diff --git "a/checkpoint-950/trainer_state.json" "b/checkpoint-950/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-950/trainer_state.json"
@@ -0,0 +1,6683 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5855161787365177,
+  "eval_steps": 500,
+  "global_step": 950,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006163328197226503,
+      "grad_norm": 7.719967365264893,
+      "learning_rate": 1.226993865030675e-07,
+      "loss": 1.9036,
+      "step": 1
+    },
+    {
+      "epoch": 0.0012326656394453005,
+      "grad_norm": 7.849653720855713,
+      "learning_rate": 2.45398773006135e-07,
+      "loss": 1.9255,
+      "step": 2
+    },
+    {
+      "epoch": 0.0018489984591679508,
+      "grad_norm": 7.692729949951172,
+      "learning_rate": 3.680981595092025e-07,
+      "loss": 1.9242,
+      "step": 3
+    },
+    {
+      "epoch": 0.002465331278890601,
+      "grad_norm": 7.635551452636719,
+      "learning_rate": 4.9079754601227e-07,
+      "loss": 1.9527,
+      "step": 4
+    },
+    {
+      "epoch": 0.0030816640986132513,
+      "grad_norm": 7.432765007019043,
+      "learning_rate": 6.134969325153375e-07,
+      "loss": 1.898,
+      "step": 5
+    },
+    {
+      "epoch": 0.0036979969183359015,
+      "grad_norm": 7.166790008544922,
+      "learning_rate": 7.36196319018405e-07,
+      "loss": 1.9389,
+      "step": 6
+    },
+    {
+      "epoch": 0.004314329738058551,
+      "grad_norm": 7.253093242645264,
+      "learning_rate": 8.588957055214725e-07,
+      "loss": 1.9287,
+      "step": 7
+    },
+    {
+      "epoch": 0.004930662557781202,
+      "grad_norm": 6.417165279388428,
+      "learning_rate": 9.8159509202454e-07,
+      "loss": 1.9146,
+      "step": 8
+    },
+    {
+      "epoch": 0.005546995377503852,
+      "grad_norm": 5.822585582733154,
+      "learning_rate": 1.1042944785276075e-06,
+      "loss": 1.8543,
+      "step": 9
+    },
+    {
+      "epoch": 0.0061633281972265025,
+      "grad_norm": 4.997036457061768,
+      "learning_rate": 1.226993865030675e-06,
+      "loss": 1.8408,
+      "step": 10
+    },
+    {
+      "epoch": 0.006779661016949152,
+      "grad_norm": 3.9310598373413086,
+      "learning_rate": 1.3496932515337425e-06,
+      "loss": 1.8958,
+      "step": 11
+    },
+    {
+      "epoch": 0.007395993836671803,
+      "grad_norm": 3.077936887741089,
+      "learning_rate": 1.47239263803681e-06,
+      "loss": 1.848,
+      "step": 12
+    },
+    {
+      "epoch": 0.008012326656394453,
+      "grad_norm": 3.142758369445801,
+      "learning_rate": 1.5950920245398775e-06,
+      "loss": 1.7819,
+      "step": 13
+    },
+    {
+      "epoch": 0.008628659476117103,
+      "grad_norm": 3.4333081245422363,
+      "learning_rate": 1.717791411042945e-06,
+      "loss": 1.8899,
+      "step": 14
+    },
+    {
+      "epoch": 0.009244992295839754,
+      "grad_norm": 3.6334469318389893,
+      "learning_rate": 1.8404907975460124e-06,
+      "loss": 1.7977,
+      "step": 15
+    },
+    {
+      "epoch": 0.009861325115562404,
+      "grad_norm": 3.958219528198242,
+      "learning_rate": 1.96319018404908e-06,
+      "loss": 1.8153,
+      "step": 16
+    },
+    {
+      "epoch": 0.010477657935285054,
+      "grad_norm": 3.7573397159576416,
+      "learning_rate": 2.085889570552147e-06,
+      "loss": 1.8357,
+      "step": 17
+    },
+    {
+      "epoch": 0.011093990755007704,
+      "grad_norm": 3.2271149158477783,
+      "learning_rate": 2.208588957055215e-06,
+      "loss": 1.7938,
+      "step": 18
+    },
+    {
+      "epoch": 0.011710323574730355,
+      "grad_norm": 2.816006660461426,
+      "learning_rate": 2.331288343558282e-06,
+      "loss": 1.8035,
+      "step": 19
+    },
+    {
+      "epoch": 0.012326656394453005,
+      "grad_norm": 2.146484375,
+      "learning_rate": 2.45398773006135e-06,
+      "loss": 1.7529,
+      "step": 20
+    },
+    {
+      "epoch": 0.012942989214175655,
+      "grad_norm": 1.7285428047180176,
+      "learning_rate": 2.5766871165644175e-06,
+      "loss": 1.7743,
+      "step": 21
+    },
+    {
+      "epoch": 0.013559322033898305,
+      "grad_norm": 1.6210649013519287,
+      "learning_rate": 2.699386503067485e-06,
+      "loss": 1.7018,
+      "step": 22
+    },
+    {
+      "epoch": 0.014175654853620955,
+      "grad_norm": 1.4267948865890503,
+      "learning_rate": 2.822085889570552e-06,
+      "loss": 1.7139,
+      "step": 23
+    },
+    {
+      "epoch": 0.014791987673343606,
+      "grad_norm": 1.6083910465240479,
+      "learning_rate": 2.94478527607362e-06,
+      "loss": 1.7585,
+      "step": 24
+    },
+    {
+      "epoch": 0.015408320493066256,
+      "grad_norm": 1.659620761871338,
+      "learning_rate": 3.0674846625766875e-06,
+      "loss": 1.7924,
+      "step": 25
+    },
+    {
+      "epoch": 0.016024653312788906,
+      "grad_norm": 1.5887559652328491,
+      "learning_rate": 3.190184049079755e-06,
+      "loss": 1.7795,
+      "step": 26
+    },
+    {
+      "epoch": 0.016640986132511557,
+      "grad_norm": 1.4531543254852295,
+      "learning_rate": 3.312883435582822e-06,
+      "loss": 1.6757,
+      "step": 27
+    },
+    {
+      "epoch": 0.017257318952234205,
+      "grad_norm": 1.3152495622634888,
+      "learning_rate": 3.43558282208589e-06,
+      "loss": 1.7308,
+      "step": 28
+    },
+    {
+      "epoch": 0.017873651771956857,
+      "grad_norm": 1.26290762424469,
+      "learning_rate": 3.5582822085889574e-06,
+      "loss": 1.7481,
+      "step": 29
+    },
+    {
+      "epoch": 0.01848998459167951,
+      "grad_norm": 1.1644504070281982,
+      "learning_rate": 3.680981595092025e-06,
+      "loss": 1.7174,
+      "step": 30
+    },
+    {
+      "epoch": 0.019106317411402157,
+      "grad_norm": 1.1975916624069214,
+      "learning_rate": 3.8036809815950928e-06,
+      "loss": 1.746,
+      "step": 31
+    },
+    {
+      "epoch": 0.019722650231124808,
+      "grad_norm": 1.130025863647461,
+      "learning_rate": 3.92638036809816e-06,
+      "loss": 1.6782,
+      "step": 32
+    },
+    {
+      "epoch": 0.020338983050847456,
+      "grad_norm": 1.1134718656539917,
+      "learning_rate": 4.049079754601227e-06,
+      "loss": 1.6995,
+      "step": 33
+    },
+    {
+      "epoch": 0.020955315870570108,
+      "grad_norm": 1.101804256439209,
+      "learning_rate": 4.171779141104294e-06,
+      "loss": 1.7914,
+      "step": 34
+    },
+    {
+      "epoch": 0.02157164869029276,
+      "grad_norm": 1.0954248905181885,
+      "learning_rate": 4.294478527607362e-06,
+      "loss": 1.7081,
+      "step": 35
+    },
+    {
+      "epoch": 0.022187981510015407,
+      "grad_norm": 1.088081955909729,
+      "learning_rate": 4.41717791411043e-06,
+      "loss": 1.7081,
+      "step": 36
+    },
+    {
+      "epoch": 0.02280431432973806,
+      "grad_norm": 1.1250042915344238,
+      "learning_rate": 4.539877300613497e-06,
+      "loss": 1.7096,
+      "step": 37
+    },
+    {
+      "epoch": 0.02342064714946071,
+      "grad_norm": 1.1025091409683228,
+      "learning_rate": 4.662576687116564e-06,
+      "loss": 1.6818,
+      "step": 38
+    },
+    {
+      "epoch": 0.02403697996918336,
+      "grad_norm": 1.1217879056930542,
+      "learning_rate": 4.785276073619632e-06,
+      "loss": 1.7492,
+      "step": 39
+    },
+    {
+      "epoch": 0.02465331278890601,
+      "grad_norm": 1.0452276468276978,
+      "learning_rate": 4.9079754601227e-06,
+      "loss": 1.6338,
+      "step": 40
+    },
+    {
+      "epoch": 0.025269645608628658,
+      "grad_norm": 1.0448682308197021,
+      "learning_rate": 5.030674846625767e-06,
+      "loss": 1.6887,
+      "step": 41
+    },
+    {
+      "epoch": 0.02588597842835131,
+      "grad_norm": 1.0155638456344604,
+      "learning_rate": 5.153374233128835e-06,
+      "loss": 1.7415,
+      "step": 42
+    },
+    {
+      "epoch": 0.02650231124807396,
+      "grad_norm": 1.0147639513015747,
+      "learning_rate": 5.276073619631902e-06,
+      "loss": 1.7461,
+      "step": 43
+    },
+    {
+      "epoch": 0.02711864406779661,
+      "grad_norm": 1.0375574827194214,
+      "learning_rate": 5.39877300613497e-06,
+      "loss": 1.7016,
+      "step": 44
+    },
+    {
+      "epoch": 0.02773497688751926,
+      "grad_norm": 1.0689653158187866,
+      "learning_rate": 5.521472392638038e-06,
+      "loss": 1.6927,
+      "step": 45
+    },
+    {
+      "epoch": 0.02835130970724191,
+      "grad_norm": 1.0986844301223755,
+      "learning_rate": 5.644171779141104e-06,
+      "loss": 1.6271,
+      "step": 46
+    },
+    {
+      "epoch": 0.02896764252696456,
+      "grad_norm": 0.9898335337638855,
+      "learning_rate": 5.766871165644172e-06,
+      "loss": 1.6789,
+      "step": 47
+    },
+    {
+      "epoch": 0.029583975346687212,
+      "grad_norm": 1.0513464212417603,
+      "learning_rate": 5.88957055214724e-06,
+      "loss": 1.7316,
+      "step": 48
+    },
+    {
+      "epoch": 0.03020030816640986,
+      "grad_norm": 0.9952815175056458,
+      "learning_rate": 6.012269938650307e-06,
+      "loss": 1.6765,
+      "step": 49
+    },
+    {
+      "epoch": 0.030816640986132512,
+      "grad_norm": 1.0385676622390747,
+      "learning_rate": 6.134969325153375e-06,
+      "loss": 1.6997,
+      "step": 50
+    },
+    {
+      "epoch": 0.03143297380585516,
+      "grad_norm": 0.9788522720336914,
+      "learning_rate": 6.257668711656443e-06,
+      "loss": 1.7689,
+      "step": 51
+    },
+    {
+      "epoch": 0.03204930662557781,
+      "grad_norm": 0.9631125330924988,
+      "learning_rate": 6.38036809815951e-06,
+      "loss": 1.7202,
+      "step": 52
+    },
+    {
+      "epoch": 0.03266563944530046,
+      "grad_norm": 0.9789916276931763,
+      "learning_rate": 6.503067484662578e-06,
+      "loss": 1.6652,
+      "step": 53
+    },
+    {
+      "epoch": 0.033281972265023115,
+      "grad_norm": 0.9792534708976746,
+      "learning_rate": 6.625766871165644e-06,
+      "loss": 1.6977,
+      "step": 54
+    },
+    {
+      "epoch": 0.03389830508474576,
+      "grad_norm": 0.9288508296012878,
+      "learning_rate": 6.748466257668712e-06,
+      "loss": 1.6672,
+      "step": 55
+    },
+    {
+      "epoch": 0.03451463790446841,
+      "grad_norm": 0.9731179475784302,
+      "learning_rate": 6.87116564417178e-06,
+      "loss": 1.685,
+      "step": 56
+    },
+    {
+      "epoch": 0.035130970724191066,
+      "grad_norm": 0.9901231527328491,
+      "learning_rate": 6.993865030674847e-06,
+      "loss": 1.7066,
+      "step": 57
+    },
+    {
+      "epoch": 0.035747303543913714,
+      "grad_norm": 0.9718102812767029,
+      "learning_rate": 7.116564417177915e-06,
+      "loss": 1.6552,
+      "step": 58
+    },
+    {
+      "epoch": 0.03636363636363636,
+      "grad_norm": 0.9616168141365051,
+      "learning_rate": 7.239263803680983e-06,
+      "loss": 1.7241,
+      "step": 59
+    },
+    {
+      "epoch": 0.03697996918335902,
+      "grad_norm": 0.9576759934425354,
+      "learning_rate": 7.36196319018405e-06,
+      "loss": 1.7291,
+      "step": 60
+    },
+    {
+      "epoch": 0.037596302003081665,
+      "grad_norm": 1.2429341077804565,
+      "learning_rate": 7.484662576687118e-06,
+      "loss": 1.7077,
+      "step": 61
+    },
+    {
+      "epoch": 0.03821263482280431,
+      "grad_norm": 0.9671909213066101,
+      "learning_rate": 7.6073619631901856e-06,
+      "loss": 1.6846,
+      "step": 62
+    },
+    {
+      "epoch": 0.03882896764252696,
+      "grad_norm": 0.9405204057693481,
+      "learning_rate": 7.730061349693252e-06,
+      "loss": 1.6772,
+      "step": 63
+    },
+    {
+      "epoch": 0.039445300462249616,
+      "grad_norm": 0.9441912174224854,
+      "learning_rate": 7.85276073619632e-06,
+      "loss": 1.6796,
+      "step": 64
+    },
+    {
+      "epoch": 0.040061633281972264,
+      "grad_norm": 0.958046555519104,
+      "learning_rate": 7.975460122699386e-06,
+      "loss": 1.7265,
+      "step": 65
+    },
+    {
+      "epoch": 0.04067796610169491,
+      "grad_norm": 0.9565371870994568,
+      "learning_rate": 8.098159509202455e-06,
+      "loss": 1.6891,
+      "step": 66
+    },
+    {
+      "epoch": 0.04129429892141757,
+      "grad_norm": 0.9451284408569336,
+      "learning_rate": 8.220858895705522e-06,
+      "loss": 1.6341,
+      "step": 67
+    },
+    {
+      "epoch": 0.041910631741140215,
+      "grad_norm": 1.0197010040283203,
+      "learning_rate": 8.343558282208589e-06,
+      "loss": 1.6611,
+      "step": 68
+    },
+    {
+      "epoch": 0.042526964560862864,
+      "grad_norm": 0.9965417981147766,
+      "learning_rate": 8.466257668711658e-06,
+      "loss": 1.7486,
+      "step": 69
+    },
+    {
+      "epoch": 0.04314329738058552,
+      "grad_norm": 0.9524909257888794,
+      "learning_rate": 8.588957055214725e-06,
+      "loss": 1.6599,
+      "step": 70
+    },
+    {
+      "epoch": 0.04375963020030817,
+      "grad_norm": 0.9577109813690186,
+      "learning_rate": 8.711656441717792e-06,
+      "loss": 1.6858,
+      "step": 71
+    },
+    {
+      "epoch": 0.044375963020030815,
+      "grad_norm": 1.0352611541748047,
+      "learning_rate": 8.83435582822086e-06,
+      "loss": 1.6938,
+      "step": 72
+    },
+    {
+      "epoch": 0.04499229583975347,
+      "grad_norm": 0.9399254322052002,
+      "learning_rate": 8.957055214723927e-06,
+      "loss": 1.6399,
+      "step": 73
+    },
+    {
+      "epoch": 0.04560862865947612,
+      "grad_norm": 1.010736107826233,
+      "learning_rate": 9.079754601226994e-06,
+      "loss": 1.6061,
+      "step": 74
+    },
+    {
+      "epoch": 0.046224961479198766,
+      "grad_norm": 1.0074188709259033,
+      "learning_rate": 9.202453987730062e-06,
+      "loss": 1.6564,
+      "step": 75
+    },
+    {
+      "epoch": 0.04684129429892142,
+      "grad_norm": 0.9153802990913391,
+      "learning_rate": 9.325153374233129e-06,
+      "loss": 1.5832,
+      "step": 76
+    },
+    {
+      "epoch": 0.04745762711864407,
+      "grad_norm": 0.9539915323257446,
+      "learning_rate": 9.447852760736197e-06,
+      "loss": 1.6375,
+      "step": 77
+    },
+    {
+      "epoch": 0.04807395993836672,
+      "grad_norm": 0.9726946949958801,
+      "learning_rate": 9.570552147239264e-06,
+      "loss": 1.6495,
+      "step": 78
+    },
+    {
+      "epoch": 0.048690292758089365,
+      "grad_norm": 0.9744021892547607,
+      "learning_rate": 9.693251533742331e-06,
+      "loss": 1.6093,
+      "step": 79
+    },
+    {
+      "epoch": 0.04930662557781202,
+      "grad_norm": 0.9683598279953003,
+      "learning_rate": 9.8159509202454e-06,
+      "loss": 1.5864,
+      "step": 80
+    },
+    {
+      "epoch": 0.04992295839753467,
+      "grad_norm": 0.9956693053245544,
+      "learning_rate": 9.938650306748467e-06,
+      "loss": 1.6495,
+      "step": 81
+    },
+    {
+      "epoch": 0.050539291217257316,
+      "grad_norm": 0.9545958042144775,
+      "learning_rate": 1.0061349693251534e-05,
+      "loss": 1.6332,
+      "step": 82
+    },
+    {
+      "epoch": 0.05115562403697997,
+      "grad_norm": 1.0290021896362305,
+      "learning_rate": 1.0184049079754601e-05,
+      "loss": 1.687,
+      "step": 83
+    },
+    {
+      "epoch": 0.05177195685670262,
+      "grad_norm": 0.9288780689239502,
+      "learning_rate": 1.030674846625767e-05,
+      "loss": 1.6079,
+      "step": 84
+    },
+    {
+      "epoch": 0.05238828967642527,
+      "grad_norm": 0.9877973198890686,
+      "learning_rate": 1.0429447852760737e-05,
+      "loss": 1.633,
+      "step": 85
+    },
+    {
+      "epoch": 0.05300462249614792,
+      "grad_norm": 1.0284079313278198,
+      "learning_rate": 1.0552147239263804e-05,
+      "loss": 1.6778,
+      "step": 86
+    },
+    {
+      "epoch": 0.05362095531587057,
+      "grad_norm": 1.0213654041290283,
+      "learning_rate": 1.0674846625766873e-05,
+      "loss": 1.6767,
+      "step": 87
+    },
+    {
+      "epoch": 0.05423728813559322,
+      "grad_norm": 0.9213134050369263,
+      "learning_rate": 1.079754601226994e-05,
+      "loss": 1.6531,
+      "step": 88
+    },
+    {
+      "epoch": 0.054853620955315874,
+      "grad_norm": 1.0577795505523682,
+      "learning_rate": 1.0920245398773005e-05,
+      "loss": 1.69,
+      "step": 89
+    },
+    {
+      "epoch": 0.05546995377503852,
+      "grad_norm": 0.947443425655365,
+      "learning_rate": 1.1042944785276076e-05,
+      "loss": 1.6391,
+      "step": 90
+    },
+    {
+      "epoch": 0.05608628659476117,
+      "grad_norm": 1.0574003458023071,
+      "learning_rate": 1.1165644171779141e-05,
+      "loss": 1.6845,
+      "step": 91
+    },
+    {
+      "epoch": 0.05670261941448382,
+      "grad_norm": 0.9256036281585693,
+      "learning_rate": 1.1288343558282208e-05,
+      "loss": 1.6158,
+      "step": 92
+    },
+    {
+      "epoch": 0.05731895223420647,
+      "grad_norm": 1.0151914358139038,
+      "learning_rate": 1.1411042944785277e-05,
+      "loss": 1.705,
+      "step": 93
+    },
+    {
+      "epoch": 0.05793528505392912,
+      "grad_norm": 0.9245960116386414,
+      "learning_rate": 1.1533742331288344e-05,
+      "loss": 1.6391,
+      "step": 94
+    },
+    {
+      "epoch": 0.05855161787365177,
+      "grad_norm": 0.9816476702690125,
+      "learning_rate": 1.1656441717791411e-05,
+      "loss": 1.6036,
+      "step": 95
+    },
+    {
+      "epoch": 0.059167950693374424,
+      "grad_norm": 0.9318041205406189,
+      "learning_rate": 1.177914110429448e-05,
+      "loss": 1.638,
+      "step": 96
+    },
+    {
+      "epoch": 0.05978428351309707,
+      "grad_norm": 0.9426379799842834,
+      "learning_rate": 1.1901840490797547e-05,
+      "loss": 1.7349,
+      "step": 97
+    },
+    {
+      "epoch": 0.06040061633281972,
+      "grad_norm": 1.1403101682662964,
+      "learning_rate": 1.2024539877300614e-05,
+      "loss": 1.7307,
+      "step": 98
+    },
+    {
+      "epoch": 0.061016949152542375,
+      "grad_norm": 1.075808048248291,
+      "learning_rate": 1.2147239263803683e-05,
+      "loss": 1.6535,
+      "step": 99
+    },
+    {
+      "epoch": 0.061633281972265024,
+      "grad_norm": 1.0147042274475098,
+      "learning_rate": 1.226993865030675e-05,
+      "loss": 1.6399,
+      "step": 100
+    },
+    {
+      "epoch": 0.06224961479198767,
+      "grad_norm": 0.9782200455665588,
+      "learning_rate": 1.2392638036809817e-05,
+      "loss": 1.6726,
+      "step": 101
+    },
+    {
+      "epoch": 0.06286594761171033,
+      "grad_norm": 1.070759892463684,
+      "learning_rate": 1.2515337423312886e-05,
+      "loss": 1.7249,
+      "step": 102
+    },
+    {
+      "epoch": 0.06348228043143297,
+      "grad_norm": 0.9782586097717285,
+      "learning_rate": 1.2638036809815953e-05,
+      "loss": 1.5952,
+      "step": 103
+    },
+    {
+      "epoch": 0.06409861325115562,
+      "grad_norm": 0.9195032119750977,
+      "learning_rate": 1.276073619631902e-05,
+      "loss": 1.6807,
+      "step": 104
+    },
+    {
+      "epoch": 0.06471494607087827,
+      "grad_norm": 1.0292532444000244,
+      "learning_rate": 1.2883435582822085e-05,
+      "loss": 1.7408,
+      "step": 105
+    },
+    {
+      "epoch": 0.06533127889060092,
+      "grad_norm": 0.9190950393676758,
+      "learning_rate": 1.3006134969325156e-05,
+      "loss": 1.6491,
+      "step": 106
+    },
+    {
+      "epoch": 0.06594761171032358,
+      "grad_norm": 0.993139386177063,
+      "learning_rate": 1.3128834355828221e-05,
+      "loss": 1.6553,
+      "step": 107
+    },
+    {
+      "epoch": 0.06656394453004623,
+      "grad_norm": 0.9514711499214172,
+      "learning_rate": 1.3251533742331288e-05,
+      "loss": 1.6865,
+      "step": 108
+    },
+    {
+      "epoch": 0.06718027734976888,
+      "grad_norm": 0.9713273644447327,
+      "learning_rate": 1.3374233128834357e-05,
+      "loss": 1.6533,
+      "step": 109
+    },
+    {
+      "epoch": 0.06779661016949153,
+      "grad_norm": 1.8169646263122559,
+      "learning_rate": 1.3496932515337424e-05,
+      "loss": 1.655,
+      "step": 110
+    },
+    {
+      "epoch": 0.06841294298921417,
+      "grad_norm": 0.9389533996582031,
+      "learning_rate": 1.3619631901840491e-05,
+      "loss": 1.6839,
+      "step": 111
+    },
+    {
+      "epoch": 0.06902927580893682,
+      "grad_norm": 0.9239947199821472,
+      "learning_rate": 1.374233128834356e-05,
+      "loss": 1.7008,
+      "step": 112
+    },
+    {
+      "epoch": 0.06964560862865947,
+      "grad_norm": 0.9407342076301575,
+      "learning_rate": 1.3865030674846627e-05,
+      "loss": 1.6727,
+      "step": 113
+    },
+    {
+      "epoch": 0.07026194144838213,
+      "grad_norm": 0.8669116497039795,
+      "learning_rate": 1.3987730061349694e-05,
+      "loss": 1.6206,
+      "step": 114
+    },
+    {
+      "epoch": 0.07087827426810478,
+      "grad_norm": 0.8853608965873718,
+      "learning_rate": 1.4110429447852763e-05,
+      "loss": 1.6105,
+      "step": 115
+    },
+    {
+      "epoch": 0.07149460708782743,
+      "grad_norm": 0.9186128377914429,
+      "learning_rate": 1.423312883435583e-05,
+      "loss": 1.6696,
+      "step": 116
+    },
+    {
+      "epoch": 0.07211093990755008,
+      "grad_norm": 0.8860874176025391,
+      "learning_rate": 1.4355828220858897e-05,
+      "loss": 1.5962,
+      "step": 117
+    },
+    {
+      "epoch": 0.07272727272727272,
+      "grad_norm": 0.8854448199272156,
+      "learning_rate": 1.4478527607361965e-05,
+      "loss": 1.6951,
+      "step": 118
+    },
+    {
+      "epoch": 0.07334360554699537,
+      "grad_norm": 0.8960251212120056,
+      "learning_rate": 1.4601226993865032e-05,
+      "loss": 1.6336,
+      "step": 119
+    },
+    {
+      "epoch": 0.07395993836671803,
+      "grad_norm": 0.8875281810760498,
+      "learning_rate": 1.47239263803681e-05,
+      "loss": 1.6505,
+      "step": 120
+    },
+    {
+      "epoch": 0.07457627118644068,
+      "grad_norm": 0.8769788146018982,
+      "learning_rate": 1.4846625766871168e-05,
+      "loss": 1.6807,
+      "step": 121
+    },
+    {
+      "epoch": 0.07519260400616333,
+      "grad_norm": 0.8698969483375549,
+      "learning_rate": 1.4969325153374235e-05,
+      "loss": 1.6754,
+      "step": 122
+    },
+    {
+      "epoch": 0.07580893682588598,
+      "grad_norm": 0.8725311160087585,
+      "learning_rate": 1.50920245398773e-05,
+      "loss": 1.6701,
+      "step": 123
+    },
+    {
+      "epoch": 0.07642526964560863,
+      "grad_norm": 0.8646256923675537,
+      "learning_rate": 1.5214723926380371e-05,
+      "loss": 1.5921,
+      "step": 124
+    },
+    {
+      "epoch": 0.07704160246533127,
+      "grad_norm": 0.9173498153686523,
+      "learning_rate": 1.5337423312883436e-05,
+      "loss": 1.6701,
+      "step": 125
+    },
+    {
+      "epoch": 0.07765793528505392,
+      "grad_norm": 0.8997793793678284,
+      "learning_rate": 1.5460122699386504e-05,
+      "loss": 1.6448,
+      "step": 126
+    },
+    {
+      "epoch": 0.07827426810477658,
+      "grad_norm": 0.9629970192909241,
+      "learning_rate": 1.5582822085889574e-05,
+      "loss": 1.6551,
+      "step": 127
+    },
+    {
+      "epoch": 0.07889060092449923,
+      "grad_norm": 0.8857690691947937,
+      "learning_rate": 1.570552147239264e-05,
+      "loss": 1.6138,
+      "step": 128
+    },
+    {
+      "epoch": 0.07950693374422188,
+      "grad_norm": 0.9150853157043457,
+      "learning_rate": 1.5828220858895708e-05,
+      "loss": 1.7048,
+      "step": 129
+    },
+    {
+      "epoch": 0.08012326656394453,
+      "grad_norm": 0.9400542974472046,
+      "learning_rate": 1.5950920245398772e-05,
+      "loss": 1.6601,
+      "step": 130
+    },
+    {
+      "epoch": 0.08073959938366718,
+      "grad_norm": 0.9279290437698364,
+      "learning_rate": 1.6073619631901842e-05,
+      "loss": 1.6426,
+      "step": 131
+    },
+    {
+      "epoch": 0.08135593220338982,
+      "grad_norm": 0.9481431841850281,
+      "learning_rate": 1.619631901840491e-05,
+      "loss": 1.6097,
+      "step": 132
+    },
+    {
+      "epoch": 0.08197226502311249,
+      "grad_norm": 0.8905608654022217,
+      "learning_rate": 1.6319018404907976e-05,
+      "loss": 1.5951,
+      "step": 133
+    },
+    {
+      "epoch": 0.08258859784283513,
+      "grad_norm": 0.9486684799194336,
+      "learning_rate": 1.6441717791411043e-05,
+      "loss": 1.6357,
+      "step": 134
+    },
+    {
+      "epoch": 0.08320493066255778,
+      "grad_norm": 0.8683281540870667,
+      "learning_rate": 1.656441717791411e-05,
+      "loss": 1.6303,
+      "step": 135
+    },
+    {
+      "epoch": 0.08382126348228043,
+      "grad_norm": 0.9099710583686829,
+      "learning_rate": 1.6687116564417178e-05,
+      "loss": 1.6775,
+      "step": 136
+    },
+    {
+      "epoch": 0.08443759630200308,
+      "grad_norm": 0.8527466654777527,
+      "learning_rate": 1.6809815950920248e-05,
+      "loss": 1.6489,
+      "step": 137
+    },
+    {
+      "epoch": 0.08505392912172573,
+      "grad_norm": 0.9052302241325378,
+      "learning_rate": 1.6932515337423315e-05,
+      "loss": 1.6709,
+      "step": 138
+    },
+    {
+      "epoch": 0.08567026194144839,
+      "grad_norm": 0.8606592416763306,
+      "learning_rate": 1.7055214723926382e-05,
+      "loss": 1.6397,
+      "step": 139
+    },
+    {
+      "epoch": 0.08628659476117104,
+      "grad_norm": 0.938488781452179,
+      "learning_rate": 1.717791411042945e-05,
+      "loss": 1.6465,
+      "step": 140
+    },
+    {
+      "epoch": 0.08690292758089369,
+      "grad_norm": 0.9017755389213562,
+      "learning_rate": 1.7300613496932516e-05,
+      "loss": 1.6894,
+      "step": 141
+    },
+    {
+      "epoch": 0.08751926040061633,
+      "grad_norm": 0.9042711853981018,
+      "learning_rate": 1.7423312883435583e-05,
+      "loss": 1.7309,
+      "step": 142
+    },
+    {
+      "epoch": 0.08813559322033898,
+      "grad_norm": 0.9422665238380432,
+      "learning_rate": 1.7546012269938654e-05,
+      "loss": 1.7347,
+      "step": 143
+    },
+    {
+      "epoch": 0.08875192604006163,
+      "grad_norm": 0.8688966631889343,
+      "learning_rate": 1.766871165644172e-05,
+      "loss": 1.7072,
+      "step": 144
+    },
+    {
+      "epoch": 0.08936825885978428,
+      "grad_norm": 0.8922861814498901,
+      "learning_rate": 1.7791411042944788e-05,
+      "loss": 1.6258,
+      "step": 145
+    },
+    {
+      "epoch": 0.08998459167950694,
+      "grad_norm": 0.8801250457763672,
+      "learning_rate": 1.7914110429447855e-05,
+      "loss": 1.6455,
+      "step": 146
+    },
+    {
+      "epoch": 0.09060092449922959,
+      "grad_norm": 0.8685930371284485,
+      "learning_rate": 1.8036809815950922e-05,
+      "loss": 1.6359,
+      "step": 147
+    },
+    {
+      "epoch": 0.09121725731895224,
+      "grad_norm": 0.832036554813385,
+      "learning_rate": 1.815950920245399e-05,
+      "loss": 1.6115,
+      "step": 148
+    },
+    {
+      "epoch": 0.09183359013867488,
+      "grad_norm": 0.983727216720581,
+      "learning_rate": 1.828220858895706e-05,
+      "loss": 1.6048,
+      "step": 149
+    },
+    {
+      "epoch": 0.09244992295839753,
+      "grad_norm": 0.8593609929084778,
+      "learning_rate": 1.8404907975460123e-05,
+      "loss": 1.6641,
+      "step": 150
+    },
+    {
+      "epoch": 0.09306625577812018,
+      "grad_norm": 0.8574192523956299,
+      "learning_rate": 1.852760736196319e-05,
+      "loss": 1.6361,
+      "step": 151
+    },
+    {
+      "epoch": 0.09368258859784284,
+      "grad_norm": 0.8460291624069214,
+      "learning_rate": 1.8650306748466257e-05,
+      "loss": 1.5913,
+      "step": 152
+    },
+    {
+      "epoch": 0.09429892141756549,
+      "grad_norm": 0.8619217872619629,
+      "learning_rate": 1.8773006134969328e-05,
+      "loss": 1.6121,
+      "step": 153
+    },
+    {
+      "epoch": 0.09491525423728814,
+      "grad_norm": 0.8337171673774719,
+      "learning_rate": 1.8895705521472395e-05,
+      "loss": 1.6558,
+      "step": 154
+    },
+    {
+      "epoch": 0.09553158705701079,
+      "grad_norm": 0.850978672504425,
+      "learning_rate": 1.9018404907975462e-05,
+      "loss": 1.6564,
+      "step": 155
+    },
+    {
+      "epoch": 0.09614791987673343,
+      "grad_norm": 0.8237566947937012,
+      "learning_rate": 1.914110429447853e-05,
+      "loss": 1.6478,
+      "step": 156
+    },
+    {
+      "epoch": 0.09676425269645608,
+      "grad_norm": 0.8377094268798828,
+      "learning_rate": 1.9263803680981596e-05,
+      "loss": 1.7155,
+      "step": 157
+    },
+    {
+      "epoch": 0.09738058551617873,
+      "grad_norm": 0.814061164855957,
+      "learning_rate": 1.9386503067484663e-05,
+      "loss": 1.5845,
+      "step": 158
+    },
+    {
+      "epoch": 0.09799691833590139,
+      "grad_norm": 0.8016124367713928,
+      "learning_rate": 1.9509202453987733e-05,
+      "loss": 1.6579,
+      "step": 159
+    },
+    {
+      "epoch": 0.09861325115562404,
+      "grad_norm": 0.8450319170951843,
+      "learning_rate": 1.96319018404908e-05,
+      "loss": 1.6359,
+      "step": 160
+    },
+    {
+      "epoch": 0.09922958397534669,
+      "grad_norm": 0.8084301352500916,
+      "learning_rate": 1.9754601226993868e-05,
+      "loss": 1.6087,
+      "step": 161
+    },
+    {
+      "epoch": 0.09984591679506934,
+      "grad_norm": 0.8020560145378113,
+      "learning_rate": 1.9877300613496935e-05,
+      "loss": 1.6117,
+      "step": 162
+    },
+    {
+      "epoch": 0.10046224961479198,
+      "grad_norm": 0.797077476978302,
+      "learning_rate": 2e-05,
+      "loss": 1.6237,
+      "step": 163
+    },
+    {
+      "epoch": 0.10107858243451463,
+      "grad_norm": 0.8158491849899292,
+      "learning_rate": 1.999997681756781e-05,
+      "loss": 1.6681,
+      "step": 164
+    },
+    {
+      "epoch": 0.1016949152542373,
+      "grad_norm": 0.8452540040016174,
+      "learning_rate": 1.999990727037872e-05,
+      "loss": 1.6788,
+      "step": 165
+    },
+    {
+      "epoch": 0.10231124807395994,
+      "grad_norm": 0.7805876731872559,
+      "learning_rate": 1.999979135875519e-05,
+      "loss": 1.626,
+      "step": 166
+    },
+    {
+      "epoch": 0.10292758089368259,
+      "grad_norm": 0.8445929884910583,
+      "learning_rate": 1.999962908323464e-05,
+      "loss": 1.6328,
+      "step": 167
+    },
+    {
+      "epoch": 0.10354391371340524,
+      "grad_norm": 0.8229109048843384,
+      "learning_rate": 1.9999420444569462e-05,
+      "loss": 1.5747,
+      "step": 168
+    },
+    {
+      "epoch": 0.10416024653312789,
+      "grad_norm": 0.8682668805122375,
+      "learning_rate": 1.9999165443727e-05,
+      "loss": 1.6392,
+      "step": 169
+    },
+    {
+      "epoch": 0.10477657935285054,
+      "grad_norm": 0.8105486035346985,
+      "learning_rate": 1.9998864081889567e-05,
+      "loss": 1.6279,
+      "step": 170
+    },
+    {
+      "epoch": 0.10539291217257318,
+      "grad_norm": 0.8182287216186523,
+      "learning_rate": 1.999851636045442e-05,
+      "loss": 1.6671,
+      "step": 171
+    },
+    {
+      "epoch": 0.10600924499229585,
+      "grad_norm": 0.8665288686752319,
+      "learning_rate": 1.9998122281033763e-05,
+      "loss": 1.6003,
+      "step": 172
+    },
+    {
+      "epoch": 0.1066255778120185,
+      "grad_norm": 0.8504026532173157,
+      "learning_rate": 1.9997681845454743e-05,
+      "loss": 1.7534,
+      "step": 173
+    },
+    {
+      "epoch": 0.10724191063174114,
+      "grad_norm": 0.7957158088684082,
+      "learning_rate": 1.9997195055759435e-05,
+      "loss": 1.5881,
+      "step": 174
+    },
+    {
+      "epoch": 0.10785824345146379,
+      "grad_norm": 0.8307750821113586,
+      "learning_rate": 1.9996661914204833e-05,
+      "loss": 1.6323,
+      "step": 175
+    },
+    {
+      "epoch": 0.10847457627118644,
+      "grad_norm": 0.8265060186386108,
+      "learning_rate": 1.9996082423262835e-05,
+      "loss": 1.6109,
+      "step": 176
+    },
+    {
+      "epoch": 0.10909090909090909,
+      "grad_norm": 0.8249815106391907,
+      "learning_rate": 1.9995456585620245e-05,
+      "loss": 1.6395,
+      "step": 177
+    },
+    {
+      "epoch": 0.10970724191063175,
+      "grad_norm": 0.790324866771698,
+      "learning_rate": 1.9994784404178755e-05,
+      "loss": 1.664,
+      "step": 178
+    },
+    {
+      "epoch": 0.1103235747303544,
+      "grad_norm": 0.8790420889854431,
+      "learning_rate": 1.9994065882054924e-05,
+      "loss": 1.6415,
+      "step": 179
+    },
+    {
+      "epoch": 0.11093990755007704,
+      "grad_norm": 0.8272804021835327,
+      "learning_rate": 1.9993301022580164e-05,
+      "loss": 1.6023,
+      "step": 180
+    },
+    {
+      "epoch": 0.11155624036979969,
+      "grad_norm": 0.7871703505516052,
+      "learning_rate": 1.9992489829300745e-05,
+      "loss": 1.6279,
+      "step": 181
+    },
+    {
+      "epoch": 0.11217257318952234,
+      "grad_norm": 0.8058778643608093,
+      "learning_rate": 1.9991632305977748e-05,
+      "loss": 1.7735,
+      "step": 182
+    },
+    {
+      "epoch": 0.11278890600924499,
+      "grad_norm": 0.7899259328842163,
+      "learning_rate": 1.9990728456587067e-05,
+      "loss": 1.6742,
+      "step": 183
+    },
+    {
+      "epoch": 0.11340523882896764,
+      "grad_norm": 0.7826928496360779,
+      "learning_rate": 1.998977828531939e-05,
+      "loss": 1.5975,
+      "step": 184
+    },
+    {
+      "epoch": 0.1140215716486903,
+      "grad_norm": 0.771249532699585,
+      "learning_rate": 1.9988781796580174e-05,
+      "loss": 1.5829,
+      "step": 185
+    },
+    {
+      "epoch": 0.11463790446841295,
+      "grad_norm": 0.7774127721786499,
+      "learning_rate": 1.998773899498962e-05,
+      "loss": 1.6149,
+      "step": 186
+    },
+    {
+      "epoch": 0.1152542372881356,
+      "grad_norm": 0.764395534992218,
+      "learning_rate": 1.9986649885382672e-05,
+      "loss": 1.5948,
+      "step": 187
+    },
+    {
+      "epoch": 0.11587057010785824,
+      "grad_norm": 0.7593989968299866,
+      "learning_rate": 1.9985514472808965e-05,
+      "loss": 1.6292,
+      "step": 188
+    },
+    {
+      "epoch": 0.11648690292758089,
+      "grad_norm": 0.7568250894546509,
+      "learning_rate": 1.9984332762532827e-05,
+      "loss": 1.6389,
+      "step": 189
+    },
+    {
+      "epoch": 0.11710323574730354,
+      "grad_norm": 0.7788695096969604,
+      "learning_rate": 1.998310476003324e-05,
+      "loss": 1.6252,
+      "step": 190
+    },
+    {
+      "epoch": 0.1177195685670262,
+      "grad_norm": 0.8050792813301086,
+      "learning_rate": 1.9981830471003825e-05,
+      "loss": 1.6863,
+      "step": 191
+    },
+    {
+      "epoch": 0.11833590138674885,
+      "grad_norm": 0.7851357460021973,
+      "learning_rate": 1.9980509901352802e-05,
+      "loss": 1.65,
+      "step": 192
+    },
+    {
+      "epoch": 0.1189522342064715,
+      "grad_norm": 0.781894862651825,
+      "learning_rate": 1.9979143057202975e-05,
+      "loss": 1.6652,
+      "step": 193
+    },
+    {
+      "epoch": 0.11956856702619414,
+      "grad_norm": 0.8047742247581482,
+      "learning_rate": 1.9977729944891697e-05,
+      "loss": 1.5493,
+      "step": 194
+    },
+    {
+      "epoch": 0.12018489984591679,
+      "grad_norm": 0.7938175797462463,
+      "learning_rate": 1.9976270570970847e-05,
+      "loss": 1.5497,
+      "step": 195
+    },
+    {
+      "epoch": 0.12080123266563944,
+      "grad_norm": 0.8171627521514893,
+      "learning_rate": 1.997476494220679e-05,
+      "loss": 1.6222,
+      "step": 196
+    },
+    {
+      "epoch": 0.12141756548536209,
+      "grad_norm": 0.7670049667358398,
+      "learning_rate": 1.9973213065580353e-05,
+      "loss": 1.66,
+      "step": 197
+    },
+    {
+      "epoch": 0.12203389830508475,
+      "grad_norm": 0.7528908252716064,
+      "learning_rate": 1.9971614948286797e-05,
+      "loss": 1.6427,
+      "step": 198
+    },
+    {
+      "epoch": 0.1226502311248074,
+      "grad_norm": 0.7718436121940613,
+      "learning_rate": 1.9969970597735764e-05,
+      "loss": 1.5954,
+      "step": 199
+    },
+    {
+      "epoch": 0.12326656394453005,
+      "grad_norm": 0.7847468256950378,
+      "learning_rate": 1.9968280021551264e-05,
+      "loss": 1.6064,
+      "step": 200
+    },
+    {
+      "epoch": 0.1238828967642527,
+      "grad_norm": 0.7835562825202942,
+      "learning_rate": 1.9966543227571634e-05,
+      "loss": 1.5742,
+      "step": 201
+    },
+    {
+      "epoch": 0.12449922958397534,
+      "grad_norm": 0.7712457180023193,
+      "learning_rate": 1.996476022384949e-05,
+      "loss": 1.6723,
+      "step": 202
+    },
+    {
+      "epoch": 0.125115562403698,
+      "grad_norm": 0.7640112638473511,
+      "learning_rate": 1.9962931018651713e-05,
+      "loss": 1.5666,
+      "step": 203
+    },
+    {
+      "epoch": 0.12573189522342065,
+      "grad_norm": 0.8142703771591187,
+      "learning_rate": 1.996105562045938e-05,
+      "loss": 1.6277,
+      "step": 204
+    },
+    {
+      "epoch": 0.1263482280431433,
+      "grad_norm": 0.7460588216781616,
+      "learning_rate": 1.9959134037967757e-05,
+      "loss": 1.6204,
+      "step": 205
+    },
+    {
+      "epoch": 0.12696456086286595,
+      "grad_norm": 0.7740151882171631,
+      "learning_rate": 1.995716628008623e-05,
+      "loss": 1.6449,
+      "step": 206
+    },
+    {
+      "epoch": 0.1275808936825886,
+      "grad_norm": 0.7314134240150452,
+      "learning_rate": 1.9955152355938278e-05,
+      "loss": 1.5764,
+      "step": 207
+    },
+    {
+      "epoch": 0.12819722650231125,
+      "grad_norm": 0.7663095593452454,
+      "learning_rate": 1.995309227486144e-05,
+      "loss": 1.6075,
+      "step": 208
+    },
+    {
+      "epoch": 0.1288135593220339,
+      "grad_norm": 0.7631840109825134,
+      "learning_rate": 1.9950986046407252e-05,
+      "loss": 1.645,
+      "step": 209
+    },
+    {
+      "epoch": 0.12942989214175654,
+      "grad_norm": 0.7416641712188721,
+      "learning_rate": 1.994883368034121e-05,
+      "loss": 1.6225,
+      "step": 210
+    },
+    {
+      "epoch": 0.1300462249614792,
+      "grad_norm": 0.769237756729126,
+      "learning_rate": 1.994663518664274e-05,
+      "loss": 1.6033,
+      "step": 211
+    },
+    {
+      "epoch": 0.13066255778120184,
+      "grad_norm": 0.7613971829414368,
+      "learning_rate": 1.994439057550512e-05,
+      "loss": 1.631,
+      "step": 212
+    },
+    {
+      "epoch": 0.13127889060092449,
+      "grad_norm": 0.7329368591308594,
+      "learning_rate": 1.9942099857335457e-05,
+      "loss": 1.6177,
+      "step": 213
+    },
+    {
+      "epoch": 0.13189522342064716,
+      "grad_norm": 0.7614597678184509,
+      "learning_rate": 1.993976304275464e-05,
+      "loss": 1.6313,
+      "step": 214
+    },
+    {
+      "epoch": 0.1325115562403698,
+      "grad_norm": 0.7219412326812744,
+      "learning_rate": 1.9937380142597277e-05,
+      "loss": 1.6037,
+      "step": 215
+    },
+    {
+      "epoch": 0.13312788906009246,
+      "grad_norm": 0.7598880529403687,
+      "learning_rate": 1.9934951167911648e-05,
+      "loss": 1.621,
+      "step": 216
+    },
+    {
+      "epoch": 0.1337442218798151,
+      "grad_norm": 0.7477957010269165,
+      "learning_rate": 1.993247612995967e-05,
+      "loss": 1.6132,
+      "step": 217
+    },
+    {
+      "epoch": 0.13436055469953775,
+      "grad_norm": 0.736220121383667,
+      "learning_rate": 1.9929955040216813e-05,
+      "loss": 1.6284,
+      "step": 218
+    },
+    {
+      "epoch": 0.1349768875192604,
+      "grad_norm": 0.7698326706886292,
+      "learning_rate": 1.9927387910372085e-05,
+      "loss": 1.6506,
+      "step": 219
+    },
+    {
+      "epoch": 0.13559322033898305,
+      "grad_norm": 0.7397698760032654,
+      "learning_rate": 1.9924774752327943e-05,
+      "loss": 1.61,
+      "step": 220
+    },
+    {
+      "epoch": 0.1362095531587057,
+      "grad_norm": 0.7629734873771667,
+      "learning_rate": 1.9922115578200256e-05,
+      "loss": 1.5915,
+      "step": 221
+    },
+    {
+      "epoch": 0.13682588597842835,
+      "grad_norm": 0.7445245385169983,
+      "learning_rate": 1.991941040031826e-05,
+      "loss": 1.6937,
+      "step": 222
+    },
+    {
+      "epoch": 0.137442218798151,
+      "grad_norm": 0.7118062973022461,
+      "learning_rate": 1.991665923122446e-05,
+      "loss": 1.5671,
+      "step": 223
+    },
+    {
+      "epoch": 0.13805855161787364,
+      "grad_norm": 0.7344526052474976,
+      "learning_rate": 1.9913862083674626e-05,
+      "loss": 1.6084,
+      "step": 224
+    },
+    {
+      "epoch": 0.1386748844375963,
+      "grad_norm": 0.748241126537323,
+      "learning_rate": 1.991101897063769e-05,
+      "loss": 1.5487,
+      "step": 225
+    },
+    {
+      "epoch": 0.13929121725731894,
+      "grad_norm": 1.3268109560012817,
+      "learning_rate": 1.9908129905295708e-05,
+      "loss": 1.6293,
+      "step": 226
+    },
+    {
+      "epoch": 0.13990755007704161,
+      "grad_norm": 0.7278845310211182,
+      "learning_rate": 1.9905194901043794e-05,
+      "loss": 1.629,
+      "step": 227
+    },
+    {
+      "epoch": 0.14052388289676426,
+      "grad_norm": 0.7867671847343445,
+      "learning_rate": 1.990221397149005e-05,
+      "loss": 1.6088,
+      "step": 228
+    },
+    {
+      "epoch": 0.1411402157164869,
+      "grad_norm": 0.7714179158210754,
+      "learning_rate": 1.9899187130455524e-05,
+      "loss": 1.6166,
+      "step": 229
+    },
+    {
+      "epoch": 0.14175654853620956,
+      "grad_norm": 0.7665597796440125,
+      "learning_rate": 1.9896114391974114e-05,
+      "loss": 1.5773,
+      "step": 230
+    },
+    {
+      "epoch": 0.1423728813559322,
+      "grad_norm": 0.810384213924408,
+      "learning_rate": 1.989299577029254e-05,
+      "loss": 1.6255,
+      "step": 231
+    },
+    {
+      "epoch": 0.14298921417565486,
+      "grad_norm": 0.7698226571083069,
+      "learning_rate": 1.988983127987024e-05,
+      "loss": 1.6249,
+      "step": 232
+    },
+    {
+      "epoch": 0.1436055469953775,
+      "grad_norm": 0.7346594333648682,
+      "learning_rate": 1.9886620935379332e-05,
+      "loss": 1.6214,
+      "step": 233
+    },
+    {
+      "epoch": 0.14422187981510015,
+      "grad_norm": 0.7804876565933228,
+      "learning_rate": 1.9883364751704544e-05,
+      "loss": 1.6361,
+      "step": 234
+    },
+    {
+      "epoch": 0.1448382126348228,
+      "grad_norm": 0.7119391560554504,
+      "learning_rate": 1.9880062743943117e-05,
+      "loss": 1.5406,
+      "step": 235
+    },
+    {
+      "epoch": 0.14545454545454545,
+      "grad_norm": 0.7504518628120422,
+      "learning_rate": 1.9876714927404774e-05,
+      "loss": 1.5745,
+      "step": 236
+    },
+    {
+      "epoch": 0.1460708782742681,
+      "grad_norm": 0.7911097407341003,
+      "learning_rate": 1.987332131761161e-05,
+      "loss": 1.6485,
+      "step": 237
+    },
+    {
+      "epoch": 0.14668721109399074,
+      "grad_norm": 0.7268603444099426,
+      "learning_rate": 1.986988193029806e-05,
+      "loss": 1.6167,
+      "step": 238
+    },
+    {
+      "epoch": 0.1473035439137134,
+      "grad_norm": 0.7688180804252625,
+      "learning_rate": 1.9866396781410793e-05,
+      "loss": 1.564,
+      "step": 239
+    },
+    {
+      "epoch": 0.14791987673343607,
+      "grad_norm": 0.7340635061264038,
+      "learning_rate": 1.9862865887108658e-05,
+      "loss": 1.5974,
+      "step": 240
+    },
+    {
+      "epoch": 0.14853620955315872,
+      "grad_norm": 0.7234272360801697,
+      "learning_rate": 1.985928926376259e-05,
+      "loss": 1.6217,
+      "step": 241
+    },
+    {
+      "epoch": 0.14915254237288136,
+      "grad_norm": 0.7681387662887573,
+      "learning_rate": 1.9855666927955565e-05,
+      "loss": 1.6267,
+      "step": 242
+    },
+    {
+      "epoch": 0.149768875192604,
+      "grad_norm": 0.7330013513565063,
+      "learning_rate": 1.985199889648249e-05,
+      "loss": 1.6124,
+      "step": 243
+    },
+    {
+      "epoch": 0.15038520801232666,
+      "grad_norm": 0.7943146824836731,
+      "learning_rate": 1.9848285186350142e-05,
+      "loss": 1.6104,
+      "step": 244
+    },
+    {
+      "epoch": 0.1510015408320493,
+      "grad_norm": 0.7412106990814209,
+      "learning_rate": 1.9844525814777083e-05,
+      "loss": 1.6685,
+      "step": 245
+    },
+    {
+      "epoch": 0.15161787365177196,
+      "grad_norm": 0.7623029351234436,
+      "learning_rate": 1.9840720799193596e-05,
+      "loss": 1.6171,
+      "step": 246
+    },
+    {
+      "epoch": 0.1522342064714946,
+      "grad_norm": 0.7107127904891968,
+      "learning_rate": 1.9836870157241585e-05,
+      "loss": 1.5725,
+      "step": 247
+    },
+    {
+      "epoch": 0.15285053929121725,
+      "grad_norm": 0.7302953004837036,
+      "learning_rate": 1.9832973906774494e-05,
+      "loss": 1.5981,
+      "step": 248
+    },
+    {
+      "epoch": 0.1534668721109399,
+      "grad_norm": 0.7209048271179199,
+      "learning_rate": 1.9829032065857235e-05,
+      "loss": 1.5362,
+      "step": 249
+    },
+    {
+      "epoch": 0.15408320493066255,
+      "grad_norm": 0.7204911112785339,
+      "learning_rate": 1.9825044652766105e-05,
+      "loss": 1.5912,
+      "step": 250
+    },
+    {
+      "epoch": 0.1546995377503852,
+      "grad_norm": 0.7284119129180908,
+      "learning_rate": 1.9821011685988686e-05,
+      "loss": 1.5918,
+      "step": 251
+    },
+    {
+      "epoch": 0.15531587057010784,
+      "grad_norm": 0.724547803401947,
+      "learning_rate": 1.9816933184223774e-05,
+      "loss": 1.7239,
+      "step": 252
+    },
+    {
+      "epoch": 0.15593220338983052,
+      "grad_norm": 0.7248793840408325,
+      "learning_rate": 1.9812809166381292e-05,
+      "loss": 1.678,
+      "step": 253
+    },
+    {
+      "epoch": 0.15654853620955317,
+      "grad_norm": 0.7166767716407776,
+      "learning_rate": 1.9808639651582187e-05,
+      "loss": 1.6679,
+      "step": 254
+    },
+    {
+      "epoch": 0.15716486902927582,
+      "grad_norm": 0.6912389397621155,
+      "learning_rate": 1.9804424659158365e-05,
+      "loss": 1.5748,
+      "step": 255
+    },
+    {
+      "epoch": 0.15778120184899846,
+      "grad_norm": 0.6941654086112976,
+      "learning_rate": 1.980016420865257e-05,
+      "loss": 1.5521,
+      "step": 256
+    },
+    {
+      "epoch": 0.1583975346687211,
+      "grad_norm": 0.7527191042900085,
+      "learning_rate": 1.9795858319818336e-05,
+      "loss": 1.6674,
+      "step": 257
+    },
+    {
+      "epoch": 0.15901386748844376,
+      "grad_norm": 0.7129676938056946,
+      "learning_rate": 1.9791507012619847e-05,
+      "loss": 1.5911,
+      "step": 258
+    },
+    {
+      "epoch": 0.1596302003081664,
+      "grad_norm": 0.7293033003807068,
+      "learning_rate": 1.9787110307231884e-05,
+      "loss": 1.6403,
+      "step": 259
+    },
+    {
+      "epoch": 0.16024653312788906,
+      "grad_norm": 0.6870817542076111,
+      "learning_rate": 1.9782668224039715e-05,
+      "loss": 1.5537,
+      "step": 260
+    },
+    {
+      "epoch": 0.1608628659476117,
+      "grad_norm": 0.731279730796814,
+      "learning_rate": 1.9778180783638997e-05,
+      "loss": 1.6001,
+      "step": 261
+    },
+    {
+      "epoch": 0.16147919876733435,
+      "grad_norm": 0.7369082570075989,
+      "learning_rate": 1.9773648006835682e-05,
+      "loss": 1.6665,
+      "step": 262
+    },
+    {
+      "epoch": 0.162095531587057,
+      "grad_norm": 0.6970654726028442,
+      "learning_rate": 1.9769069914645935e-05,
+      "loss": 1.5774,
+      "step": 263
+    },
+    {
+      "epoch": 0.16271186440677965,
+      "grad_norm": 0.7199523448944092,
+      "learning_rate": 1.9764446528296014e-05,
+      "loss": 1.6013,
+      "step": 264
+    },
+    {
+      "epoch": 0.1633281972265023,
+      "grad_norm": 0.7123932242393494,
+      "learning_rate": 1.9759777869222185e-05,
+      "loss": 1.6028,
+      "step": 265
+    },
+    {
+      "epoch": 0.16394453004622497,
+      "grad_norm": 0.7310228943824768,
+      "learning_rate": 1.975506395907063e-05,
+      "loss": 1.6614,
+      "step": 266
+    },
+    {
+      "epoch": 0.16456086286594762,
+      "grad_norm": 0.7251418232917786,
+      "learning_rate": 1.9750304819697324e-05,
+      "loss": 1.6025,
+      "step": 267
+    },
+    {
+      "epoch": 0.16517719568567027,
+      "grad_norm": 0.7043747901916504,
+      "learning_rate": 1.9745500473167952e-05,
+      "loss": 1.6135,
+      "step": 268
+    },
+    {
+      "epoch": 0.16579352850539292,
+      "grad_norm": 0.7158347368240356,
+      "learning_rate": 1.9740650941757803e-05,
+      "loss": 1.6097,
+      "step": 269
+    },
+    {
+      "epoch": 0.16640986132511557,
+      "grad_norm": 0.6973146200180054,
+      "learning_rate": 1.9735756247951665e-05,
+      "loss": 1.54,
+      "step": 270
+    },
+    {
+      "epoch": 0.1670261941448382,
+      "grad_norm": 0.7155864834785461,
+      "learning_rate": 1.9730816414443715e-05,
+      "loss": 1.5738,
+      "step": 271
+    },
+    {
+      "epoch": 0.16764252696456086,
+      "grad_norm": 0.7250121831893921,
+      "learning_rate": 1.972583146413743e-05,
+      "loss": 1.5991,
+      "step": 272
+    },
+    {
+      "epoch": 0.1682588597842835,
+      "grad_norm": 0.7630172967910767,
+      "learning_rate": 1.972080142014546e-05,
+      "loss": 1.6675,
+      "step": 273
+    },
+    {
+      "epoch": 0.16887519260400616,
+      "grad_norm": 0.711024284362793,
+      "learning_rate": 1.971572630578954e-05,
+      "loss": 1.6108,
+      "step": 274
+    },
+    {
+      "epoch": 0.1694915254237288,
+      "grad_norm": 0.7461122870445251,
+      "learning_rate": 1.971060614460036e-05,
+      "loss": 1.5841,
+      "step": 275
+    },
+    {
+      "epoch": 0.17010785824345145,
+      "grad_norm": 0.7459723949432373,
+      "learning_rate": 1.9705440960317487e-05,
+      "loss": 1.6666,
+      "step": 276
+    },
+    {
+      "epoch": 0.1707241910631741,
+      "grad_norm": 0.753265917301178,
+      "learning_rate": 1.9700230776889228e-05,
+      "loss": 1.6136,
+      "step": 277
+    },
+    {
+      "epoch": 0.17134052388289678,
+      "grad_norm": 0.7313456535339355,
+      "learning_rate": 1.969497561847252e-05,
+      "loss": 1.5842,
+      "step": 278
+    },
+    {
+      "epoch": 0.17195685670261943,
+      "grad_norm": 0.7637221813201904,
+      "learning_rate": 1.968967550943284e-05,
+      "loss": 1.5558,
+      "step": 279
+    },
+    {
+      "epoch": 0.17257318952234207,
+      "grad_norm": 0.7517579197883606,
+      "learning_rate": 1.9684330474344067e-05,
+      "loss": 1.5983,
+      "step": 280
+    },
+    {
+      "epoch": 0.17318952234206472,
+      "grad_norm": 0.7115539312362671,
+      "learning_rate": 1.967894053798839e-05,
+      "loss": 1.5945,
+      "step": 281
+    },
+    {
+      "epoch": 0.17380585516178737,
+      "grad_norm": 0.7166246175765991,
+      "learning_rate": 1.9673505725356174e-05,
+      "loss": 1.6686,
+      "step": 282
+    },
+    {
+      "epoch": 0.17442218798151002,
+      "grad_norm": 0.7135012149810791,
+      "learning_rate": 1.966802606164585e-05,
+      "loss": 1.5752,
+      "step": 283
+    },
+    {
+      "epoch": 0.17503852080123267,
+      "grad_norm": 0.7052460312843323,
+      "learning_rate": 1.966250157226381e-05,
+      "loss": 1.5281,
+      "step": 284
+    },
+    {
+      "epoch": 0.17565485362095531,
+      "grad_norm": 0.7106686234474182,
+      "learning_rate": 1.965693228282427e-05,
+      "loss": 1.6313,
+      "step": 285
+    },
+    {
+      "epoch": 0.17627118644067796,
+      "grad_norm": 0.7450364828109741,
+      "learning_rate": 1.9651318219149167e-05,
+      "loss": 1.6282,
+      "step": 286
+    },
+    {
+      "epoch": 0.1768875192604006,
+      "grad_norm": 0.7307528257369995,
+      "learning_rate": 1.964565940726803e-05,
+      "loss": 1.6233,
+      "step": 287
+    },
+    {
+      "epoch": 0.17750385208012326,
+      "grad_norm": 0.7210361957550049,
+      "learning_rate": 1.9639955873417863e-05,
+      "loss": 1.6397,
+      "step": 288
+    },
+    {
+      "epoch": 0.1781201848998459,
+      "grad_norm": 0.6987190842628479,
+      "learning_rate": 1.963420764404302e-05,
+      "loss": 1.6016,
+      "step": 289
+    },
+    {
+      "epoch": 0.17873651771956856,
+      "grad_norm": 0.7078263759613037,
+      "learning_rate": 1.9628414745795097e-05,
+      "loss": 1.6155,
+      "step": 290
+    },
+    {
+      "epoch": 0.17935285053929123,
+      "grad_norm": 0.7351784110069275,
+      "learning_rate": 1.9622577205532783e-05,
+      "loss": 1.6065,
+      "step": 291
+    },
+    {
+      "epoch": 0.17996918335901388,
+      "grad_norm": 0.7261448502540588,
+      "learning_rate": 1.9616695050321758e-05,
+      "loss": 1.6979,
+      "step": 292
+    },
+    {
+      "epoch": 0.18058551617873653,
+      "grad_norm": 0.6955497860908508,
+      "learning_rate": 1.961076830743455e-05,
+      "loss": 1.6085,
+      "step": 293
+    },
+    {
+      "epoch": 0.18120184899845918,
+      "grad_norm": 0.730446457862854,
+      "learning_rate": 1.9604797004350423e-05,
+      "loss": 1.6557,
+      "step": 294
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 0.718130350112915,
+      "learning_rate": 1.9598781168755242e-05,
+      "loss": 1.6366,
+      "step": 295
+    },
+    {
+      "epoch": 0.18243451463790447,
+      "grad_norm": 0.7395448088645935,
+      "learning_rate": 1.959272082854135e-05,
+      "loss": 1.6067,
+      "step": 296
+    },
+    {
+      "epoch": 0.18305084745762712,
+      "grad_norm": 0.7046165466308594,
+      "learning_rate": 1.9586616011807433e-05,
+      "loss": 1.5762,
+      "step": 297
+    },
+    {
+      "epoch": 0.18366718027734977,
+      "grad_norm": 0.7103109359741211,
+      "learning_rate": 1.958046674685839e-05,
+      "loss": 1.5606,
+      "step": 298
+    },
+    {
+      "epoch": 0.18428351309707242,
+      "grad_norm": 0.6921034455299377,
+      "learning_rate": 1.95742730622052e-05,
+      "loss": 1.5769,
+      "step": 299
+    },
+    {
+      "epoch": 0.18489984591679506,
+      "grad_norm": 0.701803982257843,
+      "learning_rate": 1.9568034986564806e-05,
+      "loss": 1.6225,
+      "step": 300
+    },
+    {
+      "epoch": 0.1855161787365177,
+      "grad_norm": 0.7245356440544128,
+      "learning_rate": 1.9561752548859957e-05,
+      "loss": 1.6337,
+      "step": 301
+    },
+    {
+      "epoch": 0.18613251155624036,
+      "grad_norm": 0.6958240270614624,
+      "learning_rate": 1.9555425778219087e-05,
+      "loss": 1.6244,
+      "step": 302
+    },
+    {
+      "epoch": 0.186748844375963,
+      "grad_norm": 0.7089309692382812,
+      "learning_rate": 1.9549054703976184e-05,
+      "loss": 1.6235,
+      "step": 303
+    },
+    {
+      "epoch": 0.18736517719568568,
+      "grad_norm": 0.7047931551933289,
+      "learning_rate": 1.954263935567065e-05,
+      "loss": 1.6511,
+      "step": 304
+    },
+    {
+      "epoch": 0.18798151001540833,
+      "grad_norm": 0.7393444180488586,
+      "learning_rate": 1.9536179763047158e-05,
+      "loss": 1.6015,
+      "step": 305
+    },
+    {
+      "epoch": 0.18859784283513098,
+      "grad_norm": 0.7115395665168762,
+      "learning_rate": 1.9529675956055522e-05,
+      "loss": 1.5634,
+      "step": 306
+    },
+    {
+      "epoch": 0.18921417565485363,
+      "grad_norm": 0.707436203956604,
+      "learning_rate": 1.9523127964850558e-05,
+      "loss": 1.6305,
+      "step": 307
+    },
+    {
+      "epoch": 0.18983050847457628,
+      "grad_norm": 0.6849664449691772,
+      "learning_rate": 1.951653581979193e-05,
+      "loss": 1.615,
+      "step": 308
+    },
+    {
+      "epoch": 0.19044684129429892,
+      "grad_norm": 0.7077732086181641,
+      "learning_rate": 1.9509899551444044e-05,
+      "loss": 1.5922,
+      "step": 309
+    },
+    {
+      "epoch": 0.19106317411402157,
+      "grad_norm": 0.6861093640327454,
+      "learning_rate": 1.9503219190575856e-05,
+      "loss": 1.5626,
+      "step": 310
+    },
+    {
+      "epoch": 0.19167950693374422,
+      "grad_norm": 0.7092529535293579,
+      "learning_rate": 1.949649476816077e-05,
+      "loss": 1.6065,
+      "step": 311
+    },
+    {
+      "epoch": 0.19229583975346687,
+      "grad_norm": 0.7142940759658813,
+      "learning_rate": 1.948972631537648e-05,
+      "loss": 1.612,
+      "step": 312
+    },
+    {
+      "epoch": 0.19291217257318952,
+      "grad_norm": 0.6780362129211426,
+      "learning_rate": 1.948291386360483e-05,
+      "loss": 1.5681,
+      "step": 313
+    },
+    {
+      "epoch": 0.19352850539291216,
+      "grad_norm": 0.6941437125205994,
+      "learning_rate": 1.9476057444431652e-05,
+      "loss": 1.57,
+      "step": 314
+    },
+    {
+      "epoch": 0.1941448382126348,
+      "grad_norm": 0.6761777400970459,
+      "learning_rate": 1.9469157089646652e-05,
+      "loss": 1.5651,
+      "step": 315
+    },
+    {
+      "epoch": 0.19476117103235746,
+      "grad_norm": 0.7124750018119812,
+      "learning_rate": 1.946221283124322e-05,
+      "loss": 1.5661,
+      "step": 316
+    },
+    {
+      "epoch": 0.19537750385208014,
+      "grad_norm": 0.6963821649551392,
+      "learning_rate": 1.9455224701418323e-05,
+      "loss": 1.5938,
+      "step": 317
+    },
+    {
+      "epoch": 0.19599383667180278,
+      "grad_norm": 0.7073525190353394,
+      "learning_rate": 1.9448192732572324e-05,
+      "loss": 1.5985,
+      "step": 318
+    },
+    {
+      "epoch": 0.19661016949152543,
+      "grad_norm": 0.7923668026924133,
+      "learning_rate": 1.9441116957308858e-05,
+      "loss": 1.6467,
+      "step": 319
+    },
+    {
+      "epoch": 0.19722650231124808,
+      "grad_norm": 0.6834045052528381,
+      "learning_rate": 1.9433997408434658e-05,
+      "loss": 1.585,
+      "step": 320
+    },
+    {
+      "epoch": 0.19784283513097073,
+      "grad_norm": 0.6907963156700134,
+      "learning_rate": 1.9426834118959416e-05,
+      "loss": 1.6078,
+      "step": 321
+    },
+    {
+      "epoch": 0.19845916795069338,
+      "grad_norm": 0.6976985931396484,
+      "learning_rate": 1.941962712209563e-05,
+      "loss": 1.6238,
+      "step": 322
+    },
+    {
+      "epoch": 0.19907550077041603,
+      "grad_norm": 0.7212012410163879,
+      "learning_rate": 1.941237645125843e-05,
+      "loss": 1.562,
+      "step": 323
+    },
+    {
+      "epoch": 0.19969183359013867,
+      "grad_norm": 0.7090001702308655,
+      "learning_rate": 1.9405082140065472e-05,
+      "loss": 1.6038,
+      "step": 324
+    },
+    {
+      "epoch": 0.20030816640986132,
+      "grad_norm": 0.6964325904846191,
+      "learning_rate": 1.9397744222336715e-05,
+      "loss": 1.5163,
+      "step": 325
+    },
+    {
+      "epoch": 0.20092449922958397,
+      "grad_norm": 0.7082476019859314,
+      "learning_rate": 1.9390362732094326e-05,
+      "loss": 1.5492,
+      "step": 326
+    },
+    {
+      "epoch": 0.20154083204930662,
+      "grad_norm": 0.6881728768348694,
+      "learning_rate": 1.9382937703562478e-05,
+      "loss": 1.5635,
+      "step": 327
+    },
+    {
+      "epoch": 0.20215716486902927,
+      "grad_norm": 0.6767646670341492,
+      "learning_rate": 1.9375469171167218e-05,
+      "loss": 1.5846,
+      "step": 328
+    },
+    {
+      "epoch": 0.2027734976887519,
+      "grad_norm": 0.6876394748687744,
+      "learning_rate": 1.9367957169536295e-05,
+      "loss": 1.5427,
+      "step": 329
+    },
+    {
+      "epoch": 0.2033898305084746,
+      "grad_norm": 0.7290824055671692,
+      "learning_rate": 1.9360401733499e-05,
+      "loss": 1.6104,
+      "step": 330
+    },
+    {
+      "epoch": 0.20400616332819724,
+      "grad_norm": 0.6731740832328796,
+      "learning_rate": 1.9352802898086013e-05,
+      "loss": 1.5677,
+      "step": 331
+    },
+    {
+      "epoch": 0.20462249614791989,
+      "grad_norm": 0.7473655939102173,
+      "learning_rate": 1.934516069852923e-05,
+      "loss": 1.5958,
+      "step": 332
+    },
+    {
+      "epoch": 0.20523882896764253,
+      "grad_norm": 0.7044902443885803,
+      "learning_rate": 1.9337475170261602e-05,
+      "loss": 1.5817,
+      "step": 333
+    },
+    {
+      "epoch": 0.20585516178736518,
+      "grad_norm": 0.6740896701812744,
+      "learning_rate": 1.9329746348916982e-05,
+      "loss": 1.6079,
+      "step": 334
+    },
+    {
+      "epoch": 0.20647149460708783,
+      "grad_norm": 0.6931565999984741,
+      "learning_rate": 1.9321974270329946e-05,
+      "loss": 1.6097,
+      "step": 335
+    },
+    {
+      "epoch": 0.20708782742681048,
+      "grad_norm": 0.6814671754837036,
+      "learning_rate": 1.9314158970535627e-05,
+      "loss": 1.5934,
+      "step": 336
+    },
+    {
+      "epoch": 0.20770416024653313,
+      "grad_norm": 0.687995195388794,
+      "learning_rate": 1.930630048576956e-05,
+      "loss": 1.5602,
+      "step": 337
+    },
+    {
+      "epoch": 0.20832049306625577,
+      "grad_norm": 0.6994910836219788,
+      "learning_rate": 1.9298398852467496e-05,
+      "loss": 1.5759,
+      "step": 338
+    },
+    {
+      "epoch": 0.20893682588597842,
+      "grad_norm": 0.6890596747398376,
+      "learning_rate": 1.929045410726526e-05,
+      "loss": 1.5563,
+      "step": 339
+    },
+    {
+      "epoch": 0.20955315870570107,
+      "grad_norm": 0.7027484178543091,
+      "learning_rate": 1.9282466286998548e-05,
+      "loss": 1.639,
+      "step": 340
+    },
+    {
+      "epoch": 0.21016949152542372,
+      "grad_norm": 0.7149859070777893,
+      "learning_rate": 1.9274435428702784e-05,
+      "loss": 1.6095,
+      "step": 341
+    },
+    {
+      "epoch": 0.21078582434514637,
+      "grad_norm": 0.7111890912055969,
+      "learning_rate": 1.9266361569612936e-05,
+      "loss": 1.5791,
+      "step": 342
+    },
+    {
+      "epoch": 0.21140215716486904,
+      "grad_norm": 0.7116836309432983,
+      "learning_rate": 1.9258244747163336e-05,
+      "loss": 1.6052,
+      "step": 343
+    },
+    {
+      "epoch": 0.2120184899845917,
+      "grad_norm": 0.6915231347084045,
+      "learning_rate": 1.9250084998987527e-05,
+      "loss": 1.5525,
+      "step": 344
+    },
+    {
+      "epoch": 0.21263482280431434,
+      "grad_norm": 0.6974709033966064,
+      "learning_rate": 1.9241882362918067e-05,
+      "loss": 1.5777,
+      "step": 345
+    },
+    {
+      "epoch": 0.213251155624037,
+      "grad_norm": 0.6807594299316406,
+      "learning_rate": 1.9233636876986367e-05,
+      "loss": 1.5855,
+      "step": 346
+    },
+    {
+      "epoch": 0.21386748844375963,
+      "grad_norm": 0.700409471988678,
+      "learning_rate": 1.922534857942251e-05,
+      "loss": 1.5784,
+      "step": 347
+    },
+    {
+      "epoch": 0.21448382126348228,
+      "grad_norm": 0.6881166100502014,
+      "learning_rate": 1.921701750865508e-05,
+      "loss": 1.5822,
+      "step": 348
+    },
+    {
+      "epoch": 0.21510015408320493,
+      "grad_norm": 0.6813061237335205,
+      "learning_rate": 1.920864370331097e-05,
+      "loss": 1.5237,
+      "step": 349
+    },
+    {
+      "epoch": 0.21571648690292758,
+      "grad_norm": 0.699319064617157,
+      "learning_rate": 1.9200227202215215e-05,
+      "loss": 1.6114,
+      "step": 350
+    },
+    {
+      "epoch": 0.21633281972265023,
+      "grad_norm": 0.6909384727478027,
+      "learning_rate": 1.9191768044390812e-05,
+      "loss": 1.5855,
+      "step": 351
+    },
+    {
+      "epoch": 0.21694915254237288,
+      "grad_norm": 0.7082647085189819,
+      "learning_rate": 1.918326626905852e-05,
+      "loss": 1.6683,
+      "step": 352
+    },
+    {
+      "epoch": 0.21756548536209552,
+      "grad_norm": 0.677087128162384,
+      "learning_rate": 1.917472191563672e-05,
+      "loss": 1.5157,
+      "step": 353
+    },
+    {
+      "epoch": 0.21818181818181817,
+      "grad_norm": 0.6834883093833923,
+      "learning_rate": 1.9166135023741183e-05,
+      "loss": 1.5746,
+      "step": 354
+    },
+    {
+      "epoch": 0.21879815100154082,
+      "grad_norm": 0.6941872835159302,
+      "learning_rate": 1.9157505633184915e-05,
+      "loss": 1.6262,
+      "step": 355
+    },
+    {
+      "epoch": 0.2194144838212635,
+      "grad_norm": 0.7219916582107544,
+      "learning_rate": 1.9148833783977974e-05,
+      "loss": 1.6341,
+      "step": 356
+    },
+    {
+      "epoch": 0.22003081664098614,
+      "grad_norm": 0.6782143712043762,
+      "learning_rate": 1.9140119516327268e-05,
+      "loss": 1.584,
+      "step": 357
+    },
+    {
+      "epoch": 0.2206471494607088,
+      "grad_norm": 0.7127141952514648,
+      "learning_rate": 1.913136287063638e-05,
+      "loss": 1.622,
+      "step": 358
+    },
+    {
+      "epoch": 0.22126348228043144,
+      "grad_norm": 0.6856649518013,
+      "learning_rate": 1.9122563887505382e-05,
+      "loss": 1.5578,
+      "step": 359
+    },
+    {
+      "epoch": 0.2218798151001541,
+      "grad_norm": 0.7305546402931213,
+      "learning_rate": 1.9113722607730635e-05,
+      "loss": 1.5919,
+      "step": 360
+    },
+    {
+      "epoch": 0.22249614791987674,
+      "grad_norm": 0.7384188771247864,
+      "learning_rate": 1.9104839072304617e-05,
+      "loss": 1.583,
+      "step": 361
+    },
+    {
+      "epoch": 0.22311248073959938,
+      "grad_norm": 0.7033206224441528,
+      "learning_rate": 1.909591332241572e-05,
+      "loss": 1.5924,
+      "step": 362
+    },
+    {
+      "epoch": 0.22372881355932203,
+      "grad_norm": 0.7022284269332886,
+      "learning_rate": 1.908694539944806e-05,
+      "loss": 1.5737,
+      "step": 363
+    },
+    {
+      "epoch": 0.22434514637904468,
+      "grad_norm": 0.6830691695213318,
+      "learning_rate": 1.907793534498129e-05,
+      "loss": 1.5969,
+      "step": 364
+    },
+    {
+      "epoch": 0.22496147919876733,
+      "grad_norm": 0.6887118816375732,
+      "learning_rate": 1.9068883200790407e-05,
+      "loss": 1.5703,
+      "step": 365
+    },
+    {
+      "epoch": 0.22557781201848998,
+      "grad_norm": 0.6918419003486633,
+      "learning_rate": 1.9059789008845555e-05,
+      "loss": 1.5863,
+      "step": 366
+    },
+    {
+      "epoch": 0.22619414483821262,
+      "grad_norm": 0.6840692162513733,
+      "learning_rate": 1.905065281131183e-05,
+      "loss": 1.5963,
+      "step": 367
+    },
+    {
+      "epoch": 0.22681047765793527,
+      "grad_norm": 0.7187916040420532,
+      "learning_rate": 1.904147465054909e-05,
+      "loss": 1.5624,
+      "step": 368
+    },
+    {
+      "epoch": 0.22742681047765795,
+      "grad_norm": 0.681559681892395,
+      "learning_rate": 1.903225456911175e-05,
+      "loss": 1.5976,
+      "step": 369
+    },
+    {
+      "epoch": 0.2280431432973806,
+      "grad_norm": 0.680204451084137,
+      "learning_rate": 1.902299260974859e-05,
+      "loss": 1.5661,
+      "step": 370
+    },
+    {
+      "epoch": 0.22865947611710324,
+      "grad_norm": 0.6765556931495667,
+      "learning_rate": 1.9013688815402566e-05,
+      "loss": 1.6032,
+      "step": 371
+    },
+    {
+      "epoch": 0.2292758089368259,
+      "grad_norm": 0.6932162046432495,
+      "learning_rate": 1.900434322921059e-05,
+      "loss": 1.6811,
+      "step": 372
+    },
+    {
+      "epoch": 0.22989214175654854,
+      "grad_norm": 0.6818493604660034,
+      "learning_rate": 1.8994955894503354e-05,
+      "loss": 1.5797,
+      "step": 373
+    },
+    {
+      "epoch": 0.2305084745762712,
+      "grad_norm": 0.703291118144989,
+      "learning_rate": 1.8985526854805095e-05,
+      "loss": 1.5776,
+      "step": 374
+    },
+    {
+      "epoch": 0.23112480739599384,
+      "grad_norm": 0.6792005896568298,
+      "learning_rate": 1.8976056153833434e-05,
+      "loss": 1.5505,
+      "step": 375
+    },
+    {
+      "epoch": 0.23174114021571648,
+      "grad_norm": 0.6935883164405823,
+      "learning_rate": 1.8966543835499147e-05,
+      "loss": 1.6094,
+      "step": 376
+    },
+    {
+      "epoch": 0.23235747303543913,
+      "grad_norm": 0.6876742243766785,
+      "learning_rate": 1.895698994390597e-05,
+      "loss": 1.5908,
+      "step": 377
+    },
+    {
+      "epoch": 0.23297380585516178,
+      "grad_norm": 0.6971921324729919,
+      "learning_rate": 1.894739452335039e-05,
+      "loss": 1.587,
+      "step": 378
+    },
+    {
+      "epoch": 0.23359013867488443,
+      "grad_norm": 0.6786790490150452,
+      "learning_rate": 1.8937757618321447e-05,
+      "loss": 1.5993,
+      "step": 379
+    },
+    {
+      "epoch": 0.23420647149460708,
+      "grad_norm": 0.7139135003089905,
+      "learning_rate": 1.892807927350052e-05,
+      "loss": 1.5959,
+      "step": 380
+    },
+    {
+      "epoch": 0.23482280431432973,
+      "grad_norm": 0.6749522686004639,
+      "learning_rate": 1.8918359533761118e-05,
+      "loss": 1.5887,
+      "step": 381
+    },
+    {
+      "epoch": 0.2354391371340524,
+      "grad_norm": 0.680397629737854,
+      "learning_rate": 1.8908598444168694e-05,
+      "loss": 1.6156,
+      "step": 382
+    },
+    {
+      "epoch": 0.23605546995377505,
+      "grad_norm": 0.6886288523674011,
+      "learning_rate": 1.8898796049980394e-05,
+      "loss": 1.5586,
+      "step": 383
+    },
+    {
+      "epoch": 0.2366718027734977,
+      "grad_norm": 0.6770725846290588,
+      "learning_rate": 1.8888952396644896e-05,
+      "loss": 1.5875,
+      "step": 384
+    },
+    {
+      "epoch": 0.23728813559322035,
+      "grad_norm": 0.6947922706604004,
+      "learning_rate": 1.8879067529802158e-05,
+      "loss": 1.5651,
+      "step": 385
+    },
+    {
+      "epoch": 0.237904468412943,
+      "grad_norm": 0.673457145690918,
+      "learning_rate": 1.8869141495283235e-05,
+      "loss": 1.5599,
+      "step": 386
+    },
+    {
+      "epoch": 0.23852080123266564,
+      "grad_norm": 0.6934700608253479,
+      "learning_rate": 1.885917433911005e-05,
+      "loss": 1.5688,
+      "step": 387
+    },
+    {
+      "epoch": 0.2391371340523883,
+      "grad_norm": 0.714322566986084,
+      "learning_rate": 1.8849166107495192e-05,
+      "loss": 1.5519,
+      "step": 388
+    },
+    {
+      "epoch": 0.23975346687211094,
+      "grad_norm": 0.6958152651786804,
+      "learning_rate": 1.883911684684168e-05,
+      "loss": 1.6134,
+      "step": 389
+    },
+    {
+      "epoch": 0.24036979969183359,
+      "grad_norm": 0.71065753698349,
+      "learning_rate": 1.882902660374279e-05,
+      "loss": 1.6194,
+      "step": 390
+    },
+    {
+      "epoch": 0.24098613251155623,
+      "grad_norm": 0.6988871693611145,
+      "learning_rate": 1.8818895424981786e-05,
+      "loss": 1.5894,
+      "step": 391
+    },
+    {
+      "epoch": 0.24160246533127888,
+      "grad_norm": 0.71463942527771,
+      "learning_rate": 1.8808723357531744e-05,
+      "loss": 1.6315,
+      "step": 392
+    },
+    {
+      "epoch": 0.24221879815100153,
+      "grad_norm": 0.6897231936454773,
+      "learning_rate": 1.8798510448555317e-05,
+      "loss": 1.6058,
+      "step": 393
+    },
+    {
+      "epoch": 0.24283513097072418,
+      "grad_norm": 0.6789922118186951,
+      "learning_rate": 1.8788256745404522e-05,
+      "loss": 1.6044,
+      "step": 394
+    },
+    {
+      "epoch": 0.24345146379044685,
+      "grad_norm": 0.6774851083755493,
+      "learning_rate": 1.8777962295620507e-05,
+      "loss": 1.5405,
+      "step": 395
+    },
+    {
+      "epoch": 0.2440677966101695,
+      "grad_norm": 0.6831008195877075,
+      "learning_rate": 1.8767627146933355e-05,
+      "loss": 1.5348,
+      "step": 396
+    },
+    {
+      "epoch": 0.24468412942989215,
+      "grad_norm": 0.6807131171226501,
+      "learning_rate": 1.875725134726184e-05,
+      "loss": 1.547,
+      "step": 397
+    },
+    {
+      "epoch": 0.2453004622496148,
+      "grad_norm": 0.6741266250610352,
+      "learning_rate": 1.8746834944713225e-05,
+      "loss": 1.6282,
+      "step": 398
+    },
+    {
+      "epoch": 0.24591679506933745,
+      "grad_norm": 0.6683118343353271,
+      "learning_rate": 1.8736377987583004e-05,
+      "loss": 1.5459,
+      "step": 399
+    },
+    {
+      "epoch": 0.2465331278890601,
+      "grad_norm": 0.6792947053909302,
+      "learning_rate": 1.8725880524354724e-05,
+      "loss": 1.5778,
+      "step": 400
+    },
+    {
+      "epoch": 0.24714946070878274,
+      "grad_norm": 0.69471275806427,
+      "learning_rate": 1.8715342603699736e-05,
+      "loss": 1.5562,
+      "step": 401
+    },
+    {
+      "epoch": 0.2477657935285054,
+      "grad_norm": 0.6657786965370178,
+      "learning_rate": 1.870476427447696e-05,
+      "loss": 1.5258,
+      "step": 402
+    },
+    {
+      "epoch": 0.24838212634822804,
+      "grad_norm": 0.6800218820571899,
+      "learning_rate": 1.8694145585732676e-05,
+      "loss": 1.4972,
+      "step": 403
+    },
+    {
+      "epoch": 0.2489984591679507,
+      "grad_norm": 0.6874195337295532,
+      "learning_rate": 1.8683486586700295e-05,
+      "loss": 1.6229,
+      "step": 404
+    },
+    {
+      "epoch": 0.24961479198767333,
+      "grad_norm": 0.7023640275001526,
+      "learning_rate": 1.8672787326800117e-05,
+      "loss": 1.4995,
+      "step": 405
+    },
+    {
+      "epoch": 0.250231124807396,
+      "grad_norm": 0.6925647258758545,
+      "learning_rate": 1.8662047855639118e-05,
+      "loss": 1.5911,
+      "step": 406
+    },
+    {
+      "epoch": 0.25084745762711863,
+      "grad_norm": 0.7348222136497498,
+      "learning_rate": 1.865126822301071e-05,
+      "loss": 1.5896,
+      "step": 407
+    },
+    {
+      "epoch": 0.2514637904468413,
+      "grad_norm": 0.6841826438903809,
+      "learning_rate": 1.8640448478894512e-05,
+      "loss": 1.6029,
+      "step": 408
+    },
+    {
+      "epoch": 0.2520801232665639,
+      "grad_norm": 0.690205454826355,
+      "learning_rate": 1.862958867345612e-05,
+      "loss": 1.5588,
+      "step": 409
+    },
+    {
+      "epoch": 0.2526964560862866,
+      "grad_norm": 0.6769512891769409,
+      "learning_rate": 1.8618688857046878e-05,
+      "loss": 1.5575,
+      "step": 410
+    },
+    {
+      "epoch": 0.2533127889060092,
+      "grad_norm": 0.6988481283187866,
+      "learning_rate": 1.8607749080203638e-05,
+      "loss": 1.6051,
+      "step": 411
+    },
+    {
+      "epoch": 0.2539291217257319,
+      "grad_norm": 0.7186166644096375,
+      "learning_rate": 1.859676939364852e-05,
+      "loss": 1.6317,
+      "step": 412
+    },
+    {
+      "epoch": 0.2545454545454545,
+      "grad_norm": 0.668711245059967,
+      "learning_rate": 1.8585749848288695e-05,
+      "loss": 1.5778,
+      "step": 413
+    },
+    {
+      "epoch": 0.2551617873651772,
+      "grad_norm": 0.6950772404670715,
+      "learning_rate": 1.8574690495216142e-05,
+      "loss": 1.5782,
+      "step": 414
+    },
+    {
+      "epoch": 0.25577812018489987,
+      "grad_norm": 0.6883686184883118,
+      "learning_rate": 1.8563591385707394e-05,
+      "loss": 1.575,
+      "step": 415
+    },
+    {
+      "epoch": 0.2563944530046225,
+      "grad_norm": 0.687495231628418,
+      "learning_rate": 1.855245257122332e-05,
+      "loss": 1.5573,
+      "step": 416
+    },
+    {
+      "epoch": 0.25701078582434517,
+      "grad_norm": 0.7103877067565918,
+      "learning_rate": 1.854127410340889e-05,
+      "loss": 1.5932,
+      "step": 417
+    },
+    {
+      "epoch": 0.2576271186440678,
+      "grad_norm": 0.6886118054389954,
+      "learning_rate": 1.8530056034092917e-05,
+      "loss": 1.6114,
+      "step": 418
+    },
+    {
+      "epoch": 0.25824345146379046,
+      "grad_norm": 0.6920944452285767,
+      "learning_rate": 1.851879841528782e-05,
+      "loss": 1.5686,
+      "step": 419
+    },
+    {
+      "epoch": 0.2588597842835131,
+      "grad_norm": 0.6760983467102051,
+      "learning_rate": 1.8507501299189402e-05,
+      "loss": 1.5176,
+      "step": 420
+    },
+    {
+      "epoch": 0.25947611710323576,
+      "grad_norm": 0.6778839826583862,
+      "learning_rate": 1.8496164738176592e-05,
+      "loss": 1.5329,
+      "step": 421
+    },
+    {
+      "epoch": 0.2600924499229584,
+      "grad_norm": 0.6755695939064026,
+      "learning_rate": 1.848478878481119e-05,
+      "loss": 1.503,
+      "step": 422
+    },
+    {
+      "epoch": 0.26070878274268106,
+      "grad_norm": 0.7045766115188599,
+      "learning_rate": 1.847337349183766e-05,
+      "loss": 1.6069,
+      "step": 423
+    },
+    {
+      "epoch": 0.2613251155624037,
+      "grad_norm": 0.6928831338882446,
+      "learning_rate": 1.8461918912182852e-05,
+      "loss": 1.6036,
+      "step": 424
+    },
+    {
+      "epoch": 0.26194144838212635,
+      "grad_norm": 0.7048707604408264,
+      "learning_rate": 1.8450425098955763e-05,
+      "loss": 1.6118,
+      "step": 425
+    },
+    {
+      "epoch": 0.26255778120184897,
+      "grad_norm": 0.704757571220398,
+      "learning_rate": 1.8438892105447306e-05,
+      "loss": 1.6316,
+      "step": 426
+    },
+    {
+      "epoch": 0.26317411402157165,
+      "grad_norm": 0.7133340239524841,
+      "learning_rate": 1.842731998513005e-05,
+      "loss": 1.5613,
+      "step": 427
+    },
+    {
+      "epoch": 0.2637904468412943,
+      "grad_norm": 0.6957772374153137,
+      "learning_rate": 1.841570879165797e-05,
+      "loss": 1.5847,
+      "step": 428
+    },
+    {
+      "epoch": 0.26440677966101694,
+      "grad_norm": 0.673180878162384,
+      "learning_rate": 1.8404058578866217e-05,
+      "loss": 1.5252,
+      "step": 429
+    },
+    {
+      "epoch": 0.2650231124807396,
+      "grad_norm": 0.6939311027526855,
+      "learning_rate": 1.839236940077083e-05,
+      "loss": 1.563,
+      "step": 430
+    },
+    {
+      "epoch": 0.26563944530046224,
+      "grad_norm": 0.6830098032951355,
+      "learning_rate": 1.8380641311568534e-05,
+      "loss": 1.5665,
+      "step": 431
+    },
+    {
+      "epoch": 0.2662557781201849,
+      "grad_norm": 0.6832227110862732,
+      "learning_rate": 1.8368874365636454e-05,
+      "loss": 1.6339,
+      "step": 432
+    },
+    {
+      "epoch": 0.26687211093990754,
+      "grad_norm": 0.7159360647201538,
+      "learning_rate": 1.8357068617531874e-05,
+      "loss": 1.6089,
+      "step": 433
+    },
+    {
+      "epoch": 0.2674884437596302,
+      "grad_norm": 0.6816911101341248,
+      "learning_rate": 1.8345224121991986e-05,
+      "loss": 1.5701,
+      "step": 434
+    },
+    {
+      "epoch": 0.26810477657935283,
+      "grad_norm": 0.6883374452590942,
+      "learning_rate": 1.833334093393363e-05,
+      "loss": 1.5561,
+      "step": 435
+    },
+    {
+      "epoch": 0.2687211093990755,
+      "grad_norm": 0.7020498514175415,
+      "learning_rate": 1.8321419108453054e-05,
+      "loss": 1.5848,
+      "step": 436
+    },
+    {
+      "epoch": 0.26933744221879813,
+      "grad_norm": 0.691039502620697,
+      "learning_rate": 1.8309458700825635e-05,
+      "loss": 1.6354,
+      "step": 437
+    },
+    {
+      "epoch": 0.2699537750385208,
+      "grad_norm": 0.6816720962524414,
+      "learning_rate": 1.8297459766505637e-05,
+      "loss": 1.5966,
+      "step": 438
+    },
+    {
+      "epoch": 0.2705701078582434,
+      "grad_norm": 0.6899356245994568,
+      "learning_rate": 1.8285422361125964e-05,
+      "loss": 1.5849,
+      "step": 439
+    },
+    {
+      "epoch": 0.2711864406779661,
+      "grad_norm": 0.6955758929252625,
+      "learning_rate": 1.8273346540497876e-05,
+      "loss": 1.6226,
+      "step": 440
+    },
+    {
+      "epoch": 0.2718027734976888,
+      "grad_norm": 0.6872656941413879,
+      "learning_rate": 1.8261232360610757e-05,
+      "loss": 1.5381,
+      "step": 441
+    },
+    {
+      "epoch": 0.2724191063174114,
+      "grad_norm": 0.7074469923973083,
+      "learning_rate": 1.8249079877631833e-05,
+      "loss": 1.6217,
+      "step": 442
+    },
+    {
+      "epoch": 0.2730354391371341,
+      "grad_norm": 0.6840736269950867,
+      "learning_rate": 1.823688914790593e-05,
+      "loss": 1.6315,
+      "step": 443
+    },
+    {
+      "epoch": 0.2736517719568567,
+      "grad_norm": 0.6685308218002319,
+      "learning_rate": 1.82246602279552e-05,
+      "loss": 1.536,
+      "step": 444
+    },
+    {
+      "epoch": 0.27426810477657937,
+      "grad_norm": 0.6902186870574951,
+      "learning_rate": 1.8212393174478863e-05,
+      "loss": 1.556,
+      "step": 445
+    },
+    {
+      "epoch": 0.274884437596302,
+      "grad_norm": 0.672926664352417,
+      "learning_rate": 1.8200088044352952e-05,
+      "loss": 1.508,
+      "step": 446
+    },
+    {
+      "epoch": 0.27550077041602467,
+      "grad_norm": 0.693885326385498,
+      "learning_rate": 1.8187744894630025e-05,
+      "loss": 1.5791,
+      "step": 447
+    },
+    {
+      "epoch": 0.2761171032357473,
+      "grad_norm": 0.7051182985305786,
+      "learning_rate": 1.8175363782538938e-05,
+      "loss": 1.5812,
+      "step": 448
+    },
+    {
+      "epoch": 0.27673343605546996,
+      "grad_norm": 0.6827343106269836,
+      "learning_rate": 1.8162944765484548e-05,
+      "loss": 1.5527,
+      "step": 449
+    },
+    {
+      "epoch": 0.2773497688751926,
+      "grad_norm": 0.6803579926490784,
+      "learning_rate": 1.8150487901047452e-05,
+      "loss": 1.5791,
+      "step": 450
+    },
+    {
+      "epoch": 0.27796610169491526,
+      "grad_norm": 0.6807874441146851,
+      "learning_rate": 1.813799324698374e-05,
+      "loss": 1.551,
+      "step": 451
+    },
+    {
+      "epoch": 0.2785824345146379,
+      "grad_norm": 0.6958891749382019,
+      "learning_rate": 1.8125460861224708e-05,
+      "loss": 1.5531,
+      "step": 452
+    },
+    {
+      "epoch": 0.27919876733436055,
+      "grad_norm": 0.6839728951454163,
+      "learning_rate": 1.8112890801876587e-05,
+      "loss": 1.5573,
+      "step": 453
+    },
+    {
+      "epoch": 0.27981510015408323,
+      "grad_norm": 0.6904528737068176,
+      "learning_rate": 1.8100283127220287e-05,
+      "loss": 1.5869,
+      "step": 454
+    },
+    {
+      "epoch": 0.28043143297380585,
+      "grad_norm": 0.6965783834457397,
+      "learning_rate": 1.808763789571112e-05,
+      "loss": 1.5958,
+      "step": 455
+    },
+    {
+      "epoch": 0.2810477657935285,
+      "grad_norm": 0.6977277398109436,
+      "learning_rate": 1.8074955165978538e-05,
+      "loss": 1.5924,
+      "step": 456
+    },
+    {
+      "epoch": 0.28166409861325115,
+      "grad_norm": 0.6901462078094482,
+      "learning_rate": 1.8062234996825836e-05,
+      "loss": 1.5896,
+      "step": 457
+    },
+    {
+      "epoch": 0.2822804314329738,
+      "grad_norm": 0.7124237418174744,
+      "learning_rate": 1.8049477447229914e-05,
+      "loss": 1.5563,
+      "step": 458
+    },
+    {
+      "epoch": 0.28289676425269644,
+      "grad_norm": 0.6746652722358704,
+      "learning_rate": 1.803668257634097e-05,
+      "loss": 1.5079,
+      "step": 459
+    },
+    {
+      "epoch": 0.2835130970724191,
+      "grad_norm": 0.7127771377563477,
+      "learning_rate": 1.802385044348225e-05,
+      "loss": 1.6488,
+      "step": 460
+    },
+    {
+      "epoch": 0.28412942989214174,
+      "grad_norm": 0.6728559136390686,
+      "learning_rate": 1.8010981108149772e-05,
+      "loss": 1.551,
+      "step": 461
+    },
+    {
+      "epoch": 0.2847457627118644,
+      "grad_norm": 0.6707913279533386,
+      "learning_rate": 1.799807463001203e-05,
+      "loss": 1.5679,
+      "step": 462
+    },
+    {
+      "epoch": 0.28536209553158703,
+      "grad_norm": 0.7097390294075012,
+      "learning_rate": 1.7985131068909728e-05,
+      "loss": 1.5667,
+      "step": 463
+    },
+    {
+      "epoch": 0.2859784283513097,
+      "grad_norm": 0.688072144985199,
+      "learning_rate": 1.797215048485552e-05,
+      "loss": 1.5681,
+      "step": 464
+    },
+    {
+      "epoch": 0.28659476117103233,
+      "grad_norm": 0.709757924079895,
+      "learning_rate": 1.795913293803371e-05,
+      "loss": 1.5828,
+      "step": 465
+    },
+    {
+      "epoch": 0.287211093990755,
+      "grad_norm": 0.7187873721122742,
+      "learning_rate": 1.7946078488799973e-05,
+      "loss": 1.5468,
+      "step": 466
+    },
+    {
+      "epoch": 0.2878274268104777,
+      "grad_norm": 0.7039961218833923,
+      "learning_rate": 1.793298719768108e-05,
+      "loss": 1.5224,
+      "step": 467
+    },
+    {
+      "epoch": 0.2884437596302003,
+      "grad_norm": 0.6894049048423767,
+      "learning_rate": 1.7919859125374634e-05,
+      "loss": 1.4893,
+      "step": 468
+    },
+    {
+      "epoch": 0.289060092449923,
+      "grad_norm": 0.7140965461730957,
+      "learning_rate": 1.790669433274876e-05,
+      "loss": 1.5729,
+      "step": 469
+    },
+    {
+      "epoch": 0.2896764252696456,
+      "grad_norm": 0.6724892854690552,
+      "learning_rate": 1.7893492880841843e-05,
+      "loss": 1.5117,
+      "step": 470
+    },
+    {
+      "epoch": 0.2902927580893683,
+      "grad_norm": 0.6789978742599487,
+      "learning_rate": 1.7880254830862234e-05,
+      "loss": 1.5124,
+      "step": 471
+    },
+    {
+      "epoch": 0.2909090909090909,
+      "grad_norm": 0.6928354501724243,
+      "learning_rate": 1.7866980244187973e-05,
+      "loss": 1.5321,
+      "step": 472
+    },
+    {
+      "epoch": 0.29152542372881357,
+      "grad_norm": 0.697063684463501,
+      "learning_rate": 1.78536691823665e-05,
+      "loss": 1.5868,
+      "step": 473
+    },
+    {
+      "epoch": 0.2921417565485362,
+      "grad_norm": 0.7211458086967468,
+      "learning_rate": 1.784032170711437e-05,
+      "loss": 1.5857,
+      "step": 474
+    },
+    {
+      "epoch": 0.29275808936825887,
+      "grad_norm": 0.6981489658355713,
+      "learning_rate": 1.7826937880316973e-05,
+      "loss": 1.5611,
+      "step": 475
+    },
+    {
+      "epoch": 0.2933744221879815,
+      "grad_norm": 0.6887120008468628,
+      "learning_rate": 1.7813517764028246e-05,
+      "loss": 1.6294,
+      "step": 476
+    },
+    {
+      "epoch": 0.29399075500770416,
+      "grad_norm": 0.7402843236923218,
+      "learning_rate": 1.780006142047037e-05,
+      "loss": 1.6022,
+      "step": 477
+    },
+    {
+      "epoch": 0.2946070878274268,
+      "grad_norm": 0.6611169576644897,
+      "learning_rate": 1.77865689120335e-05,
+      "loss": 1.5585,
+      "step": 478
+    },
+    {
+      "epoch": 0.29522342064714946,
+      "grad_norm": 0.697653591632843,
+      "learning_rate": 1.7773040301275467e-05,
+      "loss": 1.5447,
+      "step": 479
+    },
+    {
+      "epoch": 0.29583975346687214,
+      "grad_norm": 0.7073349952697754,
+      "learning_rate": 1.77594756509215e-05,
+      "loss": 1.5383,
+      "step": 480
+    },
+    {
+      "epoch": 0.29645608628659476,
+      "grad_norm": 0.6784787774085999,
+      "learning_rate": 1.7745875023863907e-05,
+      "loss": 1.5981,
+      "step": 481
+    },
+    {
+      "epoch": 0.29707241910631743,
+      "grad_norm": 0.6896160840988159,
+      "learning_rate": 1.7732238483161813e-05,
+      "loss": 1.5552,
+      "step": 482
+    },
+    {
+      "epoch": 0.29768875192604005,
+      "grad_norm": 0.6995619535446167,
+      "learning_rate": 1.7718566092040854e-05,
+      "loss": 1.5687,
+      "step": 483
+    },
+    {
+      "epoch": 0.2983050847457627,
+      "grad_norm": 0.6991766691207886,
+      "learning_rate": 1.770485791389289e-05,
+      "loss": 1.5024,
+      "step": 484
+    },
+    {
+      "epoch": 0.29892141756548535,
+      "grad_norm": 0.7000912427902222,
+      "learning_rate": 1.7691114012275695e-05,
+      "loss": 1.5396,
+      "step": 485
+    },
+    {
+      "epoch": 0.299537750385208,
+      "grad_norm": 0.6898724436759949,
+      "learning_rate": 1.7677334450912694e-05,
+      "loss": 1.5565,
+      "step": 486
+    },
+    {
+      "epoch": 0.30015408320493064,
+      "grad_norm": 0.6801909804344177,
+      "learning_rate": 1.7663519293692625e-05,
+      "loss": 1.6145,
+      "step": 487
+    },
+    {
+      "epoch": 0.3007704160246533,
+      "grad_norm": 0.7044969797134399,
+      "learning_rate": 1.7649668604669283e-05,
+      "loss": 1.5615,
+      "step": 488
+    },
+    {
+      "epoch": 0.30138674884437594,
+      "grad_norm": 0.7032700777053833,
+      "learning_rate": 1.76357824480612e-05,
+      "loss": 1.6474,
+      "step": 489
+    },
+    {
+      "epoch": 0.3020030816640986,
+      "grad_norm": 0.6888188123703003,
+      "learning_rate": 1.762186088825135e-05,
+      "loss": 1.5321,
+      "step": 490
+    },
+    {
+      "epoch": 0.30261941448382124,
+      "grad_norm": 0.6713204383850098,
+      "learning_rate": 1.7607903989786862e-05,
+      "loss": 1.5644,
+      "step": 491
+    },
+    {
+      "epoch": 0.3032357473035439,
+      "grad_norm": 0.7006869912147522,
+      "learning_rate": 1.7593911817378698e-05,
+      "loss": 1.5822,
+      "step": 492
+    },
+    {
+      "epoch": 0.3038520801232666,
+      "grad_norm": 0.6710270047187805,
+      "learning_rate": 1.7579884435901383e-05,
+      "loss": 1.5731,
+      "step": 493
+    },
+    {
+      "epoch": 0.3044684129429892,
+      "grad_norm": 0.6671738624572754,
+      "learning_rate": 1.7565821910392677e-05,
+      "loss": 1.5531,
+      "step": 494
+    },
+    {
+      "epoch": 0.3050847457627119,
+      "grad_norm": 0.6738025546073914,
+      "learning_rate": 1.7551724306053288e-05,
+      "loss": 1.5852,
+      "step": 495
+    },
+    {
+      "epoch": 0.3057010785824345,
+      "grad_norm": 0.7034854888916016,
+      "learning_rate": 1.753759168824657e-05,
+      "loss": 1.627,
+      "step": 496
+    },
+    {
+      "epoch": 0.3063174114021572,
+      "grad_norm": 0.6891358494758606,
+      "learning_rate": 1.752342412249821e-05,
+      "loss": 1.5522,
+      "step": 497
+    },
+    {
+      "epoch": 0.3069337442218798,
+      "grad_norm": 0.6770594716072083,
+      "learning_rate": 1.750922167449594e-05,
+      "loss": 1.6046,
+      "step": 498
+    },
+    {
+      "epoch": 0.3075500770416025,
+      "grad_norm": 0.6736787557601929,
+      "learning_rate": 1.7494984410089216e-05,
+      "loss": 1.4923,
+      "step": 499
+    },
+    {
+      "epoch": 0.3081664098613251,
+      "grad_norm": 0.6681939959526062,
+      "learning_rate": 1.7480712395288917e-05,
+      "loss": 1.5038,
+      "step": 500
+    },
+    {
+      "epoch": 0.3087827426810478,
+      "grad_norm": 0.650123655796051,
+      "learning_rate": 1.7466405696267052e-05,
+      "loss": 1.5375,
+      "step": 501
+    },
+    {
+      "epoch": 0.3093990755007704,
+      "grad_norm": 0.6611471176147461,
+      "learning_rate": 1.7452064379356435e-05,
+      "loss": 1.4954,
+      "step": 502
+    },
+    {
+      "epoch": 0.31001540832049307,
+      "grad_norm": 0.7062728404998779,
+      "learning_rate": 1.7437688511050384e-05,
+      "loss": 1.5472,
+      "step": 503
+    },
+    {
+      "epoch": 0.3106317411402157,
+      "grad_norm": 0.6930948495864868,
+      "learning_rate": 1.7423278158002422e-05,
+      "loss": 1.5964,
+      "step": 504
+    },
+    {
+      "epoch": 0.31124807395993837,
+      "grad_norm": 0.6913105845451355,
+      "learning_rate": 1.7408833387025954e-05,
+      "loss": 1.587,
+      "step": 505
+    },
+    {
+      "epoch": 0.31186440677966104,
+      "grad_norm": 0.6681187748908997,
+      "learning_rate": 1.739435426509396e-05,
+      "loss": 1.5687,
+      "step": 506
+    },
+    {
+      "epoch": 0.31248073959938366,
+      "grad_norm": 0.6620685458183289,
+      "learning_rate": 1.7379840859338705e-05,
+      "loss": 1.5263,
+      "step": 507
+    },
+    {
+      "epoch": 0.31309707241910634,
+      "grad_norm": 0.7140778303146362,
+      "learning_rate": 1.7365293237051385e-05,
+      "loss": 1.5402,
+      "step": 508
+    },
+    {
+      "epoch": 0.31371340523882896,
+      "grad_norm": 0.6758169531822205,
+      "learning_rate": 1.735071146568185e-05,
+      "loss": 1.6017,
+      "step": 509
+    },
+    {
+      "epoch": 0.31432973805855163,
+      "grad_norm": 0.6911832094192505,
+      "learning_rate": 1.7336095612838306e-05,
+      "loss": 1.5636,
+      "step": 510
+    },
+    {
+      "epoch": 0.31494607087827425,
+      "grad_norm": 0.7011511325836182,
+      "learning_rate": 1.7321445746286936e-05,
+      "loss": 1.5581,
+      "step": 511
+    },
+    {
+      "epoch": 0.31556240369799693,
+      "grad_norm": 0.6647090911865234,
+      "learning_rate": 1.7306761933951662e-05,
+      "loss": 1.583,
+      "step": 512
+    },
+    {
+      "epoch": 0.31617873651771955,
+      "grad_norm": 0.6769677996635437,
+      "learning_rate": 1.729204424391377e-05,
+      "loss": 1.5152,
+      "step": 513
+    },
+    {
+      "epoch": 0.3167950693374422,
+      "grad_norm": 0.6838423609733582,
+      "learning_rate": 1.7277292744411632e-05,
+      "loss": 1.5117,
+      "step": 514
+    },
+    {
+      "epoch": 0.31741140215716485,
+      "grad_norm": 0.711551308631897,
+      "learning_rate": 1.726250750384038e-05,
+      "loss": 1.6072,
+      "step": 515
+    },
+    {
+      "epoch": 0.3180277349768875,
+      "grad_norm": 0.7259166836738586,
+      "learning_rate": 1.724768859075158e-05,
+      "loss": 1.5218,
+      "step": 516
+    },
+    {
+      "epoch": 0.31864406779661014,
+      "grad_norm": 0.6743744611740112,
+      "learning_rate": 1.7232836073852922e-05,
+      "loss": 1.5307,
+      "step": 517
+    },
+    {
+      "epoch": 0.3192604006163328,
+      "grad_norm": 0.6786503791809082,
+      "learning_rate": 1.7217950022007898e-05,
+      "loss": 1.6026,
+      "step": 518
+    },
+    {
+      "epoch": 0.3198767334360555,
+      "grad_norm": 0.6819741725921631,
+      "learning_rate": 1.720303050423548e-05,
+      "loss": 1.6202,
+      "step": 519
+    },
+    {
+      "epoch": 0.3204930662557781,
+      "grad_norm": 0.6740456223487854,
+      "learning_rate": 1.718807758970982e-05,
+      "loss": 1.5397,
+      "step": 520
+    },
+    {
+      "epoch": 0.3211093990755008,
+      "grad_norm": 0.6742592453956604,
+      "learning_rate": 1.71730913477599e-05,
+      "loss": 1.594,
+      "step": 521
+    },
+    {
+      "epoch": 0.3217257318952234,
+      "grad_norm": 0.6774325966835022,
+      "learning_rate": 1.715807184786922e-05,
+      "loss": 1.57,
+      "step": 522
+    },
+    {
+      "epoch": 0.3223420647149461,
+      "grad_norm": 0.6940937638282776,
+      "learning_rate": 1.71430191596755e-05,
+      "loss": 1.5233,
+      "step": 523
+    },
+    {
+      "epoch": 0.3229583975346687,
+      "grad_norm": 0.662968099117279,
+      "learning_rate": 1.7127933352970318e-05,
+      "loss": 1.566,
+      "step": 524
+    },
+    {
+      "epoch": 0.3235747303543914,
+      "grad_norm": 0.6893570423126221,
+      "learning_rate": 1.7112814497698808e-05,
+      "loss": 1.481,
+      "step": 525
+    },
+    {
+      "epoch": 0.324191063174114,
+      "grad_norm": 0.6710492372512817,
+      "learning_rate": 1.7097662663959346e-05,
+      "loss": 1.5263,
+      "step": 526
+    },
+    {
+      "epoch": 0.3248073959938367,
+      "grad_norm": 0.6967582106590271,
+      "learning_rate": 1.7082477922003198e-05,
+      "loss": 1.5752,
+      "step": 527
+    },
+    {
+      "epoch": 0.3254237288135593,
+      "grad_norm": 0.6757937073707581,
+      "learning_rate": 1.7067260342234215e-05,
+      "loss": 1.557,
+      "step": 528
+    },
+    {
+      "epoch": 0.326040061633282,
+      "grad_norm": 0.698906660079956,
+      "learning_rate": 1.7052009995208503e-05,
+      "loss": 1.5779,
+      "step": 529
+    },
+    {
+      "epoch": 0.3266563944530046,
+      "grad_norm": 0.6772062182426453,
+      "learning_rate": 1.7036726951634083e-05,
+      "loss": 1.5544,
+      "step": 530
+    },
+    {
+      "epoch": 0.32727272727272727,
+      "grad_norm": 0.6687437891960144,
+      "learning_rate": 1.7021411282370586e-05,
+      "loss": 1.4842,
+      "step": 531
+    },
+    {
+      "epoch": 0.32788906009244995,
+      "grad_norm": 0.6723273396492004,
+      "learning_rate": 1.70060630584289e-05,
+      "loss": 1.4988,
+      "step": 532
+    },
+    {
+      "epoch": 0.32850539291217257,
+      "grad_norm": 0.7292004227638245,
+      "learning_rate": 1.6990682350970857e-05,
+      "loss": 1.5674,
+      "step": 533
+    },
+    {
+      "epoch": 0.32912172573189524,
+      "grad_norm": 0.6870367527008057,
+      "learning_rate": 1.69752692313089e-05,
+      "loss": 1.5785,
+      "step": 534
+    },
+    {
+      "epoch": 0.32973805855161786,
+      "grad_norm": 0.6630904078483582,
+      "learning_rate": 1.695982377090575e-05,
+      "loss": 1.475,
+      "step": 535
+    },
+    {
+      "epoch": 0.33035439137134054,
+      "grad_norm": 0.6775215268135071,
+      "learning_rate": 1.6944346041374075e-05,
+      "loss": 1.5362,
+      "step": 536
+    },
+    {
+      "epoch": 0.33097072419106316,
+      "grad_norm": 0.6786991953849792,
+      "learning_rate": 1.692883611447616e-05,
+      "loss": 1.5405,
+      "step": 537
+    },
+    {
+      "epoch": 0.33158705701078584,
+      "grad_norm": 0.6665804386138916,
+      "learning_rate": 1.691329406212356e-05,
+      "loss": 1.5367,
+      "step": 538
+    },
+    {
+      "epoch": 0.33220338983050846,
+      "grad_norm": 0.6972366571426392,
+      "learning_rate": 1.6897719956376804e-05,
+      "loss": 1.5357,
+      "step": 539
+    },
+    {
+      "epoch": 0.33281972265023113,
+      "grad_norm": 0.6892836689949036,
+      "learning_rate": 1.6882113869445013e-05,
+      "loss": 1.5211,
+      "step": 540
+    },
+    {
+      "epoch": 0.33343605546995375,
+      "grad_norm": 0.6752969026565552,
+      "learning_rate": 1.68664758736856e-05,
+      "loss": 1.5136,
+      "step": 541
+    },
+    {
+      "epoch": 0.3340523882896764,
+      "grad_norm": 0.6893889904022217,
+      "learning_rate": 1.685080604160392e-05,
+      "loss": 1.5435,
+      "step": 542
+    },
+    {
+      "epoch": 0.3346687211093991,
+      "grad_norm": 0.6779683828353882,
+      "learning_rate": 1.683510444585294e-05,
+      "loss": 1.5784,
+      "step": 543
+    },
+    {
+      "epoch": 0.3352850539291217,
+      "grad_norm": 0.7006482481956482,
+      "learning_rate": 1.6819371159232895e-05,
+      "loss": 1.5591,
+      "step": 544
+    },
+    {
+      "epoch": 0.3359013867488444,
+      "grad_norm": 0.6864523887634277,
+      "learning_rate": 1.680360625469095e-05,
+      "loss": 1.5302,
+      "step": 545
+    },
+    {
+      "epoch": 0.336517719568567,
+      "grad_norm": 0.6806911826133728,
+      "learning_rate": 1.6787809805320874e-05,
+      "loss": 1.5388,
+      "step": 546
+    },
+    {
+      "epoch": 0.3371340523882897,
+      "grad_norm": 0.6877244710922241,
+      "learning_rate": 1.677198188436269e-05,
+      "loss": 1.5988,
+      "step": 547
+    },
+    {
+      "epoch": 0.3377503852080123,
+      "grad_norm": 0.6966729164123535,
+      "learning_rate": 1.6756122565202342e-05,
+      "loss": 1.5551,
+      "step": 548
+    },
+    {
+      "epoch": 0.338366718027735,
+      "grad_norm": 0.6754442453384399,
+      "learning_rate": 1.6740231921371345e-05,
+      "loss": 1.6043,
+      "step": 549
+    },
+    {
+      "epoch": 0.3389830508474576,
+      "grad_norm": 0.6731035709381104,
+      "learning_rate": 1.6724310026546456e-05,
+      "loss": 1.5358,
+      "step": 550
+    },
+    {
+      "epoch": 0.3395993836671803,
+      "grad_norm": 0.6805406212806702,
+      "learning_rate": 1.670835695454932e-05,
+      "loss": 1.568,
+      "step": 551
+    },
+    {
+      "epoch": 0.3402157164869029,
+      "grad_norm": 0.6713092923164368,
+      "learning_rate": 1.669237277934614e-05,
+      "loss": 1.5071,
+      "step": 552
+    },
+    {
+      "epoch": 0.3408320493066256,
+      "grad_norm": 0.656065046787262,
+      "learning_rate": 1.6676357575047335e-05,
+      "loss": 1.5152,
+      "step": 553
+    },
+    {
+      "epoch": 0.3414483821263482,
+      "grad_norm": 0.6780589818954468,
+      "learning_rate": 1.666031141590717e-05,
+      "loss": 1.5066,
+      "step": 554
+    },
+    {
+      "epoch": 0.3420647149460709,
+      "grad_norm": 0.7044869661331177,
+      "learning_rate": 1.6644234376323456e-05,
+      "loss": 1.574,
+      "step": 555
+    },
+    {
+      "epoch": 0.34268104776579356,
+      "grad_norm": 0.6871955990791321,
+      "learning_rate": 1.6628126530837163e-05,
+      "loss": 1.5659,
+      "step": 556
+    },
+    {
+      "epoch": 0.3432973805855162,
+      "grad_norm": 0.6856012940406799,
+      "learning_rate": 1.66119879541321e-05,
+      "loss": 1.5672,
+      "step": 557
+    },
+    {
+      "epoch": 0.34391371340523885,
+      "grad_norm": 0.6773875951766968,
+      "learning_rate": 1.6595818721034557e-05,
+      "loss": 1.5343,
+      "step": 558
+    },
+    {
+      "epoch": 0.3445300462249615,
+      "grad_norm": 0.6893959641456604,
+      "learning_rate": 1.657961890651296e-05,
+      "loss": 1.5305,
+      "step": 559
+    },
+    {
+      "epoch": 0.34514637904468415,
+      "grad_norm": 0.6633113026618958,
+      "learning_rate": 1.6563388585677542e-05,
+      "loss": 1.513,
+      "step": 560
+    },
+    {
+      "epoch": 0.34576271186440677,
+      "grad_norm": 0.6779730319976807,
+      "learning_rate": 1.6547127833779955e-05,
+      "loss": 1.5459,
+      "step": 561
+    },
+    {
+      "epoch": 0.34637904468412944,
+      "grad_norm": 0.6723275184631348,
+      "learning_rate": 1.6530836726212957e-05,
+      "loss": 1.5661,
+      "step": 562
+    },
+    {
+      "epoch": 0.34699537750385206,
+      "grad_norm": 0.6923189163208008,
+      "learning_rate": 1.651451533851005e-05,
+      "loss": 1.564,
+      "step": 563
+    },
+    {
+      "epoch": 0.34761171032357474,
+      "grad_norm": 0.6996536254882812,
+      "learning_rate": 1.6498163746345124e-05,
+      "loss": 1.5824,
+      "step": 564
+    },
+    {
+      "epoch": 0.34822804314329736,
+      "grad_norm": 0.690564751625061,
+      "learning_rate": 1.6481782025532114e-05,
+      "loss": 1.5489,
+      "step": 565
+    },
+    {
+      "epoch": 0.34884437596302004,
+      "grad_norm": 0.6660712361335754,
+      "learning_rate": 1.6465370252024646e-05,
+      "loss": 1.5219,
+      "step": 566
+    },
+    {
+      "epoch": 0.34946070878274266,
+      "grad_norm": 0.6772855520248413,
+      "learning_rate": 1.644892850191569e-05,
+      "loss": 1.5178,
+      "step": 567
+    },
+    {
+      "epoch": 0.35007704160246533,
+      "grad_norm": 0.6799270510673523,
+      "learning_rate": 1.6432456851437192e-05,
+      "loss": 1.5482,
+      "step": 568
+    },
+    {
+      "epoch": 0.350693374422188,
+      "grad_norm": 0.694419801235199,
+      "learning_rate": 1.6415955376959738e-05,
+      "loss": 1.5399,
+      "step": 569
+    },
+    {
+      "epoch": 0.35130970724191063,
+      "grad_norm": 0.6990444660186768,
+      "learning_rate": 1.6399424154992193e-05,
+      "loss": 1.5387,
+      "step": 570
+    },
+    {
+      "epoch": 0.3519260400616333,
+      "grad_norm": 0.684523344039917,
+      "learning_rate": 1.638286326218134e-05,
+      "loss": 1.5638,
+      "step": 571
+    },
+    {
+      "epoch": 0.3525423728813559,
+      "grad_norm": 0.7013692855834961,
+      "learning_rate": 1.6366272775311533e-05,
+      "loss": 1.5641,
+      "step": 572
+    },
+    {
+      "epoch": 0.3531587057010786,
+      "grad_norm": 0.6740252375602722,
+      "learning_rate": 1.6349652771304345e-05,
+      "loss": 1.6044,
+      "step": 573
+    },
+    {
+      "epoch": 0.3537750385208012,
+      "grad_norm": 0.681363582611084,
+      "learning_rate": 1.633300332721819e-05,
+      "loss": 1.5316,
+      "step": 574
+    },
+    {
+      "epoch": 0.3543913713405239,
+      "grad_norm": 0.6800167560577393,
+      "learning_rate": 1.6316324520248003e-05,
+      "loss": 1.6052,
+      "step": 575
+    },
+    {
+      "epoch": 0.3550077041602465,
+      "grad_norm": 0.6691132187843323,
+      "learning_rate": 1.629961642772483e-05,
+      "loss": 1.538,
+      "step": 576
+    },
+    {
+      "epoch": 0.3556240369799692,
+      "grad_norm": 0.7016999125480652,
+      "learning_rate": 1.628287912711553e-05,
+      "loss": 1.5543,
+      "step": 577
+    },
+    {
+      "epoch": 0.3562403697996918,
+      "grad_norm": 0.6937313079833984,
+      "learning_rate": 1.626611269602236e-05,
+      "loss": 1.5269,
+      "step": 578
+    },
+    {
+      "epoch": 0.3568567026194145,
+      "grad_norm": 0.7040328979492188,
+      "learning_rate": 1.624931721218266e-05,
+      "loss": 1.5458,
+      "step": 579
+    },
+    {
+      "epoch": 0.3574730354391371,
+      "grad_norm": 0.6810826659202576,
+      "learning_rate": 1.6232492753468457e-05,
+      "loss": 1.4988,
+      "step": 580
+    },
+    {
+      "epoch": 0.3580893682588598,
+      "grad_norm": 0.6729612350463867,
+      "learning_rate": 1.6215639397886125e-05,
+      "loss": 1.4894,
+      "step": 581
+    },
+    {
+      "epoch": 0.35870570107858246,
+      "grad_norm": 0.6691069602966309,
+      "learning_rate": 1.6198757223576017e-05,
+      "loss": 1.5331,
+      "step": 582
+    },
+    {
+      "epoch": 0.3593220338983051,
+      "grad_norm": 0.7009195685386658,
+      "learning_rate": 1.6181846308812113e-05,
+      "loss": 1.5848,
+      "step": 583
+    },
+    {
+      "epoch": 0.35993836671802776,
+      "grad_norm": 0.6886835694313049,
+      "learning_rate": 1.6164906732001636e-05,
+      "loss": 1.5524,
+      "step": 584
+    },
+    {
+      "epoch": 0.3605546995377504,
+      "grad_norm": 0.7847486734390259,
+      "learning_rate": 1.6147938571684697e-05,
+      "loss": 1.5444,
+      "step": 585
+    },
+    {
+      "epoch": 0.36117103235747305,
+      "grad_norm": 0.6955819129943848,
+      "learning_rate": 1.6130941906533953e-05,
+      "loss": 1.6079,
+      "step": 586
+    },
+    {
+      "epoch": 0.3617873651771957,
+      "grad_norm": 0.6843573451042175,
+      "learning_rate": 1.6113916815354205e-05,
+      "loss": 1.5002,
+      "step": 587
+    },
+    {
+      "epoch": 0.36240369799691835,
+      "grad_norm": 0.7060216665267944,
+      "learning_rate": 1.6096863377082057e-05,
+      "loss": 1.5127,
+      "step": 588
+    },
+    {
+      "epoch": 0.36302003081664097,
+      "grad_norm": 0.6773853898048401,
+      "learning_rate": 1.607978167078555e-05,
+      "loss": 1.5876,
+      "step": 589
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 0.6581423282623291,
+      "learning_rate": 1.6062671775663772e-05,
+      "loss": 1.5072,
+      "step": 590
+    },
+    {
+      "epoch": 0.36425269645608627,
+      "grad_norm": 0.6983804702758789,
+      "learning_rate": 1.604553377104653e-05,
+      "loss": 1.5434,
+      "step": 591
+    },
+    {
+      "epoch": 0.36486902927580894,
+      "grad_norm": 0.7162503600120544,
+      "learning_rate": 1.6028367736393948e-05,
+      "loss": 1.5771,
+      "step": 592
+    },
+    {
+      "epoch": 0.36548536209553156,
+      "grad_norm": 0.6911599040031433,
+      "learning_rate": 1.601117375129611e-05,
+      "loss": 1.5413,
+      "step": 593
+    },
+    {
+      "epoch": 0.36610169491525424,
+      "grad_norm": 0.6881313920021057,
+      "learning_rate": 1.59939518954727e-05,
+      "loss": 1.5004,
+      "step": 594
+    },
+    {
+      "epoch": 0.3667180277349769,
+      "grad_norm": 0.6664774417877197,
+      "learning_rate": 1.5976702248772612e-05,
+      "loss": 1.4708,
+      "step": 595
+    },
+    {
+      "epoch": 0.36733436055469953,
+      "grad_norm": 0.6833895444869995,
+      "learning_rate": 1.5959424891173603e-05,
+      "loss": 1.4789,
+      "step": 596
+    },
+    {
+      "epoch": 0.3679506933744222,
+      "grad_norm": 0.6863298416137695,
+      "learning_rate": 1.594211990278191e-05,
+      "loss": 1.5537,
+      "step": 597
+    },
+    {
+      "epoch": 0.36856702619414483,
+      "grad_norm": 0.6920463442802429,
+      "learning_rate": 1.5924787363831867e-05,
+      "loss": 1.585,
+      "step": 598
+    },
+    {
+      "epoch": 0.3691833590138675,
+      "grad_norm": 0.7192078828811646,
+      "learning_rate": 1.590742735468557e-05,
+      "loss": 1.5863,
+      "step": 599
+    },
+    {
+      "epoch": 0.3697996918335901,
+      "grad_norm": 0.6896239519119263,
+      "learning_rate": 1.589003995583245e-05,
+      "loss": 1.5641,
+      "step": 600
+    },
+    {
+      "epoch": 0.3704160246533128,
+      "grad_norm": 0.7030323147773743,
+      "learning_rate": 1.5872625247888965e-05,
+      "loss": 1.5987,
+      "step": 601
+    },
+    {
+      "epoch": 0.3710323574730354,
+      "grad_norm": 0.6835032105445862,
+      "learning_rate": 1.5855183311598156e-05,
+      "loss": 1.4811,
+      "step": 602
+    },
+    {
+      "epoch": 0.3716486902927581,
+      "grad_norm": 0.6897258758544922,
+      "learning_rate": 1.583771422782933e-05,
+      "loss": 1.5531,
+      "step": 603
+    },
+    {
+      "epoch": 0.3722650231124807,
+      "grad_norm": 0.6796337366104126,
+      "learning_rate": 1.5820218077577654e-05,
+      "loss": 1.4662,
+      "step": 604
+    },
+    {
+      "epoch": 0.3728813559322034,
+      "grad_norm": 0.6810954213142395,
+      "learning_rate": 1.5802694941963796e-05,
+      "loss": 1.508,
+      "step": 605
+    },
+    {
+      "epoch": 0.373497688751926,
+      "grad_norm": 0.6683149933815002,
+      "learning_rate": 1.5785144902233535e-05,
+      "loss": 1.5335,
+      "step": 606
+    },
+    {
+      "epoch": 0.3741140215716487,
+      "grad_norm": 0.6950417160987854,
+      "learning_rate": 1.5767568039757393e-05,
+      "loss": 1.54,
+      "step": 607
+    },
+    {
+      "epoch": 0.37473035439137137,
+      "grad_norm": 0.696682333946228,
+      "learning_rate": 1.5749964436030252e-05,
+      "loss": 1.5439,
+      "step": 608
+    },
+    {
+      "epoch": 0.375346687211094,
+      "grad_norm": 0.6980884075164795,
+      "learning_rate": 1.5732334172670985e-05,
+      "loss": 1.5937,
+      "step": 609
+    },
+    {
+      "epoch": 0.37596302003081666,
+      "grad_norm": 0.6861992478370667,
+      "learning_rate": 1.5714677331422065e-05,
+      "loss": 1.5507,
+      "step": 610
+    },
+    {
+      "epoch": 0.3765793528505393,
+      "grad_norm": 0.6781518459320068,
+      "learning_rate": 1.56969939941492e-05,
+      "loss": 1.523,
+      "step": 611
+    },
+    {
+      "epoch": 0.37719568567026196,
+      "grad_norm": 0.7281219959259033,
+      "learning_rate": 1.567928424284094e-05,
+      "loss": 1.5814,
+      "step": 612
+    },
+    {
+      "epoch": 0.3778120184899846,
+      "grad_norm": 0.6696469187736511,
+      "learning_rate": 1.566154815960831e-05,
+      "loss": 1.5113,
+      "step": 613
+    },
+    {
+      "epoch": 0.37842835130970726,
+      "grad_norm": 0.6881575584411621,
+      "learning_rate": 1.5643785826684424e-05,
+      "loss": 1.4566,
+      "step": 614
+    },
+    {
+      "epoch": 0.3790446841294299,
+      "grad_norm": 0.6957029104232788,
+      "learning_rate": 1.5625997326424086e-05,
+      "loss": 1.5479,
+      "step": 615
+    },
+    {
+      "epoch": 0.37966101694915255,
+      "grad_norm": 0.6924861073493958,
+      "learning_rate": 1.5608182741303447e-05,
+      "loss": 1.55,
+      "step": 616
+    },
+    {
+      "epoch": 0.3802773497688752,
+      "grad_norm": 0.7303187847137451,
+      "learning_rate": 1.5590342153919585e-05,
+      "loss": 1.5982,
+      "step": 617
+    },
+    {
+      "epoch": 0.38089368258859785,
+      "grad_norm": 0.6760260462760925,
+      "learning_rate": 1.557247564699014e-05,
+      "loss": 1.5352,
+      "step": 618
+    },
+    {
+      "epoch": 0.38151001540832047,
+      "grad_norm": 0.6855542063713074,
+      "learning_rate": 1.5554583303352933e-05,
+      "loss": 1.5724,
+      "step": 619
+    },
+    {
+      "epoch": 0.38212634822804314,
+      "grad_norm": 0.6895236968994141,
+      "learning_rate": 1.553666520596557e-05,
+      "loss": 1.543,
+      "step": 620
+    },
+    {
+      "epoch": 0.3827426810477658,
+      "grad_norm": 0.675171434879303,
+      "learning_rate": 1.551872143790506e-05,
+      "loss": 1.5342,
+      "step": 621
+    },
+    {
+      "epoch": 0.38335901386748844,
+      "grad_norm": 0.6768494248390198,
+      "learning_rate": 1.5500752082367453e-05,
+      "loss": 1.5366,
+      "step": 622
+    },
+    {
+      "epoch": 0.3839753466872111,
+      "grad_norm": 0.7357416152954102,
+      "learning_rate": 1.5482757222667414e-05,
+      "loss": 1.5893,
+      "step": 623
+    },
+    {
+      "epoch": 0.38459167950693374,
+      "grad_norm": 0.7197828888893127,
+      "learning_rate": 1.546473694223787e-05,
+      "loss": 1.5786,
+      "step": 624
+    },
+    {
+      "epoch": 0.3852080123266564,
+      "grad_norm": 0.6956158876419067,
+      "learning_rate": 1.5446691324629607e-05,
+      "loss": 1.5808,
+      "step": 625
+    },
+    {
+      "epoch": 0.38582434514637903,
+      "grad_norm": 0.6946635246276855,
+      "learning_rate": 1.5428620453510884e-05,
+      "loss": 1.5459,
+      "step": 626
+    },
+    {
+      "epoch": 0.3864406779661017,
+      "grad_norm": 0.6987695693969727,
+      "learning_rate": 1.541052441266705e-05,
+      "loss": 1.5779,
+      "step": 627
+    },
+    {
+      "epoch": 0.38705701078582433,
+      "grad_norm": 0.6879085302352905,
+      "learning_rate": 1.539240328600015e-05,
+      "loss": 1.5481,
+      "step": 628
+    },
+    {
+      "epoch": 0.387673343605547,
+      "grad_norm": 0.7136006355285645,
+      "learning_rate": 1.5374257157528543e-05,
+      "loss": 1.5269,
+      "step": 629
+    },
+    {
+      "epoch": 0.3882896764252696,
+      "grad_norm": 0.69001305103302,
+      "learning_rate": 1.5356086111386515e-05,
+      "loss": 1.5431,
+      "step": 630
+    },
+    {
+      "epoch": 0.3889060092449923,
+      "grad_norm": 0.6977394223213196,
+      "learning_rate": 1.5337890231823868e-05,
+      "loss": 1.5606,
+      "step": 631
+    },
+    {
+      "epoch": 0.3895223420647149,
+      "grad_norm": 0.7067729234695435,
+      "learning_rate": 1.5319669603205552e-05,
+      "loss": 1.5793,
+      "step": 632
+    },
+    {
+      "epoch": 0.3901386748844376,
+      "grad_norm": 0.6843896508216858,
+      "learning_rate": 1.5301424310011265e-05,
+      "loss": 1.5327,
+      "step": 633
+    },
+    {
+      "epoch": 0.3907550077041603,
+      "grad_norm": 0.6983306407928467,
+      "learning_rate": 1.5283154436835064e-05,
+      "loss": 1.5367,
+      "step": 634
+    },
+    {
+      "epoch": 0.3913713405238829,
+      "grad_norm": 0.7010204792022705,
+      "learning_rate": 1.5264860068384962e-05,
+      "loss": 1.6084,
+      "step": 635
+    },
+    {
+      "epoch": 0.39198767334360557,
+      "grad_norm": 0.6898306608200073,
+      "learning_rate": 1.5246541289482557e-05,
+      "loss": 1.5101,
+      "step": 636
+    },
+    {
+      "epoch": 0.3926040061633282,
+      "grad_norm": 0.7169236540794373,
+      "learning_rate": 1.5228198185062617e-05,
+      "loss": 1.5689,
+      "step": 637
+    },
+    {
+      "epoch": 0.39322033898305087,
+      "grad_norm": 0.6897739768028259,
+      "learning_rate": 1.5209830840172695e-05,
+      "loss": 1.5404,
+      "step": 638
+    },
+    {
+      "epoch": 0.3938366718027735,
+      "grad_norm": 0.6967317461967468,
+      "learning_rate": 1.5191439339972735e-05,
+      "loss": 1.5491,
+      "step": 639
+    },
+    {
+      "epoch": 0.39445300462249616,
+      "grad_norm": 0.7126529216766357,
+      "learning_rate": 1.517302376973468e-05,
+      "loss": 1.4902,
+      "step": 640
+    },
+    {
+      "epoch": 0.3950693374422188,
+      "grad_norm": 0.6778983473777771,
+      "learning_rate": 1.5154584214842074e-05,
+      "loss": 1.5109,
+      "step": 641
+    },
+    {
+      "epoch": 0.39568567026194146,
+      "grad_norm": 0.6908509135246277,
+      "learning_rate": 1.513612076078966e-05,
+      "loss": 1.559,
+      "step": 642
+    },
+    {
+      "epoch": 0.3963020030816641,
+      "grad_norm": 0.7037551403045654,
+      "learning_rate": 1.5117633493183e-05,
+      "loss": 1.521,
+      "step": 643
+    },
+    {
+      "epoch": 0.39691833590138675,
+      "grad_norm": 0.6984729170799255,
+      "learning_rate": 1.5099122497738043e-05,
+      "loss": 1.5444,
+      "step": 644
+    },
+    {
+      "epoch": 0.3975346687211094,
+      "grad_norm": 0.6904637813568115,
+      "learning_rate": 1.5080587860280787e-05,
+      "loss": 1.5623,
+      "step": 645
+    },
+    {
+      "epoch": 0.39815100154083205,
+      "grad_norm": 0.6932609677314758,
+      "learning_rate": 1.5062029666746814e-05,
+      "loss": 1.5004,
+      "step": 646
+    },
+    {
+      "epoch": 0.3987673343605547,
+      "grad_norm": 0.6654030084609985,
+      "learning_rate": 1.5043448003180943e-05,
+      "loss": 1.4253,
+      "step": 647
+    },
+    {
+      "epoch": 0.39938366718027735,
+      "grad_norm": 0.7254179120063782,
+      "learning_rate": 1.5024842955736804e-05,
+      "loss": 1.6116,
+      "step": 648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7039093375205994,
+      "learning_rate": 1.5006214610676449e-05,
+      "loss": 1.5324,
+      "step": 649
+    },
+    {
+      "epoch": 0.40061633281972264,
+      "grad_norm": 0.7008185386657715,
+      "learning_rate": 1.4987563054369938e-05,
+      "loss": 1.5599,
+      "step": 650
+    },
+    {
+      "epoch": 0.4012326656394453,
+      "grad_norm": 0.7072207927703857,
+      "learning_rate": 1.4968888373294972e-05,
+      "loss": 1.5426,
+      "step": 651
+    },
+    {
+      "epoch": 0.40184899845916794,
+      "grad_norm": 0.673958420753479,
+      "learning_rate": 1.4950190654036447e-05,
+      "loss": 1.5251,
+      "step": 652
+    },
+    {
+      "epoch": 0.4024653312788906,
+      "grad_norm": 0.6863639950752258,
+      "learning_rate": 1.493146998328609e-05,
+      "loss": 1.5191,
+      "step": 653
+    },
+    {
+      "epoch": 0.40308166409861323,
+      "grad_norm": 0.6735042333602905,
+      "learning_rate": 1.4912726447842037e-05,
+      "loss": 1.5363,
+      "step": 654
+    },
+    {
+      "epoch": 0.4036979969183359,
+      "grad_norm": 0.6717848777770996,
+      "learning_rate": 1.489396013460843e-05,
+      "loss": 1.4986,
+      "step": 655
+    },
+    {
+      "epoch": 0.40431432973805853,
+      "grad_norm": 0.6853130459785461,
+      "learning_rate": 1.4875171130595032e-05,
+      "loss": 1.5753,
+      "step": 656
+    },
+    {
+      "epoch": 0.4049306625577812,
+      "grad_norm": 0.6872352957725525,
+      "learning_rate": 1.4856359522916801e-05,
+      "loss": 1.5148,
+      "step": 657
+    },
+    {
+      "epoch": 0.4055469953775038,
+      "grad_norm": 0.7115101218223572,
+      "learning_rate": 1.4837525398793505e-05,
+      "loss": 1.5684,
+      "step": 658
+    },
+    {
+      "epoch": 0.4061633281972265,
+      "grad_norm": 0.6864094734191895,
+      "learning_rate": 1.4818668845549305e-05,
+      "loss": 1.5005,
+      "step": 659
+    },
+    {
+      "epoch": 0.4067796610169492,
+      "grad_norm": 0.6788669228553772,
+      "learning_rate": 1.479978995061235e-05,
+      "loss": 1.4922,
+      "step": 660
+    },
+    {
+      "epoch": 0.4073959938366718,
+      "grad_norm": 0.6768859028816223,
+      "learning_rate": 1.478088880151438e-05,
+      "loss": 1.4836,
+      "step": 661
+    },
+    {
+      "epoch": 0.4080123266563945,
+      "grad_norm": 0.6946293711662292,
+      "learning_rate": 1.4761965485890324e-05,
+      "loss": 1.544,
+      "step": 662
+    },
+    {
+      "epoch": 0.4086286594761171,
+      "grad_norm": 0.6807865500450134,
+      "learning_rate": 1.4743020091477868e-05,
+      "loss": 1.588,
+      "step": 663
+    },
+    {
+      "epoch": 0.40924499229583977,
+      "grad_norm": 0.6681613326072693,
+      "learning_rate": 1.4724052706117089e-05,
+      "loss": 1.4861,
+      "step": 664
+    },
+    {
+      "epoch": 0.4098613251155624,
+      "grad_norm": 0.6833305954933167,
+      "learning_rate": 1.470506341775e-05,
+      "loss": 1.487,
+      "step": 665
+    },
+    {
+      "epoch": 0.41047765793528507,
+      "grad_norm": 0.6754925847053528,
+      "learning_rate": 1.4686052314420185e-05,
+      "loss": 1.541,
+      "step": 666
+    },
+    {
+      "epoch": 0.4110939907550077,
+      "grad_norm": 0.6903547048568726,
+      "learning_rate": 1.4667019484272369e-05,
+      "loss": 1.6161,
+      "step": 667
+    },
+    {
+      "epoch": 0.41171032357473036,
+      "grad_norm": 0.702776312828064,
+      "learning_rate": 1.4647965015552003e-05,
+      "loss": 1.5886,
+      "step": 668
+    },
+    {
+      "epoch": 0.412326656394453,
+      "grad_norm": 0.688215434551239,
+      "learning_rate": 1.4628888996604878e-05,
+      "loss": 1.5869,
+      "step": 669
+    },
+    {
+      "epoch": 0.41294298921417566,
+      "grad_norm": 0.688977837562561,
+      "learning_rate": 1.4609791515876696e-05,
+      "loss": 1.5499,
+      "step": 670
+    },
+    {
+      "epoch": 0.4135593220338983,
+      "grad_norm": 0.7154136300086975,
+      "learning_rate": 1.4590672661912671e-05,
+      "loss": 1.5878,
+      "step": 671
+    },
+    {
+      "epoch": 0.41417565485362096,
+      "grad_norm": 0.7083712220191956,
+      "learning_rate": 1.4571532523357101e-05,
+      "loss": 1.525,
+      "step": 672
+    },
+    {
+      "epoch": 0.41479198767334363,
+      "grad_norm": 0.70921790599823,
+      "learning_rate": 1.455237118895299e-05,
+      "loss": 1.6164,
+      "step": 673
+    },
+    {
+      "epoch": 0.41540832049306625,
+      "grad_norm": 0.6897596120834351,
+      "learning_rate": 1.4533188747541596e-05,
+      "loss": 1.5369,
+      "step": 674
+    },
+    {
+      "epoch": 0.41602465331278893,
+      "grad_norm": 0.6994243264198303,
+      "learning_rate": 1.4513985288062054e-05,
+      "loss": 1.5822,
+      "step": 675
+    },
+    {
+      "epoch": 0.41664098613251155,
+      "grad_norm": 0.6927086710929871,
+      "learning_rate": 1.4494760899550944e-05,
+      "loss": 1.5056,
+      "step": 676
+    },
+    {
+      "epoch": 0.4172573189522342,
+      "grad_norm": 0.6754436492919922,
+      "learning_rate": 1.4475515671141878e-05,
+      "loss": 1.5423,
+      "step": 677
+    },
+    {
+      "epoch": 0.41787365177195684,
+      "grad_norm": 0.6721181869506836,
+      "learning_rate": 1.4456249692065098e-05,
+      "loss": 1.5148,
+      "step": 678
+    },
+    {
+      "epoch": 0.4184899845916795,
+      "grad_norm": 0.6847098469734192,
+      "learning_rate": 1.4436963051647058e-05,
+      "loss": 1.5291,
+      "step": 679
+    },
+    {
+      "epoch": 0.41910631741140214,
+      "grad_norm": 0.721822202205658,
+      "learning_rate": 1.4417655839309998e-05,
+      "loss": 1.5548,
+      "step": 680
+    },
+    {
+      "epoch": 0.4197226502311248,
+      "grad_norm": 0.6813617944717407,
+      "learning_rate": 1.4398328144571557e-05,
+      "loss": 1.5008,
+      "step": 681
+    },
+    {
+      "epoch": 0.42033898305084744,
+      "grad_norm": 0.7119934558868408,
+      "learning_rate": 1.4378980057044316e-05,
+      "loss": 1.6075,
+      "step": 682
+    },
+    {
+      "epoch": 0.4209553158705701,
+      "grad_norm": 0.6564061045646667,
+      "learning_rate": 1.4359611666435432e-05,
+      "loss": 1.4609,
+      "step": 683
+    },
+    {
+      "epoch": 0.42157164869029273,
+      "grad_norm": 0.6707856059074402,
+      "learning_rate": 1.434022306254618e-05,
+      "loss": 1.4859,
+      "step": 684
+    },
+    {
+      "epoch": 0.4221879815100154,
+      "grad_norm": 0.6858282089233398,
+      "learning_rate": 1.4320814335271558e-05,
+      "loss": 1.5438,
+      "step": 685
+    },
+    {
+      "epoch": 0.4228043143297381,
+      "grad_norm": 0.6912311911582947,
+      "learning_rate": 1.430138557459987e-05,
+      "loss": 1.5867,
+      "step": 686
+    },
+    {
+      "epoch": 0.4234206471494607,
+      "grad_norm": 0.6980946660041809,
+      "learning_rate": 1.4281936870612301e-05,
+      "loss": 1.5645,
+      "step": 687
+    },
+    {
+      "epoch": 0.4240369799691834,
+      "grad_norm": 0.683866560459137,
+      "learning_rate": 1.4262468313482498e-05,
+      "loss": 1.6131,
+      "step": 688
+    },
+    {
+      "epoch": 0.424653312788906,
+      "grad_norm": 0.6955146193504333,
+      "learning_rate": 1.4242979993476172e-05,
+      "loss": 1.5797,
+      "step": 689
+    },
+    {
+      "epoch": 0.4252696456086287,
+      "grad_norm": 0.6703779101371765,
+      "learning_rate": 1.4223472000950649e-05,
+      "loss": 1.4769,
+      "step": 690
+    },
+    {
+      "epoch": 0.4258859784283513,
+      "grad_norm": 0.6915444731712341,
+      "learning_rate": 1.4203944426354466e-05,
+      "loss": 1.4834,
+      "step": 691
+    },
+    {
+      "epoch": 0.426502311248074,
+      "grad_norm": 0.7006959319114685,
+      "learning_rate": 1.4184397360226967e-05,
+      "loss": 1.4711,
+      "step": 692
+    },
+    {
+      "epoch": 0.4271186440677966,
+      "grad_norm": 0.6921749711036682,
+      "learning_rate": 1.4164830893197855e-05,
+      "loss": 1.5435,
+      "step": 693
+    },
+    {
+      "epoch": 0.42773497688751927,
+      "grad_norm": 0.7022824287414551,
+      "learning_rate": 1.4145245115986787e-05,
+      "loss": 1.5453,
+      "step": 694
+    },
+    {
+      "epoch": 0.4283513097072419,
+      "grad_norm": 0.7122887372970581,
+      "learning_rate": 1.4125640119402957e-05,
+      "loss": 1.5264,
+      "step": 695
+    },
+    {
+      "epoch": 0.42896764252696457,
+      "grad_norm": 0.6789283752441406,
+      "learning_rate": 1.4106015994344667e-05,
+      "loss": 1.4822,
+      "step": 696
+    },
+    {
+      "epoch": 0.4295839753466872,
+      "grad_norm": 0.6716744303703308,
+      "learning_rate": 1.40863728317989e-05,
+      "loss": 1.4625,
+      "step": 697
+    },
+    {
+      "epoch": 0.43020030816640986,
+      "grad_norm": 0.6802600622177124,
+      "learning_rate": 1.4066710722840922e-05,
+      "loss": 1.5189,
+      "step": 698
+    },
+    {
+      "epoch": 0.43081664098613254,
+      "grad_norm": 0.7036100029945374,
+      "learning_rate": 1.4047029758633823e-05,
+      "loss": 1.5863,
+      "step": 699
+    },
+    {
+      "epoch": 0.43143297380585516,
+      "grad_norm": 0.6810036301612854,
+      "learning_rate": 1.402733003042814e-05,
+      "loss": 1.5178,
+      "step": 700
+    },
+    {
+      "epoch": 0.43204930662557783,
+      "grad_norm": 0.7000814080238342,
+      "learning_rate": 1.4007611629561385e-05,
+      "loss": 1.4893,
+      "step": 701
+    },
+    {
+      "epoch": 0.43266563944530045,
+      "grad_norm": 0.6997476816177368,
+      "learning_rate": 1.3987874647457657e-05,
+      "loss": 1.5167,
+      "step": 702
+    },
+    {
+      "epoch": 0.43328197226502313,
+      "grad_norm": 0.7024034857749939,
+      "learning_rate": 1.3968119175627214e-05,
+      "loss": 1.5463,
+      "step": 703
+    },
+    {
+      "epoch": 0.43389830508474575,
+      "grad_norm": 0.7160104513168335,
+      "learning_rate": 1.3948345305666027e-05,
+      "loss": 1.5334,
+      "step": 704
+    },
+    {
+      "epoch": 0.4345146379044684,
+      "grad_norm": 0.6765588521957397,
+      "learning_rate": 1.3928553129255375e-05,
+      "loss": 1.4944,
+      "step": 705
+    },
+    {
+      "epoch": 0.43513097072419105,
+      "grad_norm": 0.6819117665290833,
+      "learning_rate": 1.390874273816142e-05,
+      "loss": 1.5785,
+      "step": 706
+    },
+    {
+      "epoch": 0.4357473035439137,
+      "grad_norm": 0.677497386932373,
+      "learning_rate": 1.3888914224234768e-05,
+      "loss": 1.5074,
+      "step": 707
+    },
+    {
+      "epoch": 0.43636363636363634,
+      "grad_norm": 0.7176629304885864,
+      "learning_rate": 1.3869067679410057e-05,
+      "loss": 1.6059,
+      "step": 708
+    },
+    {
+      "epoch": 0.436979969183359,
+      "grad_norm": 0.6862969398498535,
+      "learning_rate": 1.384920319570552e-05,
+      "loss": 1.5151,
+      "step": 709
+    },
+    {
+      "epoch": 0.43759630200308164,
+      "grad_norm": 0.6906915307044983,
+      "learning_rate": 1.3829320865222571e-05,
+      "loss": 1.4896,
+      "step": 710
+    },
+    {
+      "epoch": 0.4382126348228043,
+      "grad_norm": 0.6779195070266724,
+      "learning_rate": 1.3809420780145361e-05,
+      "loss": 1.5014,
+      "step": 711
+    },
+    {
+      "epoch": 0.438828967642527,
+      "grad_norm": 0.6874434947967529,
+      "learning_rate": 1.3789503032740368e-05,
+      "loss": 1.4992,
+      "step": 712
+    },
+    {
+      "epoch": 0.4394453004622496,
+      "grad_norm": 0.6920948624610901,
+      "learning_rate": 1.3769567715355954e-05,
+      "loss": 1.5274,
+      "step": 713
+    },
+    {
+      "epoch": 0.4400616332819723,
+      "grad_norm": 0.6916899085044861,
+      "learning_rate": 1.374961492042195e-05,
+      "loss": 1.5477,
+      "step": 714
+    },
+    {
+      "epoch": 0.4406779661016949,
+      "grad_norm": 0.7086225748062134,
+      "learning_rate": 1.3729644740449221e-05,
+      "loss": 1.5571,
+      "step": 715
+    },
+    {
+      "epoch": 0.4412942989214176,
+      "grad_norm": 0.6834512948989868,
+      "learning_rate": 1.3709657268029233e-05,
+      "loss": 1.4966,
+      "step": 716
+    },
+    {
+      "epoch": 0.4419106317411402,
+      "grad_norm": 0.6928700804710388,
+      "learning_rate": 1.3689652595833632e-05,
+      "loss": 1.5397,
+      "step": 717
+    },
+    {
+      "epoch": 0.4425269645608629,
+      "grad_norm": 0.6936474442481995,
+      "learning_rate": 1.3669630816613806e-05,
+      "loss": 1.5164,
+      "step": 718
+    },
+    {
+      "epoch": 0.4431432973805855,
+      "grad_norm": 0.6894739270210266,
+      "learning_rate": 1.3649592023200468e-05,
+      "loss": 1.5299,
+      "step": 719
+    },
+    {
+      "epoch": 0.4437596302003082,
+      "grad_norm": 0.6911756992340088,
+      "learning_rate": 1.362953630850321e-05,
+      "loss": 1.4985,
+      "step": 720
+    },
+    {
+      "epoch": 0.4443759630200308,
+      "grad_norm": 0.7013258934020996,
+      "learning_rate": 1.360946376551008e-05,
+      "loss": 1.5251,
+      "step": 721
+    },
+    {
+      "epoch": 0.44499229583975347,
+      "grad_norm": 0.6924850344657898,
+      "learning_rate": 1.358937448728715e-05,
+      "loss": 1.5015,
+      "step": 722
+    },
+    {
+      "epoch": 0.4456086286594761,
+      "grad_norm": 0.7154591083526611,
+      "learning_rate": 1.3569268566978093e-05,
+      "loss": 1.5476,
+      "step": 723
+    },
+    {
+      "epoch": 0.44622496147919877,
+      "grad_norm": 0.6930962204933167,
+      "learning_rate": 1.3549146097803728e-05,
+      "loss": 1.4837,
+      "step": 724
+    },
+    {
+      "epoch": 0.44684129429892144,
+      "grad_norm": 0.7065971493721008,
+      "learning_rate": 1.3529007173061612e-05,
+      "loss": 1.5336,
+      "step": 725
+    },
+    {
+      "epoch": 0.44745762711864406,
+      "grad_norm": 0.6910285949707031,
+      "learning_rate": 1.35088518861256e-05,
+      "loss": 1.5811,
+      "step": 726
+    },
+    {
+      "epoch": 0.44807395993836674,
+      "grad_norm": 0.6844209432601929,
+      "learning_rate": 1.34886803304454e-05,
+      "loss": 1.5298,
+      "step": 727
+    },
+    {
+      "epoch": 0.44869029275808936,
+      "grad_norm": 0.6936566233634949,
+      "learning_rate": 1.3468492599546168e-05,
+      "loss": 1.5413,
+      "step": 728
+    },
+    {
+      "epoch": 0.44930662557781204,
+      "grad_norm": 0.7082076072692871,
+      "learning_rate": 1.3448288787028032e-05,
+      "loss": 1.5828,
+      "step": 729
+    },
+    {
+      "epoch": 0.44992295839753466,
+      "grad_norm": 0.6861379742622375,
+      "learning_rate": 1.3428068986565701e-05,
+      "loss": 1.5082,
+      "step": 730
+    },
+    {
+      "epoch": 0.45053929121725733,
+      "grad_norm": 0.6852473616600037,
+      "learning_rate": 1.3407833291908005e-05,
+      "loss": 1.5264,
+      "step": 731
+    },
+    {
+      "epoch": 0.45115562403697995,
+      "grad_norm": 0.7140191197395325,
+      "learning_rate": 1.3387581796877469e-05,
+      "loss": 1.6062,
+      "step": 732
+    },
+    {
+      "epoch": 0.45177195685670263,
+      "grad_norm": 0.6729596257209778,
+      "learning_rate": 1.3367314595369872e-05,
+      "loss": 1.4818,
+      "step": 733
+    },
+    {
+      "epoch": 0.45238828967642525,
+      "grad_norm": 0.6968595385551453,
+      "learning_rate": 1.334703178135382e-05,
+      "loss": 1.558,
+      "step": 734
+    },
+    {
+      "epoch": 0.4530046224961479,
+      "grad_norm": 0.6909863352775574,
+      "learning_rate": 1.3326733448870304e-05,
+      "loss": 1.5003,
+      "step": 735
+    },
+    {
+      "epoch": 0.45362095531587054,
+      "grad_norm": 0.6805793046951294,
+      "learning_rate": 1.3306419692032273e-05,
+      "loss": 1.4863,
+      "step": 736
+    },
+    {
+      "epoch": 0.4542372881355932,
+      "grad_norm": 0.6838030219078064,
+      "learning_rate": 1.328609060502418e-05,
+      "loss": 1.5008,
+      "step": 737
+    },
+    {
+      "epoch": 0.4548536209553159,
+      "grad_norm": 0.7309401631355286,
+      "learning_rate": 1.326574628210156e-05,
+      "loss": 1.5321,
+      "step": 738
+    },
+    {
+      "epoch": 0.4554699537750385,
+      "grad_norm": 0.6883420944213867,
+      "learning_rate": 1.3245386817590594e-05,
+      "loss": 1.4785,
+      "step": 739
+    },
+    {
+      "epoch": 0.4560862865947612,
+      "grad_norm": 0.689288854598999,
+      "learning_rate": 1.3225012305887659e-05,
+      "loss": 1.4771,
+      "step": 740
+    },
+    {
+      "epoch": 0.4567026194144838,
+      "grad_norm": 0.6939703822135925,
+      "learning_rate": 1.3204622841458907e-05,
+      "loss": 1.5563,
+      "step": 741
+    },
+    {
+      "epoch": 0.4573189522342065,
+      "grad_norm": 0.702176570892334,
+      "learning_rate": 1.3184218518839812e-05,
+      "loss": 1.4959,
+      "step": 742
+    },
+    {
+      "epoch": 0.4579352850539291,
+      "grad_norm": 0.6990866661071777,
+      "learning_rate": 1.3163799432634736e-05,
+      "loss": 1.5297,
+      "step": 743
+    },
+    {
+      "epoch": 0.4585516178736518,
+      "grad_norm": 0.7106867432594299,
+      "learning_rate": 1.31433656775165e-05,
+      "loss": 1.5301,
+      "step": 744
+    },
+    {
+      "epoch": 0.4591679506933744,
+      "grad_norm": 0.707547664642334,
+      "learning_rate": 1.3122917348225928e-05,
+      "loss": 1.5439,
+      "step": 745
+    },
+    {
+      "epoch": 0.4597842835130971,
+      "grad_norm": 0.6893981099128723,
+      "learning_rate": 1.3102454539571424e-05,
+      "loss": 1.4977,
+      "step": 746
+    },
+    {
+      "epoch": 0.4604006163328197,
+      "grad_norm": 0.7004650235176086,
+      "learning_rate": 1.3081977346428523e-05,
+      "loss": 1.5819,
+      "step": 747
+    },
+    {
+      "epoch": 0.4610169491525424,
+      "grad_norm": 0.7182575464248657,
+      "learning_rate": 1.306148586373945e-05,
+      "loss": 1.5291,
+      "step": 748
+    },
+    {
+      "epoch": 0.461633281972265,
+      "grad_norm": 0.676213800907135,
+      "learning_rate": 1.3040980186512689e-05,
+      "loss": 1.4535,
+      "step": 749
+    },
+    {
+      "epoch": 0.4622496147919877,
+      "grad_norm": 0.7185645699501038,
+      "learning_rate": 1.3020460409822535e-05,
+      "loss": 1.5487,
+      "step": 750
+    },
+    {
+      "epoch": 0.46286594761171035,
+      "grad_norm": 0.6934641003608704,
+      "learning_rate": 1.2999926628808658e-05,
+      "loss": 1.5674,
+      "step": 751
+    },
+    {
+      "epoch": 0.46348228043143297,
+      "grad_norm": 0.6709468364715576,
+      "learning_rate": 1.2979378938675647e-05,
+      "loss": 1.4928,
+      "step": 752
+    },
+    {
+      "epoch": 0.46409861325115565,
+      "grad_norm": 0.6758535504341125,
+      "learning_rate": 1.2958817434692592e-05,
+      "loss": 1.5171,
+      "step": 753
+    },
+    {
+      "epoch": 0.46471494607087827,
+      "grad_norm": 0.6950028538703918,
+      "learning_rate": 1.2938242212192632e-05,
+      "loss": 1.5726,
+      "step": 754
+    },
+    {
+      "epoch": 0.46533127889060094,
+      "grad_norm": 0.6801124215126038,
+      "learning_rate": 1.2917653366572498e-05,
+      "loss": 1.485,
+      "step": 755
+    },
+    {
+      "epoch": 0.46594761171032356,
+      "grad_norm": 0.6991348266601562,
+      "learning_rate": 1.28970509932921e-05,
+      "loss": 1.5593,
+      "step": 756
+    },
+    {
+      "epoch": 0.46656394453004624,
+      "grad_norm": 0.6854835748672485,
+      "learning_rate": 1.287643518787406e-05,
+      "loss": 1.5307,
+      "step": 757
+    },
+    {
+      "epoch": 0.46718027734976886,
+      "grad_norm": 0.6962106823921204,
+      "learning_rate": 1.2855806045903283e-05,
+      "loss": 1.495,
+      "step": 758
+    },
+    {
+      "epoch": 0.46779661016949153,
+      "grad_norm": 0.7006097435951233,
+      "learning_rate": 1.2835163663026506e-05,
+      "loss": 1.5279,
+      "step": 759
+    },
+    {
+      "epoch": 0.46841294298921415,
+      "grad_norm": 0.7017481923103333,
+      "learning_rate": 1.2814508134951853e-05,
+      "loss": 1.5506,
+      "step": 760
+    },
+    {
+      "epoch": 0.46902927580893683,
+      "grad_norm": 1.2100894451141357,
+      "learning_rate": 1.2793839557448403e-05,
+      "loss": 1.463,
+      "step": 761
+    },
+    {
+      "epoch": 0.46964560862865945,
+      "grad_norm": 0.6907877326011658,
+      "learning_rate": 1.2773158026345733e-05,
+      "loss": 1.518,
+      "step": 762
+    },
+    {
+      "epoch": 0.4702619414483821,
+      "grad_norm": 0.7050778865814209,
+      "learning_rate": 1.275246363753348e-05,
+      "loss": 1.5552,
+      "step": 763
+    },
+    {
+      "epoch": 0.4708782742681048,
+      "grad_norm": 0.6761541366577148,
+      "learning_rate": 1.2731756486960907e-05,
+      "loss": 1.4784,
+      "step": 764
+    },
+    {
+      "epoch": 0.4714946070878274,
+      "grad_norm": 0.6929753422737122,
+      "learning_rate": 1.2711036670636426e-05,
+      "loss": 1.4483,
+      "step": 765
+    },
+    {
+      "epoch": 0.4721109399075501,
+      "grad_norm": 0.7061207294464111,
+      "learning_rate": 1.2690304284627186e-05,
+      "loss": 1.4819,
+      "step": 766
+    },
+    {
+      "epoch": 0.4727272727272727,
+      "grad_norm": 0.6936407685279846,
+      "learning_rate": 1.2669559425058622e-05,
+      "loss": 1.4628,
+      "step": 767
+    },
+    {
+      "epoch": 0.4733436055469954,
+      "grad_norm": 0.7063566446304321,
+      "learning_rate": 1.2648802188113982e-05,
+      "loss": 1.4798,
+      "step": 768
+    },
+    {
+      "epoch": 0.473959938366718,
+      "grad_norm": 0.690711259841919,
+      "learning_rate": 1.2628032670033921e-05,
+      "loss": 1.4731,
+      "step": 769
+    },
+    {
+      "epoch": 0.4745762711864407,
+      "grad_norm": 0.7119061350822449,
+      "learning_rate": 1.260725096711603e-05,
+      "loss": 1.5501,
+      "step": 770
+    },
+    {
+      "epoch": 0.4751926040061633,
+      "grad_norm": 0.699124276638031,
+      "learning_rate": 1.2586457175714382e-05,
+      "loss": 1.5237,
+      "step": 771
+    },
+    {
+      "epoch": 0.475808936825886,
+      "grad_norm": 0.6884123086929321,
+      "learning_rate": 1.256565139223912e-05,
+      "loss": 1.4684,
+      "step": 772
+    },
+    {
+      "epoch": 0.4764252696456086,
+      "grad_norm": 0.7211017608642578,
+      "learning_rate": 1.2544833713155976e-05,
+      "loss": 1.5732,
+      "step": 773
+    },
+    {
+      "epoch": 0.4770416024653313,
+      "grad_norm": 0.6872128844261169,
+      "learning_rate": 1.2524004234985825e-05,
+      "loss": 1.5267,
+      "step": 774
+    },
+    {
+      "epoch": 0.4776579352850539,
+      "grad_norm": 0.690025269985199,
+      "learning_rate": 1.2503163054304275e-05,
+      "loss": 1.4584,
+      "step": 775
+    },
+    {
+      "epoch": 0.4782742681047766,
+      "grad_norm": 0.6983490586280823,
+      "learning_rate": 1.2482310267741168e-05,
+      "loss": 1.5026,
+      "step": 776
+    },
+    {
+      "epoch": 0.47889060092449925,
+      "grad_norm": 0.7001078128814697,
+      "learning_rate": 1.246144597198017e-05,
+      "loss": 1.5371,
+      "step": 777
+    },
+    {
+      "epoch": 0.4795069337442219,
+      "grad_norm": 0.6846566796302795,
+      "learning_rate": 1.2440570263758306e-05,
+      "loss": 1.4678,
+      "step": 778
+    },
+    {
+      "epoch": 0.48012326656394455,
+      "grad_norm": 0.705186128616333,
+      "learning_rate": 1.2419683239865515e-05,
+      "loss": 1.5206,
+      "step": 779
+    },
+    {
+      "epoch": 0.48073959938366717,
+      "grad_norm": 0.707804799079895,
+      "learning_rate": 1.2398784997144192e-05,
+      "loss": 1.4982,
+      "step": 780
+    },
+    {
+      "epoch": 0.48135593220338985,
+      "grad_norm": 0.7467078566551208,
+      "learning_rate": 1.2377875632488765e-05,
+      "loss": 1.6151,
+      "step": 781
+    },
+    {
+      "epoch": 0.48197226502311247,
+      "grad_norm": 0.7052176594734192,
+      "learning_rate": 1.2356955242845214e-05,
+      "loss": 1.4933,
+      "step": 782
+    },
+    {
+      "epoch": 0.48258859784283514,
+      "grad_norm": 0.7082404494285583,
+      "learning_rate": 1.2336023925210647e-05,
+      "loss": 1.4995,
+      "step": 783
+    },
+    {
+      "epoch": 0.48320493066255776,
+      "grad_norm": 0.7223908305168152,
+      "learning_rate": 1.2315081776632828e-05,
+      "loss": 1.548,
+      "step": 784
+    },
+    {
+      "epoch": 0.48382126348228044,
+      "grad_norm": 0.7085282206535339,
+      "learning_rate": 1.2294128894209751e-05,
+      "loss": 1.5197,
+      "step": 785
+    },
+    {
+      "epoch": 0.48443759630200306,
+      "grad_norm": 0.6913473010063171,
+      "learning_rate": 1.227316537508917e-05,
+      "loss": 1.5403,
+      "step": 786
+    },
+    {
+      "epoch": 0.48505392912172574,
+      "grad_norm": 0.6815372705459595,
+      "learning_rate": 1.2252191316468151e-05,
+      "loss": 1.4943,
+      "step": 787
+    },
+    {
+      "epoch": 0.48567026194144836,
+      "grad_norm": 0.6794235706329346,
+      "learning_rate": 1.223120681559264e-05,
+      "loss": 1.4466,
+      "step": 788
+    },
+    {
+      "epoch": 0.48628659476117103,
+      "grad_norm": 0.742695689201355,
+      "learning_rate": 1.2210211969756987e-05,
+      "loss": 1.5392,
+      "step": 789
+    },
+    {
+      "epoch": 0.4869029275808937,
+      "grad_norm": 0.7016921043395996,
+      "learning_rate": 1.2189206876303512e-05,
+      "loss": 1.5022,
+      "step": 790
+    },
+    {
+      "epoch": 0.4875192604006163,
+      "grad_norm": 0.6898470520973206,
+      "learning_rate": 1.2168191632622044e-05,
+      "loss": 1.5099,
+      "step": 791
+    },
+    {
+      "epoch": 0.488135593220339,
+      "grad_norm": 0.6877438426017761,
+      "learning_rate": 1.2147166336149474e-05,
+      "loss": 1.4375,
+      "step": 792
+    },
+    {
+      "epoch": 0.4887519260400616,
+      "grad_norm": 0.6840031743049622,
+      "learning_rate": 1.2126131084369308e-05,
+      "loss": 1.4809,
+      "step": 793
+    },
+    {
+      "epoch": 0.4893682588597843,
+      "grad_norm": 0.6915991306304932,
+      "learning_rate": 1.2105085974811204e-05,
+      "loss": 1.5226,
+      "step": 794
+    },
+    {
+      "epoch": 0.4899845916795069,
+      "grad_norm": 0.6939176321029663,
+      "learning_rate": 1.2084031105050525e-05,
+      "loss": 1.5189,
+      "step": 795
+    },
+    {
+      "epoch": 0.4906009244992296,
+      "grad_norm": 0.6981072425842285,
+      "learning_rate": 1.206296657270789e-05,
+      "loss": 1.5454,
+      "step": 796
+    },
+    {
+      "epoch": 0.4912172573189522,
+      "grad_norm": 0.693756639957428,
+      "learning_rate": 1.204189247544872e-05,
+      "loss": 1.4886,
+      "step": 797
+    },
+    {
+      "epoch": 0.4918335901386749,
+      "grad_norm": 0.7083744406700134,
+      "learning_rate": 1.202080891098278e-05,
+      "loss": 1.4983,
+      "step": 798
+    },
+    {
+      "epoch": 0.4924499229583975,
+      "grad_norm": 0.7123251557350159,
+      "learning_rate": 1.1999715977063729e-05,
+      "loss": 1.5808,
+      "step": 799
+    },
+    {
+      "epoch": 0.4930662557781202,
+      "grad_norm": 0.7044851779937744,
+      "learning_rate": 1.1978613771488667e-05,
+      "loss": 1.5756,
+      "step": 800
+    },
+    {
+      "epoch": 0.4936825885978428,
+      "grad_norm": 0.7301934957504272,
+      "learning_rate": 1.1957502392097693e-05,
+      "loss": 1.5848,
+      "step": 801
+    },
+    {
+      "epoch": 0.4942989214175655,
+      "grad_norm": 0.7132928967475891,
+      "learning_rate": 1.1936381936773418e-05,
+      "loss": 1.5164,
+      "step": 802
+    },
+    {
+      "epoch": 0.49491525423728816,
+      "grad_norm": 0.6884498596191406,
+      "learning_rate": 1.1915252503440562e-05,
+      "loss": 1.509,
+      "step": 803
+    },
+    {
+      "epoch": 0.4955315870570108,
+      "grad_norm": 0.6921610832214355,
+      "learning_rate": 1.1894114190065442e-05,
+      "loss": 1.4744,
+      "step": 804
+    },
+    {
+      "epoch": 0.49614791987673346,
+      "grad_norm": 0.7089739441871643,
+      "learning_rate": 1.187296709465557e-05,
+      "loss": 1.5635,
+      "step": 805
+    },
+    {
+      "epoch": 0.4967642526964561,
+      "grad_norm": 0.6999155879020691,
+      "learning_rate": 1.1851811315259165e-05,
+      "loss": 1.5135,
+      "step": 806
+    },
+    {
+      "epoch": 0.49738058551617875,
+      "grad_norm": 0.7082995176315308,
+      "learning_rate": 1.1830646949964709e-05,
+      "loss": 1.5179,
+      "step": 807
+    },
+    {
+      "epoch": 0.4979969183359014,
+      "grad_norm": 0.7552445530891418,
+      "learning_rate": 1.1809474096900498e-05,
+      "loss": 1.5213,
+      "step": 808
+    },
+    {
+      "epoch": 0.49861325115562405,
+      "grad_norm": 0.6751538515090942,
+      "learning_rate": 1.1788292854234177e-05,
+      "loss": 1.4103,
+      "step": 809
+    },
+    {
+      "epoch": 0.49922958397534667,
+      "grad_norm": 0.7049698233604431,
+      "learning_rate": 1.1767103320172285e-05,
+      "loss": 1.6103,
+      "step": 810
+    },
+    {
+      "epoch": 0.49984591679506934,
+      "grad_norm": 0.7300886511802673,
+      "learning_rate": 1.1745905592959822e-05,
+      "loss": 1.5083,
+      "step": 811
+    },
+    {
+      "epoch": 0.500462249614792,
+      "grad_norm": 0.6986029148101807,
+      "learning_rate": 1.1724699770879749e-05,
+      "loss": 1.4523,
+      "step": 812
+    },
+    {
+      "epoch": 0.5010785824345146,
+      "grad_norm": 0.7011654376983643,
+      "learning_rate": 1.1703485952252577e-05,
+      "loss": 1.5229,
+      "step": 813
+    },
+    {
+      "epoch": 0.5016949152542373,
+      "grad_norm": 0.7019833922386169,
+      "learning_rate": 1.1682264235435896e-05,
+      "loss": 1.4856,
+      "step": 814
+    },
+    {
+      "epoch": 0.50231124807396,
+      "grad_norm": 0.6956936717033386,
+      "learning_rate": 1.1661034718823898e-05,
+      "loss": 1.4336,
+      "step": 815
+    },
+    {
+      "epoch": 0.5029275808936826,
+      "grad_norm": 0.6937043070793152,
+      "learning_rate": 1.1639797500846951e-05,
+      "loss": 1.4875,
+      "step": 816
+    },
+    {
+      "epoch": 0.5035439137134052,
+      "grad_norm": 0.7288076281547546,
+      "learning_rate": 1.1618552679971136e-05,
+      "loss": 1.5739,
+      "step": 817
+    },
+    {
+      "epoch": 0.5041602465331279,
+      "grad_norm": 0.7043364644050598,
+      "learning_rate": 1.1597300354697766e-05,
+      "loss": 1.5269,
+      "step": 818
+    },
+    {
+      "epoch": 0.5047765793528506,
+      "grad_norm": 0.7054306864738464,
+      "learning_rate": 1.1576040623562965e-05,
+      "loss": 1.5446,
+      "step": 819
+    },
+    {
+      "epoch": 0.5053929121725732,
+      "grad_norm": 0.6658369898796082,
+      "learning_rate": 1.1554773585137186e-05,
+      "loss": 1.4736,
+      "step": 820
+    },
+    {
+      "epoch": 0.5060092449922958,
+      "grad_norm": 0.7052777409553528,
+      "learning_rate": 1.1533499338024765e-05,
+      "loss": 1.5275,
+      "step": 821
+    },
+    {
+      "epoch": 0.5066255778120184,
+      "grad_norm": 0.6908546090126038,
+      "learning_rate": 1.1512217980863463e-05,
+      "loss": 1.4999,
+      "step": 822
+    },
+    {
+      "epoch": 0.5072419106317412,
+      "grad_norm": 0.7249552011489868,
+      "learning_rate": 1.1490929612323998e-05,
+      "loss": 1.509,
+      "step": 823
+    },
+    {
+      "epoch": 0.5078582434514638,
+      "grad_norm": 0.7018350958824158,
+      "learning_rate": 1.1469634331109606e-05,
+      "loss": 1.5244,
+      "step": 824
+    },
+    {
+      "epoch": 0.5084745762711864,
+      "grad_norm": 0.7077041864395142,
+      "learning_rate": 1.144833223595557e-05,
+      "loss": 1.5602,
+      "step": 825
+    },
+    {
+      "epoch": 0.509090909090909,
+      "grad_norm": 0.6926451325416565,
+      "learning_rate": 1.1427023425628762e-05,
+      "loss": 1.5393,
+      "step": 826
+    },
+    {
+      "epoch": 0.5097072419106318,
+      "grad_norm": 0.7010395526885986,
+      "learning_rate": 1.1405707998927196e-05,
+      "loss": 1.5072,
+      "step": 827
+    },
+    {
+      "epoch": 0.5103235747303544,
+      "grad_norm": 0.7087801098823547,
+      "learning_rate": 1.1384386054679557e-05,
+      "loss": 1.4829,
+      "step": 828
+    },
+    {
+      "epoch": 0.510939907550077,
+      "grad_norm": 0.7039773464202881,
+      "learning_rate": 1.1363057691744747e-05,
+      "loss": 1.5279,
+      "step": 829
+    },
+    {
+      "epoch": 0.5115562403697997,
+      "grad_norm": 0.7164170742034912,
+      "learning_rate": 1.1341723009011441e-05,
+      "loss": 1.5849,
+      "step": 830
+    },
+    {
+      "epoch": 0.5121725731895224,
+      "grad_norm": 0.6886072158813477,
+      "learning_rate": 1.1320382105397594e-05,
+      "loss": 1.5164,
+      "step": 831
+    },
+    {
+      "epoch": 0.512788906009245,
+      "grad_norm": 0.7064073085784912,
+      "learning_rate": 1.1299035079850022e-05,
+      "loss": 1.4717,
+      "step": 832
+    },
+    {
+      "epoch": 0.5134052388289676,
+      "grad_norm": 0.6959893703460693,
+      "learning_rate": 1.1277682031343922e-05,
+      "loss": 1.4692,
+      "step": 833
+    },
+    {
+      "epoch": 0.5140215716486903,
+      "grad_norm": 0.7060757279396057,
+      "learning_rate": 1.1256323058882414e-05,
+      "loss": 1.4552,
+      "step": 834
+    },
+    {
+      "epoch": 0.514637904468413,
+      "grad_norm": 0.6915184259414673,
+      "learning_rate": 1.1234958261496078e-05,
+      "loss": 1.5206,
+      "step": 835
+    },
+    {
+      "epoch": 0.5152542372881356,
+      "grad_norm": 0.717655599117279,
+      "learning_rate": 1.121358773824251e-05,
+      "loss": 1.5273,
+      "step": 836
+    },
+    {
+      "epoch": 0.5158705701078582,
+      "grad_norm": 0.7036489248275757,
+      "learning_rate": 1.1192211588205855e-05,
+      "loss": 1.5256,
+      "step": 837
+    },
+    {
+      "epoch": 0.5164869029275809,
+      "grad_norm": 0.7183378338813782,
+      "learning_rate": 1.117082991049634e-05,
+      "loss": 1.5777,
+      "step": 838
+    },
+    {
+      "epoch": 0.5171032357473035,
+      "grad_norm": 0.7081940770149231,
+      "learning_rate": 1.114944280424982e-05,
+      "loss": 1.5345,
+      "step": 839
+    },
+    {
+      "epoch": 0.5177195685670262,
+      "grad_norm": 0.6747557520866394,
+      "learning_rate": 1.1128050368627328e-05,
+      "loss": 1.466,
+      "step": 840
+    },
+    {
+      "epoch": 0.5183359013867489,
+      "grad_norm": 0.7015603184700012,
+      "learning_rate": 1.11066527028146e-05,
+      "loss": 1.5691,
+      "step": 841
+    },
+    {
+      "epoch": 0.5189522342064715,
+      "grad_norm": 0.7690655589103699,
+      "learning_rate": 1.1085249906021625e-05,
+      "loss": 1.5238,
+      "step": 842
+    },
+    {
+      "epoch": 0.5195685670261941,
+      "grad_norm": 0.6984804272651672,
+      "learning_rate": 1.1063842077482178e-05,
+      "loss": 1.5001,
+      "step": 843
+    },
+    {
+      "epoch": 0.5201848998459168,
+      "grad_norm": 0.6944164037704468,
+      "learning_rate": 1.1042429316453366e-05,
+      "loss": 1.4691,
+      "step": 844
+    },
+    {
+      "epoch": 0.5208012326656395,
+      "grad_norm": 0.7170011401176453,
+      "learning_rate": 1.1021011722215167e-05,
+      "loss": 1.5263,
+      "step": 845
+    },
+    {
+      "epoch": 0.5214175654853621,
+      "grad_norm": 0.7038116455078125,
+      "learning_rate": 1.099958939406996e-05,
+      "loss": 1.5281,
+      "step": 846
+    },
+    {
+      "epoch": 0.5220338983050847,
+      "grad_norm": 0.6998594403266907,
+      "learning_rate": 1.097816243134209e-05,
+      "loss": 1.5568,
+      "step": 847
+    },
+    {
+      "epoch": 0.5226502311248074,
+      "grad_norm": 0.7281408309936523,
+      "learning_rate": 1.0956730933377373e-05,
+      "loss": 1.4949,
+      "step": 848
+    },
+    {
+      "epoch": 0.5232665639445301,
+      "grad_norm": 0.6806961894035339,
+      "learning_rate": 1.0935294999542651e-05,
+      "loss": 1.4645,
+      "step": 849
+    },
+    {
+      "epoch": 0.5238828967642527,
+      "grad_norm": 0.6968141794204712,
+      "learning_rate": 1.0913854729225355e-05,
+      "loss": 1.5021,
+      "step": 850
+    },
+    {
+      "epoch": 0.5244992295839753,
+      "grad_norm": 0.6870151162147522,
+      "learning_rate": 1.0892410221832996e-05,
+      "loss": 1.4911,
+      "step": 851
+    },
+    {
+      "epoch": 0.5251155624036979,
+      "grad_norm": 0.6921217441558838,
+      "learning_rate": 1.0870961576792748e-05,
+      "loss": 1.5192,
+      "step": 852
+    },
+    {
+      "epoch": 0.5257318952234207,
+      "grad_norm": 0.6947214603424072,
+      "learning_rate": 1.0849508893550962e-05,
+      "loss": 1.5065,
+      "step": 853
+    },
+    {
+      "epoch": 0.5263482280431433,
+      "grad_norm": 0.7072527408599854,
+      "learning_rate": 1.082805227157271e-05,
+      "loss": 1.5083,
+      "step": 854
+    },
+    {
+      "epoch": 0.5269645608628659,
+      "grad_norm": 0.6911188960075378,
+      "learning_rate": 1.0806591810341331e-05,
+      "loss": 1.4197,
+      "step": 855
+    },
+    {
+      "epoch": 0.5275808936825886,
+      "grad_norm": 0.6980645060539246,
+      "learning_rate": 1.0785127609357965e-05,
+      "loss": 1.4405,
+      "step": 856
+    },
+    {
+      "epoch": 0.5281972265023113,
+      "grad_norm": 0.7004606127738953,
+      "learning_rate": 1.076365976814108e-05,
+      "loss": 1.4992,
+      "step": 857
+    },
+    {
+      "epoch": 0.5288135593220339,
+      "grad_norm": 0.7196199297904968,
+      "learning_rate": 1.0742188386226044e-05,
+      "loss": 1.5539,
+      "step": 858
+    },
+    {
+      "epoch": 0.5294298921417565,
+      "grad_norm": 0.7004305124282837,
+      "learning_rate": 1.0720713563164615e-05,
+      "loss": 1.46,
+      "step": 859
+    },
+    {
+      "epoch": 0.5300462249614792,
+      "grad_norm": 0.7296655178070068,
+      "learning_rate": 1.0699235398524528e-05,
+      "loss": 1.5443,
+      "step": 860
+    },
+    {
+      "epoch": 0.5306625577812019,
+      "grad_norm": 0.7011130452156067,
+      "learning_rate": 1.0677753991889e-05,
+      "loss": 1.4772,
+      "step": 861
+    },
+    {
+      "epoch": 0.5312788906009245,
+      "grad_norm": 0.6892836093902588,
+      "learning_rate": 1.0656269442856276e-05,
+      "loss": 1.5105,
+      "step": 862
+    },
+    {
+      "epoch": 0.5318952234206471,
+      "grad_norm": 0.7163991332054138,
+      "learning_rate": 1.063478185103918e-05,
+      "loss": 1.479,
+      "step": 863
+    },
+    {
+      "epoch": 0.5325115562403698,
+      "grad_norm": 0.6904085874557495,
+      "learning_rate": 1.0613291316064643e-05,
+      "loss": 1.4962,
+      "step": 864
+    },
+    {
+      "epoch": 0.5331278890600925,
+      "grad_norm": 0.696327805519104,
+      "learning_rate": 1.0591797937573233e-05,
+      "loss": 1.4741,
+      "step": 865
+    },
+    {
+      "epoch": 0.5337442218798151,
+      "grad_norm": 0.7276609539985657,
+      "learning_rate": 1.0570301815218717e-05,
+      "loss": 1.5051,
+      "step": 866
+    },
+    {
+      "epoch": 0.5343605546995378,
+      "grad_norm": 0.7312154173851013,
+      "learning_rate": 1.0548803048667567e-05,
+      "loss": 1.4949,
+      "step": 867
+    },
+    {
+      "epoch": 0.5349768875192604,
+      "grad_norm": 0.6947066187858582,
+      "learning_rate": 1.0527301737598523e-05,
+      "loss": 1.4803,
+      "step": 868
+    },
+    {
+      "epoch": 0.535593220338983,
+      "grad_norm": 0.7186310887336731,
+      "learning_rate": 1.0505797981702126e-05,
+      "loss": 1.5428,
+      "step": 869
+    },
+    {
+      "epoch": 0.5362095531587057,
+      "grad_norm": 0.7036899924278259,
+      "learning_rate": 1.0484291880680244e-05,
+      "loss": 1.5053,
+      "step": 870
+    },
+    {
+      "epoch": 0.5368258859784284,
+      "grad_norm": 0.6898105144500732,
+      "learning_rate": 1.0462783534245626e-05,
+      "loss": 1.5225,
+      "step": 871
+    },
+    {
+      "epoch": 0.537442218798151,
+      "grad_norm": 0.7095423340797424,
+      "learning_rate": 1.0441273042121428e-05,
+      "loss": 1.5327,
+      "step": 872
+    },
+    {
+      "epoch": 0.5380585516178736,
+      "grad_norm": 0.7025151252746582,
+      "learning_rate": 1.0419760504040754e-05,
+      "loss": 1.5009,
+      "step": 873
+    },
+    {
+      "epoch": 0.5386748844375963,
+      "grad_norm": 0.7103621959686279,
+      "learning_rate": 1.0398246019746197e-05,
+      "loss": 1.5185,
+      "step": 874
+    },
+    {
+      "epoch": 0.539291217257319,
+      "grad_norm": 0.6966539025306702,
+      "learning_rate": 1.0376729688989369e-05,
+      "loss": 1.5377,
+      "step": 875
+    },
+    {
+      "epoch": 0.5399075500770416,
+      "grad_norm": 0.6908361315727234,
+      "learning_rate": 1.0355211611530444e-05,
+      "loss": 1.4867,
+      "step": 876
+    },
+    {
+      "epoch": 0.5405238828967642,
+      "grad_norm": 0.7360613346099854,
+      "learning_rate": 1.0333691887137705e-05,
+      "loss": 1.5054,
+      "step": 877
+    },
+    {
+      "epoch": 0.5411402157164868,
+      "grad_norm": 0.7125060558319092,
+      "learning_rate": 1.0312170615587056e-05,
+      "loss": 1.488,
+      "step": 878
+    },
+    {
+      "epoch": 0.5417565485362096,
+      "grad_norm": 0.7014859318733215,
+      "learning_rate": 1.0290647896661578e-05,
+      "loss": 1.4638,
+      "step": 879
+    },
+    {
+      "epoch": 0.5423728813559322,
+      "grad_norm": 0.6928877234458923,
+      "learning_rate": 1.0269123830151072e-05,
+      "loss": 1.5271,
+      "step": 880
+    },
+    {
+      "epoch": 0.5429892141756548,
+      "grad_norm": 0.69442218542099,
+      "learning_rate": 1.0247598515851577e-05,
+      "loss": 1.4356,
+      "step": 881
+    },
+    {
+      "epoch": 0.5436055469953776,
+      "grad_norm": 0.6790909767150879,
+      "learning_rate": 1.0226072053564922e-05,
+      "loss": 1.4827,
+      "step": 882
+    },
+    {
+      "epoch": 0.5442218798151002,
+      "grad_norm": 0.7216408848762512,
+      "learning_rate": 1.0204544543098253e-05,
+      "loss": 1.508,
+      "step": 883
+    },
+    {
+      "epoch": 0.5448382126348228,
+      "grad_norm": 0.7167052626609802,
+      "learning_rate": 1.0183016084263589e-05,
+      "loss": 1.5599,
+      "step": 884
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 0.7073312997817993,
+      "learning_rate": 1.0161486776877326e-05,
+      "loss": 1.4621,
+      "step": 885
+    },
+    {
+      "epoch": 0.5460708782742681,
+      "grad_norm": 0.7041304707527161,
+      "learning_rate": 1.0139956720759816e-05,
+      "loss": 1.4712,
+      "step": 886
+    },
+    {
+      "epoch": 0.5466872110939908,
+      "grad_norm": 0.7097397446632385,
+      "learning_rate": 1.0118426015734868e-05,
+      "loss": 1.5191,
+      "step": 887
+    },
+    {
+      "epoch": 0.5473035439137134,
+      "grad_norm": 0.702149510383606,
+      "learning_rate": 1.0096894761629304e-05,
+      "loss": 1.5135,
+      "step": 888
+    },
+    {
+      "epoch": 0.547919876733436,
+      "grad_norm": 0.7111746072769165,
+      "learning_rate": 1.0075363058272491e-05,
+      "loss": 1.4854,
+      "step": 889
+    },
+    {
+      "epoch": 0.5485362095531587,
+      "grad_norm": 0.6934394240379333,
+      "learning_rate": 1.0053831005495882e-05,
+      "loss": 1.4913,
+      "step": 890
+    },
+    {
+      "epoch": 0.5491525423728814,
+      "grad_norm": 0.6981213092803955,
+      "learning_rate": 1.0032298703132544e-05,
+      "loss": 1.5221,
+      "step": 891
+    },
+    {
+      "epoch": 0.549768875192604,
+      "grad_norm": 0.6972723603248596,
+      "learning_rate": 1.001076625101671e-05,
+      "loss": 1.4649,
+      "step": 892
+    },
+    {
+      "epoch": 0.5503852080123267,
+      "grad_norm": 0.692736029624939,
+      "learning_rate": 9.989233748983294e-06,
+      "loss": 1.4676,
+      "step": 893
+    },
+    {
+      "epoch": 0.5510015408320493,
+      "grad_norm": 0.7195137739181519,
+      "learning_rate": 9.967701296867457e-06,
+      "loss": 1.4992,
+      "step": 894
+    },
+    {
+      "epoch": 0.551617873651772,
+      "grad_norm": 0.7197251915931702,
+      "learning_rate": 9.946168994504122e-06,
+      "loss": 1.5364,
+      "step": 895
+    },
+    {
+      "epoch": 0.5522342064714946,
+      "grad_norm": 0.7067369818687439,
+      "learning_rate": 9.92463694172751e-06,
+      "loss": 1.512,
+      "step": 896
+    },
+    {
+      "epoch": 0.5528505392912173,
+      "grad_norm": 0.7338434457778931,
+      "learning_rate": 9.903105238370698e-06,
+      "loss": 1.5188,
+      "step": 897
+    },
+    {
+      "epoch": 0.5534668721109399,
+      "grad_norm": 0.7214980125427246,
+      "learning_rate": 9.881573984265136e-06,
+      "loss": 1.5547,
+      "step": 898
+    },
+    {
+      "epoch": 0.5540832049306625,
+      "grad_norm": 0.6925179958343506,
+      "learning_rate": 9.860043279240186e-06,
+      "loss": 1.5145,
+      "step": 899
+    },
+    {
+      "epoch": 0.5546995377503852,
+      "grad_norm": 0.7407251596450806,
+      "learning_rate": 9.838513223122676e-06,
+      "loss": 1.4758,
+      "step": 900
+    },
+    {
+      "epoch": 0.5553158705701079,
+      "grad_norm": 0.7122443318367004,
+      "learning_rate": 9.816983915736418e-06,
+      "loss": 1.5101,
+      "step": 901
+    },
+    {
+      "epoch": 0.5559322033898305,
+      "grad_norm": 0.7500817179679871,
+      "learning_rate": 9.79545545690175e-06,
+      "loss": 1.4971,
+      "step": 902
+    },
+    {
+      "epoch": 0.5565485362095531,
+      "grad_norm": 0.7213168144226074,
+      "learning_rate": 9.773927946435083e-06,
+      "loss": 1.5557,
+      "step": 903
+    },
+    {
+      "epoch": 0.5571648690292758,
+      "grad_norm": 0.715071976184845,
+      "learning_rate": 9.752401484148426e-06,
+      "loss": 1.5035,
+      "step": 904
+    },
+    {
+      "epoch": 0.5577812018489985,
+      "grad_norm": 0.7278201580047607,
+      "learning_rate": 9.730876169848932e-06,
+      "loss": 1.5366,
+      "step": 905
+    },
+    {
+      "epoch": 0.5583975346687211,
+      "grad_norm": 0.7042462229728699,
+      "learning_rate": 9.709352103338424e-06,
+      "loss": 1.5166,
+      "step": 906
+    },
+    {
+      "epoch": 0.5590138674884437,
+      "grad_norm": 0.6888935565948486,
+      "learning_rate": 9.687829384412945e-06,
+      "loss": 1.5074,
+      "step": 907
+    },
+    {
+      "epoch": 0.5596302003081665,
+      "grad_norm": 0.7048391699790955,
+      "learning_rate": 9.6663081128623e-06,
+      "loss": 1.4729,
+      "step": 908
+    },
+    {
+      "epoch": 0.5602465331278891,
+      "grad_norm": 0.7012971639633179,
+      "learning_rate": 9.644788388469557e-06,
+      "loss": 1.4939,
+      "step": 909
+    },
+    {
+      "epoch": 0.5608628659476117,
+      "grad_norm": 0.6953085660934448,
+      "learning_rate": 9.623270311010633e-06,
+      "loss": 1.4784,
+      "step": 910
+    },
+    {
+      "epoch": 0.5614791987673343,
+      "grad_norm": 0.7051222324371338,
+      "learning_rate": 9.60175398025381e-06,
+      "loss": 1.5287,
+      "step": 911
+    },
+    {
+      "epoch": 0.562095531587057,
+      "grad_norm": 0.7343626618385315,
+      "learning_rate": 9.580239495959248e-06,
+      "loss": 1.5005,
+      "step": 912
+    },
+    {
+      "epoch": 0.5627118644067797,
+      "grad_norm": 0.7238948345184326,
+      "learning_rate": 9.558726957878574e-06,
+      "loss": 1.44,
+      "step": 913
+    },
+    {
+      "epoch": 0.5633281972265023,
+      "grad_norm": 0.7163533568382263,
+      "learning_rate": 9.53721646575438e-06,
+      "loss": 1.5159,
+      "step": 914
+    },
+    {
+      "epoch": 0.5639445300462249,
+      "grad_norm": 0.7252680063247681,
+      "learning_rate": 9.515708119319759e-06,
+      "loss": 1.5543,
+      "step": 915
+    },
+    {
+      "epoch": 0.5645608628659476,
+      "grad_norm": 0.6947450637817383,
+      "learning_rate": 9.494202018297879e-06,
+      "loss": 1.4446,
+      "step": 916
+    },
+    {
+      "epoch": 0.5651771956856703,
+      "grad_norm": 0.6983763575553894,
+      "learning_rate": 9.47269826240148e-06,
+      "loss": 1.5158,
+      "step": 917
+    },
+    {
+      "epoch": 0.5657935285053929,
+      "grad_norm": 0.710814356803894,
+      "learning_rate": 9.451196951332437e-06,
+      "loss": 1.4598,
+      "step": 918
+    },
+    {
+      "epoch": 0.5664098613251156,
+      "grad_norm": 0.6948243379592896,
+      "learning_rate": 9.429698184781284e-06,
+      "loss": 1.5017,
+      "step": 919
+    },
+    {
+      "epoch": 0.5670261941448382,
+      "grad_norm": 0.7261428833007812,
+      "learning_rate": 9.408202062426765e-06,
+      "loss": 1.4654,
+      "step": 920
+    },
+    {
+      "epoch": 0.5676425269645609,
+      "grad_norm": 0.7493246793746948,
+      "learning_rate": 9.38670868393536e-06,
+      "loss": 1.4825,
+      "step": 921
+    },
+    {
+      "epoch": 0.5682588597842835,
+      "grad_norm": 0.711744487285614,
+      "learning_rate": 9.365218148960822e-06,
+      "loss": 1.5039,
+      "step": 922
+    },
+    {
+      "epoch": 0.5688751926040062,
+      "grad_norm": 0.7111184597015381,
+      "learning_rate": 9.343730557143728e-06,
+      "loss": 1.3966,
+      "step": 923
+    },
+    {
+      "epoch": 0.5694915254237288,
+      "grad_norm": 0.7288036942481995,
+      "learning_rate": 9.322246008111007e-06,
+      "loss": 1.5065,
+      "step": 924
+    },
+    {
+      "epoch": 0.5701078582434514,
+      "grad_norm": 0.713474690914154,
+      "learning_rate": 9.300764601475474e-06,
+      "loss": 1.5008,
+      "step": 925
+    },
+    {
+      "epoch": 0.5707241910631741,
+      "grad_norm": 0.7069692015647888,
+      "learning_rate": 9.279286436835387e-06,
+      "loss": 1.511,
+      "step": 926
+    },
+    {
+      "epoch": 0.5713405238828968,
+      "grad_norm": 0.7019029855728149,
+      "learning_rate": 9.257811613773961e-06,
+      "loss": 1.4842,
+      "step": 927
+    },
+    {
+      "epoch": 0.5719568567026194,
+      "grad_norm": 0.6917131543159485,
+      "learning_rate": 9.236340231858921e-06,
+      "loss": 1.5162,
+      "step": 928
+    },
+    {
+      "epoch": 0.572573189522342,
+      "grad_norm": 0.7142927050590515,
+      "learning_rate": 9.214872390642036e-06,
+      "loss": 1.5341,
+      "step": 929
+    },
+    {
+      "epoch": 0.5731895223420647,
+      "grad_norm": 0.6957735419273376,
+      "learning_rate": 9.193408189658672e-06,
+      "loss": 1.4746,
+      "step": 930
+    },
+    {
+      "epoch": 0.5738058551617874,
+      "grad_norm": 0.7198157906532288,
+      "learning_rate": 9.171947728427294e-06,
+      "loss": 1.5021,
+      "step": 931
+    },
+    {
+      "epoch": 0.57442218798151,
+      "grad_norm": 0.706337034702301,
+      "learning_rate": 9.15049110644904e-06,
+      "loss": 1.4442,
+      "step": 932
+    },
+    {
+      "epoch": 0.5750385208012326,
+      "grad_norm": 0.70888352394104,
+      "learning_rate": 9.129038423207253e-06,
+      "loss": 1.4741,
+      "step": 933
+    },
+    {
+      "epoch": 0.5756548536209554,
+      "grad_norm": 0.7035511136054993,
+      "learning_rate": 9.107589778167006e-06,
+      "loss": 1.4447,
+      "step": 934
+    },
+    {
+      "epoch": 0.576271186440678,
+      "grad_norm": 0.7262841463088989,
+      "learning_rate": 9.086145270774648e-06,
+      "loss": 1.4885,
+      "step": 935
+    },
+    {
+      "epoch": 0.5768875192604006,
+      "grad_norm": 0.7074030637741089,
+      "learning_rate": 9.064705000457349e-06,
+      "loss": 1.499,
+      "step": 936
+    },
+    {
+      "epoch": 0.5775038520801232,
+      "grad_norm": 0.688092827796936,
+      "learning_rate": 9.043269066622633e-06,
+      "loss": 1.4529,
+      "step": 937
+    },
+    {
+      "epoch": 0.578120184899846,
+      "grad_norm": 0.7015891671180725,
+      "learning_rate": 9.021837568657912e-06,
+      "loss": 1.5045,
+      "step": 938
+    },
+    {
+      "epoch": 0.5787365177195686,
+      "grad_norm": 0.6968741416931152,
+      "learning_rate": 9.000410605930038e-06,
+      "loss": 1.5101,
+      "step": 939
+    },
+    {
+      "epoch": 0.5793528505392912,
+      "grad_norm": 0.8305391073226929,
+      "learning_rate": 8.978988277784838e-06,
+      "loss": 1.5185,
+      "step": 940
+    },
+    {
+      "epoch": 0.5799691833590138,
+      "grad_norm": 0.7060264945030212,
+      "learning_rate": 8.957570683546637e-06,
+      "loss": 1.4699,
+      "step": 941
+    },
+    {
+      "epoch": 0.5805855161787365,
+      "grad_norm": 0.702182948589325,
+      "learning_rate": 8.936157922517825e-06,
+      "loss": 1.4835,
+      "step": 942
+    },
+    {
+      "epoch": 0.5812018489984592,
+      "grad_norm": 0.7086932063102722,
+      "learning_rate": 8.91475009397838e-06,
+      "loss": 1.4811,
+      "step": 943
+    },
+    {
+      "epoch": 0.5818181818181818,
+      "grad_norm": 0.7161465883255005,
+      "learning_rate": 8.893347297185403e-06,
+      "loss": 1.5022,
+      "step": 944
+    },
+    {
+      "epoch": 0.5824345146379045,
+      "grad_norm": 0.7002130150794983,
+      "learning_rate": 8.871949631372677e-06,
+      "loss": 1.425,
+      "step": 945
+    },
+    {
+      "epoch": 0.5830508474576271,
+      "grad_norm": 0.7049859762191772,
+      "learning_rate": 8.850557195750182e-06,
+      "loss": 1.4553,
+      "step": 946
+    },
+    {
+      "epoch": 0.5836671802773498,
+      "grad_norm": 0.7008825540542603,
+      "learning_rate": 8.829170089503667e-06,
+      "loss": 1.4322,
+      "step": 947
+    },
+    {
+      "epoch": 0.5842835130970724,
+      "grad_norm": 0.7909539937973022,
+      "learning_rate": 8.807788411794147e-06,
+      "loss": 1.4421,
+      "step": 948
+    },
+    {
+      "epoch": 0.5848998459167951,
+      "grad_norm": 0.699917733669281,
+      "learning_rate": 8.786412261757491e-06,
+      "loss": 1.5129,
+      "step": 949
+    },
+    {
+      "epoch": 0.5855161787365177,
+      "grad_norm": 0.7193233370780945,
+      "learning_rate": 8.765041738503927e-06,
+      "loss": 1.5015,
+      "step": 950
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1622,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.221676915186729e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}