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": 0.4002421707303649,
+  "epoch": 0.4502724420716605,
   "eval_steps": 500,
-  "global_step": 7272,
+  "global_step": 8181,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -50911,6 +50911,6369 @@
       "learning_rate": 9.06425972327535e-06,
       "loss": 0.8406,
       "step": 7272
+    },
+    {
+      "epoch": 0.4002972095327206,
+      "grad_norm": 0.7348741292953491,
+      "learning_rate": 9.064007226547819e-06,
+      "loss": 0.8103,
+      "step": 7273
+    },
+    {
+      "epoch": 0.4003522483350762,
+      "grad_norm": 0.6561787724494934,
+      "learning_rate": 9.063754699276297e-06,
+      "loss": 0.6634,
+      "step": 7274
+    },
+    {
+      "epoch": 0.4004072871374319,
+      "grad_norm": 0.7924866080284119,
+      "learning_rate": 9.063502141462682e-06,
+      "loss": 0.6592,
+      "step": 7275
+    },
+    {
+      "epoch": 0.40046232593978753,
+      "grad_norm": 0.6873973608016968,
+      "learning_rate": 9.063249553108873e-06,
+      "loss": 0.7912,
+      "step": 7276
+    },
+    {
+      "epoch": 0.4005173647421432,
+      "grad_norm": 0.6872708797454834,
+      "learning_rate": 9.062996934216768e-06,
+      "loss": 0.732,
+      "step": 7277
+    },
+    {
+      "epoch": 0.40057240354449886,
+      "grad_norm": 0.7381585836410522,
+      "learning_rate": 9.062744284788265e-06,
+      "loss": 0.84,
+      "step": 7278
+    },
+    {
+      "epoch": 0.40062744234685455,
+      "grad_norm": 0.7885964512825012,
+      "learning_rate": 9.062491604825266e-06,
+      "loss": 0.8229,
+      "step": 7279
+    },
+    {
+      "epoch": 0.4006824811492102,
+      "grad_norm": 0.9066407680511475,
+      "learning_rate": 9.062238894329664e-06,
+      "loss": 0.7299,
+      "step": 7280
+    },
+    {
+      "epoch": 0.40073751995156587,
+      "grad_norm": 0.7694007754325867,
+      "learning_rate": 9.061986153303364e-06,
+      "loss": 0.8033,
+      "step": 7281
+    },
+    {
+      "epoch": 0.4007925587539215,
+      "grad_norm": 1.021766185760498,
+      "learning_rate": 9.061733381748263e-06,
+      "loss": 0.79,
+      "step": 7282
+    },
+    {
+      "epoch": 0.4008475975562772,
+      "grad_norm": 0.7776662111282349,
+      "learning_rate": 9.06148057966626e-06,
+      "loss": 0.8484,
+      "step": 7283
+    },
+    {
+      "epoch": 0.4009026363586328,
+      "grad_norm": 0.8646043539047241,
+      "learning_rate": 9.061227747059257e-06,
+      "loss": 0.8223,
+      "step": 7284
+    },
+    {
+      "epoch": 0.4009576751609885,
+      "grad_norm": 0.7347257733345032,
+      "learning_rate": 9.060974883929154e-06,
+      "loss": 0.8062,
+      "step": 7285
+    },
+    {
+      "epoch": 0.40101271396334415,
+      "grad_norm": 0.8233902454376221,
+      "learning_rate": 9.06072199027785e-06,
+      "loss": 0.8922,
+      "step": 7286
+    },
+    {
+      "epoch": 0.40106775276569984,
+      "grad_norm": 0.7099601030349731,
+      "learning_rate": 9.060469066107246e-06,
+      "loss": 0.7125,
+      "step": 7287
+    },
+    {
+      "epoch": 0.40112279156805547,
+      "grad_norm": 0.7549998164176941,
+      "learning_rate": 9.060216111419246e-06,
+      "loss": 0.7851,
+      "step": 7288
+    },
+    {
+      "epoch": 0.40117783037041116,
+      "grad_norm": 0.753516435623169,
+      "learning_rate": 9.059963126215748e-06,
+      "loss": 0.7831,
+      "step": 7289
+    },
+    {
+      "epoch": 0.4012328691727668,
+      "grad_norm": 0.6718429327011108,
+      "learning_rate": 9.059710110498651e-06,
+      "loss": 0.7305,
+      "step": 7290
+    },
+    {
+      "epoch": 0.4012879079751225,
+      "grad_norm": 0.6796036958694458,
+      "learning_rate": 9.05945706426986e-06,
+      "loss": 0.802,
+      "step": 7291
+    },
+    {
+      "epoch": 0.4013429467774781,
+      "grad_norm": 0.8046827912330627,
+      "learning_rate": 9.05920398753128e-06,
+      "loss": 0.7286,
+      "step": 7292
+    },
+    {
+      "epoch": 0.4013979855798338,
+      "grad_norm": 0.7518643140792847,
+      "learning_rate": 9.058950880284807e-06,
+      "loss": 0.7287,
+      "step": 7293
+    },
+    {
+      "epoch": 0.40145302438218944,
+      "grad_norm": 0.8386855125427246,
+      "learning_rate": 9.058697742532345e-06,
+      "loss": 0.8201,
+      "step": 7294
+    },
+    {
+      "epoch": 0.4015080631845451,
+      "grad_norm": 0.7780192494392395,
+      "learning_rate": 9.058444574275797e-06,
+      "loss": 0.7999,
+      "step": 7295
+    },
+    {
+      "epoch": 0.40156310198690076,
+      "grad_norm": 0.7715566754341125,
+      "learning_rate": 9.058191375517068e-06,
+      "loss": 0.732,
+      "step": 7296
+    },
+    {
+      "epoch": 0.40161814078925645,
+      "grad_norm": 0.9940280914306641,
+      "learning_rate": 9.057938146258057e-06,
+      "loss": 0.8247,
+      "step": 7297
+    },
+    {
+      "epoch": 0.4016731795916121,
+      "grad_norm": 0.7567923069000244,
+      "learning_rate": 9.05768488650067e-06,
+      "loss": 0.8254,
+      "step": 7298
+    },
+    {
+      "epoch": 0.4017282183939678,
+      "grad_norm": 0.7544496655464172,
+      "learning_rate": 9.05743159624681e-06,
+      "loss": 0.811,
+      "step": 7299
+    },
+    {
+      "epoch": 0.4017832571963234,
+      "grad_norm": 0.63368821144104,
+      "learning_rate": 9.05717827549838e-06,
+      "loss": 0.6498,
+      "step": 7300
+    },
+    {
+      "epoch": 0.4018382959986791,
+      "grad_norm": 0.7077621221542358,
+      "learning_rate": 9.056924924257284e-06,
+      "loss": 0.7401,
+      "step": 7301
+    },
+    {
+      "epoch": 0.40189333480103473,
+      "grad_norm": 0.6782366037368774,
+      "learning_rate": 9.056671542525426e-06,
+      "loss": 0.8013,
+      "step": 7302
+    },
+    {
+      "epoch": 0.40194837360339036,
+      "grad_norm": 0.6605678200721741,
+      "learning_rate": 9.056418130304709e-06,
+      "loss": 0.8038,
+      "step": 7303
+    },
+    {
+      "epoch": 0.40200341240574605,
+      "grad_norm": 0.8716840147972107,
+      "learning_rate": 9.056164687597041e-06,
+      "loss": 0.7652,
+      "step": 7304
+    },
+    {
+      "epoch": 0.4020584512081017,
+      "grad_norm": 0.8464542031288147,
+      "learning_rate": 9.055911214404325e-06,
+      "loss": 0.8663,
+      "step": 7305
+    },
+    {
+      "epoch": 0.4021134900104574,
+      "grad_norm": 0.7165409326553345,
+      "learning_rate": 9.055657710728466e-06,
+      "loss": 0.8028,
+      "step": 7306
+    },
+    {
+      "epoch": 0.402168528812813,
+      "grad_norm": 0.7313430309295654,
+      "learning_rate": 9.055404176571369e-06,
+      "loss": 0.7538,
+      "step": 7307
+    },
+    {
+      "epoch": 0.4022235676151687,
+      "grad_norm": 0.7757230401039124,
+      "learning_rate": 9.05515061193494e-06,
+      "loss": 0.9096,
+      "step": 7308
+    },
+    {
+      "epoch": 0.40227860641752433,
+      "grad_norm": 0.7178354859352112,
+      "learning_rate": 9.054897016821085e-06,
+      "loss": 0.7186,
+      "step": 7309
+    },
+    {
+      "epoch": 0.40233364521988,
+      "grad_norm": 0.8331356048583984,
+      "learning_rate": 9.054643391231708e-06,
+      "loss": 0.8724,
+      "step": 7310
+    },
+    {
+      "epoch": 0.40238868402223565,
+      "grad_norm": 0.7709757685661316,
+      "learning_rate": 9.054389735168717e-06,
+      "loss": 0.692,
+      "step": 7311
+    },
+    {
+      "epoch": 0.40244372282459134,
+      "grad_norm": 0.7393380999565125,
+      "learning_rate": 9.054136048634018e-06,
+      "loss": 0.7863,
+      "step": 7312
+    },
+    {
+      "epoch": 0.402498761626947,
+      "grad_norm": 0.7372385859489441,
+      "learning_rate": 9.053882331629518e-06,
+      "loss": 0.781,
+      "step": 7313
+    },
+    {
+      "epoch": 0.40255380042930267,
+      "grad_norm": 0.7076019048690796,
+      "learning_rate": 9.053628584157123e-06,
+      "loss": 0.7598,
+      "step": 7314
+    },
+    {
+      "epoch": 0.4026088392316583,
+      "grad_norm": 0.7465673685073853,
+      "learning_rate": 9.053374806218742e-06,
+      "loss": 0.7454,
+      "step": 7315
+    },
+    {
+      "epoch": 0.402663878034014,
+      "grad_norm": 0.7414120435714722,
+      "learning_rate": 9.05312099781628e-06,
+      "loss": 0.7135,
+      "step": 7316
+    },
+    {
+      "epoch": 0.4027189168363696,
+      "grad_norm": 0.7490748167037964,
+      "learning_rate": 9.052867158951646e-06,
+      "loss": 0.6833,
+      "step": 7317
+    },
+    {
+      "epoch": 0.4027739556387253,
+      "grad_norm": 0.8027878999710083,
+      "learning_rate": 9.052613289626747e-06,
+      "loss": 0.7466,
+      "step": 7318
+    },
+    {
+      "epoch": 0.40282899444108095,
+      "grad_norm": 0.6777862310409546,
+      "learning_rate": 9.052359389843493e-06,
+      "loss": 0.7446,
+      "step": 7319
+    },
+    {
+      "epoch": 0.40288403324343663,
+      "grad_norm": 0.9240381717681885,
+      "learning_rate": 9.052105459603787e-06,
+      "loss": 0.7801,
+      "step": 7320
+    },
+    {
+      "epoch": 0.40293907204579227,
+      "grad_norm": 0.9592602252960205,
+      "learning_rate": 9.051851498909543e-06,
+      "loss": 0.9648,
+      "step": 7321
+    },
+    {
+      "epoch": 0.40299411084814796,
+      "grad_norm": 0.8469638228416443,
+      "learning_rate": 9.051597507762669e-06,
+      "loss": 0.8303,
+      "step": 7322
+    },
+    {
+      "epoch": 0.4030491496505036,
+      "grad_norm": 0.6981443166732788,
+      "learning_rate": 9.05134348616507e-06,
+      "loss": 0.7245,
+      "step": 7323
+    },
+    {
+      "epoch": 0.4031041884528593,
+      "grad_norm": 0.7133469581604004,
+      "learning_rate": 9.05108943411866e-06,
+      "loss": 0.7763,
+      "step": 7324
+    },
+    {
+      "epoch": 0.4031592272552149,
+      "grad_norm": 0.7043703198432922,
+      "learning_rate": 9.050835351625344e-06,
+      "loss": 0.8247,
+      "step": 7325
+    },
+    {
+      "epoch": 0.4032142660575706,
+      "grad_norm": 0.6662501692771912,
+      "learning_rate": 9.050581238687036e-06,
+      "loss": 0.7669,
+      "step": 7326
+    },
+    {
+      "epoch": 0.40326930485992624,
+      "grad_norm": 0.6482356786727905,
+      "learning_rate": 9.050327095305643e-06,
+      "loss": 0.6477,
+      "step": 7327
+    },
+    {
+      "epoch": 0.4033243436622819,
+      "grad_norm": 0.7465450167655945,
+      "learning_rate": 9.050072921483076e-06,
+      "loss": 0.8053,
+      "step": 7328
+    },
+    {
+      "epoch": 0.40337938246463756,
+      "grad_norm": 0.6765472292900085,
+      "learning_rate": 9.049818717221245e-06,
+      "loss": 0.765,
+      "step": 7329
+    },
+    {
+      "epoch": 0.40343442126699325,
+      "grad_norm": 0.7098689675331116,
+      "learning_rate": 9.04956448252206e-06,
+      "loss": 0.8059,
+      "step": 7330
+    },
+    {
+      "epoch": 0.4034894600693489,
+      "grad_norm": 0.6773823499679565,
+      "learning_rate": 9.049310217387432e-06,
+      "loss": 0.6848,
+      "step": 7331
+    },
+    {
+      "epoch": 0.40354449887170457,
+      "grad_norm": 0.6884829998016357,
+      "learning_rate": 9.049055921819275e-06,
+      "loss": 0.696,
+      "step": 7332
+    },
+    {
+      "epoch": 0.4035995376740602,
+      "grad_norm": 0.662545919418335,
+      "learning_rate": 9.048801595819494e-06,
+      "loss": 0.8286,
+      "step": 7333
+    },
+    {
+      "epoch": 0.4036545764764159,
+      "grad_norm": 0.6863077878952026,
+      "learning_rate": 9.048547239390007e-06,
+      "loss": 0.7215,
+      "step": 7334
+    },
+    {
+      "epoch": 0.4037096152787715,
+      "grad_norm": 0.6982632875442505,
+      "learning_rate": 9.048292852532721e-06,
+      "loss": 0.7635,
+      "step": 7335
+    },
+    {
+      "epoch": 0.4037646540811272,
+      "grad_norm": 0.8512400984764099,
+      "learning_rate": 9.048038435249548e-06,
+      "loss": 0.6226,
+      "step": 7336
+    },
+    {
+      "epoch": 0.40381969288348285,
+      "grad_norm": 0.6952843070030212,
+      "learning_rate": 9.047783987542405e-06,
+      "loss": 0.8317,
+      "step": 7337
+    },
+    {
+      "epoch": 0.40387473168583854,
+      "grad_norm": 0.7802778482437134,
+      "learning_rate": 9.0475295094132e-06,
+      "loss": 0.8615,
+      "step": 7338
+    },
+    {
+      "epoch": 0.4039297704881942,
+      "grad_norm": 0.8783930540084839,
+      "learning_rate": 9.047275000863844e-06,
+      "loss": 0.743,
+      "step": 7339
+    },
+    {
+      "epoch": 0.40398480929054986,
+      "grad_norm": 0.7205806970596313,
+      "learning_rate": 9.047020461896256e-06,
+      "loss": 0.7953,
+      "step": 7340
+    },
+    {
+      "epoch": 0.4040398480929055,
+      "grad_norm": 0.8438451290130615,
+      "learning_rate": 9.046765892512344e-06,
+      "loss": 0.7613,
+      "step": 7341
+    },
+    {
+      "epoch": 0.4040948868952612,
+      "grad_norm": 0.7300973534584045,
+      "learning_rate": 9.046511292714021e-06,
+      "loss": 0.7856,
+      "step": 7342
+    },
+    {
+      "epoch": 0.4041499256976168,
+      "grad_norm": 0.8472041487693787,
+      "learning_rate": 9.046256662503206e-06,
+      "loss": 0.8526,
+      "step": 7343
+    },
+    {
+      "epoch": 0.4042049644999725,
+      "grad_norm": 0.789465606212616,
+      "learning_rate": 9.046002001881807e-06,
+      "loss": 0.7792,
+      "step": 7344
+    },
+    {
+      "epoch": 0.40426000330232814,
+      "grad_norm": 0.7720938920974731,
+      "learning_rate": 9.04574731085174e-06,
+      "loss": 0.8065,
+      "step": 7345
+    },
+    {
+      "epoch": 0.4043150421046838,
+      "grad_norm": 0.6968526840209961,
+      "learning_rate": 9.04549258941492e-06,
+      "loss": 0.8135,
+      "step": 7346
+    },
+    {
+      "epoch": 0.40437008090703946,
+      "grad_norm": 0.746865451335907,
+      "learning_rate": 9.04523783757326e-06,
+      "loss": 0.8216,
+      "step": 7347
+    },
+    {
+      "epoch": 0.4044251197093951,
+      "grad_norm": 0.6750560998916626,
+      "learning_rate": 9.044983055328676e-06,
+      "loss": 0.7883,
+      "step": 7348
+    },
+    {
+      "epoch": 0.4044801585117508,
+      "grad_norm": 0.6791195273399353,
+      "learning_rate": 9.044728242683081e-06,
+      "loss": 0.7721,
+      "step": 7349
+    },
+    {
+      "epoch": 0.4045351973141064,
+      "grad_norm": 0.7238358855247498,
+      "learning_rate": 9.044473399638392e-06,
+      "loss": 0.739,
+      "step": 7350
+    },
+    {
+      "epoch": 0.4045902361164621,
+      "grad_norm": 0.6793557405471802,
+      "learning_rate": 9.044218526196523e-06,
+      "loss": 0.7853,
+      "step": 7351
+    },
+    {
+      "epoch": 0.40464527491881774,
+      "grad_norm": 0.767564058303833,
+      "learning_rate": 9.043963622359392e-06,
+      "loss": 0.8158,
+      "step": 7352
+    },
+    {
+      "epoch": 0.40470031372117343,
+      "grad_norm": 0.6800708770751953,
+      "learning_rate": 9.043708688128909e-06,
+      "loss": 0.7493,
+      "step": 7353
+    },
+    {
+      "epoch": 0.40475535252352907,
+      "grad_norm": 0.75978022813797,
+      "learning_rate": 9.043453723506996e-06,
+      "loss": 0.7066,
+      "step": 7354
+    },
+    {
+      "epoch": 0.40481039132588476,
+      "grad_norm": 1.0194984674453735,
+      "learning_rate": 9.043198728495568e-06,
+      "loss": 0.6238,
+      "step": 7355
+    },
+    {
+      "epoch": 0.4048654301282404,
+      "grad_norm": 0.7102386355400085,
+      "learning_rate": 9.04294370309654e-06,
+      "loss": 0.75,
+      "step": 7356
+    },
+    {
+      "epoch": 0.4049204689305961,
+      "grad_norm": 0.8468191623687744,
+      "learning_rate": 9.04268864731183e-06,
+      "loss": 0.8095,
+      "step": 7357
+    },
+    {
+      "epoch": 0.4049755077329517,
+      "grad_norm": 0.7022871971130371,
+      "learning_rate": 9.042433561143353e-06,
+      "loss": 0.8394,
+      "step": 7358
+    },
+    {
+      "epoch": 0.4050305465353074,
+      "grad_norm": 1.1873482465744019,
+      "learning_rate": 9.042178444593028e-06,
+      "loss": 0.7863,
+      "step": 7359
+    },
+    {
+      "epoch": 0.40508558533766303,
+      "grad_norm": 0.7074940204620361,
+      "learning_rate": 9.041923297662772e-06,
+      "loss": 0.7067,
+      "step": 7360
+    },
+    {
+      "epoch": 0.4051406241400187,
+      "grad_norm": 0.7602211833000183,
+      "learning_rate": 9.041668120354503e-06,
+      "loss": 0.6594,
+      "step": 7361
+    },
+    {
+      "epoch": 0.40519566294237436,
+      "grad_norm": 0.7903324365615845,
+      "learning_rate": 9.041412912670138e-06,
+      "loss": 0.7978,
+      "step": 7362
+    },
+    {
+      "epoch": 0.40525070174473005,
+      "grad_norm": 0.7422891855239868,
+      "learning_rate": 9.041157674611595e-06,
+      "loss": 0.8162,
+      "step": 7363
+    },
+    {
+      "epoch": 0.4053057405470857,
+      "grad_norm": 0.7978767156600952,
+      "learning_rate": 9.040902406180791e-06,
+      "loss": 0.762,
+      "step": 7364
+    },
+    {
+      "epoch": 0.40536077934944137,
+      "grad_norm": 0.7719776630401611,
+      "learning_rate": 9.04064710737965e-06,
+      "loss": 0.8098,
+      "step": 7365
+    },
+    {
+      "epoch": 0.405415818151797,
+      "grad_norm": 0.8646591305732727,
+      "learning_rate": 9.040391778210083e-06,
+      "loss": 0.9372,
+      "step": 7366
+    },
+    {
+      "epoch": 0.4054708569541527,
+      "grad_norm": 0.6616937518119812,
+      "learning_rate": 9.040136418674015e-06,
+      "loss": 0.7424,
+      "step": 7367
+    },
+    {
+      "epoch": 0.4055258957565083,
+      "grad_norm": 0.7676553130149841,
+      "learning_rate": 9.039881028773363e-06,
+      "loss": 0.6327,
+      "step": 7368
+    },
+    {
+      "epoch": 0.405580934558864,
+      "grad_norm": 0.6838239431381226,
+      "learning_rate": 9.039625608510047e-06,
+      "loss": 0.7548,
+      "step": 7369
+    },
+    {
+      "epoch": 0.40563597336121965,
+      "grad_norm": 0.7476304769515991,
+      "learning_rate": 9.039370157885986e-06,
+      "loss": 0.7262,
+      "step": 7370
+    },
+    {
+      "epoch": 0.40569101216357534,
+      "grad_norm": 0.8985139727592468,
+      "learning_rate": 9.0391146769031e-06,
+      "loss": 0.7729,
+      "step": 7371
+    },
+    {
+      "epoch": 0.40574605096593097,
+      "grad_norm": 0.7840422987937927,
+      "learning_rate": 9.038859165563308e-06,
+      "loss": 0.7855,
+      "step": 7372
+    },
+    {
+      "epoch": 0.40580108976828666,
+      "grad_norm": 0.6777672171592712,
+      "learning_rate": 9.038603623868534e-06,
+      "loss": 0.7379,
+      "step": 7373
+    },
+    {
+      "epoch": 0.4058561285706423,
+      "grad_norm": 0.7226746678352356,
+      "learning_rate": 9.038348051820694e-06,
+      "loss": 0.7686,
+      "step": 7374
+    },
+    {
+      "epoch": 0.405911167372998,
+      "grad_norm": 0.7647444605827332,
+      "learning_rate": 9.038092449421713e-06,
+      "loss": 0.8859,
+      "step": 7375
+    },
+    {
+      "epoch": 0.4059662061753536,
+      "grad_norm": 0.6524979472160339,
+      "learning_rate": 9.037836816673508e-06,
+      "loss": 0.6982,
+      "step": 7376
+    },
+    {
+      "epoch": 0.4060212449777093,
+      "grad_norm": 0.7842861413955688,
+      "learning_rate": 9.037581153578004e-06,
+      "loss": 0.8099,
+      "step": 7377
+    },
+    {
+      "epoch": 0.40607628378006494,
+      "grad_norm": 0.6424387693405151,
+      "learning_rate": 9.03732546013712e-06,
+      "loss": 0.7387,
+      "step": 7378
+    },
+    {
+      "epoch": 0.40613132258242063,
+      "grad_norm": 0.8444356918334961,
+      "learning_rate": 9.037069736352779e-06,
+      "loss": 0.8813,
+      "step": 7379
+    },
+    {
+      "epoch": 0.40618636138477626,
+      "grad_norm": 0.6487529277801514,
+      "learning_rate": 9.036813982226904e-06,
+      "loss": 0.7609,
+      "step": 7380
+    },
+    {
+      "epoch": 0.40624140018713195,
+      "grad_norm": 0.7891185879707336,
+      "learning_rate": 9.036558197761413e-06,
+      "loss": 0.8589,
+      "step": 7381
+    },
+    {
+      "epoch": 0.4062964389894876,
+      "grad_norm": 0.7183120250701904,
+      "learning_rate": 9.036302382958233e-06,
+      "loss": 0.8429,
+      "step": 7382
+    },
+    {
+      "epoch": 0.4063514777918433,
+      "grad_norm": 0.6386578679084778,
+      "learning_rate": 9.036046537819283e-06,
+      "loss": 0.6955,
+      "step": 7383
+    },
+    {
+      "epoch": 0.4064065165941989,
+      "grad_norm": 0.7572369575500488,
+      "learning_rate": 9.035790662346488e-06,
+      "loss": 0.8018,
+      "step": 7384
+    },
+    {
+      "epoch": 0.4064615553965546,
+      "grad_norm": 0.7105650305747986,
+      "learning_rate": 9.035534756541771e-06,
+      "loss": 0.8527,
+      "step": 7385
+    },
+    {
+      "epoch": 0.40651659419891023,
+      "grad_norm": 0.7031856179237366,
+      "learning_rate": 9.035278820407056e-06,
+      "loss": 0.6991,
+      "step": 7386
+    },
+    {
+      "epoch": 0.4065716330012659,
+      "grad_norm": 0.7407381534576416,
+      "learning_rate": 9.035022853944266e-06,
+      "loss": 0.708,
+      "step": 7387
+    },
+    {
+      "epoch": 0.40662667180362155,
+      "grad_norm": 0.7078498601913452,
+      "learning_rate": 9.034766857155322e-06,
+      "loss": 0.7584,
+      "step": 7388
+    },
+    {
+      "epoch": 0.4066817106059772,
+      "grad_norm": 0.7643301486968994,
+      "learning_rate": 9.034510830042151e-06,
+      "loss": 0.7836,
+      "step": 7389
+    },
+    {
+      "epoch": 0.4067367494083329,
+      "grad_norm": 0.7165302038192749,
+      "learning_rate": 9.034254772606676e-06,
+      "loss": 0.7769,
+      "step": 7390
+    },
+    {
+      "epoch": 0.4067917882106885,
+      "grad_norm": 0.7442395091056824,
+      "learning_rate": 9.033998684850824e-06,
+      "loss": 0.7231,
+      "step": 7391
+    },
+    {
+      "epoch": 0.4068468270130442,
+      "grad_norm": 0.7425046563148499,
+      "learning_rate": 9.033742566776517e-06,
+      "loss": 0.7709,
+      "step": 7392
+    },
+    {
+      "epoch": 0.40690186581539983,
+      "grad_norm": 0.768419086933136,
+      "learning_rate": 9.03348641838568e-06,
+      "loss": 0.7768,
+      "step": 7393
+    },
+    {
+      "epoch": 0.4069569046177555,
+      "grad_norm": 0.6785634160041809,
+      "learning_rate": 9.03323023968024e-06,
+      "loss": 0.7468,
+      "step": 7394
+    },
+    {
+      "epoch": 0.40701194342011116,
+      "grad_norm": 0.7075444459915161,
+      "learning_rate": 9.03297403066212e-06,
+      "loss": 0.7757,
+      "step": 7395
+    },
+    {
+      "epoch": 0.40706698222246684,
+      "grad_norm": 0.7580223679542542,
+      "learning_rate": 9.032717791333247e-06,
+      "loss": 0.7311,
+      "step": 7396
+    },
+    {
+      "epoch": 0.4071220210248225,
+      "grad_norm": 0.8110041618347168,
+      "learning_rate": 9.032461521695546e-06,
+      "loss": 0.7923,
+      "step": 7397
+    },
+    {
+      "epoch": 0.40717705982717817,
+      "grad_norm": 0.7204881310462952,
+      "learning_rate": 9.032205221750945e-06,
+      "loss": 0.759,
+      "step": 7398
+    },
+    {
+      "epoch": 0.4072320986295338,
+      "grad_norm": 0.8392491340637207,
+      "learning_rate": 9.031948891501368e-06,
+      "loss": 0.8292,
+      "step": 7399
+    },
+    {
+      "epoch": 0.4072871374318895,
+      "grad_norm": 0.7134600281715393,
+      "learning_rate": 9.031692530948742e-06,
+      "loss": 0.7,
+      "step": 7400
+    },
+    {
+      "epoch": 0.4073421762342451,
+      "grad_norm": 0.6324336528778076,
+      "learning_rate": 9.031436140094995e-06,
+      "loss": 0.6964,
+      "step": 7401
+    },
+    {
+      "epoch": 0.4073972150366008,
+      "grad_norm": 0.7281947731971741,
+      "learning_rate": 9.031179718942052e-06,
+      "loss": 0.7567,
+      "step": 7402
+    },
+    {
+      "epoch": 0.40745225383895645,
+      "grad_norm": 0.8828619718551636,
+      "learning_rate": 9.030923267491842e-06,
+      "loss": 0.8139,
+      "step": 7403
+    },
+    {
+      "epoch": 0.40750729264131214,
+      "grad_norm": 0.7039986252784729,
+      "learning_rate": 9.030666785746292e-06,
+      "loss": 0.7339,
+      "step": 7404
+    },
+    {
+      "epoch": 0.40756233144366777,
+      "grad_norm": 0.7049984931945801,
+      "learning_rate": 9.030410273707331e-06,
+      "loss": 0.6842,
+      "step": 7405
+    },
+    {
+      "epoch": 0.40761737024602346,
+      "grad_norm": 0.7149737477302551,
+      "learning_rate": 9.030153731376883e-06,
+      "loss": 0.6837,
+      "step": 7406
+    },
+    {
+      "epoch": 0.4076724090483791,
+      "grad_norm": 1.0804089307785034,
+      "learning_rate": 9.029897158756878e-06,
+      "loss": 0.7726,
+      "step": 7407
+    },
+    {
+      "epoch": 0.4077274478507348,
+      "grad_norm": 0.8354909420013428,
+      "learning_rate": 9.029640555849244e-06,
+      "loss": 0.8058,
+      "step": 7408
+    },
+    {
+      "epoch": 0.4077824866530904,
+      "grad_norm": 0.7091527581214905,
+      "learning_rate": 9.029383922655914e-06,
+      "loss": 0.7636,
+      "step": 7409
+    },
+    {
+      "epoch": 0.4078375254554461,
+      "grad_norm": 0.6720988750457764,
+      "learning_rate": 9.029127259178809e-06,
+      "loss": 0.7179,
+      "step": 7410
+    },
+    {
+      "epoch": 0.40789256425780174,
+      "grad_norm": 0.685858964920044,
+      "learning_rate": 9.028870565419865e-06,
+      "loss": 0.7637,
+      "step": 7411
+    },
+    {
+      "epoch": 0.4079476030601574,
+      "grad_norm": 0.7505033016204834,
+      "learning_rate": 9.028613841381007e-06,
+      "loss": 0.7463,
+      "step": 7412
+    },
+    {
+      "epoch": 0.40800264186251306,
+      "grad_norm": 0.8801671862602234,
+      "learning_rate": 9.028357087064166e-06,
+      "loss": 0.8399,
+      "step": 7413
+    },
+    {
+      "epoch": 0.40805768066486875,
+      "grad_norm": 0.7441918849945068,
+      "learning_rate": 9.02810030247127e-06,
+      "loss": 0.7689,
+      "step": 7414
+    },
+    {
+      "epoch": 0.4081127194672244,
+      "grad_norm": 0.7410128712654114,
+      "learning_rate": 9.027843487604251e-06,
+      "loss": 0.8013,
+      "step": 7415
+    },
+    {
+      "epoch": 0.40816775826958007,
+      "grad_norm": 0.8075226545333862,
+      "learning_rate": 9.02758664246504e-06,
+      "loss": 0.7717,
+      "step": 7416
+    },
+    {
+      "epoch": 0.4082227970719357,
+      "grad_norm": 0.7985545992851257,
+      "learning_rate": 9.027329767055566e-06,
+      "loss": 0.8459,
+      "step": 7417
+    },
+    {
+      "epoch": 0.4082778358742914,
+      "grad_norm": 0.7887235283851624,
+      "learning_rate": 9.027072861377757e-06,
+      "loss": 0.8201,
+      "step": 7418
+    },
+    {
+      "epoch": 0.40833287467664703,
+      "grad_norm": 0.7876266241073608,
+      "learning_rate": 9.02681592543355e-06,
+      "loss": 0.8205,
+      "step": 7419
+    },
+    {
+      "epoch": 0.4083879134790027,
+      "grad_norm": 0.758168637752533,
+      "learning_rate": 9.02655895922487e-06,
+      "loss": 0.6619,
+      "step": 7420
+    },
+    {
+      "epoch": 0.40844295228135835,
+      "grad_norm": 0.7279811501502991,
+      "learning_rate": 9.02630196275365e-06,
+      "loss": 0.7634,
+      "step": 7421
+    },
+    {
+      "epoch": 0.40849799108371404,
+      "grad_norm": 0.7540523409843445,
+      "learning_rate": 9.026044936021822e-06,
+      "loss": 0.7819,
+      "step": 7422
+    },
+    {
+      "epoch": 0.4085530298860697,
+      "grad_norm": 0.8091018795967102,
+      "learning_rate": 9.02578787903132e-06,
+      "loss": 0.7749,
+      "step": 7423
+    },
+    {
+      "epoch": 0.40860806868842536,
+      "grad_norm": 0.7625396847724915,
+      "learning_rate": 9.025530791784074e-06,
+      "loss": 0.7635,
+      "step": 7424
+    },
+    {
+      "epoch": 0.408663107490781,
+      "grad_norm": 0.7663947939872742,
+      "learning_rate": 9.025273674282015e-06,
+      "loss": 0.8281,
+      "step": 7425
+    },
+    {
+      "epoch": 0.4087181462931367,
+      "grad_norm": 0.6672662496566772,
+      "learning_rate": 9.025016526527077e-06,
+      "loss": 0.641,
+      "step": 7426
+    },
+    {
+      "epoch": 0.4087731850954923,
+      "grad_norm": 0.7649143934249878,
+      "learning_rate": 9.024759348521193e-06,
+      "loss": 0.7462,
+      "step": 7427
+    },
+    {
+      "epoch": 0.408828223897848,
+      "grad_norm": 0.7540067434310913,
+      "learning_rate": 9.024502140266293e-06,
+      "loss": 0.8756,
+      "step": 7428
+    },
+    {
+      "epoch": 0.40888326270020364,
+      "grad_norm": 0.721615731716156,
+      "learning_rate": 9.024244901764314e-06,
+      "loss": 0.8507,
+      "step": 7429
+    },
+    {
+      "epoch": 0.40893830150255933,
+      "grad_norm": 0.6949496269226074,
+      "learning_rate": 9.023987633017186e-06,
+      "loss": 0.7021,
+      "step": 7430
+    },
+    {
+      "epoch": 0.40899334030491497,
+      "grad_norm": 0.7108990550041199,
+      "learning_rate": 9.023730334026845e-06,
+      "loss": 0.807,
+      "step": 7431
+    },
+    {
+      "epoch": 0.4090483791072706,
+      "grad_norm": 0.7606124877929688,
+      "learning_rate": 9.023473004795225e-06,
+      "loss": 0.7769,
+      "step": 7432
+    },
+    {
+      "epoch": 0.4091034179096263,
+      "grad_norm": 0.7792031764984131,
+      "learning_rate": 9.023215645324256e-06,
+      "loss": 0.728,
+      "step": 7433
+    },
+    {
+      "epoch": 0.4091584567119819,
+      "grad_norm": 0.728884756565094,
+      "learning_rate": 9.022958255615877e-06,
+      "loss": 0.7831,
+      "step": 7434
+    },
+    {
+      "epoch": 0.4092134955143376,
+      "grad_norm": 0.8196625709533691,
+      "learning_rate": 9.022700835672022e-06,
+      "loss": 0.8265,
+      "step": 7435
+    },
+    {
+      "epoch": 0.40926853431669324,
+      "grad_norm": 0.762734055519104,
+      "learning_rate": 9.022443385494621e-06,
+      "loss": 0.8028,
+      "step": 7436
+    },
+    {
+      "epoch": 0.40932357311904893,
+      "grad_norm": 0.7259558439254761,
+      "learning_rate": 9.022185905085614e-06,
+      "loss": 0.789,
+      "step": 7437
+    },
+    {
+      "epoch": 0.40937861192140457,
+      "grad_norm": 0.7402371764183044,
+      "learning_rate": 9.021928394446936e-06,
+      "loss": 0.7667,
+      "step": 7438
+    },
+    {
+      "epoch": 0.40943365072376026,
+      "grad_norm": 0.8399797677993774,
+      "learning_rate": 9.021670853580519e-06,
+      "loss": 0.8451,
+      "step": 7439
+    },
+    {
+      "epoch": 0.4094886895261159,
+      "grad_norm": 0.6439585089683533,
+      "learning_rate": 9.0214132824883e-06,
+      "loss": 0.776,
+      "step": 7440
+    },
+    {
+      "epoch": 0.4095437283284716,
+      "grad_norm": 0.6956612467765808,
+      "learning_rate": 9.021155681172215e-06,
+      "loss": 0.6921,
+      "step": 7441
+    },
+    {
+      "epoch": 0.4095987671308272,
+      "grad_norm": 0.855413556098938,
+      "learning_rate": 9.020898049634203e-06,
+      "loss": 0.8552,
+      "step": 7442
+    },
+    {
+      "epoch": 0.4096538059331829,
+      "grad_norm": 0.6690535545349121,
+      "learning_rate": 9.020640387876194e-06,
+      "loss": 0.7552,
+      "step": 7443
+    },
+    {
+      "epoch": 0.40970884473553854,
+      "grad_norm": 0.6615462899208069,
+      "learning_rate": 9.020382695900131e-06,
+      "loss": 0.8216,
+      "step": 7444
+    },
+    {
+      "epoch": 0.4097638835378942,
+      "grad_norm": 0.6975858211517334,
+      "learning_rate": 9.020124973707947e-06,
+      "loss": 0.7453,
+      "step": 7445
+    },
+    {
+      "epoch": 0.40981892234024986,
+      "grad_norm": 0.6461964249610901,
+      "learning_rate": 9.019867221301579e-06,
+      "loss": 0.656,
+      "step": 7446
+    },
+    {
+      "epoch": 0.40987396114260555,
+      "grad_norm": 0.7221645712852478,
+      "learning_rate": 9.019609438682967e-06,
+      "loss": 0.661,
+      "step": 7447
+    },
+    {
+      "epoch": 0.4099289999449612,
+      "grad_norm": 0.6785755753517151,
+      "learning_rate": 9.019351625854044e-06,
+      "loss": 0.7294,
+      "step": 7448
+    },
+    {
+      "epoch": 0.40998403874731687,
+      "grad_norm": 0.7040538787841797,
+      "learning_rate": 9.019093782816751e-06,
+      "loss": 0.8546,
+      "step": 7449
+    },
+    {
+      "epoch": 0.4100390775496725,
+      "grad_norm": 0.737922191619873,
+      "learning_rate": 9.018835909573025e-06,
+      "loss": 0.8144,
+      "step": 7450
+    },
+    {
+      "epoch": 0.4100941163520282,
+      "grad_norm": 0.6705496311187744,
+      "learning_rate": 9.018578006124802e-06,
+      "loss": 0.6937,
+      "step": 7451
+    },
+    {
+      "epoch": 0.4101491551543838,
+      "grad_norm": 0.7347431182861328,
+      "learning_rate": 9.018320072474026e-06,
+      "loss": 0.7716,
+      "step": 7452
+    },
+    {
+      "epoch": 0.4102041939567395,
+      "grad_norm": 0.7023493647575378,
+      "learning_rate": 9.018062108622631e-06,
+      "loss": 0.7295,
+      "step": 7453
+    },
+    {
+      "epoch": 0.41025923275909515,
+      "grad_norm": 0.8017870187759399,
+      "learning_rate": 9.017804114572556e-06,
+      "loss": 0.7471,
+      "step": 7454
+    },
+    {
+      "epoch": 0.41031427156145084,
+      "grad_norm": 0.9171211123466492,
+      "learning_rate": 9.01754609032574e-06,
+      "loss": 0.8262,
+      "step": 7455
+    },
+    {
+      "epoch": 0.41036931036380647,
+      "grad_norm": 0.6682952046394348,
+      "learning_rate": 9.017288035884124e-06,
+      "loss": 0.7165,
+      "step": 7456
+    },
+    {
+      "epoch": 0.41042434916616216,
+      "grad_norm": 0.9339122772216797,
+      "learning_rate": 9.017029951249648e-06,
+      "loss": 0.8618,
+      "step": 7457
+    },
+    {
+      "epoch": 0.4104793879685178,
+      "grad_norm": 0.7063136696815491,
+      "learning_rate": 9.016771836424248e-06,
+      "loss": 0.8068,
+      "step": 7458
+    },
+    {
+      "epoch": 0.4105344267708735,
+      "grad_norm": 0.6717063784599304,
+      "learning_rate": 9.016513691409867e-06,
+      "loss": 0.738,
+      "step": 7459
+    },
+    {
+      "epoch": 0.4105894655732291,
+      "grad_norm": 0.6807749271392822,
+      "learning_rate": 9.016255516208443e-06,
+      "loss": 0.7842,
+      "step": 7460
+    },
+    {
+      "epoch": 0.4106445043755848,
+      "grad_norm": 0.6990453600883484,
+      "learning_rate": 9.01599731082192e-06,
+      "loss": 0.7726,
+      "step": 7461
+    },
+    {
+      "epoch": 0.41069954317794044,
+      "grad_norm": 0.6704931259155273,
+      "learning_rate": 9.015739075252234e-06,
+      "loss": 0.7006,
+      "step": 7462
+    },
+    {
+      "epoch": 0.41075458198029613,
+      "grad_norm": 0.7162300944328308,
+      "learning_rate": 9.01548080950133e-06,
+      "loss": 0.8462,
+      "step": 7463
+    },
+    {
+      "epoch": 0.41080962078265176,
+      "grad_norm": 0.6845411658287048,
+      "learning_rate": 9.015222513571144e-06,
+      "loss": 0.7466,
+      "step": 7464
+    },
+    {
+      "epoch": 0.41086465958500745,
+      "grad_norm": 0.7146134376525879,
+      "learning_rate": 9.014964187463623e-06,
+      "loss": 0.7594,
+      "step": 7465
+    },
+    {
+      "epoch": 0.4109196983873631,
+      "grad_norm": 0.7664906978607178,
+      "learning_rate": 9.014705831180706e-06,
+      "loss": 0.8376,
+      "step": 7466
+    },
+    {
+      "epoch": 0.4109747371897188,
+      "grad_norm": 0.7319341897964478,
+      "learning_rate": 9.014447444724332e-06,
+      "loss": 0.7748,
+      "step": 7467
+    },
+    {
+      "epoch": 0.4110297759920744,
+      "grad_norm": 0.7269605398178101,
+      "learning_rate": 9.014189028096448e-06,
+      "loss": 0.6941,
+      "step": 7468
+    },
+    {
+      "epoch": 0.4110848147944301,
+      "grad_norm": 0.72607421875,
+      "learning_rate": 9.013930581298993e-06,
+      "loss": 0.7174,
+      "step": 7469
+    },
+    {
+      "epoch": 0.41113985359678573,
+      "grad_norm": 0.7385421991348267,
+      "learning_rate": 9.01367210433391e-06,
+      "loss": 0.7761,
+      "step": 7470
+    },
+    {
+      "epoch": 0.4111948923991414,
+      "grad_norm": 0.8392042517662048,
+      "learning_rate": 9.013413597203144e-06,
+      "loss": 0.7417,
+      "step": 7471
+    },
+    {
+      "epoch": 0.41124993120149705,
+      "grad_norm": 0.7454584836959839,
+      "learning_rate": 9.013155059908634e-06,
+      "loss": 0.8976,
+      "step": 7472
+    },
+    {
+      "epoch": 0.41130497000385274,
+      "grad_norm": 0.7358037829399109,
+      "learning_rate": 9.012896492452325e-06,
+      "loss": 0.7706,
+      "step": 7473
+    },
+    {
+      "epoch": 0.4113600088062084,
+      "grad_norm": 0.7454121708869934,
+      "learning_rate": 9.01263789483616e-06,
+      "loss": 0.7425,
+      "step": 7474
+    },
+    {
+      "epoch": 0.411415047608564,
+      "grad_norm": 0.7842294573783875,
+      "learning_rate": 9.012379267062081e-06,
+      "loss": 0.7739,
+      "step": 7475
+    },
+    {
+      "epoch": 0.4114700864109197,
+      "grad_norm": 0.7181714773178101,
+      "learning_rate": 9.012120609132036e-06,
+      "loss": 0.8466,
+      "step": 7476
+    },
+    {
+      "epoch": 0.41152512521327533,
+      "grad_norm": 0.7239206433296204,
+      "learning_rate": 9.011861921047966e-06,
+      "loss": 0.7493,
+      "step": 7477
+    },
+    {
+      "epoch": 0.411580164015631,
+      "grad_norm": 0.6773414611816406,
+      "learning_rate": 9.011603202811816e-06,
+      "loss": 0.7433,
+      "step": 7478
+    },
+    {
+      "epoch": 0.41163520281798666,
+      "grad_norm": 0.7770900130271912,
+      "learning_rate": 9.011344454425527e-06,
+      "loss": 0.7488,
+      "step": 7479
+    },
+    {
+      "epoch": 0.41169024162034235,
+      "grad_norm": 0.7305957674980164,
+      "learning_rate": 9.011085675891051e-06,
+      "loss": 0.7989,
+      "step": 7480
+    },
+    {
+      "epoch": 0.411745280422698,
+      "grad_norm": 0.734603762626648,
+      "learning_rate": 9.010826867210327e-06,
+      "loss": 0.805,
+      "step": 7481
+    },
+    {
+      "epoch": 0.41180031922505367,
+      "grad_norm": 0.7438979148864746,
+      "learning_rate": 9.010568028385303e-06,
+      "loss": 0.8407,
+      "step": 7482
+    },
+    {
+      "epoch": 0.4118553580274093,
+      "grad_norm": 0.6718543767929077,
+      "learning_rate": 9.01030915941792e-06,
+      "loss": 0.7575,
+      "step": 7483
+    },
+    {
+      "epoch": 0.411910396829765,
+      "grad_norm": 0.8157614469528198,
+      "learning_rate": 9.01005026031013e-06,
+      "loss": 0.8231,
+      "step": 7484
+    },
+    {
+      "epoch": 0.4119654356321206,
+      "grad_norm": 0.8927714824676514,
+      "learning_rate": 9.009791331063874e-06,
+      "loss": 0.808,
+      "step": 7485
+    },
+    {
+      "epoch": 0.4120204744344763,
+      "grad_norm": 0.7604075074195862,
+      "learning_rate": 9.009532371681101e-06,
+      "loss": 0.7505,
+      "step": 7486
+    },
+    {
+      "epoch": 0.41207551323683195,
+      "grad_norm": 0.6861944794654846,
+      "learning_rate": 9.009273382163754e-06,
+      "loss": 0.719,
+      "step": 7487
+    },
+    {
+      "epoch": 0.41213055203918764,
+      "grad_norm": 0.7043709754943848,
+      "learning_rate": 9.009014362513784e-06,
+      "loss": 0.8193,
+      "step": 7488
+    },
+    {
+      "epoch": 0.41218559084154327,
+      "grad_norm": 0.7459648847579956,
+      "learning_rate": 9.008755312733136e-06,
+      "loss": 0.8617,
+      "step": 7489
+    },
+    {
+      "epoch": 0.41224062964389896,
+      "grad_norm": 0.7272594571113586,
+      "learning_rate": 9.008496232823754e-06,
+      "loss": 0.7255,
+      "step": 7490
+    },
+    {
+      "epoch": 0.4122956684462546,
+      "grad_norm": 0.7486668229103088,
+      "learning_rate": 9.008237122787586e-06,
+      "loss": 0.6479,
+      "step": 7491
+    },
+    {
+      "epoch": 0.4123507072486103,
+      "grad_norm": 0.8149027228355408,
+      "learning_rate": 9.007977982626582e-06,
+      "loss": 0.8052,
+      "step": 7492
+    },
+    {
+      "epoch": 0.4124057460509659,
+      "grad_norm": 0.7054859399795532,
+      "learning_rate": 9.00771881234269e-06,
+      "loss": 0.8215,
+      "step": 7493
+    },
+    {
+      "epoch": 0.4124607848533216,
+      "grad_norm": 0.6840499639511108,
+      "learning_rate": 9.007459611937854e-06,
+      "loss": 0.776,
+      "step": 7494
+    },
+    {
+      "epoch": 0.41251582365567724,
+      "grad_norm": 0.7340932488441467,
+      "learning_rate": 9.007200381414026e-06,
+      "loss": 0.713,
+      "step": 7495
+    },
+    {
+      "epoch": 0.4125708624580329,
+      "grad_norm": 0.8282599449157715,
+      "learning_rate": 9.00694112077315e-06,
+      "loss": 0.7037,
+      "step": 7496
+    },
+    {
+      "epoch": 0.41262590126038856,
+      "grad_norm": 0.849588930606842,
+      "learning_rate": 9.00668183001718e-06,
+      "loss": 0.7845,
+      "step": 7497
+    },
+    {
+      "epoch": 0.41268094006274425,
+      "grad_norm": 0.8330783843994141,
+      "learning_rate": 9.00642250914806e-06,
+      "loss": 0.9049,
+      "step": 7498
+    },
+    {
+      "epoch": 0.4127359788650999,
+      "grad_norm": 0.7020101547241211,
+      "learning_rate": 9.00616315816774e-06,
+      "loss": 0.8146,
+      "step": 7499
+    },
+    {
+      "epoch": 0.4127910176674556,
+      "grad_norm": 0.7632037997245789,
+      "learning_rate": 9.005903777078173e-06,
+      "loss": 0.6629,
+      "step": 7500
+    },
+    {
+      "epoch": 0.4128460564698112,
+      "grad_norm": 0.7286840081214905,
+      "learning_rate": 9.005644365881304e-06,
+      "loss": 0.7795,
+      "step": 7501
+    },
+    {
+      "epoch": 0.4129010952721669,
+      "grad_norm": 0.710451066493988,
+      "learning_rate": 9.005384924579084e-06,
+      "loss": 0.7615,
+      "step": 7502
+    },
+    {
+      "epoch": 0.41295613407452253,
+      "grad_norm": 0.7657510042190552,
+      "learning_rate": 9.005125453173463e-06,
+      "loss": 0.8938,
+      "step": 7503
+    },
+    {
+      "epoch": 0.4130111728768782,
+      "grad_norm": 0.6978467702865601,
+      "learning_rate": 9.004865951666392e-06,
+      "loss": 0.7464,
+      "step": 7504
+    },
+    {
+      "epoch": 0.41306621167923385,
+      "grad_norm": 0.7028319835662842,
+      "learning_rate": 9.00460642005982e-06,
+      "loss": 0.7899,
+      "step": 7505
+    },
+    {
+      "epoch": 0.41312125048158954,
+      "grad_norm": 0.923951268196106,
+      "learning_rate": 9.004346858355698e-06,
+      "loss": 0.8851,
+      "step": 7506
+    },
+    {
+      "epoch": 0.4131762892839452,
+      "grad_norm": 0.7293704748153687,
+      "learning_rate": 9.004087266555978e-06,
+      "loss": 0.7594,
+      "step": 7507
+    },
+    {
+      "epoch": 0.41323132808630086,
+      "grad_norm": 0.7458868622779846,
+      "learning_rate": 9.003827644662608e-06,
+      "loss": 0.7538,
+      "step": 7508
+    },
+    {
+      "epoch": 0.4132863668886565,
+      "grad_norm": 0.6764113306999207,
+      "learning_rate": 9.003567992677543e-06,
+      "loss": 0.7303,
+      "step": 7509
+    },
+    {
+      "epoch": 0.4133414056910122,
+      "grad_norm": 0.7827350497245789,
+      "learning_rate": 9.003308310602732e-06,
+      "loss": 0.7708,
+      "step": 7510
+    },
+    {
+      "epoch": 0.4133964444933678,
+      "grad_norm": 0.7683281302452087,
+      "learning_rate": 9.003048598440127e-06,
+      "loss": 0.7971,
+      "step": 7511
+    },
+    {
+      "epoch": 0.4134514832957235,
+      "grad_norm": 0.8793813586235046,
+      "learning_rate": 9.002788856191679e-06,
+      "loss": 0.7434,
+      "step": 7512
+    },
+    {
+      "epoch": 0.41350652209807914,
+      "grad_norm": 0.6598063111305237,
+      "learning_rate": 9.002529083859343e-06,
+      "loss": 0.7082,
+      "step": 7513
+    },
+    {
+      "epoch": 0.41356156090043483,
+      "grad_norm": 0.8239839673042297,
+      "learning_rate": 9.002269281445071e-06,
+      "loss": 0.8457,
+      "step": 7514
+    },
+    {
+      "epoch": 0.41361659970279047,
+      "grad_norm": 0.7433123588562012,
+      "learning_rate": 9.002009448950812e-06,
+      "loss": 0.7399,
+      "step": 7515
+    },
+    {
+      "epoch": 0.41367163850514616,
+      "grad_norm": 0.8310487866401672,
+      "learning_rate": 9.001749586378524e-06,
+      "loss": 0.7482,
+      "step": 7516
+    },
+    {
+      "epoch": 0.4137266773075018,
+      "grad_norm": 0.7170824408531189,
+      "learning_rate": 9.001489693730155e-06,
+      "loss": 0.7856,
+      "step": 7517
+    },
+    {
+      "epoch": 0.4137817161098574,
+      "grad_norm": 0.9063520431518555,
+      "learning_rate": 9.00122977100766e-06,
+      "loss": 0.8623,
+      "step": 7518
+    },
+    {
+      "epoch": 0.4138367549122131,
+      "grad_norm": 0.8753733038902283,
+      "learning_rate": 9.000969818212996e-06,
+      "loss": 0.7875,
+      "step": 7519
+    },
+    {
+      "epoch": 0.41389179371456875,
+      "grad_norm": 0.7013519406318665,
+      "learning_rate": 9.000709835348112e-06,
+      "loss": 0.724,
+      "step": 7520
+    },
+    {
+      "epoch": 0.41394683251692443,
+      "grad_norm": 0.7385973334312439,
+      "learning_rate": 9.000449822414963e-06,
+      "loss": 0.7286,
+      "step": 7521
+    },
+    {
+      "epoch": 0.41400187131928007,
+      "grad_norm": 0.7605431079864502,
+      "learning_rate": 9.000189779415505e-06,
+      "loss": 0.728,
+      "step": 7522
+    },
+    {
+      "epoch": 0.41405691012163576,
+      "grad_norm": 0.7631710767745972,
+      "learning_rate": 8.99992970635169e-06,
+      "loss": 0.8276,
+      "step": 7523
+    },
+    {
+      "epoch": 0.4141119489239914,
+      "grad_norm": 0.8066657185554504,
+      "learning_rate": 8.999669603225477e-06,
+      "loss": 0.8319,
+      "step": 7524
+    },
+    {
+      "epoch": 0.4141669877263471,
+      "grad_norm": 0.689407229423523,
+      "learning_rate": 8.999409470038815e-06,
+      "loss": 0.6675,
+      "step": 7525
+    },
+    {
+      "epoch": 0.4142220265287027,
+      "grad_norm": 0.7391255497932434,
+      "learning_rate": 8.999149306793664e-06,
+      "loss": 0.8228,
+      "step": 7526
+    },
+    {
+      "epoch": 0.4142770653310584,
+      "grad_norm": 0.7208844423294067,
+      "learning_rate": 8.998889113491977e-06,
+      "loss": 0.7689,
+      "step": 7527
+    },
+    {
+      "epoch": 0.41433210413341404,
+      "grad_norm": 0.8278803825378418,
+      "learning_rate": 8.99862889013571e-06,
+      "loss": 0.7964,
+      "step": 7528
+    },
+    {
+      "epoch": 0.4143871429357697,
+      "grad_norm": 0.7287253141403198,
+      "learning_rate": 8.998368636726817e-06,
+      "loss": 0.7689,
+      "step": 7529
+    },
+    {
+      "epoch": 0.41444218173812536,
+      "grad_norm": 0.7159145474433899,
+      "learning_rate": 8.998108353267257e-06,
+      "loss": 0.7537,
+      "step": 7530
+    },
+    {
+      "epoch": 0.41449722054048105,
+      "grad_norm": 0.7605739235877991,
+      "learning_rate": 8.997848039758985e-06,
+      "loss": 0.7327,
+      "step": 7531
+    },
+    {
+      "epoch": 0.4145522593428367,
+      "grad_norm": 0.7290406227111816,
+      "learning_rate": 8.997587696203958e-06,
+      "loss": 0.6804,
+      "step": 7532
+    },
+    {
+      "epoch": 0.41460729814519237,
+      "grad_norm": 0.7613189816474915,
+      "learning_rate": 8.997327322604131e-06,
+      "loss": 0.7465,
+      "step": 7533
+    },
+    {
+      "epoch": 0.414662336947548,
+      "grad_norm": 0.7796703577041626,
+      "learning_rate": 8.99706691896146e-06,
+      "loss": 0.7444,
+      "step": 7534
+    },
+    {
+      "epoch": 0.4147173757499037,
+      "grad_norm": 0.8758549094200134,
+      "learning_rate": 8.996806485277904e-06,
+      "loss": 0.8586,
+      "step": 7535
+    },
+    {
+      "epoch": 0.4147724145522593,
+      "grad_norm": 0.9599420428276062,
+      "learning_rate": 8.996546021555423e-06,
+      "loss": 0.7554,
+      "step": 7536
+    },
+    {
+      "epoch": 0.414827453354615,
+      "grad_norm": 0.8216326236724854,
+      "learning_rate": 8.996285527795972e-06,
+      "loss": 0.7995,
+      "step": 7537
+    },
+    {
+      "epoch": 0.41488249215697065,
+      "grad_norm": 0.6777452230453491,
+      "learning_rate": 8.996025004001507e-06,
+      "loss": 0.7809,
+      "step": 7538
+    },
+    {
+      "epoch": 0.41493753095932634,
+      "grad_norm": 0.7354100942611694,
+      "learning_rate": 8.995764450173989e-06,
+      "loss": 0.6548,
+      "step": 7539
+    },
+    {
+      "epoch": 0.414992569761682,
+      "grad_norm": 0.7548280358314514,
+      "learning_rate": 8.995503866315373e-06,
+      "loss": 0.8308,
+      "step": 7540
+    },
+    {
+      "epoch": 0.41504760856403766,
+      "grad_norm": 0.6891447901725769,
+      "learning_rate": 8.995243252427622e-06,
+      "loss": 0.8386,
+      "step": 7541
+    },
+    {
+      "epoch": 0.4151026473663933,
+      "grad_norm": 0.6848340034484863,
+      "learning_rate": 8.99498260851269e-06,
+      "loss": 0.7587,
+      "step": 7542
+    },
+    {
+      "epoch": 0.415157686168749,
+      "grad_norm": 0.7109090685844421,
+      "learning_rate": 8.994721934572538e-06,
+      "loss": 0.6847,
+      "step": 7543
+    },
+    {
+      "epoch": 0.4152127249711046,
+      "grad_norm": 0.6708144545555115,
+      "learning_rate": 8.994461230609128e-06,
+      "loss": 0.7266,
+      "step": 7544
+    },
+    {
+      "epoch": 0.4152677637734603,
+      "grad_norm": 0.6985414028167725,
+      "learning_rate": 8.994200496624415e-06,
+      "loss": 0.7696,
+      "step": 7545
+    },
+    {
+      "epoch": 0.41532280257581594,
+      "grad_norm": 0.6989198923110962,
+      "learning_rate": 8.993939732620359e-06,
+      "loss": 0.7894,
+      "step": 7546
+    },
+    {
+      "epoch": 0.41537784137817163,
+      "grad_norm": 0.6667589545249939,
+      "learning_rate": 8.993678938598921e-06,
+      "loss": 0.7417,
+      "step": 7547
+    },
+    {
+      "epoch": 0.41543288018052726,
+      "grad_norm": 1.0692487955093384,
+      "learning_rate": 8.993418114562064e-06,
+      "loss": 0.7147,
+      "step": 7548
+    },
+    {
+      "epoch": 0.41548791898288295,
+      "grad_norm": 0.6709207892417908,
+      "learning_rate": 8.993157260511742e-06,
+      "loss": 0.7694,
+      "step": 7549
+    },
+    {
+      "epoch": 0.4155429577852386,
+      "grad_norm": 0.6714604496955872,
+      "learning_rate": 8.992896376449923e-06,
+      "loss": 0.6969,
+      "step": 7550
+    },
+    {
+      "epoch": 0.4155979965875943,
+      "grad_norm": 0.8266897201538086,
+      "learning_rate": 8.99263546237856e-06,
+      "loss": 0.8392,
+      "step": 7551
+    },
+    {
+      "epoch": 0.4156530353899499,
+      "grad_norm": 0.675188422203064,
+      "learning_rate": 8.992374518299619e-06,
+      "loss": 0.7525,
+      "step": 7552
+    },
+    {
+      "epoch": 0.4157080741923056,
+      "grad_norm": 0.7406265139579773,
+      "learning_rate": 8.992113544215059e-06,
+      "loss": 0.7895,
+      "step": 7553
+    },
+    {
+      "epoch": 0.41576311299466123,
+      "grad_norm": 0.837336003780365,
+      "learning_rate": 8.991852540126844e-06,
+      "loss": 0.7376,
+      "step": 7554
+    },
+    {
+      "epoch": 0.4158181517970169,
+      "grad_norm": 0.6774994730949402,
+      "learning_rate": 8.991591506036931e-06,
+      "loss": 0.7231,
+      "step": 7555
+    },
+    {
+      "epoch": 0.41587319059937256,
+      "grad_norm": 0.6941245794296265,
+      "learning_rate": 8.991330441947287e-06,
+      "loss": 0.7213,
+      "step": 7556
+    },
+    {
+      "epoch": 0.41592822940172824,
+      "grad_norm": 0.7588210105895996,
+      "learning_rate": 8.991069347859871e-06,
+      "loss": 0.7829,
+      "step": 7557
+    },
+    {
+      "epoch": 0.4159832682040839,
+      "grad_norm": 0.7580196857452393,
+      "learning_rate": 8.990808223776647e-06,
+      "loss": 0.7782,
+      "step": 7558
+    },
+    {
+      "epoch": 0.41603830700643957,
+      "grad_norm": 0.7597478032112122,
+      "learning_rate": 8.990547069699576e-06,
+      "loss": 0.7764,
+      "step": 7559
+    },
+    {
+      "epoch": 0.4160933458087952,
+      "grad_norm": 0.7950314283370972,
+      "learning_rate": 8.990285885630622e-06,
+      "loss": 0.7263,
+      "step": 7560
+    },
+    {
+      "epoch": 0.41614838461115083,
+      "grad_norm": 0.6962432265281677,
+      "learning_rate": 8.990024671571747e-06,
+      "loss": 0.6616,
+      "step": 7561
+    },
+    {
+      "epoch": 0.4162034234135065,
+      "grad_norm": 0.682816207408905,
+      "learning_rate": 8.989763427524915e-06,
+      "loss": 0.7862,
+      "step": 7562
+    },
+    {
+      "epoch": 0.41625846221586216,
+      "grad_norm": 0.686673104763031,
+      "learning_rate": 8.989502153492089e-06,
+      "loss": 0.8199,
+      "step": 7563
+    },
+    {
+      "epoch": 0.41631350101821785,
+      "grad_norm": 0.7954965233802795,
+      "learning_rate": 8.989240849475231e-06,
+      "loss": 0.8021,
+      "step": 7564
+    },
+    {
+      "epoch": 0.4163685398205735,
+      "grad_norm": 0.7516284584999084,
+      "learning_rate": 8.988979515476309e-06,
+      "loss": 0.7803,
+      "step": 7565
+    },
+    {
+      "epoch": 0.41642357862292917,
+      "grad_norm": 0.7148317694664001,
+      "learning_rate": 8.988718151497284e-06,
+      "loss": 0.7407,
+      "step": 7566
+    },
+    {
+      "epoch": 0.4164786174252848,
+      "grad_norm": 0.7898986339569092,
+      "learning_rate": 8.98845675754012e-06,
+      "loss": 0.8382,
+      "step": 7567
+    },
+    {
+      "epoch": 0.4165336562276405,
+      "grad_norm": 0.7014235854148865,
+      "learning_rate": 8.988195333606784e-06,
+      "loss": 0.7205,
+      "step": 7568
+    },
+    {
+      "epoch": 0.4165886950299961,
+      "grad_norm": 0.6520957350730896,
+      "learning_rate": 8.987933879699238e-06,
+      "loss": 0.7452,
+      "step": 7569
+    },
+    {
+      "epoch": 0.4166437338323518,
+      "grad_norm": 0.7462863922119141,
+      "learning_rate": 8.987672395819449e-06,
+      "loss": 0.7787,
+      "step": 7570
+    },
+    {
+      "epoch": 0.41669877263470745,
+      "grad_norm": 0.7366049885749817,
+      "learning_rate": 8.987410881969382e-06,
+      "loss": 0.7662,
+      "step": 7571
+    },
+    {
+      "epoch": 0.41675381143706314,
+      "grad_norm": 0.7732293009757996,
+      "learning_rate": 8.987149338151002e-06,
+      "loss": 0.8258,
+      "step": 7572
+    },
+    {
+      "epoch": 0.41680885023941877,
+      "grad_norm": 0.9309358596801758,
+      "learning_rate": 8.986887764366275e-06,
+      "loss": 0.6538,
+      "step": 7573
+    },
+    {
+      "epoch": 0.41686388904177446,
+      "grad_norm": 0.6976680755615234,
+      "learning_rate": 8.986626160617167e-06,
+      "loss": 0.7175,
+      "step": 7574
+    },
+    {
+      "epoch": 0.4169189278441301,
+      "grad_norm": 0.7541783452033997,
+      "learning_rate": 8.986364526905645e-06,
+      "loss": 0.8153,
+      "step": 7575
+    },
+    {
+      "epoch": 0.4169739666464858,
+      "grad_norm": 0.8968943357467651,
+      "learning_rate": 8.986102863233673e-06,
+      "loss": 0.7859,
+      "step": 7576
+    },
+    {
+      "epoch": 0.4170290054488414,
+      "grad_norm": 0.6910044550895691,
+      "learning_rate": 8.985841169603218e-06,
+      "loss": 0.8381,
+      "step": 7577
+    },
+    {
+      "epoch": 0.4170840442511971,
+      "grad_norm": 0.8944257497787476,
+      "learning_rate": 8.985579446016249e-06,
+      "loss": 0.7062,
+      "step": 7578
+    },
+    {
+      "epoch": 0.41713908305355274,
+      "grad_norm": 0.6665629744529724,
+      "learning_rate": 8.98531769247473e-06,
+      "loss": 0.7928,
+      "step": 7579
+    },
+    {
+      "epoch": 0.41719412185590843,
+      "grad_norm": 0.7642979621887207,
+      "learning_rate": 8.985055908980634e-06,
+      "loss": 0.8442,
+      "step": 7580
+    },
+    {
+      "epoch": 0.41724916065826406,
+      "grad_norm": 0.7575559020042419,
+      "learning_rate": 8.98479409553592e-06,
+      "loss": 0.795,
+      "step": 7581
+    },
+    {
+      "epoch": 0.41730419946061975,
+      "grad_norm": 0.6567206978797913,
+      "learning_rate": 8.984532252142563e-06,
+      "loss": 0.713,
+      "step": 7582
+    },
+    {
+      "epoch": 0.4173592382629754,
+      "grad_norm": 0.6677179336547852,
+      "learning_rate": 8.984270378802527e-06,
+      "loss": 0.8173,
+      "step": 7583
+    },
+    {
+      "epoch": 0.4174142770653311,
+      "grad_norm": 0.6846007704734802,
+      "learning_rate": 8.984008475517782e-06,
+      "loss": 0.7154,
+      "step": 7584
+    },
+    {
+      "epoch": 0.4174693158676867,
+      "grad_norm": 0.7758762836456299,
+      "learning_rate": 8.983746542290294e-06,
+      "loss": 0.8686,
+      "step": 7585
+    },
+    {
+      "epoch": 0.4175243546700424,
+      "grad_norm": 0.6850305199623108,
+      "learning_rate": 8.983484579122036e-06,
+      "loss": 0.7568,
+      "step": 7586
+    },
+    {
+      "epoch": 0.41757939347239803,
+      "grad_norm": 0.7165307998657227,
+      "learning_rate": 8.983222586014973e-06,
+      "loss": 0.7856,
+      "step": 7587
+    },
+    {
+      "epoch": 0.4176344322747537,
+      "grad_norm": 0.7747449278831482,
+      "learning_rate": 8.982960562971074e-06,
+      "loss": 0.8148,
+      "step": 7588
+    },
+    {
+      "epoch": 0.41768947107710935,
+      "grad_norm": 0.789235532283783,
+      "learning_rate": 8.982698509992311e-06,
+      "loss": 0.8021,
+      "step": 7589
+    },
+    {
+      "epoch": 0.41774450987946504,
+      "grad_norm": 0.664186954498291,
+      "learning_rate": 8.982436427080652e-06,
+      "loss": 0.7394,
+      "step": 7590
+    },
+    {
+      "epoch": 0.4177995486818207,
+      "grad_norm": 0.7045899033546448,
+      "learning_rate": 8.982174314238069e-06,
+      "loss": 0.7029,
+      "step": 7591
+    },
+    {
+      "epoch": 0.41785458748417637,
+      "grad_norm": 0.7569751739501953,
+      "learning_rate": 8.981912171466525e-06,
+      "loss": 0.6106,
+      "step": 7592
+    },
+    {
+      "epoch": 0.417909626286532,
+      "grad_norm": 0.7383938431739807,
+      "learning_rate": 8.981649998767998e-06,
+      "loss": 0.8163,
+      "step": 7593
+    },
+    {
+      "epoch": 0.4179646650888877,
+      "grad_norm": 0.7314342856407166,
+      "learning_rate": 8.981387796144456e-06,
+      "loss": 0.6847,
+      "step": 7594
+    },
+    {
+      "epoch": 0.4180197038912433,
+      "grad_norm": 0.7249840497970581,
+      "learning_rate": 8.981125563597867e-06,
+      "loss": 0.8025,
+      "step": 7595
+    },
+    {
+      "epoch": 0.418074742693599,
+      "grad_norm": 0.7260022759437561,
+      "learning_rate": 8.980863301130206e-06,
+      "loss": 0.7807,
+      "step": 7596
+    },
+    {
+      "epoch": 0.41812978149595464,
+      "grad_norm": 0.6249421834945679,
+      "learning_rate": 8.980601008743441e-06,
+      "loss": 0.6744,
+      "step": 7597
+    },
+    {
+      "epoch": 0.41818482029831033,
+      "grad_norm": 0.8132835626602173,
+      "learning_rate": 8.980338686439544e-06,
+      "loss": 0.7992,
+      "step": 7598
+    },
+    {
+      "epoch": 0.41823985910066597,
+      "grad_norm": 0.7279506921768188,
+      "learning_rate": 8.980076334220487e-06,
+      "loss": 0.8402,
+      "step": 7599
+    },
+    {
+      "epoch": 0.41829489790302166,
+      "grad_norm": 0.7168325781822205,
+      "learning_rate": 8.979813952088242e-06,
+      "loss": 0.9107,
+      "step": 7600
+    },
+    {
+      "epoch": 0.4183499367053773,
+      "grad_norm": 0.633661150932312,
+      "learning_rate": 8.97955154004478e-06,
+      "loss": 0.6328,
+      "step": 7601
+    },
+    {
+      "epoch": 0.418404975507733,
+      "grad_norm": 0.6770638227462769,
+      "learning_rate": 8.979289098092074e-06,
+      "loss": 0.7604,
+      "step": 7602
+    },
+    {
+      "epoch": 0.4184600143100886,
+      "grad_norm": 0.7589067816734314,
+      "learning_rate": 8.979026626232098e-06,
+      "loss": 0.7774,
+      "step": 7603
+    },
+    {
+      "epoch": 0.41851505311244425,
+      "grad_norm": 0.7116312980651855,
+      "learning_rate": 8.97876412446682e-06,
+      "loss": 0.8186,
+      "step": 7604
+    },
+    {
+      "epoch": 0.41857009191479994,
+      "grad_norm": 0.7369259595870972,
+      "learning_rate": 8.978501592798219e-06,
+      "loss": 0.6705,
+      "step": 7605
+    },
+    {
+      "epoch": 0.41862513071715557,
+      "grad_norm": 0.6201806664466858,
+      "learning_rate": 8.978239031228265e-06,
+      "loss": 0.7011,
+      "step": 7606
+    },
+    {
+      "epoch": 0.41868016951951126,
+      "grad_norm": 0.7652842998504639,
+      "learning_rate": 8.977976439758929e-06,
+      "loss": 0.8112,
+      "step": 7607
+    },
+    {
+      "epoch": 0.4187352083218669,
+      "grad_norm": 0.7214640974998474,
+      "learning_rate": 8.97771381839219e-06,
+      "loss": 0.767,
+      "step": 7608
+    },
+    {
+      "epoch": 0.4187902471242226,
+      "grad_norm": 0.8093706369400024,
+      "learning_rate": 8.977451167130015e-06,
+      "loss": 0.8112,
+      "step": 7609
+    },
+    {
+      "epoch": 0.4188452859265782,
+      "grad_norm": 0.7023005485534668,
+      "learning_rate": 8.977188485974382e-06,
+      "loss": 0.7678,
+      "step": 7610
+    },
+    {
+      "epoch": 0.4189003247289339,
+      "grad_norm": 0.8126183748245239,
+      "learning_rate": 8.976925774927267e-06,
+      "loss": 0.8207,
+      "step": 7611
+    },
+    {
+      "epoch": 0.41895536353128954,
+      "grad_norm": 0.9624595642089844,
+      "learning_rate": 8.976663033990643e-06,
+      "loss": 0.7853,
+      "step": 7612
+    },
+    {
+      "epoch": 0.4190104023336452,
+      "grad_norm": 0.7866421937942505,
+      "learning_rate": 8.976400263166483e-06,
+      "loss": 0.6319,
+      "step": 7613
+    },
+    {
+      "epoch": 0.41906544113600086,
+      "grad_norm": 0.7555810213088989,
+      "learning_rate": 8.976137462456762e-06,
+      "loss": 0.7781,
+      "step": 7614
+    },
+    {
+      "epoch": 0.41912047993835655,
+      "grad_norm": 0.7383303046226501,
+      "learning_rate": 8.975874631863457e-06,
+      "loss": 0.8152,
+      "step": 7615
+    },
+    {
+      "epoch": 0.4191755187407122,
+      "grad_norm": 0.7873355746269226,
+      "learning_rate": 8.975611771388542e-06,
+      "loss": 0.723,
+      "step": 7616
+    },
+    {
+      "epoch": 0.41923055754306787,
+      "grad_norm": 0.7265962362289429,
+      "learning_rate": 8.975348881033993e-06,
+      "loss": 0.8016,
+      "step": 7617
+    },
+    {
+      "epoch": 0.4192855963454235,
+      "grad_norm": 0.7074393033981323,
+      "learning_rate": 8.975085960801788e-06,
+      "loss": 0.7453,
+      "step": 7618
+    },
+    {
+      "epoch": 0.4193406351477792,
+      "grad_norm": 0.6975581049919128,
+      "learning_rate": 8.9748230106939e-06,
+      "loss": 0.6516,
+      "step": 7619
+    },
+    {
+      "epoch": 0.41939567395013483,
+      "grad_norm": 0.7730469107627869,
+      "learning_rate": 8.974560030712304e-06,
+      "loss": 0.7297,
+      "step": 7620
+    },
+    {
+      "epoch": 0.4194507127524905,
+      "grad_norm": 0.7289026379585266,
+      "learning_rate": 8.974297020858982e-06,
+      "loss": 0.7087,
+      "step": 7621
+    },
+    {
+      "epoch": 0.41950575155484615,
+      "grad_norm": 0.8029256463050842,
+      "learning_rate": 8.974033981135906e-06,
+      "loss": 0.7923,
+      "step": 7622
+    },
+    {
+      "epoch": 0.41956079035720184,
+      "grad_norm": 0.765312135219574,
+      "learning_rate": 8.973770911545055e-06,
+      "loss": 0.7824,
+      "step": 7623
+    },
+    {
+      "epoch": 0.4196158291595575,
+      "grad_norm": 0.7903861403465271,
+      "learning_rate": 8.973507812088404e-06,
+      "loss": 0.8207,
+      "step": 7624
+    },
+    {
+      "epoch": 0.41967086796191316,
+      "grad_norm": 0.6875497698783875,
+      "learning_rate": 8.973244682767934e-06,
+      "loss": 0.7972,
+      "step": 7625
+    },
+    {
+      "epoch": 0.4197259067642688,
+      "grad_norm": 0.7781878709793091,
+      "learning_rate": 8.972981523585617e-06,
+      "loss": 0.754,
+      "step": 7626
+    },
+    {
+      "epoch": 0.4197809455666245,
+      "grad_norm": 0.6495640873908997,
+      "learning_rate": 8.972718334543437e-06,
+      "loss": 0.6851,
+      "step": 7627
+    },
+    {
+      "epoch": 0.4198359843689801,
+      "grad_norm": 0.7610780596733093,
+      "learning_rate": 8.97245511564337e-06,
+      "loss": 0.8161,
+      "step": 7628
+    },
+    {
+      "epoch": 0.4198910231713358,
+      "grad_norm": 0.7764771580696106,
+      "learning_rate": 8.972191866887393e-06,
+      "loss": 0.8341,
+      "step": 7629
+    },
+    {
+      "epoch": 0.41994606197369144,
+      "grad_norm": 0.7709774374961853,
+      "learning_rate": 8.971928588277485e-06,
+      "loss": 0.765,
+      "step": 7630
+    },
+    {
+      "epoch": 0.42000110077604713,
+      "grad_norm": 0.8213009238243103,
+      "learning_rate": 8.971665279815625e-06,
+      "loss": 0.8971,
+      "step": 7631
+    },
+    {
+      "epoch": 0.42005613957840277,
+      "grad_norm": 0.7232406735420227,
+      "learning_rate": 8.971401941503792e-06,
+      "loss": 0.7919,
+      "step": 7632
+    },
+    {
+      "epoch": 0.42011117838075845,
+      "grad_norm": 0.7322028279304504,
+      "learning_rate": 8.971138573343964e-06,
+      "loss": 0.8167,
+      "step": 7633
+    },
+    {
+      "epoch": 0.4201662171831141,
+      "grad_norm": 0.7204442024230957,
+      "learning_rate": 8.970875175338123e-06,
+      "loss": 0.8152,
+      "step": 7634
+    },
+    {
+      "epoch": 0.4202212559854698,
+      "grad_norm": 0.7385342121124268,
+      "learning_rate": 8.970611747488246e-06,
+      "loss": 0.8204,
+      "step": 7635
+    },
+    {
+      "epoch": 0.4202762947878254,
+      "grad_norm": 0.758941113948822,
+      "learning_rate": 8.970348289796316e-06,
+      "loss": 0.8402,
+      "step": 7636
+    },
+    {
+      "epoch": 0.4203313335901811,
+      "grad_norm": 0.7331902384757996,
+      "learning_rate": 8.970084802264309e-06,
+      "loss": 0.7305,
+      "step": 7637
+    },
+    {
+      "epoch": 0.42038637239253673,
+      "grad_norm": 0.7822885513305664,
+      "learning_rate": 8.969821284894208e-06,
+      "loss": 0.8708,
+      "step": 7638
+    },
+    {
+      "epoch": 0.4204414111948924,
+      "grad_norm": 0.6625984311103821,
+      "learning_rate": 8.969557737687992e-06,
+      "loss": 0.7806,
+      "step": 7639
+    },
+    {
+      "epoch": 0.42049644999724806,
+      "grad_norm": 1.02848482131958,
+      "learning_rate": 8.969294160647645e-06,
+      "loss": 0.7176,
+      "step": 7640
+    },
+    {
+      "epoch": 0.42055148879960375,
+      "grad_norm": 0.7888724207878113,
+      "learning_rate": 8.969030553775144e-06,
+      "loss": 0.8326,
+      "step": 7641
+    },
+    {
+      "epoch": 0.4206065276019594,
+      "grad_norm": 0.7148883938789368,
+      "learning_rate": 8.968766917072472e-06,
+      "loss": 0.7405,
+      "step": 7642
+    },
+    {
+      "epoch": 0.42066156640431507,
+      "grad_norm": 0.6629698872566223,
+      "learning_rate": 8.96850325054161e-06,
+      "loss": 0.845,
+      "step": 7643
+    },
+    {
+      "epoch": 0.4207166052066707,
+      "grad_norm": 0.8414682149887085,
+      "learning_rate": 8.96823955418454e-06,
+      "loss": 1.3631,
+      "step": 7644
+    },
+    {
+      "epoch": 0.4207716440090264,
+      "grad_norm": 0.7105298638343811,
+      "learning_rate": 8.967975828003244e-06,
+      "loss": 0.6808,
+      "step": 7645
+    },
+    {
+      "epoch": 0.420826682811382,
+      "grad_norm": 0.7324852347373962,
+      "learning_rate": 8.967712071999703e-06,
+      "loss": 0.8237,
+      "step": 7646
+    },
+    {
+      "epoch": 0.42088172161373766,
+      "grad_norm": 0.737324595451355,
+      "learning_rate": 8.9674482861759e-06,
+      "loss": 0.8486,
+      "step": 7647
+    },
+    {
+      "epoch": 0.42093676041609335,
+      "grad_norm": 0.6763800382614136,
+      "learning_rate": 8.967184470533818e-06,
+      "loss": 0.72,
+      "step": 7648
+    },
+    {
+      "epoch": 0.420991799218449,
+      "grad_norm": 0.7560757994651794,
+      "learning_rate": 8.96692062507544e-06,
+      "loss": 0.7704,
+      "step": 7649
+    },
+    {
+      "epoch": 0.42104683802080467,
+      "grad_norm": 0.7289260029792786,
+      "learning_rate": 8.966656749802748e-06,
+      "loss": 0.7411,
+      "step": 7650
+    },
+    {
+      "epoch": 0.4211018768231603,
+      "grad_norm": 0.6935442686080933,
+      "learning_rate": 8.966392844717726e-06,
+      "loss": 0.7848,
+      "step": 7651
+    },
+    {
+      "epoch": 0.421156915625516,
+      "grad_norm": 0.7111918330192566,
+      "learning_rate": 8.966128909822356e-06,
+      "loss": 0.8377,
+      "step": 7652
+    },
+    {
+      "epoch": 0.4212119544278716,
+      "grad_norm": 0.8594884872436523,
+      "learning_rate": 8.965864945118625e-06,
+      "loss": 0.8227,
+      "step": 7653
+    },
+    {
+      "epoch": 0.4212669932302273,
+      "grad_norm": 0.6521008014678955,
+      "learning_rate": 8.965600950608513e-06,
+      "loss": 0.7034,
+      "step": 7654
+    },
+    {
+      "epoch": 0.42132203203258295,
+      "grad_norm": 0.6362404823303223,
+      "learning_rate": 8.965336926294007e-06,
+      "loss": 0.6712,
+      "step": 7655
+    },
+    {
+      "epoch": 0.42137707083493864,
+      "grad_norm": 0.6955040097236633,
+      "learning_rate": 8.965072872177088e-06,
+      "loss": 0.7789,
+      "step": 7656
+    },
+    {
+      "epoch": 0.42143210963729427,
+      "grad_norm": 0.7311720252037048,
+      "learning_rate": 8.964808788259745e-06,
+      "loss": 0.7522,
+      "step": 7657
+    },
+    {
+      "epoch": 0.42148714843964996,
+      "grad_norm": 0.781131386756897,
+      "learning_rate": 8.96454467454396e-06,
+      "loss": 0.7831,
+      "step": 7658
+    },
+    {
+      "epoch": 0.4215421872420056,
+      "grad_norm": 0.6740639805793762,
+      "learning_rate": 8.964280531031718e-06,
+      "loss": 0.7102,
+      "step": 7659
+    },
+    {
+      "epoch": 0.4215972260443613,
+      "grad_norm": 0.7843424677848816,
+      "learning_rate": 8.964016357725003e-06,
+      "loss": 0.8325,
+      "step": 7660
+    },
+    {
+      "epoch": 0.4216522648467169,
+      "grad_norm": 0.7833517789840698,
+      "learning_rate": 8.963752154625804e-06,
+      "loss": 0.8603,
+      "step": 7661
+    },
+    {
+      "epoch": 0.4217073036490726,
+      "grad_norm": 0.7270992994308472,
+      "learning_rate": 8.963487921736104e-06,
+      "loss": 0.745,
+      "step": 7662
+    },
+    {
+      "epoch": 0.42176234245142824,
+      "grad_norm": 0.6517582535743713,
+      "learning_rate": 8.963223659057892e-06,
+      "loss": 0.6983,
+      "step": 7663
+    },
+    {
+      "epoch": 0.42181738125378393,
+      "grad_norm": 0.6974934935569763,
+      "learning_rate": 8.962959366593149e-06,
+      "loss": 0.733,
+      "step": 7664
+    },
+    {
+      "epoch": 0.42187242005613956,
+      "grad_norm": 0.712045431137085,
+      "learning_rate": 8.962695044343865e-06,
+      "loss": 0.725,
+      "step": 7665
+    },
+    {
+      "epoch": 0.42192745885849525,
+      "grad_norm": 0.7311459183692932,
+      "learning_rate": 8.962430692312028e-06,
+      "loss": 0.8025,
+      "step": 7666
+    },
+    {
+      "epoch": 0.4219824976608509,
+      "grad_norm": 0.7439966201782227,
+      "learning_rate": 8.962166310499621e-06,
+      "loss": 0.7711,
+      "step": 7667
+    },
+    {
+      "epoch": 0.4220375364632066,
+      "grad_norm": 0.690832257270813,
+      "learning_rate": 8.961901898908632e-06,
+      "loss": 0.8414,
+      "step": 7668
+    },
+    {
+      "epoch": 0.4220925752655622,
+      "grad_norm": 0.8437964916229248,
+      "learning_rate": 8.961637457541049e-06,
+      "loss": 0.8253,
+      "step": 7669
+    },
+    {
+      "epoch": 0.4221476140679179,
+      "grad_norm": 0.7876344323158264,
+      "learning_rate": 8.96137298639886e-06,
+      "loss": 0.754,
+      "step": 7670
+    },
+    {
+      "epoch": 0.42220265287027353,
+      "grad_norm": 0.7551780343055725,
+      "learning_rate": 8.961108485484052e-06,
+      "loss": 0.8555,
+      "step": 7671
+    },
+    {
+      "epoch": 0.4222576916726292,
+      "grad_norm": 0.6867276430130005,
+      "learning_rate": 8.96084395479861e-06,
+      "loss": 0.7216,
+      "step": 7672
+    },
+    {
+      "epoch": 0.42231273047498485,
+      "grad_norm": 0.9052873849868774,
+      "learning_rate": 8.960579394344528e-06,
+      "loss": 0.7945,
+      "step": 7673
+    },
+    {
+      "epoch": 0.42236776927734054,
+      "grad_norm": 0.6731994152069092,
+      "learning_rate": 8.96031480412379e-06,
+      "loss": 0.7691,
+      "step": 7674
+    },
+    {
+      "epoch": 0.4224228080796962,
+      "grad_norm": 0.7074670195579529,
+      "learning_rate": 8.960050184138389e-06,
+      "loss": 0.8008,
+      "step": 7675
+    },
+    {
+      "epoch": 0.42247784688205187,
+      "grad_norm": 0.9482604265213013,
+      "learning_rate": 8.959785534390309e-06,
+      "loss": 0.7095,
+      "step": 7676
+    },
+    {
+      "epoch": 0.4225328856844075,
+      "grad_norm": 0.6915413737297058,
+      "learning_rate": 8.95952085488154e-06,
+      "loss": 0.6717,
+      "step": 7677
+    },
+    {
+      "epoch": 0.4225879244867632,
+      "grad_norm": 0.7565900087356567,
+      "learning_rate": 8.959256145614073e-06,
+      "loss": 0.8311,
+      "step": 7678
+    },
+    {
+      "epoch": 0.4226429632891188,
+      "grad_norm": 0.8307167887687683,
+      "learning_rate": 8.958991406589896e-06,
+      "loss": 0.8585,
+      "step": 7679
+    },
+    {
+      "epoch": 0.4226980020914745,
+      "grad_norm": 0.7955091595649719,
+      "learning_rate": 8.958726637811e-06,
+      "loss": 0.8154,
+      "step": 7680
+    },
+    {
+      "epoch": 0.42275304089383015,
+      "grad_norm": 0.7692292332649231,
+      "learning_rate": 8.958461839279376e-06,
+      "loss": 0.7965,
+      "step": 7681
+    },
+    {
+      "epoch": 0.42280807969618583,
+      "grad_norm": 0.7355942726135254,
+      "learning_rate": 8.95819701099701e-06,
+      "loss": 0.7557,
+      "step": 7682
+    },
+    {
+      "epoch": 0.42286311849854147,
+      "grad_norm": 0.8781518936157227,
+      "learning_rate": 8.957932152965895e-06,
+      "loss": 0.8033,
+      "step": 7683
+    },
+    {
+      "epoch": 0.42291815730089716,
+      "grad_norm": 0.7180802226066589,
+      "learning_rate": 8.957667265188022e-06,
+      "loss": 0.7283,
+      "step": 7684
+    },
+    {
+      "epoch": 0.4229731961032528,
+      "grad_norm": 0.6967236995697021,
+      "learning_rate": 8.95740234766538e-06,
+      "loss": 0.769,
+      "step": 7685
+    },
+    {
+      "epoch": 0.4230282349056085,
+      "grad_norm": 0.7462503910064697,
+      "learning_rate": 8.957137400399963e-06,
+      "loss": 0.8179,
+      "step": 7686
+    },
+    {
+      "epoch": 0.4230832737079641,
+      "grad_norm": 0.67714524269104,
+      "learning_rate": 8.956872423393761e-06,
+      "loss": 0.7976,
+      "step": 7687
+    },
+    {
+      "epoch": 0.4231383125103198,
+      "grad_norm": 0.8239946365356445,
+      "learning_rate": 8.956607416648763e-06,
+      "loss": 0.7946,
+      "step": 7688
+    },
+    {
+      "epoch": 0.42319335131267544,
+      "grad_norm": 0.6724610924720764,
+      "learning_rate": 8.956342380166963e-06,
+      "loss": 0.7633,
+      "step": 7689
+    },
+    {
+      "epoch": 0.42324839011503107,
+      "grad_norm": 0.744987964630127,
+      "learning_rate": 8.956077313950354e-06,
+      "loss": 0.9028,
+      "step": 7690
+    },
+    {
+      "epoch": 0.42330342891738676,
+      "grad_norm": 0.7700596451759338,
+      "learning_rate": 8.955812218000925e-06,
+      "loss": 0.8954,
+      "step": 7691
+    },
+    {
+      "epoch": 0.4233584677197424,
+      "grad_norm": 0.6952996253967285,
+      "learning_rate": 8.955547092320673e-06,
+      "loss": 0.8094,
+      "step": 7692
+    },
+    {
+      "epoch": 0.4234135065220981,
+      "grad_norm": 0.6410536766052246,
+      "learning_rate": 8.955281936911586e-06,
+      "loss": 0.6281,
+      "step": 7693
+    },
+    {
+      "epoch": 0.4234685453244537,
+      "grad_norm": 1.0939754247665405,
+      "learning_rate": 8.95501675177566e-06,
+      "loss": 0.8239,
+      "step": 7694
+    },
+    {
+      "epoch": 0.4235235841268094,
+      "grad_norm": 0.7419464588165283,
+      "learning_rate": 8.954751536914885e-06,
+      "loss": 0.8015,
+      "step": 7695
+    },
+    {
+      "epoch": 0.42357862292916504,
+      "grad_norm": 0.8171356320381165,
+      "learning_rate": 8.954486292331257e-06,
+      "loss": 0.8183,
+      "step": 7696
+    },
+    {
+      "epoch": 0.4236336617315207,
+      "grad_norm": 0.745884358882904,
+      "learning_rate": 8.95422101802677e-06,
+      "loss": 0.7457,
+      "step": 7697
+    },
+    {
+      "epoch": 0.42368870053387636,
+      "grad_norm": 0.7355740070343018,
+      "learning_rate": 8.953955714003414e-06,
+      "loss": 0.7517,
+      "step": 7698
+    },
+    {
+      "epoch": 0.42374373933623205,
+      "grad_norm": 0.7103458642959595,
+      "learning_rate": 8.953690380263186e-06,
+      "loss": 0.7306,
+      "step": 7699
+    },
+    {
+      "epoch": 0.4237987781385877,
+      "grad_norm": 0.7453970909118652,
+      "learning_rate": 8.95342501680808e-06,
+      "loss": 0.8396,
+      "step": 7700
+    },
+    {
+      "epoch": 0.4238538169409434,
+      "grad_norm": 0.7132760286331177,
+      "learning_rate": 8.953159623640088e-06,
+      "loss": 0.7861,
+      "step": 7701
+    },
+    {
+      "epoch": 0.423908855743299,
+      "grad_norm": 0.785827100276947,
+      "learning_rate": 8.952894200761209e-06,
+      "loss": 0.8681,
+      "step": 7702
+    },
+    {
+      "epoch": 0.4239638945456547,
+      "grad_norm": 0.7075281143188477,
+      "learning_rate": 8.952628748173433e-06,
+      "loss": 0.7257,
+      "step": 7703
+    },
+    {
+      "epoch": 0.42401893334801033,
+      "grad_norm": 0.8205186724662781,
+      "learning_rate": 8.952363265878758e-06,
+      "loss": 0.7361,
+      "step": 7704
+    },
+    {
+      "epoch": 0.424073972150366,
+      "grad_norm": 0.6517061591148376,
+      "learning_rate": 8.952097753879181e-06,
+      "loss": 0.7127,
+      "step": 7705
+    },
+    {
+      "epoch": 0.42412901095272165,
+      "grad_norm": 0.7252761125564575,
+      "learning_rate": 8.951832212176692e-06,
+      "loss": 0.796,
+      "step": 7706
+    },
+    {
+      "epoch": 0.42418404975507734,
+      "grad_norm": 0.6688609719276428,
+      "learning_rate": 8.951566640773292e-06,
+      "loss": 0.7698,
+      "step": 7707
+    },
+    {
+      "epoch": 0.424239088557433,
+      "grad_norm": 0.7163566946983337,
+      "learning_rate": 8.951301039670974e-06,
+      "loss": 0.8069,
+      "step": 7708
+    },
+    {
+      "epoch": 0.42429412735978866,
+      "grad_norm": 0.7027623057365417,
+      "learning_rate": 8.951035408871735e-06,
+      "loss": 0.7061,
+      "step": 7709
+    },
+    {
+      "epoch": 0.4243491661621443,
+      "grad_norm": 0.9558683037757874,
+      "learning_rate": 8.950769748377572e-06,
+      "loss": 0.926,
+      "step": 7710
+    },
+    {
+      "epoch": 0.4244042049645,
+      "grad_norm": 0.7173893451690674,
+      "learning_rate": 8.950504058190482e-06,
+      "loss": 0.7519,
+      "step": 7711
+    },
+    {
+      "epoch": 0.4244592437668556,
+      "grad_norm": 0.8481128811836243,
+      "learning_rate": 8.950238338312459e-06,
+      "loss": 0.7804,
+      "step": 7712
+    },
+    {
+      "epoch": 0.4245142825692113,
+      "grad_norm": 0.6957072615623474,
+      "learning_rate": 8.949972588745502e-06,
+      "loss": 0.611,
+      "step": 7713
+    },
+    {
+      "epoch": 0.42456932137156694,
+      "grad_norm": 0.7910122871398926,
+      "learning_rate": 8.94970680949161e-06,
+      "loss": 0.8435,
+      "step": 7714
+    },
+    {
+      "epoch": 0.42462436017392263,
+      "grad_norm": 0.8068616986274719,
+      "learning_rate": 8.949441000552777e-06,
+      "loss": 0.8658,
+      "step": 7715
+    },
+    {
+      "epoch": 0.42467939897627827,
+      "grad_norm": 0.718110978603363,
+      "learning_rate": 8.949175161931006e-06,
+      "loss": 0.7908,
+      "step": 7716
+    },
+    {
+      "epoch": 0.42473443777863396,
+      "grad_norm": 0.7329656481742859,
+      "learning_rate": 8.948909293628289e-06,
+      "loss": 0.7477,
+      "step": 7717
+    },
+    {
+      "epoch": 0.4247894765809896,
+      "grad_norm": 0.7046940326690674,
+      "learning_rate": 8.948643395646625e-06,
+      "loss": 0.7985,
+      "step": 7718
+    },
+    {
+      "epoch": 0.4248445153833453,
+      "grad_norm": 0.6699581742286682,
+      "learning_rate": 8.948377467988017e-06,
+      "loss": 0.6575,
+      "step": 7719
+    },
+    {
+      "epoch": 0.4248995541857009,
+      "grad_norm": 0.8055217266082764,
+      "learning_rate": 8.94811151065446e-06,
+      "loss": 0.7008,
+      "step": 7720
+    },
+    {
+      "epoch": 0.4249545929880566,
+      "grad_norm": 0.8374543190002441,
+      "learning_rate": 8.947845523647954e-06,
+      "loss": 0.8918,
+      "step": 7721
+    },
+    {
+      "epoch": 0.42500963179041223,
+      "grad_norm": 0.6974833607673645,
+      "learning_rate": 8.947579506970498e-06,
+      "loss": 0.8594,
+      "step": 7722
+    },
+    {
+      "epoch": 0.4250646705927679,
+      "grad_norm": 0.7466567754745483,
+      "learning_rate": 8.947313460624091e-06,
+      "loss": 0.6935,
+      "step": 7723
+    },
+    {
+      "epoch": 0.42511970939512356,
+      "grad_norm": 0.8118101358413696,
+      "learning_rate": 8.947047384610734e-06,
+      "loss": 0.8432,
+      "step": 7724
+    },
+    {
+      "epoch": 0.42517474819747925,
+      "grad_norm": 0.6885644197463989,
+      "learning_rate": 8.946781278932422e-06,
+      "loss": 0.8059,
+      "step": 7725
+    },
+    {
+      "epoch": 0.4252297869998349,
+      "grad_norm": 0.7257012128829956,
+      "learning_rate": 8.94651514359116e-06,
+      "loss": 0.8239,
+      "step": 7726
+    },
+    {
+      "epoch": 0.42528482580219057,
+      "grad_norm": 1.311591386795044,
+      "learning_rate": 8.946248978588947e-06,
+      "loss": 0.8207,
+      "step": 7727
+    },
+    {
+      "epoch": 0.4253398646045462,
+      "grad_norm": 0.7694151997566223,
+      "learning_rate": 8.945982783927784e-06,
+      "loss": 0.8948,
+      "step": 7728
+    },
+    {
+      "epoch": 0.4253949034069019,
+      "grad_norm": 0.6922980546951294,
+      "learning_rate": 8.945716559609669e-06,
+      "loss": 0.7883,
+      "step": 7729
+    },
+    {
+      "epoch": 0.4254499422092575,
+      "grad_norm": 0.7803757786750793,
+      "learning_rate": 8.945450305636605e-06,
+      "loss": 0.9166,
+      "step": 7730
+    },
+    {
+      "epoch": 0.4255049810116132,
+      "grad_norm": 0.6775311827659607,
+      "learning_rate": 8.945184022010593e-06,
+      "loss": 0.6976,
+      "step": 7731
+    },
+    {
+      "epoch": 0.42556001981396885,
+      "grad_norm": 0.7108052968978882,
+      "learning_rate": 8.944917708733634e-06,
+      "loss": 0.7763,
+      "step": 7732
+    },
+    {
+      "epoch": 0.4256150586163245,
+      "grad_norm": 0.7215770483016968,
+      "learning_rate": 8.94465136580773e-06,
+      "loss": 0.7907,
+      "step": 7733
+    },
+    {
+      "epoch": 0.42567009741868017,
+      "grad_norm": 0.6690788865089417,
+      "learning_rate": 8.944384993234881e-06,
+      "loss": 0.8403,
+      "step": 7734
+    },
+    {
+      "epoch": 0.4257251362210358,
+      "grad_norm": 0.7372478246688843,
+      "learning_rate": 8.94411859101709e-06,
+      "loss": 0.7618,
+      "step": 7735
+    },
+    {
+      "epoch": 0.4257801750233915,
+      "grad_norm": 0.9398306608200073,
+      "learning_rate": 8.94385215915636e-06,
+      "loss": 0.9043,
+      "step": 7736
+    },
+    {
+      "epoch": 0.4258352138257471,
+      "grad_norm": 0.8790311217308044,
+      "learning_rate": 8.943585697654693e-06,
+      "loss": 0.9378,
+      "step": 7737
+    },
+    {
+      "epoch": 0.4258902526281028,
+      "grad_norm": 0.7579166889190674,
+      "learning_rate": 8.943319206514091e-06,
+      "loss": 0.7913,
+      "step": 7738
+    },
+    {
+      "epoch": 0.42594529143045845,
+      "grad_norm": 0.6426860690116882,
+      "learning_rate": 8.943052685736559e-06,
+      "loss": 0.744,
+      "step": 7739
+    },
+    {
+      "epoch": 0.42600033023281414,
+      "grad_norm": 0.688117265701294,
+      "learning_rate": 8.942786135324098e-06,
+      "loss": 0.8386,
+      "step": 7740
+    },
+    {
+      "epoch": 0.4260553690351698,
+      "grad_norm": 0.7178692817687988,
+      "learning_rate": 8.94251955527871e-06,
+      "loss": 0.7937,
+      "step": 7741
+    },
+    {
+      "epoch": 0.42611040783752546,
+      "grad_norm": 0.7980415225028992,
+      "learning_rate": 8.942252945602403e-06,
+      "loss": 0.76,
+      "step": 7742
+    },
+    {
+      "epoch": 0.4261654466398811,
+      "grad_norm": 0.6858333349227905,
+      "learning_rate": 8.941986306297175e-06,
+      "loss": 0.8155,
+      "step": 7743
+    },
+    {
+      "epoch": 0.4262204854422368,
+      "grad_norm": 0.763297975063324,
+      "learning_rate": 8.941719637365037e-06,
+      "loss": 0.8003,
+      "step": 7744
+    },
+    {
+      "epoch": 0.4262755242445924,
+      "grad_norm": 0.661016047000885,
+      "learning_rate": 8.941452938807986e-06,
+      "loss": 0.6788,
+      "step": 7745
+    },
+    {
+      "epoch": 0.4263305630469481,
+      "grad_norm": 0.7168089151382446,
+      "learning_rate": 8.94118621062803e-06,
+      "loss": 0.7791,
+      "step": 7746
+    },
+    {
+      "epoch": 0.42638560184930374,
+      "grad_norm": 0.6879743337631226,
+      "learning_rate": 8.940919452827174e-06,
+      "loss": 0.7978,
+      "step": 7747
+    },
+    {
+      "epoch": 0.42644064065165943,
+      "grad_norm": 0.672298014163971,
+      "learning_rate": 8.940652665407424e-06,
+      "loss": 0.7569,
+      "step": 7748
+    },
+    {
+      "epoch": 0.42649567945401506,
+      "grad_norm": 0.7237414717674255,
+      "learning_rate": 8.940385848370782e-06,
+      "loss": 0.6788,
+      "step": 7749
+    },
+    {
+      "epoch": 0.42655071825637075,
+      "grad_norm": 0.6793895363807678,
+      "learning_rate": 8.940119001719255e-06,
+      "loss": 0.749,
+      "step": 7750
+    },
+    {
+      "epoch": 0.4266057570587264,
+      "grad_norm": 1.1172789335250854,
+      "learning_rate": 8.939852125454847e-06,
+      "loss": 0.9017,
+      "step": 7751
+    },
+    {
+      "epoch": 0.4266607958610821,
+      "grad_norm": 0.7138717770576477,
+      "learning_rate": 8.939585219579567e-06,
+      "loss": 0.8586,
+      "step": 7752
+    },
+    {
+      "epoch": 0.4267158346634377,
+      "grad_norm": 0.8678629398345947,
+      "learning_rate": 8.939318284095417e-06,
+      "loss": 0.7333,
+      "step": 7753
+    },
+    {
+      "epoch": 0.4267708734657934,
+      "grad_norm": 0.7274941802024841,
+      "learning_rate": 8.939051319004407e-06,
+      "loss": 0.8426,
+      "step": 7754
+    },
+    {
+      "epoch": 0.42682591226814903,
+      "grad_norm": 0.6845358610153198,
+      "learning_rate": 8.93878432430854e-06,
+      "loss": 0.7731,
+      "step": 7755
+    },
+    {
+      "epoch": 0.4268809510705047,
+      "grad_norm": 0.7042781710624695,
+      "learning_rate": 8.938517300009826e-06,
+      "loss": 0.6703,
+      "step": 7756
+    },
+    {
+      "epoch": 0.42693598987286036,
+      "grad_norm": 0.7147190570831299,
+      "learning_rate": 8.93825024611027e-06,
+      "loss": 0.7977,
+      "step": 7757
+    },
+    {
+      "epoch": 0.42699102867521604,
+      "grad_norm": 0.6584187150001526,
+      "learning_rate": 8.93798316261188e-06,
+      "loss": 0.716,
+      "step": 7758
+    },
+    {
+      "epoch": 0.4270460674775717,
+      "grad_norm": 0.8061439990997314,
+      "learning_rate": 8.93771604951666e-06,
+      "loss": 0.9075,
+      "step": 7759
+    },
+    {
+      "epoch": 0.42710110627992737,
+      "grad_norm": 0.6741406917572021,
+      "learning_rate": 8.937448906826622e-06,
+      "loss": 0.7828,
+      "step": 7760
+    },
+    {
+      "epoch": 0.427156145082283,
+      "grad_norm": 0.8791692852973938,
+      "learning_rate": 8.937181734543773e-06,
+      "loss": 0.7685,
+      "step": 7761
+    },
+    {
+      "epoch": 0.4272111838846387,
+      "grad_norm": 0.6804112195968628,
+      "learning_rate": 8.936914532670119e-06,
+      "loss": 0.7672,
+      "step": 7762
+    },
+    {
+      "epoch": 0.4272662226869943,
+      "grad_norm": 0.6983451843261719,
+      "learning_rate": 8.936647301207668e-06,
+      "loss": 0.8228,
+      "step": 7763
+    },
+    {
+      "epoch": 0.42732126148935,
+      "grad_norm": 0.8248929977416992,
+      "learning_rate": 8.936380040158432e-06,
+      "loss": 0.7628,
+      "step": 7764
+    },
+    {
+      "epoch": 0.42737630029170565,
+      "grad_norm": 0.8324941992759705,
+      "learning_rate": 8.936112749524415e-06,
+      "loss": 0.8125,
+      "step": 7765
+    },
+    {
+      "epoch": 0.42743133909406134,
+      "grad_norm": 0.7489150762557983,
+      "learning_rate": 8.935845429307631e-06,
+      "loss": 0.8766,
+      "step": 7766
+    },
+    {
+      "epoch": 0.42748637789641697,
+      "grad_norm": 0.7323104739189148,
+      "learning_rate": 8.935578079510083e-06,
+      "loss": 0.8607,
+      "step": 7767
+    },
+    {
+      "epoch": 0.42754141669877266,
+      "grad_norm": 0.6825152635574341,
+      "learning_rate": 8.935310700133786e-06,
+      "loss": 0.7817,
+      "step": 7768
+    },
+    {
+      "epoch": 0.4275964555011283,
+      "grad_norm": 0.8928677439689636,
+      "learning_rate": 8.935043291180748e-06,
+      "loss": 0.7621,
+      "step": 7769
+    },
+    {
+      "epoch": 0.427651494303484,
+      "grad_norm": 0.7071405649185181,
+      "learning_rate": 8.934775852652975e-06,
+      "loss": 0.7798,
+      "step": 7770
+    },
+    {
+      "epoch": 0.4277065331058396,
+      "grad_norm": 0.8225427269935608,
+      "learning_rate": 8.934508384552481e-06,
+      "loss": 0.7212,
+      "step": 7771
+    },
+    {
+      "epoch": 0.4277615719081953,
+      "grad_norm": 0.6931234002113342,
+      "learning_rate": 8.934240886881276e-06,
+      "loss": 0.7301,
+      "step": 7772
+    },
+    {
+      "epoch": 0.42781661071055094,
+      "grad_norm": 0.6901859641075134,
+      "learning_rate": 8.933973359641369e-06,
+      "loss": 0.6974,
+      "step": 7773
+    },
+    {
+      "epoch": 0.4278716495129066,
+      "grad_norm": 0.7736960649490356,
+      "learning_rate": 8.93370580283477e-06,
+      "loss": 0.6562,
+      "step": 7774
+    },
+    {
+      "epoch": 0.42792668831526226,
+      "grad_norm": 0.7363499999046326,
+      "learning_rate": 8.933438216463495e-06,
+      "loss": 0.8274,
+      "step": 7775
+    },
+    {
+      "epoch": 0.4279817271176179,
+      "grad_norm": 0.6855602860450745,
+      "learning_rate": 8.933170600529548e-06,
+      "loss": 0.7576,
+      "step": 7776
+    },
+    {
+      "epoch": 0.4280367659199736,
+      "grad_norm": 0.7641676664352417,
+      "learning_rate": 8.932902955034945e-06,
+      "loss": 0.7837,
+      "step": 7777
+    },
+    {
+      "epoch": 0.4280918047223292,
+      "grad_norm": 0.74812251329422,
+      "learning_rate": 8.932635279981695e-06,
+      "loss": 0.8402,
+      "step": 7778
+    },
+    {
+      "epoch": 0.4281468435246849,
+      "grad_norm": 0.7445259094238281,
+      "learning_rate": 8.932367575371813e-06,
+      "loss": 0.862,
+      "step": 7779
+    },
+    {
+      "epoch": 0.42820188232704054,
+      "grad_norm": 0.8977177739143372,
+      "learning_rate": 8.932099841207306e-06,
+      "loss": 0.7735,
+      "step": 7780
+    },
+    {
+      "epoch": 0.42825692112939623,
+      "grad_norm": 0.74172043800354,
+      "learning_rate": 8.93183207749019e-06,
+      "loss": 0.7053,
+      "step": 7781
+    },
+    {
+      "epoch": 0.42831195993175186,
+      "grad_norm": 0.6670083999633789,
+      "learning_rate": 8.931564284222479e-06,
+      "loss": 0.6348,
+      "step": 7782
+    },
+    {
+      "epoch": 0.42836699873410755,
+      "grad_norm": 0.7575422525405884,
+      "learning_rate": 8.93129646140618e-06,
+      "loss": 0.9354,
+      "step": 7783
+    },
+    {
+      "epoch": 0.4284220375364632,
+      "grad_norm": 0.7436977624893188,
+      "learning_rate": 8.931028609043311e-06,
+      "loss": 0.7461,
+      "step": 7784
+    },
+    {
+      "epoch": 0.4284770763388189,
+      "grad_norm": 0.7383070588111877,
+      "learning_rate": 8.930760727135882e-06,
+      "loss": 0.7629,
+      "step": 7785
+    },
+    {
+      "epoch": 0.4285321151411745,
+      "grad_norm": 0.6926067471504211,
+      "learning_rate": 8.93049281568591e-06,
+      "loss": 0.6788,
+      "step": 7786
+    },
+    {
+      "epoch": 0.4285871539435302,
+      "grad_norm": 0.7680530548095703,
+      "learning_rate": 8.930224874695404e-06,
+      "loss": 0.722,
+      "step": 7787
+    },
+    {
+      "epoch": 0.42864219274588583,
+      "grad_norm": 0.9880867004394531,
+      "learning_rate": 8.92995690416638e-06,
+      "loss": 0.833,
+      "step": 7788
+    },
+    {
+      "epoch": 0.4286972315482415,
+      "grad_norm": 0.7915430068969727,
+      "learning_rate": 8.929688904100853e-06,
+      "loss": 0.7643,
+      "step": 7789
+    },
+    {
+      "epoch": 0.42875227035059715,
+      "grad_norm": 0.6972275376319885,
+      "learning_rate": 8.929420874500836e-06,
+      "loss": 0.7697,
+      "step": 7790
+    },
+    {
+      "epoch": 0.42880730915295284,
+      "grad_norm": 0.9583331346511841,
+      "learning_rate": 8.929152815368343e-06,
+      "loss": 0.7591,
+      "step": 7791
+    },
+    {
+      "epoch": 0.4288623479553085,
+      "grad_norm": 0.7254299521446228,
+      "learning_rate": 8.928884726705388e-06,
+      "loss": 0.7913,
+      "step": 7792
+    },
+    {
+      "epoch": 0.42891738675766417,
+      "grad_norm": 0.7925865054130554,
+      "learning_rate": 8.928616608513989e-06,
+      "loss": 0.8248,
+      "step": 7793
+    },
+    {
+      "epoch": 0.4289724255600198,
+      "grad_norm": 0.9367457628250122,
+      "learning_rate": 8.928348460796157e-06,
+      "loss": 0.7767,
+      "step": 7794
+    },
+    {
+      "epoch": 0.4290274643623755,
+      "grad_norm": 0.8511868119239807,
+      "learning_rate": 8.928080283553912e-06,
+      "loss": 0.841,
+      "step": 7795
+    },
+    {
+      "epoch": 0.4290825031647311,
+      "grad_norm": 0.8518061637878418,
+      "learning_rate": 8.927812076789267e-06,
+      "loss": 0.7907,
+      "step": 7796
+    },
+    {
+      "epoch": 0.4291375419670868,
+      "grad_norm": 0.7208365797996521,
+      "learning_rate": 8.927543840504236e-06,
+      "loss": 0.7344,
+      "step": 7797
+    },
+    {
+      "epoch": 0.42919258076944244,
+      "grad_norm": 0.7541850209236145,
+      "learning_rate": 8.927275574700838e-06,
+      "loss": 0.7724,
+      "step": 7798
+    },
+    {
+      "epoch": 0.42924761957179813,
+      "grad_norm": 0.7378629446029663,
+      "learning_rate": 8.927007279381087e-06,
+      "loss": 0.7614,
+      "step": 7799
+    },
+    {
+      "epoch": 0.42930265837415377,
+      "grad_norm": 0.7358561158180237,
+      "learning_rate": 8.926738954547001e-06,
+      "loss": 0.7288,
+      "step": 7800
+    },
+    {
+      "epoch": 0.42935769717650946,
+      "grad_norm": 0.7385967969894409,
+      "learning_rate": 8.926470600200597e-06,
+      "loss": 0.7562,
+      "step": 7801
+    },
+    {
+      "epoch": 0.4294127359788651,
+      "grad_norm": 0.6904877424240112,
+      "learning_rate": 8.92620221634389e-06,
+      "loss": 0.6507,
+      "step": 7802
+    },
+    {
+      "epoch": 0.4294677747812208,
+      "grad_norm": 0.7205148935317993,
+      "learning_rate": 8.925933802978898e-06,
+      "loss": 0.7683,
+      "step": 7803
+    },
+    {
+      "epoch": 0.4295228135835764,
+      "grad_norm": 0.6830344200134277,
+      "learning_rate": 8.925665360107639e-06,
+      "loss": 0.6886,
+      "step": 7804
+    },
+    {
+      "epoch": 0.4295778523859321,
+      "grad_norm": 0.7648812532424927,
+      "learning_rate": 8.92539688773213e-06,
+      "loss": 0.7559,
+      "step": 7805
+    },
+    {
+      "epoch": 0.42963289118828774,
+      "grad_norm": 0.7819112539291382,
+      "learning_rate": 8.925128385854389e-06,
+      "loss": 0.7443,
+      "step": 7806
+    },
+    {
+      "epoch": 0.4296879299906434,
+      "grad_norm": 0.6742433309555054,
+      "learning_rate": 8.924859854476433e-06,
+      "loss": 0.7191,
+      "step": 7807
+    },
+    {
+      "epoch": 0.42974296879299906,
+      "grad_norm": 0.7368177771568298,
+      "learning_rate": 8.924591293600281e-06,
+      "loss": 0.6946,
+      "step": 7808
+    },
+    {
+      "epoch": 0.42979800759535475,
+      "grad_norm": 0.663112998008728,
+      "learning_rate": 8.924322703227953e-06,
+      "loss": 0.7405,
+      "step": 7809
+    },
+    {
+      "epoch": 0.4298530463977104,
+      "grad_norm": 0.6735410690307617,
+      "learning_rate": 8.924054083361465e-06,
+      "loss": 0.7982,
+      "step": 7810
+    },
+    {
+      "epoch": 0.42990808520006607,
+      "grad_norm": 0.7770369648933411,
+      "learning_rate": 8.923785434002834e-06,
+      "loss": 0.9179,
+      "step": 7811
+    },
+    {
+      "epoch": 0.4299631240024217,
+      "grad_norm": 0.7464482188224792,
+      "learning_rate": 8.923516755154085e-06,
+      "loss": 0.8514,
+      "step": 7812
+    },
+    {
+      "epoch": 0.4300181628047774,
+      "grad_norm": 0.9249551892280579,
+      "learning_rate": 8.923248046817235e-06,
+      "loss": 0.8287,
+      "step": 7813
+    },
+    {
+      "epoch": 0.430073201607133,
+      "grad_norm": 0.7071338891983032,
+      "learning_rate": 8.922979308994302e-06,
+      "loss": 0.7509,
+      "step": 7814
+    },
+    {
+      "epoch": 0.4301282404094887,
+      "grad_norm": 0.6910794377326965,
+      "learning_rate": 8.922710541687305e-06,
+      "loss": 0.7373,
+      "step": 7815
+    },
+    {
+      "epoch": 0.43018327921184435,
+      "grad_norm": 0.8424028158187866,
+      "learning_rate": 8.922441744898267e-06,
+      "loss": 0.741,
+      "step": 7816
+    },
+    {
+      "epoch": 0.43023831801420004,
+      "grad_norm": 0.8162125945091248,
+      "learning_rate": 8.922172918629208e-06,
+      "loss": 0.8044,
+      "step": 7817
+    },
+    {
+      "epoch": 0.43029335681655567,
+      "grad_norm": 0.7415170669555664,
+      "learning_rate": 8.921904062882145e-06,
+      "loss": 0.7427,
+      "step": 7818
+    },
+    {
+      "epoch": 0.4303483956189113,
+      "grad_norm": 1.1357808113098145,
+      "learning_rate": 8.921635177659103e-06,
+      "loss": 0.7802,
+      "step": 7819
+    },
+    {
+      "epoch": 0.430403434421267,
+      "grad_norm": 0.7039839625358582,
+      "learning_rate": 8.9213662629621e-06,
+      "loss": 0.7368,
+      "step": 7820
+    },
+    {
+      "epoch": 0.43045847322362263,
+      "grad_norm": 0.721077024936676,
+      "learning_rate": 8.921097318793157e-06,
+      "loss": 0.6575,
+      "step": 7821
+    },
+    {
+      "epoch": 0.4305135120259783,
+      "grad_norm": 0.7823510766029358,
+      "learning_rate": 8.920828345154297e-06,
+      "loss": 0.7499,
+      "step": 7822
+    },
+    {
+      "epoch": 0.43056855082833395,
+      "grad_norm": 0.6400569677352905,
+      "learning_rate": 8.920559342047539e-06,
+      "loss": 0.7091,
+      "step": 7823
+    },
+    {
+      "epoch": 0.43062358963068964,
+      "grad_norm": 0.8974951505661011,
+      "learning_rate": 8.920290309474908e-06,
+      "loss": 0.7228,
+      "step": 7824
+    },
+    {
+      "epoch": 0.4306786284330453,
+      "grad_norm": 0.8176010847091675,
+      "learning_rate": 8.920021247438426e-06,
+      "loss": 0.8852,
+      "step": 7825
+    },
+    {
+      "epoch": 0.43073366723540096,
+      "grad_norm": 0.7591422200202942,
+      "learning_rate": 8.919752155940112e-06,
+      "loss": 0.8382,
+      "step": 7826
+    },
+    {
+      "epoch": 0.4307887060377566,
+      "grad_norm": 0.7089776396751404,
+      "learning_rate": 8.919483034981988e-06,
+      "loss": 0.7188,
+      "step": 7827
+    },
+    {
+      "epoch": 0.4308437448401123,
+      "grad_norm": 0.7328840494155884,
+      "learning_rate": 8.919213884566081e-06,
+      "loss": 0.7609,
+      "step": 7828
+    },
+    {
+      "epoch": 0.4308987836424679,
+      "grad_norm": 0.6473509669303894,
+      "learning_rate": 8.918944704694411e-06,
+      "loss": 0.7027,
+      "step": 7829
+    },
+    {
+      "epoch": 0.4309538224448236,
+      "grad_norm": 0.6585624814033508,
+      "learning_rate": 8.918675495369003e-06,
+      "loss": 0.7133,
+      "step": 7830
+    },
+    {
+      "epoch": 0.43100886124717924,
+      "grad_norm": 0.7232397794723511,
+      "learning_rate": 8.918406256591876e-06,
+      "loss": 0.7458,
+      "step": 7831
+    },
+    {
+      "epoch": 0.43106390004953493,
+      "grad_norm": 0.8752645254135132,
+      "learning_rate": 8.918136988365059e-06,
+      "loss": 0.671,
+      "step": 7832
+    },
+    {
+      "epoch": 0.43111893885189057,
+      "grad_norm": 0.7890885472297668,
+      "learning_rate": 8.917867690690573e-06,
+      "loss": 0.7674,
+      "step": 7833
+    },
+    {
+      "epoch": 0.43117397765424625,
+      "grad_norm": 0.6725128293037415,
+      "learning_rate": 8.917598363570441e-06,
+      "loss": 0.7373,
+      "step": 7834
+    },
+    {
+      "epoch": 0.4312290164566019,
+      "grad_norm": 0.808897852897644,
+      "learning_rate": 8.917329007006688e-06,
+      "loss": 0.8397,
+      "step": 7835
+    },
+    {
+      "epoch": 0.4312840552589576,
+      "grad_norm": 0.7268605828285217,
+      "learning_rate": 8.91705962100134e-06,
+      "loss": 0.7957,
+      "step": 7836
+    },
+    {
+      "epoch": 0.4313390940613132,
+      "grad_norm": 0.7336069345474243,
+      "learning_rate": 8.916790205556421e-06,
+      "loss": 0.746,
+      "step": 7837
+    },
+    {
+      "epoch": 0.4313941328636689,
+      "grad_norm": 0.7380902171134949,
+      "learning_rate": 8.916520760673955e-06,
+      "loss": 0.674,
+      "step": 7838
+    },
+    {
+      "epoch": 0.43144917166602453,
+      "grad_norm": 0.8041831851005554,
+      "learning_rate": 8.916251286355967e-06,
+      "loss": 0.8392,
+      "step": 7839
+    },
+    {
+      "epoch": 0.4315042104683802,
+      "grad_norm": 0.6745681166648865,
+      "learning_rate": 8.915981782604481e-06,
+      "loss": 0.7676,
+      "step": 7840
+    },
+    {
+      "epoch": 0.43155924927073586,
+      "grad_norm": 0.6572039127349854,
+      "learning_rate": 8.915712249421526e-06,
+      "loss": 0.7471,
+      "step": 7841
+    },
+    {
+      "epoch": 0.43161428807309155,
+      "grad_norm": 0.7250062227249146,
+      "learning_rate": 8.915442686809124e-06,
+      "loss": 0.8566,
+      "step": 7842
+    },
+    {
+      "epoch": 0.4316693268754472,
+      "grad_norm": 0.7008941769599915,
+      "learning_rate": 8.915173094769306e-06,
+      "loss": 0.7876,
+      "step": 7843
+    },
+    {
+      "epoch": 0.43172436567780287,
+      "grad_norm": 0.7078337073326111,
+      "learning_rate": 8.914903473304093e-06,
+      "loss": 0.756,
+      "step": 7844
+    },
+    {
+      "epoch": 0.4317794044801585,
+      "grad_norm": 0.7822949886322021,
+      "learning_rate": 8.914633822415513e-06,
+      "loss": 0.9423,
+      "step": 7845
+    },
+    {
+      "epoch": 0.4318344432825142,
+      "grad_norm": 0.6707580089569092,
+      "learning_rate": 8.914364142105593e-06,
+      "loss": 0.639,
+      "step": 7846
+    },
+    {
+      "epoch": 0.4318894820848698,
+      "grad_norm": 0.7868423461914062,
+      "learning_rate": 8.914094432376362e-06,
+      "loss": 0.7768,
+      "step": 7847
+    },
+    {
+      "epoch": 0.4319445208872255,
+      "grad_norm": 0.6147592067718506,
+      "learning_rate": 8.913824693229845e-06,
+      "loss": 0.6693,
+      "step": 7848
+    },
+    {
+      "epoch": 0.43199955968958115,
+      "grad_norm": 0.6901249885559082,
+      "learning_rate": 8.913554924668067e-06,
+      "loss": 0.7779,
+      "step": 7849
+    },
+    {
+      "epoch": 0.43205459849193684,
+      "grad_norm": 0.7062137126922607,
+      "learning_rate": 8.913285126693058e-06,
+      "loss": 0.7951,
+      "step": 7850
+    },
+    {
+      "epoch": 0.43210963729429247,
+      "grad_norm": 0.6363390684127808,
+      "learning_rate": 8.913015299306846e-06,
+      "loss": 0.6723,
+      "step": 7851
+    },
+    {
+      "epoch": 0.43216467609664816,
+      "grad_norm": 0.7168677449226379,
+      "learning_rate": 8.912745442511459e-06,
+      "loss": 0.7442,
+      "step": 7852
+    },
+    {
+      "epoch": 0.4322197148990038,
+      "grad_norm": 0.7347995042800903,
+      "learning_rate": 8.912475556308925e-06,
+      "loss": 0.8361,
+      "step": 7853
+    },
+    {
+      "epoch": 0.4322747537013595,
+      "grad_norm": 0.683777391910553,
+      "learning_rate": 8.91220564070127e-06,
+      "loss": 0.7583,
+      "step": 7854
+    },
+    {
+      "epoch": 0.4323297925037151,
+      "grad_norm": 0.7436330914497375,
+      "learning_rate": 8.911935695690527e-06,
+      "loss": 0.8414,
+      "step": 7855
+    },
+    {
+      "epoch": 0.4323848313060708,
+      "grad_norm": 0.7748109102249146,
+      "learning_rate": 8.911665721278721e-06,
+      "loss": 0.7812,
+      "step": 7856
+    },
+    {
+      "epoch": 0.43243987010842644,
+      "grad_norm": 0.7984411120414734,
+      "learning_rate": 8.911395717467883e-06,
+      "loss": 0.6845,
+      "step": 7857
+    },
+    {
+      "epoch": 0.4324949089107821,
+      "grad_norm": 0.680144727230072,
+      "learning_rate": 8.911125684260042e-06,
+      "loss": 0.7156,
+      "step": 7858
+    },
+    {
+      "epoch": 0.43254994771313776,
+      "grad_norm": 0.7738325595855713,
+      "learning_rate": 8.910855621657228e-06,
+      "loss": 0.7295,
+      "step": 7859
+    },
+    {
+      "epoch": 0.43260498651549345,
+      "grad_norm": 0.7276971340179443,
+      "learning_rate": 8.910585529661469e-06,
+      "loss": 0.7982,
+      "step": 7860
+    },
+    {
+      "epoch": 0.4326600253178491,
+      "grad_norm": 0.7655037641525269,
+      "learning_rate": 8.910315408274796e-06,
+      "loss": 0.8416,
+      "step": 7861
+    },
+    {
+      "epoch": 0.4327150641202047,
+      "grad_norm": 0.7220892906188965,
+      "learning_rate": 8.910045257499238e-06,
+      "loss": 0.8002,
+      "step": 7862
+    },
+    {
+      "epoch": 0.4327701029225604,
+      "grad_norm": 0.6255655884742737,
+      "learning_rate": 8.90977507733683e-06,
+      "loss": 0.6477,
+      "step": 7863
+    },
+    {
+      "epoch": 0.43282514172491604,
+      "grad_norm": 0.649472713470459,
+      "learning_rate": 8.909504867789594e-06,
+      "loss": 0.6838,
+      "step": 7864
+    },
+    {
+      "epoch": 0.43288018052727173,
+      "grad_norm": 0.6915234923362732,
+      "learning_rate": 8.909234628859568e-06,
+      "loss": 0.7146,
+      "step": 7865
+    },
+    {
+      "epoch": 0.43293521932962736,
+      "grad_norm": 0.7120145559310913,
+      "learning_rate": 8.908964360548783e-06,
+      "loss": 0.7782,
+      "step": 7866
+    },
+    {
+      "epoch": 0.43299025813198305,
+      "grad_norm": 0.8125410079956055,
+      "learning_rate": 8.908694062859267e-06,
+      "loss": 0.7514,
+      "step": 7867
+    },
+    {
+      "epoch": 0.4330452969343387,
+      "grad_norm": 0.6821436882019043,
+      "learning_rate": 8.908423735793053e-06,
+      "loss": 0.8074,
+      "step": 7868
+    },
+    {
+      "epoch": 0.4331003357366944,
+      "grad_norm": 0.8079590201377869,
+      "learning_rate": 8.908153379352171e-06,
+      "loss": 0.7932,
+      "step": 7869
+    },
+    {
+      "epoch": 0.43315537453905,
+      "grad_norm": 0.676013708114624,
+      "learning_rate": 8.907882993538655e-06,
+      "loss": 0.6611,
+      "step": 7870
+    },
+    {
+      "epoch": 0.4332104133414057,
+      "grad_norm": 0.706624448299408,
+      "learning_rate": 8.907612578354537e-06,
+      "loss": 0.8241,
+      "step": 7871
+    },
+    {
+      "epoch": 0.43326545214376133,
+      "grad_norm": 0.6533300876617432,
+      "learning_rate": 8.907342133801848e-06,
+      "loss": 0.6969,
+      "step": 7872
+    },
+    {
+      "epoch": 0.433320490946117,
+      "grad_norm": 0.6778282523155212,
+      "learning_rate": 8.907071659882622e-06,
+      "loss": 0.6877,
+      "step": 7873
+    },
+    {
+      "epoch": 0.43337552974847265,
+      "grad_norm": 0.7068879008293152,
+      "learning_rate": 8.906801156598892e-06,
+      "loss": 0.7912,
+      "step": 7874
+    },
+    {
+      "epoch": 0.43343056855082834,
+      "grad_norm": 0.6620263457298279,
+      "learning_rate": 8.90653062395269e-06,
+      "loss": 0.7317,
+      "step": 7875
+    },
+    {
+      "epoch": 0.433485607353184,
+      "grad_norm": 0.7084807753562927,
+      "learning_rate": 8.906260061946049e-06,
+      "loss": 0.7268,
+      "step": 7876
+    },
+    {
+      "epoch": 0.43354064615553967,
+      "grad_norm": 0.7899147272109985,
+      "learning_rate": 8.905989470581003e-06,
+      "loss": 0.8258,
+      "step": 7877
+    },
+    {
+      "epoch": 0.4335956849578953,
+      "grad_norm": 0.6657128930091858,
+      "learning_rate": 8.905718849859585e-06,
+      "loss": 0.6564,
+      "step": 7878
+    },
+    {
+      "epoch": 0.433650723760251,
+      "grad_norm": 0.8737723231315613,
+      "learning_rate": 8.905448199783831e-06,
+      "loss": 0.8646,
+      "step": 7879
+    },
+    {
+      "epoch": 0.4337057625626066,
+      "grad_norm": 0.7517673969268799,
+      "learning_rate": 8.905177520355775e-06,
+      "loss": 0.7658,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4337608013649623,
+      "grad_norm": 0.6724270582199097,
+      "learning_rate": 8.904906811577447e-06,
+      "loss": 0.7509,
+      "step": 7881
+    },
+    {
+      "epoch": 0.43381584016731795,
+      "grad_norm": 0.6490511894226074,
+      "learning_rate": 8.904636073450885e-06,
+      "loss": 0.7282,
+      "step": 7882
+    },
+    {
+      "epoch": 0.43387087896967363,
+      "grad_norm": 0.73885178565979,
+      "learning_rate": 8.904365305978126e-06,
+      "loss": 0.7575,
+      "step": 7883
+    },
+    {
+      "epoch": 0.43392591777202927,
+      "grad_norm": 0.6823462843894958,
+      "learning_rate": 8.9040945091612e-06,
+      "loss": 0.7566,
+      "step": 7884
+    },
+    {
+      "epoch": 0.43398095657438496,
+      "grad_norm": 0.6705971956253052,
+      "learning_rate": 8.903823683002146e-06,
+      "loss": 0.7726,
+      "step": 7885
+    },
+    {
+      "epoch": 0.4340359953767406,
+      "grad_norm": 0.6898428201675415,
+      "learning_rate": 8.903552827502998e-06,
+      "loss": 0.7545,
+      "step": 7886
+    },
+    {
+      "epoch": 0.4340910341790963,
+      "grad_norm": 0.810357928276062,
+      "learning_rate": 8.90328194266579e-06,
+      "loss": 0.8883,
+      "step": 7887
+    },
+    {
+      "epoch": 0.4341460729814519,
+      "grad_norm": 0.6505162119865417,
+      "learning_rate": 8.903011028492563e-06,
+      "loss": 0.7205,
+      "step": 7888
+    },
+    {
+      "epoch": 0.4342011117838076,
+      "grad_norm": 0.8401693105697632,
+      "learning_rate": 8.902740084985348e-06,
+      "loss": 0.8105,
+      "step": 7889
+    },
+    {
+      "epoch": 0.43425615058616324,
+      "grad_norm": 0.7151880860328674,
+      "learning_rate": 8.902469112146183e-06,
+      "loss": 0.7748,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4343111893885189,
+      "grad_norm": 0.7257007956504822,
+      "learning_rate": 8.902198109977107e-06,
+      "loss": 0.7818,
+      "step": 7891
+    },
+    {
+      "epoch": 0.43436622819087456,
+      "grad_norm": 0.786691427230835,
+      "learning_rate": 8.901927078480153e-06,
+      "loss": 0.8527,
+      "step": 7892
+    },
+    {
+      "epoch": 0.43442126699323025,
+      "grad_norm": 0.7420910596847534,
+      "learning_rate": 8.901656017657358e-06,
+      "loss": 0.7087,
+      "step": 7893
+    },
+    {
+      "epoch": 0.4344763057955859,
+      "grad_norm": 0.6713958978652954,
+      "learning_rate": 8.901384927510763e-06,
+      "loss": 0.7366,
+      "step": 7894
+    },
+    {
+      "epoch": 0.43453134459794157,
+      "grad_norm": 1.0276658535003662,
+      "learning_rate": 8.901113808042402e-06,
+      "loss": 0.7462,
+      "step": 7895
+    },
+    {
+      "epoch": 0.4345863834002972,
+      "grad_norm": 0.7207444906234741,
+      "learning_rate": 8.900842659254314e-06,
+      "loss": 0.6777,
+      "step": 7896
+    },
+    {
+      "epoch": 0.4346414222026529,
+      "grad_norm": 0.7581979036331177,
+      "learning_rate": 8.900571481148538e-06,
+      "loss": 0.8081,
+      "step": 7897
+    },
+    {
+      "epoch": 0.4346964610050085,
+      "grad_norm": 0.9224075675010681,
+      "learning_rate": 8.90030027372711e-06,
+      "loss": 0.892,
+      "step": 7898
+    },
+    {
+      "epoch": 0.4347514998073642,
+      "grad_norm": 0.6844260096549988,
+      "learning_rate": 8.900029036992069e-06,
+      "loss": 0.8063,
+      "step": 7899
+    },
+    {
+      "epoch": 0.43480653860971985,
+      "grad_norm": 0.7008691430091858,
+      "learning_rate": 8.899757770945453e-06,
+      "loss": 0.6998,
+      "step": 7900
+    },
+    {
+      "epoch": 0.43486157741207554,
+      "grad_norm": 0.7311949729919434,
+      "learning_rate": 8.899486475589303e-06,
+      "loss": 0.7724,
+      "step": 7901
+    },
+    {
+      "epoch": 0.4349166162144312,
+      "grad_norm": 0.7441468238830566,
+      "learning_rate": 8.899215150925656e-06,
+      "loss": 0.7728,
+      "step": 7902
+    },
+    {
+      "epoch": 0.43497165501678686,
+      "grad_norm": 0.7405179142951965,
+      "learning_rate": 8.89894379695655e-06,
+      "loss": 0.8267,
+      "step": 7903
+    },
+    {
+      "epoch": 0.4350266938191425,
+      "grad_norm": 0.6967620253562927,
+      "learning_rate": 8.898672413684029e-06,
+      "loss": 0.7284,
+      "step": 7904
+    },
+    {
+      "epoch": 0.43508173262149813,
+      "grad_norm": 0.8979219794273376,
+      "learning_rate": 8.898401001110127e-06,
+      "loss": 0.8267,
+      "step": 7905
+    },
+    {
+      "epoch": 0.4351367714238538,
+      "grad_norm": 0.7905356884002686,
+      "learning_rate": 8.898129559236888e-06,
+      "loss": 0.8011,
+      "step": 7906
+    },
+    {
+      "epoch": 0.43519181022620945,
+      "grad_norm": 0.6740859150886536,
+      "learning_rate": 8.897858088066351e-06,
+      "loss": 0.6597,
+      "step": 7907
+    },
+    {
+      "epoch": 0.43524684902856514,
+      "grad_norm": 0.7451572418212891,
+      "learning_rate": 8.897586587600555e-06,
+      "loss": 0.7466,
+      "step": 7908
+    },
+    {
+      "epoch": 0.4353018878309208,
+      "grad_norm": 0.7726565003395081,
+      "learning_rate": 8.897315057841542e-06,
+      "loss": 0.7873,
+      "step": 7909
+    },
+    {
+      "epoch": 0.43535692663327646,
+      "grad_norm": 0.8348171710968018,
+      "learning_rate": 8.897043498791354e-06,
+      "loss": 0.7583,
+      "step": 7910
+    },
+    {
+      "epoch": 0.4354119654356321,
+      "grad_norm": 0.6714087724685669,
+      "learning_rate": 8.896771910452027e-06,
+      "loss": 0.7909,
+      "step": 7911
+    },
+    {
+      "epoch": 0.4354670042379878,
+      "grad_norm": 0.7397969365119934,
+      "learning_rate": 8.896500292825607e-06,
+      "loss": 0.7734,
+      "step": 7912
+    },
+    {
+      "epoch": 0.4355220430403434,
+      "grad_norm": 0.6806391477584839,
+      "learning_rate": 8.896228645914133e-06,
+      "loss": 0.7898,
+      "step": 7913
+    },
+    {
+      "epoch": 0.4355770818426991,
+      "grad_norm": 0.7135224342346191,
+      "learning_rate": 8.89595696971965e-06,
+      "loss": 0.7453,
+      "step": 7914
+    },
+    {
+      "epoch": 0.43563212064505474,
+      "grad_norm": 0.8275992274284363,
+      "learning_rate": 8.895685264244195e-06,
+      "loss": 0.7326,
+      "step": 7915
+    },
+    {
+      "epoch": 0.43568715944741043,
+      "grad_norm": 0.7254159450531006,
+      "learning_rate": 8.895413529489813e-06,
+      "loss": 0.7523,
+      "step": 7916
+    },
+    {
+      "epoch": 0.43574219824976607,
+      "grad_norm": 0.8060647249221802,
+      "learning_rate": 8.895141765458546e-06,
+      "loss": 0.7878,
+      "step": 7917
+    },
+    {
+      "epoch": 0.43579723705212176,
+      "grad_norm": 0.7007316946983337,
+      "learning_rate": 8.894869972152435e-06,
+      "loss": 0.7837,
+      "step": 7918
+    },
+    {
+      "epoch": 0.4358522758544774,
+      "grad_norm": 0.6874841451644897,
+      "learning_rate": 8.894598149573524e-06,
+      "loss": 0.7773,
+      "step": 7919
+    },
+    {
+      "epoch": 0.4359073146568331,
+      "grad_norm": 0.7557696104049683,
+      "learning_rate": 8.894326297723856e-06,
+      "loss": 0.6905,
+      "step": 7920
+    },
+    {
+      "epoch": 0.4359623534591887,
+      "grad_norm": 0.7589512467384338,
+      "learning_rate": 8.894054416605475e-06,
+      "loss": 0.8292,
+      "step": 7921
+    },
+    {
+      "epoch": 0.4360173922615444,
+      "grad_norm": 0.9062818884849548,
+      "learning_rate": 8.893782506220424e-06,
+      "loss": 0.9149,
+      "step": 7922
+    },
+    {
+      "epoch": 0.43607243106390003,
+      "grad_norm": 0.7553420662879944,
+      "learning_rate": 8.893510566570744e-06,
+      "loss": 0.7256,
+      "step": 7923
+    },
+    {
+      "epoch": 0.4361274698662557,
+      "grad_norm": 0.7130489349365234,
+      "learning_rate": 8.89323859765848e-06,
+      "loss": 0.7375,
+      "step": 7924
+    },
+    {
+      "epoch": 0.43618250866861136,
+      "grad_norm": 0.6234793066978455,
+      "learning_rate": 8.89296659948568e-06,
+      "loss": 0.716,
+      "step": 7925
+    },
+    {
+      "epoch": 0.43623754747096705,
+      "grad_norm": 0.7527539134025574,
+      "learning_rate": 8.892694572054383e-06,
+      "loss": 0.7884,
+      "step": 7926
+    },
+    {
+      "epoch": 0.4362925862733227,
+      "grad_norm": 0.7677647471427917,
+      "learning_rate": 8.892422515366636e-06,
+      "loss": 0.7136,
+      "step": 7927
+    },
+    {
+      "epoch": 0.43634762507567837,
+      "grad_norm": 0.7212143540382385,
+      "learning_rate": 8.892150429424484e-06,
+      "loss": 0.8113,
+      "step": 7928
+    },
+    {
+      "epoch": 0.436402663878034,
+      "grad_norm": 0.6735568046569824,
+      "learning_rate": 8.89187831422997e-06,
+      "loss": 0.6472,
+      "step": 7929
+    },
+    {
+      "epoch": 0.4364577026803897,
+      "grad_norm": 0.7120702862739563,
+      "learning_rate": 8.891606169785141e-06,
+      "loss": 0.8032,
+      "step": 7930
+    },
+    {
+      "epoch": 0.4365127414827453,
+      "grad_norm": 0.679499089717865,
+      "learning_rate": 8.891333996092041e-06,
+      "loss": 0.7366,
+      "step": 7931
+    },
+    {
+      "epoch": 0.436567780285101,
+      "grad_norm": 0.7774114012718201,
+      "learning_rate": 8.891061793152718e-06,
+      "loss": 0.7917,
+      "step": 7932
+    },
+    {
+      "epoch": 0.43662281908745665,
+      "grad_norm": 0.6951174139976501,
+      "learning_rate": 8.890789560969216e-06,
+      "loss": 0.7518,
+      "step": 7933
+    },
+    {
+      "epoch": 0.43667785788981234,
+      "grad_norm": 0.7645227909088135,
+      "learning_rate": 8.89051729954358e-06,
+      "loss": 0.7787,
+      "step": 7934
+    },
+    {
+      "epoch": 0.43673289669216797,
+      "grad_norm": 0.7127084732055664,
+      "learning_rate": 8.890245008877857e-06,
+      "loss": 0.8137,
+      "step": 7935
+    },
+    {
+      "epoch": 0.43678793549452366,
+      "grad_norm": 0.7541413903236389,
+      "learning_rate": 8.889972688974095e-06,
+      "loss": 0.776,
+      "step": 7936
+    },
+    {
+      "epoch": 0.4368429742968793,
+      "grad_norm": 0.690963625907898,
+      "learning_rate": 8.889700339834339e-06,
+      "loss": 0.7691,
+      "step": 7937
+    },
+    {
+      "epoch": 0.436898013099235,
+      "grad_norm": 0.750221848487854,
+      "learning_rate": 8.889427961460636e-06,
+      "loss": 0.7831,
+      "step": 7938
+    },
+    {
+      "epoch": 0.4369530519015906,
+      "grad_norm": 0.7255545854568481,
+      "learning_rate": 8.889155553855035e-06,
+      "loss": 0.7831,
+      "step": 7939
+    },
+    {
+      "epoch": 0.4370080907039463,
+      "grad_norm": 0.7187026143074036,
+      "learning_rate": 8.88888311701958e-06,
+      "loss": 0.792,
+      "step": 7940
+    },
+    {
+      "epoch": 0.43706312950630194,
+      "grad_norm": 0.8313350081443787,
+      "learning_rate": 8.888610650956322e-06,
+      "loss": 0.706,
+      "step": 7941
+    },
+    {
+      "epoch": 0.43711816830865763,
+      "grad_norm": 0.8083454370498657,
+      "learning_rate": 8.888338155667307e-06,
+      "loss": 0.7857,
+      "step": 7942
+    },
+    {
+      "epoch": 0.43717320711101326,
+      "grad_norm": 0.8200840353965759,
+      "learning_rate": 8.888065631154583e-06,
+      "loss": 0.8601,
+      "step": 7943
+    },
+    {
+      "epoch": 0.43722824591336895,
+      "grad_norm": 0.7503816485404968,
+      "learning_rate": 8.887793077420198e-06,
+      "loss": 0.7744,
+      "step": 7944
+    },
+    {
+      "epoch": 0.4372832847157246,
+      "grad_norm": 0.7466493248939514,
+      "learning_rate": 8.887520494466202e-06,
+      "loss": 0.7818,
+      "step": 7945
+    },
+    {
+      "epoch": 0.4373383235180803,
+      "grad_norm": 0.728118360042572,
+      "learning_rate": 8.887247882294641e-06,
+      "loss": 0.7157,
+      "step": 7946
+    },
+    {
+      "epoch": 0.4373933623204359,
+      "grad_norm": 0.9199670553207397,
+      "learning_rate": 8.886975240907568e-06,
+      "loss": 0.8283,
+      "step": 7947
+    },
+    {
+      "epoch": 0.43744840112279154,
+      "grad_norm": 0.735584557056427,
+      "learning_rate": 8.886702570307027e-06,
+      "loss": 0.6588,
+      "step": 7948
+    },
+    {
+      "epoch": 0.43750343992514723,
+      "grad_norm": 0.8619036674499512,
+      "learning_rate": 8.886429870495072e-06,
+      "loss": 0.7269,
+      "step": 7949
+    },
+    {
+      "epoch": 0.43755847872750286,
+      "grad_norm": 0.7304830551147461,
+      "learning_rate": 8.886157141473747e-06,
+      "loss": 0.6725,
+      "step": 7950
+    },
+    {
+      "epoch": 0.43761351752985855,
+      "grad_norm": 0.7669086456298828,
+      "learning_rate": 8.885884383245109e-06,
+      "loss": 0.6957,
+      "step": 7951
+    },
+    {
+      "epoch": 0.4376685563322142,
+      "grad_norm": 0.7558299899101257,
+      "learning_rate": 8.885611595811203e-06,
+      "loss": 0.8159,
+      "step": 7952
+    },
+    {
+      "epoch": 0.4377235951345699,
+      "grad_norm": 0.7661786079406738,
+      "learning_rate": 8.88533877917408e-06,
+      "loss": 0.764,
+      "step": 7953
+    },
+    {
+      "epoch": 0.4377786339369255,
+      "grad_norm": 0.7461101412773132,
+      "learning_rate": 8.88506593333579e-06,
+      "loss": 0.7544,
+      "step": 7954
+    },
+    {
+      "epoch": 0.4378336727392812,
+      "grad_norm": 0.7989180088043213,
+      "learning_rate": 8.884793058298387e-06,
+      "loss": 0.6913,
+      "step": 7955
+    },
+    {
+      "epoch": 0.43788871154163683,
+      "grad_norm": 0.7964022755622864,
+      "learning_rate": 8.884520154063917e-06,
+      "loss": 0.7339,
+      "step": 7956
+    },
+    {
+      "epoch": 0.4379437503439925,
+      "grad_norm": 0.7278034687042236,
+      "learning_rate": 8.884247220634433e-06,
+      "loss": 0.8477,
+      "step": 7957
+    },
+    {
+      "epoch": 0.43799878914634816,
+      "grad_norm": 0.7294753789901733,
+      "learning_rate": 8.883974258011988e-06,
+      "loss": 0.8412,
+      "step": 7958
+    },
+    {
+      "epoch": 0.43805382794870384,
+      "grad_norm": 0.665734589099884,
+      "learning_rate": 8.88370126619863e-06,
+      "loss": 0.7838,
+      "step": 7959
+    },
+    {
+      "epoch": 0.4381088667510595,
+      "grad_norm": 0.6984216570854187,
+      "learning_rate": 8.883428245196414e-06,
+      "loss": 0.7657,
+      "step": 7960
+    },
+    {
+      "epoch": 0.43816390555341517,
+      "grad_norm": 0.8048402070999146,
+      "learning_rate": 8.883155195007393e-06,
+      "loss": 0.7553,
+      "step": 7961
+    },
+    {
+      "epoch": 0.4382189443557708,
+      "grad_norm": 0.7145794630050659,
+      "learning_rate": 8.882882115633616e-06,
+      "loss": 0.6583,
+      "step": 7962
+    },
+    {
+      "epoch": 0.4382739831581265,
+      "grad_norm": 0.7073546648025513,
+      "learning_rate": 8.882609007077135e-06,
+      "loss": 0.7869,
+      "step": 7963
+    },
+    {
+      "epoch": 0.4383290219604821,
+      "grad_norm": 0.8300859928131104,
+      "learning_rate": 8.882335869340004e-06,
+      "loss": 0.773,
+      "step": 7964
+    },
+    {
+      "epoch": 0.4383840607628378,
+      "grad_norm": 0.8343188762664795,
+      "learning_rate": 8.882062702424276e-06,
+      "loss": 0.6743,
+      "step": 7965
+    },
+    {
+      "epoch": 0.43843909956519345,
+      "grad_norm": 0.7106530666351318,
+      "learning_rate": 8.881789506332007e-06,
+      "loss": 0.7414,
+      "step": 7966
+    },
+    {
+      "epoch": 0.43849413836754914,
+      "grad_norm": 0.7015630602836609,
+      "learning_rate": 8.881516281065244e-06,
+      "loss": 0.7434,
+      "step": 7967
+    },
+    {
+      "epoch": 0.43854917716990477,
+      "grad_norm": 0.8106673955917358,
+      "learning_rate": 8.881243026626044e-06,
+      "loss": 0.7741,
+      "step": 7968
+    },
+    {
+      "epoch": 0.43860421597226046,
+      "grad_norm": 0.8181495070457458,
+      "learning_rate": 8.88096974301646e-06,
+      "loss": 0.8046,
+      "step": 7969
+    },
+    {
+      "epoch": 0.4386592547746161,
+      "grad_norm": 0.7767857313156128,
+      "learning_rate": 8.880696430238546e-06,
+      "loss": 0.8586,
+      "step": 7970
+    },
+    {
+      "epoch": 0.4387142935769718,
+      "grad_norm": 0.7257522940635681,
+      "learning_rate": 8.880423088294359e-06,
+      "loss": 0.7799,
+      "step": 7971
+    },
+    {
+      "epoch": 0.4387693323793274,
+      "grad_norm": 0.6896021366119385,
+      "learning_rate": 8.880149717185948e-06,
+      "loss": 0.8178,
+      "step": 7972
+    },
+    {
+      "epoch": 0.4388243711816831,
+      "grad_norm": 0.7646406292915344,
+      "learning_rate": 8.879876316915372e-06,
+      "loss": 0.8754,
+      "step": 7973
+    },
+    {
+      "epoch": 0.43887940998403874,
+      "grad_norm": 0.8043848872184753,
+      "learning_rate": 8.879602887484684e-06,
+      "loss": 0.8562,
+      "step": 7974
+    },
+    {
+      "epoch": 0.4389344487863944,
+      "grad_norm": 0.6727305054664612,
+      "learning_rate": 8.879329428895937e-06,
+      "loss": 0.6168,
+      "step": 7975
+    },
+    {
+      "epoch": 0.43898948758875006,
+      "grad_norm": 0.7634731531143188,
+      "learning_rate": 8.87905594115119e-06,
+      "loss": 0.857,
+      "step": 7976
+    },
+    {
+      "epoch": 0.43904452639110575,
+      "grad_norm": 0.6544492244720459,
+      "learning_rate": 8.878782424252497e-06,
+      "loss": 0.6302,
+      "step": 7977
+    },
+    {
+      "epoch": 0.4390995651934614,
+      "grad_norm": 0.8126636743545532,
+      "learning_rate": 8.878508878201915e-06,
+      "loss": 0.7823,
+      "step": 7978
+    },
+    {
+      "epoch": 0.43915460399581707,
+      "grad_norm": 0.7235779166221619,
+      "learning_rate": 8.878235303001497e-06,
+      "loss": 0.7527,
+      "step": 7979
+    },
+    {
+      "epoch": 0.4392096427981727,
+      "grad_norm": 0.6961055397987366,
+      "learning_rate": 8.8779616986533e-06,
+      "loss": 0.7383,
+      "step": 7980
+    },
+    {
+      "epoch": 0.4392646816005284,
+      "grad_norm": 0.7684490084648132,
+      "learning_rate": 8.877688065159382e-06,
+      "loss": 0.8009,
+      "step": 7981
+    },
+    {
+      "epoch": 0.43931972040288403,
+      "grad_norm": 0.7897803783416748,
+      "learning_rate": 8.877414402521797e-06,
+      "loss": 0.7561,
+      "step": 7982
+    },
+    {
+      "epoch": 0.4393747592052397,
+      "grad_norm": 0.7877688407897949,
+      "learning_rate": 8.877140710742606e-06,
+      "loss": 0.7949,
+      "step": 7983
+    },
+    {
+      "epoch": 0.43942979800759535,
+      "grad_norm": 0.8341611623764038,
+      "learning_rate": 8.876866989823862e-06,
+      "loss": 0.7585,
+      "step": 7984
+    },
+    {
+      "epoch": 0.43948483680995104,
+      "grad_norm": 0.7663636207580566,
+      "learning_rate": 8.876593239767622e-06,
+      "loss": 0.771,
+      "step": 7985
+    },
+    {
+      "epoch": 0.4395398756123067,
+      "grad_norm": 0.6824129223823547,
+      "learning_rate": 8.876319460575946e-06,
+      "loss": 0.7852,
+      "step": 7986
+    },
+    {
+      "epoch": 0.43959491441466236,
+      "grad_norm": 0.6533854007720947,
+      "learning_rate": 8.876045652250891e-06,
+      "loss": 0.723,
+      "step": 7987
+    },
+    {
+      "epoch": 0.439649953217018,
+      "grad_norm": 0.7174259424209595,
+      "learning_rate": 8.875771814794515e-06,
+      "loss": 0.749,
+      "step": 7988
+    },
+    {
+      "epoch": 0.4397049920193737,
+      "grad_norm": 0.8585928678512573,
+      "learning_rate": 8.875497948208875e-06,
+      "loss": 0.6727,
+      "step": 7989
+    },
+    {
+      "epoch": 0.4397600308217293,
+      "grad_norm": 0.7558062672615051,
+      "learning_rate": 8.875224052496029e-06,
+      "loss": 0.7929,
+      "step": 7990
+    },
+    {
+      "epoch": 0.43981506962408495,
+      "grad_norm": 0.7063853144645691,
+      "learning_rate": 8.874950127658037e-06,
+      "loss": 0.7397,
+      "step": 7991
+    },
+    {
+      "epoch": 0.43987010842644064,
+      "grad_norm": 0.7165526747703552,
+      "learning_rate": 8.874676173696956e-06,
+      "loss": 0.7678,
+      "step": 7992
+    },
+    {
+      "epoch": 0.4399251472287963,
+      "grad_norm": 0.7657830715179443,
+      "learning_rate": 8.874402190614847e-06,
+      "loss": 0.8318,
+      "step": 7993
+    },
+    {
+      "epoch": 0.43998018603115197,
+      "grad_norm": 0.7776834964752197,
+      "learning_rate": 8.874128178413769e-06,
+      "loss": 0.8589,
+      "step": 7994
+    },
+    {
+      "epoch": 0.4400352248335076,
+      "grad_norm": 0.6805633306503296,
+      "learning_rate": 8.873854137095778e-06,
+      "loss": 0.7009,
+      "step": 7995
+    },
+    {
+      "epoch": 0.4400902636358633,
+      "grad_norm": 0.6962490677833557,
+      "learning_rate": 8.87358006666294e-06,
+      "loss": 0.7896,
+      "step": 7996
+    },
+    {
+      "epoch": 0.4401453024382189,
+      "grad_norm": 0.611610472202301,
+      "learning_rate": 8.873305967117307e-06,
+      "loss": 0.5993,
+      "step": 7997
+    },
+    {
+      "epoch": 0.4402003412405746,
+      "grad_norm": 0.7442964911460876,
+      "learning_rate": 8.873031838460946e-06,
+      "loss": 0.8277,
+      "step": 7998
+    },
+    {
+      "epoch": 0.44025538004293024,
+      "grad_norm": 0.6858734488487244,
+      "learning_rate": 8.872757680695914e-06,
+      "loss": 0.8064,
+      "step": 7999
+    },
+    {
+      "epoch": 0.44031041884528593,
+      "grad_norm": 0.6654849052429199,
+      "learning_rate": 8.872483493824273e-06,
+      "loss": 0.7408,
+      "step": 8000
+    },
+    {
+      "epoch": 0.44036545764764157,
+      "grad_norm": 0.8241575956344604,
+      "learning_rate": 8.87220927784808e-06,
+      "loss": 0.8819,
+      "step": 8001
+    },
+    {
+      "epoch": 0.44042049644999726,
+      "grad_norm": 0.7078573107719421,
+      "learning_rate": 8.8719350327694e-06,
+      "loss": 0.7709,
+      "step": 8002
+    },
+    {
+      "epoch": 0.4404755352523529,
+      "grad_norm": 0.7369210720062256,
+      "learning_rate": 8.871660758590292e-06,
+      "loss": 0.7867,
+      "step": 8003
+    },
+    {
+      "epoch": 0.4405305740547086,
+      "grad_norm": 0.7206673622131348,
+      "learning_rate": 8.87138645531282e-06,
+      "loss": 0.8697,
+      "step": 8004
+    },
+    {
+      "epoch": 0.4405856128570642,
+      "grad_norm": 0.8370183706283569,
+      "learning_rate": 8.871112122939041e-06,
+      "loss": 0.7201,
+      "step": 8005
+    },
+    {
+      "epoch": 0.4406406516594199,
+      "grad_norm": 0.8015196323394775,
+      "learning_rate": 8.870837761471023e-06,
+      "loss": 0.774,
+      "step": 8006
+    },
+    {
+      "epoch": 0.44069569046177554,
+      "grad_norm": 0.730185329914093,
+      "learning_rate": 8.870563370910821e-06,
+      "loss": 0.7153,
+      "step": 8007
+    },
+    {
+      "epoch": 0.4407507292641312,
+      "grad_norm": 0.6719930768013,
+      "learning_rate": 8.870288951260503e-06,
+      "loss": 0.7949,
+      "step": 8008
+    },
+    {
+      "epoch": 0.44080576806648686,
+      "grad_norm": 0.7614291906356812,
+      "learning_rate": 8.870014502522128e-06,
+      "loss": 0.7143,
+      "step": 8009
+    },
+    {
+      "epoch": 0.44086080686884255,
+      "grad_norm": 0.7438056468963623,
+      "learning_rate": 8.86974002469776e-06,
+      "loss": 0.6859,
+      "step": 8010
+    },
+    {
+      "epoch": 0.4409158456711982,
+      "grad_norm": 0.759903073310852,
+      "learning_rate": 8.869465517789463e-06,
+      "loss": 0.8095,
+      "step": 8011
+    },
+    {
+      "epoch": 0.44097088447355387,
+      "grad_norm": 0.7622823119163513,
+      "learning_rate": 8.869190981799298e-06,
+      "loss": 0.786,
+      "step": 8012
+    },
+    {
+      "epoch": 0.4410259232759095,
+      "grad_norm": 0.677003800868988,
+      "learning_rate": 8.86891641672933e-06,
+      "loss": 0.7074,
+      "step": 8013
+    },
+    {
+      "epoch": 0.4410809620782652,
+      "grad_norm": 0.9258451461791992,
+      "learning_rate": 8.86864182258162e-06,
+      "loss": 0.7218,
+      "step": 8014
+    },
+    {
+      "epoch": 0.4411360008806208,
+      "grad_norm": 0.7027828693389893,
+      "learning_rate": 8.868367199358236e-06,
+      "loss": 0.7654,
+      "step": 8015
+    },
+    {
+      "epoch": 0.4411910396829765,
+      "grad_norm": 0.8279967308044434,
+      "learning_rate": 8.868092547061239e-06,
+      "loss": 0.8969,
+      "step": 8016
+    },
+    {
+      "epoch": 0.44124607848533215,
+      "grad_norm": 0.7366079688072205,
+      "learning_rate": 8.867817865692693e-06,
+      "loss": 0.8421,
+      "step": 8017
+    },
+    {
+      "epoch": 0.44130111728768784,
+      "grad_norm": 0.7548787593841553,
+      "learning_rate": 8.867543155254665e-06,
+      "loss": 0.79,
+      "step": 8018
+    },
+    {
+      "epoch": 0.44135615609004347,
+      "grad_norm": 0.7558487057685852,
+      "learning_rate": 8.867268415749215e-06,
+      "loss": 0.8461,
+      "step": 8019
+    },
+    {
+      "epoch": 0.44141119489239916,
+      "grad_norm": 0.6413403153419495,
+      "learning_rate": 8.866993647178413e-06,
+      "loss": 0.6811,
+      "step": 8020
+    },
+    {
+      "epoch": 0.4414662336947548,
+      "grad_norm": 0.9251089692115784,
+      "learning_rate": 8.86671884954432e-06,
+      "loss": 0.868,
+      "step": 8021
+    },
+    {
+      "epoch": 0.4415212724971105,
+      "grad_norm": 0.7920099496841431,
+      "learning_rate": 8.866444022849006e-06,
+      "loss": 0.8131,
+      "step": 8022
+    },
+    {
+      "epoch": 0.4415763112994661,
+      "grad_norm": 0.8738380670547485,
+      "learning_rate": 8.866169167094532e-06,
+      "loss": 0.857,
+      "step": 8023
+    },
+    {
+      "epoch": 0.4416313501018218,
+      "grad_norm": 0.7181336283683777,
+      "learning_rate": 8.865894282282965e-06,
+      "loss": 0.7869,
+      "step": 8024
+    },
+    {
+      "epoch": 0.44168638890417744,
+      "grad_norm": 0.8003776669502258,
+      "learning_rate": 8.865619368416373e-06,
+      "loss": 0.8874,
+      "step": 8025
+    },
+    {
+      "epoch": 0.44174142770653313,
+      "grad_norm": 0.7186623215675354,
+      "learning_rate": 8.86534442549682e-06,
+      "loss": 0.7931,
+      "step": 8026
+    },
+    {
+      "epoch": 0.44179646650888876,
+      "grad_norm": 0.7006831765174866,
+      "learning_rate": 8.865069453526371e-06,
+      "loss": 0.7046,
+      "step": 8027
+    },
+    {
+      "epoch": 0.44185150531124445,
+      "grad_norm": 0.7394786477088928,
+      "learning_rate": 8.864794452507097e-06,
+      "loss": 0.685,
+      "step": 8028
+    },
+    {
+      "epoch": 0.4419065441136001,
+      "grad_norm": 0.7512097358703613,
+      "learning_rate": 8.864519422441062e-06,
+      "loss": 0.8047,
+      "step": 8029
+    },
+    {
+      "epoch": 0.4419615829159558,
+      "grad_norm": 0.6866902709007263,
+      "learning_rate": 8.864244363330333e-06,
+      "loss": 0.7099,
+      "step": 8030
+    },
+    {
+      "epoch": 0.4420166217183114,
+      "grad_norm": 0.7316723465919495,
+      "learning_rate": 8.863969275176978e-06,
+      "loss": 0.7767,
+      "step": 8031
+    },
+    {
+      "epoch": 0.4420716605206671,
+      "grad_norm": 0.7103593349456787,
+      "learning_rate": 8.863694157983064e-06,
+      "loss": 0.7832,
+      "step": 8032
+    },
+    {
+      "epoch": 0.44212669932302273,
+      "grad_norm": 0.6922749876976013,
+      "learning_rate": 8.863419011750659e-06,
+      "loss": 0.7833,
+      "step": 8033
+    },
+    {
+      "epoch": 0.44218173812537837,
+      "grad_norm": 0.7989425659179688,
+      "learning_rate": 8.863143836481831e-06,
+      "loss": 0.8651,
+      "step": 8034
+    },
+    {
+      "epoch": 0.44223677692773405,
+      "grad_norm": 0.6765440702438354,
+      "learning_rate": 8.862868632178648e-06,
+      "loss": 0.7858,
+      "step": 8035
+    },
+    {
+      "epoch": 0.4422918157300897,
+      "grad_norm": 0.670767068862915,
+      "learning_rate": 8.862593398843178e-06,
+      "loss": 0.6789,
+      "step": 8036
+    },
+    {
+      "epoch": 0.4423468545324454,
+      "grad_norm": 0.7556853294372559,
+      "learning_rate": 8.86231813647749e-06,
+      "loss": 0.8036,
+      "step": 8037
+    },
+    {
+      "epoch": 0.442401893334801,
+      "grad_norm": 0.788690984249115,
+      "learning_rate": 8.862042845083654e-06,
+      "loss": 0.8355,
+      "step": 8038
+    },
+    {
+      "epoch": 0.4424569321371567,
+      "grad_norm": 0.8439056873321533,
+      "learning_rate": 8.861767524663736e-06,
+      "loss": 0.7327,
+      "step": 8039
+    },
+    {
+      "epoch": 0.44251197093951233,
+      "grad_norm": 0.7101821899414062,
+      "learning_rate": 8.861492175219808e-06,
+      "loss": 0.8303,
+      "step": 8040
+    },
+    {
+      "epoch": 0.442567009741868,
+      "grad_norm": 0.741680383682251,
+      "learning_rate": 8.861216796753937e-06,
+      "loss": 0.7377,
+      "step": 8041
+    },
+    {
+      "epoch": 0.44262204854422366,
+      "grad_norm": 0.7588099837303162,
+      "learning_rate": 8.860941389268196e-06,
+      "loss": 0.8217,
+      "step": 8042
+    },
+    {
+      "epoch": 0.44267708734657935,
+      "grad_norm": 0.7654829025268555,
+      "learning_rate": 8.860665952764654e-06,
+      "loss": 0.8416,
+      "step": 8043
+    },
+    {
+      "epoch": 0.442732126148935,
+      "grad_norm": 0.7025987505912781,
+      "learning_rate": 8.860390487245378e-06,
+      "loss": 0.7312,
+      "step": 8044
+    },
+    {
+      "epoch": 0.44278716495129067,
+      "grad_norm": 0.7206251621246338,
+      "learning_rate": 8.860114992712441e-06,
+      "loss": 0.7522,
+      "step": 8045
+    },
+    {
+      "epoch": 0.4428422037536463,
+      "grad_norm": 0.7041749954223633,
+      "learning_rate": 8.859839469167912e-06,
+      "loss": 0.746,
+      "step": 8046
+    },
+    {
+      "epoch": 0.442897242556002,
+      "grad_norm": 0.6941862106323242,
+      "learning_rate": 8.859563916613864e-06,
+      "loss": 0.7692,
+      "step": 8047
+    },
+    {
+      "epoch": 0.4429522813583576,
+      "grad_norm": 0.6897740364074707,
+      "learning_rate": 8.859288335052367e-06,
+      "loss": 0.7963,
+      "step": 8048
+    },
+    {
+      "epoch": 0.4430073201607133,
+      "grad_norm": 0.6744545698165894,
+      "learning_rate": 8.859012724485492e-06,
+      "loss": 0.7647,
+      "step": 8049
+    },
+    {
+      "epoch": 0.44306235896306895,
+      "grad_norm": 0.7899364829063416,
+      "learning_rate": 8.858737084915309e-06,
+      "loss": 0.8373,
+      "step": 8050
+    },
+    {
+      "epoch": 0.44311739776542464,
+      "grad_norm": 0.806016743183136,
+      "learning_rate": 8.85846141634389e-06,
+      "loss": 0.7871,
+      "step": 8051
+    },
+    {
+      "epoch": 0.44317243656778027,
+      "grad_norm": 0.7444993257522583,
+      "learning_rate": 8.85818571877331e-06,
+      "loss": 0.8099,
+      "step": 8052
+    },
+    {
+      "epoch": 0.44322747537013596,
+      "grad_norm": 0.772735059261322,
+      "learning_rate": 8.85790999220564e-06,
+      "loss": 0.7113,
+      "step": 8053
+    },
+    {
+      "epoch": 0.4432825141724916,
+      "grad_norm": 0.7743984460830688,
+      "learning_rate": 8.85763423664295e-06,
+      "loss": 0.8935,
+      "step": 8054
+    },
+    {
+      "epoch": 0.4433375529748473,
+      "grad_norm": 0.6751214265823364,
+      "learning_rate": 8.857358452087313e-06,
+      "loss": 0.6769,
+      "step": 8055
+    },
+    {
+      "epoch": 0.4433925917772029,
+      "grad_norm": 0.6921005845069885,
+      "learning_rate": 8.857082638540803e-06,
+      "loss": 0.7071,
+      "step": 8056
+    },
+    {
+      "epoch": 0.4434476305795586,
+      "grad_norm": 0.7884092330932617,
+      "learning_rate": 8.856806796005491e-06,
+      "loss": 0.7919,
+      "step": 8057
+    },
+    {
+      "epoch": 0.44350266938191424,
+      "grad_norm": 0.6522679924964905,
+      "learning_rate": 8.856530924483452e-06,
+      "loss": 0.7449,
+      "step": 8058
+    },
+    {
+      "epoch": 0.4435577081842699,
+      "grad_norm": 0.7172590494155884,
+      "learning_rate": 8.85625502397676e-06,
+      "loss": 0.7306,
+      "step": 8059
+    },
+    {
+      "epoch": 0.44361274698662556,
+      "grad_norm": 0.698658287525177,
+      "learning_rate": 8.855979094487488e-06,
+      "loss": 0.803,
+      "step": 8060
+    },
+    {
+      "epoch": 0.44366778578898125,
+      "grad_norm": 0.685589075088501,
+      "learning_rate": 8.855703136017708e-06,
+      "loss": 0.763,
+      "step": 8061
+    },
+    {
+      "epoch": 0.4437228245913369,
+      "grad_norm": 0.8259774446487427,
+      "learning_rate": 8.855427148569495e-06,
+      "loss": 0.811,
+      "step": 8062
+    },
+    {
+      "epoch": 0.4437778633936926,
+      "grad_norm": 0.6976660490036011,
+      "learning_rate": 8.855151132144926e-06,
+      "loss": 0.7345,
+      "step": 8063
+    },
+    {
+      "epoch": 0.4438329021960482,
+      "grad_norm": 0.7696738243103027,
+      "learning_rate": 8.854875086746071e-06,
+      "loss": 0.823,
+      "step": 8064
+    },
+    {
+      "epoch": 0.4438879409984039,
+      "grad_norm": 0.6627930998802185,
+      "learning_rate": 8.854599012375006e-06,
+      "loss": 0.7455,
+      "step": 8065
+    },
+    {
+      "epoch": 0.44394297980075953,
+      "grad_norm": 0.7492700815200806,
+      "learning_rate": 8.854322909033809e-06,
+      "loss": 0.8195,
+      "step": 8066
+    },
+    {
+      "epoch": 0.4439980186031152,
+      "grad_norm": 0.8335888981819153,
+      "learning_rate": 8.85404677672455e-06,
+      "loss": 0.7683,
+      "step": 8067
+    },
+    {
+      "epoch": 0.44405305740547085,
+      "grad_norm": 0.7448242902755737,
+      "learning_rate": 8.853770615449309e-06,
+      "loss": 0.8352,
+      "step": 8068
+    },
+    {
+      "epoch": 0.44410809620782654,
+      "grad_norm": 0.700616180896759,
+      "learning_rate": 8.853494425210158e-06,
+      "loss": 0.7892,
+      "step": 8069
+    },
+    {
+      "epoch": 0.4441631350101822,
+      "grad_norm": 0.6959284543991089,
+      "learning_rate": 8.853218206009176e-06,
+      "loss": 0.6944,
+      "step": 8070
+    },
+    {
+      "epoch": 0.44421817381253786,
+      "grad_norm": 0.7507375478744507,
+      "learning_rate": 8.852941957848438e-06,
+      "loss": 0.8921,
+      "step": 8071
+    },
+    {
+      "epoch": 0.4442732126148935,
+      "grad_norm": 0.7843918204307556,
+      "learning_rate": 8.852665680730019e-06,
+      "loss": 0.816,
+      "step": 8072
+    },
+    {
+      "epoch": 0.4443282514172492,
+      "grad_norm": 0.8702702522277832,
+      "learning_rate": 8.852389374655995e-06,
+      "loss": 0.8191,
+      "step": 8073
+    },
+    {
+      "epoch": 0.4443832902196048,
+      "grad_norm": 0.6784317493438721,
+      "learning_rate": 8.852113039628445e-06,
+      "loss": 0.7726,
+      "step": 8074
+    },
+    {
+      "epoch": 0.4444383290219605,
+      "grad_norm": 0.724530041217804,
+      "learning_rate": 8.851836675649443e-06,
+      "loss": 0.8214,
+      "step": 8075
+    },
+    {
+      "epoch": 0.44449336782431614,
+      "grad_norm": 0.9814287424087524,
+      "learning_rate": 8.851560282721067e-06,
+      "loss": 0.8368,
+      "step": 8076
+    },
+    {
+      "epoch": 0.4445484066266718,
+      "grad_norm": 0.6606815457344055,
+      "learning_rate": 8.851283860845398e-06,
+      "loss": 0.7772,
+      "step": 8077
+    },
+    {
+      "epoch": 0.44460344542902747,
+      "grad_norm": 0.6910951137542725,
+      "learning_rate": 8.851007410024507e-06,
+      "loss": 0.7007,
+      "step": 8078
+    },
+    {
+      "epoch": 0.4446584842313831,
+      "grad_norm": 0.6764300465583801,
+      "learning_rate": 8.850730930260479e-06,
+      "loss": 0.7265,
+      "step": 8079
+    },
+    {
+      "epoch": 0.4447135230337388,
+      "grad_norm": 0.669622004032135,
+      "learning_rate": 8.850454421555386e-06,
+      "loss": 0.7551,
+      "step": 8080
+    },
+    {
+      "epoch": 0.4447685618360944,
+      "grad_norm": 0.7068240642547607,
+      "learning_rate": 8.850177883911307e-06,
+      "loss": 0.8358,
+      "step": 8081
+    },
+    {
+      "epoch": 0.4448236006384501,
+      "grad_norm": 0.7100360989570618,
+      "learning_rate": 8.849901317330324e-06,
+      "loss": 0.7074,
+      "step": 8082
+    },
+    {
+      "epoch": 0.44487863944080575,
+      "grad_norm": 0.7510328888893127,
+      "learning_rate": 8.849624721814511e-06,
+      "loss": 0.6654,
+      "step": 8083
+    },
+    {
+      "epoch": 0.44493367824316143,
+      "grad_norm": 0.8106432557106018,
+      "learning_rate": 8.849348097365951e-06,
+      "loss": 0.6944,
+      "step": 8084
+    },
+    {
+      "epoch": 0.44498871704551707,
+      "grad_norm": 0.6852346062660217,
+      "learning_rate": 8.84907144398672e-06,
+      "loss": 0.7203,
+      "step": 8085
+    },
+    {
+      "epoch": 0.44504375584787276,
+      "grad_norm": 0.8495593667030334,
+      "learning_rate": 8.848794761678898e-06,
+      "loss": 0.7918,
+      "step": 8086
+    },
+    {
+      "epoch": 0.4450987946502284,
+      "grad_norm": 0.7110981941223145,
+      "learning_rate": 8.848518050444565e-06,
+      "loss": 0.8176,
+      "step": 8087
+    },
+    {
+      "epoch": 0.4451538334525841,
+      "grad_norm": 0.7740922570228577,
+      "learning_rate": 8.8482413102858e-06,
+      "loss": 0.7573,
+      "step": 8088
+    },
+    {
+      "epoch": 0.4452088722549397,
+      "grad_norm": 0.9645134806632996,
+      "learning_rate": 8.847964541204685e-06,
+      "loss": 0.7842,
+      "step": 8089
+    },
+    {
+      "epoch": 0.4452639110572954,
+      "grad_norm": 0.767621636390686,
+      "learning_rate": 8.847687743203299e-06,
+      "loss": 0.8182,
+      "step": 8090
+    },
+    {
+      "epoch": 0.44531894985965104,
+      "grad_norm": 0.6842975616455078,
+      "learning_rate": 8.84741091628372e-06,
+      "loss": 0.7795,
+      "step": 8091
+    },
+    {
+      "epoch": 0.4453739886620067,
+      "grad_norm": 0.768644392490387,
+      "learning_rate": 8.847134060448032e-06,
+      "loss": 0.7363,
+      "step": 8092
+    },
+    {
+      "epoch": 0.44542902746436236,
+      "grad_norm": 0.6813824772834778,
+      "learning_rate": 8.846857175698314e-06,
+      "loss": 0.7601,
+      "step": 8093
+    },
+    {
+      "epoch": 0.44548406626671805,
+      "grad_norm": 0.8608306646347046,
+      "learning_rate": 8.846580262036645e-06,
+      "loss": 0.8205,
+      "step": 8094
+    },
+    {
+      "epoch": 0.4455391050690737,
+      "grad_norm": 0.6917694807052612,
+      "learning_rate": 8.84630331946511e-06,
+      "loss": 0.7207,
+      "step": 8095
+    },
+    {
+      "epoch": 0.44559414387142937,
+      "grad_norm": 0.6777203679084778,
+      "learning_rate": 8.84602634798579e-06,
+      "loss": 0.6939,
+      "step": 8096
+    },
+    {
+      "epoch": 0.445649182673785,
+      "grad_norm": 0.7249894142150879,
+      "learning_rate": 8.845749347600764e-06,
+      "loss": 0.7918,
+      "step": 8097
+    },
+    {
+      "epoch": 0.4457042214761407,
+      "grad_norm": 0.7446995973587036,
+      "learning_rate": 8.845472318312116e-06,
+      "loss": 0.7379,
+      "step": 8098
+    },
+    {
+      "epoch": 0.4457592602784963,
+      "grad_norm": 0.8245479464530945,
+      "learning_rate": 8.845195260121927e-06,
+      "loss": 0.8532,
+      "step": 8099
+    },
+    {
+      "epoch": 0.445814299080852,
+      "grad_norm": 0.7160329818725586,
+      "learning_rate": 8.84491817303228e-06,
+      "loss": 0.7042,
+      "step": 8100
+    },
+    {
+      "epoch": 0.44586933788320765,
+      "grad_norm": 0.8056026101112366,
+      "learning_rate": 8.844641057045257e-06,
+      "loss": 0.8581,
+      "step": 8101
+    },
+    {
+      "epoch": 0.44592437668556334,
+      "grad_norm": 0.7257886528968811,
+      "learning_rate": 8.84436391216294e-06,
+      "loss": 0.7297,
+      "step": 8102
+    },
+    {
+      "epoch": 0.445979415487919,
+      "grad_norm": 0.7400404810905457,
+      "learning_rate": 8.844086738387415e-06,
+      "loss": 0.7703,
+      "step": 8103
+    },
+    {
+      "epoch": 0.44603445429027466,
+      "grad_norm": 0.665271520614624,
+      "learning_rate": 8.843809535720763e-06,
+      "loss": 0.7769,
+      "step": 8104
+    },
+    {
+      "epoch": 0.4460894930926303,
+      "grad_norm": 0.7041043639183044,
+      "learning_rate": 8.843532304165066e-06,
+      "loss": 0.7995,
+      "step": 8105
+    },
+    {
+      "epoch": 0.446144531894986,
+      "grad_norm": 0.8517841100692749,
+      "learning_rate": 8.84325504372241e-06,
+      "loss": 0.8239,
+      "step": 8106
+    },
+    {
+      "epoch": 0.4461995706973416,
+      "grad_norm": 0.7045741677284241,
+      "learning_rate": 8.842977754394877e-06,
+      "loss": 0.7982,
+      "step": 8107
+    },
+    {
+      "epoch": 0.4462546094996973,
+      "grad_norm": 0.7056185007095337,
+      "learning_rate": 8.842700436184552e-06,
+      "loss": 0.8003,
+      "step": 8108
+    },
+    {
+      "epoch": 0.44630964830205294,
+      "grad_norm": 0.9042232632637024,
+      "learning_rate": 8.842423089093519e-06,
+      "loss": 0.7534,
+      "step": 8109
+    },
+    {
+      "epoch": 0.44636468710440863,
+      "grad_norm": 0.8584854602813721,
+      "learning_rate": 8.842145713123863e-06,
+      "loss": 0.7759,
+      "step": 8110
+    },
+    {
+      "epoch": 0.44641972590676426,
+      "grad_norm": 0.7333530187606812,
+      "learning_rate": 8.841868308277668e-06,
+      "loss": 0.7218,
+      "step": 8111
+    },
+    {
+      "epoch": 0.44647476470911995,
+      "grad_norm": 0.7866941094398499,
+      "learning_rate": 8.84159087455702e-06,
+      "loss": 0.7016,
+      "step": 8112
+    },
+    {
+      "epoch": 0.4465298035114756,
+      "grad_norm": 0.7785252928733826,
+      "learning_rate": 8.841313411964001e-06,
+      "loss": 0.8232,
+      "step": 8113
+    },
+    {
+      "epoch": 0.4465848423138313,
+      "grad_norm": 0.7060698866844177,
+      "learning_rate": 8.841035920500702e-06,
+      "loss": 0.6987,
+      "step": 8114
+    },
+    {
+      "epoch": 0.4466398811161869,
+      "grad_norm": 0.7211717963218689,
+      "learning_rate": 8.840758400169203e-06,
+      "loss": 0.8604,
+      "step": 8115
+    },
+    {
+      "epoch": 0.4466949199185426,
+      "grad_norm": 0.979678213596344,
+      "learning_rate": 8.840480850971593e-06,
+      "loss": 0.9028,
+      "step": 8116
+    },
+    {
+      "epoch": 0.44674995872089823,
+      "grad_norm": 0.6595104336738586,
+      "learning_rate": 8.840203272909957e-06,
+      "loss": 0.6899,
+      "step": 8117
+    },
+    {
+      "epoch": 0.4468049975232539,
+      "grad_norm": 0.6392405033111572,
+      "learning_rate": 8.83992566598638e-06,
+      "loss": 0.7729,
+      "step": 8118
+    },
+    {
+      "epoch": 0.44686003632560956,
+      "grad_norm": 1.1084040403366089,
+      "learning_rate": 8.839648030202949e-06,
+      "loss": 0.822,
+      "step": 8119
+    },
+    {
+      "epoch": 0.4469150751279652,
+      "grad_norm": 0.7024106383323669,
+      "learning_rate": 8.839370365561754e-06,
+      "loss": 0.7615,
+      "step": 8120
+    },
+    {
+      "epoch": 0.4469701139303209,
+      "grad_norm": 0.7204060554504395,
+      "learning_rate": 8.839092672064878e-06,
+      "loss": 0.7527,
+      "step": 8121
+    },
+    {
+      "epoch": 0.4470251527326765,
+      "grad_norm": 0.7307723760604858,
+      "learning_rate": 8.838814949714407e-06,
+      "loss": 0.8139,
+      "step": 8122
+    },
+    {
+      "epoch": 0.4470801915350322,
+      "grad_norm": 0.824034571647644,
+      "learning_rate": 8.838537198512434e-06,
+      "loss": 0.8299,
+      "step": 8123
+    },
+    {
+      "epoch": 0.44713523033738783,
+      "grad_norm": 0.6603747606277466,
+      "learning_rate": 8.83825941846104e-06,
+      "loss": 0.6762,
+      "step": 8124
+    },
+    {
+      "epoch": 0.4471902691397435,
+      "grad_norm": 0.7403088808059692,
+      "learning_rate": 8.837981609562316e-06,
+      "loss": 0.716,
+      "step": 8125
+    },
+    {
+      "epoch": 0.44724530794209916,
+      "grad_norm": 0.742173969745636,
+      "learning_rate": 8.837703771818351e-06,
+      "loss": 0.7672,
+      "step": 8126
+    },
+    {
+      "epoch": 0.44730034674445485,
+      "grad_norm": 0.7158839106559753,
+      "learning_rate": 8.837425905231232e-06,
+      "loss": 0.6941,
+      "step": 8127
+    },
+    {
+      "epoch": 0.4473553855468105,
+      "grad_norm": 0.7659464478492737,
+      "learning_rate": 8.837148009803044e-06,
+      "loss": 0.7293,
+      "step": 8128
+    },
+    {
+      "epoch": 0.44741042434916617,
+      "grad_norm": 0.8681113719940186,
+      "learning_rate": 8.836870085535882e-06,
+      "loss": 0.8647,
+      "step": 8129
+    },
+    {
+      "epoch": 0.4474654631515218,
+      "grad_norm": 0.7117272615432739,
+      "learning_rate": 8.83659213243183e-06,
+      "loss": 0.8035,
+      "step": 8130
+    },
+    {
+      "epoch": 0.4475205019538775,
+      "grad_norm": 0.8220957517623901,
+      "learning_rate": 8.836314150492978e-06,
+      "loss": 0.6978,
+      "step": 8131
+    },
+    {
+      "epoch": 0.4475755407562331,
+      "grad_norm": 0.7045003175735474,
+      "learning_rate": 8.836036139721418e-06,
+      "loss": 0.747,
+      "step": 8132
+    },
+    {
+      "epoch": 0.4476305795585888,
+      "grad_norm": 0.6833191514015198,
+      "learning_rate": 8.835758100119235e-06,
+      "loss": 0.7604,
+      "step": 8133
+    },
+    {
+      "epoch": 0.44768561836094445,
+      "grad_norm": 0.7305697798728943,
+      "learning_rate": 8.835480031688521e-06,
+      "loss": 0.7301,
+      "step": 8134
+    },
+    {
+      "epoch": 0.44774065716330014,
+      "grad_norm": 0.7266964912414551,
+      "learning_rate": 8.835201934431366e-06,
+      "loss": 0.7675,
+      "step": 8135
+    },
+    {
+      "epoch": 0.44779569596565577,
+      "grad_norm": 0.6822015047073364,
+      "learning_rate": 8.834923808349861e-06,
+      "loss": 0.8226,
+      "step": 8136
+    },
+    {
+      "epoch": 0.44785073476801146,
+      "grad_norm": 0.7443515062332153,
+      "learning_rate": 8.834645653446095e-06,
+      "loss": 0.9289,
+      "step": 8137
+    },
+    {
+      "epoch": 0.4479057735703671,
+      "grad_norm": 0.7337210178375244,
+      "learning_rate": 8.834367469722158e-06,
+      "loss": 0.7758,
+      "step": 8138
+    },
+    {
+      "epoch": 0.4479608123727228,
+      "grad_norm": 0.6794925332069397,
+      "learning_rate": 8.83408925718014e-06,
+      "loss": 0.8426,
+      "step": 8139
+    },
+    {
+      "epoch": 0.4480158511750784,
+      "grad_norm": 0.7808265089988708,
+      "learning_rate": 8.833811015822135e-06,
+      "loss": 0.8464,
+      "step": 8140
+    },
+    {
+      "epoch": 0.4480708899774341,
+      "grad_norm": 0.7837018370628357,
+      "learning_rate": 8.833532745650234e-06,
+      "loss": 0.8722,
+      "step": 8141
+    },
+    {
+      "epoch": 0.44812592877978974,
+      "grad_norm": 0.9218140840530396,
+      "learning_rate": 8.833254446666526e-06,
+      "loss": 0.7981,
+      "step": 8142
+    },
+    {
+      "epoch": 0.44818096758214543,
+      "grad_norm": 0.7980387806892395,
+      "learning_rate": 8.832976118873103e-06,
+      "loss": 0.7705,
+      "step": 8143
+    },
+    {
+      "epoch": 0.44823600638450106,
+      "grad_norm": 0.7354007363319397,
+      "learning_rate": 8.832697762272057e-06,
+      "loss": 0.8286,
+      "step": 8144
+    },
+    {
+      "epoch": 0.44829104518685675,
+      "grad_norm": 0.7006223201751709,
+      "learning_rate": 8.832419376865482e-06,
+      "loss": 0.7107,
+      "step": 8145
+    },
+    {
+      "epoch": 0.4483460839892124,
+      "grad_norm": 0.7838212847709656,
+      "learning_rate": 8.83214096265547e-06,
+      "loss": 0.7676,
+      "step": 8146
+    },
+    {
+      "epoch": 0.4484011227915681,
+      "grad_norm": 0.7768213748931885,
+      "learning_rate": 8.83186251964411e-06,
+      "loss": 0.8689,
+      "step": 8147
+    },
+    {
+      "epoch": 0.4484561615939237,
+      "grad_norm": 0.7451630234718323,
+      "learning_rate": 8.831584047833497e-06,
+      "loss": 0.8625,
+      "step": 8148
+    },
+    {
+      "epoch": 0.4485112003962794,
+      "grad_norm": 0.7573269605636597,
+      "learning_rate": 8.831305547225725e-06,
+      "loss": 0.7357,
+      "step": 8149
+    },
+    {
+      "epoch": 0.44856623919863503,
+      "grad_norm": 0.6884848475456238,
+      "learning_rate": 8.831027017822886e-06,
+      "loss": 0.7306,
+      "step": 8150
+    },
+    {
+      "epoch": 0.4486212780009907,
+      "grad_norm": 0.7715907096862793,
+      "learning_rate": 8.830748459627073e-06,
+      "loss": 0.8311,
+      "step": 8151
+    },
+    {
+      "epoch": 0.44867631680334635,
+      "grad_norm": 0.6919859647750854,
+      "learning_rate": 8.83046987264038e-06,
+      "loss": 0.845,
+      "step": 8152
+    },
+    {
+      "epoch": 0.44873135560570204,
+      "grad_norm": 0.7066411972045898,
+      "learning_rate": 8.830191256864902e-06,
+      "loss": 0.7554,
+      "step": 8153
+    },
+    {
+      "epoch": 0.4487863944080577,
+      "grad_norm": 0.754196047782898,
+      "learning_rate": 8.829912612302729e-06,
+      "loss": 0.7396,
+      "step": 8154
+    },
+    {
+      "epoch": 0.44884143321041337,
+      "grad_norm": 0.7612286806106567,
+      "learning_rate": 8.82963393895596e-06,
+      "loss": 0.8154,
+      "step": 8155
+    },
+    {
+      "epoch": 0.448896472012769,
+      "grad_norm": 0.8576892614364624,
+      "learning_rate": 8.829355236826688e-06,
+      "loss": 0.7395,
+      "step": 8156
+    },
+    {
+      "epoch": 0.4489515108151247,
+      "grad_norm": 0.6813738346099854,
+      "learning_rate": 8.829076505917005e-06,
+      "loss": 0.7661,
+      "step": 8157
+    },
+    {
+      "epoch": 0.4490065496174803,
+      "grad_norm": 0.7453964948654175,
+      "learning_rate": 8.828797746229009e-06,
+      "loss": 0.8221,
+      "step": 8158
+    },
+    {
+      "epoch": 0.449061588419836,
+      "grad_norm": 0.7546728849411011,
+      "learning_rate": 8.828518957764795e-06,
+      "loss": 0.7717,
+      "step": 8159
+    },
+    {
+      "epoch": 0.44911662722219164,
+      "grad_norm": 0.8270652890205383,
+      "learning_rate": 8.828240140526456e-06,
+      "loss": 0.7582,
+      "step": 8160
+    },
+    {
+      "epoch": 0.44917166602454733,
+      "grad_norm": 0.8188696503639221,
+      "learning_rate": 8.827961294516089e-06,
+      "loss": 0.8841,
+      "step": 8161
+    },
+    {
+      "epoch": 0.44922670482690297,
+      "grad_norm": 0.9101365208625793,
+      "learning_rate": 8.82768241973579e-06,
+      "loss": 0.7099,
+      "step": 8162
+    },
+    {
+      "epoch": 0.4492817436292586,
+      "grad_norm": 0.6749762892723083,
+      "learning_rate": 8.827403516187656e-06,
+      "loss": 0.7766,
+      "step": 8163
+    },
+    {
+      "epoch": 0.4493367824316143,
+      "grad_norm": 1.1351534128189087,
+      "learning_rate": 8.827124583873781e-06,
+      "loss": 0.7536,
+      "step": 8164
+    },
+    {
+      "epoch": 0.4493918212339699,
+      "grad_norm": 0.8729487061500549,
+      "learning_rate": 8.826845622796261e-06,
+      "loss": 0.8613,
+      "step": 8165
+    },
+    {
+      "epoch": 0.4494468600363256,
+      "grad_norm": 0.7495871782302856,
+      "learning_rate": 8.826566632957193e-06,
+      "loss": 0.8365,
+      "step": 8166
+    },
+    {
+      "epoch": 0.44950189883868125,
+      "grad_norm": 0.6414516568183899,
+      "learning_rate": 8.826287614358677e-06,
+      "loss": 0.6574,
+      "step": 8167
+    },
+    {
+      "epoch": 0.44955693764103694,
+      "grad_norm": 0.6954017281532288,
+      "learning_rate": 8.826008567002805e-06,
+      "loss": 0.7857,
+      "step": 8168
+    },
+    {
+      "epoch": 0.44961197644339257,
+      "grad_norm": 0.7199459075927734,
+      "learning_rate": 8.825729490891678e-06,
+      "loss": 0.8585,
+      "step": 8169
+    },
+    {
+      "epoch": 0.44966701524574826,
+      "grad_norm": 0.8245406746864319,
+      "learning_rate": 8.825450386027392e-06,
+      "loss": 0.7238,
+      "step": 8170
+    },
+    {
+      "epoch": 0.4497220540481039,
+      "grad_norm": 0.6348667740821838,
+      "learning_rate": 8.825171252412044e-06,
+      "loss": 0.6991,
+      "step": 8171
+    },
+    {
+      "epoch": 0.4497770928504596,
+      "grad_norm": 0.6304741501808167,
+      "learning_rate": 8.824892090047734e-06,
+      "loss": 0.7101,
+      "step": 8172
+    },
+    {
+      "epoch": 0.4498321316528152,
+      "grad_norm": 0.7088820338249207,
+      "learning_rate": 8.82461289893656e-06,
+      "loss": 0.8217,
+      "step": 8173
+    },
+    {
+      "epoch": 0.4498871704551709,
+      "grad_norm": 0.7570851445198059,
+      "learning_rate": 8.824333679080617e-06,
+      "loss": 0.8029,
+      "step": 8174
+    },
+    {
+      "epoch": 0.44994220925752654,
+      "grad_norm": 0.7544378042221069,
+      "learning_rate": 8.824054430482007e-06,
+      "loss": 0.777,
+      "step": 8175
+    },
+    {
+      "epoch": 0.4499972480598822,
+      "grad_norm": 0.8226260542869568,
+      "learning_rate": 8.823775153142827e-06,
+      "loss": 0.8391,
+      "step": 8176
+    },
+    {
+      "epoch": 0.45005228686223786,
+      "grad_norm": 0.6861422061920166,
+      "learning_rate": 8.823495847065176e-06,
+      "loss": 0.7491,
+      "step": 8177
+    },
+    {
+      "epoch": 0.45010732566459355,
+      "grad_norm": 0.6643275618553162,
+      "learning_rate": 8.823216512251153e-06,
+      "loss": 0.6773,
+      "step": 8178
+    },
+    {
+      "epoch": 0.4501623644669492,
+      "grad_norm": 0.8201391100883484,
+      "learning_rate": 8.82293714870286e-06,
+      "loss": 0.8065,
+      "step": 8179
+    },
+    {
+      "epoch": 0.45021740326930487,
+      "grad_norm": 0.7783405780792236,
+      "learning_rate": 8.822657756422394e-06,
+      "loss": 0.7884,
+      "step": 8180
+    },
+    {
+      "epoch": 0.4502724420716605,
+      "grad_norm": 0.720745861530304,
+      "learning_rate": 8.822378335411856e-06,
+      "loss": 0.765,
+      "step": 8181
     }
   ],
   "logging_steps": 1,
@@ -50930,7 +57293,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 2.1460117712975954e+19,
+  "total_flos": 2.414263242709795e+19,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null