diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": 0.2650785446166992,
   "best_model_checkpoint": "./w2v-bert-2.0-hausa_naijavoices_100h/checkpoint-10000",
-  "epoch": 36.312849162011176,
+  "epoch": 39.10614525139665,
   "eval_steps": 1000,
-  "global_step": 13000,
+  "global_step": 14000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -91137,6 +91137,7016 @@
       "eval_steps_per_second": 0.628,
       "eval_wer": 0.34800703890952767,
       "step": 13000
+    },
+    {
+      "epoch": 36.315642458100555,
+      "grad_norm": 0.5103163719177246,
+      "learning_rate": 2.8493482309124767e-05,
+      "loss": 0.1163,
+      "step": 13001
+    },
+    {
+      "epoch": 36.31843575418994,
+      "grad_norm": 0.45563486218452454,
+      "learning_rate": 2.8493109869646184e-05,
+      "loss": 0.0957,
+      "step": 13002
+    },
+    {
+      "epoch": 36.32122905027933,
+      "grad_norm": 0.6634702682495117,
+      "learning_rate": 2.8492737430167597e-05,
+      "loss": 0.086,
+      "step": 13003
+    },
+    {
+      "epoch": 36.324022346368714,
+      "grad_norm": 0.3813391327857971,
+      "learning_rate": 2.8492364990689013e-05,
+      "loss": 0.0841,
+      "step": 13004
+    },
+    {
+      "epoch": 36.3268156424581,
+      "grad_norm": 0.4647751748561859,
+      "learning_rate": 2.849199255121043e-05,
+      "loss": 0.0837,
+      "step": 13005
+    },
+    {
+      "epoch": 36.32960893854749,
+      "grad_norm": 0.5191593170166016,
+      "learning_rate": 2.8491620111731843e-05,
+      "loss": 0.0903,
+      "step": 13006
+    },
+    {
+      "epoch": 36.332402234636874,
+      "grad_norm": 2.6562979221343994,
+      "learning_rate": 2.849124767225326e-05,
+      "loss": 0.0852,
+      "step": 13007
+    },
+    {
+      "epoch": 36.33519553072626,
+      "grad_norm": 1.221608281135559,
+      "learning_rate": 2.8490875232774676e-05,
+      "loss": 0.102,
+      "step": 13008
+    },
+    {
+      "epoch": 36.33798882681564,
+      "grad_norm": 1.1749480962753296,
+      "learning_rate": 2.8490502793296093e-05,
+      "loss": 0.0789,
+      "step": 13009
+    },
+    {
+      "epoch": 36.340782122905026,
+      "grad_norm": 0.8152073621749878,
+      "learning_rate": 2.8490130353817502e-05,
+      "loss": 0.082,
+      "step": 13010
+    },
+    {
+      "epoch": 36.34357541899441,
+      "grad_norm": 0.6769988536834717,
+      "learning_rate": 2.848975791433892e-05,
+      "loss": 0.0732,
+      "step": 13011
+    },
+    {
+      "epoch": 36.3463687150838,
+      "grad_norm": 0.7155337929725647,
+      "learning_rate": 2.8489385474860336e-05,
+      "loss": 0.0835,
+      "step": 13012
+    },
+    {
+      "epoch": 36.349162011173185,
+      "grad_norm": 0.9940541386604309,
+      "learning_rate": 2.8489013035381752e-05,
+      "loss": 0.1017,
+      "step": 13013
+    },
+    {
+      "epoch": 36.35195530726257,
+      "grad_norm": 0.3431513011455536,
+      "learning_rate": 2.848864059590317e-05,
+      "loss": 0.1231,
+      "step": 13014
+    },
+    {
+      "epoch": 36.35474860335196,
+      "grad_norm": 0.4082512855529785,
+      "learning_rate": 2.8488268156424582e-05,
+      "loss": 0.1119,
+      "step": 13015
+    },
+    {
+      "epoch": 36.357541899441344,
+      "grad_norm": 0.6857700347900391,
+      "learning_rate": 2.8487895716946e-05,
+      "loss": 0.1125,
+      "step": 13016
+    },
+    {
+      "epoch": 36.36033519553072,
+      "grad_norm": 0.4404042661190033,
+      "learning_rate": 2.848752327746741e-05,
+      "loss": 0.1183,
+      "step": 13017
+    },
+    {
+      "epoch": 36.36312849162011,
+      "grad_norm": 0.4675848186016083,
+      "learning_rate": 2.8487150837988828e-05,
+      "loss": 0.122,
+      "step": 13018
+    },
+    {
+      "epoch": 36.365921787709496,
+      "grad_norm": 0.7972007393836975,
+      "learning_rate": 2.848677839851024e-05,
+      "loss": 0.1244,
+      "step": 13019
+    },
+    {
+      "epoch": 36.36871508379888,
+      "grad_norm": 0.8320142030715942,
+      "learning_rate": 2.8486405959031658e-05,
+      "loss": 0.1171,
+      "step": 13020
+    },
+    {
+      "epoch": 36.37150837988827,
+      "grad_norm": 0.9292674660682678,
+      "learning_rate": 2.8486033519553074e-05,
+      "loss": 0.1069,
+      "step": 13021
+    },
+    {
+      "epoch": 36.374301675977655,
+      "grad_norm": 0.4998623728752136,
+      "learning_rate": 2.848566108007449e-05,
+      "loss": 0.1004,
+      "step": 13022
+    },
+    {
+      "epoch": 36.37709497206704,
+      "grad_norm": 1.3461363315582275,
+      "learning_rate": 2.8485288640595904e-05,
+      "loss": 0.0973,
+      "step": 13023
+    },
+    {
+      "epoch": 36.37988826815643,
+      "grad_norm": 0.49873998761177063,
+      "learning_rate": 2.8484916201117317e-05,
+      "loss": 0.1177,
+      "step": 13024
+    },
+    {
+      "epoch": 36.38268156424581,
+      "grad_norm": 0.5666747689247131,
+      "learning_rate": 2.8484543761638734e-05,
+      "loss": 0.1083,
+      "step": 13025
+    },
+    {
+      "epoch": 36.385474860335194,
+      "grad_norm": 1.3581769466400146,
+      "learning_rate": 2.848417132216015e-05,
+      "loss": 0.0955,
+      "step": 13026
+    },
+    {
+      "epoch": 36.38826815642458,
+      "grad_norm": 0.6231361627578735,
+      "learning_rate": 2.8483798882681567e-05,
+      "loss": 0.0989,
+      "step": 13027
+    },
+    {
+      "epoch": 36.391061452513966,
+      "grad_norm": 0.5649336576461792,
+      "learning_rate": 2.848342644320298e-05,
+      "loss": 0.1005,
+      "step": 13028
+    },
+    {
+      "epoch": 36.39385474860335,
+      "grad_norm": 0.5340343117713928,
+      "learning_rate": 2.8483054003724396e-05,
+      "loss": 0.0914,
+      "step": 13029
+    },
+    {
+      "epoch": 36.39664804469274,
+      "grad_norm": 0.6382849216461182,
+      "learning_rate": 2.848268156424581e-05,
+      "loss": 0.0874,
+      "step": 13030
+    },
+    {
+      "epoch": 36.399441340782126,
+      "grad_norm": 0.7184431552886963,
+      "learning_rate": 2.8482309124767226e-05,
+      "loss": 0.0749,
+      "step": 13031
+    },
+    {
+      "epoch": 36.40223463687151,
+      "grad_norm": 0.7611850500106812,
+      "learning_rate": 2.8481936685288643e-05,
+      "loss": 0.064,
+      "step": 13032
+    },
+    {
+      "epoch": 36.40502793296089,
+      "grad_norm": 0.8529970645904541,
+      "learning_rate": 2.8481564245810056e-05,
+      "loss": 0.084,
+      "step": 13033
+    },
+    {
+      "epoch": 36.40782122905028,
+      "grad_norm": 0.6734018325805664,
+      "learning_rate": 2.8481191806331472e-05,
+      "loss": 0.0931,
+      "step": 13034
+    },
+    {
+      "epoch": 36.410614525139664,
+      "grad_norm": 0.7495433688163757,
+      "learning_rate": 2.848081936685289e-05,
+      "loss": 0.0937,
+      "step": 13035
+    },
+    {
+      "epoch": 36.41340782122905,
+      "grad_norm": 0.7531351447105408,
+      "learning_rate": 2.8480446927374305e-05,
+      "loss": 0.0887,
+      "step": 13036
+    },
+    {
+      "epoch": 36.41620111731844,
+      "grad_norm": 0.9327375888824463,
+      "learning_rate": 2.8480074487895715e-05,
+      "loss": 0.0894,
+      "step": 13037
+    },
+    {
+      "epoch": 36.41899441340782,
+      "grad_norm": 2.5498523712158203,
+      "learning_rate": 2.847970204841713e-05,
+      "loss": 0.1179,
+      "step": 13038
+    },
+    {
+      "epoch": 36.42178770949721,
+      "grad_norm": 0.732311487197876,
+      "learning_rate": 2.8479329608938548e-05,
+      "loss": 0.1288,
+      "step": 13039
+    },
+    {
+      "epoch": 36.42458100558659,
+      "grad_norm": 0.3687518239021301,
+      "learning_rate": 2.8478957169459965e-05,
+      "loss": 0.1117,
+      "step": 13040
+    },
+    {
+      "epoch": 36.427374301675975,
+      "grad_norm": 0.41522008180618286,
+      "learning_rate": 2.847858472998138e-05,
+      "loss": 0.0986,
+      "step": 13041
+    },
+    {
+      "epoch": 36.43016759776536,
+      "grad_norm": 1.1092604398727417,
+      "learning_rate": 2.8478212290502794e-05,
+      "loss": 0.1044,
+      "step": 13042
+    },
+    {
+      "epoch": 36.43296089385475,
+      "grad_norm": 0.5997103452682495,
+      "learning_rate": 2.8477839851024208e-05,
+      "loss": 0.1477,
+      "step": 13043
+    },
+    {
+      "epoch": 36.435754189944134,
+      "grad_norm": 1.1083749532699585,
+      "learning_rate": 2.8477467411545624e-05,
+      "loss": 0.1203,
+      "step": 13044
+    },
+    {
+      "epoch": 36.43854748603352,
+      "grad_norm": 0.6279087662696838,
+      "learning_rate": 2.847709497206704e-05,
+      "loss": 0.1205,
+      "step": 13045
+    },
+    {
+      "epoch": 36.44134078212291,
+      "grad_norm": 0.35380157828330994,
+      "learning_rate": 2.8476722532588454e-05,
+      "loss": 0.0955,
+      "step": 13046
+    },
+    {
+      "epoch": 36.444134078212294,
+      "grad_norm": 0.44671377539634705,
+      "learning_rate": 2.847635009310987e-05,
+      "loss": 0.1007,
+      "step": 13047
+    },
+    {
+      "epoch": 36.44692737430167,
+      "grad_norm": 0.5153653621673584,
+      "learning_rate": 2.8475977653631287e-05,
+      "loss": 0.0959,
+      "step": 13048
+    },
+    {
+      "epoch": 36.44972067039106,
+      "grad_norm": 0.7488757967948914,
+      "learning_rate": 2.8475605214152703e-05,
+      "loss": 0.0933,
+      "step": 13049
+    },
+    {
+      "epoch": 36.452513966480446,
+      "grad_norm": 1.1935946941375732,
+      "learning_rate": 2.8475232774674116e-05,
+      "loss": 0.1124,
+      "step": 13050
+    },
+    {
+      "epoch": 36.45530726256983,
+      "grad_norm": 0.9721536636352539,
+      "learning_rate": 2.847486033519553e-05,
+      "loss": 0.0827,
+      "step": 13051
+    },
+    {
+      "epoch": 36.45810055865922,
+      "grad_norm": 0.5015149712562561,
+      "learning_rate": 2.8474487895716946e-05,
+      "loss": 0.0967,
+      "step": 13052
+    },
+    {
+      "epoch": 36.460893854748605,
+      "grad_norm": 0.9974826574325562,
+      "learning_rate": 2.8474115456238363e-05,
+      "loss": 0.0924,
+      "step": 13053
+    },
+    {
+      "epoch": 36.46368715083799,
+      "grad_norm": 0.8104403614997864,
+      "learning_rate": 2.847374301675978e-05,
+      "loss": 0.0931,
+      "step": 13054
+    },
+    {
+      "epoch": 36.46648044692738,
+      "grad_norm": 0.632412314414978,
+      "learning_rate": 2.8473370577281192e-05,
+      "loss": 0.0825,
+      "step": 13055
+    },
+    {
+      "epoch": 36.46927374301676,
+      "grad_norm": 0.8117548227310181,
+      "learning_rate": 2.847299813780261e-05,
+      "loss": 0.084,
+      "step": 13056
+    },
+    {
+      "epoch": 36.47206703910614,
+      "grad_norm": 0.728563129901886,
+      "learning_rate": 2.8472625698324022e-05,
+      "loss": 0.091,
+      "step": 13057
+    },
+    {
+      "epoch": 36.47486033519553,
+      "grad_norm": 0.6499229669570923,
+      "learning_rate": 2.847225325884544e-05,
+      "loss": 0.1043,
+      "step": 13058
+    },
+    {
+      "epoch": 36.477653631284916,
+      "grad_norm": 0.5659500360488892,
+      "learning_rate": 2.8471880819366852e-05,
+      "loss": 0.0817,
+      "step": 13059
+    },
+    {
+      "epoch": 36.4804469273743,
+      "grad_norm": 1.1683993339538574,
+      "learning_rate": 2.8471508379888268e-05,
+      "loss": 0.0788,
+      "step": 13060
+    },
+    {
+      "epoch": 36.48324022346369,
+      "grad_norm": 1.722424030303955,
+      "learning_rate": 2.8471135940409685e-05,
+      "loss": 0.0754,
+      "step": 13061
+    },
+    {
+      "epoch": 36.486033519553075,
+      "grad_norm": 1.2399303913116455,
+      "learning_rate": 2.84707635009311e-05,
+      "loss": 0.117,
+      "step": 13062
+    },
+    {
+      "epoch": 36.48882681564246,
+      "grad_norm": 1.3291460275650024,
+      "learning_rate": 2.8470391061452514e-05,
+      "loss": 0.1093,
+      "step": 13063
+    },
+    {
+      "epoch": 36.49162011173184,
+      "grad_norm": 0.41691839694976807,
+      "learning_rate": 2.8470018621973928e-05,
+      "loss": 0.1385,
+      "step": 13064
+    },
+    {
+      "epoch": 36.49441340782123,
+      "grad_norm": 0.8629980087280273,
+      "learning_rate": 2.8469646182495344e-05,
+      "loss": 0.1391,
+      "step": 13065
+    },
+    {
+      "epoch": 36.497206703910614,
+      "grad_norm": 0.5989841818809509,
+      "learning_rate": 2.846927374301676e-05,
+      "loss": 0.1405,
+      "step": 13066
+    },
+    {
+      "epoch": 36.5,
+      "grad_norm": 0.4392365515232086,
+      "learning_rate": 2.8468901303538177e-05,
+      "loss": 0.1256,
+      "step": 13067
+    },
+    {
+      "epoch": 36.502793296089386,
+      "grad_norm": 0.41069284081459045,
+      "learning_rate": 2.846852886405959e-05,
+      "loss": 0.1267,
+      "step": 13068
+    },
+    {
+      "epoch": 36.50558659217877,
+      "grad_norm": 1.3699334859848022,
+      "learning_rate": 2.8468156424581007e-05,
+      "loss": 0.1088,
+      "step": 13069
+    },
+    {
+      "epoch": 36.50837988826816,
+      "grad_norm": 0.3773321807384491,
+      "learning_rate": 2.846778398510242e-05,
+      "loss": 0.1207,
+      "step": 13070
+    },
+    {
+      "epoch": 36.51117318435754,
+      "grad_norm": 0.3941558599472046,
+      "learning_rate": 2.8467411545623837e-05,
+      "loss": 0.0972,
+      "step": 13071
+    },
+    {
+      "epoch": 36.513966480446925,
+      "grad_norm": 0.5023883581161499,
+      "learning_rate": 2.8467039106145253e-05,
+      "loss": 0.0949,
+      "step": 13072
+    },
+    {
+      "epoch": 36.51675977653631,
+      "grad_norm": 0.6170898079872131,
+      "learning_rate": 2.8466666666666666e-05,
+      "loss": 0.1054,
+      "step": 13073
+    },
+    {
+      "epoch": 36.5195530726257,
+      "grad_norm": 0.4747655689716339,
+      "learning_rate": 2.8466294227188083e-05,
+      "loss": 0.1049,
+      "step": 13074
+    },
+    {
+      "epoch": 36.522346368715084,
+      "grad_norm": 0.6342079043388367,
+      "learning_rate": 2.84659217877095e-05,
+      "loss": 0.1078,
+      "step": 13075
+    },
+    {
+      "epoch": 36.52513966480447,
+      "grad_norm": 0.9232836961746216,
+      "learning_rate": 2.8465549348230916e-05,
+      "loss": 0.0895,
+      "step": 13076
+    },
+    {
+      "epoch": 36.52793296089386,
+      "grad_norm": 0.6280178427696228,
+      "learning_rate": 2.8465176908752326e-05,
+      "loss": 0.1007,
+      "step": 13077
+    },
+    {
+      "epoch": 36.53072625698324,
+      "grad_norm": 1.9618819952011108,
+      "learning_rate": 2.8464804469273742e-05,
+      "loss": 0.0875,
+      "step": 13078
+    },
+    {
+      "epoch": 36.53351955307262,
+      "grad_norm": 0.5735899209976196,
+      "learning_rate": 2.846443202979516e-05,
+      "loss": 0.0871,
+      "step": 13079
+    },
+    {
+      "epoch": 36.53631284916201,
+      "grad_norm": 0.7038599848747253,
+      "learning_rate": 2.8464059590316575e-05,
+      "loss": 0.0882,
+      "step": 13080
+    },
+    {
+      "epoch": 36.539106145251395,
+      "grad_norm": 0.5991548895835876,
+      "learning_rate": 2.8463687150837992e-05,
+      "loss": 0.0794,
+      "step": 13081
+    },
+    {
+      "epoch": 36.54189944134078,
+      "grad_norm": 0.7051090002059937,
+      "learning_rate": 2.8463314711359405e-05,
+      "loss": 0.0823,
+      "step": 13082
+    },
+    {
+      "epoch": 36.54469273743017,
+      "grad_norm": 0.5633767247200012,
+      "learning_rate": 2.8462942271880818e-05,
+      "loss": 0.0807,
+      "step": 13083
+    },
+    {
+      "epoch": 36.547486033519554,
+      "grad_norm": 0.7850005626678467,
+      "learning_rate": 2.8462569832402235e-05,
+      "loss": 0.0999,
+      "step": 13084
+    },
+    {
+      "epoch": 36.55027932960894,
+      "grad_norm": 0.6875097751617432,
+      "learning_rate": 2.846219739292365e-05,
+      "loss": 0.0797,
+      "step": 13085
+    },
+    {
+      "epoch": 36.55307262569833,
+      "grad_norm": 0.8078372478485107,
+      "learning_rate": 2.8461824953445064e-05,
+      "loss": 0.0774,
+      "step": 13086
+    },
+    {
+      "epoch": 36.555865921787706,
+      "grad_norm": 0.9205446243286133,
+      "learning_rate": 2.846145251396648e-05,
+      "loss": 0.0913,
+      "step": 13087
+    },
+    {
+      "epoch": 36.55865921787709,
+      "grad_norm": 1.0804953575134277,
+      "learning_rate": 2.8461080074487897e-05,
+      "loss": 0.1215,
+      "step": 13088
+    },
+    {
+      "epoch": 36.56145251396648,
+      "grad_norm": 0.5169636607170105,
+      "learning_rate": 2.8460707635009314e-05,
+      "loss": 0.12,
+      "step": 13089
+    },
+    {
+      "epoch": 36.564245810055866,
+      "grad_norm": 0.7846399545669556,
+      "learning_rate": 2.8460335195530727e-05,
+      "loss": 0.1233,
+      "step": 13090
+    },
+    {
+      "epoch": 36.56703910614525,
+      "grad_norm": 0.6060928702354431,
+      "learning_rate": 2.845996275605214e-05,
+      "loss": 0.1051,
+      "step": 13091
+    },
+    {
+      "epoch": 36.56983240223464,
+      "grad_norm": 0.9782649278640747,
+      "learning_rate": 2.8459590316573557e-05,
+      "loss": 0.1298,
+      "step": 13092
+    },
+    {
+      "epoch": 36.572625698324025,
+      "grad_norm": 0.5207920074462891,
+      "learning_rate": 2.8459217877094973e-05,
+      "loss": 0.1315,
+      "step": 13093
+    },
+    {
+      "epoch": 36.57541899441341,
+      "grad_norm": 0.5926817059516907,
+      "learning_rate": 2.845884543761639e-05,
+      "loss": 0.1254,
+      "step": 13094
+    },
+    {
+      "epoch": 36.57821229050279,
+      "grad_norm": 0.707974374294281,
+      "learning_rate": 2.8458472998137803e-05,
+      "loss": 0.1204,
+      "step": 13095
+    },
+    {
+      "epoch": 36.58100558659218,
+      "grad_norm": 0.5093036890029907,
+      "learning_rate": 2.845810055865922e-05,
+      "loss": 0.1044,
+      "step": 13096
+    },
+    {
+      "epoch": 36.58379888268156,
+      "grad_norm": 0.47131848335266113,
+      "learning_rate": 2.8457728119180633e-05,
+      "loss": 0.1029,
+      "step": 13097
+    },
+    {
+      "epoch": 36.58659217877095,
+      "grad_norm": 0.7507746815681458,
+      "learning_rate": 2.845735567970205e-05,
+      "loss": 0.1113,
+      "step": 13098
+    },
+    {
+      "epoch": 36.589385474860336,
+      "grad_norm": 0.445652037858963,
+      "learning_rate": 2.8456983240223466e-05,
+      "loss": 0.0843,
+      "step": 13099
+    },
+    {
+      "epoch": 36.59217877094972,
+      "grad_norm": 0.6210559606552124,
+      "learning_rate": 2.845661080074488e-05,
+      "loss": 0.0881,
+      "step": 13100
+    },
+    {
+      "epoch": 36.59497206703911,
+      "grad_norm": 0.5198919773101807,
+      "learning_rate": 2.8456238361266295e-05,
+      "loss": 0.099,
+      "step": 13101
+    },
+    {
+      "epoch": 36.59776536312849,
+      "grad_norm": 0.7758316397666931,
+      "learning_rate": 2.8455865921787712e-05,
+      "loss": 0.1087,
+      "step": 13102
+    },
+    {
+      "epoch": 36.600558659217874,
+      "grad_norm": 0.36870887875556946,
+      "learning_rate": 2.8455493482309125e-05,
+      "loss": 0.0796,
+      "step": 13103
+    },
+    {
+      "epoch": 36.60335195530726,
+      "grad_norm": 4.285307884216309,
+      "learning_rate": 2.8455121042830538e-05,
+      "loss": 0.0923,
+      "step": 13104
+    },
+    {
+      "epoch": 36.60614525139665,
+      "grad_norm": 1.537385106086731,
+      "learning_rate": 2.8454748603351955e-05,
+      "loss": 0.1028,
+      "step": 13105
+    },
+    {
+      "epoch": 36.608938547486034,
+      "grad_norm": 0.9062882661819458,
+      "learning_rate": 2.845437616387337e-05,
+      "loss": 0.0798,
+      "step": 13106
+    },
+    {
+      "epoch": 36.61173184357542,
+      "grad_norm": 1.2549960613250732,
+      "learning_rate": 2.8454003724394788e-05,
+      "loss": 0.0865,
+      "step": 13107
+    },
+    {
+      "epoch": 36.614525139664806,
+      "grad_norm": 0.4961925148963928,
+      "learning_rate": 2.8453631284916204e-05,
+      "loss": 0.0696,
+      "step": 13108
+    },
+    {
+      "epoch": 36.61731843575419,
+      "grad_norm": 1.2841949462890625,
+      "learning_rate": 2.8453258845437618e-05,
+      "loss": 0.0937,
+      "step": 13109
+    },
+    {
+      "epoch": 36.62011173184357,
+      "grad_norm": 0.7004602551460266,
+      "learning_rate": 2.845288640595903e-05,
+      "loss": 0.0846,
+      "step": 13110
+    },
+    {
+      "epoch": 36.62290502793296,
+      "grad_norm": 1.0388610363006592,
+      "learning_rate": 2.8452513966480447e-05,
+      "loss": 0.0798,
+      "step": 13111
+    },
+    {
+      "epoch": 36.625698324022345,
+      "grad_norm": 0.8924124240875244,
+      "learning_rate": 2.8452141527001864e-05,
+      "loss": 0.0911,
+      "step": 13112
+    },
+    {
+      "epoch": 36.62849162011173,
+      "grad_norm": 0.9572955369949341,
+      "learning_rate": 2.8451769087523277e-05,
+      "loss": 0.1039,
+      "step": 13113
+    },
+    {
+      "epoch": 36.63128491620112,
+      "grad_norm": 0.5310201644897461,
+      "learning_rate": 2.8451396648044693e-05,
+      "loss": 0.1167,
+      "step": 13114
+    },
+    {
+      "epoch": 36.634078212290504,
+      "grad_norm": 0.4729738235473633,
+      "learning_rate": 2.845102420856611e-05,
+      "loss": 0.1421,
+      "step": 13115
+    },
+    {
+      "epoch": 36.63687150837989,
+      "grad_norm": 0.3305386006832123,
+      "learning_rate": 2.8450651769087526e-05,
+      "loss": 0.1175,
+      "step": 13116
+    },
+    {
+      "epoch": 36.63966480446928,
+      "grad_norm": 0.49460217356681824,
+      "learning_rate": 2.845027932960894e-05,
+      "loss": 0.1306,
+      "step": 13117
+    },
+    {
+      "epoch": 36.642458100558656,
+      "grad_norm": 0.3999815881252289,
+      "learning_rate": 2.8449906890130353e-05,
+      "loss": 0.1112,
+      "step": 13118
+    },
+    {
+      "epoch": 36.64525139664804,
+      "grad_norm": 0.6515106558799744,
+      "learning_rate": 2.844953445065177e-05,
+      "loss": 0.1082,
+      "step": 13119
+    },
+    {
+      "epoch": 36.64804469273743,
+      "grad_norm": 1.4159810543060303,
+      "learning_rate": 2.8449162011173186e-05,
+      "loss": 0.1204,
+      "step": 13120
+    },
+    {
+      "epoch": 36.650837988826815,
+      "grad_norm": 0.7723098993301392,
+      "learning_rate": 2.8448789571694602e-05,
+      "loss": 0.1074,
+      "step": 13121
+    },
+    {
+      "epoch": 36.6536312849162,
+      "grad_norm": 0.9828838109970093,
+      "learning_rate": 2.8448417132216016e-05,
+      "loss": 0.1013,
+      "step": 13122
+    },
+    {
+      "epoch": 36.65642458100559,
+      "grad_norm": 0.5460968613624573,
+      "learning_rate": 2.844804469273743e-05,
+      "loss": 0.1136,
+      "step": 13123
+    },
+    {
+      "epoch": 36.659217877094974,
+      "grad_norm": 0.4706367552280426,
+      "learning_rate": 2.8447672253258845e-05,
+      "loss": 0.1149,
+      "step": 13124
+    },
+    {
+      "epoch": 36.66201117318436,
+      "grad_norm": 0.38748401403427124,
+      "learning_rate": 2.8447299813780262e-05,
+      "loss": 0.0912,
+      "step": 13125
+    },
+    {
+      "epoch": 36.66480446927374,
+      "grad_norm": 0.45856836438179016,
+      "learning_rate": 2.8446927374301678e-05,
+      "loss": 0.0865,
+      "step": 13126
+    },
+    {
+      "epoch": 36.667597765363126,
+      "grad_norm": 0.7570058703422546,
+      "learning_rate": 2.844655493482309e-05,
+      "loss": 0.1041,
+      "step": 13127
+    },
+    {
+      "epoch": 36.67039106145251,
+      "grad_norm": 0.821748673915863,
+      "learning_rate": 2.8446182495344508e-05,
+      "loss": 0.1071,
+      "step": 13128
+    },
+    {
+      "epoch": 36.6731843575419,
+      "grad_norm": 0.5184983611106873,
+      "learning_rate": 2.8445810055865925e-05,
+      "loss": 0.0797,
+      "step": 13129
+    },
+    {
+      "epoch": 36.675977653631286,
+      "grad_norm": 0.3808233439922333,
+      "learning_rate": 2.8445437616387338e-05,
+      "loss": 0.0871,
+      "step": 13130
+    },
+    {
+      "epoch": 36.67877094972067,
+      "grad_norm": 0.6422951221466064,
+      "learning_rate": 2.844506517690875e-05,
+      "loss": 0.0887,
+      "step": 13131
+    },
+    {
+      "epoch": 36.68156424581006,
+      "grad_norm": 0.684080958366394,
+      "learning_rate": 2.8444692737430167e-05,
+      "loss": 0.0795,
+      "step": 13132
+    },
+    {
+      "epoch": 36.684357541899445,
+      "grad_norm": 0.7174074053764343,
+      "learning_rate": 2.8444320297951584e-05,
+      "loss": 0.0931,
+      "step": 13133
+    },
+    {
+      "epoch": 36.687150837988824,
+      "grad_norm": 0.7364905476570129,
+      "learning_rate": 2.8443947858473e-05,
+      "loss": 0.0802,
+      "step": 13134
+    },
+    {
+      "epoch": 36.68994413407821,
+      "grad_norm": 0.7372745871543884,
+      "learning_rate": 2.8443575418994417e-05,
+      "loss": 0.0972,
+      "step": 13135
+    },
+    {
+      "epoch": 36.6927374301676,
+      "grad_norm": 0.7332994341850281,
+      "learning_rate": 2.844320297951583e-05,
+      "loss": 0.0594,
+      "step": 13136
+    },
+    {
+      "epoch": 36.69553072625698,
+      "grad_norm": 7.079324245452881,
+      "learning_rate": 2.8442830540037243e-05,
+      "loss": 0.0908,
+      "step": 13137
+    },
+    {
+      "epoch": 36.69832402234637,
+      "grad_norm": 1.1419892311096191,
+      "learning_rate": 2.844245810055866e-05,
+      "loss": 0.1062,
+      "step": 13138
+    },
+    {
+      "epoch": 36.701117318435756,
+      "grad_norm": 0.4915393590927124,
+      "learning_rate": 2.8442085661080076e-05,
+      "loss": 0.1254,
+      "step": 13139
+    },
+    {
+      "epoch": 36.70391061452514,
+      "grad_norm": 0.355114221572876,
+      "learning_rate": 2.844171322160149e-05,
+      "loss": 0.1235,
+      "step": 13140
+    },
+    {
+      "epoch": 36.70670391061452,
+      "grad_norm": 1.367867350578308,
+      "learning_rate": 2.8441340782122906e-05,
+      "loss": 0.1279,
+      "step": 13141
+    },
+    {
+      "epoch": 36.70949720670391,
+      "grad_norm": 0.5479897856712341,
+      "learning_rate": 2.8440968342644323e-05,
+      "loss": 0.1014,
+      "step": 13142
+    },
+    {
+      "epoch": 36.712290502793294,
+      "grad_norm": 0.44179767370224,
+      "learning_rate": 2.8440595903165736e-05,
+      "loss": 0.1249,
+      "step": 13143
+    },
+    {
+      "epoch": 36.71508379888268,
+      "grad_norm": 0.4458467960357666,
+      "learning_rate": 2.8440223463687152e-05,
+      "loss": 0.0995,
+      "step": 13144
+    },
+    {
+      "epoch": 36.71787709497207,
+      "grad_norm": 1.1193277835845947,
+      "learning_rate": 2.8439851024208565e-05,
+      "loss": 0.1329,
+      "step": 13145
+    },
+    {
+      "epoch": 36.720670391061454,
+      "grad_norm": 0.6375119686126709,
+      "learning_rate": 2.8439478584729982e-05,
+      "loss": 0.0915,
+      "step": 13146
+    },
+    {
+      "epoch": 36.72346368715084,
+      "grad_norm": 0.4403560161590576,
+      "learning_rate": 2.84391061452514e-05,
+      "loss": 0.1027,
+      "step": 13147
+    },
+    {
+      "epoch": 36.726256983240226,
+      "grad_norm": 0.6501517295837402,
+      "learning_rate": 2.8438733705772815e-05,
+      "loss": 0.1,
+      "step": 13148
+    },
+    {
+      "epoch": 36.729050279329606,
+      "grad_norm": 0.4135424792766571,
+      "learning_rate": 2.8438361266294228e-05,
+      "loss": 0.1057,
+      "step": 13149
+    },
+    {
+      "epoch": 36.73184357541899,
+      "grad_norm": 0.8687627911567688,
+      "learning_rate": 2.843798882681564e-05,
+      "loss": 0.1108,
+      "step": 13150
+    },
+    {
+      "epoch": 36.73463687150838,
+      "grad_norm": 0.4155297577381134,
+      "learning_rate": 2.8437616387337058e-05,
+      "loss": 0.0879,
+      "step": 13151
+    },
+    {
+      "epoch": 36.737430167597765,
+      "grad_norm": 0.5047573447227478,
+      "learning_rate": 2.8437243947858474e-05,
+      "loss": 0.101,
+      "step": 13152
+    },
+    {
+      "epoch": 36.74022346368715,
+      "grad_norm": 0.6942926645278931,
+      "learning_rate": 2.8436871508379887e-05,
+      "loss": 0.0892,
+      "step": 13153
+    },
+    {
+      "epoch": 36.74301675977654,
+      "grad_norm": 1.0379856824874878,
+      "learning_rate": 2.8436499068901304e-05,
+      "loss": 0.0922,
+      "step": 13154
+    },
+    {
+      "epoch": 36.745810055865924,
+      "grad_norm": 0.4566536247730255,
+      "learning_rate": 2.843612662942272e-05,
+      "loss": 0.0717,
+      "step": 13155
+    },
+    {
+      "epoch": 36.74860335195531,
+      "grad_norm": 0.9123262763023376,
+      "learning_rate": 2.8435754189944137e-05,
+      "loss": 0.0781,
+      "step": 13156
+    },
+    {
+      "epoch": 36.75139664804469,
+      "grad_norm": 0.4724482297897339,
+      "learning_rate": 2.843538175046555e-05,
+      "loss": 0.0946,
+      "step": 13157
+    },
+    {
+      "epoch": 36.754189944134076,
+      "grad_norm": 0.7829778790473938,
+      "learning_rate": 2.8435009310986963e-05,
+      "loss": 0.0967,
+      "step": 13158
+    },
+    {
+      "epoch": 36.75698324022346,
+      "grad_norm": 0.9873928427696228,
+      "learning_rate": 2.843463687150838e-05,
+      "loss": 0.0709,
+      "step": 13159
+    },
+    {
+      "epoch": 36.75977653631285,
+      "grad_norm": 0.6474883556365967,
+      "learning_rate": 2.8434264432029796e-05,
+      "loss": 0.0921,
+      "step": 13160
+    },
+    {
+      "epoch": 36.762569832402235,
+      "grad_norm": 0.6852738857269287,
+      "learning_rate": 2.8433891992551213e-05,
+      "loss": 0.0933,
+      "step": 13161
+    },
+    {
+      "epoch": 36.76536312849162,
+      "grad_norm": 0.5250356793403625,
+      "learning_rate": 2.8433519553072626e-05,
+      "loss": 0.0797,
+      "step": 13162
+    },
+    {
+      "epoch": 36.76815642458101,
+      "grad_norm": 1.3000695705413818,
+      "learning_rate": 2.843314711359404e-05,
+      "loss": 0.101,
+      "step": 13163
+    },
+    {
+      "epoch": 36.770949720670394,
+      "grad_norm": 0.4818575084209442,
+      "learning_rate": 2.8432774674115456e-05,
+      "loss": 0.132,
+      "step": 13164
+    },
+    {
+      "epoch": 36.773743016759774,
+      "grad_norm": 0.3751843273639679,
+      "learning_rate": 2.8432402234636872e-05,
+      "loss": 0.1106,
+      "step": 13165
+    },
+    {
+      "epoch": 36.77653631284916,
+      "grad_norm": 0.5293797254562378,
+      "learning_rate": 2.843202979515829e-05,
+      "loss": 0.1372,
+      "step": 13166
+    },
+    {
+      "epoch": 36.779329608938546,
+      "grad_norm": 0.47639548778533936,
+      "learning_rate": 2.8431657355679702e-05,
+      "loss": 0.1118,
+      "step": 13167
+    },
+    {
+      "epoch": 36.78212290502793,
+      "grad_norm": 0.4971713423728943,
+      "learning_rate": 2.843128491620112e-05,
+      "loss": 0.1132,
+      "step": 13168
+    },
+    {
+      "epoch": 36.78491620111732,
+      "grad_norm": 0.6892354488372803,
+      "learning_rate": 2.8430912476722535e-05,
+      "loss": 0.1185,
+      "step": 13169
+    },
+    {
+      "epoch": 36.787709497206706,
+      "grad_norm": 0.3397558629512787,
+      "learning_rate": 2.8430540037243948e-05,
+      "loss": 0.0944,
+      "step": 13170
+    },
+    {
+      "epoch": 36.79050279329609,
+      "grad_norm": 0.4257424473762512,
+      "learning_rate": 2.843016759776536e-05,
+      "loss": 0.1035,
+      "step": 13171
+    },
+    {
+      "epoch": 36.79329608938548,
+      "grad_norm": 0.412949800491333,
+      "learning_rate": 2.8429795158286778e-05,
+      "loss": 0.1024,
+      "step": 13172
+    },
+    {
+      "epoch": 36.79608938547486,
+      "grad_norm": 0.7454365491867065,
+      "learning_rate": 2.8429422718808194e-05,
+      "loss": 0.1367,
+      "step": 13173
+    },
+    {
+      "epoch": 36.798882681564244,
+      "grad_norm": 1.1098469495773315,
+      "learning_rate": 2.842905027932961e-05,
+      "loss": 0.1223,
+      "step": 13174
+    },
+    {
+      "epoch": 36.80167597765363,
+      "grad_norm": 0.5613962411880493,
+      "learning_rate": 2.8428677839851028e-05,
+      "loss": 0.1224,
+      "step": 13175
+    },
+    {
+      "epoch": 36.80446927374302,
+      "grad_norm": 0.46567198634147644,
+      "learning_rate": 2.842830540037244e-05,
+      "loss": 0.1028,
+      "step": 13176
+    },
+    {
+      "epoch": 36.8072625698324,
+      "grad_norm": 0.5464855432510376,
+      "learning_rate": 2.8427932960893854e-05,
+      "loss": 0.0952,
+      "step": 13177
+    },
+    {
+      "epoch": 36.81005586592179,
+      "grad_norm": 0.5916465520858765,
+      "learning_rate": 2.842756052141527e-05,
+      "loss": 0.0911,
+      "step": 13178
+    },
+    {
+      "epoch": 36.812849162011176,
+      "grad_norm": 0.6826441884040833,
+      "learning_rate": 2.8427188081936687e-05,
+      "loss": 0.1084,
+      "step": 13179
+    },
+    {
+      "epoch": 36.815642458100555,
+      "grad_norm": 0.7281197309494019,
+      "learning_rate": 2.84268156424581e-05,
+      "loss": 0.1074,
+      "step": 13180
+    },
+    {
+      "epoch": 36.81843575418994,
+      "grad_norm": 0.4663384258747101,
+      "learning_rate": 2.8426443202979517e-05,
+      "loss": 0.0977,
+      "step": 13181
+    },
+    {
+      "epoch": 36.82122905027933,
+      "grad_norm": 0.8397631645202637,
+      "learning_rate": 2.8426070763500933e-05,
+      "loss": 0.0837,
+      "step": 13182
+    },
+    {
+      "epoch": 36.824022346368714,
+      "grad_norm": 0.8874335885047913,
+      "learning_rate": 2.8425698324022346e-05,
+      "loss": 0.0988,
+      "step": 13183
+    },
+    {
+      "epoch": 36.8268156424581,
+      "grad_norm": 0.44335997104644775,
+      "learning_rate": 2.8425325884543763e-05,
+      "loss": 0.0917,
+      "step": 13184
+    },
+    {
+      "epoch": 36.82960893854749,
+      "grad_norm": 1.6380071640014648,
+      "learning_rate": 2.8424953445065176e-05,
+      "loss": 0.1009,
+      "step": 13185
+    },
+    {
+      "epoch": 36.832402234636874,
+      "grad_norm": 0.7539939284324646,
+      "learning_rate": 2.8424581005586593e-05,
+      "loss": 0.0974,
+      "step": 13186
+    },
+    {
+      "epoch": 36.83519553072626,
+      "grad_norm": 1.3203480243682861,
+      "learning_rate": 2.842420856610801e-05,
+      "loss": 0.109,
+      "step": 13187
+    },
+    {
+      "epoch": 36.83798882681564,
+      "grad_norm": 1.7456378936767578,
+      "learning_rate": 2.8423836126629426e-05,
+      "loss": 0.1004,
+      "step": 13188
+    },
+    {
+      "epoch": 36.840782122905026,
+      "grad_norm": 0.521722674369812,
+      "learning_rate": 2.842346368715084e-05,
+      "loss": 0.1203,
+      "step": 13189
+    },
+    {
+      "epoch": 36.84357541899441,
+      "grad_norm": 0.7139871120452881,
+      "learning_rate": 2.8423091247672252e-05,
+      "loss": 0.1281,
+      "step": 13190
+    },
+    {
+      "epoch": 36.8463687150838,
+      "grad_norm": 0.6199519038200378,
+      "learning_rate": 2.842271880819367e-05,
+      "loss": 0.1215,
+      "step": 13191
+    },
+    {
+      "epoch": 36.849162011173185,
+      "grad_norm": 0.37283897399902344,
+      "learning_rate": 2.8422346368715085e-05,
+      "loss": 0.1176,
+      "step": 13192
+    },
+    {
+      "epoch": 36.85195530726257,
+      "grad_norm": 0.3728958070278168,
+      "learning_rate": 2.84219739292365e-05,
+      "loss": 0.0929,
+      "step": 13193
+    },
+    {
+      "epoch": 36.85474860335196,
+      "grad_norm": 0.4494488537311554,
+      "learning_rate": 2.8421601489757915e-05,
+      "loss": 0.1108,
+      "step": 13194
+    },
+    {
+      "epoch": 36.857541899441344,
+      "grad_norm": 0.48986729979515076,
+      "learning_rate": 2.842122905027933e-05,
+      "loss": 0.1142,
+      "step": 13195
+    },
+    {
+      "epoch": 36.86033519553072,
+      "grad_norm": 0.43761172890663147,
+      "learning_rate": 2.8420856610800744e-05,
+      "loss": 0.1285,
+      "step": 13196
+    },
+    {
+      "epoch": 36.86312849162011,
+      "grad_norm": 0.674289345741272,
+      "learning_rate": 2.842048417132216e-05,
+      "loss": 0.1029,
+      "step": 13197
+    },
+    {
+      "epoch": 36.865921787709496,
+      "grad_norm": 0.5453643798828125,
+      "learning_rate": 2.8420111731843574e-05,
+      "loss": 0.1072,
+      "step": 13198
+    },
+    {
+      "epoch": 36.86871508379888,
+      "grad_norm": 0.4174542725086212,
+      "learning_rate": 2.841973929236499e-05,
+      "loss": 0.1048,
+      "step": 13199
+    },
+    {
+      "epoch": 36.87150837988827,
+      "grad_norm": 0.39227601885795593,
+      "learning_rate": 2.8419366852886407e-05,
+      "loss": 0.1043,
+      "step": 13200
+    },
+    {
+      "epoch": 36.874301675977655,
+      "grad_norm": 0.4821287989616394,
+      "learning_rate": 2.8418994413407824e-05,
+      "loss": 0.1172,
+      "step": 13201
+    },
+    {
+      "epoch": 36.87709497206704,
+      "grad_norm": 0.3642721474170685,
+      "learning_rate": 2.841862197392924e-05,
+      "loss": 0.0936,
+      "step": 13202
+    },
+    {
+      "epoch": 36.87988826815643,
+      "grad_norm": 0.5393512845039368,
+      "learning_rate": 2.841824953445065e-05,
+      "loss": 0.0887,
+      "step": 13203
+    },
+    {
+      "epoch": 36.88268156424581,
+      "grad_norm": 1.2354297637939453,
+      "learning_rate": 2.8417877094972066e-05,
+      "loss": 0.1027,
+      "step": 13204
+    },
+    {
+      "epoch": 36.885474860335194,
+      "grad_norm": 0.46873167157173157,
+      "learning_rate": 2.8417504655493483e-05,
+      "loss": 0.0888,
+      "step": 13205
+    },
+    {
+      "epoch": 36.88826815642458,
+      "grad_norm": 0.5780599117279053,
+      "learning_rate": 2.84171322160149e-05,
+      "loss": 0.1048,
+      "step": 13206
+    },
+    {
+      "epoch": 36.891061452513966,
+      "grad_norm": 0.7402818202972412,
+      "learning_rate": 2.8416759776536313e-05,
+      "loss": 0.1128,
+      "step": 13207
+    },
+    {
+      "epoch": 36.89385474860335,
+      "grad_norm": 0.5356085300445557,
+      "learning_rate": 2.841638733705773e-05,
+      "loss": 0.0932,
+      "step": 13208
+    },
+    {
+      "epoch": 36.89664804469274,
+      "grad_norm": 0.8292642831802368,
+      "learning_rate": 2.8416014897579146e-05,
+      "loss": 0.0829,
+      "step": 13209
+    },
+    {
+      "epoch": 36.899441340782126,
+      "grad_norm": 1.1126779317855835,
+      "learning_rate": 2.841564245810056e-05,
+      "loss": 0.097,
+      "step": 13210
+    },
+    {
+      "epoch": 36.90223463687151,
+      "grad_norm": 0.5330525636672974,
+      "learning_rate": 2.8415270018621975e-05,
+      "loss": 0.0742,
+      "step": 13211
+    },
+    {
+      "epoch": 36.90502793296089,
+      "grad_norm": 0.6607159376144409,
+      "learning_rate": 2.841489757914339e-05,
+      "loss": 0.0895,
+      "step": 13212
+    },
+    {
+      "epoch": 36.90782122905028,
+      "grad_norm": 1.2516320943832397,
+      "learning_rate": 2.8414525139664805e-05,
+      "loss": 0.1351,
+      "step": 13213
+    },
+    {
+      "epoch": 36.910614525139664,
+      "grad_norm": 0.37802550196647644,
+      "learning_rate": 2.841415270018622e-05,
+      "loss": 0.1214,
+      "step": 13214
+    },
+    {
+      "epoch": 36.91340782122905,
+      "grad_norm": 0.4019375443458557,
+      "learning_rate": 2.8413780260707638e-05,
+      "loss": 0.1308,
+      "step": 13215
+    },
+    {
+      "epoch": 36.91620111731844,
+      "grad_norm": 0.4334688186645508,
+      "learning_rate": 2.8413407821229048e-05,
+      "loss": 0.1254,
+      "step": 13216
+    },
+    {
+      "epoch": 36.91899441340782,
+      "grad_norm": 0.4634140133857727,
+      "learning_rate": 2.8413035381750464e-05,
+      "loss": 0.1257,
+      "step": 13217
+    },
+    {
+      "epoch": 36.92178770949721,
+      "grad_norm": 0.709736168384552,
+      "learning_rate": 2.841266294227188e-05,
+      "loss": 0.1236,
+      "step": 13218
+    },
+    {
+      "epoch": 36.92458100558659,
+      "grad_norm": 3.1148626804351807,
+      "learning_rate": 2.8412290502793298e-05,
+      "loss": 0.1265,
+      "step": 13219
+    },
+    {
+      "epoch": 36.927374301675975,
+      "grad_norm": 0.3892812430858612,
+      "learning_rate": 2.8411918063314714e-05,
+      "loss": 0.1122,
+      "step": 13220
+    },
+    {
+      "epoch": 36.93016759776536,
+      "grad_norm": 0.499705970287323,
+      "learning_rate": 2.8411545623836127e-05,
+      "loss": 0.0959,
+      "step": 13221
+    },
+    {
+      "epoch": 36.93296089385475,
+      "grad_norm": 0.6239756941795349,
+      "learning_rate": 2.8411173184357544e-05,
+      "loss": 0.1202,
+      "step": 13222
+    },
+    {
+      "epoch": 36.935754189944134,
+      "grad_norm": 0.6576144695281982,
+      "learning_rate": 2.8410800744878957e-05,
+      "loss": 0.0878,
+      "step": 13223
+    },
+    {
+      "epoch": 36.93854748603352,
+      "grad_norm": 0.7083747982978821,
+      "learning_rate": 2.8410428305400373e-05,
+      "loss": 0.0938,
+      "step": 13224
+    },
+    {
+      "epoch": 36.94134078212291,
+      "grad_norm": 1.140073299407959,
+      "learning_rate": 2.8410055865921787e-05,
+      "loss": 0.103,
+      "step": 13225
+    },
+    {
+      "epoch": 36.944134078212294,
+      "grad_norm": 0.5083725452423096,
+      "learning_rate": 2.8409683426443203e-05,
+      "loss": 0.0934,
+      "step": 13226
+    },
+    {
+      "epoch": 36.94692737430167,
+      "grad_norm": 0.7583737373352051,
+      "learning_rate": 2.840931098696462e-05,
+      "loss": 0.0941,
+      "step": 13227
+    },
+    {
+      "epoch": 36.94972067039106,
+      "grad_norm": 0.5554139018058777,
+      "learning_rate": 2.8408938547486036e-05,
+      "loss": 0.084,
+      "step": 13228
+    },
+    {
+      "epoch": 36.952513966480446,
+      "grad_norm": 0.7402909994125366,
+      "learning_rate": 2.8408566108007453e-05,
+      "loss": 0.0973,
+      "step": 13229
+    },
+    {
+      "epoch": 36.95530726256983,
+      "grad_norm": 0.7091984152793884,
+      "learning_rate": 2.8408193668528862e-05,
+      "loss": 0.0993,
+      "step": 13230
+    },
+    {
+      "epoch": 36.95810055865922,
+      "grad_norm": 0.7565969228744507,
+      "learning_rate": 2.840782122905028e-05,
+      "loss": 0.1019,
+      "step": 13231
+    },
+    {
+      "epoch": 36.960893854748605,
+      "grad_norm": 0.49960198998451233,
+      "learning_rate": 2.8407448789571696e-05,
+      "loss": 0.0865,
+      "step": 13232
+    },
+    {
+      "epoch": 36.96368715083799,
+      "grad_norm": 0.7727753520011902,
+      "learning_rate": 2.8407076350093112e-05,
+      "loss": 0.0817,
+      "step": 13233
+    },
+    {
+      "epoch": 36.96648044692738,
+      "grad_norm": 0.8217834234237671,
+      "learning_rate": 2.8406703910614525e-05,
+      "loss": 0.0905,
+      "step": 13234
+    },
+    {
+      "epoch": 36.96927374301676,
+      "grad_norm": 0.6883734464645386,
+      "learning_rate": 2.8406331471135942e-05,
+      "loss": 0.0864,
+      "step": 13235
+    },
+    {
+      "epoch": 36.97206703910614,
+      "grad_norm": 0.5650118589401245,
+      "learning_rate": 2.8405959031657355e-05,
+      "loss": 0.0866,
+      "step": 13236
+    },
+    {
+      "epoch": 36.97486033519553,
+      "grad_norm": 1.497057318687439,
+      "learning_rate": 2.840558659217877e-05,
+      "loss": 0.0874,
+      "step": 13237
+    },
+    {
+      "epoch": 36.977653631284916,
+      "grad_norm": 0.9132478833198547,
+      "learning_rate": 2.8405214152700188e-05,
+      "loss": 0.1133,
+      "step": 13238
+    },
+    {
+      "epoch": 36.9804469273743,
+      "grad_norm": 1.1537809371948242,
+      "learning_rate": 2.84048417132216e-05,
+      "loss": 0.133,
+      "step": 13239
+    },
+    {
+      "epoch": 36.98324022346369,
+      "grad_norm": 0.7350137829780579,
+      "learning_rate": 2.8404469273743018e-05,
+      "loss": 0.1315,
+      "step": 13240
+    },
+    {
+      "epoch": 36.986033519553075,
+      "grad_norm": 0.9472057223320007,
+      "learning_rate": 2.8404096834264434e-05,
+      "loss": 0.1178,
+      "step": 13241
+    },
+    {
+      "epoch": 36.98882681564246,
+      "grad_norm": 0.42533066868782043,
+      "learning_rate": 2.840372439478585e-05,
+      "loss": 0.0875,
+      "step": 13242
+    },
+    {
+      "epoch": 36.99162011173184,
+      "grad_norm": 0.6571608185768127,
+      "learning_rate": 2.840335195530726e-05,
+      "loss": 0.1159,
+      "step": 13243
+    },
+    {
+      "epoch": 36.99441340782123,
+      "grad_norm": 1.0682183504104614,
+      "learning_rate": 2.8402979515828677e-05,
+      "loss": 0.1039,
+      "step": 13244
+    },
+    {
+      "epoch": 36.997206703910614,
+      "grad_norm": 1.3828575611114502,
+      "learning_rate": 2.8402607076350094e-05,
+      "loss": 0.0998,
+      "step": 13245
+    },
+    {
+      "epoch": 37.0,
+      "grad_norm": 0.8961092233657837,
+      "learning_rate": 2.840223463687151e-05,
+      "loss": 0.0814,
+      "step": 13246
+    },
+    {
+      "epoch": 37.002793296089386,
+      "grad_norm": 0.74761563539505,
+      "learning_rate": 2.8401862197392923e-05,
+      "loss": 0.1385,
+      "step": 13247
+    },
+    {
+      "epoch": 37.00558659217877,
+      "grad_norm": 0.37932631373405457,
+      "learning_rate": 2.840148975791434e-05,
+      "loss": 0.1189,
+      "step": 13248
+    },
+    {
+      "epoch": 37.00837988826816,
+      "grad_norm": 0.43780866265296936,
+      "learning_rate": 2.8401117318435756e-05,
+      "loss": 0.1135,
+      "step": 13249
+    },
+    {
+      "epoch": 37.01117318435754,
+      "grad_norm": 0.3957724869251251,
+      "learning_rate": 2.840074487895717e-05,
+      "loss": 0.1077,
+      "step": 13250
+    },
+    {
+      "epoch": 37.013966480446925,
+      "grad_norm": 0.9262292981147766,
+      "learning_rate": 2.8400372439478586e-05,
+      "loss": 0.1378,
+      "step": 13251
+    },
+    {
+      "epoch": 37.01675977653631,
+      "grad_norm": 0.46259424090385437,
+      "learning_rate": 2.84e-05,
+      "loss": 0.1103,
+      "step": 13252
+    },
+    {
+      "epoch": 37.0195530726257,
+      "grad_norm": 0.5878164768218994,
+      "learning_rate": 2.8399627560521416e-05,
+      "loss": 0.1142,
+      "step": 13253
+    },
+    {
+      "epoch": 37.022346368715084,
+      "grad_norm": 0.6430202722549438,
+      "learning_rate": 2.8399255121042832e-05,
+      "loss": 0.1042,
+      "step": 13254
+    },
+    {
+      "epoch": 37.02513966480447,
+      "grad_norm": 0.4991042912006378,
+      "learning_rate": 2.839888268156425e-05,
+      "loss": 0.0864,
+      "step": 13255
+    },
+    {
+      "epoch": 37.02793296089386,
+      "grad_norm": 0.45447537302970886,
+      "learning_rate": 2.839851024208566e-05,
+      "loss": 0.123,
+      "step": 13256
+    },
+    {
+      "epoch": 37.03072625698324,
+      "grad_norm": 0.37202754616737366,
+      "learning_rate": 2.8398137802607075e-05,
+      "loss": 0.0912,
+      "step": 13257
+    },
+    {
+      "epoch": 37.03351955307262,
+      "grad_norm": 0.5839473605155945,
+      "learning_rate": 2.839776536312849e-05,
+      "loss": 0.0897,
+      "step": 13258
+    },
+    {
+      "epoch": 37.03631284916201,
+      "grad_norm": 0.3872879445552826,
+      "learning_rate": 2.8397392923649908e-05,
+      "loss": 0.0834,
+      "step": 13259
+    },
+    {
+      "epoch": 37.039106145251395,
+      "grad_norm": 0.910811185836792,
+      "learning_rate": 2.8397020484171325e-05,
+      "loss": 0.0976,
+      "step": 13260
+    },
+    {
+      "epoch": 37.04189944134078,
+      "grad_norm": 0.5208591222763062,
+      "learning_rate": 2.8396648044692738e-05,
+      "loss": 0.0856,
+      "step": 13261
+    },
+    {
+      "epoch": 37.04469273743017,
+      "grad_norm": 0.35912981629371643,
+      "learning_rate": 2.8396275605214154e-05,
+      "loss": 0.0799,
+      "step": 13262
+    },
+    {
+      "epoch": 37.047486033519554,
+      "grad_norm": 0.6543576717376709,
+      "learning_rate": 2.8395903165735567e-05,
+      "loss": 0.0988,
+      "step": 13263
+    },
+    {
+      "epoch": 37.05027932960894,
+      "grad_norm": 1.3767013549804688,
+      "learning_rate": 2.8395530726256984e-05,
+      "loss": 0.0887,
+      "step": 13264
+    },
+    {
+      "epoch": 37.05307262569833,
+      "grad_norm": 0.424280047416687,
+      "learning_rate": 2.8395158286778397e-05,
+      "loss": 0.0784,
+      "step": 13265
+    },
+    {
+      "epoch": 37.055865921787706,
+      "grad_norm": 0.49617505073547363,
+      "learning_rate": 2.8394785847299814e-05,
+      "loss": 0.0757,
+      "step": 13266
+    },
+    {
+      "epoch": 37.05865921787709,
+      "grad_norm": 0.46867772936820984,
+      "learning_rate": 2.839441340782123e-05,
+      "loss": 0.0662,
+      "step": 13267
+    },
+    {
+      "epoch": 37.06145251396648,
+      "grad_norm": 0.684806764125824,
+      "learning_rate": 2.8394040968342647e-05,
+      "loss": 0.0799,
+      "step": 13268
+    },
+    {
+      "epoch": 37.064245810055866,
+      "grad_norm": 0.6075620651245117,
+      "learning_rate": 2.8393668528864063e-05,
+      "loss": 0.0686,
+      "step": 13269
+    },
+    {
+      "epoch": 37.06703910614525,
+      "grad_norm": 1.639358401298523,
+      "learning_rate": 2.8393296089385473e-05,
+      "loss": 0.0766,
+      "step": 13270
+    },
+    {
+      "epoch": 37.06983240223464,
+      "grad_norm": 0.9062066674232483,
+      "learning_rate": 2.839292364990689e-05,
+      "loss": 0.0985,
+      "step": 13271
+    },
+    {
+      "epoch": 37.072625698324025,
+      "grad_norm": 0.5305954217910767,
+      "learning_rate": 2.8392551210428306e-05,
+      "loss": 0.143,
+      "step": 13272
+    },
+    {
+      "epoch": 37.07541899441341,
+      "grad_norm": 0.38107144832611084,
+      "learning_rate": 2.8392178770949723e-05,
+      "loss": 0.1104,
+      "step": 13273
+    },
+    {
+      "epoch": 37.07821229050279,
+      "grad_norm": 0.5925142168998718,
+      "learning_rate": 2.8391806331471136e-05,
+      "loss": 0.12,
+      "step": 13274
+    },
+    {
+      "epoch": 37.08100558659218,
+      "grad_norm": 0.45266959071159363,
+      "learning_rate": 2.8391433891992552e-05,
+      "loss": 0.1267,
+      "step": 13275
+    },
+    {
+      "epoch": 37.08379888268156,
+      "grad_norm": 0.48352015018463135,
+      "learning_rate": 2.8391061452513965e-05,
+      "loss": 0.0999,
+      "step": 13276
+    },
+    {
+      "epoch": 37.08659217877095,
+      "grad_norm": 0.388950377702713,
+      "learning_rate": 2.8390689013035382e-05,
+      "loss": 0.1228,
+      "step": 13277
+    },
+    {
+      "epoch": 37.089385474860336,
+      "grad_norm": 0.5266644358634949,
+      "learning_rate": 2.83903165735568e-05,
+      "loss": 0.1099,
+      "step": 13278
+    },
+    {
+      "epoch": 37.09217877094972,
+      "grad_norm": 0.5096754431724548,
+      "learning_rate": 2.8389944134078212e-05,
+      "loss": 0.1007,
+      "step": 13279
+    },
+    {
+      "epoch": 37.09497206703911,
+      "grad_norm": 0.6119211912155151,
+      "learning_rate": 2.8389571694599628e-05,
+      "loss": 0.1055,
+      "step": 13280
+    },
+    {
+      "epoch": 37.097765363128495,
+      "grad_norm": 0.44630804657936096,
+      "learning_rate": 2.8389199255121045e-05,
+      "loss": 0.1079,
+      "step": 13281
+    },
+    {
+      "epoch": 37.100558659217874,
+      "grad_norm": 0.4275634288787842,
+      "learning_rate": 2.838882681564246e-05,
+      "loss": 0.1094,
+      "step": 13282
+    },
+    {
+      "epoch": 37.10335195530726,
+      "grad_norm": 0.4561696946620941,
+      "learning_rate": 2.838845437616387e-05,
+      "loss": 0.0932,
+      "step": 13283
+    },
+    {
+      "epoch": 37.10614525139665,
+      "grad_norm": 0.5482856631278992,
+      "learning_rate": 2.8388081936685288e-05,
+      "loss": 0.0911,
+      "step": 13284
+    },
+    {
+      "epoch": 37.108938547486034,
+      "grad_norm": 0.9299963116645813,
+      "learning_rate": 2.8387709497206704e-05,
+      "loss": 0.0902,
+      "step": 13285
+    },
+    {
+      "epoch": 37.11173184357542,
+      "grad_norm": 0.4009720981121063,
+      "learning_rate": 2.838733705772812e-05,
+      "loss": 0.0824,
+      "step": 13286
+    },
+    {
+      "epoch": 37.114525139664806,
+      "grad_norm": 0.9100480079650879,
+      "learning_rate": 2.8386964618249537e-05,
+      "loss": 0.0713,
+      "step": 13287
+    },
+    {
+      "epoch": 37.11731843575419,
+      "grad_norm": 0.6844253540039062,
+      "learning_rate": 2.838659217877095e-05,
+      "loss": 0.0746,
+      "step": 13288
+    },
+    {
+      "epoch": 37.12011173184357,
+      "grad_norm": 0.5541043877601624,
+      "learning_rate": 2.8386219739292367e-05,
+      "loss": 0.075,
+      "step": 13289
+    },
+    {
+      "epoch": 37.12290502793296,
+      "grad_norm": 2.5420081615448,
+      "learning_rate": 2.838584729981378e-05,
+      "loss": 0.0753,
+      "step": 13290
+    },
+    {
+      "epoch": 37.125698324022345,
+      "grad_norm": 0.4890208840370178,
+      "learning_rate": 2.8385474860335197e-05,
+      "loss": 0.074,
+      "step": 13291
+    },
+    {
+      "epoch": 37.12849162011173,
+      "grad_norm": 0.865363359451294,
+      "learning_rate": 2.838510242085661e-05,
+      "loss": 0.082,
+      "step": 13292
+    },
+    {
+      "epoch": 37.13128491620112,
+      "grad_norm": 1.2129333019256592,
+      "learning_rate": 2.8384729981378026e-05,
+      "loss": 0.1015,
+      "step": 13293
+    },
+    {
+      "epoch": 37.134078212290504,
+      "grad_norm": 1.9394524097442627,
+      "learning_rate": 2.8384357541899443e-05,
+      "loss": 0.0688,
+      "step": 13294
+    },
+    {
+      "epoch": 37.13687150837989,
+      "grad_norm": 1.8177286386489868,
+      "learning_rate": 2.838398510242086e-05,
+      "loss": 0.0943,
+      "step": 13295
+    },
+    {
+      "epoch": 37.13966480446928,
+      "grad_norm": 0.7867361307144165,
+      "learning_rate": 2.8383612662942272e-05,
+      "loss": 0.1062,
+      "step": 13296
+    },
+    {
+      "epoch": 37.142458100558656,
+      "grad_norm": 0.5121132731437683,
+      "learning_rate": 2.8383240223463686e-05,
+      "loss": 0.1339,
+      "step": 13297
+    },
+    {
+      "epoch": 37.14525139664804,
+      "grad_norm": 0.46477118134498596,
+      "learning_rate": 2.8382867783985102e-05,
+      "loss": 0.1315,
+      "step": 13298
+    },
+    {
+      "epoch": 37.14804469273743,
+      "grad_norm": 0.4613207280635834,
+      "learning_rate": 2.838249534450652e-05,
+      "loss": 0.1268,
+      "step": 13299
+    },
+    {
+      "epoch": 37.150837988826815,
+      "grad_norm": 1.3940175771713257,
+      "learning_rate": 2.8382122905027935e-05,
+      "loss": 0.1207,
+      "step": 13300
+    },
+    {
+      "epoch": 37.1536312849162,
+      "grad_norm": 0.4325193762779236,
+      "learning_rate": 2.838175046554935e-05,
+      "loss": 0.0965,
+      "step": 13301
+    },
+    {
+      "epoch": 37.15642458100559,
+      "grad_norm": 0.415004163980484,
+      "learning_rate": 2.8381378026070765e-05,
+      "loss": 0.1054,
+      "step": 13302
+    },
+    {
+      "epoch": 37.159217877094974,
+      "grad_norm": 0.3888610005378723,
+      "learning_rate": 2.8381005586592178e-05,
+      "loss": 0.1094,
+      "step": 13303
+    },
+    {
+      "epoch": 37.16201117318436,
+      "grad_norm": 0.5196507573127747,
+      "learning_rate": 2.8380633147113595e-05,
+      "loss": 0.112,
+      "step": 13304
+    },
+    {
+      "epoch": 37.16480446927374,
+      "grad_norm": 0.49852314591407776,
+      "learning_rate": 2.838026070763501e-05,
+      "loss": 0.0795,
+      "step": 13305
+    },
+    {
+      "epoch": 37.167597765363126,
+      "grad_norm": 0.7088412046432495,
+      "learning_rate": 2.8379888268156424e-05,
+      "loss": 0.0973,
+      "step": 13306
+    },
+    {
+      "epoch": 37.17039106145251,
+      "grad_norm": 0.3519684076309204,
+      "learning_rate": 2.837951582867784e-05,
+      "loss": 0.0835,
+      "step": 13307
+    },
+    {
+      "epoch": 37.1731843575419,
+      "grad_norm": 0.6783422827720642,
+      "learning_rate": 2.8379143389199257e-05,
+      "loss": 0.0806,
+      "step": 13308
+    },
+    {
+      "epoch": 37.175977653631286,
+      "grad_norm": 0.7427796125411987,
+      "learning_rate": 2.8378770949720674e-05,
+      "loss": 0.0869,
+      "step": 13309
+    },
+    {
+      "epoch": 37.17877094972067,
+      "grad_norm": 0.6161222457885742,
+      "learning_rate": 2.8378398510242084e-05,
+      "loss": 0.0739,
+      "step": 13310
+    },
+    {
+      "epoch": 37.18156424581006,
+      "grad_norm": 0.5002378225326538,
+      "learning_rate": 2.83780260707635e-05,
+      "loss": 0.0853,
+      "step": 13311
+    },
+    {
+      "epoch": 37.184357541899445,
+      "grad_norm": 0.4288215637207031,
+      "learning_rate": 2.8377653631284917e-05,
+      "loss": 0.0827,
+      "step": 13312
+    },
+    {
+      "epoch": 37.187150837988824,
+      "grad_norm": 0.5347750782966614,
+      "learning_rate": 2.8377281191806333e-05,
+      "loss": 0.0902,
+      "step": 13313
+    },
+    {
+      "epoch": 37.18994413407821,
+      "grad_norm": 1.5278944969177246,
+      "learning_rate": 2.837690875232775e-05,
+      "loss": 0.0627,
+      "step": 13314
+    },
+    {
+      "epoch": 37.1927374301676,
+      "grad_norm": 0.489658385515213,
+      "learning_rate": 2.8376536312849163e-05,
+      "loss": 0.0789,
+      "step": 13315
+    },
+    {
+      "epoch": 37.19553072625698,
+      "grad_norm": 0.5824589133262634,
+      "learning_rate": 2.8376163873370576e-05,
+      "loss": 0.0842,
+      "step": 13316
+    },
+    {
+      "epoch": 37.19832402234637,
+      "grad_norm": 0.7069962620735168,
+      "learning_rate": 2.8375791433891993e-05,
+      "loss": 0.0673,
+      "step": 13317
+    },
+    {
+      "epoch": 37.201117318435756,
+      "grad_norm": 0.6756222248077393,
+      "learning_rate": 2.837541899441341e-05,
+      "loss": 0.0781,
+      "step": 13318
+    },
+    {
+      "epoch": 37.20391061452514,
+      "grad_norm": 0.5991140007972717,
+      "learning_rate": 2.8375046554934822e-05,
+      "loss": 0.0648,
+      "step": 13319
+    },
+    {
+      "epoch": 37.20670391061452,
+      "grad_norm": 1.3049763441085815,
+      "learning_rate": 2.837467411545624e-05,
+      "loss": 0.0709,
+      "step": 13320
+    },
+    {
+      "epoch": 37.20949720670391,
+      "grad_norm": 1.0473449230194092,
+      "learning_rate": 2.8374301675977655e-05,
+      "loss": 0.0869,
+      "step": 13321
+    },
+    {
+      "epoch": 37.212290502793294,
+      "grad_norm": 0.6897624135017395,
+      "learning_rate": 2.8373929236499072e-05,
+      "loss": 0.1259,
+      "step": 13322
+    },
+    {
+      "epoch": 37.21508379888268,
+      "grad_norm": 0.5800973176956177,
+      "learning_rate": 2.8373556797020485e-05,
+      "loss": 0.1242,
+      "step": 13323
+    },
+    {
+      "epoch": 37.21787709497207,
+      "grad_norm": 0.3458525538444519,
+      "learning_rate": 2.8373184357541898e-05,
+      "loss": 0.1159,
+      "step": 13324
+    },
+    {
+      "epoch": 37.220670391061454,
+      "grad_norm": 0.36441630125045776,
+      "learning_rate": 2.8372811918063315e-05,
+      "loss": 0.1088,
+      "step": 13325
+    },
+    {
+      "epoch": 37.22346368715084,
+      "grad_norm": 0.48520901799201965,
+      "learning_rate": 2.837243947858473e-05,
+      "loss": 0.111,
+      "step": 13326
+    },
+    {
+      "epoch": 37.226256983240226,
+      "grad_norm": 0.5218999981880188,
+      "learning_rate": 2.8372067039106148e-05,
+      "loss": 0.0923,
+      "step": 13327
+    },
+    {
+      "epoch": 37.229050279329606,
+      "grad_norm": 0.5178213715553284,
+      "learning_rate": 2.837169459962756e-05,
+      "loss": 0.1068,
+      "step": 13328
+    },
+    {
+      "epoch": 37.23184357541899,
+      "grad_norm": 0.4273589849472046,
+      "learning_rate": 2.8371322160148978e-05,
+      "loss": 0.086,
+      "step": 13329
+    },
+    {
+      "epoch": 37.23463687150838,
+      "grad_norm": 0.32568174600601196,
+      "learning_rate": 2.837094972067039e-05,
+      "loss": 0.0961,
+      "step": 13330
+    },
+    {
+      "epoch": 37.237430167597765,
+      "grad_norm": 0.43901607394218445,
+      "learning_rate": 2.8370577281191807e-05,
+      "loss": 0.0957,
+      "step": 13331
+    },
+    {
+      "epoch": 37.24022346368715,
+      "grad_norm": 0.4905693233013153,
+      "learning_rate": 2.8370204841713224e-05,
+      "loss": 0.1086,
+      "step": 13332
+    },
+    {
+      "epoch": 37.24301675977654,
+      "grad_norm": 0.6996617317199707,
+      "learning_rate": 2.8369832402234637e-05,
+      "loss": 0.0883,
+      "step": 13333
+    },
+    {
+      "epoch": 37.245810055865924,
+      "grad_norm": 0.424746572971344,
+      "learning_rate": 2.8369459962756053e-05,
+      "loss": 0.1,
+      "step": 13334
+    },
+    {
+      "epoch": 37.24860335195531,
+      "grad_norm": 0.6246015429496765,
+      "learning_rate": 2.836908752327747e-05,
+      "loss": 0.0867,
+      "step": 13335
+    },
+    {
+      "epoch": 37.25139664804469,
+      "grad_norm": 0.7625470161437988,
+      "learning_rate": 2.8368715083798883e-05,
+      "loss": 0.091,
+      "step": 13336
+    },
+    {
+      "epoch": 37.254189944134076,
+      "grad_norm": 0.7978317141532898,
+      "learning_rate": 2.8368342644320296e-05,
+      "loss": 0.0841,
+      "step": 13337
+    },
+    {
+      "epoch": 37.25698324022346,
+      "grad_norm": 0.6554155945777893,
+      "learning_rate": 2.8367970204841713e-05,
+      "loss": 0.092,
+      "step": 13338
+    },
+    {
+      "epoch": 37.25977653631285,
+      "grad_norm": 0.9054871797561646,
+      "learning_rate": 2.836759776536313e-05,
+      "loss": 0.1006,
+      "step": 13339
+    },
+    {
+      "epoch": 37.262569832402235,
+      "grad_norm": 0.7611244916915894,
+      "learning_rate": 2.8367225325884546e-05,
+      "loss": 0.0789,
+      "step": 13340
+    },
+    {
+      "epoch": 37.26536312849162,
+      "grad_norm": 0.748970091342926,
+      "learning_rate": 2.836685288640596e-05,
+      "loss": 0.0634,
+      "step": 13341
+    },
+    {
+      "epoch": 37.26815642458101,
+      "grad_norm": 0.8926631808280945,
+      "learning_rate": 2.8366480446927376e-05,
+      "loss": 0.0833,
+      "step": 13342
+    },
+    {
+      "epoch": 37.270949720670394,
+      "grad_norm": 0.6337838768959045,
+      "learning_rate": 2.836610800744879e-05,
+      "loss": 0.0817,
+      "step": 13343
+    },
+    {
+      "epoch": 37.273743016759774,
+      "grad_norm": 0.4261005222797394,
+      "learning_rate": 2.8365735567970205e-05,
+      "loss": 0.0568,
+      "step": 13344
+    },
+    {
+      "epoch": 37.27653631284916,
+      "grad_norm": 0.7921914458274841,
+      "learning_rate": 2.8365363128491622e-05,
+      "loss": 0.0658,
+      "step": 13345
+    },
+    {
+      "epoch": 37.279329608938546,
+      "grad_norm": 1.3355543613433838,
+      "learning_rate": 2.8364990689013035e-05,
+      "loss": 0.1009,
+      "step": 13346
+    },
+    {
+      "epoch": 37.28212290502793,
+      "grad_norm": 0.6163201332092285,
+      "learning_rate": 2.836461824953445e-05,
+      "loss": 0.1288,
+      "step": 13347
+    },
+    {
+      "epoch": 37.28491620111732,
+      "grad_norm": 0.60809725522995,
+      "learning_rate": 2.8364245810055868e-05,
+      "loss": 0.1338,
+      "step": 13348
+    },
+    {
+      "epoch": 37.287709497206706,
+      "grad_norm": 0.4188610315322876,
+      "learning_rate": 2.8363873370577284e-05,
+      "loss": 0.1319,
+      "step": 13349
+    },
+    {
+      "epoch": 37.29050279329609,
+      "grad_norm": 0.6678484678268433,
+      "learning_rate": 2.8363500931098694e-05,
+      "loss": 0.1344,
+      "step": 13350
+    },
+    {
+      "epoch": 37.29329608938548,
+      "grad_norm": 0.3436428904533386,
+      "learning_rate": 2.836312849162011e-05,
+      "loss": 0.1105,
+      "step": 13351
+    },
+    {
+      "epoch": 37.29608938547486,
+      "grad_norm": 0.4123501479625702,
+      "learning_rate": 2.8362756052141527e-05,
+      "loss": 0.0983,
+      "step": 13352
+    },
+    {
+      "epoch": 37.298882681564244,
+      "grad_norm": 0.6006595492362976,
+      "learning_rate": 2.8362383612662944e-05,
+      "loss": 0.1112,
+      "step": 13353
+    },
+    {
+      "epoch": 37.30167597765363,
+      "grad_norm": 0.46118563413619995,
+      "learning_rate": 2.836201117318436e-05,
+      "loss": 0.0929,
+      "step": 13354
+    },
+    {
+      "epoch": 37.30446927374302,
+      "grad_norm": 0.3846932053565979,
+      "learning_rate": 2.8361638733705774e-05,
+      "loss": 0.1011,
+      "step": 13355
+    },
+    {
+      "epoch": 37.3072625698324,
+      "grad_norm": 0.4042539596557617,
+      "learning_rate": 2.8361266294227187e-05,
+      "loss": 0.0863,
+      "step": 13356
+    },
+    {
+      "epoch": 37.31005586592179,
+      "grad_norm": 0.6703089475631714,
+      "learning_rate": 2.8360893854748603e-05,
+      "loss": 0.089,
+      "step": 13357
+    },
+    {
+      "epoch": 37.312849162011176,
+      "grad_norm": 0.8210147619247437,
+      "learning_rate": 2.836052141527002e-05,
+      "loss": 0.0928,
+      "step": 13358
+    },
+    {
+      "epoch": 37.315642458100555,
+      "grad_norm": 0.48430782556533813,
+      "learning_rate": 2.8360148975791433e-05,
+      "loss": 0.089,
+      "step": 13359
+    },
+    {
+      "epoch": 37.31843575418994,
+      "grad_norm": 0.5108228921890259,
+      "learning_rate": 2.835977653631285e-05,
+      "loss": 0.1011,
+      "step": 13360
+    },
+    {
+      "epoch": 37.32122905027933,
+      "grad_norm": 0.4103139340877533,
+      "learning_rate": 2.8359404096834266e-05,
+      "loss": 0.0889,
+      "step": 13361
+    },
+    {
+      "epoch": 37.324022346368714,
+      "grad_norm": 0.5112071633338928,
+      "learning_rate": 2.8359031657355683e-05,
+      "loss": 0.0867,
+      "step": 13362
+    },
+    {
+      "epoch": 37.3268156424581,
+      "grad_norm": 0.6700937747955322,
+      "learning_rate": 2.8358659217877096e-05,
+      "loss": 0.081,
+      "step": 13363
+    },
+    {
+      "epoch": 37.32960893854749,
+      "grad_norm": 1.1024699211120605,
+      "learning_rate": 2.835828677839851e-05,
+      "loss": 0.093,
+      "step": 13364
+    },
+    {
+      "epoch": 37.332402234636874,
+      "grad_norm": 0.5729321837425232,
+      "learning_rate": 2.8357914338919925e-05,
+      "loss": 0.072,
+      "step": 13365
+    },
+    {
+      "epoch": 37.33519553072626,
+      "grad_norm": 0.5340880155563354,
+      "learning_rate": 2.8357541899441342e-05,
+      "loss": 0.0751,
+      "step": 13366
+    },
+    {
+      "epoch": 37.33798882681564,
+      "grad_norm": 0.6916103363037109,
+      "learning_rate": 2.835716945996276e-05,
+      "loss": 0.0762,
+      "step": 13367
+    },
+    {
+      "epoch": 37.340782122905026,
+      "grad_norm": 0.693001925945282,
+      "learning_rate": 2.835679702048417e-05,
+      "loss": 0.0707,
+      "step": 13368
+    },
+    {
+      "epoch": 37.34357541899441,
+      "grad_norm": 0.8475944995880127,
+      "learning_rate": 2.8356424581005588e-05,
+      "loss": 0.0819,
+      "step": 13369
+    },
+    {
+      "epoch": 37.3463687150838,
+      "grad_norm": 1.7849116325378418,
+      "learning_rate": 2.8356052141527e-05,
+      "loss": 0.0717,
+      "step": 13370
+    },
+    {
+      "epoch": 37.349162011173185,
+      "grad_norm": 1.116263747215271,
+      "learning_rate": 2.8355679702048418e-05,
+      "loss": 0.1079,
+      "step": 13371
+    },
+    {
+      "epoch": 37.35195530726257,
+      "grad_norm": 0.4863429069519043,
+      "learning_rate": 2.8355307262569834e-05,
+      "loss": 0.1154,
+      "step": 13372
+    },
+    {
+      "epoch": 37.35474860335196,
+      "grad_norm": 0.4517359435558319,
+      "learning_rate": 2.8354934823091247e-05,
+      "loss": 0.0943,
+      "step": 13373
+    },
+    {
+      "epoch": 37.357541899441344,
+      "grad_norm": 0.36375024914741516,
+      "learning_rate": 2.8354562383612664e-05,
+      "loss": 0.1108,
+      "step": 13374
+    },
+    {
+      "epoch": 37.36033519553072,
+      "grad_norm": 0.685482919216156,
+      "learning_rate": 2.835418994413408e-05,
+      "loss": 0.0988,
+      "step": 13375
+    },
+    {
+      "epoch": 37.36312849162011,
+      "grad_norm": 0.4224201738834381,
+      "learning_rate": 2.8353817504655494e-05,
+      "loss": 0.1146,
+      "step": 13376
+    },
+    {
+      "epoch": 37.365921787709496,
+      "grad_norm": 0.5666427612304688,
+      "learning_rate": 2.8353445065176907e-05,
+      "loss": 0.1056,
+      "step": 13377
+    },
+    {
+      "epoch": 37.36871508379888,
+      "grad_norm": 0.4976658225059509,
+      "learning_rate": 2.8353072625698323e-05,
+      "loss": 0.1166,
+      "step": 13378
+    },
+    {
+      "epoch": 37.37150837988827,
+      "grad_norm": 1.321623682975769,
+      "learning_rate": 2.835270018621974e-05,
+      "loss": 0.0955,
+      "step": 13379
+    },
+    {
+      "epoch": 37.374301675977655,
+      "grad_norm": 0.47414878010749817,
+      "learning_rate": 2.8352327746741156e-05,
+      "loss": 0.0955,
+      "step": 13380
+    },
+    {
+      "epoch": 37.37709497206704,
+      "grad_norm": 0.6215054392814636,
+      "learning_rate": 2.8351955307262573e-05,
+      "loss": 0.11,
+      "step": 13381
+    },
+    {
+      "epoch": 37.37988826815643,
+      "grad_norm": 0.4070114493370056,
+      "learning_rate": 2.8351582867783986e-05,
+      "loss": 0.1018,
+      "step": 13382
+    },
+    {
+      "epoch": 37.38268156424581,
+      "grad_norm": 0.38315749168395996,
+      "learning_rate": 2.83512104283054e-05,
+      "loss": 0.0971,
+      "step": 13383
+    },
+    {
+      "epoch": 37.385474860335194,
+      "grad_norm": 0.522167444229126,
+      "learning_rate": 2.8350837988826816e-05,
+      "loss": 0.0904,
+      "step": 13384
+    },
+    {
+      "epoch": 37.38826815642458,
+      "grad_norm": 0.3863961696624756,
+      "learning_rate": 2.8350465549348232e-05,
+      "loss": 0.0726,
+      "step": 13385
+    },
+    {
+      "epoch": 37.391061452513966,
+      "grad_norm": 1.890020489692688,
+      "learning_rate": 2.8350093109869645e-05,
+      "loss": 0.0855,
+      "step": 13386
+    },
+    {
+      "epoch": 37.39385474860335,
+      "grad_norm": 0.49436914920806885,
+      "learning_rate": 2.8349720670391062e-05,
+      "loss": 0.0852,
+      "step": 13387
+    },
+    {
+      "epoch": 37.39664804469274,
+      "grad_norm": 0.6537649035453796,
+      "learning_rate": 2.834934823091248e-05,
+      "loss": 0.068,
+      "step": 13388
+    },
+    {
+      "epoch": 37.399441340782126,
+      "grad_norm": 0.5225786566734314,
+      "learning_rate": 2.8348975791433895e-05,
+      "loss": 0.0778,
+      "step": 13389
+    },
+    {
+      "epoch": 37.40223463687151,
+      "grad_norm": 0.6574114561080933,
+      "learning_rate": 2.8348603351955308e-05,
+      "loss": 0.068,
+      "step": 13390
+    },
+    {
+      "epoch": 37.40502793296089,
+      "grad_norm": 0.6743513941764832,
+      "learning_rate": 2.834823091247672e-05,
+      "loss": 0.0831,
+      "step": 13391
+    },
+    {
+      "epoch": 37.40782122905028,
+      "grad_norm": 0.46237507462501526,
+      "learning_rate": 2.8347858472998138e-05,
+      "loss": 0.0859,
+      "step": 13392
+    },
+    {
+      "epoch": 37.410614525139664,
+      "grad_norm": 0.529401957988739,
+      "learning_rate": 2.8347486033519554e-05,
+      "loss": 0.0712,
+      "step": 13393
+    },
+    {
+      "epoch": 37.41340782122905,
+      "grad_norm": 0.7206148505210876,
+      "learning_rate": 2.834711359404097e-05,
+      "loss": 0.088,
+      "step": 13394
+    },
+    {
+      "epoch": 37.41620111731844,
+      "grad_norm": 0.673270583152771,
+      "learning_rate": 2.8346741154562384e-05,
+      "loss": 0.0656,
+      "step": 13395
+    },
+    {
+      "epoch": 37.41899441340782,
+      "grad_norm": 3.8371360301971436,
+      "learning_rate": 2.8346368715083797e-05,
+      "loss": 0.1028,
+      "step": 13396
+    },
+    {
+      "epoch": 37.42178770949721,
+      "grad_norm": 1.7300827503204346,
+      "learning_rate": 2.8345996275605214e-05,
+      "loss": 0.1199,
+      "step": 13397
+    },
+    {
+      "epoch": 37.42458100558659,
+      "grad_norm": 0.36664679646492004,
+      "learning_rate": 2.834562383612663e-05,
+      "loss": 0.1263,
+      "step": 13398
+    },
+    {
+      "epoch": 37.427374301675975,
+      "grad_norm": 0.3068758547306061,
+      "learning_rate": 2.8345251396648047e-05,
+      "loss": 0.0956,
+      "step": 13399
+    },
+    {
+      "epoch": 37.43016759776536,
+      "grad_norm": 0.503147304058075,
+      "learning_rate": 2.834487895716946e-05,
+      "loss": 0.1094,
+      "step": 13400
+    },
+    {
+      "epoch": 37.43296089385475,
+      "grad_norm": 0.7151728868484497,
+      "learning_rate": 2.8344506517690877e-05,
+      "loss": 0.1028,
+      "step": 13401
+    },
+    {
+      "epoch": 37.435754189944134,
+      "grad_norm": 0.7435734868049622,
+      "learning_rate": 2.8344134078212293e-05,
+      "loss": 0.1255,
+      "step": 13402
+    },
+    {
+      "epoch": 37.43854748603352,
+      "grad_norm": 0.472522109746933,
+      "learning_rate": 2.8343761638733706e-05,
+      "loss": 0.1107,
+      "step": 13403
+    },
+    {
+      "epoch": 37.44134078212291,
+      "grad_norm": 0.37457749247550964,
+      "learning_rate": 2.834338919925512e-05,
+      "loss": 0.0905,
+      "step": 13404
+    },
+    {
+      "epoch": 37.444134078212294,
+      "grad_norm": 0.4437308609485626,
+      "learning_rate": 2.8343016759776536e-05,
+      "loss": 0.0931,
+      "step": 13405
+    },
+    {
+      "epoch": 37.44692737430167,
+      "grad_norm": 0.4241126775741577,
+      "learning_rate": 2.8342644320297952e-05,
+      "loss": 0.0853,
+      "step": 13406
+    },
+    {
+      "epoch": 37.44972067039106,
+      "grad_norm": 0.7503301501274109,
+      "learning_rate": 2.834227188081937e-05,
+      "loss": 0.1058,
+      "step": 13407
+    },
+    {
+      "epoch": 37.452513966480446,
+      "grad_norm": 0.43189507722854614,
+      "learning_rate": 2.8341899441340786e-05,
+      "loss": 0.0872,
+      "step": 13408
+    },
+    {
+      "epoch": 37.45530726256983,
+      "grad_norm": 0.5886490941047668,
+      "learning_rate": 2.83415270018622e-05,
+      "loss": 0.0919,
+      "step": 13409
+    },
+    {
+      "epoch": 37.45810055865922,
+      "grad_norm": 0.6694846153259277,
+      "learning_rate": 2.8341154562383612e-05,
+      "loss": 0.1075,
+      "step": 13410
+    },
+    {
+      "epoch": 37.460893854748605,
+      "grad_norm": 0.5880787968635559,
+      "learning_rate": 2.834078212290503e-05,
+      "loss": 0.0863,
+      "step": 13411
+    },
+    {
+      "epoch": 37.46368715083799,
+      "grad_norm": 0.6287997364997864,
+      "learning_rate": 2.8340409683426445e-05,
+      "loss": 0.0732,
+      "step": 13412
+    },
+    {
+      "epoch": 37.46648044692738,
+      "grad_norm": 0.43379098176956177,
+      "learning_rate": 2.8340037243947858e-05,
+      "loss": 0.0837,
+      "step": 13413
+    },
+    {
+      "epoch": 37.46927374301676,
+      "grad_norm": 0.432059645652771,
+      "learning_rate": 2.8339664804469275e-05,
+      "loss": 0.0781,
+      "step": 13414
+    },
+    {
+      "epoch": 37.47206703910614,
+      "grad_norm": 0.5733720660209656,
+      "learning_rate": 2.833929236499069e-05,
+      "loss": 0.0967,
+      "step": 13415
+    },
+    {
+      "epoch": 37.47486033519553,
+      "grad_norm": 0.5086501240730286,
+      "learning_rate": 2.8338919925512104e-05,
+      "loss": 0.079,
+      "step": 13416
+    },
+    {
+      "epoch": 37.477653631284916,
+      "grad_norm": 1.2448936700820923,
+      "learning_rate": 2.833854748603352e-05,
+      "loss": 0.0589,
+      "step": 13417
+    },
+    {
+      "epoch": 37.4804469273743,
+      "grad_norm": 0.6683579683303833,
+      "learning_rate": 2.8338175046554934e-05,
+      "loss": 0.0611,
+      "step": 13418
+    },
+    {
+      "epoch": 37.48324022346369,
+      "grad_norm": 0.5866195559501648,
+      "learning_rate": 2.833780260707635e-05,
+      "loss": 0.0747,
+      "step": 13419
+    },
+    {
+      "epoch": 37.486033519553075,
+      "grad_norm": 1.6199088096618652,
+      "learning_rate": 2.8337430167597767e-05,
+      "loss": 0.0792,
+      "step": 13420
+    },
+    {
+      "epoch": 37.48882681564246,
+      "grad_norm": 1.4266573190689087,
+      "learning_rate": 2.8337057728119184e-05,
+      "loss": 0.0934,
+      "step": 13421
+    },
+    {
+      "epoch": 37.49162011173184,
+      "grad_norm": 1.9309886693954468,
+      "learning_rate": 2.8336685288640597e-05,
+      "loss": 0.1278,
+      "step": 13422
+    },
+    {
+      "epoch": 37.49441340782123,
+      "grad_norm": 0.704800009727478,
+      "learning_rate": 2.833631284916201e-05,
+      "loss": 0.1105,
+      "step": 13423
+    },
+    {
+      "epoch": 37.497206703910614,
+      "grad_norm": 0.5257163047790527,
+      "learning_rate": 2.8335940409683426e-05,
+      "loss": 0.122,
+      "step": 13424
+    },
+    {
+      "epoch": 37.5,
+      "grad_norm": 0.5658948421478271,
+      "learning_rate": 2.8335567970204843e-05,
+      "loss": 0.102,
+      "step": 13425
+    },
+    {
+      "epoch": 37.502793296089386,
+      "grad_norm": 1.2854669094085693,
+      "learning_rate": 2.833519553072626e-05,
+      "loss": 0.1175,
+      "step": 13426
+    },
+    {
+      "epoch": 37.50558659217877,
+      "grad_norm": 0.4650685787200928,
+      "learning_rate": 2.8334823091247673e-05,
+      "loss": 0.1125,
+      "step": 13427
+    },
+    {
+      "epoch": 37.50837988826816,
+      "grad_norm": 0.41859397292137146,
+      "learning_rate": 2.833445065176909e-05,
+      "loss": 0.1228,
+      "step": 13428
+    },
+    {
+      "epoch": 37.51117318435754,
+      "grad_norm": 0.8538311719894409,
+      "learning_rate": 2.8334078212290506e-05,
+      "loss": 0.0994,
+      "step": 13429
+    },
+    {
+      "epoch": 37.513966480446925,
+      "grad_norm": 0.7266781330108643,
+      "learning_rate": 2.833370577281192e-05,
+      "loss": 0.0919,
+      "step": 13430
+    },
+    {
+      "epoch": 37.51675977653631,
+      "grad_norm": 0.46352332830429077,
+      "learning_rate": 2.8333333333333332e-05,
+      "loss": 0.0963,
+      "step": 13431
+    },
+    {
+      "epoch": 37.5195530726257,
+      "grad_norm": 0.6529415249824524,
+      "learning_rate": 2.833296089385475e-05,
+      "loss": 0.0961,
+      "step": 13432
+    },
+    {
+      "epoch": 37.522346368715084,
+      "grad_norm": 0.618046224117279,
+      "learning_rate": 2.8332588454376165e-05,
+      "loss": 0.0979,
+      "step": 13433
+    },
+    {
+      "epoch": 37.52513966480447,
+      "grad_norm": 0.5989570021629333,
+      "learning_rate": 2.833221601489758e-05,
+      "loss": 0.096,
+      "step": 13434
+    },
+    {
+      "epoch": 37.52793296089386,
+      "grad_norm": 0.5621271729469299,
+      "learning_rate": 2.8331843575418995e-05,
+      "loss": 0.086,
+      "step": 13435
+    },
+    {
+      "epoch": 37.53072625698324,
+      "grad_norm": 0.4390662908554077,
+      "learning_rate": 2.8331471135940408e-05,
+      "loss": 0.0627,
+      "step": 13436
+    },
+    {
+      "epoch": 37.53351955307262,
+      "grad_norm": 0.6179614663124084,
+      "learning_rate": 2.8331098696461824e-05,
+      "loss": 0.0813,
+      "step": 13437
+    },
+    {
+      "epoch": 37.53631284916201,
+      "grad_norm": 0.4908084571361542,
+      "learning_rate": 2.833072625698324e-05,
+      "loss": 0.0697,
+      "step": 13438
+    },
+    {
+      "epoch": 37.539106145251395,
+      "grad_norm": 0.4476589262485504,
+      "learning_rate": 2.8330353817504657e-05,
+      "loss": 0.0832,
+      "step": 13439
+    },
+    {
+      "epoch": 37.54189944134078,
+      "grad_norm": 0.47580498456954956,
+      "learning_rate": 2.832998137802607e-05,
+      "loss": 0.0648,
+      "step": 13440
+    },
+    {
+      "epoch": 37.54469273743017,
+      "grad_norm": 0.5279530882835388,
+      "learning_rate": 2.8329608938547487e-05,
+      "loss": 0.081,
+      "step": 13441
+    },
+    {
+      "epoch": 37.547486033519554,
+      "grad_norm": 0.5034967064857483,
+      "learning_rate": 2.8329236499068904e-05,
+      "loss": 0.0704,
+      "step": 13442
+    },
+    {
+      "epoch": 37.55027932960894,
+      "grad_norm": 0.7681204080581665,
+      "learning_rate": 2.8328864059590317e-05,
+      "loss": 0.0901,
+      "step": 13443
+    },
+    {
+      "epoch": 37.55307262569833,
+      "grad_norm": 1.0770174264907837,
+      "learning_rate": 2.832849162011173e-05,
+      "loss": 0.0817,
+      "step": 13444
+    },
+    {
+      "epoch": 37.555865921787706,
+      "grad_norm": 0.6479112505912781,
+      "learning_rate": 2.8328119180633147e-05,
+      "loss": 0.0739,
+      "step": 13445
+    },
+    {
+      "epoch": 37.55865921787709,
+      "grad_norm": 2.0961380004882812,
+      "learning_rate": 2.8327746741154563e-05,
+      "loss": 0.1462,
+      "step": 13446
+    },
+    {
+      "epoch": 37.56145251396648,
+      "grad_norm": 0.4542543888092041,
+      "learning_rate": 2.832737430167598e-05,
+      "loss": 0.1317,
+      "step": 13447
+    },
+    {
+      "epoch": 37.564245810055866,
+      "grad_norm": 0.42304617166519165,
+      "learning_rate": 2.8327001862197396e-05,
+      "loss": 0.1034,
+      "step": 13448
+    },
+    {
+      "epoch": 37.56703910614525,
+      "grad_norm": 2.2587661743164062,
+      "learning_rate": 2.832662942271881e-05,
+      "loss": 0.1152,
+      "step": 13449
+    },
+    {
+      "epoch": 37.56983240223464,
+      "grad_norm": 0.6085516214370728,
+      "learning_rate": 2.8326256983240222e-05,
+      "loss": 0.1089,
+      "step": 13450
+    },
+    {
+      "epoch": 37.572625698324025,
+      "grad_norm": 0.7562211751937866,
+      "learning_rate": 2.832588454376164e-05,
+      "loss": 0.139,
+      "step": 13451
+    },
+    {
+      "epoch": 37.57541899441341,
+      "grad_norm": 0.7050812244415283,
+      "learning_rate": 2.8325512104283056e-05,
+      "loss": 0.1198,
+      "step": 13452
+    },
+    {
+      "epoch": 37.57821229050279,
+      "grad_norm": 2.613920211791992,
+      "learning_rate": 2.832513966480447e-05,
+      "loss": 0.0951,
+      "step": 13453
+    },
+    {
+      "epoch": 37.58100558659218,
+      "grad_norm": 0.5703511238098145,
+      "learning_rate": 2.8324767225325885e-05,
+      "loss": 0.1243,
+      "step": 13454
+    },
+    {
+      "epoch": 37.58379888268156,
+      "grad_norm": 0.6454934477806091,
+      "learning_rate": 2.8324394785847302e-05,
+      "loss": 0.1023,
+      "step": 13455
+    },
+    {
+      "epoch": 37.58659217877095,
+      "grad_norm": 0.364760160446167,
+      "learning_rate": 2.8324022346368715e-05,
+      "loss": 0.1019,
+      "step": 13456
+    },
+    {
+      "epoch": 37.589385474860336,
+      "grad_norm": 0.5709313154220581,
+      "learning_rate": 2.832364990689013e-05,
+      "loss": 0.098,
+      "step": 13457
+    },
+    {
+      "epoch": 37.59217877094972,
+      "grad_norm": 0.5588530898094177,
+      "learning_rate": 2.8323277467411545e-05,
+      "loss": 0.0932,
+      "step": 13458
+    },
+    {
+      "epoch": 37.59497206703911,
+      "grad_norm": 0.7134435176849365,
+      "learning_rate": 2.832290502793296e-05,
+      "loss": 0.0909,
+      "step": 13459
+    },
+    {
+      "epoch": 37.59776536312849,
+      "grad_norm": 0.5087964534759521,
+      "learning_rate": 2.8322532588454378e-05,
+      "loss": 0.0866,
+      "step": 13460
+    },
+    {
+      "epoch": 37.600558659217874,
+      "grad_norm": 2.167469024658203,
+      "learning_rate": 2.8322160148975794e-05,
+      "loss": 0.0873,
+      "step": 13461
+    },
+    {
+      "epoch": 37.60335195530726,
+      "grad_norm": 0.5431972146034241,
+      "learning_rate": 2.8321787709497207e-05,
+      "loss": 0.0714,
+      "step": 13462
+    },
+    {
+      "epoch": 37.60614525139665,
+      "grad_norm": 0.6218196153640747,
+      "learning_rate": 2.832141527001862e-05,
+      "loss": 0.0896,
+      "step": 13463
+    },
+    {
+      "epoch": 37.608938547486034,
+      "grad_norm": 0.6366216540336609,
+      "learning_rate": 2.8321042830540037e-05,
+      "loss": 0.0856,
+      "step": 13464
+    },
+    {
+      "epoch": 37.61173184357542,
+      "grad_norm": 0.6926112771034241,
+      "learning_rate": 2.8320670391061454e-05,
+      "loss": 0.083,
+      "step": 13465
+    },
+    {
+      "epoch": 37.614525139664806,
+      "grad_norm": 1.5821130275726318,
+      "learning_rate": 2.832029795158287e-05,
+      "loss": 0.0878,
+      "step": 13466
+    },
+    {
+      "epoch": 37.61731843575419,
+      "grad_norm": 0.8636652827262878,
+      "learning_rate": 2.8319925512104283e-05,
+      "loss": 0.0942,
+      "step": 13467
+    },
+    {
+      "epoch": 37.62011173184357,
+      "grad_norm": 0.61641526222229,
+      "learning_rate": 2.83195530726257e-05,
+      "loss": 0.0624,
+      "step": 13468
+    },
+    {
+      "epoch": 37.62290502793296,
+      "grad_norm": 0.6826140880584717,
+      "learning_rate": 2.8319180633147116e-05,
+      "loss": 0.0932,
+      "step": 13469
+    },
+    {
+      "epoch": 37.625698324022345,
+      "grad_norm": 1.7921136617660522,
+      "learning_rate": 2.831880819366853e-05,
+      "loss": 0.0746,
+      "step": 13470
+    },
+    {
+      "epoch": 37.62849162011173,
+      "grad_norm": 1.4691777229309082,
+      "learning_rate": 2.8318435754189943e-05,
+      "loss": 0.0733,
+      "step": 13471
+    },
+    {
+      "epoch": 37.63128491620112,
+      "grad_norm": 2.1894166469573975,
+      "learning_rate": 2.831806331471136e-05,
+      "loss": 0.1993,
+      "step": 13472
+    },
+    {
+      "epoch": 37.634078212290504,
+      "grad_norm": 0.7527551054954529,
+      "learning_rate": 2.8317690875232776e-05,
+      "loss": 0.1275,
+      "step": 13473
+    },
+    {
+      "epoch": 37.63687150837989,
+      "grad_norm": 0.49783098697662354,
+      "learning_rate": 2.8317318435754192e-05,
+      "loss": 0.1285,
+      "step": 13474
+    },
+    {
+      "epoch": 37.63966480446928,
+      "grad_norm": 0.5186768174171448,
+      "learning_rate": 2.831694599627561e-05,
+      "loss": 0.1161,
+      "step": 13475
+    },
+    {
+      "epoch": 37.642458100558656,
+      "grad_norm": 0.4457460641860962,
+      "learning_rate": 2.831657355679702e-05,
+      "loss": 0.1206,
+      "step": 13476
+    },
+    {
+      "epoch": 37.64525139664804,
+      "grad_norm": 0.4691607356071472,
+      "learning_rate": 2.8316201117318435e-05,
+      "loss": 0.1025,
+      "step": 13477
+    },
+    {
+      "epoch": 37.64804469273743,
+      "grad_norm": 0.4640813171863556,
+      "learning_rate": 2.831582867783985e-05,
+      "loss": 0.0891,
+      "step": 13478
+    },
+    {
+      "epoch": 37.650837988826815,
+      "grad_norm": 0.5140835642814636,
+      "learning_rate": 2.8315456238361268e-05,
+      "loss": 0.0927,
+      "step": 13479
+    },
+    {
+      "epoch": 37.6536312849162,
+      "grad_norm": 0.53755122423172,
+      "learning_rate": 2.831508379888268e-05,
+      "loss": 0.1013,
+      "step": 13480
+    },
+    {
+      "epoch": 37.65642458100559,
+      "grad_norm": 0.5389484167098999,
+      "learning_rate": 2.8314711359404098e-05,
+      "loss": 0.0797,
+      "step": 13481
+    },
+    {
+      "epoch": 37.659217877094974,
+      "grad_norm": 0.45718449354171753,
+      "learning_rate": 2.8314338919925514e-05,
+      "loss": 0.0992,
+      "step": 13482
+    },
+    {
+      "epoch": 37.66201117318436,
+      "grad_norm": 0.4195377230644226,
+      "learning_rate": 2.8313966480446927e-05,
+      "loss": 0.0825,
+      "step": 13483
+    },
+    {
+      "epoch": 37.66480446927374,
+      "grad_norm": 0.6377631425857544,
+      "learning_rate": 2.8313594040968344e-05,
+      "loss": 0.088,
+      "step": 13484
+    },
+    {
+      "epoch": 37.667597765363126,
+      "grad_norm": 0.6691291332244873,
+      "learning_rate": 2.8313221601489757e-05,
+      "loss": 0.121,
+      "step": 13485
+    },
+    {
+      "epoch": 37.67039106145251,
+      "grad_norm": 0.5260879993438721,
+      "learning_rate": 2.8312849162011174e-05,
+      "loss": 0.0961,
+      "step": 13486
+    },
+    {
+      "epoch": 37.6731843575419,
+      "grad_norm": 0.5460242033004761,
+      "learning_rate": 2.831247672253259e-05,
+      "loss": 0.0994,
+      "step": 13487
+    },
+    {
+      "epoch": 37.675977653631286,
+      "grad_norm": 2.0270771980285645,
+      "learning_rate": 2.8312104283054007e-05,
+      "loss": 0.0847,
+      "step": 13488
+    },
+    {
+      "epoch": 37.67877094972067,
+      "grad_norm": 0.4517759084701538,
+      "learning_rate": 2.831173184357542e-05,
+      "loss": 0.0899,
+      "step": 13489
+    },
+    {
+      "epoch": 37.68156424581006,
+      "grad_norm": 0.40351763367652893,
+      "learning_rate": 2.8311359404096833e-05,
+      "loss": 0.0745,
+      "step": 13490
+    },
+    {
+      "epoch": 37.684357541899445,
+      "grad_norm": 0.5966194272041321,
+      "learning_rate": 2.831098696461825e-05,
+      "loss": 0.0867,
+      "step": 13491
+    },
+    {
+      "epoch": 37.687150837988824,
+      "grad_norm": 0.5589469075202942,
+      "learning_rate": 2.8310614525139666e-05,
+      "loss": 0.0737,
+      "step": 13492
+    },
+    {
+      "epoch": 37.68994413407821,
+      "grad_norm": 1.3126776218414307,
+      "learning_rate": 2.8310242085661083e-05,
+      "loss": 0.0851,
+      "step": 13493
+    },
+    {
+      "epoch": 37.6927374301676,
+      "grad_norm": 1.3814119100570679,
+      "learning_rate": 2.8309869646182496e-05,
+      "loss": 0.0877,
+      "step": 13494
+    },
+    {
+      "epoch": 37.69553072625698,
+      "grad_norm": 1.1588068008422852,
+      "learning_rate": 2.8309497206703912e-05,
+      "loss": 0.0776,
+      "step": 13495
+    },
+    {
+      "epoch": 37.69832402234637,
+      "grad_norm": 0.6683679223060608,
+      "learning_rate": 2.8309124767225325e-05,
+      "loss": 0.0942,
+      "step": 13496
+    },
+    {
+      "epoch": 37.701117318435756,
+      "grad_norm": 0.5763810873031616,
+      "learning_rate": 2.8308752327746742e-05,
+      "loss": 0.1245,
+      "step": 13497
+    },
+    {
+      "epoch": 37.70391061452514,
+      "grad_norm": 0.3733951449394226,
+      "learning_rate": 2.8308379888268155e-05,
+      "loss": 0.1116,
+      "step": 13498
+    },
+    {
+      "epoch": 37.70670391061452,
+      "grad_norm": 0.5441080331802368,
+      "learning_rate": 2.830800744878957e-05,
+      "loss": 0.1028,
+      "step": 13499
+    },
+    {
+      "epoch": 37.70949720670391,
+      "grad_norm": 0.721950888633728,
+      "learning_rate": 2.8307635009310988e-05,
+      "loss": 0.1203,
+      "step": 13500
+    },
+    {
+      "epoch": 37.712290502793294,
+      "grad_norm": 1.3386579751968384,
+      "learning_rate": 2.8307262569832405e-05,
+      "loss": 0.1179,
+      "step": 13501
+    },
+    {
+      "epoch": 37.71508379888268,
+      "grad_norm": 0.4122866988182068,
+      "learning_rate": 2.830689013035382e-05,
+      "loss": 0.0993,
+      "step": 13502
+    },
+    {
+      "epoch": 37.71787709497207,
+      "grad_norm": 0.8496657609939575,
+      "learning_rate": 2.830651769087523e-05,
+      "loss": 0.1042,
+      "step": 13503
+    },
+    {
+      "epoch": 37.720670391061454,
+      "grad_norm": 0.42184126377105713,
+      "learning_rate": 2.8306145251396648e-05,
+      "loss": 0.1155,
+      "step": 13504
+    },
+    {
+      "epoch": 37.72346368715084,
+      "grad_norm": 0.4997349977493286,
+      "learning_rate": 2.8305772811918064e-05,
+      "loss": 0.0942,
+      "step": 13505
+    },
+    {
+      "epoch": 37.726256983240226,
+      "grad_norm": 0.8263510465621948,
+      "learning_rate": 2.830540037243948e-05,
+      "loss": 0.0947,
+      "step": 13506
+    },
+    {
+      "epoch": 37.729050279329606,
+      "grad_norm": 0.4532104730606079,
+      "learning_rate": 2.8305027932960894e-05,
+      "loss": 0.093,
+      "step": 13507
+    },
+    {
+      "epoch": 37.73184357541899,
+      "grad_norm": 0.6759839653968811,
+      "learning_rate": 2.830465549348231e-05,
+      "loss": 0.101,
+      "step": 13508
+    },
+    {
+      "epoch": 37.73463687150838,
+      "grad_norm": 0.47681787610054016,
+      "learning_rate": 2.8304283054003727e-05,
+      "loss": 0.0886,
+      "step": 13509
+    },
+    {
+      "epoch": 37.737430167597765,
+      "grad_norm": 0.7991171479225159,
+      "learning_rate": 2.830391061452514e-05,
+      "loss": 0.0772,
+      "step": 13510
+    },
+    {
+      "epoch": 37.74022346368715,
+      "grad_norm": 0.5240347981452942,
+      "learning_rate": 2.8303538175046557e-05,
+      "loss": 0.0871,
+      "step": 13511
+    },
+    {
+      "epoch": 37.74301675977654,
+      "grad_norm": 1.0193067789077759,
+      "learning_rate": 2.830316573556797e-05,
+      "loss": 0.0934,
+      "step": 13512
+    },
+    {
+      "epoch": 37.745810055865924,
+      "grad_norm": 0.5378578305244446,
+      "learning_rate": 2.8302793296089386e-05,
+      "loss": 0.074,
+      "step": 13513
+    },
+    {
+      "epoch": 37.74860335195531,
+      "grad_norm": 0.5372198820114136,
+      "learning_rate": 2.8302420856610803e-05,
+      "loss": 0.0699,
+      "step": 13514
+    },
+    {
+      "epoch": 37.75139664804469,
+      "grad_norm": 1.436480164527893,
+      "learning_rate": 2.830204841713222e-05,
+      "loss": 0.084,
+      "step": 13515
+    },
+    {
+      "epoch": 37.754189944134076,
+      "grad_norm": 0.9659476280212402,
+      "learning_rate": 2.830167597765363e-05,
+      "loss": 0.0992,
+      "step": 13516
+    },
+    {
+      "epoch": 37.75698324022346,
+      "grad_norm": 0.6202499270439148,
+      "learning_rate": 2.8301303538175046e-05,
+      "loss": 0.089,
+      "step": 13517
+    },
+    {
+      "epoch": 37.75977653631285,
+      "grad_norm": 0.5778876543045044,
+      "learning_rate": 2.8300931098696462e-05,
+      "loss": 0.0689,
+      "step": 13518
+    },
+    {
+      "epoch": 37.762569832402235,
+      "grad_norm": 0.6211645007133484,
+      "learning_rate": 2.830055865921788e-05,
+      "loss": 0.0678,
+      "step": 13519
+    },
+    {
+      "epoch": 37.76536312849162,
+      "grad_norm": 1.2204548120498657,
+      "learning_rate": 2.8300186219739295e-05,
+      "loss": 0.0869,
+      "step": 13520
+    },
+    {
+      "epoch": 37.76815642458101,
+      "grad_norm": 1.3301259279251099,
+      "learning_rate": 2.829981378026071e-05,
+      "loss": 0.1025,
+      "step": 13521
+    },
+    {
+      "epoch": 37.770949720670394,
+      "grad_norm": 0.6454889178276062,
+      "learning_rate": 2.8299441340782125e-05,
+      "loss": 0.1185,
+      "step": 13522
+    },
+    {
+      "epoch": 37.773743016759774,
+      "grad_norm": 0.678033709526062,
+      "learning_rate": 2.8299068901303538e-05,
+      "loss": 0.1458,
+      "step": 13523
+    },
+    {
+      "epoch": 37.77653631284916,
+      "grad_norm": 1.3174045085906982,
+      "learning_rate": 2.8298696461824955e-05,
+      "loss": 0.1204,
+      "step": 13524
+    },
+    {
+      "epoch": 37.779329608938546,
+      "grad_norm": 0.5161803960800171,
+      "learning_rate": 2.8298324022346368e-05,
+      "loss": 0.0941,
+      "step": 13525
+    },
+    {
+      "epoch": 37.78212290502793,
+      "grad_norm": 3.0800626277923584,
+      "learning_rate": 2.8297951582867784e-05,
+      "loss": 0.1096,
+      "step": 13526
+    },
+    {
+      "epoch": 37.78491620111732,
+      "grad_norm": 0.6972345113754272,
+      "learning_rate": 2.82975791433892e-05,
+      "loss": 0.1123,
+      "step": 13527
+    },
+    {
+      "epoch": 37.787709497206706,
+      "grad_norm": 1.100494384765625,
+      "learning_rate": 2.8297206703910617e-05,
+      "loss": 0.1191,
+      "step": 13528
+    },
+    {
+      "epoch": 37.79050279329609,
+      "grad_norm": 0.4266567826271057,
+      "learning_rate": 2.829683426443203e-05,
+      "loss": 0.1074,
+      "step": 13529
+    },
+    {
+      "epoch": 37.79329608938548,
+      "grad_norm": 1.1610013246536255,
+      "learning_rate": 2.8296461824953444e-05,
+      "loss": 0.1241,
+      "step": 13530
+    },
+    {
+      "epoch": 37.79608938547486,
+      "grad_norm": 0.9091783165931702,
+      "learning_rate": 2.829608938547486e-05,
+      "loss": 0.097,
+      "step": 13531
+    },
+    {
+      "epoch": 37.798882681564244,
+      "grad_norm": 0.601504385471344,
+      "learning_rate": 2.8295716945996277e-05,
+      "loss": 0.1166,
+      "step": 13532
+    },
+    {
+      "epoch": 37.80167597765363,
+      "grad_norm": 0.9661040306091309,
+      "learning_rate": 2.8295344506517693e-05,
+      "loss": 0.098,
+      "step": 13533
+    },
+    {
+      "epoch": 37.80446927374302,
+      "grad_norm": 0.47214701771736145,
+      "learning_rate": 2.8294972067039106e-05,
+      "loss": 0.0988,
+      "step": 13534
+    },
+    {
+      "epoch": 37.8072625698324,
+      "grad_norm": 0.49803289771080017,
+      "learning_rate": 2.8294599627560523e-05,
+      "loss": 0.0838,
+      "step": 13535
+    },
+    {
+      "epoch": 37.81005586592179,
+      "grad_norm": 0.5302631258964539,
+      "learning_rate": 2.8294227188081936e-05,
+      "loss": 0.0729,
+      "step": 13536
+    },
+    {
+      "epoch": 37.812849162011176,
+      "grad_norm": 0.8785151839256287,
+      "learning_rate": 2.8293854748603353e-05,
+      "loss": 0.0772,
+      "step": 13537
+    },
+    {
+      "epoch": 37.815642458100555,
+      "grad_norm": 0.7076191306114197,
+      "learning_rate": 2.8293482309124766e-05,
+      "loss": 0.0923,
+      "step": 13538
+    },
+    {
+      "epoch": 37.81843575418994,
+      "grad_norm": 0.60430508852005,
+      "learning_rate": 2.8293109869646182e-05,
+      "loss": 0.0891,
+      "step": 13539
+    },
+    {
+      "epoch": 37.82122905027933,
+      "grad_norm": 1.816399335861206,
+      "learning_rate": 2.82927374301676e-05,
+      "loss": 0.0803,
+      "step": 13540
+    },
+    {
+      "epoch": 37.824022346368714,
+      "grad_norm": 0.838579535484314,
+      "learning_rate": 2.8292364990689015e-05,
+      "loss": 0.1021,
+      "step": 13541
+    },
+    {
+      "epoch": 37.8268156424581,
+      "grad_norm": 0.48321452736854553,
+      "learning_rate": 2.8291992551210432e-05,
+      "loss": 0.0738,
+      "step": 13542
+    },
+    {
+      "epoch": 37.82960893854749,
+      "grad_norm": 2.8692708015441895,
+      "learning_rate": 2.829162011173184e-05,
+      "loss": 0.0862,
+      "step": 13543
+    },
+    {
+      "epoch": 37.832402234636874,
+      "grad_norm": 0.587476909160614,
+      "learning_rate": 2.8291247672253258e-05,
+      "loss": 0.0766,
+      "step": 13544
+    },
+    {
+      "epoch": 37.83519553072626,
+      "grad_norm": 1.1105986833572388,
+      "learning_rate": 2.8290875232774675e-05,
+      "loss": 0.072,
+      "step": 13545
+    },
+    {
+      "epoch": 37.83798882681564,
+      "grad_norm": 2.053555488586426,
+      "learning_rate": 2.829050279329609e-05,
+      "loss": 0.115,
+      "step": 13546
+    },
+    {
+      "epoch": 37.840782122905026,
+      "grad_norm": 0.5315178632736206,
+      "learning_rate": 2.8290130353817504e-05,
+      "loss": 0.1307,
+      "step": 13547
+    },
+    {
+      "epoch": 37.84357541899441,
+      "grad_norm": 0.47529634833335876,
+      "learning_rate": 2.828975791433892e-05,
+      "loss": 0.1158,
+      "step": 13548
+    },
+    {
+      "epoch": 37.8463687150838,
+      "grad_norm": 0.3512713313102722,
+      "learning_rate": 2.8289385474860337e-05,
+      "loss": 0.1268,
+      "step": 13549
+    },
+    {
+      "epoch": 37.849162011173185,
+      "grad_norm": 0.44737356901168823,
+      "learning_rate": 2.828901303538175e-05,
+      "loss": 0.1351,
+      "step": 13550
+    },
+    {
+      "epoch": 37.85195530726257,
+      "grad_norm": 0.6788381338119507,
+      "learning_rate": 2.8288640595903167e-05,
+      "loss": 0.1074,
+      "step": 13551
+    },
+    {
+      "epoch": 37.85474860335196,
+      "grad_norm": 0.9020258188247681,
+      "learning_rate": 2.828826815642458e-05,
+      "loss": 0.1038,
+      "step": 13552
+    },
+    {
+      "epoch": 37.857541899441344,
+      "grad_norm": 0.4978260397911072,
+      "learning_rate": 2.8287895716945997e-05,
+      "loss": 0.1096,
+      "step": 13553
+    },
+    {
+      "epoch": 37.86033519553072,
+      "grad_norm": 0.9137853980064392,
+      "learning_rate": 2.8287523277467413e-05,
+      "loss": 0.0982,
+      "step": 13554
+    },
+    {
+      "epoch": 37.86312849162011,
+      "grad_norm": 0.5086469650268555,
+      "learning_rate": 2.828715083798883e-05,
+      "loss": 0.1034,
+      "step": 13555
+    },
+    {
+      "epoch": 37.865921787709496,
+      "grad_norm": 0.7379629015922546,
+      "learning_rate": 2.828677839851024e-05,
+      "loss": 0.1085,
+      "step": 13556
+    },
+    {
+      "epoch": 37.86871508379888,
+      "grad_norm": 1.3744231462478638,
+      "learning_rate": 2.8286405959031656e-05,
+      "loss": 0.0843,
+      "step": 13557
+    },
+    {
+      "epoch": 37.87150837988827,
+      "grad_norm": 0.49118366837501526,
+      "learning_rate": 2.8286033519553073e-05,
+      "loss": 0.1005,
+      "step": 13558
+    },
+    {
+      "epoch": 37.874301675977655,
+      "grad_norm": 0.5433399081230164,
+      "learning_rate": 2.828566108007449e-05,
+      "loss": 0.0903,
+      "step": 13559
+    },
+    {
+      "epoch": 37.87709497206704,
+      "grad_norm": 0.4455433487892151,
+      "learning_rate": 2.8285288640595906e-05,
+      "loss": 0.0804,
+      "step": 13560
+    },
+    {
+      "epoch": 37.87988826815643,
+      "grad_norm": 0.52919602394104,
+      "learning_rate": 2.828491620111732e-05,
+      "loss": 0.087,
+      "step": 13561
+    },
+    {
+      "epoch": 37.88268156424581,
+      "grad_norm": 0.5937462449073792,
+      "learning_rate": 2.8284543761638735e-05,
+      "loss": 0.0739,
+      "step": 13562
+    },
+    {
+      "epoch": 37.885474860335194,
+      "grad_norm": 0.5406915545463562,
+      "learning_rate": 2.828417132216015e-05,
+      "loss": 0.0722,
+      "step": 13563
+    },
+    {
+      "epoch": 37.88826815642458,
+      "grad_norm": 0.6284484267234802,
+      "learning_rate": 2.8283798882681565e-05,
+      "loss": 0.0586,
+      "step": 13564
+    },
+    {
+      "epoch": 37.891061452513966,
+      "grad_norm": 1.0538344383239746,
+      "learning_rate": 2.828342644320298e-05,
+      "loss": 0.1021,
+      "step": 13565
+    },
+    {
+      "epoch": 37.89385474860335,
+      "grad_norm": 0.4975854754447937,
+      "learning_rate": 2.8283054003724395e-05,
+      "loss": 0.08,
+      "step": 13566
+    },
+    {
+      "epoch": 37.89664804469274,
+      "grad_norm": 0.5422156453132629,
+      "learning_rate": 2.828268156424581e-05,
+      "loss": 0.0784,
+      "step": 13567
+    },
+    {
+      "epoch": 37.899441340782126,
+      "grad_norm": 2.4349567890167236,
+      "learning_rate": 2.8282309124767228e-05,
+      "loss": 0.0901,
+      "step": 13568
+    },
+    {
+      "epoch": 37.90223463687151,
+      "grad_norm": 6.845480918884277,
+      "learning_rate": 2.8281936685288644e-05,
+      "loss": 0.0814,
+      "step": 13569
+    },
+    {
+      "epoch": 37.90502793296089,
+      "grad_norm": 1.056230902671814,
+      "learning_rate": 2.8281564245810054e-05,
+      "loss": 0.1114,
+      "step": 13570
+    },
+    {
+      "epoch": 37.90782122905028,
+      "grad_norm": 1.1407275199890137,
+      "learning_rate": 2.828119180633147e-05,
+      "loss": 0.1104,
+      "step": 13571
+    },
+    {
+      "epoch": 37.910614525139664,
+      "grad_norm": 1.0816618204116821,
+      "learning_rate": 2.8280819366852887e-05,
+      "loss": 0.1578,
+      "step": 13572
+    },
+    {
+      "epoch": 37.91340782122905,
+      "grad_norm": 0.6514893770217896,
+      "learning_rate": 2.8280446927374304e-05,
+      "loss": 0.1263,
+      "step": 13573
+    },
+    {
+      "epoch": 37.91620111731844,
+      "grad_norm": 0.3552793562412262,
+      "learning_rate": 2.8280074487895717e-05,
+      "loss": 0.1194,
+      "step": 13574
+    },
+    {
+      "epoch": 37.91899441340782,
+      "grad_norm": 0.46572446823120117,
+      "learning_rate": 2.8279702048417134e-05,
+      "loss": 0.1245,
+      "step": 13575
+    },
+    {
+      "epoch": 37.92178770949721,
+      "grad_norm": 0.40307238698005676,
+      "learning_rate": 2.8279329608938547e-05,
+      "loss": 0.1126,
+      "step": 13576
+    },
+    {
+      "epoch": 37.92458100558659,
+      "grad_norm": 0.931151270866394,
+      "learning_rate": 2.8278957169459963e-05,
+      "loss": 0.1095,
+      "step": 13577
+    },
+    {
+      "epoch": 37.927374301675975,
+      "grad_norm": 0.5486778616905212,
+      "learning_rate": 2.827858472998138e-05,
+      "loss": 0.1035,
+      "step": 13578
+    },
+    {
+      "epoch": 37.93016759776536,
+      "grad_norm": 0.2694908380508423,
+      "learning_rate": 2.8278212290502793e-05,
+      "loss": 0.0905,
+      "step": 13579
+    },
+    {
+      "epoch": 37.93296089385475,
+      "grad_norm": 0.449146568775177,
+      "learning_rate": 2.827783985102421e-05,
+      "loss": 0.1,
+      "step": 13580
+    },
+    {
+      "epoch": 37.935754189944134,
+      "grad_norm": 0.754401445388794,
+      "learning_rate": 2.8277467411545626e-05,
+      "loss": 0.1009,
+      "step": 13581
+    },
+    {
+      "epoch": 37.93854748603352,
+      "grad_norm": 0.639654815196991,
+      "learning_rate": 2.8277094972067042e-05,
+      "loss": 0.0942,
+      "step": 13582
+    },
+    {
+      "epoch": 37.94134078212291,
+      "grad_norm": 0.9236147403717041,
+      "learning_rate": 2.8276722532588452e-05,
+      "loss": 0.0889,
+      "step": 13583
+    },
+    {
+      "epoch": 37.944134078212294,
+      "grad_norm": 1.026513695716858,
+      "learning_rate": 2.827635009310987e-05,
+      "loss": 0.0894,
+      "step": 13584
+    },
+    {
+      "epoch": 37.94692737430167,
+      "grad_norm": 0.43466776609420776,
+      "learning_rate": 2.8275977653631285e-05,
+      "loss": 0.0806,
+      "step": 13585
+    },
+    {
+      "epoch": 37.94972067039106,
+      "grad_norm": 0.4979175925254822,
+      "learning_rate": 2.8275605214152702e-05,
+      "loss": 0.0805,
+      "step": 13586
+    },
+    {
+      "epoch": 37.952513966480446,
+      "grad_norm": 0.5099448561668396,
+      "learning_rate": 2.827523277467412e-05,
+      "loss": 0.0756,
+      "step": 13587
+    },
+    {
+      "epoch": 37.95530726256983,
+      "grad_norm": 0.818507969379425,
+      "learning_rate": 2.827486033519553e-05,
+      "loss": 0.0806,
+      "step": 13588
+    },
+    {
+      "epoch": 37.95810055865922,
+      "grad_norm": 0.5272806882858276,
+      "learning_rate": 2.8274487895716948e-05,
+      "loss": 0.0785,
+      "step": 13589
+    },
+    {
+      "epoch": 37.960893854748605,
+      "grad_norm": 0.7429022192955017,
+      "learning_rate": 2.827411545623836e-05,
+      "loss": 0.0851,
+      "step": 13590
+    },
+    {
+      "epoch": 37.96368715083799,
+      "grad_norm": 1.2446386814117432,
+      "learning_rate": 2.8273743016759778e-05,
+      "loss": 0.0571,
+      "step": 13591
+    },
+    {
+      "epoch": 37.96648044692738,
+      "grad_norm": 0.5699859857559204,
+      "learning_rate": 2.827337057728119e-05,
+      "loss": 0.0644,
+      "step": 13592
+    },
+    {
+      "epoch": 37.96927374301676,
+      "grad_norm": 0.9957475662231445,
+      "learning_rate": 2.8272998137802607e-05,
+      "loss": 0.0865,
+      "step": 13593
+    },
+    {
+      "epoch": 37.97206703910614,
+      "grad_norm": 0.5944711565971375,
+      "learning_rate": 2.8272625698324024e-05,
+      "loss": 0.0766,
+      "step": 13594
+    },
+    {
+      "epoch": 37.97486033519553,
+      "grad_norm": 0.7629641890525818,
+      "learning_rate": 2.827225325884544e-05,
+      "loss": 0.0716,
+      "step": 13595
+    },
+    {
+      "epoch": 37.977653631284916,
+      "grad_norm": 1.5692315101623535,
+      "learning_rate": 2.8271880819366854e-05,
+      "loss": 0.0926,
+      "step": 13596
+    },
+    {
+      "epoch": 37.9804469273743,
+      "grad_norm": 0.5939427018165588,
+      "learning_rate": 2.8271508379888267e-05,
+      "loss": 0.1423,
+      "step": 13597
+    },
+    {
+      "epoch": 37.98324022346369,
+      "grad_norm": 0.5194096565246582,
+      "learning_rate": 2.8271135940409683e-05,
+      "loss": 0.1203,
+      "step": 13598
+    },
+    {
+      "epoch": 37.986033519553075,
+      "grad_norm": 0.6489980220794678,
+      "learning_rate": 2.82707635009311e-05,
+      "loss": 0.099,
+      "step": 13599
+    },
+    {
+      "epoch": 37.98882681564246,
+      "grad_norm": 1.8672524690628052,
+      "learning_rate": 2.8270391061452516e-05,
+      "loss": 0.1208,
+      "step": 13600
+    },
+    {
+      "epoch": 37.99162011173184,
+      "grad_norm": 0.8116322755813599,
+      "learning_rate": 2.827001862197393e-05,
+      "loss": 0.0923,
+      "step": 13601
+    },
+    {
+      "epoch": 37.99441340782123,
+      "grad_norm": 0.7780147194862366,
+      "learning_rate": 2.8269646182495346e-05,
+      "loss": 0.0892,
+      "step": 13602
+    },
+    {
+      "epoch": 37.997206703910614,
+      "grad_norm": 1.1590795516967773,
+      "learning_rate": 2.826927374301676e-05,
+      "loss": 0.0803,
+      "step": 13603
+    },
+    {
+      "epoch": 38.0,
+      "grad_norm": 0.8958903551101685,
+      "learning_rate": 2.8268901303538176e-05,
+      "loss": 0.0854,
+      "step": 13604
+    },
+    {
+      "epoch": 38.002793296089386,
+      "grad_norm": 0.7198182344436646,
+      "learning_rate": 2.8268528864059592e-05,
+      "loss": 0.1597,
+      "step": 13605
+    },
+    {
+      "epoch": 38.00558659217877,
+      "grad_norm": 0.893574059009552,
+      "learning_rate": 2.8268156424581005e-05,
+      "loss": 0.1118,
+      "step": 13606
+    },
+    {
+      "epoch": 38.00837988826816,
+      "grad_norm": 0.37087881565093994,
+      "learning_rate": 2.8267783985102422e-05,
+      "loss": 0.1334,
+      "step": 13607
+    },
+    {
+      "epoch": 38.01117318435754,
+      "grad_norm": 0.3942200839519501,
+      "learning_rate": 2.826741154562384e-05,
+      "loss": 0.1025,
+      "step": 13608
+    },
+    {
+      "epoch": 38.013966480446925,
+      "grad_norm": 0.6616092920303345,
+      "learning_rate": 2.8267039106145255e-05,
+      "loss": 0.1067,
+      "step": 13609
+    },
+    {
+      "epoch": 38.01675977653631,
+      "grad_norm": 1.5576893091201782,
+      "learning_rate": 2.8266666666666665e-05,
+      "loss": 0.1003,
+      "step": 13610
+    },
+    {
+      "epoch": 38.0195530726257,
+      "grad_norm": 0.5423663258552551,
+      "learning_rate": 2.826629422718808e-05,
+      "loss": 0.0944,
+      "step": 13611
+    },
+    {
+      "epoch": 38.022346368715084,
+      "grad_norm": 0.7358835339546204,
+      "learning_rate": 2.8265921787709498e-05,
+      "loss": 0.1091,
+      "step": 13612
+    },
+    {
+      "epoch": 38.02513966480447,
+      "grad_norm": 1.841209053993225,
+      "learning_rate": 2.8265549348230914e-05,
+      "loss": 0.1033,
+      "step": 13613
+    },
+    {
+      "epoch": 38.02793296089386,
+      "grad_norm": 0.5449988842010498,
+      "learning_rate": 2.826517690875233e-05,
+      "loss": 0.1002,
+      "step": 13614
+    },
+    {
+      "epoch": 38.03072625698324,
+      "grad_norm": 1.0527795553207397,
+      "learning_rate": 2.8264804469273744e-05,
+      "loss": 0.0782,
+      "step": 13615
+    },
+    {
+      "epoch": 38.03351955307262,
+      "grad_norm": 0.8708153367042542,
+      "learning_rate": 2.8264432029795157e-05,
+      "loss": 0.0923,
+      "step": 13616
+    },
+    {
+      "epoch": 38.03631284916201,
+      "grad_norm": 0.5480820536613464,
+      "learning_rate": 2.8264059590316574e-05,
+      "loss": 0.0742,
+      "step": 13617
+    },
+    {
+      "epoch": 38.039106145251395,
+      "grad_norm": 0.9960919618606567,
+      "learning_rate": 2.826368715083799e-05,
+      "loss": 0.0832,
+      "step": 13618
+    },
+    {
+      "epoch": 38.04189944134078,
+      "grad_norm": 0.5023010969161987,
+      "learning_rate": 2.8263314711359403e-05,
+      "loss": 0.0854,
+      "step": 13619
+    },
+    {
+      "epoch": 38.04469273743017,
+      "grad_norm": 0.4977894127368927,
+      "learning_rate": 2.826294227188082e-05,
+      "loss": 0.08,
+      "step": 13620
+    },
+    {
+      "epoch": 38.047486033519554,
+      "grad_norm": 0.5609042048454285,
+      "learning_rate": 2.8262569832402237e-05,
+      "loss": 0.063,
+      "step": 13621
+    },
+    {
+      "epoch": 38.05027932960894,
+      "grad_norm": 1.672320008277893,
+      "learning_rate": 2.8262197392923653e-05,
+      "loss": 0.0659,
+      "step": 13622
+    },
+    {
+      "epoch": 38.05307262569833,
+      "grad_norm": 0.5528787970542908,
+      "learning_rate": 2.8261824953445066e-05,
+      "loss": 0.0887,
+      "step": 13623
+    },
+    {
+      "epoch": 38.055865921787706,
+      "grad_norm": 1.5771676301956177,
+      "learning_rate": 2.826145251396648e-05,
+      "loss": 0.0774,
+      "step": 13624
+    },
+    {
+      "epoch": 38.05865921787709,
+      "grad_norm": 0.7770093083381653,
+      "learning_rate": 2.8261080074487896e-05,
+      "loss": 0.0618,
+      "step": 13625
+    },
+    {
+      "epoch": 38.06145251396648,
+      "grad_norm": 0.7451122403144836,
+      "learning_rate": 2.8260707635009312e-05,
+      "loss": 0.0577,
+      "step": 13626
+    },
+    {
+      "epoch": 38.064245810055866,
+      "grad_norm": 0.6932182312011719,
+      "learning_rate": 2.826033519553073e-05,
+      "loss": 0.0737,
+      "step": 13627
+    },
+    {
+      "epoch": 38.06703910614525,
+      "grad_norm": 1.5139777660369873,
+      "learning_rate": 2.8259962756052142e-05,
+      "loss": 0.0694,
+      "step": 13628
+    },
+    {
+      "epoch": 38.06983240223464,
+      "grad_norm": 1.4735448360443115,
+      "learning_rate": 2.8259590316573555e-05,
+      "loss": 0.0989,
+      "step": 13629
+    },
+    {
+      "epoch": 38.072625698324025,
+      "grad_norm": 1.1085277795791626,
+      "learning_rate": 2.8259217877094972e-05,
+      "loss": 0.1209,
+      "step": 13630
+    },
+    {
+      "epoch": 38.07541899441341,
+      "grad_norm": 0.3986848294734955,
+      "learning_rate": 2.825884543761639e-05,
+      "loss": 0.1104,
+      "step": 13631
+    },
+    {
+      "epoch": 38.07821229050279,
+      "grad_norm": 0.44649389386177063,
+      "learning_rate": 2.82584729981378e-05,
+      "loss": 0.1067,
+      "step": 13632
+    },
+    {
+      "epoch": 38.08100558659218,
+      "grad_norm": 0.689566433429718,
+      "learning_rate": 2.8258100558659218e-05,
+      "loss": 0.1,
+      "step": 13633
+    },
+    {
+      "epoch": 38.08379888268156,
+      "grad_norm": 1.3156629800796509,
+      "learning_rate": 2.8257728119180635e-05,
+      "loss": 0.117,
+      "step": 13634
+    },
+    {
+      "epoch": 38.08659217877095,
+      "grad_norm": 0.5990111827850342,
+      "learning_rate": 2.825735567970205e-05,
+      "loss": 0.0831,
+      "step": 13635
+    },
+    {
+      "epoch": 38.089385474860336,
+      "grad_norm": 0.5798774361610413,
+      "learning_rate": 2.8256983240223464e-05,
+      "loss": 0.0903,
+      "step": 13636
+    },
+    {
+      "epoch": 38.09217877094972,
+      "grad_norm": 0.5133525729179382,
+      "learning_rate": 2.8256610800744877e-05,
+      "loss": 0.0883,
+      "step": 13637
+    },
+    {
+      "epoch": 38.09497206703911,
+      "grad_norm": 0.4218067526817322,
+      "learning_rate": 2.8256238361266294e-05,
+      "loss": 0.0781,
+      "step": 13638
+    },
+    {
+      "epoch": 38.097765363128495,
+      "grad_norm": 0.4693346917629242,
+      "learning_rate": 2.825586592178771e-05,
+      "loss": 0.0987,
+      "step": 13639
+    },
+    {
+      "epoch": 38.100558659217874,
+      "grad_norm": 0.5241656303405762,
+      "learning_rate": 2.8255493482309127e-05,
+      "loss": 0.0797,
+      "step": 13640
+    },
+    {
+      "epoch": 38.10335195530726,
+      "grad_norm": 1.5795722007751465,
+      "learning_rate": 2.825512104283054e-05,
+      "loss": 0.0955,
+      "step": 13641
+    },
+    {
+      "epoch": 38.10614525139665,
+      "grad_norm": 1.8565741777420044,
+      "learning_rate": 2.8254748603351957e-05,
+      "loss": 0.0908,
+      "step": 13642
+    },
+    {
+      "epoch": 38.108938547486034,
+      "grad_norm": 0.6266420483589172,
+      "learning_rate": 2.825437616387337e-05,
+      "loss": 0.0917,
+      "step": 13643
+    },
+    {
+      "epoch": 38.11173184357542,
+      "grad_norm": 0.49125924706459045,
+      "learning_rate": 2.8254003724394786e-05,
+      "loss": 0.0922,
+      "step": 13644
+    },
+    {
+      "epoch": 38.114525139664806,
+      "grad_norm": 1.1133333444595337,
+      "learning_rate": 2.8253631284916203e-05,
+      "loss": 0.0682,
+      "step": 13645
+    },
+    {
+      "epoch": 38.11731843575419,
+      "grad_norm": 0.5136995911598206,
+      "learning_rate": 2.8253258845437616e-05,
+      "loss": 0.0777,
+      "step": 13646
+    },
+    {
+      "epoch": 38.12011173184357,
+      "grad_norm": 0.6037970185279846,
+      "learning_rate": 2.8252886405959033e-05,
+      "loss": 0.0809,
+      "step": 13647
+    },
+    {
+      "epoch": 38.12290502793296,
+      "grad_norm": 0.4830288290977478,
+      "learning_rate": 2.825251396648045e-05,
+      "loss": 0.0784,
+      "step": 13648
+    },
+    {
+      "epoch": 38.125698324022345,
+      "grad_norm": 2.425668478012085,
+      "learning_rate": 2.8252141527001862e-05,
+      "loss": 0.0558,
+      "step": 13649
+    },
+    {
+      "epoch": 38.12849162011173,
+      "grad_norm": 0.6688857674598694,
+      "learning_rate": 2.8251769087523275e-05,
+      "loss": 0.078,
+      "step": 13650
+    },
+    {
+      "epoch": 38.13128491620112,
+      "grad_norm": 1.0375055074691772,
+      "learning_rate": 2.8251396648044692e-05,
+      "loss": 0.0747,
+      "step": 13651
+    },
+    {
+      "epoch": 38.134078212290504,
+      "grad_norm": 0.7346288561820984,
+      "learning_rate": 2.825102420856611e-05,
+      "loss": 0.0718,
+      "step": 13652
+    },
+    {
+      "epoch": 38.13687150837989,
+      "grad_norm": 0.8996039628982544,
+      "learning_rate": 2.8250651769087525e-05,
+      "loss": 0.0792,
+      "step": 13653
+    },
+    {
+      "epoch": 38.13966480446928,
+      "grad_norm": 0.8932623863220215,
+      "learning_rate": 2.825027932960894e-05,
+      "loss": 0.0857,
+      "step": 13654
+    },
+    {
+      "epoch": 38.142458100558656,
+      "grad_norm": 0.378711462020874,
+      "learning_rate": 2.8249906890130355e-05,
+      "loss": 0.1138,
+      "step": 13655
+    },
+    {
+      "epoch": 38.14525139664804,
+      "grad_norm": 0.49095889925956726,
+      "learning_rate": 2.8249534450651768e-05,
+      "loss": 0.1177,
+      "step": 13656
+    },
+    {
+      "epoch": 38.14804469273743,
+      "grad_norm": 0.3813093304634094,
+      "learning_rate": 2.8249162011173184e-05,
+      "loss": 0.1115,
+      "step": 13657
+    },
+    {
+      "epoch": 38.150837988826815,
+      "grad_norm": 1.2772905826568604,
+      "learning_rate": 2.82487895716946e-05,
+      "loss": 0.1208,
+      "step": 13658
+    },
+    {
+      "epoch": 38.1536312849162,
+      "grad_norm": 0.4416792392730713,
+      "learning_rate": 2.8248417132216014e-05,
+      "loss": 0.0938,
+      "step": 13659
+    },
+    {
+      "epoch": 38.15642458100559,
+      "grad_norm": 0.6753383278846741,
+      "learning_rate": 2.824804469273743e-05,
+      "loss": 0.105,
+      "step": 13660
+    },
+    {
+      "epoch": 38.159217877094974,
+      "grad_norm": 0.6458865404129028,
+      "learning_rate": 2.8247672253258847e-05,
+      "loss": 0.1056,
+      "step": 13661
+    },
+    {
+      "epoch": 38.16201117318436,
+      "grad_norm": 0.559935986995697,
+      "learning_rate": 2.8247299813780264e-05,
+      "loss": 0.101,
+      "step": 13662
+    },
+    {
+      "epoch": 38.16480446927374,
+      "grad_norm": 0.5054547786712646,
+      "learning_rate": 2.8246927374301677e-05,
+      "loss": 0.0761,
+      "step": 13663
+    },
+    {
+      "epoch": 38.167597765363126,
+      "grad_norm": 0.45961886644363403,
+      "learning_rate": 2.824655493482309e-05,
+      "loss": 0.0965,
+      "step": 13664
+    },
+    {
+      "epoch": 38.17039106145251,
+      "grad_norm": 0.4697854518890381,
+      "learning_rate": 2.8246182495344507e-05,
+      "loss": 0.0898,
+      "step": 13665
+    },
+    {
+      "epoch": 38.1731843575419,
+      "grad_norm": 0.577907145023346,
+      "learning_rate": 2.8245810055865923e-05,
+      "loss": 0.0934,
+      "step": 13666
+    },
+    {
+      "epoch": 38.175977653631286,
+      "grad_norm": 0.4840467870235443,
+      "learning_rate": 2.824543761638734e-05,
+      "loss": 0.0776,
+      "step": 13667
+    },
+    {
+      "epoch": 38.17877094972067,
+      "grad_norm": 0.8989623785018921,
+      "learning_rate": 2.8245065176908753e-05,
+      "loss": 0.0881,
+      "step": 13668
+    },
+    {
+      "epoch": 38.18156424581006,
+      "grad_norm": 0.5477580428123474,
+      "learning_rate": 2.8244692737430166e-05,
+      "loss": 0.0836,
+      "step": 13669
+    },
+    {
+      "epoch": 38.184357541899445,
+      "grad_norm": 0.7397028803825378,
+      "learning_rate": 2.8244320297951582e-05,
+      "loss": 0.068,
+      "step": 13670
+    },
+    {
+      "epoch": 38.187150837988824,
+      "grad_norm": 0.5579710602760315,
+      "learning_rate": 2.8243947858473e-05,
+      "loss": 0.0811,
+      "step": 13671
+    },
+    {
+      "epoch": 38.18994413407821,
+      "grad_norm": 0.40642082691192627,
+      "learning_rate": 2.8243575418994415e-05,
+      "loss": 0.0666,
+      "step": 13672
+    },
+    {
+      "epoch": 38.1927374301676,
+      "grad_norm": 0.541372537612915,
+      "learning_rate": 2.824320297951583e-05,
+      "loss": 0.062,
+      "step": 13673
+    },
+    {
+      "epoch": 38.19553072625698,
+      "grad_norm": 0.4557761251926422,
+      "learning_rate": 2.8242830540037245e-05,
+      "loss": 0.0749,
+      "step": 13674
+    },
+    {
+      "epoch": 38.19832402234637,
+      "grad_norm": 2.782489538192749,
+      "learning_rate": 2.8242458100558662e-05,
+      "loss": 0.0729,
+      "step": 13675
+    },
+    {
+      "epoch": 38.201117318435756,
+      "grad_norm": 0.8448912501335144,
+      "learning_rate": 2.8242085661080075e-05,
+      "loss": 0.075,
+      "step": 13676
+    },
+    {
+      "epoch": 38.20391061452514,
+      "grad_norm": 0.615928053855896,
+      "learning_rate": 2.8241713221601488e-05,
+      "loss": 0.057,
+      "step": 13677
+    },
+    {
+      "epoch": 38.20670391061452,
+      "grad_norm": 0.7912046313285828,
+      "learning_rate": 2.8241340782122905e-05,
+      "loss": 0.0759,
+      "step": 13678
+    },
+    {
+      "epoch": 38.20949720670391,
+      "grad_norm": 1.0558451414108276,
+      "learning_rate": 2.824096834264432e-05,
+      "loss": 0.0853,
+      "step": 13679
+    },
+    {
+      "epoch": 38.212290502793294,
+      "grad_norm": 0.37920260429382324,
+      "learning_rate": 2.8240595903165738e-05,
+      "loss": 0.1139,
+      "step": 13680
+    },
+    {
+      "epoch": 38.21508379888268,
+      "grad_norm": 0.694470226764679,
+      "learning_rate": 2.8240223463687154e-05,
+      "loss": 0.1109,
+      "step": 13681
+    },
+    {
+      "epoch": 38.21787709497207,
+      "grad_norm": 0.4372653663158417,
+      "learning_rate": 2.8239851024208567e-05,
+      "loss": 0.1277,
+      "step": 13682
+    },
+    {
+      "epoch": 38.220670391061454,
+      "grad_norm": 0.5218306183815002,
+      "learning_rate": 2.823947858472998e-05,
+      "loss": 0.1122,
+      "step": 13683
+    },
+    {
+      "epoch": 38.22346368715084,
+      "grad_norm": 0.39648476243019104,
+      "learning_rate": 2.8239106145251397e-05,
+      "loss": 0.1024,
+      "step": 13684
+    },
+    {
+      "epoch": 38.226256983240226,
+      "grad_norm": 0.49528464674949646,
+      "learning_rate": 2.8238733705772814e-05,
+      "loss": 0.107,
+      "step": 13685
+    },
+    {
+      "epoch": 38.229050279329606,
+      "grad_norm": 0.4020237326622009,
+      "learning_rate": 2.8238361266294227e-05,
+      "loss": 0.09,
+      "step": 13686
+    },
+    {
+      "epoch": 38.23184357541899,
+      "grad_norm": 0.9500442743301392,
+      "learning_rate": 2.8237988826815643e-05,
+      "loss": 0.0947,
+      "step": 13687
+    },
+    {
+      "epoch": 38.23463687150838,
+      "grad_norm": 1.2254416942596436,
+      "learning_rate": 2.823761638733706e-05,
+      "loss": 0.1017,
+      "step": 13688
+    },
+    {
+      "epoch": 38.237430167597765,
+      "grad_norm": 0.573433518409729,
+      "learning_rate": 2.8237243947858473e-05,
+      "loss": 0.0968,
+      "step": 13689
+    },
+    {
+      "epoch": 38.24022346368715,
+      "grad_norm": 0.4361385405063629,
+      "learning_rate": 2.823687150837989e-05,
+      "loss": 0.0667,
+      "step": 13690
+    },
+    {
+      "epoch": 38.24301675977654,
+      "grad_norm": 0.9446371793746948,
+      "learning_rate": 2.8236499068901303e-05,
+      "loss": 0.0832,
+      "step": 13691
+    },
+    {
+      "epoch": 38.245810055865924,
+      "grad_norm": 0.9464927911758423,
+      "learning_rate": 2.823612662942272e-05,
+      "loss": 0.0936,
+      "step": 13692
+    },
+    {
+      "epoch": 38.24860335195531,
+      "grad_norm": 0.7366499900817871,
+      "learning_rate": 2.8235754189944136e-05,
+      "loss": 0.0939,
+      "step": 13693
+    },
+    {
+      "epoch": 38.25139664804469,
+      "grad_norm": 0.6689229011535645,
+      "learning_rate": 2.8235381750465552e-05,
+      "loss": 0.0875,
+      "step": 13694
+    },
+    {
+      "epoch": 38.254189944134076,
+      "grad_norm": 0.6350197792053223,
+      "learning_rate": 2.8235009310986965e-05,
+      "loss": 0.0712,
+      "step": 13695
+    },
+    {
+      "epoch": 38.25698324022346,
+      "grad_norm": 0.38538724184036255,
+      "learning_rate": 2.823463687150838e-05,
+      "loss": 0.0752,
+      "step": 13696
+    },
+    {
+      "epoch": 38.25977653631285,
+      "grad_norm": 1.7887818813323975,
+      "learning_rate": 2.8234264432029795e-05,
+      "loss": 0.0693,
+      "step": 13697
+    },
+    {
+      "epoch": 38.262569832402235,
+      "grad_norm": 1.0628087520599365,
+      "learning_rate": 2.823389199255121e-05,
+      "loss": 0.0703,
+      "step": 13698
+    },
+    {
+      "epoch": 38.26536312849162,
+      "grad_norm": 0.6115090847015381,
+      "learning_rate": 2.8233519553072628e-05,
+      "loss": 0.0882,
+      "step": 13699
+    },
+    {
+      "epoch": 38.26815642458101,
+      "grad_norm": 0.5248382091522217,
+      "learning_rate": 2.823314711359404e-05,
+      "loss": 0.0687,
+      "step": 13700
+    },
+    {
+      "epoch": 38.270949720670394,
+      "grad_norm": 0.67684406042099,
+      "learning_rate": 2.8232774674115458e-05,
+      "loss": 0.0637,
+      "step": 13701
+    },
+    {
+      "epoch": 38.273743016759774,
+      "grad_norm": 1.2188458442687988,
+      "learning_rate": 2.8232402234636874e-05,
+      "loss": 0.066,
+      "step": 13702
+    },
+    {
+      "epoch": 38.27653631284916,
+      "grad_norm": 0.8558314442634583,
+      "learning_rate": 2.8232029795158287e-05,
+      "loss": 0.0658,
+      "step": 13703
+    },
+    {
+      "epoch": 38.279329608938546,
+      "grad_norm": 0.9805386066436768,
+      "learning_rate": 2.82316573556797e-05,
+      "loss": 0.0978,
+      "step": 13704
+    },
+    {
+      "epoch": 38.28212290502793,
+      "grad_norm": 0.578081488609314,
+      "learning_rate": 2.8231284916201117e-05,
+      "loss": 0.1263,
+      "step": 13705
+    },
+    {
+      "epoch": 38.28491620111732,
+      "grad_norm": 0.49524828791618347,
+      "learning_rate": 2.8230912476722534e-05,
+      "loss": 0.0998,
+      "step": 13706
+    },
+    {
+      "epoch": 38.287709497206706,
+      "grad_norm": 0.8349334001541138,
+      "learning_rate": 2.823054003724395e-05,
+      "loss": 0.1322,
+      "step": 13707
+    },
+    {
+      "epoch": 38.29050279329609,
+      "grad_norm": 0.9149228930473328,
+      "learning_rate": 2.8230167597765367e-05,
+      "loss": 0.1065,
+      "step": 13708
+    },
+    {
+      "epoch": 38.29329608938548,
+      "grad_norm": 1.0362051725387573,
+      "learning_rate": 2.8229795158286776e-05,
+      "loss": 0.1195,
+      "step": 13709
+    },
+    {
+      "epoch": 38.29608938547486,
+      "grad_norm": 0.8216106295585632,
+      "learning_rate": 2.8229422718808193e-05,
+      "loss": 0.1023,
+      "step": 13710
+    },
+    {
+      "epoch": 38.298882681564244,
+      "grad_norm": 0.5101133584976196,
+      "learning_rate": 2.822905027932961e-05,
+      "loss": 0.1119,
+      "step": 13711
+    },
+    {
+      "epoch": 38.30167597765363,
+      "grad_norm": 1.2717500925064087,
+      "learning_rate": 2.8228677839851026e-05,
+      "loss": 0.1024,
+      "step": 13712
+    },
+    {
+      "epoch": 38.30446927374302,
+      "grad_norm": 0.4609566032886505,
+      "learning_rate": 2.822830540037244e-05,
+      "loss": 0.0885,
+      "step": 13713
+    },
+    {
+      "epoch": 38.3072625698324,
+      "grad_norm": 0.8391607999801636,
+      "learning_rate": 2.8227932960893856e-05,
+      "loss": 0.0964,
+      "step": 13714
+    },
+    {
+      "epoch": 38.31005586592179,
+      "grad_norm": 1.4443435668945312,
+      "learning_rate": 2.8227560521415272e-05,
+      "loss": 0.081,
+      "step": 13715
+    },
+    {
+      "epoch": 38.312849162011176,
+      "grad_norm": 0.8448764085769653,
+      "learning_rate": 2.8227188081936685e-05,
+      "loss": 0.1022,
+      "step": 13716
+    },
+    {
+      "epoch": 38.315642458100555,
+      "grad_norm": 0.4799998104572296,
+      "learning_rate": 2.8226815642458102e-05,
+      "loss": 0.0798,
+      "step": 13717
+    },
+    {
+      "epoch": 38.31843575418994,
+      "grad_norm": 0.45563554763793945,
+      "learning_rate": 2.8226443202979515e-05,
+      "loss": 0.0934,
+      "step": 13718
+    },
+    {
+      "epoch": 38.32122905027933,
+      "grad_norm": 0.5778693556785583,
+      "learning_rate": 2.822607076350093e-05,
+      "loss": 0.0816,
+      "step": 13719
+    },
+    {
+      "epoch": 38.324022346368714,
+      "grad_norm": 0.4454435110092163,
+      "learning_rate": 2.8225698324022348e-05,
+      "loss": 0.0824,
+      "step": 13720
+    },
+    {
+      "epoch": 38.3268156424581,
+      "grad_norm": 0.5092951655387878,
+      "learning_rate": 2.8225325884543765e-05,
+      "loss": 0.0774,
+      "step": 13721
+    },
+    {
+      "epoch": 38.32960893854749,
+      "grad_norm": 2.418661117553711,
+      "learning_rate": 2.8224953445065178e-05,
+      "loss": 0.0668,
+      "step": 13722
+    },
+    {
+      "epoch": 38.332402234636874,
+      "grad_norm": 0.5884087681770325,
+      "learning_rate": 2.822458100558659e-05,
+      "loss": 0.0876,
+      "step": 13723
+    },
+    {
+      "epoch": 38.33519553072626,
+      "grad_norm": 0.46387264132499695,
+      "learning_rate": 2.8224208566108008e-05,
+      "loss": 0.0781,
+      "step": 13724
+    },
+    {
+      "epoch": 38.33798882681564,
+      "grad_norm": 0.5995131134986877,
+      "learning_rate": 2.8223836126629424e-05,
+      "loss": 0.0644,
+      "step": 13725
+    },
+    {
+      "epoch": 38.340782122905026,
+      "grad_norm": 0.6904237866401672,
+      "learning_rate": 2.8223463687150837e-05,
+      "loss": 0.0881,
+      "step": 13726
+    },
+    {
+      "epoch": 38.34357541899441,
+      "grad_norm": 1.140262484550476,
+      "learning_rate": 2.8223091247672254e-05,
+      "loss": 0.0847,
+      "step": 13727
+    },
+    {
+      "epoch": 38.3463687150838,
+      "grad_norm": 0.8256446123123169,
+      "learning_rate": 2.822271880819367e-05,
+      "loss": 0.0815,
+      "step": 13728
+    },
+    {
+      "epoch": 38.349162011173185,
+      "grad_norm": 1.4070671796798706,
+      "learning_rate": 2.8222346368715083e-05,
+      "loss": 0.092,
+      "step": 13729
+    },
+    {
+      "epoch": 38.35195530726257,
+      "grad_norm": 0.3839598298072815,
+      "learning_rate": 2.82219739292365e-05,
+      "loss": 0.1188,
+      "step": 13730
+    },
+    {
+      "epoch": 38.35474860335196,
+      "grad_norm": 0.4576537609100342,
+      "learning_rate": 2.8221601489757913e-05,
+      "loss": 0.1068,
+      "step": 13731
+    },
+    {
+      "epoch": 38.357541899441344,
+      "grad_norm": 0.454610675573349,
+      "learning_rate": 2.822122905027933e-05,
+      "loss": 0.117,
+      "step": 13732
+    },
+    {
+      "epoch": 38.36033519553072,
+      "grad_norm": 0.3502713143825531,
+      "learning_rate": 2.8220856610800746e-05,
+      "loss": 0.1086,
+      "step": 13733
+    },
+    {
+      "epoch": 38.36312849162011,
+      "grad_norm": 0.4660640060901642,
+      "learning_rate": 2.8220484171322163e-05,
+      "loss": 0.101,
+      "step": 13734
+    },
+    {
+      "epoch": 38.365921787709496,
+      "grad_norm": 0.5614672303199768,
+      "learning_rate": 2.8220111731843576e-05,
+      "loss": 0.1131,
+      "step": 13735
+    },
+    {
+      "epoch": 38.36871508379888,
+      "grad_norm": 0.5471348166465759,
+      "learning_rate": 2.821973929236499e-05,
+      "loss": 0.1126,
+      "step": 13736
+    },
+    {
+      "epoch": 38.37150837988827,
+      "grad_norm": 0.4048161208629608,
+      "learning_rate": 2.8219366852886406e-05,
+      "loss": 0.1018,
+      "step": 13737
+    },
+    {
+      "epoch": 38.374301675977655,
+      "grad_norm": 0.4353809952735901,
+      "learning_rate": 2.8218994413407822e-05,
+      "loss": 0.0969,
+      "step": 13738
+    },
+    {
+      "epoch": 38.37709497206704,
+      "grad_norm": 0.6568073034286499,
+      "learning_rate": 2.821862197392924e-05,
+      "loss": 0.0982,
+      "step": 13739
+    },
+    {
+      "epoch": 38.37988826815643,
+      "grad_norm": 0.6703651547431946,
+      "learning_rate": 2.8218249534450652e-05,
+      "loss": 0.0996,
+      "step": 13740
+    },
+    {
+      "epoch": 38.38268156424581,
+      "grad_norm": 0.5202404260635376,
+      "learning_rate": 2.821787709497207e-05,
+      "loss": 0.077,
+      "step": 13741
+    },
+    {
+      "epoch": 38.385474860335194,
+      "grad_norm": 1.2995223999023438,
+      "learning_rate": 2.8217504655493485e-05,
+      "loss": 0.0766,
+      "step": 13742
+    },
+    {
+      "epoch": 38.38826815642458,
+      "grad_norm": 0.3537974953651428,
+      "learning_rate": 2.8217132216014898e-05,
+      "loss": 0.0695,
+      "step": 13743
+    },
+    {
+      "epoch": 38.391061452513966,
+      "grad_norm": 0.6345534324645996,
+      "learning_rate": 2.821675977653631e-05,
+      "loss": 0.0658,
+      "step": 13744
+    },
+    {
+      "epoch": 38.39385474860335,
+      "grad_norm": 0.5241303443908691,
+      "learning_rate": 2.8216387337057728e-05,
+      "loss": 0.0929,
+      "step": 13745
+    },
+    {
+      "epoch": 38.39664804469274,
+      "grad_norm": 0.687155544757843,
+      "learning_rate": 2.8216014897579144e-05,
+      "loss": 0.0966,
+      "step": 13746
+    },
+    {
+      "epoch": 38.399441340782126,
+      "grad_norm": 0.9251924753189087,
+      "learning_rate": 2.821564245810056e-05,
+      "loss": 0.0783,
+      "step": 13747
+    },
+    {
+      "epoch": 38.40223463687151,
+      "grad_norm": 0.5184628963470459,
+      "learning_rate": 2.8215270018621977e-05,
+      "loss": 0.0647,
+      "step": 13748
+    },
+    {
+      "epoch": 38.40502793296089,
+      "grad_norm": 0.8536320328712463,
+      "learning_rate": 2.8214897579143387e-05,
+      "loss": 0.0717,
+      "step": 13749
+    },
+    {
+      "epoch": 38.40782122905028,
+      "grad_norm": 0.48444679379463196,
+      "learning_rate": 2.8214525139664804e-05,
+      "loss": 0.0751,
+      "step": 13750
+    },
+    {
+      "epoch": 38.410614525139664,
+      "grad_norm": 0.9786247611045837,
+      "learning_rate": 2.821415270018622e-05,
+      "loss": 0.0753,
+      "step": 13751
+    },
+    {
+      "epoch": 38.41340782122905,
+      "grad_norm": 0.5685248374938965,
+      "learning_rate": 2.8213780260707637e-05,
+      "loss": 0.0634,
+      "step": 13752
+    },
+    {
+      "epoch": 38.41620111731844,
+      "grad_norm": 0.9955393671989441,
+      "learning_rate": 2.821340782122905e-05,
+      "loss": 0.0654,
+      "step": 13753
+    },
+    {
+      "epoch": 38.41899441340782,
+      "grad_norm": 0.8188115954399109,
+      "learning_rate": 2.8213035381750466e-05,
+      "loss": 0.0809,
+      "step": 13754
+    },
+    {
+      "epoch": 38.42178770949721,
+      "grad_norm": 0.38793066143989563,
+      "learning_rate": 2.8212662942271883e-05,
+      "loss": 0.1298,
+      "step": 13755
+    },
+    {
+      "epoch": 38.42458100558659,
+      "grad_norm": 0.7562322020530701,
+      "learning_rate": 2.8212290502793296e-05,
+      "loss": 0.1211,
+      "step": 13756
+    },
+    {
+      "epoch": 38.427374301675975,
+      "grad_norm": 0.9052409529685974,
+      "learning_rate": 2.8211918063314713e-05,
+      "loss": 0.0966,
+      "step": 13757
+    },
+    {
+      "epoch": 38.43016759776536,
+      "grad_norm": 0.6667492389678955,
+      "learning_rate": 2.8211545623836126e-05,
+      "loss": 0.1246,
+      "step": 13758
+    },
+    {
+      "epoch": 38.43296089385475,
+      "grad_norm": 0.618563711643219,
+      "learning_rate": 2.8211173184357542e-05,
+      "loss": 0.1239,
+      "step": 13759
+    },
+    {
+      "epoch": 38.435754189944134,
+      "grad_norm": 0.48272815346717834,
+      "learning_rate": 2.821080074487896e-05,
+      "loss": 0.0999,
+      "step": 13760
+    },
+    {
+      "epoch": 38.43854748603352,
+      "grad_norm": 0.4821487367153168,
+      "learning_rate": 2.8210428305400375e-05,
+      "loss": 0.1096,
+      "step": 13761
+    },
+    {
+      "epoch": 38.44134078212291,
+      "grad_norm": 0.41666385531425476,
+      "learning_rate": 2.821005586592179e-05,
+      "loss": 0.0921,
+      "step": 13762
+    },
+    {
+      "epoch": 38.444134078212294,
+      "grad_norm": 0.5778366327285767,
+      "learning_rate": 2.82096834264432e-05,
+      "loss": 0.1046,
+      "step": 13763
+    },
+    {
+      "epoch": 38.44692737430167,
+      "grad_norm": 0.4795728027820587,
+      "learning_rate": 2.8209310986964618e-05,
+      "loss": 0.0929,
+      "step": 13764
+    },
+    {
+      "epoch": 38.44972067039106,
+      "grad_norm": 0.627427875995636,
+      "learning_rate": 2.8208938547486035e-05,
+      "loss": 0.0938,
+      "step": 13765
+    },
+    {
+      "epoch": 38.452513966480446,
+      "grad_norm": 1.396582841873169,
+      "learning_rate": 2.820856610800745e-05,
+      "loss": 0.0865,
+      "step": 13766
+    },
+    {
+      "epoch": 38.45530726256983,
+      "grad_norm": 0.6990799307823181,
+      "learning_rate": 2.8208193668528864e-05,
+      "loss": 0.0727,
+      "step": 13767
+    },
+    {
+      "epoch": 38.45810055865922,
+      "grad_norm": 0.41566789150238037,
+      "learning_rate": 2.820782122905028e-05,
+      "loss": 0.0746,
+      "step": 13768
+    },
+    {
+      "epoch": 38.460893854748605,
+      "grad_norm": 1.025338888168335,
+      "learning_rate": 2.8207448789571694e-05,
+      "loss": 0.0679,
+      "step": 13769
+    },
+    {
+      "epoch": 38.46368715083799,
+      "grad_norm": 0.9662831425666809,
+      "learning_rate": 2.820707635009311e-05,
+      "loss": 0.084,
+      "step": 13770
+    },
+    {
+      "epoch": 38.46648044692738,
+      "grad_norm": 0.534064769744873,
+      "learning_rate": 2.8206703910614524e-05,
+      "loss": 0.067,
+      "step": 13771
+    },
+    {
+      "epoch": 38.46927374301676,
+      "grad_norm": 0.6542330384254456,
+      "learning_rate": 2.820633147113594e-05,
+      "loss": 0.0653,
+      "step": 13772
+    },
+    {
+      "epoch": 38.47206703910614,
+      "grad_norm": 0.7994427680969238,
+      "learning_rate": 2.8205959031657357e-05,
+      "loss": 0.0705,
+      "step": 13773
+    },
+    {
+      "epoch": 38.47486033519553,
+      "grad_norm": 0.8987044095993042,
+      "learning_rate": 2.8205586592178773e-05,
+      "loss": 0.0683,
+      "step": 13774
+    },
+    {
+      "epoch": 38.477653631284916,
+      "grad_norm": 1.035936951637268,
+      "learning_rate": 2.820521415270019e-05,
+      "loss": 0.0574,
+      "step": 13775
+    },
+    {
+      "epoch": 38.4804469273743,
+      "grad_norm": 0.48982343077659607,
+      "learning_rate": 2.82048417132216e-05,
+      "loss": 0.0704,
+      "step": 13776
+    },
+    {
+      "epoch": 38.48324022346369,
+      "grad_norm": 1.3583160638809204,
+      "learning_rate": 2.8204469273743016e-05,
+      "loss": 0.0734,
+      "step": 13777
+    },
+    {
+      "epoch": 38.486033519553075,
+      "grad_norm": 2.6069440841674805,
+      "learning_rate": 2.8204096834264433e-05,
+      "loss": 0.0759,
+      "step": 13778
+    },
+    {
+      "epoch": 38.48882681564246,
+      "grad_norm": 0.9855948686599731,
+      "learning_rate": 2.820372439478585e-05,
+      "loss": 0.0954,
+      "step": 13779
+    },
+    {
+      "epoch": 38.49162011173184,
+      "grad_norm": 0.4216324985027313,
+      "learning_rate": 2.8203351955307262e-05,
+      "loss": 0.1209,
+      "step": 13780
+    },
+    {
+      "epoch": 38.49441340782123,
+      "grad_norm": 0.5342094898223877,
+      "learning_rate": 2.820297951582868e-05,
+      "loss": 0.1108,
+      "step": 13781
+    },
+    {
+      "epoch": 38.497206703910614,
+      "grad_norm": 0.45024871826171875,
+      "learning_rate": 2.8202607076350095e-05,
+      "loss": 0.109,
+      "step": 13782
+    },
+    {
+      "epoch": 38.5,
+      "grad_norm": 0.807569682598114,
+      "learning_rate": 2.820223463687151e-05,
+      "loss": 0.1157,
+      "step": 13783
+    },
+    {
+      "epoch": 38.502793296089386,
+      "grad_norm": 0.4081205129623413,
+      "learning_rate": 2.8201862197392925e-05,
+      "loss": 0.1132,
+      "step": 13784
+    },
+    {
+      "epoch": 38.50558659217877,
+      "grad_norm": 0.5807958245277405,
+      "learning_rate": 2.8201489757914338e-05,
+      "loss": 0.0973,
+      "step": 13785
+    },
+    {
+      "epoch": 38.50837988826816,
+      "grad_norm": 0.4793674647808075,
+      "learning_rate": 2.8201117318435755e-05,
+      "loss": 0.0997,
+      "step": 13786
+    },
+    {
+      "epoch": 38.51117318435754,
+      "grad_norm": 0.5050674676895142,
+      "learning_rate": 2.820074487895717e-05,
+      "loss": 0.0974,
+      "step": 13787
+    },
+    {
+      "epoch": 38.513966480446925,
+      "grad_norm": 0.39423036575317383,
+      "learning_rate": 2.8200372439478588e-05,
+      "loss": 0.0904,
+      "step": 13788
+    },
+    {
+      "epoch": 38.51675977653631,
+      "grad_norm": 1.1291871070861816,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 0.104,
+      "step": 13789
+    },
+    {
+      "epoch": 38.5195530726257,
+      "grad_norm": 0.47780680656433105,
+      "learning_rate": 2.8199627560521414e-05,
+      "loss": 0.0994,
+      "step": 13790
+    },
+    {
+      "epoch": 38.522346368715084,
+      "grad_norm": 1.103088140487671,
+      "learning_rate": 2.819925512104283e-05,
+      "loss": 0.0869,
+      "step": 13791
+    },
+    {
+      "epoch": 38.52513966480447,
+      "grad_norm": 0.7918819189071655,
+      "learning_rate": 2.8198882681564247e-05,
+      "loss": 0.0948,
+      "step": 13792
+    },
+    {
+      "epoch": 38.52793296089386,
+      "grad_norm": 1.0708072185516357,
+      "learning_rate": 2.8198510242085664e-05,
+      "loss": 0.0765,
+      "step": 13793
+    },
+    {
+      "epoch": 38.53072625698324,
+      "grad_norm": 0.7510428428649902,
+      "learning_rate": 2.8198137802607077e-05,
+      "loss": 0.0889,
+      "step": 13794
+    },
+    {
+      "epoch": 38.53351955307262,
+      "grad_norm": 0.503798246383667,
+      "learning_rate": 2.8197765363128493e-05,
+      "loss": 0.0804,
+      "step": 13795
+    },
+    {
+      "epoch": 38.53631284916201,
+      "grad_norm": 0.4742206633090973,
+      "learning_rate": 2.8197392923649907e-05,
+      "loss": 0.081,
+      "step": 13796
+    },
+    {
+      "epoch": 38.539106145251395,
+      "grad_norm": 1.0509443283081055,
+      "learning_rate": 2.8197020484171323e-05,
+      "loss": 0.076,
+      "step": 13797
+    },
+    {
+      "epoch": 38.54189944134078,
+      "grad_norm": 1.1594491004943848,
+      "learning_rate": 2.8196648044692736e-05,
+      "loss": 0.0838,
+      "step": 13798
+    },
+    {
+      "epoch": 38.54469273743017,
+      "grad_norm": 0.6252346634864807,
+      "learning_rate": 2.8196275605214153e-05,
+      "loss": 0.0907,
+      "step": 13799
+    },
+    {
+      "epoch": 38.547486033519554,
+      "grad_norm": 8.080602645874023,
+      "learning_rate": 2.819590316573557e-05,
+      "loss": 0.0713,
+      "step": 13800
+    },
+    {
+      "epoch": 38.55027932960894,
+      "grad_norm": 0.5671552419662476,
+      "learning_rate": 2.8195530726256986e-05,
+      "loss": 0.0737,
+      "step": 13801
+    },
+    {
+      "epoch": 38.55307262569833,
+      "grad_norm": 0.8001420497894287,
+      "learning_rate": 2.8195158286778402e-05,
+      "loss": 0.0651,
+      "step": 13802
+    },
+    {
+      "epoch": 38.555865921787706,
+      "grad_norm": 0.5933485627174377,
+      "learning_rate": 2.8194785847299812e-05,
+      "loss": 0.0728,
+      "step": 13803
+    },
+    {
+      "epoch": 38.55865921787709,
+      "grad_norm": 1.5367540121078491,
+      "learning_rate": 2.819441340782123e-05,
+      "loss": 0.0816,
+      "step": 13804
+    },
+    {
+      "epoch": 38.56145251396648,
+      "grad_norm": 0.5064272880554199,
+      "learning_rate": 2.8194040968342645e-05,
+      "loss": 0.1266,
+      "step": 13805
+    },
+    {
+      "epoch": 38.564245810055866,
+      "grad_norm": 0.4853374660015106,
+      "learning_rate": 2.8193668528864062e-05,
+      "loss": 0.134,
+      "step": 13806
+    },
+    {
+      "epoch": 38.56703910614525,
+      "grad_norm": 0.6735977530479431,
+      "learning_rate": 2.8193296089385475e-05,
+      "loss": 0.1075,
+      "step": 13807
+    },
+    {
+      "epoch": 38.56983240223464,
+      "grad_norm": 0.6326329708099365,
+      "learning_rate": 2.819292364990689e-05,
+      "loss": 0.1032,
+      "step": 13808
+    },
+    {
+      "epoch": 38.572625698324025,
+      "grad_norm": 0.4763261675834656,
+      "learning_rate": 2.8192551210428305e-05,
+      "loss": 0.1095,
+      "step": 13809
+    },
+    {
+      "epoch": 38.57541899441341,
+      "grad_norm": 0.5325655341148376,
+      "learning_rate": 2.819217877094972e-05,
+      "loss": 0.0995,
+      "step": 13810
+    },
+    {
+      "epoch": 38.57821229050279,
+      "grad_norm": 0.8144444823265076,
+      "learning_rate": 2.8191806331471138e-05,
+      "loss": 0.1076,
+      "step": 13811
+    },
+    {
+      "epoch": 38.58100558659218,
+      "grad_norm": 1.1911715269088745,
+      "learning_rate": 2.819143389199255e-05,
+      "loss": 0.1014,
+      "step": 13812
+    },
+    {
+      "epoch": 38.58379888268156,
+      "grad_norm": 0.39264824986457825,
+      "learning_rate": 2.8191061452513967e-05,
+      "loss": 0.0828,
+      "step": 13813
+    },
+    {
+      "epoch": 38.58659217877095,
+      "grad_norm": 0.7393624782562256,
+      "learning_rate": 2.8190689013035384e-05,
+      "loss": 0.1007,
+      "step": 13814
+    },
+    {
+      "epoch": 38.589385474860336,
+      "grad_norm": 0.6280959248542786,
+      "learning_rate": 2.81903165735568e-05,
+      "loss": 0.0924,
+      "step": 13815
+    },
+    {
+      "epoch": 38.59217877094972,
+      "grad_norm": 0.6185868978500366,
+      "learning_rate": 2.818994413407821e-05,
+      "loss": 0.0936,
+      "step": 13816
+    },
+    {
+      "epoch": 38.59497206703911,
+      "grad_norm": 0.678554356098175,
+      "learning_rate": 2.8189571694599627e-05,
+      "loss": 0.0812,
+      "step": 13817
+    },
+    {
+      "epoch": 38.59776536312849,
+      "grad_norm": 0.6086649894714355,
+      "learning_rate": 2.8189199255121043e-05,
+      "loss": 0.0808,
+      "step": 13818
+    },
+    {
+      "epoch": 38.600558659217874,
+      "grad_norm": 0.5977902412414551,
+      "learning_rate": 2.818882681564246e-05,
+      "loss": 0.0703,
+      "step": 13819
+    },
+    {
+      "epoch": 38.60335195530726,
+      "grad_norm": 0.729066014289856,
+      "learning_rate": 2.8188454376163873e-05,
+      "loss": 0.0817,
+      "step": 13820
+    },
+    {
+      "epoch": 38.60614525139665,
+      "grad_norm": 0.4781682789325714,
+      "learning_rate": 2.818808193668529e-05,
+      "loss": 0.0712,
+      "step": 13821
+    },
+    {
+      "epoch": 38.608938547486034,
+      "grad_norm": 0.64860999584198,
+      "learning_rate": 2.8187709497206706e-05,
+      "loss": 0.0797,
+      "step": 13822
+    },
+    {
+      "epoch": 38.61173184357542,
+      "grad_norm": 1.1487290859222412,
+      "learning_rate": 2.818733705772812e-05,
+      "loss": 0.0681,
+      "step": 13823
+    },
+    {
+      "epoch": 38.614525139664806,
+      "grad_norm": 0.7879559993743896,
+      "learning_rate": 2.8186964618249536e-05,
+      "loss": 0.0826,
+      "step": 13824
+    },
+    {
+      "epoch": 38.61731843575419,
+      "grad_norm": 0.5300081372261047,
+      "learning_rate": 2.818659217877095e-05,
+      "loss": 0.0761,
+      "step": 13825
+    },
+    {
+      "epoch": 38.62011173184357,
+      "grad_norm": 0.8959120512008667,
+      "learning_rate": 2.8186219739292365e-05,
+      "loss": 0.0746,
+      "step": 13826
+    },
+    {
+      "epoch": 38.62290502793296,
+      "grad_norm": 1.5305004119873047,
+      "learning_rate": 2.8185847299813782e-05,
+      "loss": 0.0786,
+      "step": 13827
+    },
+    {
+      "epoch": 38.625698324022345,
+      "grad_norm": 0.9396858215332031,
+      "learning_rate": 2.81854748603352e-05,
+      "loss": 0.0861,
+      "step": 13828
+    },
+    {
+      "epoch": 38.62849162011173,
+      "grad_norm": 6.650175094604492,
+      "learning_rate": 2.8185102420856608e-05,
+      "loss": 0.0934,
+      "step": 13829
+    },
+    {
+      "epoch": 38.63128491620112,
+      "grad_norm": 0.46983224153518677,
+      "learning_rate": 2.8184729981378025e-05,
+      "loss": 0.1453,
+      "step": 13830
+    },
+    {
+      "epoch": 38.634078212290504,
+      "grad_norm": 0.42784979939460754,
+      "learning_rate": 2.818435754189944e-05,
+      "loss": 0.1195,
+      "step": 13831
+    },
+    {
+      "epoch": 38.63687150837989,
+      "grad_norm": 0.42154332995414734,
+      "learning_rate": 2.8183985102420858e-05,
+      "loss": 0.1004,
+      "step": 13832
+    },
+    {
+      "epoch": 38.63966480446928,
+      "grad_norm": 0.3716309070587158,
+      "learning_rate": 2.8183612662942274e-05,
+      "loss": 0.1004,
+      "step": 13833
+    },
+    {
+      "epoch": 38.642458100558656,
+      "grad_norm": 0.4281313419342041,
+      "learning_rate": 2.8183240223463688e-05,
+      "loss": 0.1035,
+      "step": 13834
+    },
+    {
+      "epoch": 38.64525139664804,
+      "grad_norm": 0.7026533484458923,
+      "learning_rate": 2.8182867783985104e-05,
+      "loss": 0.1159,
+      "step": 13835
+    },
+    {
+      "epoch": 38.64804469273743,
+      "grad_norm": 0.47632020711898804,
+      "learning_rate": 2.8182495344506517e-05,
+      "loss": 0.1036,
+      "step": 13836
+    },
+    {
+      "epoch": 38.650837988826815,
+      "grad_norm": 0.3790722191333771,
+      "learning_rate": 2.8182122905027934e-05,
+      "loss": 0.0978,
+      "step": 13837
+    },
+    {
+      "epoch": 38.6536312849162,
+      "grad_norm": 1.067626714706421,
+      "learning_rate": 2.8181750465549347e-05,
+      "loss": 0.0944,
+      "step": 13838
+    },
+    {
+      "epoch": 38.65642458100559,
+      "grad_norm": 0.43713054060935974,
+      "learning_rate": 2.8181378026070763e-05,
+      "loss": 0.0967,
+      "step": 13839
+    },
+    {
+      "epoch": 38.659217877094974,
+      "grad_norm": 0.4869667887687683,
+      "learning_rate": 2.818100558659218e-05,
+      "loss": 0.1297,
+      "step": 13840
+    },
+    {
+      "epoch": 38.66201117318436,
+      "grad_norm": 0.6247203946113586,
+      "learning_rate": 2.8180633147113597e-05,
+      "loss": 0.083,
+      "step": 13841
+    },
+    {
+      "epoch": 38.66480446927374,
+      "grad_norm": 0.5846632122993469,
+      "learning_rate": 2.8180260707635013e-05,
+      "loss": 0.0756,
+      "step": 13842
+    },
+    {
+      "epoch": 38.667597765363126,
+      "grad_norm": 0.569110631942749,
+      "learning_rate": 2.8179888268156423e-05,
+      "loss": 0.0884,
+      "step": 13843
+    },
+    {
+      "epoch": 38.67039106145251,
+      "grad_norm": 0.5299766063690186,
+      "learning_rate": 2.817951582867784e-05,
+      "loss": 0.079,
+      "step": 13844
+    },
+    {
+      "epoch": 38.6731843575419,
+      "grad_norm": 0.4069921672344208,
+      "learning_rate": 2.8179143389199256e-05,
+      "loss": 0.0792,
+      "step": 13845
+    },
+    {
+      "epoch": 38.675977653631286,
+      "grad_norm": 0.5567750334739685,
+      "learning_rate": 2.8178770949720672e-05,
+      "loss": 0.0842,
+      "step": 13846
+    },
+    {
+      "epoch": 38.67877094972067,
+      "grad_norm": 1.068578839302063,
+      "learning_rate": 2.8178398510242086e-05,
+      "loss": 0.0707,
+      "step": 13847
+    },
+    {
+      "epoch": 38.68156424581006,
+      "grad_norm": 0.6553998589515686,
+      "learning_rate": 2.8178026070763502e-05,
+      "loss": 0.0668,
+      "step": 13848
+    },
+    {
+      "epoch": 38.684357541899445,
+      "grad_norm": 1.1579440832138062,
+      "learning_rate": 2.8177653631284915e-05,
+      "loss": 0.0767,
+      "step": 13849
+    },
+    {
+      "epoch": 38.687150837988824,
+      "grad_norm": 0.6024009585380554,
+      "learning_rate": 2.8177281191806332e-05,
+      "loss": 0.0665,
+      "step": 13850
+    },
+    {
+      "epoch": 38.68994413407821,
+      "grad_norm": 0.5937240719795227,
+      "learning_rate": 2.817690875232775e-05,
+      "loss": 0.0685,
+      "step": 13851
+    },
+    {
+      "epoch": 38.6927374301676,
+      "grad_norm": 0.9592479467391968,
+      "learning_rate": 2.817653631284916e-05,
+      "loss": 0.0723,
+      "step": 13852
+    },
+    {
+      "epoch": 38.69553072625698,
+      "grad_norm": 0.8652563095092773,
+      "learning_rate": 2.8176163873370578e-05,
+      "loss": 0.0864,
+      "step": 13853
+    },
+    {
+      "epoch": 38.69832402234637,
+      "grad_norm": 1.0788688659667969,
+      "learning_rate": 2.8175791433891995e-05,
+      "loss": 0.0815,
+      "step": 13854
+    },
+    {
+      "epoch": 38.701117318435756,
+      "grad_norm": 0.6012893319129944,
+      "learning_rate": 2.817541899441341e-05,
+      "loss": 0.125,
+      "step": 13855
+    },
+    {
+      "epoch": 38.70391061452514,
+      "grad_norm": 0.5100256204605103,
+      "learning_rate": 2.817504655493482e-05,
+      "loss": 0.1371,
+      "step": 13856
+    },
+    {
+      "epoch": 38.70670391061452,
+      "grad_norm": 0.3431745171546936,
+      "learning_rate": 2.8174674115456237e-05,
+      "loss": 0.1087,
+      "step": 13857
+    },
+    {
+      "epoch": 38.70949720670391,
+      "grad_norm": 0.6963807940483093,
+      "learning_rate": 2.8174301675977654e-05,
+      "loss": 0.1034,
+      "step": 13858
+    },
+    {
+      "epoch": 38.712290502793294,
+      "grad_norm": 0.7737045884132385,
+      "learning_rate": 2.817392923649907e-05,
+      "loss": 0.1246,
+      "step": 13859
+    },
+    {
+      "epoch": 38.71508379888268,
+      "grad_norm": 0.7375789284706116,
+      "learning_rate": 2.8173556797020487e-05,
+      "loss": 0.1079,
+      "step": 13860
+    },
+    {
+      "epoch": 38.71787709497207,
+      "grad_norm": 0.6823984980583191,
+      "learning_rate": 2.81731843575419e-05,
+      "loss": 0.1127,
+      "step": 13861
+    },
+    {
+      "epoch": 38.720670391061454,
+      "grad_norm": 0.37735646963119507,
+      "learning_rate": 2.8172811918063317e-05,
+      "loss": 0.0995,
+      "step": 13862
+    },
+    {
+      "epoch": 38.72346368715084,
+      "grad_norm": 0.5369113683700562,
+      "learning_rate": 2.817243947858473e-05,
+      "loss": 0.0871,
+      "step": 13863
+    },
+    {
+      "epoch": 38.726256983240226,
+      "grad_norm": 0.7061443328857422,
+      "learning_rate": 2.8172067039106146e-05,
+      "loss": 0.0957,
+      "step": 13864
+    },
+    {
+      "epoch": 38.729050279329606,
+      "grad_norm": 0.6290752291679382,
+      "learning_rate": 2.817169459962756e-05,
+      "loss": 0.1146,
+      "step": 13865
+    },
+    {
+      "epoch": 38.73184357541899,
+      "grad_norm": 0.6431031823158264,
+      "learning_rate": 2.8171322160148976e-05,
+      "loss": 0.1022,
+      "step": 13866
+    },
+    {
+      "epoch": 38.73463687150838,
+      "grad_norm": 0.5127524733543396,
+      "learning_rate": 2.8170949720670393e-05,
+      "loss": 0.0906,
+      "step": 13867
+    },
+    {
+      "epoch": 38.737430167597765,
+      "grad_norm": 0.6714181303977966,
+      "learning_rate": 2.817057728119181e-05,
+      "loss": 0.0906,
+      "step": 13868
+    },
+    {
+      "epoch": 38.74022346368715,
+      "grad_norm": 0.6430686116218567,
+      "learning_rate": 2.8170204841713222e-05,
+      "loss": 0.0862,
+      "step": 13869
+    },
+    {
+      "epoch": 38.74301675977654,
+      "grad_norm": 0.5403818488121033,
+      "learning_rate": 2.8169832402234635e-05,
+      "loss": 0.0742,
+      "step": 13870
+    },
+    {
+      "epoch": 38.745810055865924,
+      "grad_norm": 0.678941547870636,
+      "learning_rate": 2.8169459962756052e-05,
+      "loss": 0.0797,
+      "step": 13871
+    },
+    {
+      "epoch": 38.74860335195531,
+      "grad_norm": 0.6076549291610718,
+      "learning_rate": 2.816908752327747e-05,
+      "loss": 0.0872,
+      "step": 13872
+    },
+    {
+      "epoch": 38.75139664804469,
+      "grad_norm": 1.3735816478729248,
+      "learning_rate": 2.8168715083798885e-05,
+      "loss": 0.1057,
+      "step": 13873
+    },
+    {
+      "epoch": 38.754189944134076,
+      "grad_norm": 1.2291306257247925,
+      "learning_rate": 2.8168342644320298e-05,
+      "loss": 0.0818,
+      "step": 13874
+    },
+    {
+      "epoch": 38.75698324022346,
+      "grad_norm": 0.5107596516609192,
+      "learning_rate": 2.8167970204841715e-05,
+      "loss": 0.064,
+      "step": 13875
+    },
+    {
+      "epoch": 38.75977653631285,
+      "grad_norm": 0.7181755900382996,
+      "learning_rate": 2.8167597765363128e-05,
+      "loss": 0.0756,
+      "step": 13876
+    },
+    {
+      "epoch": 38.762569832402235,
+      "grad_norm": 0.790803074836731,
+      "learning_rate": 2.8167225325884544e-05,
+      "loss": 0.092,
+      "step": 13877
+    },
+    {
+      "epoch": 38.76536312849162,
+      "grad_norm": 0.9664952158927917,
+      "learning_rate": 2.816685288640596e-05,
+      "loss": 0.0977,
+      "step": 13878
+    },
+    {
+      "epoch": 38.76815642458101,
+      "grad_norm": 2.558032989501953,
+      "learning_rate": 2.8166480446927374e-05,
+      "loss": 0.0898,
+      "step": 13879
+    },
+    {
+      "epoch": 38.770949720670394,
+      "grad_norm": 0.33781322836875916,
+      "learning_rate": 2.816610800744879e-05,
+      "loss": 0.1236,
+      "step": 13880
+    },
+    {
+      "epoch": 38.773743016759774,
+      "grad_norm": 0.3987351357936859,
+      "learning_rate": 2.8165735567970207e-05,
+      "loss": 0.1282,
+      "step": 13881
+    },
+    {
+      "epoch": 38.77653631284916,
+      "grad_norm": 0.7068341374397278,
+      "learning_rate": 2.8165363128491624e-05,
+      "loss": 0.1016,
+      "step": 13882
+    },
+    {
+      "epoch": 38.779329608938546,
+      "grad_norm": 0.35489025712013245,
+      "learning_rate": 2.8164990689013033e-05,
+      "loss": 0.1154,
+      "step": 13883
+    },
+    {
+      "epoch": 38.78212290502793,
+      "grad_norm": 0.7438069581985474,
+      "learning_rate": 2.816461824953445e-05,
+      "loss": 0.1131,
+      "step": 13884
+    },
+    {
+      "epoch": 38.78491620111732,
+      "grad_norm": 0.9405038952827454,
+      "learning_rate": 2.8164245810055866e-05,
+      "loss": 0.117,
+      "step": 13885
+    },
+    {
+      "epoch": 38.787709497206706,
+      "grad_norm": 0.49241313338279724,
+      "learning_rate": 2.8163873370577283e-05,
+      "loss": 0.1054,
+      "step": 13886
+    },
+    {
+      "epoch": 38.79050279329609,
+      "grad_norm": 0.4505404531955719,
+      "learning_rate": 2.81635009310987e-05,
+      "loss": 0.0912,
+      "step": 13887
+    },
+    {
+      "epoch": 38.79329608938548,
+      "grad_norm": 0.539735734462738,
+      "learning_rate": 2.8163128491620113e-05,
+      "loss": 0.0984,
+      "step": 13888
+    },
+    {
+      "epoch": 38.79608938547486,
+      "grad_norm": 0.45133382081985474,
+      "learning_rate": 2.8162756052141526e-05,
+      "loss": 0.0825,
+      "step": 13889
+    },
+    {
+      "epoch": 38.798882681564244,
+      "grad_norm": 1.4024449586868286,
+      "learning_rate": 2.8162383612662942e-05,
+      "loss": 0.097,
+      "step": 13890
+    },
+    {
+      "epoch": 38.80167597765363,
+      "grad_norm": 0.7537744641304016,
+      "learning_rate": 2.816201117318436e-05,
+      "loss": 0.0805,
+      "step": 13891
+    },
+    {
+      "epoch": 38.80446927374302,
+      "grad_norm": 0.401835173368454,
+      "learning_rate": 2.8161638733705772e-05,
+      "loss": 0.0813,
+      "step": 13892
+    },
+    {
+      "epoch": 38.8072625698324,
+      "grad_norm": 0.42641526460647583,
+      "learning_rate": 2.816126629422719e-05,
+      "loss": 0.0802,
+      "step": 13893
+    },
+    {
+      "epoch": 38.81005586592179,
+      "grad_norm": 0.6978179216384888,
+      "learning_rate": 2.8160893854748605e-05,
+      "loss": 0.0812,
+      "step": 13894
+    },
+    {
+      "epoch": 38.812849162011176,
+      "grad_norm": 0.34789130091667175,
+      "learning_rate": 2.816052141527002e-05,
+      "loss": 0.0638,
+      "step": 13895
+    },
+    {
+      "epoch": 38.815642458100555,
+      "grad_norm": 0.6026569604873657,
+      "learning_rate": 2.8160148975791435e-05,
+      "loss": 0.064,
+      "step": 13896
+    },
+    {
+      "epoch": 38.81843575418994,
+      "grad_norm": 0.44745948910713196,
+      "learning_rate": 2.8159776536312848e-05,
+      "loss": 0.0839,
+      "step": 13897
+    },
+    {
+      "epoch": 38.82122905027933,
+      "grad_norm": 0.6541919112205505,
+      "learning_rate": 2.8159404096834265e-05,
+      "loss": 0.0789,
+      "step": 13898
+    },
+    {
+      "epoch": 38.824022346368714,
+      "grad_norm": 0.5804952383041382,
+      "learning_rate": 2.815903165735568e-05,
+      "loss": 0.0668,
+      "step": 13899
+    },
+    {
+      "epoch": 38.8268156424581,
+      "grad_norm": 0.7965481877326965,
+      "learning_rate": 2.8158659217877098e-05,
+      "loss": 0.0809,
+      "step": 13900
+    },
+    {
+      "epoch": 38.82960893854749,
+      "grad_norm": 0.732953667640686,
+      "learning_rate": 2.815828677839851e-05,
+      "loss": 0.0766,
+      "step": 13901
+    },
+    {
+      "epoch": 38.832402234636874,
+      "grad_norm": 0.8889634013175964,
+      "learning_rate": 2.8157914338919927e-05,
+      "loss": 0.0866,
+      "step": 13902
+    },
+    {
+      "epoch": 38.83519553072626,
+      "grad_norm": 0.752293050289154,
+      "learning_rate": 2.815754189944134e-05,
+      "loss": 0.074,
+      "step": 13903
+    },
+    {
+      "epoch": 38.83798882681564,
+      "grad_norm": 1.1037317514419556,
+      "learning_rate": 2.8157169459962757e-05,
+      "loss": 0.088,
+      "step": 13904
+    },
+    {
+      "epoch": 38.840782122905026,
+      "grad_norm": 0.6492994427680969,
+      "learning_rate": 2.8156797020484173e-05,
+      "loss": 0.1333,
+      "step": 13905
+    },
+    {
+      "epoch": 38.84357541899441,
+      "grad_norm": 0.7365188002586365,
+      "learning_rate": 2.8156424581005587e-05,
+      "loss": 0.1376,
+      "step": 13906
+    },
+    {
+      "epoch": 38.8463687150838,
+      "grad_norm": NaN,
+      "learning_rate": 2.8156424581005587e-05,
+      "loss": 0.1063,
+      "step": 13907
+    },
+    {
+      "epoch": 38.849162011173185,
+      "grad_norm": 0.850227415561676,
+      "learning_rate": 2.8156052141527003e-05,
+      "loss": 0.1292,
+      "step": 13908
+    },
+    {
+      "epoch": 38.85195530726257,
+      "grad_norm": 0.6937223672866821,
+      "learning_rate": 2.815567970204842e-05,
+      "loss": 0.1142,
+      "step": 13909
+    },
+    {
+      "epoch": 38.85474860335196,
+      "grad_norm": 0.5119539499282837,
+      "learning_rate": 2.8155307262569833e-05,
+      "loss": 0.1076,
+      "step": 13910
+    },
+    {
+      "epoch": 38.857541899441344,
+      "grad_norm": 0.6546431183815002,
+      "learning_rate": 2.8154934823091246e-05,
+      "loss": 0.1022,
+      "step": 13911
+    },
+    {
+      "epoch": 38.86033519553072,
+      "grad_norm": 1.2861254215240479,
+      "learning_rate": 2.8154562383612663e-05,
+      "loss": 0.0916,
+      "step": 13912
+    },
+    {
+      "epoch": 38.86312849162011,
+      "grad_norm": 0.3550238609313965,
+      "learning_rate": 2.815418994413408e-05,
+      "loss": 0.093,
+      "step": 13913
+    },
+    {
+      "epoch": 38.865921787709496,
+      "grad_norm": 0.595964252948761,
+      "learning_rate": 2.8153817504655496e-05,
+      "loss": 0.0908,
+      "step": 13914
+    },
+    {
+      "epoch": 38.86871508379888,
+      "grad_norm": 0.5790852308273315,
+      "learning_rate": 2.815344506517691e-05,
+      "loss": 0.1004,
+      "step": 13915
+    },
+    {
+      "epoch": 38.87150837988827,
+      "grad_norm": 0.5495722889900208,
+      "learning_rate": 2.8153072625698325e-05,
+      "loss": 0.0834,
+      "step": 13916
+    },
+    {
+      "epoch": 38.874301675977655,
+      "grad_norm": 0.9984388947486877,
+      "learning_rate": 2.815270018621974e-05,
+      "loss": 0.0855,
+      "step": 13917
+    },
+    {
+      "epoch": 38.87709497206704,
+      "grad_norm": 0.5843257308006287,
+      "learning_rate": 2.8152327746741155e-05,
+      "loss": 0.0896,
+      "step": 13918
+    },
+    {
+      "epoch": 38.87988826815643,
+      "grad_norm": 0.6011370420455933,
+      "learning_rate": 2.815195530726257e-05,
+      "loss": 0.0821,
+      "step": 13919
+    },
+    {
+      "epoch": 38.88268156424581,
+      "grad_norm": 0.4790153205394745,
+      "learning_rate": 2.8151582867783985e-05,
+      "loss": 0.0833,
+      "step": 13920
+    },
+    {
+      "epoch": 38.885474860335194,
+      "grad_norm": 0.6771031022071838,
+      "learning_rate": 2.81512104283054e-05,
+      "loss": 0.0998,
+      "step": 13921
+    },
+    {
+      "epoch": 38.88826815642458,
+      "grad_norm": 0.9446739554405212,
+      "learning_rate": 2.8150837988826818e-05,
+      "loss": 0.0745,
+      "step": 13922
+    },
+    {
+      "epoch": 38.891061452513966,
+      "grad_norm": 0.49410757422447205,
+      "learning_rate": 2.8150465549348234e-05,
+      "loss": 0.0633,
+      "step": 13923
+    },
+    {
+      "epoch": 38.89385474860335,
+      "grad_norm": 0.48964133858680725,
+      "learning_rate": 2.8150093109869644e-05,
+      "loss": 0.0612,
+      "step": 13924
+    },
+    {
+      "epoch": 38.89664804469274,
+      "grad_norm": 0.581567645072937,
+      "learning_rate": 2.814972067039106e-05,
+      "loss": 0.0676,
+      "step": 13925
+    },
+    {
+      "epoch": 38.899441340782126,
+      "grad_norm": 1.5354546308517456,
+      "learning_rate": 2.8149348230912477e-05,
+      "loss": 0.0603,
+      "step": 13926
+    },
+    {
+      "epoch": 38.90223463687151,
+      "grad_norm": 1.0759246349334717,
+      "learning_rate": 2.8148975791433894e-05,
+      "loss": 0.0595,
+      "step": 13927
+    },
+    {
+      "epoch": 38.90502793296089,
+      "grad_norm": 0.7140375971794128,
+      "learning_rate": 2.814860335195531e-05,
+      "loss": 0.0801,
+      "step": 13928
+    },
+    {
+      "epoch": 38.90782122905028,
+      "grad_norm": 0.9975626468658447,
+      "learning_rate": 2.8148230912476723e-05,
+      "loss": 0.0932,
+      "step": 13929
+    },
+    {
+      "epoch": 38.910614525139664,
+      "grad_norm": 0.6339572668075562,
+      "learning_rate": 2.8147858472998136e-05,
+      "loss": 0.1189,
+      "step": 13930
+    },
+    {
+      "epoch": 38.91340782122905,
+      "grad_norm": 0.43222346901893616,
+      "learning_rate": 2.8147486033519553e-05,
+      "loss": 0.115,
+      "step": 13931
+    },
+    {
+      "epoch": 38.91620111731844,
+      "grad_norm": 0.7280413508415222,
+      "learning_rate": 2.814711359404097e-05,
+      "loss": 0.0914,
+      "step": 13932
+    },
+    {
+      "epoch": 38.91899441340782,
+      "grad_norm": 0.5688527822494507,
+      "learning_rate": 2.8146741154562383e-05,
+      "loss": 0.1085,
+      "step": 13933
+    },
+    {
+      "epoch": 38.92178770949721,
+      "grad_norm": 0.7641273140907288,
+      "learning_rate": 2.81463687150838e-05,
+      "loss": 0.1243,
+      "step": 13934
+    },
+    {
+      "epoch": 38.92458100558659,
+      "grad_norm": 0.4794342815876007,
+      "learning_rate": 2.8145996275605216e-05,
+      "loss": 0.1112,
+      "step": 13935
+    },
+    {
+      "epoch": 38.927374301675975,
+      "grad_norm": 1.7034844160079956,
+      "learning_rate": 2.8145623836126632e-05,
+      "loss": 0.1209,
+      "step": 13936
+    },
+    {
+      "epoch": 38.93016759776536,
+      "grad_norm": 0.4324362576007843,
+      "learning_rate": 2.8145251396648045e-05,
+      "loss": 0.1074,
+      "step": 13937
+    },
+    {
+      "epoch": 38.93296089385475,
+      "grad_norm": 1.1344329118728638,
+      "learning_rate": 2.814487895716946e-05,
+      "loss": 0.1,
+      "step": 13938
+    },
+    {
+      "epoch": 38.935754189944134,
+      "grad_norm": 0.7757636308670044,
+      "learning_rate": 2.8144506517690875e-05,
+      "loss": 0.1044,
+      "step": 13939
+    },
+    {
+      "epoch": 38.93854748603352,
+      "grad_norm": 0.5023086071014404,
+      "learning_rate": 2.814413407821229e-05,
+      "loss": 0.1031,
+      "step": 13940
+    },
+    {
+      "epoch": 38.94134078212291,
+      "grad_norm": 0.4075654149055481,
+      "learning_rate": 2.8143761638733708e-05,
+      "loss": 0.1082,
+      "step": 13941
+    },
+    {
+      "epoch": 38.944134078212294,
+      "grad_norm": 0.48283904790878296,
+      "learning_rate": 2.814338919925512e-05,
+      "loss": 0.0905,
+      "step": 13942
+    },
+    {
+      "epoch": 38.94692737430167,
+      "grad_norm": 0.8485958576202393,
+      "learning_rate": 2.8143016759776538e-05,
+      "loss": 0.1032,
+      "step": 13943
+    },
+    {
+      "epoch": 38.94972067039106,
+      "grad_norm": 0.6102892756462097,
+      "learning_rate": 2.814264432029795e-05,
+      "loss": 0.0903,
+      "step": 13944
+    },
+    {
+      "epoch": 38.952513966480446,
+      "grad_norm": 0.5736841559410095,
+      "learning_rate": 2.8142271880819368e-05,
+      "loss": 0.0699,
+      "step": 13945
+    },
+    {
+      "epoch": 38.95530726256983,
+      "grad_norm": 0.7148399353027344,
+      "learning_rate": 2.8141899441340784e-05,
+      "loss": 0.0906,
+      "step": 13946
+    },
+    {
+      "epoch": 38.95810055865922,
+      "grad_norm": 1.0956953763961792,
+      "learning_rate": 2.8141527001862197e-05,
+      "loss": 0.0947,
+      "step": 13947
+    },
+    {
+      "epoch": 38.960893854748605,
+      "grad_norm": 0.6016318202018738,
+      "learning_rate": 2.8141154562383614e-05,
+      "loss": 0.0669,
+      "step": 13948
+    },
+    {
+      "epoch": 38.96368715083799,
+      "grad_norm": 0.7711247801780701,
+      "learning_rate": 2.814078212290503e-05,
+      "loss": 0.0866,
+      "step": 13949
+    },
+    {
+      "epoch": 38.96648044692738,
+      "grad_norm": 0.5966675281524658,
+      "learning_rate": 2.8140409683426443e-05,
+      "loss": 0.0763,
+      "step": 13950
+    },
+    {
+      "epoch": 38.96927374301676,
+      "grad_norm": 0.6873340606689453,
+      "learning_rate": 2.8140037243947857e-05,
+      "loss": 0.0739,
+      "step": 13951
+    },
+    {
+      "epoch": 38.97206703910614,
+      "grad_norm": 1.223775029182434,
+      "learning_rate": 2.8139664804469273e-05,
+      "loss": 0.0636,
+      "step": 13952
+    },
+    {
+      "epoch": 38.97486033519553,
+      "grad_norm": 0.8691296577453613,
+      "learning_rate": 2.813929236499069e-05,
+      "loss": 0.0678,
+      "step": 13953
+    },
+    {
+      "epoch": 38.977653631284916,
+      "grad_norm": 1.0352729558944702,
+      "learning_rate": 2.8138919925512106e-05,
+      "loss": 0.0936,
+      "step": 13954
+    },
+    {
+      "epoch": 38.9804469273743,
+      "grad_norm": 0.8331159353256226,
+      "learning_rate": 2.8138547486033523e-05,
+      "loss": 0.1105,
+      "step": 13955
+    },
+    {
+      "epoch": 38.98324022346369,
+      "grad_norm": 0.5664482712745667,
+      "learning_rate": 2.8138175046554936e-05,
+      "loss": 0.1048,
+      "step": 13956
+    },
+    {
+      "epoch": 38.986033519553075,
+      "grad_norm": 0.41837936639785767,
+      "learning_rate": 2.813780260707635e-05,
+      "loss": 0.0925,
+      "step": 13957
+    },
+    {
+      "epoch": 38.98882681564246,
+      "grad_norm": 3.002607583999634,
+      "learning_rate": 2.8137430167597766e-05,
+      "loss": 0.0986,
+      "step": 13958
+    },
+    {
+      "epoch": 38.99162011173184,
+      "grad_norm": 0.7762014865875244,
+      "learning_rate": 2.8137057728119182e-05,
+      "loss": 0.0654,
+      "step": 13959
+    },
+    {
+      "epoch": 38.99441340782123,
+      "grad_norm": 1.0032436847686768,
+      "learning_rate": 2.8136685288640595e-05,
+      "loss": 0.0858,
+      "step": 13960
+    },
+    {
+      "epoch": 38.997206703910614,
+      "grad_norm": 1.0900036096572876,
+      "learning_rate": 2.8136312849162012e-05,
+      "loss": 0.0764,
+      "step": 13961
+    },
+    {
+      "epoch": 39.0,
+      "grad_norm": 0.7696616649627686,
+      "learning_rate": 2.813594040968343e-05,
+      "loss": 0.089,
+      "step": 13962
+    },
+    {
+      "epoch": 39.002793296089386,
+      "grad_norm": 0.6514900922775269,
+      "learning_rate": 2.8135567970204845e-05,
+      "loss": 0.1198,
+      "step": 13963
+    },
+    {
+      "epoch": 39.00558659217877,
+      "grad_norm": 0.35125869512557983,
+      "learning_rate": 2.8135195530726258e-05,
+      "loss": 0.1265,
+      "step": 13964
+    },
+    {
+      "epoch": 39.00837988826816,
+      "grad_norm": 0.33939993381500244,
+      "learning_rate": 2.813482309124767e-05,
+      "loss": 0.1047,
+      "step": 13965
+    },
+    {
+      "epoch": 39.01117318435754,
+      "grad_norm": 0.4291524291038513,
+      "learning_rate": 2.8134450651769088e-05,
+      "loss": 0.1084,
+      "step": 13966
+    },
+    {
+      "epoch": 39.013966480446925,
+      "grad_norm": 1.307497501373291,
+      "learning_rate": 2.8134078212290504e-05,
+      "loss": 0.1234,
+      "step": 13967
+    },
+    {
+      "epoch": 39.01675977653631,
+      "grad_norm": 0.4758478105068207,
+      "learning_rate": 2.813370577281192e-05,
+      "loss": 0.0846,
+      "step": 13968
+    },
+    {
+      "epoch": 39.0195530726257,
+      "grad_norm": 0.7991107702255249,
+      "learning_rate": 2.8133333333333334e-05,
+      "loss": 0.0974,
+      "step": 13969
+    },
+    {
+      "epoch": 39.022346368715084,
+      "grad_norm": 0.39669328927993774,
+      "learning_rate": 2.8132960893854747e-05,
+      "loss": 0.0913,
+      "step": 13970
+    },
+    {
+      "epoch": 39.02513966480447,
+      "grad_norm": 0.5983617305755615,
+      "learning_rate": 2.8132588454376164e-05,
+      "loss": 0.0935,
+      "step": 13971
+    },
+    {
+      "epoch": 39.02793296089386,
+      "grad_norm": 1.2277920246124268,
+      "learning_rate": 2.813221601489758e-05,
+      "loss": 0.0805,
+      "step": 13972
+    },
+    {
+      "epoch": 39.03072625698324,
+      "grad_norm": 0.3706749975681305,
+      "learning_rate": 2.8131843575418997e-05,
+      "loss": 0.0657,
+      "step": 13973
+    },
+    {
+      "epoch": 39.03351955307262,
+      "grad_norm": 1.1726256608963013,
+      "learning_rate": 2.813147113594041e-05,
+      "loss": 0.0962,
+      "step": 13974
+    },
+    {
+      "epoch": 39.03631284916201,
+      "grad_norm": 0.4656217396259308,
+      "learning_rate": 2.8131098696461826e-05,
+      "loss": 0.0754,
+      "step": 13975
+    },
+    {
+      "epoch": 39.039106145251395,
+      "grad_norm": 0.6613001823425293,
+      "learning_rate": 2.8130726256983243e-05,
+      "loss": 0.0779,
+      "step": 13976
+    },
+    {
+      "epoch": 39.04189944134078,
+      "grad_norm": 0.6982211470603943,
+      "learning_rate": 2.8130353817504656e-05,
+      "loss": 0.0752,
+      "step": 13977
+    },
+    {
+      "epoch": 39.04469273743017,
+      "grad_norm": 0.5721521377563477,
+      "learning_rate": 2.812998137802607e-05,
+      "loss": 0.0859,
+      "step": 13978
+    },
+    {
+      "epoch": 39.047486033519554,
+      "grad_norm": 0.9946912527084351,
+      "learning_rate": 2.8129608938547486e-05,
+      "loss": 0.0697,
+      "step": 13979
+    },
+    {
+      "epoch": 39.05027932960894,
+      "grad_norm": 0.7086393237113953,
+      "learning_rate": 2.8129236499068902e-05,
+      "loss": 0.0738,
+      "step": 13980
+    },
+    {
+      "epoch": 39.05307262569833,
+      "grad_norm": 0.6088959574699402,
+      "learning_rate": 2.812886405959032e-05,
+      "loss": 0.0728,
+      "step": 13981
+    },
+    {
+      "epoch": 39.055865921787706,
+      "grad_norm": 0.47745397686958313,
+      "learning_rate": 2.8128491620111735e-05,
+      "loss": 0.0543,
+      "step": 13982
+    },
+    {
+      "epoch": 39.05865921787709,
+      "grad_norm": 0.475727915763855,
+      "learning_rate": 2.812811918063315e-05,
+      "loss": 0.0632,
+      "step": 13983
+    },
+    {
+      "epoch": 39.06145251396648,
+      "grad_norm": 0.6240501403808594,
+      "learning_rate": 2.812774674115456e-05,
+      "loss": 0.0712,
+      "step": 13984
+    },
+    {
+      "epoch": 39.064245810055866,
+      "grad_norm": 0.5536040663719177,
+      "learning_rate": 2.8127374301675978e-05,
+      "loss": 0.0542,
+      "step": 13985
+    },
+    {
+      "epoch": 39.06703910614525,
+      "grad_norm": 0.9809906482696533,
+      "learning_rate": 2.8127001862197395e-05,
+      "loss": 0.0697,
+      "step": 13986
+    },
+    {
+      "epoch": 39.06983240223464,
+      "grad_norm": 5.530771732330322,
+      "learning_rate": 2.8126629422718808e-05,
+      "loss": 0.0735,
+      "step": 13987
+    },
+    {
+      "epoch": 39.072625698324025,
+      "grad_norm": 0.5011076927185059,
+      "learning_rate": 2.8126256983240224e-05,
+      "loss": 0.1168,
+      "step": 13988
+    },
+    {
+      "epoch": 39.07541899441341,
+      "grad_norm": 0.763650119304657,
+      "learning_rate": 2.812588454376164e-05,
+      "loss": 0.1218,
+      "step": 13989
+    },
+    {
+      "epoch": 39.07821229050279,
+      "grad_norm": 0.36942729353904724,
+      "learning_rate": 2.8125512104283054e-05,
+      "loss": 0.1008,
+      "step": 13990
+    },
+    {
+      "epoch": 39.08100558659218,
+      "grad_norm": 0.9547336101531982,
+      "learning_rate": 2.812513966480447e-05,
+      "loss": 0.1045,
+      "step": 13991
+    },
+    {
+      "epoch": 39.08379888268156,
+      "grad_norm": 1.6202845573425293,
+      "learning_rate": 2.8124767225325884e-05,
+      "loss": 0.0881,
+      "step": 13992
+    },
+    {
+      "epoch": 39.08659217877095,
+      "grad_norm": 0.40806204080581665,
+      "learning_rate": 2.81243947858473e-05,
+      "loss": 0.1054,
+      "step": 13993
+    },
+    {
+      "epoch": 39.089385474860336,
+      "grad_norm": 0.73966383934021,
+      "learning_rate": 2.8124022346368717e-05,
+      "loss": 0.094,
+      "step": 13994
+    },
+    {
+      "epoch": 39.09217877094972,
+      "grad_norm": 0.43949905037879944,
+      "learning_rate": 2.8123649906890133e-05,
+      "loss": 0.1135,
+      "step": 13995
+    },
+    {
+      "epoch": 39.09497206703911,
+      "grad_norm": 0.4388486444950104,
+      "learning_rate": 2.8123277467411546e-05,
+      "loss": 0.0768,
+      "step": 13996
+    },
+    {
+      "epoch": 39.097765363128495,
+      "grad_norm": 1.4556699991226196,
+      "learning_rate": 2.812290502793296e-05,
+      "loss": 0.1005,
+      "step": 13997
+    },
+    {
+      "epoch": 39.100558659217874,
+      "grad_norm": 0.4349512755870819,
+      "learning_rate": 2.8122532588454376e-05,
+      "loss": 0.0869,
+      "step": 13998
+    },
+    {
+      "epoch": 39.10335195530726,
+      "grad_norm": 0.5730288624763489,
+      "learning_rate": 2.8122160148975793e-05,
+      "loss": 0.1073,
+      "step": 13999
+    },
+    {
+      "epoch": 39.10614525139665,
+      "grad_norm": 0.6371434926986694,
+      "learning_rate": 2.812178770949721e-05,
+      "loss": 0.066,
+      "step": 14000
+    },
+    {
+      "epoch": 39.10614525139665,
+      "eval_cer": 0.19417258573096352,
+      "eval_loss": 0.33140841126441956,
+      "eval_runtime": 24.1764,
+      "eval_samples_per_second": 187.704,
+      "eval_steps_per_second": 0.62,
+      "eval_wer": 0.3525879165386442,
+      "step": 14000
     }
   ],
   "logging_steps": 1.0,
@@ -91151,7 +98161,7 @@
         "early_stopping_threshold": 0.0
       },
       "attributes": {
-        "early_stopping_patience_counter": 3
+        "early_stopping_patience_counter": 4
       }
     },
     "TrainerControl": {
@@ -91165,7 +98175,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4.0153217257554156e+20,
+  "total_flos": 4.324168153867349e+20,
   "train_batch_size": 160,
   "trial_name": null,
   "trial_params": null