814,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5144440047487139,
+      "grad_norm": 136.8212890625,
+      "learning_rate": 2.2856784814467939e-07,
+      "loss": 4.0562,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5164226355362089,
+      "grad_norm": 139.30670166015625,
+      "learning_rate": 2.2848540410769167e-07,
+      "loss": 4.1199,
+      "step": 6525
+    },
+    {
+      "epoch": 0.518401266323704,
+      "grad_norm": 194.90414428710938,
+      "learning_rate": 2.28402960070704e-07,
+      "loss": 4.0562,
+      "step": 6550
+    },
+    {
+      "epoch": 0.520379897111199,
+      "grad_norm": 103.54257202148438,
+      "learning_rate": 2.283205160337163e-07,
+      "loss": 4.0797,
+      "step": 6575
+    },
+    {
+      "epoch": 0.5223585278986941,
+      "grad_norm": 101.63102722167969,
+      "learning_rate": 2.2823807199672863e-07,
+      "loss": 4.0591,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5243371586861891,
+      "grad_norm": 104.28479766845703,
+      "learning_rate": 2.2815562795974092e-07,
+      "loss": 3.8991,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 166.01107788085938,
+      "learning_rate": 2.2807318392275323e-07,
+      "loss": 4.1011,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5282944202611792,
+      "grad_norm": 154.64959716796875,
+      "learning_rate": 2.2799073988576554e-07,
+      "loss": 3.9283,
+      "step": 6675
+    },
+    {
+      "epoch": 0.5302730510486743,
+      "grad_norm": 96.0099868774414,
+      "learning_rate": 2.2790829584877782e-07,
+      "loss": 3.8247,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5322516818361693,
+      "grad_norm": 120.90514373779297,
+      "learning_rate": 2.2782585181179014e-07,
+      "loss": 4.0629,
+      "step": 6725
+    },
+    {
+      "epoch": 0.5342303126236644,
+      "grad_norm": 106.48863983154297,
+      "learning_rate": 2.2774340777480245e-07,
+      "loss": 4.0127,
+      "step": 6750
+    },
+    {
+      "epoch": 0.5362089434111594,
+      "grad_norm": 113.17047882080078,
+      "learning_rate": 2.2766096373781476e-07,
+      "loss": 4.03,
+      "step": 6775
+    },
+    {
+      "epoch": 0.5381875741986545,
+      "grad_norm": 130.6500701904297,
+      "learning_rate": 2.2757851970082707e-07,
+      "loss": 4.0192,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5401662049861495,
+      "grad_norm": 142.3747100830078,
+      "learning_rate": 2.2749607566383938e-07,
+      "loss": 4.1507,
+      "step": 6825
+    },
+    {
+      "epoch": 0.5421448357736446,
+      "grad_norm": 125.88548278808594,
+      "learning_rate": 2.2741363162685167e-07,
+      "loss": 4.2026,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5441234665611397,
+      "grad_norm": 156.44570922851562,
+      "learning_rate": 2.27331187589864e-07,
+      "loss": 4.1063,
+      "step": 6875
+    },
+    {
+      "epoch": 0.5461020973486348,
+      "grad_norm": 150.82635498046875,
+      "learning_rate": 2.272487435528763e-07,
+      "loss": 4.0477,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5480807281361298,
+      "grad_norm": 170.67994689941406,
+      "learning_rate": 2.271662995158886e-07,
+      "loss": 4.1398,
+      "step": 6925
+    },
+    {
+      "epoch": 0.5500593589236249,
+      "grad_norm": 114.224609375,
+      "learning_rate": 2.2708385547890091e-07,
+      "loss": 4.0906,
+      "step": 6950
+    },
+    {
+      "epoch": 0.55203798971112,
+      "grad_norm": 135.5966033935547,
+      "learning_rate": 2.2700141144191323e-07,
+      "loss": 3.872,
+      "step": 6975
+    },
+    {
+      "epoch": 0.554016620498615,
+      "grad_norm": 120.73974609375,
+      "learning_rate": 2.2691896740492554e-07,
+      "loss": 3.9762,
+      "step": 7000
+    },
+    {
+      "epoch": 0.55599525128611,
+      "grad_norm": 107.66891479492188,
+      "learning_rate": 2.2683652336793782e-07,
+      "loss": 4.0551,
+      "step": 7025
+    },
+    {
+      "epoch": 0.5579738820736051,
+      "grad_norm": 107.60162353515625,
+      "learning_rate": 2.2675407933095013e-07,
+      "loss": 3.973,
+      "step": 7050
+    },
+    {
+      "epoch": 0.5599525128611001,
+      "grad_norm": 118.88258361816406,
+      "learning_rate": 2.2667163529396245e-07,
+      "loss": 3.9864,
+      "step": 7075
+    },
+    {
+      "epoch": 0.5619311436485952,
+      "grad_norm": 148.85667419433594,
+      "learning_rate": 2.2658919125697476e-07,
+      "loss": 3.9409,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5639097744360902,
+      "grad_norm": 148.57321166992188,
+      "learning_rate": 2.2650674721998704e-07,
+      "loss": 3.9611,
+      "step": 7125
+    },
+    {
+      "epoch": 0.5658884052235853,
+      "grad_norm": 172.39999389648438,
+      "learning_rate": 2.2642430318299938e-07,
+      "loss": 3.97,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5678670360110804,
+      "grad_norm": 120.57051086425781,
+      "learning_rate": 2.2634185914601167e-07,
+      "loss": 3.9352,
+      "step": 7175
+    },
+    {
+      "epoch": 0.5698456667985754,
+      "grad_norm": 143.2531280517578,
+      "learning_rate": 2.26259415109024e-07,
+      "loss": 3.9686,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5718242975860705,
+      "grad_norm": 123.57396697998047,
+      "learning_rate": 2.261769710720363e-07,
+      "loss": 4.0855,
+      "step": 7225
+    },
+    {
+      "epoch": 0.5738029283735655,
+      "grad_norm": 115.12631225585938,
+      "learning_rate": 2.260945270350486e-07,
+      "loss": 3.9754,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5757815591610606,
+      "grad_norm": 114.95091247558594,
+      "learning_rate": 2.260120829980609e-07,
+      "loss": 3.8981,
+      "step": 7275
+    },
+    {
+      "epoch": 0.5777601899485556,
+      "grad_norm": 105.46833038330078,
+      "learning_rate": 2.2592963896107322e-07,
+      "loss": 3.9452,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5797388207360507,
+      "grad_norm": 132.89012145996094,
+      "learning_rate": 2.258471949240855e-07,
+      "loss": 4.0808,
+      "step": 7325
+    },
+    {
+      "epoch": 0.5817174515235457,
+      "grad_norm": 143.6460418701172,
+      "learning_rate": 2.2576475088709782e-07,
+      "loss": 4.0108,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5836960823110408,
+      "grad_norm": 130.83352661132812,
+      "learning_rate": 2.2568230685011013e-07,
+      "loss": 3.9701,
+      "step": 7375
+    },
+    {
+      "epoch": 0.5856747130985358,
+      "grad_norm": 111.10405731201172,
+      "learning_rate": 2.2559986281312244e-07,
+      "loss": 4.3162,
+      "step": 7400
+    },
+    {
+      "epoch": 0.5876533438860309,
+      "grad_norm": 163.31959533691406,
+      "learning_rate": 2.2551741877613476e-07,
+      "loss": 3.9096,
+      "step": 7425
+    },
+    {
+      "epoch": 0.5896319746735259,
+      "grad_norm": 134.72927856445312,
+      "learning_rate": 2.2543497473914704e-07,
+      "loss": 3.8888,
+      "step": 7450
+    },
+    {
+      "epoch": 0.591610605461021,
+      "grad_norm": 124.26619720458984,
+      "learning_rate": 2.2535253070215938e-07,
+      "loss": 4.0683,
+      "step": 7475
+    },
+    {
+      "epoch": 0.593589236248516,
+      "grad_norm": 106.8174057006836,
+      "learning_rate": 2.2527008666517166e-07,
+      "loss": 4.0547,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5955678670360111,
+      "grad_norm": 108.11019897460938,
+      "learning_rate": 2.2518764262818398e-07,
+      "loss": 3.9728,
+      "step": 7525
+    },
+    {
+      "epoch": 0.5975464978235061,
+      "grad_norm": 117.44151306152344,
+      "learning_rate": 2.251051985911963e-07,
+      "loss": 4.0569,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5995251286110012,
+      "grad_norm": 106.18008422851562,
+      "learning_rate": 2.250227545542086e-07,
+      "loss": 3.9042,
+      "step": 7575
+    },
+    {
+      "epoch": 0.6015037593984962,
+      "grad_norm": 88.67406463623047,
+      "learning_rate": 2.249403105172209e-07,
+      "loss": 4.0719,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6034823901859913,
+      "grad_norm": 111.12770080566406,
+      "learning_rate": 2.248578664802332e-07,
+      "loss": 3.9749,
+      "step": 7625
+    },
+    {
+      "epoch": 0.6054610209734863,
+      "grad_norm": 119.26530456542969,
+      "learning_rate": 2.247754224432455e-07,
+      "loss": 3.9832,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6074396517609814,
+      "grad_norm": 157.9289093017578,
+      "learning_rate": 2.2469297840625782e-07,
+      "loss": 3.9538,
+      "step": 7675
+    },
+    {
+      "epoch": 0.6094182825484764,
+      "grad_norm": 122.70995330810547,
+      "learning_rate": 2.2461053436927013e-07,
+      "loss": 3.8497,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6113969133359715,
+      "grad_norm": 142.41835021972656,
+      "learning_rate": 2.2452809033228242e-07,
+      "loss": 3.9172,
+      "step": 7725
+    },
+    {
+      "epoch": 0.6133755441234665,
+      "grad_norm": 128.31825256347656,
+      "learning_rate": 2.2444564629529475e-07,
+      "loss": 3.7326,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6153541749109616,
+      "grad_norm": 142.67408752441406,
+      "learning_rate": 2.2436320225830704e-07,
+      "loss": 3.8782,
+      "step": 7775
+    },
+    {
+      "epoch": 0.6173328056984567,
+      "grad_norm": 145.0731658935547,
+      "learning_rate": 2.2428075822131938e-07,
+      "loss": 4.024,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6193114364859518,
+      "grad_norm": 187.09068298339844,
+      "learning_rate": 2.2419831418433166e-07,
+      "loss": 3.8939,
+      "step": 7825
+    },
+    {
+      "epoch": 0.6212900672734468,
+      "grad_norm": 122.93965148925781,
+      "learning_rate": 2.2411587014734397e-07,
+      "loss": 4.0373,
+      "step": 7850
+    },
+    {
+      "epoch": 0.6232686980609419,
+      "grad_norm": 152.1845245361328,
+      "learning_rate": 2.2403342611035628e-07,
+      "loss": 4.1168,
+      "step": 7875
+    },
+    {
+      "epoch": 0.6252473288484369,
+      "grad_norm": 100.07666778564453,
+      "learning_rate": 2.239509820733686e-07,
+      "loss": 3.9718,
+      "step": 7900
+    },
+    {
+      "epoch": 0.627225959635932,
+      "grad_norm": 130.85479736328125,
+      "learning_rate": 2.2386853803638088e-07,
+      "loss": 4.0301,
+      "step": 7925
+    },
+    {
+      "epoch": 0.629204590423427,
+      "grad_norm": 123.073974609375,
+      "learning_rate": 2.237860939993932e-07,
+      "loss": 4.0104,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6311832212109221,
+      "grad_norm": 168.19808959960938,
+      "learning_rate": 2.237036499624055e-07,
+      "loss": 3.9421,
+      "step": 7975
+    },
+    {
+      "epoch": 0.6331618519984171,
+      "grad_norm": 118.69593811035156,
+      "learning_rate": 2.2362120592541782e-07,
+      "loss": 3.8238,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6351404827859122,
+      "grad_norm": 192.9334259033203,
+      "learning_rate": 2.2353876188843013e-07,
+      "loss": 3.8227,
+      "step": 8025
+    },
+    {
+      "epoch": 0.6371191135734072,
+      "grad_norm": 103.11824035644531,
+      "learning_rate": 2.2345631785144241e-07,
+      "loss": 3.9359,
+      "step": 8050
+    },
+    {
+      "epoch": 0.6390977443609023,
+      "grad_norm": 129.3599090576172,
+      "learning_rate": 2.2337387381445475e-07,
+      "loss": 4.0562,
+      "step": 8075
+    },
+    {
+      "epoch": 0.6410763751483973,
+      "grad_norm": 124.06795501708984,
+      "learning_rate": 2.2329142977746704e-07,
+      "loss": 4.1502,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6430550059358924,
+      "grad_norm": 113.18289184570312,
+      "learning_rate": 2.2320898574047935e-07,
+      "loss": 4.0059,
+      "step": 8125
+    },
+    {
+      "epoch": 0.6450336367233874,
+      "grad_norm": 117.89970397949219,
+      "learning_rate": 2.2312654170349166e-07,
+      "loss": 4.0162,
+      "step": 8150
+    },
+    {
+      "epoch": 0.6470122675108825,
+      "grad_norm": 109.6517105102539,
+      "learning_rate": 2.2304409766650397e-07,
+      "loss": 3.9979,
+      "step": 8175
+    },
+    {
+      "epoch": 0.6489908982983775,
+      "grad_norm": 123.35499572753906,
+      "learning_rate": 2.2296165362951628e-07,
+      "loss": 3.9273,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6509695290858726,
+      "grad_norm": 141.97459411621094,
+      "learning_rate": 2.228792095925286e-07,
+      "loss": 4.0558,
+      "step": 8225
+    },
+    {
+      "epoch": 0.6529481598733676,
+      "grad_norm": 159.06973266601562,
+      "learning_rate": 2.2279676555554088e-07,
+      "loss": 3.8374,
+      "step": 8250
+    },
+    {
+      "epoch": 0.6549267906608627,
+      "grad_norm": 120.933837890625,
+      "learning_rate": 2.227143215185532e-07,
+      "loss": 3.8412,
+      "step": 8275
+    },
+    {
+      "epoch": 0.6569054214483577,
+      "grad_norm": 106.266357421875,
+      "learning_rate": 2.226318774815655e-07,
+      "loss": 3.8757,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6588840522358528,
+      "grad_norm": 138.7765655517578,
+      "learning_rate": 2.225494334445778e-07,
+      "loss": 4.2284,
+      "step": 8325
+    },
+    {
+      "epoch": 0.6608626830233478,
+      "grad_norm": 120.76045989990234,
+      "learning_rate": 2.2246698940759013e-07,
+      "loss": 3.9127,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6628413138108429,
+      "grad_norm": 117.31808471679688,
+      "learning_rate": 2.223845453706024e-07,
+      "loss": 3.7577,
+      "step": 8375
+    },
+    {
+      "epoch": 0.6648199445983379,
+      "grad_norm": 108.21405029296875,
+      "learning_rate": 2.2230210133361475e-07,
+      "loss": 4.1095,
+      "step": 8400
+    },
+    {
+      "epoch": 0.666798575385833,
+      "grad_norm": 126.65251159667969,
+      "learning_rate": 2.2221965729662703e-07,
+      "loss": 3.9047,
+      "step": 8425
+    },
+    {
+      "epoch": 0.668777206173328,
+      "grad_norm": 135.06512451171875,
+      "learning_rate": 2.2213721325963935e-07,
+      "loss": 3.9267,
+      "step": 8450
+    },
+    {
+      "epoch": 0.6707558369608231,
+      "grad_norm": 150.37025451660156,
+      "learning_rate": 2.2205476922265166e-07,
+      "loss": 3.9662,
+      "step": 8475
+    },
+    {
+      "epoch": 0.6727344677483181,
+      "grad_norm": 138.01531982421875,
+      "learning_rate": 2.2197232518566397e-07,
+      "loss": 3.8107,
+      "step": 8500
+    },
+    {
+      "epoch": 0.6747130985358132,
+      "grad_norm": 130.35153198242188,
+      "learning_rate": 2.2188988114867625e-07,
+      "loss": 3.8068,
+      "step": 8525
+    },
+    {
+      "epoch": 0.6766917293233082,
+      "grad_norm": 161.9180145263672,
+      "learning_rate": 2.218074371116886e-07,
+      "loss": 4.0893,
+      "step": 8550
+    },
+    {
+      "epoch": 0.6786703601108033,
+      "grad_norm": 165.08409118652344,
+      "learning_rate": 2.2172499307470088e-07,
+      "loss": 3.8419,
+      "step": 8575
+    },
+    {
+      "epoch": 0.6806489908982983,
+      "grad_norm": 153.2915496826172,
+      "learning_rate": 2.216425490377132e-07,
+      "loss": 3.9302,
+      "step": 8600
+    },
+    {
+      "epoch": 0.6826276216857934,
+      "grad_norm": 153.20138549804688,
+      "learning_rate": 2.215601050007255e-07,
+      "loss": 3.9947,
+      "step": 8625
+    },
+    {
+      "epoch": 0.6846062524732884,
+      "grad_norm": 124.32341003417969,
+      "learning_rate": 2.2147766096373779e-07,
+      "loss": 3.8241,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6865848832607835,
+      "grad_norm": 209.813232421875,
+      "learning_rate": 2.2139521692675012e-07,
+      "loss": 3.917,
+      "step": 8675
+    },
+    {
+      "epoch": 0.6885635140482786,
+      "grad_norm": 116.88125610351562,
+      "learning_rate": 2.213127728897624e-07,
+      "loss": 3.9474,
+      "step": 8700
+    },
+    {
+      "epoch": 0.6905421448357737,
+      "grad_norm": 178.58721923828125,
+      "learning_rate": 2.2123032885277472e-07,
+      "loss": 3.9247,
+      "step": 8725
+    },
+    {
+      "epoch": 0.6925207756232687,
+      "grad_norm": 123.67437744140625,
+      "learning_rate": 2.2114788481578703e-07,
+      "loss": 3.9548,
+      "step": 8750
+    },
+    {
+      "epoch": 0.6944994064107638,
+      "grad_norm": 154.60626220703125,
+      "learning_rate": 2.2106544077879934e-07,
+      "loss": 3.8239,
+      "step": 8775
+    },
+    {
+      "epoch": 0.6964780371982588,
+      "grad_norm": 141.65699768066406,
+      "learning_rate": 2.2098299674181166e-07,
+      "loss": 3.8458,
+      "step": 8800
+    },
+    {
+      "epoch": 0.6984566679857539,
+      "grad_norm": 112.9280776977539,
+      "learning_rate": 2.2090055270482397e-07,
+      "loss": 3.9648,
+      "step": 8825
+    },
+    {
+      "epoch": 0.700435298773249,
+      "grad_norm": 154.57643127441406,
+      "learning_rate": 2.2081810866783625e-07,
+      "loss": 4.0078,
+      "step": 8850
+    },
+    {
+      "epoch": 0.702413929560744,
+      "grad_norm": 129.07418823242188,
+      "learning_rate": 2.207356646308486e-07,
+      "loss": 4.0276,
+      "step": 8875
+    },
+    {
+      "epoch": 0.704392560348239,
+      "grad_norm": 113.59859466552734,
+      "learning_rate": 2.2065322059386088e-07,
+      "loss": 4.1596,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7063711911357341,
+      "grad_norm": 136.26283264160156,
+      "learning_rate": 2.2057077655687316e-07,
+      "loss": 3.948,
+      "step": 8925
+    },
+    {
+      "epoch": 0.7083498219232292,
+      "grad_norm": 118.27870178222656,
+      "learning_rate": 2.204883325198855e-07,
+      "loss": 3.7611,
+      "step": 8950
+    },
+    {
+      "epoch": 0.7103284527107242,
+      "grad_norm": 159.56643676757812,
+      "learning_rate": 2.2040588848289778e-07,
+      "loss": 4.0215,
+      "step": 8975
+    },
+    {
+      "epoch": 0.7123070834982193,
+      "grad_norm": 125.84573364257812,
+      "learning_rate": 2.2032344444591012e-07,
+      "loss": 4.0046,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 148.7548065185547,
+      "learning_rate": 2.202410004089224e-07,
+      "loss": 4.0033,
+      "step": 9025
+    },
+    {
+      "epoch": 0.7162643450732094,
+      "grad_norm": 109.41517639160156,
+      "learning_rate": 2.2015855637193472e-07,
+      "loss": 3.9298,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7182429758607044,
+      "grad_norm": 112.52848815917969,
+      "learning_rate": 2.2007611233494703e-07,
+      "loss": 4.1003,
+      "step": 9075
+    },
+    {
+      "epoch": 0.7202216066481995,
+      "grad_norm": 114.72808074951172,
+      "learning_rate": 2.1999366829795934e-07,
+      "loss": 4.1112,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7222002374356945,
+      "grad_norm": 144.11619567871094,
+      "learning_rate": 2.1991122426097163e-07,
+      "loss": 3.764,
+      "step": 9125
+    },
+    {
+      "epoch": 0.7241788682231896,
+      "grad_norm": 118.64055633544922,
+      "learning_rate": 2.1982878022398396e-07,
+      "loss": 3.9262,
+      "step": 9150
+    },
+    {
+      "epoch": 0.7261574990106846,
+      "grad_norm": 166.79525756835938,
+      "learning_rate": 2.1974633618699625e-07,
+      "loss": 4.0518,
+      "step": 9175
+    },
+    {
+      "epoch": 0.7281361297981797,
+      "grad_norm": 128.2512969970703,
+      "learning_rate": 2.1966389215000856e-07,
+      "loss": 3.952,
+      "step": 9200
+    },
+    {
+      "epoch": 0.7301147605856747,
+      "grad_norm": 143.56414794921875,
+      "learning_rate": 2.1958144811302087e-07,
+      "loss": 3.7034,
+      "step": 9225
+    },
+    {
+      "epoch": 0.7320933913731698,
+      "grad_norm": 120.13394165039062,
+      "learning_rate": 2.1949900407603316e-07,
+      "loss": 3.7839,
+      "step": 9250
+    },
+    {
+      "epoch": 0.7340720221606648,
+      "grad_norm": 148.74070739746094,
+      "learning_rate": 2.194165600390455e-07,
+      "loss": 3.8871,
+      "step": 9275
+    },
+    {
+      "epoch": 0.7360506529481599,
+      "grad_norm": 148.17022705078125,
+      "learning_rate": 2.1933411600205778e-07,
+      "loss": 3.7486,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7380292837356549,
+      "grad_norm": 112.7260513305664,
+      "learning_rate": 2.192516719650701e-07,
+      "loss": 3.9436,
+      "step": 9325
+    },
+    {
+      "epoch": 0.74000791452315,
+      "grad_norm": 131.4718780517578,
+      "learning_rate": 2.191692279280824e-07,
+      "loss": 4.1101,
+      "step": 9350
+    },
+    {
+      "epoch": 0.741986545310645,
+      "grad_norm": 106.73101043701172,
+      "learning_rate": 2.1908678389109472e-07,
+      "loss": 3.9285,
+      "step": 9375
+    },
+    {
+      "epoch": 0.7439651760981401,
+      "grad_norm": 120.58040618896484,
+      "learning_rate": 2.1900433985410703e-07,
+      "loss": 3.8471,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7459438068856351,
+      "grad_norm": 135.69512939453125,
+      "learning_rate": 2.1892189581711934e-07,
+      "loss": 3.7629,
+      "step": 9425
+    },
+    {
+      "epoch": 0.7479224376731302,
+      "grad_norm": 125.78627014160156,
+      "learning_rate": 2.1883945178013162e-07,
+      "loss": 4.0646,
+      "step": 9450
+    },
+    {
+      "epoch": 0.7499010684606252,
+      "grad_norm": 150.2305145263672,
+      "learning_rate": 2.1875700774314396e-07,
+      "loss": 3.9361,
+      "step": 9475
+    },
+    {
+      "epoch": 0.7518796992481203,
+      "grad_norm": 95.4436264038086,
+      "learning_rate": 2.1867456370615625e-07,
+      "loss": 3.7688,
+      "step": 9500
+    },
+    {
+      "epoch": 0.7538583300356153,
+      "grad_norm": 141.27809143066406,
+      "learning_rate": 2.1859211966916853e-07,
+      "loss": 4.0217,
+      "step": 9525
+    },
+    {
+      "epoch": 0.7558369608231104,
+      "grad_norm": 133.8254852294922,
+      "learning_rate": 2.1850967563218087e-07,
+      "loss": 4.0683,
+      "step": 9550
+    },
+    {
+      "epoch": 0.7578155916106054,
+      "grad_norm": 139.919189453125,
+      "learning_rate": 2.1842723159519316e-07,
+      "loss": 3.9958,
+      "step": 9575
+    },
+    {
+      "epoch": 0.7597942223981005,
+      "grad_norm": 173.58946228027344,
+      "learning_rate": 2.183447875582055e-07,
+      "loss": 3.9474,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7617728531855956,
+      "grad_norm": 107.07398223876953,
+      "learning_rate": 2.1826234352121778e-07,
+      "loss": 3.8308,
+      "step": 9625
+    },
+    {
+      "epoch": 0.7637514839730907,
+      "grad_norm": 124.00753784179688,
+      "learning_rate": 2.181798994842301e-07,
+      "loss": 3.8218,
+      "step": 9650
+    },
+    {
+      "epoch": 0.7657301147605857,
+      "grad_norm": 138.23736572265625,
+      "learning_rate": 2.180974554472424e-07,
+      "loss": 3.7296,
+      "step": 9675
+    },
+    {
+      "epoch": 0.7677087455480808,
+      "grad_norm": 128.9496612548828,
+      "learning_rate": 2.1801501141025471e-07,
+      "loss": 3.9163,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7696873763355758,
+      "grad_norm": 108.07875061035156,
+      "learning_rate": 2.17932567373267e-07,
+      "loss": 3.9408,
+      "step": 9725
+    },
+    {
+      "epoch": 0.7716660071230709,
+      "grad_norm": 126.18501281738281,
+      "learning_rate": 2.1785012333627934e-07,
+      "loss": 4.1993,
+      "step": 9750
+    },
+    {
+      "epoch": 0.7736446379105659,
+      "grad_norm": 144.8102264404297,
+      "learning_rate": 2.1776767929929162e-07,
+      "loss": 3.877,
+      "step": 9775
+    },
+    {
+      "epoch": 0.775623268698061,
+      "grad_norm": 118.8504638671875,
+      "learning_rate": 2.1768523526230396e-07,
+      "loss": 3.9788,
+      "step": 9800
+    },
+    {
+      "epoch": 0.777601899485556,
+      "grad_norm": 127.45133209228516,
+      "learning_rate": 2.1760279122531625e-07,
+      "loss": 3.8987,
+      "step": 9825
+    },
+    {
+      "epoch": 0.7795805302730511,
+      "grad_norm": 134.95892333984375,
+      "learning_rate": 2.1752034718832853e-07,
+      "loss": 3.8251,
+      "step": 9850
+    },
+    {
+      "epoch": 0.7815591610605461,
+      "grad_norm": 124.00614929199219,
+      "learning_rate": 2.1743790315134087e-07,
+      "loss": 3.6875,
+      "step": 9875
+    },
+    {
+      "epoch": 0.7835377918480412,
+      "grad_norm": 126.81105041503906,
+      "learning_rate": 2.1735545911435315e-07,
+      "loss": 3.7447,
+      "step": 9900
+    },
+    {
+      "epoch": 0.7855164226355362,
+      "grad_norm": 106.54443359375,
+      "learning_rate": 2.172730150773655e-07,
+      "loss": 3.9051,
+      "step": 9925
+    },
+    {
+      "epoch": 0.7874950534230313,
+      "grad_norm": 156.5098876953125,
+      "learning_rate": 2.1719057104037778e-07,
+      "loss": 3.9587,
+      "step": 9950
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 128.83648681640625,
+      "learning_rate": 2.171081270033901e-07,
+      "loss": 4.0755,
+      "step": 9975
+    },
+    {
+      "epoch": 0.7914523149980214,
+      "grad_norm": 131.54664611816406,
+      "learning_rate": 2.170256829664024e-07,
+      "loss": 4.0072,
+      "step": 10000
+    },
+    {
+      "epoch": 0.7914523149980214,
+      "eval_loss": 3.8681728839874268,
+      "eval_runtime": 9.5698,
+      "eval_samples_per_second": 264.165,
+      "eval_steps_per_second": 33.021,
+      "step": 10000
+    },
+    {
+      "epoch": 0.7934309457855164,
+      "grad_norm": 130.07545471191406,
+      "learning_rate": 2.169432389294147e-07,
+      "loss": 3.7883,
+      "step": 10025
+    },
+    {
+      "epoch": 0.7954095765730115,
+      "grad_norm": 135.5163116455078,
+      "learning_rate": 2.16860794892427e-07,
+      "loss": 3.8028,
+      "step": 10050
+    },
+    {
+      "epoch": 0.7973882073605065,
+      "grad_norm": 130.33010864257812,
+      "learning_rate": 2.1677835085543934e-07,
+      "loss": 3.9204,
+      "step": 10075
+    },
+    {
+      "epoch": 0.7993668381480016,
+      "grad_norm": 131.47166442871094,
+      "learning_rate": 2.1669590681845162e-07,
+      "loss": 3.7961,
+      "step": 10100
+    },
+    {
+      "epoch": 0.8013454689354966,
+      "grad_norm": 126.78470611572266,
+      "learning_rate": 2.1661346278146396e-07,
+      "loss": 3.8249,
+      "step": 10125
+    },
+    {
+      "epoch": 0.8033240997229917,
+      "grad_norm": 188.20448303222656,
+      "learning_rate": 2.1653101874447624e-07,
+      "loss": 3.778,
+      "step": 10150
+    },
+    {
+      "epoch": 0.8053027305104867,
+      "grad_norm": 137.7311553955078,
+      "learning_rate": 2.1644857470748853e-07,
+      "loss": 3.9346,
+      "step": 10175
+    },
+    {
+      "epoch": 0.8072813612979818,
+      "grad_norm": 89.07192993164062,
+      "learning_rate": 2.1636613067050087e-07,
+      "loss": 3.7072,
+      "step": 10200
+    },
+    {
+      "epoch": 0.8092599920854768,
+      "grad_norm": 96.4891357421875,
+      "learning_rate": 2.1628368663351315e-07,
+      "loss": 3.8609,
+      "step": 10225
+    },
+    {
+      "epoch": 0.8112386228729719,
+      "grad_norm": 144.67372131347656,
+      "learning_rate": 2.1620124259652546e-07,
+      "loss": 3.6157,
+      "step": 10250
+    },
+    {
+      "epoch": 0.8132172536604669,
+      "grad_norm": 115.51387023925781,
+      "learning_rate": 2.1611879855953778e-07,
+      "loss": 3.7226,
+      "step": 10275
+    },
+    {
+      "epoch": 0.815195884447962,
+      "grad_norm": 99.25152587890625,
+      "learning_rate": 2.160363545225501e-07,
+      "loss": 3.9583,
+      "step": 10300
+    },
+    {
+      "epoch": 0.817174515235457,
+      "grad_norm": 129.25799560546875,
+      "learning_rate": 2.1595391048556237e-07,
+      "loss": 3.9272,
+      "step": 10325
+    },
+    {
+      "epoch": 0.8191531460229521,
+      "grad_norm": 142.94081115722656,
+      "learning_rate": 2.158714664485747e-07,
+      "loss": 4.0403,
+      "step": 10350
+    },
+    {
+      "epoch": 0.8211317768104471,
+      "grad_norm": 138.73974609375,
+      "learning_rate": 2.15789022411587e-07,
+      "loss": 3.9837,
+      "step": 10375
+    },
+    {
+      "epoch": 0.8231104075979422,
+      "grad_norm": 128.65940856933594,
+      "learning_rate": 2.1570657837459933e-07,
+      "loss": 3.9414,
+      "step": 10400
+    },
+    {
+      "epoch": 0.8250890383854372,
+      "grad_norm": 98.70417022705078,
+      "learning_rate": 2.1562413433761162e-07,
+      "loss": 3.9875,
+      "step": 10425
+    },
+    {
+      "epoch": 0.8270676691729323,
+      "grad_norm": 95.58749389648438,
+      "learning_rate": 2.155416903006239e-07,
+      "loss": 4.0485,
+      "step": 10450
+    },
+    {
+      "epoch": 0.8290462999604273,
+      "grad_norm": 115.98033905029297,
+      "learning_rate": 2.1545924626363624e-07,
+      "loss": 3.9302,
+      "step": 10475
+    },
+    {
+      "epoch": 0.8310249307479224,
+      "grad_norm": 115.73841094970703,
+      "learning_rate": 2.1537680222664853e-07,
+      "loss": 3.8931,
+      "step": 10500
+    },
+    {
+      "epoch": 0.8330035615354174,
+      "grad_norm": 120.9448013305664,
+      "learning_rate": 2.1529435818966086e-07,
+      "loss": 3.9164,
+      "step": 10525
+    },
+    {
+      "epoch": 0.8349821923229126,
+      "grad_norm": 108.95042419433594,
+      "learning_rate": 2.1521191415267315e-07,
+      "loss": 3.8218,
+      "step": 10550
+    },
+    {
+      "epoch": 0.8369608231104076,
+      "grad_norm": 134.7786407470703,
+      "learning_rate": 2.1512947011568546e-07,
+      "loss": 3.8017,
+      "step": 10575
+    },
+    {
+      "epoch": 0.8389394538979027,
+      "grad_norm": 135.17352294921875,
+      "learning_rate": 2.1504702607869777e-07,
+      "loss": 3.7896,
+      "step": 10600
+    },
+    {
+      "epoch": 0.8409180846853977,
+      "grad_norm": 107.60875701904297,
+      "learning_rate": 2.1496458204171008e-07,
+      "loss": 3.8653,
+      "step": 10625
+    },
+    {
+      "epoch": 0.8428967154728928,
+      "grad_norm": 106.85396575927734,
+      "learning_rate": 2.1488213800472237e-07,
+      "loss": 3.7688,
+      "step": 10650
+    },
+    {
+      "epoch": 0.8448753462603878,
+      "grad_norm": 154.32797241210938,
+      "learning_rate": 2.147996939677347e-07,
+      "loss": 3.7928,
+      "step": 10675
+    },
+    {
+      "epoch": 0.8468539770478829,
+      "grad_norm": 105.45416259765625,
+      "learning_rate": 2.14717249930747e-07,
+      "loss": 3.9773,
+      "step": 10700
+    },
+    {
+      "epoch": 0.848832607835378,
+      "grad_norm": 130.95082092285156,
+      "learning_rate": 2.1463480589375933e-07,
+      "loss": 3.9007,
+      "step": 10725
+    },
+    {
+      "epoch": 0.850811238622873,
+      "grad_norm": 111.79964447021484,
+      "learning_rate": 2.1455236185677162e-07,
+      "loss": 3.8503,
+      "step": 10750
+    },
+    {
+      "epoch": 0.852789869410368,
+      "grad_norm": 111.66275787353516,
+      "learning_rate": 2.144699178197839e-07,
+      "loss": 4.1484,
+      "step": 10775
+    },
+    {
+      "epoch": 0.8547685001978631,
+      "grad_norm": 105.59733581542969,
+      "learning_rate": 2.1438747378279624e-07,
+      "loss": 3.8871,
+      "step": 10800
+    },
+    {
+      "epoch": 0.8567471309853582,
+      "grad_norm": 100.40887451171875,
+      "learning_rate": 2.1430502974580852e-07,
+      "loss": 3.8456,
+      "step": 10825
+    },
+    {
+      "epoch": 0.8587257617728532,
+      "grad_norm": 129.26177978515625,
+      "learning_rate": 2.1422258570882084e-07,
+      "loss": 3.673,
+      "step": 10850
+    },
+    {
+      "epoch": 0.8607043925603483,
+      "grad_norm": 144.7709197998047,
+      "learning_rate": 2.1414014167183315e-07,
+      "loss": 3.7465,
+      "step": 10875
+    },
+    {
+      "epoch": 0.8626830233478433,
+      "grad_norm": 137.7886505126953,
+      "learning_rate": 2.1405769763484546e-07,
+      "loss": 3.8944,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8646616541353384,
+      "grad_norm": 117.49480438232422,
+      "learning_rate": 2.1397525359785774e-07,
+      "loss": 3.7995,
+      "step": 10925
+    },
+    {
+      "epoch": 0.8666402849228334,
+      "grad_norm": 168.4739990234375,
+      "learning_rate": 2.1389280956087008e-07,
+      "loss": 4.0448,
+      "step": 10950
+    },
+    {
+      "epoch": 0.8686189157103285,
+      "grad_norm": 133.91720581054688,
+      "learning_rate": 2.1381036552388237e-07,
+      "loss": 3.6778,
+      "step": 10975
+    },
+    {
+      "epoch": 0.8705975464978235,
+      "grad_norm": 100.49728393554688,
+      "learning_rate": 2.137279214868947e-07,
+      "loss": 3.9251,
+      "step": 11000
+    },
+    {
+      "epoch": 0.8725761772853186,
+      "grad_norm": 189.62327575683594,
+      "learning_rate": 2.13645477449907e-07,
+      "loss": 3.8032,
+      "step": 11025
+    },
+    {
+      "epoch": 0.8745548080728136,
+      "grad_norm": 108.47595977783203,
+      "learning_rate": 2.135630334129193e-07,
+      "loss": 3.92,
+      "step": 11050
+    },
+    {
+      "epoch": 0.8765334388603087,
+      "grad_norm": 150.95584106445312,
+      "learning_rate": 2.1348058937593161e-07,
+      "loss": 3.9559,
+      "step": 11075
+    },
+    {
+      "epoch": 0.8785120696478037,
+      "grad_norm": 126.86693572998047,
+      "learning_rate": 2.133981453389439e-07,
+      "loss": 3.9322,
+      "step": 11100
+    },
+    {
+      "epoch": 0.8804907004352988,
+      "grad_norm": 158.07049560546875,
+      "learning_rate": 2.1331570130195624e-07,
+      "loss": 3.7498,
+      "step": 11125
+    },
+    {
+      "epoch": 0.8824693312227938,
+      "grad_norm": 156.98699951171875,
+      "learning_rate": 2.1323325726496852e-07,
+      "loss": 3.9679,
+      "step": 11150
+    },
+    {
+      "epoch": 0.8844479620102889,
+      "grad_norm": 91.1626968383789,
+      "learning_rate": 2.1315081322798083e-07,
+      "loss": 3.9669,
+      "step": 11175
+    },
+    {
+      "epoch": 0.8864265927977839,
+      "grad_norm": 145.4545135498047,
+      "learning_rate": 2.1306836919099315e-07,
+      "loss": 3.9038,
+      "step": 11200
+    },
+    {
+      "epoch": 0.888405223585279,
+      "grad_norm": 134.78793334960938,
+      "learning_rate": 2.1298592515400546e-07,
+      "loss": 3.9122,
+      "step": 11225
+    },
+    {
+      "epoch": 0.890383854372774,
+      "grad_norm": 132.1155242919922,
+      "learning_rate": 2.1290348111701774e-07,
+      "loss": 4.0016,
+      "step": 11250
+    },
+    {
+      "epoch": 0.8923624851602691,
+      "grad_norm": 116.72430419921875,
+      "learning_rate": 2.1282103708003008e-07,
+      "loss": 3.9878,
+      "step": 11275
+    },
+    {
+      "epoch": 0.8943411159477641,
+      "grad_norm": 117.89616394042969,
+      "learning_rate": 2.1273859304304237e-07,
+      "loss": 3.7834,
+      "step": 11300
+    },
+    {
+      "epoch": 0.8963197467352592,
+      "grad_norm": 151.65277099609375,
+      "learning_rate": 2.126561490060547e-07,
+      "loss": 3.9696,
+      "step": 11325
+    },
+    {
+      "epoch": 0.8982983775227542,
+      "grad_norm": 139.39405822753906,
+      "learning_rate": 2.12573704969067e-07,
+      "loss": 3.9585,
+      "step": 11350
+    },
+    {
+      "epoch": 0.9002770083102493,
+      "grad_norm": 123.25849914550781,
+      "learning_rate": 2.124912609320793e-07,
+      "loss": 3.816,
+      "step": 11375
+    },
+    {
+      "epoch": 0.9022556390977443,
+      "grad_norm": 148.54562377929688,
+      "learning_rate": 2.124088168950916e-07,
+      "loss": 3.8433,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9042342698852394,
+      "grad_norm": 135.60752868652344,
+      "learning_rate": 2.123263728581039e-07,
+      "loss": 3.7851,
+      "step": 11425
+    },
+    {
+      "epoch": 0.9062129006727345,
+      "grad_norm": 129.5711212158203,
+      "learning_rate": 2.122439288211162e-07,
+      "loss": 3.7133,
+      "step": 11450
+    },
+    {
+      "epoch": 0.9081915314602296,
+      "grad_norm": 136.88392639160156,
+      "learning_rate": 2.1216148478412852e-07,
+      "loss": 3.9864,
+      "step": 11475
+    },
+    {
+      "epoch": 0.9101701622477246,
+      "grad_norm": 148.6637725830078,
+      "learning_rate": 2.1207904074714083e-07,
+      "loss": 3.6906,
+      "step": 11500
+    },
+    {
+      "epoch": 0.9121487930352197,
+      "grad_norm": 164.6747283935547,
+      "learning_rate": 2.1199659671015314e-07,
+      "loss": 3.8543,
+      "step": 11525
+    },
+    {
+      "epoch": 0.9141274238227147,
+      "grad_norm": 114.64603424072266,
+      "learning_rate": 2.1191415267316546e-07,
+      "loss": 3.6973,
+      "step": 11550
+    },
+    {
+      "epoch": 0.9161060546102098,
+      "grad_norm": 129.53265380859375,
+      "learning_rate": 2.1183170863617774e-07,
+      "loss": 3.5896,
+      "step": 11575
+    },
+    {
+      "epoch": 0.9180846853977048,
+      "grad_norm": 145.63973999023438,
+      "learning_rate": 2.1174926459919008e-07,
+      "loss": 3.963,
+      "step": 11600
+    },
+    {
+      "epoch": 0.9200633161851999,
+      "grad_norm": 132.06729125976562,
+      "learning_rate": 2.1166682056220236e-07,
+      "loss": 3.6703,
+      "step": 11625
+    },
+    {
+      "epoch": 0.9220419469726949,
+      "grad_norm": 129.76583862304688,
+      "learning_rate": 2.1158437652521468e-07,
+      "loss": 3.6965,
+      "step": 11650
+    },
+    {
+      "epoch": 0.92402057776019,
+      "grad_norm": 115.21681213378906,
+      "learning_rate": 2.1150193248822699e-07,
+      "loss": 3.8711,
+      "step": 11675
+    },
+    {
+      "epoch": 0.925999208547685,
+      "grad_norm": 85.92119598388672,
+      "learning_rate": 2.114194884512393e-07,
+      "loss": 3.8377,
+      "step": 11700
+    },
+    {
+      "epoch": 0.9279778393351801,
+      "grad_norm": 117.11397552490234,
+      "learning_rate": 2.113370444142516e-07,
+      "loss": 3.7965,
+      "step": 11725
+    },
+    {
+      "epoch": 0.9299564701226751,
+      "grad_norm": 113.50032806396484,
+      "learning_rate": 2.112546003772639e-07,
+      "loss": 3.7684,
+      "step": 11750
+    },
+    {
+      "epoch": 0.9319351009101702,
+      "grad_norm": 107.85367584228516,
+      "learning_rate": 2.111721563402762e-07,
+      "loss": 3.6988,
+      "step": 11775
+    },
+    {
+      "epoch": 0.9339137316976652,
+      "grad_norm": 121.42901611328125,
+      "learning_rate": 2.1108971230328852e-07,
+      "loss": 3.6745,
+      "step": 11800
+    },
+    {
+      "epoch": 0.9358923624851603,
+      "grad_norm": 139.971923828125,
+      "learning_rate": 2.1100726826630083e-07,
+      "loss": 3.8403,
+      "step": 11825
+    },
+    {
+      "epoch": 0.9378709932726553,
+      "grad_norm": 126.40741729736328,
+      "learning_rate": 2.1092482422931312e-07,
+      "loss": 3.7846,
+      "step": 11850
+    },
+    {
+      "epoch": 0.9398496240601504,
+      "grad_norm": 110.96858215332031,
+      "learning_rate": 2.1084238019232545e-07,
+      "loss": 3.9256,
+      "step": 11875
+    },
+    {
+      "epoch": 0.9418282548476454,
+      "grad_norm": 107.72772979736328,
+      "learning_rate": 2.1075993615533774e-07,
+      "loss": 3.7888,
+      "step": 11900
+    },
+    {
+      "epoch": 0.9438068856351405,
+      "grad_norm": 94.70952606201172,
+      "learning_rate": 2.1067749211835008e-07,
+      "loss": 3.8496,
+      "step": 11925
+    },
+    {
+      "epoch": 0.9457855164226355,
+      "grad_norm": 178.84500122070312,
+      "learning_rate": 2.1059504808136236e-07,
+      "loss": 3.737,
+      "step": 11950
+    },
+    {
+      "epoch": 0.9477641472101306,
+      "grad_norm": 179.0364227294922,
+      "learning_rate": 2.1051260404437467e-07,
+      "loss": 3.7584,
+      "step": 11975
+    },
+    {
+      "epoch": 0.9497427779976256,
+      "grad_norm": 141.7266845703125,
+      "learning_rate": 2.1043016000738698e-07,
+      "loss": 3.7948,
+      "step": 12000
+    },
+    {
+      "epoch": 0.9517214087851207,
+      "grad_norm": 120.56340026855469,
+      "learning_rate": 2.1034771597039927e-07,
+      "loss": 3.8951,
+      "step": 12025
+    },
+    {
+      "epoch": 0.9537000395726157,
+      "grad_norm": 113.90511322021484,
+      "learning_rate": 2.1026527193341158e-07,
+      "loss": 3.8823,
+      "step": 12050
+    },
+    {
+      "epoch": 0.9556786703601108,
+      "grad_norm": 103.57383728027344,
+      "learning_rate": 2.101828278964239e-07,
+      "loss": 3.7853,
+      "step": 12075
+    },
+    {
+      "epoch": 0.9576573011476058,
+      "grad_norm": 111.0738525390625,
+      "learning_rate": 2.101003838594362e-07,
+      "loss": 3.8394,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9596359319351009,
+      "grad_norm": 130.9629364013672,
+      "learning_rate": 2.1001793982244852e-07,
+      "loss": 3.7752,
+      "step": 12125
+    },
+    {
+      "epoch": 0.9616145627225959,
+      "grad_norm": 98.29790496826172,
+      "learning_rate": 2.0993549578546083e-07,
+      "loss": 3.8332,
+      "step": 12150
+    },
+    {
+      "epoch": 0.963593193510091,
+      "grad_norm": 116.66043853759766,
+      "learning_rate": 2.098530517484731e-07,
+      "loss": 3.964,
+      "step": 12175
+    },
+    {
+      "epoch": 0.965571824297586,
+      "grad_norm": 179.0226287841797,
+      "learning_rate": 2.0977060771148545e-07,
+      "loss": 3.8848,
+      "step": 12200
+    },
+    {
+      "epoch": 0.9675504550850811,
+      "grad_norm": 100.40991973876953,
+      "learning_rate": 2.0968816367449774e-07,
+      "loss": 3.8461,
+      "step": 12225
+    },
+    {
+      "epoch": 0.9695290858725761,
+      "grad_norm": 117.44912719726562,
+      "learning_rate": 2.0960571963751005e-07,
+      "loss": 3.723,
+      "step": 12250
+    },
+    {
+      "epoch": 0.9715077166600712,
+      "grad_norm": 106.5282974243164,
+      "learning_rate": 2.0952327560052236e-07,
+      "loss": 3.8557,
+      "step": 12275
+    },
+    {
+      "epoch": 0.9734863474475662,
+      "grad_norm": 111.7376480102539,
+      "learning_rate": 2.0944083156353467e-07,
+      "loss": 3.8505,
+      "step": 12300
+    },
+    {
+      "epoch": 0.9754649782350613,
+      "grad_norm": 185.3255157470703,
+      "learning_rate": 2.0935838752654698e-07,
+      "loss": 3.9339,
+      "step": 12325
+    },
+    {
+      "epoch": 0.9774436090225563,
+      "grad_norm": 128.0303955078125,
+      "learning_rate": 2.0927594348955927e-07,
+      "loss": 3.9595,
+      "step": 12350
+    },
+    {
+      "epoch": 0.9794222398100515,
+      "grad_norm": 117.01809692382812,
+      "learning_rate": 2.0919349945257158e-07,
+      "loss": 3.8385,
+      "step": 12375
+    },
+    {
+      "epoch": 0.9814008705975465,
+      "grad_norm": 132.46775817871094,
+      "learning_rate": 2.091110554155839e-07,
+      "loss": 3.6976,
+      "step": 12400
+    },
+    {
+      "epoch": 0.9833795013850416,
+      "grad_norm": 125.96062469482422,
+      "learning_rate": 2.090286113785962e-07,
+      "loss": 3.9652,
+      "step": 12425
+    },
+    {
+      "epoch": 0.9853581321725366,
+      "grad_norm": 172.13600158691406,
+      "learning_rate": 2.089461673416085e-07,
+      "loss": 3.9026,
+      "step": 12450
+    },
+    {
+      "epoch": 0.9873367629600317,
+      "grad_norm": 107.95278930664062,
+      "learning_rate": 2.0886372330462083e-07,
+      "loss": 3.8457,
+      "step": 12475
+    },
+    {
+      "epoch": 0.9893153937475268,
+      "grad_norm": 196.08303833007812,
+      "learning_rate": 2.087812792676331e-07,
+      "loss": 3.8319,
+      "step": 12500
+    },
+    {
+      "epoch": 0.9912940245350218,
+      "grad_norm": 119.07249450683594,
+      "learning_rate": 2.0869883523064545e-07,
+      "loss": 3.7464,
+      "step": 12525
+    },
+    {
+      "epoch": 0.9932726553225169,
+      "grad_norm": 118.91383361816406,
+      "learning_rate": 2.0861639119365773e-07,
+      "loss": 3.7734,
+      "step": 12550
+    },
+    {
+      "epoch": 0.9952512861100119,
+      "grad_norm": 122.37627410888672,
+      "learning_rate": 2.0853394715667005e-07,
+      "loss": 3.9263,
+      "step": 12575
+    },
+    {
+      "epoch": 0.997229916897507,
+      "grad_norm": 101.24188232421875,
+      "learning_rate": 2.0845150311968236e-07,
+      "loss": 3.6865,
+      "step": 12600
+    },
+    {
+      "epoch": 0.999208547685002,
+      "grad_norm": 253.74789428710938,
+      "learning_rate": 2.0836905908269467e-07,
+      "loss": 3.857,
+      "step": 12625
+    },
+    {
+      "epoch": 1.001187178472497,
+      "grad_norm": 121.46297454833984,
+      "learning_rate": 2.0828661504570695e-07,
+      "loss": 3.9145,
+      "step": 12650
+    },
+    {
+      "epoch": 1.003165809259992,
+      "grad_norm": 115.41104125976562,
+      "learning_rate": 2.0820417100871927e-07,
+      "loss": 3.9441,
+      "step": 12675
+    },
+    {
+      "epoch": 1.0051444400474872,
+      "grad_norm": 105.03096008300781,
+      "learning_rate": 2.0812172697173158e-07,
+      "loss": 3.6926,
+      "step": 12700
+    },
+    {
+      "epoch": 1.007123070834982,
+      "grad_norm": 98.86603546142578,
+      "learning_rate": 2.080392829347439e-07,
+      "loss": 3.5524,
+      "step": 12725
+    },
+    {
+      "epoch": 1.0091017016224773,
+      "grad_norm": 120.75003814697266,
+      "learning_rate": 2.079568388977562e-07,
+      "loss": 3.8494,
+      "step": 12750
+    },
+    {
+      "epoch": 1.0110803324099722,
+      "grad_norm": 86.03683471679688,
+      "learning_rate": 2.0787439486076849e-07,
+      "loss": 3.7635,
+      "step": 12775
+    },
+    {
+      "epoch": 1.0130589631974674,
+      "grad_norm": 132.28945922851562,
+      "learning_rate": 2.0779195082378082e-07,
+      "loss": 4.007,
+      "step": 12800
+    },
+    {
+      "epoch": 1.0150375939849625,
+      "grad_norm": 129.51051330566406,
+      "learning_rate": 2.077095067867931e-07,
+      "loss": 3.6065,
+      "step": 12825
+    },
+    {
+      "epoch": 1.0170162247724575,
+      "grad_norm": 103.38876342773438,
+      "learning_rate": 2.0762706274980542e-07,
+      "loss": 3.7508,
+      "step": 12850
+    },
+    {
+      "epoch": 1.0189948555599526,
+      "grad_norm": 163.7755584716797,
+      "learning_rate": 2.0754461871281773e-07,
+      "loss": 3.7519,
+      "step": 12875
+    },
+    {
+      "epoch": 1.0209734863474476,
+      "grad_norm": 123.19217681884766,
+      "learning_rate": 2.0746217467583004e-07,
+      "loss": 3.8441,
+      "step": 12900
+    },
+    {
+      "epoch": 1.0229521171349427,
+      "grad_norm": 122.48994445800781,
+      "learning_rate": 2.0737973063884236e-07,
+      "loss": 3.8664,
+      "step": 12925
+    },
+    {
+      "epoch": 1.0249307479224377,
+      "grad_norm": 97.31283569335938,
+      "learning_rate": 2.0729728660185467e-07,
+      "loss": 3.9103,
+      "step": 12950
+    },
+    {
+      "epoch": 1.0269093787099328,
+      "grad_norm": 178.35073852539062,
+      "learning_rate": 2.0721484256486695e-07,
+      "loss": 3.8718,
+      "step": 12975
+    },
+    {
+      "epoch": 1.0288880094974278,
+      "grad_norm": 124.15072631835938,
+      "learning_rate": 2.0713239852787926e-07,
+      "loss": 3.7962,
+      "step": 13000
+    },
+    {
+      "epoch": 1.030866640284923,
+      "grad_norm": 141.29360961914062,
+      "learning_rate": 2.0704995449089158e-07,
+      "loss": 3.8207,
+      "step": 13025
+    },
+    {
+      "epoch": 1.0328452710724179,
+      "grad_norm": 117.48384857177734,
+      "learning_rate": 2.0696751045390386e-07,
+      "loss": 4.0525,
+      "step": 13050
+    },
+    {
+      "epoch": 1.034823901859913,
+      "grad_norm": 124.6396713256836,
+      "learning_rate": 2.068850664169162e-07,
+      "loss": 3.8003,
+      "step": 13075
+    },
+    {
+      "epoch": 1.036802532647408,
+      "grad_norm": 164.6786346435547,
+      "learning_rate": 2.0680262237992848e-07,
+      "loss": 3.6338,
+      "step": 13100
+    },
+    {
+      "epoch": 1.0387811634349031,
+      "grad_norm": 144.25732421875,
+      "learning_rate": 2.0672017834294082e-07,
+      "loss": 3.9446,
+      "step": 13125
+    },
+    {
+      "epoch": 1.040759794222398,
+      "grad_norm": 130.24571228027344,
+      "learning_rate": 2.066377343059531e-07,
+      "loss": 3.9012,
+      "step": 13150
+    },
+    {
+      "epoch": 1.0427384250098932,
+      "grad_norm": 122.11434173583984,
+      "learning_rate": 2.0655529026896542e-07,
+      "loss": 3.7799,
+      "step": 13175
+    },
+    {
+      "epoch": 1.0447170557973882,
+      "grad_norm": 161.62745666503906,
+      "learning_rate": 2.0647284623197773e-07,
+      "loss": 3.8881,
+      "step": 13200
+    },
+    {
+      "epoch": 1.0466956865848833,
+      "grad_norm": 129.13072204589844,
+      "learning_rate": 2.0639040219499004e-07,
+      "loss": 3.8295,
+      "step": 13225
+    },
+    {
+      "epoch": 1.0486743173723783,
+      "grad_norm": 153.86805725097656,
+      "learning_rate": 2.0630795815800233e-07,
+      "loss": 3.7328,
+      "step": 13250
+    },
+    {
+      "epoch": 1.0506529481598734,
+      "grad_norm": 173.9022979736328,
+      "learning_rate": 2.0622551412101464e-07,
+      "loss": 3.9866,
+      "step": 13275
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 104.5372543334961,
+      "learning_rate": 2.0614307008402695e-07,
+      "loss": 3.6121,
+      "step": 13300
+    },
+    {
+      "epoch": 1.0546102097348635,
+      "grad_norm": 115.93688201904297,
+      "learning_rate": 2.0606062604703926e-07,
+      "loss": 3.8851,
+      "step": 13325
+    },
+    {
+      "epoch": 1.0565888405223585,
+      "grad_norm": 140.26748657226562,
+      "learning_rate": 2.0597818201005157e-07,
+      "loss": 3.9924,
+      "step": 13350
+    },
+    {
+      "epoch": 1.0585674713098536,
+      "grad_norm": 122.87477111816406,
+      "learning_rate": 2.0589573797306386e-07,
+      "loss": 3.7768,
+      "step": 13375
+    },
+    {
+      "epoch": 1.0605461020973486,
+      "grad_norm": 105.3402099609375,
+      "learning_rate": 2.058132939360762e-07,
+      "loss": 3.6371,
+      "step": 13400
+    },
+    {
+      "epoch": 1.0625247328848437,
+      "grad_norm": 112.1316146850586,
+      "learning_rate": 2.0573084989908848e-07,
+      "loss": 3.8185,
+      "step": 13425
+    },
+    {
+      "epoch": 1.0645033636723387,
+      "grad_norm": 105.67694854736328,
+      "learning_rate": 2.056484058621008e-07,
+      "loss": 3.7202,
+      "step": 13450
+    },
+    {
+      "epoch": 1.0664819944598338,
+      "grad_norm": 160.9032440185547,
+      "learning_rate": 2.055659618251131e-07,
+      "loss": 3.6819,
+      "step": 13475
+    },
+    {
+      "epoch": 1.0684606252473288,
+      "grad_norm": 197.6964569091797,
+      "learning_rate": 2.0548351778812542e-07,
+      "loss": 3.914,
+      "step": 13500
+    },
+    {
+      "epoch": 1.070439256034824,
+      "grad_norm": 100.68509674072266,
+      "learning_rate": 2.0540107375113773e-07,
+      "loss": 3.7281,
+      "step": 13525
+    },
+    {
+      "epoch": 1.0724178868223189,
+      "grad_norm": 121.31140899658203,
+      "learning_rate": 2.0531862971415004e-07,
+      "loss": 3.7413,
+      "step": 13550
+    },
+    {
+      "epoch": 1.074396517609814,
+      "grad_norm": 176.4757080078125,
+      "learning_rate": 2.0523618567716232e-07,
+      "loss": 4.1066,
+      "step": 13575
+    },
+    {
+      "epoch": 1.076375148397309,
+      "grad_norm": 98.44918823242188,
+      "learning_rate": 2.0515374164017464e-07,
+      "loss": 3.6721,
+      "step": 13600
+    },
+    {
+      "epoch": 1.0783537791848041,
+      "grad_norm": 114.9334487915039,
+      "learning_rate": 2.0507129760318695e-07,
+      "loss": 3.7888,
+      "step": 13625
+    },
+    {
+      "epoch": 1.080332409972299,
+      "grad_norm": 172.9710235595703,
+      "learning_rate": 2.0498885356619923e-07,
+      "loss": 3.8811,
+      "step": 13650
+    },
+    {
+      "epoch": 1.0823110407597942,
+      "grad_norm": 77.68281555175781,
+      "learning_rate": 2.0490640952921157e-07,
+      "loss": 3.7865,
+      "step": 13675
+    },
+    {
+      "epoch": 1.0842896715472894,
+      "grad_norm": 117.03499603271484,
+      "learning_rate": 2.0482396549222386e-07,
+      "loss": 3.6545,
+      "step": 13700
+    },
+    {
+      "epoch": 1.0862683023347843,
+      "grad_norm": 111.88362121582031,
+      "learning_rate": 2.047415214552362e-07,
+      "loss": 3.7833,
+      "step": 13725
+    },
+    {
+      "epoch": 1.0882469331222793,
+      "grad_norm": 177.05654907226562,
+      "learning_rate": 2.0465907741824848e-07,
+      "loss": 3.8182,
+      "step": 13750
+    },
+    {
+      "epoch": 1.0902255639097744,
+      "grad_norm": 110.29389953613281,
+      "learning_rate": 2.045766333812608e-07,
+      "loss": 3.571,
+      "step": 13775
+    },
+    {
+      "epoch": 1.0922041946972696,
+      "grad_norm": 81.29917907714844,
+      "learning_rate": 2.044941893442731e-07,
+      "loss": 3.7003,
+      "step": 13800
+    },
+    {
+      "epoch": 1.0941828254847645,
+      "grad_norm": 130.32589721679688,
+      "learning_rate": 2.0441174530728541e-07,
+      "loss": 3.965,
+      "step": 13825
+    },
+    {
+      "epoch": 1.0961614562722597,
+      "grad_norm": 134.07106018066406,
+      "learning_rate": 2.043293012702977e-07,
+      "loss": 4.0457,
+      "step": 13850
+    },
+    {
+      "epoch": 1.0981400870597546,
+      "grad_norm": 204.80294799804688,
+      "learning_rate": 2.0424685723331004e-07,
+      "loss": 3.5721,
+      "step": 13875
+    },
+    {
+      "epoch": 1.1001187178472498,
+      "grad_norm": 157.00570678710938,
+      "learning_rate": 2.0416441319632232e-07,
+      "loss": 3.8119,
+      "step": 13900
+    },
+    {
+      "epoch": 1.1020973486347447,
+      "grad_norm": 118.97384643554688,
+      "learning_rate": 2.0408196915933463e-07,
+      "loss": 3.6172,
+      "step": 13925
+    },
+    {
+      "epoch": 1.10407597942224,
+      "grad_norm": 147.43740844726562,
+      "learning_rate": 2.0399952512234695e-07,
+      "loss": 3.8369,
+      "step": 13950
+    },
+    {
+      "epoch": 1.1060546102097348,
+      "grad_norm": 116.36746215820312,
+      "learning_rate": 2.0391708108535923e-07,
+      "loss": 3.8669,
+      "step": 13975
+    },
+    {
+      "epoch": 1.10803324099723,
+      "grad_norm": 119.54280090332031,
+      "learning_rate": 2.0383463704837157e-07,
+      "loss": 3.8281,
+      "step": 14000
+    },
+    {
+      "epoch": 1.110011871784725,
+      "grad_norm": 161.6844024658203,
+      "learning_rate": 2.0375219301138385e-07,
+      "loss": 3.632,
+      "step": 14025
+    },
+    {
+      "epoch": 1.11199050257222,
+      "grad_norm": 115.4676284790039,
+      "learning_rate": 2.0366974897439617e-07,
+      "loss": 3.6665,
+      "step": 14050
+    },
+    {
+      "epoch": 1.113969133359715,
+      "grad_norm": 149.53504943847656,
+      "learning_rate": 2.0358730493740848e-07,
+      "loss": 3.8808,
+      "step": 14075
+    },
+    {
+      "epoch": 1.1159477641472102,
+      "grad_norm": 108.68144989013672,
+      "learning_rate": 2.035048609004208e-07,
+      "loss": 3.763,
+      "step": 14100
+    },
+    {
+      "epoch": 1.1179263949347051,
+      "grad_norm": 140.69000244140625,
+      "learning_rate": 2.034224168634331e-07,
+      "loss": 3.8483,
+      "step": 14125
+    },
+    {
+      "epoch": 1.1199050257222003,
+      "grad_norm": 116.24478149414062,
+      "learning_rate": 2.033399728264454e-07,
+      "loss": 3.6232,
+      "step": 14150
+    },
+    {
+      "epoch": 1.1218836565096952,
+      "grad_norm": 151.31192016601562,
+      "learning_rate": 2.032575287894577e-07,
+      "loss": 3.9463,
+      "step": 14175
+    },
+    {
+      "epoch": 1.1238622872971904,
+      "grad_norm": 104.94828796386719,
+      "learning_rate": 2.0317508475247004e-07,
+      "loss": 3.8141,
+      "step": 14200
+    },
+    {
+      "epoch": 1.1258409180846853,
+      "grad_norm": 174.0530548095703,
+      "learning_rate": 2.0309264071548232e-07,
+      "loss": 3.7278,
+      "step": 14225
+    },
+    {
+      "epoch": 1.1278195488721805,
+      "grad_norm": 123.31523895263672,
+      "learning_rate": 2.030101966784946e-07,
+      "loss": 3.6594,
+      "step": 14250
+    },
+    {
+      "epoch": 1.1297981796596754,
+      "grad_norm": 199.5244140625,
+      "learning_rate": 2.0292775264150694e-07,
+      "loss": 3.7397,
+      "step": 14275
+    },
+    {
+      "epoch": 1.1317768104471706,
+      "grad_norm": 121.85034942626953,
+      "learning_rate": 2.0284530860451923e-07,
+      "loss": 3.7268,
+      "step": 14300
+    },
+    {
+      "epoch": 1.1337554412346655,
+      "grad_norm": 100.23025512695312,
+      "learning_rate": 2.0276286456753157e-07,
+      "loss": 3.6129,
+      "step": 14325
+    },
+    {
+      "epoch": 1.1357340720221607,
+      "grad_norm": 194.6273651123047,
+      "learning_rate": 2.0268042053054385e-07,
+      "loss": 3.8529,
+      "step": 14350
+    },
+    {
+      "epoch": 1.1377127028096556,
+      "grad_norm": 86.26454162597656,
+      "learning_rate": 2.0259797649355616e-07,
+      "loss": 3.7558,
+      "step": 14375
+    },
+    {
+      "epoch": 1.1396913335971508,
+      "grad_norm": 94.16887664794922,
+      "learning_rate": 2.0251553245656847e-07,
+      "loss": 3.743,
+      "step": 14400
+    },
+    {
+      "epoch": 1.1416699643846457,
+      "grad_norm": 112.39076232910156,
+      "learning_rate": 2.0243308841958079e-07,
+      "loss": 3.6799,
+      "step": 14425
+    },
+    {
+      "epoch": 1.143648595172141,
+      "grad_norm": 146.2572479248047,
+      "learning_rate": 2.0235064438259307e-07,
+      "loss": 3.8062,
+      "step": 14450
+    },
+    {
+      "epoch": 1.1456272259596358,
+      "grad_norm": 100.57402801513672,
+      "learning_rate": 2.022682003456054e-07,
+      "loss": 3.7005,
+      "step": 14475
+    },
+    {
+      "epoch": 1.147605856747131,
+      "grad_norm": 110.64191436767578,
+      "learning_rate": 2.021857563086177e-07,
+      "loss": 4.007,
+      "step": 14500
+    },
+    {
+      "epoch": 1.149584487534626,
+      "grad_norm": 182.17724609375,
+      "learning_rate": 2.0210331227163003e-07,
+      "loss": 3.782,
+      "step": 14525
+    },
+    {
+      "epoch": 1.151563118322121,
+      "grad_norm": 135.72979736328125,
+      "learning_rate": 2.0202086823464232e-07,
+      "loss": 3.6697,
+      "step": 14550
+    },
+    {
+      "epoch": 1.1535417491096163,
+      "grad_norm": 101.66889953613281,
+      "learning_rate": 2.019384241976546e-07,
+      "loss": 3.7594,
+      "step": 14575
+    },
+    {
+      "epoch": 1.1555203798971112,
+      "grad_norm": 160.48109436035156,
+      "learning_rate": 2.0185598016066694e-07,
+      "loss": 3.7643,
+      "step": 14600
+    },
+    {
+      "epoch": 1.1574990106846061,
+      "grad_norm": 104.34113311767578,
+      "learning_rate": 2.0177353612367923e-07,
+      "loss": 3.9202,
+      "step": 14625
+    },
+    {
+      "epoch": 1.1594776414721013,
+      "grad_norm": 89.21700286865234,
+      "learning_rate": 2.0169109208669154e-07,
+      "loss": 3.5803,
+      "step": 14650
+    },
+    {
+      "epoch": 1.1614562722595965,
+      "grad_norm": 111.87966918945312,
+      "learning_rate": 2.0160864804970385e-07,
+      "loss": 3.5052,
+      "step": 14675
+    },
+    {
+      "epoch": 1.1634349030470914,
+      "grad_norm": 134.26589965820312,
+      "learning_rate": 2.0152620401271616e-07,
+      "loss": 3.7542,
+      "step": 14700
+    },
+    {
+      "epoch": 1.1654135338345863,
+      "grad_norm": 93.41925811767578,
+      "learning_rate": 2.0144375997572847e-07,
+      "loss": 3.8865,
+      "step": 14725
+    },
+    {
+      "epoch": 1.1673921646220815,
+      "grad_norm": 121.87728881835938,
+      "learning_rate": 2.0136131593874078e-07,
+      "loss": 3.9852,
+      "step": 14750
+    },
+    {
+      "epoch": 1.1693707954095767,
+      "grad_norm": 129.56564331054688,
+      "learning_rate": 2.0127887190175307e-07,
+      "loss": 3.6025,
+      "step": 14775
+    },
+    {
+      "epoch": 1.1713494261970716,
+      "grad_norm": 138.36422729492188,
+      "learning_rate": 2.011964278647654e-07,
+      "loss": 3.9298,
+      "step": 14800
+    },
+    {
+      "epoch": 1.1733280569845668,
+      "grad_norm": 115.63604736328125,
+      "learning_rate": 2.011139838277777e-07,
+      "loss": 3.811,
+      "step": 14825
+    },
+    {
+      "epoch": 1.1753066877720617,
+      "grad_norm": 128.9251708984375,
+      "learning_rate": 2.0103153979078998e-07,
+      "loss": 4.013,
+      "step": 14850
+    },
+    {
+      "epoch": 1.1772853185595569,
+      "grad_norm": 115.66397857666016,
+      "learning_rate": 2.0094909575380232e-07,
+      "loss": 3.9727,
+      "step": 14875
+    },
+    {
+      "epoch": 1.1792639493470518,
+      "grad_norm": 105.60704040527344,
+      "learning_rate": 2.008666517168146e-07,
+      "loss": 4.0072,
+      "step": 14900
+    },
+    {
+      "epoch": 1.181242580134547,
+      "grad_norm": 127.380859375,
+      "learning_rate": 2.0078420767982694e-07,
+      "loss": 3.763,
+      "step": 14925
+    },
+    {
+      "epoch": 1.183221210922042,
+      "grad_norm": 87.45874786376953,
+      "learning_rate": 2.0070176364283922e-07,
+      "loss": 3.6529,
+      "step": 14950
+    },
+    {
+      "epoch": 1.185199841709537,
+      "grad_norm": 257.6990966796875,
+      "learning_rate": 2.0061931960585154e-07,
+      "loss": 3.8049,
+      "step": 14975
+    },
+    {
+      "epoch": 1.187178472497032,
+      "grad_norm": 111.80567169189453,
+      "learning_rate": 2.0053687556886385e-07,
+      "loss": 3.9666,
+      "step": 15000
+    },
+    {
+      "epoch": 1.1891571032845272,
+      "grad_norm": 98.00049591064453,
+      "learning_rate": 2.0045443153187616e-07,
+      "loss": 3.773,
+      "step": 15025
+    },
+    {
+      "epoch": 1.1911357340720221,
+      "grad_norm": 117.19742584228516,
+      "learning_rate": 2.0037198749488844e-07,
+      "loss": 3.8228,
+      "step": 15050
+    },
+    {
+      "epoch": 1.1931143648595173,
+      "grad_norm": 99.34817504882812,
+      "learning_rate": 2.0028954345790078e-07,
+      "loss": 3.8045,
+      "step": 15075
+    },
+    {
+      "epoch": 1.1950929956470122,
+      "grad_norm": 193.61734008789062,
+      "learning_rate": 2.0020709942091307e-07,
+      "loss": 3.7231,
+      "step": 15100
+    },
+    {
+      "epoch": 1.1970716264345074,
+      "grad_norm": 110.1533203125,
+      "learning_rate": 2.001246553839254e-07,
+      "loss": 3.7013,
+      "step": 15125
+    },
+    {
+      "epoch": 1.1990502572220023,
+      "grad_norm": 81.41429901123047,
+      "learning_rate": 2.000422113469377e-07,
+      "loss": 3.5668,
+      "step": 15150
+    },
+    {
+      "epoch": 1.2010288880094975,
+      "grad_norm": 103.47528839111328,
+      "learning_rate": 1.9995976730994998e-07,
+      "loss": 3.6627,
+      "step": 15175
+    },
+    {
+      "epoch": 1.2030075187969924,
+      "grad_norm": 179.40489196777344,
+      "learning_rate": 1.9987732327296231e-07,
+      "loss": 3.8307,
+      "step": 15200
+    },
+    {
+      "epoch": 1.2049861495844876,
+      "grad_norm": 108.21305084228516,
+      "learning_rate": 1.997948792359746e-07,
+      "loss": 3.7619,
+      "step": 15225
+    },
+    {
+      "epoch": 1.2069647803719825,
+      "grad_norm": 206.6326904296875,
+      "learning_rate": 1.997124351989869e-07,
+      "loss": 3.7036,
+      "step": 15250
+    },
+    {
+      "epoch": 1.2089434111594777,
+      "grad_norm": 115.525390625,
+      "learning_rate": 1.9962999116199922e-07,
+      "loss": 3.823,
+      "step": 15275
+    },
+    {
+      "epoch": 1.2109220419469726,
+      "grad_norm": 124.59696197509766,
+      "learning_rate": 1.9954754712501153e-07,
+      "loss": 3.7983,
+      "step": 15300
+    },
+    {
+      "epoch": 1.2129006727344678,
+      "grad_norm": 107.92992401123047,
+      "learning_rate": 1.9946510308802385e-07,
+      "loss": 3.7677,
+      "step": 15325
+    },
+    {
+      "epoch": 1.2148793035219627,
+      "grad_norm": 164.7980194091797,
+      "learning_rate": 1.9938265905103616e-07,
+      "loss": 3.7692,
+      "step": 15350
+    },
+    {
+      "epoch": 1.2168579343094579,
+      "grad_norm": 95.48210906982422,
+      "learning_rate": 1.9930021501404844e-07,
+      "loss": 3.8568,
+      "step": 15375
+    },
+    {
+      "epoch": 1.2188365650969528,
+      "grad_norm": 172.18484497070312,
+      "learning_rate": 1.9921777097706078e-07,
+      "loss": 3.596,
+      "step": 15400
+    },
+    {
+      "epoch": 1.220815195884448,
+      "grad_norm": 146.89479064941406,
+      "learning_rate": 1.9913532694007307e-07,
+      "loss": 3.7992,
+      "step": 15425
+    },
+    {
+      "epoch": 1.222793826671943,
+      "grad_norm": 106.06961822509766,
+      "learning_rate": 1.9905288290308538e-07,
+      "loss": 3.5689,
+      "step": 15450
+    },
+    {
+      "epoch": 1.224772457459438,
+      "grad_norm": 124.81307220458984,
+      "learning_rate": 1.989704388660977e-07,
+      "loss": 3.8072,
+      "step": 15475
+    },
+    {
+      "epoch": 1.226751088246933,
+      "grad_norm": 144.33468627929688,
+      "learning_rate": 1.9888799482910997e-07,
+      "loss": 3.7889,
+      "step": 15500
+    },
+    {
+      "epoch": 1.2287297190344282,
+      "grad_norm": 108.416015625,
+      "learning_rate": 1.988055507921223e-07,
+      "loss": 3.7345,
+      "step": 15525
+    },
+    {
+      "epoch": 1.2307083498219233,
+      "grad_norm": 118.40252685546875,
+      "learning_rate": 1.987231067551346e-07,
+      "loss": 3.9306,
+      "step": 15550
+    },
+    {
+      "epoch": 1.2326869806094183,
+      "grad_norm": 121.7845230102539,
+      "learning_rate": 1.986406627181469e-07,
+      "loss": 3.6828,
+      "step": 15575
+    },
+    {
+      "epoch": 1.2346656113969132,
+      "grad_norm": 126.77731323242188,
+      "learning_rate": 1.9855821868115922e-07,
+      "loss": 3.8064,
+      "step": 15600
+    },
+    {
+      "epoch": 1.2366442421844084,
+      "grad_norm": 128.55198669433594,
+      "learning_rate": 1.9847577464417153e-07,
+      "loss": 3.952,
+      "step": 15625
+    },
+    {
+      "epoch": 1.2386228729719035,
+      "grad_norm": 99.97991180419922,
+      "learning_rate": 1.9839333060718382e-07,
+      "loss": 3.8026,
+      "step": 15650
+    },
+    {
+      "epoch": 1.2406015037593985,
+      "grad_norm": 109.74803924560547,
+      "learning_rate": 1.9831088657019615e-07,
+      "loss": 3.8499,
+      "step": 15675
+    },
+    {
+      "epoch": 1.2425801345468936,
+      "grad_norm": 125.961669921875,
+      "learning_rate": 1.9822844253320844e-07,
+      "loss": 3.8659,
+      "step": 15700
+    },
+    {
+      "epoch": 1.2445587653343886,
+      "grad_norm": 106.30962371826172,
+      "learning_rate": 1.9814599849622078e-07,
+      "loss": 3.7085,
+      "step": 15725
+    },
+    {
+      "epoch": 1.2465373961218837,
+      "grad_norm": 101.37333679199219,
+      "learning_rate": 1.9806355445923306e-07,
+      "loss": 3.6006,
+      "step": 15750
+    },
+    {
+      "epoch": 1.2485160269093787,
+      "grad_norm": 117.87637329101562,
+      "learning_rate": 1.9798111042224537e-07,
+      "loss": 3.7463,
+      "step": 15775
+    },
+    {
+      "epoch": 1.2504946576968738,
+      "grad_norm": 158.502685546875,
+      "learning_rate": 1.9789866638525769e-07,
+      "loss": 3.7699,
+      "step": 15800
+    },
+    {
+      "epoch": 1.2524732884843688,
+      "grad_norm": 213.62936401367188,
+      "learning_rate": 1.9781622234826997e-07,
+      "loss": 3.9892,
+      "step": 15825
+    },
+    {
+      "epoch": 1.254451919271864,
+      "grad_norm": 120.2110595703125,
+      "learning_rate": 1.9773377831128228e-07,
+      "loss": 3.6972,
+      "step": 15850
+    },
+    {
+      "epoch": 1.2564305500593589,
+      "grad_norm": 115.94564819335938,
+      "learning_rate": 1.976513342742946e-07,
+      "loss": 3.6659,
+      "step": 15875
+    },
+    {
+      "epoch": 1.258409180846854,
+      "grad_norm": 100.15674591064453,
+      "learning_rate": 1.975688902373069e-07,
+      "loss": 3.6585,
+      "step": 15900
+    },
+    {
+      "epoch": 1.260387811634349,
+      "grad_norm": 126.27003479003906,
+      "learning_rate": 1.9748644620031922e-07,
+      "loss": 3.793,
+      "step": 15925
+    },
+    {
+      "epoch": 1.2623664424218441,
+      "grad_norm": 122.43645477294922,
+      "learning_rate": 1.9740400216333153e-07,
+      "loss": 3.7562,
+      "step": 15950
+    },
+    {
+      "epoch": 1.264345073209339,
+      "grad_norm": 121.08145904541016,
+      "learning_rate": 1.9732155812634381e-07,
+      "loss": 3.759,
+      "step": 15975
+    },
+    {
+      "epoch": 1.2663237039968342,
+      "grad_norm": 116.81194305419922,
+      "learning_rate": 1.9723911408935615e-07,
+      "loss": 3.6303,
+      "step": 16000
+    },
+    {
+      "epoch": 1.2683023347843292,
+      "grad_norm": 107.8984146118164,
+      "learning_rate": 1.9715667005236844e-07,
+      "loss": 3.7525,
+      "step": 16025
+    },
+    {
+      "epoch": 1.2702809655718243,
+      "grad_norm": 114.40152740478516,
+      "learning_rate": 1.9707422601538075e-07,
+      "loss": 3.6044,
+      "step": 16050
+    },
+    {
+      "epoch": 1.2722595963593193,
+      "grad_norm": 128.336181640625,
+      "learning_rate": 1.9699178197839306e-07,
+      "loss": 3.6883,
+      "step": 16075
+    },
+    {
+      "epoch": 1.2742382271468145,
+      "grad_norm": 177.9373321533203,
+      "learning_rate": 1.9690933794140535e-07,
+      "loss": 3.7305,
+      "step": 16100
+    },
+    {
+      "epoch": 1.2762168579343094,
+      "grad_norm": 137.6163787841797,
+      "learning_rate": 1.9682689390441768e-07,
+      "loss": 3.4276,
+      "step": 16125
+    },
+    {
+      "epoch": 1.2781954887218046,
+      "grad_norm": 93.84378814697266,
+      "learning_rate": 1.9674444986742997e-07,
+      "loss": 3.954,
+      "step": 16150
+    },
+    {
+      "epoch": 1.2801741195092995,
+      "grad_norm": 114.7217788696289,
+      "learning_rate": 1.9666200583044228e-07,
+      "loss": 3.9032,
+      "step": 16175
+    },
+    {
+      "epoch": 1.2821527502967947,
+      "grad_norm": 127.64576721191406,
+      "learning_rate": 1.965795617934546e-07,
+      "loss": 3.9561,
+      "step": 16200
+    },
+    {
+      "epoch": 1.2841313810842896,
+      "grad_norm": 147.38531494140625,
+      "learning_rate": 1.964971177564669e-07,
+      "loss": 3.7639,
+      "step": 16225
+    },
+    {
+      "epoch": 1.2861100118717848,
+      "grad_norm": 106.52879333496094,
+      "learning_rate": 1.964146737194792e-07,
+      "loss": 3.6835,
+      "step": 16250
+    },
+    {
+      "epoch": 1.2880886426592797,
+      "grad_norm": 117.00553131103516,
+      "learning_rate": 1.9633222968249153e-07,
+      "loss": 3.5749,
+      "step": 16275
+    },
+    {
+      "epoch": 1.2900672734467749,
+      "grad_norm": 145.50201416015625,
+      "learning_rate": 1.962497856455038e-07,
+      "loss": 3.8452,
+      "step": 16300
+    },
+    {
+      "epoch": 1.29204590423427,
+      "grad_norm": 109.88056945800781,
+      "learning_rate": 1.9616734160851615e-07,
+      "loss": 3.7797,
+      "step": 16325
+    },
+    {
+      "epoch": 1.294024535021765,
+      "grad_norm": 98.50968170166016,
+      "learning_rate": 1.9608489757152844e-07,
+      "loss": 3.9423,
+      "step": 16350
+    },
+    {
+      "epoch": 1.29600316580926,
+      "grad_norm": 137.3463134765625,
+      "learning_rate": 1.9600245353454075e-07,
+      "loss": 3.8175,
+      "step": 16375
+    },
+    {
+      "epoch": 1.297981796596755,
+      "grad_norm": 104.44812774658203,
+      "learning_rate": 1.9592000949755306e-07,
+      "loss": 3.8284,
+      "step": 16400
+    },
+    {
+      "epoch": 1.2999604273842502,
+      "grad_norm": 134.79293823242188,
+      "learning_rate": 1.9583756546056534e-07,
+      "loss": 3.8295,
+      "step": 16425
+    },
+    {
+      "epoch": 1.3019390581717452,
+      "grad_norm": 136.05784606933594,
+      "learning_rate": 1.9575512142357766e-07,
+      "loss": 3.9577,
+      "step": 16450
+    },
+    {
+      "epoch": 1.30391768895924,
+      "grad_norm": 162.8216094970703,
+      "learning_rate": 1.9567267738658997e-07,
+      "loss": 3.9843,
+      "step": 16475
+    },
+    {
+      "epoch": 1.3058963197467353,
+      "grad_norm": 132.25741577148438,
+      "learning_rate": 1.9559023334960228e-07,
+      "loss": 3.8045,
+      "step": 16500
+    },
+    {
+      "epoch": 1.3078749505342304,
+      "grad_norm": 140.72642517089844,
+      "learning_rate": 1.955077893126146e-07,
+      "loss": 3.8217,
+      "step": 16525
+    },
+    {
+      "epoch": 1.3098535813217254,
+      "grad_norm": 119.49310302734375,
+      "learning_rate": 1.954253452756269e-07,
+      "loss": 3.6075,
+      "step": 16550
+    },
+    {
+      "epoch": 1.3118322121092203,
+      "grad_norm": 189.51280212402344,
+      "learning_rate": 1.953429012386392e-07,
+      "loss": 3.9516,
+      "step": 16575
+    },
+    {
+      "epoch": 1.3138108428967155,
+      "grad_norm": 106.88180541992188,
+      "learning_rate": 1.9526045720165153e-07,
+      "loss": 3.9774,
+      "step": 16600
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 136.42471313476562,
+      "learning_rate": 1.951780131646638e-07,
+      "loss": 3.7752,
+      "step": 16625
+    },
+    {
+      "epoch": 1.3177681044717056,
+      "grad_norm": 134.68948364257812,
+      "learning_rate": 1.9509556912767612e-07,
+      "loss": 3.6695,
+      "step": 16650
+    },
+    {
+      "epoch": 1.3197467352592005,
+      "grad_norm": 88.88003540039062,
+      "learning_rate": 1.9501312509068843e-07,
+      "loss": 3.9039,
+      "step": 16675
+    },
+    {
+      "epoch": 1.3217253660466957,
+      "grad_norm": 106.34395599365234,
+      "learning_rate": 1.9493068105370075e-07,
+      "loss": 3.6726,
+      "step": 16700
+    },
+    {
+      "epoch": 1.3237039968341908,
+      "grad_norm": 94.92339324951172,
+      "learning_rate": 1.9484823701671306e-07,
+      "loss": 3.7362,
+      "step": 16725
+    },
+    {
+      "epoch": 1.3256826276216858,
+      "grad_norm": 131.17221069335938,
+      "learning_rate": 1.9476579297972534e-07,
+      "loss": 3.6668,
+      "step": 16750
+    },
+    {
+      "epoch": 1.327661258409181,
+      "grad_norm": 115.50166320800781,
+      "learning_rate": 1.9468334894273765e-07,
+      "loss": 3.9881,
+      "step": 16775
+    },
+    {
+      "epoch": 1.3296398891966759,
+      "grad_norm": 146.72767639160156,
+      "learning_rate": 1.9460090490574997e-07,
+      "loss": 3.8375,
+      "step": 16800
+    },
+    {
+      "epoch": 1.331618519984171,
+      "grad_norm": 125.41651153564453,
+      "learning_rate": 1.9451846086876228e-07,
+      "loss": 3.7523,
+      "step": 16825
+    },
+    {
+      "epoch": 1.333597150771666,
+      "grad_norm": 148.9279327392578,
+      "learning_rate": 1.9443601683177456e-07,
+      "loss": 3.7209,
+      "step": 16850
+    },
+    {
+      "epoch": 1.3355757815591611,
+      "grad_norm": 132.95213317871094,
+      "learning_rate": 1.943535727947869e-07,
+      "loss": 3.8043,
+      "step": 16875
+    },
+    {
+      "epoch": 1.337554412346656,
+      "grad_norm": 121.32286834716797,
+      "learning_rate": 1.9427112875779919e-07,
+      "loss": 3.8463,
+      "step": 16900
+    },
+    {
+      "epoch": 1.3395330431341512,
+      "grad_norm": 104.13847351074219,
+      "learning_rate": 1.9418868472081152e-07,
+      "loss": 3.6028,
+      "step": 16925
+    },
+    {
+      "epoch": 1.3415116739216462,
+      "grad_norm": 116.32896423339844,
+      "learning_rate": 1.941062406838238e-07,
+      "loss": 3.6065,
+      "step": 16950
+    },
+    {
+      "epoch": 1.3434903047091413,
+      "grad_norm": 123.75446319580078,
+      "learning_rate": 1.9402379664683612e-07,
+      "loss": 3.8383,
+      "step": 16975
+    },
+    {
+      "epoch": 1.3454689354966363,
+      "grad_norm": 146.36181640625,
+      "learning_rate": 1.9394135260984843e-07,
+      "loss": 3.7784,
+      "step": 17000
+    },
+    {
+      "epoch": 1.3474475662841314,
+      "grad_norm": 153.75225830078125,
+      "learning_rate": 1.9385890857286074e-07,
+      "loss": 3.8691,
+      "step": 17025
+    },
+    {
+      "epoch": 1.3494261970716264,
+      "grad_norm": 144.79983520507812,
+      "learning_rate": 1.9377646453587303e-07,
+      "loss": 3.7902,
+      "step": 17050
+    },
+    {
+      "epoch": 1.3514048278591215,
+      "grad_norm": 107.72118377685547,
+      "learning_rate": 1.9369402049888534e-07,
+      "loss": 3.7424,
+      "step": 17075
+    },
+    {
+      "epoch": 1.3533834586466165,
+      "grad_norm": 107.33843994140625,
+      "learning_rate": 1.9361157646189765e-07,
+      "loss": 3.7423,
+      "step": 17100
+    },
+    {
+      "epoch": 1.3553620894341116,
+      "grad_norm": 154.79566955566406,
+      "learning_rate": 1.9352913242490996e-07,
+      "loss": 3.5984,
+      "step": 17125
+    },
+    {
+      "epoch": 1.3573407202216066,
+      "grad_norm": 118.29646301269531,
+      "learning_rate": 1.9344668838792227e-07,
+      "loss": 3.6184,
+      "step": 17150
+    },
+    {
+      "epoch": 1.3593193510091017,
+      "grad_norm": 194.39981079101562,
+      "learning_rate": 1.9336424435093456e-07,
+      "loss": 3.7566,
+      "step": 17175
+    },
+    {
+      "epoch": 1.361297981796597,
+      "grad_norm": 126.4017333984375,
+      "learning_rate": 1.932818003139469e-07,
+      "loss": 3.7378,
+      "step": 17200
+    },
+    {
+      "epoch": 1.3632766125840918,
+      "grad_norm": 154.46499633789062,
+      "learning_rate": 1.9319935627695918e-07,
+      "loss": 3.575,
+      "step": 17225
+    },
+    {
+      "epoch": 1.3652552433715868,
+      "grad_norm": 121.02112579345703,
+      "learning_rate": 1.9311691223997152e-07,
+      "loss": 3.5395,
+      "step": 17250
+    },
+    {
+      "epoch": 1.367233874159082,
+      "grad_norm": 112.81676483154297,
+      "learning_rate": 1.930344682029838e-07,
+      "loss": 3.7668,
+      "step": 17275
+    },
+    {
+      "epoch": 1.369212504946577,
+      "grad_norm": 128.57958984375,
+      "learning_rate": 1.9295202416599612e-07,
+      "loss": 3.6508,
+      "step": 17300
+    },
+    {
+      "epoch": 1.371191135734072,
+      "grad_norm": 100.09276580810547,
+      "learning_rate": 1.9286958012900843e-07,
+      "loss": 3.568,
+      "step": 17325
+    },
+    {
+      "epoch": 1.373169766521567,
+      "grad_norm": 102.0973892211914,
+      "learning_rate": 1.9278713609202074e-07,
+      "loss": 3.6707,
+      "step": 17350
+    },
+    {
+      "epoch": 1.3751483973090621,
+      "grad_norm": 180.7505645751953,
+      "learning_rate": 1.9270469205503303e-07,
+      "loss": 3.6053,
+      "step": 17375
+    },
+    {
+      "epoch": 1.3771270280965573,
+      "grad_norm": 141.00286865234375,
+      "learning_rate": 1.9262224801804534e-07,
+      "loss": 3.802,
+      "step": 17400
+    },
+    {
+      "epoch": 1.3791056588840522,
+      "grad_norm": 105.14612579345703,
+      "learning_rate": 1.9253980398105765e-07,
+      "loss": 3.8125,
+      "step": 17425
+    },
+    {
+      "epoch": 1.3810842896715472,
+      "grad_norm": 134.7199249267578,
+      "learning_rate": 1.9245735994406993e-07,
+      "loss": 3.6837,
+      "step": 17450
+    },
+    {
+      "epoch": 1.3830629204590423,
+      "grad_norm": 161.79367065429688,
+      "learning_rate": 1.9237491590708227e-07,
+      "loss": 3.4173,
+      "step": 17475
+    },
+    {
+      "epoch": 1.3850415512465375,
+      "grad_norm": 90.2686996459961,
+      "learning_rate": 1.9229247187009456e-07,
+      "loss": 3.7035,
+      "step": 17500
+    },
+    {
+      "epoch": 1.3870201820340324,
+      "grad_norm": 138.9261016845703,
+      "learning_rate": 1.922100278331069e-07,
+      "loss": 3.7375,
+      "step": 17525
+    },
+    {
+      "epoch": 1.3889988128215274,
+      "grad_norm": 99.18257141113281,
+      "learning_rate": 1.9212758379611918e-07,
+      "loss": 3.7228,
+      "step": 17550
+    },
+    {
+      "epoch": 1.3909774436090225,
+      "grad_norm": 108.61868286132812,
+      "learning_rate": 1.920451397591315e-07,
+      "loss": 3.8505,
+      "step": 17575
+    },
+    {
+      "epoch": 1.3929560743965177,
+      "grad_norm": 89.02510070800781,
+      "learning_rate": 1.919626957221438e-07,
+      "loss": 3.7366,
+      "step": 17600
+    },
+    {
+      "epoch": 1.3949347051840126,
+      "grad_norm": 183.8499755859375,
+      "learning_rate": 1.9188025168515612e-07,
+      "loss": 3.7842,
+      "step": 17625
+    },
+    {
+      "epoch": 1.3969133359715078,
+      "grad_norm": 111.4132080078125,
+      "learning_rate": 1.917978076481684e-07,
+      "loss": 4.0444,
+      "step": 17650
+    },
+    {
+      "epoch": 1.3988919667590027,
+      "grad_norm": 114.52892303466797,
+      "learning_rate": 1.917153636111807e-07,
+      "loss": 3.6568,
+      "step": 17675
+    },
+    {
+      "epoch": 1.400870597546498,
+      "grad_norm": 108.65158081054688,
+      "learning_rate": 1.9163291957419302e-07,
+      "loss": 3.7598,
+      "step": 17700
+    },
+    {
+      "epoch": 1.4028492283339928,
+      "grad_norm": 118.24126434326172,
+      "learning_rate": 1.9155047553720534e-07,
+      "loss": 3.5768,
+      "step": 17725
+    },
+    {
+      "epoch": 1.404827859121488,
+      "grad_norm": 116.43769073486328,
+      "learning_rate": 1.9146803150021765e-07,
+      "loss": 3.6195,
+      "step": 17750
+    },
+    {
+      "epoch": 1.406806489908983,
+      "grad_norm": 96.20083618164062,
+      "learning_rate": 1.9138558746322993e-07,
+      "loss": 3.7732,
+      "step": 17775
+    },
+    {
+      "epoch": 1.408785120696478,
+      "grad_norm": 106.93498229980469,
+      "learning_rate": 1.9130314342624227e-07,
+      "loss": 3.5733,
+      "step": 17800
+    },
+    {
+      "epoch": 1.410763751483973,
+      "grad_norm": 126.78841400146484,
+      "learning_rate": 1.9122069938925456e-07,
+      "loss": 3.4865,
+      "step": 17825
+    },
+    {
+      "epoch": 1.4127423822714682,
+      "grad_norm": 161.895263671875,
+      "learning_rate": 1.911382553522669e-07,
+      "loss": 3.7985,
+      "step": 17850
+    },
+    {
+      "epoch": 1.4147210130589631,
+      "grad_norm": 137.6743927001953,
+      "learning_rate": 1.9105581131527918e-07,
+      "loss": 3.5315,
+      "step": 17875
+    },
+    {
+      "epoch": 1.4166996438464583,
+      "grad_norm": 81.24407196044922,
+      "learning_rate": 1.909733672782915e-07,
+      "loss": 3.6007,
+      "step": 17900
+    },
+    {
+      "epoch": 1.4186782746339532,
+      "grad_norm": 109.42854309082031,
+      "learning_rate": 1.908909232413038e-07,
+      "loss": 3.5264,
+      "step": 17925
+    },
+    {
+      "epoch": 1.4206569054214484,
+      "grad_norm": 113.56609344482422,
+      "learning_rate": 1.9080847920431611e-07,
+      "loss": 3.7304,
+      "step": 17950
+    },
+    {
+      "epoch": 1.4226355362089433,
+      "grad_norm": 148.14793395996094,
+      "learning_rate": 1.907260351673284e-07,
+      "loss": 3.6923,
+      "step": 17975
+    },
+    {
+      "epoch": 1.4246141669964385,
+      "grad_norm": 141.70896911621094,
+      "learning_rate": 1.906435911303407e-07,
+      "loss": 3.813,
+      "step": 18000
+    },
+    {
+      "epoch": 1.4265927977839334,
+      "grad_norm": 107.8005142211914,
+      "learning_rate": 1.9056114709335302e-07,
+      "loss": 3.8025,
+      "step": 18025
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 119.71678924560547,
+      "learning_rate": 1.904787030563653e-07,
+      "loss": 3.6632,
+      "step": 18050
+    },
+    {
+      "epoch": 1.4305500593589235,
+      "grad_norm": 89.46141815185547,
+      "learning_rate": 1.9039625901937765e-07,
+      "loss": 3.5954,
+      "step": 18075
+    },
+    {
+      "epoch": 1.4325286901464187,
+      "grad_norm": 94.6341552734375,
+      "learning_rate": 1.9031381498238993e-07,
+      "loss": 3.7922,
+      "step": 18100
+    },
+    {
+      "epoch": 1.4345073209339136,
+      "grad_norm": 119.52442932128906,
+      "learning_rate": 1.9023137094540227e-07,
+      "loss": 3.8677,
+      "step": 18125
+    },
+    {
+      "epoch": 1.4364859517214088,
+      "grad_norm": 85.59278869628906,
+      "learning_rate": 1.9014892690841455e-07,
+      "loss": 3.7679,
+      "step": 18150
+    },
+    {
+      "epoch": 1.438464582508904,
+      "grad_norm": 140.89190673828125,
+      "learning_rate": 1.9006648287142687e-07,
+      "loss": 3.7617,
+      "step": 18175
+    },
+    {
+      "epoch": 1.440443213296399,
+      "grad_norm": 114.29212188720703,
+      "learning_rate": 1.8998403883443918e-07,
+      "loss": 3.9376,
+      "step": 18200
+    },
+    {
+      "epoch": 1.4424218440838938,
+      "grad_norm": 118.00497436523438,
+      "learning_rate": 1.899015947974515e-07,
+      "loss": 3.7841,
+      "step": 18225
+    },
+    {
+      "epoch": 1.444400474871389,
+      "grad_norm": 140.94200134277344,
+      "learning_rate": 1.8981915076046377e-07,
+      "loss": 3.694,
+      "step": 18250
+    },
+    {
+      "epoch": 1.4463791056588842,
+      "grad_norm": 153.66397094726562,
+      "learning_rate": 1.897367067234761e-07,
+      "loss": 3.7603,
+      "step": 18275
+    },
+    {
+      "epoch": 1.448357736446379,
+      "grad_norm": 128.06732177734375,
+      "learning_rate": 1.896542626864884e-07,
+      "loss": 3.7799,
+      "step": 18300
+    },
+    {
+      "epoch": 1.450336367233874,
+      "grad_norm": 112.23291778564453,
+      "learning_rate": 1.895718186495007e-07,
+      "loss": 3.7655,
+      "step": 18325
+    },
+    {
+      "epoch": 1.4523149980213692,
+      "grad_norm": 91.07350158691406,
+      "learning_rate": 1.8948937461251302e-07,
+      "loss": 3.643,
+      "step": 18350
+    },
+    {
+      "epoch": 1.4542936288088644,
+      "grad_norm": 98.65179443359375,
+      "learning_rate": 1.894069305755253e-07,
+      "loss": 3.7346,
+      "step": 18375
+    },
+    {
+      "epoch": 1.4562722595963593,
+      "grad_norm": 118.8671875,
+      "learning_rate": 1.8932448653853764e-07,
+      "loss": 3.7361,
+      "step": 18400
+    },
+    {
+      "epoch": 1.4582508903838542,
+      "grad_norm": 121.17200469970703,
+      "learning_rate": 1.8924204250154993e-07,
+      "loss": 3.7823,
+      "step": 18425
+    },
+    {
+      "epoch": 1.4602295211713494,
+      "grad_norm": 91.53225708007812,
+      "learning_rate": 1.8915959846456227e-07,
+      "loss": 3.659,
+      "step": 18450
+    },
+    {
+      "epoch": 1.4622081519588446,
+      "grad_norm": 140.78236389160156,
+      "learning_rate": 1.8907715442757455e-07,
+      "loss": 3.8495,
+      "step": 18475
+    },
+    {
+      "epoch": 1.4641867827463395,
+      "grad_norm": 136.43206787109375,
+      "learning_rate": 1.8899471039058686e-07,
+      "loss": 3.8468,
+      "step": 18500
+    },
+    {
+      "epoch": 1.4661654135338344,
+      "grad_norm": 106.83531951904297,
+      "learning_rate": 1.8891226635359917e-07,
+      "loss": 3.6799,
+      "step": 18525
+    },
+    {
+      "epoch": 1.4681440443213296,
+      "grad_norm": 132.8029327392578,
+      "learning_rate": 1.8882982231661149e-07,
+      "loss": 3.7441,
+      "step": 18550
+    },
+    {
+      "epoch": 1.4701226751088248,
+      "grad_norm": 125.74601745605469,
+      "learning_rate": 1.8874737827962377e-07,
+      "loss": 3.586,
+      "step": 18575
+    },
+    {
+      "epoch": 1.4721013058963197,
+      "grad_norm": 126.61151123046875,
+      "learning_rate": 1.886649342426361e-07,
+      "loss": 3.7686,
+      "step": 18600
+    },
+    {
+      "epoch": 1.4740799366838149,
+      "grad_norm": 127.62877655029297,
+      "learning_rate": 1.885824902056484e-07,
+      "loss": 3.74,
+      "step": 18625
+    },
+    {
+      "epoch": 1.4760585674713098,
+      "grad_norm": 134.3148193359375,
+      "learning_rate": 1.8850004616866068e-07,
+      "loss": 3.8815,
+      "step": 18650
+    },
+    {
+      "epoch": 1.478037198258805,
+      "grad_norm": 167.0948028564453,
+      "learning_rate": 1.8841760213167302e-07,
+      "loss": 3.7258,
+      "step": 18675
+    },
+    {
+      "epoch": 1.4800158290463,
+      "grad_norm": 103.17122650146484,
+      "learning_rate": 1.883351580946853e-07,
+      "loss": 3.7211,
+      "step": 18700
+    },
+    {
+      "epoch": 1.481994459833795,
+      "grad_norm": 98.9283218383789,
+      "learning_rate": 1.8825271405769764e-07,
+      "loss": 3.6401,
+      "step": 18725
+    },
+    {
+      "epoch": 1.48397309062129,
+      "grad_norm": 110.53363800048828,
+      "learning_rate": 1.8817027002070993e-07,
+      "loss": 3.84,
+      "step": 18750
+    },
+    {
+      "epoch": 1.4859517214087852,
+      "grad_norm": 115.60739135742188,
+      "learning_rate": 1.8808782598372224e-07,
+      "loss": 3.9043,
+      "step": 18775
+    },
+    {
+      "epoch": 1.4879303521962801,
+      "grad_norm": 130.4183807373047,
+      "learning_rate": 1.8800538194673455e-07,
+      "loss": 3.6778,
+      "step": 18800
+    },
+    {
+      "epoch": 1.4899089829837753,
+      "grad_norm": 94.30120086669922,
+      "learning_rate": 1.8792293790974686e-07,
+      "loss": 3.5853,
+      "step": 18825
+    },
+    {
+      "epoch": 1.4918876137712702,
+      "grad_norm": 102.75646209716797,
+      "learning_rate": 1.8784049387275917e-07,
+      "loss": 4.0326,
+      "step": 18850
+    },
+    {
+      "epoch": 1.4938662445587654,
+      "grad_norm": 153.1327362060547,
+      "learning_rate": 1.8775804983577148e-07,
+      "loss": 3.8785,
+      "step": 18875
+    },
+    {
+      "epoch": 1.4958448753462603,
+      "grad_norm": 100.03012084960938,
+      "learning_rate": 1.8767560579878377e-07,
+      "loss": 3.692,
+      "step": 18900
+    },
+    {
+      "epoch": 1.4978235061337555,
+      "grad_norm": 128.64657592773438,
+      "learning_rate": 1.8759316176179608e-07,
+      "loss": 3.7799,
+      "step": 18925
+    },
+    {
+      "epoch": 1.4998021369212504,
+      "grad_norm": 99.55155944824219,
+      "learning_rate": 1.875107177248084e-07,
+      "loss": 3.7112,
+      "step": 18950
+    },
+    {
+      "epoch": 1.5017807677087456,
+      "grad_norm": 171.94178771972656,
+      "learning_rate": 1.8742827368782068e-07,
+      "loss": 3.7663,
+      "step": 18975
+    },
+    {
+      "epoch": 1.5037593984962405,
+      "grad_norm": 107.23114013671875,
+      "learning_rate": 1.8734582965083302e-07,
+      "loss": 3.6127,
+      "step": 19000
+    },
+    {
+      "epoch": 1.5057380292837357,
+      "grad_norm": 101.01669311523438,
+      "learning_rate": 1.872633856138453e-07,
+      "loss": 3.5948,
+      "step": 19025
+    },
+    {
+      "epoch": 1.5077166600712308,
+      "grad_norm": 158.38211059570312,
+      "learning_rate": 1.8718094157685764e-07,
+      "loss": 3.5552,
+      "step": 19050
+    },
+    {
+      "epoch": 1.5096952908587258,
+      "grad_norm": 117.05270385742188,
+      "learning_rate": 1.8709849753986992e-07,
+      "loss": 3.7912,
+      "step": 19075
+    },
+    {
+      "epoch": 1.5116739216462207,
+      "grad_norm": 117.87310028076172,
+      "learning_rate": 1.8701605350288224e-07,
+      "loss": 3.7512,
+      "step": 19100
+    },
+    {
+      "epoch": 1.5136525524337159,
+      "grad_norm": 106.46678161621094,
+      "learning_rate": 1.8693360946589455e-07,
+      "loss": 3.7715,
+      "step": 19125
+    },
+    {
+      "epoch": 1.515631183221211,
+      "grad_norm": 107.43925476074219,
+      "learning_rate": 1.8685116542890686e-07,
+      "loss": 3.7395,
+      "step": 19150
+    },
+    {
+      "epoch": 1.517609814008706,
+      "grad_norm": 101.0313491821289,
+      "learning_rate": 1.8676872139191914e-07,
+      "loss": 3.6996,
+      "step": 19175
+    },
+    {
+      "epoch": 1.519588444796201,
+      "grad_norm": 179.96051025390625,
+      "learning_rate": 1.8668627735493148e-07,
+      "loss": 3.6133,
+      "step": 19200
+    },
+    {
+      "epoch": 1.521567075583696,
+      "grad_norm": 90.7691879272461,
+      "learning_rate": 1.8660383331794377e-07,
+      "loss": 3.6053,
+      "step": 19225
+    },
+    {
+      "epoch": 1.5235457063711912,
+      "grad_norm": 137.12637329101562,
+      "learning_rate": 1.8652138928095605e-07,
+      "loss": 3.6739,
+      "step": 19250
+    },
+    {
+      "epoch": 1.5255243371586862,
+      "grad_norm": 138.26014709472656,
+      "learning_rate": 1.864389452439684e-07,
+      "loss": 3.8328,
+      "step": 19275
+    },
+    {
+      "epoch": 1.5275029679461811,
+      "grad_norm": 94.83419799804688,
+      "learning_rate": 1.8635650120698068e-07,
+      "loss": 3.7555,
+      "step": 19300
+    },
+    {
+      "epoch": 1.5294815987336763,
+      "grad_norm": 105.08883666992188,
+      "learning_rate": 1.8627405716999301e-07,
+      "loss": 3.8474,
+      "step": 19325
+    },
+    {
+      "epoch": 1.5314602295211714,
+      "grad_norm": 124.04432678222656,
+      "learning_rate": 1.861916131330053e-07,
+      "loss": 3.5361,
+      "step": 19350
+    },
+    {
+      "epoch": 1.5334388603086664,
+      "grad_norm": 97.83289337158203,
+      "learning_rate": 1.861091690960176e-07,
+      "loss": 3.6283,
+      "step": 19375
+    },
+    {
+      "epoch": 1.5354174910961613,
+      "grad_norm": 110.26960754394531,
+      "learning_rate": 1.8602672505902992e-07,
+      "loss": 3.7872,
+      "step": 19400
+    },
+    {
+      "epoch": 1.5373961218836565,
+      "grad_norm": 115.99392700195312,
+      "learning_rate": 1.8594428102204223e-07,
+      "loss": 3.6191,
+      "step": 19425
+    },
+    {
+      "epoch": 1.5393747526711516,
+      "grad_norm": 105.29808044433594,
+      "learning_rate": 1.8586183698505454e-07,
+      "loss": 3.6837,
+      "step": 19450
+    },
+    {
+      "epoch": 1.5413533834586466,
+      "grad_norm": 100.55680847167969,
+      "learning_rate": 1.8577939294806686e-07,
+      "loss": 3.7439,
+      "step": 19475
+    },
+    {
+      "epoch": 1.5433320142461415,
+      "grad_norm": 117.83526611328125,
+      "learning_rate": 1.8569694891107914e-07,
+      "loss": 3.5938,
+      "step": 19500
+    },
+    {
+      "epoch": 1.5453106450336367,
+      "grad_norm": 84.95865631103516,
+      "learning_rate": 1.8561450487409148e-07,
+      "loss": 3.7425,
+      "step": 19525
+    },
+    {
+      "epoch": 1.5472892758211318,
+      "grad_norm": 128.07754516601562,
+      "learning_rate": 1.8553206083710376e-07,
+      "loss": 3.7379,
+      "step": 19550
+    },
+    {
+      "epoch": 1.5492679066086268,
+      "grad_norm": 103.81727600097656,
+      "learning_rate": 1.8544961680011605e-07,
+      "loss": 3.8121,
+      "step": 19575
+    },
+    {
+      "epoch": 1.5512465373961217,
+      "grad_norm": 113.71387481689453,
+      "learning_rate": 1.853671727631284e-07,
+      "loss": 3.5854,
+      "step": 19600
+    },
+    {
+      "epoch": 1.5532251681836169,
+      "grad_norm": 106.94216918945312,
+      "learning_rate": 1.8528472872614067e-07,
+      "loss": 3.7723,
+      "step": 19625
+    },
+    {
+      "epoch": 1.555203798971112,
+      "grad_norm": 113.77094268798828,
+      "learning_rate": 1.85202284689153e-07,
+      "loss": 3.7218,
+      "step": 19650
+    },
+    {
+      "epoch": 1.557182429758607,
+      "grad_norm": 102.69156646728516,
+      "learning_rate": 1.851198406521653e-07,
+      "loss": 3.5478,
+      "step": 19675
+    },
+    {
+      "epoch": 1.5591610605461022,
+      "grad_norm": 146.14686584472656,
+      "learning_rate": 1.850373966151776e-07,
+      "loss": 3.5888,
+      "step": 19700
+    },
+    {
+      "epoch": 1.5611396913335973,
+      "grad_norm": 117.55293273925781,
+      "learning_rate": 1.8495495257818992e-07,
+      "loss": 3.5802,
+      "step": 19725
+    },
+    {
+      "epoch": 1.5631183221210923,
+      "grad_norm": 116.557861328125,
+      "learning_rate": 1.8487250854120223e-07,
+      "loss": 3.6588,
+      "step": 19750
+    },
+    {
+      "epoch": 1.5650969529085872,
+      "grad_norm": 143.797607421875,
+      "learning_rate": 1.8479006450421452e-07,
+      "loss": 3.7888,
+      "step": 19775
+    },
+    {
+      "epoch": 1.5670755836960824,
+      "grad_norm": 90.89989471435547,
+      "learning_rate": 1.8470762046722685e-07,
+      "loss": 3.7565,
+      "step": 19800
+    },
+    {
+      "epoch": 1.5690542144835775,
+      "grad_norm": 93.96920776367188,
+      "learning_rate": 1.8462517643023914e-07,
+      "loss": 3.7576,
+      "step": 19825
+    },
+    {
+      "epoch": 1.5710328452710725,
+      "grad_norm": 146.9014129638672,
+      "learning_rate": 1.8454273239325148e-07,
+      "loss": 3.7507,
+      "step": 19850
+    },
+    {
+      "epoch": 1.5730114760585674,
+      "grad_norm": 119.38358306884766,
+      "learning_rate": 1.8446028835626376e-07,
+      "loss": 3.6588,
+      "step": 19875
+    },
+    {
+      "epoch": 1.5749901068460626,
+      "grad_norm": 113.38998413085938,
+      "learning_rate": 1.8437784431927605e-07,
+      "loss": 3.7866,
+      "step": 19900
+    },
+    {
+      "epoch": 1.5769687376335577,
+      "grad_norm": 103.051513671875,
+      "learning_rate": 1.8429540028228839e-07,
+      "loss": 3.6443,
+      "step": 19925
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 94.77304077148438,
+      "learning_rate": 1.8421295624530067e-07,
+      "loss": 3.6966,
+      "step": 19950
+    },
+    {
+      "epoch": 1.5809259992085476,
+      "grad_norm": 167.68310546875,
+      "learning_rate": 1.8413051220831298e-07,
+      "loss": 3.684,
+      "step": 19975
+    },
+    {
+      "epoch": 1.5829046299960428,
+      "grad_norm": 96.78482055664062,
+      "learning_rate": 1.840480681713253e-07,
+      "loss": 3.733,
+      "step": 20000
+    },
+    {
+      "epoch": 1.5829046299960428,
+      "eval_loss": 3.6955573558807373,
+      "eval_runtime": 9.5434,
+      "eval_samples_per_second": 264.894,
+      "eval_steps_per_second": 33.112,
+      "step": 20000
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 75810,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 10000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 71534592000000.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}