diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.10006054268259122,
+  "epoch": 0.15009081402388685,
   "eval_steps": 500,
-  "global_step": 1818,
+  "global_step": 2727,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -12733,6 +12733,6369 @@
       "learning_rate": 9.94464527995885e-06,
       "loss": 0.8529,
       "step": 1818
+    },
+    {
+      "epoch": 0.10011558148494688,
+      "grad_norm": 0.9318063259124756,
+      "learning_rate": 9.944580939660501e-06,
+      "loss": 0.8978,
+      "step": 1819
+    },
+    {
+      "epoch": 0.10017062028730254,
+      "grad_norm": 0.847023069858551,
+      "learning_rate": 9.944516562200004e-06,
+      "loss": 0.8007,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1002256590896582,
+      "grad_norm": 0.8817011117935181,
+      "learning_rate": 9.944452147577844e-06,
+      "loss": 0.8819,
+      "step": 1821
+    },
+    {
+      "epoch": 0.10028069789201387,
+      "grad_norm": 0.8560144901275635,
+      "learning_rate": 9.944387695794505e-06,
+      "loss": 0.8219,
+      "step": 1822
+    },
+    {
+      "epoch": 0.10033573669436953,
+      "grad_norm": 0.9358342885971069,
+      "learning_rate": 9.944323206850472e-06,
+      "loss": 0.8533,
+      "step": 1823
+    },
+    {
+      "epoch": 0.10039077549672519,
+      "grad_norm": 0.8327087163925171,
+      "learning_rate": 9.94425868074623e-06,
+      "loss": 0.8359,
+      "step": 1824
+    },
+    {
+      "epoch": 0.10044581429908085,
+      "grad_norm": 1.0590367317199707,
+      "learning_rate": 9.944194117482263e-06,
+      "loss": 0.9659,
+      "step": 1825
+    },
+    {
+      "epoch": 0.10050085310143651,
+      "grad_norm": 0.8739829063415527,
+      "learning_rate": 9.944129517059055e-06,
+      "loss": 0.7868,
+      "step": 1826
+    },
+    {
+      "epoch": 0.10055589190379217,
+      "grad_norm": 0.8465235233306885,
+      "learning_rate": 9.944064879477093e-06,
+      "loss": 0.8554,
+      "step": 1827
+    },
+    {
+      "epoch": 0.10061093070614784,
+      "grad_norm": 0.9068321585655212,
+      "learning_rate": 9.944000204736864e-06,
+      "loss": 0.8648,
+      "step": 1828
+    },
+    {
+      "epoch": 0.1006659695085035,
+      "grad_norm": 0.8308066725730896,
+      "learning_rate": 9.943935492838853e-06,
+      "loss": 0.8471,
+      "step": 1829
+    },
+    {
+      "epoch": 0.10072100831085916,
+      "grad_norm": 0.9973901510238647,
+      "learning_rate": 9.943870743783545e-06,
+      "loss": 0.9398,
+      "step": 1830
+    },
+    {
+      "epoch": 0.10077604711321482,
+      "grad_norm": 0.8532593250274658,
+      "learning_rate": 9.94380595757143e-06,
+      "loss": 0.9001,
+      "step": 1831
+    },
+    {
+      "epoch": 0.10083108591557048,
+      "grad_norm": 0.8571139574050903,
+      "learning_rate": 9.94374113420299e-06,
+      "loss": 0.85,
+      "step": 1832
+    },
+    {
+      "epoch": 0.10088612471792614,
+      "grad_norm": 0.905624508857727,
+      "learning_rate": 9.943676273678717e-06,
+      "loss": 0.9587,
+      "step": 1833
+    },
+    {
+      "epoch": 0.1009411635202818,
+      "grad_norm": 1.0224663019180298,
+      "learning_rate": 9.943611375999097e-06,
+      "loss": 0.8236,
+      "step": 1834
+    },
+    {
+      "epoch": 0.10099620232263747,
+      "grad_norm": 0.8900588154792786,
+      "learning_rate": 9.943546441164615e-06,
+      "loss": 0.877,
+      "step": 1835
+    },
+    {
+      "epoch": 0.10105124112499313,
+      "grad_norm": 0.8852938413619995,
+      "learning_rate": 9.943481469175765e-06,
+      "loss": 0.9521,
+      "step": 1836
+    },
+    {
+      "epoch": 0.10110627992734877,
+      "grad_norm": 0.9249371290206909,
+      "learning_rate": 9.943416460033027e-06,
+      "loss": 0.8541,
+      "step": 1837
+    },
+    {
+      "epoch": 0.10116131872970444,
+      "grad_norm": 0.8533583283424377,
+      "learning_rate": 9.943351413736897e-06,
+      "loss": 0.8571,
+      "step": 1838
+    },
+    {
+      "epoch": 0.1012163575320601,
+      "grad_norm": 0.743800699710846,
+      "learning_rate": 9.94328633028786e-06,
+      "loss": 0.749,
+      "step": 1839
+    },
+    {
+      "epoch": 0.10127139633441576,
+      "grad_norm": 0.7836641669273376,
+      "learning_rate": 9.943221209686407e-06,
+      "loss": 0.8237,
+      "step": 1840
+    },
+    {
+      "epoch": 0.10132643513677142,
+      "grad_norm": 0.800782322883606,
+      "learning_rate": 9.943156051933024e-06,
+      "loss": 0.8323,
+      "step": 1841
+    },
+    {
+      "epoch": 0.10138147393912708,
+      "grad_norm": 0.7531478404998779,
+      "learning_rate": 9.943090857028206e-06,
+      "loss": 0.8041,
+      "step": 1842
+    },
+    {
+      "epoch": 0.10143651274148274,
+      "grad_norm": 0.9837996959686279,
+      "learning_rate": 9.94302562497244e-06,
+      "loss": 0.8084,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1014915515438384,
+      "grad_norm": 0.8038331866264343,
+      "learning_rate": 9.942960355766216e-06,
+      "loss": 0.8454,
+      "step": 1844
+    },
+    {
+      "epoch": 0.10154659034619407,
+      "grad_norm": 0.7822145819664001,
+      "learning_rate": 9.942895049410024e-06,
+      "loss": 0.8137,
+      "step": 1845
+    },
+    {
+      "epoch": 0.10160162914854973,
+      "grad_norm": 0.8222663998603821,
+      "learning_rate": 9.942829705904358e-06,
+      "loss": 0.8981,
+      "step": 1846
+    },
+    {
+      "epoch": 0.10165666795090539,
+      "grad_norm": 1.0095717906951904,
+      "learning_rate": 9.942764325249707e-06,
+      "loss": 0.9159,
+      "step": 1847
+    },
+    {
+      "epoch": 0.10171170675326105,
+      "grad_norm": 0.8264054656028748,
+      "learning_rate": 9.942698907446561e-06,
+      "loss": 0.9233,
+      "step": 1848
+    },
+    {
+      "epoch": 0.10176674555561671,
+      "grad_norm": 0.8244288563728333,
+      "learning_rate": 9.942633452495414e-06,
+      "loss": 0.8507,
+      "step": 1849
+    },
+    {
+      "epoch": 0.10182178435797237,
+      "grad_norm": 0.8457715511322021,
+      "learning_rate": 9.942567960396755e-06,
+      "loss": 0.7897,
+      "step": 1850
+    },
+    {
+      "epoch": 0.10187682316032803,
+      "grad_norm": 0.8356698155403137,
+      "learning_rate": 9.94250243115108e-06,
+      "loss": 0.7927,
+      "step": 1851
+    },
+    {
+      "epoch": 0.1019318619626837,
+      "grad_norm": 0.8251230716705322,
+      "learning_rate": 9.94243686475888e-06,
+      "loss": 0.8977,
+      "step": 1852
+    },
+    {
+      "epoch": 0.10198690076503936,
+      "grad_norm": 0.8370125889778137,
+      "learning_rate": 9.942371261220647e-06,
+      "loss": 0.8204,
+      "step": 1853
+    },
+    {
+      "epoch": 0.10204193956739502,
+      "grad_norm": 1.6722066402435303,
+      "learning_rate": 9.942305620536876e-06,
+      "loss": 0.9284,
+      "step": 1854
+    },
+    {
+      "epoch": 0.10209697836975068,
+      "grad_norm": 0.8424906730651855,
+      "learning_rate": 9.942239942708057e-06,
+      "loss": 0.833,
+      "step": 1855
+    },
+    {
+      "epoch": 0.10215201717210634,
+      "grad_norm": 0.7475115656852722,
+      "learning_rate": 9.942174227734686e-06,
+      "loss": 0.6158,
+      "step": 1856
+    },
+    {
+      "epoch": 0.102207055974462,
+      "grad_norm": 0.8652095198631287,
+      "learning_rate": 9.942108475617256e-06,
+      "loss": 0.8781,
+      "step": 1857
+    },
+    {
+      "epoch": 0.10226209477681765,
+      "grad_norm": 1.0621691942214966,
+      "learning_rate": 9.942042686356263e-06,
+      "loss": 1.0276,
+      "step": 1858
+    },
+    {
+      "epoch": 0.10231713357917331,
+      "grad_norm": 1.113357424736023,
+      "learning_rate": 9.941976859952199e-06,
+      "loss": 0.8799,
+      "step": 1859
+    },
+    {
+      "epoch": 0.10237217238152897,
+      "grad_norm": 0.9153568148612976,
+      "learning_rate": 9.94191099640556e-06,
+      "loss": 0.7988,
+      "step": 1860
+    },
+    {
+      "epoch": 0.10242721118388463,
+      "grad_norm": 0.9217341542243958,
+      "learning_rate": 9.941845095716842e-06,
+      "loss": 0.7785,
+      "step": 1861
+    },
+    {
+      "epoch": 0.1024822499862403,
+      "grad_norm": 0.8702190518379211,
+      "learning_rate": 9.941779157886538e-06,
+      "loss": 0.7648,
+      "step": 1862
+    },
+    {
+      "epoch": 0.10253728878859596,
+      "grad_norm": 0.8609822988510132,
+      "learning_rate": 9.941713182915144e-06,
+      "loss": 0.9095,
+      "step": 1863
+    },
+    {
+      "epoch": 0.10259232759095162,
+      "grad_norm": 0.7766719460487366,
+      "learning_rate": 9.941647170803157e-06,
+      "loss": 0.6984,
+      "step": 1864
+    },
+    {
+      "epoch": 0.10264736639330728,
+      "grad_norm": 0.8497375249862671,
+      "learning_rate": 9.941581121551074e-06,
+      "loss": 0.9161,
+      "step": 1865
+    },
+    {
+      "epoch": 0.10270240519566294,
+      "grad_norm": 0.8007600903511047,
+      "learning_rate": 9.941515035159388e-06,
+      "loss": 0.8099,
+      "step": 1866
+    },
+    {
+      "epoch": 0.1027574439980186,
+      "grad_norm": 0.7932959794998169,
+      "learning_rate": 9.941448911628599e-06,
+      "loss": 0.8049,
+      "step": 1867
+    },
+    {
+      "epoch": 0.10281248280037426,
+      "grad_norm": 1.3169244527816772,
+      "learning_rate": 9.941382750959203e-06,
+      "loss": 0.8601,
+      "step": 1868
+    },
+    {
+      "epoch": 0.10286752160272992,
+      "grad_norm": 0.8011140823364258,
+      "learning_rate": 9.941316553151696e-06,
+      "loss": 0.8397,
+      "step": 1869
+    },
+    {
+      "epoch": 0.10292256040508559,
+      "grad_norm": 0.811210572719574,
+      "learning_rate": 9.941250318206577e-06,
+      "loss": 0.7863,
+      "step": 1870
+    },
+    {
+      "epoch": 0.10297759920744125,
+      "grad_norm": 0.8172751665115356,
+      "learning_rate": 9.941184046124342e-06,
+      "loss": 0.8114,
+      "step": 1871
+    },
+    {
+      "epoch": 0.10303263800979691,
+      "grad_norm": 0.8072887063026428,
+      "learning_rate": 9.941117736905493e-06,
+      "loss": 0.8928,
+      "step": 1872
+    },
+    {
+      "epoch": 0.10308767681215257,
+      "grad_norm": 0.9111380577087402,
+      "learning_rate": 9.941051390550524e-06,
+      "loss": 0.866,
+      "step": 1873
+    },
+    {
+      "epoch": 0.10314271561450823,
+      "grad_norm": 0.8158383369445801,
+      "learning_rate": 9.940985007059936e-06,
+      "loss": 0.7805,
+      "step": 1874
+    },
+    {
+      "epoch": 0.1031977544168639,
+      "grad_norm": 0.8858961462974548,
+      "learning_rate": 9.940918586434226e-06,
+      "loss": 0.8424,
+      "step": 1875
+    },
+    {
+      "epoch": 0.10325279321921955,
+      "grad_norm": 0.8835182189941406,
+      "learning_rate": 9.940852128673895e-06,
+      "loss": 0.7816,
+      "step": 1876
+    },
+    {
+      "epoch": 0.10330783202157522,
+      "grad_norm": 1.044227123260498,
+      "learning_rate": 9.940785633779444e-06,
+      "loss": 0.8952,
+      "step": 1877
+    },
+    {
+      "epoch": 0.10336287082393088,
+      "grad_norm": 0.8255050778388977,
+      "learning_rate": 9.940719101751367e-06,
+      "loss": 0.8215,
+      "step": 1878
+    },
+    {
+      "epoch": 0.10341790962628654,
+      "grad_norm": 0.8561689257621765,
+      "learning_rate": 9.940652532590172e-06,
+      "loss": 0.9686,
+      "step": 1879
+    },
+    {
+      "epoch": 0.10347294842864219,
+      "grad_norm": 0.8798959255218506,
+      "learning_rate": 9.94058592629635e-06,
+      "loss": 0.8993,
+      "step": 1880
+    },
+    {
+      "epoch": 0.10352798723099785,
+      "grad_norm": 0.9292098879814148,
+      "learning_rate": 9.940519282870411e-06,
+      "loss": 0.8536,
+      "step": 1881
+    },
+    {
+      "epoch": 0.10358302603335351,
+      "grad_norm": 0.8865400552749634,
+      "learning_rate": 9.940452602312851e-06,
+      "loss": 0.8024,
+      "step": 1882
+    },
+    {
+      "epoch": 0.10363806483570917,
+      "grad_norm": 0.8985510468482971,
+      "learning_rate": 9.94038588462417e-06,
+      "loss": 0.7748,
+      "step": 1883
+    },
+    {
+      "epoch": 0.10369310363806483,
+      "grad_norm": 0.9973617196083069,
+      "learning_rate": 9.940319129804872e-06,
+      "loss": 0.875,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1037481424404205,
+      "grad_norm": 0.8615350723266602,
+      "learning_rate": 9.940252337855458e-06,
+      "loss": 0.904,
+      "step": 1885
+    },
+    {
+      "epoch": 0.10380318124277615,
+      "grad_norm": 0.8752412796020508,
+      "learning_rate": 9.940185508776429e-06,
+      "loss": 0.8735,
+      "step": 1886
+    },
+    {
+      "epoch": 0.10385822004513182,
+      "grad_norm": 0.8639446496963501,
+      "learning_rate": 9.94011864256829e-06,
+      "loss": 0.7952,
+      "step": 1887
+    },
+    {
+      "epoch": 0.10391325884748748,
+      "grad_norm": 0.7932116389274597,
+      "learning_rate": 9.94005173923154e-06,
+      "loss": 0.8721,
+      "step": 1888
+    },
+    {
+      "epoch": 0.10396829764984314,
+      "grad_norm": 0.8573791980743408,
+      "learning_rate": 9.939984798766685e-06,
+      "loss": 0.9271,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1040233364521988,
+      "grad_norm": 0.9080122113227844,
+      "learning_rate": 9.939917821174225e-06,
+      "loss": 0.8991,
+      "step": 1890
+    },
+    {
+      "epoch": 0.10407837525455446,
+      "grad_norm": 0.7883808612823486,
+      "learning_rate": 9.939850806454664e-06,
+      "loss": 0.6895,
+      "step": 1891
+    },
+    {
+      "epoch": 0.10413341405691012,
+      "grad_norm": 0.8067768216133118,
+      "learning_rate": 9.93978375460851e-06,
+      "loss": 0.835,
+      "step": 1892
+    },
+    {
+      "epoch": 0.10418845285926578,
+      "grad_norm": 0.8756459951400757,
+      "learning_rate": 9.939716665636262e-06,
+      "loss": 0.8144,
+      "step": 1893
+    },
+    {
+      "epoch": 0.10424349166162145,
+      "grad_norm": 0.8056700825691223,
+      "learning_rate": 9.939649539538425e-06,
+      "loss": 0.7454,
+      "step": 1894
+    },
+    {
+      "epoch": 0.10429853046397711,
+      "grad_norm": 1.0756300687789917,
+      "learning_rate": 9.939582376315505e-06,
+      "loss": 0.8096,
+      "step": 1895
+    },
+    {
+      "epoch": 0.10435356926633277,
+      "grad_norm": 0.8938102126121521,
+      "learning_rate": 9.939515175968006e-06,
+      "loss": 0.7496,
+      "step": 1896
+    },
+    {
+      "epoch": 0.10440860806868843,
+      "grad_norm": 0.9371656775474548,
+      "learning_rate": 9.939447938496434e-06,
+      "loss": 0.9817,
+      "step": 1897
+    },
+    {
+      "epoch": 0.10446364687104409,
+      "grad_norm": 1.0216082334518433,
+      "learning_rate": 9.939380663901292e-06,
+      "loss": 0.8804,
+      "step": 1898
+    },
+    {
+      "epoch": 0.10451868567339975,
+      "grad_norm": 0.8791126012802124,
+      "learning_rate": 9.939313352183088e-06,
+      "loss": 0.7811,
+      "step": 1899
+    },
+    {
+      "epoch": 0.10457372447575541,
+      "grad_norm": 0.9925445914268494,
+      "learning_rate": 9.939246003342326e-06,
+      "loss": 0.8892,
+      "step": 1900
+    },
+    {
+      "epoch": 0.10462876327811106,
+      "grad_norm": 1.0459916591644287,
+      "learning_rate": 9.939178617379514e-06,
+      "loss": 0.7938,
+      "step": 1901
+    },
+    {
+      "epoch": 0.10468380208046672,
+      "grad_norm": 0.9103816747665405,
+      "learning_rate": 9.93911119429516e-06,
+      "loss": 0.8282,
+      "step": 1902
+    },
+    {
+      "epoch": 0.10473884088282238,
+      "grad_norm": 0.9602296352386475,
+      "learning_rate": 9.939043734089764e-06,
+      "loss": 0.919,
+      "step": 1903
+    },
+    {
+      "epoch": 0.10479387968517805,
+      "grad_norm": 0.9529246687889099,
+      "learning_rate": 9.93897623676384e-06,
+      "loss": 0.9469,
+      "step": 1904
+    },
+    {
+      "epoch": 0.10484891848753371,
+      "grad_norm": 0.9619705080986023,
+      "learning_rate": 9.938908702317893e-06,
+      "loss": 0.9371,
+      "step": 1905
+    },
+    {
+      "epoch": 0.10490395728988937,
+      "grad_norm": 1.0106935501098633,
+      "learning_rate": 9.938841130752428e-06,
+      "loss": 0.7502,
+      "step": 1906
+    },
+    {
+      "epoch": 0.10495899609224503,
+      "grad_norm": 0.913985013961792,
+      "learning_rate": 9.938773522067957e-06,
+      "loss": 0.8172,
+      "step": 1907
+    },
+    {
+      "epoch": 0.10501403489460069,
+      "grad_norm": 0.9474983215332031,
+      "learning_rate": 9.938705876264985e-06,
+      "loss": 0.8999,
+      "step": 1908
+    },
+    {
+      "epoch": 0.10506907369695635,
+      "grad_norm": 0.9185097813606262,
+      "learning_rate": 9.938638193344024e-06,
+      "loss": 0.8976,
+      "step": 1909
+    },
+    {
+      "epoch": 0.10512411249931201,
+      "grad_norm": 0.7633675932884216,
+      "learning_rate": 9.938570473305578e-06,
+      "loss": 0.7777,
+      "step": 1910
+    },
+    {
+      "epoch": 0.10517915130166768,
+      "grad_norm": 0.9547691345214844,
+      "learning_rate": 9.938502716150159e-06,
+      "loss": 0.8154,
+      "step": 1911
+    },
+    {
+      "epoch": 0.10523419010402334,
+      "grad_norm": 0.8556191921234131,
+      "learning_rate": 9.938434921878275e-06,
+      "loss": 0.828,
+      "step": 1912
+    },
+    {
+      "epoch": 0.105289228906379,
+      "grad_norm": 0.9826140999794006,
+      "learning_rate": 9.938367090490437e-06,
+      "loss": 0.8085,
+      "step": 1913
+    },
+    {
+      "epoch": 0.10534426770873466,
+      "grad_norm": 0.8610432744026184,
+      "learning_rate": 9.938299221987154e-06,
+      "loss": 0.9103,
+      "step": 1914
+    },
+    {
+      "epoch": 0.10539930651109032,
+      "grad_norm": 0.8383543491363525,
+      "learning_rate": 9.938231316368934e-06,
+      "loss": 0.8182,
+      "step": 1915
+    },
+    {
+      "epoch": 0.10545434531344598,
+      "grad_norm": 0.8552964925765991,
+      "learning_rate": 9.93816337363629e-06,
+      "loss": 0.8024,
+      "step": 1916
+    },
+    {
+      "epoch": 0.10550938411580164,
+      "grad_norm": 0.9255730509757996,
+      "learning_rate": 9.938095393789732e-06,
+      "loss": 0.8566,
+      "step": 1917
+    },
+    {
+      "epoch": 0.1055644229181573,
+      "grad_norm": 0.9882987141609192,
+      "learning_rate": 9.938027376829774e-06,
+      "loss": 0.7119,
+      "step": 1918
+    },
+    {
+      "epoch": 0.10561946172051297,
+      "grad_norm": 1.139404535293579,
+      "learning_rate": 9.93795932275692e-06,
+      "loss": 0.8839,
+      "step": 1919
+    },
+    {
+      "epoch": 0.10567450052286863,
+      "grad_norm": 1.004782795906067,
+      "learning_rate": 9.937891231571686e-06,
+      "loss": 0.904,
+      "step": 1920
+    },
+    {
+      "epoch": 0.10572953932522429,
+      "grad_norm": 0.8437260389328003,
+      "learning_rate": 9.937823103274585e-06,
+      "loss": 0.7942,
+      "step": 1921
+    },
+    {
+      "epoch": 0.10578457812757995,
+      "grad_norm": 1.1388722658157349,
+      "learning_rate": 9.937754937866127e-06,
+      "loss": 0.9491,
+      "step": 1922
+    },
+    {
+      "epoch": 0.1058396169299356,
+      "grad_norm": 0.9266740083694458,
+      "learning_rate": 9.937686735346823e-06,
+      "loss": 0.9067,
+      "step": 1923
+    },
+    {
+      "epoch": 0.10589465573229126,
+      "grad_norm": 0.7536123991012573,
+      "learning_rate": 9.93761849571719e-06,
+      "loss": 0.6533,
+      "step": 1924
+    },
+    {
+      "epoch": 0.10594969453464692,
+      "grad_norm": 0.8781737089157104,
+      "learning_rate": 9.937550218977737e-06,
+      "loss": 0.8319,
+      "step": 1925
+    },
+    {
+      "epoch": 0.10600473333700258,
+      "grad_norm": 0.8577924966812134,
+      "learning_rate": 9.937481905128976e-06,
+      "loss": 0.8604,
+      "step": 1926
+    },
+    {
+      "epoch": 0.10605977213935824,
+      "grad_norm": 0.8351713418960571,
+      "learning_rate": 9.937413554171424e-06,
+      "loss": 0.946,
+      "step": 1927
+    },
+    {
+      "epoch": 0.1061148109417139,
+      "grad_norm": 0.971491813659668,
+      "learning_rate": 9.937345166105594e-06,
+      "loss": 0.7383,
+      "step": 1928
+    },
+    {
+      "epoch": 0.10616984974406957,
+      "grad_norm": 0.8020079731941223,
+      "learning_rate": 9.937276740932001e-06,
+      "loss": 0.7468,
+      "step": 1929
+    },
+    {
+      "epoch": 0.10622488854642523,
+      "grad_norm": 0.9057347178459167,
+      "learning_rate": 9.937208278651153e-06,
+      "loss": 0.8223,
+      "step": 1930
+    },
+    {
+      "epoch": 0.10627992734878089,
+      "grad_norm": 0.8384734392166138,
+      "learning_rate": 9.937139779263574e-06,
+      "loss": 0.8773,
+      "step": 1931
+    },
+    {
+      "epoch": 0.10633496615113655,
+      "grad_norm": 0.8732065558433533,
+      "learning_rate": 9.93707124276977e-06,
+      "loss": 0.8265,
+      "step": 1932
+    },
+    {
+      "epoch": 0.10639000495349221,
+      "grad_norm": 0.8744868040084839,
+      "learning_rate": 9.937002669170264e-06,
+      "loss": 0.8497,
+      "step": 1933
+    },
+    {
+      "epoch": 0.10644504375584787,
+      "grad_norm": 0.8589879870414734,
+      "learning_rate": 9.936934058465564e-06,
+      "loss": 0.8116,
+      "step": 1934
+    },
+    {
+      "epoch": 0.10650008255820353,
+      "grad_norm": 0.8614563941955566,
+      "learning_rate": 9.936865410656192e-06,
+      "loss": 0.7823,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1065551213605592,
+      "grad_norm": 0.8381434082984924,
+      "learning_rate": 9.93679672574266e-06,
+      "loss": 0.7889,
+      "step": 1936
+    },
+    {
+      "epoch": 0.10661016016291486,
+      "grad_norm": 0.9834293127059937,
+      "learning_rate": 9.936728003725484e-06,
+      "loss": 0.8358,
+      "step": 1937
+    },
+    {
+      "epoch": 0.10666519896527052,
+      "grad_norm": 0.8461851477622986,
+      "learning_rate": 9.936659244605184e-06,
+      "loss": 0.8408,
+      "step": 1938
+    },
+    {
+      "epoch": 0.10672023776762618,
+      "grad_norm": 1.0186371803283691,
+      "learning_rate": 9.936590448382273e-06,
+      "loss": 0.8118,
+      "step": 1939
+    },
+    {
+      "epoch": 0.10677527656998184,
+      "grad_norm": 0.866321325302124,
+      "learning_rate": 9.93652161505727e-06,
+      "loss": 0.8696,
+      "step": 1940
+    },
+    {
+      "epoch": 0.1068303153723375,
+      "grad_norm": 0.9179622530937195,
+      "learning_rate": 9.936452744630692e-06,
+      "loss": 0.8419,
+      "step": 1941
+    },
+    {
+      "epoch": 0.10688535417469316,
+      "grad_norm": 0.8250496983528137,
+      "learning_rate": 9.936383837103057e-06,
+      "loss": 0.8511,
+      "step": 1942
+    },
+    {
+      "epoch": 0.10694039297704883,
+      "grad_norm": 0.8475700616836548,
+      "learning_rate": 9.936314892474883e-06,
+      "loss": 0.8404,
+      "step": 1943
+    },
+    {
+      "epoch": 0.10699543177940447,
+      "grad_norm": 0.774334192276001,
+      "learning_rate": 9.936245910746684e-06,
+      "loss": 0.7461,
+      "step": 1944
+    },
+    {
+      "epoch": 0.10705047058176013,
+      "grad_norm": 0.9313948154449463,
+      "learning_rate": 9.936176891918986e-06,
+      "loss": 0.8486,
+      "step": 1945
+    },
+    {
+      "epoch": 0.1071055093841158,
+      "grad_norm": 0.8784124255180359,
+      "learning_rate": 9.936107835992304e-06,
+      "loss": 0.84,
+      "step": 1946
+    },
+    {
+      "epoch": 0.10716054818647146,
+      "grad_norm": 0.9087465405464172,
+      "learning_rate": 9.936038742967154e-06,
+      "loss": 0.9012,
+      "step": 1947
+    },
+    {
+      "epoch": 0.10721558698882712,
+      "grad_norm": 0.8462012410163879,
+      "learning_rate": 9.93596961284406e-06,
+      "loss": 0.9193,
+      "step": 1948
+    },
+    {
+      "epoch": 0.10727062579118278,
+      "grad_norm": 0.8984553813934326,
+      "learning_rate": 9.935900445623538e-06,
+      "loss": 0.781,
+      "step": 1949
+    },
+    {
+      "epoch": 0.10732566459353844,
+      "grad_norm": 0.9197295308113098,
+      "learning_rate": 9.935831241306111e-06,
+      "loss": 0.8861,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1073807033958941,
+      "grad_norm": 0.8452801704406738,
+      "learning_rate": 9.935761999892296e-06,
+      "loss": 0.8649,
+      "step": 1951
+    },
+    {
+      "epoch": 0.10743574219824976,
+      "grad_norm": 0.8047192096710205,
+      "learning_rate": 9.935692721382618e-06,
+      "loss": 0.8704,
+      "step": 1952
+    },
+    {
+      "epoch": 0.10749078100060543,
+      "grad_norm": 0.9536359906196594,
+      "learning_rate": 9.935623405777593e-06,
+      "loss": 0.7803,
+      "step": 1953
+    },
+    {
+      "epoch": 0.10754581980296109,
+      "grad_norm": 0.8215291500091553,
+      "learning_rate": 9.935554053077744e-06,
+      "loss": 0.8247,
+      "step": 1954
+    },
+    {
+      "epoch": 0.10760085860531675,
+      "grad_norm": 0.9261930584907532,
+      "learning_rate": 9.93548466328359e-06,
+      "loss": 0.8594,
+      "step": 1955
+    },
+    {
+      "epoch": 0.10765589740767241,
+      "grad_norm": 0.7973492741584778,
+      "learning_rate": 9.935415236395656e-06,
+      "loss": 0.7464,
+      "step": 1956
+    },
+    {
+      "epoch": 0.10771093621002807,
+      "grad_norm": 0.9328988790512085,
+      "learning_rate": 9.935345772414463e-06,
+      "loss": 0.8472,
+      "step": 1957
+    },
+    {
+      "epoch": 0.10776597501238373,
+      "grad_norm": 0.9490759968757629,
+      "learning_rate": 9.935276271340532e-06,
+      "loss": 0.806,
+      "step": 1958
+    },
+    {
+      "epoch": 0.1078210138147394,
+      "grad_norm": 0.9149925112724304,
+      "learning_rate": 9.935206733174385e-06,
+      "loss": 0.8741,
+      "step": 1959
+    },
+    {
+      "epoch": 0.10787605261709506,
+      "grad_norm": 1.0074039697647095,
+      "learning_rate": 9.935137157916546e-06,
+      "loss": 0.8493,
+      "step": 1960
+    },
+    {
+      "epoch": 0.10793109141945072,
+      "grad_norm": 0.8783678412437439,
+      "learning_rate": 9.935067545567535e-06,
+      "loss": 0.8132,
+      "step": 1961
+    },
+    {
+      "epoch": 0.10798613022180638,
+      "grad_norm": 0.8273885250091553,
+      "learning_rate": 9.934997896127879e-06,
+      "loss": 0.7448,
+      "step": 1962
+    },
+    {
+      "epoch": 0.10804116902416204,
+      "grad_norm": 0.761947512626648,
+      "learning_rate": 9.9349282095981e-06,
+      "loss": 0.7933,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1080962078265177,
+      "grad_norm": 0.814809262752533,
+      "learning_rate": 9.934858485978722e-06,
+      "loss": 0.7551,
+      "step": 1964
+    },
+    {
+      "epoch": 0.10815124662887336,
+      "grad_norm": 0.8108895421028137,
+      "learning_rate": 9.934788725270266e-06,
+      "loss": 0.6787,
+      "step": 1965
+    },
+    {
+      "epoch": 0.10820628543122901,
+      "grad_norm": 0.8669139742851257,
+      "learning_rate": 9.934718927473262e-06,
+      "loss": 0.8395,
+      "step": 1966
+    },
+    {
+      "epoch": 0.10826132423358467,
+      "grad_norm": 0.9093756079673767,
+      "learning_rate": 9.93464909258823e-06,
+      "loss": 0.8341,
+      "step": 1967
+    },
+    {
+      "epoch": 0.10831636303594033,
+      "grad_norm": 0.8923841714859009,
+      "learning_rate": 9.934579220615697e-06,
+      "loss": 0.9422,
+      "step": 1968
+    },
+    {
+      "epoch": 0.108371401838296,
+      "grad_norm": 0.850429117679596,
+      "learning_rate": 9.934509311556186e-06,
+      "loss": 0.8446,
+      "step": 1969
+    },
+    {
+      "epoch": 0.10842644064065166,
+      "grad_norm": 0.8762460350990295,
+      "learning_rate": 9.934439365410224e-06,
+      "loss": 0.7788,
+      "step": 1970
+    },
+    {
+      "epoch": 0.10848147944300732,
+      "grad_norm": 0.9700387716293335,
+      "learning_rate": 9.934369382178338e-06,
+      "loss": 0.8455,
+      "step": 1971
+    },
+    {
+      "epoch": 0.10853651824536298,
+      "grad_norm": 0.8003185987472534,
+      "learning_rate": 9.934299361861053e-06,
+      "loss": 0.8026,
+      "step": 1972
+    },
+    {
+      "epoch": 0.10859155704771864,
+      "grad_norm": 0.9626984596252441,
+      "learning_rate": 9.934229304458893e-06,
+      "loss": 0.8219,
+      "step": 1973
+    },
+    {
+      "epoch": 0.1086465958500743,
+      "grad_norm": 0.8722280859947205,
+      "learning_rate": 9.934159209972386e-06,
+      "loss": 0.8866,
+      "step": 1974
+    },
+    {
+      "epoch": 0.10870163465242996,
+      "grad_norm": 0.838736355304718,
+      "learning_rate": 9.934089078402061e-06,
+      "loss": 0.7723,
+      "step": 1975
+    },
+    {
+      "epoch": 0.10875667345478562,
+      "grad_norm": 0.8373032808303833,
+      "learning_rate": 9.934018909748443e-06,
+      "loss": 0.9003,
+      "step": 1976
+    },
+    {
+      "epoch": 0.10881171225714129,
+      "grad_norm": 0.8704653978347778,
+      "learning_rate": 9.93394870401206e-06,
+      "loss": 0.8926,
+      "step": 1977
+    },
+    {
+      "epoch": 0.10886675105949695,
+      "grad_norm": 0.8088163733482361,
+      "learning_rate": 9.933878461193437e-06,
+      "loss": 0.8059,
+      "step": 1978
+    },
+    {
+      "epoch": 0.10892178986185261,
+      "grad_norm": 0.856421947479248,
+      "learning_rate": 9.933808181293108e-06,
+      "loss": 0.8447,
+      "step": 1979
+    },
+    {
+      "epoch": 0.10897682866420827,
+      "grad_norm": 0.9676237106323242,
+      "learning_rate": 9.933737864311595e-06,
+      "loss": 0.9009,
+      "step": 1980
+    },
+    {
+      "epoch": 0.10903186746656393,
+      "grad_norm": 0.7955103516578674,
+      "learning_rate": 9.933667510249428e-06,
+      "loss": 0.881,
+      "step": 1981
+    },
+    {
+      "epoch": 0.10908690626891959,
+      "grad_norm": 0.7935854196548462,
+      "learning_rate": 9.933597119107136e-06,
+      "loss": 0.8773,
+      "step": 1982
+    },
+    {
+      "epoch": 0.10914194507127525,
+      "grad_norm": 0.7726008296012878,
+      "learning_rate": 9.933526690885251e-06,
+      "loss": 0.8133,
+      "step": 1983
+    },
+    {
+      "epoch": 0.10919698387363092,
+      "grad_norm": 0.8577712178230286,
+      "learning_rate": 9.9334562255843e-06,
+      "loss": 0.7455,
+      "step": 1984
+    },
+    {
+      "epoch": 0.10925202267598658,
+      "grad_norm": 0.9996447563171387,
+      "learning_rate": 9.933385723204812e-06,
+      "loss": 0.7312,
+      "step": 1985
+    },
+    {
+      "epoch": 0.10930706147834224,
+      "grad_norm": 0.9600629806518555,
+      "learning_rate": 9.933315183747318e-06,
+      "loss": 0.8792,
+      "step": 1986
+    },
+    {
+      "epoch": 0.10936210028069789,
+      "grad_norm": 0.9126206636428833,
+      "learning_rate": 9.933244607212347e-06,
+      "loss": 1.0023,
+      "step": 1987
+    },
+    {
+      "epoch": 0.10941713908305355,
+      "grad_norm": 0.774153470993042,
+      "learning_rate": 9.93317399360043e-06,
+      "loss": 0.7877,
+      "step": 1988
+    },
+    {
+      "epoch": 0.10947217788540921,
+      "grad_norm": 0.848495364189148,
+      "learning_rate": 9.933103342912096e-06,
+      "loss": 0.8825,
+      "step": 1989
+    },
+    {
+      "epoch": 0.10952721668776487,
+      "grad_norm": 0.806408166885376,
+      "learning_rate": 9.933032655147881e-06,
+      "loss": 0.7389,
+      "step": 1990
+    },
+    {
+      "epoch": 0.10958225549012053,
+      "grad_norm": 0.8579222559928894,
+      "learning_rate": 9.932961930308312e-06,
+      "loss": 0.8283,
+      "step": 1991
+    },
+    {
+      "epoch": 0.10963729429247619,
+      "grad_norm": 0.7548109292984009,
+      "learning_rate": 9.93289116839392e-06,
+      "loss": 0.7971,
+      "step": 1992
+    },
+    {
+      "epoch": 0.10969233309483185,
+      "grad_norm": 0.7954711318016052,
+      "learning_rate": 9.93282036940524e-06,
+      "loss": 0.849,
+      "step": 1993
+    },
+    {
+      "epoch": 0.10974737189718752,
+      "grad_norm": 0.7911425232887268,
+      "learning_rate": 9.932749533342802e-06,
+      "loss": 0.86,
+      "step": 1994
+    },
+    {
+      "epoch": 0.10980241069954318,
+      "grad_norm": 0.8505094051361084,
+      "learning_rate": 9.932678660207141e-06,
+      "loss": 0.7871,
+      "step": 1995
+    },
+    {
+      "epoch": 0.10985744950189884,
+      "grad_norm": 0.809612512588501,
+      "learning_rate": 9.932607749998784e-06,
+      "loss": 0.8337,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1099124883042545,
+      "grad_norm": 0.738523006439209,
+      "learning_rate": 9.93253680271827e-06,
+      "loss": 0.7634,
+      "step": 1997
+    },
+    {
+      "epoch": 0.10996752710661016,
+      "grad_norm": 0.8434372544288635,
+      "learning_rate": 9.932465818366128e-06,
+      "loss": 0.7987,
+      "step": 1998
+    },
+    {
+      "epoch": 0.11002256590896582,
+      "grad_norm": 0.8068081140518188,
+      "learning_rate": 9.932394796942895e-06,
+      "loss": 0.9496,
+      "step": 1999
+    },
+    {
+      "epoch": 0.11007760471132148,
+      "grad_norm": 0.754342794418335,
+      "learning_rate": 9.932323738449103e-06,
+      "loss": 0.7355,
+      "step": 2000
+    },
+    {
+      "epoch": 0.11013264351367714,
+      "grad_norm": 0.8830806612968445,
+      "learning_rate": 9.932252642885285e-06,
+      "loss": 0.8458,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1101876823160328,
+      "grad_norm": 0.9915485978126526,
+      "learning_rate": 9.932181510251977e-06,
+      "loss": 0.8116,
+      "step": 2002
+    },
+    {
+      "epoch": 0.11024272111838847,
+      "grad_norm": 0.858368992805481,
+      "learning_rate": 9.932110340549712e-06,
+      "loss": 0.8354,
+      "step": 2003
+    },
+    {
+      "epoch": 0.11029775992074413,
+      "grad_norm": 0.8591521382331848,
+      "learning_rate": 9.932039133779028e-06,
+      "loss": 0.8316,
+      "step": 2004
+    },
+    {
+      "epoch": 0.11035279872309979,
+      "grad_norm": 0.8714838624000549,
+      "learning_rate": 9.931967889940455e-06,
+      "loss": 0.8106,
+      "step": 2005
+    },
+    {
+      "epoch": 0.11040783752545545,
+      "grad_norm": 0.8082797527313232,
+      "learning_rate": 9.931896609034534e-06,
+      "loss": 0.7762,
+      "step": 2006
+    },
+    {
+      "epoch": 0.11046287632781111,
+      "grad_norm": 0.9226199984550476,
+      "learning_rate": 9.931825291061797e-06,
+      "loss": 0.8641,
+      "step": 2007
+    },
+    {
+      "epoch": 0.11051791513016677,
+      "grad_norm": 0.8883050680160522,
+      "learning_rate": 9.931753936022783e-06,
+      "loss": 0.9014,
+      "step": 2008
+    },
+    {
+      "epoch": 0.11057295393252242,
+      "grad_norm": 0.9024807810783386,
+      "learning_rate": 9.931682543918024e-06,
+      "loss": 0.9085,
+      "step": 2009
+    },
+    {
+      "epoch": 0.11062799273487808,
+      "grad_norm": 0.8381460905075073,
+      "learning_rate": 9.931611114748062e-06,
+      "loss": 0.8043,
+      "step": 2010
+    },
+    {
+      "epoch": 0.11068303153723374,
+      "grad_norm": 1.1222339868545532,
+      "learning_rate": 9.931539648513429e-06,
+      "loss": 0.8388,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1107380703395894,
+      "grad_norm": 0.9710868000984192,
+      "learning_rate": 9.931468145214665e-06,
+      "loss": 0.8934,
+      "step": 2012
+    },
+    {
+      "epoch": 0.11079310914194507,
+      "grad_norm": 0.9821141958236694,
+      "learning_rate": 9.931396604852304e-06,
+      "loss": 0.931,
+      "step": 2013
+    },
+    {
+      "epoch": 0.11084814794430073,
+      "grad_norm": 1.0658717155456543,
+      "learning_rate": 9.931325027426889e-06,
+      "loss": 0.9032,
+      "step": 2014
+    },
+    {
+      "epoch": 0.11090318674665639,
+      "grad_norm": 0.8836946487426758,
+      "learning_rate": 9.931253412938956e-06,
+      "loss": 0.9131,
+      "step": 2015
+    },
+    {
+      "epoch": 0.11095822554901205,
+      "grad_norm": 0.8438361883163452,
+      "learning_rate": 9.93118176138904e-06,
+      "loss": 0.8674,
+      "step": 2016
+    },
+    {
+      "epoch": 0.11101326435136771,
+      "grad_norm": 0.928142786026001,
+      "learning_rate": 9.93111007277768e-06,
+      "loss": 0.8882,
+      "step": 2017
+    },
+    {
+      "epoch": 0.11106830315372337,
+      "grad_norm": 0.9176276922225952,
+      "learning_rate": 9.93103834710542e-06,
+      "loss": 0.8904,
+      "step": 2018
+    },
+    {
+      "epoch": 0.11112334195607904,
+      "grad_norm": 1.0462889671325684,
+      "learning_rate": 9.930966584372795e-06,
+      "loss": 0.8029,
+      "step": 2019
+    },
+    {
+      "epoch": 0.1111783807584347,
+      "grad_norm": 0.7627375721931458,
+      "learning_rate": 9.930894784580344e-06,
+      "loss": 0.8474,
+      "step": 2020
+    },
+    {
+      "epoch": 0.11123341956079036,
+      "grad_norm": 1.0545588731765747,
+      "learning_rate": 9.93082294772861e-06,
+      "loss": 0.7985,
+      "step": 2021
+    },
+    {
+      "epoch": 0.11128845836314602,
+      "grad_norm": 0.9752298593521118,
+      "learning_rate": 9.93075107381813e-06,
+      "loss": 0.8725,
+      "step": 2022
+    },
+    {
+      "epoch": 0.11134349716550168,
+      "grad_norm": 0.8403159379959106,
+      "learning_rate": 9.930679162849444e-06,
+      "loss": 0.8854,
+      "step": 2023
+    },
+    {
+      "epoch": 0.11139853596785734,
+      "grad_norm": 0.8879380226135254,
+      "learning_rate": 9.930607214823094e-06,
+      "loss": 0.7269,
+      "step": 2024
+    },
+    {
+      "epoch": 0.111453574770213,
+      "grad_norm": 0.907256543636322,
+      "learning_rate": 9.930535229739618e-06,
+      "loss": 0.8145,
+      "step": 2025
+    },
+    {
+      "epoch": 0.11150861357256867,
+      "grad_norm": 1.1066968441009521,
+      "learning_rate": 9.93046320759956e-06,
+      "loss": 0.9281,
+      "step": 2026
+    },
+    {
+      "epoch": 0.11156365237492433,
+      "grad_norm": 0.9226258397102356,
+      "learning_rate": 9.930391148403462e-06,
+      "loss": 0.9048,
+      "step": 2027
+    },
+    {
+      "epoch": 0.11161869117727999,
+      "grad_norm": 0.9652156829833984,
+      "learning_rate": 9.930319052151862e-06,
+      "loss": 0.9321,
+      "step": 2028
+    },
+    {
+      "epoch": 0.11167372997963565,
+      "grad_norm": 0.9102638363838196,
+      "learning_rate": 9.930246918845305e-06,
+      "loss": 0.8169,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1117287687819913,
+      "grad_norm": 0.7765716314315796,
+      "learning_rate": 9.93017474848433e-06,
+      "loss": 0.7691,
+      "step": 2030
+    },
+    {
+      "epoch": 0.11178380758434696,
+      "grad_norm": 0.9053775072097778,
+      "learning_rate": 9.930102541069484e-06,
+      "loss": 0.782,
+      "step": 2031
+    },
+    {
+      "epoch": 0.11183884638670262,
+      "grad_norm": 0.8892827033996582,
+      "learning_rate": 9.930030296601306e-06,
+      "loss": 0.8575,
+      "step": 2032
+    },
+    {
+      "epoch": 0.11189388518905828,
+      "grad_norm": 0.8947604894638062,
+      "learning_rate": 9.929958015080339e-06,
+      "loss": 0.8607,
+      "step": 2033
+    },
+    {
+      "epoch": 0.11194892399141394,
+      "grad_norm": 0.8936871290206909,
+      "learning_rate": 9.929885696507127e-06,
+      "loss": 0.8111,
+      "step": 2034
+    },
+    {
+      "epoch": 0.1120039627937696,
+      "grad_norm": 0.9579165577888489,
+      "learning_rate": 9.929813340882214e-06,
+      "loss": 0.911,
+      "step": 2035
+    },
+    {
+      "epoch": 0.11205900159612527,
+      "grad_norm": 0.7885386347770691,
+      "learning_rate": 9.929740948206146e-06,
+      "loss": 0.8074,
+      "step": 2036
+    },
+    {
+      "epoch": 0.11211404039848093,
+      "grad_norm": 0.817939281463623,
+      "learning_rate": 9.929668518479462e-06,
+      "loss": 0.8451,
+      "step": 2037
+    },
+    {
+      "epoch": 0.11216907920083659,
+      "grad_norm": 0.8695761561393738,
+      "learning_rate": 9.92959605170271e-06,
+      "loss": 0.7158,
+      "step": 2038
+    },
+    {
+      "epoch": 0.11222411800319225,
+      "grad_norm": 0.8569639325141907,
+      "learning_rate": 9.929523547876433e-06,
+      "loss": 0.8568,
+      "step": 2039
+    },
+    {
+      "epoch": 0.11227915680554791,
+      "grad_norm": 0.8569897413253784,
+      "learning_rate": 9.929451007001176e-06,
+      "loss": 0.8971,
+      "step": 2040
+    },
+    {
+      "epoch": 0.11233419560790357,
+      "grad_norm": 0.8520069718360901,
+      "learning_rate": 9.929378429077487e-06,
+      "loss": 0.9027,
+      "step": 2041
+    },
+    {
+      "epoch": 0.11238923441025923,
+      "grad_norm": 0.9338961839675903,
+      "learning_rate": 9.929305814105907e-06,
+      "loss": 0.8646,
+      "step": 2042
+    },
+    {
+      "epoch": 0.1124442732126149,
+      "grad_norm": 0.8497192859649658,
+      "learning_rate": 9.929233162086985e-06,
+      "loss": 0.9068,
+      "step": 2043
+    },
+    {
+      "epoch": 0.11249931201497056,
+      "grad_norm": 0.8570863008499146,
+      "learning_rate": 9.929160473021267e-06,
+      "loss": 0.962,
+      "step": 2044
+    },
+    {
+      "epoch": 0.11255435081732622,
+      "grad_norm": 0.9072359800338745,
+      "learning_rate": 9.929087746909296e-06,
+      "loss": 0.8454,
+      "step": 2045
+    },
+    {
+      "epoch": 0.11260938961968188,
+      "grad_norm": 0.7920698523521423,
+      "learning_rate": 9.929014983751623e-06,
+      "loss": 0.8031,
+      "step": 2046
+    },
+    {
+      "epoch": 0.11266442842203754,
+      "grad_norm": 1.0180169343948364,
+      "learning_rate": 9.928942183548791e-06,
+      "loss": 0.7759,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1127194672243932,
+      "grad_norm": 0.8746892809867859,
+      "learning_rate": 9.928869346301351e-06,
+      "loss": 0.9038,
+      "step": 2048
+    },
+    {
+      "epoch": 0.11277450602674886,
+      "grad_norm": 0.8283438086509705,
+      "learning_rate": 9.928796472009846e-06,
+      "loss": 0.8883,
+      "step": 2049
+    },
+    {
+      "epoch": 0.11282954482910452,
+      "grad_norm": 1.321917176246643,
+      "learning_rate": 9.928723560674828e-06,
+      "loss": 0.835,
+      "step": 2050
+    },
+    {
+      "epoch": 0.11288458363146017,
+      "grad_norm": 0.9356202483177185,
+      "learning_rate": 9.928650612296841e-06,
+      "loss": 0.8077,
+      "step": 2051
+    },
+    {
+      "epoch": 0.11293962243381583,
+      "grad_norm": 0.8493767380714417,
+      "learning_rate": 9.928577626876439e-06,
+      "loss": 0.8295,
+      "step": 2052
+    },
+    {
+      "epoch": 0.1129946612361715,
+      "grad_norm": 0.784818708896637,
+      "learning_rate": 9.928504604414164e-06,
+      "loss": 0.8322,
+      "step": 2053
+    },
+    {
+      "epoch": 0.11304970003852716,
+      "grad_norm": 0.9095364809036255,
+      "learning_rate": 9.928431544910567e-06,
+      "loss": 0.8757,
+      "step": 2054
+    },
+    {
+      "epoch": 0.11310473884088282,
+      "grad_norm": 0.8889689445495605,
+      "learning_rate": 9.9283584483662e-06,
+      "loss": 0.8583,
+      "step": 2055
+    },
+    {
+      "epoch": 0.11315977764323848,
+      "grad_norm": 0.8702652454376221,
+      "learning_rate": 9.928285314781607e-06,
+      "loss": 0.8414,
+      "step": 2056
+    },
+    {
+      "epoch": 0.11321481644559414,
+      "grad_norm": 0.8531168699264526,
+      "learning_rate": 9.928212144157342e-06,
+      "loss": 0.7844,
+      "step": 2057
+    },
+    {
+      "epoch": 0.1132698552479498,
+      "grad_norm": 1.0250271558761597,
+      "learning_rate": 9.928138936493956e-06,
+      "loss": 0.8766,
+      "step": 2058
+    },
+    {
+      "epoch": 0.11332489405030546,
+      "grad_norm": 0.7963449358940125,
+      "learning_rate": 9.928065691791996e-06,
+      "loss": 0.8166,
+      "step": 2059
+    },
+    {
+      "epoch": 0.11337993285266112,
+      "grad_norm": 1.1033011674880981,
+      "learning_rate": 9.927992410052013e-06,
+      "loss": 0.8748,
+      "step": 2060
+    },
+    {
+      "epoch": 0.11343497165501679,
+      "grad_norm": 0.8760959506034851,
+      "learning_rate": 9.927919091274558e-06,
+      "loss": 0.8623,
+      "step": 2061
+    },
+    {
+      "epoch": 0.11349001045737245,
+      "grad_norm": 1.1783028841018677,
+      "learning_rate": 9.927845735460182e-06,
+      "loss": 0.9144,
+      "step": 2062
+    },
+    {
+      "epoch": 0.11354504925972811,
+      "grad_norm": 0.8868625164031982,
+      "learning_rate": 9.927772342609437e-06,
+      "loss": 0.8614,
+      "step": 2063
+    },
+    {
+      "epoch": 0.11360008806208377,
+      "grad_norm": 0.8784704804420471,
+      "learning_rate": 9.927698912722874e-06,
+      "loss": 0.7802,
+      "step": 2064
+    },
+    {
+      "epoch": 0.11365512686443943,
+      "grad_norm": 1.0090643167495728,
+      "learning_rate": 9.927625445801046e-06,
+      "loss": 0.8876,
+      "step": 2065
+    },
+    {
+      "epoch": 0.1137101656667951,
+      "grad_norm": 0.7624390721321106,
+      "learning_rate": 9.927551941844502e-06,
+      "loss": 0.794,
+      "step": 2066
+    },
+    {
+      "epoch": 0.11376520446915075,
+      "grad_norm": 0.7814189791679382,
+      "learning_rate": 9.927478400853798e-06,
+      "loss": 0.8176,
+      "step": 2067
+    },
+    {
+      "epoch": 0.11382024327150642,
+      "grad_norm": 0.876338541507721,
+      "learning_rate": 9.927404822829486e-06,
+      "loss": 0.8634,
+      "step": 2068
+    },
+    {
+      "epoch": 0.11387528207386208,
+      "grad_norm": 0.7931430339813232,
+      "learning_rate": 9.927331207772117e-06,
+      "loss": 0.8012,
+      "step": 2069
+    },
+    {
+      "epoch": 0.11393032087621774,
+      "grad_norm": 1.0064504146575928,
+      "learning_rate": 9.927257555682246e-06,
+      "loss": 0.8321,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1139853596785734,
+      "grad_norm": 0.8233053684234619,
+      "learning_rate": 9.927183866560425e-06,
+      "loss": 0.8004,
+      "step": 2071
+    },
+    {
+      "epoch": 0.11404039848092906,
+      "grad_norm": 1.0106632709503174,
+      "learning_rate": 9.927110140407211e-06,
+      "loss": 0.8627,
+      "step": 2072
+    },
+    {
+      "epoch": 0.11409543728328471,
+      "grad_norm": 0.8262843489646912,
+      "learning_rate": 9.927036377223155e-06,
+      "loss": 0.737,
+      "step": 2073
+    },
+    {
+      "epoch": 0.11415047608564037,
+      "grad_norm": 0.9349029660224915,
+      "learning_rate": 9.926962577008813e-06,
+      "loss": 0.9049,
+      "step": 2074
+    },
+    {
+      "epoch": 0.11420551488799603,
+      "grad_norm": 0.8689929842948914,
+      "learning_rate": 9.926888739764739e-06,
+      "loss": 0.7858,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1142605536903517,
+      "grad_norm": 0.8442347645759583,
+      "learning_rate": 9.926814865491487e-06,
+      "loss": 0.8145,
+      "step": 2076
+    },
+    {
+      "epoch": 0.11431559249270735,
+      "grad_norm": 0.9143397212028503,
+      "learning_rate": 9.926740954189615e-06,
+      "loss": 0.8025,
+      "step": 2077
+    },
+    {
+      "epoch": 0.11437063129506302,
+      "grad_norm": 1.293251395225525,
+      "learning_rate": 9.926667005859676e-06,
+      "loss": 1.0256,
+      "step": 2078
+    },
+    {
+      "epoch": 0.11442567009741868,
+      "grad_norm": 0.9661351442337036,
+      "learning_rate": 9.926593020502226e-06,
+      "loss": 0.991,
+      "step": 2079
+    },
+    {
+      "epoch": 0.11448070889977434,
+      "grad_norm": 0.8110861778259277,
+      "learning_rate": 9.926518998117823e-06,
+      "loss": 0.7129,
+      "step": 2080
+    },
+    {
+      "epoch": 0.11453574770213,
+      "grad_norm": 0.8351119160652161,
+      "learning_rate": 9.92644493870702e-06,
+      "loss": 0.8894,
+      "step": 2081
+    },
+    {
+      "epoch": 0.11459078650448566,
+      "grad_norm": 0.8492733240127563,
+      "learning_rate": 9.926370842270377e-06,
+      "loss": 0.8039,
+      "step": 2082
+    },
+    {
+      "epoch": 0.11464582530684132,
+      "grad_norm": 0.895353376865387,
+      "learning_rate": 9.92629670880845e-06,
+      "loss": 0.8743,
+      "step": 2083
+    },
+    {
+      "epoch": 0.11470086410919698,
+      "grad_norm": 0.7871271967887878,
+      "learning_rate": 9.926222538321795e-06,
+      "loss": 0.8426,
+      "step": 2084
+    },
+    {
+      "epoch": 0.11475590291155265,
+      "grad_norm": 0.8904643058776855,
+      "learning_rate": 9.92614833081097e-06,
+      "loss": 0.8454,
+      "step": 2085
+    },
+    {
+      "epoch": 0.11481094171390831,
+      "grad_norm": 0.9166308641433716,
+      "learning_rate": 9.926074086276532e-06,
+      "loss": 0.9162,
+      "step": 2086
+    },
+    {
+      "epoch": 0.11486598051626397,
+      "grad_norm": 0.8730728626251221,
+      "learning_rate": 9.92599980471904e-06,
+      "loss": 0.8524,
+      "step": 2087
+    },
+    {
+      "epoch": 0.11492101931861963,
+      "grad_norm": 0.7932829260826111,
+      "learning_rate": 9.925925486139052e-06,
+      "loss": 0.7838,
+      "step": 2088
+    },
+    {
+      "epoch": 0.11497605812097529,
+      "grad_norm": 1.0033760070800781,
+      "learning_rate": 9.925851130537127e-06,
+      "loss": 0.8746,
+      "step": 2089
+    },
+    {
+      "epoch": 0.11503109692333095,
+      "grad_norm": 0.7783192992210388,
+      "learning_rate": 9.925776737913823e-06,
+      "loss": 0.7308,
+      "step": 2090
+    },
+    {
+      "epoch": 0.11508613572568661,
+      "grad_norm": 0.8441587686538696,
+      "learning_rate": 9.925702308269702e-06,
+      "loss": 0.7933,
+      "step": 2091
+    },
+    {
+      "epoch": 0.11514117452804228,
+      "grad_norm": 0.9433023929595947,
+      "learning_rate": 9.925627841605319e-06,
+      "loss": 0.7857,
+      "step": 2092
+    },
+    {
+      "epoch": 0.11519621333039794,
+      "grad_norm": 0.8958256244659424,
+      "learning_rate": 9.925553337921235e-06,
+      "loss": 0.9116,
+      "step": 2093
+    },
+    {
+      "epoch": 0.11525125213275358,
+      "grad_norm": 0.7610845565795898,
+      "learning_rate": 9.925478797218011e-06,
+      "loss": 0.8006,
+      "step": 2094
+    },
+    {
+      "epoch": 0.11530629093510925,
+      "grad_norm": 0.7977023720741272,
+      "learning_rate": 9.925404219496207e-06,
+      "loss": 0.8068,
+      "step": 2095
+    },
+    {
+      "epoch": 0.11536132973746491,
+      "grad_norm": 0.8087283372879028,
+      "learning_rate": 9.925329604756383e-06,
+      "loss": 0.7968,
+      "step": 2096
+    },
+    {
+      "epoch": 0.11541636853982057,
+      "grad_norm": 1.1066477298736572,
+      "learning_rate": 9.925254952999102e-06,
+      "loss": 0.8167,
+      "step": 2097
+    },
+    {
+      "epoch": 0.11547140734217623,
+      "grad_norm": 0.7806832194328308,
+      "learning_rate": 9.925180264224921e-06,
+      "loss": 0.8069,
+      "step": 2098
+    },
+    {
+      "epoch": 0.11552644614453189,
+      "grad_norm": 0.7745190858840942,
+      "learning_rate": 9.925105538434406e-06,
+      "loss": 0.7968,
+      "step": 2099
+    },
+    {
+      "epoch": 0.11558148494688755,
+      "grad_norm": 0.9045543074607849,
+      "learning_rate": 9.925030775628113e-06,
+      "loss": 0.8417,
+      "step": 2100
+    },
+    {
+      "epoch": 0.11563652374924321,
+      "grad_norm": 1.2962623834609985,
+      "learning_rate": 9.924955975806608e-06,
+      "loss": 0.8162,
+      "step": 2101
+    },
+    {
+      "epoch": 0.11569156255159888,
+      "grad_norm": 0.8571485877037048,
+      "learning_rate": 9.924881138970453e-06,
+      "loss": 0.8581,
+      "step": 2102
+    },
+    {
+      "epoch": 0.11574660135395454,
+      "grad_norm": 0.8326650857925415,
+      "learning_rate": 9.92480626512021e-06,
+      "loss": 0.8438,
+      "step": 2103
+    },
+    {
+      "epoch": 0.1158016401563102,
+      "grad_norm": 0.7973701357841492,
+      "learning_rate": 9.924731354256441e-06,
+      "loss": 0.8337,
+      "step": 2104
+    },
+    {
+      "epoch": 0.11585667895866586,
+      "grad_norm": 0.8614075779914856,
+      "learning_rate": 9.924656406379708e-06,
+      "loss": 0.8275,
+      "step": 2105
+    },
+    {
+      "epoch": 0.11591171776102152,
+      "grad_norm": 0.7911350131034851,
+      "learning_rate": 9.924581421490577e-06,
+      "loss": 0.8032,
+      "step": 2106
+    },
+    {
+      "epoch": 0.11596675656337718,
+      "grad_norm": 0.8763116598129272,
+      "learning_rate": 9.92450639958961e-06,
+      "loss": 0.8725,
+      "step": 2107
+    },
+    {
+      "epoch": 0.11602179536573284,
+      "grad_norm": 0.9754133224487305,
+      "learning_rate": 9.92443134067737e-06,
+      "loss": 0.9115,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1160768341680885,
+      "grad_norm": 0.7783731818199158,
+      "learning_rate": 9.924356244754425e-06,
+      "loss": 0.8223,
+      "step": 2109
+    },
+    {
+      "epoch": 0.11613187297044417,
+      "grad_norm": 0.865301787853241,
+      "learning_rate": 9.924281111821335e-06,
+      "loss": 0.8053,
+      "step": 2110
+    },
+    {
+      "epoch": 0.11618691177279983,
+      "grad_norm": 0.8654297590255737,
+      "learning_rate": 9.924205941878666e-06,
+      "loss": 0.716,
+      "step": 2111
+    },
+    {
+      "epoch": 0.11624195057515549,
+      "grad_norm": 0.7646550536155701,
+      "learning_rate": 9.924130734926982e-06,
+      "loss": 0.8027,
+      "step": 2112
+    },
+    {
+      "epoch": 0.11629698937751115,
+      "grad_norm": 0.810587465763092,
+      "learning_rate": 9.924055490966851e-06,
+      "loss": 0.7416,
+      "step": 2113
+    },
+    {
+      "epoch": 0.11635202817986681,
+      "grad_norm": 0.8610082268714905,
+      "learning_rate": 9.923980209998838e-06,
+      "loss": 0.8527,
+      "step": 2114
+    },
+    {
+      "epoch": 0.11640706698222247,
+      "grad_norm": 0.8409233689308167,
+      "learning_rate": 9.923904892023506e-06,
+      "loss": 0.8169,
+      "step": 2115
+    },
+    {
+      "epoch": 0.11646210578457812,
+      "grad_norm": 0.7786587476730347,
+      "learning_rate": 9.923829537041425e-06,
+      "loss": 0.6897,
+      "step": 2116
+    },
+    {
+      "epoch": 0.11651714458693378,
+      "grad_norm": 0.852908730506897,
+      "learning_rate": 9.923754145053158e-06,
+      "loss": 0.7821,
+      "step": 2117
+    },
+    {
+      "epoch": 0.11657218338928944,
+      "grad_norm": 0.9130391478538513,
+      "learning_rate": 9.923678716059273e-06,
+      "loss": 1.0377,
+      "step": 2118
+    },
+    {
+      "epoch": 0.1166272221916451,
+      "grad_norm": 0.8371701240539551,
+      "learning_rate": 9.923603250060336e-06,
+      "loss": 0.8312,
+      "step": 2119
+    },
+    {
+      "epoch": 0.11668226099400077,
+      "grad_norm": 0.8045756220817566,
+      "learning_rate": 9.923527747056916e-06,
+      "loss": 0.7971,
+      "step": 2120
+    },
+    {
+      "epoch": 0.11673729979635643,
+      "grad_norm": 0.8832160234451294,
+      "learning_rate": 9.923452207049577e-06,
+      "loss": 0.7362,
+      "step": 2121
+    },
+    {
+      "epoch": 0.11679233859871209,
+      "grad_norm": 0.8253088593482971,
+      "learning_rate": 9.923376630038893e-06,
+      "loss": 0.8177,
+      "step": 2122
+    },
+    {
+      "epoch": 0.11684737740106775,
+      "grad_norm": 0.7953168749809265,
+      "learning_rate": 9.923301016025424e-06,
+      "loss": 0.7053,
+      "step": 2123
+    },
+    {
+      "epoch": 0.11690241620342341,
+      "grad_norm": 0.7256457805633545,
+      "learning_rate": 9.923225365009745e-06,
+      "loss": 0.7554,
+      "step": 2124
+    },
+    {
+      "epoch": 0.11695745500577907,
+      "grad_norm": 0.9896693229675293,
+      "learning_rate": 9.923149676992424e-06,
+      "loss": 0.8285,
+      "step": 2125
+    },
+    {
+      "epoch": 0.11701249380813473,
+      "grad_norm": 0.7846312522888184,
+      "learning_rate": 9.923073951974023e-06,
+      "loss": 0.7527,
+      "step": 2126
+    },
+    {
+      "epoch": 0.1170675326104904,
+      "grad_norm": 0.8949825167655945,
+      "learning_rate": 9.92299818995512e-06,
+      "loss": 0.8545,
+      "step": 2127
+    },
+    {
+      "epoch": 0.11712257141284606,
+      "grad_norm": 1.0023548603057861,
+      "learning_rate": 9.922922390936278e-06,
+      "loss": 0.7668,
+      "step": 2128
+    },
+    {
+      "epoch": 0.11717761021520172,
+      "grad_norm": 0.8663881421089172,
+      "learning_rate": 9.92284655491807e-06,
+      "loss": 0.8073,
+      "step": 2129
+    },
+    {
+      "epoch": 0.11723264901755738,
+      "grad_norm": 0.8274385929107666,
+      "learning_rate": 9.922770681901064e-06,
+      "loss": 0.9002,
+      "step": 2130
+    },
+    {
+      "epoch": 0.11728768781991304,
+      "grad_norm": 0.8508959412574768,
+      "learning_rate": 9.922694771885832e-06,
+      "loss": 0.9325,
+      "step": 2131
+    },
+    {
+      "epoch": 0.1173427266222687,
+      "grad_norm": 0.8176792860031128,
+      "learning_rate": 9.922618824872946e-06,
+      "loss": 0.8415,
+      "step": 2132
+    },
+    {
+      "epoch": 0.11739776542462436,
+      "grad_norm": 0.770951509475708,
+      "learning_rate": 9.922542840862971e-06,
+      "loss": 0.8051,
+      "step": 2133
+    },
+    {
+      "epoch": 0.11745280422698003,
+      "grad_norm": 0.8558167219161987,
+      "learning_rate": 9.922466819856484e-06,
+      "loss": 0.85,
+      "step": 2134
+    },
+    {
+      "epoch": 0.11750784302933569,
+      "grad_norm": 0.8288151025772095,
+      "learning_rate": 9.922390761854053e-06,
+      "loss": 0.8141,
+      "step": 2135
+    },
+    {
+      "epoch": 0.11756288183169135,
+      "grad_norm": 0.8220882415771484,
+      "learning_rate": 9.922314666856252e-06,
+      "loss": 0.8109,
+      "step": 2136
+    },
+    {
+      "epoch": 0.117617920634047,
+      "grad_norm": 0.7875000238418579,
+      "learning_rate": 9.92223853486365e-06,
+      "loss": 0.9085,
+      "step": 2137
+    },
+    {
+      "epoch": 0.11767295943640266,
+      "grad_norm": 0.8052374124526978,
+      "learning_rate": 9.922162365876822e-06,
+      "loss": 0.8785,
+      "step": 2138
+    },
+    {
+      "epoch": 0.11772799823875832,
+      "grad_norm": 1.0311180353164673,
+      "learning_rate": 9.922086159896338e-06,
+      "loss": 0.9112,
+      "step": 2139
+    },
+    {
+      "epoch": 0.11778303704111398,
+      "grad_norm": 0.943911075592041,
+      "learning_rate": 9.922009916922773e-06,
+      "loss": 0.8332,
+      "step": 2140
+    },
+    {
+      "epoch": 0.11783807584346964,
+      "grad_norm": 0.8156648278236389,
+      "learning_rate": 9.921933636956697e-06,
+      "loss": 0.8837,
+      "step": 2141
+    },
+    {
+      "epoch": 0.1178931146458253,
+      "grad_norm": 0.860292375087738,
+      "learning_rate": 9.921857319998688e-06,
+      "loss": 0.7963,
+      "step": 2142
+    },
+    {
+      "epoch": 0.11794815344818096,
+      "grad_norm": 0.8861456513404846,
+      "learning_rate": 9.921780966049315e-06,
+      "loss": 0.8335,
+      "step": 2143
+    },
+    {
+      "epoch": 0.11800319225053663,
+      "grad_norm": 0.793533205986023,
+      "learning_rate": 9.921704575109155e-06,
+      "loss": 0.7881,
+      "step": 2144
+    },
+    {
+      "epoch": 0.11805823105289229,
+      "grad_norm": 0.8039320111274719,
+      "learning_rate": 9.921628147178781e-06,
+      "loss": 0.8369,
+      "step": 2145
+    },
+    {
+      "epoch": 0.11811326985524795,
+      "grad_norm": 0.8785450458526611,
+      "learning_rate": 9.921551682258765e-06,
+      "loss": 0.7981,
+      "step": 2146
+    },
+    {
+      "epoch": 0.11816830865760361,
+      "grad_norm": 0.810251772403717,
+      "learning_rate": 9.921475180349687e-06,
+      "loss": 0.7926,
+      "step": 2147
+    },
+    {
+      "epoch": 0.11822334745995927,
+      "grad_norm": 0.8470801115036011,
+      "learning_rate": 9.921398641452117e-06,
+      "loss": 0.8061,
+      "step": 2148
+    },
+    {
+      "epoch": 0.11827838626231493,
+      "grad_norm": 0.8147469162940979,
+      "learning_rate": 9.921322065566633e-06,
+      "loss": 0.7906,
+      "step": 2149
+    },
+    {
+      "epoch": 0.1183334250646706,
+      "grad_norm": 0.8792327046394348,
+      "learning_rate": 9.92124545269381e-06,
+      "loss": 0.9025,
+      "step": 2150
+    },
+    {
+      "epoch": 0.11838846386702626,
+      "grad_norm": 0.794607400894165,
+      "learning_rate": 9.921168802834223e-06,
+      "loss": 0.8284,
+      "step": 2151
+    },
+    {
+      "epoch": 0.11844350266938192,
+      "grad_norm": 0.8601556420326233,
+      "learning_rate": 9.921092115988447e-06,
+      "loss": 0.8196,
+      "step": 2152
+    },
+    {
+      "epoch": 0.11849854147173758,
+      "grad_norm": 0.786967933177948,
+      "learning_rate": 9.921015392157062e-06,
+      "loss": 0.8744,
+      "step": 2153
+    },
+    {
+      "epoch": 0.11855358027409324,
+      "grad_norm": 0.8481432199478149,
+      "learning_rate": 9.920938631340641e-06,
+      "loss": 0.7206,
+      "step": 2154
+    },
+    {
+      "epoch": 0.1186086190764489,
+      "grad_norm": 0.8025142550468445,
+      "learning_rate": 9.920861833539765e-06,
+      "loss": 0.8126,
+      "step": 2155
+    },
+    {
+      "epoch": 0.11866365787880456,
+      "grad_norm": 0.9853057265281677,
+      "learning_rate": 9.920784998755006e-06,
+      "loss": 0.8883,
+      "step": 2156
+    },
+    {
+      "epoch": 0.11871869668116022,
+      "grad_norm": 1.0008476972579956,
+      "learning_rate": 9.920708126986947e-06,
+      "loss": 0.9326,
+      "step": 2157
+    },
+    {
+      "epoch": 0.11877373548351589,
+      "grad_norm": 0.837347686290741,
+      "learning_rate": 9.920631218236161e-06,
+      "loss": 0.9002,
+      "step": 2158
+    },
+    {
+      "epoch": 0.11882877428587153,
+      "grad_norm": 0.7866735458374023,
+      "learning_rate": 9.920554272503227e-06,
+      "loss": 0.765,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1188838130882272,
+      "grad_norm": 0.8714935779571533,
+      "learning_rate": 9.920477289788726e-06,
+      "loss": 1.0294,
+      "step": 2160
+    },
+    {
+      "epoch": 0.11893885189058286,
+      "grad_norm": 1.0671826601028442,
+      "learning_rate": 9.920400270093234e-06,
+      "loss": 0.8341,
+      "step": 2161
+    },
+    {
+      "epoch": 0.11899389069293852,
+      "grad_norm": 0.8594604134559631,
+      "learning_rate": 9.92032321341733e-06,
+      "loss": 0.8731,
+      "step": 2162
+    },
+    {
+      "epoch": 0.11904892949529418,
+      "grad_norm": 0.8387738466262817,
+      "learning_rate": 9.920246119761597e-06,
+      "loss": 0.7898,
+      "step": 2163
+    },
+    {
+      "epoch": 0.11910396829764984,
+      "grad_norm": 0.8957195281982422,
+      "learning_rate": 9.920168989126608e-06,
+      "loss": 0.8475,
+      "step": 2164
+    },
+    {
+      "epoch": 0.1191590071000055,
+      "grad_norm": 0.8224207162857056,
+      "learning_rate": 9.920091821512948e-06,
+      "loss": 0.7944,
+      "step": 2165
+    },
+    {
+      "epoch": 0.11921404590236116,
+      "grad_norm": 1.0309031009674072,
+      "learning_rate": 9.920014616921192e-06,
+      "loss": 0.8992,
+      "step": 2166
+    },
+    {
+      "epoch": 0.11926908470471682,
+      "grad_norm": 0.7300832271575928,
+      "learning_rate": 9.919937375351925e-06,
+      "loss": 0.7016,
+      "step": 2167
+    },
+    {
+      "epoch": 0.11932412350707249,
+      "grad_norm": 0.7565537691116333,
+      "learning_rate": 9.919860096805724e-06,
+      "loss": 0.8113,
+      "step": 2168
+    },
+    {
+      "epoch": 0.11937916230942815,
+      "grad_norm": 1.0101505517959595,
+      "learning_rate": 9.919782781283174e-06,
+      "loss": 0.8765,
+      "step": 2169
+    },
+    {
+      "epoch": 0.11943420111178381,
+      "grad_norm": 0.8369461894035339,
+      "learning_rate": 9.919705428784852e-06,
+      "loss": 0.8248,
+      "step": 2170
+    },
+    {
+      "epoch": 0.11948923991413947,
+      "grad_norm": 0.8106105327606201,
+      "learning_rate": 9.919628039311342e-06,
+      "loss": 0.8585,
+      "step": 2171
+    },
+    {
+      "epoch": 0.11954427871649513,
+      "grad_norm": 0.7863745093345642,
+      "learning_rate": 9.919550612863224e-06,
+      "loss": 0.8393,
+      "step": 2172
+    },
+    {
+      "epoch": 0.11959931751885079,
+      "grad_norm": 0.8664719462394714,
+      "learning_rate": 9.919473149441081e-06,
+      "loss": 0.8882,
+      "step": 2173
+    },
+    {
+      "epoch": 0.11965435632120645,
+      "grad_norm": 0.6977574825286865,
+      "learning_rate": 9.919395649045494e-06,
+      "loss": 0.7264,
+      "step": 2174
+    },
+    {
+      "epoch": 0.11970939512356212,
+      "grad_norm": 0.8000102639198303,
+      "learning_rate": 9.919318111677045e-06,
+      "loss": 0.7828,
+      "step": 2175
+    },
+    {
+      "epoch": 0.11976443392591778,
+      "grad_norm": 0.868228018283844,
+      "learning_rate": 9.91924053733632e-06,
+      "loss": 0.7904,
+      "step": 2176
+    },
+    {
+      "epoch": 0.11981947272827344,
+      "grad_norm": 0.839080274105072,
+      "learning_rate": 9.9191629260239e-06,
+      "loss": 0.7663,
+      "step": 2177
+    },
+    {
+      "epoch": 0.1198745115306291,
+      "grad_norm": 0.8222747445106506,
+      "learning_rate": 9.919085277740366e-06,
+      "loss": 0.7208,
+      "step": 2178
+    },
+    {
+      "epoch": 0.11992955033298476,
+      "grad_norm": 1.4550986289978027,
+      "learning_rate": 9.919007592486304e-06,
+      "loss": 0.8154,
+      "step": 2179
+    },
+    {
+      "epoch": 0.11998458913534041,
+      "grad_norm": 0.9110257625579834,
+      "learning_rate": 9.9189298702623e-06,
+      "loss": 0.8134,
+      "step": 2180
+    },
+    {
+      "epoch": 0.12003962793769607,
+      "grad_norm": 0.84796142578125,
+      "learning_rate": 9.918852111068935e-06,
+      "loss": 0.8074,
+      "step": 2181
+    },
+    {
+      "epoch": 0.12009466674005173,
+      "grad_norm": 0.8134179711341858,
+      "learning_rate": 9.918774314906793e-06,
+      "loss": 0.6335,
+      "step": 2182
+    },
+    {
+      "epoch": 0.12014970554240739,
+      "grad_norm": 0.8481448888778687,
+      "learning_rate": 9.918696481776461e-06,
+      "loss": 0.8804,
+      "step": 2183
+    },
+    {
+      "epoch": 0.12020474434476305,
+      "grad_norm": 0.88057941198349,
+      "learning_rate": 9.918618611678523e-06,
+      "loss": 0.9326,
+      "step": 2184
+    },
+    {
+      "epoch": 0.12025978314711872,
+      "grad_norm": 0.8435977697372437,
+      "learning_rate": 9.918540704613564e-06,
+      "loss": 0.8141,
+      "step": 2185
+    },
+    {
+      "epoch": 0.12031482194947438,
+      "grad_norm": 0.8186982870101929,
+      "learning_rate": 9.918462760582169e-06,
+      "loss": 0.837,
+      "step": 2186
+    },
+    {
+      "epoch": 0.12036986075183004,
+      "grad_norm": 0.887783944606781,
+      "learning_rate": 9.918384779584924e-06,
+      "loss": 0.8062,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1204248995541857,
+      "grad_norm": 0.9368415474891663,
+      "learning_rate": 9.918306761622417e-06,
+      "loss": 1.0098,
+      "step": 2188
+    },
+    {
+      "epoch": 0.12047993835654136,
+      "grad_norm": 0.8443986773490906,
+      "learning_rate": 9.918228706695232e-06,
+      "loss": 0.8178,
+      "step": 2189
+    },
+    {
+      "epoch": 0.12053497715889702,
+      "grad_norm": 0.7897284626960754,
+      "learning_rate": 9.918150614803956e-06,
+      "loss": 0.8013,
+      "step": 2190
+    },
+    {
+      "epoch": 0.12059001596125268,
+      "grad_norm": 0.886012077331543,
+      "learning_rate": 9.91807248594918e-06,
+      "loss": 0.8141,
+      "step": 2191
+    },
+    {
+      "epoch": 0.12064505476360834,
+      "grad_norm": 0.8585757613182068,
+      "learning_rate": 9.917994320131484e-06,
+      "loss": 0.8381,
+      "step": 2192
+    },
+    {
+      "epoch": 0.120700093565964,
+      "grad_norm": 1.6192269325256348,
+      "learning_rate": 9.917916117351459e-06,
+      "loss": 0.9082,
+      "step": 2193
+    },
+    {
+      "epoch": 0.12075513236831967,
+      "grad_norm": 1.160414457321167,
+      "learning_rate": 9.917837877609695e-06,
+      "loss": 0.8673,
+      "step": 2194
+    },
+    {
+      "epoch": 0.12081017117067533,
+      "grad_norm": 0.8363412022590637,
+      "learning_rate": 9.917759600906775e-06,
+      "loss": 0.816,
+      "step": 2195
+    },
+    {
+      "epoch": 0.12086520997303099,
+      "grad_norm": 0.8344097137451172,
+      "learning_rate": 9.917681287243292e-06,
+      "loss": 0.8629,
+      "step": 2196
+    },
+    {
+      "epoch": 0.12092024877538665,
+      "grad_norm": 0.9817582368850708,
+      "learning_rate": 9.917602936619834e-06,
+      "loss": 0.8106,
+      "step": 2197
+    },
+    {
+      "epoch": 0.12097528757774231,
+      "grad_norm": 0.8828088641166687,
+      "learning_rate": 9.917524549036987e-06,
+      "loss": 0.8465,
+      "step": 2198
+    },
+    {
+      "epoch": 0.12103032638009797,
+      "grad_norm": 0.8428277969360352,
+      "learning_rate": 9.917446124495344e-06,
+      "loss": 0.7721,
+      "step": 2199
+    },
+    {
+      "epoch": 0.12108536518245364,
+      "grad_norm": 0.8748664855957031,
+      "learning_rate": 9.917367662995489e-06,
+      "loss": 0.8679,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1211404039848093,
+      "grad_norm": 0.8652347922325134,
+      "learning_rate": 9.917289164538018e-06,
+      "loss": 0.8906,
+      "step": 2201
+    },
+    {
+      "epoch": 0.12119544278716494,
+      "grad_norm": 1.157142162322998,
+      "learning_rate": 9.917210629123518e-06,
+      "loss": 0.9046,
+      "step": 2202
+    },
+    {
+      "epoch": 0.1212504815895206,
+      "grad_norm": 0.8186333179473877,
+      "learning_rate": 9.917132056752576e-06,
+      "loss": 0.8494,
+      "step": 2203
+    },
+    {
+      "epoch": 0.12130552039187627,
+      "grad_norm": 0.7769078612327576,
+      "learning_rate": 9.917053447425788e-06,
+      "loss": 0.8018,
+      "step": 2204
+    },
+    {
+      "epoch": 0.12136055919423193,
+      "grad_norm": 0.9190469980239868,
+      "learning_rate": 9.916974801143742e-06,
+      "loss": 0.8206,
+      "step": 2205
+    },
+    {
+      "epoch": 0.12141559799658759,
+      "grad_norm": 1.2200725078582764,
+      "learning_rate": 9.91689611790703e-06,
+      "loss": 0.9109,
+      "step": 2206
+    },
+    {
+      "epoch": 0.12147063679894325,
+      "grad_norm": 0.7902093529701233,
+      "learning_rate": 9.916817397716243e-06,
+      "loss": 0.8314,
+      "step": 2207
+    },
+    {
+      "epoch": 0.12152567560129891,
+      "grad_norm": 0.8160610198974609,
+      "learning_rate": 9.91673864057197e-06,
+      "loss": 0.8605,
+      "step": 2208
+    },
+    {
+      "epoch": 0.12158071440365457,
+      "grad_norm": 0.833163857460022,
+      "learning_rate": 9.916659846474807e-06,
+      "loss": 0.8125,
+      "step": 2209
+    },
+    {
+      "epoch": 0.12163575320601024,
+      "grad_norm": 0.776314377784729,
+      "learning_rate": 9.916581015425346e-06,
+      "loss": 0.8137,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1216907920083659,
+      "grad_norm": 0.8525915145874023,
+      "learning_rate": 9.916502147424178e-06,
+      "loss": 0.8703,
+      "step": 2211
+    },
+    {
+      "epoch": 0.12174583081072156,
+      "grad_norm": 0.8268684148788452,
+      "learning_rate": 9.916423242471895e-06,
+      "loss": 0.7775,
+      "step": 2212
+    },
+    {
+      "epoch": 0.12180086961307722,
+      "grad_norm": 0.8717706799507141,
+      "learning_rate": 9.916344300569091e-06,
+      "loss": 0.8002,
+      "step": 2213
+    },
+    {
+      "epoch": 0.12185590841543288,
+      "grad_norm": 0.9499961137771606,
+      "learning_rate": 9.91626532171636e-06,
+      "loss": 0.8861,
+      "step": 2214
+    },
+    {
+      "epoch": 0.12191094721778854,
+      "grad_norm": 0.9521885514259338,
+      "learning_rate": 9.916186305914296e-06,
+      "loss": 0.7602,
+      "step": 2215
+    },
+    {
+      "epoch": 0.1219659860201442,
+      "grad_norm": 0.8945447206497192,
+      "learning_rate": 9.916107253163488e-06,
+      "loss": 0.8603,
+      "step": 2216
+    },
+    {
+      "epoch": 0.12202102482249987,
+      "grad_norm": 0.8232392072677612,
+      "learning_rate": 9.916028163464536e-06,
+      "loss": 0.8419,
+      "step": 2217
+    },
+    {
+      "epoch": 0.12207606362485553,
+      "grad_norm": 0.8183467984199524,
+      "learning_rate": 9.915949036818032e-06,
+      "loss": 0.9038,
+      "step": 2218
+    },
+    {
+      "epoch": 0.12213110242721119,
+      "grad_norm": 0.7805467247962952,
+      "learning_rate": 9.915869873224571e-06,
+      "loss": 0.7313,
+      "step": 2219
+    },
+    {
+      "epoch": 0.12218614122956685,
+      "grad_norm": 0.838101327419281,
+      "learning_rate": 9.915790672684749e-06,
+      "loss": 0.7973,
+      "step": 2220
+    },
+    {
+      "epoch": 0.12224118003192251,
+      "grad_norm": 0.7795171141624451,
+      "learning_rate": 9.915711435199158e-06,
+      "loss": 0.7796,
+      "step": 2221
+    },
+    {
+      "epoch": 0.12229621883427817,
+      "grad_norm": 0.7971234917640686,
+      "learning_rate": 9.915632160768398e-06,
+      "loss": 0.8309,
+      "step": 2222
+    },
+    {
+      "epoch": 0.12235125763663382,
+      "grad_norm": 0.8543851375579834,
+      "learning_rate": 9.915552849393061e-06,
+      "loss": 0.7826,
+      "step": 2223
+    },
+    {
+      "epoch": 0.12240629643898948,
+      "grad_norm": 0.9315086007118225,
+      "learning_rate": 9.915473501073744e-06,
+      "loss": 0.9294,
+      "step": 2224
+    },
+    {
+      "epoch": 0.12246133524134514,
+      "grad_norm": 0.8794427514076233,
+      "learning_rate": 9.915394115811046e-06,
+      "loss": 0.8968,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1225163740437008,
+      "grad_norm": 0.9499204754829407,
+      "learning_rate": 9.91531469360556e-06,
+      "loss": 0.9841,
+      "step": 2226
+    },
+    {
+      "epoch": 0.12257141284605647,
+      "grad_norm": 0.9233788251876831,
+      "learning_rate": 9.915235234457885e-06,
+      "loss": 0.7794,
+      "step": 2227
+    },
+    {
+      "epoch": 0.12262645164841213,
+      "grad_norm": 0.8971870541572571,
+      "learning_rate": 9.915155738368618e-06,
+      "loss": 0.919,
+      "step": 2228
+    },
+    {
+      "epoch": 0.12268149045076779,
+      "grad_norm": 0.8122105002403259,
+      "learning_rate": 9.915076205338356e-06,
+      "loss": 0.8227,
+      "step": 2229
+    },
+    {
+      "epoch": 0.12273652925312345,
+      "grad_norm": 0.7878004908561707,
+      "learning_rate": 9.914996635367696e-06,
+      "loss": 0.7622,
+      "step": 2230
+    },
+    {
+      "epoch": 0.12279156805547911,
+      "grad_norm": 0.8229606747627258,
+      "learning_rate": 9.914917028457238e-06,
+      "loss": 0.8265,
+      "step": 2231
+    },
+    {
+      "epoch": 0.12284660685783477,
+      "grad_norm": 0.8972312808036804,
+      "learning_rate": 9.914837384607578e-06,
+      "loss": 0.8914,
+      "step": 2232
+    },
+    {
+      "epoch": 0.12290164566019043,
+      "grad_norm": 0.762922465801239,
+      "learning_rate": 9.914757703819318e-06,
+      "loss": 0.6853,
+      "step": 2233
+    },
+    {
+      "epoch": 0.1229566844625461,
+      "grad_norm": 0.8949442505836487,
+      "learning_rate": 9.914677986093054e-06,
+      "loss": 0.8303,
+      "step": 2234
+    },
+    {
+      "epoch": 0.12301172326490176,
+      "grad_norm": 1.0220820903778076,
+      "learning_rate": 9.914598231429384e-06,
+      "loss": 1.0027,
+      "step": 2235
+    },
+    {
+      "epoch": 0.12306676206725742,
+      "grad_norm": 0.8265436887741089,
+      "learning_rate": 9.914518439828911e-06,
+      "loss": 0.8317,
+      "step": 2236
+    },
+    {
+      "epoch": 0.12312180086961308,
+      "grad_norm": 0.780444324016571,
+      "learning_rate": 9.914438611292231e-06,
+      "loss": 0.756,
+      "step": 2237
+    },
+    {
+      "epoch": 0.12317683967196874,
+      "grad_norm": 0.8569482564926147,
+      "learning_rate": 9.914358745819948e-06,
+      "loss": 0.8126,
+      "step": 2238
+    },
+    {
+      "epoch": 0.1232318784743244,
+      "grad_norm": 0.8167145848274231,
+      "learning_rate": 9.91427884341266e-06,
+      "loss": 0.8345,
+      "step": 2239
+    },
+    {
+      "epoch": 0.12328691727668006,
+      "grad_norm": 0.7915990948677063,
+      "learning_rate": 9.914198904070967e-06,
+      "loss": 0.7416,
+      "step": 2240
+    },
+    {
+      "epoch": 0.12334195607903573,
+      "grad_norm": 0.8568083047866821,
+      "learning_rate": 9.91411892779547e-06,
+      "loss": 0.8329,
+      "step": 2241
+    },
+    {
+      "epoch": 0.12339699488139139,
+      "grad_norm": 1.1727303266525269,
+      "learning_rate": 9.914038914586772e-06,
+      "loss": 0.8421,
+      "step": 2242
+    },
+    {
+      "epoch": 0.12345203368374705,
+      "grad_norm": 0.8706398010253906,
+      "learning_rate": 9.913958864445472e-06,
+      "loss": 0.9013,
+      "step": 2243
+    },
+    {
+      "epoch": 0.12350707248610271,
+      "grad_norm": 0.8376144170761108,
+      "learning_rate": 9.913878777372173e-06,
+      "loss": 0.8456,
+      "step": 2244
+    },
+    {
+      "epoch": 0.12356211128845836,
+      "grad_norm": 0.8388974070549011,
+      "learning_rate": 9.913798653367478e-06,
+      "loss": 0.787,
+      "step": 2245
+    },
+    {
+      "epoch": 0.12361715009081402,
+      "grad_norm": 0.8625446557998657,
+      "learning_rate": 9.913718492431984e-06,
+      "loss": 0.7758,
+      "step": 2246
+    },
+    {
+      "epoch": 0.12367218889316968,
+      "grad_norm": 0.8805570006370544,
+      "learning_rate": 9.913638294566299e-06,
+      "loss": 0.8755,
+      "step": 2247
+    },
+    {
+      "epoch": 0.12372722769552534,
+      "grad_norm": 0.8102611899375916,
+      "learning_rate": 9.913558059771025e-06,
+      "loss": 0.8495,
+      "step": 2248
+    },
+    {
+      "epoch": 0.123782266497881,
+      "grad_norm": 0.8506311774253845,
+      "learning_rate": 9.913477788046762e-06,
+      "loss": 0.7413,
+      "step": 2249
+    },
+    {
+      "epoch": 0.12383730530023666,
+      "grad_norm": 1.0789196491241455,
+      "learning_rate": 9.913397479394116e-06,
+      "loss": 0.8993,
+      "step": 2250
+    },
+    {
+      "epoch": 0.12389234410259232,
+      "grad_norm": 1.5664849281311035,
+      "learning_rate": 9.91331713381369e-06,
+      "loss": 0.8322,
+      "step": 2251
+    },
+    {
+      "epoch": 0.12394738290494799,
+      "grad_norm": 1.1347390413284302,
+      "learning_rate": 9.913236751306085e-06,
+      "loss": 0.8756,
+      "step": 2252
+    },
+    {
+      "epoch": 0.12400242170730365,
+      "grad_norm": 0.8111063241958618,
+      "learning_rate": 9.913156331871911e-06,
+      "loss": 0.831,
+      "step": 2253
+    },
+    {
+      "epoch": 0.12405746050965931,
+      "grad_norm": 0.817812979221344,
+      "learning_rate": 9.913075875511769e-06,
+      "loss": 0.8531,
+      "step": 2254
+    },
+    {
+      "epoch": 0.12411249931201497,
+      "grad_norm": 0.7678318619728088,
+      "learning_rate": 9.912995382226263e-06,
+      "loss": 0.8028,
+      "step": 2255
+    },
+    {
+      "epoch": 0.12416753811437063,
+      "grad_norm": 0.8207805156707764,
+      "learning_rate": 9.912914852015998e-06,
+      "loss": 0.8856,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1242225769167263,
+      "grad_norm": 0.978484570980072,
+      "learning_rate": 9.912834284881582e-06,
+      "loss": 0.933,
+      "step": 2257
+    },
+    {
+      "epoch": 0.12427761571908195,
+      "grad_norm": 0.9215858578681946,
+      "learning_rate": 9.912753680823617e-06,
+      "loss": 0.7771,
+      "step": 2258
+    },
+    {
+      "epoch": 0.12433265452143762,
+      "grad_norm": 0.8542179465293884,
+      "learning_rate": 9.91267303984271e-06,
+      "loss": 0.8652,
+      "step": 2259
+    },
+    {
+      "epoch": 0.12438769332379328,
+      "grad_norm": 0.7985575199127197,
+      "learning_rate": 9.912592361939469e-06,
+      "loss": 0.7011,
+      "step": 2260
+    },
+    {
+      "epoch": 0.12444273212614894,
+      "grad_norm": 0.8868670463562012,
+      "learning_rate": 9.912511647114498e-06,
+      "loss": 0.8222,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1244977709285046,
+      "grad_norm": 0.7966209650039673,
+      "learning_rate": 9.912430895368405e-06,
+      "loss": 0.776,
+      "step": 2262
+    },
+    {
+      "epoch": 0.12455280973086026,
+      "grad_norm": 0.7844830751419067,
+      "learning_rate": 9.912350106701796e-06,
+      "loss": 0.7513,
+      "step": 2263
+    },
+    {
+      "epoch": 0.12460784853321592,
+      "grad_norm": 0.7788559794425964,
+      "learning_rate": 9.912269281115278e-06,
+      "loss": 0.8517,
+      "step": 2264
+    },
+    {
+      "epoch": 0.12466288733557158,
+      "grad_norm": 0.778225839138031,
+      "learning_rate": 9.912188418609461e-06,
+      "loss": 0.7504,
+      "step": 2265
+    },
+    {
+      "epoch": 0.12471792613792723,
+      "grad_norm": 0.7955968976020813,
+      "learning_rate": 9.912107519184947e-06,
+      "loss": 0.8152,
+      "step": 2266
+    },
+    {
+      "epoch": 0.1247729649402829,
+      "grad_norm": 1.1202566623687744,
+      "learning_rate": 9.912026582842352e-06,
+      "loss": 0.9325,
+      "step": 2267
+    },
+    {
+      "epoch": 0.12482800374263855,
+      "grad_norm": 0.9762749671936035,
+      "learning_rate": 9.911945609582279e-06,
+      "loss": 0.9027,
+      "step": 2268
+    },
+    {
+      "epoch": 0.12488304254499422,
+      "grad_norm": 0.8311051726341248,
+      "learning_rate": 9.911864599405336e-06,
+      "loss": 0.838,
+      "step": 2269
+    },
+    {
+      "epoch": 0.12493808134734988,
+      "grad_norm": 1.0136815309524536,
+      "learning_rate": 9.911783552312134e-06,
+      "loss": 0.9288,
+      "step": 2270
+    },
+    {
+      "epoch": 0.12499312014970554,
+      "grad_norm": 0.7960494160652161,
+      "learning_rate": 9.911702468303282e-06,
+      "loss": 0.8007,
+      "step": 2271
+    },
+    {
+      "epoch": 0.1250481589520612,
+      "grad_norm": 0.9980880618095398,
+      "learning_rate": 9.911621347379388e-06,
+      "loss": 0.8613,
+      "step": 2272
+    },
+    {
+      "epoch": 0.12510319775441686,
+      "grad_norm": 0.8916807770729065,
+      "learning_rate": 9.911540189541065e-06,
+      "loss": 0.8783,
+      "step": 2273
+    },
+    {
+      "epoch": 0.12515823655677252,
+      "grad_norm": 0.9455892443656921,
+      "learning_rate": 9.911458994788919e-06,
+      "loss": 0.8676,
+      "step": 2274
+    },
+    {
+      "epoch": 0.12521327535912818,
+      "grad_norm": 0.7649906277656555,
+      "learning_rate": 9.911377763123561e-06,
+      "loss": 0.7763,
+      "step": 2275
+    },
+    {
+      "epoch": 0.12526831416148385,
+      "grad_norm": 0.8971202373504639,
+      "learning_rate": 9.911296494545604e-06,
+      "loss": 0.9022,
+      "step": 2276
+    },
+    {
+      "epoch": 0.1253233529638395,
+      "grad_norm": 0.833678126335144,
+      "learning_rate": 9.911215189055657e-06,
+      "loss": 0.8401,
+      "step": 2277
+    },
+    {
+      "epoch": 0.12537839176619517,
+      "grad_norm": 0.8967958688735962,
+      "learning_rate": 9.911133846654331e-06,
+      "loss": 0.8678,
+      "step": 2278
+    },
+    {
+      "epoch": 0.12543343056855083,
+      "grad_norm": 0.8195546865463257,
+      "learning_rate": 9.911052467342239e-06,
+      "loss": 0.842,
+      "step": 2279
+    },
+    {
+      "epoch": 0.1254884693709065,
+      "grad_norm": 1.095815896987915,
+      "learning_rate": 9.910971051119988e-06,
+      "loss": 0.845,
+      "step": 2280
+    },
+    {
+      "epoch": 0.12554350817326215,
+      "grad_norm": 0.9452629685401917,
+      "learning_rate": 9.910889597988197e-06,
+      "loss": 0.8971,
+      "step": 2281
+    },
+    {
+      "epoch": 0.12559854697561781,
+      "grad_norm": 0.9872332215309143,
+      "learning_rate": 9.910808107947471e-06,
+      "loss": 0.7994,
+      "step": 2282
+    },
+    {
+      "epoch": 0.12565358577797348,
+      "grad_norm": 0.7761966586112976,
+      "learning_rate": 9.910726580998427e-06,
+      "loss": 0.7791,
+      "step": 2283
+    },
+    {
+      "epoch": 0.12570862458032914,
+      "grad_norm": 0.8950315713882446,
+      "learning_rate": 9.910645017141678e-06,
+      "loss": 0.8499,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1257636633826848,
+      "grad_norm": 0.8796371221542358,
+      "learning_rate": 9.910563416377834e-06,
+      "loss": 0.8587,
+      "step": 2285
+    },
+    {
+      "epoch": 0.12581870218504046,
+      "grad_norm": 0.8291982412338257,
+      "learning_rate": 9.91048177870751e-06,
+      "loss": 0.9166,
+      "step": 2286
+    },
+    {
+      "epoch": 0.12587374098739612,
+      "grad_norm": 0.758369505405426,
+      "learning_rate": 9.91040010413132e-06,
+      "loss": 0.8305,
+      "step": 2287
+    },
+    {
+      "epoch": 0.12592877978975178,
+      "grad_norm": 0.8775640726089478,
+      "learning_rate": 9.910318392649876e-06,
+      "loss": 0.8513,
+      "step": 2288
+    },
+    {
+      "epoch": 0.12598381859210744,
+      "grad_norm": 0.8581671118736267,
+      "learning_rate": 9.910236644263796e-06,
+      "loss": 0.8134,
+      "step": 2289
+    },
+    {
+      "epoch": 0.1260388573944631,
+      "grad_norm": 0.8570736050605774,
+      "learning_rate": 9.910154858973689e-06,
+      "loss": 0.826,
+      "step": 2290
+    },
+    {
+      "epoch": 0.12609389619681877,
+      "grad_norm": 0.8712487816810608,
+      "learning_rate": 9.910073036780173e-06,
+      "loss": 0.8042,
+      "step": 2291
+    },
+    {
+      "epoch": 0.12614893499917443,
+      "grad_norm": 0.7584837675094604,
+      "learning_rate": 9.909991177683862e-06,
+      "loss": 0.7715,
+      "step": 2292
+    },
+    {
+      "epoch": 0.1262039738015301,
+      "grad_norm": 0.8618917465209961,
+      "learning_rate": 9.909909281685373e-06,
+      "loss": 0.8755,
+      "step": 2293
+    },
+    {
+      "epoch": 0.12625901260388575,
+      "grad_norm": 0.9530277848243713,
+      "learning_rate": 9.90982734878532e-06,
+      "loss": 0.8538,
+      "step": 2294
+    },
+    {
+      "epoch": 0.1263140514062414,
+      "grad_norm": 0.8394436836242676,
+      "learning_rate": 9.909745378984319e-06,
+      "loss": 0.8401,
+      "step": 2295
+    },
+    {
+      "epoch": 0.12636909020859707,
+      "grad_norm": 0.8224034309387207,
+      "learning_rate": 9.909663372282984e-06,
+      "loss": 0.7201,
+      "step": 2296
+    },
+    {
+      "epoch": 0.12642412901095273,
+      "grad_norm": 0.8215349912643433,
+      "learning_rate": 9.909581328681934e-06,
+      "loss": 0.8824,
+      "step": 2297
+    },
+    {
+      "epoch": 0.12647916781330837,
+      "grad_norm": 0.839389443397522,
+      "learning_rate": 9.909499248181786e-06,
+      "loss": 0.8056,
+      "step": 2298
+    },
+    {
+      "epoch": 0.12653420661566403,
+      "grad_norm": 0.9440048933029175,
+      "learning_rate": 9.909417130783156e-06,
+      "loss": 0.908,
+      "step": 2299
+    },
+    {
+      "epoch": 0.1265892454180197,
+      "grad_norm": 0.8336486220359802,
+      "learning_rate": 9.90933497648666e-06,
+      "loss": 0.8382,
+      "step": 2300
+    },
+    {
+      "epoch": 0.12664428422037535,
+      "grad_norm": 1.1541366577148438,
+      "learning_rate": 9.909252785292918e-06,
+      "loss": 0.8782,
+      "step": 2301
+    },
+    {
+      "epoch": 0.12669932302273101,
+      "grad_norm": 0.8730320334434509,
+      "learning_rate": 9.909170557202545e-06,
+      "loss": 0.7687,
+      "step": 2302
+    },
+    {
+      "epoch": 0.12675436182508668,
+      "grad_norm": 0.9927527904510498,
+      "learning_rate": 9.90908829221616e-06,
+      "loss": 0.8134,
+      "step": 2303
+    },
+    {
+      "epoch": 0.12680940062744234,
+      "grad_norm": 0.9521791338920593,
+      "learning_rate": 9.909005990334381e-06,
+      "loss": 0.9187,
+      "step": 2304
+    },
+    {
+      "epoch": 0.126864439429798,
+      "grad_norm": 0.8012455701828003,
+      "learning_rate": 9.908923651557828e-06,
+      "loss": 0.8581,
+      "step": 2305
+    },
+    {
+      "epoch": 0.12691947823215366,
+      "grad_norm": 0.8882689476013184,
+      "learning_rate": 9.90884127588712e-06,
+      "loss": 0.9317,
+      "step": 2306
+    },
+    {
+      "epoch": 0.12697451703450932,
+      "grad_norm": 0.8408340215682983,
+      "learning_rate": 9.908758863322872e-06,
+      "loss": 0.8444,
+      "step": 2307
+    },
+    {
+      "epoch": 0.12702955583686498,
+      "grad_norm": 0.7856307029724121,
+      "learning_rate": 9.908676413865709e-06,
+      "loss": 0.8457,
+      "step": 2308
+    },
+    {
+      "epoch": 0.12708459463922064,
+      "grad_norm": 0.9459167718887329,
+      "learning_rate": 9.908593927516247e-06,
+      "loss": 0.8153,
+      "step": 2309
+    },
+    {
+      "epoch": 0.1271396334415763,
+      "grad_norm": 0.8629655838012695,
+      "learning_rate": 9.908511404275107e-06,
+      "loss": 0.8279,
+      "step": 2310
+    },
+    {
+      "epoch": 0.12719467224393197,
+      "grad_norm": 1.2012875080108643,
+      "learning_rate": 9.90842884414291e-06,
+      "loss": 1.4388,
+      "step": 2311
+    },
+    {
+      "epoch": 0.12724971104628763,
+      "grad_norm": 1.20725417137146,
+      "learning_rate": 9.908346247120274e-06,
+      "loss": 0.8704,
+      "step": 2312
+    },
+    {
+      "epoch": 0.1273047498486433,
+      "grad_norm": 0.8152929544448853,
+      "learning_rate": 9.908263613207822e-06,
+      "loss": 0.8618,
+      "step": 2313
+    },
+    {
+      "epoch": 0.12735978865099895,
+      "grad_norm": 0.8400965332984924,
+      "learning_rate": 9.908180942406175e-06,
+      "loss": 0.7881,
+      "step": 2314
+    },
+    {
+      "epoch": 0.1274148274533546,
+      "grad_norm": 0.8856974840164185,
+      "learning_rate": 9.908098234715956e-06,
+      "loss": 0.9073,
+      "step": 2315
+    },
+    {
+      "epoch": 0.12746986625571027,
+      "grad_norm": 0.8708439469337463,
+      "learning_rate": 9.908015490137782e-06,
+      "loss": 0.8099,
+      "step": 2316
+    },
+    {
+      "epoch": 0.12752490505806593,
+      "grad_norm": 0.8632444143295288,
+      "learning_rate": 9.907932708672277e-06,
+      "loss": 0.8472,
+      "step": 2317
+    },
+    {
+      "epoch": 0.1275799438604216,
+      "grad_norm": 0.8977149128913879,
+      "learning_rate": 9.907849890320062e-06,
+      "loss": 0.8878,
+      "step": 2318
+    },
+    {
+      "epoch": 0.12763498266277726,
+      "grad_norm": 0.8589425086975098,
+      "learning_rate": 9.907767035081765e-06,
+      "loss": 0.7905,
+      "step": 2319
+    },
+    {
+      "epoch": 0.12769002146513292,
+      "grad_norm": 0.9873501062393188,
+      "learning_rate": 9.907684142958002e-06,
+      "loss": 0.9002,
+      "step": 2320
+    },
+    {
+      "epoch": 0.12774506026748858,
+      "grad_norm": 0.8963840007781982,
+      "learning_rate": 9.9076012139494e-06,
+      "loss": 0.92,
+      "step": 2321
+    },
+    {
+      "epoch": 0.12780009906984424,
+      "grad_norm": 0.7933574318885803,
+      "learning_rate": 9.90751824805658e-06,
+      "loss": 0.7664,
+      "step": 2322
+    },
+    {
+      "epoch": 0.1278551378721999,
+      "grad_norm": 0.9660933017730713,
+      "learning_rate": 9.907435245280167e-06,
+      "loss": 0.9162,
+      "step": 2323
+    },
+    {
+      "epoch": 0.12791017667455556,
+      "grad_norm": 0.8698949217796326,
+      "learning_rate": 9.907352205620783e-06,
+      "loss": 0.7988,
+      "step": 2324
+    },
+    {
+      "epoch": 0.12796521547691123,
+      "grad_norm": 0.9077615141868591,
+      "learning_rate": 9.907269129079055e-06,
+      "loss": 0.8581,
+      "step": 2325
+    },
+    {
+      "epoch": 0.1280202542792669,
+      "grad_norm": 0.9128179550170898,
+      "learning_rate": 9.907186015655607e-06,
+      "loss": 0.8552,
+      "step": 2326
+    },
+    {
+      "epoch": 0.12807529308162255,
+      "grad_norm": 0.9321265816688538,
+      "learning_rate": 9.907102865351062e-06,
+      "loss": 0.889,
+      "step": 2327
+    },
+    {
+      "epoch": 0.1281303318839782,
+      "grad_norm": 0.9687464833259583,
+      "learning_rate": 9.907019678166044e-06,
+      "loss": 0.7944,
+      "step": 2328
+    },
+    {
+      "epoch": 0.12818537068633387,
+      "grad_norm": 0.862223207950592,
+      "learning_rate": 9.90693645410118e-06,
+      "loss": 0.7699,
+      "step": 2329
+    },
+    {
+      "epoch": 0.12824040948868953,
+      "grad_norm": 0.9662127494812012,
+      "learning_rate": 9.906853193157095e-06,
+      "loss": 0.7818,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1282954482910452,
+      "grad_norm": 0.8008295297622681,
+      "learning_rate": 9.906769895334413e-06,
+      "loss": 0.8443,
+      "step": 2331
+    },
+    {
+      "epoch": 0.12835048709340086,
+      "grad_norm": 0.8638464212417603,
+      "learning_rate": 9.906686560633765e-06,
+      "loss": 0.8438,
+      "step": 2332
+    },
+    {
+      "epoch": 0.12840552589575652,
+      "grad_norm": 0.9215866327285767,
+      "learning_rate": 9.906603189055773e-06,
+      "loss": 0.7481,
+      "step": 2333
+    },
+    {
+      "epoch": 0.12846056469811218,
+      "grad_norm": 0.7926739454269409,
+      "learning_rate": 9.906519780601066e-06,
+      "loss": 0.7404,
+      "step": 2334
+    },
+    {
+      "epoch": 0.12851560350046784,
+      "grad_norm": 0.9590242505073547,
+      "learning_rate": 9.906436335270268e-06,
+      "loss": 0.8319,
+      "step": 2335
+    },
+    {
+      "epoch": 0.1285706423028235,
+      "grad_norm": 1.0300076007843018,
+      "learning_rate": 9.906352853064009e-06,
+      "loss": 0.8635,
+      "step": 2336
+    },
+    {
+      "epoch": 0.12862568110517916,
+      "grad_norm": 0.8401443958282471,
+      "learning_rate": 9.906269333982915e-06,
+      "loss": 0.9584,
+      "step": 2337
+    },
+    {
+      "epoch": 0.12868071990753482,
+      "grad_norm": 0.8144069910049438,
+      "learning_rate": 9.906185778027613e-06,
+      "loss": 0.7375,
+      "step": 2338
+    },
+    {
+      "epoch": 0.12873575870989049,
+      "grad_norm": 0.8513948917388916,
+      "learning_rate": 9.906102185198733e-06,
+      "loss": 0.8353,
+      "step": 2339
+    },
+    {
+      "epoch": 0.12879079751224615,
+      "grad_norm": 0.8243077397346497,
+      "learning_rate": 9.906018555496903e-06,
+      "loss": 0.8665,
+      "step": 2340
+    },
+    {
+      "epoch": 0.12884583631460178,
+      "grad_norm": 0.8699066042900085,
+      "learning_rate": 9.905934888922749e-06,
+      "loss": 0.8537,
+      "step": 2341
+    },
+    {
+      "epoch": 0.12890087511695744,
+      "grad_norm": 1.0980210304260254,
+      "learning_rate": 9.905851185476902e-06,
+      "loss": 0.8887,
+      "step": 2342
+    },
+    {
+      "epoch": 0.1289559139193131,
+      "grad_norm": 0.8189190030097961,
+      "learning_rate": 9.905767445159992e-06,
+      "loss": 0.8467,
+      "step": 2343
+    },
+    {
+      "epoch": 0.12901095272166876,
+      "grad_norm": 0.8273541331291199,
+      "learning_rate": 9.905683667972645e-06,
+      "loss": 0.8701,
+      "step": 2344
+    },
+    {
+      "epoch": 0.12906599152402443,
+      "grad_norm": 0.8987969160079956,
+      "learning_rate": 9.905599853915496e-06,
+      "loss": 0.909,
+      "step": 2345
+    },
+    {
+      "epoch": 0.1291210303263801,
+      "grad_norm": 0.818268895149231,
+      "learning_rate": 9.905516002989168e-06,
+      "loss": 0.7946,
+      "step": 2346
+    },
+    {
+      "epoch": 0.12917606912873575,
+      "grad_norm": 0.7401725053787231,
+      "learning_rate": 9.905432115194296e-06,
+      "loss": 0.7006,
+      "step": 2347
+    },
+    {
+      "epoch": 0.1292311079310914,
+      "grad_norm": 0.8263179659843445,
+      "learning_rate": 9.905348190531511e-06,
+      "loss": 0.7768,
+      "step": 2348
+    },
+    {
+      "epoch": 0.12928614673344707,
+      "grad_norm": 0.9241918921470642,
+      "learning_rate": 9.90526422900144e-06,
+      "loss": 0.8593,
+      "step": 2349
+    },
+    {
+      "epoch": 0.12934118553580273,
+      "grad_norm": 0.7804501056671143,
+      "learning_rate": 9.905180230604718e-06,
+      "loss": 0.7607,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1293962243381584,
+      "grad_norm": 0.9408491253852844,
+      "learning_rate": 9.905096195341973e-06,
+      "loss": 0.8906,
+      "step": 2351
+    },
+    {
+      "epoch": 0.12945126314051406,
+      "grad_norm": 1.0356301069259644,
+      "learning_rate": 9.905012123213838e-06,
+      "loss": 0.8051,
+      "step": 2352
+    },
+    {
+      "epoch": 0.12950630194286972,
+      "grad_norm": 0.8546886444091797,
+      "learning_rate": 9.904928014220945e-06,
+      "loss": 0.7543,
+      "step": 2353
+    },
+    {
+      "epoch": 0.12956134074522538,
+      "grad_norm": 0.9229897856712341,
+      "learning_rate": 9.904843868363927e-06,
+      "loss": 0.8823,
+      "step": 2354
+    },
+    {
+      "epoch": 0.12961637954758104,
+      "grad_norm": 0.8364199995994568,
+      "learning_rate": 9.904759685643414e-06,
+      "loss": 0.8825,
+      "step": 2355
+    },
+    {
+      "epoch": 0.1296714183499367,
+      "grad_norm": 0.9092077016830444,
+      "learning_rate": 9.90467546606004e-06,
+      "loss": 0.8721,
+      "step": 2356
+    },
+    {
+      "epoch": 0.12972645715229236,
+      "grad_norm": 1.042973518371582,
+      "learning_rate": 9.904591209614441e-06,
+      "loss": 0.7984,
+      "step": 2357
+    },
+    {
+      "epoch": 0.12978149595464802,
+      "grad_norm": 0.7262618541717529,
+      "learning_rate": 9.904506916307243e-06,
+      "loss": 0.6721,
+      "step": 2358
+    },
+    {
+      "epoch": 0.12983653475700369,
+      "grad_norm": 0.7562826871871948,
+      "learning_rate": 9.904422586139086e-06,
+      "loss": 0.7702,
+      "step": 2359
+    },
+    {
+      "epoch": 0.12989157355935935,
+      "grad_norm": 0.8821595907211304,
+      "learning_rate": 9.904338219110603e-06,
+      "loss": 0.8555,
+      "step": 2360
+    },
+    {
+      "epoch": 0.129946612361715,
+      "grad_norm": 1.0340098142623901,
+      "learning_rate": 9.904253815222424e-06,
+      "loss": 0.9004,
+      "step": 2361
+    },
+    {
+      "epoch": 0.13000165116407067,
+      "grad_norm": 0.8533693552017212,
+      "learning_rate": 9.904169374475188e-06,
+      "loss": 0.836,
+      "step": 2362
+    },
+    {
+      "epoch": 0.13005668996642633,
+      "grad_norm": 0.8564199805259705,
+      "learning_rate": 9.904084896869528e-06,
+      "loss": 0.9281,
+      "step": 2363
+    },
+    {
+      "epoch": 0.130111728768782,
+      "grad_norm": 0.7817538976669312,
+      "learning_rate": 9.904000382406079e-06,
+      "loss": 0.7444,
+      "step": 2364
+    },
+    {
+      "epoch": 0.13016676757113765,
+      "grad_norm": 1.1420893669128418,
+      "learning_rate": 9.903915831085473e-06,
+      "loss": 0.9116,
+      "step": 2365
+    },
+    {
+      "epoch": 0.13022180637349332,
+      "grad_norm": 0.9671920537948608,
+      "learning_rate": 9.903831242908351e-06,
+      "loss": 0.899,
+      "step": 2366
+    },
+    {
+      "epoch": 0.13027684517584898,
+      "grad_norm": 0.8528717756271362,
+      "learning_rate": 9.903746617875345e-06,
+      "loss": 0.7231,
+      "step": 2367
+    },
+    {
+      "epoch": 0.13033188397820464,
+      "grad_norm": 0.786960244178772,
+      "learning_rate": 9.903661955987091e-06,
+      "loss": 0.7997,
+      "step": 2368
+    },
+    {
+      "epoch": 0.1303869227805603,
+      "grad_norm": 0.941683292388916,
+      "learning_rate": 9.903577257244228e-06,
+      "loss": 0.9127,
+      "step": 2369
+    },
+    {
+      "epoch": 0.13044196158291596,
+      "grad_norm": 0.886900007724762,
+      "learning_rate": 9.903492521647391e-06,
+      "loss": 0.9086,
+      "step": 2370
+    },
+    {
+      "epoch": 0.13049700038527162,
+      "grad_norm": 0.9924801588058472,
+      "learning_rate": 9.903407749197216e-06,
+      "loss": 0.9055,
+      "step": 2371
+    },
+    {
+      "epoch": 0.13055203918762728,
+      "grad_norm": 0.6998724341392517,
+      "learning_rate": 9.903322939894342e-06,
+      "loss": 0.6972,
+      "step": 2372
+    },
+    {
+      "epoch": 0.13060707798998294,
+      "grad_norm": 0.8448702096939087,
+      "learning_rate": 9.903238093739404e-06,
+      "loss": 0.7862,
+      "step": 2373
+    },
+    {
+      "epoch": 0.1306621167923386,
+      "grad_norm": 0.8557441830635071,
+      "learning_rate": 9.90315321073304e-06,
+      "loss": 0.8364,
+      "step": 2374
+    },
+    {
+      "epoch": 0.13071715559469427,
+      "grad_norm": 0.7978441119194031,
+      "learning_rate": 9.903068290875892e-06,
+      "loss": 0.7671,
+      "step": 2375
+    },
+    {
+      "epoch": 0.13077219439704993,
+      "grad_norm": 0.781315803527832,
+      "learning_rate": 9.902983334168594e-06,
+      "loss": 0.7963,
+      "step": 2376
+    },
+    {
+      "epoch": 0.1308272331994056,
+      "grad_norm": 0.7326155304908752,
+      "learning_rate": 9.902898340611785e-06,
+      "loss": 0.8,
+      "step": 2377
+    },
+    {
+      "epoch": 0.13088227200176125,
+      "grad_norm": 0.7693139314651489,
+      "learning_rate": 9.902813310206105e-06,
+      "loss": 0.8459,
+      "step": 2378
+    },
+    {
+      "epoch": 0.1309373108041169,
+      "grad_norm": 0.9441308975219727,
+      "learning_rate": 9.902728242952191e-06,
+      "loss": 0.8519,
+      "step": 2379
+    },
+    {
+      "epoch": 0.13099234960647257,
+      "grad_norm": 0.8350616693496704,
+      "learning_rate": 9.902643138850686e-06,
+      "loss": 0.876,
+      "step": 2380
+    },
+    {
+      "epoch": 0.13104738840882824,
+      "grad_norm": 0.8675554394721985,
+      "learning_rate": 9.902557997902227e-06,
+      "loss": 0.8172,
+      "step": 2381
+    },
+    {
+      "epoch": 0.1311024272111839,
+      "grad_norm": 0.9618930220603943,
+      "learning_rate": 9.902472820107454e-06,
+      "loss": 0.8852,
+      "step": 2382
+    },
+    {
+      "epoch": 0.13115746601353956,
+      "grad_norm": 0.862341046333313,
+      "learning_rate": 9.902387605467007e-06,
+      "loss": 0.9256,
+      "step": 2383
+    },
+    {
+      "epoch": 0.1312125048158952,
+      "grad_norm": 0.8749859929084778,
+      "learning_rate": 9.902302353981527e-06,
+      "loss": 0.8809,
+      "step": 2384
+    },
+    {
+      "epoch": 0.13126754361825085,
+      "grad_norm": 0.9061958193778992,
+      "learning_rate": 9.902217065651657e-06,
+      "loss": 0.779,
+      "step": 2385
+    },
+    {
+      "epoch": 0.13132258242060652,
+      "grad_norm": 0.8909298777580261,
+      "learning_rate": 9.902131740478033e-06,
+      "loss": 0.8203,
+      "step": 2386
+    },
+    {
+      "epoch": 0.13137762122296218,
+      "grad_norm": 0.8507269024848938,
+      "learning_rate": 9.902046378461302e-06,
+      "loss": 0.776,
+      "step": 2387
+    },
+    {
+      "epoch": 0.13143266002531784,
+      "grad_norm": 0.9577299356460571,
+      "learning_rate": 9.901960979602101e-06,
+      "loss": 0.8104,
+      "step": 2388
+    },
+    {
+      "epoch": 0.1314876988276735,
+      "grad_norm": 0.9244948625564575,
+      "learning_rate": 9.901875543901074e-06,
+      "loss": 0.9035,
+      "step": 2389
+    },
+    {
+      "epoch": 0.13154273763002916,
+      "grad_norm": 0.7534334063529968,
+      "learning_rate": 9.901790071358861e-06,
+      "loss": 0.7262,
+      "step": 2390
+    },
+    {
+      "epoch": 0.13159777643238482,
+      "grad_norm": 0.8920090198516846,
+      "learning_rate": 9.901704561976106e-06,
+      "loss": 0.932,
+      "step": 2391
+    },
+    {
+      "epoch": 0.13165281523474048,
+      "grad_norm": 0.8524243235588074,
+      "learning_rate": 9.901619015753455e-06,
+      "loss": 0.8107,
+      "step": 2392
+    },
+    {
+      "epoch": 0.13170785403709614,
+      "grad_norm": 0.8170381784439087,
+      "learning_rate": 9.901533432691543e-06,
+      "loss": 0.8814,
+      "step": 2393
+    },
+    {
+      "epoch": 0.1317628928394518,
+      "grad_norm": 0.8281697034835815,
+      "learning_rate": 9.90144781279102e-06,
+      "loss": 0.8221,
+      "step": 2394
+    },
+    {
+      "epoch": 0.13181793164180747,
+      "grad_norm": 0.9283351302146912,
+      "learning_rate": 9.901362156052528e-06,
+      "loss": 0.8346,
+      "step": 2395
+    },
+    {
+      "epoch": 0.13187297044416313,
+      "grad_norm": 0.8331275582313538,
+      "learning_rate": 9.901276462476708e-06,
+      "loss": 0.7498,
+      "step": 2396
+    },
+    {
+      "epoch": 0.1319280092465188,
+      "grad_norm": 0.8427191972732544,
+      "learning_rate": 9.901190732064207e-06,
+      "loss": 0.8265,
+      "step": 2397
+    },
+    {
+      "epoch": 0.13198304804887445,
+      "grad_norm": 0.8510351777076721,
+      "learning_rate": 9.901104964815669e-06,
+      "loss": 0.8369,
+      "step": 2398
+    },
+    {
+      "epoch": 0.1320380868512301,
+      "grad_norm": 0.8468914031982422,
+      "learning_rate": 9.901019160731738e-06,
+      "loss": 0.8585,
+      "step": 2399
+    },
+    {
+      "epoch": 0.13209312565358577,
+      "grad_norm": 0.8302182555198669,
+      "learning_rate": 9.900933319813058e-06,
+      "loss": 0.8611,
+      "step": 2400
+    },
+    {
+      "epoch": 0.13214816445594144,
+      "grad_norm": 0.8527448773384094,
+      "learning_rate": 9.900847442060277e-06,
+      "loss": 0.899,
+      "step": 2401
+    },
+    {
+      "epoch": 0.1322032032582971,
+      "grad_norm": 0.8354688286781311,
+      "learning_rate": 9.900761527474037e-06,
+      "loss": 0.8083,
+      "step": 2402
+    },
+    {
+      "epoch": 0.13225824206065276,
+      "grad_norm": 0.8612173795700073,
+      "learning_rate": 9.900675576054986e-06,
+      "loss": 0.8124,
+      "step": 2403
+    },
+    {
+      "epoch": 0.13231328086300842,
+      "grad_norm": 0.7424876689910889,
+      "learning_rate": 9.900589587803767e-06,
+      "loss": 0.6884,
+      "step": 2404
+    },
+    {
+      "epoch": 0.13236831966536408,
+      "grad_norm": 0.8431115746498108,
+      "learning_rate": 9.90050356272103e-06,
+      "loss": 0.9575,
+      "step": 2405
+    },
+    {
+      "epoch": 0.13242335846771974,
+      "grad_norm": 0.7958092093467712,
+      "learning_rate": 9.90041750080742e-06,
+      "loss": 0.7608,
+      "step": 2406
+    },
+    {
+      "epoch": 0.1324783972700754,
+      "grad_norm": 0.926258385181427,
+      "learning_rate": 9.900331402063583e-06,
+      "loss": 0.9072,
+      "step": 2407
+    },
+    {
+      "epoch": 0.13253343607243107,
+      "grad_norm": 0.7952526807785034,
+      "learning_rate": 9.900245266490169e-06,
+      "loss": 0.8001,
+      "step": 2408
+    },
+    {
+      "epoch": 0.13258847487478673,
+      "grad_norm": 0.8309933543205261,
+      "learning_rate": 9.900159094087822e-06,
+      "loss": 0.9154,
+      "step": 2409
+    },
+    {
+      "epoch": 0.1326435136771424,
+      "grad_norm": 0.858007550239563,
+      "learning_rate": 9.90007288485719e-06,
+      "loss": 0.855,
+      "step": 2410
+    },
+    {
+      "epoch": 0.13269855247949805,
+      "grad_norm": 0.9513822197914124,
+      "learning_rate": 9.899986638798923e-06,
+      "loss": 0.8162,
+      "step": 2411
+    },
+    {
+      "epoch": 0.1327535912818537,
+      "grad_norm": 0.8387427926063538,
+      "learning_rate": 9.899900355913668e-06,
+      "loss": 0.8955,
+      "step": 2412
+    },
+    {
+      "epoch": 0.13280863008420937,
+      "grad_norm": 0.7727940678596497,
+      "learning_rate": 9.899814036202073e-06,
+      "loss": 0.6765,
+      "step": 2413
+    },
+    {
+      "epoch": 0.13286366888656503,
+      "grad_norm": 0.7760928869247437,
+      "learning_rate": 9.899727679664788e-06,
+      "loss": 0.7179,
+      "step": 2414
+    },
+    {
+      "epoch": 0.1329187076889207,
+      "grad_norm": 0.7798073887825012,
+      "learning_rate": 9.899641286302462e-06,
+      "loss": 0.8541,
+      "step": 2415
+    },
+    {
+      "epoch": 0.13297374649127636,
+      "grad_norm": 0.8302769660949707,
+      "learning_rate": 9.899554856115743e-06,
+      "loss": 0.8925,
+      "step": 2416
+    },
+    {
+      "epoch": 0.13302878529363202,
+      "grad_norm": 0.8300751447677612,
+      "learning_rate": 9.89946838910528e-06,
+      "loss": 0.7489,
+      "step": 2417
+    },
+    {
+      "epoch": 0.13308382409598768,
+      "grad_norm": 0.8032094240188599,
+      "learning_rate": 9.899381885271725e-06,
+      "loss": 0.811,
+      "step": 2418
+    },
+    {
+      "epoch": 0.13313886289834334,
+      "grad_norm": 5.237870216369629,
+      "learning_rate": 9.899295344615727e-06,
+      "loss": 0.7609,
+      "step": 2419
+    },
+    {
+      "epoch": 0.133193901700699,
+      "grad_norm": 0.8145740628242493,
+      "learning_rate": 9.899208767137935e-06,
+      "loss": 0.8435,
+      "step": 2420
+    },
+    {
+      "epoch": 0.13324894050305466,
+      "grad_norm": 0.9716018438339233,
+      "learning_rate": 9.899122152839004e-06,
+      "loss": 0.7924,
+      "step": 2421
+    },
+    {
+      "epoch": 0.13330397930541033,
+      "grad_norm": 0.7846183776855469,
+      "learning_rate": 9.899035501719582e-06,
+      "loss": 0.8941,
+      "step": 2422
+    },
+    {
+      "epoch": 0.133359018107766,
+      "grad_norm": 0.7653689980506897,
+      "learning_rate": 9.89894881378032e-06,
+      "loss": 0.811,
+      "step": 2423
+    },
+    {
+      "epoch": 0.13341405691012165,
+      "grad_norm": 0.8221875429153442,
+      "learning_rate": 9.89886208902187e-06,
+      "loss": 0.8131,
+      "step": 2424
+    },
+    {
+      "epoch": 0.1334690957124773,
+      "grad_norm": 0.7422335147857666,
+      "learning_rate": 9.898775327444885e-06,
+      "loss": 0.6366,
+      "step": 2425
+    },
+    {
+      "epoch": 0.13352413451483297,
+      "grad_norm": 0.8072695136070251,
+      "learning_rate": 9.898688529050014e-06,
+      "loss": 0.7989,
+      "step": 2426
+    },
+    {
+      "epoch": 0.1335791733171886,
+      "grad_norm": 0.7717600464820862,
+      "learning_rate": 9.898601693837911e-06,
+      "loss": 0.7524,
+      "step": 2427
+    },
+    {
+      "epoch": 0.13363421211954427,
+      "grad_norm": 0.8070919513702393,
+      "learning_rate": 9.898514821809231e-06,
+      "loss": 0.7724,
+      "step": 2428
+    },
+    {
+      "epoch": 0.13368925092189993,
+      "grad_norm": 0.8184726238250732,
+      "learning_rate": 9.898427912964624e-06,
+      "loss": 0.845,
+      "step": 2429
+    },
+    {
+      "epoch": 0.1337442897242556,
+      "grad_norm": 0.8168759346008301,
+      "learning_rate": 9.898340967304744e-06,
+      "loss": 0.8377,
+      "step": 2430
+    },
+    {
+      "epoch": 0.13379932852661125,
+      "grad_norm": 0.8701872825622559,
+      "learning_rate": 9.898253984830244e-06,
+      "loss": 0.908,
+      "step": 2431
+    },
+    {
+      "epoch": 0.1338543673289669,
+      "grad_norm": 0.8092133402824402,
+      "learning_rate": 9.898166965541779e-06,
+      "loss": 0.866,
+      "step": 2432
+    },
+    {
+      "epoch": 0.13390940613132257,
+      "grad_norm": 0.8337095975875854,
+      "learning_rate": 9.898079909440002e-06,
+      "loss": 0.8622,
+      "step": 2433
+    },
+    {
+      "epoch": 0.13396444493367823,
+      "grad_norm": 1.1016209125518799,
+      "learning_rate": 9.897992816525567e-06,
+      "loss": 0.8486,
+      "step": 2434
+    },
+    {
+      "epoch": 0.1340194837360339,
+      "grad_norm": 0.8136518597602844,
+      "learning_rate": 9.89790568679913e-06,
+      "loss": 0.8681,
+      "step": 2435
+    },
+    {
+      "epoch": 0.13407452253838956,
+      "grad_norm": 0.8202341794967651,
+      "learning_rate": 9.897818520261344e-06,
+      "loss": 0.9144,
+      "step": 2436
+    },
+    {
+      "epoch": 0.13412956134074522,
+      "grad_norm": 0.8836861848831177,
+      "learning_rate": 9.897731316912866e-06,
+      "loss": 0.8643,
+      "step": 2437
+    },
+    {
+      "epoch": 0.13418460014310088,
+      "grad_norm": 0.9040210247039795,
+      "learning_rate": 9.89764407675435e-06,
+      "loss": 0.7681,
+      "step": 2438
+    },
+    {
+      "epoch": 0.13423963894545654,
+      "grad_norm": 0.8762359619140625,
+      "learning_rate": 9.897556799786452e-06,
+      "loss": 0.8765,
+      "step": 2439
+    },
+    {
+      "epoch": 0.1342946777478122,
+      "grad_norm": 0.8859462738037109,
+      "learning_rate": 9.897469486009827e-06,
+      "loss": 0.9051,
+      "step": 2440
+    },
+    {
+      "epoch": 0.13434971655016786,
+      "grad_norm": 0.7727539539337158,
+      "learning_rate": 9.897382135425134e-06,
+      "loss": 0.7397,
+      "step": 2441
+    },
+    {
+      "epoch": 0.13440475535252353,
+      "grad_norm": 0.9018967151641846,
+      "learning_rate": 9.897294748033028e-06,
+      "loss": 0.8542,
+      "step": 2442
+    },
+    {
+      "epoch": 0.1344597941548792,
+      "grad_norm": 0.8228337168693542,
+      "learning_rate": 9.897207323834165e-06,
+      "loss": 0.7585,
+      "step": 2443
+    },
+    {
+      "epoch": 0.13451483295723485,
+      "grad_norm": 0.7509974241256714,
+      "learning_rate": 9.897119862829203e-06,
+      "loss": 0.7285,
+      "step": 2444
+    },
+    {
+      "epoch": 0.1345698717595905,
+      "grad_norm": 0.9225835800170898,
+      "learning_rate": 9.897032365018797e-06,
+      "loss": 0.8352,
+      "step": 2445
+    },
+    {
+      "epoch": 0.13462491056194617,
+      "grad_norm": 0.800981879234314,
+      "learning_rate": 9.896944830403609e-06,
+      "loss": 0.7352,
+      "step": 2446
+    },
+    {
+      "epoch": 0.13467994936430183,
+      "grad_norm": 0.8263673186302185,
+      "learning_rate": 9.896857258984294e-06,
+      "loss": 0.8426,
+      "step": 2447
+    },
+    {
+      "epoch": 0.1347349881666575,
+      "grad_norm": 0.8857110738754272,
+      "learning_rate": 9.89676965076151e-06,
+      "loss": 0.8078,
+      "step": 2448
+    },
+    {
+      "epoch": 0.13479002696901315,
+      "grad_norm": 0.8637158274650574,
+      "learning_rate": 9.896682005735916e-06,
+      "loss": 0.8688,
+      "step": 2449
+    },
+    {
+      "epoch": 0.13484506577136882,
+      "grad_norm": 0.9050095081329346,
+      "learning_rate": 9.89659432390817e-06,
+      "loss": 0.831,
+      "step": 2450
+    },
+    {
+      "epoch": 0.13490010457372448,
+      "grad_norm": 0.829757034778595,
+      "learning_rate": 9.896506605278933e-06,
+      "loss": 0.8095,
+      "step": 2451
+    },
+    {
+      "epoch": 0.13495514337608014,
+      "grad_norm": 0.8910449743270874,
+      "learning_rate": 9.896418849848864e-06,
+      "loss": 0.9134,
+      "step": 2452
+    },
+    {
+      "epoch": 0.1350101821784358,
+      "grad_norm": 0.8856307864189148,
+      "learning_rate": 9.89633105761862e-06,
+      "loss": 0.8171,
+      "step": 2453
+    },
+    {
+      "epoch": 0.13506522098079146,
+      "grad_norm": 0.8159938454627991,
+      "learning_rate": 9.896243228588864e-06,
+      "loss": 0.8205,
+      "step": 2454
+    },
+    {
+      "epoch": 0.13512025978314712,
+      "grad_norm": 0.8200929760932922,
+      "learning_rate": 9.896155362760254e-06,
+      "loss": 0.7529,
+      "step": 2455
+    },
+    {
+      "epoch": 0.13517529858550278,
+      "grad_norm": 0.7591279149055481,
+      "learning_rate": 9.89606746013345e-06,
+      "loss": 0.8205,
+      "step": 2456
+    },
+    {
+      "epoch": 0.13523033738785845,
+      "grad_norm": 0.8598676323890686,
+      "learning_rate": 9.895979520709114e-06,
+      "loss": 0.8212,
+      "step": 2457
+    },
+    {
+      "epoch": 0.1352853761902141,
+      "grad_norm": 0.7290365099906921,
+      "learning_rate": 9.895891544487905e-06,
+      "loss": 0.7893,
+      "step": 2458
+    },
+    {
+      "epoch": 0.13534041499256977,
+      "grad_norm": 0.8040594458580017,
+      "learning_rate": 9.895803531470487e-06,
+      "loss": 0.8358,
+      "step": 2459
+    },
+    {
+      "epoch": 0.13539545379492543,
+      "grad_norm": 0.9286525249481201,
+      "learning_rate": 9.895715481657522e-06,
+      "loss": 0.8104,
+      "step": 2460
+    },
+    {
+      "epoch": 0.1354504925972811,
+      "grad_norm": 0.843054473400116,
+      "learning_rate": 9.895627395049668e-06,
+      "loss": 0.7872,
+      "step": 2461
+    },
+    {
+      "epoch": 0.13550553139963675,
+      "grad_norm": 0.7894387245178223,
+      "learning_rate": 9.895539271647588e-06,
+      "loss": 0.8615,
+      "step": 2462
+    },
+    {
+      "epoch": 0.13556057020199241,
+      "grad_norm": 0.9185294508934021,
+      "learning_rate": 9.895451111451948e-06,
+      "loss": 0.8732,
+      "step": 2463
+    },
+    {
+      "epoch": 0.13561560900434808,
+      "grad_norm": 0.8586474657058716,
+      "learning_rate": 9.895362914463405e-06,
+      "loss": 0.9658,
+      "step": 2464
+    },
+    {
+      "epoch": 0.13567064780670374,
+      "grad_norm": 0.8810474276542664,
+      "learning_rate": 9.895274680682628e-06,
+      "loss": 0.8622,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1357256866090594,
+      "grad_norm": 0.8862990736961365,
+      "learning_rate": 9.895186410110273e-06,
+      "loss": 0.916,
+      "step": 2466
+    },
+    {
+      "epoch": 0.13578072541141506,
+      "grad_norm": 0.7916743159294128,
+      "learning_rate": 9.89509810274701e-06,
+      "loss": 0.837,
+      "step": 2467
+    },
+    {
+      "epoch": 0.13583576421377072,
+      "grad_norm": 0.9063515663146973,
+      "learning_rate": 9.8950097585935e-06,
+      "loss": 0.8065,
+      "step": 2468
+    },
+    {
+      "epoch": 0.13589080301612638,
+      "grad_norm": 0.7656043767929077,
+      "learning_rate": 9.894921377650405e-06,
+      "loss": 0.7064,
+      "step": 2469
+    },
+    {
+      "epoch": 0.13594584181848202,
+      "grad_norm": 1.0630278587341309,
+      "learning_rate": 9.894832959918392e-06,
+      "loss": 0.8168,
+      "step": 2470
+    },
+    {
+      "epoch": 0.13600088062083768,
+      "grad_norm": 0.9118956923484802,
+      "learning_rate": 9.894744505398126e-06,
+      "loss": 0.8972,
+      "step": 2471
+    },
+    {
+      "epoch": 0.13605591942319334,
+      "grad_norm": 0.8989213705062866,
+      "learning_rate": 9.89465601409027e-06,
+      "loss": 0.8374,
+      "step": 2472
+    },
+    {
+      "epoch": 0.136110958225549,
+      "grad_norm": 0.9398229718208313,
+      "learning_rate": 9.894567485995489e-06,
+      "loss": 0.8956,
+      "step": 2473
+    },
+    {
+      "epoch": 0.13616599702790466,
+      "grad_norm": 0.7980280518531799,
+      "learning_rate": 9.894478921114449e-06,
+      "loss": 0.8055,
+      "step": 2474
+    },
+    {
+      "epoch": 0.13622103583026032,
+      "grad_norm": 0.8910034894943237,
+      "learning_rate": 9.894390319447816e-06,
+      "loss": 0.8371,
+      "step": 2475
+    },
+    {
+      "epoch": 0.13627607463261598,
+      "grad_norm": 0.7848070859909058,
+      "learning_rate": 9.894301680996255e-06,
+      "loss": 0.8024,
+      "step": 2476
+    },
+    {
+      "epoch": 0.13633111343497165,
+      "grad_norm": 0.8538175821304321,
+      "learning_rate": 9.894213005760434e-06,
+      "loss": 0.8819,
+      "step": 2477
+    },
+    {
+      "epoch": 0.1363861522373273,
+      "grad_norm": 0.7885367274284363,
+      "learning_rate": 9.894124293741017e-06,
+      "loss": 0.7916,
+      "step": 2478
+    },
+    {
+      "epoch": 0.13644119103968297,
+      "grad_norm": 0.8555673956871033,
+      "learning_rate": 9.894035544938672e-06,
+      "loss": 0.8521,
+      "step": 2479
+    },
+    {
+      "epoch": 0.13649622984203863,
+      "grad_norm": 0.8104771971702576,
+      "learning_rate": 9.893946759354066e-06,
+      "loss": 0.8437,
+      "step": 2480
+    },
+    {
+      "epoch": 0.1365512686443943,
+      "grad_norm": 0.9131864309310913,
+      "learning_rate": 9.893857936987866e-06,
+      "loss": 0.8123,
+      "step": 2481
+    },
+    {
+      "epoch": 0.13660630744674995,
+      "grad_norm": 0.9414293766021729,
+      "learning_rate": 9.893769077840739e-06,
+      "loss": 0.7897,
+      "step": 2482
+    },
+    {
+      "epoch": 0.13666134624910561,
+      "grad_norm": 0.823265016078949,
+      "learning_rate": 9.893680181913355e-06,
+      "loss": 0.847,
+      "step": 2483
+    },
+    {
+      "epoch": 0.13671638505146128,
+      "grad_norm": 0.82098788022995,
+      "learning_rate": 9.89359124920638e-06,
+      "loss": 0.7823,
+      "step": 2484
+    },
+    {
+      "epoch": 0.13677142385381694,
+      "grad_norm": 0.817551851272583,
+      "learning_rate": 9.893502279720483e-06,
+      "loss": 0.8084,
+      "step": 2485
+    },
+    {
+      "epoch": 0.1368264626561726,
+      "grad_norm": 1.0722150802612305,
+      "learning_rate": 9.893413273456333e-06,
+      "loss": 0.7394,
+      "step": 2486
+    },
+    {
+      "epoch": 0.13688150145852826,
+      "grad_norm": 0.8045433759689331,
+      "learning_rate": 9.893324230414598e-06,
+      "loss": 0.7528,
+      "step": 2487
+    },
+    {
+      "epoch": 0.13693654026088392,
+      "grad_norm": 0.8694071173667908,
+      "learning_rate": 9.893235150595949e-06,
+      "loss": 0.803,
+      "step": 2488
+    },
+    {
+      "epoch": 0.13699157906323958,
+      "grad_norm": 0.8238615989685059,
+      "learning_rate": 9.893146034001054e-06,
+      "loss": 0.7909,
+      "step": 2489
+    },
+    {
+      "epoch": 0.13704661786559524,
+      "grad_norm": 0.7782405018806458,
+      "learning_rate": 9.893056880630583e-06,
+      "loss": 0.6859,
+      "step": 2490
+    },
+    {
+      "epoch": 0.1371016566679509,
+      "grad_norm": 0.7865599989891052,
+      "learning_rate": 9.892967690485207e-06,
+      "loss": 0.7982,
+      "step": 2491
+    },
+    {
+      "epoch": 0.13715669547030657,
+      "grad_norm": 0.768120288848877,
+      "learning_rate": 9.892878463565595e-06,
+      "loss": 0.8234,
+      "step": 2492
+    },
+    {
+      "epoch": 0.13721173427266223,
+      "grad_norm": 0.812493085861206,
+      "learning_rate": 9.89278919987242e-06,
+      "loss": 0.9152,
+      "step": 2493
+    },
+    {
+      "epoch": 0.1372667730750179,
+      "grad_norm": 0.7256335616111755,
+      "learning_rate": 9.892699899406348e-06,
+      "loss": 0.6703,
+      "step": 2494
+    },
+    {
+      "epoch": 0.13732181187737355,
+      "grad_norm": 0.8022804260253906,
+      "learning_rate": 9.892610562168054e-06,
+      "loss": 0.7918,
+      "step": 2495
+    },
+    {
+      "epoch": 0.1373768506797292,
+      "grad_norm": 0.8204907774925232,
+      "learning_rate": 9.89252118815821e-06,
+      "loss": 0.9094,
+      "step": 2496
+    },
+    {
+      "epoch": 0.13743188948208487,
+      "grad_norm": 0.9986788630485535,
+      "learning_rate": 9.892431777377484e-06,
+      "loss": 0.8921,
+      "step": 2497
+    },
+    {
+      "epoch": 0.13748692828444053,
+      "grad_norm": 0.7937983870506287,
+      "learning_rate": 9.892342329826554e-06,
+      "loss": 0.8048,
+      "step": 2498
+    },
+    {
+      "epoch": 0.1375419670867962,
+      "grad_norm": 0.9295744895935059,
+      "learning_rate": 9.892252845506086e-06,
+      "loss": 0.755,
+      "step": 2499
+    },
+    {
+      "epoch": 0.13759700588915186,
+      "grad_norm": 0.7920984625816345,
+      "learning_rate": 9.892163324416757e-06,
+      "loss": 0.7603,
+      "step": 2500
+    },
+    {
+      "epoch": 0.13765204469150752,
+      "grad_norm": 0.9229464530944824,
+      "learning_rate": 9.892073766559236e-06,
+      "loss": 0.8115,
+      "step": 2501
+    },
+    {
+      "epoch": 0.13770708349386318,
+      "grad_norm": 0.8205353021621704,
+      "learning_rate": 9.8919841719342e-06,
+      "loss": 0.8357,
+      "step": 2502
+    },
+    {
+      "epoch": 0.13776212229621884,
+      "grad_norm": 0.86461341381073,
+      "learning_rate": 9.891894540542318e-06,
+      "loss": 0.748,
+      "step": 2503
+    },
+    {
+      "epoch": 0.1378171610985745,
+      "grad_norm": 0.767145574092865,
+      "learning_rate": 9.891804872384267e-06,
+      "loss": 0.7404,
+      "step": 2504
+    },
+    {
+      "epoch": 0.13787219990093016,
+      "grad_norm": 0.7492040991783142,
+      "learning_rate": 9.891715167460721e-06,
+      "loss": 0.6958,
+      "step": 2505
+    },
+    {
+      "epoch": 0.13792723870328583,
+      "grad_norm": 0.8643150329589844,
+      "learning_rate": 9.891625425772353e-06,
+      "loss": 0.8408,
+      "step": 2506
+    },
+    {
+      "epoch": 0.1379822775056415,
+      "grad_norm": 0.8026981353759766,
+      "learning_rate": 9.891535647319838e-06,
+      "loss": 0.7895,
+      "step": 2507
+    },
+    {
+      "epoch": 0.13803731630799715,
+      "grad_norm": 1.2780394554138184,
+      "learning_rate": 9.89144583210385e-06,
+      "loss": 0.9113,
+      "step": 2508
+    },
+    {
+      "epoch": 0.1380923551103528,
+      "grad_norm": 0.8476191163063049,
+      "learning_rate": 9.891355980125064e-06,
+      "loss": 0.8224,
+      "step": 2509
+    },
+    {
+      "epoch": 0.13814739391270847,
+      "grad_norm": 1.048682689666748,
+      "learning_rate": 9.891266091384157e-06,
+      "loss": 0.8913,
+      "step": 2510
+    },
+    {
+      "epoch": 0.13820243271506413,
+      "grad_norm": 1.0314993858337402,
+      "learning_rate": 9.891176165881801e-06,
+      "loss": 0.8315,
+      "step": 2511
+    },
+    {
+      "epoch": 0.1382574715174198,
+      "grad_norm": 0.9500058889389038,
+      "learning_rate": 9.891086203618676e-06,
+      "loss": 0.9185,
+      "step": 2512
+    },
+    {
+      "epoch": 0.13831251031977543,
+      "grad_norm": 0.7860653400421143,
+      "learning_rate": 9.890996204595457e-06,
+      "loss": 0.804,
+      "step": 2513
+    },
+    {
+      "epoch": 0.1383675491221311,
+      "grad_norm": 0.8354741930961609,
+      "learning_rate": 9.89090616881282e-06,
+      "loss": 0.8214,
+      "step": 2514
+    },
+    {
+      "epoch": 0.13842258792448675,
+      "grad_norm": 0.9115905165672302,
+      "learning_rate": 9.890816096271438e-06,
+      "loss": 0.8801,
+      "step": 2515
+    },
+    {
+      "epoch": 0.1384776267268424,
+      "grad_norm": 0.8852075338363647,
+      "learning_rate": 9.890725986971994e-06,
+      "loss": 0.8821,
+      "step": 2516
+    },
+    {
+      "epoch": 0.13853266552919807,
+      "grad_norm": 0.804314374923706,
+      "learning_rate": 9.890635840915164e-06,
+      "loss": 0.8412,
+      "step": 2517
+    },
+    {
+      "epoch": 0.13858770433155373,
+      "grad_norm": 0.8242805600166321,
+      "learning_rate": 9.890545658101623e-06,
+      "loss": 0.8447,
+      "step": 2518
+    },
+    {
+      "epoch": 0.1386427431339094,
+      "grad_norm": 0.8385655879974365,
+      "learning_rate": 9.890455438532048e-06,
+      "loss": 0.8161,
+      "step": 2519
+    },
+    {
+      "epoch": 0.13869778193626506,
+      "grad_norm": 0.7950524687767029,
+      "learning_rate": 9.89036518220712e-06,
+      "loss": 0.8024,
+      "step": 2520
+    },
+    {
+      "epoch": 0.13875282073862072,
+      "grad_norm": 1.0031861066818237,
+      "learning_rate": 9.890274889127518e-06,
+      "loss": 0.8399,
+      "step": 2521
+    },
+    {
+      "epoch": 0.13880785954097638,
+      "grad_norm": 0.8403242230415344,
+      "learning_rate": 9.890184559293917e-06,
+      "loss": 0.8115,
+      "step": 2522
+    },
+    {
+      "epoch": 0.13886289834333204,
+      "grad_norm": 0.8389976024627686,
+      "learning_rate": 9.890094192706998e-06,
+      "loss": 0.9573,
+      "step": 2523
+    },
+    {
+      "epoch": 0.1389179371456877,
+      "grad_norm": 0.8408516645431519,
+      "learning_rate": 9.890003789367442e-06,
+      "loss": 0.8572,
+      "step": 2524
+    },
+    {
+      "epoch": 0.13897297594804336,
+      "grad_norm": 0.7607787251472473,
+      "learning_rate": 9.889913349275925e-06,
+      "loss": 0.8119,
+      "step": 2525
+    },
+    {
+      "epoch": 0.13902801475039903,
+      "grad_norm": 0.7696373462677002,
+      "learning_rate": 9.889822872433127e-06,
+      "loss": 0.8287,
+      "step": 2526
+    },
+    {
+      "epoch": 0.1390830535527547,
+      "grad_norm": 0.8518380522727966,
+      "learning_rate": 9.889732358839732e-06,
+      "loss": 0.9008,
+      "step": 2527
+    },
+    {
+      "epoch": 0.13913809235511035,
+      "grad_norm": 0.8851314783096313,
+      "learning_rate": 9.889641808496416e-06,
+      "loss": 0.8148,
+      "step": 2528
+    },
+    {
+      "epoch": 0.139193131157466,
+      "grad_norm": 0.9245797395706177,
+      "learning_rate": 9.889551221403862e-06,
+      "loss": 0.846,
+      "step": 2529
+    },
+    {
+      "epoch": 0.13924816995982167,
+      "grad_norm": 0.8445762991905212,
+      "learning_rate": 9.889460597562748e-06,
+      "loss": 0.8306,
+      "step": 2530
+    },
+    {
+      "epoch": 0.13930320876217733,
+      "grad_norm": 0.9149277806282043,
+      "learning_rate": 9.88936993697376e-06,
+      "loss": 0.8033,
+      "step": 2531
+    },
+    {
+      "epoch": 0.139358247564533,
+      "grad_norm": 0.894666850566864,
+      "learning_rate": 9.889279239637572e-06,
+      "loss": 0.8299,
+      "step": 2532
+    },
+    {
+      "epoch": 0.13941328636688866,
+      "grad_norm": 1.2897371053695679,
+      "learning_rate": 9.889188505554871e-06,
+      "loss": 0.7776,
+      "step": 2533
+    },
+    {
+      "epoch": 0.13946832516924432,
+      "grad_norm": 0.8927022218704224,
+      "learning_rate": 9.889097734726341e-06,
+      "loss": 0.8706,
+      "step": 2534
+    },
+    {
+      "epoch": 0.13952336397159998,
+      "grad_norm": 0.7688571214675903,
+      "learning_rate": 9.889006927152658e-06,
+      "loss": 0.8191,
+      "step": 2535
+    },
+    {
+      "epoch": 0.13957840277395564,
+      "grad_norm": 0.926671028137207,
+      "learning_rate": 9.88891608283451e-06,
+      "loss": 0.7489,
+      "step": 2536
+    },
+    {
+      "epoch": 0.1396334415763113,
+      "grad_norm": 0.8316965699195862,
+      "learning_rate": 9.888825201772577e-06,
+      "loss": 0.7783,
+      "step": 2537
+    },
+    {
+      "epoch": 0.13968848037866696,
+      "grad_norm": 0.8619750738143921,
+      "learning_rate": 9.88873428396754e-06,
+      "loss": 0.8269,
+      "step": 2538
+    },
+    {
+      "epoch": 0.13974351918102262,
+      "grad_norm": 0.8588540554046631,
+      "learning_rate": 9.888643329420086e-06,
+      "loss": 0.8133,
+      "step": 2539
+    },
+    {
+      "epoch": 0.13979855798337829,
+      "grad_norm": 0.7947841882705688,
+      "learning_rate": 9.8885523381309e-06,
+      "loss": 0.8041,
+      "step": 2540
+    },
+    {
+      "epoch": 0.13985359678573395,
+      "grad_norm": 0.8440257906913757,
+      "learning_rate": 9.888461310100661e-06,
+      "loss": 0.8324,
+      "step": 2541
+    },
+    {
+      "epoch": 0.1399086355880896,
+      "grad_norm": 0.7842260003089905,
+      "learning_rate": 9.888370245330055e-06,
+      "loss": 0.8031,
+      "step": 2542
+    },
+    {
+      "epoch": 0.13996367439044527,
+      "grad_norm": 0.8108223080635071,
+      "learning_rate": 9.888279143819768e-06,
+      "loss": 0.7998,
+      "step": 2543
+    },
+    {
+      "epoch": 0.14001871319280093,
+      "grad_norm": 0.9748625159263611,
+      "learning_rate": 9.888188005570482e-06,
+      "loss": 0.9553,
+      "step": 2544
+    },
+    {
+      "epoch": 0.1400737519951566,
+      "grad_norm": 0.8465562462806702,
+      "learning_rate": 9.888096830582883e-06,
+      "loss": 0.7884,
+      "step": 2545
+    },
+    {
+      "epoch": 0.14012879079751225,
+      "grad_norm": 0.9339833855628967,
+      "learning_rate": 9.88800561885766e-06,
+      "loss": 0.8135,
+      "step": 2546
+    },
+    {
+      "epoch": 0.14018382959986792,
+      "grad_norm": 0.7749297022819519,
+      "learning_rate": 9.887914370395492e-06,
+      "loss": 0.8411,
+      "step": 2547
+    },
+    {
+      "epoch": 0.14023886840222358,
+      "grad_norm": 0.862606942653656,
+      "learning_rate": 9.887823085197068e-06,
+      "loss": 0.7631,
+      "step": 2548
+    },
+    {
+      "epoch": 0.14029390720457924,
+      "grad_norm": 1.3383793830871582,
+      "learning_rate": 9.887731763263076e-06,
+      "loss": 0.7979,
+      "step": 2549
+    },
+    {
+      "epoch": 0.1403489460069349,
+      "grad_norm": 0.8092008233070374,
+      "learning_rate": 9.887640404594199e-06,
+      "loss": 0.7566,
+      "step": 2550
+    },
+    {
+      "epoch": 0.14040398480929056,
+      "grad_norm": 0.9233745336532593,
+      "learning_rate": 9.887549009191126e-06,
+      "loss": 0.8954,
+      "step": 2551
+    },
+    {
+      "epoch": 0.14045902361164622,
+      "grad_norm": 0.8533664345741272,
+      "learning_rate": 9.887457577054542e-06,
+      "loss": 0.8311,
+      "step": 2552
+    },
+    {
+      "epoch": 0.14051406241400188,
+      "grad_norm": 0.7679287791252136,
+      "learning_rate": 9.887366108185135e-06,
+      "loss": 0.7641,
+      "step": 2553
+    },
+    {
+      "epoch": 0.14056910121635754,
+      "grad_norm": 0.7998354434967041,
+      "learning_rate": 9.887274602583594e-06,
+      "loss": 0.7759,
+      "step": 2554
+    },
+    {
+      "epoch": 0.1406241400187132,
+      "grad_norm": 0.8877138495445251,
+      "learning_rate": 9.887183060250605e-06,
+      "loss": 0.8928,
+      "step": 2555
+    },
+    {
+      "epoch": 0.14067917882106884,
+      "grad_norm": 0.8022066354751587,
+      "learning_rate": 9.887091481186855e-06,
+      "loss": 0.8233,
+      "step": 2556
+    },
+    {
+      "epoch": 0.1407342176234245,
+      "grad_norm": 0.8419097065925598,
+      "learning_rate": 9.886999865393035e-06,
+      "loss": 0.8044,
+      "step": 2557
+    },
+    {
+      "epoch": 0.14078925642578016,
+      "grad_norm": 0.9581286311149597,
+      "learning_rate": 9.88690821286983e-06,
+      "loss": 0.8531,
+      "step": 2558
+    },
+    {
+      "epoch": 0.14084429522813582,
+      "grad_norm": 0.894851803779602,
+      "learning_rate": 9.886816523617933e-06,
+      "loss": 0.8594,
+      "step": 2559
+    },
+    {
+      "epoch": 0.14089933403049149,
+      "grad_norm": 0.7813432812690735,
+      "learning_rate": 9.886724797638032e-06,
+      "loss": 0.7311,
+      "step": 2560
+    },
+    {
+      "epoch": 0.14095437283284715,
+      "grad_norm": 0.8194118142127991,
+      "learning_rate": 9.886633034930814e-06,
+      "loss": 0.8067,
+      "step": 2561
+    },
+    {
+      "epoch": 0.1410094116352028,
+      "grad_norm": 0.8091121912002563,
+      "learning_rate": 9.88654123549697e-06,
+      "loss": 0.7558,
+      "step": 2562
+    },
+    {
+      "epoch": 0.14106445043755847,
+      "grad_norm": 0.8334764242172241,
+      "learning_rate": 9.88644939933719e-06,
+      "loss": 0.8375,
+      "step": 2563
+    },
+    {
+      "epoch": 0.14111948923991413,
+      "grad_norm": 0.8283817768096924,
+      "learning_rate": 9.886357526452166e-06,
+      "loss": 0.7839,
+      "step": 2564
+    },
+    {
+      "epoch": 0.1411745280422698,
+      "grad_norm": 0.8708772659301758,
+      "learning_rate": 9.886265616842585e-06,
+      "loss": 0.8193,
+      "step": 2565
+    },
+    {
+      "epoch": 0.14122956684462545,
+      "grad_norm": 0.9883641600608826,
+      "learning_rate": 9.886173670509141e-06,
+      "loss": 0.9409,
+      "step": 2566
+    },
+    {
+      "epoch": 0.14128460564698112,
+      "grad_norm": 0.8601766228675842,
+      "learning_rate": 9.886081687452523e-06,
+      "loss": 0.9391,
+      "step": 2567
+    },
+    {
+      "epoch": 0.14133964444933678,
+      "grad_norm": 0.8729620575904846,
+      "learning_rate": 9.885989667673422e-06,
+      "loss": 0.8372,
+      "step": 2568
+    },
+    {
+      "epoch": 0.14139468325169244,
+      "grad_norm": 0.7899564504623413,
+      "learning_rate": 9.885897611172532e-06,
+      "loss": 0.7773,
+      "step": 2569
+    },
+    {
+      "epoch": 0.1414497220540481,
+      "grad_norm": 0.8120512962341309,
+      "learning_rate": 9.885805517950542e-06,
+      "loss": 0.887,
+      "step": 2570
+    },
+    {
+      "epoch": 0.14150476085640376,
+      "grad_norm": 0.8475256562232971,
+      "learning_rate": 9.885713388008148e-06,
+      "loss": 0.7935,
+      "step": 2571
+    },
+    {
+      "epoch": 0.14155979965875942,
+      "grad_norm": 0.7669919729232788,
+      "learning_rate": 9.885621221346038e-06,
+      "loss": 0.7728,
+      "step": 2572
+    },
+    {
+      "epoch": 0.14161483846111508,
+      "grad_norm": 0.8298916220664978,
+      "learning_rate": 9.885529017964906e-06,
+      "loss": 0.7723,
+      "step": 2573
+    },
+    {
+      "epoch": 0.14166987726347074,
+      "grad_norm": 0.8630721569061279,
+      "learning_rate": 9.885436777865447e-06,
+      "loss": 0.8395,
+      "step": 2574
+    },
+    {
+      "epoch": 0.1417249160658264,
+      "grad_norm": 0.7566008567810059,
+      "learning_rate": 9.885344501048352e-06,
+      "loss": 0.806,
+      "step": 2575
+    },
+    {
+      "epoch": 0.14177995486818207,
+      "grad_norm": 0.7870769500732422,
+      "learning_rate": 9.885252187514316e-06,
+      "loss": 0.7683,
+      "step": 2576
+    },
+    {
+      "epoch": 0.14183499367053773,
+      "grad_norm": 0.879648745059967,
+      "learning_rate": 9.885159837264033e-06,
+      "loss": 0.8472,
+      "step": 2577
+    },
+    {
+      "epoch": 0.1418900324728934,
+      "grad_norm": 0.76839280128479,
+      "learning_rate": 9.885067450298196e-06,
+      "loss": 0.8534,
+      "step": 2578
+    },
+    {
+      "epoch": 0.14194507127524905,
+      "grad_norm": 0.8268701434135437,
+      "learning_rate": 9.884975026617498e-06,
+      "loss": 0.7799,
+      "step": 2579
+    },
+    {
+      "epoch": 0.1420001100776047,
+      "grad_norm": 0.8226090669631958,
+      "learning_rate": 9.884882566222638e-06,
+      "loss": 0.6756,
+      "step": 2580
+    },
+    {
+      "epoch": 0.14205514887996037,
+      "grad_norm": 0.8299756050109863,
+      "learning_rate": 9.884790069114307e-06,
+      "loss": 0.734,
+      "step": 2581
+    },
+    {
+      "epoch": 0.14211018768231604,
+      "grad_norm": 0.8241812586784363,
+      "learning_rate": 9.8846975352932e-06,
+      "loss": 0.8335,
+      "step": 2582
+    },
+    {
+      "epoch": 0.1421652264846717,
+      "grad_norm": 0.8458926677703857,
+      "learning_rate": 9.884604964760016e-06,
+      "loss": 0.7376,
+      "step": 2583
+    },
+    {
+      "epoch": 0.14222026528702736,
+      "grad_norm": 0.876966655254364,
+      "learning_rate": 9.884512357515447e-06,
+      "loss": 0.9414,
+      "step": 2584
+    },
+    {
+      "epoch": 0.14227530408938302,
+      "grad_norm": 0.770252525806427,
+      "learning_rate": 9.88441971356019e-06,
+      "loss": 0.8312,
+      "step": 2585
+    },
+    {
+      "epoch": 0.14233034289173868,
+      "grad_norm": 0.7883023023605347,
+      "learning_rate": 9.884327032894945e-06,
+      "loss": 0.8568,
+      "step": 2586
+    },
+    {
+      "epoch": 0.14238538169409434,
+      "grad_norm": 0.9092289209365845,
+      "learning_rate": 9.884234315520405e-06,
+      "loss": 0.9078,
+      "step": 2587
+    },
+    {
+      "epoch": 0.14244042049645,
+      "grad_norm": 0.7946531176567078,
+      "learning_rate": 9.884141561437266e-06,
+      "loss": 0.6895,
+      "step": 2588
+    },
+    {
+      "epoch": 0.14249545929880567,
+      "grad_norm": 0.7791070342063904,
+      "learning_rate": 9.884048770646227e-06,
+      "loss": 0.6984,
+      "step": 2589
+    },
+    {
+      "epoch": 0.14255049810116133,
+      "grad_norm": 0.7775537371635437,
+      "learning_rate": 9.883955943147982e-06,
+      "loss": 0.7568,
+      "step": 2590
+    },
+    {
+      "epoch": 0.142605536903517,
+      "grad_norm": 0.7735158801078796,
+      "learning_rate": 9.883863078943234e-06,
+      "loss": 0.8215,
+      "step": 2591
+    },
+    {
+      "epoch": 0.14266057570587265,
+      "grad_norm": 0.881365180015564,
+      "learning_rate": 9.88377017803268e-06,
+      "loss": 0.8817,
+      "step": 2592
+    },
+    {
+      "epoch": 0.1427156145082283,
+      "grad_norm": 0.8643443584442139,
+      "learning_rate": 9.883677240417014e-06,
+      "loss": 0.8024,
+      "step": 2593
+    },
+    {
+      "epoch": 0.14277065331058397,
+      "grad_norm": 0.885713517665863,
+      "learning_rate": 9.883584266096938e-06,
+      "loss": 0.7612,
+      "step": 2594
+    },
+    {
+      "epoch": 0.14282569211293963,
+      "grad_norm": 0.771340012550354,
+      "learning_rate": 9.88349125507315e-06,
+      "loss": 0.8293,
+      "step": 2595
+    },
+    {
+      "epoch": 0.1428807309152953,
+      "grad_norm": 0.8284093737602234,
+      "learning_rate": 9.88339820734635e-06,
+      "loss": 0.8539,
+      "step": 2596
+    },
+    {
+      "epoch": 0.14293576971765096,
+      "grad_norm": 0.9597725868225098,
+      "learning_rate": 9.883305122917233e-06,
+      "loss": 0.9054,
+      "step": 2597
+    },
+    {
+      "epoch": 0.14299080852000662,
+      "grad_norm": 0.7552937269210815,
+      "learning_rate": 9.883212001786504e-06,
+      "loss": 0.8047,
+      "step": 2598
+    },
+    {
+      "epoch": 0.14304584732236225,
+      "grad_norm": 0.8008492588996887,
+      "learning_rate": 9.883118843954861e-06,
+      "loss": 0.802,
+      "step": 2599
+    },
+    {
+      "epoch": 0.1431008861247179,
+      "grad_norm": 0.8169753551483154,
+      "learning_rate": 9.883025649423003e-06,
+      "loss": 0.8837,
+      "step": 2600
+    },
+    {
+      "epoch": 0.14315592492707357,
+      "grad_norm": 0.8521036505699158,
+      "learning_rate": 9.882932418191632e-06,
+      "loss": 0.8266,
+      "step": 2601
+    },
+    {
+      "epoch": 0.14321096372942924,
+      "grad_norm": 0.8647341728210449,
+      "learning_rate": 9.882839150261449e-06,
+      "loss": 0.8949,
+      "step": 2602
+    },
+    {
+      "epoch": 0.1432660025317849,
+      "grad_norm": 0.9236162304878235,
+      "learning_rate": 9.882745845633153e-06,
+      "loss": 0.8474,
+      "step": 2603
+    },
+    {
+      "epoch": 0.14332104133414056,
+      "grad_norm": 0.8422677516937256,
+      "learning_rate": 9.882652504307445e-06,
+      "loss": 0.8396,
+      "step": 2604
+    },
+    {
+      "epoch": 0.14337608013649622,
+      "grad_norm": 0.902036190032959,
+      "learning_rate": 9.88255912628503e-06,
+      "loss": 0.8075,
+      "step": 2605
+    },
+    {
+      "epoch": 0.14343111893885188,
+      "grad_norm": 0.8972339630126953,
+      "learning_rate": 9.882465711566605e-06,
+      "loss": 0.8143,
+      "step": 2606
+    },
+    {
+      "epoch": 0.14348615774120754,
+      "grad_norm": 0.8025243282318115,
+      "learning_rate": 9.882372260152877e-06,
+      "loss": 0.771,
+      "step": 2607
+    },
+    {
+      "epoch": 0.1435411965435632,
+      "grad_norm": 0.8260911107063293,
+      "learning_rate": 9.882278772044545e-06,
+      "loss": 0.7679,
+      "step": 2608
+    },
+    {
+      "epoch": 0.14359623534591887,
+      "grad_norm": 0.8069774508476257,
+      "learning_rate": 9.882185247242313e-06,
+      "loss": 0.8489,
+      "step": 2609
+    },
+    {
+      "epoch": 0.14365127414827453,
+      "grad_norm": 0.8702567219734192,
+      "learning_rate": 9.882091685746883e-06,
+      "loss": 0.9258,
+      "step": 2610
+    },
+    {
+      "epoch": 0.1437063129506302,
+      "grad_norm": 0.8841683268547058,
+      "learning_rate": 9.881998087558959e-06,
+      "loss": 0.7858,
+      "step": 2611
+    },
+    {
+      "epoch": 0.14376135175298585,
+      "grad_norm": 0.7302986979484558,
+      "learning_rate": 9.881904452679246e-06,
+      "loss": 0.7339,
+      "step": 2612
+    },
+    {
+      "epoch": 0.1438163905553415,
+      "grad_norm": 0.7852466106414795,
+      "learning_rate": 9.881810781108442e-06,
+      "loss": 0.8397,
+      "step": 2613
+    },
+    {
+      "epoch": 0.14387142935769717,
+      "grad_norm": 0.7986249327659607,
+      "learning_rate": 9.881717072847258e-06,
+      "loss": 0.7573,
+      "step": 2614
+    },
+    {
+      "epoch": 0.14392646816005283,
+      "grad_norm": 0.750000536441803,
+      "learning_rate": 9.881623327896395e-06,
+      "loss": 0.7128,
+      "step": 2615
+    },
+    {
+      "epoch": 0.1439815069624085,
+      "grad_norm": 0.8796436786651611,
+      "learning_rate": 9.881529546256557e-06,
+      "loss": 0.9364,
+      "step": 2616
+    },
+    {
+      "epoch": 0.14403654576476416,
+      "grad_norm": 0.8621297478675842,
+      "learning_rate": 9.881435727928449e-06,
+      "loss": 0.9323,
+      "step": 2617
+    },
+    {
+      "epoch": 0.14409158456711982,
+      "grad_norm": 0.8213173151016235,
+      "learning_rate": 9.881341872912777e-06,
+      "loss": 0.7746,
+      "step": 2618
+    },
+    {
+      "epoch": 0.14414662336947548,
+      "grad_norm": 0.7761938571929932,
+      "learning_rate": 9.881247981210247e-06,
+      "loss": 0.8065,
+      "step": 2619
+    },
+    {
+      "epoch": 0.14420166217183114,
+      "grad_norm": 0.8333988785743713,
+      "learning_rate": 9.881154052821564e-06,
+      "loss": 0.8727,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1442567009741868,
+      "grad_norm": 0.7263909578323364,
+      "learning_rate": 9.881060087747433e-06,
+      "loss": 0.8194,
+      "step": 2621
+    },
+    {
+      "epoch": 0.14431173977654246,
+      "grad_norm": 0.7472667098045349,
+      "learning_rate": 9.880966085988562e-06,
+      "loss": 0.77,
+      "step": 2622
+    },
+    {
+      "epoch": 0.14436677857889813,
+      "grad_norm": 0.7999943494796753,
+      "learning_rate": 9.880872047545656e-06,
+      "loss": 0.7936,
+      "step": 2623
+    },
+    {
+      "epoch": 0.1444218173812538,
+      "grad_norm": 0.8359610438346863,
+      "learning_rate": 9.88077797241942e-06,
+      "loss": 0.7946,
+      "step": 2624
+    },
+    {
+      "epoch": 0.14447685618360945,
+      "grad_norm": 0.8666403889656067,
+      "learning_rate": 9.880683860610566e-06,
+      "loss": 0.8152,
+      "step": 2625
+    },
+    {
+      "epoch": 0.1445318949859651,
+      "grad_norm": 0.7883741855621338,
+      "learning_rate": 9.880589712119797e-06,
+      "loss": 0.7972,
+      "step": 2626
+    },
+    {
+      "epoch": 0.14458693378832077,
+      "grad_norm": 0.8048827648162842,
+      "learning_rate": 9.880495526947824e-06,
+      "loss": 0.8221,
+      "step": 2627
+    },
+    {
+      "epoch": 0.14464197259067643,
+      "grad_norm": 0.718292236328125,
+      "learning_rate": 9.88040130509535e-06,
+      "loss": 0.7648,
+      "step": 2628
+    },
+    {
+      "epoch": 0.1446970113930321,
+      "grad_norm": 0.7748421430587769,
+      "learning_rate": 9.880307046563088e-06,
+      "loss": 0.8146,
+      "step": 2629
+    },
+    {
+      "epoch": 0.14475205019538775,
+      "grad_norm": 0.8015987873077393,
+      "learning_rate": 9.880212751351745e-06,
+      "loss": 0.7935,
+      "step": 2630
+    },
+    {
+      "epoch": 0.14480708899774342,
+      "grad_norm": 0.7628459930419922,
+      "learning_rate": 9.88011841946203e-06,
+      "loss": 0.7469,
+      "step": 2631
+    },
+    {
+      "epoch": 0.14486212780009908,
+      "grad_norm": 0.7152888774871826,
+      "learning_rate": 9.88002405089465e-06,
+      "loss": 0.7721,
+      "step": 2632
+    },
+    {
+      "epoch": 0.14491716660245474,
+      "grad_norm": 0.8075545430183411,
+      "learning_rate": 9.879929645650315e-06,
+      "loss": 0.8799,
+      "step": 2633
+    },
+    {
+      "epoch": 0.1449722054048104,
+      "grad_norm": 0.7981964945793152,
+      "learning_rate": 9.879835203729736e-06,
+      "loss": 0.8265,
+      "step": 2634
+    },
+    {
+      "epoch": 0.14502724420716606,
+      "grad_norm": 0.7699866890907288,
+      "learning_rate": 9.879740725133623e-06,
+      "loss": 0.8489,
+      "step": 2635
+    },
+    {
+      "epoch": 0.14508228300952172,
+      "grad_norm": 0.7991634011268616,
+      "learning_rate": 9.879646209862682e-06,
+      "loss": 0.8754,
+      "step": 2636
+    },
+    {
+      "epoch": 0.14513732181187738,
+      "grad_norm": 0.8284991383552551,
+      "learning_rate": 9.879551657917628e-06,
+      "loss": 0.811,
+      "step": 2637
+    },
+    {
+      "epoch": 0.14519236061423305,
+      "grad_norm": 0.9189227819442749,
+      "learning_rate": 9.87945706929917e-06,
+      "loss": 0.8486,
+      "step": 2638
+    },
+    {
+      "epoch": 0.1452473994165887,
+      "grad_norm": 0.8599026799201965,
+      "learning_rate": 9.879362444008018e-06,
+      "loss": 0.8383,
+      "step": 2639
+    },
+    {
+      "epoch": 0.14530243821894437,
+      "grad_norm": 0.8764603137969971,
+      "learning_rate": 9.879267782044885e-06,
+      "loss": 0.7918,
+      "step": 2640
+    },
+    {
+      "epoch": 0.14535747702130003,
+      "grad_norm": 0.8061341047286987,
+      "learning_rate": 9.87917308341048e-06,
+      "loss": 0.8292,
+      "step": 2641
+    },
+    {
+      "epoch": 0.14541251582365566,
+      "grad_norm": 1.031220555305481,
+      "learning_rate": 9.879078348105518e-06,
+      "loss": 0.6612,
+      "step": 2642
+    },
+    {
+      "epoch": 0.14546755462601133,
+      "grad_norm": 1.014491319656372,
+      "learning_rate": 9.878983576130708e-06,
+      "loss": 0.8512,
+      "step": 2643
+    },
+    {
+      "epoch": 0.145522593428367,
+      "grad_norm": 0.8365896940231323,
+      "learning_rate": 9.878888767486764e-06,
+      "loss": 0.7995,
+      "step": 2644
+    },
+    {
+      "epoch": 0.14557763223072265,
+      "grad_norm": 0.8086197972297668,
+      "learning_rate": 9.878793922174397e-06,
+      "loss": 0.8069,
+      "step": 2645
+    },
+    {
+      "epoch": 0.1456326710330783,
+      "grad_norm": 0.8075234889984131,
+      "learning_rate": 9.878699040194322e-06,
+      "loss": 0.8415,
+      "step": 2646
+    },
+    {
+      "epoch": 0.14568770983543397,
+      "grad_norm": 0.9413748979568481,
+      "learning_rate": 9.87860412154725e-06,
+      "loss": 0.7811,
+      "step": 2647
+    },
+    {
+      "epoch": 0.14574274863778963,
+      "grad_norm": 0.7744552493095398,
+      "learning_rate": 9.878509166233895e-06,
+      "loss": 0.7983,
+      "step": 2648
+    },
+    {
+      "epoch": 0.1457977874401453,
+      "grad_norm": 0.8184664845466614,
+      "learning_rate": 9.878414174254974e-06,
+      "loss": 0.8052,
+      "step": 2649
+    },
+    {
+      "epoch": 0.14585282624250095,
+      "grad_norm": 0.928814172744751,
+      "learning_rate": 9.878319145611195e-06,
+      "loss": 0.7695,
+      "step": 2650
+    },
+    {
+      "epoch": 0.14590786504485662,
+      "grad_norm": 0.9623318314552307,
+      "learning_rate": 9.878224080303276e-06,
+      "loss": 0.9025,
+      "step": 2651
+    },
+    {
+      "epoch": 0.14596290384721228,
+      "grad_norm": 0.866538405418396,
+      "learning_rate": 9.87812897833193e-06,
+      "loss": 0.7895,
+      "step": 2652
+    },
+    {
+      "epoch": 0.14601794264956794,
+      "grad_norm": 0.9248599410057068,
+      "learning_rate": 9.878033839697874e-06,
+      "loss": 0.8532,
+      "step": 2653
+    },
+    {
+      "epoch": 0.1460729814519236,
+      "grad_norm": 0.7866301536560059,
+      "learning_rate": 9.87793866440182e-06,
+      "loss": 0.8724,
+      "step": 2654
+    },
+    {
+      "epoch": 0.14612802025427926,
+      "grad_norm": 0.8471634387969971,
+      "learning_rate": 9.877843452444485e-06,
+      "loss": 0.9184,
+      "step": 2655
+    },
+    {
+      "epoch": 0.14618305905663492,
+      "grad_norm": 0.7367103695869446,
+      "learning_rate": 9.877748203826585e-06,
+      "loss": 0.7328,
+      "step": 2656
+    },
+    {
+      "epoch": 0.14623809785899058,
+      "grad_norm": 0.95980304479599,
+      "learning_rate": 9.877652918548834e-06,
+      "loss": 0.9274,
+      "step": 2657
+    },
+    {
+      "epoch": 0.14629313666134625,
+      "grad_norm": 1.0511064529418945,
+      "learning_rate": 9.87755759661195e-06,
+      "loss": 0.8223,
+      "step": 2658
+    },
+    {
+      "epoch": 0.1463481754637019,
+      "grad_norm": 0.7616510391235352,
+      "learning_rate": 9.877462238016649e-06,
+      "loss": 0.7473,
+      "step": 2659
+    },
+    {
+      "epoch": 0.14640321426605757,
+      "grad_norm": 0.7814056873321533,
+      "learning_rate": 9.877366842763647e-06,
+      "loss": 0.8898,
+      "step": 2660
+    },
+    {
+      "epoch": 0.14645825306841323,
+      "grad_norm": 0.8707298636436462,
+      "learning_rate": 9.877271410853662e-06,
+      "loss": 0.8792,
+      "step": 2661
+    },
+    {
+      "epoch": 0.1465132918707689,
+      "grad_norm": 0.8618701696395874,
+      "learning_rate": 9.877175942287409e-06,
+      "loss": 0.8761,
+      "step": 2662
+    },
+    {
+      "epoch": 0.14656833067312455,
+      "grad_norm": 0.9437732100486755,
+      "learning_rate": 9.877080437065609e-06,
+      "loss": 0.7922,
+      "step": 2663
+    },
+    {
+      "epoch": 0.14662336947548021,
+      "grad_norm": 0.9465780258178711,
+      "learning_rate": 9.876984895188976e-06,
+      "loss": 0.8449,
+      "step": 2664
+    },
+    {
+      "epoch": 0.14667840827783588,
+      "grad_norm": 0.7149911522865295,
+      "learning_rate": 9.876889316658233e-06,
+      "loss": 0.6408,
+      "step": 2665
+    },
+    {
+      "epoch": 0.14673344708019154,
+      "grad_norm": 0.9996811151504517,
+      "learning_rate": 9.876793701474092e-06,
+      "loss": 0.9324,
+      "step": 2666
+    },
+    {
+      "epoch": 0.1467884858825472,
+      "grad_norm": 0.7941329479217529,
+      "learning_rate": 9.876698049637277e-06,
+      "loss": 0.8115,
+      "step": 2667
+    },
+    {
+      "epoch": 0.14684352468490286,
+      "grad_norm": 0.754175066947937,
+      "learning_rate": 9.876602361148504e-06,
+      "loss": 0.7709,
+      "step": 2668
+    },
+    {
+      "epoch": 0.14689856348725852,
+      "grad_norm": 0.7867946624755859,
+      "learning_rate": 9.876506636008494e-06,
+      "loss": 0.8578,
+      "step": 2669
+    },
+    {
+      "epoch": 0.14695360228961418,
+      "grad_norm": 0.7441185116767883,
+      "learning_rate": 9.876410874217965e-06,
+      "loss": 0.8491,
+      "step": 2670
+    },
+    {
+      "epoch": 0.14700864109196984,
+      "grad_norm": 0.8414027690887451,
+      "learning_rate": 9.876315075777638e-06,
+      "loss": 0.8404,
+      "step": 2671
+    },
+    {
+      "epoch": 0.1470636798943255,
+      "grad_norm": 0.7911489009857178,
+      "learning_rate": 9.876219240688231e-06,
+      "loss": 0.8606,
+      "step": 2672
+    },
+    {
+      "epoch": 0.14711871869668117,
+      "grad_norm": 0.8601381778717041,
+      "learning_rate": 9.876123368950465e-06,
+      "loss": 0.7753,
+      "step": 2673
+    },
+    {
+      "epoch": 0.14717375749903683,
+      "grad_norm": 0.8672378659248352,
+      "learning_rate": 9.876027460565062e-06,
+      "loss": 0.7763,
+      "step": 2674
+    },
+    {
+      "epoch": 0.1472287963013925,
+      "grad_norm": 0.7192933559417725,
+      "learning_rate": 9.875931515532742e-06,
+      "loss": 0.7681,
+      "step": 2675
+    },
+    {
+      "epoch": 0.14728383510374815,
+      "grad_norm": 0.7483426332473755,
+      "learning_rate": 9.875835533854226e-06,
+      "loss": 0.8129,
+      "step": 2676
+    },
+    {
+      "epoch": 0.1473388739061038,
+      "grad_norm": 0.8883694410324097,
+      "learning_rate": 9.875739515530235e-06,
+      "loss": 0.8912,
+      "step": 2677
+    },
+    {
+      "epoch": 0.14739391270845947,
+      "grad_norm": 0.8440148234367371,
+      "learning_rate": 9.87564346056149e-06,
+      "loss": 0.8411,
+      "step": 2678
+    },
+    {
+      "epoch": 0.14744895151081513,
+      "grad_norm": 0.8916668891906738,
+      "learning_rate": 9.875547368948715e-06,
+      "loss": 0.8484,
+      "step": 2679
+    },
+    {
+      "epoch": 0.1475039903131708,
+      "grad_norm": 0.805258572101593,
+      "learning_rate": 9.875451240692631e-06,
+      "loss": 0.8172,
+      "step": 2680
+    },
+    {
+      "epoch": 0.14755902911552646,
+      "grad_norm": 0.8322305679321289,
+      "learning_rate": 9.87535507579396e-06,
+      "loss": 0.809,
+      "step": 2681
+    },
+    {
+      "epoch": 0.14761406791788212,
+      "grad_norm": 0.7320597767829895,
+      "learning_rate": 9.875258874253424e-06,
+      "loss": 0.7346,
+      "step": 2682
+    },
+    {
+      "epoch": 0.14766910672023778,
+      "grad_norm": 1.018036127090454,
+      "learning_rate": 9.875162636071749e-06,
+      "loss": 0.931,
+      "step": 2683
+    },
+    {
+      "epoch": 0.14772414552259344,
+      "grad_norm": 0.8601503968238831,
+      "learning_rate": 9.875066361249657e-06,
+      "loss": 0.7689,
+      "step": 2684
+    },
+    {
+      "epoch": 0.14777918432494908,
+      "grad_norm": 0.8478472232818604,
+      "learning_rate": 9.87497004978787e-06,
+      "loss": 0.9545,
+      "step": 2685
+    },
+    {
+      "epoch": 0.14783422312730474,
+      "grad_norm": 0.7510890364646912,
+      "learning_rate": 9.874873701687115e-06,
+      "loss": 0.7794,
+      "step": 2686
+    },
+    {
+      "epoch": 0.1478892619296604,
+      "grad_norm": 0.8226999044418335,
+      "learning_rate": 9.874777316948112e-06,
+      "loss": 0.8477,
+      "step": 2687
+    },
+    {
+      "epoch": 0.14794430073201606,
+      "grad_norm": 0.8284991979598999,
+      "learning_rate": 9.874680895571588e-06,
+      "loss": 0.8498,
+      "step": 2688
+    },
+    {
+      "epoch": 0.14799933953437172,
+      "grad_norm": 0.9007356762886047,
+      "learning_rate": 9.874584437558267e-06,
+      "loss": 0.8526,
+      "step": 2689
+    },
+    {
+      "epoch": 0.14805437833672738,
+      "grad_norm": 0.8770126104354858,
+      "learning_rate": 9.874487942908877e-06,
+      "loss": 0.844,
+      "step": 2690
+    },
+    {
+      "epoch": 0.14810941713908304,
+      "grad_norm": 1.1561466455459595,
+      "learning_rate": 9.874391411624138e-06,
+      "loss": 0.976,
+      "step": 2691
+    },
+    {
+      "epoch": 0.1481644559414387,
+      "grad_norm": 0.8162640929222107,
+      "learning_rate": 9.874294843704777e-06,
+      "loss": 0.8581,
+      "step": 2692
+    },
+    {
+      "epoch": 0.14821949474379437,
+      "grad_norm": 0.8308132290840149,
+      "learning_rate": 9.874198239151522e-06,
+      "loss": 0.8303,
+      "step": 2693
+    },
+    {
+      "epoch": 0.14827453354615003,
+      "grad_norm": 0.771024227142334,
+      "learning_rate": 9.874101597965098e-06,
+      "loss": 0.8351,
+      "step": 2694
+    },
+    {
+      "epoch": 0.1483295723485057,
+      "grad_norm": 0.7588162422180176,
+      "learning_rate": 9.874004920146232e-06,
+      "loss": 0.7858,
+      "step": 2695
+    },
+    {
+      "epoch": 0.14838461115086135,
+      "grad_norm": 0.8282446265220642,
+      "learning_rate": 9.873908205695648e-06,
+      "loss": 0.8465,
+      "step": 2696
+    },
+    {
+      "epoch": 0.148439649953217,
+      "grad_norm": 0.8342786431312561,
+      "learning_rate": 9.873811454614076e-06,
+      "loss": 0.8688,
+      "step": 2697
+    },
+    {
+      "epoch": 0.14849468875557267,
+      "grad_norm": 0.7957108020782471,
+      "learning_rate": 9.87371466690224e-06,
+      "loss": 0.8381,
+      "step": 2698
+    },
+    {
+      "epoch": 0.14854972755792833,
+      "grad_norm": 0.8763726353645325,
+      "learning_rate": 9.87361784256087e-06,
+      "loss": 0.8922,
+      "step": 2699
+    },
+    {
+      "epoch": 0.148604766360284,
+      "grad_norm": 0.7760055661201477,
+      "learning_rate": 9.873520981590693e-06,
+      "loss": 0.8384,
+      "step": 2700
+    },
+    {
+      "epoch": 0.14865980516263966,
+      "grad_norm": 0.9691097736358643,
+      "learning_rate": 9.873424083992436e-06,
+      "loss": 0.8581,
+      "step": 2701
+    },
+    {
+      "epoch": 0.14871484396499532,
+      "grad_norm": 0.9072558879852295,
+      "learning_rate": 9.87332714976683e-06,
+      "loss": 0.8942,
+      "step": 2702
+    },
+    {
+      "epoch": 0.14876988276735098,
+      "grad_norm": 0.8961714506149292,
+      "learning_rate": 9.8732301789146e-06,
+      "loss": 0.8062,
+      "step": 2703
+    },
+    {
+      "epoch": 0.14882492156970664,
+      "grad_norm": 1.4835050106048584,
+      "learning_rate": 9.873133171436477e-06,
+      "loss": 0.886,
+      "step": 2704
+    },
+    {
+      "epoch": 0.1488799603720623,
+      "grad_norm": 0.8153702616691589,
+      "learning_rate": 9.87303612733319e-06,
+      "loss": 0.8369,
+      "step": 2705
+    },
+    {
+      "epoch": 0.14893499917441796,
+      "grad_norm": 0.8755800724029541,
+      "learning_rate": 9.872939046605467e-06,
+      "loss": 0.7591,
+      "step": 2706
+    },
+    {
+      "epoch": 0.14899003797677363,
+      "grad_norm": 0.8173243403434753,
+      "learning_rate": 9.872841929254038e-06,
+      "loss": 0.8626,
+      "step": 2707
+    },
+    {
+      "epoch": 0.1490450767791293,
+      "grad_norm": 0.7891639471054077,
+      "learning_rate": 9.872744775279634e-06,
+      "loss": 0.737,
+      "step": 2708
+    },
+    {
+      "epoch": 0.14910011558148495,
+      "grad_norm": 1.0270631313323975,
+      "learning_rate": 9.872647584682985e-06,
+      "loss": 0.9202,
+      "step": 2709
+    },
+    {
+      "epoch": 0.1491551543838406,
+      "grad_norm": 0.7736123204231262,
+      "learning_rate": 9.872550357464822e-06,
+      "loss": 0.7835,
+      "step": 2710
+    },
+    {
+      "epoch": 0.14921019318619627,
+      "grad_norm": 0.7791550159454346,
+      "learning_rate": 9.872453093625873e-06,
+      "loss": 0.8375,
+      "step": 2711
+    },
+    {
+      "epoch": 0.14926523198855193,
+      "grad_norm": 0.8410583734512329,
+      "learning_rate": 9.872355793166872e-06,
+      "loss": 0.877,
+      "step": 2712
+    },
+    {
+      "epoch": 0.1493202707909076,
+      "grad_norm": 0.8277738094329834,
+      "learning_rate": 9.87225845608855e-06,
+      "loss": 0.7255,
+      "step": 2713
+    },
+    {
+      "epoch": 0.14937530959326326,
+      "grad_norm": 0.8617290258407593,
+      "learning_rate": 9.872161082391635e-06,
+      "loss": 0.7885,
+      "step": 2714
+    },
+    {
+      "epoch": 0.14943034839561892,
+      "grad_norm": 0.8866406679153442,
+      "learning_rate": 9.872063672076864e-06,
+      "loss": 0.8621,
+      "step": 2715
+    },
+    {
+      "epoch": 0.14948538719797458,
+      "grad_norm": 0.7581049799919128,
+      "learning_rate": 9.871966225144964e-06,
+      "loss": 0.8177,
+      "step": 2716
+    },
+    {
+      "epoch": 0.14954042600033024,
+      "grad_norm": 0.833696722984314,
+      "learning_rate": 9.871868741596673e-06,
+      "loss": 0.8382,
+      "step": 2717
+    },
+    {
+      "epoch": 0.1495954648026859,
+      "grad_norm": 1.0857365131378174,
+      "learning_rate": 9.871771221432718e-06,
+      "loss": 0.9254,
+      "step": 2718
+    },
+    {
+      "epoch": 0.14965050360504156,
+      "grad_norm": 0.7622446417808533,
+      "learning_rate": 9.871673664653837e-06,
+      "loss": 0.832,
+      "step": 2719
+    },
+    {
+      "epoch": 0.14970554240739722,
+      "grad_norm": 0.7436832785606384,
+      "learning_rate": 9.871576071260758e-06,
+      "loss": 0.7642,
+      "step": 2720
+    },
+    {
+      "epoch": 0.14976058120975289,
+      "grad_norm": 0.8547641634941101,
+      "learning_rate": 9.87147844125422e-06,
+      "loss": 0.7584,
+      "step": 2721
+    },
+    {
+      "epoch": 0.14981562001210855,
+      "grad_norm": 0.7634096145629883,
+      "learning_rate": 9.871380774634953e-06,
+      "loss": 0.8332,
+      "step": 2722
+    },
+    {
+      "epoch": 0.1498706588144642,
+      "grad_norm": 0.7949081063270569,
+      "learning_rate": 9.871283071403692e-06,
+      "loss": 0.7812,
+      "step": 2723
+    },
+    {
+      "epoch": 0.14992569761681987,
+      "grad_norm": 0.8089914321899414,
+      "learning_rate": 9.871185331561171e-06,
+      "loss": 0.8503,
+      "step": 2724
+    },
+    {
+      "epoch": 0.14998073641917553,
+      "grad_norm": 0.8451627492904663,
+      "learning_rate": 9.871087555108125e-06,
+      "loss": 0.9101,
+      "step": 2725
+    },
+    {
+      "epoch": 0.1500357752215312,
+      "grad_norm": 0.8399865627288818,
+      "learning_rate": 9.87098974204529e-06,
+      "loss": 0.8222,
+      "step": 2726
+    },
+    {
+      "epoch": 0.15009081402388685,
+      "grad_norm": 0.7786773443222046,
+      "learning_rate": 9.870891892373397e-06,
+      "loss": 0.8069,
+      "step": 2727
     }
   ],
   "logging_steps": 1,
@@ -12752,7 +19115,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 5.365029428243988e+18,
+  "total_flos": 8.047544142365983e+18,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null