{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 9820,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000203685692607737,
      "grad_norm": 0.43429237604141235,
      "learning_rate": 4.23728813559322e-06,
      "loss": 3.0715341567993164,
      "step": 2
    },
    {
      "epoch": 0.000407371385215474,
      "grad_norm": 0.3986178934574127,
      "learning_rate": 1.2711864406779663e-05,
      "loss": 3.0496439933776855,
      "step": 4
    },
    {
      "epoch": 0.000611057077823211,
      "grad_norm": 0.347873330116272,
      "learning_rate": 2.11864406779661e-05,
      "loss": 3.0722804069519043,
      "step": 6
    },
    {
      "epoch": 0.000814742770430948,
      "grad_norm": 0.38114133477211,
      "learning_rate": 2.9661016949152544e-05,
      "loss": 3.0841379165649414,
      "step": 8
    },
    {
      "epoch": 0.001018428463038685,
      "grad_norm": 0.3545224964618683,
      "learning_rate": 3.813559322033899e-05,
      "loss": 2.963942527770996,
      "step": 10
    },
    {
      "epoch": 0.001222114155646422,
      "grad_norm": 0.3599280118942261,
      "learning_rate": 4.6610169491525425e-05,
      "loss": 3.152094841003418,
      "step": 12
    },
    {
      "epoch": 0.001425799848254159,
      "grad_norm": 0.3866329491138458,
      "learning_rate": 5.508474576271186e-05,
      "loss": 3.1587839126586914,
      "step": 14
    },
    {
      "epoch": 0.001629485540861896,
      "grad_norm": 0.38113707304000854,
      "learning_rate": 6.35593220338983e-05,
      "loss": 3.074766159057617,
      "step": 16
    },
    {
      "epoch": 0.001833171233469633,
      "grad_norm": 0.4228023886680603,
      "learning_rate": 7.203389830508474e-05,
      "loss": 3.1017208099365234,
      "step": 18
    },
    {
      "epoch": 0.00203685692607737,
      "grad_norm": 0.36811915040016174,
      "learning_rate": 8.050847457627118e-05,
      "loss": 2.957829236984253,
      "step": 20
    },
    {
      "epoch": 0.002240542618685107,
      "grad_norm": 0.31414222717285156,
      "learning_rate": 8.898305084745763e-05,
      "loss": 2.90093994140625,
      "step": 22
    },
    {
      "epoch": 0.002444228311292844,
      "grad_norm": 0.3228512108325958,
      "learning_rate": 9.745762711864407e-05,
      "loss": 3.069493293762207,
      "step": 24
    },
    {
      "epoch": 0.002647914003900581,
      "grad_norm": 0.3249562084674835,
      "learning_rate": 0.0001059322033898305,
      "loss": 3.09674072265625,
      "step": 26
    },
    {
      "epoch": 0.002851599696508318,
      "grad_norm": 0.3768482506275177,
      "learning_rate": 0.00011440677966101696,
      "loss": 3.070322036743164,
      "step": 28
    },
    {
      "epoch": 0.003055285389116055,
      "grad_norm": 0.32275235652923584,
      "learning_rate": 0.0001228813559322034,
      "loss": 2.8644938468933105,
      "step": 30
    },
    {
      "epoch": 0.003258971081723792,
      "grad_norm": 0.34401702880859375,
      "learning_rate": 0.00013135593220338986,
      "loss": 3.1347928047180176,
      "step": 32
    },
    {
      "epoch": 0.003462656774331529,
      "grad_norm": 0.3586300313472748,
      "learning_rate": 0.00013983050847457627,
      "loss": 3.106912136077881,
      "step": 34
    },
    {
      "epoch": 0.003666342466939266,
      "grad_norm": 0.3450547754764557,
      "learning_rate": 0.0001483050847457627,
      "loss": 3.0832133293151855,
      "step": 36
    },
    {
      "epoch": 0.003870028159547003,
      "grad_norm": 0.34801945090293884,
      "learning_rate": 0.00015677966101694914,
      "loss": 2.9645042419433594,
      "step": 38
    },
    {
      "epoch": 0.00407371385215474,
      "grad_norm": 0.34780916571617126,
      "learning_rate": 0.0001652542372881356,
      "loss": 3.014695167541504,
      "step": 40
    },
    {
      "epoch": 0.004277399544762477,
      "grad_norm": 0.3741895854473114,
      "learning_rate": 0.00017372881355932204,
      "loss": 3.1716694831848145,
      "step": 42
    },
    {
      "epoch": 0.004481085237370214,
      "grad_norm": 0.3800599277019501,
      "learning_rate": 0.00018220338983050845,
      "loss": 2.9408786296844482,
      "step": 44
    },
    {
      "epoch": 0.004684770929977951,
      "grad_norm": 0.31223005056381226,
      "learning_rate": 0.00019067796610169492,
      "loss": 3.08722186088562,
      "step": 46
    },
    {
      "epoch": 0.004888456622585688,
      "grad_norm": 0.3792102634906769,
      "learning_rate": 0.00019915254237288136,
      "loss": 3.1528396606445312,
      "step": 48
    },
    {
      "epoch": 0.005092142315193425,
      "grad_norm": 0.3707769215106964,
      "learning_rate": 0.0002076271186440678,
      "loss": 3.0891048908233643,
      "step": 50
    },
    {
      "epoch": 0.005295828007801162,
      "grad_norm": 0.33640429377555847,
      "learning_rate": 0.00021610169491525426,
      "loss": 3.0794899463653564,
      "step": 52
    },
    {
      "epoch": 0.005499513700408899,
      "grad_norm": 0.29960381984710693,
      "learning_rate": 0.0002245762711864407,
      "loss": 2.9251627922058105,
      "step": 54
    },
    {
      "epoch": 0.005703199393016636,
      "grad_norm": 0.3139283061027527,
      "learning_rate": 0.0002330508474576271,
      "loss": 2.988818645477295,
      "step": 56
    },
    {
      "epoch": 0.005906885085624373,
      "grad_norm": 0.3302924335002899,
      "learning_rate": 0.00024152542372881357,
      "loss": 2.958843946456909,
      "step": 58
    },
    {
      "epoch": 0.00611057077823211,
      "grad_norm": 0.32685577869415283,
      "learning_rate": 0.00025,
      "loss": 3.157499313354492,
      "step": 60
    },
    {
      "epoch": 0.006314256470839847,
      "grad_norm": 0.3420144319534302,
      "learning_rate": 0.00025847457627118644,
      "loss": 3.0464529991149902,
      "step": 62
    },
    {
      "epoch": 0.006517942163447584,
      "grad_norm": 0.3040681481361389,
      "learning_rate": 0.0002669491525423729,
      "loss": 2.9510767459869385,
      "step": 64
    },
    {
      "epoch": 0.006721627856055321,
      "grad_norm": 0.3473084270954132,
      "learning_rate": 0.0002754237288135593,
      "loss": 3.094024896621704,
      "step": 66
    },
    {
      "epoch": 0.006925313548663058,
      "grad_norm": 0.2590670883655548,
      "learning_rate": 0.0002838983050847458,
      "loss": 3.0465497970581055,
      "step": 68
    },
    {
      "epoch": 0.0071289992412707955,
      "grad_norm": 0.3464723229408264,
      "learning_rate": 0.0002923728813559322,
      "loss": 2.9268321990966797,
      "step": 70
    },
    {
      "epoch": 0.007332684933878532,
      "grad_norm": 0.2758452892303467,
      "learning_rate": 0.00030084745762711863,
      "loss": 2.918442964553833,
      "step": 72
    },
    {
      "epoch": 0.007536370626486269,
      "grad_norm": 0.30435070395469666,
      "learning_rate": 0.0003093220338983051,
      "loss": 3.0483314990997314,
      "step": 74
    },
    {
      "epoch": 0.007740056319094006,
      "grad_norm": 0.256240576505661,
      "learning_rate": 0.0003177966101694915,
      "loss": 2.9276201725006104,
      "step": 76
    },
    {
      "epoch": 0.007943742011701744,
      "grad_norm": 0.31318631768226624,
      "learning_rate": 0.000326271186440678,
      "loss": 3.086130142211914,
      "step": 78
    },
    {
      "epoch": 0.00814742770430948,
      "grad_norm": 0.31463009119033813,
      "learning_rate": 0.00033474576271186443,
      "loss": 3.006633758544922,
      "step": 80
    },
    {
      "epoch": 0.008351113396917216,
      "grad_norm": 0.2976848781108856,
      "learning_rate": 0.0003432203389830508,
      "loss": 2.82028865814209,
      "step": 82
    },
    {
      "epoch": 0.008554799089524954,
      "grad_norm": 0.35044899582862854,
      "learning_rate": 0.0003516949152542373,
      "loss": 3.1081886291503906,
      "step": 84
    },
    {
      "epoch": 0.00875848478213269,
      "grad_norm": 0.2826573848724365,
      "learning_rate": 0.00036016949152542374,
      "loss": 2.9417290687561035,
      "step": 86
    },
    {
      "epoch": 0.008962170474740428,
      "grad_norm": 0.35906901955604553,
      "learning_rate": 0.0003686440677966102,
      "loss": 3.0192437171936035,
      "step": 88
    },
    {
      "epoch": 0.009165856167348165,
      "grad_norm": 0.2926231026649475,
      "learning_rate": 0.0003771186440677966,
      "loss": 2.9435534477233887,
      "step": 90
    },
    {
      "epoch": 0.009369541859955902,
      "grad_norm": 0.29269522428512573,
      "learning_rate": 0.0003855932203389831,
      "loss": 3.122875213623047,
      "step": 92
    },
    {
      "epoch": 0.009573227552563639,
      "grad_norm": 0.2618827521800995,
      "learning_rate": 0.0003940677966101695,
      "loss": 2.7958455085754395,
      "step": 94
    },
    {
      "epoch": 0.009776913245171376,
      "grad_norm": 0.23626284301280975,
      "learning_rate": 0.00040254237288135593,
      "loss": 3.0104169845581055,
      "step": 96
    },
    {
      "epoch": 0.009980598937779113,
      "grad_norm": 0.3233124613761902,
      "learning_rate": 0.0004110169491525424,
      "loss": 3.1561360359191895,
      "step": 98
    },
    {
      "epoch": 0.01018428463038685,
      "grad_norm": 0.25498658418655396,
      "learning_rate": 0.0004194915254237288,
      "loss": 2.9825079441070557,
      "step": 100
    },
    {
      "epoch": 0.010387970322994588,
      "grad_norm": 0.26352331042289734,
      "learning_rate": 0.0004279661016949153,
      "loss": 3.002692699432373,
      "step": 102
    },
    {
      "epoch": 0.010591656015602325,
      "grad_norm": 0.25789007544517517,
      "learning_rate": 0.00043644067796610173,
      "loss": 2.7518932819366455,
      "step": 104
    },
    {
      "epoch": 0.010795341708210062,
      "grad_norm": 0.22616815567016602,
      "learning_rate": 0.0004449152542372881,
      "loss": 2.8698978424072266,
      "step": 106
    },
    {
      "epoch": 0.010999027400817799,
      "grad_norm": 0.2553200125694275,
      "learning_rate": 0.0004533898305084746,
      "loss": 2.7725863456726074,
      "step": 108
    },
    {
      "epoch": 0.011202713093425534,
      "grad_norm": 0.24017801880836487,
      "learning_rate": 0.00046186440677966104,
      "loss": 2.923917770385742,
      "step": 110
    },
    {
      "epoch": 0.011406398786033271,
      "grad_norm": 0.2561245262622833,
      "learning_rate": 0.0004703389830508475,
      "loss": 3.086453676223755,
      "step": 112
    },
    {
      "epoch": 0.011610084478641008,
      "grad_norm": 0.2561284601688385,
      "learning_rate": 0.0004788135593220339,
      "loss": 3.0930871963500977,
      "step": 114
    },
    {
      "epoch": 0.011813770171248746,
      "grad_norm": 0.2632646858692169,
      "learning_rate": 0.0004872881355932203,
      "loss": 3.094892978668213,
      "step": 116
    },
    {
      "epoch": 0.012017455863856483,
      "grad_norm": 0.2755280137062073,
      "learning_rate": 0.0004957627118644068,
      "loss": 2.8941264152526855,
      "step": 118
    },
    {
      "epoch": 0.01222114155646422,
      "grad_norm": 0.24211177229881287,
      "learning_rate": 0.0005042372881355932,
      "loss": 3.0071117877960205,
      "step": 120
    },
    {
      "epoch": 0.012424827249071957,
      "grad_norm": 0.21835361421108246,
      "learning_rate": 0.0005127118644067797,
      "loss": 2.7603557109832764,
      "step": 122
    },
    {
      "epoch": 0.012628512941679694,
      "grad_norm": 0.29309555888175964,
      "learning_rate": 0.0005211864406779662,
      "loss": 2.9338154792785645,
      "step": 124
    },
    {
      "epoch": 0.012832198634287431,
      "grad_norm": 0.23978236317634583,
      "learning_rate": 0.0005296610169491525,
      "loss": 2.859168529510498,
      "step": 126
    },
    {
      "epoch": 0.013035884326895168,
      "grad_norm": 0.24559195339679718,
      "learning_rate": 0.000538135593220339,
      "loss": 2.802013397216797,
      "step": 128
    },
    {
      "epoch": 0.013239570019502905,
      "grad_norm": 0.21800199151039124,
      "learning_rate": 0.0005466101694915254,
      "loss": 2.9812464714050293,
      "step": 130
    },
    {
      "epoch": 0.013443255712110642,
      "grad_norm": 0.2281453013420105,
      "learning_rate": 0.0005550847457627119,
      "loss": 2.9414823055267334,
      "step": 132
    },
    {
      "epoch": 0.01364694140471838,
      "grad_norm": 0.24012310802936554,
      "learning_rate": 0.0005635593220338984,
      "loss": 2.8748555183410645,
      "step": 134
    },
    {
      "epoch": 0.013850627097326117,
      "grad_norm": 0.23962637782096863,
      "learning_rate": 0.0005720338983050848,
      "loss": 3.017672061920166,
      "step": 136
    },
    {
      "epoch": 0.014054312789933854,
      "grad_norm": 0.2184254676103592,
      "learning_rate": 0.0005805084745762712,
      "loss": 2.839954376220703,
      "step": 138
    },
    {
      "epoch": 0.014257998482541591,
      "grad_norm": 0.2688688039779663,
      "learning_rate": 0.0005889830508474576,
      "loss": 2.832872152328491,
      "step": 140
    },
    {
      "epoch": 0.014461684175149326,
      "grad_norm": 0.22779379785060883,
      "learning_rate": 0.000597457627118644,
      "loss": 3.046627998352051,
      "step": 142
    },
    {
      "epoch": 0.014665369867757063,
      "grad_norm": 0.214941143989563,
      "learning_rate": 0.0006059322033898306,
      "loss": 2.678877592086792,
      "step": 144
    },
    {
      "epoch": 0.0148690555603648,
      "grad_norm": 0.2540000081062317,
      "learning_rate": 0.000614406779661017,
      "loss": 2.876096248626709,
      "step": 146
    },
    {
      "epoch": 0.015072741252972538,
      "grad_norm": 0.2230798453092575,
      "learning_rate": 0.0006228813559322035,
      "loss": 2.939134120941162,
      "step": 148
    },
    {
      "epoch": 0.015276426945580275,
      "grad_norm": 0.2185341715812683,
      "learning_rate": 0.0006313559322033898,
      "loss": 2.83620548248291,
      "step": 150
    },
    {
      "epoch": 0.015480112638188012,
      "grad_norm": 0.2135150283575058,
      "learning_rate": 0.0006398305084745762,
      "loss": 2.7651500701904297,
      "step": 152
    },
    {
      "epoch": 0.01568379833079575,
      "grad_norm": 0.19527095556259155,
      "learning_rate": 0.0006483050847457628,
      "loss": 2.8379552364349365,
      "step": 154
    },
    {
      "epoch": 0.015887484023403488,
      "grad_norm": 0.20186032354831696,
      "learning_rate": 0.0006567796610169492,
      "loss": 2.899538278579712,
      "step": 156
    },
    {
      "epoch": 0.01609116971601122,
      "grad_norm": 0.21385957300662994,
      "learning_rate": 0.0006652542372881356,
      "loss": 2.954371929168701,
      "step": 158
    },
    {
      "epoch": 0.01629485540861896,
      "grad_norm": 0.24298499524593353,
      "learning_rate": 0.0006737288135593221,
      "loss": 3.0152151584625244,
      "step": 160
    },
    {
      "epoch": 0.016498541101226696,
      "grad_norm": 0.23230932652950287,
      "learning_rate": 0.0006822033898305084,
      "loss": 2.9243545532226562,
      "step": 162
    },
    {
      "epoch": 0.016702226793834433,
      "grad_norm": 0.22520692646503448,
      "learning_rate": 0.000690677966101695,
      "loss": 3.0186548233032227,
      "step": 164
    },
    {
      "epoch": 0.01690591248644217,
      "grad_norm": 0.19831904768943787,
      "learning_rate": 0.0006991525423728814,
      "loss": 2.9022178649902344,
      "step": 166
    },
    {
      "epoch": 0.017109598179049907,
      "grad_norm": 0.2311016321182251,
      "learning_rate": 0.0007076271186440678,
      "loss": 3.0346760749816895,
      "step": 168
    },
    {
      "epoch": 0.017313283871657644,
      "grad_norm": 0.19864687323570251,
      "learning_rate": 0.0007161016949152543,
      "loss": 2.8734421730041504,
      "step": 170
    },
    {
      "epoch": 0.01751696956426538,
      "grad_norm": 0.21467842161655426,
      "learning_rate": 0.0007245762711864406,
      "loss": 2.9142980575561523,
      "step": 172
    },
    {
      "epoch": 0.01772065525687312,
      "grad_norm": 0.20455697178840637,
      "learning_rate": 0.0007330508474576271,
      "loss": 2.9625816345214844,
      "step": 174
    },
    {
      "epoch": 0.017924340949480855,
      "grad_norm": 0.25027167797088623,
      "learning_rate": 0.0007415254237288136,
      "loss": 2.8275511264801025,
      "step": 176
    },
    {
      "epoch": 0.018128026642088593,
      "grad_norm": 0.2081494778394699,
      "learning_rate": 0.00075,
      "loss": 2.79408597946167,
      "step": 178
    },
    {
      "epoch": 0.01833171233469633,
      "grad_norm": 0.19708451628684998,
      "learning_rate": 0.0007584745762711865,
      "loss": 2.7728185653686523,
      "step": 180
    },
    {
      "epoch": 0.018535398027304067,
      "grad_norm": 0.25516965985298157,
      "learning_rate": 0.0007669491525423729,
      "loss": 2.8916430473327637,
      "step": 182
    },
    {
      "epoch": 0.018739083719911804,
      "grad_norm": 0.20391100645065308,
      "learning_rate": 0.0007754237288135594,
      "loss": 2.9890096187591553,
      "step": 184
    },
    {
      "epoch": 0.01894276941251954,
      "grad_norm": 0.21976011991500854,
      "learning_rate": 0.0007838983050847458,
      "loss": 3.0404508113861084,
      "step": 186
    },
    {
      "epoch": 0.019146455105127278,
      "grad_norm": 0.20210285484790802,
      "learning_rate": 0.0007923728813559322,
      "loss": 2.758718967437744,
      "step": 188
    },
    {
      "epoch": 0.019350140797735015,
      "grad_norm": 0.2007555365562439,
      "learning_rate": 0.0008008474576271186,
      "loss": 2.8846611976623535,
      "step": 190
    },
    {
      "epoch": 0.019553826490342752,
      "grad_norm": 0.1996457278728485,
      "learning_rate": 0.0008093220338983051,
      "loss": 2.7096285820007324,
      "step": 192
    },
    {
      "epoch": 0.01975751218295049,
      "grad_norm": 0.20696577429771423,
      "learning_rate": 0.0008177966101694916,
      "loss": 2.7052865028381348,
      "step": 194
    },
    {
      "epoch": 0.019961197875558227,
      "grad_norm": 0.19655631482601166,
      "learning_rate": 0.000826271186440678,
      "loss": 2.7481884956359863,
      "step": 196
    },
    {
      "epoch": 0.020164883568165964,
      "grad_norm": 0.19629375636577606,
      "learning_rate": 0.0008347457627118644,
      "loss": 2.833449363708496,
      "step": 198
    },
    {
      "epoch": 0.0203685692607737,
      "grad_norm": 0.1913364678621292,
      "learning_rate": 0.0008432203389830508,
      "loss": 2.769970417022705,
      "step": 200
    },
    {
      "epoch": 0.020572254953381438,
      "grad_norm": 0.19474193453788757,
      "learning_rate": 0.0008516949152542373,
      "loss": 2.765143394470215,
      "step": 202
    },
    {
      "epoch": 0.020775940645989175,
      "grad_norm": 0.18722805380821228,
      "learning_rate": 0.0008601694915254238,
      "loss": 2.770843505859375,
      "step": 204
    },
    {
      "epoch": 0.020979626338596912,
      "grad_norm": 0.21279126405715942,
      "learning_rate": 0.0008686440677966102,
      "loss": 2.8868601322174072,
      "step": 206
    },
    {
      "epoch": 0.02118331203120465,
      "grad_norm": 0.1949380338191986,
      "learning_rate": 0.0008771186440677966,
      "loss": 2.9213500022888184,
      "step": 208
    },
    {
      "epoch": 0.021386997723812386,
      "grad_norm": 0.20765450596809387,
      "learning_rate": 0.000885593220338983,
      "loss": 2.824416160583496,
      "step": 210
    },
    {
      "epoch": 0.021590683416420123,
      "grad_norm": 0.2169354110956192,
      "learning_rate": 0.0008940677966101694,
      "loss": 2.8556411266326904,
      "step": 212
    },
    {
      "epoch": 0.02179436910902786,
      "grad_norm": 0.21635349094867706,
      "learning_rate": 0.000902542372881356,
      "loss": 2.8351492881774902,
      "step": 214
    },
    {
      "epoch": 0.021998054801635598,
      "grad_norm": 0.1571572721004486,
      "learning_rate": 0.0009110169491525424,
      "loss": 2.787085771560669,
      "step": 216
    },
    {
      "epoch": 0.022201740494243335,
      "grad_norm": 0.22742828726768494,
      "learning_rate": 0.0009194915254237289,
      "loss": 3.0058445930480957,
      "step": 218
    },
    {
      "epoch": 0.02240542618685107,
      "grad_norm": 0.19112920761108398,
      "learning_rate": 0.0009279661016949152,
      "loss": 2.9633638858795166,
      "step": 220
    },
    {
      "epoch": 0.022609111879458806,
      "grad_norm": 0.18683850765228271,
      "learning_rate": 0.0009364406779661016,
      "loss": 2.7496559619903564,
      "step": 222
    },
    {
      "epoch": 0.022812797572066543,
      "grad_norm": 0.20445257425308228,
      "learning_rate": 0.0009449152542372882,
      "loss": 2.748857021331787,
      "step": 224
    },
    {
      "epoch": 0.02301648326467428,
      "grad_norm": 0.16937677562236786,
      "learning_rate": 0.0009533898305084746,
      "loss": 2.7546653747558594,
      "step": 226
    },
    {
      "epoch": 0.023220168957282017,
      "grad_norm": 0.19685238599777222,
      "learning_rate": 0.000961864406779661,
      "loss": 2.786573648452759,
      "step": 228
    },
    {
      "epoch": 0.023423854649889754,
      "grad_norm": 0.1848354935646057,
      "learning_rate": 0.0009703389830508475,
      "loss": 2.715834379196167,
      "step": 230
    },
    {
      "epoch": 0.02362754034249749,
      "grad_norm": 0.17445173859596252,
      "learning_rate": 0.0009788135593220338,
      "loss": 2.732861280441284,
      "step": 232
    },
    {
      "epoch": 0.023831226035105228,
      "grad_norm": 0.22051791846752167,
      "learning_rate": 0.0009872881355932205,
      "loss": 2.7958149909973145,
      "step": 234
    },
    {
      "epoch": 0.024034911727712965,
      "grad_norm": 0.1891220510005951,
      "learning_rate": 0.000995762711864407,
      "loss": 2.7374048233032227,
      "step": 236
    },
    {
      "epoch": 0.024238597420320702,
      "grad_norm": 0.20292770862579346,
      "learning_rate": 0.001,
      "loss": 2.7357590198516846,
      "step": 238
    },
    {
      "epoch": 0.02444228311292844,
      "grad_norm": 0.2155386209487915,
      "learning_rate": 0.001,
      "loss": 2.700511932373047,
      "step": 240
    },
    {
      "epoch": 0.024645968805536177,
      "grad_norm": 0.2047884613275528,
      "learning_rate": 0.001,
      "loss": 2.77866792678833,
      "step": 242
    },
    {
      "epoch": 0.024849654498143914,
      "grad_norm": 0.20039160549640656,
      "learning_rate": 0.001,
      "loss": 2.701127529144287,
      "step": 244
    },
    {
      "epoch": 0.02505334019075165,
      "grad_norm": 0.19082506000995636,
      "learning_rate": 0.001,
      "loss": 2.734724521636963,
      "step": 246
    },
    {
      "epoch": 0.025257025883359388,
      "grad_norm": 0.19759756326675415,
      "learning_rate": 0.001,
      "loss": 2.7374839782714844,
      "step": 248
    },
    {
      "epoch": 0.025460711575967125,
      "grad_norm": 0.18879933655261993,
      "learning_rate": 0.001,
      "loss": 2.8644447326660156,
      "step": 250
    },
    {
      "epoch": 0.025664397268574862,
      "grad_norm": 0.17626875638961792,
      "learning_rate": 0.001,
      "loss": 2.6527795791625977,
      "step": 252
    },
    {
      "epoch": 0.0258680829611826,
      "grad_norm": 0.22223737835884094,
      "learning_rate": 0.001,
      "loss": 2.6607208251953125,
      "step": 254
    },
    {
      "epoch": 0.026071768653790336,
      "grad_norm": 0.16813570261001587,
      "learning_rate": 0.001,
      "loss": 2.6267848014831543,
      "step": 256
    },
    {
      "epoch": 0.026275454346398074,
      "grad_norm": 0.21426071226596832,
      "learning_rate": 0.001,
      "loss": 2.765974998474121,
      "step": 258
    },
    {
      "epoch": 0.02647914003900581,
      "grad_norm": 0.2328093945980072,
      "learning_rate": 0.001,
      "loss": 2.731281280517578,
      "step": 260
    },
    {
      "epoch": 0.026682825731613548,
      "grad_norm": 0.20797553658485413,
      "learning_rate": 0.001,
      "loss": 2.702108383178711,
      "step": 262
    },
    {
      "epoch": 0.026886511424221285,
      "grad_norm": 0.1850983202457428,
      "learning_rate": 0.001,
      "loss": 2.6462900638580322,
      "step": 264
    },
    {
      "epoch": 0.027090197116829022,
      "grad_norm": 0.20260952413082123,
      "learning_rate": 0.001,
      "loss": 2.583885669708252,
      "step": 266
    },
    {
      "epoch": 0.02729388280943676,
      "grad_norm": 0.21707405149936676,
      "learning_rate": 0.001,
      "loss": 2.757481575012207,
      "step": 268
    },
    {
      "epoch": 0.027497568502044496,
      "grad_norm": 0.2196391522884369,
      "learning_rate": 0.001,
      "loss": 2.7673392295837402,
      "step": 270
    },
    {
      "epoch": 0.027701254194652233,
      "grad_norm": 0.1920921504497528,
      "learning_rate": 0.001,
      "loss": 2.684785842895508,
      "step": 272
    },
    {
      "epoch": 0.02790493988725997,
      "grad_norm": 0.21097669005393982,
      "learning_rate": 0.001,
      "loss": 2.859036684036255,
      "step": 274
    },
    {
      "epoch": 0.028108625579867708,
      "grad_norm": 0.18593129515647888,
      "learning_rate": 0.001,
      "loss": 2.72040057182312,
      "step": 276
    },
    {
      "epoch": 0.028312311272475445,
      "grad_norm": 0.18707594275474548,
      "learning_rate": 0.001,
      "loss": 2.5428624153137207,
      "step": 278
    },
    {
      "epoch": 0.028515996965083182,
      "grad_norm": 0.20339910686016083,
      "learning_rate": 0.001,
      "loss": 2.7140541076660156,
      "step": 280
    },
    {
      "epoch": 0.028719682657690915,
      "grad_norm": 0.16796834766864777,
      "learning_rate": 0.001,
      "loss": 2.654017686843872,
      "step": 282
    },
    {
      "epoch": 0.028923368350298653,
      "grad_norm": 0.1726291924715042,
      "learning_rate": 0.001,
      "loss": 2.5497474670410156,
      "step": 284
    },
    {
      "epoch": 0.02912705404290639,
      "grad_norm": 0.18665650486946106,
      "learning_rate": 0.001,
      "loss": 2.586742877960205,
      "step": 286
    },
    {
      "epoch": 0.029330739735514127,
      "grad_norm": 0.17324943840503693,
      "learning_rate": 0.001,
      "loss": 2.7045912742614746,
      "step": 288
    },
    {
      "epoch": 0.029534425428121864,
      "grad_norm": 0.14653468132019043,
      "learning_rate": 0.001,
      "loss": 2.5070672035217285,
      "step": 290
    },
    {
      "epoch": 0.0297381111207296,
      "grad_norm": 0.16369901597499847,
      "learning_rate": 0.001,
      "loss": 2.642507791519165,
      "step": 292
    },
    {
      "epoch": 0.029941796813337338,
      "grad_norm": 0.20849809050559998,
      "learning_rate": 0.001,
      "loss": 2.6750869750976562,
      "step": 294
    },
    {
      "epoch": 0.030145482505945075,
      "grad_norm": 0.19255299866199493,
      "learning_rate": 0.001,
      "loss": 2.610416889190674,
      "step": 296
    },
    {
      "epoch": 0.030349168198552812,
      "grad_norm": 0.16760413348674774,
      "learning_rate": 0.001,
      "loss": 2.559685468673706,
      "step": 298
    },
    {
      "epoch": 0.03055285389116055,
      "grad_norm": 0.18125088512897491,
      "learning_rate": 0.001,
      "loss": 2.687553882598877,
      "step": 300
    },
    {
      "epoch": 0.030756539583768287,
      "grad_norm": 0.17401932179927826,
      "learning_rate": 0.001,
      "loss": 2.6689815521240234,
      "step": 302
    },
    {
      "epoch": 0.030960225276376024,
      "grad_norm": 0.20022936165332794,
      "learning_rate": 0.001,
      "loss": 2.6125288009643555,
      "step": 304
    },
    {
      "epoch": 0.03116391096898376,
      "grad_norm": 0.18032217025756836,
      "learning_rate": 0.001,
      "loss": 2.6641383171081543,
      "step": 306
    },
    {
      "epoch": 0.0313675966615915,
      "grad_norm": 0.20606161653995514,
      "learning_rate": 0.001,
      "loss": 2.611727237701416,
      "step": 308
    },
    {
      "epoch": 0.03157128235419924,
      "grad_norm": 0.18587076663970947,
      "learning_rate": 0.001,
      "loss": 2.569898843765259,
      "step": 310
    },
    {
      "epoch": 0.031774968046806976,
      "grad_norm": 0.17147186398506165,
      "learning_rate": 0.001,
      "loss": 2.617847204208374,
      "step": 312
    },
    {
      "epoch": 0.031978653739414706,
      "grad_norm": 0.16324499249458313,
      "learning_rate": 0.001,
      "loss": 2.4753494262695312,
      "step": 314
    },
    {
      "epoch": 0.03218233943202244,
      "grad_norm": 0.16740012168884277,
      "learning_rate": 0.001,
      "loss": 2.444758653640747,
      "step": 316
    },
    {
      "epoch": 0.03238602512463018,
      "grad_norm": 0.18180949985980988,
      "learning_rate": 0.001,
      "loss": 2.5795812606811523,
      "step": 318
    },
    {
      "epoch": 0.03258971081723792,
      "grad_norm": 0.23419760167598724,
      "learning_rate": 0.001,
      "loss": 2.552910566329956,
      "step": 320
    },
    {
      "epoch": 0.032793396509845654,
      "grad_norm": 0.1857665330171585,
      "learning_rate": 0.001,
      "loss": 2.615581512451172,
      "step": 322
    },
    {
      "epoch": 0.03299708220245339,
      "grad_norm": 0.17672771215438843,
      "learning_rate": 0.001,
      "loss": 2.569040060043335,
      "step": 324
    },
    {
      "epoch": 0.03320076789506113,
      "grad_norm": 0.18069970607757568,
      "learning_rate": 0.001,
      "loss": 2.5750274658203125,
      "step": 326
    },
    {
      "epoch": 0.033404453587668866,
      "grad_norm": 0.20245902240276337,
      "learning_rate": 0.001,
      "loss": 2.5230045318603516,
      "step": 328
    },
    {
      "epoch": 0.0336081392802766,
      "grad_norm": 0.17096543312072754,
      "learning_rate": 0.001,
      "loss": 2.5007877349853516,
      "step": 330
    },
    {
      "epoch": 0.03381182497288434,
      "grad_norm": 0.25001198053359985,
      "learning_rate": 0.001,
      "loss": 2.5213730335235596,
      "step": 332
    },
    {
      "epoch": 0.03401551066549208,
      "grad_norm": 0.17875534296035767,
      "learning_rate": 0.001,
      "loss": 2.61470890045166,
      "step": 334
    },
    {
      "epoch": 0.034219196358099814,
      "grad_norm": 0.17580968141555786,
      "learning_rate": 0.001,
      "loss": 2.6579861640930176,
      "step": 336
    },
    {
      "epoch": 0.03442288205070755,
      "grad_norm": 0.18410705029964447,
      "learning_rate": 0.001,
      "loss": 2.612981081008911,
      "step": 338
    },
    {
      "epoch": 0.03462656774331529,
      "grad_norm": 0.19346310198307037,
      "learning_rate": 0.001,
      "loss": 2.612034559249878,
      "step": 340
    },
    {
      "epoch": 0.034830253435923025,
      "grad_norm": 0.173202246427536,
      "learning_rate": 0.001,
      "loss": 2.4768123626708984,
      "step": 342
    },
    {
      "epoch": 0.03503393912853076,
      "grad_norm": 0.19127140939235687,
      "learning_rate": 0.001,
      "loss": 2.560167074203491,
      "step": 344
    },
    {
      "epoch": 0.0352376248211385,
      "grad_norm": 0.15092049539089203,
      "learning_rate": 0.001,
      "loss": 2.436901330947876,
      "step": 346
    },
    {
      "epoch": 0.03544131051374624,
      "grad_norm": 0.1530880481004715,
      "learning_rate": 0.001,
      "loss": 2.469222068786621,
      "step": 348
    },
    {
      "epoch": 0.035644996206353974,
      "grad_norm": 0.17987480759620667,
      "learning_rate": 0.001,
      "loss": 2.6302077770233154,
      "step": 350
    },
    {
      "epoch": 0.03584868189896171,
      "grad_norm": 0.16851210594177246,
      "learning_rate": 0.001,
      "loss": 2.496203660964966,
      "step": 352
    },
    {
      "epoch": 0.03605236759156945,
      "grad_norm": 0.1411920189857483,
      "learning_rate": 0.001,
      "loss": 2.591665029525757,
      "step": 354
    },
    {
      "epoch": 0.036256053284177185,
      "grad_norm": 0.19852329790592194,
      "learning_rate": 0.001,
      "loss": 2.5722689628601074,
      "step": 356
    },
    {
      "epoch": 0.03645973897678492,
      "grad_norm": 0.1617630273103714,
      "learning_rate": 0.001,
      "loss": 2.5056533813476562,
      "step": 358
    },
    {
      "epoch": 0.03666342466939266,
      "grad_norm": 0.15730296075344086,
      "learning_rate": 0.001,
      "loss": 2.492058753967285,
      "step": 360
    },
    {
      "epoch": 0.036867110362000396,
      "grad_norm": 0.16067196428775787,
      "learning_rate": 0.001,
      "loss": 2.5618321895599365,
      "step": 362
    },
    {
      "epoch": 0.037070796054608134,
      "grad_norm": 0.1827927529811859,
      "learning_rate": 0.001,
      "loss": 2.560427188873291,
      "step": 364
    },
    {
      "epoch": 0.03727448174721587,
      "grad_norm": 0.14876431226730347,
      "learning_rate": 0.001,
      "loss": 2.620274543762207,
      "step": 366
    },
    {
      "epoch": 0.03747816743982361,
      "grad_norm": 0.17947666347026825,
      "learning_rate": 0.001,
      "loss": 2.593306541442871,
      "step": 368
    },
    {
      "epoch": 0.037681853132431345,
      "grad_norm": 0.14291419088840485,
      "learning_rate": 0.001,
      "loss": 2.481203556060791,
      "step": 370
    },
    {
      "epoch": 0.03788553882503908,
      "grad_norm": 0.1438884139060974,
      "learning_rate": 0.001,
      "loss": 2.610198497772217,
      "step": 372
    },
    {
      "epoch": 0.03808922451764682,
      "grad_norm": 0.1857542246580124,
      "learning_rate": 0.001,
      "loss": 2.651867628097534,
      "step": 374
    },
    {
      "epoch": 0.038292910210254556,
      "grad_norm": 0.1620441973209381,
      "learning_rate": 0.001,
      "loss": 2.520944356918335,
      "step": 376
    },
    {
      "epoch": 0.03849659590286229,
      "grad_norm": 0.1617012917995453,
      "learning_rate": 0.001,
      "loss": 2.5907726287841797,
      "step": 378
    },
    {
      "epoch": 0.03870028159547003,
      "grad_norm": 0.14825578033924103,
      "learning_rate": 0.001,
      "loss": 2.5959296226501465,
      "step": 380
    },
    {
      "epoch": 0.03890396728807777,
      "grad_norm": 0.15478894114494324,
      "learning_rate": 0.001,
      "loss": 2.4500818252563477,
      "step": 382
    },
    {
      "epoch": 0.039107652980685505,
      "grad_norm": 0.16744601726531982,
      "learning_rate": 0.001,
      "loss": 2.464383125305176,
      "step": 384
    },
    {
      "epoch": 0.03931133867329324,
      "grad_norm": 0.18705719709396362,
      "learning_rate": 0.001,
      "loss": 2.6304311752319336,
      "step": 386
    },
    {
      "epoch": 0.03951502436590098,
      "grad_norm": 0.17762768268585205,
      "learning_rate": 0.001,
      "loss": 2.555314540863037,
      "step": 388
    },
    {
      "epoch": 0.039718710058508716,
      "grad_norm": 0.16939027607440948,
      "learning_rate": 0.001,
      "loss": 2.366931200027466,
      "step": 390
    },
    {
      "epoch": 0.03992239575111645,
      "grad_norm": 0.14292941987514496,
      "learning_rate": 0.001,
      "loss": 2.393606424331665,
      "step": 392
    },
    {
      "epoch": 0.04012608144372419,
      "grad_norm": 0.20613864064216614,
      "learning_rate": 0.001,
      "loss": 2.5478148460388184,
      "step": 394
    },
    {
      "epoch": 0.04032976713633193,
      "grad_norm": 0.1518314629793167,
      "learning_rate": 0.001,
      "loss": 2.541569709777832,
      "step": 396
    },
    {
      "epoch": 0.040533452828939665,
      "grad_norm": 0.1368105411529541,
      "learning_rate": 0.001,
      "loss": 2.4851884841918945,
      "step": 398
    },
    {
      "epoch": 0.0407371385215474,
      "grad_norm": 0.1745341718196869,
      "learning_rate": 0.001,
      "loss": 2.5417652130126953,
      "step": 400
    },
    {
      "epoch": 0.04094082421415514,
      "grad_norm": 0.14014804363250732,
      "learning_rate": 0.001,
      "loss": 2.5273385047912598,
      "step": 402
    },
    {
      "epoch": 0.041144509906762876,
      "grad_norm": 0.15301305055618286,
      "learning_rate": 0.001,
      "loss": 2.4864182472229004,
      "step": 404
    },
    {
      "epoch": 0.04134819559937061,
      "grad_norm": 0.14361804723739624,
      "learning_rate": 0.001,
      "loss": 2.375804901123047,
      "step": 406
    },
    {
      "epoch": 0.04155188129197835,
      "grad_norm": 0.13209913671016693,
      "learning_rate": 0.001,
      "loss": 2.48667573928833,
      "step": 408
    },
    {
      "epoch": 0.04175556698458609,
      "grad_norm": 0.16684380173683167,
      "learning_rate": 0.001,
      "loss": 2.4732260704040527,
      "step": 410
    },
    {
      "epoch": 0.041959252677193824,
      "grad_norm": 0.17294998466968536,
      "learning_rate": 0.001,
      "loss": 2.5970282554626465,
      "step": 412
    },
    {
      "epoch": 0.04216293836980156,
      "grad_norm": 0.13331954181194305,
      "learning_rate": 0.001,
      "loss": 2.447071075439453,
      "step": 414
    },
    {
      "epoch": 0.0423666240624093,
      "grad_norm": 0.15602345764636993,
      "learning_rate": 0.001,
      "loss": 2.5255327224731445,
      "step": 416
    },
    {
      "epoch": 0.042570309755017036,
      "grad_norm": 0.14769864082336426,
      "learning_rate": 0.001,
      "loss": 2.6446990966796875,
      "step": 418
    },
    {
      "epoch": 0.04277399544762477,
      "grad_norm": 0.16930250823497772,
      "learning_rate": 0.001,
      "loss": 2.5860612392425537,
      "step": 420
    },
    {
      "epoch": 0.04297768114023251,
      "grad_norm": 0.14852115511894226,
      "learning_rate": 0.001,
      "loss": 2.482539415359497,
      "step": 422
    },
    {
      "epoch": 0.04318136683284025,
      "grad_norm": 0.16319233179092407,
      "learning_rate": 0.001,
      "loss": 2.4694812297821045,
      "step": 424
    },
    {
      "epoch": 0.043385052525447984,
      "grad_norm": 0.14504896104335785,
      "learning_rate": 0.001,
      "loss": 2.565683364868164,
      "step": 426
    },
    {
      "epoch": 0.04358873821805572,
      "grad_norm": 0.18446308374404907,
      "learning_rate": 0.001,
      "loss": 2.4253687858581543,
      "step": 428
    },
    {
      "epoch": 0.04379242391066346,
      "grad_norm": 0.1722123622894287,
      "learning_rate": 0.001,
      "loss": 2.502716064453125,
      "step": 430
    },
    {
      "epoch": 0.043996109603271195,
      "grad_norm": 0.17189428210258484,
      "learning_rate": 0.001,
      "loss": 2.503190040588379,
      "step": 432
    },
    {
      "epoch": 0.04419979529587893,
      "grad_norm": 0.13091914355754852,
      "learning_rate": 0.001,
      "loss": 2.4718127250671387,
      "step": 434
    },
    {
      "epoch": 0.04440348098848667,
      "grad_norm": 0.14693881571292877,
      "learning_rate": 0.001,
      "loss": 2.544137477874756,
      "step": 436
    },
    {
      "epoch": 0.0446071666810944,
      "grad_norm": 0.20538637042045593,
      "learning_rate": 0.001,
      "loss": 2.4231510162353516,
      "step": 438
    },
    {
      "epoch": 0.04481085237370214,
      "grad_norm": 0.22402171790599823,
      "learning_rate": 0.001,
      "loss": 2.5782904624938965,
      "step": 440
    },
    {
      "epoch": 0.045014538066309874,
      "grad_norm": 0.1384379267692566,
      "learning_rate": 0.001,
      "loss": 2.4766011238098145,
      "step": 442
    },
    {
      "epoch": 0.04521822375891761,
      "grad_norm": 0.1364443004131317,
      "learning_rate": 0.001,
      "loss": 2.502331256866455,
      "step": 444
    },
    {
      "epoch": 0.04542190945152535,
      "grad_norm": 0.14493145048618317,
      "learning_rate": 0.001,
      "loss": 2.4864392280578613,
      "step": 446
    },
    {
      "epoch": 0.045625595144133085,
      "grad_norm": 0.13503649830818176,
      "learning_rate": 0.001,
      "loss": 2.486527919769287,
      "step": 448
    },
    {
      "epoch": 0.04582928083674082,
      "grad_norm": 0.14300934970378876,
      "learning_rate": 0.001,
      "loss": 2.335270404815674,
      "step": 450
    },
    {
      "epoch": 0.04603296652934856,
      "grad_norm": 0.12717053294181824,
      "learning_rate": 0.001,
      "loss": 2.5808048248291016,
      "step": 452
    },
    {
      "epoch": 0.0462366522219563,
      "grad_norm": 0.14120681583881378,
      "learning_rate": 0.001,
      "loss": 2.570100784301758,
      "step": 454
    },
    {
      "epoch": 0.046440337914564034,
      "grad_norm": 0.17605650424957275,
      "learning_rate": 0.001,
      "loss": 2.311452865600586,
      "step": 456
    },
    {
      "epoch": 0.04664402360717177,
      "grad_norm": 0.20320197939872742,
      "learning_rate": 0.001,
      "loss": 2.486205577850342,
      "step": 458
    },
    {
      "epoch": 0.04684770929977951,
      "grad_norm": 0.13210153579711914,
      "learning_rate": 0.001,
      "loss": 2.298549175262451,
      "step": 460
    },
    {
      "epoch": 0.047051394992387245,
      "grad_norm": 0.14128974080085754,
      "learning_rate": 0.001,
      "loss": 2.401202440261841,
      "step": 462
    },
    {
      "epoch": 0.04725508068499498,
      "grad_norm": 0.12276806682348251,
      "learning_rate": 0.001,
      "loss": 2.325411558151245,
      "step": 464
    },
    {
      "epoch": 0.04745876637760272,
      "grad_norm": 0.1508936583995819,
      "learning_rate": 0.001,
      "loss": 2.426778793334961,
      "step": 466
    },
    {
      "epoch": 0.047662452070210456,
      "grad_norm": 0.1815779209136963,
      "learning_rate": 0.001,
      "loss": 2.590831995010376,
      "step": 468
    },
    {
      "epoch": 0.047866137762818194,
      "grad_norm": 0.1449596881866455,
      "learning_rate": 0.001,
      "loss": 2.492798328399658,
      "step": 470
    },
    {
      "epoch": 0.04806982345542593,
      "grad_norm": 0.1448836475610733,
      "learning_rate": 0.001,
      "loss": 2.4654526710510254,
      "step": 472
    },
    {
      "epoch": 0.04827350914803367,
      "grad_norm": 0.181374654173851,
      "learning_rate": 0.001,
      "loss": 2.3260984420776367,
      "step": 474
    },
    {
      "epoch": 0.048477194840641405,
      "grad_norm": 0.17633110284805298,
      "learning_rate": 0.001,
      "loss": 2.442765712738037,
      "step": 476
    },
    {
      "epoch": 0.04868088053324914,
      "grad_norm": 0.12500132620334625,
      "learning_rate": 0.001,
      "loss": 2.425706386566162,
      "step": 478
    },
    {
      "epoch": 0.04888456622585688,
      "grad_norm": 0.20653249323368073,
      "learning_rate": 0.001,
      "loss": 2.525418281555176,
      "step": 480
    },
    {
      "epoch": 0.049088251918464616,
      "grad_norm": 0.16679427027702332,
      "learning_rate": 0.001,
      "loss": 2.500777244567871,
      "step": 482
    },
    {
      "epoch": 0.04929193761107235,
      "grad_norm": 0.17499476671218872,
      "learning_rate": 0.001,
      "loss": 2.363821029663086,
      "step": 484
    },
    {
      "epoch": 0.04949562330368009,
      "grad_norm": 0.1404733955860138,
      "learning_rate": 0.001,
      "loss": 2.5179927349090576,
      "step": 486
    },
    {
      "epoch": 0.04969930899628783,
      "grad_norm": 0.21139004826545715,
      "learning_rate": 0.001,
      "loss": 2.444701671600342,
      "step": 488
    },
    {
      "epoch": 0.049902994688895565,
      "grad_norm": 0.18328775465488434,
      "learning_rate": 0.001,
      "loss": 2.479682445526123,
      "step": 490
    },
    {
      "epoch": 0.0501066803815033,
      "grad_norm": 0.14166665077209473,
      "learning_rate": 0.001,
      "loss": 2.391298294067383,
      "step": 492
    },
    {
      "epoch": 0.05031036607411104,
      "grad_norm": 0.16102235019207,
      "learning_rate": 0.001,
      "loss": 2.464491367340088,
      "step": 494
    },
    {
      "epoch": 0.050514051766718776,
      "grad_norm": 0.16111540794372559,
      "learning_rate": 0.001,
      "loss": 2.532060384750366,
      "step": 496
    },
    {
      "epoch": 0.05071773745932651,
      "grad_norm": 0.19084085524082184,
      "learning_rate": 0.001,
      "loss": 2.4772276878356934,
      "step": 498
    },
    {
      "epoch": 0.05092142315193425,
      "grad_norm": 0.13296407461166382,
      "learning_rate": 0.001,
      "loss": 2.480391025543213,
      "step": 500
    },
    {
      "epoch": 0.05112510884454199,
      "grad_norm": 0.1665947139263153,
      "learning_rate": 0.001,
      "loss": 2.3465399742126465,
      "step": 502
    },
    {
      "epoch": 0.051328794537149725,
      "grad_norm": 0.13415436446666718,
      "learning_rate": 0.001,
      "loss": 2.3161702156066895,
      "step": 504
    },
    {
      "epoch": 0.05153248022975746,
      "grad_norm": 0.2139132171869278,
      "learning_rate": 0.001,
      "loss": 2.527242660522461,
      "step": 506
    },
    {
      "epoch": 0.0517361659223652,
      "grad_norm": 0.15039341151714325,
      "learning_rate": 0.001,
      "loss": 2.4237122535705566,
      "step": 508
    },
    {
      "epoch": 0.051939851614972936,
      "grad_norm": 0.13356785476207733,
      "learning_rate": 0.001,
      "loss": 2.526890754699707,
      "step": 510
    },
    {
      "epoch": 0.05214353730758067,
      "grad_norm": 0.14575345814228058,
      "learning_rate": 0.001,
      "loss": 2.351581573486328,
      "step": 512
    },
    {
      "epoch": 0.05234722300018841,
      "grad_norm": 0.17870698869228363,
      "learning_rate": 0.001,
      "loss": 2.2642135620117188,
      "step": 514
    },
    {
      "epoch": 0.05255090869279615,
      "grad_norm": 0.1305026113986969,
      "learning_rate": 0.001,
      "loss": 2.3467187881469727,
      "step": 516
    },
    {
      "epoch": 0.052754594385403884,
      "grad_norm": 0.1529131978750229,
      "learning_rate": 0.001,
      "loss": 2.4729580879211426,
      "step": 518
    },
    {
      "epoch": 0.05295828007801162,
      "grad_norm": 0.12487910687923431,
      "learning_rate": 0.001,
      "loss": 2.414607524871826,
      "step": 520
    },
    {
      "epoch": 0.05316196577061936,
      "grad_norm": 0.21933351457118988,
      "learning_rate": 0.001,
      "loss": 2.3251729011535645,
      "step": 522
    },
    {
      "epoch": 0.053365651463227096,
      "grad_norm": 0.1373833566904068,
      "learning_rate": 0.001,
      "loss": 2.494901657104492,
      "step": 524
    },
    {
      "epoch": 0.05356933715583483,
      "grad_norm": 0.13093040883541107,
      "learning_rate": 0.001,
      "loss": 2.433558225631714,
      "step": 526
    },
    {
      "epoch": 0.05377302284844257,
      "grad_norm": 0.1696823388338089,
      "learning_rate": 0.001,
      "loss": 2.408080577850342,
      "step": 528
    },
    {
      "epoch": 0.05397670854105031,
      "grad_norm": 0.20331719517707825,
      "learning_rate": 0.001,
      "loss": 2.349245071411133,
      "step": 530
    },
    {
      "epoch": 0.054180394233658044,
      "grad_norm": 0.21336638927459717,
      "learning_rate": 0.001,
      "loss": 2.5095975399017334,
      "step": 532
    },
    {
      "epoch": 0.05438407992626578,
      "grad_norm": 0.1489047110080719,
      "learning_rate": 0.001,
      "loss": 2.2968223094940186,
      "step": 534
    },
    {
      "epoch": 0.05458776561887352,
      "grad_norm": 0.15516585111618042,
      "learning_rate": 0.001,
      "loss": 2.3000330924987793,
      "step": 536
    },
    {
      "epoch": 0.054791451311481255,
      "grad_norm": 0.15010711550712585,
      "learning_rate": 0.001,
      "loss": 2.447078227996826,
      "step": 538
    },
    {
      "epoch": 0.05499513700408899,
      "grad_norm": 0.24178047478199005,
      "learning_rate": 0.001,
      "loss": 2.5308611392974854,
      "step": 540
    },
    {
      "epoch": 0.05519882269669673,
      "grad_norm": 0.1438474804162979,
      "learning_rate": 0.001,
      "loss": 2.370006561279297,
      "step": 542
    },
    {
      "epoch": 0.05540250838930447,
      "grad_norm": 0.19887539744377136,
      "learning_rate": 0.001,
      "loss": 2.304811954498291,
      "step": 544
    },
    {
      "epoch": 0.055606194081912204,
      "grad_norm": 0.1604827493429184,
      "learning_rate": 0.001,
      "loss": 2.4291768074035645,
      "step": 546
    },
    {
      "epoch": 0.05580987977451994,
      "grad_norm": 0.12520171701908112,
      "learning_rate": 0.001,
      "loss": 2.3078784942626953,
      "step": 548
    },
    {
      "epoch": 0.05601356546712768,
      "grad_norm": 0.1354176104068756,
      "learning_rate": 0.001,
      "loss": 2.4032387733459473,
      "step": 550
    },
    {
      "epoch": 0.056217251159735415,
      "grad_norm": 0.2157740294933319,
      "learning_rate": 0.001,
      "loss": 2.432861328125,
      "step": 552
    },
    {
      "epoch": 0.05642093685234315,
      "grad_norm": 0.13985057175159454,
      "learning_rate": 0.001,
      "loss": 2.4224905967712402,
      "step": 554
    },
    {
      "epoch": 0.05662462254495089,
      "grad_norm": 0.11958431452512741,
      "learning_rate": 0.001,
      "loss": 2.53584623336792,
      "step": 556
    },
    {
      "epoch": 0.05682830823755863,
      "grad_norm": 0.16192640364170074,
      "learning_rate": 0.001,
      "loss": 2.4648146629333496,
      "step": 558
    },
    {
      "epoch": 0.057031993930166364,
      "grad_norm": 0.12312827259302139,
      "learning_rate": 0.001,
      "loss": 2.394575357437134,
      "step": 560
    },
    {
      "epoch": 0.057235679622774094,
      "grad_norm": 0.10764305293560028,
      "learning_rate": 0.001,
      "loss": 2.3491954803466797,
      "step": 562
    },
    {
      "epoch": 0.05743936531538183,
      "grad_norm": 0.21056392788887024,
      "learning_rate": 0.001,
      "loss": 2.4286513328552246,
      "step": 564
    },
    {
      "epoch": 0.05764305100798957,
      "grad_norm": 0.16411584615707397,
      "learning_rate": 0.001,
      "loss": 2.399359941482544,
      "step": 566
    },
    {
      "epoch": 0.057846736700597305,
      "grad_norm": 0.22644242644309998,
      "learning_rate": 0.001,
      "loss": 2.387659788131714,
      "step": 568
    },
    {
      "epoch": 0.05805042239320504,
      "grad_norm": 0.17329514026641846,
      "learning_rate": 0.001,
      "loss": 2.3412699699401855,
      "step": 570
    },
    {
      "epoch": 0.05825410808581278,
      "grad_norm": 0.14542576670646667,
      "learning_rate": 0.001,
      "loss": 2.585819721221924,
      "step": 572
    },
    {
      "epoch": 0.058457793778420516,
      "grad_norm": 0.1785878986120224,
      "learning_rate": 0.001,
      "loss": 2.3262405395507812,
      "step": 574
    },
    {
      "epoch": 0.058661479471028254,
      "grad_norm": 0.18529710173606873,
      "learning_rate": 0.001,
      "loss": 2.4284236431121826,
      "step": 576
    },
    {
      "epoch": 0.05886516516363599,
      "grad_norm": 0.1265268474817276,
      "learning_rate": 0.001,
      "loss": 2.3213372230529785,
      "step": 578
    },
    {
      "epoch": 0.05906885085624373,
      "grad_norm": 0.15799376368522644,
      "learning_rate": 0.001,
      "loss": 2.4859187602996826,
      "step": 580
    },
    {
      "epoch": 0.059272536548851465,
      "grad_norm": 0.1474652886390686,
      "learning_rate": 0.001,
      "loss": 2.412809371948242,
      "step": 582
    },
    {
      "epoch": 0.0594762222414592,
      "grad_norm": 0.14701229333877563,
      "learning_rate": 0.001,
      "loss": 2.4214015007019043,
      "step": 584
    },
    {
      "epoch": 0.05967990793406694,
      "grad_norm": 0.15289834141731262,
      "learning_rate": 0.001,
      "loss": 2.3893816471099854,
      "step": 586
    },
    {
      "epoch": 0.059883593626674676,
      "grad_norm": 0.13821718096733093,
      "learning_rate": 0.001,
      "loss": 2.222414493560791,
      "step": 588
    },
    {
      "epoch": 0.06008727931928241,
      "grad_norm": 0.17063651978969574,
      "learning_rate": 0.001,
      "loss": 2.4244956970214844,
      "step": 590
    },
    {
      "epoch": 0.06029096501189015,
      "grad_norm": 0.14678941667079926,
      "learning_rate": 0.001,
      "loss": 2.386418104171753,
      "step": 592
    },
    {
      "epoch": 0.06049465070449789,
      "grad_norm": 0.15014195442199707,
      "learning_rate": 0.001,
      "loss": 2.452552556991577,
      "step": 594
    },
    {
      "epoch": 0.060698336397105625,
      "grad_norm": 0.15659426152706146,
      "learning_rate": 0.001,
      "loss": 2.478659152984619,
      "step": 596
    },
    {
      "epoch": 0.06090202208971336,
      "grad_norm": 0.21840929985046387,
      "learning_rate": 0.001,
      "loss": 2.5052907466888428,
      "step": 598
    },
    {
      "epoch": 0.0611057077823211,
      "grad_norm": 0.15946412086486816,
      "learning_rate": 0.001,
      "loss": 2.4904353618621826,
      "step": 600
    },
    {
      "epoch": 0.061309393474928836,
      "grad_norm": 0.24483181536197662,
      "learning_rate": 0.001,
      "loss": 2.4524784088134766,
      "step": 602
    },
    {
      "epoch": 0.06151307916753657,
      "grad_norm": 0.12643180787563324,
      "learning_rate": 0.001,
      "loss": 2.438936710357666,
      "step": 604
    },
    {
      "epoch": 0.06171676486014431,
      "grad_norm": 0.21820737421512604,
      "learning_rate": 0.001,
      "loss": 2.3311002254486084,
      "step": 606
    },
    {
      "epoch": 0.06192045055275205,
      "grad_norm": 0.14396512508392334,
      "learning_rate": 0.001,
      "loss": 2.239259719848633,
      "step": 608
    },
    {
      "epoch": 0.062124136245359785,
      "grad_norm": 0.16121049225330353,
      "learning_rate": 0.001,
      "loss": 2.4931247234344482,
      "step": 610
    },
    {
      "epoch": 0.06232782193796752,
      "grad_norm": 0.11628785729408264,
      "learning_rate": 0.001,
      "loss": 2.3528120517730713,
      "step": 612
    },
    {
      "epoch": 0.06253150763057526,
      "grad_norm": 0.14725114405155182,
      "learning_rate": 0.001,
      "loss": 2.390408992767334,
      "step": 614
    },
    {
      "epoch": 0.062735193323183,
      "grad_norm": 0.15125161409378052,
      "learning_rate": 0.001,
      "loss": 2.490048408508301,
      "step": 616
    },
    {
      "epoch": 0.06293887901579073,
      "grad_norm": 0.18672005832195282,
      "learning_rate": 0.001,
      "loss": 2.3881595134735107,
      "step": 618
    },
    {
      "epoch": 0.06314256470839848,
      "grad_norm": 0.23136253654956818,
      "learning_rate": 0.001,
      "loss": 2.5726869106292725,
      "step": 620
    },
    {
      "epoch": 0.06334625040100621,
      "grad_norm": 0.16482077538967133,
      "learning_rate": 0.001,
      "loss": 2.30220365524292,
      "step": 622
    },
    {
      "epoch": 0.06354993609361395,
      "grad_norm": 0.16276246309280396,
      "learning_rate": 0.001,
      "loss": 2.386591911315918,
      "step": 624
    },
    {
      "epoch": 0.06375362178622168,
      "grad_norm": 0.14680305123329163,
      "learning_rate": 0.001,
      "loss": 2.4549002647399902,
      "step": 626
    },
    {
      "epoch": 0.06395730747882941,
      "grad_norm": 0.1576865017414093,
      "learning_rate": 0.001,
      "loss": 2.5257983207702637,
      "step": 628
    },
    {
      "epoch": 0.06416099317143716,
      "grad_norm": 0.12920598685741425,
      "learning_rate": 0.001,
      "loss": 2.2668979167938232,
      "step": 630
    },
    {
      "epoch": 0.06436467886404489,
      "grad_norm": 0.2532174289226532,
      "learning_rate": 0.001,
      "loss": 2.521669864654541,
      "step": 632
    },
    {
      "epoch": 0.06456836455665263,
      "grad_norm": 0.16507922112941742,
      "learning_rate": 0.001,
      "loss": 2.4533538818359375,
      "step": 634
    },
    {
      "epoch": 0.06477205024926036,
      "grad_norm": 0.21064133942127228,
      "learning_rate": 0.001,
      "loss": 2.3523426055908203,
      "step": 636
    },
    {
      "epoch": 0.0649757359418681,
      "grad_norm": 0.1625257283449173,
      "learning_rate": 0.001,
      "loss": 2.3525571823120117,
      "step": 638
    },
    {
      "epoch": 0.06517942163447583,
      "grad_norm": 0.1415996253490448,
      "learning_rate": 0.001,
      "loss": 2.466803550720215,
      "step": 640
    },
    {
      "epoch": 0.06538310732708358,
      "grad_norm": 0.14383377134799957,
      "learning_rate": 0.001,
      "loss": 2.391633987426758,
      "step": 642
    },
    {
      "epoch": 0.06558679301969131,
      "grad_norm": 0.11341153830289841,
      "learning_rate": 0.001,
      "loss": 2.470705986022949,
      "step": 644
    },
    {
      "epoch": 0.06579047871229905,
      "grad_norm": 0.14039863646030426,
      "learning_rate": 0.001,
      "loss": 2.36773681640625,
      "step": 646
    },
    {
      "epoch": 0.06599416440490678,
      "grad_norm": 0.16056828200817108,
      "learning_rate": 0.001,
      "loss": 2.4659290313720703,
      "step": 648
    },
    {
      "epoch": 0.06619785009751453,
      "grad_norm": 0.16848772764205933,
      "learning_rate": 0.001,
      "loss": 2.3277876377105713,
      "step": 650
    },
    {
      "epoch": 0.06640153579012226,
      "grad_norm": 0.14960509538650513,
      "learning_rate": 0.001,
      "loss": 2.412219524383545,
      "step": 652
    },
    {
      "epoch": 0.06660522148273,
      "grad_norm": 0.171751007437706,
      "learning_rate": 0.001,
      "loss": 2.3823647499084473,
      "step": 654
    },
    {
      "epoch": 0.06680890717533773,
      "grad_norm": 0.13371272385120392,
      "learning_rate": 0.001,
      "loss": 2.3756704330444336,
      "step": 656
    },
    {
      "epoch": 0.06701259286794548,
      "grad_norm": 0.15660886466503143,
      "learning_rate": 0.001,
      "loss": 2.4957213401794434,
      "step": 658
    },
    {
      "epoch": 0.0672162785605532,
      "grad_norm": 0.1245158463716507,
      "learning_rate": 0.001,
      "loss": 2.388338088989258,
      "step": 660
    },
    {
      "epoch": 0.06741996425316095,
      "grad_norm": 0.21906982362270355,
      "learning_rate": 0.001,
      "loss": 2.315791130065918,
      "step": 662
    },
    {
      "epoch": 0.06762364994576868,
      "grad_norm": 0.15058840811252594,
      "learning_rate": 0.001,
      "loss": 2.301079273223877,
      "step": 664
    },
    {
      "epoch": 0.06782733563837642,
      "grad_norm": 0.14141669869422913,
      "learning_rate": 0.001,
      "loss": 2.3011112213134766,
      "step": 666
    },
    {
      "epoch": 0.06803102133098415,
      "grad_norm": 0.15706506371498108,
      "learning_rate": 0.001,
      "loss": 2.427041530609131,
      "step": 668
    },
    {
      "epoch": 0.0682347070235919,
      "grad_norm": 0.12748289108276367,
      "learning_rate": 0.001,
      "loss": 2.4868903160095215,
      "step": 670
    },
    {
      "epoch": 0.06843839271619963,
      "grad_norm": 0.18001021444797516,
      "learning_rate": 0.001,
      "loss": 2.2647552490234375,
      "step": 672
    },
    {
      "epoch": 0.06864207840880737,
      "grad_norm": 0.13664653897285461,
      "learning_rate": 0.001,
      "loss": 2.3283309936523438,
      "step": 674
    },
    {
      "epoch": 0.0688457641014151,
      "grad_norm": 0.17604152858257294,
      "learning_rate": 0.001,
      "loss": 2.44325852394104,
      "step": 676
    },
    {
      "epoch": 0.06904944979402285,
      "grad_norm": 0.14027665555477142,
      "learning_rate": 0.001,
      "loss": 2.187713146209717,
      "step": 678
    },
    {
      "epoch": 0.06925313548663058,
      "grad_norm": 0.15569831430912018,
      "learning_rate": 0.001,
      "loss": 2.32820987701416,
      "step": 680
    },
    {
      "epoch": 0.06945682117923832,
      "grad_norm": 0.15393580496311188,
      "learning_rate": 0.001,
      "loss": 2.4309194087982178,
      "step": 682
    },
    {
      "epoch": 0.06966050687184605,
      "grad_norm": 0.10522426664829254,
      "learning_rate": 0.001,
      "loss": 2.399233341217041,
      "step": 684
    },
    {
      "epoch": 0.0698641925644538,
      "grad_norm": 0.14497016370296478,
      "learning_rate": 0.001,
      "loss": 2.3406460285186768,
      "step": 686
    },
    {
      "epoch": 0.07006787825706152,
      "grad_norm": 0.2184428870677948,
      "learning_rate": 0.001,
      "loss": 2.4863924980163574,
      "step": 688
    },
    {
      "epoch": 0.07027156394966927,
      "grad_norm": 0.13810864090919495,
      "learning_rate": 0.001,
      "loss": 2.3335256576538086,
      "step": 690
    },
    {
      "epoch": 0.070475249642277,
      "grad_norm": 0.20958462357521057,
      "learning_rate": 0.001,
      "loss": 2.313250780105591,
      "step": 692
    },
    {
      "epoch": 0.07067893533488474,
      "grad_norm": 0.14061829447746277,
      "learning_rate": 0.001,
      "loss": 2.371455192565918,
      "step": 694
    },
    {
      "epoch": 0.07088262102749247,
      "grad_norm": 0.1286882609128952,
      "learning_rate": 0.001,
      "loss": 2.5236775875091553,
      "step": 696
    },
    {
      "epoch": 0.07108630672010022,
      "grad_norm": 0.16081297397613525,
      "learning_rate": 0.001,
      "loss": 2.2586677074432373,
      "step": 698
    },
    {
      "epoch": 0.07128999241270795,
      "grad_norm": 0.1672571897506714,
      "learning_rate": 0.001,
      "loss": 2.3151655197143555,
      "step": 700
    },
    {
      "epoch": 0.07149367810531569,
      "grad_norm": 0.21381475031375885,
      "learning_rate": 0.001,
      "loss": 2.4687323570251465,
      "step": 702
    },
    {
      "epoch": 0.07169736379792342,
      "grad_norm": 0.13332071900367737,
      "learning_rate": 0.001,
      "loss": 2.3806028366088867,
      "step": 704
    },
    {
      "epoch": 0.07190104949053117,
      "grad_norm": 0.1734728068113327,
      "learning_rate": 0.001,
      "loss": 2.4498543739318848,
      "step": 706
    },
    {
      "epoch": 0.0721047351831389,
      "grad_norm": 0.1472063511610031,
      "learning_rate": 0.001,
      "loss": 2.3393473625183105,
      "step": 708
    },
    {
      "epoch": 0.07230842087574664,
      "grad_norm": 0.14142705500125885,
      "learning_rate": 0.001,
      "loss": 2.3803839683532715,
      "step": 710
    },
    {
      "epoch": 0.07251210656835437,
      "grad_norm": 0.1276078075170517,
      "learning_rate": 0.001,
      "loss": 2.4736897945404053,
      "step": 712
    },
    {
      "epoch": 0.07271579226096211,
      "grad_norm": 0.12908488512039185,
      "learning_rate": 0.001,
      "loss": 2.4540164470672607,
      "step": 714
    },
    {
      "epoch": 0.07291947795356984,
      "grad_norm": 0.20256392657756805,
      "learning_rate": 0.001,
      "loss": 2.3653182983398438,
      "step": 716
    },
    {
      "epoch": 0.07312316364617759,
      "grad_norm": 0.146226167678833,
      "learning_rate": 0.001,
      "loss": 2.296677350997925,
      "step": 718
    },
    {
      "epoch": 0.07332684933878532,
      "grad_norm": 0.1283259391784668,
      "learning_rate": 0.001,
      "loss": 2.4530141353607178,
      "step": 720
    },
    {
      "epoch": 0.07353053503139306,
      "grad_norm": 0.14761316776275635,
      "learning_rate": 0.001,
      "loss": 2.38979434967041,
      "step": 722
    },
    {
      "epoch": 0.07373422072400079,
      "grad_norm": 0.12775765359401703,
      "learning_rate": 0.001,
      "loss": 2.3454208374023438,
      "step": 724
    },
    {
      "epoch": 0.07393790641660854,
      "grad_norm": 0.1561470925807953,
      "learning_rate": 0.001,
      "loss": 2.3907265663146973,
      "step": 726
    },
    {
      "epoch": 0.07414159210921627,
      "grad_norm": 0.14256912469863892,
      "learning_rate": 0.001,
      "loss": 2.464344024658203,
      "step": 728
    },
    {
      "epoch": 0.07434527780182401,
      "grad_norm": 0.1297033727169037,
      "learning_rate": 0.001,
      "loss": 2.4586777687072754,
      "step": 730
    },
    {
      "epoch": 0.07454896349443174,
      "grad_norm": 0.15857358276844025,
      "learning_rate": 0.001,
      "loss": 2.4187262058258057,
      "step": 732
    },
    {
      "epoch": 0.07475264918703949,
      "grad_norm": 0.1440712958574295,
      "learning_rate": 0.001,
      "loss": 2.3810861110687256,
      "step": 734
    },
    {
      "epoch": 0.07495633487964722,
      "grad_norm": 0.15531261265277863,
      "learning_rate": 0.001,
      "loss": 2.434206485748291,
      "step": 736
    },
    {
      "epoch": 0.07516002057225496,
      "grad_norm": 0.15755966305732727,
      "learning_rate": 0.001,
      "loss": 2.401261806488037,
      "step": 738
    },
    {
      "epoch": 0.07536370626486269,
      "grad_norm": 0.1356690526008606,
      "learning_rate": 0.001,
      "loss": 2.217540979385376,
      "step": 740
    },
    {
      "epoch": 0.07556739195747043,
      "grad_norm": 0.141010120511055,
      "learning_rate": 0.001,
      "loss": 2.3790860176086426,
      "step": 742
    },
    {
      "epoch": 0.07577107765007816,
      "grad_norm": 0.13231390714645386,
      "learning_rate": 0.001,
      "loss": 2.4473717212677,
      "step": 744
    },
    {
      "epoch": 0.07597476334268591,
      "grad_norm": 0.19353404641151428,
      "learning_rate": 0.001,
      "loss": 2.37355375289917,
      "step": 746
    },
    {
      "epoch": 0.07617844903529364,
      "grad_norm": 0.16636745631694794,
      "learning_rate": 0.001,
      "loss": 2.414081573486328,
      "step": 748
    },
    {
      "epoch": 0.07638213472790137,
      "grad_norm": 0.13192428648471832,
      "learning_rate": 0.001,
      "loss": 2.324626922607422,
      "step": 750
    },
    {
      "epoch": 0.07658582042050911,
      "grad_norm": 0.14182987809181213,
      "learning_rate": 0.001,
      "loss": 2.3529624938964844,
      "step": 752
    },
    {
      "epoch": 0.07678950611311684,
      "grad_norm": 0.13347235321998596,
      "learning_rate": 0.001,
      "loss": 2.401195526123047,
      "step": 754
    },
    {
      "epoch": 0.07699319180572459,
      "grad_norm": 0.1635565161705017,
      "learning_rate": 0.001,
      "loss": 2.365734815597534,
      "step": 756
    },
    {
      "epoch": 0.07719687749833232,
      "grad_norm": 0.13783858716487885,
      "learning_rate": 0.001,
      "loss": 2.4385738372802734,
      "step": 758
    },
    {
      "epoch": 0.07740056319094006,
      "grad_norm": 0.12309624254703522,
      "learning_rate": 0.001,
      "loss": 2.252499580383301,
      "step": 760
    },
    {
      "epoch": 0.07760424888354779,
      "grad_norm": 0.11199089884757996,
      "learning_rate": 0.001,
      "loss": 2.3232645988464355,
      "step": 762
    },
    {
      "epoch": 0.07780793457615554,
      "grad_norm": 0.1180214211344719,
      "learning_rate": 0.001,
      "loss": 2.313728094100952,
      "step": 764
    },
    {
      "epoch": 0.07801162026876327,
      "grad_norm": 0.16183336079120636,
      "learning_rate": 0.001,
      "loss": 2.445600986480713,
      "step": 766
    },
    {
      "epoch": 0.07821530596137101,
      "grad_norm": 0.12845246493816376,
      "learning_rate": 0.001,
      "loss": 2.370180130004883,
      "step": 768
    },
    {
      "epoch": 0.07841899165397874,
      "grad_norm": 0.11891917139291763,
      "learning_rate": 0.001,
      "loss": 2.368825912475586,
      "step": 770
    },
    {
      "epoch": 0.07862267734658648,
      "grad_norm": 0.15114830434322357,
      "learning_rate": 0.001,
      "loss": 2.2580504417419434,
      "step": 772
    },
    {
      "epoch": 0.07882636303919421,
      "grad_norm": 0.13465386629104614,
      "learning_rate": 0.001,
      "loss": 2.3601551055908203,
      "step": 774
    },
    {
      "epoch": 0.07903004873180196,
      "grad_norm": 0.14674389362335205,
      "learning_rate": 0.001,
      "loss": 2.3438944816589355,
      "step": 776
    },
    {
      "epoch": 0.07923373442440969,
      "grad_norm": 0.14450907707214355,
      "learning_rate": 0.001,
      "loss": 2.421693801879883,
      "step": 778
    },
    {
      "epoch": 0.07943742011701743,
      "grad_norm": 0.17574162781238556,
      "learning_rate": 0.001,
      "loss": 2.3767805099487305,
      "step": 780
    },
    {
      "epoch": 0.07964110580962516,
      "grad_norm": 0.1861622929573059,
      "learning_rate": 0.001,
      "loss": 2.2643611431121826,
      "step": 782
    },
    {
      "epoch": 0.0798447915022329,
      "grad_norm": 0.13955058157444,
      "learning_rate": 0.001,
      "loss": 2.519105911254883,
      "step": 784
    },
    {
      "epoch": 0.08004847719484064,
      "grad_norm": 0.1529414802789688,
      "learning_rate": 0.001,
      "loss": 2.49379301071167,
      "step": 786
    },
    {
      "epoch": 0.08025216288744838,
      "grad_norm": 0.14161188900470734,
      "learning_rate": 0.001,
      "loss": 2.345775604248047,
      "step": 788
    },
    {
      "epoch": 0.08045584858005611,
      "grad_norm": 0.16425299644470215,
      "learning_rate": 0.001,
      "loss": 2.2722764015197754,
      "step": 790
    },
    {
      "epoch": 0.08065953427266385,
      "grad_norm": 0.14209797978401184,
      "learning_rate": 0.001,
      "loss": 2.29223370552063,
      "step": 792
    },
    {
      "epoch": 0.08086321996527158,
      "grad_norm": 0.17363210022449493,
      "learning_rate": 0.001,
      "loss": 2.3616294860839844,
      "step": 794
    },
    {
      "epoch": 0.08106690565787933,
      "grad_norm": 0.13007752597332,
      "learning_rate": 0.001,
      "loss": 2.290221691131592,
      "step": 796
    },
    {
      "epoch": 0.08127059135048706,
      "grad_norm": 0.12231266498565674,
      "learning_rate": 0.001,
      "loss": 2.28560733795166,
      "step": 798
    },
    {
      "epoch": 0.0814742770430948,
      "grad_norm": 0.12175382673740387,
      "learning_rate": 0.001,
      "loss": 2.2132177352905273,
      "step": 800
    },
    {
      "epoch": 0.08167796273570253,
      "grad_norm": 0.16265690326690674,
      "learning_rate": 0.001,
      "loss": 2.4972333908081055,
      "step": 802
    },
    {
      "epoch": 0.08188164842831028,
      "grad_norm": 0.16150617599487305,
      "learning_rate": 0.001,
      "loss": 2.364105701446533,
      "step": 804
    },
    {
      "epoch": 0.08208533412091801,
      "grad_norm": 0.1564147025346756,
      "learning_rate": 0.001,
      "loss": 2.3739280700683594,
      "step": 806
    },
    {
      "epoch": 0.08228901981352575,
      "grad_norm": 0.15049463510513306,
      "learning_rate": 0.001,
      "loss": 2.301060438156128,
      "step": 808
    },
    {
      "epoch": 0.08249270550613348,
      "grad_norm": 0.10063830763101578,
      "learning_rate": 0.001,
      "loss": 2.297070026397705,
      "step": 810
    },
    {
      "epoch": 0.08269639119874123,
      "grad_norm": 0.11229169368743896,
      "learning_rate": 0.001,
      "loss": 2.336510181427002,
      "step": 812
    },
    {
      "epoch": 0.08290007689134896,
      "grad_norm": 0.16400857269763947,
      "learning_rate": 0.001,
      "loss": 2.3421037197113037,
      "step": 814
    },
    {
      "epoch": 0.0831037625839567,
      "grad_norm": 0.13292518258094788,
      "learning_rate": 0.001,
      "loss": 2.337923526763916,
      "step": 816
    },
    {
      "epoch": 0.08330744827656443,
      "grad_norm": 0.11681003868579865,
      "learning_rate": 0.001,
      "loss": 2.2108919620513916,
      "step": 818
    },
    {
      "epoch": 0.08351113396917217,
      "grad_norm": 0.14619381725788116,
      "learning_rate": 0.001,
      "loss": 2.298758029937744,
      "step": 820
    },
    {
      "epoch": 0.0837148196617799,
      "grad_norm": 0.15487033128738403,
      "learning_rate": 0.001,
      "loss": 2.307450532913208,
      "step": 822
    },
    {
      "epoch": 0.08391850535438765,
      "grad_norm": 0.1401037722826004,
      "learning_rate": 0.001,
      "loss": 2.2012481689453125,
      "step": 824
    },
    {
      "epoch": 0.08412219104699538,
      "grad_norm": 0.11236076802015305,
      "learning_rate": 0.001,
      "loss": 2.327569007873535,
      "step": 826
    },
    {
      "epoch": 0.08432587673960312,
      "grad_norm": 0.14327991008758545,
      "learning_rate": 0.001,
      "loss": 2.3115854263305664,
      "step": 828
    },
    {
      "epoch": 0.08452956243221085,
      "grad_norm": 0.15697742998600006,
      "learning_rate": 0.001,
      "loss": 2.353227138519287,
      "step": 830
    },
    {
      "epoch": 0.0847332481248186,
      "grad_norm": 0.13949653506278992,
      "learning_rate": 0.001,
      "loss": 2.395277500152588,
      "step": 832
    },
    {
      "epoch": 0.08493693381742633,
      "grad_norm": 0.1968502700328827,
      "learning_rate": 0.001,
      "loss": 2.3596739768981934,
      "step": 834
    },
    {
      "epoch": 0.08514061951003407,
      "grad_norm": 0.11924322694540024,
      "learning_rate": 0.001,
      "loss": 2.263094902038574,
      "step": 836
    },
    {
      "epoch": 0.0853443052026418,
      "grad_norm": 0.1290053278207779,
      "learning_rate": 0.001,
      "loss": 2.2881202697753906,
      "step": 838
    },
    {
      "epoch": 0.08554799089524955,
      "grad_norm": 0.13687236607074738,
      "learning_rate": 0.001,
      "loss": 2.3096022605895996,
      "step": 840
    },
    {
      "epoch": 0.08575167658785728,
      "grad_norm": 0.15709546208381653,
      "learning_rate": 0.001,
      "loss": 2.3753647804260254,
      "step": 842
    },
    {
      "epoch": 0.08595536228046502,
      "grad_norm": 0.15128524601459503,
      "learning_rate": 0.001,
      "loss": 2.2196359634399414,
      "step": 844
    },
    {
      "epoch": 0.08615904797307275,
      "grad_norm": 0.17012596130371094,
      "learning_rate": 0.001,
      "loss": 2.3221240043640137,
      "step": 846
    },
    {
      "epoch": 0.0863627336656805,
      "grad_norm": 0.17292873561382294,
      "learning_rate": 0.001,
      "loss": 2.4005184173583984,
      "step": 848
    },
    {
      "epoch": 0.08656641935828822,
      "grad_norm": 0.13018584251403809,
      "learning_rate": 0.001,
      "loss": 2.2153921127319336,
      "step": 850
    },
    {
      "epoch": 0.08677010505089597,
      "grad_norm": 0.10669287294149399,
      "learning_rate": 0.001,
      "loss": 2.194291591644287,
      "step": 852
    },
    {
      "epoch": 0.0869737907435037,
      "grad_norm": 0.14020614326000214,
      "learning_rate": 0.001,
      "loss": 2.2735891342163086,
      "step": 854
    },
    {
      "epoch": 0.08717747643611144,
      "grad_norm": 0.12752361595630646,
      "learning_rate": 0.001,
      "loss": 2.215221405029297,
      "step": 856
    },
    {
      "epoch": 0.08738116212871917,
      "grad_norm": 0.12103628367185593,
      "learning_rate": 0.001,
      "loss": 2.2759177684783936,
      "step": 858
    },
    {
      "epoch": 0.08758484782132692,
      "grad_norm": 0.13997580111026764,
      "learning_rate": 0.001,
      "loss": 2.403286933898926,
      "step": 860
    },
    {
      "epoch": 0.08778853351393465,
      "grad_norm": 0.09950102865695953,
      "learning_rate": 0.001,
      "loss": 2.2898263931274414,
      "step": 862
    },
    {
      "epoch": 0.08799221920654239,
      "grad_norm": 0.10788457095623016,
      "learning_rate": 0.001,
      "loss": 2.4227848052978516,
      "step": 864
    },
    {
      "epoch": 0.08819590489915012,
      "grad_norm": 0.21043109893798828,
      "learning_rate": 0.001,
      "loss": 2.429198980331421,
      "step": 866
    },
    {
      "epoch": 0.08839959059175787,
      "grad_norm": 0.15814237296581268,
      "learning_rate": 0.001,
      "loss": 2.4179248809814453,
      "step": 868
    },
    {
      "epoch": 0.0886032762843656,
      "grad_norm": 0.1744883954524994,
      "learning_rate": 0.001,
      "loss": 2.3965072631835938,
      "step": 870
    },
    {
      "epoch": 0.08880696197697334,
      "grad_norm": 0.1552916318178177,
      "learning_rate": 0.001,
      "loss": 2.328907012939453,
      "step": 872
    },
    {
      "epoch": 0.08901064766958107,
      "grad_norm": 0.13742396235466003,
      "learning_rate": 0.001,
      "loss": 2.3239312171936035,
      "step": 874
    },
    {
      "epoch": 0.0892143333621888,
      "grad_norm": 0.14262227714061737,
      "learning_rate": 0.001,
      "loss": 2.3565616607666016,
      "step": 876
    },
    {
      "epoch": 0.08941801905479654,
      "grad_norm": 0.17221017181873322,
      "learning_rate": 0.001,
      "loss": 2.415496826171875,
      "step": 878
    },
    {
      "epoch": 0.08962170474740427,
      "grad_norm": 0.19875116646289825,
      "learning_rate": 0.001,
      "loss": 2.2481584548950195,
      "step": 880
    },
    {
      "epoch": 0.08982539044001202,
      "grad_norm": 0.1284896582365036,
      "learning_rate": 0.001,
      "loss": 2.4492640495300293,
      "step": 882
    },
    {
      "epoch": 0.09002907613261975,
      "grad_norm": 0.13184677064418793,
      "learning_rate": 0.001,
      "loss": 2.372671127319336,
      "step": 884
    },
    {
      "epoch": 0.09023276182522749,
      "grad_norm": 0.1906169205904007,
      "learning_rate": 0.001,
      "loss": 2.4158291816711426,
      "step": 886
    },
    {
      "epoch": 0.09043644751783522,
      "grad_norm": 0.1401611864566803,
      "learning_rate": 0.001,
      "loss": 2.3385462760925293,
      "step": 888
    },
    {
      "epoch": 0.09064013321044297,
      "grad_norm": 0.12353956699371338,
      "learning_rate": 0.001,
      "loss": 2.278907299041748,
      "step": 890
    },
    {
      "epoch": 0.0908438189030507,
      "grad_norm": 0.16807407140731812,
      "learning_rate": 0.001,
      "loss": 2.2827811241149902,
      "step": 892
    },
    {
      "epoch": 0.09104750459565844,
      "grad_norm": 0.14958223700523376,
      "learning_rate": 0.001,
      "loss": 2.3473262786865234,
      "step": 894
    },
    {
      "epoch": 0.09125119028826617,
      "grad_norm": 0.12488524615764618,
      "learning_rate": 0.001,
      "loss": 2.3199753761291504,
      "step": 896
    },
    {
      "epoch": 0.09145487598087391,
      "grad_norm": 0.11561881005764008,
      "learning_rate": 0.001,
      "loss": 2.3396992683410645,
      "step": 898
    },
    {
      "epoch": 0.09165856167348164,
      "grad_norm": 0.13839057087898254,
      "learning_rate": 0.001,
      "loss": 2.261535167694092,
      "step": 900
    },
    {
      "epoch": 0.09186224736608939,
      "grad_norm": 0.14394783973693848,
      "learning_rate": 0.001,
      "loss": 2.2484734058380127,
      "step": 902
    },
    {
      "epoch": 0.09206593305869712,
      "grad_norm": 0.12262673676013947,
      "learning_rate": 0.001,
      "loss": 2.3997769355773926,
      "step": 904
    },
    {
      "epoch": 0.09226961875130486,
      "grad_norm": 0.15846256911754608,
      "learning_rate": 0.001,
      "loss": 2.482158660888672,
      "step": 906
    },
    {
      "epoch": 0.0924733044439126,
      "grad_norm": 0.1403331607580185,
      "learning_rate": 0.001,
      "loss": 2.3149425983428955,
      "step": 908
    },
    {
      "epoch": 0.09267699013652034,
      "grad_norm": 0.11590701341629028,
      "learning_rate": 0.001,
      "loss": 2.252936840057373,
      "step": 910
    },
    {
      "epoch": 0.09288067582912807,
      "grad_norm": 0.15002688765525818,
      "learning_rate": 0.001,
      "loss": 2.465951919555664,
      "step": 912
    },
    {
      "epoch": 0.09308436152173581,
      "grad_norm": 0.11549942940473557,
      "learning_rate": 0.001,
      "loss": 2.3846688270568848,
      "step": 914
    },
    {
      "epoch": 0.09328804721434354,
      "grad_norm": 0.2140442132949829,
      "learning_rate": 0.001,
      "loss": 2.3112986087799072,
      "step": 916
    },
    {
      "epoch": 0.09349173290695129,
      "grad_norm": 0.11461041867733002,
      "learning_rate": 0.001,
      "loss": 2.34951114654541,
      "step": 918
    },
    {
      "epoch": 0.09369541859955902,
      "grad_norm": 0.14297766983509064,
      "learning_rate": 0.001,
      "loss": 2.3200302124023438,
      "step": 920
    },
    {
      "epoch": 0.09389910429216676,
      "grad_norm": 0.13446763157844543,
      "learning_rate": 0.001,
      "loss": 2.2369260787963867,
      "step": 922
    },
    {
      "epoch": 0.09410278998477449,
      "grad_norm": 0.15914899110794067,
      "learning_rate": 0.001,
      "loss": 2.3423070907592773,
      "step": 924
    },
    {
      "epoch": 0.09430647567738223,
      "grad_norm": 0.1342618763446808,
      "learning_rate": 0.001,
      "loss": 2.372291088104248,
      "step": 926
    },
    {
      "epoch": 0.09451016136998996,
      "grad_norm": 0.13107901811599731,
      "learning_rate": 0.001,
      "loss": 2.4062089920043945,
      "step": 928
    },
    {
      "epoch": 0.09471384706259771,
      "grad_norm": 0.12248536199331284,
      "learning_rate": 0.001,
      "loss": 2.302046298980713,
      "step": 930
    },
    {
      "epoch": 0.09491753275520544,
      "grad_norm": 0.19349296391010284,
      "learning_rate": 0.001,
      "loss": 2.462143659591675,
      "step": 932
    },
    {
      "epoch": 0.09512121844781318,
      "grad_norm": 0.13010412454605103,
      "learning_rate": 0.001,
      "loss": 2.3128662109375,
      "step": 934
    },
    {
      "epoch": 0.09532490414042091,
      "grad_norm": 0.1216125339269638,
      "learning_rate": 0.001,
      "loss": 2.4690804481506348,
      "step": 936
    },
    {
      "epoch": 0.09552858983302866,
      "grad_norm": 0.13656845688819885,
      "learning_rate": 0.001,
      "loss": 2.3659753799438477,
      "step": 938
    },
    {
      "epoch": 0.09573227552563639,
      "grad_norm": 0.12095683813095093,
      "learning_rate": 0.001,
      "loss": 2.232006072998047,
      "step": 940
    },
    {
      "epoch": 0.09593596121824413,
      "grad_norm": 0.1748407930135727,
      "learning_rate": 0.001,
      "loss": 2.3408350944519043,
      "step": 942
    },
    {
      "epoch": 0.09613964691085186,
      "grad_norm": 0.17006203532218933,
      "learning_rate": 0.001,
      "loss": 2.312584161758423,
      "step": 944
    },
    {
      "epoch": 0.0963433326034596,
      "grad_norm": 0.14773087203502655,
      "learning_rate": 0.001,
      "loss": 2.4049830436706543,
      "step": 946
    },
    {
      "epoch": 0.09654701829606734,
      "grad_norm": 0.11357466131448746,
      "learning_rate": 0.001,
      "loss": 2.3488221168518066,
      "step": 948
    },
    {
      "epoch": 0.09675070398867508,
      "grad_norm": 0.1419863998889923,
      "learning_rate": 0.001,
      "loss": 2.396393299102783,
      "step": 950
    },
    {
      "epoch": 0.09695438968128281,
      "grad_norm": 0.1624799519777298,
      "learning_rate": 0.001,
      "loss": 2.202622652053833,
      "step": 952
    },
    {
      "epoch": 0.09715807537389055,
      "grad_norm": 0.1133020743727684,
      "learning_rate": 0.001,
      "loss": 2.319488048553467,
      "step": 954
    },
    {
      "epoch": 0.09736176106649828,
      "grad_norm": 0.11021844297647476,
      "learning_rate": 0.001,
      "loss": 2.2013115882873535,
      "step": 956
    },
    {
      "epoch": 0.09756544675910603,
      "grad_norm": 0.10054001212120056,
      "learning_rate": 0.001,
      "loss": 2.143434524536133,
      "step": 958
    },
    {
      "epoch": 0.09776913245171376,
      "grad_norm": 0.1423506736755371,
      "learning_rate": 0.001,
      "loss": 2.254807949066162,
      "step": 960
    },
    {
      "epoch": 0.0979728181443215,
      "grad_norm": 0.12533621490001678,
      "learning_rate": 0.001,
      "loss": 2.329463005065918,
      "step": 962
    },
    {
      "epoch": 0.09817650383692923,
      "grad_norm": 0.1275789886713028,
      "learning_rate": 0.001,
      "loss": 2.331271171569824,
      "step": 964
    },
    {
      "epoch": 0.09838018952953698,
      "grad_norm": 0.16389058530330658,
      "learning_rate": 0.001,
      "loss": 2.401937246322632,
      "step": 966
    },
    {
      "epoch": 0.0985838752221447,
      "grad_norm": 0.13876675069332123,
      "learning_rate": 0.001,
      "loss": 2.3546090126037598,
      "step": 968
    },
    {
      "epoch": 0.09878756091475245,
      "grad_norm": 0.13486288487911224,
      "learning_rate": 0.001,
      "loss": 2.3958654403686523,
      "step": 970
    },
    {
      "epoch": 0.09899124660736018,
      "grad_norm": 0.14998584985733032,
      "learning_rate": 0.001,
      "loss": 2.268312692642212,
      "step": 972
    },
    {
      "epoch": 0.09919493229996793,
      "grad_norm": 0.131804421544075,
      "learning_rate": 0.001,
      "loss": 2.2712173461914062,
      "step": 974
    },
    {
      "epoch": 0.09939861799257566,
      "grad_norm": 0.19584639370441437,
      "learning_rate": 0.001,
      "loss": 2.324697971343994,
      "step": 976
    },
    {
      "epoch": 0.0996023036851834,
      "grad_norm": 0.14130058884620667,
      "learning_rate": 0.001,
      "loss": 2.2887425422668457,
      "step": 978
    },
    {
      "epoch": 0.09980598937779113,
      "grad_norm": 0.1212717592716217,
      "learning_rate": 0.001,
      "loss": 2.1976208686828613,
      "step": 980
    },
    {
      "epoch": 0.10000967507039887,
      "grad_norm": 0.1271243393421173,
      "learning_rate": 0.001,
      "loss": 2.3817524909973145,
      "step": 982
    },
    {
      "epoch": 0.1002133607630066,
      "grad_norm": 0.17316672205924988,
      "learning_rate": 0.001,
      "loss": 2.3972630500793457,
      "step": 984
    },
    {
      "epoch": 0.10041704645561435,
      "grad_norm": 0.11812806129455566,
      "learning_rate": 0.001,
      "loss": 2.3125171661376953,
      "step": 986
    },
    {
      "epoch": 0.10062073214822208,
      "grad_norm": 0.15972073376178741,
      "learning_rate": 0.001,
      "loss": 2.298402786254883,
      "step": 988
    },
    {
      "epoch": 0.10082441784082982,
      "grad_norm": 0.23898038268089294,
      "learning_rate": 0.001,
      "loss": 2.414161205291748,
      "step": 990
    },
    {
      "epoch": 0.10102810353343755,
      "grad_norm": 0.1334601640701294,
      "learning_rate": 0.001,
      "loss": 2.42509126663208,
      "step": 992
    },
    {
      "epoch": 0.1012317892260453,
      "grad_norm": 0.132346048951149,
      "learning_rate": 0.001,
      "loss": 2.3624181747436523,
      "step": 994
    },
    {
      "epoch": 0.10143547491865303,
      "grad_norm": 0.138170525431633,
      "learning_rate": 0.001,
      "loss": 2.426797389984131,
      "step": 996
    },
    {
      "epoch": 0.10163916061126077,
      "grad_norm": 0.1954992711544037,
      "learning_rate": 0.001,
      "loss": 2.3347153663635254,
      "step": 998
    },
    {
      "epoch": 0.1018428463038685,
      "grad_norm": 0.14430329203605652,
      "learning_rate": 0.001,
      "loss": 2.258366584777832,
      "step": 1000
    },
    {
      "epoch": 0.10204653199647623,
      "grad_norm": 0.1310466080904007,
      "learning_rate": 0.001,
      "loss": 2.3448519706726074,
      "step": 1002
    },
    {
      "epoch": 0.10225021768908397,
      "grad_norm": 0.18919570744037628,
      "learning_rate": 0.001,
      "loss": 2.2650671005249023,
      "step": 1004
    },
    {
      "epoch": 0.1024539033816917,
      "grad_norm": 0.10913380235433578,
      "learning_rate": 0.001,
      "loss": 2.364968776702881,
      "step": 1006
    },
    {
      "epoch": 0.10265758907429945,
      "grad_norm": 0.15841281414031982,
      "learning_rate": 0.001,
      "loss": 2.2783215045928955,
      "step": 1008
    },
    {
      "epoch": 0.10286127476690718,
      "grad_norm": 0.13265272974967957,
      "learning_rate": 0.001,
      "loss": 2.3723952770233154,
      "step": 1010
    },
    {
      "epoch": 0.10306496045951492,
      "grad_norm": 0.12433423101902008,
      "learning_rate": 0.001,
      "loss": 2.350348472595215,
      "step": 1012
    },
    {
      "epoch": 0.10326864615212265,
      "grad_norm": 0.14462880790233612,
      "learning_rate": 0.001,
      "loss": 2.422945737838745,
      "step": 1014
    },
    {
      "epoch": 0.1034723318447304,
      "grad_norm": 0.19107630848884583,
      "learning_rate": 0.001,
      "loss": 2.376389265060425,
      "step": 1016
    },
    {
      "epoch": 0.10367601753733813,
      "grad_norm": 0.1590726226568222,
      "learning_rate": 0.001,
      "loss": 2.377300262451172,
      "step": 1018
    },
    {
      "epoch": 0.10387970322994587,
      "grad_norm": 0.1677413284778595,
      "learning_rate": 0.001,
      "loss": 2.1984310150146484,
      "step": 1020
    },
    {
      "epoch": 0.1040833889225536,
      "grad_norm": 0.11651123315095901,
      "learning_rate": 0.001,
      "loss": 2.3237266540527344,
      "step": 1022
    },
    {
      "epoch": 0.10428707461516135,
      "grad_norm": 0.1055523231625557,
      "learning_rate": 0.001,
      "loss": 2.344867706298828,
      "step": 1024
    },
    {
      "epoch": 0.10449076030776908,
      "grad_norm": 0.14594456553459167,
      "learning_rate": 0.001,
      "loss": 2.2657675743103027,
      "step": 1026
    },
    {
      "epoch": 0.10469444600037682,
      "grad_norm": 0.12498893588781357,
      "learning_rate": 0.001,
      "loss": 2.3380961418151855,
      "step": 1028
    },
    {
      "epoch": 0.10489813169298455,
      "grad_norm": 0.11735816299915314,
      "learning_rate": 0.001,
      "loss": 2.2514801025390625,
      "step": 1030
    },
    {
      "epoch": 0.1051018173855923,
      "grad_norm": 0.11617386341094971,
      "learning_rate": 0.001,
      "loss": 2.3522465229034424,
      "step": 1032
    },
    {
      "epoch": 0.10530550307820002,
      "grad_norm": 0.1427600085735321,
      "learning_rate": 0.001,
      "loss": 2.240522623062134,
      "step": 1034
    },
    {
      "epoch": 0.10550918877080777,
      "grad_norm": 0.13073092699050903,
      "learning_rate": 0.001,
      "loss": 2.371950149536133,
      "step": 1036
    },
    {
      "epoch": 0.1057128744634155,
      "grad_norm": 0.12549246847629547,
      "learning_rate": 0.001,
      "loss": 2.334606647491455,
      "step": 1038
    },
    {
      "epoch": 0.10591656015602324,
      "grad_norm": 0.13150936365127563,
      "learning_rate": 0.001,
      "loss": 2.4346208572387695,
      "step": 1040
    },
    {
      "epoch": 0.10612024584863097,
      "grad_norm": 0.10769376903772354,
      "learning_rate": 0.001,
      "loss": 2.3629727363586426,
      "step": 1042
    },
    {
      "epoch": 0.10632393154123872,
      "grad_norm": 0.15533798933029175,
      "learning_rate": 0.001,
      "loss": 2.162998676300049,
      "step": 1044
    },
    {
      "epoch": 0.10652761723384645,
      "grad_norm": 0.18196429312229156,
      "learning_rate": 0.001,
      "loss": 2.410029888153076,
      "step": 1046
    },
    {
      "epoch": 0.10673130292645419,
      "grad_norm": 0.14360062777996063,
      "learning_rate": 0.001,
      "loss": 2.4534988403320312,
      "step": 1048
    },
    {
      "epoch": 0.10693498861906192,
      "grad_norm": 0.12310464680194855,
      "learning_rate": 0.001,
      "loss": 2.3116860389709473,
      "step": 1050
    },
    {
      "epoch": 0.10713867431166967,
      "grad_norm": 0.13358059525489807,
      "learning_rate": 0.001,
      "loss": 2.132629871368408,
      "step": 1052
    },
    {
      "epoch": 0.1073423600042774,
      "grad_norm": 0.13491131365299225,
      "learning_rate": 0.001,
      "loss": 2.340111255645752,
      "step": 1054
    },
    {
      "epoch": 0.10754604569688514,
      "grad_norm": 0.12104333937168121,
      "learning_rate": 0.001,
      "loss": 2.444775104522705,
      "step": 1056
    },
    {
      "epoch": 0.10774973138949287,
      "grad_norm": 0.16234225034713745,
      "learning_rate": 0.001,
      "loss": 2.2439401149749756,
      "step": 1058
    },
    {
      "epoch": 0.10795341708210061,
      "grad_norm": 0.18181288242340088,
      "learning_rate": 0.001,
      "loss": 2.3053174018859863,
      "step": 1060
    },
    {
      "epoch": 0.10815710277470834,
      "grad_norm": 0.12882298231124878,
      "learning_rate": 0.001,
      "loss": 2.3111460208892822,
      "step": 1062
    },
    {
      "epoch": 0.10836078846731609,
      "grad_norm": 0.14855574071407318,
      "learning_rate": 0.001,
      "loss": 2.2714507579803467,
      "step": 1064
    },
    {
      "epoch": 0.10856447415992382,
      "grad_norm": 0.13528132438659668,
      "learning_rate": 0.001,
      "loss": 2.2530250549316406,
      "step": 1066
    },
    {
      "epoch": 0.10876815985253156,
      "grad_norm": 0.1697559356689453,
      "learning_rate": 0.001,
      "loss": 2.3411383628845215,
      "step": 1068
    },
    {
      "epoch": 0.10897184554513929,
      "grad_norm": 0.12618792057037354,
      "learning_rate": 0.001,
      "loss": 2.3647549152374268,
      "step": 1070
    },
    {
      "epoch": 0.10917553123774704,
      "grad_norm": 0.1760406196117401,
      "learning_rate": 0.001,
      "loss": 2.396852493286133,
      "step": 1072
    },
    {
      "epoch": 0.10937921693035477,
      "grad_norm": 0.11971203237771988,
      "learning_rate": 0.001,
      "loss": 2.234952926635742,
      "step": 1074
    },
    {
      "epoch": 0.10958290262296251,
      "grad_norm": 0.12949073314666748,
      "learning_rate": 0.001,
      "loss": 2.144731283187866,
      "step": 1076
    },
    {
      "epoch": 0.10978658831557024,
      "grad_norm": 0.15076152980327606,
      "learning_rate": 0.001,
      "loss": 2.2908332347869873,
      "step": 1078
    },
    {
      "epoch": 0.10999027400817799,
      "grad_norm": 0.18728488683700562,
      "learning_rate": 0.001,
      "loss": 2.2711992263793945,
      "step": 1080
    },
    {
      "epoch": 0.11019395970078572,
      "grad_norm": 0.180471271276474,
      "learning_rate": 0.001,
      "loss": 2.2513864040374756,
      "step": 1082
    },
    {
      "epoch": 0.11039764539339346,
      "grad_norm": 0.14394719898700714,
      "learning_rate": 0.001,
      "loss": 2.2796318531036377,
      "step": 1084
    },
    {
      "epoch": 0.11060133108600119,
      "grad_norm": 0.12234708666801453,
      "learning_rate": 0.001,
      "loss": 2.336668014526367,
      "step": 1086
    },
    {
      "epoch": 0.11080501677860893,
      "grad_norm": 0.11426885426044464,
      "learning_rate": 0.001,
      "loss": 2.2299487590789795,
      "step": 1088
    },
    {
      "epoch": 0.11100870247121666,
      "grad_norm": 0.12293100357055664,
      "learning_rate": 0.001,
      "loss": 2.286055564880371,
      "step": 1090
    },
    {
      "epoch": 0.11121238816382441,
      "grad_norm": 0.1630256026983261,
      "learning_rate": 0.001,
      "loss": 2.3499207496643066,
      "step": 1092
    },
    {
      "epoch": 0.11141607385643214,
      "grad_norm": 0.1850396692752838,
      "learning_rate": 0.001,
      "loss": 2.2983670234680176,
      "step": 1094
    },
    {
      "epoch": 0.11161975954903988,
      "grad_norm": 0.14822565019130707,
      "learning_rate": 0.001,
      "loss": 2.2744078636169434,
      "step": 1096
    },
    {
      "epoch": 0.11182344524164761,
      "grad_norm": 0.1359357386827469,
      "learning_rate": 0.001,
      "loss": 2.3856053352355957,
      "step": 1098
    },
    {
      "epoch": 0.11202713093425536,
      "grad_norm": 0.14370974898338318,
      "learning_rate": 0.001,
      "loss": 2.3769493103027344,
      "step": 1100
    },
    {
      "epoch": 0.11223081662686309,
      "grad_norm": 0.15565572679042816,
      "learning_rate": 0.001,
      "loss": 2.4240214824676514,
      "step": 1102
    },
    {
      "epoch": 0.11243450231947083,
      "grad_norm": 0.1395672708749771,
      "learning_rate": 0.001,
      "loss": 2.4519031047821045,
      "step": 1104
    },
    {
      "epoch": 0.11263818801207856,
      "grad_norm": 0.1402001827955246,
      "learning_rate": 0.001,
      "loss": 2.3516674041748047,
      "step": 1106
    },
    {
      "epoch": 0.1128418737046863,
      "grad_norm": 0.1812765747308731,
      "learning_rate": 0.001,
      "loss": 2.3173036575317383,
      "step": 1108
    },
    {
      "epoch": 0.11304555939729403,
      "grad_norm": 0.1807703822851181,
      "learning_rate": 0.001,
      "loss": 2.3959858417510986,
      "step": 1110
    },
    {
      "epoch": 0.11324924508990178,
      "grad_norm": 0.10616767406463623,
      "learning_rate": 0.001,
      "loss": 2.303649663925171,
      "step": 1112
    },
    {
      "epoch": 0.11345293078250951,
      "grad_norm": 0.12623263895511627,
      "learning_rate": 0.001,
      "loss": 2.2013654708862305,
      "step": 1114
    },
    {
      "epoch": 0.11365661647511725,
      "grad_norm": 0.1335139125585556,
      "learning_rate": 0.001,
      "loss": 2.3189024925231934,
      "step": 1116
    },
    {
      "epoch": 0.11386030216772498,
      "grad_norm": 0.13938729465007782,
      "learning_rate": 0.001,
      "loss": 2.3767733573913574,
      "step": 1118
    },
    {
      "epoch": 0.11406398786033273,
      "grad_norm": 0.15226581692695618,
      "learning_rate": 0.001,
      "loss": 2.2362565994262695,
      "step": 1120
    },
    {
      "epoch": 0.11426767355294046,
      "grad_norm": 0.1332261860370636,
      "learning_rate": 0.001,
      "loss": 2.390932083129883,
      "step": 1122
    },
    {
      "epoch": 0.11447135924554819,
      "grad_norm": 0.16166451573371887,
      "learning_rate": 0.001,
      "loss": 2.284475564956665,
      "step": 1124
    },
    {
      "epoch": 0.11467504493815593,
      "grad_norm": 0.1606854796409607,
      "learning_rate": 0.001,
      "loss": 2.2420718669891357,
      "step": 1126
    },
    {
      "epoch": 0.11487873063076366,
      "grad_norm": 0.14837868511676788,
      "learning_rate": 0.001,
      "loss": 2.367478370666504,
      "step": 1128
    },
    {
      "epoch": 0.1150824163233714,
      "grad_norm": 0.1660929173231125,
      "learning_rate": 0.001,
      "loss": 2.353435516357422,
      "step": 1130
    },
    {
      "epoch": 0.11528610201597914,
      "grad_norm": 0.14714336395263672,
      "learning_rate": 0.001,
      "loss": 2.240561008453369,
      "step": 1132
    },
    {
      "epoch": 0.11548978770858688,
      "grad_norm": 0.1696758270263672,
      "learning_rate": 0.001,
      "loss": 2.401780605316162,
      "step": 1134
    },
    {
      "epoch": 0.11569347340119461,
      "grad_norm": 0.11708437651395798,
      "learning_rate": 0.001,
      "loss": 2.3840692043304443,
      "step": 1136
    },
    {
      "epoch": 0.11589715909380235,
      "grad_norm": 0.15054014325141907,
      "learning_rate": 0.001,
      "loss": 2.322364568710327,
      "step": 1138
    },
    {
      "epoch": 0.11610084478641008,
      "grad_norm": 0.11653526872396469,
      "learning_rate": 0.001,
      "loss": 2.338688611984253,
      "step": 1140
    },
    {
      "epoch": 0.11630453047901783,
      "grad_norm": 0.165032297372818,
      "learning_rate": 0.001,
      "loss": 2.2469890117645264,
      "step": 1142
    },
    {
      "epoch": 0.11650821617162556,
      "grad_norm": 0.13528777658939362,
      "learning_rate": 0.001,
      "loss": 2.330409288406372,
      "step": 1144
    },
    {
      "epoch": 0.1167119018642333,
      "grad_norm": 0.1406550258398056,
      "learning_rate": 0.001,
      "loss": 2.341418743133545,
      "step": 1146
    },
    {
      "epoch": 0.11691558755684103,
      "grad_norm": 0.13800764083862305,
      "learning_rate": 0.001,
      "loss": 2.1987316608428955,
      "step": 1148
    },
    {
      "epoch": 0.11711927324944878,
      "grad_norm": 0.17280232906341553,
      "learning_rate": 0.001,
      "loss": 2.3652126789093018,
      "step": 1150
    },
    {
      "epoch": 0.11732295894205651,
      "grad_norm": 0.1110534518957138,
      "learning_rate": 0.001,
      "loss": 2.3910956382751465,
      "step": 1152
    },
    {
      "epoch": 0.11752664463466425,
      "grad_norm": 0.1631549745798111,
      "learning_rate": 0.001,
      "loss": 2.2832326889038086,
      "step": 1154
    },
    {
      "epoch": 0.11773033032727198,
      "grad_norm": 0.10658882558345795,
      "learning_rate": 0.001,
      "loss": 2.296307325363159,
      "step": 1156
    },
    {
      "epoch": 0.11793401601987973,
      "grad_norm": 0.11634773015975952,
      "learning_rate": 0.001,
      "loss": 2.2748093605041504,
      "step": 1158
    },
    {
      "epoch": 0.11813770171248746,
      "grad_norm": 0.16808545589447021,
      "learning_rate": 0.001,
      "loss": 2.4028244018554688,
      "step": 1160
    },
    {
      "epoch": 0.1183413874050952,
      "grad_norm": 0.17511147260665894,
      "learning_rate": 0.001,
      "loss": 2.3691062927246094,
      "step": 1162
    },
    {
      "epoch": 0.11854507309770293,
      "grad_norm": 0.14348070323467255,
      "learning_rate": 0.001,
      "loss": 2.2933926582336426,
      "step": 1164
    },
    {
      "epoch": 0.11874875879031067,
      "grad_norm": 0.16089175641536713,
      "learning_rate": 0.001,
      "loss": 2.2251906394958496,
      "step": 1166
    },
    {
      "epoch": 0.1189524444829184,
      "grad_norm": 0.12851308286190033,
      "learning_rate": 0.001,
      "loss": 2.24068021774292,
      "step": 1168
    },
    {
      "epoch": 0.11915613017552615,
      "grad_norm": 0.11311781406402588,
      "learning_rate": 0.001,
      "loss": 2.2718887329101562,
      "step": 1170
    },
    {
      "epoch": 0.11935981586813388,
      "grad_norm": 0.12160394340753555,
      "learning_rate": 0.001,
      "loss": 2.34944224357605,
      "step": 1172
    },
    {
      "epoch": 0.11956350156074162,
      "grad_norm": 0.126570463180542,
      "learning_rate": 0.001,
      "loss": 2.355952739715576,
      "step": 1174
    },
    {
      "epoch": 0.11976718725334935,
      "grad_norm": 0.13494451344013214,
      "learning_rate": 0.001,
      "loss": 2.307828903198242,
      "step": 1176
    },
    {
      "epoch": 0.1199708729459571,
      "grad_norm": 0.13781823217868805,
      "learning_rate": 0.001,
      "loss": 2.324371576309204,
      "step": 1178
    },
    {
      "epoch": 0.12017455863856483,
      "grad_norm": 0.10533767193555832,
      "learning_rate": 0.001,
      "loss": 2.445134401321411,
      "step": 1180
    },
    {
      "epoch": 0.12037824433117257,
      "grad_norm": 0.15252140164375305,
      "learning_rate": 0.001,
      "loss": 2.2274982929229736,
      "step": 1182
    },
    {
      "epoch": 0.1205819300237803,
      "grad_norm": 0.12289360165596008,
      "learning_rate": 0.001,
      "loss": 2.1794257164001465,
      "step": 1184
    },
    {
      "epoch": 0.12078561571638805,
      "grad_norm": 0.10404551029205322,
      "learning_rate": 0.001,
      "loss": 2.279928207397461,
      "step": 1186
    },
    {
      "epoch": 0.12098930140899578,
      "grad_norm": 0.1342879831790924,
      "learning_rate": 0.001,
      "loss": 2.3192224502563477,
      "step": 1188
    },
    {
      "epoch": 0.12119298710160352,
      "grad_norm": 0.12558764219284058,
      "learning_rate": 0.001,
      "loss": 2.1747608184814453,
      "step": 1190
    },
    {
      "epoch": 0.12139667279421125,
      "grad_norm": 0.16525360941886902,
      "learning_rate": 0.001,
      "loss": 2.357774496078491,
      "step": 1192
    },
    {
      "epoch": 0.121600358486819,
      "grad_norm": 0.12379719316959381,
      "learning_rate": 0.001,
      "loss": 2.299389123916626,
      "step": 1194
    },
    {
      "epoch": 0.12180404417942672,
      "grad_norm": 0.11043840646743774,
      "learning_rate": 0.001,
      "loss": 2.255770444869995,
      "step": 1196
    },
    {
      "epoch": 0.12200772987203447,
      "grad_norm": 0.11381401866674423,
      "learning_rate": 0.001,
      "loss": 2.214700698852539,
      "step": 1198
    },
    {
      "epoch": 0.1222114155646422,
      "grad_norm": 0.13698233664035797,
      "learning_rate": 0.001,
      "loss": 2.3045594692230225,
      "step": 1200
    },
    {
      "epoch": 0.12241510125724994,
      "grad_norm": 0.13085117936134338,
      "learning_rate": 0.001,
      "loss": 2.2918033599853516,
      "step": 1202
    },
    {
      "epoch": 0.12261878694985767,
      "grad_norm": 0.14579962193965912,
      "learning_rate": 0.001,
      "loss": 2.2729244232177734,
      "step": 1204
    },
    {
      "epoch": 0.12282247264246542,
      "grad_norm": 0.20603346824645996,
      "learning_rate": 0.001,
      "loss": 2.295858383178711,
      "step": 1206
    },
    {
      "epoch": 0.12302615833507315,
      "grad_norm": 0.14502312242984772,
      "learning_rate": 0.001,
      "loss": 2.269056558609009,
      "step": 1208
    },
    {
      "epoch": 0.12322984402768089,
      "grad_norm": 0.13535895943641663,
      "learning_rate": 0.001,
      "loss": 2.392310857772827,
      "step": 1210
    },
    {
      "epoch": 0.12343352972028862,
      "grad_norm": 0.15947061777114868,
      "learning_rate": 0.001,
      "loss": 2.45267391204834,
      "step": 1212
    },
    {
      "epoch": 0.12363721541289636,
      "grad_norm": 0.15757353603839874,
      "learning_rate": 0.001,
      "loss": 2.2670321464538574,
      "step": 1214
    },
    {
      "epoch": 0.1238409011055041,
      "grad_norm": 0.12380511313676834,
      "learning_rate": 0.001,
      "loss": 2.3104138374328613,
      "step": 1216
    },
    {
      "epoch": 0.12404458679811184,
      "grad_norm": 0.15357719361782074,
      "learning_rate": 0.001,
      "loss": 2.380150318145752,
      "step": 1218
    },
    {
      "epoch": 0.12424827249071957,
      "grad_norm": 0.15045972168445587,
      "learning_rate": 0.001,
      "loss": 2.3635616302490234,
      "step": 1220
    },
    {
      "epoch": 0.12445195818332731,
      "grad_norm": 0.1298726499080658,
      "learning_rate": 0.001,
      "loss": 2.399935483932495,
      "step": 1222
    },
    {
      "epoch": 0.12465564387593504,
      "grad_norm": 0.12366219609975815,
      "learning_rate": 0.001,
      "loss": 2.280714511871338,
      "step": 1224
    },
    {
      "epoch": 0.12485932956854279,
      "grad_norm": 0.12203608453273773,
      "learning_rate": 0.001,
      "loss": 2.3415088653564453,
      "step": 1226
    },
    {
      "epoch": 0.12506301526115052,
      "grad_norm": 0.1977177858352661,
      "learning_rate": 0.001,
      "loss": 2.3016042709350586,
      "step": 1228
    },
    {
      "epoch": 0.12526670095375825,
      "grad_norm": 0.13178561627864838,
      "learning_rate": 0.001,
      "loss": 2.4792494773864746,
      "step": 1230
    },
    {
      "epoch": 0.125470386646366,
      "grad_norm": 0.11507956683635712,
      "learning_rate": 0.001,
      "loss": 2.2157139778137207,
      "step": 1232
    },
    {
      "epoch": 0.12567407233897374,
      "grad_norm": 0.14747236669063568,
      "learning_rate": 0.001,
      "loss": 2.2386555671691895,
      "step": 1234
    },
    {
      "epoch": 0.12587775803158147,
      "grad_norm": 0.14165477454662323,
      "learning_rate": 0.001,
      "loss": 2.359828472137451,
      "step": 1236
    },
    {
      "epoch": 0.1260814437241892,
      "grad_norm": 0.13601887226104736,
      "learning_rate": 0.001,
      "loss": 2.2934937477111816,
      "step": 1238
    },
    {
      "epoch": 0.12628512941679695,
      "grad_norm": 0.1508026272058487,
      "learning_rate": 0.001,
      "loss": 2.4417662620544434,
      "step": 1240
    },
    {
      "epoch": 0.12648881510940468,
      "grad_norm": 0.12827517092227936,
      "learning_rate": 0.001,
      "loss": 2.409933567047119,
      "step": 1242
    },
    {
      "epoch": 0.12669250080201241,
      "grad_norm": 0.1395314633846283,
      "learning_rate": 0.001,
      "loss": 2.2863316535949707,
      "step": 1244
    },
    {
      "epoch": 0.12689618649462014,
      "grad_norm": 0.11949114501476288,
      "learning_rate": 0.001,
      "loss": 2.356004476547241,
      "step": 1246
    },
    {
      "epoch": 0.1270998721872279,
      "grad_norm": 0.13378994166851044,
      "learning_rate": 0.001,
      "loss": 2.435274362564087,
      "step": 1248
    },
    {
      "epoch": 0.12730355787983563,
      "grad_norm": 0.12122593075037003,
      "learning_rate": 0.001,
      "loss": 2.3828868865966797,
      "step": 1250
    },
    {
      "epoch": 0.12750724357244336,
      "grad_norm": 0.10411356389522552,
      "learning_rate": 0.001,
      "loss": 2.204752206802368,
      "step": 1252
    },
    {
      "epoch": 0.1277109292650511,
      "grad_norm": 0.11348672211170197,
      "learning_rate": 0.001,
      "loss": 2.3470256328582764,
      "step": 1254
    },
    {
      "epoch": 0.12791461495765882,
      "grad_norm": 0.177280455827713,
      "learning_rate": 0.001,
      "loss": 2.292320728302002,
      "step": 1256
    },
    {
      "epoch": 0.12811830065026658,
      "grad_norm": 0.12782837450504303,
      "learning_rate": 0.001,
      "loss": 2.3086230754852295,
      "step": 1258
    },
    {
      "epoch": 0.1283219863428743,
      "grad_norm": 0.15036098659038544,
      "learning_rate": 0.001,
      "loss": 2.268972873687744,
      "step": 1260
    },
    {
      "epoch": 0.12852567203548204,
      "grad_norm": 0.11857707798480988,
      "learning_rate": 0.001,
      "loss": 2.3153960704803467,
      "step": 1262
    },
    {
      "epoch": 0.12872935772808977,
      "grad_norm": 0.12831400334835052,
      "learning_rate": 0.001,
      "loss": 2.266936779022217,
      "step": 1264
    },
    {
      "epoch": 0.12893304342069753,
      "grad_norm": 0.12454640120267868,
      "learning_rate": 0.001,
      "loss": 2.3218774795532227,
      "step": 1266
    },
    {
      "epoch": 0.12913672911330526,
      "grad_norm": 0.16310855746269226,
      "learning_rate": 0.001,
      "loss": 2.297043800354004,
      "step": 1268
    },
    {
      "epoch": 0.129340414805913,
      "grad_norm": 0.13950838148593903,
      "learning_rate": 0.001,
      "loss": 2.2700228691101074,
      "step": 1270
    },
    {
      "epoch": 0.12954410049852072,
      "grad_norm": 0.11999953538179398,
      "learning_rate": 0.001,
      "loss": 2.2313079833984375,
      "step": 1272
    },
    {
      "epoch": 0.12974778619112848,
      "grad_norm": 0.18597012758255005,
      "learning_rate": 0.001,
      "loss": 2.189657688140869,
      "step": 1274
    },
    {
      "epoch": 0.1299514718837362,
      "grad_norm": 0.11764484643936157,
      "learning_rate": 0.001,
      "loss": 2.2684998512268066,
      "step": 1276
    },
    {
      "epoch": 0.13015515757634394,
      "grad_norm": 0.8403294086456299,
      "learning_rate": 0.001,
      "loss": 2.160092830657959,
      "step": 1278
    },
    {
      "epoch": 0.13035884326895167,
      "grad_norm": 0.25753694772720337,
      "learning_rate": 0.001,
      "loss": 2.470325469970703,
      "step": 1280
    },
    {
      "epoch": 0.13056252896155943,
      "grad_norm": 0.13458745181560516,
      "learning_rate": 0.001,
      "loss": 2.256683826446533,
      "step": 1282
    },
    {
      "epoch": 0.13076621465416716,
      "grad_norm": 0.12703561782836914,
      "learning_rate": 0.001,
      "loss": 2.3370580673217773,
      "step": 1284
    },
    {
      "epoch": 0.1309699003467749,
      "grad_norm": 0.1511157602071762,
      "learning_rate": 0.001,
      "loss": 2.2443575859069824,
      "step": 1286
    },
    {
      "epoch": 0.13117358603938262,
      "grad_norm": 0.12478653341531754,
      "learning_rate": 0.001,
      "loss": 2.3122358322143555,
      "step": 1288
    },
    {
      "epoch": 0.13137727173199037,
      "grad_norm": 0.12047602981328964,
      "learning_rate": 0.001,
      "loss": 2.3490350246429443,
      "step": 1290
    },
    {
      "epoch": 0.1315809574245981,
      "grad_norm": 0.1821681708097458,
      "learning_rate": 0.001,
      "loss": 2.410033941268921,
      "step": 1292
    },
    {
      "epoch": 0.13178464311720584,
      "grad_norm": 0.16469165682792664,
      "learning_rate": 0.001,
      "loss": 2.2634449005126953,
      "step": 1294
    },
    {
      "epoch": 0.13198832880981357,
      "grad_norm": 0.14360667765140533,
      "learning_rate": 0.001,
      "loss": 2.287686824798584,
      "step": 1296
    },
    {
      "epoch": 0.13219201450242132,
      "grad_norm": 0.11358577758073807,
      "learning_rate": 0.001,
      "loss": 2.333216667175293,
      "step": 1298
    },
    {
      "epoch": 0.13239570019502905,
      "grad_norm": 0.10490573942661285,
      "learning_rate": 0.001,
      "loss": 2.1581897735595703,
      "step": 1300
    },
    {
      "epoch": 0.13259938588763678,
      "grad_norm": 0.15009844303131104,
      "learning_rate": 0.001,
      "loss": 2.3686699867248535,
      "step": 1302
    },
    {
      "epoch": 0.1328030715802445,
      "grad_norm": 0.11042007803916931,
      "learning_rate": 0.001,
      "loss": 2.4477574825286865,
      "step": 1304
    },
    {
      "epoch": 0.13300675727285227,
      "grad_norm": 0.19411614537239075,
      "learning_rate": 0.001,
      "loss": 2.318124532699585,
      "step": 1306
    },
    {
      "epoch": 0.13321044296546,
      "grad_norm": 0.1321125626564026,
      "learning_rate": 0.001,
      "loss": 2.258729934692383,
      "step": 1308
    },
    {
      "epoch": 0.13341412865806773,
      "grad_norm": 0.20469620823860168,
      "learning_rate": 0.001,
      "loss": 2.3881263732910156,
      "step": 1310
    },
    {
      "epoch": 0.13361781435067546,
      "grad_norm": 0.1385854333639145,
      "learning_rate": 0.001,
      "loss": 2.3442578315734863,
      "step": 1312
    },
    {
      "epoch": 0.13382150004328322,
      "grad_norm": 0.13223320245742798,
      "learning_rate": 0.001,
      "loss": 2.2125403881073,
      "step": 1314
    },
    {
      "epoch": 0.13402518573589095,
      "grad_norm": 0.13467329740524292,
      "learning_rate": 0.001,
      "loss": 2.2173047065734863,
      "step": 1316
    },
    {
      "epoch": 0.13422887142849868,
      "grad_norm": 0.14170202612876892,
      "learning_rate": 0.001,
      "loss": 2.224923610687256,
      "step": 1318
    },
    {
      "epoch": 0.1344325571211064,
      "grad_norm": 0.11134914308786392,
      "learning_rate": 0.001,
      "loss": 2.3736720085144043,
      "step": 1320
    },
    {
      "epoch": 0.13463624281371417,
      "grad_norm": 0.12863759696483612,
      "learning_rate": 0.001,
      "loss": 2.2664079666137695,
      "step": 1322
    },
    {
      "epoch": 0.1348399285063219,
      "grad_norm": 0.11852700263261795,
      "learning_rate": 0.001,
      "loss": 2.2379565238952637,
      "step": 1324
    },
    {
      "epoch": 0.13504361419892963,
      "grad_norm": 0.136606827378273,
      "learning_rate": 0.001,
      "loss": 2.401174545288086,
      "step": 1326
    },
    {
      "epoch": 0.13524729989153736,
      "grad_norm": 0.13214415311813354,
      "learning_rate": 0.001,
      "loss": 2.3109569549560547,
      "step": 1328
    },
    {
      "epoch": 0.13545098558414512,
      "grad_norm": 0.13786061108112335,
      "learning_rate": 0.001,
      "loss": 2.2040538787841797,
      "step": 1330
    },
    {
      "epoch": 0.13565467127675285,
      "grad_norm": 0.1231464296579361,
      "learning_rate": 0.001,
      "loss": 2.262195110321045,
      "step": 1332
    },
    {
      "epoch": 0.13585835696936058,
      "grad_norm": 0.16777579486370087,
      "learning_rate": 0.001,
      "loss": 2.23669171333313,
      "step": 1334
    },
    {
      "epoch": 0.1360620426619683,
      "grad_norm": 0.13115210831165314,
      "learning_rate": 0.001,
      "loss": 2.3366878032684326,
      "step": 1336
    },
    {
      "epoch": 0.13626572835457607,
      "grad_norm": 0.11647990345954895,
      "learning_rate": 0.001,
      "loss": 2.3501110076904297,
      "step": 1338
    },
    {
      "epoch": 0.1364694140471838,
      "grad_norm": 0.11065904796123505,
      "learning_rate": 0.001,
      "loss": 2.297945261001587,
      "step": 1340
    },
    {
      "epoch": 0.13667309973979153,
      "grad_norm": 0.13199475407600403,
      "learning_rate": 0.001,
      "loss": 2.268505573272705,
      "step": 1342
    },
    {
      "epoch": 0.13687678543239926,
      "grad_norm": 0.11747533082962036,
      "learning_rate": 0.001,
      "loss": 2.319788932800293,
      "step": 1344
    },
    {
      "epoch": 0.13708047112500701,
      "grad_norm": 0.13678120076656342,
      "learning_rate": 0.001,
      "loss": 2.2595109939575195,
      "step": 1346
    },
    {
      "epoch": 0.13728415681761474,
      "grad_norm": 0.1080140694975853,
      "learning_rate": 0.001,
      "loss": 2.3086862564086914,
      "step": 1348
    },
    {
      "epoch": 0.13748784251022247,
      "grad_norm": 0.12435734272003174,
      "learning_rate": 0.001,
      "loss": 2.2998476028442383,
      "step": 1350
    },
    {
      "epoch": 0.1376915282028302,
      "grad_norm": 0.1304766833782196,
      "learning_rate": 0.001,
      "loss": 2.2706236839294434,
      "step": 1352
    },
    {
      "epoch": 0.13789521389543796,
      "grad_norm": 0.1449604034423828,
      "learning_rate": 0.001,
      "loss": 2.1634068489074707,
      "step": 1354
    },
    {
      "epoch": 0.1380988995880457,
      "grad_norm": 0.11719675362110138,
      "learning_rate": 0.001,
      "loss": 2.285316228866577,
      "step": 1356
    },
    {
      "epoch": 0.13830258528065342,
      "grad_norm": 0.1367073655128479,
      "learning_rate": 0.001,
      "loss": 2.1408684253692627,
      "step": 1358
    },
    {
      "epoch": 0.13850627097326115,
      "grad_norm": 0.13647928833961487,
      "learning_rate": 0.001,
      "loss": 2.2553510665893555,
      "step": 1360
    },
    {
      "epoch": 0.1387099566658689,
      "grad_norm": 0.10615210235118866,
      "learning_rate": 0.001,
      "loss": 2.150329113006592,
      "step": 1362
    },
    {
      "epoch": 0.13891364235847664,
      "grad_norm": 0.12040876597166061,
      "learning_rate": 0.001,
      "loss": 2.3858461380004883,
      "step": 1364
    },
    {
      "epoch": 0.13911732805108437,
      "grad_norm": 0.18915456533432007,
      "learning_rate": 0.001,
      "loss": 2.2814602851867676,
      "step": 1366
    },
    {
      "epoch": 0.1393210137436921,
      "grad_norm": 0.1295802742242813,
      "learning_rate": 0.001,
      "loss": 2.198582887649536,
      "step": 1368
    },
    {
      "epoch": 0.13952469943629986,
      "grad_norm": 0.14024437963962555,
      "learning_rate": 0.001,
      "loss": 2.213052272796631,
      "step": 1370
    },
    {
      "epoch": 0.1397283851289076,
      "grad_norm": 0.19377948343753815,
      "learning_rate": 0.001,
      "loss": 2.260023593902588,
      "step": 1372
    },
    {
      "epoch": 0.13993207082151532,
      "grad_norm": 0.22245225310325623,
      "learning_rate": 0.001,
      "loss": 2.2867884635925293,
      "step": 1374
    },
    {
      "epoch": 0.14013575651412305,
      "grad_norm": 0.13909383118152618,
      "learning_rate": 0.001,
      "loss": 2.281022071838379,
      "step": 1376
    },
    {
      "epoch": 0.14033944220673078,
      "grad_norm": 0.10922561585903168,
      "learning_rate": 0.001,
      "loss": 2.22629451751709,
      "step": 1378
    },
    {
      "epoch": 0.14054312789933854,
      "grad_norm": 0.11606261134147644,
      "learning_rate": 0.001,
      "loss": 2.2350687980651855,
      "step": 1380
    },
    {
      "epoch": 0.14074681359194627,
      "grad_norm": 0.1505487561225891,
      "learning_rate": 0.001,
      "loss": 2.354034423828125,
      "step": 1382
    },
    {
      "epoch": 0.140950499284554,
      "grad_norm": 0.10743935406208038,
      "learning_rate": 0.001,
      "loss": 2.3579516410827637,
      "step": 1384
    },
    {
      "epoch": 0.14115418497716173,
      "grad_norm": 0.11736391484737396,
      "learning_rate": 0.001,
      "loss": 2.1971263885498047,
      "step": 1386
    },
    {
      "epoch": 0.1413578706697695,
      "grad_norm": 0.14929978549480438,
      "learning_rate": 0.001,
      "loss": 2.232727289199829,
      "step": 1388
    },
    {
      "epoch": 0.14156155636237722,
      "grad_norm": 0.14550326764583588,
      "learning_rate": 0.001,
      "loss": 2.292724132537842,
      "step": 1390
    },
    {
      "epoch": 0.14176524205498495,
      "grad_norm": 0.14482112228870392,
      "learning_rate": 0.001,
      "loss": 2.297009229660034,
      "step": 1392
    },
    {
      "epoch": 0.14196892774759268,
      "grad_norm": 0.11566905677318573,
      "learning_rate": 0.001,
      "loss": 2.2508625984191895,
      "step": 1394
    },
    {
      "epoch": 0.14217261344020043,
      "grad_norm": 0.11954132467508316,
      "learning_rate": 0.001,
      "loss": 2.251127243041992,
      "step": 1396
    },
    {
      "epoch": 0.14237629913280817,
      "grad_norm": 0.17162640392780304,
      "learning_rate": 0.001,
      "loss": 2.4085464477539062,
      "step": 1398
    },
    {
      "epoch": 0.1425799848254159,
      "grad_norm": 0.14266429841518402,
      "learning_rate": 0.001,
      "loss": 2.1826343536376953,
      "step": 1400
    },
    {
      "epoch": 0.14278367051802363,
      "grad_norm": 0.13260005414485931,
      "learning_rate": 0.001,
      "loss": 2.3074026107788086,
      "step": 1402
    },
    {
      "epoch": 0.14298735621063138,
      "grad_norm": 0.2018645703792572,
      "learning_rate": 0.001,
      "loss": 2.3938217163085938,
      "step": 1404
    },
    {
      "epoch": 0.1431910419032391,
      "grad_norm": 0.16205783188343048,
      "learning_rate": 0.001,
      "loss": 2.323729991912842,
      "step": 1406
    },
    {
      "epoch": 0.14339472759584684,
      "grad_norm": 0.1310798078775406,
      "learning_rate": 0.001,
      "loss": 2.1933913230895996,
      "step": 1408
    },
    {
      "epoch": 0.14359841328845457,
      "grad_norm": 0.11988269537687302,
      "learning_rate": 0.001,
      "loss": 2.1748721599578857,
      "step": 1410
    },
    {
      "epoch": 0.14380209898106233,
      "grad_norm": 0.1221393346786499,
      "learning_rate": 0.001,
      "loss": 2.1556668281555176,
      "step": 1412
    },
    {
      "epoch": 0.14400578467367006,
      "grad_norm": 0.1125149354338646,
      "learning_rate": 0.001,
      "loss": 2.1856586933135986,
      "step": 1414
    },
    {
      "epoch": 0.1442094703662778,
      "grad_norm": 0.2065027356147766,
      "learning_rate": 0.001,
      "loss": 2.3425517082214355,
      "step": 1416
    },
    {
      "epoch": 0.14441315605888552,
      "grad_norm": 0.1465022712945938,
      "learning_rate": 0.001,
      "loss": 2.3948893547058105,
      "step": 1418
    },
    {
      "epoch": 0.14461684175149328,
      "grad_norm": 0.1188584715127945,
      "learning_rate": 0.001,
      "loss": 2.3004002571105957,
      "step": 1420
    },
    {
      "epoch": 0.144820527444101,
      "grad_norm": 0.12648388743400574,
      "learning_rate": 0.001,
      "loss": 2.2132301330566406,
      "step": 1422
    },
    {
      "epoch": 0.14502421313670874,
      "grad_norm": 0.12114394456148148,
      "learning_rate": 0.001,
      "loss": 2.1825499534606934,
      "step": 1424
    },
    {
      "epoch": 0.14522789882931647,
      "grad_norm": 0.12594999372959137,
      "learning_rate": 0.001,
      "loss": 2.351299524307251,
      "step": 1426
    },
    {
      "epoch": 0.14543158452192423,
      "grad_norm": 0.10092118382453918,
      "learning_rate": 0.001,
      "loss": 2.1911075115203857,
      "step": 1428
    },
    {
      "epoch": 0.14563527021453196,
      "grad_norm": 0.1303112655878067,
      "learning_rate": 0.001,
      "loss": 2.3925914764404297,
      "step": 1430
    },
    {
      "epoch": 0.1458389559071397,
      "grad_norm": 0.13757909834384918,
      "learning_rate": 0.001,
      "loss": 2.3756937980651855,
      "step": 1432
    },
    {
      "epoch": 0.14604264159974742,
      "grad_norm": 0.10290458798408508,
      "learning_rate": 0.001,
      "loss": 2.280261754989624,
      "step": 1434
    },
    {
      "epoch": 0.14624632729235518,
      "grad_norm": 0.11752485483884811,
      "learning_rate": 0.001,
      "loss": 2.2450215816497803,
      "step": 1436
    },
    {
      "epoch": 0.1464500129849629,
      "grad_norm": 0.14681889116764069,
      "learning_rate": 0.001,
      "loss": 2.2488200664520264,
      "step": 1438
    },
    {
      "epoch": 0.14665369867757064,
      "grad_norm": 0.1787908971309662,
      "learning_rate": 0.001,
      "loss": 2.27052640914917,
      "step": 1440
    },
    {
      "epoch": 0.14685738437017837,
      "grad_norm": 0.11947617679834366,
      "learning_rate": 0.001,
      "loss": 2.20919132232666,
      "step": 1442
    },
    {
      "epoch": 0.14706107006278613,
      "grad_norm": 0.11338362842798233,
      "learning_rate": 0.001,
      "loss": 2.3115060329437256,
      "step": 1444
    },
    {
      "epoch": 0.14726475575539386,
      "grad_norm": 0.12206569314002991,
      "learning_rate": 0.001,
      "loss": 2.284231185913086,
      "step": 1446
    },
    {
      "epoch": 0.14746844144800159,
      "grad_norm": 0.2112078219652176,
      "learning_rate": 0.001,
      "loss": 2.282350540161133,
      "step": 1448
    },
    {
      "epoch": 0.14767212714060932,
      "grad_norm": 0.16579729318618774,
      "learning_rate": 0.001,
      "loss": 2.325324058532715,
      "step": 1450
    },
    {
      "epoch": 0.14787581283321707,
      "grad_norm": 0.11571720242500305,
      "learning_rate": 0.001,
      "loss": 2.365626096725464,
      "step": 1452
    },
    {
      "epoch": 0.1480794985258248,
      "grad_norm": 0.16553634405136108,
      "learning_rate": 0.001,
      "loss": 2.3686938285827637,
      "step": 1454
    },
    {
      "epoch": 0.14828318421843253,
      "grad_norm": 0.14248521625995636,
      "learning_rate": 0.001,
      "loss": 2.1547675132751465,
      "step": 1456
    },
    {
      "epoch": 0.14848686991104026,
      "grad_norm": 0.11390973627567291,
      "learning_rate": 0.001,
      "loss": 2.1701176166534424,
      "step": 1458
    },
    {
      "epoch": 0.14869055560364802,
      "grad_norm": 0.19162440299987793,
      "learning_rate": 0.001,
      "loss": 2.1794161796569824,
      "step": 1460
    },
    {
      "epoch": 0.14889424129625575,
      "grad_norm": 0.11151053011417389,
      "learning_rate": 0.001,
      "loss": 2.3425257205963135,
      "step": 1462
    },
    {
      "epoch": 0.14909792698886348,
      "grad_norm": 0.1483498215675354,
      "learning_rate": 0.001,
      "loss": 2.1221303939819336,
      "step": 1464
    },
    {
      "epoch": 0.1493016126814712,
      "grad_norm": 0.14340192079544067,
      "learning_rate": 0.001,
      "loss": 2.2426605224609375,
      "step": 1466
    },
    {
      "epoch": 0.14950529837407897,
      "grad_norm": 0.18132853507995605,
      "learning_rate": 0.001,
      "loss": 2.282252788543701,
      "step": 1468
    },
    {
      "epoch": 0.1497089840666867,
      "grad_norm": 0.14692775905132294,
      "learning_rate": 0.001,
      "loss": 2.2678937911987305,
      "step": 1470
    },
    {
      "epoch": 0.14991266975929443,
      "grad_norm": 0.15779425203800201,
      "learning_rate": 0.001,
      "loss": 2.1876890659332275,
      "step": 1472
    },
    {
      "epoch": 0.15011635545190216,
      "grad_norm": 0.12640318274497986,
      "learning_rate": 0.001,
      "loss": 2.4198780059814453,
      "step": 1474
    },
    {
      "epoch": 0.15032004114450992,
      "grad_norm": 0.12446163594722748,
      "learning_rate": 0.001,
      "loss": 2.2879490852355957,
      "step": 1476
    },
    {
      "epoch": 0.15052372683711765,
      "grad_norm": 0.15538619458675385,
      "learning_rate": 0.001,
      "loss": 2.275510787963867,
      "step": 1478
    },
    {
      "epoch": 0.15072741252972538,
      "grad_norm": 0.1539137214422226,
      "learning_rate": 0.001,
      "loss": 2.4395718574523926,
      "step": 1480
    },
    {
      "epoch": 0.1509310982223331,
      "grad_norm": 0.12473942339420319,
      "learning_rate": 0.001,
      "loss": 2.310051918029785,
      "step": 1482
    },
    {
      "epoch": 0.15113478391494087,
      "grad_norm": 0.1005767360329628,
      "learning_rate": 0.001,
      "loss": 2.238264799118042,
      "step": 1484
    },
    {
      "epoch": 0.1513384696075486,
      "grad_norm": 0.14604520797729492,
      "learning_rate": 0.001,
      "loss": 2.2654194831848145,
      "step": 1486
    },
    {
      "epoch": 0.15154215530015633,
      "grad_norm": 0.11982761323451996,
      "learning_rate": 0.001,
      "loss": 2.250889301300049,
      "step": 1488
    },
    {
      "epoch": 0.15174584099276406,
      "grad_norm": 0.13738788664340973,
      "learning_rate": 0.001,
      "loss": 2.2431583404541016,
      "step": 1490
    },
    {
      "epoch": 0.15194952668537182,
      "grad_norm": 0.12340154498815536,
      "learning_rate": 0.001,
      "loss": 2.2782888412475586,
      "step": 1492
    },
    {
      "epoch": 0.15215321237797955,
      "grad_norm": 0.17523355782032013,
      "learning_rate": 0.001,
      "loss": 2.3537983894348145,
      "step": 1494
    },
    {
      "epoch": 0.15235689807058728,
      "grad_norm": 0.16608746349811554,
      "learning_rate": 0.001,
      "loss": 2.3930535316467285,
      "step": 1496
    },
    {
      "epoch": 0.152560583763195,
      "grad_norm": 0.20769719779491425,
      "learning_rate": 0.001,
      "loss": 2.3184003829956055,
      "step": 1498
    },
    {
      "epoch": 0.15276426945580274,
      "grad_norm": 0.14790062606334686,
      "learning_rate": 0.001,
      "loss": 2.2837750911712646,
      "step": 1500
    },
    {
      "epoch": 0.1529679551484105,
      "grad_norm": 0.1415153443813324,
      "learning_rate": 0.001,
      "loss": 2.4199743270874023,
      "step": 1502
    },
    {
      "epoch": 0.15317164084101823,
      "grad_norm": 0.15071113407611847,
      "learning_rate": 0.001,
      "loss": 2.190336227416992,
      "step": 1504
    },
    {
      "epoch": 0.15337532653362596,
      "grad_norm": 0.14500774443149567,
      "learning_rate": 0.001,
      "loss": 2.2240800857543945,
      "step": 1506
    },
    {
      "epoch": 0.15357901222623369,
      "grad_norm": 0.13779789209365845,
      "learning_rate": 0.001,
      "loss": 2.3525876998901367,
      "step": 1508
    },
    {
      "epoch": 0.15378269791884144,
      "grad_norm": 0.16249080002307892,
      "learning_rate": 0.001,
      "loss": 2.30896258354187,
      "step": 1510
    },
    {
      "epoch": 0.15398638361144917,
      "grad_norm": 0.1434853971004486,
      "learning_rate": 0.001,
      "loss": 2.3443779945373535,
      "step": 1512
    },
    {
      "epoch": 0.1541900693040569,
      "grad_norm": 0.1501772254705429,
      "learning_rate": 0.001,
      "loss": 2.347278594970703,
      "step": 1514
    },
    {
      "epoch": 0.15439375499666463,
      "grad_norm": 0.13574911653995514,
      "learning_rate": 0.001,
      "loss": 2.1997270584106445,
      "step": 1516
    },
    {
      "epoch": 0.1545974406892724,
      "grad_norm": 0.12275908887386322,
      "learning_rate": 0.001,
      "loss": 2.232667922973633,
      "step": 1518
    },
    {
      "epoch": 0.15480112638188012,
      "grad_norm": 0.16242709755897522,
      "learning_rate": 0.001,
      "loss": 2.416630983352661,
      "step": 1520
    },
    {
      "epoch": 0.15500481207448785,
      "grad_norm": 0.12594132125377655,
      "learning_rate": 0.001,
      "loss": 2.3942675590515137,
      "step": 1522
    },
    {
      "epoch": 0.15520849776709558,
      "grad_norm": 0.12508217990398407,
      "learning_rate": 0.001,
      "loss": 2.4326539039611816,
      "step": 1524
    },
    {
      "epoch": 0.15541218345970334,
      "grad_norm": 0.11348503828048706,
      "learning_rate": 0.001,
      "loss": 2.3928873538970947,
      "step": 1526
    },
    {
      "epoch": 0.15561586915231107,
      "grad_norm": 0.14756476879119873,
      "learning_rate": 0.001,
      "loss": 2.293936014175415,
      "step": 1528
    },
    {
      "epoch": 0.1558195548449188,
      "grad_norm": 0.14252212643623352,
      "learning_rate": 0.001,
      "loss": 2.436924695968628,
      "step": 1530
    },
    {
      "epoch": 0.15602324053752653,
      "grad_norm": 0.11945221573114395,
      "learning_rate": 0.001,
      "loss": 2.202279567718506,
      "step": 1532
    },
    {
      "epoch": 0.1562269262301343,
      "grad_norm": 0.1497989147901535,
      "learning_rate": 0.001,
      "loss": 2.297797203063965,
      "step": 1534
    },
    {
      "epoch": 0.15643061192274202,
      "grad_norm": 0.17367427051067352,
      "learning_rate": 0.001,
      "loss": 2.2256722450256348,
      "step": 1536
    },
    {
      "epoch": 0.15663429761534975,
      "grad_norm": 0.11395201832056046,
      "learning_rate": 0.001,
      "loss": 2.1031737327575684,
      "step": 1538
    },
    {
      "epoch": 0.15683798330795748,
      "grad_norm": 0.14142853021621704,
      "learning_rate": 0.001,
      "loss": 2.182234764099121,
      "step": 1540
    },
    {
      "epoch": 0.15704166900056524,
      "grad_norm": 0.19108781218528748,
      "learning_rate": 0.001,
      "loss": 2.316662311553955,
      "step": 1542
    },
    {
      "epoch": 0.15724535469317297,
      "grad_norm": 0.11160020530223846,
      "learning_rate": 0.001,
      "loss": 2.282100200653076,
      "step": 1544
    },
    {
      "epoch": 0.1574490403857807,
      "grad_norm": 0.13792738318443298,
      "learning_rate": 0.001,
      "loss": 2.2409024238586426,
      "step": 1546
    },
    {
      "epoch": 0.15765272607838843,
      "grad_norm": 0.13835802674293518,
      "learning_rate": 0.001,
      "loss": 2.2772560119628906,
      "step": 1548
    },
    {
      "epoch": 0.15785641177099619,
      "grad_norm": 0.12570512294769287,
      "learning_rate": 0.001,
      "loss": 2.289435625076294,
      "step": 1550
    },
    {
      "epoch": 0.15806009746360392,
      "grad_norm": 0.14798298478126526,
      "learning_rate": 0.001,
      "loss": 2.4096624851226807,
      "step": 1552
    },
    {
      "epoch": 0.15826378315621165,
      "grad_norm": 0.12917469441890717,
      "learning_rate": 0.001,
      "loss": 2.3700122833251953,
      "step": 1554
    },
    {
      "epoch": 0.15846746884881938,
      "grad_norm": 0.14884668588638306,
      "learning_rate": 0.001,
      "loss": 2.398839235305786,
      "step": 1556
    },
    {
      "epoch": 0.15867115454142713,
      "grad_norm": 0.11651086062192917,
      "learning_rate": 0.001,
      "loss": 2.3477072715759277,
      "step": 1558
    },
    {
      "epoch": 0.15887484023403486,
      "grad_norm": 0.14477771520614624,
      "learning_rate": 0.001,
      "loss": 2.2315664291381836,
      "step": 1560
    },
    {
      "epoch": 0.1590785259266426,
      "grad_norm": 0.13825131952762604,
      "learning_rate": 0.001,
      "loss": 2.38710355758667,
      "step": 1562
    },
    {
      "epoch": 0.15928221161925032,
      "grad_norm": 0.12688004970550537,
      "learning_rate": 0.001,
      "loss": 2.2716541290283203,
      "step": 1564
    },
    {
      "epoch": 0.15948589731185808,
      "grad_norm": 0.14630845189094543,
      "learning_rate": 0.001,
      "loss": 2.322012424468994,
      "step": 1566
    },
    {
      "epoch": 0.1596895830044658,
      "grad_norm": 0.13042543828487396,
      "learning_rate": 0.001,
      "loss": 2.2893457412719727,
      "step": 1568
    },
    {
      "epoch": 0.15989326869707354,
      "grad_norm": 0.12550748884677887,
      "learning_rate": 0.001,
      "loss": 2.2664923667907715,
      "step": 1570
    },
    {
      "epoch": 0.16009695438968127,
      "grad_norm": 0.11710084974765778,
      "learning_rate": 0.001,
      "loss": 2.265880823135376,
      "step": 1572
    },
    {
      "epoch": 0.16030064008228903,
      "grad_norm": 0.11775349825620651,
      "learning_rate": 0.001,
      "loss": 2.279970645904541,
      "step": 1574
    },
    {
      "epoch": 0.16050432577489676,
      "grad_norm": 0.11084197461605072,
      "learning_rate": 0.001,
      "loss": 2.345181941986084,
      "step": 1576
    },
    {
      "epoch": 0.1607080114675045,
      "grad_norm": 0.12854976952075958,
      "learning_rate": 0.001,
      "loss": 2.4217028617858887,
      "step": 1578
    },
    {
      "epoch": 0.16091169716011222,
      "grad_norm": 0.11781428009271622,
      "learning_rate": 0.001,
      "loss": 2.2615160942077637,
      "step": 1580
    },
    {
      "epoch": 0.16111538285271998,
      "grad_norm": 0.11380614340305328,
      "learning_rate": 0.001,
      "loss": 2.3148841857910156,
      "step": 1582
    },
    {
      "epoch": 0.1613190685453277,
      "grad_norm": 0.16926519572734833,
      "learning_rate": 0.001,
      "loss": 2.283989906311035,
      "step": 1584
    },
    {
      "epoch": 0.16152275423793544,
      "grad_norm": 0.16043107211589813,
      "learning_rate": 0.001,
      "loss": 2.2514662742614746,
      "step": 1586
    },
    {
      "epoch": 0.16172643993054317,
      "grad_norm": 0.14825080335140228,
      "learning_rate": 0.001,
      "loss": 2.25325345993042,
      "step": 1588
    },
    {
      "epoch": 0.16193012562315093,
      "grad_norm": 0.14392808079719543,
      "learning_rate": 0.001,
      "loss": 2.3794808387756348,
      "step": 1590
    },
    {
      "epoch": 0.16213381131575866,
      "grad_norm": 0.1436644196510315,
      "learning_rate": 0.001,
      "loss": 2.3162455558776855,
      "step": 1592
    },
    {
      "epoch": 0.1623374970083664,
      "grad_norm": 0.12363376468420029,
      "learning_rate": 0.001,
      "loss": 2.205590009689331,
      "step": 1594
    },
    {
      "epoch": 0.16254118270097412,
      "grad_norm": 0.14481106400489807,
      "learning_rate": 0.001,
      "loss": 2.2604000568389893,
      "step": 1596
    },
    {
      "epoch": 0.16274486839358188,
      "grad_norm": 0.14503900706768036,
      "learning_rate": 0.001,
      "loss": 2.357645273208618,
      "step": 1598
    },
    {
      "epoch": 0.1629485540861896,
      "grad_norm": 0.15864428877830505,
      "learning_rate": 0.001,
      "loss": 2.241450786590576,
      "step": 1600
    },
    {
      "epoch": 0.16315223977879734,
      "grad_norm": 0.12539087235927582,
      "learning_rate": 0.001,
      "loss": 2.226886034011841,
      "step": 1602
    },
    {
      "epoch": 0.16335592547140507,
      "grad_norm": 0.15284046530723572,
      "learning_rate": 0.001,
      "loss": 2.3101062774658203,
      "step": 1604
    },
    {
      "epoch": 0.16355961116401282,
      "grad_norm": 0.11110225319862366,
      "learning_rate": 0.001,
      "loss": 2.348165512084961,
      "step": 1606
    },
    {
      "epoch": 0.16376329685662055,
      "grad_norm": 0.11495605856180191,
      "learning_rate": 0.001,
      "loss": 2.254300594329834,
      "step": 1608
    },
    {
      "epoch": 0.16396698254922829,
      "grad_norm": 0.15379373729228973,
      "learning_rate": 0.001,
      "loss": 2.246784210205078,
      "step": 1610
    },
    {
      "epoch": 0.16417066824183602,
      "grad_norm": 0.12673284113407135,
      "learning_rate": 0.001,
      "loss": 2.394547700881958,
      "step": 1612
    },
    {
      "epoch": 0.16437435393444377,
      "grad_norm": 0.12681229412555695,
      "learning_rate": 0.001,
      "loss": 2.281219244003296,
      "step": 1614
    },
    {
      "epoch": 0.1645780396270515,
      "grad_norm": 0.11926347017288208,
      "learning_rate": 0.001,
      "loss": 2.386408805847168,
      "step": 1616
    },
    {
      "epoch": 0.16478172531965923,
      "grad_norm": 0.12719887495040894,
      "learning_rate": 0.001,
      "loss": 2.179245948791504,
      "step": 1618
    },
    {
      "epoch": 0.16498541101226696,
      "grad_norm": 0.13935737311840057,
      "learning_rate": 0.001,
      "loss": 2.3050761222839355,
      "step": 1620
    },
    {
      "epoch": 0.16518909670487472,
      "grad_norm": 0.12195385247468948,
      "learning_rate": 0.001,
      "loss": 2.2655844688415527,
      "step": 1622
    },
    {
      "epoch": 0.16539278239748245,
      "grad_norm": 0.1171225756406784,
      "learning_rate": 0.001,
      "loss": 2.1812992095947266,
      "step": 1624
    },
    {
      "epoch": 0.16559646809009018,
      "grad_norm": 0.14381477236747742,
      "learning_rate": 0.001,
      "loss": 2.2905373573303223,
      "step": 1626
    },
    {
      "epoch": 0.1658001537826979,
      "grad_norm": 0.13103318214416504,
      "learning_rate": 0.001,
      "loss": 2.213222026824951,
      "step": 1628
    },
    {
      "epoch": 0.16600383947530564,
      "grad_norm": 0.16448645293712616,
      "learning_rate": 0.001,
      "loss": 2.299635887145996,
      "step": 1630
    },
    {
      "epoch": 0.1662075251679134,
      "grad_norm": 0.1739826649427414,
      "learning_rate": 0.001,
      "loss": 2.271181583404541,
      "step": 1632
    },
    {
      "epoch": 0.16641121086052113,
      "grad_norm": 0.23825599253177643,
      "learning_rate": 0.001,
      "loss": 2.3851051330566406,
      "step": 1634
    },
    {
      "epoch": 0.16661489655312886,
      "grad_norm": 0.16053061187267303,
      "learning_rate": 0.001,
      "loss": 2.1737558841705322,
      "step": 1636
    },
    {
      "epoch": 0.1668185822457366,
      "grad_norm": 0.11751782149076462,
      "learning_rate": 0.001,
      "loss": 2.365385055541992,
      "step": 1638
    },
    {
      "epoch": 0.16702226793834435,
      "grad_norm": 0.2115192413330078,
      "learning_rate": 0.001,
      "loss": 2.292003631591797,
      "step": 1640
    },
    {
      "epoch": 0.16722595363095208,
      "grad_norm": 0.12495293468236923,
      "learning_rate": 0.001,
      "loss": 2.3000550270080566,
      "step": 1642
    },
    {
      "epoch": 0.1674296393235598,
      "grad_norm": 0.13344384729862213,
      "learning_rate": 0.001,
      "loss": 2.2220003604888916,
      "step": 1644
    },
    {
      "epoch": 0.16763332501616754,
      "grad_norm": 0.12182826548814774,
      "learning_rate": 0.001,
      "loss": 2.2011256217956543,
      "step": 1646
    },
    {
      "epoch": 0.1678370107087753,
      "grad_norm": 0.15800395607948303,
      "learning_rate": 0.001,
      "loss": 2.1917614936828613,
      "step": 1648
    },
    {
      "epoch": 0.16804069640138303,
      "grad_norm": 0.13424475491046906,
      "learning_rate": 0.001,
      "loss": 2.271029233932495,
      "step": 1650
    },
    {
      "epoch": 0.16824438209399076,
      "grad_norm": 0.16282255947589874,
      "learning_rate": 0.001,
      "loss": 2.3509764671325684,
      "step": 1652
    },
    {
      "epoch": 0.1684480677865985,
      "grad_norm": 0.11741949617862701,
      "learning_rate": 0.001,
      "loss": 2.106837749481201,
      "step": 1654
    },
    {
      "epoch": 0.16865175347920625,
      "grad_norm": 0.1644708663225174,
      "learning_rate": 0.001,
      "loss": 2.2175421714782715,
      "step": 1656
    },
    {
      "epoch": 0.16885543917181398,
      "grad_norm": 0.18072333931922913,
      "learning_rate": 0.001,
      "loss": 2.4644646644592285,
      "step": 1658
    },
    {
      "epoch": 0.1690591248644217,
      "grad_norm": 0.12914173305034637,
      "learning_rate": 0.001,
      "loss": 2.2569591999053955,
      "step": 1660
    },
    {
      "epoch": 0.16926281055702944,
      "grad_norm": 0.11493448168039322,
      "learning_rate": 0.001,
      "loss": 2.195399522781372,
      "step": 1662
    },
    {
      "epoch": 0.1694664962496372,
      "grad_norm": 0.1798621416091919,
      "learning_rate": 0.001,
      "loss": 2.3416361808776855,
      "step": 1664
    },
    {
      "epoch": 0.16967018194224492,
      "grad_norm": 0.14603568613529205,
      "learning_rate": 0.001,
      "loss": 2.329366683959961,
      "step": 1666
    },
    {
      "epoch": 0.16987386763485265,
      "grad_norm": 0.1269005388021469,
      "learning_rate": 0.001,
      "loss": 2.333205223083496,
      "step": 1668
    },
    {
      "epoch": 0.17007755332746038,
      "grad_norm": 0.19661493599414825,
      "learning_rate": 0.001,
      "loss": 2.322256088256836,
      "step": 1670
    },
    {
      "epoch": 0.17028123902006814,
      "grad_norm": 0.16902798414230347,
      "learning_rate": 0.001,
      "loss": 2.3290462493896484,
      "step": 1672
    },
    {
      "epoch": 0.17048492471267587,
      "grad_norm": 0.12297827005386353,
      "learning_rate": 0.001,
      "loss": 2.3565680980682373,
      "step": 1674
    },
    {
      "epoch": 0.1706886104052836,
      "grad_norm": 0.1306285858154297,
      "learning_rate": 0.001,
      "loss": 2.2724099159240723,
      "step": 1676
    },
    {
      "epoch": 0.17089229609789133,
      "grad_norm": 0.10821755975484848,
      "learning_rate": 0.001,
      "loss": 2.2606139183044434,
      "step": 1678
    },
    {
      "epoch": 0.1710959817904991,
      "grad_norm": 0.16018140316009521,
      "learning_rate": 0.001,
      "loss": 2.1930341720581055,
      "step": 1680
    },
    {
      "epoch": 0.17129966748310682,
      "grad_norm": 0.11929906904697418,
      "learning_rate": 0.001,
      "loss": 2.19176983833313,
      "step": 1682
    },
    {
      "epoch": 0.17150335317571455,
      "grad_norm": 0.12529197335243225,
      "learning_rate": 0.001,
      "loss": 2.2159552574157715,
      "step": 1684
    },
    {
      "epoch": 0.17170703886832228,
      "grad_norm": 0.12740343809127808,
      "learning_rate": 0.001,
      "loss": 2.1699676513671875,
      "step": 1686
    },
    {
      "epoch": 0.17191072456093004,
      "grad_norm": 0.13094380497932434,
      "learning_rate": 0.001,
      "loss": 2.3254294395446777,
      "step": 1688
    },
    {
      "epoch": 0.17211441025353777,
      "grad_norm": 0.11066184192895889,
      "learning_rate": 0.001,
      "loss": 2.279755115509033,
      "step": 1690
    },
    {
      "epoch": 0.1723180959461455,
      "grad_norm": 0.11544082313776016,
      "learning_rate": 0.001,
      "loss": 2.4295830726623535,
      "step": 1692
    },
    {
      "epoch": 0.17252178163875323,
      "grad_norm": 0.11252087354660034,
      "learning_rate": 0.001,
      "loss": 2.3428802490234375,
      "step": 1694
    },
    {
      "epoch": 0.172725467331361,
      "grad_norm": 0.14396661520004272,
      "learning_rate": 0.001,
      "loss": 2.310354471206665,
      "step": 1696
    },
    {
      "epoch": 0.17292915302396872,
      "grad_norm": 0.11711877584457397,
      "learning_rate": 0.001,
      "loss": 2.257307529449463,
      "step": 1698
    },
    {
      "epoch": 0.17313283871657645,
      "grad_norm": 0.10243598371744156,
      "learning_rate": 0.001,
      "loss": 2.225754737854004,
      "step": 1700
    },
    {
      "epoch": 0.17333652440918418,
      "grad_norm": 0.13412359356880188,
      "learning_rate": 0.001,
      "loss": 2.290280818939209,
      "step": 1702
    },
    {
      "epoch": 0.17354021010179194,
      "grad_norm": 0.14657658338546753,
      "learning_rate": 0.001,
      "loss": 2.319303035736084,
      "step": 1704
    },
    {
      "epoch": 0.17374389579439967,
      "grad_norm": 0.17576320469379425,
      "learning_rate": 0.001,
      "loss": 2.1556406021118164,
      "step": 1706
    },
    {
      "epoch": 0.1739475814870074,
      "grad_norm": 0.12034939229488373,
      "learning_rate": 0.001,
      "loss": 2.218989372253418,
      "step": 1708
    },
    {
      "epoch": 0.17415126717961513,
      "grad_norm": 0.13272996246814728,
      "learning_rate": 0.001,
      "loss": 2.2968735694885254,
      "step": 1710
    },
    {
      "epoch": 0.17435495287222288,
      "grad_norm": 0.12485348433256149,
      "learning_rate": 0.001,
      "loss": 2.3677115440368652,
      "step": 1712
    },
    {
      "epoch": 0.17455863856483061,
      "grad_norm": 0.1644013226032257,
      "learning_rate": 0.001,
      "loss": 2.284097194671631,
      "step": 1714
    },
    {
      "epoch": 0.17476232425743835,
      "grad_norm": 0.13006234169006348,
      "learning_rate": 0.001,
      "loss": 2.31014347076416,
      "step": 1716
    },
    {
      "epoch": 0.17496600995004608,
      "grad_norm": 0.1224561482667923,
      "learning_rate": 0.001,
      "loss": 2.3670620918273926,
      "step": 1718
    },
    {
      "epoch": 0.17516969564265383,
      "grad_norm": 0.12992897629737854,
      "learning_rate": 0.001,
      "loss": 2.178412437438965,
      "step": 1720
    },
    {
      "epoch": 0.17537338133526156,
      "grad_norm": 0.12214700877666473,
      "learning_rate": 0.001,
      "loss": 2.320553779602051,
      "step": 1722
    },
    {
      "epoch": 0.1755770670278693,
      "grad_norm": 0.1568002551794052,
      "learning_rate": 0.001,
      "loss": 2.3351945877075195,
      "step": 1724
    },
    {
      "epoch": 0.17578075272047702,
      "grad_norm": 0.15210723876953125,
      "learning_rate": 0.001,
      "loss": 2.218327522277832,
      "step": 1726
    },
    {
      "epoch": 0.17598443841308478,
      "grad_norm": 0.13396842777729034,
      "learning_rate": 0.001,
      "loss": 2.3767080307006836,
      "step": 1728
    },
    {
      "epoch": 0.1761881241056925,
      "grad_norm": 0.10916212201118469,
      "learning_rate": 0.001,
      "loss": 2.2568793296813965,
      "step": 1730
    },
    {
      "epoch": 0.17639180979830024,
      "grad_norm": 0.14060238003730774,
      "learning_rate": 0.001,
      "loss": 2.150402069091797,
      "step": 1732
    },
    {
      "epoch": 0.17659549549090797,
      "grad_norm": 0.12852467596530914,
      "learning_rate": 0.001,
      "loss": 2.139565944671631,
      "step": 1734
    },
    {
      "epoch": 0.17679918118351573,
      "grad_norm": 0.1313311606645584,
      "learning_rate": 0.001,
      "loss": 2.3456919193267822,
      "step": 1736
    },
    {
      "epoch": 0.17700286687612346,
      "grad_norm": 0.13520671427249908,
      "learning_rate": 0.001,
      "loss": 2.259895086288452,
      "step": 1738
    },
    {
      "epoch": 0.1772065525687312,
      "grad_norm": 0.1742531657218933,
      "learning_rate": 0.001,
      "loss": 2.2307538986206055,
      "step": 1740
    },
    {
      "epoch": 0.17741023826133892,
      "grad_norm": 0.12394128739833832,
      "learning_rate": 0.001,
      "loss": 2.3418262004852295,
      "step": 1742
    },
    {
      "epoch": 0.17761392395394668,
      "grad_norm": 0.1508609801530838,
      "learning_rate": 0.001,
      "loss": 2.179624557495117,
      "step": 1744
    },
    {
      "epoch": 0.1778176096465544,
      "grad_norm": 0.13064666092395782,
      "learning_rate": 0.001,
      "loss": 2.2286763191223145,
      "step": 1746
    },
    {
      "epoch": 0.17802129533916214,
      "grad_norm": 0.15075881779193878,
      "learning_rate": 0.001,
      "loss": 2.3078582286834717,
      "step": 1748
    },
    {
      "epoch": 0.17822498103176987,
      "grad_norm": 0.14731241762638092,
      "learning_rate": 0.001,
      "loss": 2.2993414402008057,
      "step": 1750
    },
    {
      "epoch": 0.1784286667243776,
      "grad_norm": 0.15912532806396484,
      "learning_rate": 0.001,
      "loss": 2.2270431518554688,
      "step": 1752
    },
    {
      "epoch": 0.17863235241698536,
      "grad_norm": 0.12861213088035583,
      "learning_rate": 0.001,
      "loss": 2.3897876739501953,
      "step": 1754
    },
    {
      "epoch": 0.1788360381095931,
      "grad_norm": 0.16583751142024994,
      "learning_rate": 0.001,
      "loss": 2.3085989952087402,
      "step": 1756
    },
    {
      "epoch": 0.17903972380220082,
      "grad_norm": 0.20208525657653809,
      "learning_rate": 0.001,
      "loss": 2.221642017364502,
      "step": 1758
    },
    {
      "epoch": 0.17924340949480855,
      "grad_norm": 0.13053898513317108,
      "learning_rate": 0.001,
      "loss": 2.250598430633545,
      "step": 1760
    },
    {
      "epoch": 0.1794470951874163,
      "grad_norm": 0.12090899050235748,
      "learning_rate": 0.001,
      "loss": 2.3353374004364014,
      "step": 1762
    },
    {
      "epoch": 0.17965078088002404,
      "grad_norm": 0.11142735183238983,
      "learning_rate": 0.001,
      "loss": 2.318586587905884,
      "step": 1764
    },
    {
      "epoch": 0.17985446657263177,
      "grad_norm": 0.1642095446586609,
      "learning_rate": 0.001,
      "loss": 2.2444489002227783,
      "step": 1766
    },
    {
      "epoch": 0.1800581522652395,
      "grad_norm": 0.112493135035038,
      "learning_rate": 0.001,
      "loss": 2.182068347930908,
      "step": 1768
    },
    {
      "epoch": 0.18026183795784725,
      "grad_norm": 0.1195751428604126,
      "learning_rate": 0.001,
      "loss": 2.258422374725342,
      "step": 1770
    },
    {
      "epoch": 0.18046552365045498,
      "grad_norm": 0.11072104424238205,
      "learning_rate": 0.001,
      "loss": 2.219205379486084,
      "step": 1772
    },
    {
      "epoch": 0.18066920934306271,
      "grad_norm": 0.11943604052066803,
      "learning_rate": 0.001,
      "loss": 2.3309390544891357,
      "step": 1774
    },
    {
      "epoch": 0.18087289503567044,
      "grad_norm": 0.13090580701828003,
      "learning_rate": 0.001,
      "loss": 2.411932945251465,
      "step": 1776
    },
    {
      "epoch": 0.1810765807282782,
      "grad_norm": 0.15213090181350708,
      "learning_rate": 0.001,
      "loss": 2.2222065925598145,
      "step": 1778
    },
    {
      "epoch": 0.18128026642088593,
      "grad_norm": 0.11946600675582886,
      "learning_rate": 0.001,
      "loss": 2.2404463291168213,
      "step": 1780
    },
    {
      "epoch": 0.18148395211349366,
      "grad_norm": 0.13827477395534515,
      "learning_rate": 0.001,
      "loss": 2.2047629356384277,
      "step": 1782
    },
    {
      "epoch": 0.1816876378061014,
      "grad_norm": 0.14407865703105927,
      "learning_rate": 0.001,
      "loss": 2.2788329124450684,
      "step": 1784
    },
    {
      "epoch": 0.18189132349870915,
      "grad_norm": 0.10818200558423996,
      "learning_rate": 0.001,
      "loss": 2.2631936073303223,
      "step": 1786
    },
    {
      "epoch": 0.18209500919131688,
      "grad_norm": 0.14322280883789062,
      "learning_rate": 0.001,
      "loss": 2.2802233695983887,
      "step": 1788
    },
    {
      "epoch": 0.1822986948839246,
      "grad_norm": 0.156350240111351,
      "learning_rate": 0.001,
      "loss": 2.162168502807617,
      "step": 1790
    },
    {
      "epoch": 0.18250238057653234,
      "grad_norm": 0.1306649148464203,
      "learning_rate": 0.001,
      "loss": 2.309302806854248,
      "step": 1792
    },
    {
      "epoch": 0.1827060662691401,
      "grad_norm": 0.13662590086460114,
      "learning_rate": 0.001,
      "loss": 2.310072898864746,
      "step": 1794
    },
    {
      "epoch": 0.18290975196174783,
      "grad_norm": 0.10258147120475769,
      "learning_rate": 0.001,
      "loss": 2.1784932613372803,
      "step": 1796
    },
    {
      "epoch": 0.18311343765435556,
      "grad_norm": 0.15004298090934753,
      "learning_rate": 0.001,
      "loss": 2.267864227294922,
      "step": 1798
    },
    {
      "epoch": 0.1833171233469633,
      "grad_norm": 0.1482090950012207,
      "learning_rate": 0.001,
      "loss": 2.2371299266815186,
      "step": 1800
    },
    {
      "epoch": 0.18352080903957105,
      "grad_norm": 0.12361263483762741,
      "learning_rate": 0.001,
      "loss": 2.2280044555664062,
      "step": 1802
    },
    {
      "epoch": 0.18372449473217878,
      "grad_norm": 0.20645122230052948,
      "learning_rate": 0.001,
      "loss": 2.324251651763916,
      "step": 1804
    },
    {
      "epoch": 0.1839281804247865,
      "grad_norm": 0.1301790475845337,
      "learning_rate": 0.001,
      "loss": 2.3393445014953613,
      "step": 1806
    },
    {
      "epoch": 0.18413186611739424,
      "grad_norm": 0.17419618368148804,
      "learning_rate": 0.001,
      "loss": 2.272359848022461,
      "step": 1808
    },
    {
      "epoch": 0.184335551810002,
      "grad_norm": 0.17688395082950592,
      "learning_rate": 0.001,
      "loss": 2.254687786102295,
      "step": 1810
    },
    {
      "epoch": 0.18453923750260973,
      "grad_norm": 0.14215746521949768,
      "learning_rate": 0.001,
      "loss": 2.25285005569458,
      "step": 1812
    },
    {
      "epoch": 0.18474292319521746,
      "grad_norm": 0.13047850131988525,
      "learning_rate": 0.001,
      "loss": 2.3001708984375,
      "step": 1814
    },
    {
      "epoch": 0.1849466088878252,
      "grad_norm": 0.14951179921627045,
      "learning_rate": 0.001,
      "loss": 2.155695676803589,
      "step": 1816
    },
    {
      "epoch": 0.18515029458043294,
      "grad_norm": 0.16679415106773376,
      "learning_rate": 0.001,
      "loss": 2.211757183074951,
      "step": 1818
    },
    {
      "epoch": 0.18535398027304067,
      "grad_norm": 0.11148359626531601,
      "learning_rate": 0.001,
      "loss": 2.245142936706543,
      "step": 1820
    },
    {
      "epoch": 0.1855576659656484,
      "grad_norm": 0.15451306104660034,
      "learning_rate": 0.001,
      "loss": 2.1444811820983887,
      "step": 1822
    },
    {
      "epoch": 0.18576135165825614,
      "grad_norm": 0.138282909989357,
      "learning_rate": 0.001,
      "loss": 2.1940650939941406,
      "step": 1824
    },
    {
      "epoch": 0.1859650373508639,
      "grad_norm": 0.12361491471529007,
      "learning_rate": 0.001,
      "loss": 2.2921230792999268,
      "step": 1826
    },
    {
      "epoch": 0.18616872304347162,
      "grad_norm": 0.12076527625322342,
      "learning_rate": 0.001,
      "loss": 2.408809185028076,
      "step": 1828
    },
    {
      "epoch": 0.18637240873607935,
      "grad_norm": 0.1443057805299759,
      "learning_rate": 0.001,
      "loss": 2.162017583847046,
      "step": 1830
    },
    {
      "epoch": 0.18657609442868708,
      "grad_norm": 0.16993238031864166,
      "learning_rate": 0.001,
      "loss": 2.3511500358581543,
      "step": 1832
    },
    {
      "epoch": 0.18677978012129484,
      "grad_norm": 0.1373995989561081,
      "learning_rate": 0.001,
      "loss": 2.2945899963378906,
      "step": 1834
    },
    {
      "epoch": 0.18698346581390257,
      "grad_norm": 0.14401763677597046,
      "learning_rate": 0.001,
      "loss": 2.250800132751465,
      "step": 1836
    },
    {
      "epoch": 0.1871871515065103,
      "grad_norm": 0.12688805162906647,
      "learning_rate": 0.001,
      "loss": 2.2637619972229004,
      "step": 1838
    },
    {
      "epoch": 0.18739083719911803,
      "grad_norm": 0.13894347846508026,
      "learning_rate": 0.001,
      "loss": 2.3058338165283203,
      "step": 1840
    },
    {
      "epoch": 0.1875945228917258,
      "grad_norm": 0.15748387575149536,
      "learning_rate": 0.001,
      "loss": 2.318121910095215,
      "step": 1842
    },
    {
      "epoch": 0.18779820858433352,
      "grad_norm": 0.13450486958026886,
      "learning_rate": 0.001,
      "loss": 2.2128782272338867,
      "step": 1844
    },
    {
      "epoch": 0.18800189427694125,
      "grad_norm": 0.09953590482473373,
      "learning_rate": 0.001,
      "loss": 2.1747329235076904,
      "step": 1846
    },
    {
      "epoch": 0.18820557996954898,
      "grad_norm": 0.12754009664058685,
      "learning_rate": 0.001,
      "loss": 2.1000475883483887,
      "step": 1848
    },
    {
      "epoch": 0.18840926566215674,
      "grad_norm": 0.1944160908460617,
      "learning_rate": 0.001,
      "loss": 2.248335838317871,
      "step": 1850
    },
    {
      "epoch": 0.18861295135476447,
      "grad_norm": 0.14382530748844147,
      "learning_rate": 0.001,
      "loss": 2.2490291595458984,
      "step": 1852
    },
    {
      "epoch": 0.1888166370473722,
      "grad_norm": 0.14138129353523254,
      "learning_rate": 0.001,
      "loss": 2.312286376953125,
      "step": 1854
    },
    {
      "epoch": 0.18902032273997993,
      "grad_norm": 0.15458998084068298,
      "learning_rate": 0.001,
      "loss": 2.268916606903076,
      "step": 1856
    },
    {
      "epoch": 0.1892240084325877,
      "grad_norm": 0.11406920850276947,
      "learning_rate": 0.001,
      "loss": 2.211608409881592,
      "step": 1858
    },
    {
      "epoch": 0.18942769412519542,
      "grad_norm": 0.17857547104358673,
      "learning_rate": 0.001,
      "loss": 2.28853702545166,
      "step": 1860
    },
    {
      "epoch": 0.18963137981780315,
      "grad_norm": 0.1451644003391266,
      "learning_rate": 0.001,
      "loss": 2.3160321712493896,
      "step": 1862
    },
    {
      "epoch": 0.18983506551041088,
      "grad_norm": 0.13216717541217804,
      "learning_rate": 0.001,
      "loss": 2.3635177612304688,
      "step": 1864
    },
    {
      "epoch": 0.19003875120301864,
      "grad_norm": 0.13350705802440643,
      "learning_rate": 0.001,
      "loss": 2.3776588439941406,
      "step": 1866
    },
    {
      "epoch": 0.19024243689562637,
      "grad_norm": 0.11901355534791946,
      "learning_rate": 0.001,
      "loss": 2.2876992225646973,
      "step": 1868
    },
    {
      "epoch": 0.1904461225882341,
      "grad_norm": 0.16949307918548584,
      "learning_rate": 0.001,
      "loss": 2.283496856689453,
      "step": 1870
    },
    {
      "epoch": 0.19064980828084183,
      "grad_norm": 0.21688564121723175,
      "learning_rate": 0.001,
      "loss": 2.0967283248901367,
      "step": 1872
    },
    {
      "epoch": 0.19085349397344956,
      "grad_norm": 0.12751726806163788,
      "learning_rate": 0.001,
      "loss": 2.133613348007202,
      "step": 1874
    },
    {
      "epoch": 0.19105717966605731,
      "grad_norm": 0.16183775663375854,
      "learning_rate": 0.001,
      "loss": 2.3090200424194336,
      "step": 1876
    },
    {
      "epoch": 0.19126086535866504,
      "grad_norm": 0.147927388548851,
      "learning_rate": 0.001,
      "loss": 2.2997188568115234,
      "step": 1878
    },
    {
      "epoch": 0.19146455105127277,
      "grad_norm": 0.1989961564540863,
      "learning_rate": 0.001,
      "loss": 2.2083957195281982,
      "step": 1880
    },
    {
      "epoch": 0.1916682367438805,
      "grad_norm": 0.19957345724105835,
      "learning_rate": 0.001,
      "loss": 2.202188014984131,
      "step": 1882
    },
    {
      "epoch": 0.19187192243648826,
      "grad_norm": 0.18530404567718506,
      "learning_rate": 0.001,
      "loss": 2.246824026107788,
      "step": 1884
    },
    {
      "epoch": 0.192075608129096,
      "grad_norm": 0.20355533063411713,
      "learning_rate": 0.001,
      "loss": 2.321110486984253,
      "step": 1886
    },
    {
      "epoch": 0.19227929382170372,
      "grad_norm": 0.17978130280971527,
      "learning_rate": 0.001,
      "loss": 2.324754238128662,
      "step": 1888
    },
    {
      "epoch": 0.19248297951431145,
      "grad_norm": 0.1350458562374115,
      "learning_rate": 0.001,
      "loss": 2.2518486976623535,
      "step": 1890
    },
    {
      "epoch": 0.1926866652069192,
      "grad_norm": 0.1311439722776413,
      "learning_rate": 0.001,
      "loss": 2.3041610717773438,
      "step": 1892
    },
    {
      "epoch": 0.19289035089952694,
      "grad_norm": 0.13782720267772675,
      "learning_rate": 0.001,
      "loss": 2.2900140285491943,
      "step": 1894
    },
    {
      "epoch": 0.19309403659213467,
      "grad_norm": 0.1125819981098175,
      "learning_rate": 0.001,
      "loss": 2.1185519695281982,
      "step": 1896
    },
    {
      "epoch": 0.1932977222847424,
      "grad_norm": 0.1338050812482834,
      "learning_rate": 0.001,
      "loss": 2.2318191528320312,
      "step": 1898
    },
    {
      "epoch": 0.19350140797735016,
      "grad_norm": 0.15599524974822998,
      "learning_rate": 0.001,
      "loss": 2.2573604583740234,
      "step": 1900
    },
    {
      "epoch": 0.1937050936699579,
      "grad_norm": 0.17383265495300293,
      "learning_rate": 0.001,
      "loss": 2.288689613342285,
      "step": 1902
    },
    {
      "epoch": 0.19390877936256562,
      "grad_norm": 0.12204976379871368,
      "learning_rate": 0.001,
      "loss": 2.2188026905059814,
      "step": 1904
    },
    {
      "epoch": 0.19411246505517335,
      "grad_norm": 0.11854109168052673,
      "learning_rate": 0.001,
      "loss": 2.180716037750244,
      "step": 1906
    },
    {
      "epoch": 0.1943161507477811,
      "grad_norm": 0.12776939570903778,
      "learning_rate": 0.001,
      "loss": 2.3203916549682617,
      "step": 1908
    },
    {
      "epoch": 0.19451983644038884,
      "grad_norm": 0.16124294698238373,
      "learning_rate": 0.001,
      "loss": 2.2740416526794434,
      "step": 1910
    },
    {
      "epoch": 0.19472352213299657,
      "grad_norm": 0.1658787876367569,
      "learning_rate": 0.001,
      "loss": 2.3415675163269043,
      "step": 1912
    },
    {
      "epoch": 0.1949272078256043,
      "grad_norm": 0.110081747174263,
      "learning_rate": 0.001,
      "loss": 2.2325198650360107,
      "step": 1914
    },
    {
      "epoch": 0.19513089351821206,
      "grad_norm": 0.12876591086387634,
      "learning_rate": 0.001,
      "loss": 2.227674961090088,
      "step": 1916
    },
    {
      "epoch": 0.1953345792108198,
      "grad_norm": 0.1319383978843689,
      "learning_rate": 0.001,
      "loss": 2.2369399070739746,
      "step": 1918
    },
    {
      "epoch": 0.19553826490342752,
      "grad_norm": 0.13540108501911163,
      "learning_rate": 0.001,
      "loss": 2.3262979984283447,
      "step": 1920
    },
    {
      "epoch": 0.19574195059603525,
      "grad_norm": 0.17047101259231567,
      "learning_rate": 0.001,
      "loss": 2.2330522537231445,
      "step": 1922
    },
    {
      "epoch": 0.195945636288643,
      "grad_norm": 0.15342363715171814,
      "learning_rate": 0.001,
      "loss": 2.42228364944458,
      "step": 1924
    },
    {
      "epoch": 0.19614932198125073,
      "grad_norm": 0.20931433141231537,
      "learning_rate": 0.001,
      "loss": 2.2376961708068848,
      "step": 1926
    },
    {
      "epoch": 0.19635300767385847,
      "grad_norm": 0.13761970400810242,
      "learning_rate": 0.001,
      "loss": 2.3241734504699707,
      "step": 1928
    },
    {
      "epoch": 0.1965566933664662,
      "grad_norm": 0.10777649283409119,
      "learning_rate": 0.001,
      "loss": 2.2623276710510254,
      "step": 1930
    },
    {
      "epoch": 0.19676037905907395,
      "grad_norm": 0.19385802745819092,
      "learning_rate": 0.001,
      "loss": 2.126744270324707,
      "step": 1932
    },
    {
      "epoch": 0.19696406475168168,
      "grad_norm": 0.18003562092781067,
      "learning_rate": 0.001,
      "loss": 2.1843791007995605,
      "step": 1934
    },
    {
      "epoch": 0.1971677504442894,
      "grad_norm": 0.13125939667224884,
      "learning_rate": 0.001,
      "loss": 2.3373396396636963,
      "step": 1936
    },
    {
      "epoch": 0.19737143613689714,
      "grad_norm": 0.20455387234687805,
      "learning_rate": 0.001,
      "loss": 2.175537109375,
      "step": 1938
    },
    {
      "epoch": 0.1975751218295049,
      "grad_norm": 0.1598929464817047,
      "learning_rate": 0.001,
      "loss": 2.2872838973999023,
      "step": 1940
    },
    {
      "epoch": 0.19777880752211263,
      "grad_norm": 0.11708422005176544,
      "learning_rate": 0.001,
      "loss": 2.2122445106506348,
      "step": 1942
    },
    {
      "epoch": 0.19798249321472036,
      "grad_norm": 0.11887247115373611,
      "learning_rate": 0.001,
      "loss": 2.219590187072754,
      "step": 1944
    },
    {
      "epoch": 0.1981861789073281,
      "grad_norm": 0.15619325637817383,
      "learning_rate": 0.001,
      "loss": 2.268925189971924,
      "step": 1946
    },
    {
      "epoch": 0.19838986459993585,
      "grad_norm": 0.11586815118789673,
      "learning_rate": 0.001,
      "loss": 2.234114170074463,
      "step": 1948
    },
    {
      "epoch": 0.19859355029254358,
      "grad_norm": 0.11454984545707703,
      "learning_rate": 0.001,
      "loss": 2.309263229370117,
      "step": 1950
    },
    {
      "epoch": 0.1987972359851513,
      "grad_norm": 0.12585331499576569,
      "learning_rate": 0.001,
      "loss": 2.2635698318481445,
      "step": 1952
    },
    {
      "epoch": 0.19900092167775904,
      "grad_norm": 0.1262817680835724,
      "learning_rate": 0.001,
      "loss": 2.196417808532715,
      "step": 1954
    },
    {
      "epoch": 0.1992046073703668,
      "grad_norm": 0.22268332540988922,
      "learning_rate": 0.001,
      "loss": 2.3261237144470215,
      "step": 1956
    },
    {
      "epoch": 0.19940829306297453,
      "grad_norm": 0.12685826420783997,
      "learning_rate": 0.001,
      "loss": 2.114393711090088,
      "step": 1958
    },
    {
      "epoch": 0.19961197875558226,
      "grad_norm": 0.12530820071697235,
      "learning_rate": 0.001,
      "loss": 2.313584327697754,
      "step": 1960
    },
    {
      "epoch": 0.19981566444819,
      "grad_norm": 0.18668481707572937,
      "learning_rate": 0.001,
      "loss": 2.2548890113830566,
      "step": 1962
    },
    {
      "epoch": 0.20001935014079775,
      "grad_norm": 0.11121103167533875,
      "learning_rate": 0.001,
      "loss": 2.350402355194092,
      "step": 1964
    },
    {
      "epoch": 0.20022303583340548,
      "grad_norm": 0.1332654058933258,
      "learning_rate": 0.001,
      "loss": 2.3518030643463135,
      "step": 1966
    },
    {
      "epoch": 0.2004267215260132,
      "grad_norm": 0.13542044162750244,
      "learning_rate": 0.001,
      "loss": 2.362475872039795,
      "step": 1968
    },
    {
      "epoch": 0.20063040721862094,
      "grad_norm": 0.10843665152788162,
      "learning_rate": 0.001,
      "loss": 2.342334508895874,
      "step": 1970
    },
    {
      "epoch": 0.2008340929112287,
      "grad_norm": 0.14251796901226044,
      "learning_rate": 0.001,
      "loss": 2.289336919784546,
      "step": 1972
    },
    {
      "epoch": 0.20103777860383643,
      "grad_norm": 0.18810075521469116,
      "learning_rate": 0.001,
      "loss": 2.2337474822998047,
      "step": 1974
    },
    {
      "epoch": 0.20124146429644416,
      "grad_norm": 0.13369949162006378,
      "learning_rate": 0.001,
      "loss": 2.2457265853881836,
      "step": 1976
    },
    {
      "epoch": 0.20144514998905189,
      "grad_norm": 0.13221774995326996,
      "learning_rate": 0.001,
      "loss": 2.2601826190948486,
      "step": 1978
    },
    {
      "epoch": 0.20164883568165964,
      "grad_norm": 0.129745751619339,
      "learning_rate": 0.001,
      "loss": 2.415891170501709,
      "step": 1980
    },
    {
      "epoch": 0.20185252137426737,
      "grad_norm": 0.1163729876279831,
      "learning_rate": 0.001,
      "loss": 2.1851000785827637,
      "step": 1982
    },
    {
      "epoch": 0.2020562070668751,
      "grad_norm": 0.17379450798034668,
      "learning_rate": 0.001,
      "loss": 2.3865816593170166,
      "step": 1984
    },
    {
      "epoch": 0.20225989275948283,
      "grad_norm": 0.12209579348564148,
      "learning_rate": 0.001,
      "loss": 2.319667100906372,
      "step": 1986
    },
    {
      "epoch": 0.2024635784520906,
      "grad_norm": 0.15141528844833374,
      "learning_rate": 0.001,
      "loss": 2.3105239868164062,
      "step": 1988
    },
    {
      "epoch": 0.20266726414469832,
      "grad_norm": 0.13487720489501953,
      "learning_rate": 0.001,
      "loss": 2.3425450325012207,
      "step": 1990
    },
    {
      "epoch": 0.20287094983730605,
      "grad_norm": 0.11849024146795273,
      "learning_rate": 0.001,
      "loss": 2.327389717102051,
      "step": 1992
    },
    {
      "epoch": 0.20307463552991378,
      "grad_norm": 0.12061267346143723,
      "learning_rate": 0.001,
      "loss": 2.346524953842163,
      "step": 1994
    },
    {
      "epoch": 0.20327832122252154,
      "grad_norm": 0.14630869030952454,
      "learning_rate": 0.001,
      "loss": 2.277308464050293,
      "step": 1996
    },
    {
      "epoch": 0.20348200691512927,
      "grad_norm": 0.15986157953739166,
      "learning_rate": 0.001,
      "loss": 2.306529998779297,
      "step": 1998
    },
    {
      "epoch": 0.203685692607737,
      "grad_norm": 0.12603764235973358,
      "learning_rate": 0.001,
      "loss": 2.1614506244659424,
      "step": 2000
    },
    {
      "epoch": 0.20388937830034473,
      "grad_norm": 0.12749092280864716,
      "learning_rate": 0.001,
      "loss": 2.2275943756103516,
      "step": 2002
    },
    {
      "epoch": 0.20409306399295246,
      "grad_norm": 0.1311330646276474,
      "learning_rate": 0.001,
      "loss": 2.215738296508789,
      "step": 2004
    },
    {
      "epoch": 0.20429674968556022,
      "grad_norm": 0.10149648040533066,
      "learning_rate": 0.001,
      "loss": 2.3582026958465576,
      "step": 2006
    },
    {
      "epoch": 0.20450043537816795,
      "grad_norm": 0.1210395023226738,
      "learning_rate": 0.001,
      "loss": 2.2181241512298584,
      "step": 2008
    },
    {
      "epoch": 0.20470412107077568,
      "grad_norm": 0.13598351180553436,
      "learning_rate": 0.001,
      "loss": 2.167415142059326,
      "step": 2010
    },
    {
      "epoch": 0.2049078067633834,
      "grad_norm": 0.14506720006465912,
      "learning_rate": 0.001,
      "loss": 2.2836151123046875,
      "step": 2012
    },
    {
      "epoch": 0.20511149245599117,
      "grad_norm": 0.13037912547588348,
      "learning_rate": 0.001,
      "loss": 2.2564404010772705,
      "step": 2014
    },
    {
      "epoch": 0.2053151781485989,
      "grad_norm": 0.11640441417694092,
      "learning_rate": 0.001,
      "loss": 2.2599525451660156,
      "step": 2016
    },
    {
      "epoch": 0.20551886384120663,
      "grad_norm": 0.13348421454429626,
      "learning_rate": 0.001,
      "loss": 2.2407541275024414,
      "step": 2018
    },
    {
      "epoch": 0.20572254953381436,
      "grad_norm": 0.12687475979328156,
      "learning_rate": 0.001,
      "loss": 2.316174030303955,
      "step": 2020
    },
    {
      "epoch": 0.20592623522642212,
      "grad_norm": 0.13885043561458588,
      "learning_rate": 0.001,
      "loss": 2.2986114025115967,
      "step": 2022
    },
    {
      "epoch": 0.20612992091902985,
      "grad_norm": 0.12649931013584137,
      "learning_rate": 0.001,
      "loss": 2.3572874069213867,
      "step": 2024
    },
    {
      "epoch": 0.20633360661163758,
      "grad_norm": 0.13342230021953583,
      "learning_rate": 0.001,
      "loss": 2.0838122367858887,
      "step": 2026
    },
    {
      "epoch": 0.2065372923042453,
      "grad_norm": 0.13612042367458344,
      "learning_rate": 0.001,
      "loss": 2.2671127319335938,
      "step": 2028
    },
    {
      "epoch": 0.20674097799685306,
      "grad_norm": 0.14538002014160156,
      "learning_rate": 0.001,
      "loss": 2.2662739753723145,
      "step": 2030
    },
    {
      "epoch": 0.2069446636894608,
      "grad_norm": 0.12667714059352875,
      "learning_rate": 0.001,
      "loss": 2.2172627449035645,
      "step": 2032
    },
    {
      "epoch": 0.20714834938206853,
      "grad_norm": 0.12291267514228821,
      "learning_rate": 0.001,
      "loss": 2.1739583015441895,
      "step": 2034
    },
    {
      "epoch": 0.20735203507467626,
      "grad_norm": 0.11600970476865768,
      "learning_rate": 0.001,
      "loss": 2.2439427375793457,
      "step": 2036
    },
    {
      "epoch": 0.207555720767284,
      "grad_norm": 0.13937845826148987,
      "learning_rate": 0.001,
      "loss": 2.1982436180114746,
      "step": 2038
    },
    {
      "epoch": 0.20775940645989174,
      "grad_norm": 0.11657320708036423,
      "learning_rate": 0.001,
      "loss": 2.2145957946777344,
      "step": 2040
    },
    {
      "epoch": 0.20796309215249947,
      "grad_norm": 0.1060621589422226,
      "learning_rate": 0.001,
      "loss": 2.325204372406006,
      "step": 2042
    },
    {
      "epoch": 0.2081667778451072,
      "grad_norm": 0.14791932702064514,
      "learning_rate": 0.001,
      "loss": 2.3769519329071045,
      "step": 2044
    },
    {
      "epoch": 0.20837046353771496,
      "grad_norm": 0.16861262917518616,
      "learning_rate": 0.001,
      "loss": 2.255207061767578,
      "step": 2046
    },
    {
      "epoch": 0.2085741492303227,
      "grad_norm": 0.12092728167772293,
      "learning_rate": 0.001,
      "loss": 2.2588930130004883,
      "step": 2048
    },
    {
      "epoch": 0.20877783492293042,
      "grad_norm": 0.12038931250572205,
      "learning_rate": 0.001,
      "loss": 2.318528652191162,
      "step": 2050
    },
    {
      "epoch": 0.20898152061553815,
      "grad_norm": 0.12795990705490112,
      "learning_rate": 0.001,
      "loss": 2.304769992828369,
      "step": 2052
    },
    {
      "epoch": 0.2091852063081459,
      "grad_norm": 0.12296199053525925,
      "learning_rate": 0.001,
      "loss": 2.290985584259033,
      "step": 2054
    },
    {
      "epoch": 0.20938889200075364,
      "grad_norm": 0.15992136299610138,
      "learning_rate": 0.001,
      "loss": 2.3338170051574707,
      "step": 2056
    },
    {
      "epoch": 0.20959257769336137,
      "grad_norm": 0.17864419519901276,
      "learning_rate": 0.001,
      "loss": 2.1608638763427734,
      "step": 2058
    },
    {
      "epoch": 0.2097962633859691,
      "grad_norm": 0.13432948291301727,
      "learning_rate": 0.001,
      "loss": 2.2765345573425293,
      "step": 2060
    },
    {
      "epoch": 0.20999994907857686,
      "grad_norm": 0.2173418402671814,
      "learning_rate": 0.001,
      "loss": 2.237825870513916,
      "step": 2062
    },
    {
      "epoch": 0.2102036347711846,
      "grad_norm": 0.2099733054637909,
      "learning_rate": 0.001,
      "loss": 2.219252109527588,
      "step": 2064
    },
    {
      "epoch": 0.21040732046379232,
      "grad_norm": 0.12270800024271011,
      "learning_rate": 0.001,
      "loss": 2.1198623180389404,
      "step": 2066
    },
    {
      "epoch": 0.21061100615640005,
      "grad_norm": 0.1364654302597046,
      "learning_rate": 0.001,
      "loss": 2.3696980476379395,
      "step": 2068
    },
    {
      "epoch": 0.2108146918490078,
      "grad_norm": 0.22015328705310822,
      "learning_rate": 0.001,
      "loss": 2.193979263305664,
      "step": 2070
    },
    {
      "epoch": 0.21101837754161554,
      "grad_norm": 0.13526102900505066,
      "learning_rate": 0.001,
      "loss": 2.296517848968506,
      "step": 2072
    },
    {
      "epoch": 0.21122206323422327,
      "grad_norm": 0.11848942190408707,
      "learning_rate": 0.001,
      "loss": 2.171509265899658,
      "step": 2074
    },
    {
      "epoch": 0.211425748926831,
      "grad_norm": 0.1455782949924469,
      "learning_rate": 0.001,
      "loss": 2.246652126312256,
      "step": 2076
    },
    {
      "epoch": 0.21162943461943876,
      "grad_norm": 0.2020053267478943,
      "learning_rate": 0.001,
      "loss": 2.212353467941284,
      "step": 2078
    },
    {
      "epoch": 0.21183312031204649,
      "grad_norm": 0.14376763999462128,
      "learning_rate": 0.001,
      "loss": 2.4191670417785645,
      "step": 2080
    },
    {
      "epoch": 0.21203680600465422,
      "grad_norm": 0.19415634870529175,
      "learning_rate": 0.001,
      "loss": 2.2870001792907715,
      "step": 2082
    },
    {
      "epoch": 0.21224049169726195,
      "grad_norm": 0.18984787166118622,
      "learning_rate": 0.001,
      "loss": 2.220033645629883,
      "step": 2084
    },
    {
      "epoch": 0.2124441773898697,
      "grad_norm": 0.1426749974489212,
      "learning_rate": 0.001,
      "loss": 2.190608501434326,
      "step": 2086
    },
    {
      "epoch": 0.21264786308247743,
      "grad_norm": 0.1331908255815506,
      "learning_rate": 0.001,
      "loss": 2.175689458847046,
      "step": 2088
    },
    {
      "epoch": 0.21285154877508516,
      "grad_norm": 0.09822644293308258,
      "learning_rate": 0.001,
      "loss": 2.078017234802246,
      "step": 2090
    },
    {
      "epoch": 0.2130552344676929,
      "grad_norm": 0.11721331626176834,
      "learning_rate": 0.001,
      "loss": 2.1472275257110596,
      "step": 2092
    },
    {
      "epoch": 0.21325892016030065,
      "grad_norm": 0.13697774708271027,
      "learning_rate": 0.001,
      "loss": 2.119791269302368,
      "step": 2094
    },
    {
      "epoch": 0.21346260585290838,
      "grad_norm": 0.12550288438796997,
      "learning_rate": 0.001,
      "loss": 2.255849838256836,
      "step": 2096
    },
    {
      "epoch": 0.2136662915455161,
      "grad_norm": 0.1347886025905609,
      "learning_rate": 0.001,
      "loss": 2.1878976821899414,
      "step": 2098
    },
    {
      "epoch": 0.21386997723812384,
      "grad_norm": 0.15933649241924286,
      "learning_rate": 0.001,
      "loss": 2.376943826675415,
      "step": 2100
    },
    {
      "epoch": 0.2140736629307316,
      "grad_norm": 0.14798781275749207,
      "learning_rate": 0.001,
      "loss": 2.406372547149658,
      "step": 2102
    },
    {
      "epoch": 0.21427734862333933,
      "grad_norm": 0.11847440153360367,
      "learning_rate": 0.001,
      "loss": 2.2349047660827637,
      "step": 2104
    },
    {
      "epoch": 0.21448103431594706,
      "grad_norm": 0.14966922998428345,
      "learning_rate": 0.001,
      "loss": 2.199357271194458,
      "step": 2106
    },
    {
      "epoch": 0.2146847200085548,
      "grad_norm": 0.16093970835208893,
      "learning_rate": 0.001,
      "loss": 2.360997438430786,
      "step": 2108
    },
    {
      "epoch": 0.21488840570116255,
      "grad_norm": 0.16777126491069794,
      "learning_rate": 0.001,
      "loss": 2.127255916595459,
      "step": 2110
    },
    {
      "epoch": 0.21509209139377028,
      "grad_norm": 0.18444634974002838,
      "learning_rate": 0.001,
      "loss": 2.208926200866699,
      "step": 2112
    },
    {
      "epoch": 0.215295777086378,
      "grad_norm": 0.1910533308982849,
      "learning_rate": 0.001,
      "loss": 2.2258739471435547,
      "step": 2114
    },
    {
      "epoch": 0.21549946277898574,
      "grad_norm": 0.16046257317066193,
      "learning_rate": 0.001,
      "loss": 2.3034329414367676,
      "step": 2116
    },
    {
      "epoch": 0.2157031484715935,
      "grad_norm": 0.1671079695224762,
      "learning_rate": 0.001,
      "loss": 2.203211784362793,
      "step": 2118
    },
    {
      "epoch": 0.21590683416420123,
      "grad_norm": 0.13861027359962463,
      "learning_rate": 0.001,
      "loss": 2.215731620788574,
      "step": 2120
    },
    {
      "epoch": 0.21611051985680896,
      "grad_norm": 0.14537538588047028,
      "learning_rate": 0.001,
      "loss": 2.1655242443084717,
      "step": 2122
    },
    {
      "epoch": 0.2163142055494167,
      "grad_norm": 0.13189592957496643,
      "learning_rate": 0.001,
      "loss": 2.2886962890625,
      "step": 2124
    },
    {
      "epoch": 0.21651789124202442,
      "grad_norm": 0.12383344769477844,
      "learning_rate": 0.001,
      "loss": 2.3259522914886475,
      "step": 2126
    },
    {
      "epoch": 0.21672157693463218,
      "grad_norm": 0.13056810200214386,
      "learning_rate": 0.001,
      "loss": 2.2502012252807617,
      "step": 2128
    },
    {
      "epoch": 0.2169252626272399,
      "grad_norm": 0.17766277492046356,
      "learning_rate": 0.001,
      "loss": 2.281503438949585,
      "step": 2130
    },
    {
      "epoch": 0.21712894831984764,
      "grad_norm": 0.14903143048286438,
      "learning_rate": 0.001,
      "loss": 2.2841367721557617,
      "step": 2132
    },
    {
      "epoch": 0.21733263401245537,
      "grad_norm": 0.12021240592002869,
      "learning_rate": 0.001,
      "loss": 2.2492244243621826,
      "step": 2134
    },
    {
      "epoch": 0.21753631970506312,
      "grad_norm": 0.11126147955656052,
      "learning_rate": 0.001,
      "loss": 2.2749483585357666,
      "step": 2136
    },
    {
      "epoch": 0.21774000539767085,
      "grad_norm": 0.10774312913417816,
      "learning_rate": 0.001,
      "loss": 2.187044143676758,
      "step": 2138
    },
    {
      "epoch": 0.21794369109027859,
      "grad_norm": 0.1213623657822609,
      "learning_rate": 0.001,
      "loss": 2.2154085636138916,
      "step": 2140
    },
    {
      "epoch": 0.21814737678288632,
      "grad_norm": 0.14465981721878052,
      "learning_rate": 0.001,
      "loss": 2.1029224395751953,
      "step": 2142
    },
    {
      "epoch": 0.21835106247549407,
      "grad_norm": 0.10764388740062714,
      "learning_rate": 0.001,
      "loss": 2.2928218841552734,
      "step": 2144
    },
    {
      "epoch": 0.2185547481681018,
      "grad_norm": 0.1555708646774292,
      "learning_rate": 0.001,
      "loss": 2.3006021976470947,
      "step": 2146
    },
    {
      "epoch": 0.21875843386070953,
      "grad_norm": 0.13192640244960785,
      "learning_rate": 0.001,
      "loss": 2.152228355407715,
      "step": 2148
    },
    {
      "epoch": 0.21896211955331726,
      "grad_norm": 0.116376593708992,
      "learning_rate": 0.001,
      "loss": 2.290051221847534,
      "step": 2150
    },
    {
      "epoch": 0.21916580524592502,
      "grad_norm": 0.15078386664390564,
      "learning_rate": 0.001,
      "loss": 2.3026695251464844,
      "step": 2152
    },
    {
      "epoch": 0.21936949093853275,
      "grad_norm": 0.14644142985343933,
      "learning_rate": 0.001,
      "loss": 2.1502597332000732,
      "step": 2154
    },
    {
      "epoch": 0.21957317663114048,
      "grad_norm": 0.1318386048078537,
      "learning_rate": 0.001,
      "loss": 2.1251206398010254,
      "step": 2156
    },
    {
      "epoch": 0.2197768623237482,
      "grad_norm": 0.12190419435501099,
      "learning_rate": 0.001,
      "loss": 2.2428131103515625,
      "step": 2158
    },
    {
      "epoch": 0.21998054801635597,
      "grad_norm": 0.144894078373909,
      "learning_rate": 0.001,
      "loss": 2.134995937347412,
      "step": 2160
    },
    {
      "epoch": 0.2201842337089637,
      "grad_norm": 0.11810851097106934,
      "learning_rate": 0.001,
      "loss": 2.366105794906616,
      "step": 2162
    },
    {
      "epoch": 0.22038791940157143,
      "grad_norm": 0.14622870087623596,
      "learning_rate": 0.001,
      "loss": 2.210442066192627,
      "step": 2164
    },
    {
      "epoch": 0.22059160509417916,
      "grad_norm": 0.11801375448703766,
      "learning_rate": 0.001,
      "loss": 2.2921509742736816,
      "step": 2166
    },
    {
      "epoch": 0.22079529078678692,
      "grad_norm": 0.17492353916168213,
      "learning_rate": 0.001,
      "loss": 2.2734193801879883,
      "step": 2168
    },
    {
      "epoch": 0.22099897647939465,
      "grad_norm": 0.13734686374664307,
      "learning_rate": 0.001,
      "loss": 2.3175535202026367,
      "step": 2170
    },
    {
      "epoch": 0.22120266217200238,
      "grad_norm": 0.16846592724323273,
      "learning_rate": 0.001,
      "loss": 2.186642646789551,
      "step": 2172
    },
    {
      "epoch": 0.2214063478646101,
      "grad_norm": 0.13702936470508575,
      "learning_rate": 0.001,
      "loss": 2.229022979736328,
      "step": 2174
    },
    {
      "epoch": 0.22161003355721787,
      "grad_norm": 0.14587877690792084,
      "learning_rate": 0.001,
      "loss": 2.234549045562744,
      "step": 2176
    },
    {
      "epoch": 0.2218137192498256,
      "grad_norm": 0.13199906051158905,
      "learning_rate": 0.001,
      "loss": 2.260359287261963,
      "step": 2178
    },
    {
      "epoch": 0.22201740494243333,
      "grad_norm": 0.12365327775478363,
      "learning_rate": 0.001,
      "loss": 2.2838587760925293,
      "step": 2180
    },
    {
      "epoch": 0.22222109063504106,
      "grad_norm": 0.14833196997642517,
      "learning_rate": 0.001,
      "loss": 2.2439093589782715,
      "step": 2182
    },
    {
      "epoch": 0.22242477632764882,
      "grad_norm": 0.12094587832689285,
      "learning_rate": 0.001,
      "loss": 2.3266594409942627,
      "step": 2184
    },
    {
      "epoch": 0.22262846202025655,
      "grad_norm": 0.14344653487205505,
      "learning_rate": 0.001,
      "loss": 2.2581090927124023,
      "step": 2186
    },
    {
      "epoch": 0.22283214771286428,
      "grad_norm": 0.16069138050079346,
      "learning_rate": 0.001,
      "loss": 2.2626309394836426,
      "step": 2188
    },
    {
      "epoch": 0.223035833405472,
      "grad_norm": 0.1205589771270752,
      "learning_rate": 0.001,
      "loss": 2.2711329460144043,
      "step": 2190
    },
    {
      "epoch": 0.22323951909807976,
      "grad_norm": 0.13031691312789917,
      "learning_rate": 0.001,
      "loss": 2.270057201385498,
      "step": 2192
    },
    {
      "epoch": 0.2234432047906875,
      "grad_norm": 0.10588197410106659,
      "learning_rate": 0.001,
      "loss": 2.179744243621826,
      "step": 2194
    },
    {
      "epoch": 0.22364689048329522,
      "grad_norm": 0.11193032562732697,
      "learning_rate": 0.001,
      "loss": 2.2466163635253906,
      "step": 2196
    },
    {
      "epoch": 0.22385057617590295,
      "grad_norm": 0.1477038413286209,
      "learning_rate": 0.001,
      "loss": 2.1967272758483887,
      "step": 2198
    },
    {
      "epoch": 0.2240542618685107,
      "grad_norm": 0.14351655542850494,
      "learning_rate": 0.001,
      "loss": 2.2262516021728516,
      "step": 2200
    },
    {
      "epoch": 0.22425794756111844,
      "grad_norm": 0.1721118539571762,
      "learning_rate": 0.001,
      "loss": 2.2458155155181885,
      "step": 2202
    },
    {
      "epoch": 0.22446163325372617,
      "grad_norm": 0.10241930931806564,
      "learning_rate": 0.001,
      "loss": 2.2978739738464355,
      "step": 2204
    },
    {
      "epoch": 0.2246653189463339,
      "grad_norm": 0.16029687225818634,
      "learning_rate": 0.001,
      "loss": 2.17799711227417,
      "step": 2206
    },
    {
      "epoch": 0.22486900463894166,
      "grad_norm": 0.172994464635849,
      "learning_rate": 0.001,
      "loss": 2.331993579864502,
      "step": 2208
    },
    {
      "epoch": 0.2250726903315494,
      "grad_norm": 0.12230091542005539,
      "learning_rate": 0.001,
      "loss": 2.237166404724121,
      "step": 2210
    },
    {
      "epoch": 0.22527637602415712,
      "grad_norm": 0.14420680701732635,
      "learning_rate": 0.001,
      "loss": 2.3947150707244873,
      "step": 2212
    },
    {
      "epoch": 0.22548006171676485,
      "grad_norm": 0.1573219895362854,
      "learning_rate": 0.001,
      "loss": 2.278330087661743,
      "step": 2214
    },
    {
      "epoch": 0.2256837474093726,
      "grad_norm": 0.10855304449796677,
      "learning_rate": 0.001,
      "loss": 2.2428178787231445,
      "step": 2216
    },
    {
      "epoch": 0.22588743310198034,
      "grad_norm": 0.1076224222779274,
      "learning_rate": 0.001,
      "loss": 2.3099775314331055,
      "step": 2218
    },
    {
      "epoch": 0.22609111879458807,
      "grad_norm": 0.1327914148569107,
      "learning_rate": 0.001,
      "loss": 2.2350735664367676,
      "step": 2220
    },
    {
      "epoch": 0.2262948044871958,
      "grad_norm": 0.1599942445755005,
      "learning_rate": 0.001,
      "loss": 2.3115594387054443,
      "step": 2222
    },
    {
      "epoch": 0.22649849017980356,
      "grad_norm": 0.12186815589666367,
      "learning_rate": 0.001,
      "loss": 2.142502784729004,
      "step": 2224
    },
    {
      "epoch": 0.2267021758724113,
      "grad_norm": 0.11893671751022339,
      "learning_rate": 0.001,
      "loss": 2.2509889602661133,
      "step": 2226
    },
    {
      "epoch": 0.22690586156501902,
      "grad_norm": 0.12350459396839142,
      "learning_rate": 0.001,
      "loss": 2.241121530532837,
      "step": 2228
    },
    {
      "epoch": 0.22710954725762675,
      "grad_norm": 0.13305187225341797,
      "learning_rate": 0.001,
      "loss": 2.3953404426574707,
      "step": 2230
    },
    {
      "epoch": 0.2273132329502345,
      "grad_norm": 0.11953561753034592,
      "learning_rate": 0.001,
      "loss": 2.2393741607666016,
      "step": 2232
    },
    {
      "epoch": 0.22751691864284224,
      "grad_norm": 0.1264950931072235,
      "learning_rate": 0.001,
      "loss": 2.3280367851257324,
      "step": 2234
    },
    {
      "epoch": 0.22772060433544997,
      "grad_norm": 0.15629138052463531,
      "learning_rate": 0.001,
      "loss": 2.3063459396362305,
      "step": 2236
    },
    {
      "epoch": 0.2279242900280577,
      "grad_norm": 0.1287769228219986,
      "learning_rate": 0.001,
      "loss": 2.316145896911621,
      "step": 2238
    },
    {
      "epoch": 0.22812797572066545,
      "grad_norm": 0.10872452706098557,
      "learning_rate": 0.001,
      "loss": 2.0702037811279297,
      "step": 2240
    },
    {
      "epoch": 0.22833166141327318,
      "grad_norm": 0.13038697838783264,
      "learning_rate": 0.001,
      "loss": 2.2753920555114746,
      "step": 2242
    },
    {
      "epoch": 0.22853534710588091,
      "grad_norm": 0.12785492837429047,
      "learning_rate": 0.001,
      "loss": 2.1782665252685547,
      "step": 2244
    },
    {
      "epoch": 0.22873903279848865,
      "grad_norm": 0.1103779599070549,
      "learning_rate": 0.001,
      "loss": 2.1625418663024902,
      "step": 2246
    },
    {
      "epoch": 0.22894271849109638,
      "grad_norm": 0.1287870854139328,
      "learning_rate": 0.001,
      "loss": 2.1845321655273438,
      "step": 2248
    },
    {
      "epoch": 0.22914640418370413,
      "grad_norm": 0.1187158152461052,
      "learning_rate": 0.001,
      "loss": 2.261803388595581,
      "step": 2250
    },
    {
      "epoch": 0.22935008987631186,
      "grad_norm": 0.12079809606075287,
      "learning_rate": 0.001,
      "loss": 2.1906609535217285,
      "step": 2252
    },
    {
      "epoch": 0.2295537755689196,
      "grad_norm": 0.14951969683170319,
      "learning_rate": 0.001,
      "loss": 2.2447075843811035,
      "step": 2254
    },
    {
      "epoch": 0.22975746126152732,
      "grad_norm": 0.1196364238858223,
      "learning_rate": 0.001,
      "loss": 2.325568675994873,
      "step": 2256
    },
    {
      "epoch": 0.22996114695413508,
      "grad_norm": 0.1726703643798828,
      "learning_rate": 0.001,
      "loss": 2.339211940765381,
      "step": 2258
    },
    {
      "epoch": 0.2301648326467428,
      "grad_norm": 0.17774413526058197,
      "learning_rate": 0.001,
      "loss": 2.2623257637023926,
      "step": 2260
    },
    {
      "epoch": 0.23036851833935054,
      "grad_norm": 0.13105565309524536,
      "learning_rate": 0.001,
      "loss": 2.3285000324249268,
      "step": 2262
    },
    {
      "epoch": 0.23057220403195827,
      "grad_norm": 0.11584489792585373,
      "learning_rate": 0.001,
      "loss": 2.1701455116271973,
      "step": 2264
    },
    {
      "epoch": 0.23077588972456603,
      "grad_norm": 0.12426963448524475,
      "learning_rate": 0.001,
      "loss": 2.1661179065704346,
      "step": 2266
    },
    {
      "epoch": 0.23097957541717376,
      "grad_norm": 0.13072186708450317,
      "learning_rate": 0.001,
      "loss": 2.2079944610595703,
      "step": 2268
    },
    {
      "epoch": 0.2311832611097815,
      "grad_norm": 0.11987145245075226,
      "learning_rate": 0.001,
      "loss": 2.238759994506836,
      "step": 2270
    },
    {
      "epoch": 0.23138694680238922,
      "grad_norm": 0.1274658888578415,
      "learning_rate": 0.001,
      "loss": 2.2876806259155273,
      "step": 2272
    },
    {
      "epoch": 0.23159063249499698,
      "grad_norm": 0.13418513536453247,
      "learning_rate": 0.001,
      "loss": 2.3080406188964844,
      "step": 2274
    },
    {
      "epoch": 0.2317943181876047,
      "grad_norm": 0.11713685095310211,
      "learning_rate": 0.001,
      "loss": 2.122743606567383,
      "step": 2276
    },
    {
      "epoch": 0.23199800388021244,
      "grad_norm": 0.12808719277381897,
      "learning_rate": 0.001,
      "loss": 2.261695384979248,
      "step": 2278
    },
    {
      "epoch": 0.23220168957282017,
      "grad_norm": 0.14203794300556183,
      "learning_rate": 0.001,
      "loss": 2.342231273651123,
      "step": 2280
    },
    {
      "epoch": 0.23240537526542793,
      "grad_norm": 0.12997564673423767,
      "learning_rate": 0.001,
      "loss": 2.220278263092041,
      "step": 2282
    },
    {
      "epoch": 0.23260906095803566,
      "grad_norm": 0.14943812787532806,
      "learning_rate": 0.001,
      "loss": 2.249023914337158,
      "step": 2284
    },
    {
      "epoch": 0.2328127466506434,
      "grad_norm": 0.19298604130744934,
      "learning_rate": 0.001,
      "loss": 2.246429443359375,
      "step": 2286
    },
    {
      "epoch": 0.23301643234325112,
      "grad_norm": 0.12968875467777252,
      "learning_rate": 0.001,
      "loss": 2.220607042312622,
      "step": 2288
    },
    {
      "epoch": 0.23322011803585888,
      "grad_norm": 0.1168389841914177,
      "learning_rate": 0.001,
      "loss": 2.2542152404785156,
      "step": 2290
    },
    {
      "epoch": 0.2334238037284666,
      "grad_norm": 0.17614257335662842,
      "learning_rate": 0.001,
      "loss": 2.2829504013061523,
      "step": 2292
    },
    {
      "epoch": 0.23362748942107434,
      "grad_norm": 0.15027718245983124,
      "learning_rate": 0.001,
      "loss": 2.3181204795837402,
      "step": 2294
    },
    {
      "epoch": 0.23383117511368207,
      "grad_norm": 0.13256654143333435,
      "learning_rate": 0.001,
      "loss": 2.176738739013672,
      "step": 2296
    },
    {
      "epoch": 0.23403486080628982,
      "grad_norm": 0.1239066869020462,
      "learning_rate": 0.001,
      "loss": 2.3277997970581055,
      "step": 2298
    },
    {
      "epoch": 0.23423854649889755,
      "grad_norm": 0.1111677810549736,
      "learning_rate": 0.001,
      "loss": 2.2314040660858154,
      "step": 2300
    },
    {
      "epoch": 0.23444223219150528,
      "grad_norm": 0.1241074949502945,
      "learning_rate": 0.001,
      "loss": 2.3606104850769043,
      "step": 2302
    },
    {
      "epoch": 0.23464591788411301,
      "grad_norm": 0.12904836237430573,
      "learning_rate": 0.001,
      "loss": 2.134746551513672,
      "step": 2304
    },
    {
      "epoch": 0.23484960357672077,
      "grad_norm": 0.10937532037496567,
      "learning_rate": 0.001,
      "loss": 2.235872507095337,
      "step": 2306
    },
    {
      "epoch": 0.2350532892693285,
      "grad_norm": 0.11524411290884018,
      "learning_rate": 0.001,
      "loss": 2.2225921154022217,
      "step": 2308
    },
    {
      "epoch": 0.23525697496193623,
      "grad_norm": 0.1091182678937912,
      "learning_rate": 0.001,
      "loss": 2.249697208404541,
      "step": 2310
    },
    {
      "epoch": 0.23546066065454396,
      "grad_norm": 0.09452769160270691,
      "learning_rate": 0.001,
      "loss": 2.1702475547790527,
      "step": 2312
    },
    {
      "epoch": 0.23566434634715172,
      "grad_norm": 0.1309838891029358,
      "learning_rate": 0.001,
      "loss": 2.3923091888427734,
      "step": 2314
    },
    {
      "epoch": 0.23586803203975945,
      "grad_norm": 0.13221074640750885,
      "learning_rate": 0.001,
      "loss": 2.2691121101379395,
      "step": 2316
    },
    {
      "epoch": 0.23607171773236718,
      "grad_norm": 0.13891996443271637,
      "learning_rate": 0.001,
      "loss": 2.258054256439209,
      "step": 2318
    },
    {
      "epoch": 0.2362754034249749,
      "grad_norm": 0.12255372107028961,
      "learning_rate": 0.001,
      "loss": 2.3149542808532715,
      "step": 2320
    },
    {
      "epoch": 0.23647908911758267,
      "grad_norm": 0.13695189356803894,
      "learning_rate": 0.001,
      "loss": 2.2831039428710938,
      "step": 2322
    },
    {
      "epoch": 0.2366827748101904,
      "grad_norm": 0.12873858213424683,
      "learning_rate": 0.001,
      "loss": 2.1129136085510254,
      "step": 2324
    },
    {
      "epoch": 0.23688646050279813,
      "grad_norm": 0.11141201853752136,
      "learning_rate": 0.001,
      "loss": 2.1697378158569336,
      "step": 2326
    },
    {
      "epoch": 0.23709014619540586,
      "grad_norm": 0.11890513449907303,
      "learning_rate": 0.001,
      "loss": 2.2880191802978516,
      "step": 2328
    },
    {
      "epoch": 0.23729383188801362,
      "grad_norm": 0.10421814769506454,
      "learning_rate": 0.001,
      "loss": 2.152355194091797,
      "step": 2330
    },
    {
      "epoch": 0.23749751758062135,
      "grad_norm": 0.14245902001857758,
      "learning_rate": 0.001,
      "loss": 2.274724006652832,
      "step": 2332
    },
    {
      "epoch": 0.23770120327322908,
      "grad_norm": 0.1268816888332367,
      "learning_rate": 0.001,
      "loss": 2.3017592430114746,
      "step": 2334
    },
    {
      "epoch": 0.2379048889658368,
      "grad_norm": 0.14744152128696442,
      "learning_rate": 0.001,
      "loss": 2.2755537033081055,
      "step": 2336
    },
    {
      "epoch": 0.23810857465844457,
      "grad_norm": 0.14433081448078156,
      "learning_rate": 0.001,
      "loss": 2.187593936920166,
      "step": 2338
    },
    {
      "epoch": 0.2383122603510523,
      "grad_norm": 0.134014293551445,
      "learning_rate": 0.001,
      "loss": 2.13706636428833,
      "step": 2340
    },
    {
      "epoch": 0.23851594604366003,
      "grad_norm": 0.1468515694141388,
      "learning_rate": 0.001,
      "loss": 2.2837612628936768,
      "step": 2342
    },
    {
      "epoch": 0.23871963173626776,
      "grad_norm": 0.13836735486984253,
      "learning_rate": 0.001,
      "loss": 2.2482898235321045,
      "step": 2344
    },
    {
      "epoch": 0.23892331742887551,
      "grad_norm": 0.10638038069009781,
      "learning_rate": 0.001,
      "loss": 2.148932695388794,
      "step": 2346
    },
    {
      "epoch": 0.23912700312148324,
      "grad_norm": 0.16863828897476196,
      "learning_rate": 0.001,
      "loss": 2.2510690689086914,
      "step": 2348
    },
    {
      "epoch": 0.23933068881409097,
      "grad_norm": 0.12372192740440369,
      "learning_rate": 0.001,
      "loss": 2.160264492034912,
      "step": 2350
    },
    {
      "epoch": 0.2395343745066987,
      "grad_norm": 0.12689048051834106,
      "learning_rate": 0.001,
      "loss": 2.220759868621826,
      "step": 2352
    },
    {
      "epoch": 0.23973806019930646,
      "grad_norm": 0.1720622330904007,
      "learning_rate": 0.001,
      "loss": 2.120234251022339,
      "step": 2354
    },
    {
      "epoch": 0.2399417458919142,
      "grad_norm": 0.16328337788581848,
      "learning_rate": 0.001,
      "loss": 2.2486424446105957,
      "step": 2356
    },
    {
      "epoch": 0.24014543158452192,
      "grad_norm": 0.13333815336227417,
      "learning_rate": 0.001,
      "loss": 2.2202744483947754,
      "step": 2358
    },
    {
      "epoch": 0.24034911727712965,
      "grad_norm": 0.11074129492044449,
      "learning_rate": 0.001,
      "loss": 2.1337339878082275,
      "step": 2360
    },
    {
      "epoch": 0.2405528029697374,
      "grad_norm": 0.11352615058422089,
      "learning_rate": 0.001,
      "loss": 2.1236324310302734,
      "step": 2362
    },
    {
      "epoch": 0.24075648866234514,
      "grad_norm": 0.12343548983335495,
      "learning_rate": 0.001,
      "loss": 2.2955195903778076,
      "step": 2364
    },
    {
      "epoch": 0.24096017435495287,
      "grad_norm": 0.1231144592165947,
      "learning_rate": 0.001,
      "loss": 2.1093971729278564,
      "step": 2366
    },
    {
      "epoch": 0.2411638600475606,
      "grad_norm": 0.11531911045312881,
      "learning_rate": 0.001,
      "loss": 2.217890501022339,
      "step": 2368
    },
    {
      "epoch": 0.24136754574016836,
      "grad_norm": 0.12205381691455841,
      "learning_rate": 0.001,
      "loss": 2.039658784866333,
      "step": 2370
    },
    {
      "epoch": 0.2415712314327761,
      "grad_norm": 0.1533801555633545,
      "learning_rate": 0.001,
      "loss": 2.228492259979248,
      "step": 2372
    },
    {
      "epoch": 0.24177491712538382,
      "grad_norm": 0.12235800921916962,
      "learning_rate": 0.001,
      "loss": 2.305098056793213,
      "step": 2374
    },
    {
      "epoch": 0.24197860281799155,
      "grad_norm": 0.13657255470752716,
      "learning_rate": 0.001,
      "loss": 2.1344902515411377,
      "step": 2376
    },
    {
      "epoch": 0.24218228851059928,
      "grad_norm": 0.15733499825000763,
      "learning_rate": 0.001,
      "loss": 2.3524599075317383,
      "step": 2378
    },
    {
      "epoch": 0.24238597420320704,
      "grad_norm": 0.14680233597755432,
      "learning_rate": 0.001,
      "loss": 2.272952079772949,
      "step": 2380
    },
    {
      "epoch": 0.24258965989581477,
      "grad_norm": 0.12580932676792145,
      "learning_rate": 0.001,
      "loss": 2.330237627029419,
      "step": 2382
    },
    {
      "epoch": 0.2427933455884225,
      "grad_norm": 0.11925962567329407,
      "learning_rate": 0.001,
      "loss": 2.2972898483276367,
      "step": 2384
    },
    {
      "epoch": 0.24299703128103023,
      "grad_norm": 0.10948441177606583,
      "learning_rate": 0.001,
      "loss": 2.2953498363494873,
      "step": 2386
    },
    {
      "epoch": 0.243200716973638,
      "grad_norm": 0.13907292485237122,
      "learning_rate": 0.001,
      "loss": 2.3546230792999268,
      "step": 2388
    },
    {
      "epoch": 0.24340440266624572,
      "grad_norm": 0.17448817193508148,
      "learning_rate": 0.001,
      "loss": 2.4091057777404785,
      "step": 2390
    },
    {
      "epoch": 0.24360808835885345,
      "grad_norm": 0.13512557744979858,
      "learning_rate": 0.001,
      "loss": 2.1673364639282227,
      "step": 2392
    },
    {
      "epoch": 0.24381177405146118,
      "grad_norm": 0.12703931331634521,
      "learning_rate": 0.001,
      "loss": 2.3570284843444824,
      "step": 2394
    },
    {
      "epoch": 0.24401545974406894,
      "grad_norm": 0.1330791711807251,
      "learning_rate": 0.001,
      "loss": 2.3375802040100098,
      "step": 2396
    },
    {
      "epoch": 0.24421914543667667,
      "grad_norm": 0.12423437088727951,
      "learning_rate": 0.001,
      "loss": 2.2006168365478516,
      "step": 2398
    },
    {
      "epoch": 0.2444228311292844,
      "grad_norm": 0.116011843085289,
      "learning_rate": 0.001,
      "loss": 2.3071022033691406,
      "step": 2400
    },
    {
      "epoch": 0.24462651682189213,
      "grad_norm": 0.12046250700950623,
      "learning_rate": 0.001,
      "loss": 2.2343902587890625,
      "step": 2402
    },
    {
      "epoch": 0.24483020251449988,
      "grad_norm": 0.1241685226559639,
      "learning_rate": 0.001,
      "loss": 2.3002641201019287,
      "step": 2404
    },
    {
      "epoch": 0.24503388820710761,
      "grad_norm": 0.13112609088420868,
      "learning_rate": 0.001,
      "loss": 2.2365260124206543,
      "step": 2406
    },
    {
      "epoch": 0.24523757389971534,
      "grad_norm": 0.12894590198993683,
      "learning_rate": 0.001,
      "loss": 2.2513551712036133,
      "step": 2408
    },
    {
      "epoch": 0.24544125959232307,
      "grad_norm": 0.1860557645559311,
      "learning_rate": 0.001,
      "loss": 2.249741554260254,
      "step": 2410
    },
    {
      "epoch": 0.24564494528493083,
      "grad_norm": 0.17300473153591156,
      "learning_rate": 0.001,
      "loss": 2.2818374633789062,
      "step": 2412
    },
    {
      "epoch": 0.24584863097753856,
      "grad_norm": 0.1064365804195404,
      "learning_rate": 0.001,
      "loss": 2.167818307876587,
      "step": 2414
    },
    {
      "epoch": 0.2460523166701463,
      "grad_norm": 0.11387424916028976,
      "learning_rate": 0.001,
      "loss": 2.2182374000549316,
      "step": 2416
    },
    {
      "epoch": 0.24625600236275402,
      "grad_norm": 0.11281440407037735,
      "learning_rate": 0.001,
      "loss": 2.2971351146698,
      "step": 2418
    },
    {
      "epoch": 0.24645968805536178,
      "grad_norm": 0.14176355302333832,
      "learning_rate": 0.001,
      "loss": 2.28466796875,
      "step": 2420
    },
    {
      "epoch": 0.2466633737479695,
      "grad_norm": 0.12202437967061996,
      "learning_rate": 0.001,
      "loss": 2.2136001586914062,
      "step": 2422
    },
    {
      "epoch": 0.24686705944057724,
      "grad_norm": 0.1930209994316101,
      "learning_rate": 0.001,
      "loss": 2.3153481483459473,
      "step": 2424
    },
    {
      "epoch": 0.24707074513318497,
      "grad_norm": 0.1590680330991745,
      "learning_rate": 0.001,
      "loss": 2.169844627380371,
      "step": 2426
    },
    {
      "epoch": 0.24727443082579273,
      "grad_norm": 0.11763955652713776,
      "learning_rate": 0.001,
      "loss": 2.2070846557617188,
      "step": 2428
    },
    {
      "epoch": 0.24747811651840046,
      "grad_norm": 0.15771201252937317,
      "learning_rate": 0.001,
      "loss": 2.281949996948242,
      "step": 2430
    },
    {
      "epoch": 0.2476818022110082,
      "grad_norm": 0.11038631945848465,
      "learning_rate": 0.001,
      "loss": 2.1789164543151855,
      "step": 2432
    },
    {
      "epoch": 0.24788548790361592,
      "grad_norm": 0.1112900897860527,
      "learning_rate": 0.001,
      "loss": 2.150944232940674,
      "step": 2434
    },
    {
      "epoch": 0.24808917359622368,
      "grad_norm": 0.10558906197547913,
      "learning_rate": 0.001,
      "loss": 2.3076682090759277,
      "step": 2436
    },
    {
      "epoch": 0.2482928592888314,
      "grad_norm": 0.14115585386753082,
      "learning_rate": 0.001,
      "loss": 2.2640576362609863,
      "step": 2438
    },
    {
      "epoch": 0.24849654498143914,
      "grad_norm": 0.1521279513835907,
      "learning_rate": 0.001,
      "loss": 2.297089099884033,
      "step": 2440
    },
    {
      "epoch": 0.24870023067404687,
      "grad_norm": 0.10184413194656372,
      "learning_rate": 0.001,
      "loss": 2.2579843997955322,
      "step": 2442
    },
    {
      "epoch": 0.24890391636665463,
      "grad_norm": 0.10858550667762756,
      "learning_rate": 0.001,
      "loss": 2.1578266620635986,
      "step": 2444
    },
    {
      "epoch": 0.24910760205926236,
      "grad_norm": 0.133353129029274,
      "learning_rate": 0.001,
      "loss": 2.2211272716522217,
      "step": 2446
    },
    {
      "epoch": 0.2493112877518701,
      "grad_norm": 0.13719642162322998,
      "learning_rate": 0.001,
      "loss": 2.223576545715332,
      "step": 2448
    },
    {
      "epoch": 0.24951497344447782,
      "grad_norm": 0.14417602121829987,
      "learning_rate": 0.001,
      "loss": 2.232917308807373,
      "step": 2450
    },
    {
      "epoch": 0.24971865913708557,
      "grad_norm": 0.19275414943695068,
      "learning_rate": 0.001,
      "loss": 2.2598061561584473,
      "step": 2452
    },
    {
      "epoch": 0.2499223448296933,
      "grad_norm": 0.11563865095376968,
      "learning_rate": 0.001,
      "loss": 2.0549275875091553,
      "step": 2454
    },
    {
      "epoch": 0.25012603052230103,
      "grad_norm": 0.12827977538108826,
      "learning_rate": 0.001,
      "loss": 2.158107280731201,
      "step": 2456
    },
    {
      "epoch": 0.2503297162149088,
      "grad_norm": 0.1570153385400772,
      "learning_rate": 0.001,
      "loss": 2.2887916564941406,
      "step": 2458
    },
    {
      "epoch": 0.2505334019075165,
      "grad_norm": 0.1329558789730072,
      "learning_rate": 0.001,
      "loss": 2.3444855213165283,
      "step": 2460
    },
    {
      "epoch": 0.25073708760012425,
      "grad_norm": 0.1116601973772049,
      "learning_rate": 0.001,
      "loss": 2.202824354171753,
      "step": 2462
    },
    {
      "epoch": 0.250940773292732,
      "grad_norm": 0.13204801082611084,
      "learning_rate": 0.001,
      "loss": 2.1498966217041016,
      "step": 2464
    },
    {
      "epoch": 0.2511444589853397,
      "grad_norm": 0.1589539498090744,
      "learning_rate": 0.001,
      "loss": 2.1422924995422363,
      "step": 2466
    },
    {
      "epoch": 0.25134814467794747,
      "grad_norm": 0.11851079761981964,
      "learning_rate": 0.001,
      "loss": 2.159518003463745,
      "step": 2468
    },
    {
      "epoch": 0.2515518303705552,
      "grad_norm": 0.14626146852970123,
      "learning_rate": 0.001,
      "loss": 2.22283673286438,
      "step": 2470
    },
    {
      "epoch": 0.25175551606316293,
      "grad_norm": 0.14713218808174133,
      "learning_rate": 0.001,
      "loss": 2.311976909637451,
      "step": 2472
    },
    {
      "epoch": 0.2519592017557707,
      "grad_norm": 0.1259654015302658,
      "learning_rate": 0.001,
      "loss": 2.3853487968444824,
      "step": 2474
    },
    {
      "epoch": 0.2521628874483784,
      "grad_norm": 0.10935270041227341,
      "learning_rate": 0.001,
      "loss": 2.216487407684326,
      "step": 2476
    },
    {
      "epoch": 0.25236657314098615,
      "grad_norm": 0.11416438221931458,
      "learning_rate": 0.001,
      "loss": 2.1839990615844727,
      "step": 2478
    },
    {
      "epoch": 0.2525702588335939,
      "grad_norm": 0.11564593762159348,
      "learning_rate": 0.001,
      "loss": 2.2234673500061035,
      "step": 2480
    },
    {
      "epoch": 0.2527739445262016,
      "grad_norm": 0.15390987694263458,
      "learning_rate": 0.001,
      "loss": 2.238325834274292,
      "step": 2482
    },
    {
      "epoch": 0.25297763021880937,
      "grad_norm": 0.12190257012844086,
      "learning_rate": 0.001,
      "loss": 2.231517791748047,
      "step": 2484
    },
    {
      "epoch": 0.25318131591141707,
      "grad_norm": 0.12369199097156525,
      "learning_rate": 0.001,
      "loss": 2.1562795639038086,
      "step": 2486
    },
    {
      "epoch": 0.25338500160402483,
      "grad_norm": 0.1637447327375412,
      "learning_rate": 0.001,
      "loss": 2.193091630935669,
      "step": 2488
    },
    {
      "epoch": 0.2535886872966326,
      "grad_norm": 0.11535680294036865,
      "learning_rate": 0.001,
      "loss": 2.1275601387023926,
      "step": 2490
    },
    {
      "epoch": 0.2537923729892403,
      "grad_norm": 0.1322251260280609,
      "learning_rate": 0.001,
      "loss": 2.124635696411133,
      "step": 2492
    },
    {
      "epoch": 0.25399605868184805,
      "grad_norm": 0.12674516439437866,
      "learning_rate": 0.001,
      "loss": 2.380678653717041,
      "step": 2494
    },
    {
      "epoch": 0.2541997443744558,
      "grad_norm": 0.1229044646024704,
      "learning_rate": 0.001,
      "loss": 2.294072389602661,
      "step": 2496
    },
    {
      "epoch": 0.2544034300670635,
      "grad_norm": 0.1499728113412857,
      "learning_rate": 0.001,
      "loss": 2.4008021354675293,
      "step": 2498
    },
    {
      "epoch": 0.25460711575967127,
      "grad_norm": 0.14515431225299835,
      "learning_rate": 0.001,
      "loss": 2.353992462158203,
      "step": 2500
    },
    {
      "epoch": 0.25481080145227897,
      "grad_norm": 0.1462417095899582,
      "learning_rate": 0.001,
      "loss": 2.164869785308838,
      "step": 2502
    },
    {
      "epoch": 0.2550144871448867,
      "grad_norm": 0.11557002365589142,
      "learning_rate": 0.001,
      "loss": 2.2584660053253174,
      "step": 2504
    },
    {
      "epoch": 0.2552181728374945,
      "grad_norm": 0.14306089282035828,
      "learning_rate": 0.001,
      "loss": 2.2077407836914062,
      "step": 2506
    },
    {
      "epoch": 0.2554218585301022,
      "grad_norm": 0.15302926301956177,
      "learning_rate": 0.001,
      "loss": 2.244406223297119,
      "step": 2508
    },
    {
      "epoch": 0.25562554422270994,
      "grad_norm": 0.12399857491254807,
      "learning_rate": 0.001,
      "loss": 2.2401511669158936,
      "step": 2510
    },
    {
      "epoch": 0.25582922991531765,
      "grad_norm": 0.12772302329540253,
      "learning_rate": 0.001,
      "loss": 2.099644184112549,
      "step": 2512
    },
    {
      "epoch": 0.2560329156079254,
      "grad_norm": 0.1432444006204605,
      "learning_rate": 0.001,
      "loss": 2.3525986671447754,
      "step": 2514
    },
    {
      "epoch": 0.25623660130053316,
      "grad_norm": 0.13702404499053955,
      "learning_rate": 0.001,
      "loss": 2.3260555267333984,
      "step": 2516
    },
    {
      "epoch": 0.25644028699314086,
      "grad_norm": 0.1140730157494545,
      "learning_rate": 0.001,
      "loss": 2.2621805667877197,
      "step": 2518
    },
    {
      "epoch": 0.2566439726857486,
      "grad_norm": 0.11614888906478882,
      "learning_rate": 0.001,
      "loss": 2.1644275188446045,
      "step": 2520
    },
    {
      "epoch": 0.2568476583783564,
      "grad_norm": 0.1176842525601387,
      "learning_rate": 0.001,
      "loss": 2.1955811977386475,
      "step": 2522
    },
    {
      "epoch": 0.2570513440709641,
      "grad_norm": 0.11319168657064438,
      "learning_rate": 0.001,
      "loss": 2.209890842437744,
      "step": 2524
    },
    {
      "epoch": 0.25725502976357184,
      "grad_norm": 0.13947781920433044,
      "learning_rate": 0.001,
      "loss": 2.2336792945861816,
      "step": 2526
    },
    {
      "epoch": 0.25745871545617954,
      "grad_norm": 0.2605893313884735,
      "learning_rate": 0.001,
      "loss": 2.3205220699310303,
      "step": 2528
    },
    {
      "epoch": 0.2576624011487873,
      "grad_norm": 0.11903031170368195,
      "learning_rate": 0.001,
      "loss": 2.265829563140869,
      "step": 2530
    },
    {
      "epoch": 0.25786608684139506,
      "grad_norm": 0.15078675746917725,
      "learning_rate": 0.001,
      "loss": 2.216618061065674,
      "step": 2532
    },
    {
      "epoch": 0.25806977253400276,
      "grad_norm": 0.1161879375576973,
      "learning_rate": 0.001,
      "loss": 2.296602725982666,
      "step": 2534
    },
    {
      "epoch": 0.2582734582266105,
      "grad_norm": 0.12012317031621933,
      "learning_rate": 0.001,
      "loss": 2.2389254570007324,
      "step": 2536
    },
    {
      "epoch": 0.2584771439192183,
      "grad_norm": 0.1462898701429367,
      "learning_rate": 0.001,
      "loss": 2.378206491470337,
      "step": 2538
    },
    {
      "epoch": 0.258680829611826,
      "grad_norm": 0.11691492795944214,
      "learning_rate": 0.001,
      "loss": 2.207993984222412,
      "step": 2540
    },
    {
      "epoch": 0.25888451530443374,
      "grad_norm": 0.11594048887491226,
      "learning_rate": 0.001,
      "loss": 2.3116958141326904,
      "step": 2542
    },
    {
      "epoch": 0.25908820099704144,
      "grad_norm": 0.1071212887763977,
      "learning_rate": 0.001,
      "loss": 2.2215981483459473,
      "step": 2544
    },
    {
      "epoch": 0.2592918866896492,
      "grad_norm": 0.13799069821834564,
      "learning_rate": 0.001,
      "loss": 2.171743869781494,
      "step": 2546
    },
    {
      "epoch": 0.25949557238225696,
      "grad_norm": 0.10537232458591461,
      "learning_rate": 0.001,
      "loss": 2.177974224090576,
      "step": 2548
    },
    {
      "epoch": 0.25969925807486466,
      "grad_norm": 0.16781821846961975,
      "learning_rate": 0.001,
      "loss": 2.211303472518921,
      "step": 2550
    },
    {
      "epoch": 0.2599029437674724,
      "grad_norm": 0.12037919461727142,
      "learning_rate": 0.001,
      "loss": 2.1052298545837402,
      "step": 2552
    },
    {
      "epoch": 0.2601066294600802,
      "grad_norm": 0.11407136917114258,
      "learning_rate": 0.001,
      "loss": 2.2459144592285156,
      "step": 2554
    },
    {
      "epoch": 0.2603103151526879,
      "grad_norm": 0.12179651856422424,
      "learning_rate": 0.001,
      "loss": 2.2908711433410645,
      "step": 2556
    },
    {
      "epoch": 0.26051400084529563,
      "grad_norm": 0.13243801891803741,
      "learning_rate": 0.001,
      "loss": 2.249368190765381,
      "step": 2558
    },
    {
      "epoch": 0.26071768653790334,
      "grad_norm": 0.14861999452114105,
      "learning_rate": 0.001,
      "loss": 2.1337122917175293,
      "step": 2560
    },
    {
      "epoch": 0.2609213722305111,
      "grad_norm": 0.11577431112527847,
      "learning_rate": 0.001,
      "loss": 2.3043885231018066,
      "step": 2562
    },
    {
      "epoch": 0.26112505792311885,
      "grad_norm": 0.14580883085727692,
      "learning_rate": 0.001,
      "loss": 2.2461435794830322,
      "step": 2564
    },
    {
      "epoch": 0.26132874361572656,
      "grad_norm": 0.1088109016418457,
      "learning_rate": 0.001,
      "loss": 2.1832947731018066,
      "step": 2566
    },
    {
      "epoch": 0.2615324293083343,
      "grad_norm": 0.11399704217910767,
      "learning_rate": 0.001,
      "loss": 2.242194175720215,
      "step": 2568
    },
    {
      "epoch": 0.26173611500094207,
      "grad_norm": 0.1680019348859787,
      "learning_rate": 0.001,
      "loss": 2.2239298820495605,
      "step": 2570
    },
    {
      "epoch": 0.2619398006935498,
      "grad_norm": 0.14158180356025696,
      "learning_rate": 0.001,
      "loss": 2.2023544311523438,
      "step": 2572
    },
    {
      "epoch": 0.26214348638615753,
      "grad_norm": 0.12665045261383057,
      "learning_rate": 0.001,
      "loss": 2.3767213821411133,
      "step": 2574
    },
    {
      "epoch": 0.26234717207876523,
      "grad_norm": 0.1360526829957962,
      "learning_rate": 0.001,
      "loss": 2.1772518157958984,
      "step": 2576
    },
    {
      "epoch": 0.262550857771373,
      "grad_norm": 0.11608131974935532,
      "learning_rate": 0.001,
      "loss": 2.171983480453491,
      "step": 2578
    },
    {
      "epoch": 0.26275454346398075,
      "grad_norm": 0.11201854050159454,
      "learning_rate": 0.001,
      "loss": 2.215081214904785,
      "step": 2580
    },
    {
      "epoch": 0.26295822915658845,
      "grad_norm": 0.16245776414871216,
      "learning_rate": 0.001,
      "loss": 2.1455609798431396,
      "step": 2582
    },
    {
      "epoch": 0.2631619148491962,
      "grad_norm": 0.10045953840017319,
      "learning_rate": 0.001,
      "loss": 2.0799124240875244,
      "step": 2584
    },
    {
      "epoch": 0.26336560054180397,
      "grad_norm": 0.11225317418575287,
      "learning_rate": 0.001,
      "loss": 2.281095027923584,
      "step": 2586
    },
    {
      "epoch": 0.26356928623441167,
      "grad_norm": 0.11138707399368286,
      "learning_rate": 0.001,
      "loss": 2.293059825897217,
      "step": 2588
    },
    {
      "epoch": 0.26377297192701943,
      "grad_norm": 0.13241741061210632,
      "learning_rate": 0.001,
      "loss": 2.152883529663086,
      "step": 2590
    },
    {
      "epoch": 0.26397665761962713,
      "grad_norm": 0.14563027024269104,
      "learning_rate": 0.001,
      "loss": 2.2672085762023926,
      "step": 2592
    },
    {
      "epoch": 0.2641803433122349,
      "grad_norm": 0.13751880824565887,
      "learning_rate": 0.001,
      "loss": 2.278979778289795,
      "step": 2594
    },
    {
      "epoch": 0.26438402900484265,
      "grad_norm": 0.13011743128299713,
      "learning_rate": 0.001,
      "loss": 2.1860084533691406,
      "step": 2596
    },
    {
      "epoch": 0.26458771469745035,
      "grad_norm": 0.11266366392374039,
      "learning_rate": 0.001,
      "loss": 2.1849236488342285,
      "step": 2598
    },
    {
      "epoch": 0.2647914003900581,
      "grad_norm": 0.127931147813797,
      "learning_rate": 0.001,
      "loss": 2.1793103218078613,
      "step": 2600
    },
    {
      "epoch": 0.26499508608266586,
      "grad_norm": 0.10890716314315796,
      "learning_rate": 0.001,
      "loss": 2.3062689304351807,
      "step": 2602
    },
    {
      "epoch": 0.26519877177527357,
      "grad_norm": 0.16053281724452972,
      "learning_rate": 0.001,
      "loss": 2.2867627143859863,
      "step": 2604
    },
    {
      "epoch": 0.2654024574678813,
      "grad_norm": 0.11953164637088776,
      "learning_rate": 0.001,
      "loss": 2.4339542388916016,
      "step": 2606
    },
    {
      "epoch": 0.265606143160489,
      "grad_norm": 0.16321510076522827,
      "learning_rate": 0.001,
      "loss": 2.152247428894043,
      "step": 2608
    },
    {
      "epoch": 0.2658098288530968,
      "grad_norm": 0.18862083554267883,
      "learning_rate": 0.001,
      "loss": 2.3639888763427734,
      "step": 2610
    },
    {
      "epoch": 0.26601351454570454,
      "grad_norm": 0.10077431797981262,
      "learning_rate": 0.001,
      "loss": 2.0608482360839844,
      "step": 2612
    },
    {
      "epoch": 0.26621720023831225,
      "grad_norm": 0.13922734558582306,
      "learning_rate": 0.001,
      "loss": 2.2325422763824463,
      "step": 2614
    },
    {
      "epoch": 0.26642088593092,
      "grad_norm": 0.12039919942617416,
      "learning_rate": 0.001,
      "loss": 2.156432628631592,
      "step": 2616
    },
    {
      "epoch": 0.26662457162352776,
      "grad_norm": 0.11588136106729507,
      "learning_rate": 0.001,
      "loss": 2.2495508193969727,
      "step": 2618
    },
    {
      "epoch": 0.26682825731613546,
      "grad_norm": 0.1091311052441597,
      "learning_rate": 0.001,
      "loss": 2.2025794982910156,
      "step": 2620
    },
    {
      "epoch": 0.2670319430087432,
      "grad_norm": 0.14030663669109344,
      "learning_rate": 0.001,
      "loss": 2.197568893432617,
      "step": 2622
    },
    {
      "epoch": 0.2672356287013509,
      "grad_norm": 0.12266213446855545,
      "learning_rate": 0.001,
      "loss": 2.289583683013916,
      "step": 2624
    },
    {
      "epoch": 0.2674393143939587,
      "grad_norm": 0.18210247159004211,
      "learning_rate": 0.001,
      "loss": 2.1881961822509766,
      "step": 2626
    },
    {
      "epoch": 0.26764300008656644,
      "grad_norm": 0.11354729533195496,
      "learning_rate": 0.001,
      "loss": 2.3412721157073975,
      "step": 2628
    },
    {
      "epoch": 0.26784668577917414,
      "grad_norm": 0.115841805934906,
      "learning_rate": 0.001,
      "loss": 2.2849974632263184,
      "step": 2630
    },
    {
      "epoch": 0.2680503714717819,
      "grad_norm": 0.12043590098619461,
      "learning_rate": 0.001,
      "loss": 2.232790470123291,
      "step": 2632
    },
    {
      "epoch": 0.2682540571643896,
      "grad_norm": 0.1375025510787964,
      "learning_rate": 0.001,
      "loss": 2.1107709407806396,
      "step": 2634
    },
    {
      "epoch": 0.26845774285699736,
      "grad_norm": 0.13393042981624603,
      "learning_rate": 0.001,
      "loss": 2.335998058319092,
      "step": 2636
    },
    {
      "epoch": 0.2686614285496051,
      "grad_norm": 0.13254795968532562,
      "learning_rate": 0.001,
      "loss": 2.3033809661865234,
      "step": 2638
    },
    {
      "epoch": 0.2688651142422128,
      "grad_norm": 0.12843671441078186,
      "learning_rate": 0.001,
      "loss": 2.333094596862793,
      "step": 2640
    },
    {
      "epoch": 0.2690687999348206,
      "grad_norm": 0.16022805869579315,
      "learning_rate": 0.001,
      "loss": 2.263850688934326,
      "step": 2642
    },
    {
      "epoch": 0.26927248562742834,
      "grad_norm": 0.11230074614286423,
      "learning_rate": 0.001,
      "loss": 2.2642784118652344,
      "step": 2644
    },
    {
      "epoch": 0.26947617132003604,
      "grad_norm": 0.13527797162532806,
      "learning_rate": 0.001,
      "loss": 2.193479061126709,
      "step": 2646
    },
    {
      "epoch": 0.2696798570126438,
      "grad_norm": 0.1293148547410965,
      "learning_rate": 0.001,
      "loss": 2.265923500061035,
      "step": 2648
    },
    {
      "epoch": 0.2698835427052515,
      "grad_norm": 0.12284532934427261,
      "learning_rate": 0.001,
      "loss": 2.208169460296631,
      "step": 2650
    },
    {
      "epoch": 0.27008722839785926,
      "grad_norm": 0.12729403376579285,
      "learning_rate": 0.001,
      "loss": 2.37559175491333,
      "step": 2652
    },
    {
      "epoch": 0.270290914090467,
      "grad_norm": 0.14475327730178833,
      "learning_rate": 0.001,
      "loss": 2.1450417041778564,
      "step": 2654
    },
    {
      "epoch": 0.2704945997830747,
      "grad_norm": 0.1124851256608963,
      "learning_rate": 0.001,
      "loss": 2.154197931289673,
      "step": 2656
    },
    {
      "epoch": 0.2706982854756825,
      "grad_norm": 0.11966680735349655,
      "learning_rate": 0.001,
      "loss": 2.3112502098083496,
      "step": 2658
    },
    {
      "epoch": 0.27090197116829023,
      "grad_norm": 0.10652370005846024,
      "learning_rate": 0.001,
      "loss": 2.2478370666503906,
      "step": 2660
    },
    {
      "epoch": 0.27110565686089794,
      "grad_norm": 0.12037665396928787,
      "learning_rate": 0.001,
      "loss": 2.200302839279175,
      "step": 2662
    },
    {
      "epoch": 0.2713093425535057,
      "grad_norm": 0.12574388086795807,
      "learning_rate": 0.001,
      "loss": 2.282327890396118,
      "step": 2664
    },
    {
      "epoch": 0.2715130282461134,
      "grad_norm": 0.12543271481990814,
      "learning_rate": 0.001,
      "loss": 2.125049591064453,
      "step": 2666
    },
    {
      "epoch": 0.27171671393872115,
      "grad_norm": 0.11538044363260269,
      "learning_rate": 0.001,
      "loss": 2.1202750205993652,
      "step": 2668
    },
    {
      "epoch": 0.2719203996313289,
      "grad_norm": 0.187924325466156,
      "learning_rate": 0.001,
      "loss": 2.2101354598999023,
      "step": 2670
    },
    {
      "epoch": 0.2721240853239366,
      "grad_norm": 0.14874625205993652,
      "learning_rate": 0.001,
      "loss": 2.3126096725463867,
      "step": 2672
    },
    {
      "epoch": 0.2723277710165444,
      "grad_norm": 0.14120592176914215,
      "learning_rate": 0.001,
      "loss": 2.2418274879455566,
      "step": 2674
    },
    {
      "epoch": 0.27253145670915213,
      "grad_norm": 0.1680409014225006,
      "learning_rate": 0.001,
      "loss": 2.1097233295440674,
      "step": 2676
    },
    {
      "epoch": 0.27273514240175983,
      "grad_norm": 0.1910484880208969,
      "learning_rate": 0.001,
      "loss": 2.317147731781006,
      "step": 2678
    },
    {
      "epoch": 0.2729388280943676,
      "grad_norm": 0.11034401506185532,
      "learning_rate": 0.001,
      "loss": 2.3379340171813965,
      "step": 2680
    },
    {
      "epoch": 0.2731425137869753,
      "grad_norm": 0.15769143402576447,
      "learning_rate": 0.001,
      "loss": 2.2032222747802734,
      "step": 2682
    },
    {
      "epoch": 0.27334619947958305,
      "grad_norm": 0.14189593493938446,
      "learning_rate": 0.001,
      "loss": 2.2042741775512695,
      "step": 2684
    },
    {
      "epoch": 0.2735498851721908,
      "grad_norm": 0.13151733577251434,
      "learning_rate": 0.001,
      "loss": 2.3284454345703125,
      "step": 2686
    },
    {
      "epoch": 0.2737535708647985,
      "grad_norm": 0.1409992128610611,
      "learning_rate": 0.001,
      "loss": 2.24107027053833,
      "step": 2688
    },
    {
      "epoch": 0.27395725655740627,
      "grad_norm": 0.1207733228802681,
      "learning_rate": 0.001,
      "loss": 2.1382782459259033,
      "step": 2690
    },
    {
      "epoch": 0.27416094225001403,
      "grad_norm": 0.18676437437534332,
      "learning_rate": 0.001,
      "loss": 2.124037027359009,
      "step": 2692
    },
    {
      "epoch": 0.27436462794262173,
      "grad_norm": 0.1174948513507843,
      "learning_rate": 0.001,
      "loss": 2.190953016281128,
      "step": 2694
    },
    {
      "epoch": 0.2745683136352295,
      "grad_norm": 0.18644596636295319,
      "learning_rate": 0.001,
      "loss": 2.288017749786377,
      "step": 2696
    },
    {
      "epoch": 0.2747719993278372,
      "grad_norm": 0.17893360555171967,
      "learning_rate": 0.001,
      "loss": 2.196091413497925,
      "step": 2698
    },
    {
      "epoch": 0.27497568502044495,
      "grad_norm": 0.12849412858486176,
      "learning_rate": 0.001,
      "loss": 2.234558582305908,
      "step": 2700
    },
    {
      "epoch": 0.2751793707130527,
      "grad_norm": 0.1161603033542633,
      "learning_rate": 0.001,
      "loss": 2.1817355155944824,
      "step": 2702
    },
    {
      "epoch": 0.2753830564056604,
      "grad_norm": 0.14464206993579865,
      "learning_rate": 0.001,
      "loss": 2.1868491172790527,
      "step": 2704
    },
    {
      "epoch": 0.27558674209826817,
      "grad_norm": 0.15228553116321564,
      "learning_rate": 0.001,
      "loss": 2.2760324478149414,
      "step": 2706
    },
    {
      "epoch": 0.2757904277908759,
      "grad_norm": 0.13040511310100555,
      "learning_rate": 0.001,
      "loss": 2.264056444168091,
      "step": 2708
    },
    {
      "epoch": 0.2759941134834836,
      "grad_norm": 0.1410803347826004,
      "learning_rate": 0.001,
      "loss": 2.3007102012634277,
      "step": 2710
    },
    {
      "epoch": 0.2761977991760914,
      "grad_norm": 0.16221722960472107,
      "learning_rate": 0.001,
      "loss": 2.198531150817871,
      "step": 2712
    },
    {
      "epoch": 0.2764014848686991,
      "grad_norm": 0.11859212070703506,
      "learning_rate": 0.001,
      "loss": 2.175461769104004,
      "step": 2714
    },
    {
      "epoch": 0.27660517056130685,
      "grad_norm": 0.13209441304206848,
      "learning_rate": 0.001,
      "loss": 2.2433438301086426,
      "step": 2716
    },
    {
      "epoch": 0.2768088562539146,
      "grad_norm": 0.16816027462482452,
      "learning_rate": 0.001,
      "loss": 2.2644829750061035,
      "step": 2718
    },
    {
      "epoch": 0.2770125419465223,
      "grad_norm": 0.12626010179519653,
      "learning_rate": 0.001,
      "loss": 2.136946201324463,
      "step": 2720
    },
    {
      "epoch": 0.27721622763913006,
      "grad_norm": 0.10263483226299286,
      "learning_rate": 0.001,
      "loss": 2.080695629119873,
      "step": 2722
    },
    {
      "epoch": 0.2774199133317378,
      "grad_norm": 0.13575968146324158,
      "learning_rate": 0.001,
      "loss": 2.120053768157959,
      "step": 2724
    },
    {
      "epoch": 0.2776235990243455,
      "grad_norm": 0.13683274388313293,
      "learning_rate": 0.001,
      "loss": 2.3494646549224854,
      "step": 2726
    },
    {
      "epoch": 0.2778272847169533,
      "grad_norm": 0.1262296587228775,
      "learning_rate": 0.001,
      "loss": 2.201550006866455,
      "step": 2728
    },
    {
      "epoch": 0.278030970409561,
      "grad_norm": 0.12005586922168732,
      "learning_rate": 0.001,
      "loss": 2.250321626663208,
      "step": 2730
    },
    {
      "epoch": 0.27823465610216874,
      "grad_norm": 0.12513411045074463,
      "learning_rate": 0.001,
      "loss": 2.275740146636963,
      "step": 2732
    },
    {
      "epoch": 0.2784383417947765,
      "grad_norm": 0.12613843381404877,
      "learning_rate": 0.001,
      "loss": 2.195952892303467,
      "step": 2734
    },
    {
      "epoch": 0.2786420274873842,
      "grad_norm": 0.13212727010250092,
      "learning_rate": 0.001,
      "loss": 2.2753710746765137,
      "step": 2736
    },
    {
      "epoch": 0.27884571317999196,
      "grad_norm": 0.11923087388277054,
      "learning_rate": 0.001,
      "loss": 2.1646358966827393,
      "step": 2738
    },
    {
      "epoch": 0.2790493988725997,
      "grad_norm": 0.11962882429361343,
      "learning_rate": 0.001,
      "loss": 2.190734386444092,
      "step": 2740
    },
    {
      "epoch": 0.2792530845652074,
      "grad_norm": 0.11848846077919006,
      "learning_rate": 0.001,
      "loss": 2.2708523273468018,
      "step": 2742
    },
    {
      "epoch": 0.2794567702578152,
      "grad_norm": 0.13870587944984436,
      "learning_rate": 0.001,
      "loss": 2.334907054901123,
      "step": 2744
    },
    {
      "epoch": 0.2796604559504229,
      "grad_norm": 0.12782078981399536,
      "learning_rate": 0.001,
      "loss": 2.426814556121826,
      "step": 2746
    },
    {
      "epoch": 0.27986414164303064,
      "grad_norm": 0.1474793404340744,
      "learning_rate": 0.001,
      "loss": 2.2919235229492188,
      "step": 2748
    },
    {
      "epoch": 0.2800678273356384,
      "grad_norm": 0.15307286381721497,
      "learning_rate": 0.001,
      "loss": 2.1787068843841553,
      "step": 2750
    },
    {
      "epoch": 0.2802715130282461,
      "grad_norm": 0.21244055032730103,
      "learning_rate": 0.001,
      "loss": 2.166435480117798,
      "step": 2752
    },
    {
      "epoch": 0.28047519872085386,
      "grad_norm": 0.17185427248477936,
      "learning_rate": 0.001,
      "loss": 2.158736228942871,
      "step": 2754
    },
    {
      "epoch": 0.28067888441346156,
      "grad_norm": 0.11628913134336472,
      "learning_rate": 0.001,
      "loss": 2.369392156600952,
      "step": 2756
    },
    {
      "epoch": 0.2808825701060693,
      "grad_norm": 0.13531611859798431,
      "learning_rate": 0.001,
      "loss": 2.221059799194336,
      "step": 2758
    },
    {
      "epoch": 0.2810862557986771,
      "grad_norm": 0.10144199430942535,
      "learning_rate": 0.001,
      "loss": 2.1945834159851074,
      "step": 2760
    },
    {
      "epoch": 0.2812899414912848,
      "grad_norm": 0.12250307202339172,
      "learning_rate": 0.001,
      "loss": 2.1723380088806152,
      "step": 2762
    },
    {
      "epoch": 0.28149362718389254,
      "grad_norm": 0.20078040659427643,
      "learning_rate": 0.001,
      "loss": 2.2696096897125244,
      "step": 2764
    },
    {
      "epoch": 0.2816973128765003,
      "grad_norm": 0.1168745830655098,
      "learning_rate": 0.001,
      "loss": 2.2752878665924072,
      "step": 2766
    },
    {
      "epoch": 0.281900998569108,
      "grad_norm": 0.13565461337566376,
      "learning_rate": 0.001,
      "loss": 2.1242947578430176,
      "step": 2768
    },
    {
      "epoch": 0.28210468426171575,
      "grad_norm": 0.10728311538696289,
      "learning_rate": 0.001,
      "loss": 2.1414921283721924,
      "step": 2770
    },
    {
      "epoch": 0.28230836995432346,
      "grad_norm": 0.15147177875041962,
      "learning_rate": 0.001,
      "loss": 2.3697502613067627,
      "step": 2772
    },
    {
      "epoch": 0.2825120556469312,
      "grad_norm": 0.158035010099411,
      "learning_rate": 0.001,
      "loss": 2.3428702354431152,
      "step": 2774
    },
    {
      "epoch": 0.282715741339539,
      "grad_norm": 0.15726393461227417,
      "learning_rate": 0.001,
      "loss": 2.1368656158447266,
      "step": 2776
    },
    {
      "epoch": 0.2829194270321467,
      "grad_norm": 0.1746235191822052,
      "learning_rate": 0.001,
      "loss": 2.1313743591308594,
      "step": 2778
    },
    {
      "epoch": 0.28312311272475443,
      "grad_norm": 0.11663658916950226,
      "learning_rate": 0.001,
      "loss": 2.355541944503784,
      "step": 2780
    },
    {
      "epoch": 0.2833267984173622,
      "grad_norm": 0.12054722011089325,
      "learning_rate": 0.001,
      "loss": 2.1068291664123535,
      "step": 2782
    },
    {
      "epoch": 0.2835304841099699,
      "grad_norm": 0.1535932868719101,
      "learning_rate": 0.001,
      "loss": 2.2932066917419434,
      "step": 2784
    },
    {
      "epoch": 0.28373416980257765,
      "grad_norm": 0.190424844622612,
      "learning_rate": 0.001,
      "loss": 2.265789747238159,
      "step": 2786
    },
    {
      "epoch": 0.28393785549518535,
      "grad_norm": 0.11368729919195175,
      "learning_rate": 0.001,
      "loss": 2.2834763526916504,
      "step": 2788
    },
    {
      "epoch": 0.2841415411877931,
      "grad_norm": 0.21125321090221405,
      "learning_rate": 0.001,
      "loss": 2.3401148319244385,
      "step": 2790
    },
    {
      "epoch": 0.28434522688040087,
      "grad_norm": 0.14262887835502625,
      "learning_rate": 0.001,
      "loss": 2.116389751434326,
      "step": 2792
    },
    {
      "epoch": 0.28454891257300857,
      "grad_norm": 0.12400005757808685,
      "learning_rate": 0.001,
      "loss": 2.2731122970581055,
      "step": 2794
    },
    {
      "epoch": 0.28475259826561633,
      "grad_norm": 0.13200870156288147,
      "learning_rate": 0.001,
      "loss": 2.3588614463806152,
      "step": 2796
    },
    {
      "epoch": 0.2849562839582241,
      "grad_norm": 0.15799500048160553,
      "learning_rate": 0.001,
      "loss": 2.207251787185669,
      "step": 2798
    },
    {
      "epoch": 0.2851599696508318,
      "grad_norm": 0.12848299741744995,
      "learning_rate": 0.001,
      "loss": 2.300833225250244,
      "step": 2800
    },
    {
      "epoch": 0.28536365534343955,
      "grad_norm": 0.12269412726163864,
      "learning_rate": 0.001,
      "loss": 2.191100597381592,
      "step": 2802
    },
    {
      "epoch": 0.28556734103604725,
      "grad_norm": 0.1329207420349121,
      "learning_rate": 0.001,
      "loss": 2.1269264221191406,
      "step": 2804
    },
    {
      "epoch": 0.285771026728655,
      "grad_norm": 0.14722250401973724,
      "learning_rate": 0.001,
      "loss": 2.0932044982910156,
      "step": 2806
    },
    {
      "epoch": 0.28597471242126277,
      "grad_norm": 0.16723033785820007,
      "learning_rate": 0.001,
      "loss": 2.320561408996582,
      "step": 2808
    },
    {
      "epoch": 0.28617839811387047,
      "grad_norm": 0.1321042776107788,
      "learning_rate": 0.001,
      "loss": 2.280291795730591,
      "step": 2810
    },
    {
      "epoch": 0.2863820838064782,
      "grad_norm": 0.1455741822719574,
      "learning_rate": 0.001,
      "loss": 2.1647353172302246,
      "step": 2812
    },
    {
      "epoch": 0.286585769499086,
      "grad_norm": 0.12242356687784195,
      "learning_rate": 0.001,
      "loss": 2.126441240310669,
      "step": 2814
    },
    {
      "epoch": 0.2867894551916937,
      "grad_norm": 0.15124623477458954,
      "learning_rate": 0.001,
      "loss": 2.094339370727539,
      "step": 2816
    },
    {
      "epoch": 0.28699314088430145,
      "grad_norm": 0.15034911036491394,
      "learning_rate": 0.001,
      "loss": 2.0807104110717773,
      "step": 2818
    },
    {
      "epoch": 0.28719682657690915,
      "grad_norm": 0.12756751477718353,
      "learning_rate": 0.001,
      "loss": 2.2227602005004883,
      "step": 2820
    },
    {
      "epoch": 0.2874005122695169,
      "grad_norm": 0.14330913126468658,
      "learning_rate": 0.001,
      "loss": 2.2062745094299316,
      "step": 2822
    },
    {
      "epoch": 0.28760419796212466,
      "grad_norm": 0.12252888083457947,
      "learning_rate": 0.001,
      "loss": 2.273329496383667,
      "step": 2824
    },
    {
      "epoch": 0.28780788365473237,
      "grad_norm": 0.13872644305229187,
      "learning_rate": 0.001,
      "loss": 2.199389696121216,
      "step": 2826
    },
    {
      "epoch": 0.2880115693473401,
      "grad_norm": 0.15420307219028473,
      "learning_rate": 0.001,
      "loss": 2.340616226196289,
      "step": 2828
    },
    {
      "epoch": 0.2882152550399479,
      "grad_norm": 0.14609819650650024,
      "learning_rate": 0.001,
      "loss": 2.2589943408966064,
      "step": 2830
    },
    {
      "epoch": 0.2884189407325556,
      "grad_norm": 0.12604065239429474,
      "learning_rate": 0.001,
      "loss": 2.3338418006896973,
      "step": 2832
    },
    {
      "epoch": 0.28862262642516334,
      "grad_norm": 0.10743481665849686,
      "learning_rate": 0.001,
      "loss": 2.1535425186157227,
      "step": 2834
    },
    {
      "epoch": 0.28882631211777104,
      "grad_norm": 0.12625005841255188,
      "learning_rate": 0.001,
      "loss": 2.217900276184082,
      "step": 2836
    },
    {
      "epoch": 0.2890299978103788,
      "grad_norm": 0.11961814016103745,
      "learning_rate": 0.001,
      "loss": 2.202756881713867,
      "step": 2838
    },
    {
      "epoch": 0.28923368350298656,
      "grad_norm": 0.13313569128513336,
      "learning_rate": 0.001,
      "loss": 2.177443504333496,
      "step": 2840
    },
    {
      "epoch": 0.28943736919559426,
      "grad_norm": 0.13315175473690033,
      "learning_rate": 0.001,
      "loss": 2.159515619277954,
      "step": 2842
    },
    {
      "epoch": 0.289641054888202,
      "grad_norm": 0.1236295998096466,
      "learning_rate": 0.001,
      "loss": 2.443605899810791,
      "step": 2844
    },
    {
      "epoch": 0.2898447405808098,
      "grad_norm": 0.11543665826320648,
      "learning_rate": 0.001,
      "loss": 2.161893606185913,
      "step": 2846
    },
    {
      "epoch": 0.2900484262734175,
      "grad_norm": 0.13089220225811005,
      "learning_rate": 0.001,
      "loss": 2.27069354057312,
      "step": 2848
    },
    {
      "epoch": 0.29025211196602524,
      "grad_norm": 0.10956435650587082,
      "learning_rate": 0.001,
      "loss": 2.2586164474487305,
      "step": 2850
    },
    {
      "epoch": 0.29045579765863294,
      "grad_norm": 0.11617504805326462,
      "learning_rate": 0.001,
      "loss": 2.3115572929382324,
      "step": 2852
    },
    {
      "epoch": 0.2906594833512407,
      "grad_norm": 0.10935907065868378,
      "learning_rate": 0.001,
      "loss": 2.3303041458129883,
      "step": 2854
    },
    {
      "epoch": 0.29086316904384846,
      "grad_norm": 0.1102304458618164,
      "learning_rate": 0.001,
      "loss": 2.1151766777038574,
      "step": 2856
    },
    {
      "epoch": 0.29106685473645616,
      "grad_norm": 0.12016117572784424,
      "learning_rate": 0.001,
      "loss": 2.160602569580078,
      "step": 2858
    },
    {
      "epoch": 0.2912705404290639,
      "grad_norm": 0.14624397456645966,
      "learning_rate": 0.001,
      "loss": 2.324522018432617,
      "step": 2860
    },
    {
      "epoch": 0.2914742261216717,
      "grad_norm": 0.1222170814871788,
      "learning_rate": 0.001,
      "loss": 2.2811951637268066,
      "step": 2862
    },
    {
      "epoch": 0.2916779118142794,
      "grad_norm": 0.11914020031690598,
      "learning_rate": 0.001,
      "loss": 2.1276369094848633,
      "step": 2864
    },
    {
      "epoch": 0.29188159750688714,
      "grad_norm": 0.13370655477046967,
      "learning_rate": 0.001,
      "loss": 2.35481333732605,
      "step": 2866
    },
    {
      "epoch": 0.29208528319949484,
      "grad_norm": 0.11148037761449814,
      "learning_rate": 0.001,
      "loss": 2.3246665000915527,
      "step": 2868
    },
    {
      "epoch": 0.2922889688921026,
      "grad_norm": 0.11772003024816513,
      "learning_rate": 0.001,
      "loss": 2.1601624488830566,
      "step": 2870
    },
    {
      "epoch": 0.29249265458471035,
      "grad_norm": 0.12403424084186554,
      "learning_rate": 0.001,
      "loss": 2.165870189666748,
      "step": 2872
    },
    {
      "epoch": 0.29269634027731806,
      "grad_norm": 0.133452907204628,
      "learning_rate": 0.001,
      "loss": 2.1149988174438477,
      "step": 2874
    },
    {
      "epoch": 0.2929000259699258,
      "grad_norm": 0.12772035598754883,
      "learning_rate": 0.001,
      "loss": 2.2353968620300293,
      "step": 2876
    },
    {
      "epoch": 0.2931037116625335,
      "grad_norm": 0.11632455885410309,
      "learning_rate": 0.001,
      "loss": 2.0372705459594727,
      "step": 2878
    },
    {
      "epoch": 0.2933073973551413,
      "grad_norm": 0.16389203071594238,
      "learning_rate": 0.001,
      "loss": 2.2925047874450684,
      "step": 2880
    },
    {
      "epoch": 0.29351108304774903,
      "grad_norm": 0.14279712736606598,
      "learning_rate": 0.001,
      "loss": 2.1901612281799316,
      "step": 2882
    },
    {
      "epoch": 0.29371476874035674,
      "grad_norm": 0.1168428286910057,
      "learning_rate": 0.001,
      "loss": 2.2323126792907715,
      "step": 2884
    },
    {
      "epoch": 0.2939184544329645,
      "grad_norm": 0.13500447571277618,
      "learning_rate": 0.001,
      "loss": 2.2118797302246094,
      "step": 2886
    },
    {
      "epoch": 0.29412214012557225,
      "grad_norm": 0.1528005748987198,
      "learning_rate": 0.001,
      "loss": 2.1339292526245117,
      "step": 2888
    },
    {
      "epoch": 0.29432582581817995,
      "grad_norm": 0.12431622296571732,
      "learning_rate": 0.001,
      "loss": 2.261763095855713,
      "step": 2890
    },
    {
      "epoch": 0.2945295115107877,
      "grad_norm": 0.16758611798286438,
      "learning_rate": 0.001,
      "loss": 2.2349653244018555,
      "step": 2892
    },
    {
      "epoch": 0.2947331972033954,
      "grad_norm": 0.1324918121099472,
      "learning_rate": 0.001,
      "loss": 2.3428568840026855,
      "step": 2894
    },
    {
      "epoch": 0.29493688289600317,
      "grad_norm": 0.1220194548368454,
      "learning_rate": 0.001,
      "loss": 2.0143094062805176,
      "step": 2896
    },
    {
      "epoch": 0.29514056858861093,
      "grad_norm": 0.19227296113967896,
      "learning_rate": 0.001,
      "loss": 2.097236156463623,
      "step": 2898
    },
    {
      "epoch": 0.29534425428121863,
      "grad_norm": 0.13902077078819275,
      "learning_rate": 0.001,
      "loss": 2.2687830924987793,
      "step": 2900
    },
    {
      "epoch": 0.2955479399738264,
      "grad_norm": 0.13509751856327057,
      "learning_rate": 0.001,
      "loss": 2.2959647178649902,
      "step": 2902
    },
    {
      "epoch": 0.29575162566643415,
      "grad_norm": 0.17153140902519226,
      "learning_rate": 0.001,
      "loss": 2.3824918270111084,
      "step": 2904
    },
    {
      "epoch": 0.29595531135904185,
      "grad_norm": 0.13108821213245392,
      "learning_rate": 0.001,
      "loss": 2.208404064178467,
      "step": 2906
    },
    {
      "epoch": 0.2961589970516496,
      "grad_norm": 0.13402928411960602,
      "learning_rate": 0.001,
      "loss": 2.111860752105713,
      "step": 2908
    },
    {
      "epoch": 0.2963626827442573,
      "grad_norm": 0.13126176595687866,
      "learning_rate": 0.001,
      "loss": 2.133674383163452,
      "step": 2910
    },
    {
      "epoch": 0.29656636843686507,
      "grad_norm": 0.2155710756778717,
      "learning_rate": 0.001,
      "loss": 2.2037229537963867,
      "step": 2912
    },
    {
      "epoch": 0.2967700541294728,
      "grad_norm": 0.13246358931064606,
      "learning_rate": 0.001,
      "loss": 2.3067359924316406,
      "step": 2914
    },
    {
      "epoch": 0.29697373982208053,
      "grad_norm": 0.1517786830663681,
      "learning_rate": 0.001,
      "loss": 2.1954948902130127,
      "step": 2916
    },
    {
      "epoch": 0.2971774255146883,
      "grad_norm": 0.21202431619167328,
      "learning_rate": 0.001,
      "loss": 2.288029193878174,
      "step": 2918
    },
    {
      "epoch": 0.29738111120729604,
      "grad_norm": 0.1436876654624939,
      "learning_rate": 0.001,
      "loss": 2.2964606285095215,
      "step": 2920
    },
    {
      "epoch": 0.29758479689990375,
      "grad_norm": 0.10918133705854416,
      "learning_rate": 0.001,
      "loss": 2.073823928833008,
      "step": 2922
    },
    {
      "epoch": 0.2977884825925115,
      "grad_norm": 0.13685692846775055,
      "learning_rate": 0.001,
      "loss": 2.282309055328369,
      "step": 2924
    },
    {
      "epoch": 0.2979921682851192,
      "grad_norm": 0.15041635930538177,
      "learning_rate": 0.001,
      "loss": 2.103414535522461,
      "step": 2926
    },
    {
      "epoch": 0.29819585397772697,
      "grad_norm": 0.118439219892025,
      "learning_rate": 0.001,
      "loss": 2.2054033279418945,
      "step": 2928
    },
    {
      "epoch": 0.2983995396703347,
      "grad_norm": 0.123115673661232,
      "learning_rate": 0.001,
      "loss": 2.221282958984375,
      "step": 2930
    },
    {
      "epoch": 0.2986032253629424,
      "grad_norm": 0.13006238639354706,
      "learning_rate": 0.001,
      "loss": 2.173612117767334,
      "step": 2932
    },
    {
      "epoch": 0.2988069110555502,
      "grad_norm": 0.12799493968486786,
      "learning_rate": 0.001,
      "loss": 2.340790271759033,
      "step": 2934
    },
    {
      "epoch": 0.29901059674815794,
      "grad_norm": 0.1385725885629654,
      "learning_rate": 0.001,
      "loss": 2.2663733959198,
      "step": 2936
    },
    {
      "epoch": 0.29921428244076564,
      "grad_norm": 0.11950669437646866,
      "learning_rate": 0.001,
      "loss": 2.164724349975586,
      "step": 2938
    },
    {
      "epoch": 0.2994179681333734,
      "grad_norm": 0.11369793862104416,
      "learning_rate": 0.001,
      "loss": 2.096268653869629,
      "step": 2940
    },
    {
      "epoch": 0.2996216538259811,
      "grad_norm": 0.13515204191207886,
      "learning_rate": 0.001,
      "loss": 2.3154380321502686,
      "step": 2942
    },
    {
      "epoch": 0.29982533951858886,
      "grad_norm": 0.1111815869808197,
      "learning_rate": 0.001,
      "loss": 2.045447826385498,
      "step": 2944
    },
    {
      "epoch": 0.3000290252111966,
      "grad_norm": 0.10586021095514297,
      "learning_rate": 0.001,
      "loss": 2.274158477783203,
      "step": 2946
    },
    {
      "epoch": 0.3002327109038043,
      "grad_norm": 0.15888042747974396,
      "learning_rate": 0.001,
      "loss": 2.2025914192199707,
      "step": 2948
    },
    {
      "epoch": 0.3004363965964121,
      "grad_norm": 0.09648504108190536,
      "learning_rate": 0.001,
      "loss": 2.1580305099487305,
      "step": 2950
    },
    {
      "epoch": 0.30064008228901984,
      "grad_norm": 0.10784848779439926,
      "learning_rate": 0.001,
      "loss": 2.193531036376953,
      "step": 2952
    },
    {
      "epoch": 0.30084376798162754,
      "grad_norm": 0.10904043912887573,
      "learning_rate": 0.001,
      "loss": 2.054264783859253,
      "step": 2954
    },
    {
      "epoch": 0.3010474536742353,
      "grad_norm": 0.1313980668783188,
      "learning_rate": 0.001,
      "loss": 2.1839940547943115,
      "step": 2956
    },
    {
      "epoch": 0.301251139366843,
      "grad_norm": 0.11450493335723877,
      "learning_rate": 0.001,
      "loss": 2.204195976257324,
      "step": 2958
    },
    {
      "epoch": 0.30145482505945076,
      "grad_norm": 0.13880035281181335,
      "learning_rate": 0.001,
      "loss": 2.197450637817383,
      "step": 2960
    },
    {
      "epoch": 0.3016585107520585,
      "grad_norm": 0.11303418129682541,
      "learning_rate": 0.001,
      "loss": 2.1346285343170166,
      "step": 2962
    },
    {
      "epoch": 0.3018621964446662,
      "grad_norm": 0.17563775181770325,
      "learning_rate": 0.001,
      "loss": 2.2230074405670166,
      "step": 2964
    },
    {
      "epoch": 0.302065882137274,
      "grad_norm": 0.12805888056755066,
      "learning_rate": 0.001,
      "loss": 2.150618314743042,
      "step": 2966
    },
    {
      "epoch": 0.30226956782988174,
      "grad_norm": 0.1233203113079071,
      "learning_rate": 0.001,
      "loss": 2.217440605163574,
      "step": 2968
    },
    {
      "epoch": 0.30247325352248944,
      "grad_norm": 0.1267346441745758,
      "learning_rate": 0.001,
      "loss": 2.2972168922424316,
      "step": 2970
    },
    {
      "epoch": 0.3026769392150972,
      "grad_norm": 0.18360374867916107,
      "learning_rate": 0.001,
      "loss": 2.07957124710083,
      "step": 2972
    },
    {
      "epoch": 0.3028806249077049,
      "grad_norm": 0.1287737488746643,
      "learning_rate": 0.001,
      "loss": 2.2120866775512695,
      "step": 2974
    },
    {
      "epoch": 0.30308431060031266,
      "grad_norm": 0.1911492496728897,
      "learning_rate": 0.001,
      "loss": 2.300553798675537,
      "step": 2976
    },
    {
      "epoch": 0.3032879962929204,
      "grad_norm": 0.12429459393024445,
      "learning_rate": 0.001,
      "loss": 2.235351800918579,
      "step": 2978
    },
    {
      "epoch": 0.3034916819855281,
      "grad_norm": 0.16596931219100952,
      "learning_rate": 0.001,
      "loss": 2.2125768661499023,
      "step": 2980
    },
    {
      "epoch": 0.3036953676781359,
      "grad_norm": 0.1448335498571396,
      "learning_rate": 0.001,
      "loss": 2.3619933128356934,
      "step": 2982
    },
    {
      "epoch": 0.30389905337074363,
      "grad_norm": 0.13478198647499084,
      "learning_rate": 0.001,
      "loss": 2.267805576324463,
      "step": 2984
    },
    {
      "epoch": 0.30410273906335133,
      "grad_norm": 0.11998455971479416,
      "learning_rate": 0.001,
      "loss": 2.0567948818206787,
      "step": 2986
    },
    {
      "epoch": 0.3043064247559591,
      "grad_norm": 0.11893241107463837,
      "learning_rate": 0.001,
      "loss": 2.366260051727295,
      "step": 2988
    },
    {
      "epoch": 0.3045101104485668,
      "grad_norm": 0.11435852199792862,
      "learning_rate": 0.001,
      "loss": 2.218113899230957,
      "step": 2990
    },
    {
      "epoch": 0.30471379614117455,
      "grad_norm": 0.16522185504436493,
      "learning_rate": 0.001,
      "loss": 2.347424268722534,
      "step": 2992
    },
    {
      "epoch": 0.3049174818337823,
      "grad_norm": 0.1241653636097908,
      "learning_rate": 0.001,
      "loss": 2.264617919921875,
      "step": 2994
    },
    {
      "epoch": 0.30512116752639,
      "grad_norm": 0.134572371840477,
      "learning_rate": 0.001,
      "loss": 2.095726490020752,
      "step": 2996
    },
    {
      "epoch": 0.30532485321899777,
      "grad_norm": 0.13727058470249176,
      "learning_rate": 0.001,
      "loss": 2.267194986343384,
      "step": 2998
    },
    {
      "epoch": 0.3055285389116055,
      "grad_norm": 0.14657095074653625,
      "learning_rate": 0.001,
      "loss": 2.186218738555908,
      "step": 3000
    },
    {
      "epoch": 0.30573222460421323,
      "grad_norm": 0.13491271436214447,
      "learning_rate": 0.001,
      "loss": 2.2306082248687744,
      "step": 3002
    },
    {
      "epoch": 0.305935910296821,
      "grad_norm": 0.13055455684661865,
      "learning_rate": 0.001,
      "loss": 2.1990504264831543,
      "step": 3004
    },
    {
      "epoch": 0.3061395959894287,
      "grad_norm": 0.1288089007139206,
      "learning_rate": 0.001,
      "loss": 2.3285932540893555,
      "step": 3006
    },
    {
      "epoch": 0.30634328168203645,
      "grad_norm": 0.11277665942907333,
      "learning_rate": 0.001,
      "loss": 2.323044776916504,
      "step": 3008
    },
    {
      "epoch": 0.3065469673746442,
      "grad_norm": 0.11544152349233627,
      "learning_rate": 0.001,
      "loss": 2.222487688064575,
      "step": 3010
    },
    {
      "epoch": 0.3067506530672519,
      "grad_norm": 0.11497961729764938,
      "learning_rate": 0.001,
      "loss": 2.122617721557617,
      "step": 3012
    },
    {
      "epoch": 0.30695433875985967,
      "grad_norm": 0.1645895093679428,
      "learning_rate": 0.001,
      "loss": 2.2004668712615967,
      "step": 3014
    },
    {
      "epoch": 0.30715802445246737,
      "grad_norm": 0.1156701073050499,
      "learning_rate": 0.001,
      "loss": 2.174050807952881,
      "step": 3016
    },
    {
      "epoch": 0.30736171014507513,
      "grad_norm": 0.14424480497837067,
      "learning_rate": 0.001,
      "loss": 2.27053165435791,
      "step": 3018
    },
    {
      "epoch": 0.3075653958376829,
      "grad_norm": 0.16335907578468323,
      "learning_rate": 0.001,
      "loss": 2.137019157409668,
      "step": 3020
    },
    {
      "epoch": 0.3077690815302906,
      "grad_norm": 0.12381400913000107,
      "learning_rate": 0.001,
      "loss": 2.272343158721924,
      "step": 3022
    },
    {
      "epoch": 0.30797276722289835,
      "grad_norm": 0.21364425122737885,
      "learning_rate": 0.001,
      "loss": 2.251906394958496,
      "step": 3024
    },
    {
      "epoch": 0.3081764529155061,
      "grad_norm": 0.12200777232646942,
      "learning_rate": 0.001,
      "loss": 2.161190986633301,
      "step": 3026
    },
    {
      "epoch": 0.3083801386081138,
      "grad_norm": 0.12086398154497147,
      "learning_rate": 0.001,
      "loss": 2.2517032623291016,
      "step": 3028
    },
    {
      "epoch": 0.30858382430072157,
      "grad_norm": 0.1413571983575821,
      "learning_rate": 0.001,
      "loss": 2.274062395095825,
      "step": 3030
    },
    {
      "epoch": 0.30878750999332927,
      "grad_norm": 0.12189828604459763,
      "learning_rate": 0.001,
      "loss": 2.189202308654785,
      "step": 3032
    },
    {
      "epoch": 0.308991195685937,
      "grad_norm": 0.13868200778961182,
      "learning_rate": 0.001,
      "loss": 2.206498622894287,
      "step": 3034
    },
    {
      "epoch": 0.3091948813785448,
      "grad_norm": 0.1735059767961502,
      "learning_rate": 0.001,
      "loss": 2.3375368118286133,
      "step": 3036
    },
    {
      "epoch": 0.3093985670711525,
      "grad_norm": 0.13681675493717194,
      "learning_rate": 0.001,
      "loss": 2.2417216300964355,
      "step": 3038
    },
    {
      "epoch": 0.30960225276376024,
      "grad_norm": 0.14370331168174744,
      "learning_rate": 0.001,
      "loss": 2.2766690254211426,
      "step": 3040
    },
    {
      "epoch": 0.309805938456368,
      "grad_norm": 0.1866101622581482,
      "learning_rate": 0.001,
      "loss": 2.248917579650879,
      "step": 3042
    },
    {
      "epoch": 0.3100096241489757,
      "grad_norm": 0.1270902156829834,
      "learning_rate": 0.001,
      "loss": 2.2278358936309814,
      "step": 3044
    },
    {
      "epoch": 0.31021330984158346,
      "grad_norm": 0.13034552335739136,
      "learning_rate": 0.001,
      "loss": 2.322420120239258,
      "step": 3046
    },
    {
      "epoch": 0.31041699553419116,
      "grad_norm": 0.13077622652053833,
      "learning_rate": 0.001,
      "loss": 2.2807278633117676,
      "step": 3048
    },
    {
      "epoch": 0.3106206812267989,
      "grad_norm": 0.10664664208889008,
      "learning_rate": 0.001,
      "loss": 2.1908605098724365,
      "step": 3050
    },
    {
      "epoch": 0.3108243669194067,
      "grad_norm": 0.1463399976491928,
      "learning_rate": 0.001,
      "loss": 2.2205777168273926,
      "step": 3052
    },
    {
      "epoch": 0.3110280526120144,
      "grad_norm": 0.1185903400182724,
      "learning_rate": 0.001,
      "loss": 2.242265462875366,
      "step": 3054
    },
    {
      "epoch": 0.31123173830462214,
      "grad_norm": 0.13022932410240173,
      "learning_rate": 0.001,
      "loss": 2.258598804473877,
      "step": 3056
    },
    {
      "epoch": 0.3114354239972299,
      "grad_norm": 0.118958979845047,
      "learning_rate": 0.001,
      "loss": 2.3303723335266113,
      "step": 3058
    },
    {
      "epoch": 0.3116391096898376,
      "grad_norm": 0.12233783304691315,
      "learning_rate": 0.001,
      "loss": 2.2570457458496094,
      "step": 3060
    },
    {
      "epoch": 0.31184279538244536,
      "grad_norm": 0.11822043359279633,
      "learning_rate": 0.001,
      "loss": 2.1745357513427734,
      "step": 3062
    },
    {
      "epoch": 0.31204648107505306,
      "grad_norm": 0.12646500766277313,
      "learning_rate": 0.001,
      "loss": 2.0567641258239746,
      "step": 3064
    },
    {
      "epoch": 0.3122501667676608,
      "grad_norm": 0.12569080293178558,
      "learning_rate": 0.001,
      "loss": 2.3920884132385254,
      "step": 3066
    },
    {
      "epoch": 0.3124538524602686,
      "grad_norm": 0.20610778033733368,
      "learning_rate": 0.001,
      "loss": 2.2216758728027344,
      "step": 3068
    },
    {
      "epoch": 0.3126575381528763,
      "grad_norm": 0.16336123645305634,
      "learning_rate": 0.001,
      "loss": 2.201632499694824,
      "step": 3070
    },
    {
      "epoch": 0.31286122384548404,
      "grad_norm": 0.11999719589948654,
      "learning_rate": 0.001,
      "loss": 2.1262941360473633,
      "step": 3072
    },
    {
      "epoch": 0.3130649095380918,
      "grad_norm": 0.14314936101436615,
      "learning_rate": 0.001,
      "loss": 2.272801399230957,
      "step": 3074
    },
    {
      "epoch": 0.3132685952306995,
      "grad_norm": 0.14245890080928802,
      "learning_rate": 0.001,
      "loss": 2.137526512145996,
      "step": 3076
    },
    {
      "epoch": 0.31347228092330726,
      "grad_norm": 0.11790891736745834,
      "learning_rate": 0.001,
      "loss": 2.101440906524658,
      "step": 3078
    },
    {
      "epoch": 0.31367596661591496,
      "grad_norm": 0.14476323127746582,
      "learning_rate": 0.001,
      "loss": 2.207820415496826,
      "step": 3080
    },
    {
      "epoch": 0.3138796523085227,
      "grad_norm": 0.10720179975032806,
      "learning_rate": 0.001,
      "loss": 2.226987361907959,
      "step": 3082
    },
    {
      "epoch": 0.3140833380011305,
      "grad_norm": 0.13871949911117554,
      "learning_rate": 0.001,
      "loss": 2.263669013977051,
      "step": 3084
    },
    {
      "epoch": 0.3142870236937382,
      "grad_norm": 0.11941425502300262,
      "learning_rate": 0.001,
      "loss": 2.201568841934204,
      "step": 3086
    },
    {
      "epoch": 0.31449070938634593,
      "grad_norm": 0.12205391377210617,
      "learning_rate": 0.001,
      "loss": 2.187673568725586,
      "step": 3088
    },
    {
      "epoch": 0.3146943950789537,
      "grad_norm": 0.1880333125591278,
      "learning_rate": 0.001,
      "loss": 2.310145854949951,
      "step": 3090
    },
    {
      "epoch": 0.3148980807715614,
      "grad_norm": 0.17234188318252563,
      "learning_rate": 0.001,
      "loss": 2.3456978797912598,
      "step": 3092
    },
    {
      "epoch": 0.31510176646416915,
      "grad_norm": 0.1232944205403328,
      "learning_rate": 0.001,
      "loss": 2.183500289916992,
      "step": 3094
    },
    {
      "epoch": 0.31530545215677686,
      "grad_norm": 0.12123943120241165,
      "learning_rate": 0.001,
      "loss": 2.13847017288208,
      "step": 3096
    },
    {
      "epoch": 0.3155091378493846,
      "grad_norm": 0.1613069474697113,
      "learning_rate": 0.001,
      "loss": 2.2106881141662598,
      "step": 3098
    },
    {
      "epoch": 0.31571282354199237,
      "grad_norm": 0.17888714373111725,
      "learning_rate": 0.001,
      "loss": 2.2726352214813232,
      "step": 3100
    },
    {
      "epoch": 0.3159165092346001,
      "grad_norm": 0.16037239134311676,
      "learning_rate": 0.001,
      "loss": 2.249880313873291,
      "step": 3102
    },
    {
      "epoch": 0.31612019492720783,
      "grad_norm": 0.14775200188159943,
      "learning_rate": 0.001,
      "loss": 2.1007044315338135,
      "step": 3104
    },
    {
      "epoch": 0.3163238806198156,
      "grad_norm": 0.1476036161184311,
      "learning_rate": 0.001,
      "loss": 2.3558197021484375,
      "step": 3106
    },
    {
      "epoch": 0.3165275663124233,
      "grad_norm": 0.17594276368618011,
      "learning_rate": 0.001,
      "loss": 2.1924970149993896,
      "step": 3108
    },
    {
      "epoch": 0.31673125200503105,
      "grad_norm": 0.12669704854488373,
      "learning_rate": 0.001,
      "loss": 2.1994948387145996,
      "step": 3110
    },
    {
      "epoch": 0.31693493769763875,
      "grad_norm": 0.12177923321723938,
      "learning_rate": 0.001,
      "loss": 2.161025047302246,
      "step": 3112
    },
    {
      "epoch": 0.3171386233902465,
      "grad_norm": 0.1219155490398407,
      "learning_rate": 0.001,
      "loss": 2.2487399578094482,
      "step": 3114
    },
    {
      "epoch": 0.31734230908285427,
      "grad_norm": 0.15910372138023376,
      "learning_rate": 0.001,
      "loss": 2.2906126976013184,
      "step": 3116
    },
    {
      "epoch": 0.31754599477546197,
      "grad_norm": 0.12473728507757187,
      "learning_rate": 0.001,
      "loss": 2.2033638954162598,
      "step": 3118
    },
    {
      "epoch": 0.31774968046806973,
      "grad_norm": 0.20003828406333923,
      "learning_rate": 0.001,
      "loss": 2.1814651489257812,
      "step": 3120
    },
    {
      "epoch": 0.3179533661606775,
      "grad_norm": 0.130088210105896,
      "learning_rate": 0.001,
      "loss": 2.3408236503601074,
      "step": 3122
    },
    {
      "epoch": 0.3181570518532852,
      "grad_norm": 0.13847662508487701,
      "learning_rate": 0.001,
      "loss": 2.2156031131744385,
      "step": 3124
    },
    {
      "epoch": 0.31836073754589295,
      "grad_norm": 0.10316331684589386,
      "learning_rate": 0.001,
      "loss": 2.1038782596588135,
      "step": 3126
    },
    {
      "epoch": 0.31856442323850065,
      "grad_norm": 0.14215460419654846,
      "learning_rate": 0.001,
      "loss": 2.12893009185791,
      "step": 3128
    },
    {
      "epoch": 0.3187681089311084,
      "grad_norm": 0.15275093913078308,
      "learning_rate": 0.001,
      "loss": 2.190680503845215,
      "step": 3130
    },
    {
      "epoch": 0.31897179462371616,
      "grad_norm": 0.1366843283176422,
      "learning_rate": 0.001,
      "loss": 2.3228023052215576,
      "step": 3132
    },
    {
      "epoch": 0.31917548031632387,
      "grad_norm": 0.15855936706066132,
      "learning_rate": 0.001,
      "loss": 2.130465507507324,
      "step": 3134
    },
    {
      "epoch": 0.3193791660089316,
      "grad_norm": 0.20910046994686127,
      "learning_rate": 0.001,
      "loss": 2.3001272678375244,
      "step": 3136
    },
    {
      "epoch": 0.3195828517015393,
      "grad_norm": 0.11284060776233673,
      "learning_rate": 0.001,
      "loss": 2.2090420722961426,
      "step": 3138
    },
    {
      "epoch": 0.3197865373941471,
      "grad_norm": 0.12254350632429123,
      "learning_rate": 0.001,
      "loss": 2.157954216003418,
      "step": 3140
    },
    {
      "epoch": 0.31999022308675484,
      "grad_norm": 0.14089658856391907,
      "learning_rate": 0.001,
      "loss": 2.030580520629883,
      "step": 3142
    },
    {
      "epoch": 0.32019390877936255,
      "grad_norm": 0.17857369780540466,
      "learning_rate": 0.001,
      "loss": 2.192897319793701,
      "step": 3144
    },
    {
      "epoch": 0.3203975944719703,
      "grad_norm": 0.12865512073040009,
      "learning_rate": 0.001,
      "loss": 2.2639245986938477,
      "step": 3146
    },
    {
      "epoch": 0.32060128016457806,
      "grad_norm": 0.15388204157352448,
      "learning_rate": 0.001,
      "loss": 2.2917118072509766,
      "step": 3148
    },
    {
      "epoch": 0.32080496585718576,
      "grad_norm": 0.12142860889434814,
      "learning_rate": 0.001,
      "loss": 2.1318869590759277,
      "step": 3150
    },
    {
      "epoch": 0.3210086515497935,
      "grad_norm": 0.13087767362594604,
      "learning_rate": 0.001,
      "loss": 2.2002780437469482,
      "step": 3152
    },
    {
      "epoch": 0.3212123372424012,
      "grad_norm": 0.11320862174034119,
      "learning_rate": 0.001,
      "loss": 2.0286622047424316,
      "step": 3154
    },
    {
      "epoch": 0.321416022935009,
      "grad_norm": 0.15085753798484802,
      "learning_rate": 0.001,
      "loss": 2.194484233856201,
      "step": 3156
    },
    {
      "epoch": 0.32161970862761674,
      "grad_norm": 0.13279789686203003,
      "learning_rate": 0.001,
      "loss": 2.2933077812194824,
      "step": 3158
    },
    {
      "epoch": 0.32182339432022444,
      "grad_norm": 0.12445057928562164,
      "learning_rate": 0.001,
      "loss": 2.2416422367095947,
      "step": 3160
    },
    {
      "epoch": 0.3220270800128322,
      "grad_norm": 0.13437704741954803,
      "learning_rate": 0.001,
      "loss": 2.3522067070007324,
      "step": 3162
    },
    {
      "epoch": 0.32223076570543996,
      "grad_norm": 0.17572377622127533,
      "learning_rate": 0.001,
      "loss": 2.3481035232543945,
      "step": 3164
    },
    {
      "epoch": 0.32243445139804766,
      "grad_norm": 0.14414077997207642,
      "learning_rate": 0.001,
      "loss": 2.2533445358276367,
      "step": 3166
    },
    {
      "epoch": 0.3226381370906554,
      "grad_norm": 0.11754142493009567,
      "learning_rate": 0.001,
      "loss": 2.215257167816162,
      "step": 3168
    },
    {
      "epoch": 0.3228418227832631,
      "grad_norm": 0.13035739958286285,
      "learning_rate": 0.001,
      "loss": 2.0732533931732178,
      "step": 3170
    },
    {
      "epoch": 0.3230455084758709,
      "grad_norm": 0.1412942260503769,
      "learning_rate": 0.001,
      "loss": 2.1362390518188477,
      "step": 3172
    },
    {
      "epoch": 0.32324919416847864,
      "grad_norm": 0.1152598038315773,
      "learning_rate": 0.001,
      "loss": 2.3266968727111816,
      "step": 3174
    },
    {
      "epoch": 0.32345287986108634,
      "grad_norm": 0.1267794519662857,
      "learning_rate": 0.001,
      "loss": 2.191272735595703,
      "step": 3176
    },
    {
      "epoch": 0.3236565655536941,
      "grad_norm": 0.13127602636814117,
      "learning_rate": 0.001,
      "loss": 2.1627800464630127,
      "step": 3178
    },
    {
      "epoch": 0.32386025124630186,
      "grad_norm": 0.12332537770271301,
      "learning_rate": 0.001,
      "loss": 2.266486644744873,
      "step": 3180
    },
    {
      "epoch": 0.32406393693890956,
      "grad_norm": 0.12937723100185394,
      "learning_rate": 0.001,
      "loss": 2.2301063537597656,
      "step": 3182
    },
    {
      "epoch": 0.3242676226315173,
      "grad_norm": 0.16026873886585236,
      "learning_rate": 0.001,
      "loss": 2.40329647064209,
      "step": 3184
    },
    {
      "epoch": 0.324471308324125,
      "grad_norm": 0.12421286106109619,
      "learning_rate": 0.001,
      "loss": 2.2058820724487305,
      "step": 3186
    },
    {
      "epoch": 0.3246749940167328,
      "grad_norm": 0.15131467580795288,
      "learning_rate": 0.001,
      "loss": 2.291151762008667,
      "step": 3188
    },
    {
      "epoch": 0.32487867970934053,
      "grad_norm": 0.12873317301273346,
      "learning_rate": 0.001,
      "loss": 2.162569046020508,
      "step": 3190
    },
    {
      "epoch": 0.32508236540194824,
      "grad_norm": 0.1402159333229065,
      "learning_rate": 0.001,
      "loss": 2.35141658782959,
      "step": 3192
    },
    {
      "epoch": 0.325286051094556,
      "grad_norm": 0.17426125705242157,
      "learning_rate": 0.001,
      "loss": 2.2792086601257324,
      "step": 3194
    },
    {
      "epoch": 0.32548973678716375,
      "grad_norm": 0.1834244728088379,
      "learning_rate": 0.001,
      "loss": 2.283576011657715,
      "step": 3196
    },
    {
      "epoch": 0.32569342247977145,
      "grad_norm": 0.13337835669517517,
      "learning_rate": 0.001,
      "loss": 2.215167999267578,
      "step": 3198
    },
    {
      "epoch": 0.3258971081723792,
      "grad_norm": 0.13495489954948425,
      "learning_rate": 0.001,
      "loss": 2.258345603942871,
      "step": 3200
    },
    {
      "epoch": 0.3261007938649869,
      "grad_norm": 0.12048188596963882,
      "learning_rate": 0.001,
      "loss": 2.3053717613220215,
      "step": 3202
    },
    {
      "epoch": 0.3263044795575947,
      "grad_norm": 0.12214597314596176,
      "learning_rate": 0.001,
      "loss": 2.3298816680908203,
      "step": 3204
    },
    {
      "epoch": 0.32650816525020243,
      "grad_norm": 0.12435226887464523,
      "learning_rate": 0.001,
      "loss": 2.3785743713378906,
      "step": 3206
    },
    {
      "epoch": 0.32671185094281013,
      "grad_norm": 0.1430298238992691,
      "learning_rate": 0.001,
      "loss": 2.2507057189941406,
      "step": 3208
    },
    {
      "epoch": 0.3269155366354179,
      "grad_norm": 0.11805085837841034,
      "learning_rate": 0.001,
      "loss": 2.031937837600708,
      "step": 3210
    },
    {
      "epoch": 0.32711922232802565,
      "grad_norm": 0.12675049901008606,
      "learning_rate": 0.001,
      "loss": 2.172717571258545,
      "step": 3212
    },
    {
      "epoch": 0.32732290802063335,
      "grad_norm": 0.14166893064975739,
      "learning_rate": 0.001,
      "loss": 2.1159873008728027,
      "step": 3214
    },
    {
      "epoch": 0.3275265937132411,
      "grad_norm": 0.1296423375606537,
      "learning_rate": 0.001,
      "loss": 2.1951351165771484,
      "step": 3216
    },
    {
      "epoch": 0.3277302794058488,
      "grad_norm": 0.1276164948940277,
      "learning_rate": 0.001,
      "loss": 2.240252733230591,
      "step": 3218
    },
    {
      "epoch": 0.32793396509845657,
      "grad_norm": 0.10781542956829071,
      "learning_rate": 0.001,
      "loss": 2.0860447883605957,
      "step": 3220
    },
    {
      "epoch": 0.32813765079106433,
      "grad_norm": 0.12937945127487183,
      "learning_rate": 0.001,
      "loss": 2.3495073318481445,
      "step": 3222
    },
    {
      "epoch": 0.32834133648367203,
      "grad_norm": 0.15405693650245667,
      "learning_rate": 0.001,
      "loss": 2.2434699535369873,
      "step": 3224
    },
    {
      "epoch": 0.3285450221762798,
      "grad_norm": 0.1259496808052063,
      "learning_rate": 0.001,
      "loss": 2.1116504669189453,
      "step": 3226
    },
    {
      "epoch": 0.32874870786888755,
      "grad_norm": 0.14775384962558746,
      "learning_rate": 0.001,
      "loss": 2.1709330081939697,
      "step": 3228
    },
    {
      "epoch": 0.32895239356149525,
      "grad_norm": 0.13158324360847473,
      "learning_rate": 0.001,
      "loss": 2.415889263153076,
      "step": 3230
    },
    {
      "epoch": 0.329156079254103,
      "grad_norm": 0.17860354483127594,
      "learning_rate": 0.001,
      "loss": 2.403841018676758,
      "step": 3232
    },
    {
      "epoch": 0.3293597649467107,
      "grad_norm": 0.14081436395645142,
      "learning_rate": 0.001,
      "loss": 2.22745418548584,
      "step": 3234
    },
    {
      "epoch": 0.32956345063931847,
      "grad_norm": 0.20054569840431213,
      "learning_rate": 0.001,
      "loss": 2.0769691467285156,
      "step": 3236
    },
    {
      "epoch": 0.3297671363319262,
      "grad_norm": 0.11941264569759369,
      "learning_rate": 0.001,
      "loss": 2.127598285675049,
      "step": 3238
    },
    {
      "epoch": 0.3299708220245339,
      "grad_norm": 0.18203867971897125,
      "learning_rate": 0.001,
      "loss": 2.3120503425598145,
      "step": 3240
    },
    {
      "epoch": 0.3301745077171417,
      "grad_norm": 0.12091968208551407,
      "learning_rate": 0.001,
      "loss": 2.2296524047851562,
      "step": 3242
    },
    {
      "epoch": 0.33037819340974944,
      "grad_norm": 0.12841914594173431,
      "learning_rate": 0.001,
      "loss": 2.2675323486328125,
      "step": 3244
    },
    {
      "epoch": 0.33058187910235715,
      "grad_norm": 0.13421010971069336,
      "learning_rate": 0.001,
      "loss": 2.267707347869873,
      "step": 3246
    },
    {
      "epoch": 0.3307855647949649,
      "grad_norm": 0.10711365193128586,
      "learning_rate": 0.001,
      "loss": 2.128004550933838,
      "step": 3248
    },
    {
      "epoch": 0.3309892504875726,
      "grad_norm": 0.11908148974180222,
      "learning_rate": 0.001,
      "loss": 2.1449875831604004,
      "step": 3250
    },
    {
      "epoch": 0.33119293618018036,
      "grad_norm": 0.12044800817966461,
      "learning_rate": 0.001,
      "loss": 2.210219144821167,
      "step": 3252
    },
    {
      "epoch": 0.3313966218727881,
      "grad_norm": 0.1381181925535202,
      "learning_rate": 0.001,
      "loss": 2.159575939178467,
      "step": 3254
    },
    {
      "epoch": 0.3316003075653958,
      "grad_norm": 0.11904972791671753,
      "learning_rate": 0.001,
      "loss": 2.177422285079956,
      "step": 3256
    },
    {
      "epoch": 0.3318039932580036,
      "grad_norm": 0.14340455830097198,
      "learning_rate": 0.001,
      "loss": 2.3136234283447266,
      "step": 3258
    },
    {
      "epoch": 0.3320076789506113,
      "grad_norm": 0.17739500105381012,
      "learning_rate": 0.001,
      "loss": 2.2294065952301025,
      "step": 3260
    },
    {
      "epoch": 0.33221136464321904,
      "grad_norm": 0.12360537797212601,
      "learning_rate": 0.001,
      "loss": 2.40341854095459,
      "step": 3262
    },
    {
      "epoch": 0.3324150503358268,
      "grad_norm": 0.13499204814434052,
      "learning_rate": 0.001,
      "loss": 2.122408628463745,
      "step": 3264
    },
    {
      "epoch": 0.3326187360284345,
      "grad_norm": 0.18163251876831055,
      "learning_rate": 0.001,
      "loss": 2.3846468925476074,
      "step": 3266
    },
    {
      "epoch": 0.33282242172104226,
      "grad_norm": 0.1586497575044632,
      "learning_rate": 0.001,
      "loss": 2.2566685676574707,
      "step": 3268
    },
    {
      "epoch": 0.33302610741365,
      "grad_norm": 0.11256980895996094,
      "learning_rate": 0.001,
      "loss": 2.1704487800598145,
      "step": 3270
    },
    {
      "epoch": 0.3332297931062577,
      "grad_norm": 0.1385485678911209,
      "learning_rate": 0.001,
      "loss": 2.2655396461486816,
      "step": 3272
    },
    {
      "epoch": 0.3334334787988655,
      "grad_norm": 0.10709425806999207,
      "learning_rate": 0.001,
      "loss": 2.2765681743621826,
      "step": 3274
    },
    {
      "epoch": 0.3336371644914732,
      "grad_norm": 0.15259429812431335,
      "learning_rate": 0.001,
      "loss": 2.1788430213928223,
      "step": 3276
    },
    {
      "epoch": 0.33384085018408094,
      "grad_norm": 0.12258775532245636,
      "learning_rate": 0.001,
      "loss": 2.3327431678771973,
      "step": 3278
    },
    {
      "epoch": 0.3340445358766887,
      "grad_norm": 0.12670549750328064,
      "learning_rate": 0.001,
      "loss": 2.2298102378845215,
      "step": 3280
    },
    {
      "epoch": 0.3342482215692964,
      "grad_norm": 0.15333110094070435,
      "learning_rate": 0.001,
      "loss": 2.240304470062256,
      "step": 3282
    },
    {
      "epoch": 0.33445190726190416,
      "grad_norm": 0.14585888385772705,
      "learning_rate": 0.001,
      "loss": 2.297560691833496,
      "step": 3284
    },
    {
      "epoch": 0.3346555929545119,
      "grad_norm": 0.12096761167049408,
      "learning_rate": 0.001,
      "loss": 2.1680421829223633,
      "step": 3286
    },
    {
      "epoch": 0.3348592786471196,
      "grad_norm": 0.1129307746887207,
      "learning_rate": 0.001,
      "loss": 2.341413974761963,
      "step": 3288
    },
    {
      "epoch": 0.3350629643397274,
      "grad_norm": 0.11568030714988708,
      "learning_rate": 0.001,
      "loss": 2.1954855918884277,
      "step": 3290
    },
    {
      "epoch": 0.3352666500323351,
      "grad_norm": 0.12840788066387177,
      "learning_rate": 0.001,
      "loss": 2.2861838340759277,
      "step": 3292
    },
    {
      "epoch": 0.33547033572494284,
      "grad_norm": 0.12425125390291214,
      "learning_rate": 0.001,
      "loss": 2.235463857650757,
      "step": 3294
    },
    {
      "epoch": 0.3356740214175506,
      "grad_norm": 0.1464596390724182,
      "learning_rate": 0.001,
      "loss": 2.2366340160369873,
      "step": 3296
    },
    {
      "epoch": 0.3358777071101583,
      "grad_norm": 0.12105492502450943,
      "learning_rate": 0.001,
      "loss": 2.284456968307495,
      "step": 3298
    },
    {
      "epoch": 0.33608139280276605,
      "grad_norm": 0.12726210057735443,
      "learning_rate": 0.001,
      "loss": 2.321445941925049,
      "step": 3300
    },
    {
      "epoch": 0.3362850784953738,
      "grad_norm": 0.11900043487548828,
      "learning_rate": 0.001,
      "loss": 2.1984283924102783,
      "step": 3302
    },
    {
      "epoch": 0.3364887641879815,
      "grad_norm": 0.17053425312042236,
      "learning_rate": 0.001,
      "loss": 2.2712202072143555,
      "step": 3304
    },
    {
      "epoch": 0.3366924498805893,
      "grad_norm": 0.1417412906885147,
      "learning_rate": 0.001,
      "loss": 2.241527557373047,
      "step": 3306
    },
    {
      "epoch": 0.336896135573197,
      "grad_norm": 0.11831735074520111,
      "learning_rate": 0.001,
      "loss": 2.218235969543457,
      "step": 3308
    },
    {
      "epoch": 0.33709982126580473,
      "grad_norm": 0.10921081900596619,
      "learning_rate": 0.001,
      "loss": 2.076932907104492,
      "step": 3310
    },
    {
      "epoch": 0.3373035069584125,
      "grad_norm": 0.13989761471748352,
      "learning_rate": 0.001,
      "loss": 2.2790465354919434,
      "step": 3312
    },
    {
      "epoch": 0.3375071926510202,
      "grad_norm": 0.12579354643821716,
      "learning_rate": 0.001,
      "loss": 2.2928168773651123,
      "step": 3314
    },
    {
      "epoch": 0.33771087834362795,
      "grad_norm": 0.11148692667484283,
      "learning_rate": 0.001,
      "loss": 2.1613969802856445,
      "step": 3316
    },
    {
      "epoch": 0.3379145640362357,
      "grad_norm": 0.106688492000103,
      "learning_rate": 0.001,
      "loss": 2.2643911838531494,
      "step": 3318
    },
    {
      "epoch": 0.3381182497288434,
      "grad_norm": 0.12407828867435455,
      "learning_rate": 0.001,
      "loss": 2.1439857482910156,
      "step": 3320
    },
    {
      "epoch": 0.33832193542145117,
      "grad_norm": 0.11777269095182419,
      "learning_rate": 0.001,
      "loss": 2.301845073699951,
      "step": 3322
    },
    {
      "epoch": 0.33852562111405887,
      "grad_norm": 0.16216471791267395,
      "learning_rate": 0.001,
      "loss": 2.317795991897583,
      "step": 3324
    },
    {
      "epoch": 0.33872930680666663,
      "grad_norm": 0.15358373522758484,
      "learning_rate": 0.001,
      "loss": 2.34394907951355,
      "step": 3326
    },
    {
      "epoch": 0.3389329924992744,
      "grad_norm": 0.15510942041873932,
      "learning_rate": 0.001,
      "loss": 2.063309669494629,
      "step": 3328
    },
    {
      "epoch": 0.3391366781918821,
      "grad_norm": 0.13730761408805847,
      "learning_rate": 0.001,
      "loss": 2.344235420227051,
      "step": 3330
    },
    {
      "epoch": 0.33934036388448985,
      "grad_norm": 0.12268651276826859,
      "learning_rate": 0.001,
      "loss": 2.1566734313964844,
      "step": 3332
    },
    {
      "epoch": 0.3395440495770976,
      "grad_norm": 0.11201062798500061,
      "learning_rate": 0.001,
      "loss": 2.321176052093506,
      "step": 3334
    },
    {
      "epoch": 0.3397477352697053,
      "grad_norm": 0.11148251593112946,
      "learning_rate": 0.001,
      "loss": 2.2463111877441406,
      "step": 3336
    },
    {
      "epoch": 0.33995142096231307,
      "grad_norm": 0.15328234434127808,
      "learning_rate": 0.001,
      "loss": 2.2150890827178955,
      "step": 3338
    },
    {
      "epoch": 0.34015510665492077,
      "grad_norm": 0.12060753256082535,
      "learning_rate": 0.001,
      "loss": 2.2593460083007812,
      "step": 3340
    },
    {
      "epoch": 0.3403587923475285,
      "grad_norm": 0.15244564414024353,
      "learning_rate": 0.001,
      "loss": 2.2068536281585693,
      "step": 3342
    },
    {
      "epoch": 0.3405624780401363,
      "grad_norm": 0.1992480456829071,
      "learning_rate": 0.001,
      "loss": 2.2674362659454346,
      "step": 3344
    },
    {
      "epoch": 0.340766163732744,
      "grad_norm": 0.13807161152362823,
      "learning_rate": 0.001,
      "loss": 2.1386144161224365,
      "step": 3346
    },
    {
      "epoch": 0.34096984942535175,
      "grad_norm": 0.16374462842941284,
      "learning_rate": 0.001,
      "loss": 2.1343448162078857,
      "step": 3348
    },
    {
      "epoch": 0.3411735351179595,
      "grad_norm": 0.21406373381614685,
      "learning_rate": 0.001,
      "loss": 2.2142913341522217,
      "step": 3350
    },
    {
      "epoch": 0.3413772208105672,
      "grad_norm": 0.10800709575414658,
      "learning_rate": 0.001,
      "loss": 2.1894657611846924,
      "step": 3352
    },
    {
      "epoch": 0.34158090650317496,
      "grad_norm": 0.1559574455022812,
      "learning_rate": 0.001,
      "loss": 2.3224587440490723,
      "step": 3354
    },
    {
      "epoch": 0.34178459219578267,
      "grad_norm": 0.10454726964235306,
      "learning_rate": 0.001,
      "loss": 2.2975058555603027,
      "step": 3356
    },
    {
      "epoch": 0.3419882778883904,
      "grad_norm": 0.11455735564231873,
      "learning_rate": 0.001,
      "loss": 2.1378173828125,
      "step": 3358
    },
    {
      "epoch": 0.3421919635809982,
      "grad_norm": 0.12441648542881012,
      "learning_rate": 0.001,
      "loss": 2.1160242557525635,
      "step": 3360
    },
    {
      "epoch": 0.3423956492736059,
      "grad_norm": 0.12808935344219208,
      "learning_rate": 0.001,
      "loss": 2.2875115871429443,
      "step": 3362
    },
    {
      "epoch": 0.34259933496621364,
      "grad_norm": 0.13400490581989288,
      "learning_rate": 0.001,
      "loss": 2.2607433795928955,
      "step": 3364
    },
    {
      "epoch": 0.3428030206588214,
      "grad_norm": 0.15365806221961975,
      "learning_rate": 0.001,
      "loss": 2.2491941452026367,
      "step": 3366
    },
    {
      "epoch": 0.3430067063514291,
      "grad_norm": 0.11202633380889893,
      "learning_rate": 0.001,
      "loss": 2.2208032608032227,
      "step": 3368
    },
    {
      "epoch": 0.34321039204403686,
      "grad_norm": 0.12362506985664368,
      "learning_rate": 0.001,
      "loss": 2.225914478302002,
      "step": 3370
    },
    {
      "epoch": 0.34341407773664456,
      "grad_norm": 0.1497492641210556,
      "learning_rate": 0.001,
      "loss": 2.3328592777252197,
      "step": 3372
    },
    {
      "epoch": 0.3436177634292523,
      "grad_norm": 0.1446647346019745,
      "learning_rate": 0.001,
      "loss": 2.3135414123535156,
      "step": 3374
    },
    {
      "epoch": 0.3438214491218601,
      "grad_norm": 0.15581881999969482,
      "learning_rate": 0.001,
      "loss": 2.2201285362243652,
      "step": 3376
    },
    {
      "epoch": 0.3440251348144678,
      "grad_norm": 0.16309510171413422,
      "learning_rate": 0.001,
      "loss": 2.1942343711853027,
      "step": 3378
    },
    {
      "epoch": 0.34422882050707554,
      "grad_norm": 0.13959437608718872,
      "learning_rate": 0.001,
      "loss": 2.216275215148926,
      "step": 3380
    },
    {
      "epoch": 0.34443250619968324,
      "grad_norm": 0.12913787364959717,
      "learning_rate": 0.001,
      "loss": 2.0881664752960205,
      "step": 3382
    },
    {
      "epoch": 0.344636191892291,
      "grad_norm": 0.1351737678050995,
      "learning_rate": 0.001,
      "loss": 2.3026084899902344,
      "step": 3384
    },
    {
      "epoch": 0.34483987758489876,
      "grad_norm": 0.1502687931060791,
      "learning_rate": 0.001,
      "loss": 2.212451934814453,
      "step": 3386
    },
    {
      "epoch": 0.34504356327750646,
      "grad_norm": 0.11581950634717941,
      "learning_rate": 0.001,
      "loss": 2.2169382572174072,
      "step": 3388
    },
    {
      "epoch": 0.3452472489701142,
      "grad_norm": 0.12402158975601196,
      "learning_rate": 0.001,
      "loss": 2.2259278297424316,
      "step": 3390
    },
    {
      "epoch": 0.345450934662722,
      "grad_norm": 0.13077907264232635,
      "learning_rate": 0.001,
      "loss": 2.2257838249206543,
      "step": 3392
    },
    {
      "epoch": 0.3456546203553297,
      "grad_norm": 0.13757625222206116,
      "learning_rate": 0.001,
      "loss": 2.0628676414489746,
      "step": 3394
    },
    {
      "epoch": 0.34585830604793744,
      "grad_norm": 0.12965792417526245,
      "learning_rate": 0.001,
      "loss": 2.2818679809570312,
      "step": 3396
    },
    {
      "epoch": 0.34606199174054514,
      "grad_norm": 0.12401019036769867,
      "learning_rate": 0.001,
      "loss": 2.237403392791748,
      "step": 3398
    },
    {
      "epoch": 0.3462656774331529,
      "grad_norm": 0.11685767024755478,
      "learning_rate": 0.001,
      "loss": 2.1435723304748535,
      "step": 3400
    },
    {
      "epoch": 0.34646936312576065,
      "grad_norm": 0.13462549448013306,
      "learning_rate": 0.001,
      "loss": 2.22341251373291,
      "step": 3402
    },
    {
      "epoch": 0.34667304881836836,
      "grad_norm": 0.1125277429819107,
      "learning_rate": 0.001,
      "loss": 2.2102389335632324,
      "step": 3404
    },
    {
      "epoch": 0.3468767345109761,
      "grad_norm": 0.11834315210580826,
      "learning_rate": 0.001,
      "loss": 1.9741604328155518,
      "step": 3406
    },
    {
      "epoch": 0.3470804202035839,
      "grad_norm": 0.1330326348543167,
      "learning_rate": 0.001,
      "loss": 2.2921180725097656,
      "step": 3408
    },
    {
      "epoch": 0.3472841058961916,
      "grad_norm": 0.13048245012760162,
      "learning_rate": 0.001,
      "loss": 2.314887046813965,
      "step": 3410
    },
    {
      "epoch": 0.34748779158879933,
      "grad_norm": 0.139661505818367,
      "learning_rate": 0.001,
      "loss": 2.3506202697753906,
      "step": 3412
    },
    {
      "epoch": 0.34769147728140704,
      "grad_norm": 0.1188085675239563,
      "learning_rate": 0.001,
      "loss": 2.329133987426758,
      "step": 3414
    },
    {
      "epoch": 0.3478951629740148,
      "grad_norm": 0.1271413266658783,
      "learning_rate": 0.001,
      "loss": 2.246154546737671,
      "step": 3416
    },
    {
      "epoch": 0.34809884866662255,
      "grad_norm": 0.11312456429004669,
      "learning_rate": 0.001,
      "loss": 2.2862706184387207,
      "step": 3418
    },
    {
      "epoch": 0.34830253435923025,
      "grad_norm": 0.13578253984451294,
      "learning_rate": 0.001,
      "loss": 2.1927378177642822,
      "step": 3420
    },
    {
      "epoch": 0.348506220051838,
      "grad_norm": 0.11440230160951614,
      "learning_rate": 0.001,
      "loss": 2.324124813079834,
      "step": 3422
    },
    {
      "epoch": 0.34870990574444577,
      "grad_norm": 0.14648713171482086,
      "learning_rate": 0.001,
      "loss": 2.311583995819092,
      "step": 3424
    },
    {
      "epoch": 0.34891359143705347,
      "grad_norm": 0.17335565388202667,
      "learning_rate": 0.001,
      "loss": 2.2224926948547363,
      "step": 3426
    },
    {
      "epoch": 0.34911727712966123,
      "grad_norm": 0.11348274350166321,
      "learning_rate": 0.001,
      "loss": 2.148806571960449,
      "step": 3428
    },
    {
      "epoch": 0.34932096282226893,
      "grad_norm": 0.14697636663913727,
      "learning_rate": 0.001,
      "loss": 2.278208017349243,
      "step": 3430
    },
    {
      "epoch": 0.3495246485148767,
      "grad_norm": 0.09793109446763992,
      "learning_rate": 0.001,
      "loss": 2.271514415740967,
      "step": 3432
    },
    {
      "epoch": 0.34972833420748445,
      "grad_norm": 0.13675279915332794,
      "learning_rate": 0.001,
      "loss": 2.1934309005737305,
      "step": 3434
    },
    {
      "epoch": 0.34993201990009215,
      "grad_norm": 0.14020921289920807,
      "learning_rate": 0.001,
      "loss": 2.262037754058838,
      "step": 3436
    },
    {
      "epoch": 0.3501357055926999,
      "grad_norm": 0.12382035702466965,
      "learning_rate": 0.001,
      "loss": 2.3760738372802734,
      "step": 3438
    },
    {
      "epoch": 0.35033939128530767,
      "grad_norm": 0.13710129261016846,
      "learning_rate": 0.001,
      "loss": 2.0710654258728027,
      "step": 3440
    },
    {
      "epoch": 0.35054307697791537,
      "grad_norm": 0.11160828173160553,
      "learning_rate": 0.001,
      "loss": 2.3554530143737793,
      "step": 3442
    },
    {
      "epoch": 0.3507467626705231,
      "grad_norm": 0.15076620876789093,
      "learning_rate": 0.001,
      "loss": 2.1770100593566895,
      "step": 3444
    },
    {
      "epoch": 0.35095044836313083,
      "grad_norm": 0.11376726627349854,
      "learning_rate": 0.001,
      "loss": 2.1799774169921875,
      "step": 3446
    },
    {
      "epoch": 0.3511541340557386,
      "grad_norm": 0.11222446709871292,
      "learning_rate": 0.001,
      "loss": 2.1980838775634766,
      "step": 3448
    },
    {
      "epoch": 0.35135781974834635,
      "grad_norm": 0.1252325028181076,
      "learning_rate": 0.001,
      "loss": 2.175074815750122,
      "step": 3450
    },
    {
      "epoch": 0.35156150544095405,
      "grad_norm": 0.11914262175559998,
      "learning_rate": 0.001,
      "loss": 2.142869710922241,
      "step": 3452
    },
    {
      "epoch": 0.3517651911335618,
      "grad_norm": 0.17598606646060944,
      "learning_rate": 0.001,
      "loss": 2.1280109882354736,
      "step": 3454
    },
    {
      "epoch": 0.35196887682616956,
      "grad_norm": 0.1375114619731903,
      "learning_rate": 0.001,
      "loss": 2.1938817501068115,
      "step": 3456
    },
    {
      "epoch": 0.35217256251877727,
      "grad_norm": 0.12302275002002716,
      "learning_rate": 0.001,
      "loss": 2.3656857013702393,
      "step": 3458
    },
    {
      "epoch": 0.352376248211385,
      "grad_norm": 0.11187437176704407,
      "learning_rate": 0.001,
      "loss": 2.187161922454834,
      "step": 3460
    },
    {
      "epoch": 0.3525799339039927,
      "grad_norm": 0.11826854944229126,
      "learning_rate": 0.001,
      "loss": 2.156195640563965,
      "step": 3462
    },
    {
      "epoch": 0.3527836195966005,
      "grad_norm": 0.12288734316825867,
      "learning_rate": 0.001,
      "loss": 2.2573280334472656,
      "step": 3464
    },
    {
      "epoch": 0.35298730528920824,
      "grad_norm": 0.1305217295885086,
      "learning_rate": 0.001,
      "loss": 2.1970863342285156,
      "step": 3466
    },
    {
      "epoch": 0.35319099098181594,
      "grad_norm": 0.16518492996692657,
      "learning_rate": 0.001,
      "loss": 2.224515438079834,
      "step": 3468
    },
    {
      "epoch": 0.3533946766744237,
      "grad_norm": 0.1567467749118805,
      "learning_rate": 0.001,
      "loss": 2.3085126876831055,
      "step": 3470
    },
    {
      "epoch": 0.35359836236703146,
      "grad_norm": 0.12635159492492676,
      "learning_rate": 0.001,
      "loss": 2.3603787422180176,
      "step": 3472
    },
    {
      "epoch": 0.35380204805963916,
      "grad_norm": 0.11152909696102142,
      "learning_rate": 0.001,
      "loss": 2.144360065460205,
      "step": 3474
    },
    {
      "epoch": 0.3540057337522469,
      "grad_norm": 0.15233413875102997,
      "learning_rate": 0.001,
      "loss": 2.1906566619873047,
      "step": 3476
    },
    {
      "epoch": 0.3542094194448546,
      "grad_norm": 0.10214392095804214,
      "learning_rate": 0.001,
      "loss": 2.1637210845947266,
      "step": 3478
    },
    {
      "epoch": 0.3544131051374624,
      "grad_norm": 0.13837546110153198,
      "learning_rate": 0.001,
      "loss": 2.1627817153930664,
      "step": 3480
    },
    {
      "epoch": 0.35461679083007014,
      "grad_norm": 0.14367908239364624,
      "learning_rate": 0.001,
      "loss": 2.3208844661712646,
      "step": 3482
    },
    {
      "epoch": 0.35482047652267784,
      "grad_norm": 0.1294812709093094,
      "learning_rate": 0.001,
      "loss": 2.1451575756073,
      "step": 3484
    },
    {
      "epoch": 0.3550241622152856,
      "grad_norm": 0.13918744027614594,
      "learning_rate": 0.001,
      "loss": 2.0939278602600098,
      "step": 3486
    },
    {
      "epoch": 0.35522784790789336,
      "grad_norm": 0.20184211432933807,
      "learning_rate": 0.001,
      "loss": 2.1806681156158447,
      "step": 3488
    },
    {
      "epoch": 0.35543153360050106,
      "grad_norm": 0.13120372593402863,
      "learning_rate": 0.001,
      "loss": 2.327425003051758,
      "step": 3490
    },
    {
      "epoch": 0.3556352192931088,
      "grad_norm": 0.12535788118839264,
      "learning_rate": 0.001,
      "loss": 2.063114643096924,
      "step": 3492
    },
    {
      "epoch": 0.3558389049857165,
      "grad_norm": 0.17534857988357544,
      "learning_rate": 0.001,
      "loss": 2.224891185760498,
      "step": 3494
    },
    {
      "epoch": 0.3560425906783243,
      "grad_norm": 0.13031789660453796,
      "learning_rate": 0.001,
      "loss": 2.2582292556762695,
      "step": 3496
    },
    {
      "epoch": 0.35624627637093204,
      "grad_norm": 0.13159000873565674,
      "learning_rate": 0.001,
      "loss": 2.271498680114746,
      "step": 3498
    },
    {
      "epoch": 0.35644996206353974,
      "grad_norm": 0.12399876117706299,
      "learning_rate": 0.001,
      "loss": 2.2224297523498535,
      "step": 3500
    },
    {
      "epoch": 0.3566536477561475,
      "grad_norm": 0.14281699061393738,
      "learning_rate": 0.001,
      "loss": 2.2346043586730957,
      "step": 3502
    },
    {
      "epoch": 0.3568573334487552,
      "grad_norm": 0.12313243746757507,
      "learning_rate": 0.001,
      "loss": 2.285038948059082,
      "step": 3504
    },
    {
      "epoch": 0.35706101914136296,
      "grad_norm": 0.14311948418617249,
      "learning_rate": 0.001,
      "loss": 2.166233539581299,
      "step": 3506
    },
    {
      "epoch": 0.3572647048339707,
      "grad_norm": 0.12527725100517273,
      "learning_rate": 0.001,
      "loss": 2.084888458251953,
      "step": 3508
    },
    {
      "epoch": 0.3574683905265784,
      "grad_norm": 0.14751815795898438,
      "learning_rate": 0.001,
      "loss": 2.101661205291748,
      "step": 3510
    },
    {
      "epoch": 0.3576720762191862,
      "grad_norm": 0.12159454077482224,
      "learning_rate": 0.001,
      "loss": 2.186204671859741,
      "step": 3512
    },
    {
      "epoch": 0.35787576191179393,
      "grad_norm": 0.12828876078128815,
      "learning_rate": 0.001,
      "loss": 2.146432876586914,
      "step": 3514
    },
    {
      "epoch": 0.35807944760440164,
      "grad_norm": 0.17304585874080658,
      "learning_rate": 0.001,
      "loss": 2.2300801277160645,
      "step": 3516
    },
    {
      "epoch": 0.3582831332970094,
      "grad_norm": 0.1713605374097824,
      "learning_rate": 0.001,
      "loss": 2.134331703186035,
      "step": 3518
    },
    {
      "epoch": 0.3584868189896171,
      "grad_norm": 0.1493527740240097,
      "learning_rate": 0.001,
      "loss": 2.3100733757019043,
      "step": 3520
    },
    {
      "epoch": 0.35869050468222485,
      "grad_norm": 0.20885471999645233,
      "learning_rate": 0.001,
      "loss": 2.2061233520507812,
      "step": 3522
    },
    {
      "epoch": 0.3588941903748326,
      "grad_norm": 0.11212380975484848,
      "learning_rate": 0.001,
      "loss": 2.3083043098449707,
      "step": 3524
    },
    {
      "epoch": 0.3590978760674403,
      "grad_norm": 0.14110633730888367,
      "learning_rate": 0.001,
      "loss": 2.137645721435547,
      "step": 3526
    },
    {
      "epoch": 0.35930156176004807,
      "grad_norm": 0.1346132606267929,
      "learning_rate": 0.001,
      "loss": 2.1230106353759766,
      "step": 3528
    },
    {
      "epoch": 0.35950524745265583,
      "grad_norm": 0.11617866903543472,
      "learning_rate": 0.001,
      "loss": 2.184844493865967,
      "step": 3530
    },
    {
      "epoch": 0.35970893314526353,
      "grad_norm": 0.12836550176143646,
      "learning_rate": 0.001,
      "loss": 2.166102409362793,
      "step": 3532
    },
    {
      "epoch": 0.3599126188378713,
      "grad_norm": 0.12120220065116882,
      "learning_rate": 0.001,
      "loss": 2.257739543914795,
      "step": 3534
    },
    {
      "epoch": 0.360116304530479,
      "grad_norm": 0.1548142284154892,
      "learning_rate": 0.001,
      "loss": 2.286558151245117,
      "step": 3536
    },
    {
      "epoch": 0.36031999022308675,
      "grad_norm": 0.15226969122886658,
      "learning_rate": 0.001,
      "loss": 2.2948756217956543,
      "step": 3538
    },
    {
      "epoch": 0.3605236759156945,
      "grad_norm": 0.15626177191734314,
      "learning_rate": 0.001,
      "loss": 2.1712329387664795,
      "step": 3540
    },
    {
      "epoch": 0.3607273616083022,
      "grad_norm": 0.14793770015239716,
      "learning_rate": 0.001,
      "loss": 2.2965450286865234,
      "step": 3542
    },
    {
      "epoch": 0.36093104730090997,
      "grad_norm": 0.1480514258146286,
      "learning_rate": 0.001,
      "loss": 2.2480740547180176,
      "step": 3544
    },
    {
      "epoch": 0.3611347329935177,
      "grad_norm": 0.12531131505966187,
      "learning_rate": 0.001,
      "loss": 2.1405270099639893,
      "step": 3546
    },
    {
      "epoch": 0.36133841868612543,
      "grad_norm": 0.10371244698762894,
      "learning_rate": 0.001,
      "loss": 2.1949281692504883,
      "step": 3548
    },
    {
      "epoch": 0.3615421043787332,
      "grad_norm": 0.12973830103874207,
      "learning_rate": 0.001,
      "loss": 2.2924373149871826,
      "step": 3550
    },
    {
      "epoch": 0.3617457900713409,
      "grad_norm": 0.12010840326547623,
      "learning_rate": 0.001,
      "loss": 2.2210941314697266,
      "step": 3552
    },
    {
      "epoch": 0.36194947576394865,
      "grad_norm": 0.11944655328989029,
      "learning_rate": 0.001,
      "loss": 2.1628408432006836,
      "step": 3554
    },
    {
      "epoch": 0.3621531614565564,
      "grad_norm": 0.11929278075695038,
      "learning_rate": 0.001,
      "loss": 2.321613311767578,
      "step": 3556
    },
    {
      "epoch": 0.3623568471491641,
      "grad_norm": 0.1630430817604065,
      "learning_rate": 0.001,
      "loss": 2.1024322509765625,
      "step": 3558
    },
    {
      "epoch": 0.36256053284177187,
      "grad_norm": 0.11645588278770447,
      "learning_rate": 0.001,
      "loss": 2.2172842025756836,
      "step": 3560
    },
    {
      "epoch": 0.3627642185343796,
      "grad_norm": 0.11888258159160614,
      "learning_rate": 0.001,
      "loss": 2.1390767097473145,
      "step": 3562
    },
    {
      "epoch": 0.3629679042269873,
      "grad_norm": 0.12021204084157944,
      "learning_rate": 0.001,
      "loss": 2.211909770965576,
      "step": 3564
    },
    {
      "epoch": 0.3631715899195951,
      "grad_norm": 0.1186222955584526,
      "learning_rate": 0.001,
      "loss": 2.1866793632507324,
      "step": 3566
    },
    {
      "epoch": 0.3633752756122028,
      "grad_norm": 0.13954369723796844,
      "learning_rate": 0.001,
      "loss": 2.2162885665893555,
      "step": 3568
    },
    {
      "epoch": 0.36357896130481054,
      "grad_norm": 0.11827221512794495,
      "learning_rate": 0.001,
      "loss": 2.1529245376586914,
      "step": 3570
    },
    {
      "epoch": 0.3637826469974183,
      "grad_norm": 0.13028810918331146,
      "learning_rate": 0.001,
      "loss": 2.2999093532562256,
      "step": 3572
    },
    {
      "epoch": 0.363986332690026,
      "grad_norm": 0.12348278611898422,
      "learning_rate": 0.001,
      "loss": 2.300081729888916,
      "step": 3574
    },
    {
      "epoch": 0.36419001838263376,
      "grad_norm": 0.13068823516368866,
      "learning_rate": 0.001,
      "loss": 2.1208910942077637,
      "step": 3576
    },
    {
      "epoch": 0.3643937040752415,
      "grad_norm": 0.12682175636291504,
      "learning_rate": 0.001,
      "loss": 2.2576401233673096,
      "step": 3578
    },
    {
      "epoch": 0.3645973897678492,
      "grad_norm": 0.1343180537223816,
      "learning_rate": 0.001,
      "loss": 2.1144509315490723,
      "step": 3580
    },
    {
      "epoch": 0.364801075460457,
      "grad_norm": 0.11417565494775772,
      "learning_rate": 0.001,
      "loss": 2.2772815227508545,
      "step": 3582
    },
    {
      "epoch": 0.3650047611530647,
      "grad_norm": 0.12847739458084106,
      "learning_rate": 0.001,
      "loss": 2.231179714202881,
      "step": 3584
    },
    {
      "epoch": 0.36520844684567244,
      "grad_norm": 0.13191384077072144,
      "learning_rate": 0.001,
      "loss": 2.256033420562744,
      "step": 3586
    },
    {
      "epoch": 0.3654121325382802,
      "grad_norm": 0.12267439812421799,
      "learning_rate": 0.001,
      "loss": 2.2703378200531006,
      "step": 3588
    },
    {
      "epoch": 0.3656158182308879,
      "grad_norm": 0.10720360279083252,
      "learning_rate": 0.001,
      "loss": 2.1396660804748535,
      "step": 3590
    },
    {
      "epoch": 0.36581950392349566,
      "grad_norm": 0.12530778348445892,
      "learning_rate": 0.001,
      "loss": 2.2958426475524902,
      "step": 3592
    },
    {
      "epoch": 0.3660231896161034,
      "grad_norm": 0.11066869646310806,
      "learning_rate": 0.001,
      "loss": 2.217876672744751,
      "step": 3594
    },
    {
      "epoch": 0.3662268753087111,
      "grad_norm": 0.1444341093301773,
      "learning_rate": 0.001,
      "loss": 2.201347827911377,
      "step": 3596
    },
    {
      "epoch": 0.3664305610013189,
      "grad_norm": 0.16365212202072144,
      "learning_rate": 0.001,
      "loss": 2.0908889770507812,
      "step": 3598
    },
    {
      "epoch": 0.3666342466939266,
      "grad_norm": 0.12200598418712616,
      "learning_rate": 0.001,
      "loss": 2.169760227203369,
      "step": 3600
    },
    {
      "epoch": 0.36683793238653434,
      "grad_norm": 0.1339927762746811,
      "learning_rate": 0.001,
      "loss": 2.4054689407348633,
      "step": 3602
    },
    {
      "epoch": 0.3670416180791421,
      "grad_norm": 0.12095937132835388,
      "learning_rate": 0.001,
      "loss": 2.2685601711273193,
      "step": 3604
    },
    {
      "epoch": 0.3672453037717498,
      "grad_norm": 0.12811778485774994,
      "learning_rate": 0.001,
      "loss": 2.1034913063049316,
      "step": 3606
    },
    {
      "epoch": 0.36744898946435756,
      "grad_norm": 0.12413855642080307,
      "learning_rate": 0.001,
      "loss": 2.2220330238342285,
      "step": 3608
    },
    {
      "epoch": 0.3676526751569653,
      "grad_norm": 0.11905890703201294,
      "learning_rate": 0.001,
      "loss": 2.2348618507385254,
      "step": 3610
    },
    {
      "epoch": 0.367856360849573,
      "grad_norm": 0.16517575085163116,
      "learning_rate": 0.001,
      "loss": 2.1427841186523438,
      "step": 3612
    },
    {
      "epoch": 0.3680600465421808,
      "grad_norm": 0.12454784661531448,
      "learning_rate": 0.001,
      "loss": 2.2734227180480957,
      "step": 3614
    },
    {
      "epoch": 0.3682637322347885,
      "grad_norm": 0.11832301318645477,
      "learning_rate": 0.001,
      "loss": 2.2306952476501465,
      "step": 3616
    },
    {
      "epoch": 0.36846741792739623,
      "grad_norm": 0.17544803023338318,
      "learning_rate": 0.001,
      "loss": 2.2873363494873047,
      "step": 3618
    },
    {
      "epoch": 0.368671103620004,
      "grad_norm": 0.2149296998977661,
      "learning_rate": 0.001,
      "loss": 2.3324732780456543,
      "step": 3620
    },
    {
      "epoch": 0.3688747893126117,
      "grad_norm": 0.14849837124347687,
      "learning_rate": 0.001,
      "loss": 2.12753963470459,
      "step": 3622
    },
    {
      "epoch": 0.36907847500521945,
      "grad_norm": 0.1655086725950241,
      "learning_rate": 0.001,
      "loss": 2.208394765853882,
      "step": 3624
    },
    {
      "epoch": 0.36928216069782716,
      "grad_norm": 0.13763630390167236,
      "learning_rate": 0.001,
      "loss": 2.3052632808685303,
      "step": 3626
    },
    {
      "epoch": 0.3694858463904349,
      "grad_norm": 0.17215770483016968,
      "learning_rate": 0.001,
      "loss": 2.177114963531494,
      "step": 3628
    },
    {
      "epoch": 0.36968953208304267,
      "grad_norm": 0.12396519631147385,
      "learning_rate": 0.001,
      "loss": 2.207113027572632,
      "step": 3630
    },
    {
      "epoch": 0.3698932177756504,
      "grad_norm": 0.1406824290752411,
      "learning_rate": 0.001,
      "loss": 2.193310260772705,
      "step": 3632
    },
    {
      "epoch": 0.37009690346825813,
      "grad_norm": 0.2175106257200241,
      "learning_rate": 0.001,
      "loss": 2.328104019165039,
      "step": 3634
    },
    {
      "epoch": 0.3703005891608659,
      "grad_norm": 0.12080927938222885,
      "learning_rate": 0.001,
      "loss": 2.253566265106201,
      "step": 3636
    },
    {
      "epoch": 0.3705042748534736,
      "grad_norm": 0.12280062586069107,
      "learning_rate": 0.001,
      "loss": 2.2220849990844727,
      "step": 3638
    },
    {
      "epoch": 0.37070796054608135,
      "grad_norm": 0.13374705612659454,
      "learning_rate": 0.001,
      "loss": 2.206608772277832,
      "step": 3640
    },
    {
      "epoch": 0.37091164623868905,
      "grad_norm": 0.1464224010705948,
      "learning_rate": 0.001,
      "loss": 2.1871023178100586,
      "step": 3642
    },
    {
      "epoch": 0.3711153319312968,
      "grad_norm": 0.10993880033493042,
      "learning_rate": 0.001,
      "loss": 2.1627659797668457,
      "step": 3644
    },
    {
      "epoch": 0.37131901762390457,
      "grad_norm": 0.1275756061077118,
      "learning_rate": 0.001,
      "loss": 2.2842798233032227,
      "step": 3646
    },
    {
      "epoch": 0.37152270331651227,
      "grad_norm": 0.12706775963306427,
      "learning_rate": 0.001,
      "loss": 2.073704242706299,
      "step": 3648
    },
    {
      "epoch": 0.37172638900912003,
      "grad_norm": 0.17831000685691833,
      "learning_rate": 0.001,
      "loss": 2.2445874214172363,
      "step": 3650
    },
    {
      "epoch": 0.3719300747017278,
      "grad_norm": 0.11816295236349106,
      "learning_rate": 0.001,
      "loss": 2.3057661056518555,
      "step": 3652
    },
    {
      "epoch": 0.3721337603943355,
      "grad_norm": 0.14674772322177887,
      "learning_rate": 0.001,
      "loss": 2.022184133529663,
      "step": 3654
    },
    {
      "epoch": 0.37233744608694325,
      "grad_norm": 0.21128718554973602,
      "learning_rate": 0.001,
      "loss": 2.213656425476074,
      "step": 3656
    },
    {
      "epoch": 0.37254113177955095,
      "grad_norm": 0.16424496471881866,
      "learning_rate": 0.001,
      "loss": 2.204690933227539,
      "step": 3658
    },
    {
      "epoch": 0.3727448174721587,
      "grad_norm": 0.10061439871788025,
      "learning_rate": 0.001,
      "loss": 2.1947317123413086,
      "step": 3660
    },
    {
      "epoch": 0.37294850316476647,
      "grad_norm": 0.14796066284179688,
      "learning_rate": 0.001,
      "loss": 2.184986114501953,
      "step": 3662
    },
    {
      "epoch": 0.37315218885737417,
      "grad_norm": 0.192653626203537,
      "learning_rate": 0.001,
      "loss": 2.1062207221984863,
      "step": 3664
    },
    {
      "epoch": 0.3733558745499819,
      "grad_norm": 0.17699939012527466,
      "learning_rate": 0.001,
      "loss": 2.1638264656066895,
      "step": 3666
    },
    {
      "epoch": 0.3735595602425897,
      "grad_norm": 0.19161616265773773,
      "learning_rate": 0.001,
      "loss": 2.2993345260620117,
      "step": 3668
    },
    {
      "epoch": 0.3737632459351974,
      "grad_norm": 0.18388746678829193,
      "learning_rate": 0.001,
      "loss": 2.177663803100586,
      "step": 3670
    },
    {
      "epoch": 0.37396693162780514,
      "grad_norm": 0.13052256405353546,
      "learning_rate": 0.001,
      "loss": 2.0990395545959473,
      "step": 3672
    },
    {
      "epoch": 0.37417061732041285,
      "grad_norm": 0.1441250741481781,
      "learning_rate": 0.001,
      "loss": 2.2130026817321777,
      "step": 3674
    },
    {
      "epoch": 0.3743743030130206,
      "grad_norm": 0.12692983448505402,
      "learning_rate": 0.001,
      "loss": 2.389270305633545,
      "step": 3676
    },
    {
      "epoch": 0.37457798870562836,
      "grad_norm": 0.11333245038986206,
      "learning_rate": 0.001,
      "loss": 2.228152275085449,
      "step": 3678
    },
    {
      "epoch": 0.37478167439823606,
      "grad_norm": 0.13410510122776031,
      "learning_rate": 0.001,
      "loss": 2.2453980445861816,
      "step": 3680
    },
    {
      "epoch": 0.3749853600908438,
      "grad_norm": 0.13257771730422974,
      "learning_rate": 0.001,
      "loss": 2.1572608947753906,
      "step": 3682
    },
    {
      "epoch": 0.3751890457834516,
      "grad_norm": 0.1120760589838028,
      "learning_rate": 0.001,
      "loss": 2.1871235370635986,
      "step": 3684
    },
    {
      "epoch": 0.3753927314760593,
      "grad_norm": 0.1270664483308792,
      "learning_rate": 0.001,
      "loss": 2.171825885772705,
      "step": 3686
    },
    {
      "epoch": 0.37559641716866704,
      "grad_norm": 0.13181950151920319,
      "learning_rate": 0.001,
      "loss": 2.262918472290039,
      "step": 3688
    },
    {
      "epoch": 0.37580010286127474,
      "grad_norm": 0.12882135808467865,
      "learning_rate": 0.001,
      "loss": 2.1677327156066895,
      "step": 3690
    },
    {
      "epoch": 0.3760037885538825,
      "grad_norm": 0.15123261511325836,
      "learning_rate": 0.001,
      "loss": 2.1990318298339844,
      "step": 3692
    },
    {
      "epoch": 0.37620747424649026,
      "grad_norm": 0.11365501582622528,
      "learning_rate": 0.001,
      "loss": 2.1302285194396973,
      "step": 3694
    },
    {
      "epoch": 0.37641115993909796,
      "grad_norm": 0.1635390818119049,
      "learning_rate": 0.001,
      "loss": 2.28334379196167,
      "step": 3696
    },
    {
      "epoch": 0.3766148456317057,
      "grad_norm": 0.1165168508887291,
      "learning_rate": 0.001,
      "loss": 2.121373176574707,
      "step": 3698
    },
    {
      "epoch": 0.3768185313243135,
      "grad_norm": 0.12626992166042328,
      "learning_rate": 0.001,
      "loss": 2.0674617290496826,
      "step": 3700
    },
    {
      "epoch": 0.3770222170169212,
      "grad_norm": 0.12576761841773987,
      "learning_rate": 0.001,
      "loss": 2.2214322090148926,
      "step": 3702
    },
    {
      "epoch": 0.37722590270952894,
      "grad_norm": 0.11442302912473679,
      "learning_rate": 0.001,
      "loss": 2.210869312286377,
      "step": 3704
    },
    {
      "epoch": 0.37742958840213664,
      "grad_norm": 0.11342886090278625,
      "learning_rate": 0.001,
      "loss": 2.1229491233825684,
      "step": 3706
    },
    {
      "epoch": 0.3776332740947444,
      "grad_norm": 0.12016986310482025,
      "learning_rate": 0.001,
      "loss": 2.2149548530578613,
      "step": 3708
    },
    {
      "epoch": 0.37783695978735216,
      "grad_norm": 0.14021353423595428,
      "learning_rate": 0.001,
      "loss": 2.1446681022644043,
      "step": 3710
    },
    {
      "epoch": 0.37804064547995986,
      "grad_norm": 0.17110289633274078,
      "learning_rate": 0.001,
      "loss": 2.1752538681030273,
      "step": 3712
    },
    {
      "epoch": 0.3782443311725676,
      "grad_norm": 0.10583534091711044,
      "learning_rate": 0.001,
      "loss": 2.1433825492858887,
      "step": 3714
    },
    {
      "epoch": 0.3784480168651754,
      "grad_norm": 0.1057182028889656,
      "learning_rate": 0.001,
      "loss": 2.1997122764587402,
      "step": 3716
    },
    {
      "epoch": 0.3786517025577831,
      "grad_norm": 0.14835745096206665,
      "learning_rate": 0.001,
      "loss": 2.227942705154419,
      "step": 3718
    },
    {
      "epoch": 0.37885538825039083,
      "grad_norm": 0.13983874022960663,
      "learning_rate": 0.001,
      "loss": 2.139832019805908,
      "step": 3720
    },
    {
      "epoch": 0.37905907394299854,
      "grad_norm": 0.1081908568739891,
      "learning_rate": 0.001,
      "loss": 2.186354160308838,
      "step": 3722
    },
    {
      "epoch": 0.3792627596356063,
      "grad_norm": 0.1303173005580902,
      "learning_rate": 0.001,
      "loss": 2.261294364929199,
      "step": 3724
    },
    {
      "epoch": 0.37946644532821405,
      "grad_norm": 0.1151830404996872,
      "learning_rate": 0.001,
      "loss": 2.1910512447357178,
      "step": 3726
    },
    {
      "epoch": 0.37967013102082176,
      "grad_norm": 0.1598437875509262,
      "learning_rate": 0.001,
      "loss": 2.2650532722473145,
      "step": 3728
    },
    {
      "epoch": 0.3798738167134295,
      "grad_norm": 0.1348625272512436,
      "learning_rate": 0.001,
      "loss": 2.2618372440338135,
      "step": 3730
    },
    {
      "epoch": 0.38007750240603727,
      "grad_norm": 0.15862815082073212,
      "learning_rate": 0.001,
      "loss": 2.161686897277832,
      "step": 3732
    },
    {
      "epoch": 0.380281188098645,
      "grad_norm": 0.2144019454717636,
      "learning_rate": 0.001,
      "loss": 2.241938352584839,
      "step": 3734
    },
    {
      "epoch": 0.38048487379125273,
      "grad_norm": 0.17520995438098907,
      "learning_rate": 0.001,
      "loss": 2.2364273071289062,
      "step": 3736
    },
    {
      "epoch": 0.38068855948386043,
      "grad_norm": 0.12301590293645859,
      "learning_rate": 0.001,
      "loss": 2.1582586765289307,
      "step": 3738
    },
    {
      "epoch": 0.3808922451764682,
      "grad_norm": 0.16844573616981506,
      "learning_rate": 0.001,
      "loss": 2.1512451171875,
      "step": 3740
    },
    {
      "epoch": 0.38109593086907595,
      "grad_norm": 0.13104769587516785,
      "learning_rate": 0.001,
      "loss": 2.2368664741516113,
      "step": 3742
    },
    {
      "epoch": 0.38129961656168365,
      "grad_norm": 0.1358935832977295,
      "learning_rate": 0.001,
      "loss": 2.152050495147705,
      "step": 3744
    },
    {
      "epoch": 0.3815033022542914,
      "grad_norm": 0.13237006962299347,
      "learning_rate": 0.001,
      "loss": 2.2478222846984863,
      "step": 3746
    },
    {
      "epoch": 0.3817069879468991,
      "grad_norm": 0.12153082340955734,
      "learning_rate": 0.001,
      "loss": 2.219362497329712,
      "step": 3748
    },
    {
      "epoch": 0.38191067363950687,
      "grad_norm": 0.10859730839729309,
      "learning_rate": 0.001,
      "loss": 2.1928820610046387,
      "step": 3750
    },
    {
      "epoch": 0.38211435933211463,
      "grad_norm": 0.11205179244279861,
      "learning_rate": 0.001,
      "loss": 2.129220485687256,
      "step": 3752
    },
    {
      "epoch": 0.38231804502472233,
      "grad_norm": 0.12309841066598892,
      "learning_rate": 0.001,
      "loss": 2.273061513900757,
      "step": 3754
    },
    {
      "epoch": 0.3825217307173301,
      "grad_norm": 0.16733293235301971,
      "learning_rate": 0.001,
      "loss": 2.1769073009490967,
      "step": 3756
    },
    {
      "epoch": 0.38272541640993785,
      "grad_norm": 0.12014417350292206,
      "learning_rate": 0.001,
      "loss": 2.185882806777954,
      "step": 3758
    },
    {
      "epoch": 0.38292910210254555,
      "grad_norm": 0.13894550502300262,
      "learning_rate": 0.001,
      "loss": 2.2381861209869385,
      "step": 3760
    },
    {
      "epoch": 0.3831327877951533,
      "grad_norm": 0.1324264407157898,
      "learning_rate": 0.001,
      "loss": 2.1068241596221924,
      "step": 3762
    },
    {
      "epoch": 0.383336473487761,
      "grad_norm": 0.14980508387088776,
      "learning_rate": 0.001,
      "loss": 2.1060242652893066,
      "step": 3764
    },
    {
      "epoch": 0.38354015918036877,
      "grad_norm": 0.13755924999713898,
      "learning_rate": 0.001,
      "loss": 2.1736552715301514,
      "step": 3766
    },
    {
      "epoch": 0.3837438448729765,
      "grad_norm": 0.11933133751153946,
      "learning_rate": 0.001,
      "loss": 2.2253258228302,
      "step": 3768
    },
    {
      "epoch": 0.3839475305655842,
      "grad_norm": 0.11654568463563919,
      "learning_rate": 0.001,
      "loss": 2.194042444229126,
      "step": 3770
    },
    {
      "epoch": 0.384151216258192,
      "grad_norm": 0.120018370449543,
      "learning_rate": 0.001,
      "loss": 2.2425174713134766,
      "step": 3772
    },
    {
      "epoch": 0.38435490195079974,
      "grad_norm": 0.12910117208957672,
      "learning_rate": 0.001,
      "loss": 2.301889419555664,
      "step": 3774
    },
    {
      "epoch": 0.38455858764340745,
      "grad_norm": 0.14505784213542938,
      "learning_rate": 0.001,
      "loss": 2.2044858932495117,
      "step": 3776
    },
    {
      "epoch": 0.3847622733360152,
      "grad_norm": 0.10886447131633759,
      "learning_rate": 0.001,
      "loss": 2.184316635131836,
      "step": 3778
    },
    {
      "epoch": 0.3849659590286229,
      "grad_norm": 0.11740448325872421,
      "learning_rate": 0.001,
      "loss": 2.1325430870056152,
      "step": 3780
    },
    {
      "epoch": 0.38516964472123066,
      "grad_norm": 0.12306894361972809,
      "learning_rate": 0.001,
      "loss": 2.0760128498077393,
      "step": 3782
    },
    {
      "epoch": 0.3853733304138384,
      "grad_norm": 0.17106038331985474,
      "learning_rate": 0.001,
      "loss": 2.2451210021972656,
      "step": 3784
    },
    {
      "epoch": 0.3855770161064461,
      "grad_norm": 0.09727507084608078,
      "learning_rate": 0.001,
      "loss": 2.1599514484405518,
      "step": 3786
    },
    {
      "epoch": 0.3857807017990539,
      "grad_norm": 0.1376086175441742,
      "learning_rate": 0.001,
      "loss": 2.1848597526550293,
      "step": 3788
    },
    {
      "epoch": 0.38598438749166164,
      "grad_norm": 0.14288882911205292,
      "learning_rate": 0.001,
      "loss": 2.260162115097046,
      "step": 3790
    },
    {
      "epoch": 0.38618807318426934,
      "grad_norm": 0.1187720075249672,
      "learning_rate": 0.001,
      "loss": 2.252530097961426,
      "step": 3792
    },
    {
      "epoch": 0.3863917588768771,
      "grad_norm": 0.11900303512811661,
      "learning_rate": 0.001,
      "loss": 2.0526862144470215,
      "step": 3794
    },
    {
      "epoch": 0.3865954445694848,
      "grad_norm": 0.19291703402996063,
      "learning_rate": 0.001,
      "loss": 2.19840145111084,
      "step": 3796
    },
    {
      "epoch": 0.38679913026209256,
      "grad_norm": 0.10849013179540634,
      "learning_rate": 0.001,
      "loss": 2.1716816425323486,
      "step": 3798
    },
    {
      "epoch": 0.3870028159547003,
      "grad_norm": 0.19231291115283966,
      "learning_rate": 0.001,
      "loss": 2.2237918376922607,
      "step": 3800
    },
    {
      "epoch": 0.387206501647308,
      "grad_norm": 0.12965287268161774,
      "learning_rate": 0.001,
      "loss": 2.127094268798828,
      "step": 3802
    },
    {
      "epoch": 0.3874101873399158,
      "grad_norm": 0.14623716473579407,
      "learning_rate": 0.001,
      "loss": 2.2709500789642334,
      "step": 3804
    },
    {
      "epoch": 0.38761387303252354,
      "grad_norm": 0.11677421629428864,
      "learning_rate": 0.001,
      "loss": 2.20546293258667,
      "step": 3806
    },
    {
      "epoch": 0.38781755872513124,
      "grad_norm": 0.13427220284938812,
      "learning_rate": 0.001,
      "loss": 2.3063790798187256,
      "step": 3808
    },
    {
      "epoch": 0.388021244417739,
      "grad_norm": 0.1175718829035759,
      "learning_rate": 0.001,
      "loss": 2.188828468322754,
      "step": 3810
    },
    {
      "epoch": 0.3882249301103467,
      "grad_norm": 0.11965592205524445,
      "learning_rate": 0.001,
      "loss": 2.235511302947998,
      "step": 3812
    },
    {
      "epoch": 0.38842861580295446,
      "grad_norm": 0.13712123036384583,
      "learning_rate": 0.001,
      "loss": 2.1606624126434326,
      "step": 3814
    },
    {
      "epoch": 0.3886323014955622,
      "grad_norm": 0.12492693215608597,
      "learning_rate": 0.001,
      "loss": 2.334223508834839,
      "step": 3816
    },
    {
      "epoch": 0.3888359871881699,
      "grad_norm": 0.11826645582914352,
      "learning_rate": 0.001,
      "loss": 2.1097614765167236,
      "step": 3818
    },
    {
      "epoch": 0.3890396728807777,
      "grad_norm": 0.16619405150413513,
      "learning_rate": 0.001,
      "loss": 2.24014949798584,
      "step": 3820
    },
    {
      "epoch": 0.38924335857338543,
      "grad_norm": 0.10866484045982361,
      "learning_rate": 0.001,
      "loss": 2.175229549407959,
      "step": 3822
    },
    {
      "epoch": 0.38944704426599314,
      "grad_norm": 0.12609688937664032,
      "learning_rate": 0.001,
      "loss": 1.996896743774414,
      "step": 3824
    },
    {
      "epoch": 0.3896507299586009,
      "grad_norm": 0.139070063829422,
      "learning_rate": 0.001,
      "loss": 2.2142958641052246,
      "step": 3826
    },
    {
      "epoch": 0.3898544156512086,
      "grad_norm": 0.14254498481750488,
      "learning_rate": 0.001,
      "loss": 2.26951265335083,
      "step": 3828
    },
    {
      "epoch": 0.39005810134381635,
      "grad_norm": 0.11034014821052551,
      "learning_rate": 0.001,
      "loss": 2.1882972717285156,
      "step": 3830
    },
    {
      "epoch": 0.3902617870364241,
      "grad_norm": 0.12613648176193237,
      "learning_rate": 0.001,
      "loss": 2.2379891872406006,
      "step": 3832
    },
    {
      "epoch": 0.3904654727290318,
      "grad_norm": 0.15869705379009247,
      "learning_rate": 0.001,
      "loss": 2.24969744682312,
      "step": 3834
    },
    {
      "epoch": 0.3906691584216396,
      "grad_norm": 0.125254824757576,
      "learning_rate": 0.001,
      "loss": 1.9954802989959717,
      "step": 3836
    },
    {
      "epoch": 0.39087284411424733,
      "grad_norm": 0.162289559841156,
      "learning_rate": 0.001,
      "loss": 2.2378034591674805,
      "step": 3838
    },
    {
      "epoch": 0.39107652980685503,
      "grad_norm": 0.12688472867012024,
      "learning_rate": 0.001,
      "loss": 2.250728130340576,
      "step": 3840
    },
    {
      "epoch": 0.3912802154994628,
      "grad_norm": 0.12841850519180298,
      "learning_rate": 0.001,
      "loss": 2.248108386993408,
      "step": 3842
    },
    {
      "epoch": 0.3914839011920705,
      "grad_norm": 0.123918816447258,
      "learning_rate": 0.001,
      "loss": 2.153231143951416,
      "step": 3844
    },
    {
      "epoch": 0.39168758688467825,
      "grad_norm": 0.12733371555805206,
      "learning_rate": 0.001,
      "loss": 2.1420934200286865,
      "step": 3846
    },
    {
      "epoch": 0.391891272577286,
      "grad_norm": 0.12671947479248047,
      "learning_rate": 0.001,
      "loss": 2.3961753845214844,
      "step": 3848
    },
    {
      "epoch": 0.3920949582698937,
      "grad_norm": 0.10436997562646866,
      "learning_rate": 0.001,
      "loss": 2.1273133754730225,
      "step": 3850
    },
    {
      "epoch": 0.39229864396250147,
      "grad_norm": 0.12703107297420502,
      "learning_rate": 0.001,
      "loss": 2.265791893005371,
      "step": 3852
    },
    {
      "epoch": 0.39250232965510923,
      "grad_norm": 0.12814073264598846,
      "learning_rate": 0.001,
      "loss": 2.161802053451538,
      "step": 3854
    },
    {
      "epoch": 0.39270601534771693,
      "grad_norm": 0.11687880754470825,
      "learning_rate": 0.001,
      "loss": 2.0546135902404785,
      "step": 3856
    },
    {
      "epoch": 0.3929097010403247,
      "grad_norm": 0.11961738020181656,
      "learning_rate": 0.001,
      "loss": 2.1122546195983887,
      "step": 3858
    },
    {
      "epoch": 0.3931133867329324,
      "grad_norm": 0.12406721711158752,
      "learning_rate": 0.001,
      "loss": 2.275146961212158,
      "step": 3860
    },
    {
      "epoch": 0.39331707242554015,
      "grad_norm": 0.12284654378890991,
      "learning_rate": 0.001,
      "loss": 2.2611098289489746,
      "step": 3862
    },
    {
      "epoch": 0.3935207581181479,
      "grad_norm": 0.13419759273529053,
      "learning_rate": 0.001,
      "loss": 2.303778886795044,
      "step": 3864
    },
    {
      "epoch": 0.3937244438107556,
      "grad_norm": 0.11473444849252701,
      "learning_rate": 0.001,
      "loss": 2.218348979949951,
      "step": 3866
    },
    {
      "epoch": 0.39392812950336337,
      "grad_norm": 0.14503128826618195,
      "learning_rate": 0.001,
      "loss": 2.141663074493408,
      "step": 3868
    },
    {
      "epoch": 0.3941318151959711,
      "grad_norm": 0.14251893758773804,
      "learning_rate": 0.001,
      "loss": 2.238205671310425,
      "step": 3870
    },
    {
      "epoch": 0.3943355008885788,
      "grad_norm": 0.121541328728199,
      "learning_rate": 0.001,
      "loss": 2.337467670440674,
      "step": 3872
    },
    {
      "epoch": 0.3945391865811866,
      "grad_norm": 0.15428981184959412,
      "learning_rate": 0.001,
      "loss": 2.1983578205108643,
      "step": 3874
    },
    {
      "epoch": 0.3947428722737943,
      "grad_norm": 0.13934384286403656,
      "learning_rate": 0.001,
      "loss": 2.198909282684326,
      "step": 3876
    },
    {
      "epoch": 0.39494655796640205,
      "grad_norm": 0.15244385600090027,
      "learning_rate": 0.001,
      "loss": 2.273414373397827,
      "step": 3878
    },
    {
      "epoch": 0.3951502436590098,
      "grad_norm": 0.12597526609897614,
      "learning_rate": 0.001,
      "loss": 2.2298169136047363,
      "step": 3880
    },
    {
      "epoch": 0.3953539293516175,
      "grad_norm": 0.13279296457767487,
      "learning_rate": 0.001,
      "loss": 2.2835330963134766,
      "step": 3882
    },
    {
      "epoch": 0.39555761504422526,
      "grad_norm": 0.1485043615102768,
      "learning_rate": 0.001,
      "loss": 2.1471104621887207,
      "step": 3884
    },
    {
      "epoch": 0.39576130073683297,
      "grad_norm": 0.11506547033786774,
      "learning_rate": 0.001,
      "loss": 2.0998849868774414,
      "step": 3886
    },
    {
      "epoch": 0.3959649864294407,
      "grad_norm": 0.13874870538711548,
      "learning_rate": 0.001,
      "loss": 2.209096670150757,
      "step": 3888
    },
    {
      "epoch": 0.3961686721220485,
      "grad_norm": 0.13944384455680847,
      "learning_rate": 0.001,
      "loss": 2.2507729530334473,
      "step": 3890
    },
    {
      "epoch": 0.3963723578146562,
      "grad_norm": 0.12651324272155762,
      "learning_rate": 0.001,
      "loss": 2.146869659423828,
      "step": 3892
    },
    {
      "epoch": 0.39657604350726394,
      "grad_norm": 0.14104998111724854,
      "learning_rate": 0.001,
      "loss": 2.2991650104522705,
      "step": 3894
    },
    {
      "epoch": 0.3967797291998717,
      "grad_norm": 0.15078434348106384,
      "learning_rate": 0.001,
      "loss": 2.0582215785980225,
      "step": 3896
    },
    {
      "epoch": 0.3969834148924794,
      "grad_norm": 0.15589860081672668,
      "learning_rate": 0.001,
      "loss": 2.1256115436553955,
      "step": 3898
    },
    {
      "epoch": 0.39718710058508716,
      "grad_norm": 0.12559843063354492,
      "learning_rate": 0.001,
      "loss": 2.338071346282959,
      "step": 3900
    },
    {
      "epoch": 0.39739078627769486,
      "grad_norm": 0.1314842253923416,
      "learning_rate": 0.001,
      "loss": 2.256688117980957,
      "step": 3902
    },
    {
      "epoch": 0.3975944719703026,
      "grad_norm": 0.12948621809482574,
      "learning_rate": 0.001,
      "loss": 2.1566288471221924,
      "step": 3904
    },
    {
      "epoch": 0.3977981576629104,
      "grad_norm": 0.1768215447664261,
      "learning_rate": 0.001,
      "loss": 2.105041265487671,
      "step": 3906
    },
    {
      "epoch": 0.3980018433555181,
      "grad_norm": 0.12003377825021744,
      "learning_rate": 0.001,
      "loss": 2.2268145084381104,
      "step": 3908
    },
    {
      "epoch": 0.39820552904812584,
      "grad_norm": 0.12539687752723694,
      "learning_rate": 0.001,
      "loss": 2.1561403274536133,
      "step": 3910
    },
    {
      "epoch": 0.3984092147407336,
      "grad_norm": 0.16727659106254578,
      "learning_rate": 0.001,
      "loss": 2.1793789863586426,
      "step": 3912
    },
    {
      "epoch": 0.3986129004333413,
      "grad_norm": 0.10681556910276413,
      "learning_rate": 0.001,
      "loss": 2.175724983215332,
      "step": 3914
    },
    {
      "epoch": 0.39881658612594906,
      "grad_norm": 0.11855163425207138,
      "learning_rate": 0.001,
      "loss": 2.148562431335449,
      "step": 3916
    },
    {
      "epoch": 0.39902027181855676,
      "grad_norm": 0.15125872194766998,
      "learning_rate": 0.001,
      "loss": 2.203880548477173,
      "step": 3918
    },
    {
      "epoch": 0.3992239575111645,
      "grad_norm": 0.15261207520961761,
      "learning_rate": 0.001,
      "loss": 2.302281618118286,
      "step": 3920
    },
    {
      "epoch": 0.3994276432037723,
      "grad_norm": 0.13579753041267395,
      "learning_rate": 0.001,
      "loss": 2.2890148162841797,
      "step": 3922
    },
    {
      "epoch": 0.39963132889638,
      "grad_norm": 0.15956062078475952,
      "learning_rate": 0.001,
      "loss": 2.347428560256958,
      "step": 3924
    },
    {
      "epoch": 0.39983501458898774,
      "grad_norm": 0.12418688833713531,
      "learning_rate": 0.001,
      "loss": 2.1371448040008545,
      "step": 3926
    },
    {
      "epoch": 0.4000387002815955,
      "grad_norm": 0.14112235605716705,
      "learning_rate": 0.001,
      "loss": 2.138265609741211,
      "step": 3928
    },
    {
      "epoch": 0.4002423859742032,
      "grad_norm": 0.13630786538124084,
      "learning_rate": 0.001,
      "loss": 2.3567728996276855,
      "step": 3930
    },
    {
      "epoch": 0.40044607166681095,
      "grad_norm": 0.12442337721586227,
      "learning_rate": 0.001,
      "loss": 2.147681713104248,
      "step": 3932
    },
    {
      "epoch": 0.40064975735941866,
      "grad_norm": 0.1255188286304474,
      "learning_rate": 0.001,
      "loss": 2.102877140045166,
      "step": 3934
    },
    {
      "epoch": 0.4008534430520264,
      "grad_norm": 0.10882001370191574,
      "learning_rate": 0.001,
      "loss": 2.22833251953125,
      "step": 3936
    },
    {
      "epoch": 0.4010571287446342,
      "grad_norm": 0.14413021504878998,
      "learning_rate": 0.001,
      "loss": 2.2805018424987793,
      "step": 3938
    },
    {
      "epoch": 0.4012608144372419,
      "grad_norm": 0.16915743052959442,
      "learning_rate": 0.001,
      "loss": 2.276557683944702,
      "step": 3940
    },
    {
      "epoch": 0.40146450012984963,
      "grad_norm": 0.1425083875656128,
      "learning_rate": 0.001,
      "loss": 2.2889440059661865,
      "step": 3942
    },
    {
      "epoch": 0.4016681858224574,
      "grad_norm": 0.12284961342811584,
      "learning_rate": 0.001,
      "loss": 2.2834134101867676,
      "step": 3944
    },
    {
      "epoch": 0.4018718715150651,
      "grad_norm": 0.16576682031154633,
      "learning_rate": 0.001,
      "loss": 2.2051684856414795,
      "step": 3946
    },
    {
      "epoch": 0.40207555720767285,
      "grad_norm": 0.14463412761688232,
      "learning_rate": 0.001,
      "loss": 2.1574690341949463,
      "step": 3948
    },
    {
      "epoch": 0.40227924290028055,
      "grad_norm": 0.13224251568317413,
      "learning_rate": 0.001,
      "loss": 2.2559897899627686,
      "step": 3950
    },
    {
      "epoch": 0.4024829285928883,
      "grad_norm": 0.12155351787805557,
      "learning_rate": 0.001,
      "loss": 2.3226184844970703,
      "step": 3952
    },
    {
      "epoch": 0.40268661428549607,
      "grad_norm": 0.11629988253116608,
      "learning_rate": 0.001,
      "loss": 2.267862558364868,
      "step": 3954
    },
    {
      "epoch": 0.40289029997810377,
      "grad_norm": 0.12121924012899399,
      "learning_rate": 0.001,
      "loss": 2.268049955368042,
      "step": 3956
    },
    {
      "epoch": 0.40309398567071153,
      "grad_norm": 0.10794980078935623,
      "learning_rate": 0.001,
      "loss": 2.214012861251831,
      "step": 3958
    },
    {
      "epoch": 0.4032976713633193,
      "grad_norm": 0.09973394870758057,
      "learning_rate": 0.001,
      "loss": 2.162726879119873,
      "step": 3960
    },
    {
      "epoch": 0.403501357055927,
      "grad_norm": 0.12381641566753387,
      "learning_rate": 0.001,
      "loss": 2.2478909492492676,
      "step": 3962
    },
    {
      "epoch": 0.40370504274853475,
      "grad_norm": 0.12303980439901352,
      "learning_rate": 0.001,
      "loss": 2.20383358001709,
      "step": 3964
    },
    {
      "epoch": 0.40390872844114245,
      "grad_norm": 0.1427641212940216,
      "learning_rate": 0.001,
      "loss": 2.166118621826172,
      "step": 3966
    },
    {
      "epoch": 0.4041124141337502,
      "grad_norm": 0.12834836542606354,
      "learning_rate": 0.001,
      "loss": 2.1520910263061523,
      "step": 3968
    },
    {
      "epoch": 0.40431609982635797,
      "grad_norm": 0.15708638727664948,
      "learning_rate": 0.001,
      "loss": 2.2303786277770996,
      "step": 3970
    },
    {
      "epoch": 0.40451978551896567,
      "grad_norm": 0.1256009340286255,
      "learning_rate": 0.001,
      "loss": 2.185821771621704,
      "step": 3972
    },
    {
      "epoch": 0.4047234712115734,
      "grad_norm": 0.1487552970647812,
      "learning_rate": 0.001,
      "loss": 2.1567461490631104,
      "step": 3974
    },
    {
      "epoch": 0.4049271569041812,
      "grad_norm": 0.1453331857919693,
      "learning_rate": 0.001,
      "loss": 2.222637176513672,
      "step": 3976
    },
    {
      "epoch": 0.4051308425967889,
      "grad_norm": 0.1543055772781372,
      "learning_rate": 0.001,
      "loss": 2.2868423461914062,
      "step": 3978
    },
    {
      "epoch": 0.40533452828939665,
      "grad_norm": 0.15814676880836487,
      "learning_rate": 0.001,
      "loss": 2.2537899017333984,
      "step": 3980
    },
    {
      "epoch": 0.40553821398200435,
      "grad_norm": 0.15745288133621216,
      "learning_rate": 0.001,
      "loss": 2.209897041320801,
      "step": 3982
    },
    {
      "epoch": 0.4057418996746121,
      "grad_norm": 0.13929052650928497,
      "learning_rate": 0.001,
      "loss": 2.324138641357422,
      "step": 3984
    },
    {
      "epoch": 0.40594558536721986,
      "grad_norm": 0.12905153632164001,
      "learning_rate": 0.001,
      "loss": 2.1976261138916016,
      "step": 3986
    },
    {
      "epoch": 0.40614927105982757,
      "grad_norm": 0.1111786738038063,
      "learning_rate": 0.001,
      "loss": 2.212836265563965,
      "step": 3988
    },
    {
      "epoch": 0.4063529567524353,
      "grad_norm": 0.13105790317058563,
      "learning_rate": 0.001,
      "loss": 2.227750778198242,
      "step": 3990
    },
    {
      "epoch": 0.4065566424450431,
      "grad_norm": 0.10568863153457642,
      "learning_rate": 0.001,
      "loss": 2.035661220550537,
      "step": 3992
    },
    {
      "epoch": 0.4067603281376508,
      "grad_norm": 0.14333495497703552,
      "learning_rate": 0.001,
      "loss": 2.195624828338623,
      "step": 3994
    },
    {
      "epoch": 0.40696401383025854,
      "grad_norm": 0.14519420266151428,
      "learning_rate": 0.001,
      "loss": 2.2171239852905273,
      "step": 3996
    },
    {
      "epoch": 0.40716769952286624,
      "grad_norm": 0.13415592908859253,
      "learning_rate": 0.001,
      "loss": 2.1655707359313965,
      "step": 3998
    },
    {
      "epoch": 0.407371385215474,
      "grad_norm": 0.13211412727832794,
      "learning_rate": 0.001,
      "loss": 2.1199684143066406,
      "step": 4000
    },
    {
      "epoch": 0.40757507090808176,
      "grad_norm": 0.11821920424699783,
      "learning_rate": 0.001,
      "loss": 2.3044564723968506,
      "step": 4002
    },
    {
      "epoch": 0.40777875660068946,
      "grad_norm": 0.10921493172645569,
      "learning_rate": 0.001,
      "loss": 2.080066680908203,
      "step": 4004
    },
    {
      "epoch": 0.4079824422932972,
      "grad_norm": 0.12105704098939896,
      "learning_rate": 0.001,
      "loss": 2.1876745223999023,
      "step": 4006
    },
    {
      "epoch": 0.4081861279859049,
      "grad_norm": 0.16559045016765594,
      "learning_rate": 0.001,
      "loss": 2.1563096046447754,
      "step": 4008
    },
    {
      "epoch": 0.4083898136785127,
      "grad_norm": 0.1319076120853424,
      "learning_rate": 0.001,
      "loss": 2.1712636947631836,
      "step": 4010
    },
    {
      "epoch": 0.40859349937112044,
      "grad_norm": 0.11300410330295563,
      "learning_rate": 0.001,
      "loss": 2.1742453575134277,
      "step": 4012
    },
    {
      "epoch": 0.40879718506372814,
      "grad_norm": 0.11520037800073624,
      "learning_rate": 0.001,
      "loss": 2.1897411346435547,
      "step": 4014
    },
    {
      "epoch": 0.4090008707563359,
      "grad_norm": 0.12819744646549225,
      "learning_rate": 0.001,
      "loss": 2.090991735458374,
      "step": 4016
    },
    {
      "epoch": 0.40920455644894366,
      "grad_norm": 0.14020076394081116,
      "learning_rate": 0.001,
      "loss": 2.1483850479125977,
      "step": 4018
    },
    {
      "epoch": 0.40940824214155136,
      "grad_norm": 0.1369781792163849,
      "learning_rate": 0.001,
      "loss": 2.158473491668701,
      "step": 4020
    },
    {
      "epoch": 0.4096119278341591,
      "grad_norm": 0.13765113055706024,
      "learning_rate": 0.001,
      "loss": 2.182034492492676,
      "step": 4022
    },
    {
      "epoch": 0.4098156135267668,
      "grad_norm": 0.1326528638601303,
      "learning_rate": 0.001,
      "loss": 2.143368721008301,
      "step": 4024
    },
    {
      "epoch": 0.4100192992193746,
      "grad_norm": 0.104697585105896,
      "learning_rate": 0.001,
      "loss": 2.213972568511963,
      "step": 4026
    },
    {
      "epoch": 0.41022298491198234,
      "grad_norm": 0.2335682064294815,
      "learning_rate": 0.001,
      "loss": 2.2350687980651855,
      "step": 4028
    },
    {
      "epoch": 0.41042667060459004,
      "grad_norm": 0.13026103377342224,
      "learning_rate": 0.001,
      "loss": 2.2919516563415527,
      "step": 4030
    },
    {
      "epoch": 0.4106303562971978,
      "grad_norm": 0.13204964995384216,
      "learning_rate": 0.001,
      "loss": 2.2775168418884277,
      "step": 4032
    },
    {
      "epoch": 0.41083404198980555,
      "grad_norm": 0.16885264217853546,
      "learning_rate": 0.001,
      "loss": 2.20158052444458,
      "step": 4034
    },
    {
      "epoch": 0.41103772768241326,
      "grad_norm": 0.12631377577781677,
      "learning_rate": 0.001,
      "loss": 2.232248306274414,
      "step": 4036
    },
    {
      "epoch": 0.411241413375021,
      "grad_norm": 0.12087277323007584,
      "learning_rate": 0.001,
      "loss": 2.0278446674346924,
      "step": 4038
    },
    {
      "epoch": 0.4114450990676287,
      "grad_norm": 0.11374824494123459,
      "learning_rate": 0.001,
      "loss": 2.150179624557495,
      "step": 4040
    },
    {
      "epoch": 0.4116487847602365,
      "grad_norm": 0.11980725824832916,
      "learning_rate": 0.001,
      "loss": 2.178225517272949,
      "step": 4042
    },
    {
      "epoch": 0.41185247045284423,
      "grad_norm": 0.11329711973667145,
      "learning_rate": 0.001,
      "loss": 2.2083945274353027,
      "step": 4044
    },
    {
      "epoch": 0.41205615614545194,
      "grad_norm": 0.11443198472261429,
      "learning_rate": 0.001,
      "loss": 2.1106669902801514,
      "step": 4046
    },
    {
      "epoch": 0.4122598418380597,
      "grad_norm": 0.12176729738712311,
      "learning_rate": 0.001,
      "loss": 2.2621235847473145,
      "step": 4048
    },
    {
      "epoch": 0.41246352753066745,
      "grad_norm": 0.11784686148166656,
      "learning_rate": 0.001,
      "loss": 2.245405673980713,
      "step": 4050
    },
    {
      "epoch": 0.41266721322327515,
      "grad_norm": 0.14302575588226318,
      "learning_rate": 0.001,
      "loss": 2.256922721862793,
      "step": 4052
    },
    {
      "epoch": 0.4128708989158829,
      "grad_norm": 0.1184011772274971,
      "learning_rate": 0.001,
      "loss": 2.2725577354431152,
      "step": 4054
    },
    {
      "epoch": 0.4130745846084906,
      "grad_norm": 0.12380281835794449,
      "learning_rate": 0.001,
      "loss": 2.257274627685547,
      "step": 4056
    },
    {
      "epoch": 0.41327827030109837,
      "grad_norm": 0.17460589110851288,
      "learning_rate": 0.001,
      "loss": 2.2021567821502686,
      "step": 4058
    },
    {
      "epoch": 0.41348195599370613,
      "grad_norm": 0.1132892295718193,
      "learning_rate": 0.001,
      "loss": 2.136749744415283,
      "step": 4060
    },
    {
      "epoch": 0.41368564168631383,
      "grad_norm": 0.10061421245336533,
      "learning_rate": 0.001,
      "loss": 2.2540698051452637,
      "step": 4062
    },
    {
      "epoch": 0.4138893273789216,
      "grad_norm": 0.12785644829273224,
      "learning_rate": 0.001,
      "loss": 2.262770414352417,
      "step": 4064
    },
    {
      "epoch": 0.41409301307152935,
      "grad_norm": 0.11318495869636536,
      "learning_rate": 0.001,
      "loss": 2.201420783996582,
      "step": 4066
    },
    {
      "epoch": 0.41429669876413705,
      "grad_norm": 0.11267337948083878,
      "learning_rate": 0.001,
      "loss": 2.288902997970581,
      "step": 4068
    },
    {
      "epoch": 0.4145003844567448,
      "grad_norm": 0.13486526906490326,
      "learning_rate": 0.001,
      "loss": 2.110752820968628,
      "step": 4070
    },
    {
      "epoch": 0.4147040701493525,
      "grad_norm": 0.16018477082252502,
      "learning_rate": 0.001,
      "loss": 2.2451491355895996,
      "step": 4072
    },
    {
      "epoch": 0.41490775584196027,
      "grad_norm": 0.1452580839395523,
      "learning_rate": 0.001,
      "loss": 2.1350951194763184,
      "step": 4074
    },
    {
      "epoch": 0.415111441534568,
      "grad_norm": 0.15626394748687744,
      "learning_rate": 0.001,
      "loss": 2.0829038619995117,
      "step": 4076
    },
    {
      "epoch": 0.41531512722717573,
      "grad_norm": 0.1235283613204956,
      "learning_rate": 0.001,
      "loss": 2.2602500915527344,
      "step": 4078
    },
    {
      "epoch": 0.4155188129197835,
      "grad_norm": 0.167628213763237,
      "learning_rate": 0.001,
      "loss": 2.299485683441162,
      "step": 4080
    },
    {
      "epoch": 0.41572249861239124,
      "grad_norm": 0.14180535078048706,
      "learning_rate": 0.001,
      "loss": 2.1941943168640137,
      "step": 4082
    },
    {
      "epoch": 0.41592618430499895,
      "grad_norm": 0.11531230807304382,
      "learning_rate": 0.001,
      "loss": 2.275689125061035,
      "step": 4084
    },
    {
      "epoch": 0.4161298699976067,
      "grad_norm": 0.12500078976154327,
      "learning_rate": 0.001,
      "loss": 2.234076499938965,
      "step": 4086
    },
    {
      "epoch": 0.4163335556902144,
      "grad_norm": 0.14937688410282135,
      "learning_rate": 0.001,
      "loss": 2.1778955459594727,
      "step": 4088
    },
    {
      "epoch": 0.41653724138282217,
      "grad_norm": 0.23520752787590027,
      "learning_rate": 0.001,
      "loss": 2.2685680389404297,
      "step": 4090
    },
    {
      "epoch": 0.4167409270754299,
      "grad_norm": 0.14653757214546204,
      "learning_rate": 0.001,
      "loss": 2.0472235679626465,
      "step": 4092
    },
    {
      "epoch": 0.4169446127680376,
      "grad_norm": 0.11923457682132721,
      "learning_rate": 0.001,
      "loss": 2.2321295738220215,
      "step": 4094
    },
    {
      "epoch": 0.4171482984606454,
      "grad_norm": 0.13525165617465973,
      "learning_rate": 0.001,
      "loss": 2.17832612991333,
      "step": 4096
    },
    {
      "epoch": 0.41735198415325314,
      "grad_norm": 0.11429329216480255,
      "learning_rate": 0.001,
      "loss": 2.3208627700805664,
      "step": 4098
    },
    {
      "epoch": 0.41755566984586084,
      "grad_norm": 0.13810522854328156,
      "learning_rate": 0.001,
      "loss": 2.1860547065734863,
      "step": 4100
    },
    {
      "epoch": 0.4177593555384686,
      "grad_norm": 0.1656692922115326,
      "learning_rate": 0.001,
      "loss": 2.261390447616577,
      "step": 4102
    },
    {
      "epoch": 0.4179630412310763,
      "grad_norm": 0.10786822438240051,
      "learning_rate": 0.001,
      "loss": 2.236171007156372,
      "step": 4104
    },
    {
      "epoch": 0.41816672692368406,
      "grad_norm": 0.11479002237319946,
      "learning_rate": 0.001,
      "loss": 2.2370052337646484,
      "step": 4106
    },
    {
      "epoch": 0.4183704126162918,
      "grad_norm": 0.12531642615795135,
      "learning_rate": 0.001,
      "loss": 2.2474379539489746,
      "step": 4108
    },
    {
      "epoch": 0.4185740983088995,
      "grad_norm": 0.13567128777503967,
      "learning_rate": 0.001,
      "loss": 2.3014869689941406,
      "step": 4110
    },
    {
      "epoch": 0.4187777840015073,
      "grad_norm": 0.15669982135295868,
      "learning_rate": 0.001,
      "loss": 2.2179782390594482,
      "step": 4112
    },
    {
      "epoch": 0.41898146969411504,
      "grad_norm": 0.11240307241678238,
      "learning_rate": 0.001,
      "loss": 2.224113941192627,
      "step": 4114
    },
    {
      "epoch": 0.41918515538672274,
      "grad_norm": 0.1360252946615219,
      "learning_rate": 0.001,
      "loss": 2.208083152770996,
      "step": 4116
    },
    {
      "epoch": 0.4193888410793305,
      "grad_norm": 0.11606542021036148,
      "learning_rate": 0.001,
      "loss": 2.2967026233673096,
      "step": 4118
    },
    {
      "epoch": 0.4195925267719382,
      "grad_norm": 0.13782162964344025,
      "learning_rate": 0.001,
      "loss": 2.2680647373199463,
      "step": 4120
    },
    {
      "epoch": 0.41979621246454596,
      "grad_norm": 0.1258581429719925,
      "learning_rate": 0.001,
      "loss": 2.1320130825042725,
      "step": 4122
    },
    {
      "epoch": 0.4199998981571537,
      "grad_norm": 0.12317005544900894,
      "learning_rate": 0.001,
      "loss": 2.075666904449463,
      "step": 4124
    },
    {
      "epoch": 0.4202035838497614,
      "grad_norm": 0.11177681386470795,
      "learning_rate": 0.001,
      "loss": 2.208695888519287,
      "step": 4126
    },
    {
      "epoch": 0.4204072695423692,
      "grad_norm": 0.10535506159067154,
      "learning_rate": 0.001,
      "loss": 2.062082529067993,
      "step": 4128
    },
    {
      "epoch": 0.4206109552349769,
      "grad_norm": 0.1450238823890686,
      "learning_rate": 0.001,
      "loss": 2.297680377960205,
      "step": 4130
    },
    {
      "epoch": 0.42081464092758464,
      "grad_norm": 0.14394697546958923,
      "learning_rate": 0.001,
      "loss": 2.2634053230285645,
      "step": 4132
    },
    {
      "epoch": 0.4210183266201924,
      "grad_norm": 0.1193581372499466,
      "learning_rate": 0.001,
      "loss": 2.1792120933532715,
      "step": 4134
    },
    {
      "epoch": 0.4212220123128001,
      "grad_norm": 0.10828955471515656,
      "learning_rate": 0.001,
      "loss": 2.197519302368164,
      "step": 4136
    },
    {
      "epoch": 0.42142569800540786,
      "grad_norm": 0.14661340415477753,
      "learning_rate": 0.001,
      "loss": 2.3023147583007812,
      "step": 4138
    },
    {
      "epoch": 0.4216293836980156,
      "grad_norm": 0.12358168512582779,
      "learning_rate": 0.001,
      "loss": 2.2646687030792236,
      "step": 4140
    },
    {
      "epoch": 0.4218330693906233,
      "grad_norm": 0.1282944679260254,
      "learning_rate": 0.001,
      "loss": 2.177133798599243,
      "step": 4142
    },
    {
      "epoch": 0.4220367550832311,
      "grad_norm": 0.11844725906848907,
      "learning_rate": 0.001,
      "loss": 2.174407482147217,
      "step": 4144
    },
    {
      "epoch": 0.4222404407758388,
      "grad_norm": 0.11349865794181824,
      "learning_rate": 0.001,
      "loss": 2.351041793823242,
      "step": 4146
    },
    {
      "epoch": 0.42244412646844653,
      "grad_norm": 0.11026953905820847,
      "learning_rate": 0.001,
      "loss": 2.1562111377716064,
      "step": 4148
    },
    {
      "epoch": 0.4226478121610543,
      "grad_norm": 0.13089168071746826,
      "learning_rate": 0.001,
      "loss": 2.1656084060668945,
      "step": 4150
    },
    {
      "epoch": 0.422851497853662,
      "grad_norm": 0.13755269348621368,
      "learning_rate": 0.001,
      "loss": 2.1976048946380615,
      "step": 4152
    },
    {
      "epoch": 0.42305518354626975,
      "grad_norm": 0.12838555872440338,
      "learning_rate": 0.001,
      "loss": 2.2098851203918457,
      "step": 4154
    },
    {
      "epoch": 0.4232588692388775,
      "grad_norm": 0.13100522756576538,
      "learning_rate": 0.001,
      "loss": 2.192282199859619,
      "step": 4156
    },
    {
      "epoch": 0.4234625549314852,
      "grad_norm": 0.10611054301261902,
      "learning_rate": 0.001,
      "loss": 2.183992862701416,
      "step": 4158
    },
    {
      "epoch": 0.42366624062409297,
      "grad_norm": 0.12371915578842163,
      "learning_rate": 0.001,
      "loss": 2.2089614868164062,
      "step": 4160
    },
    {
      "epoch": 0.4238699263167007,
      "grad_norm": 0.1168675422668457,
      "learning_rate": 0.001,
      "loss": 2.21104097366333,
      "step": 4162
    },
    {
      "epoch": 0.42407361200930843,
      "grad_norm": 0.12465793639421463,
      "learning_rate": 0.001,
      "loss": 2.2150745391845703,
      "step": 4164
    },
    {
      "epoch": 0.4242772977019162,
      "grad_norm": 0.11862754076719284,
      "learning_rate": 0.0009998231966053747,
      "loss": 2.162632942199707,
      "step": 4166
    },
    {
      "epoch": 0.4244809833945239,
      "grad_norm": 0.11651857942342758,
      "learning_rate": 0.0009994695898161246,
      "loss": 2.123439311981201,
      "step": 4168
    },
    {
      "epoch": 0.42468466908713165,
      "grad_norm": 0.11577537655830383,
      "learning_rate": 0.0009991159830268742,
      "loss": 2.2830214500427246,
      "step": 4170
    },
    {
      "epoch": 0.4248883547797394,
      "grad_norm": 0.1239021047949791,
      "learning_rate": 0.0009987623762376238,
      "loss": 2.246725559234619,
      "step": 4172
    },
    {
      "epoch": 0.4250920404723471,
      "grad_norm": 0.13939408957958221,
      "learning_rate": 0.0009984087694483734,
      "loss": 2.1188125610351562,
      "step": 4174
    },
    {
      "epoch": 0.42529572616495487,
      "grad_norm": 0.10905332118272781,
      "learning_rate": 0.000998055162659123,
      "loss": 2.359175443649292,
      "step": 4176
    },
    {
      "epoch": 0.42549941185756257,
      "grad_norm": 0.11835073679685593,
      "learning_rate": 0.0009977015558698728,
      "loss": 2.1062569618225098,
      "step": 4178
    },
    {
      "epoch": 0.42570309755017033,
      "grad_norm": 0.12922857701778412,
      "learning_rate": 0.0009973479490806222,
      "loss": 2.207723617553711,
      "step": 4180
    },
    {
      "epoch": 0.4259067832427781,
      "grad_norm": 0.13281698524951935,
      "learning_rate": 0.000996994342291372,
      "loss": 2.1048521995544434,
      "step": 4182
    },
    {
      "epoch": 0.4261104689353858,
      "grad_norm": 0.1781170517206192,
      "learning_rate": 0.0009966407355021217,
      "loss": 2.2977778911590576,
      "step": 4184
    },
    {
      "epoch": 0.42631415462799355,
      "grad_norm": 0.11545410752296448,
      "learning_rate": 0.0009962871287128713,
      "loss": 2.2334914207458496,
      "step": 4186
    },
    {
      "epoch": 0.4265178403206013,
      "grad_norm": 0.11393097788095474,
      "learning_rate": 0.000995933521923621,
      "loss": 2.222301483154297,
      "step": 4188
    },
    {
      "epoch": 0.426721526013209,
      "grad_norm": 0.11272786557674408,
      "learning_rate": 0.0009955799151343705,
      "loss": 2.276956558227539,
      "step": 4190
    },
    {
      "epoch": 0.42692521170581677,
      "grad_norm": 0.14239895343780518,
      "learning_rate": 0.0009952263083451204,
      "loss": 2.2302489280700684,
      "step": 4192
    },
    {
      "epoch": 0.42712889739842447,
      "grad_norm": 0.13381977379322052,
      "learning_rate": 0.0009948727015558698,
      "loss": 2.2422430515289307,
      "step": 4194
    },
    {
      "epoch": 0.4273325830910322,
      "grad_norm": 0.1054474413394928,
      "learning_rate": 0.0009945190947666196,
      "loss": 2.10978364944458,
      "step": 4196
    },
    {
      "epoch": 0.42753626878364,
      "grad_norm": 0.11890409886837006,
      "learning_rate": 0.0009941654879773692,
      "loss": 2.2442421913146973,
      "step": 4198
    },
    {
      "epoch": 0.4277399544762477,
      "grad_norm": 0.17273283004760742,
      "learning_rate": 0.0009938118811881188,
      "loss": 2.158938407897949,
      "step": 4200
    },
    {
      "epoch": 0.42794364016885544,
      "grad_norm": 0.1469118446111679,
      "learning_rate": 0.0009934582743988684,
      "loss": 2.2233972549438477,
      "step": 4202
    },
    {
      "epoch": 0.4281473258614632,
      "grad_norm": 0.1263427585363388,
      "learning_rate": 0.000993104667609618,
      "loss": 2.1840033531188965,
      "step": 4204
    },
    {
      "epoch": 0.4283510115540709,
      "grad_norm": 0.13873954117298126,
      "learning_rate": 0.0009927510608203679,
      "loss": 2.3403615951538086,
      "step": 4206
    },
    {
      "epoch": 0.42855469724667866,
      "grad_norm": 0.1893831342458725,
      "learning_rate": 0.0009923974540311173,
      "loss": 2.1420304775238037,
      "step": 4208
    },
    {
      "epoch": 0.42875838293928636,
      "grad_norm": 0.11971204727888107,
      "learning_rate": 0.0009920438472418671,
      "loss": 2.156568765640259,
      "step": 4210
    },
    {
      "epoch": 0.4289620686318941,
      "grad_norm": 0.11398480832576752,
      "learning_rate": 0.0009916902404526167,
      "loss": 2.142240524291992,
      "step": 4212
    },
    {
      "epoch": 0.4291657543245019,
      "grad_norm": 0.14814355969429016,
      "learning_rate": 0.0009913366336633663,
      "loss": 2.2014827728271484,
      "step": 4214
    },
    {
      "epoch": 0.4293694400171096,
      "grad_norm": 0.14438605308532715,
      "learning_rate": 0.000990983026874116,
      "loss": 2.284620523452759,
      "step": 4216
    },
    {
      "epoch": 0.42957312570971734,
      "grad_norm": 0.11435892432928085,
      "learning_rate": 0.0009906294200848656,
      "loss": 2.195075511932373,
      "step": 4218
    },
    {
      "epoch": 0.4297768114023251,
      "grad_norm": 0.18329235911369324,
      "learning_rate": 0.0009902758132956154,
      "loss": 2.2264962196350098,
      "step": 4220
    },
    {
      "epoch": 0.4299804970949328,
      "grad_norm": 0.17036262154579163,
      "learning_rate": 0.000989922206506365,
      "loss": 2.233908176422119,
      "step": 4222
    },
    {
      "epoch": 0.43018418278754056,
      "grad_norm": 0.11597452312707901,
      "learning_rate": 0.0009895685997171146,
      "loss": 2.2492361068725586,
      "step": 4224
    },
    {
      "epoch": 0.43038786848014826,
      "grad_norm": 0.12845131754875183,
      "learning_rate": 0.0009892149929278643,
      "loss": 2.1409029960632324,
      "step": 4226
    },
    {
      "epoch": 0.430591554172756,
      "grad_norm": 0.142255038022995,
      "learning_rate": 0.0009888613861386139,
      "loss": 2.2832655906677246,
      "step": 4228
    },
    {
      "epoch": 0.4307952398653638,
      "grad_norm": 0.12171827256679535,
      "learning_rate": 0.0009885077793493635,
      "loss": 2.164909839630127,
      "step": 4230
    },
    {
      "epoch": 0.4309989255579715,
      "grad_norm": 0.1148129403591156,
      "learning_rate": 0.000988154172560113,
      "loss": 2.2611358165740967,
      "step": 4232
    },
    {
      "epoch": 0.43120261125057924,
      "grad_norm": 0.1508050560951233,
      "learning_rate": 0.000987800565770863,
      "loss": 2.241694688796997,
      "step": 4234
    },
    {
      "epoch": 0.431406296943187,
      "grad_norm": 0.12122002243995667,
      "learning_rate": 0.0009874469589816125,
      "loss": 2.1215121746063232,
      "step": 4236
    },
    {
      "epoch": 0.4316099826357947,
      "grad_norm": 0.11889788508415222,
      "learning_rate": 0.0009870933521923622,
      "loss": 2.3226418495178223,
      "step": 4238
    },
    {
      "epoch": 0.43181366832840246,
      "grad_norm": 0.1672208309173584,
      "learning_rate": 0.0009867397454031118,
      "loss": 2.28900146484375,
      "step": 4240
    },
    {
      "epoch": 0.43201735402101016,
      "grad_norm": 0.1263466626405716,
      "learning_rate": 0.0009863861386138614,
      "loss": 2.2135705947875977,
      "step": 4242
    },
    {
      "epoch": 0.4322210397136179,
      "grad_norm": 0.1357012540102005,
      "learning_rate": 0.000986032531824611,
      "loss": 2.096940755844116,
      "step": 4244
    },
    {
      "epoch": 0.4324247254062257,
      "grad_norm": 0.12798616290092468,
      "learning_rate": 0.0009856789250353606,
      "loss": 2.2253332138061523,
      "step": 4246
    },
    {
      "epoch": 0.4326284110988334,
      "grad_norm": 0.1374821811914444,
      "learning_rate": 0.0009853253182461105,
      "loss": 2.1667656898498535,
      "step": 4248
    },
    {
      "epoch": 0.43283209679144113,
      "grad_norm": 0.1020515114068985,
      "learning_rate": 0.00098497171145686,
      "loss": 2.3298535346984863,
      "step": 4250
    },
    {
      "epoch": 0.43303578248404884,
      "grad_norm": 0.13508231937885284,
      "learning_rate": 0.0009846181046676097,
      "loss": 2.233224630355835,
      "step": 4252
    },
    {
      "epoch": 0.4332394681766566,
      "grad_norm": 0.11288771033287048,
      "learning_rate": 0.0009842644978783593,
      "loss": 2.2855472564697266,
      "step": 4254
    },
    {
      "epoch": 0.43344315386926435,
      "grad_norm": 0.1412867307662964,
      "learning_rate": 0.000983910891089109,
      "loss": 2.2136459350585938,
      "step": 4256
    },
    {
      "epoch": 0.43364683956187206,
      "grad_norm": 0.17472001910209656,
      "learning_rate": 0.0009835572842998585,
      "loss": 2.2373223304748535,
      "step": 4258
    },
    {
      "epoch": 0.4338505252544798,
      "grad_norm": 0.10899540781974792,
      "learning_rate": 0.0009832036775106081,
      "loss": 2.114041328430176,
      "step": 4260
    },
    {
      "epoch": 0.43405421094708757,
      "grad_norm": 0.14947561919689178,
      "learning_rate": 0.000982850070721358,
      "loss": 2.330435276031494,
      "step": 4262
    },
    {
      "epoch": 0.4342578966396953,
      "grad_norm": 0.13055101037025452,
      "learning_rate": 0.0009824964639321076,
      "loss": 1.9940898418426514,
      "step": 4264
    },
    {
      "epoch": 0.43446158233230303,
      "grad_norm": 0.12248721718788147,
      "learning_rate": 0.0009821428571428572,
      "loss": 2.2250773906707764,
      "step": 4266
    },
    {
      "epoch": 0.43466526802491073,
      "grad_norm": 0.14312873780727386,
      "learning_rate": 0.0009817892503536068,
      "loss": 2.152599811553955,
      "step": 4268
    },
    {
      "epoch": 0.4348689537175185,
      "grad_norm": 0.1489303857088089,
      "learning_rate": 0.0009814356435643564,
      "loss": 2.0902323722839355,
      "step": 4270
    },
    {
      "epoch": 0.43507263941012625,
      "grad_norm": 0.1320357620716095,
      "learning_rate": 0.000981082036775106,
      "loss": 2.154132604598999,
      "step": 4272
    },
    {
      "epoch": 0.43527632510273395,
      "grad_norm": 0.14320127665996552,
      "learning_rate": 0.0009807284299858557,
      "loss": 2.138770818710327,
      "step": 4274
    },
    {
      "epoch": 0.4354800107953417,
      "grad_norm": 0.12237825989723206,
      "learning_rate": 0.0009803748231966055,
      "loss": 2.190579891204834,
      "step": 4276
    },
    {
      "epoch": 0.43568369648794947,
      "grad_norm": 0.1189723014831543,
      "learning_rate": 0.0009800212164073551,
      "loss": 2.231762409210205,
      "step": 4278
    },
    {
      "epoch": 0.43588738218055717,
      "grad_norm": 0.1552569717168808,
      "learning_rate": 0.0009796676096181047,
      "loss": 2.219942569732666,
      "step": 4280
    },
    {
      "epoch": 0.43609106787316493,
      "grad_norm": 0.14445845782756805,
      "learning_rate": 0.0009793140028288543,
      "loss": 2.2041656970977783,
      "step": 4282
    },
    {
      "epoch": 0.43629475356577263,
      "grad_norm": 0.11309345066547394,
      "learning_rate": 0.000978960396039604,
      "loss": 2.2544217109680176,
      "step": 4284
    },
    {
      "epoch": 0.4364984392583804,
      "grad_norm": 0.11555808037519455,
      "learning_rate": 0.0009786067892503536,
      "loss": 2.1852335929870605,
      "step": 4286
    },
    {
      "epoch": 0.43670212495098815,
      "grad_norm": 0.12599119544029236,
      "learning_rate": 0.0009782531824611032,
      "loss": 2.219628095626831,
      "step": 4288
    },
    {
      "epoch": 0.43690581064359585,
      "grad_norm": 0.10205013304948807,
      "learning_rate": 0.000977899575671853,
      "loss": 2.225231170654297,
      "step": 4290
    },
    {
      "epoch": 0.4371094963362036,
      "grad_norm": 0.11422756314277649,
      "learning_rate": 0.0009775459688826026,
      "loss": 2.1353530883789062,
      "step": 4292
    },
    {
      "epoch": 0.43731318202881136,
      "grad_norm": 0.18837063014507294,
      "learning_rate": 0.0009771923620933523,
      "loss": 2.1743812561035156,
      "step": 4294
    },
    {
      "epoch": 0.43751686772141907,
      "grad_norm": 0.11036856472492218,
      "learning_rate": 0.0009768387553041019,
      "loss": 2.293053388595581,
      "step": 4296
    },
    {
      "epoch": 0.4377205534140268,
      "grad_norm": 0.14470547437667847,
      "learning_rate": 0.0009764851485148515,
      "loss": 2.2285213470458984,
      "step": 4298
    },
    {
      "epoch": 0.4379242391066345,
      "grad_norm": 0.12040532380342484,
      "learning_rate": 0.0009761315417256011,
      "loss": 2.1392831802368164,
      "step": 4300
    },
    {
      "epoch": 0.4381279247992423,
      "grad_norm": 0.12665696442127228,
      "learning_rate": 0.0009757779349363508,
      "loss": 2.043128728866577,
      "step": 4302
    },
    {
      "epoch": 0.43833161049185004,
      "grad_norm": 0.11461839824914932,
      "learning_rate": 0.0009754243281471004,
      "loss": 2.1160292625427246,
      "step": 4304
    },
    {
      "epoch": 0.43853529618445775,
      "grad_norm": 0.1151294857263565,
      "learning_rate": 0.0009750707213578502,
      "loss": 2.1871838569641113,
      "step": 4306
    },
    {
      "epoch": 0.4387389818770655,
      "grad_norm": 0.13056394457817078,
      "learning_rate": 0.0009747171145685997,
      "loss": 2.2960352897644043,
      "step": 4308
    },
    {
      "epoch": 0.43894266756967326,
      "grad_norm": 0.12113704532384872,
      "learning_rate": 0.0009743635077793494,
      "loss": 2.1452252864837646,
      "step": 4310
    },
    {
      "epoch": 0.43914635326228096,
      "grad_norm": 0.113646499812603,
      "learning_rate": 0.000974009900990099,
      "loss": 2.261399269104004,
      "step": 4312
    },
    {
      "epoch": 0.4393500389548887,
      "grad_norm": 0.11756783723831177,
      "learning_rate": 0.0009736562942008486,
      "loss": 2.150895595550537,
      "step": 4314
    },
    {
      "epoch": 0.4395537246474964,
      "grad_norm": 0.15133149921894073,
      "learning_rate": 0.0009733026874115983,
      "loss": 2.252561330795288,
      "step": 4316
    },
    {
      "epoch": 0.4397574103401042,
      "grad_norm": 0.11054699122905731,
      "learning_rate": 0.000972949080622348,
      "loss": 2.166551113128662,
      "step": 4318
    },
    {
      "epoch": 0.43996109603271194,
      "grad_norm": 0.11198518425226212,
      "learning_rate": 0.0009725954738330977,
      "loss": 2.3004539012908936,
      "step": 4320
    },
    {
      "epoch": 0.44016478172531964,
      "grad_norm": 0.15852196514606476,
      "learning_rate": 0.0009722418670438473,
      "loss": 2.0939407348632812,
      "step": 4322
    },
    {
      "epoch": 0.4403684674179274,
      "grad_norm": 0.11138569563627243,
      "learning_rate": 0.0009718882602545969,
      "loss": 2.2390646934509277,
      "step": 4324
    },
    {
      "epoch": 0.44057215311053516,
      "grad_norm": 0.13162268698215485,
      "learning_rate": 0.0009715346534653465,
      "loss": 2.240595817565918,
      "step": 4326
    },
    {
      "epoch": 0.44077583880314286,
      "grad_norm": 0.11763791739940643,
      "learning_rate": 0.0009711810466760961,
      "loss": 2.1473238468170166,
      "step": 4328
    },
    {
      "epoch": 0.4409795244957506,
      "grad_norm": 0.15132999420166016,
      "learning_rate": 0.0009708274398868459,
      "loss": 2.350248336791992,
      "step": 4330
    },
    {
      "epoch": 0.4411832101883583,
      "grad_norm": 0.11212446540594101,
      "learning_rate": 0.0009704738330975955,
      "loss": 2.2141082286834717,
      "step": 4332
    },
    {
      "epoch": 0.4413868958809661,
      "grad_norm": 0.11372709274291992,
      "learning_rate": 0.0009701202263083452,
      "loss": 2.188967227935791,
      "step": 4334
    },
    {
      "epoch": 0.44159058157357384,
      "grad_norm": 0.1144038587808609,
      "learning_rate": 0.0009697666195190948,
      "loss": 2.17398738861084,
      "step": 4336
    },
    {
      "epoch": 0.44179426726618154,
      "grad_norm": 0.11627436429262161,
      "learning_rate": 0.0009694130127298444,
      "loss": 2.0848448276519775,
      "step": 4338
    },
    {
      "epoch": 0.4419979529587893,
      "grad_norm": 0.12132667005062103,
      "learning_rate": 0.000969059405940594,
      "loss": 2.1416335105895996,
      "step": 4340
    },
    {
      "epoch": 0.44220163865139706,
      "grad_norm": 0.12726832926273346,
      "learning_rate": 0.0009687057991513437,
      "loss": 2.2278149127960205,
      "step": 4342
    },
    {
      "epoch": 0.44240532434400476,
      "grad_norm": 0.12594659626483917,
      "learning_rate": 0.0009683521923620934,
      "loss": 2.166984796524048,
      "step": 4344
    },
    {
      "epoch": 0.4426090100366125,
      "grad_norm": 0.11619829386472702,
      "learning_rate": 0.000967998585572843,
      "loss": 2.1566598415374756,
      "step": 4346
    },
    {
      "epoch": 0.4428126957292202,
      "grad_norm": 0.11630271375179291,
      "learning_rate": 0.0009676449787835927,
      "loss": 2.310114622116089,
      "step": 4348
    },
    {
      "epoch": 0.443016381421828,
      "grad_norm": 0.11392687261104584,
      "learning_rate": 0.0009672913719943423,
      "loss": 2.1387219429016113,
      "step": 4350
    },
    {
      "epoch": 0.44322006711443573,
      "grad_norm": 0.11599093675613403,
      "learning_rate": 0.000966937765205092,
      "loss": 2.1854066848754883,
      "step": 4352
    },
    {
      "epoch": 0.44342375280704344,
      "grad_norm": 0.14233894646167755,
      "learning_rate": 0.0009665841584158416,
      "loss": 2.2717671394348145,
      "step": 4354
    },
    {
      "epoch": 0.4436274384996512,
      "grad_norm": 0.10979294031858444,
      "learning_rate": 0.0009662305516265912,
      "loss": 2.1178131103515625,
      "step": 4356
    },
    {
      "epoch": 0.44383112419225895,
      "grad_norm": 0.1332825869321823,
      "learning_rate": 0.0009658769448373409,
      "loss": 2.3106472492218018,
      "step": 4358
    },
    {
      "epoch": 0.44403480988486665,
      "grad_norm": 0.12598037719726562,
      "learning_rate": 0.0009655233380480905,
      "loss": 2.2123067378997803,
      "step": 4360
    },
    {
      "epoch": 0.4442384955774744,
      "grad_norm": 0.11085667461156845,
      "learning_rate": 0.0009651697312588403,
      "loss": 2.27787446975708,
      "step": 4362
    },
    {
      "epoch": 0.4444421812700821,
      "grad_norm": 0.10593260824680328,
      "learning_rate": 0.0009648161244695899,
      "loss": 2.1992437839508057,
      "step": 4364
    },
    {
      "epoch": 0.4446458669626899,
      "grad_norm": 0.11951328814029694,
      "learning_rate": 0.0009644625176803395,
      "loss": 2.2590012550354004,
      "step": 4366
    },
    {
      "epoch": 0.44484955265529763,
      "grad_norm": 0.11717348545789719,
      "learning_rate": 0.0009641089108910891,
      "loss": 2.206866502761841,
      "step": 4368
    },
    {
      "epoch": 0.44505323834790533,
      "grad_norm": 0.11078006029129028,
      "learning_rate": 0.0009637553041018387,
      "loss": 2.1674516201019287,
      "step": 4370
    },
    {
      "epoch": 0.4452569240405131,
      "grad_norm": 0.11901329457759857,
      "learning_rate": 0.0009634016973125884,
      "loss": 2.3275575637817383,
      "step": 4372
    },
    {
      "epoch": 0.4454606097331208,
      "grad_norm": 0.12545901536941528,
      "learning_rate": 0.000963048090523338,
      "loss": 2.1839704513549805,
      "step": 4374
    },
    {
      "epoch": 0.44566429542572855,
      "grad_norm": 0.10773919522762299,
      "learning_rate": 0.0009626944837340878,
      "loss": 2.0161609649658203,
      "step": 4376
    },
    {
      "epoch": 0.4458679811183363,
      "grad_norm": 0.13316288590431213,
      "learning_rate": 0.0009623408769448374,
      "loss": 2.3217997550964355,
      "step": 4378
    },
    {
      "epoch": 0.446071666810944,
      "grad_norm": 0.1481369286775589,
      "learning_rate": 0.000961987270155587,
      "loss": 2.169943332672119,
      "step": 4380
    },
    {
      "epoch": 0.44627535250355177,
      "grad_norm": 0.13584120571613312,
      "learning_rate": 0.0009616336633663366,
      "loss": 2.246072769165039,
      "step": 4382
    },
    {
      "epoch": 0.44647903819615953,
      "grad_norm": 0.15168927609920502,
      "learning_rate": 0.0009612800565770862,
      "loss": 2.2057156562805176,
      "step": 4384
    },
    {
      "epoch": 0.44668272388876723,
      "grad_norm": 0.11657849699258804,
      "learning_rate": 0.000960926449787836,
      "loss": 2.278423309326172,
      "step": 4386
    },
    {
      "epoch": 0.446886409581375,
      "grad_norm": 0.11838920414447784,
      "learning_rate": 0.0009605728429985856,
      "loss": 1.998152256011963,
      "step": 4388
    },
    {
      "epoch": 0.4470900952739827,
      "grad_norm": 0.18011502921581268,
      "learning_rate": 0.0009602192362093353,
      "loss": 2.292894124984741,
      "step": 4390
    },
    {
      "epoch": 0.44729378096659045,
      "grad_norm": 0.1484185755252838,
      "learning_rate": 0.0009598656294200849,
      "loss": 2.2178573608398438,
      "step": 4392
    },
    {
      "epoch": 0.4474974666591982,
      "grad_norm": 0.12086524069309235,
      "learning_rate": 0.0009595120226308345,
      "loss": 2.2623579502105713,
      "step": 4394
    },
    {
      "epoch": 0.4477011523518059,
      "grad_norm": 0.1690281629562378,
      "learning_rate": 0.0009591584158415841,
      "loss": 2.320298194885254,
      "step": 4396
    },
    {
      "epoch": 0.44790483804441367,
      "grad_norm": 0.15073975920677185,
      "learning_rate": 0.0009588048090523338,
      "loss": 2.1139626502990723,
      "step": 4398
    },
    {
      "epoch": 0.4481085237370214,
      "grad_norm": 0.13161656260490417,
      "learning_rate": 0.0009584512022630835,
      "loss": 2.252819538116455,
      "step": 4400
    },
    {
      "epoch": 0.4483122094296291,
      "grad_norm": 0.14897215366363525,
      "learning_rate": 0.0009580975954738331,
      "loss": 2.2076938152313232,
      "step": 4402
    },
    {
      "epoch": 0.4485158951222369,
      "grad_norm": 0.20672737061977386,
      "learning_rate": 0.0009577439886845828,
      "loss": 2.2520596981048584,
      "step": 4404
    },
    {
      "epoch": 0.4487195808148446,
      "grad_norm": 0.12532411515712738,
      "learning_rate": 0.0009573903818953324,
      "loss": 2.208359956741333,
      "step": 4406
    },
    {
      "epoch": 0.44892326650745235,
      "grad_norm": 0.11771109700202942,
      "learning_rate": 0.000957036775106082,
      "loss": 2.162736415863037,
      "step": 4408
    },
    {
      "epoch": 0.4491269522000601,
      "grad_norm": 0.17962761223316193,
      "learning_rate": 0.0009566831683168317,
      "loss": 2.2087392807006836,
      "step": 4410
    },
    {
      "epoch": 0.4493306378926678,
      "grad_norm": 0.11363206058740616,
      "learning_rate": 0.0009563295615275813,
      "loss": 2.1970815658569336,
      "step": 4412
    },
    {
      "epoch": 0.44953432358527556,
      "grad_norm": 0.1161012277007103,
      "learning_rate": 0.000955975954738331,
      "loss": 2.201359272003174,
      "step": 4414
    },
    {
      "epoch": 0.4497380092778833,
      "grad_norm": 0.1562829613685608,
      "learning_rate": 0.0009556223479490806,
      "loss": 2.1833038330078125,
      "step": 4416
    },
    {
      "epoch": 0.449941694970491,
      "grad_norm": 0.12501488626003265,
      "learning_rate": 0.0009552687411598303,
      "loss": 2.2777647972106934,
      "step": 4418
    },
    {
      "epoch": 0.4501453806630988,
      "grad_norm": 0.1460924744606018,
      "learning_rate": 0.00095491513437058,
      "loss": 2.3600897789001465,
      "step": 4420
    },
    {
      "epoch": 0.4503490663557065,
      "grad_norm": 0.1303459256887436,
      "learning_rate": 0.0009545615275813297,
      "loss": 2.1813454627990723,
      "step": 4422
    },
    {
      "epoch": 0.45055275204831424,
      "grad_norm": 0.12850989401340485,
      "learning_rate": 0.0009542079207920792,
      "loss": 2.2656354904174805,
      "step": 4424
    },
    {
      "epoch": 0.450756437740922,
      "grad_norm": 0.13071349263191223,
      "learning_rate": 0.0009538543140028288,
      "loss": 2.276623249053955,
      "step": 4426
    },
    {
      "epoch": 0.4509601234335297,
      "grad_norm": 0.11832624673843384,
      "learning_rate": 0.0009535007072135785,
      "loss": 2.14739990234375,
      "step": 4428
    },
    {
      "epoch": 0.45116380912613746,
      "grad_norm": 0.11867155879735947,
      "learning_rate": 0.0009531471004243281,
      "loss": 2.1903433799743652,
      "step": 4430
    },
    {
      "epoch": 0.4513674948187452,
      "grad_norm": 0.11709476262331009,
      "learning_rate": 0.0009527934936350779,
      "loss": 2.203371047973633,
      "step": 4432
    },
    {
      "epoch": 0.4515711805113529,
      "grad_norm": 0.1188405305147171,
      "learning_rate": 0.0009524398868458275,
      "loss": 2.1705784797668457,
      "step": 4434
    },
    {
      "epoch": 0.4517748662039607,
      "grad_norm": 0.11651252955198288,
      "learning_rate": 0.0009520862800565772,
      "loss": 2.209197521209717,
      "step": 4436
    },
    {
      "epoch": 0.4519785518965684,
      "grad_norm": 0.1263955682516098,
      "learning_rate": 0.0009517326732673267,
      "loss": 2.2215709686279297,
      "step": 4438
    },
    {
      "epoch": 0.45218223758917614,
      "grad_norm": 0.11652659624814987,
      "learning_rate": 0.0009513790664780763,
      "loss": 2.1509881019592285,
      "step": 4440
    },
    {
      "epoch": 0.4523859232817839,
      "grad_norm": 0.11473740637302399,
      "learning_rate": 0.000951025459688826,
      "loss": 2.2191076278686523,
      "step": 4442
    },
    {
      "epoch": 0.4525896089743916,
      "grad_norm": 0.1386810541152954,
      "learning_rate": 0.0009506718528995757,
      "loss": 2.2249457836151123,
      "step": 4444
    },
    {
      "epoch": 0.45279329466699936,
      "grad_norm": 0.12152288854122162,
      "learning_rate": 0.0009503182461103254,
      "loss": 2.0329904556274414,
      "step": 4446
    },
    {
      "epoch": 0.4529969803596071,
      "grad_norm": 0.2049301564693451,
      "learning_rate": 0.000949964639321075,
      "loss": 2.2301712036132812,
      "step": 4448
    },
    {
      "epoch": 0.4532006660522148,
      "grad_norm": 0.11563771963119507,
      "learning_rate": 0.0009496110325318247,
      "loss": 2.157712459564209,
      "step": 4450
    },
    {
      "epoch": 0.4534043517448226,
      "grad_norm": 0.13029226660728455,
      "learning_rate": 0.0009492574257425742,
      "loss": 2.167325019836426,
      "step": 4452
    },
    {
      "epoch": 0.4536080374374303,
      "grad_norm": 0.16012638807296753,
      "learning_rate": 0.0009489038189533238,
      "loss": 2.1951656341552734,
      "step": 4454
    },
    {
      "epoch": 0.45381172313003804,
      "grad_norm": 0.14239563047885895,
      "learning_rate": 0.0009485502121640736,
      "loss": 2.196239948272705,
      "step": 4456
    },
    {
      "epoch": 0.4540154088226458,
      "grad_norm": 0.1368158608675003,
      "learning_rate": 0.0009481966053748232,
      "loss": 2.3310964107513428,
      "step": 4458
    },
    {
      "epoch": 0.4542190945152535,
      "grad_norm": 0.11941059678792953,
      "learning_rate": 0.0009478429985855729,
      "loss": 2.2303993701934814,
      "step": 4460
    },
    {
      "epoch": 0.45442278020786125,
      "grad_norm": 0.11550739407539368,
      "learning_rate": 0.0009474893917963225,
      "loss": 2.232518196105957,
      "step": 4462
    },
    {
      "epoch": 0.454626465900469,
      "grad_norm": 0.1612805873155594,
      "learning_rate": 0.0009471357850070722,
      "loss": 2.2875092029571533,
      "step": 4464
    },
    {
      "epoch": 0.4548301515930767,
      "grad_norm": 0.139884352684021,
      "learning_rate": 0.0009467821782178218,
      "loss": 2.242872714996338,
      "step": 4466
    },
    {
      "epoch": 0.4550338372856845,
      "grad_norm": 0.12729321420192719,
      "learning_rate": 0.0009464285714285714,
      "loss": 2.256532907485962,
      "step": 4468
    },
    {
      "epoch": 0.4552375229782922,
      "grad_norm": 0.1498723030090332,
      "learning_rate": 0.0009460749646393211,
      "loss": 2.1910200119018555,
      "step": 4470
    },
    {
      "epoch": 0.45544120867089993,
      "grad_norm": 0.11553414165973663,
      "learning_rate": 0.0009457213578500707,
      "loss": 2.1738436222076416,
      "step": 4472
    },
    {
      "epoch": 0.4556448943635077,
      "grad_norm": 0.11114083975553513,
      "learning_rate": 0.0009453677510608204,
      "loss": 2.148408889770508,
      "step": 4474
    },
    {
      "epoch": 0.4558485800561154,
      "grad_norm": 0.11569429188966751,
      "learning_rate": 0.00094501414427157,
      "loss": 2.125425100326538,
      "step": 4476
    },
    {
      "epoch": 0.45605226574872315,
      "grad_norm": 0.13036134839057922,
      "learning_rate": 0.0009446605374823198,
      "loss": 2.1818900108337402,
      "step": 4478
    },
    {
      "epoch": 0.4562559514413309,
      "grad_norm": 0.16318650543689728,
      "learning_rate": 0.0009443069306930693,
      "loss": 2.113551378250122,
      "step": 4480
    },
    {
      "epoch": 0.4564596371339386,
      "grad_norm": 0.1188010573387146,
      "learning_rate": 0.0009439533239038189,
      "loss": 2.118746757507324,
      "step": 4482
    },
    {
      "epoch": 0.45666332282654637,
      "grad_norm": 0.1337585300207138,
      "learning_rate": 0.0009435997171145686,
      "loss": 2.317044258117676,
      "step": 4484
    },
    {
      "epoch": 0.45686700851915407,
      "grad_norm": 0.10741809755563736,
      "learning_rate": 0.0009432461103253182,
      "loss": 2.2648725509643555,
      "step": 4486
    },
    {
      "epoch": 0.45707069421176183,
      "grad_norm": 0.1303325891494751,
      "learning_rate": 0.000942892503536068,
      "loss": 2.079585075378418,
      "step": 4488
    },
    {
      "epoch": 0.4572743799043696,
      "grad_norm": 0.13622227311134338,
      "learning_rate": 0.0009425388967468176,
      "loss": 2.187987804412842,
      "step": 4490
    },
    {
      "epoch": 0.4574780655969773,
      "grad_norm": 0.13003699481487274,
      "learning_rate": 0.0009421852899575673,
      "loss": 2.205852508544922,
      "step": 4492
    },
    {
      "epoch": 0.45768175128958505,
      "grad_norm": 0.15368115901947021,
      "learning_rate": 0.0009418316831683168,
      "loss": 2.2658562660217285,
      "step": 4494
    },
    {
      "epoch": 0.45788543698219275,
      "grad_norm": 0.1095602959394455,
      "learning_rate": 0.0009414780763790664,
      "loss": 2.312866687774658,
      "step": 4496
    },
    {
      "epoch": 0.4580891226748005,
      "grad_norm": 0.10826533287763596,
      "learning_rate": 0.0009411244695898161,
      "loss": 2.1380162239074707,
      "step": 4498
    },
    {
      "epoch": 0.45829280836740827,
      "grad_norm": 0.132846400141716,
      "learning_rate": 0.0009407708628005658,
      "loss": 2.0847649574279785,
      "step": 4500
    },
    {
      "epoch": 0.45849649406001597,
      "grad_norm": 0.149509996175766,
      "learning_rate": 0.0009404172560113155,
      "loss": 2.360865592956543,
      "step": 4502
    },
    {
      "epoch": 0.4587001797526237,
      "grad_norm": 0.109349824488163,
      "learning_rate": 0.0009400636492220651,
      "loss": 2.1312148571014404,
      "step": 4504
    },
    {
      "epoch": 0.4589038654452315,
      "grad_norm": 0.1262819916009903,
      "learning_rate": 0.0009397100424328148,
      "loss": 2.2583096027374268,
      "step": 4506
    },
    {
      "epoch": 0.4591075511378392,
      "grad_norm": 0.13042666018009186,
      "learning_rate": 0.0009393564356435643,
      "loss": 2.2028722763061523,
      "step": 4508
    },
    {
      "epoch": 0.45931123683044695,
      "grad_norm": 0.12320356070995331,
      "learning_rate": 0.0009390028288543139,
      "loss": 2.2139830589294434,
      "step": 4510
    },
    {
      "epoch": 0.45951492252305465,
      "grad_norm": 0.11396641284227371,
      "learning_rate": 0.0009386492220650637,
      "loss": 2.1654281616210938,
      "step": 4512
    },
    {
      "epoch": 0.4597186082156624,
      "grad_norm": 0.1073664203286171,
      "learning_rate": 0.0009382956152758133,
      "loss": 2.2789506912231445,
      "step": 4514
    },
    {
      "epoch": 0.45992229390827016,
      "grad_norm": 0.12553894519805908,
      "learning_rate": 0.000937942008486563,
      "loss": 2.2695937156677246,
      "step": 4516
    },
    {
      "epoch": 0.46012597960087787,
      "grad_norm": 0.12091857194900513,
      "learning_rate": 0.0009375884016973126,
      "loss": 2.2762794494628906,
      "step": 4518
    },
    {
      "epoch": 0.4603296652934856,
      "grad_norm": 0.10877951979637146,
      "learning_rate": 0.0009372347949080623,
      "loss": 2.1890954971313477,
      "step": 4520
    },
    {
      "epoch": 0.4605333509860934,
      "grad_norm": 0.1450071632862091,
      "learning_rate": 0.000936881188118812,
      "loss": 2.3168787956237793,
      "step": 4522
    },
    {
      "epoch": 0.4607370366787011,
      "grad_norm": 0.11485420912504196,
      "learning_rate": 0.0009365275813295615,
      "loss": 2.184518337249756,
      "step": 4524
    },
    {
      "epoch": 0.46094072237130884,
      "grad_norm": 0.1416822373867035,
      "learning_rate": 0.0009361739745403112,
      "loss": 2.2299935817718506,
      "step": 4526
    },
    {
      "epoch": 0.46114440806391654,
      "grad_norm": 0.1386720985174179,
      "learning_rate": 0.0009358203677510608,
      "loss": 2.1448328495025635,
      "step": 4528
    },
    {
      "epoch": 0.4613480937565243,
      "grad_norm": 0.11416134238243103,
      "learning_rate": 0.0009354667609618105,
      "loss": 2.1203298568725586,
      "step": 4530
    },
    {
      "epoch": 0.46155177944913206,
      "grad_norm": 0.13752859830856323,
      "learning_rate": 0.0009351131541725601,
      "loss": 2.148169755935669,
      "step": 4532
    },
    {
      "epoch": 0.46175546514173976,
      "grad_norm": 0.16493940353393555,
      "learning_rate": 0.0009347595473833099,
      "loss": 2.2548978328704834,
      "step": 4534
    },
    {
      "epoch": 0.4619591508343475,
      "grad_norm": 0.14364545047283173,
      "learning_rate": 0.0009344059405940595,
      "loss": 2.228271484375,
      "step": 4536
    },
    {
      "epoch": 0.4621628365269553,
      "grad_norm": 0.11524691432714462,
      "learning_rate": 0.000934052333804809,
      "loss": 2.2563278675079346,
      "step": 4538
    },
    {
      "epoch": 0.462366522219563,
      "grad_norm": 0.1620008945465088,
      "learning_rate": 0.0009336987270155587,
      "loss": 2.2421956062316895,
      "step": 4540
    },
    {
      "epoch": 0.46257020791217074,
      "grad_norm": 0.12945519387722015,
      "learning_rate": 0.0009333451202263083,
      "loss": 2.222079277038574,
      "step": 4542
    },
    {
      "epoch": 0.46277389360477844,
      "grad_norm": 0.13271155953407288,
      "learning_rate": 0.000932991513437058,
      "loss": 2.117818593978882,
      "step": 4544
    },
    {
      "epoch": 0.4629775792973862,
      "grad_norm": 0.1185743510723114,
      "learning_rate": 0.0009326379066478077,
      "loss": 2.2372612953186035,
      "step": 4546
    },
    {
      "epoch": 0.46318126498999396,
      "grad_norm": 0.13853168487548828,
      "learning_rate": 0.0009322842998585574,
      "loss": 2.168034076690674,
      "step": 4548
    },
    {
      "epoch": 0.46338495068260166,
      "grad_norm": 0.12362273782491684,
      "learning_rate": 0.000931930693069307,
      "loss": 2.260960102081299,
      "step": 4550
    },
    {
      "epoch": 0.4635886363752094,
      "grad_norm": 0.13363778591156006,
      "learning_rate": 0.0009315770862800565,
      "loss": 2.0831398963928223,
      "step": 4552
    },
    {
      "epoch": 0.4637923220678172,
      "grad_norm": 0.11144138127565384,
      "learning_rate": 0.0009312234794908062,
      "loss": 2.1213834285736084,
      "step": 4554
    },
    {
      "epoch": 0.4639960077604249,
      "grad_norm": 0.1225903332233429,
      "learning_rate": 0.0009308698727015558,
      "loss": 2.2389798164367676,
      "step": 4556
    },
    {
      "epoch": 0.46419969345303264,
      "grad_norm": 0.12594036757946014,
      "learning_rate": 0.0009305162659123056,
      "loss": 2.2691171169281006,
      "step": 4558
    },
    {
      "epoch": 0.46440337914564034,
      "grad_norm": 0.14439961314201355,
      "learning_rate": 0.0009301626591230552,
      "loss": 2.322134256362915,
      "step": 4560
    },
    {
      "epoch": 0.4646070648382481,
      "grad_norm": 0.12116039544343948,
      "learning_rate": 0.0009298090523338049,
      "loss": 2.2643065452575684,
      "step": 4562
    },
    {
      "epoch": 0.46481075053085585,
      "grad_norm": 0.10569262504577637,
      "learning_rate": 0.0009294554455445545,
      "loss": 2.1556806564331055,
      "step": 4564
    },
    {
      "epoch": 0.46501443622346356,
      "grad_norm": 0.13451336324214935,
      "learning_rate": 0.000929101838755304,
      "loss": 2.277400493621826,
      "step": 4566
    },
    {
      "epoch": 0.4652181219160713,
      "grad_norm": 0.1093977838754654,
      "learning_rate": 0.0009287482319660538,
      "loss": 2.2632699012756348,
      "step": 4568
    },
    {
      "epoch": 0.4654218076086791,
      "grad_norm": 0.1303178071975708,
      "learning_rate": 0.0009283946251768034,
      "loss": 2.1009511947631836,
      "step": 4570
    },
    {
      "epoch": 0.4656254933012868,
      "grad_norm": 0.12130264937877655,
      "learning_rate": 0.0009280410183875531,
      "loss": 2.2588562965393066,
      "step": 4572
    },
    {
      "epoch": 0.46582917899389453,
      "grad_norm": 0.138875812292099,
      "learning_rate": 0.0009276874115983027,
      "loss": 2.2083396911621094,
      "step": 4574
    },
    {
      "epoch": 0.46603286468650224,
      "grad_norm": 0.11544418334960938,
      "learning_rate": 0.0009273338048090524,
      "loss": 2.1770074367523193,
      "step": 4576
    },
    {
      "epoch": 0.46623655037911,
      "grad_norm": 0.11872239410877228,
      "learning_rate": 0.000926980198019802,
      "loss": 2.114316940307617,
      "step": 4578
    },
    {
      "epoch": 0.46644023607171775,
      "grad_norm": 0.10941053926944733,
      "learning_rate": 0.0009266265912305516,
      "loss": 2.191135883331299,
      "step": 4580
    },
    {
      "epoch": 0.46664392176432545,
      "grad_norm": 0.18240197002887726,
      "learning_rate": 0.0009262729844413013,
      "loss": 2.240601062774658,
      "step": 4582
    },
    {
      "epoch": 0.4668476074569332,
      "grad_norm": 0.16108688712120056,
      "learning_rate": 0.0009259193776520509,
      "loss": 2.089489698410034,
      "step": 4584
    },
    {
      "epoch": 0.46705129314954097,
      "grad_norm": 0.12510554492473602,
      "learning_rate": 0.0009255657708628006,
      "loss": 2.145279884338379,
      "step": 4586
    },
    {
      "epoch": 0.46725497884214867,
      "grad_norm": 0.17591729760169983,
      "learning_rate": 0.0009252121640735502,
      "loss": 2.2700133323669434,
      "step": 4588
    },
    {
      "epoch": 0.46745866453475643,
      "grad_norm": 0.13197676837444305,
      "learning_rate": 0.0009248585572843,
      "loss": 2.037383794784546,
      "step": 4590
    },
    {
      "epoch": 0.46766235022736413,
      "grad_norm": 0.15310487151145935,
      "learning_rate": 0.0009245049504950496,
      "loss": 2.2234151363372803,
      "step": 4592
    },
    {
      "epoch": 0.4678660359199719,
      "grad_norm": 0.21063393354415894,
      "learning_rate": 0.0009241513437057991,
      "loss": 2.2171249389648438,
      "step": 4594
    },
    {
      "epoch": 0.46806972161257965,
      "grad_norm": 0.1246064305305481,
      "learning_rate": 0.0009237977369165488,
      "loss": 2.133751153945923,
      "step": 4596
    },
    {
      "epoch": 0.46827340730518735,
      "grad_norm": 0.1334415078163147,
      "learning_rate": 0.0009234441301272984,
      "loss": 2.283499240875244,
      "step": 4598
    },
    {
      "epoch": 0.4684770929977951,
      "grad_norm": 0.19017153978347778,
      "learning_rate": 0.0009230905233380481,
      "loss": 2.266451358795166,
      "step": 4600
    },
    {
      "epoch": 0.46868077869040287,
      "grad_norm": 0.17900705337524414,
      "learning_rate": 0.0009227369165487978,
      "loss": 2.2166764736175537,
      "step": 4602
    },
    {
      "epoch": 0.46888446438301057,
      "grad_norm": 0.11171939969062805,
      "learning_rate": 0.0009223833097595475,
      "loss": 2.082451343536377,
      "step": 4604
    },
    {
      "epoch": 0.4690881500756183,
      "grad_norm": 0.1073419600725174,
      "learning_rate": 0.0009220297029702971,
      "loss": 2.0717008113861084,
      "step": 4606
    },
    {
      "epoch": 0.46929183576822603,
      "grad_norm": 0.12128672748804092,
      "learning_rate": 0.0009216760961810466,
      "loss": 2.160865306854248,
      "step": 4608
    },
    {
      "epoch": 0.4694955214608338,
      "grad_norm": 0.14447112381458282,
      "learning_rate": 0.0009213224893917963,
      "loss": 2.195280075073242,
      "step": 4610
    },
    {
      "epoch": 0.46969920715344154,
      "grad_norm": 0.147905632853508,
      "learning_rate": 0.0009209688826025459,
      "loss": 2.271442413330078,
      "step": 4612
    },
    {
      "epoch": 0.46990289284604925,
      "grad_norm": 0.11903314292430878,
      "learning_rate": 0.0009206152758132957,
      "loss": 2.2420997619628906,
      "step": 4614
    },
    {
      "epoch": 0.470106578538657,
      "grad_norm": 0.12563195824623108,
      "learning_rate": 0.0009202616690240453,
      "loss": 2.2566580772399902,
      "step": 4616
    },
    {
      "epoch": 0.47031026423126476,
      "grad_norm": 0.11203783005475998,
      "learning_rate": 0.000919908062234795,
      "loss": 2.1690585613250732,
      "step": 4618
    },
    {
      "epoch": 0.47051394992387247,
      "grad_norm": 0.13423675298690796,
      "learning_rate": 0.0009195544554455446,
      "loss": 2.0896193981170654,
      "step": 4620
    },
    {
      "epoch": 0.4707176356164802,
      "grad_norm": 0.16609716415405273,
      "learning_rate": 0.0009192008486562942,
      "loss": 2.2711873054504395,
      "step": 4622
    },
    {
      "epoch": 0.4709213213090879,
      "grad_norm": 0.11692589521408081,
      "learning_rate": 0.0009188472418670438,
      "loss": 2.147109031677246,
      "step": 4624
    },
    {
      "epoch": 0.4711250070016957,
      "grad_norm": 0.12507468461990356,
      "learning_rate": 0.0009184936350777935,
      "loss": 2.15876841545105,
      "step": 4626
    },
    {
      "epoch": 0.47132869269430344,
      "grad_norm": 0.14635099470615387,
      "learning_rate": 0.0009181400282885432,
      "loss": 2.2498507499694824,
      "step": 4628
    },
    {
      "epoch": 0.47153237838691114,
      "grad_norm": 0.1396169513463974,
      "learning_rate": 0.0009177864214992928,
      "loss": 2.241702079772949,
      "step": 4630
    },
    {
      "epoch": 0.4717360640795189,
      "grad_norm": 0.1083715558052063,
      "learning_rate": 0.0009174328147100425,
      "loss": 2.183156967163086,
      "step": 4632
    },
    {
      "epoch": 0.4719397497721266,
      "grad_norm": 0.10736341029405594,
      "learning_rate": 0.0009170792079207921,
      "loss": 2.1250216960906982,
      "step": 4634
    },
    {
      "epoch": 0.47214343546473436,
      "grad_norm": 0.11951309442520142,
      "learning_rate": 0.0009167256011315417,
      "loss": 2.271010398864746,
      "step": 4636
    },
    {
      "epoch": 0.4723471211573421,
      "grad_norm": 0.12312625348567963,
      "learning_rate": 0.0009163719943422914,
      "loss": 2.041473388671875,
      "step": 4638
    },
    {
      "epoch": 0.4725508068499498,
      "grad_norm": 0.12153839319944382,
      "learning_rate": 0.000916018387553041,
      "loss": 2.297431230545044,
      "step": 4640
    },
    {
      "epoch": 0.4727544925425576,
      "grad_norm": 0.1217447966337204,
      "learning_rate": 0.0009156647807637907,
      "loss": 2.1084144115448,
      "step": 4642
    },
    {
      "epoch": 0.47295817823516534,
      "grad_norm": 0.11667545139789581,
      "learning_rate": 0.0009153111739745403,
      "loss": 2.2127456665039062,
      "step": 4644
    },
    {
      "epoch": 0.47316186392777304,
      "grad_norm": 0.12402203679084778,
      "learning_rate": 0.00091495756718529,
      "loss": 2.1745026111602783,
      "step": 4646
    },
    {
      "epoch": 0.4733655496203808,
      "grad_norm": 0.12920121848583221,
      "learning_rate": 0.0009146039603960397,
      "loss": 2.2726244926452637,
      "step": 4648
    },
    {
      "epoch": 0.4735692353129885,
      "grad_norm": 0.14174428582191467,
      "learning_rate": 0.0009142503536067893,
      "loss": 2.243546962738037,
      "step": 4650
    },
    {
      "epoch": 0.47377292100559626,
      "grad_norm": 0.12071418762207031,
      "learning_rate": 0.0009138967468175389,
      "loss": 2.2711455821990967,
      "step": 4652
    },
    {
      "epoch": 0.473976606698204,
      "grad_norm": 0.12173871695995331,
      "learning_rate": 0.0009135431400282885,
      "loss": 2.041210651397705,
      "step": 4654
    },
    {
      "epoch": 0.4741802923908117,
      "grad_norm": 0.1421707570552826,
      "learning_rate": 0.0009131895332390382,
      "loss": 2.224005699157715,
      "step": 4656
    },
    {
      "epoch": 0.4743839780834195,
      "grad_norm": 0.11143438518047333,
      "learning_rate": 0.0009128359264497878,
      "loss": 2.262967824935913,
      "step": 4658
    },
    {
      "epoch": 0.47458766377602724,
      "grad_norm": 0.1207672730088234,
      "learning_rate": 0.0009124823196605376,
      "loss": 2.249025821685791,
      "step": 4660
    },
    {
      "epoch": 0.47479134946863494,
      "grad_norm": 0.11522098630666733,
      "learning_rate": 0.0009121287128712872,
      "loss": 2.2892255783081055,
      "step": 4662
    },
    {
      "epoch": 0.4749950351612427,
      "grad_norm": 0.10159576684236526,
      "learning_rate": 0.0009117751060820368,
      "loss": 2.223292112350464,
      "step": 4664
    },
    {
      "epoch": 0.4751987208538504,
      "grad_norm": 0.15874212980270386,
      "learning_rate": 0.0009114214992927864,
      "loss": 2.3133318424224854,
      "step": 4666
    },
    {
      "epoch": 0.47540240654645816,
      "grad_norm": 0.11143343895673752,
      "learning_rate": 0.000911067892503536,
      "loss": 2.2569117546081543,
      "step": 4668
    },
    {
      "epoch": 0.4756060922390659,
      "grad_norm": 0.1572769582271576,
      "learning_rate": 0.0009107142857142857,
      "loss": 2.356480121612549,
      "step": 4670
    },
    {
      "epoch": 0.4758097779316736,
      "grad_norm": 0.11646872758865356,
      "learning_rate": 0.0009103606789250354,
      "loss": 2.153618574142456,
      "step": 4672
    },
    {
      "epoch": 0.4760134636242814,
      "grad_norm": 0.1093180775642395,
      "learning_rate": 0.0009100070721357851,
      "loss": 2.2985758781433105,
      "step": 4674
    },
    {
      "epoch": 0.47621714931688913,
      "grad_norm": 0.12663964927196503,
      "learning_rate": 0.0009096534653465347,
      "loss": 2.120077133178711,
      "step": 4676
    },
    {
      "epoch": 0.47642083500949683,
      "grad_norm": 0.10697914659976959,
      "learning_rate": 0.0009092998585572843,
      "loss": 2.1009583473205566,
      "step": 4678
    },
    {
      "epoch": 0.4766245207021046,
      "grad_norm": 0.1185147687792778,
      "learning_rate": 0.0009089462517680339,
      "loss": 2.1521239280700684,
      "step": 4680
    },
    {
      "epoch": 0.4768282063947123,
      "grad_norm": 0.12941861152648926,
      "learning_rate": 0.0009085926449787835,
      "loss": 2.1537222862243652,
      "step": 4682
    },
    {
      "epoch": 0.47703189208732005,
      "grad_norm": 0.12805911898612976,
      "learning_rate": 0.0009082390381895333,
      "loss": 2.3298213481903076,
      "step": 4684
    },
    {
      "epoch": 0.4772355777799278,
      "grad_norm": 0.12878035008907318,
      "learning_rate": 0.0009078854314002829,
      "loss": 2.2663540840148926,
      "step": 4686
    },
    {
      "epoch": 0.4774392634725355,
      "grad_norm": 0.14399686455726624,
      "learning_rate": 0.0009075318246110326,
      "loss": 2.165882110595703,
      "step": 4688
    },
    {
      "epoch": 0.47764294916514327,
      "grad_norm": 0.18520574271678925,
      "learning_rate": 0.0009071782178217822,
      "loss": 2.157289981842041,
      "step": 4690
    },
    {
      "epoch": 0.47784663485775103,
      "grad_norm": 0.1315893679857254,
      "learning_rate": 0.0009068246110325318,
      "loss": 2.270570755004883,
      "step": 4692
    },
    {
      "epoch": 0.47805032055035873,
      "grad_norm": 0.1279367357492447,
      "learning_rate": 0.0009064710042432815,
      "loss": 2.1966519355773926,
      "step": 4694
    },
    {
      "epoch": 0.4782540062429665,
      "grad_norm": 0.14573341608047485,
      "learning_rate": 0.0009061173974540311,
      "loss": 2.252955913543701,
      "step": 4696
    },
    {
      "epoch": 0.4784576919355742,
      "grad_norm": 0.11123565584421158,
      "learning_rate": 0.0009057637906647808,
      "loss": 2.3083770275115967,
      "step": 4698
    },
    {
      "epoch": 0.47866137762818195,
      "grad_norm": 0.12670496106147766,
      "learning_rate": 0.0009054101838755304,
      "loss": 2.1851253509521484,
      "step": 4700
    },
    {
      "epoch": 0.4788650633207897,
      "grad_norm": 0.13188141584396362,
      "learning_rate": 0.0009050565770862801,
      "loss": 2.2099075317382812,
      "step": 4702
    },
    {
      "epoch": 0.4790687490133974,
      "grad_norm": 0.09896662831306458,
      "learning_rate": 0.0009047029702970297,
      "loss": 2.035644054412842,
      "step": 4704
    },
    {
      "epoch": 0.47927243470600517,
      "grad_norm": 0.17076706886291504,
      "learning_rate": 0.0009043493635077794,
      "loss": 2.3483753204345703,
      "step": 4706
    },
    {
      "epoch": 0.4794761203986129,
      "grad_norm": 0.11471794545650482,
      "learning_rate": 0.000903995756718529,
      "loss": 2.1785130500793457,
      "step": 4708
    },
    {
      "epoch": 0.47967980609122063,
      "grad_norm": 0.1200699731707573,
      "learning_rate": 0.0009036421499292786,
      "loss": 2.1699445247650146,
      "step": 4710
    },
    {
      "epoch": 0.4798834917838284,
      "grad_norm": 0.11952436715364456,
      "learning_rate": 0.0009032885431400283,
      "loss": 2.0885415077209473,
      "step": 4712
    },
    {
      "epoch": 0.4800871774764361,
      "grad_norm": 0.18040212988853455,
      "learning_rate": 0.0009029349363507779,
      "loss": 2.222677230834961,
      "step": 4714
    },
    {
      "epoch": 0.48029086316904385,
      "grad_norm": 0.15156938135623932,
      "learning_rate": 0.0009025813295615277,
      "loss": 2.183131217956543,
      "step": 4716
    },
    {
      "epoch": 0.4804945488616516,
      "grad_norm": 0.13971301913261414,
      "learning_rate": 0.0009022277227722773,
      "loss": 2.1118955612182617,
      "step": 4718
    },
    {
      "epoch": 0.4806982345542593,
      "grad_norm": 0.16743847727775574,
      "learning_rate": 0.0009018741159830269,
      "loss": 2.149616241455078,
      "step": 4720
    },
    {
      "epoch": 0.48090192024686707,
      "grad_norm": 0.12019199132919312,
      "learning_rate": 0.0009015205091937766,
      "loss": 2.1799120903015137,
      "step": 4722
    },
    {
      "epoch": 0.4811056059394748,
      "grad_norm": 0.09825186431407928,
      "learning_rate": 0.0009011669024045261,
      "loss": 1.9995477199554443,
      "step": 4724
    },
    {
      "epoch": 0.4813092916320825,
      "grad_norm": 0.15617519617080688,
      "learning_rate": 0.0009008132956152758,
      "loss": 2.1660380363464355,
      "step": 4726
    },
    {
      "epoch": 0.4815129773246903,
      "grad_norm": 0.1688273698091507,
      "learning_rate": 0.0009004596888260255,
      "loss": 2.29870343208313,
      "step": 4728
    },
    {
      "epoch": 0.481716663017298,
      "grad_norm": 0.1358804553747177,
      "learning_rate": 0.0009001060820367752,
      "loss": 2.2847900390625,
      "step": 4730
    },
    {
      "epoch": 0.48192034870990574,
      "grad_norm": 0.1450638473033905,
      "learning_rate": 0.0008997524752475248,
      "loss": 2.121986150741577,
      "step": 4732
    },
    {
      "epoch": 0.4821240344025135,
      "grad_norm": 0.13609428703784943,
      "learning_rate": 0.0008993988684582744,
      "loss": 2.3344759941101074,
      "step": 4734
    },
    {
      "epoch": 0.4823277200951212,
      "grad_norm": 0.11523746699094772,
      "learning_rate": 0.0008990452616690241,
      "loss": 2.2089929580688477,
      "step": 4736
    },
    {
      "epoch": 0.48253140578772896,
      "grad_norm": 0.12423786520957947,
      "learning_rate": 0.0008986916548797736,
      "loss": 2.365591049194336,
      "step": 4738
    },
    {
      "epoch": 0.4827350914803367,
      "grad_norm": 0.13455383479595184,
      "learning_rate": 0.0008983380480905234,
      "loss": 2.1496877670288086,
      "step": 4740
    },
    {
      "epoch": 0.4829387771729444,
      "grad_norm": 0.1088300421833992,
      "learning_rate": 0.000897984441301273,
      "loss": 2.2358715534210205,
      "step": 4742
    },
    {
      "epoch": 0.4831424628655522,
      "grad_norm": 0.10394295305013657,
      "learning_rate": 0.0008976308345120227,
      "loss": 2.2629523277282715,
      "step": 4744
    },
    {
      "epoch": 0.4833461485581599,
      "grad_norm": 0.11890991032123566,
      "learning_rate": 0.0008972772277227723,
      "loss": 2.300826072692871,
      "step": 4746
    },
    {
      "epoch": 0.48354983425076764,
      "grad_norm": 0.10701869428157806,
      "learning_rate": 0.0008969236209335219,
      "loss": 2.0974245071411133,
      "step": 4748
    },
    {
      "epoch": 0.4837535199433754,
      "grad_norm": 0.12070364505052567,
      "learning_rate": 0.0008965700141442717,
      "loss": 2.255922555923462,
      "step": 4750
    },
    {
      "epoch": 0.4839572056359831,
      "grad_norm": 0.1318737268447876,
      "learning_rate": 0.0008962164073550212,
      "loss": 2.2108776569366455,
      "step": 4752
    },
    {
      "epoch": 0.48416089132859086,
      "grad_norm": 0.12049932777881622,
      "learning_rate": 0.0008958628005657709,
      "loss": 2.270115852355957,
      "step": 4754
    },
    {
      "epoch": 0.48436457702119856,
      "grad_norm": 0.11139902472496033,
      "learning_rate": 0.0008955091937765205,
      "loss": 2.2738208770751953,
      "step": 4756
    },
    {
      "epoch": 0.4845682627138063,
      "grad_norm": 0.12910480797290802,
      "learning_rate": 0.0008951555869872702,
      "loss": 2.0943491458892822,
      "step": 4758
    },
    {
      "epoch": 0.4847719484064141,
      "grad_norm": 0.12974999845027924,
      "learning_rate": 0.0008948019801980198,
      "loss": 2.145627021789551,
      "step": 4760
    },
    {
      "epoch": 0.4849756340990218,
      "grad_norm": 0.11960352212190628,
      "learning_rate": 0.0008944483734087695,
      "loss": 2.1442575454711914,
      "step": 4762
    },
    {
      "epoch": 0.48517931979162954,
      "grad_norm": 0.15973255038261414,
      "learning_rate": 0.0008940947666195192,
      "loss": 2.1510133743286133,
      "step": 4764
    },
    {
      "epoch": 0.4853830054842373,
      "grad_norm": 0.12670853734016418,
      "learning_rate": 0.0008937411598302687,
      "loss": 2.19404935836792,
      "step": 4766
    },
    {
      "epoch": 0.485586691176845,
      "grad_norm": 0.11740083247423172,
      "learning_rate": 0.0008933875530410184,
      "loss": 2.1237831115722656,
      "step": 4768
    },
    {
      "epoch": 0.48579037686945276,
      "grad_norm": 0.16061419248580933,
      "learning_rate": 0.000893033946251768,
      "loss": 2.2734453678131104,
      "step": 4770
    },
    {
      "epoch": 0.48599406256206046,
      "grad_norm": 0.1250544935464859,
      "learning_rate": 0.0008926803394625177,
      "loss": 2.1865363121032715,
      "step": 4772
    },
    {
      "epoch": 0.4861977482546682,
      "grad_norm": 0.11283563822507858,
      "learning_rate": 0.0008923267326732674,
      "loss": 2.1140050888061523,
      "step": 4774
    },
    {
      "epoch": 0.486401433947276,
      "grad_norm": 0.11459698528051376,
      "learning_rate": 0.000891973125884017,
      "loss": 2.11739444732666,
      "step": 4776
    },
    {
      "epoch": 0.4866051196398837,
      "grad_norm": 0.12477733194828033,
      "learning_rate": 0.0008916195190947667,
      "loss": 2.2733802795410156,
      "step": 4778
    },
    {
      "epoch": 0.48680880533249143,
      "grad_norm": 0.12662135064601898,
      "learning_rate": 0.0008912659123055162,
      "loss": 2.222902774810791,
      "step": 4780
    },
    {
      "epoch": 0.4870124910250992,
      "grad_norm": 0.1317615807056427,
      "learning_rate": 0.0008909123055162659,
      "loss": 2.181884288787842,
      "step": 4782
    },
    {
      "epoch": 0.4872161767177069,
      "grad_norm": 0.14360646903514862,
      "learning_rate": 0.0008905586987270155,
      "loss": 2.2596731185913086,
      "step": 4784
    },
    {
      "epoch": 0.48741986241031465,
      "grad_norm": 0.10456658899784088,
      "learning_rate": 0.0008902050919377653,
      "loss": 2.2897191047668457,
      "step": 4786
    },
    {
      "epoch": 0.48762354810292236,
      "grad_norm": 0.12053926289081573,
      "learning_rate": 0.0008898514851485149,
      "loss": 2.251544237136841,
      "step": 4788
    },
    {
      "epoch": 0.4878272337955301,
      "grad_norm": 0.12110774964094162,
      "learning_rate": 0.0008894978783592645,
      "loss": 2.016080856323242,
      "step": 4790
    },
    {
      "epoch": 0.48803091948813787,
      "grad_norm": 0.15052752196788788,
      "learning_rate": 0.0008891442715700142,
      "loss": 2.1524555683135986,
      "step": 4792
    },
    {
      "epoch": 0.4882346051807456,
      "grad_norm": 0.12412973493337631,
      "learning_rate": 0.0008887906647807637,
      "loss": 1.9916410446166992,
      "step": 4794
    },
    {
      "epoch": 0.48843829087335333,
      "grad_norm": 0.17985013127326965,
      "learning_rate": 0.0008884370579915135,
      "loss": 2.2016749382019043,
      "step": 4796
    },
    {
      "epoch": 0.4886419765659611,
      "grad_norm": 0.10758385062217712,
      "learning_rate": 0.0008880834512022631,
      "loss": 2.0514492988586426,
      "step": 4798
    },
    {
      "epoch": 0.4888456622585688,
      "grad_norm": 0.13457533717155457,
      "learning_rate": 0.0008877298444130128,
      "loss": 2.2183165550231934,
      "step": 4800
    },
    {
      "epoch": 0.48904934795117655,
      "grad_norm": 0.13166062533855438,
      "learning_rate": 0.0008873762376237624,
      "loss": 2.349989175796509,
      "step": 4802
    },
    {
      "epoch": 0.48925303364378425,
      "grad_norm": 0.14633508026599884,
      "learning_rate": 0.000887022630834512,
      "loss": 2.3118934631347656,
      "step": 4804
    },
    {
      "epoch": 0.489456719336392,
      "grad_norm": 0.1214480921626091,
      "learning_rate": 0.0008866690240452617,
      "loss": 2.195310354232788,
      "step": 4806
    },
    {
      "epoch": 0.48966040502899977,
      "grad_norm": 0.10843229293823242,
      "learning_rate": 0.0008863154172560113,
      "loss": 2.0960402488708496,
      "step": 4808
    },
    {
      "epoch": 0.48986409072160747,
      "grad_norm": 0.10975231230258942,
      "learning_rate": 0.000885961810466761,
      "loss": 2.150974988937378,
      "step": 4810
    },
    {
      "epoch": 0.49006777641421523,
      "grad_norm": 0.12202282249927521,
      "learning_rate": 0.0008856082036775106,
      "loss": 2.2255401611328125,
      "step": 4812
    },
    {
      "epoch": 0.490271462106823,
      "grad_norm": 0.13704191148281097,
      "learning_rate": 0.0008852545968882603,
      "loss": 2.2655138969421387,
      "step": 4814
    },
    {
      "epoch": 0.4904751477994307,
      "grad_norm": 0.12641043961048126,
      "learning_rate": 0.0008849009900990099,
      "loss": 2.1809802055358887,
      "step": 4816
    },
    {
      "epoch": 0.49067883349203845,
      "grad_norm": 0.13467901945114136,
      "learning_rate": 0.0008845473833097595,
      "loss": 2.265690803527832,
      "step": 4818
    },
    {
      "epoch": 0.49088251918464615,
      "grad_norm": 0.14265763759613037,
      "learning_rate": 0.0008841937765205093,
      "loss": 2.1960744857788086,
      "step": 4820
    },
    {
      "epoch": 0.4910862048772539,
      "grad_norm": 0.132966086268425,
      "learning_rate": 0.0008838401697312589,
      "loss": 2.211818218231201,
      "step": 4822
    },
    {
      "epoch": 0.49128989056986166,
      "grad_norm": 0.11924374848604202,
      "learning_rate": 0.0008834865629420085,
      "loss": 2.141209125518799,
      "step": 4824
    },
    {
      "epoch": 0.49149357626246937,
      "grad_norm": 0.14768558740615845,
      "learning_rate": 0.0008831329561527581,
      "loss": 2.255418300628662,
      "step": 4826
    },
    {
      "epoch": 0.4916972619550771,
      "grad_norm": 0.11781346797943115,
      "learning_rate": 0.0008827793493635078,
      "loss": 2.16153621673584,
      "step": 4828
    },
    {
      "epoch": 0.4919009476476849,
      "grad_norm": 0.13667428493499756,
      "learning_rate": 0.0008824257425742575,
      "loss": 2.231654644012451,
      "step": 4830
    },
    {
      "epoch": 0.4921046333402926,
      "grad_norm": 0.18724015355110168,
      "learning_rate": 0.0008820721357850071,
      "loss": 2.1190671920776367,
      "step": 4832
    },
    {
      "epoch": 0.49230831903290034,
      "grad_norm": 0.10730881243944168,
      "learning_rate": 0.0008817185289957568,
      "loss": 2.1985535621643066,
      "step": 4834
    },
    {
      "epoch": 0.49251200472550805,
      "grad_norm": 0.11654356122016907,
      "learning_rate": 0.0008813649222065064,
      "loss": 2.1441972255706787,
      "step": 4836
    },
    {
      "epoch": 0.4927156904181158,
      "grad_norm": 0.17904196679592133,
      "learning_rate": 0.000881011315417256,
      "loss": 2.273231267929077,
      "step": 4838
    },
    {
      "epoch": 0.49291937611072356,
      "grad_norm": 0.13681036233901978,
      "learning_rate": 0.0008806577086280056,
      "loss": 2.1762032508850098,
      "step": 4840
    },
    {
      "epoch": 0.49312306180333126,
      "grad_norm": 0.1366841346025467,
      "learning_rate": 0.0008803041018387554,
      "loss": 2.1838529109954834,
      "step": 4842
    },
    {
      "epoch": 0.493326747495939,
      "grad_norm": 0.11755393445491791,
      "learning_rate": 0.000879950495049505,
      "loss": 2.2875030040740967,
      "step": 4844
    },
    {
      "epoch": 0.4935304331885468,
      "grad_norm": 0.21625502407550812,
      "learning_rate": 0.0008795968882602546,
      "loss": 2.234309196472168,
      "step": 4846
    },
    {
      "epoch": 0.4937341188811545,
      "grad_norm": 0.12441275268793106,
      "learning_rate": 0.0008792432814710043,
      "loss": 2.179758071899414,
      "step": 4848
    },
    {
      "epoch": 0.49393780457376224,
      "grad_norm": 0.18150700628757477,
      "learning_rate": 0.0008788896746817539,
      "loss": 2.3365015983581543,
      "step": 4850
    },
    {
      "epoch": 0.49414149026636994,
      "grad_norm": 0.12365284562110901,
      "learning_rate": 0.0008785360678925035,
      "loss": 2.1963324546813965,
      "step": 4852
    },
    {
      "epoch": 0.4943451759589777,
      "grad_norm": 0.11641213297843933,
      "learning_rate": 0.0008781824611032532,
      "loss": 2.0649592876434326,
      "step": 4854
    },
    {
      "epoch": 0.49454886165158546,
      "grad_norm": 0.16246701776981354,
      "learning_rate": 0.0008778288543140029,
      "loss": 2.3083534240722656,
      "step": 4856
    },
    {
      "epoch": 0.49475254734419316,
      "grad_norm": 0.11017903685569763,
      "learning_rate": 0.0008774752475247525,
      "loss": 2.2137868404388428,
      "step": 4858
    },
    {
      "epoch": 0.4949562330368009,
      "grad_norm": 0.16723176836967468,
      "learning_rate": 0.0008771216407355021,
      "loss": 2.2528328895568848,
      "step": 4860
    },
    {
      "epoch": 0.4951599187294087,
      "grad_norm": 0.15064918994903564,
      "learning_rate": 0.0008767680339462518,
      "loss": 2.2863686084747314,
      "step": 4862
    },
    {
      "epoch": 0.4953636044220164,
      "grad_norm": 0.14373983442783356,
      "learning_rate": 0.0008764144271570014,
      "loss": 2.2215733528137207,
      "step": 4864
    },
    {
      "epoch": 0.49556729011462414,
      "grad_norm": 0.12899497151374817,
      "learning_rate": 0.0008760608203677511,
      "loss": 2.208493709564209,
      "step": 4866
    },
    {
      "epoch": 0.49577097580723184,
      "grad_norm": 0.11388517916202545,
      "learning_rate": 0.0008757072135785007,
      "loss": 2.022732734680176,
      "step": 4868
    },
    {
      "epoch": 0.4959746614998396,
      "grad_norm": 0.12080218642950058,
      "learning_rate": 0.0008753536067892504,
      "loss": 2.283005475997925,
      "step": 4870
    },
    {
      "epoch": 0.49617834719244736,
      "grad_norm": 0.11749328672885895,
      "learning_rate": 0.000875,
      "loss": 2.219442367553711,
      "step": 4872
    },
    {
      "epoch": 0.49638203288505506,
      "grad_norm": 0.11574358493089676,
      "learning_rate": 0.0008746463932107496,
      "loss": 2.1388936042785645,
      "step": 4874
    },
    {
      "epoch": 0.4965857185776628,
      "grad_norm": 0.16069753468036652,
      "learning_rate": 0.0008742927864214994,
      "loss": 2.187541961669922,
      "step": 4876
    },
    {
      "epoch": 0.4967894042702705,
      "grad_norm": 0.1355918049812317,
      "learning_rate": 0.000873939179632249,
      "loss": 2.1153206825256348,
      "step": 4878
    },
    {
      "epoch": 0.4969930899628783,
      "grad_norm": 0.13793177902698517,
      "learning_rate": 0.0008735855728429986,
      "loss": 2.145829200744629,
      "step": 4880
    },
    {
      "epoch": 0.49719677565548603,
      "grad_norm": 0.19802600145339966,
      "learning_rate": 0.0008732319660537482,
      "loss": 2.2374377250671387,
      "step": 4882
    },
    {
      "epoch": 0.49740046134809374,
      "grad_norm": 0.13155874609947205,
      "learning_rate": 0.0008728783592644979,
      "loss": 2.2675070762634277,
      "step": 4884
    },
    {
      "epoch": 0.4976041470407015,
      "grad_norm": 0.13055989146232605,
      "learning_rate": 0.0008725247524752475,
      "loss": 2.155550956726074,
      "step": 4886
    },
    {
      "epoch": 0.49780783273330925,
      "grad_norm": 0.16608566045761108,
      "learning_rate": 0.0008721711456859972,
      "loss": 2.138080596923828,
      "step": 4888
    },
    {
      "epoch": 0.49801151842591695,
      "grad_norm": 0.13015252351760864,
      "learning_rate": 0.0008718175388967469,
      "loss": 2.043149948120117,
      "step": 4890
    },
    {
      "epoch": 0.4982152041185247,
      "grad_norm": 0.12644249200820923,
      "learning_rate": 0.0008714639321074965,
      "loss": 2.1886301040649414,
      "step": 4892
    },
    {
      "epoch": 0.4984188898111324,
      "grad_norm": 0.11616521328687668,
      "learning_rate": 0.0008711103253182461,
      "loss": 2.172900438308716,
      "step": 4894
    },
    {
      "epoch": 0.4986225755037402,
      "grad_norm": 0.131742462515831,
      "learning_rate": 0.0008707567185289957,
      "loss": 2.175194501876831,
      "step": 4896
    },
    {
      "epoch": 0.49882626119634793,
      "grad_norm": 0.15420173108577728,
      "learning_rate": 0.0008704031117397454,
      "loss": 2.171351432800293,
      "step": 4898
    },
    {
      "epoch": 0.49902994688895563,
      "grad_norm": 0.11715130507946014,
      "learning_rate": 0.0008700495049504951,
      "loss": 2.103121280670166,
      "step": 4900
    },
    {
      "epoch": 0.4992336325815634,
      "grad_norm": 0.1276465356349945,
      "learning_rate": 0.0008696958981612447,
      "loss": 2.1572413444519043,
      "step": 4902
    },
    {
      "epoch": 0.49943731827417115,
      "grad_norm": 0.14851492643356323,
      "learning_rate": 0.0008693422913719944,
      "loss": 2.2446117401123047,
      "step": 4904
    },
    {
      "epoch": 0.49964100396677885,
      "grad_norm": 0.16749806702136993,
      "learning_rate": 0.000868988684582744,
      "loss": 2.2180745601654053,
      "step": 4906
    },
    {
      "epoch": 0.4998446896593866,
      "grad_norm": 0.13275262713432312,
      "learning_rate": 0.0008686350777934936,
      "loss": 2.238365650177002,
      "step": 4908
    },
    {
      "epoch": 0.5000483753519943,
      "grad_norm": 0.12023303657770157,
      "learning_rate": 0.0008682814710042432,
      "loss": 2.0071849822998047,
      "step": 4910
    },
    {
      "epoch": 0.5002520610446021,
      "grad_norm": 0.10800334811210632,
      "learning_rate": 0.000867927864214993,
      "loss": 2.1183738708496094,
      "step": 4912
    },
    {
      "epoch": 0.5004557467372098,
      "grad_norm": 0.13783392310142517,
      "learning_rate": 0.0008675742574257426,
      "loss": 2.212221622467041,
      "step": 4914
    },
    {
      "epoch": 0.5006594324298176,
      "grad_norm": 0.11800715327262878,
      "learning_rate": 0.0008672206506364922,
      "loss": 2.0681071281433105,
      "step": 4916
    },
    {
      "epoch": 0.5008631181224252,
      "grad_norm": 0.12124183773994446,
      "learning_rate": 0.0008668670438472419,
      "loss": 2.3058724403381348,
      "step": 4918
    },
    {
      "epoch": 0.501066803815033,
      "grad_norm": 0.14865583181381226,
      "learning_rate": 0.0008665134370579915,
      "loss": 2.313575267791748,
      "step": 4920
    },
    {
      "epoch": 0.5012704895076407,
      "grad_norm": 0.12178020924329758,
      "learning_rate": 0.0008661598302687412,
      "loss": 2.2611684799194336,
      "step": 4922
    },
    {
      "epoch": 0.5014741752002485,
      "grad_norm": 0.11182554066181183,
      "learning_rate": 0.0008658062234794908,
      "loss": 2.2161929607391357,
      "step": 4924
    },
    {
      "epoch": 0.5016778608928563,
      "grad_norm": 0.12803375720977783,
      "learning_rate": 0.0008654526166902405,
      "loss": 2.2218289375305176,
      "step": 4926
    },
    {
      "epoch": 0.501881546585464,
      "grad_norm": 0.12626463174819946,
      "learning_rate": 0.0008650990099009901,
      "loss": 2.0726730823516846,
      "step": 4928
    },
    {
      "epoch": 0.5020852322780717,
      "grad_norm": 0.12647219002246857,
      "learning_rate": 0.0008647454031117397,
      "loss": 2.1301116943359375,
      "step": 4930
    },
    {
      "epoch": 0.5022889179706794,
      "grad_norm": 0.14045177400112152,
      "learning_rate": 0.0008643917963224894,
      "loss": 2.2914271354675293,
      "step": 4932
    },
    {
      "epoch": 0.5024926036632872,
      "grad_norm": 0.12012903392314911,
      "learning_rate": 0.0008640381895332391,
      "loss": 2.2886650562286377,
      "step": 4934
    },
    {
      "epoch": 0.5026962893558949,
      "grad_norm": 0.14497384428977966,
      "learning_rate": 0.0008636845827439888,
      "loss": 2.1270246505737305,
      "step": 4936
    },
    {
      "epoch": 0.5028999750485027,
      "grad_norm": 0.14201490581035614,
      "learning_rate": 0.0008633309759547383,
      "loss": 2.128908157348633,
      "step": 4938
    },
    {
      "epoch": 0.5031036607411103,
      "grad_norm": 0.13806842267513275,
      "learning_rate": 0.000862977369165488,
      "loss": 2.186222553253174,
      "step": 4940
    },
    {
      "epoch": 0.5033073464337181,
      "grad_norm": 0.19249582290649414,
      "learning_rate": 0.0008626237623762376,
      "loss": 2.1915698051452637,
      "step": 4942
    },
    {
      "epoch": 0.5035110321263259,
      "grad_norm": 0.11891204863786697,
      "learning_rate": 0.0008622701555869872,
      "loss": 2.1132044792175293,
      "step": 4944
    },
    {
      "epoch": 0.5037147178189336,
      "grad_norm": 0.13233911991119385,
      "learning_rate": 0.000861916548797737,
      "loss": 2.172250270843506,
      "step": 4946
    },
    {
      "epoch": 0.5039184035115414,
      "grad_norm": 0.17398637533187866,
      "learning_rate": 0.0008615629420084866,
      "loss": 2.2885708808898926,
      "step": 4948
    },
    {
      "epoch": 0.504122089204149,
      "grad_norm": 0.14322367310523987,
      "learning_rate": 0.0008612093352192363,
      "loss": 2.281339645385742,
      "step": 4950
    },
    {
      "epoch": 0.5043257748967568,
      "grad_norm": 0.12107861787080765,
      "learning_rate": 0.0008608557284299858,
      "loss": 2.1631810665130615,
      "step": 4952
    },
    {
      "epoch": 0.5045294605893645,
      "grad_norm": 0.11637546867132187,
      "learning_rate": 0.0008605021216407355,
      "loss": 2.121302604675293,
      "step": 4954
    },
    {
      "epoch": 0.5047331462819723,
      "grad_norm": 0.11281871050596237,
      "learning_rate": 0.0008601485148514852,
      "loss": 2.25927996635437,
      "step": 4956
    },
    {
      "epoch": 0.5049368319745801,
      "grad_norm": 0.14516855776309967,
      "learning_rate": 0.0008597949080622348,
      "loss": 2.048182964324951,
      "step": 4958
    },
    {
      "epoch": 0.5051405176671878,
      "grad_norm": 0.11057573556900024,
      "learning_rate": 0.0008594413012729845,
      "loss": 2.042949676513672,
      "step": 4960
    },
    {
      "epoch": 0.5053442033597955,
      "grad_norm": 0.11559516936540604,
      "learning_rate": 0.0008590876944837341,
      "loss": 2.138516902923584,
      "step": 4962
    },
    {
      "epoch": 0.5055478890524032,
      "grad_norm": 0.12866032123565674,
      "learning_rate": 0.0008587340876944838,
      "loss": 2.2007784843444824,
      "step": 4964
    },
    {
      "epoch": 0.505751574745011,
      "grad_norm": 0.11582763493061066,
      "learning_rate": 0.0008583804809052333,
      "loss": 2.175652027130127,
      "step": 4966
    },
    {
      "epoch": 0.5059552604376187,
      "grad_norm": 0.1259339600801468,
      "learning_rate": 0.0008580268741159831,
      "loss": 2.2329771518707275,
      "step": 4968
    },
    {
      "epoch": 0.5061589461302265,
      "grad_norm": 0.12424679845571518,
      "learning_rate": 0.0008576732673267327,
      "loss": 2.2211313247680664,
      "step": 4970
    },
    {
      "epoch": 0.5063626318228341,
      "grad_norm": 0.11080215871334076,
      "learning_rate": 0.0008573196605374823,
      "loss": 2.210732936859131,
      "step": 4972
    },
    {
      "epoch": 0.5065663175154419,
      "grad_norm": 0.10492751002311707,
      "learning_rate": 0.000856966053748232,
      "loss": 2.2811591625213623,
      "step": 4974
    },
    {
      "epoch": 0.5067700032080497,
      "grad_norm": 0.10982757061719894,
      "learning_rate": 0.0008566124469589816,
      "loss": 2.307586193084717,
      "step": 4976
    },
    {
      "epoch": 0.5069736889006574,
      "grad_norm": 0.12710316479206085,
      "learning_rate": 0.0008562588401697314,
      "loss": 2.119844913482666,
      "step": 4978
    },
    {
      "epoch": 0.5071773745932652,
      "grad_norm": 0.10244240611791611,
      "learning_rate": 0.0008559052333804809,
      "loss": 2.1466193199157715,
      "step": 4980
    },
    {
      "epoch": 0.5073810602858728,
      "grad_norm": 0.12233993411064148,
      "learning_rate": 0.0008555516265912306,
      "loss": 2.2796826362609863,
      "step": 4982
    },
    {
      "epoch": 0.5075847459784806,
      "grad_norm": 0.12066663801670074,
      "learning_rate": 0.0008551980198019802,
      "loss": 2.223788261413574,
      "step": 4984
    },
    {
      "epoch": 0.5077884316710883,
      "grad_norm": 0.12710580229759216,
      "learning_rate": 0.0008548444130127298,
      "loss": 2.262063980102539,
      "step": 4986
    },
    {
      "epoch": 0.5079921173636961,
      "grad_norm": 0.16390135884284973,
      "learning_rate": 0.0008544908062234795,
      "loss": 2.243265151977539,
      "step": 4988
    },
    {
      "epoch": 0.5081958030563039,
      "grad_norm": 0.11351998150348663,
      "learning_rate": 0.0008541371994342292,
      "loss": 2.1079492568969727,
      "step": 4990
    },
    {
      "epoch": 0.5083994887489116,
      "grad_norm": 0.14687949419021606,
      "learning_rate": 0.0008537835926449789,
      "loss": 2.193711280822754,
      "step": 4992
    },
    {
      "epoch": 0.5086031744415193,
      "grad_norm": 0.14203138649463654,
      "learning_rate": 0.0008534299858557284,
      "loss": 2.1728384494781494,
      "step": 4994
    },
    {
      "epoch": 0.508806860134127,
      "grad_norm": 0.127712219953537,
      "learning_rate": 0.0008530763790664781,
      "loss": 2.1585443019866943,
      "step": 4996
    },
    {
      "epoch": 0.5090105458267348,
      "grad_norm": 0.21903108060359955,
      "learning_rate": 0.0008527227722772277,
      "loss": 2.2555124759674072,
      "step": 4998
    },
    {
      "epoch": 0.5092142315193425,
      "grad_norm": 0.13355471193790436,
      "learning_rate": 0.0008523691654879773,
      "loss": 2.2137441635131836,
      "step": 5000
    },
    {
      "epoch": 0.5094179172119503,
      "grad_norm": 0.12607721984386444,
      "learning_rate": 0.0008520155586987271,
      "loss": 2.2624385356903076,
      "step": 5002
    },
    {
      "epoch": 0.5096216029045579,
      "grad_norm": 0.12369659543037415,
      "learning_rate": 0.0008516619519094767,
      "loss": 2.157174825668335,
      "step": 5004
    },
    {
      "epoch": 0.5098252885971657,
      "grad_norm": 0.16258826851844788,
      "learning_rate": 0.0008513083451202264,
      "loss": 2.1680660247802734,
      "step": 5006
    },
    {
      "epoch": 0.5100289742897735,
      "grad_norm": 0.11050062626600266,
      "learning_rate": 0.0008509547383309759,
      "loss": 2.1131086349487305,
      "step": 5008
    },
    {
      "epoch": 0.5102326599823812,
      "grad_norm": 0.13122889399528503,
      "learning_rate": 0.0008506011315417256,
      "loss": 2.170982599258423,
      "step": 5010
    },
    {
      "epoch": 0.510436345674989,
      "grad_norm": 0.10916507244110107,
      "learning_rate": 0.0008502475247524752,
      "loss": 2.1027419567108154,
      "step": 5012
    },
    {
      "epoch": 0.5106400313675966,
      "grad_norm": 0.13671652972698212,
      "learning_rate": 0.0008498939179632249,
      "loss": 2.1125004291534424,
      "step": 5014
    },
    {
      "epoch": 0.5108437170602044,
      "grad_norm": 0.12384458631277084,
      "learning_rate": 0.0008495403111739746,
      "loss": 2.2215585708618164,
      "step": 5016
    },
    {
      "epoch": 0.5110474027528121,
      "grad_norm": 0.12740764021873474,
      "learning_rate": 0.0008491867043847242,
      "loss": 2.3330631256103516,
      "step": 5018
    },
    {
      "epoch": 0.5112510884454199,
      "grad_norm": 0.11171647161245346,
      "learning_rate": 0.0008488330975954739,
      "loss": 2.1949644088745117,
      "step": 5020
    },
    {
      "epoch": 0.5114547741380276,
      "grad_norm": 0.12620732188224792,
      "learning_rate": 0.0008484794908062234,
      "loss": 2.1888999938964844,
      "step": 5022
    },
    {
      "epoch": 0.5116584598306353,
      "grad_norm": 0.11070707440376282,
      "learning_rate": 0.0008481258840169732,
      "loss": 2.074064254760742,
      "step": 5024
    },
    {
      "epoch": 0.511862145523243,
      "grad_norm": 0.1060396358370781,
      "learning_rate": 0.0008477722772277228,
      "loss": 2.1421425342559814,
      "step": 5026
    },
    {
      "epoch": 0.5120658312158508,
      "grad_norm": 0.17912282049655914,
      "learning_rate": 0.0008474186704384724,
      "loss": 2.257978916168213,
      "step": 5028
    },
    {
      "epoch": 0.5122695169084586,
      "grad_norm": 0.1068357452750206,
      "learning_rate": 0.0008470650636492221,
      "loss": 2.097294330596924,
      "step": 5030
    },
    {
      "epoch": 0.5124732026010663,
      "grad_norm": 0.15793606638908386,
      "learning_rate": 0.0008467114568599717,
      "loss": 2.034804344177246,
      "step": 5032
    },
    {
      "epoch": 0.5126768882936741,
      "grad_norm": 0.1358993947505951,
      "learning_rate": 0.0008463578500707214,
      "loss": 2.298523187637329,
      "step": 5034
    },
    {
      "epoch": 0.5128805739862817,
      "grad_norm": 0.10322975367307663,
      "learning_rate": 0.0008460042432814711,
      "loss": 2.1065778732299805,
      "step": 5036
    },
    {
      "epoch": 0.5130842596788895,
      "grad_norm": 0.1718066930770874,
      "learning_rate": 0.0008456506364922207,
      "loss": 2.2068498134613037,
      "step": 5038
    },
    {
      "epoch": 0.5132879453714972,
      "grad_norm": 0.1205066442489624,
      "learning_rate": 0.0008452970297029703,
      "loss": 2.112915515899658,
      "step": 5040
    },
    {
      "epoch": 0.513491631064105,
      "grad_norm": 0.1057831197977066,
      "learning_rate": 0.0008449434229137199,
      "loss": 2.143975019454956,
      "step": 5042
    },
    {
      "epoch": 0.5136953167567128,
      "grad_norm": 0.12325635552406311,
      "learning_rate": 0.0008445898161244696,
      "loss": 2.188964366912842,
      "step": 5044
    },
    {
      "epoch": 0.5138990024493204,
      "grad_norm": 0.10831008106470108,
      "learning_rate": 0.0008442362093352192,
      "loss": 2.17036771774292,
      "step": 5046
    },
    {
      "epoch": 0.5141026881419282,
      "grad_norm": 0.11841947585344315,
      "learning_rate": 0.000843882602545969,
      "loss": 2.1844377517700195,
      "step": 5048
    },
    {
      "epoch": 0.5143063738345359,
      "grad_norm": 0.11877504736185074,
      "learning_rate": 0.0008435289957567186,
      "loss": 2.138848304748535,
      "step": 5050
    },
    {
      "epoch": 0.5145100595271437,
      "grad_norm": 0.13538970053195953,
      "learning_rate": 0.0008431753889674682,
      "loss": 2.0996203422546387,
      "step": 5052
    },
    {
      "epoch": 0.5147137452197514,
      "grad_norm": 0.10899586975574493,
      "learning_rate": 0.0008428217821782178,
      "loss": 2.1536927223205566,
      "step": 5054
    },
    {
      "epoch": 0.5149174309123591,
      "grad_norm": 0.10650411993265152,
      "learning_rate": 0.0008424681753889674,
      "loss": 2.200711250305176,
      "step": 5056
    },
    {
      "epoch": 0.5151211166049668,
      "grad_norm": 0.11705128103494644,
      "learning_rate": 0.0008421145685997172,
      "loss": 2.1684212684631348,
      "step": 5058
    },
    {
      "epoch": 0.5153248022975746,
      "grad_norm": 0.1113666370511055,
      "learning_rate": 0.0008417609618104668,
      "loss": 2.203496217727661,
      "step": 5060
    },
    {
      "epoch": 0.5155284879901824,
      "grad_norm": 0.1223960593342781,
      "learning_rate": 0.0008414073550212165,
      "loss": 2.1488654613494873,
      "step": 5062
    },
    {
      "epoch": 0.5157321736827901,
      "grad_norm": 0.1194973811507225,
      "learning_rate": 0.0008410537482319661,
      "loss": 2.161210536956787,
      "step": 5064
    },
    {
      "epoch": 0.5159358593753979,
      "grad_norm": 0.10832330584526062,
      "learning_rate": 0.0008407001414427157,
      "loss": 2.2489004135131836,
      "step": 5066
    },
    {
      "epoch": 0.5161395450680055,
      "grad_norm": 0.10324008017778397,
      "learning_rate": 0.0008403465346534653,
      "loss": 2.126166343688965,
      "step": 5068
    },
    {
      "epoch": 0.5163432307606133,
      "grad_norm": 0.16477029025554657,
      "learning_rate": 0.000839992927864215,
      "loss": 2.310744285583496,
      "step": 5070
    },
    {
      "epoch": 0.516546916453221,
      "grad_norm": 0.1232154369354248,
      "learning_rate": 0.0008396393210749647,
      "loss": 2.126646041870117,
      "step": 5072
    },
    {
      "epoch": 0.5167506021458288,
      "grad_norm": 0.12828287482261658,
      "learning_rate": 0.0008392857142857143,
      "loss": 2.174060821533203,
      "step": 5074
    },
    {
      "epoch": 0.5169542878384366,
      "grad_norm": 0.18438594043254852,
      "learning_rate": 0.000838932107496464,
      "loss": 2.2895898818969727,
      "step": 5076
    },
    {
      "epoch": 0.5171579735310442,
      "grad_norm": 0.12678952515125275,
      "learning_rate": 0.0008385785007072136,
      "loss": 2.2002785205841064,
      "step": 5078
    },
    {
      "epoch": 0.517361659223652,
      "grad_norm": 0.11629480868577957,
      "learning_rate": 0.0008382248939179632,
      "loss": 2.1650257110595703,
      "step": 5080
    },
    {
      "epoch": 0.5175653449162597,
      "grad_norm": 0.14037558436393738,
      "learning_rate": 0.0008378712871287129,
      "loss": 2.22664737701416,
      "step": 5082
    },
    {
      "epoch": 0.5177690306088675,
      "grad_norm": 0.18306361138820648,
      "learning_rate": 0.0008375176803394625,
      "loss": 2.2177486419677734,
      "step": 5084
    },
    {
      "epoch": 0.5179727163014752,
      "grad_norm": 0.11023323982954025,
      "learning_rate": 0.0008371640735502122,
      "loss": 2.1271634101867676,
      "step": 5086
    },
    {
      "epoch": 0.5181764019940829,
      "grad_norm": 0.14761903882026672,
      "learning_rate": 0.0008368104667609618,
      "loss": 2.2252233028411865,
      "step": 5088
    },
    {
      "epoch": 0.5183800876866906,
      "grad_norm": 0.10682391375303268,
      "learning_rate": 0.0008364568599717115,
      "loss": 2.1482062339782715,
      "step": 5090
    },
    {
      "epoch": 0.5185837733792984,
      "grad_norm": 0.12667791545391083,
      "learning_rate": 0.0008361032531824611,
      "loss": 2.166778326034546,
      "step": 5092
    },
    {
      "epoch": 0.5187874590719062,
      "grad_norm": 0.12551379203796387,
      "learning_rate": 0.0008357496463932108,
      "loss": 2.1661009788513184,
      "step": 5094
    },
    {
      "epoch": 0.5189911447645139,
      "grad_norm": 0.12127218395471573,
      "learning_rate": 0.0008353960396039604,
      "loss": 2.339672327041626,
      "step": 5096
    },
    {
      "epoch": 0.5191948304571217,
      "grad_norm": 0.1296466886997223,
      "learning_rate": 0.00083504243281471,
      "loss": 2.3050570487976074,
      "step": 5098
    },
    {
      "epoch": 0.5193985161497293,
      "grad_norm": 0.2258855104446411,
      "learning_rate": 0.0008346888260254597,
      "loss": 2.049459934234619,
      "step": 5100
    },
    {
      "epoch": 0.5196022018423371,
      "grad_norm": 0.12435401231050491,
      "learning_rate": 0.0008343352192362093,
      "loss": 2.1512506008148193,
      "step": 5102
    },
    {
      "epoch": 0.5198058875349448,
      "grad_norm": 0.14348949491977692,
      "learning_rate": 0.0008339816124469591,
      "loss": 2.1663951873779297,
      "step": 5104
    },
    {
      "epoch": 0.5200095732275526,
      "grad_norm": 0.11625627428293228,
      "learning_rate": 0.0008336280056577087,
      "loss": 2.174859046936035,
      "step": 5106
    },
    {
      "epoch": 0.5202132589201603,
      "grad_norm": 0.1265910118818283,
      "learning_rate": 0.0008332743988684583,
      "loss": 2.3072218894958496,
      "step": 5108
    },
    {
      "epoch": 0.520416944612768,
      "grad_norm": 0.13347850739955902,
      "learning_rate": 0.0008329207920792079,
      "loss": 2.4160280227661133,
      "step": 5110
    },
    {
      "epoch": 0.5206206303053758,
      "grad_norm": 0.12800730764865875,
      "learning_rate": 0.0008325671852899575,
      "loss": 2.0977985858917236,
      "step": 5112
    },
    {
      "epoch": 0.5208243159979835,
      "grad_norm": 0.12365824729204178,
      "learning_rate": 0.0008322135785007072,
      "loss": 2.1059398651123047,
      "step": 5114
    },
    {
      "epoch": 0.5210280016905913,
      "grad_norm": 0.16394305229187012,
      "learning_rate": 0.0008318599717114569,
      "loss": 2.170620918273926,
      "step": 5116
    },
    {
      "epoch": 0.521231687383199,
      "grad_norm": 0.11819277703762054,
      "learning_rate": 0.0008315063649222066,
      "loss": 2.06512451171875,
      "step": 5118
    },
    {
      "epoch": 0.5214353730758067,
      "grad_norm": 0.11292362958192825,
      "learning_rate": 0.0008311527581329562,
      "loss": 2.1826839447021484,
      "step": 5120
    },
    {
      "epoch": 0.5216390587684144,
      "grad_norm": 0.14442545175552368,
      "learning_rate": 0.0008307991513437058,
      "loss": 2.203705072402954,
      "step": 5122
    },
    {
      "epoch": 0.5218427444610222,
      "grad_norm": 0.14532631635665894,
      "learning_rate": 0.0008304455445544554,
      "loss": 2.099902391433716,
      "step": 5124
    },
    {
      "epoch": 0.52204643015363,
      "grad_norm": 0.10644169896841049,
      "learning_rate": 0.000830091937765205,
      "loss": 2.2501120567321777,
      "step": 5126
    },
    {
      "epoch": 0.5222501158462377,
      "grad_norm": 0.12702259421348572,
      "learning_rate": 0.0008297383309759548,
      "loss": 2.18904185295105,
      "step": 5128
    },
    {
      "epoch": 0.5224538015388454,
      "grad_norm": 0.1601700633764267,
      "learning_rate": 0.0008293847241867044,
      "loss": 2.219494104385376,
      "step": 5130
    },
    {
      "epoch": 0.5226574872314531,
      "grad_norm": 0.10143869370222092,
      "learning_rate": 0.0008290311173974541,
      "loss": 2.105084180831909,
      "step": 5132
    },
    {
      "epoch": 0.5228611729240609,
      "grad_norm": 0.13775479793548584,
      "learning_rate": 0.0008286775106082037,
      "loss": 2.0944643020629883,
      "step": 5134
    },
    {
      "epoch": 0.5230648586166686,
      "grad_norm": 0.10659192502498627,
      "learning_rate": 0.0008283239038189534,
      "loss": 2.1481099128723145,
      "step": 5136
    },
    {
      "epoch": 0.5232685443092764,
      "grad_norm": 0.16377289593219757,
      "learning_rate": 0.000827970297029703,
      "loss": 2.216259717941284,
      "step": 5138
    },
    {
      "epoch": 0.5234722300018841,
      "grad_norm": 0.1260034143924713,
      "learning_rate": 0.0008276166902404526,
      "loss": 2.189362049102783,
      "step": 5140
    },
    {
      "epoch": 0.5236759156944918,
      "grad_norm": 0.12827739119529724,
      "learning_rate": 0.0008272630834512023,
      "loss": 2.1100661754608154,
      "step": 5142
    },
    {
      "epoch": 0.5238796013870995,
      "grad_norm": 0.1254393309354782,
      "learning_rate": 0.0008269094766619519,
      "loss": 2.098095417022705,
      "step": 5144
    },
    {
      "epoch": 0.5240832870797073,
      "grad_norm": 0.11370775103569031,
      "learning_rate": 0.0008265558698727016,
      "loss": 2.1878561973571777,
      "step": 5146
    },
    {
      "epoch": 0.5242869727723151,
      "grad_norm": 0.10757619142532349,
      "learning_rate": 0.0008262022630834512,
      "loss": 2.230466365814209,
      "step": 5148
    },
    {
      "epoch": 0.5244906584649228,
      "grad_norm": 0.11732494831085205,
      "learning_rate": 0.000825848656294201,
      "loss": 2.2152819633483887,
      "step": 5150
    },
    {
      "epoch": 0.5246943441575305,
      "grad_norm": 0.1259903460741043,
      "learning_rate": 0.0008254950495049505,
      "loss": 2.190828323364258,
      "step": 5152
    },
    {
      "epoch": 0.5248980298501382,
      "grad_norm": 0.12724237143993378,
      "learning_rate": 0.0008251414427157001,
      "loss": 2.2166922092437744,
      "step": 5154
    },
    {
      "epoch": 0.525101715542746,
      "grad_norm": 0.13432280719280243,
      "learning_rate": 0.0008247878359264498,
      "loss": 2.0902011394500732,
      "step": 5156
    },
    {
      "epoch": 0.5253054012353537,
      "grad_norm": 0.11277803778648376,
      "learning_rate": 0.0008244342291371994,
      "loss": 2.297497272491455,
      "step": 5158
    },
    {
      "epoch": 0.5255090869279615,
      "grad_norm": 0.11834308505058289,
      "learning_rate": 0.0008240806223479491,
      "loss": 2.150355100631714,
      "step": 5160
    },
    {
      "epoch": 0.5257127726205691,
      "grad_norm": 0.1385759860277176,
      "learning_rate": 0.0008237270155586988,
      "loss": 2.320197105407715,
      "step": 5162
    },
    {
      "epoch": 0.5259164583131769,
      "grad_norm": 0.13366083800792694,
      "learning_rate": 0.0008233734087694485,
      "loss": 2.198103427886963,
      "step": 5164
    },
    {
      "epoch": 0.5261201440057847,
      "grad_norm": 0.11068776994943619,
      "learning_rate": 0.000823019801980198,
      "loss": 2.149693489074707,
      "step": 5166
    },
    {
      "epoch": 0.5263238296983924,
      "grad_norm": 0.09877748787403107,
      "learning_rate": 0.0008226661951909476,
      "loss": 2.2081663608551025,
      "step": 5168
    },
    {
      "epoch": 0.5265275153910002,
      "grad_norm": 0.20159295201301575,
      "learning_rate": 0.0008223125884016973,
      "loss": 2.3439087867736816,
      "step": 5170
    },
    {
      "epoch": 0.5267312010836079,
      "grad_norm": 0.11299239844083786,
      "learning_rate": 0.0008219589816124469,
      "loss": 2.18491268157959,
      "step": 5172
    },
    {
      "epoch": 0.5269348867762156,
      "grad_norm": 0.16257445514202118,
      "learning_rate": 0.0008216053748231967,
      "loss": 2.2650375366210938,
      "step": 5174
    },
    {
      "epoch": 0.5271385724688233,
      "grad_norm": 0.13896779716014862,
      "learning_rate": 0.0008212517680339463,
      "loss": 2.1850805282592773,
      "step": 5176
    },
    {
      "epoch": 0.5273422581614311,
      "grad_norm": 0.1404113918542862,
      "learning_rate": 0.000820898161244696,
      "loss": 2.196833610534668,
      "step": 5178
    },
    {
      "epoch": 0.5275459438540389,
      "grad_norm": 0.14703549444675446,
      "learning_rate": 0.0008205445544554455,
      "loss": 2.084629774093628,
      "step": 5180
    },
    {
      "epoch": 0.5277496295466466,
      "grad_norm": 0.11251578480005264,
      "learning_rate": 0.0008201909476661951,
      "loss": 2.320702075958252,
      "step": 5182
    },
    {
      "epoch": 0.5279533152392543,
      "grad_norm": 0.11673443019390106,
      "learning_rate": 0.0008198373408769449,
      "loss": 2.2357163429260254,
      "step": 5184
    },
    {
      "epoch": 0.528157000931862,
      "grad_norm": 0.11428599059581757,
      "learning_rate": 0.0008194837340876945,
      "loss": 2.124011516571045,
      "step": 5186
    },
    {
      "epoch": 0.5283606866244698,
      "grad_norm": 0.11357028037309647,
      "learning_rate": 0.0008191301272984442,
      "loss": 2.154174566268921,
      "step": 5188
    },
    {
      "epoch": 0.5285643723170775,
      "grad_norm": 0.12181391566991806,
      "learning_rate": 0.0008187765205091938,
      "loss": 2.1929473876953125,
      "step": 5190
    },
    {
      "epoch": 0.5287680580096853,
      "grad_norm": 0.13221541047096252,
      "learning_rate": 0.0008184229137199435,
      "loss": 2.2528162002563477,
      "step": 5192
    },
    {
      "epoch": 0.5289717437022929,
      "grad_norm": 0.12194789946079254,
      "learning_rate": 0.000818069306930693,
      "loss": 2.2212376594543457,
      "step": 5194
    },
    {
      "epoch": 0.5291754293949007,
      "grad_norm": 0.11862846463918686,
      "learning_rate": 0.0008177157001414427,
      "loss": 2.129556655883789,
      "step": 5196
    },
    {
      "epoch": 0.5293791150875085,
      "grad_norm": 0.13440634310245514,
      "learning_rate": 0.0008173620933521924,
      "loss": 2.105417490005493,
      "step": 5198
    },
    {
      "epoch": 0.5295828007801162,
      "grad_norm": 0.12092459946870804,
      "learning_rate": 0.000817008486562942,
      "loss": 2.2677931785583496,
      "step": 5200
    },
    {
      "epoch": 0.529786486472724,
      "grad_norm": 0.1394500583410263,
      "learning_rate": 0.0008166548797736917,
      "loss": 2.260242223739624,
      "step": 5202
    },
    {
      "epoch": 0.5299901721653317,
      "grad_norm": 0.17945902049541473,
      "learning_rate": 0.0008163012729844413,
      "loss": 2.3567299842834473,
      "step": 5204
    },
    {
      "epoch": 0.5301938578579394,
      "grad_norm": 0.13393515348434448,
      "learning_rate": 0.000815947666195191,
      "loss": 2.176961898803711,
      "step": 5206
    },
    {
      "epoch": 0.5303975435505471,
      "grad_norm": 0.12032050639390945,
      "learning_rate": 0.0008155940594059406,
      "loss": 2.227409601211548,
      "step": 5208
    },
    {
      "epoch": 0.5306012292431549,
      "grad_norm": 0.10939647257328033,
      "learning_rate": 0.0008152404526166902,
      "loss": 2.284618854522705,
      "step": 5210
    },
    {
      "epoch": 0.5308049149357627,
      "grad_norm": 0.16582220792770386,
      "learning_rate": 0.0008148868458274399,
      "loss": 2.064558506011963,
      "step": 5212
    },
    {
      "epoch": 0.5310086006283704,
      "grad_norm": 0.10561855137348175,
      "learning_rate": 0.0008145332390381895,
      "loss": 2.1174240112304688,
      "step": 5214
    },
    {
      "epoch": 0.531212286320978,
      "grad_norm": 0.1343773454427719,
      "learning_rate": 0.0008141796322489392,
      "loss": 2.2669007778167725,
      "step": 5216
    },
    {
      "epoch": 0.5314159720135858,
      "grad_norm": 0.11250844597816467,
      "learning_rate": 0.0008138260254596889,
      "loss": 2.1945903301239014,
      "step": 5218
    },
    {
      "epoch": 0.5316196577061936,
      "grad_norm": 0.1587960124015808,
      "learning_rate": 0.0008134724186704386,
      "loss": 2.0644850730895996,
      "step": 5220
    },
    {
      "epoch": 0.5318233433988013,
      "grad_norm": 0.16016189754009247,
      "learning_rate": 0.0008131188118811881,
      "loss": 2.2601006031036377,
      "step": 5222
    },
    {
      "epoch": 0.5320270290914091,
      "grad_norm": 0.18519634008407593,
      "learning_rate": 0.0008127652050919377,
      "loss": 2.1467206478118896,
      "step": 5224
    },
    {
      "epoch": 0.5322307147840167,
      "grad_norm": 0.13050612807273865,
      "learning_rate": 0.0008124115983026874,
      "loss": 2.028193473815918,
      "step": 5226
    },
    {
      "epoch": 0.5324344004766245,
      "grad_norm": 0.11968324333429337,
      "learning_rate": 0.000812057991513437,
      "loss": 2.2259373664855957,
      "step": 5228
    },
    {
      "epoch": 0.5326380861692322,
      "grad_norm": 0.14200636744499207,
      "learning_rate": 0.0008117043847241868,
      "loss": 2.2132818698883057,
      "step": 5230
    },
    {
      "epoch": 0.53284177186184,
      "grad_norm": 0.11914645880460739,
      "learning_rate": 0.0008113507779349364,
      "loss": 2.091172456741333,
      "step": 5232
    },
    {
      "epoch": 0.5330454575544478,
      "grad_norm": 0.13462969660758972,
      "learning_rate": 0.0008109971711456861,
      "loss": 2.2007384300231934,
      "step": 5234
    },
    {
      "epoch": 0.5332491432470555,
      "grad_norm": 0.11606946587562561,
      "learning_rate": 0.0008106435643564357,
      "loss": 2.2109384536743164,
      "step": 5236
    },
    {
      "epoch": 0.5334528289396632,
      "grad_norm": 0.13977740705013275,
      "learning_rate": 0.0008102899575671852,
      "loss": 2.2918996810913086,
      "step": 5238
    },
    {
      "epoch": 0.5336565146322709,
      "grad_norm": 0.11626478284597397,
      "learning_rate": 0.0008099363507779349,
      "loss": 2.0695905685424805,
      "step": 5240
    },
    {
      "epoch": 0.5338602003248787,
      "grad_norm": 0.11896582692861557,
      "learning_rate": 0.0008095827439886846,
      "loss": 2.181711196899414,
      "step": 5242
    },
    {
      "epoch": 0.5340638860174864,
      "grad_norm": 0.14181528985500336,
      "learning_rate": 0.0008092291371994343,
      "loss": 2.275689125061035,
      "step": 5244
    },
    {
      "epoch": 0.5342675717100942,
      "grad_norm": 0.12781667709350586,
      "learning_rate": 0.0008088755304101839,
      "loss": 2.172219753265381,
      "step": 5246
    },
    {
      "epoch": 0.5344712574027018,
      "grad_norm": 0.11594336479902267,
      "learning_rate": 0.0008085219236209336,
      "loss": 2.1980910301208496,
      "step": 5248
    },
    {
      "epoch": 0.5346749430953096,
      "grad_norm": 0.13171595335006714,
      "learning_rate": 0.0008081683168316832,
      "loss": 2.2718465328216553,
      "step": 5250
    },
    {
      "epoch": 0.5348786287879174,
      "grad_norm": 0.13006249070167542,
      "learning_rate": 0.0008078147100424327,
      "loss": 2.0673365592956543,
      "step": 5252
    },
    {
      "epoch": 0.5350823144805251,
      "grad_norm": 0.11449848115444183,
      "learning_rate": 0.0008074611032531825,
      "loss": 2.250624179840088,
      "step": 5254
    },
    {
      "epoch": 0.5352860001731329,
      "grad_norm": 0.14292436838150024,
      "learning_rate": 0.0008071074964639321,
      "loss": 2.206151008605957,
      "step": 5256
    },
    {
      "epoch": 0.5354896858657405,
      "grad_norm": 0.14645226299762726,
      "learning_rate": 0.0008067538896746818,
      "loss": 2.178520679473877,
      "step": 5258
    },
    {
      "epoch": 0.5356933715583483,
      "grad_norm": 0.1323077529668808,
      "learning_rate": 0.0008064002828854314,
      "loss": 2.115074872970581,
      "step": 5260
    },
    {
      "epoch": 0.535897057250956,
      "grad_norm": 0.17293553054332733,
      "learning_rate": 0.0008060466760961811,
      "loss": 2.1672232151031494,
      "step": 5262
    },
    {
      "epoch": 0.5361007429435638,
      "grad_norm": 0.11623018980026245,
      "learning_rate": 0.0008056930693069308,
      "loss": 2.128838062286377,
      "step": 5264
    },
    {
      "epoch": 0.5363044286361716,
      "grad_norm": 0.13231678307056427,
      "learning_rate": 0.0008053394625176803,
      "loss": 2.2204246520996094,
      "step": 5266
    },
    {
      "epoch": 0.5365081143287792,
      "grad_norm": 0.14964033663272858,
      "learning_rate": 0.00080498585572843,
      "loss": 2.2563235759735107,
      "step": 5268
    },
    {
      "epoch": 0.536711800021387,
      "grad_norm": 0.1569688767194748,
      "learning_rate": 0.0008046322489391796,
      "loss": 2.061260461807251,
      "step": 5270
    },
    {
      "epoch": 0.5369154857139947,
      "grad_norm": 0.1322789490222931,
      "learning_rate": 0.0008042786421499293,
      "loss": 1.9513612985610962,
      "step": 5272
    },
    {
      "epoch": 0.5371191714066025,
      "grad_norm": 0.13683097064495087,
      "learning_rate": 0.0008039250353606789,
      "loss": 2.367429256439209,
      "step": 5274
    },
    {
      "epoch": 0.5373228570992102,
      "grad_norm": 0.13814878463745117,
      "learning_rate": 0.0008035714285714287,
      "loss": 2.116466999053955,
      "step": 5276
    },
    {
      "epoch": 0.537526542791818,
      "grad_norm": 0.17293033003807068,
      "learning_rate": 0.0008032178217821783,
      "loss": 2.247734546661377,
      "step": 5278
    },
    {
      "epoch": 0.5377302284844256,
      "grad_norm": 0.13070423901081085,
      "learning_rate": 0.0008028642149929278,
      "loss": 2.2992098331451416,
      "step": 5280
    },
    {
      "epoch": 0.5379339141770334,
      "grad_norm": 0.12358853220939636,
      "learning_rate": 0.0008025106082036775,
      "loss": 2.264075756072998,
      "step": 5282
    },
    {
      "epoch": 0.5381375998696412,
      "grad_norm": 0.1677423119544983,
      "learning_rate": 0.0008021570014144271,
      "loss": 2.1622865200042725,
      "step": 5284
    },
    {
      "epoch": 0.5383412855622489,
      "grad_norm": 0.12399329990148544,
      "learning_rate": 0.0008018033946251768,
      "loss": 2.237581968307495,
      "step": 5286
    },
    {
      "epoch": 0.5385449712548567,
      "grad_norm": 0.11848261207342148,
      "learning_rate": 0.0008014497878359265,
      "loss": 2.143383026123047,
      "step": 5288
    },
    {
      "epoch": 0.5387486569474643,
      "grad_norm": 0.13752037286758423,
      "learning_rate": 0.0008010961810466762,
      "loss": 2.315373182296753,
      "step": 5290
    },
    {
      "epoch": 0.5389523426400721,
      "grad_norm": 0.11655480414628983,
      "learning_rate": 0.0008007425742574258,
      "loss": 2.3120508193969727,
      "step": 5292
    },
    {
      "epoch": 0.5391560283326798,
      "grad_norm": 0.10640902817249298,
      "learning_rate": 0.0008003889674681753,
      "loss": 2.2024996280670166,
      "step": 5294
    },
    {
      "epoch": 0.5393597140252876,
      "grad_norm": 0.12252594530582428,
      "learning_rate": 0.000800035360678925,
      "loss": 2.2933223247528076,
      "step": 5296
    },
    {
      "epoch": 0.5395633997178954,
      "grad_norm": 0.14120176434516907,
      "learning_rate": 0.0007996817538896746,
      "loss": 2.2180628776550293,
      "step": 5298
    },
    {
      "epoch": 0.539767085410503,
      "grad_norm": 0.141900897026062,
      "learning_rate": 0.0007993281471004244,
      "loss": 2.281520366668701,
      "step": 5300
    },
    {
      "epoch": 0.5399707711031108,
      "grad_norm": 0.11652051657438278,
      "learning_rate": 0.000798974540311174,
      "loss": 2.3110671043395996,
      "step": 5302
    },
    {
      "epoch": 0.5401744567957185,
      "grad_norm": 0.1494811475276947,
      "learning_rate": 0.0007986209335219237,
      "loss": 2.1971747875213623,
      "step": 5304
    },
    {
      "epoch": 0.5403781424883263,
      "grad_norm": 0.18974202871322632,
      "learning_rate": 0.0007982673267326733,
      "loss": 2.139044761657715,
      "step": 5306
    },
    {
      "epoch": 0.540581828180934,
      "grad_norm": 0.12465883791446686,
      "learning_rate": 0.0007979137199434228,
      "loss": 2.184131622314453,
      "step": 5308
    },
    {
      "epoch": 0.5407855138735418,
      "grad_norm": 0.11915626376867294,
      "learning_rate": 0.0007975601131541726,
      "loss": 2.209097146987915,
      "step": 5310
    },
    {
      "epoch": 0.5409891995661494,
      "grad_norm": 0.11856343597173691,
      "learning_rate": 0.0007972065063649222,
      "loss": 2.2876620292663574,
      "step": 5312
    },
    {
      "epoch": 0.5411928852587572,
      "grad_norm": 0.1093050092458725,
      "learning_rate": 0.0007968528995756719,
      "loss": 2.231363296508789,
      "step": 5314
    },
    {
      "epoch": 0.541396570951365,
      "grad_norm": 0.12352319806814194,
      "learning_rate": 0.0007964992927864215,
      "loss": 2.0712504386901855,
      "step": 5316
    },
    {
      "epoch": 0.5416002566439727,
      "grad_norm": 0.1495259702205658,
      "learning_rate": 0.0007961456859971712,
      "loss": 2.1918773651123047,
      "step": 5318
    },
    {
      "epoch": 0.5418039423365805,
      "grad_norm": 0.12062849849462509,
      "learning_rate": 0.0007957920792079208,
      "loss": 2.303311824798584,
      "step": 5320
    },
    {
      "epoch": 0.5420076280291881,
      "grad_norm": 0.10673106461763382,
      "learning_rate": 0.0007954384724186704,
      "loss": 2.0955119132995605,
      "step": 5322
    },
    {
      "epoch": 0.5422113137217959,
      "grad_norm": 0.12796102464199066,
      "learning_rate": 0.0007950848656294201,
      "loss": 2.165935754776001,
      "step": 5324
    },
    {
      "epoch": 0.5424149994144036,
      "grad_norm": 0.12113739550113678,
      "learning_rate": 0.0007947312588401697,
      "loss": 2.181358575820923,
      "step": 5326
    },
    {
      "epoch": 0.5426186851070114,
      "grad_norm": 0.1037791296839714,
      "learning_rate": 0.0007943776520509194,
      "loss": 2.037444591522217,
      "step": 5328
    },
    {
      "epoch": 0.5428223707996191,
      "grad_norm": 0.13072717189788818,
      "learning_rate": 0.000794024045261669,
      "loss": 2.267014265060425,
      "step": 5330
    },
    {
      "epoch": 0.5430260564922268,
      "grad_norm": 0.12374146282672882,
      "learning_rate": 0.0007936704384724188,
      "loss": 2.179870128631592,
      "step": 5332
    },
    {
      "epoch": 0.5432297421848346,
      "grad_norm": 0.14277249574661255,
      "learning_rate": 0.0007933168316831684,
      "loss": 2.101255416870117,
      "step": 5334
    },
    {
      "epoch": 0.5434334278774423,
      "grad_norm": 0.12442793697118759,
      "learning_rate": 0.000792963224893918,
      "loss": 2.268251657485962,
      "step": 5336
    },
    {
      "epoch": 0.5436371135700501,
      "grad_norm": 0.12436573952436447,
      "learning_rate": 0.0007926096181046676,
      "loss": 2.090402126312256,
      "step": 5338
    },
    {
      "epoch": 0.5438407992626578,
      "grad_norm": 0.1925506591796875,
      "learning_rate": 0.0007922560113154172,
      "loss": 2.272948741912842,
      "step": 5340
    },
    {
      "epoch": 0.5440444849552656,
      "grad_norm": 0.10873495042324066,
      "learning_rate": 0.0007919024045261669,
      "loss": 2.1838059425354004,
      "step": 5342
    },
    {
      "epoch": 0.5442481706478732,
      "grad_norm": 0.10734472423791885,
      "learning_rate": 0.0007915487977369166,
      "loss": 2.1729161739349365,
      "step": 5344
    },
    {
      "epoch": 0.544451856340481,
      "grad_norm": 0.11548444628715515,
      "learning_rate": 0.0007911951909476663,
      "loss": 2.2454967498779297,
      "step": 5346
    },
    {
      "epoch": 0.5446555420330887,
      "grad_norm": 0.10911649465560913,
      "learning_rate": 0.0007908415841584159,
      "loss": 2.1299619674682617,
      "step": 5348
    },
    {
      "epoch": 0.5448592277256965,
      "grad_norm": 0.11679762601852417,
      "learning_rate": 0.0007904879773691655,
      "loss": 2.1421523094177246,
      "step": 5350
    },
    {
      "epoch": 0.5450629134183043,
      "grad_norm": 0.11888381093740463,
      "learning_rate": 0.0007901343705799151,
      "loss": 2.219864845275879,
      "step": 5352
    },
    {
      "epoch": 0.5452665991109119,
      "grad_norm": 0.09826654940843582,
      "learning_rate": 0.0007897807637906647,
      "loss": 2.044268846511841,
      "step": 5354
    },
    {
      "epoch": 0.5454702848035197,
      "grad_norm": 0.11918283253908157,
      "learning_rate": 0.0007894271570014145,
      "loss": 2.1726956367492676,
      "step": 5356
    },
    {
      "epoch": 0.5456739704961274,
      "grad_norm": 0.12886692583560944,
      "learning_rate": 0.0007890735502121641,
      "loss": 2.312138080596924,
      "step": 5358
    },
    {
      "epoch": 0.5458776561887352,
      "grad_norm": 0.11691565811634064,
      "learning_rate": 0.0007887199434229138,
      "loss": 2.1417927742004395,
      "step": 5360
    },
    {
      "epoch": 0.5460813418813429,
      "grad_norm": 0.1122870072722435,
      "learning_rate": 0.0007883663366336634,
      "loss": 2.0286943912506104,
      "step": 5362
    },
    {
      "epoch": 0.5462850275739506,
      "grad_norm": 0.12238631397485733,
      "learning_rate": 0.000788012729844413,
      "loss": 2.246967315673828,
      "step": 5364
    },
    {
      "epoch": 0.5464887132665583,
      "grad_norm": 0.1875850111246109,
      "learning_rate": 0.0007876591230551626,
      "loss": 2.148320198059082,
      "step": 5366
    },
    {
      "epoch": 0.5466923989591661,
      "grad_norm": 0.11179684847593307,
      "learning_rate": 0.0007873055162659123,
      "loss": 2.2059571743011475,
      "step": 5368
    },
    {
      "epoch": 0.5468960846517739,
      "grad_norm": 0.11295250058174133,
      "learning_rate": 0.000786951909476662,
      "loss": 2.225703239440918,
      "step": 5370
    },
    {
      "epoch": 0.5470997703443816,
      "grad_norm": 0.20771226286888123,
      "learning_rate": 0.0007865983026874116,
      "loss": 2.25679087638855,
      "step": 5372
    },
    {
      "epoch": 0.5473034560369893,
      "grad_norm": 0.12048251926898956,
      "learning_rate": 0.0007862446958981613,
      "loss": 2.1028501987457275,
      "step": 5374
    },
    {
      "epoch": 0.547507141729597,
      "grad_norm": 0.11392420530319214,
      "learning_rate": 0.0007858910891089109,
      "loss": 2.274704694747925,
      "step": 5376
    },
    {
      "epoch": 0.5477108274222048,
      "grad_norm": 0.12857267260551453,
      "learning_rate": 0.0007855374823196606,
      "loss": 2.1960277557373047,
      "step": 5378
    },
    {
      "epoch": 0.5479145131148125,
      "grad_norm": 0.10829721391201019,
      "learning_rate": 0.0007851838755304102,
      "loss": 2.15274977684021,
      "step": 5380
    },
    {
      "epoch": 0.5481181988074203,
      "grad_norm": 0.11893216520547867,
      "learning_rate": 0.0007848302687411598,
      "loss": 2.277278423309326,
      "step": 5382
    },
    {
      "epoch": 0.5483218845000281,
      "grad_norm": 0.10940895229578018,
      "learning_rate": 0.0007844766619519095,
      "loss": 2.2065067291259766,
      "step": 5384
    },
    {
      "epoch": 0.5485255701926357,
      "grad_norm": 0.1107495129108429,
      "learning_rate": 0.0007841230551626591,
      "loss": 1.988950490951538,
      "step": 5386
    },
    {
      "epoch": 0.5487292558852435,
      "grad_norm": 0.1373213827610016,
      "learning_rate": 0.0007837694483734088,
      "loss": 2.063143730163574,
      "step": 5388
    },
    {
      "epoch": 0.5489329415778512,
      "grad_norm": 0.10859141498804092,
      "learning_rate": 0.0007834158415841585,
      "loss": 2.1011669635772705,
      "step": 5390
    },
    {
      "epoch": 0.549136627270459,
      "grad_norm": 0.1453084498643875,
      "learning_rate": 0.0007830622347949081,
      "loss": 2.3189151287078857,
      "step": 5392
    },
    {
      "epoch": 0.5493403129630667,
      "grad_norm": 0.21547430753707886,
      "learning_rate": 0.0007827086280056577,
      "loss": 2.242112636566162,
      "step": 5394
    },
    {
      "epoch": 0.5495439986556744,
      "grad_norm": 0.12134073674678802,
      "learning_rate": 0.0007823550212164073,
      "loss": 2.243335723876953,
      "step": 5396
    },
    {
      "epoch": 0.5497476843482821,
      "grad_norm": 0.13093750178813934,
      "learning_rate": 0.000782001414427157,
      "loss": 2.253917694091797,
      "step": 5398
    },
    {
      "epoch": 0.5499513700408899,
      "grad_norm": 0.12389874458312988,
      "learning_rate": 0.0007816478076379066,
      "loss": 2.243673324584961,
      "step": 5400
    },
    {
      "epoch": 0.5501550557334977,
      "grad_norm": 0.12094106525182724,
      "learning_rate": 0.0007812942008486564,
      "loss": 2.246340751647949,
      "step": 5402
    },
    {
      "epoch": 0.5503587414261054,
      "grad_norm": 0.1240958571434021,
      "learning_rate": 0.000780940594059406,
      "loss": 2.087345600128174,
      "step": 5404
    },
    {
      "epoch": 0.5505624271187131,
      "grad_norm": 0.12217339873313904,
      "learning_rate": 0.0007805869872701556,
      "loss": 2.1736154556274414,
      "step": 5406
    },
    {
      "epoch": 0.5507661128113208,
      "grad_norm": 0.12415944039821625,
      "learning_rate": 0.0007802333804809052,
      "loss": 2.222632884979248,
      "step": 5408
    },
    {
      "epoch": 0.5509697985039286,
      "grad_norm": 0.12799328565597534,
      "learning_rate": 0.0007798797736916548,
      "loss": 2.200870990753174,
      "step": 5410
    },
    {
      "epoch": 0.5511734841965363,
      "grad_norm": 0.11702173203229904,
      "learning_rate": 0.0007795261669024046,
      "loss": 2.2857747077941895,
      "step": 5412
    },
    {
      "epoch": 0.5513771698891441,
      "grad_norm": 0.12212134152650833,
      "learning_rate": 0.0007791725601131542,
      "loss": 2.064748764038086,
      "step": 5414
    },
    {
      "epoch": 0.5515808555817518,
      "grad_norm": 0.13553938269615173,
      "learning_rate": 0.0007788189533239039,
      "loss": 2.230475902557373,
      "step": 5416
    },
    {
      "epoch": 0.5517845412743595,
      "grad_norm": 0.14421720802783966,
      "learning_rate": 0.0007784653465346535,
      "loss": 2.0755066871643066,
      "step": 5418
    },
    {
      "epoch": 0.5519882269669673,
      "grad_norm": 0.15038596093654633,
      "learning_rate": 0.0007781117397454031,
      "loss": 2.349822521209717,
      "step": 5420
    },
    {
      "epoch": 0.552191912659575,
      "grad_norm": 0.11526555567979813,
      "learning_rate": 0.0007777581329561527,
      "loss": 2.1431446075439453,
      "step": 5422
    },
    {
      "epoch": 0.5523955983521828,
      "grad_norm": 0.10479426383972168,
      "learning_rate": 0.0007774045261669024,
      "loss": 2.1232223510742188,
      "step": 5424
    },
    {
      "epoch": 0.5525992840447905,
      "grad_norm": 0.10967665910720825,
      "learning_rate": 0.0007770509193776521,
      "loss": 2.203411340713501,
      "step": 5426
    },
    {
      "epoch": 0.5528029697373982,
      "grad_norm": 0.10369928181171417,
      "learning_rate": 0.0007766973125884017,
      "loss": 2.1811447143554688,
      "step": 5428
    },
    {
      "epoch": 0.5530066554300059,
      "grad_norm": 0.12151315063238144,
      "learning_rate": 0.0007763437057991514,
      "loss": 2.2590713500976562,
      "step": 5430
    },
    {
      "epoch": 0.5532103411226137,
      "grad_norm": 0.1407015025615692,
      "learning_rate": 0.000775990099009901,
      "loss": 2.1766867637634277,
      "step": 5432
    },
    {
      "epoch": 0.5534140268152214,
      "grad_norm": 0.10782373696565628,
      "learning_rate": 0.0007756364922206506,
      "loss": 2.2598071098327637,
      "step": 5434
    },
    {
      "epoch": 0.5536177125078292,
      "grad_norm": 0.12348472326993942,
      "learning_rate": 0.0007752828854314004,
      "loss": 2.1248276233673096,
      "step": 5436
    },
    {
      "epoch": 0.5538213982004369,
      "grad_norm": 0.14899663627147675,
      "learning_rate": 0.0007749292786421499,
      "loss": 2.2195653915405273,
      "step": 5438
    },
    {
      "epoch": 0.5540250838930446,
      "grad_norm": 0.10735470056533813,
      "learning_rate": 0.0007745756718528996,
      "loss": 2.125946044921875,
      "step": 5440
    },
    {
      "epoch": 0.5542287695856524,
      "grad_norm": 0.13414768874645233,
      "learning_rate": 0.0007742220650636492,
      "loss": 2.138725757598877,
      "step": 5442
    },
    {
      "epoch": 0.5544324552782601,
      "grad_norm": 0.11430314183235168,
      "learning_rate": 0.0007738684582743989,
      "loss": 2.265775203704834,
      "step": 5444
    },
    {
      "epoch": 0.5546361409708679,
      "grad_norm": 0.11863192915916443,
      "learning_rate": 0.0007735148514851486,
      "loss": 2.2101926803588867,
      "step": 5446
    },
    {
      "epoch": 0.5548398266634756,
      "grad_norm": 0.12555073201656342,
      "learning_rate": 0.0007731612446958982,
      "loss": 2.1564784049987793,
      "step": 5448
    },
    {
      "epoch": 0.5550435123560833,
      "grad_norm": 0.12202906608581543,
      "learning_rate": 0.0007728076379066479,
      "loss": 2.277492046356201,
      "step": 5450
    },
    {
      "epoch": 0.555247198048691,
      "grad_norm": 0.1245780661702156,
      "learning_rate": 0.0007724540311173974,
      "loss": 2.208643674850464,
      "step": 5452
    },
    {
      "epoch": 0.5554508837412988,
      "grad_norm": 0.12742605805397034,
      "learning_rate": 0.0007721004243281471,
      "loss": 2.2650575637817383,
      "step": 5454
    },
    {
      "epoch": 0.5556545694339066,
      "grad_norm": 0.18629691004753113,
      "learning_rate": 0.0007717468175388967,
      "loss": 2.1824657917022705,
      "step": 5456
    },
    {
      "epoch": 0.5558582551265143,
      "grad_norm": 0.14174389839172363,
      "learning_rate": 0.0007713932107496465,
      "loss": 2.2312045097351074,
      "step": 5458
    },
    {
      "epoch": 0.556061940819122,
      "grad_norm": 0.14883537590503693,
      "learning_rate": 0.0007710396039603961,
      "loss": 1.9981105327606201,
      "step": 5460
    },
    {
      "epoch": 0.5562656265117297,
      "grad_norm": 0.24698324501514435,
      "learning_rate": 0.0007706859971711457,
      "loss": 2.079042911529541,
      "step": 5462
    },
    {
      "epoch": 0.5564693122043375,
      "grad_norm": 0.11522044986486435,
      "learning_rate": 0.0007703323903818954,
      "loss": 2.243058443069458,
      "step": 5464
    },
    {
      "epoch": 0.5566729978969452,
      "grad_norm": 0.11745002120733261,
      "learning_rate": 0.0007699787835926449,
      "loss": 2.1751461029052734,
      "step": 5466
    },
    {
      "epoch": 0.556876683589553,
      "grad_norm": 0.12648165225982666,
      "learning_rate": 0.0007696251768033946,
      "loss": 2.1258010864257812,
      "step": 5468
    },
    {
      "epoch": 0.5570803692821606,
      "grad_norm": 0.17080917954444885,
      "learning_rate": 0.0007692715700141443,
      "loss": 2.1963648796081543,
      "step": 5470
    },
    {
      "epoch": 0.5572840549747684,
      "grad_norm": 0.1391838937997818,
      "learning_rate": 0.000768917963224894,
      "loss": 2.2353696823120117,
      "step": 5472
    },
    {
      "epoch": 0.5574877406673762,
      "grad_norm": 0.16418743133544922,
      "learning_rate": 0.0007685643564356436,
      "loss": 2.17199969291687,
      "step": 5474
    },
    {
      "epoch": 0.5576914263599839,
      "grad_norm": 0.13869188725948334,
      "learning_rate": 0.0007682107496463932,
      "loss": 2.184211254119873,
      "step": 5476
    },
    {
      "epoch": 0.5578951120525917,
      "grad_norm": 0.18197417259216309,
      "learning_rate": 0.0007678571428571429,
      "loss": 2.2816295623779297,
      "step": 5478
    },
    {
      "epoch": 0.5580987977451994,
      "grad_norm": 0.12123195081949234,
      "learning_rate": 0.0007675035360678924,
      "loss": 2.211876392364502,
      "step": 5480
    },
    {
      "epoch": 0.5583024834378071,
      "grad_norm": 0.12275488674640656,
      "learning_rate": 0.0007671499292786422,
      "loss": 2.184711456298828,
      "step": 5482
    },
    {
      "epoch": 0.5585061691304148,
      "grad_norm": 0.1449512541294098,
      "learning_rate": 0.0007667963224893918,
      "loss": 2.199005126953125,
      "step": 5484
    },
    {
      "epoch": 0.5587098548230226,
      "grad_norm": 0.12805058062076569,
      "learning_rate": 0.0007664427157001415,
      "loss": 2.160581588745117,
      "step": 5486
    },
    {
      "epoch": 0.5589135405156304,
      "grad_norm": 0.12123409658670425,
      "learning_rate": 0.0007660891089108911,
      "loss": 2.1456246376037598,
      "step": 5488
    },
    {
      "epoch": 0.5591172262082381,
      "grad_norm": 0.13756166398525238,
      "learning_rate": 0.0007657355021216407,
      "loss": 2.2305946350097656,
      "step": 5490
    },
    {
      "epoch": 0.5593209119008458,
      "grad_norm": 0.1444600224494934,
      "learning_rate": 0.0007653818953323905,
      "loss": 2.309861660003662,
      "step": 5492
    },
    {
      "epoch": 0.5595245975934535,
      "grad_norm": 0.12502315640449524,
      "learning_rate": 0.00076502828854314,
      "loss": 2.116710662841797,
      "step": 5494
    },
    {
      "epoch": 0.5597282832860613,
      "grad_norm": 0.11468110978603363,
      "learning_rate": 0.0007646746817538897,
      "loss": 2.125654697418213,
      "step": 5496
    },
    {
      "epoch": 0.559931968978669,
      "grad_norm": 0.13653726875782013,
      "learning_rate": 0.0007643210749646393,
      "loss": 2.36659836769104,
      "step": 5498
    },
    {
      "epoch": 0.5601356546712768,
      "grad_norm": 0.12316544353961945,
      "learning_rate": 0.000763967468175389,
      "loss": 2.2145047187805176,
      "step": 5500
    },
    {
      "epoch": 0.5603393403638844,
      "grad_norm": 0.10237129032611847,
      "learning_rate": 0.0007636138613861386,
      "loss": 2.195134162902832,
      "step": 5502
    },
    {
      "epoch": 0.5605430260564922,
      "grad_norm": 0.1693531721830368,
      "learning_rate": 0.0007632602545968883,
      "loss": 2.1791422367095947,
      "step": 5504
    },
    {
      "epoch": 0.5607467117491,
      "grad_norm": 0.1264389008283615,
      "learning_rate": 0.000762906647807638,
      "loss": 2.2286808490753174,
      "step": 5506
    },
    {
      "epoch": 0.5609503974417077,
      "grad_norm": 0.09869515150785446,
      "learning_rate": 0.0007625530410183875,
      "loss": 2.167285442352295,
      "step": 5508
    },
    {
      "epoch": 0.5611540831343155,
      "grad_norm": 0.11070269346237183,
      "learning_rate": 0.0007621994342291372,
      "loss": 2.241633892059326,
      "step": 5510
    },
    {
      "epoch": 0.5613577688269231,
      "grad_norm": 0.17682640254497528,
      "learning_rate": 0.0007618458274398868,
      "loss": 2.249135732650757,
      "step": 5512
    },
    {
      "epoch": 0.5615614545195309,
      "grad_norm": 0.1398433893918991,
      "learning_rate": 0.0007614922206506365,
      "loss": 2.1333439350128174,
      "step": 5514
    },
    {
      "epoch": 0.5617651402121386,
      "grad_norm": 0.14497871696949005,
      "learning_rate": 0.0007611386138613862,
      "loss": 2.048008680343628,
      "step": 5516
    },
    {
      "epoch": 0.5619688259047464,
      "grad_norm": 0.13310587406158447,
      "learning_rate": 0.0007607850070721358,
      "loss": 2.1080758571624756,
      "step": 5518
    },
    {
      "epoch": 0.5621725115973542,
      "grad_norm": 0.12043895572423935,
      "learning_rate": 0.0007604314002828855,
      "loss": 2.254199981689453,
      "step": 5520
    },
    {
      "epoch": 0.5623761972899619,
      "grad_norm": 0.16091550886631012,
      "learning_rate": 0.000760077793493635,
      "loss": 2.1024727821350098,
      "step": 5522
    },
    {
      "epoch": 0.5625798829825696,
      "grad_norm": 0.13031473755836487,
      "learning_rate": 0.0007597241867043847,
      "loss": 2.108842372894287,
      "step": 5524
    },
    {
      "epoch": 0.5627835686751773,
      "grad_norm": 0.10596181452274323,
      "learning_rate": 0.0007593705799151343,
      "loss": 2.0818216800689697,
      "step": 5526
    },
    {
      "epoch": 0.5629872543677851,
      "grad_norm": 0.12174347788095474,
      "learning_rate": 0.0007590169731258841,
      "loss": 2.078129768371582,
      "step": 5528
    },
    {
      "epoch": 0.5631909400603928,
      "grad_norm": 0.16686631739139557,
      "learning_rate": 0.0007586633663366337,
      "loss": 2.076770782470703,
      "step": 5530
    },
    {
      "epoch": 0.5633946257530006,
      "grad_norm": 0.14236228168010712,
      "learning_rate": 0.0007583097595473833,
      "loss": 2.329357624053955,
      "step": 5532
    },
    {
      "epoch": 0.5635983114456082,
      "grad_norm": 0.19253623485565186,
      "learning_rate": 0.000757956152758133,
      "loss": 2.218338966369629,
      "step": 5534
    },
    {
      "epoch": 0.563801997138216,
      "grad_norm": 0.15200982987880707,
      "learning_rate": 0.0007576025459688826,
      "loss": 2.2943201065063477,
      "step": 5536
    },
    {
      "epoch": 0.5640056828308238,
      "grad_norm": 0.09773068130016327,
      "learning_rate": 0.0007572489391796323,
      "loss": 2.1221587657928467,
      "step": 5538
    },
    {
      "epoch": 0.5642093685234315,
      "grad_norm": 0.11191946268081665,
      "learning_rate": 0.0007568953323903819,
      "loss": 2.1315903663635254,
      "step": 5540
    },
    {
      "epoch": 0.5644130542160393,
      "grad_norm": 0.11184900254011154,
      "learning_rate": 0.0007565417256011316,
      "loss": 2.206247329711914,
      "step": 5542
    },
    {
      "epoch": 0.5646167399086469,
      "grad_norm": 0.12547892332077026,
      "learning_rate": 0.0007561881188118812,
      "loss": 2.2616868019104004,
      "step": 5544
    },
    {
      "epoch": 0.5648204256012547,
      "grad_norm": 0.11614930629730225,
      "learning_rate": 0.0007558345120226308,
      "loss": 2.156707763671875,
      "step": 5546
    },
    {
      "epoch": 0.5650241112938624,
      "grad_norm": 0.12440509349107742,
      "learning_rate": 0.0007554809052333805,
      "loss": 2.2248895168304443,
      "step": 5548
    },
    {
      "epoch": 0.5652277969864702,
      "grad_norm": 0.12463603168725967,
      "learning_rate": 0.0007551272984441302,
      "loss": 2.169832706451416,
      "step": 5550
    },
    {
      "epoch": 0.565431482679078,
      "grad_norm": 0.10741818696260452,
      "learning_rate": 0.0007547736916548798,
      "loss": 2.042513847351074,
      "step": 5552
    },
    {
      "epoch": 0.5656351683716857,
      "grad_norm": 0.18778938055038452,
      "learning_rate": 0.0007544200848656294,
      "loss": 2.132786750793457,
      "step": 5554
    },
    {
      "epoch": 0.5658388540642934,
      "grad_norm": 0.1224108636379242,
      "learning_rate": 0.0007540664780763791,
      "loss": 2.1636898517608643,
      "step": 5556
    },
    {
      "epoch": 0.5660425397569011,
      "grad_norm": 0.12876828014850616,
      "learning_rate": 0.0007537128712871287,
      "loss": 2.0716824531555176,
      "step": 5558
    },
    {
      "epoch": 0.5662462254495089,
      "grad_norm": 0.13119415938854218,
      "learning_rate": 0.0007533592644978783,
      "loss": 2.115713119506836,
      "step": 5560
    },
    {
      "epoch": 0.5664499111421166,
      "grad_norm": 0.12610168755054474,
      "learning_rate": 0.0007530056577086281,
      "loss": 2.246218681335449,
      "step": 5562
    },
    {
      "epoch": 0.5666535968347244,
      "grad_norm": 0.11753132939338684,
      "learning_rate": 0.0007526520509193777,
      "loss": 2.213016986846924,
      "step": 5564
    },
    {
      "epoch": 0.566857282527332,
      "grad_norm": 0.10466507077217102,
      "learning_rate": 0.0007522984441301273,
      "loss": 2.2209348678588867,
      "step": 5566
    },
    {
      "epoch": 0.5670609682199398,
      "grad_norm": 0.15010905265808105,
      "learning_rate": 0.0007519448373408769,
      "loss": 2.1415112018585205,
      "step": 5568
    },
    {
      "epoch": 0.5672646539125475,
      "grad_norm": 0.11446774750947952,
      "learning_rate": 0.0007515912305516266,
      "loss": 2.2216758728027344,
      "step": 5570
    },
    {
      "epoch": 0.5674683396051553,
      "grad_norm": 0.13279752433300018,
      "learning_rate": 0.0007512376237623763,
      "loss": 2.256660223007202,
      "step": 5572
    },
    {
      "epoch": 0.5676720252977631,
      "grad_norm": 0.12397786974906921,
      "learning_rate": 0.0007508840169731259,
      "loss": 2.1019246578216553,
      "step": 5574
    },
    {
      "epoch": 0.5678757109903707,
      "grad_norm": 0.11810149997472763,
      "learning_rate": 0.0007505304101838756,
      "loss": 2.1080479621887207,
      "step": 5576
    },
    {
      "epoch": 0.5680793966829785,
      "grad_norm": 0.12142546474933624,
      "learning_rate": 0.0007501768033946252,
      "loss": 2.097409725189209,
      "step": 5578
    },
    {
      "epoch": 0.5682830823755862,
      "grad_norm": 0.15604248642921448,
      "learning_rate": 0.0007498231966053748,
      "loss": 2.0749728679656982,
      "step": 5580
    },
    {
      "epoch": 0.568486768068194,
      "grad_norm": 0.123172327876091,
      "learning_rate": 0.0007494695898161244,
      "loss": 2.122084856033325,
      "step": 5582
    },
    {
      "epoch": 0.5686904537608017,
      "grad_norm": 0.11206883937120438,
      "learning_rate": 0.0007491159830268742,
      "loss": 2.2141404151916504,
      "step": 5584
    },
    {
      "epoch": 0.5688941394534095,
      "grad_norm": 0.10153889656066895,
      "learning_rate": 0.0007487623762376238,
      "loss": 2.2096760272979736,
      "step": 5586
    },
    {
      "epoch": 0.5690978251460171,
      "grad_norm": 0.17403686046600342,
      "learning_rate": 0.0007484087694483734,
      "loss": 2.1911303997039795,
      "step": 5588
    },
    {
      "epoch": 0.5693015108386249,
      "grad_norm": 0.1021316796541214,
      "learning_rate": 0.0007480551626591231,
      "loss": 2.1196603775024414,
      "step": 5590
    },
    {
      "epoch": 0.5695051965312327,
      "grad_norm": 0.1253519505262375,
      "learning_rate": 0.0007477015558698727,
      "loss": 2.1641898155212402,
      "step": 5592
    },
    {
      "epoch": 0.5697088822238404,
      "grad_norm": 0.11909963935613632,
      "learning_rate": 0.0007473479490806223,
      "loss": 2.2786197662353516,
      "step": 5594
    },
    {
      "epoch": 0.5699125679164482,
      "grad_norm": 0.11770684272050858,
      "learning_rate": 0.000746994342291372,
      "loss": 2.2450907230377197,
      "step": 5596
    },
    {
      "epoch": 0.5701162536090558,
      "grad_norm": 0.11919151246547699,
      "learning_rate": 0.0007466407355021217,
      "loss": 2.061859607696533,
      "step": 5598
    },
    {
      "epoch": 0.5703199393016636,
      "grad_norm": 0.11989054828882217,
      "learning_rate": 0.0007462871287128713,
      "loss": 2.0634613037109375,
      "step": 5600
    },
    {
      "epoch": 0.5705236249942713,
      "grad_norm": 0.12115888297557831,
      "learning_rate": 0.0007459335219236209,
      "loss": 2.3590641021728516,
      "step": 5602
    },
    {
      "epoch": 0.5707273106868791,
      "grad_norm": 0.12585215270519257,
      "learning_rate": 0.0007455799151343706,
      "loss": 2.1100969314575195,
      "step": 5604
    },
    {
      "epoch": 0.5709309963794869,
      "grad_norm": 0.13560695946216583,
      "learning_rate": 0.0007452263083451203,
      "loss": 2.1116816997528076,
      "step": 5606
    },
    {
      "epoch": 0.5711346820720945,
      "grad_norm": 0.12365266680717468,
      "learning_rate": 0.0007448727015558699,
      "loss": 2.097095012664795,
      "step": 5608
    },
    {
      "epoch": 0.5713383677647023,
      "grad_norm": 0.11030678451061249,
      "learning_rate": 0.0007445190947666195,
      "loss": 2.007403612136841,
      "step": 5610
    },
    {
      "epoch": 0.57154205345731,
      "grad_norm": 0.11775430291891098,
      "learning_rate": 0.0007441654879773692,
      "loss": 2.286966323852539,
      "step": 5612
    },
    {
      "epoch": 0.5717457391499178,
      "grad_norm": 0.12257828563451767,
      "learning_rate": 0.0007438118811881188,
      "loss": 2.1636693477630615,
      "step": 5614
    },
    {
      "epoch": 0.5719494248425255,
      "grad_norm": 0.1327316164970398,
      "learning_rate": 0.0007434582743988684,
      "loss": 2.2045092582702637,
      "step": 5616
    },
    {
      "epoch": 0.5721531105351333,
      "grad_norm": 0.10372668504714966,
      "learning_rate": 0.0007431046676096182,
      "loss": 2.0279693603515625,
      "step": 5618
    },
    {
      "epoch": 0.5723567962277409,
      "grad_norm": 0.13285857439041138,
      "learning_rate": 0.0007427510608203678,
      "loss": 2.3299806118011475,
      "step": 5620
    },
    {
      "epoch": 0.5725604819203487,
      "grad_norm": 0.12562139332294464,
      "learning_rate": 0.0007423974540311174,
      "loss": 2.2645797729492188,
      "step": 5622
    },
    {
      "epoch": 0.5727641676129565,
      "grad_norm": 0.12214555591344833,
      "learning_rate": 0.000742043847241867,
      "loss": 2.2804007530212402,
      "step": 5624
    },
    {
      "epoch": 0.5729678533055642,
      "grad_norm": 0.12079360336065292,
      "learning_rate": 0.0007416902404526167,
      "loss": 2.2410755157470703,
      "step": 5626
    },
    {
      "epoch": 0.573171538998172,
      "grad_norm": 0.12311611324548721,
      "learning_rate": 0.0007413366336633663,
      "loss": 2.042483329772949,
      "step": 5628
    },
    {
      "epoch": 0.5733752246907796,
      "grad_norm": 0.1292836219072342,
      "learning_rate": 0.000740983026874116,
      "loss": 2.2727954387664795,
      "step": 5630
    },
    {
      "epoch": 0.5735789103833874,
      "grad_norm": 0.1301746815443039,
      "learning_rate": 0.0007406294200848657,
      "loss": 2.1457064151763916,
      "step": 5632
    },
    {
      "epoch": 0.5737825960759951,
      "grad_norm": 0.104102224111557,
      "learning_rate": 0.0007402758132956153,
      "loss": 2.076968193054199,
      "step": 5634
    },
    {
      "epoch": 0.5739862817686029,
      "grad_norm": 0.11525720357894897,
      "learning_rate": 0.000739922206506365,
      "loss": 2.312845468521118,
      "step": 5636
    },
    {
      "epoch": 0.5741899674612106,
      "grad_norm": 0.12652763724327087,
      "learning_rate": 0.0007395685997171145,
      "loss": 2.227560043334961,
      "step": 5638
    },
    {
      "epoch": 0.5743936531538183,
      "grad_norm": 0.12284594029188156,
      "learning_rate": 0.0007392149929278643,
      "loss": 2.018963098526001,
      "step": 5640
    },
    {
      "epoch": 0.574597338846426,
      "grad_norm": 0.1133703663945198,
      "learning_rate": 0.0007388613861386139,
      "loss": 2.284489631652832,
      "step": 5642
    },
    {
      "epoch": 0.5748010245390338,
      "grad_norm": 0.17900599539279938,
      "learning_rate": 0.0007385077793493635,
      "loss": 2.352895975112915,
      "step": 5644
    },
    {
      "epoch": 0.5750047102316416,
      "grad_norm": 0.13084252178668976,
      "learning_rate": 0.0007381541725601132,
      "loss": 2.2882752418518066,
      "step": 5646
    },
    {
      "epoch": 0.5752083959242493,
      "grad_norm": 0.12346002459526062,
      "learning_rate": 0.0007378005657708628,
      "loss": 2.0930776596069336,
      "step": 5648
    },
    {
      "epoch": 0.575412081616857,
      "grad_norm": 0.14437732100486755,
      "learning_rate": 0.0007374469589816125,
      "loss": 2.0397298336029053,
      "step": 5650
    },
    {
      "epoch": 0.5756157673094647,
      "grad_norm": 0.13602668046951294,
      "learning_rate": 0.000737093352192362,
      "loss": 2.111401319503784,
      "step": 5652
    },
    {
      "epoch": 0.5758194530020725,
      "grad_norm": 0.11576951295137405,
      "learning_rate": 0.0007367397454031118,
      "loss": 2.060234546661377,
      "step": 5654
    },
    {
      "epoch": 0.5760231386946802,
      "grad_norm": 0.14814341068267822,
      "learning_rate": 0.0007363861386138614,
      "loss": 2.2040693759918213,
      "step": 5656
    },
    {
      "epoch": 0.576226824387288,
      "grad_norm": 0.13323092460632324,
      "learning_rate": 0.000736032531824611,
      "loss": 2.215704917907715,
      "step": 5658
    },
    {
      "epoch": 0.5764305100798958,
      "grad_norm": 0.1304558962583542,
      "learning_rate": 0.0007356789250353607,
      "loss": 2.0584301948547363,
      "step": 5660
    },
    {
      "epoch": 0.5766341957725034,
      "grad_norm": 0.11538539081811905,
      "learning_rate": 0.0007353253182461103,
      "loss": 2.047435998916626,
      "step": 5662
    },
    {
      "epoch": 0.5768378814651112,
      "grad_norm": 0.11744815856218338,
      "learning_rate": 0.0007349717114568601,
      "loss": 2.000859022140503,
      "step": 5664
    },
    {
      "epoch": 0.5770415671577189,
      "grad_norm": 0.13838565349578857,
      "learning_rate": 0.0007346181046676096,
      "loss": 2.26808762550354,
      "step": 5666
    },
    {
      "epoch": 0.5772452528503267,
      "grad_norm": 0.12291118502616882,
      "learning_rate": 0.0007342644978783593,
      "loss": 2.219853162765503,
      "step": 5668
    },
    {
      "epoch": 0.5774489385429344,
      "grad_norm": 0.11459051072597504,
      "learning_rate": 0.0007339108910891089,
      "loss": 2.30008602142334,
      "step": 5670
    },
    {
      "epoch": 0.5776526242355421,
      "grad_norm": 0.11127769947052002,
      "learning_rate": 0.0007335572842998585,
      "loss": 2.097968578338623,
      "step": 5672
    },
    {
      "epoch": 0.5778563099281498,
      "grad_norm": 0.1433795541524887,
      "learning_rate": 0.0007332036775106083,
      "loss": 2.275020122528076,
      "step": 5674
    },
    {
      "epoch": 0.5780599956207576,
      "grad_norm": 0.11843910068273544,
      "learning_rate": 0.0007328500707213579,
      "loss": 2.136817216873169,
      "step": 5676
    },
    {
      "epoch": 0.5782636813133654,
      "grad_norm": 0.12363915145397186,
      "learning_rate": 0.0007324964639321076,
      "loss": 2.180633544921875,
      "step": 5678
    },
    {
      "epoch": 0.5784673670059731,
      "grad_norm": 0.1321050524711609,
      "learning_rate": 0.0007321428571428571,
      "loss": 2.256535530090332,
      "step": 5680
    },
    {
      "epoch": 0.5786710526985808,
      "grad_norm": 0.10442063957452774,
      "learning_rate": 0.0007317892503536068,
      "loss": 2.2629504203796387,
      "step": 5682
    },
    {
      "epoch": 0.5788747383911885,
      "grad_norm": 0.11953689903020859,
      "learning_rate": 0.0007314356435643564,
      "loss": 1.976685881614685,
      "step": 5684
    },
    {
      "epoch": 0.5790784240837963,
      "grad_norm": 0.1100454106926918,
      "learning_rate": 0.000731082036775106,
      "loss": 2.048116445541382,
      "step": 5686
    },
    {
      "epoch": 0.579282109776404,
      "grad_norm": 0.10882994532585144,
      "learning_rate": 0.0007307284299858558,
      "loss": 2.152884006500244,
      "step": 5688
    },
    {
      "epoch": 0.5794857954690118,
      "grad_norm": 0.1338670253753662,
      "learning_rate": 0.0007303748231966054,
      "loss": 2.0614571571350098,
      "step": 5690
    },
    {
      "epoch": 0.5796894811616196,
      "grad_norm": 0.12294940650463104,
      "learning_rate": 0.0007300212164073551,
      "loss": 2.2152326107025146,
      "step": 5692
    },
    {
      "epoch": 0.5798931668542272,
      "grad_norm": 0.12312138825654984,
      "learning_rate": 0.0007296676096181046,
      "loss": 2.198913097381592,
      "step": 5694
    },
    {
      "epoch": 0.580096852546835,
      "grad_norm": 0.0906839445233345,
      "learning_rate": 0.0007293140028288543,
      "loss": 2.0192747116088867,
      "step": 5696
    },
    {
      "epoch": 0.5803005382394427,
      "grad_norm": 0.11505401879549026,
      "learning_rate": 0.000728960396039604,
      "loss": 2.101466178894043,
      "step": 5698
    },
    {
      "epoch": 0.5805042239320505,
      "grad_norm": 0.10482259094715118,
      "learning_rate": 0.0007286067892503536,
      "loss": 2.1325016021728516,
      "step": 5700
    },
    {
      "epoch": 0.5807079096246582,
      "grad_norm": 0.1126074492931366,
      "learning_rate": 0.0007282531824611033,
      "loss": 2.028740167617798,
      "step": 5702
    },
    {
      "epoch": 0.5809115953172659,
      "grad_norm": 0.12492992728948593,
      "learning_rate": 0.0007278995756718529,
      "loss": 2.2029075622558594,
      "step": 5704
    },
    {
      "epoch": 0.5811152810098736,
      "grad_norm": 0.1306290626525879,
      "learning_rate": 0.0007275459688826026,
      "loss": 2.2290592193603516,
      "step": 5706
    },
    {
      "epoch": 0.5813189667024814,
      "grad_norm": 0.11664531379938126,
      "learning_rate": 0.0007271923620933521,
      "loss": 2.061734199523926,
      "step": 5708
    },
    {
      "epoch": 0.5815226523950892,
      "grad_norm": 0.12610818445682526,
      "learning_rate": 0.0007268387553041019,
      "loss": 2.245943069458008,
      "step": 5710
    },
    {
      "epoch": 0.5817263380876969,
      "grad_norm": 0.1164708286523819,
      "learning_rate": 0.0007264851485148515,
      "loss": 2.25504732131958,
      "step": 5712
    },
    {
      "epoch": 0.5819300237803046,
      "grad_norm": 0.10686644911766052,
      "learning_rate": 0.0007261315417256011,
      "loss": 2.2719407081604004,
      "step": 5714
    },
    {
      "epoch": 0.5821337094729123,
      "grad_norm": 0.15834969282150269,
      "learning_rate": 0.0007257779349363508,
      "loss": 2.2463080883026123,
      "step": 5716
    },
    {
      "epoch": 0.5823373951655201,
      "grad_norm": 0.1196233406662941,
      "learning_rate": 0.0007254243281471004,
      "loss": 2.190795421600342,
      "step": 5718
    },
    {
      "epoch": 0.5825410808581278,
      "grad_norm": 0.1104876697063446,
      "learning_rate": 0.0007250707213578502,
      "loss": 2.109438896179199,
      "step": 5720
    },
    {
      "epoch": 0.5827447665507356,
      "grad_norm": 0.11524586379528046,
      "learning_rate": 0.0007247171145685997,
      "loss": 2.051980972290039,
      "step": 5722
    },
    {
      "epoch": 0.5829484522433434,
      "grad_norm": 0.12571680545806885,
      "learning_rate": 0.0007243635077793494,
      "loss": 2.22143816947937,
      "step": 5724
    },
    {
      "epoch": 0.583152137935951,
      "grad_norm": 0.16392545402050018,
      "learning_rate": 0.000724009900990099,
      "loss": 2.080940008163452,
      "step": 5726
    },
    {
      "epoch": 0.5833558236285588,
      "grad_norm": 0.12662646174430847,
      "learning_rate": 0.0007236562942008486,
      "loss": 2.263650417327881,
      "step": 5728
    },
    {
      "epoch": 0.5835595093211665,
      "grad_norm": 0.10644856840372086,
      "learning_rate": 0.0007233026874115983,
      "loss": 2.286713123321533,
      "step": 5730
    },
    {
      "epoch": 0.5837631950137743,
      "grad_norm": 0.11860610544681549,
      "learning_rate": 0.000722949080622348,
      "loss": 2.0691041946411133,
      "step": 5732
    },
    {
      "epoch": 0.583966880706382,
      "grad_norm": 0.1357344090938568,
      "learning_rate": 0.0007225954738330977,
      "loss": 2.115696430206299,
      "step": 5734
    },
    {
      "epoch": 0.5841705663989897,
      "grad_norm": 0.10321585088968277,
      "learning_rate": 0.0007222418670438473,
      "loss": 2.1780788898468018,
      "step": 5736
    },
    {
      "epoch": 0.5843742520915974,
      "grad_norm": 0.14070770144462585,
      "learning_rate": 0.0007218882602545969,
      "loss": 2.0064010620117188,
      "step": 5738
    },
    {
      "epoch": 0.5845779377842052,
      "grad_norm": 0.12550899386405945,
      "learning_rate": 0.0007215346534653465,
      "loss": 2.2525646686553955,
      "step": 5740
    },
    {
      "epoch": 0.584781623476813,
      "grad_norm": 0.1286008208990097,
      "learning_rate": 0.0007211810466760961,
      "loss": 2.199039936065674,
      "step": 5742
    },
    {
      "epoch": 0.5849853091694207,
      "grad_norm": 0.11532837152481079,
      "learning_rate": 0.0007208274398868459,
      "loss": 2.1299610137939453,
      "step": 5744
    },
    {
      "epoch": 0.5851889948620284,
      "grad_norm": 0.13768477737903595,
      "learning_rate": 0.0007204738330975955,
      "loss": 2.1538450717926025,
      "step": 5746
    },
    {
      "epoch": 0.5853926805546361,
      "grad_norm": 0.13155896961688995,
      "learning_rate": 0.0007201202263083452,
      "loss": 2.1262364387512207,
      "step": 5748
    },
    {
      "epoch": 0.5855963662472439,
      "grad_norm": 0.12181179970502853,
      "learning_rate": 0.0007197666195190948,
      "loss": 2.0381970405578613,
      "step": 5750
    },
    {
      "epoch": 0.5858000519398516,
      "grad_norm": 0.14482003450393677,
      "learning_rate": 0.0007194130127298444,
      "loss": 2.161560535430908,
      "step": 5752
    },
    {
      "epoch": 0.5860037376324594,
      "grad_norm": 0.1338665783405304,
      "learning_rate": 0.000719059405940594,
      "loss": 2.2674694061279297,
      "step": 5754
    },
    {
      "epoch": 0.586207423325067,
      "grad_norm": 0.12887023389339447,
      "learning_rate": 0.0007187057991513437,
      "loss": 2.1345934867858887,
      "step": 5756
    },
    {
      "epoch": 0.5864111090176748,
      "grad_norm": 0.12575219571590424,
      "learning_rate": 0.0007183521923620934,
      "loss": 2.227172374725342,
      "step": 5758
    },
    {
      "epoch": 0.5866147947102825,
      "grad_norm": 0.12753604352474213,
      "learning_rate": 0.000717998585572843,
      "loss": 2.2868831157684326,
      "step": 5760
    },
    {
      "epoch": 0.5868184804028903,
      "grad_norm": 0.16653795540332794,
      "learning_rate": 0.0007176449787835927,
      "loss": 2.128230571746826,
      "step": 5762
    },
    {
      "epoch": 0.5870221660954981,
      "grad_norm": 0.14026276767253876,
      "learning_rate": 0.0007172913719943423,
      "loss": 2.1803252696990967,
      "step": 5764
    },
    {
      "epoch": 0.5872258517881058,
      "grad_norm": 0.14044813811779022,
      "learning_rate": 0.000716937765205092,
      "loss": 2.224703073501587,
      "step": 5766
    },
    {
      "epoch": 0.5874295374807135,
      "grad_norm": 0.12035451084375381,
      "learning_rate": 0.0007165841584158416,
      "loss": 2.173762321472168,
      "step": 5768
    },
    {
      "epoch": 0.5876332231733212,
      "grad_norm": 0.15759555995464325,
      "learning_rate": 0.0007162305516265912,
      "loss": 2.1230368614196777,
      "step": 5770
    },
    {
      "epoch": 0.587836908865929,
      "grad_norm": 0.14997094869613647,
      "learning_rate": 0.0007158769448373409,
      "loss": 2.2385635375976562,
      "step": 5772
    },
    {
      "epoch": 0.5880405945585367,
      "grad_norm": 0.12262008339166641,
      "learning_rate": 0.0007155233380480905,
      "loss": 2.1380510330200195,
      "step": 5774
    },
    {
      "epoch": 0.5882442802511445,
      "grad_norm": 0.1245741918683052,
      "learning_rate": 0.0007151697312588402,
      "loss": 2.101010322570801,
      "step": 5776
    },
    {
      "epoch": 0.5884479659437521,
      "grad_norm": 0.12575244903564453,
      "learning_rate": 0.0007148161244695899,
      "loss": 2.227003335952759,
      "step": 5778
    },
    {
      "epoch": 0.5886516516363599,
      "grad_norm": 0.12538304924964905,
      "learning_rate": 0.0007144625176803395,
      "loss": 2.067056655883789,
      "step": 5780
    },
    {
      "epoch": 0.5888553373289677,
      "grad_norm": 0.12415127456188202,
      "learning_rate": 0.0007141089108910891,
      "loss": 2.156764030456543,
      "step": 5782
    },
    {
      "epoch": 0.5890590230215754,
      "grad_norm": 0.10758251696825027,
      "learning_rate": 0.0007137553041018387,
      "loss": 2.140799045562744,
      "step": 5784
    },
    {
      "epoch": 0.5892627087141832,
      "grad_norm": 0.11932411789894104,
      "learning_rate": 0.0007134016973125884,
      "loss": 2.259995460510254,
      "step": 5786
    },
    {
      "epoch": 0.5894663944067908,
      "grad_norm": 0.11605558544397354,
      "learning_rate": 0.000713048090523338,
      "loss": 2.1587512493133545,
      "step": 5788
    },
    {
      "epoch": 0.5896700800993986,
      "grad_norm": 0.10753548890352249,
      "learning_rate": 0.0007126944837340878,
      "loss": 2.131270408630371,
      "step": 5790
    },
    {
      "epoch": 0.5898737657920063,
      "grad_norm": 0.1256030797958374,
      "learning_rate": 0.0007123408769448374,
      "loss": 2.087528705596924,
      "step": 5792
    },
    {
      "epoch": 0.5900774514846141,
      "grad_norm": 0.10056347399950027,
      "learning_rate": 0.000711987270155587,
      "loss": 2.1330227851867676,
      "step": 5794
    },
    {
      "epoch": 0.5902811371772219,
      "grad_norm": 0.12868225574493408,
      "learning_rate": 0.0007116336633663366,
      "loss": 2.1285815238952637,
      "step": 5796
    },
    {
      "epoch": 0.5904848228698296,
      "grad_norm": 0.14323009550571442,
      "learning_rate": 0.0007112800565770862,
      "loss": 2.1502928733825684,
      "step": 5798
    },
    {
      "epoch": 0.5906885085624373,
      "grad_norm": 0.14026352763175964,
      "learning_rate": 0.000710926449787836,
      "loss": 2.337285280227661,
      "step": 5800
    },
    {
      "epoch": 0.590892194255045,
      "grad_norm": 0.13254085183143616,
      "learning_rate": 0.0007105728429985856,
      "loss": 2.258711814880371,
      "step": 5802
    },
    {
      "epoch": 0.5910958799476528,
      "grad_norm": 0.12581101059913635,
      "learning_rate": 0.0007102192362093353,
      "loss": 2.1116037368774414,
      "step": 5804
    },
    {
      "epoch": 0.5912995656402605,
      "grad_norm": 0.1658536046743393,
      "learning_rate": 0.0007098656294200849,
      "loss": 2.2243106365203857,
      "step": 5806
    },
    {
      "epoch": 0.5915032513328683,
      "grad_norm": 0.15318162739276886,
      "learning_rate": 0.0007095120226308345,
      "loss": 2.2848992347717285,
      "step": 5808
    },
    {
      "epoch": 0.5917069370254759,
      "grad_norm": 0.11438168585300446,
      "learning_rate": 0.0007091584158415841,
      "loss": 2.032163381576538,
      "step": 5810
    },
    {
      "epoch": 0.5919106227180837,
      "grad_norm": 0.15281060338020325,
      "learning_rate": 0.0007088048090523338,
      "loss": 2.310468912124634,
      "step": 5812
    },
    {
      "epoch": 0.5921143084106915,
      "grad_norm": 0.12696152925491333,
      "learning_rate": 0.0007084512022630835,
      "loss": 2.1293540000915527,
      "step": 5814
    },
    {
      "epoch": 0.5923179941032992,
      "grad_norm": 0.11996600031852722,
      "learning_rate": 0.0007080975954738331,
      "loss": 2.146444082260132,
      "step": 5816
    },
    {
      "epoch": 0.592521679795907,
      "grad_norm": 0.12982875108718872,
      "learning_rate": 0.0007077439886845828,
      "loss": 2.2628679275512695,
      "step": 5818
    },
    {
      "epoch": 0.5927253654885146,
      "grad_norm": 0.13845832645893097,
      "learning_rate": 0.0007073903818953324,
      "loss": 2.2842679023742676,
      "step": 5820
    },
    {
      "epoch": 0.5929290511811224,
      "grad_norm": 0.12359467893838882,
      "learning_rate": 0.000707036775106082,
      "loss": 2.1193108558654785,
      "step": 5822
    },
    {
      "epoch": 0.5931327368737301,
      "grad_norm": 0.1370133012533188,
      "learning_rate": 0.0007066831683168317,
      "loss": 2.043365716934204,
      "step": 5824
    },
    {
      "epoch": 0.5933364225663379,
      "grad_norm": 0.13645340502262115,
      "learning_rate": 0.0007063295615275813,
      "loss": 2.2149620056152344,
      "step": 5826
    },
    {
      "epoch": 0.5935401082589457,
      "grad_norm": 0.13821366429328918,
      "learning_rate": 0.000705975954738331,
      "loss": 2.1222686767578125,
      "step": 5828
    },
    {
      "epoch": 0.5937437939515534,
      "grad_norm": 0.11535792052745819,
      "learning_rate": 0.0007056223479490806,
      "loss": 2.037485361099243,
      "step": 5830
    },
    {
      "epoch": 0.5939474796441611,
      "grad_norm": 0.13120539486408234,
      "learning_rate": 0.0007052687411598303,
      "loss": 2.226588249206543,
      "step": 5832
    },
    {
      "epoch": 0.5941511653367688,
      "grad_norm": 0.12012495845556259,
      "learning_rate": 0.00070491513437058,
      "loss": 2.239220142364502,
      "step": 5834
    },
    {
      "epoch": 0.5943548510293766,
      "grad_norm": 0.14491599798202515,
      "learning_rate": 0.0007045615275813297,
      "loss": 2.215017795562744,
      "step": 5836
    },
    {
      "epoch": 0.5945585367219843,
      "grad_norm": 0.1036931574344635,
      "learning_rate": 0.0007042079207920792,
      "loss": 2.057384967803955,
      "step": 5838
    },
    {
      "epoch": 0.5947622224145921,
      "grad_norm": 0.13469237089157104,
      "learning_rate": 0.0007038543140028288,
      "loss": 2.252002716064453,
      "step": 5840
    },
    {
      "epoch": 0.5949659081071997,
      "grad_norm": 0.12333661317825317,
      "learning_rate": 0.0007035007072135785,
      "loss": 2.129938840866089,
      "step": 5842
    },
    {
      "epoch": 0.5951695937998075,
      "grad_norm": 0.1126076802611351,
      "learning_rate": 0.0007031471004243281,
      "loss": 2.2039663791656494,
      "step": 5844
    },
    {
      "epoch": 0.5953732794924153,
      "grad_norm": 0.14256829023361206,
      "learning_rate": 0.0007027934936350779,
      "loss": 2.1440048217773438,
      "step": 5846
    },
    {
      "epoch": 0.595576965185023,
      "grad_norm": 0.154927060008049,
      "learning_rate": 0.0007024398868458275,
      "loss": 2.3388547897338867,
      "step": 5848
    },
    {
      "epoch": 0.5957806508776308,
      "grad_norm": 0.12084995210170746,
      "learning_rate": 0.0007020862800565772,
      "loss": 2.116058349609375,
      "step": 5850
    },
    {
      "epoch": 0.5959843365702384,
      "grad_norm": 0.17354971170425415,
      "learning_rate": 0.0007017326732673267,
      "loss": 2.0893397331237793,
      "step": 5852
    },
    {
      "epoch": 0.5961880222628462,
      "grad_norm": 0.1491243690252304,
      "learning_rate": 0.0007013790664780763,
      "loss": 2.2560250759124756,
      "step": 5854
    },
    {
      "epoch": 0.5963917079554539,
      "grad_norm": 0.11963433027267456,
      "learning_rate": 0.000701025459688826,
      "loss": 2.1822943687438965,
      "step": 5856
    },
    {
      "epoch": 0.5965953936480617,
      "grad_norm": 0.18143661320209503,
      "learning_rate": 0.0007006718528995757,
      "loss": 2.164369583129883,
      "step": 5858
    },
    {
      "epoch": 0.5967990793406694,
      "grad_norm": 0.13938595354557037,
      "learning_rate": 0.0007003182461103254,
      "loss": 2.2488672733306885,
      "step": 5860
    },
    {
      "epoch": 0.5970027650332772,
      "grad_norm": 0.13489268720149994,
      "learning_rate": 0.000699964639321075,
      "loss": 2.2433359622955322,
      "step": 5862
    },
    {
      "epoch": 0.5972064507258849,
      "grad_norm": 0.12128165364265442,
      "learning_rate": 0.0006996110325318247,
      "loss": 2.2040345668792725,
      "step": 5864
    },
    {
      "epoch": 0.5974101364184926,
      "grad_norm": 0.1320391446352005,
      "learning_rate": 0.0006992574257425742,
      "loss": 2.2673254013061523,
      "step": 5866
    },
    {
      "epoch": 0.5976138221111004,
      "grad_norm": 0.1891942322254181,
      "learning_rate": 0.0006989038189533238,
      "loss": 2.2457022666931152,
      "step": 5868
    },
    {
      "epoch": 0.5978175078037081,
      "grad_norm": 0.12701493501663208,
      "learning_rate": 0.0006985502121640736,
      "loss": 2.2249765396118164,
      "step": 5870
    },
    {
      "epoch": 0.5980211934963159,
      "grad_norm": 0.12259265035390854,
      "learning_rate": 0.0006981966053748232,
      "loss": 2.2566633224487305,
      "step": 5872
    },
    {
      "epoch": 0.5982248791889235,
      "grad_norm": 0.12629735469818115,
      "learning_rate": 0.0006978429985855729,
      "loss": 2.1290676593780518,
      "step": 5874
    },
    {
      "epoch": 0.5984285648815313,
      "grad_norm": 0.11066842079162598,
      "learning_rate": 0.0006974893917963225,
      "loss": 2.248239278793335,
      "step": 5876
    },
    {
      "epoch": 0.598632250574139,
      "grad_norm": 0.10732977837324142,
      "learning_rate": 0.0006971357850070722,
      "loss": 2.036114454269409,
      "step": 5878
    },
    {
      "epoch": 0.5988359362667468,
      "grad_norm": 0.12938185036182404,
      "learning_rate": 0.0006967821782178218,
      "loss": 2.182767868041992,
      "step": 5880
    },
    {
      "epoch": 0.5990396219593546,
      "grad_norm": 0.11534034460783005,
      "learning_rate": 0.0006964285714285714,
      "loss": 2.2511093616485596,
      "step": 5882
    },
    {
      "epoch": 0.5992433076519622,
      "grad_norm": 0.12999317049980164,
      "learning_rate": 0.0006960749646393211,
      "loss": 2.2557802200317383,
      "step": 5884
    },
    {
      "epoch": 0.59944699334457,
      "grad_norm": 0.1293525993824005,
      "learning_rate": 0.0006957213578500707,
      "loss": 2.1462244987487793,
      "step": 5886
    },
    {
      "epoch": 0.5996506790371777,
      "grad_norm": 0.10800858587026596,
      "learning_rate": 0.0006953677510608204,
      "loss": 2.091259479522705,
      "step": 5888
    },
    {
      "epoch": 0.5998543647297855,
      "grad_norm": 0.11408194154500961,
      "learning_rate": 0.00069501414427157,
      "loss": 2.2213294506073,
      "step": 5890
    },
    {
      "epoch": 0.6000580504223932,
      "grad_norm": 0.19715900719165802,
      "learning_rate": 0.0006946605374823198,
      "loss": 2.130643367767334,
      "step": 5892
    },
    {
      "epoch": 0.6002617361150009,
      "grad_norm": 0.12957103550434113,
      "learning_rate": 0.0006943069306930693,
      "loss": 2.2938973903656006,
      "step": 5894
    },
    {
      "epoch": 0.6004654218076086,
      "grad_norm": 0.11463894695043564,
      "learning_rate": 0.0006939533239038189,
      "loss": 2.1532740592956543,
      "step": 5896
    },
    {
      "epoch": 0.6006691075002164,
      "grad_norm": 0.14020101726055145,
      "learning_rate": 0.0006935997171145686,
      "loss": 2.0930893421173096,
      "step": 5898
    },
    {
      "epoch": 0.6008727931928242,
      "grad_norm": 0.12350431084632874,
      "learning_rate": 0.0006932461103253182,
      "loss": 2.2487239837646484,
      "step": 5900
    },
    {
      "epoch": 0.6010764788854319,
      "grad_norm": 0.10158339887857437,
      "learning_rate": 0.000692892503536068,
      "loss": 1.998267650604248,
      "step": 5902
    },
    {
      "epoch": 0.6012801645780397,
      "grad_norm": 0.12207748740911484,
      "learning_rate": 0.0006925388967468176,
      "loss": 2.177238702774048,
      "step": 5904
    },
    {
      "epoch": 0.6014838502706473,
      "grad_norm": 0.10963020473718643,
      "learning_rate": 0.0006921852899575673,
      "loss": 2.1312599182128906,
      "step": 5906
    },
    {
      "epoch": 0.6016875359632551,
      "grad_norm": 0.12284676730632782,
      "learning_rate": 0.0006918316831683168,
      "loss": 2.2571730613708496,
      "step": 5908
    },
    {
      "epoch": 0.6018912216558628,
      "grad_norm": 0.11777108907699585,
      "learning_rate": 0.0006914780763790664,
      "loss": 1.9476723670959473,
      "step": 5910
    },
    {
      "epoch": 0.6020949073484706,
      "grad_norm": 0.11942917108535767,
      "learning_rate": 0.0006911244695898161,
      "loss": 2.2475385665893555,
      "step": 5912
    },
    {
      "epoch": 0.6022985930410784,
      "grad_norm": 0.13617755472660065,
      "learning_rate": 0.0006907708628005657,
      "loss": 2.3211159706115723,
      "step": 5914
    },
    {
      "epoch": 0.602502278733686,
      "grad_norm": 0.13237358629703522,
      "learning_rate": 0.0006904172560113155,
      "loss": 2.229127883911133,
      "step": 5916
    },
    {
      "epoch": 0.6027059644262938,
      "grad_norm": 0.12165000289678574,
      "learning_rate": 0.0006900636492220651,
      "loss": 2.162569522857666,
      "step": 5918
    },
    {
      "epoch": 0.6029096501189015,
      "grad_norm": 0.16229882836341858,
      "learning_rate": 0.0006897100424328148,
      "loss": 2.162285566329956,
      "step": 5920
    },
    {
      "epoch": 0.6031133358115093,
      "grad_norm": 0.17291904985904694,
      "learning_rate": 0.0006893564356435643,
      "loss": 2.307204246520996,
      "step": 5922
    },
    {
      "epoch": 0.603317021504117,
      "grad_norm": 0.11836862564086914,
      "learning_rate": 0.0006890028288543139,
      "loss": 2.1268110275268555,
      "step": 5924
    },
    {
      "epoch": 0.6035207071967247,
      "grad_norm": 0.10857526957988739,
      "learning_rate": 0.0006886492220650637,
      "loss": 2.0886805057525635,
      "step": 5926
    },
    {
      "epoch": 0.6037243928893324,
      "grad_norm": 0.13450521230697632,
      "learning_rate": 0.0006882956152758133,
      "loss": 2.2853922843933105,
      "step": 5928
    },
    {
      "epoch": 0.6039280785819402,
      "grad_norm": 0.11961854994297028,
      "learning_rate": 0.000687942008486563,
      "loss": 2.207688093185425,
      "step": 5930
    },
    {
      "epoch": 0.604131764274548,
      "grad_norm": 0.11995051056146622,
      "learning_rate": 0.0006875884016973126,
      "loss": 2.139657497406006,
      "step": 5932
    },
    {
      "epoch": 0.6043354499671557,
      "grad_norm": 0.12403984367847443,
      "learning_rate": 0.0006872347949080623,
      "loss": 2.2460837364196777,
      "step": 5934
    },
    {
      "epoch": 0.6045391356597635,
      "grad_norm": 0.13402248919010162,
      "learning_rate": 0.000686881188118812,
      "loss": 2.254303455352783,
      "step": 5936
    },
    {
      "epoch": 0.6047428213523711,
      "grad_norm": 0.12209654599428177,
      "learning_rate": 0.0006865275813295615,
      "loss": 2.119863510131836,
      "step": 5938
    },
    {
      "epoch": 0.6049465070449789,
      "grad_norm": 0.13662000000476837,
      "learning_rate": 0.0006861739745403112,
      "loss": 2.200282335281372,
      "step": 5940
    },
    {
      "epoch": 0.6051501927375866,
      "grad_norm": 0.15376532077789307,
      "learning_rate": 0.0006858203677510608,
      "loss": 2.0868234634399414,
      "step": 5942
    },
    {
      "epoch": 0.6053538784301944,
      "grad_norm": 0.12053542584180832,
      "learning_rate": 0.0006854667609618105,
      "loss": 2.2881765365600586,
      "step": 5944
    },
    {
      "epoch": 0.6055575641228021,
      "grad_norm": 0.11602432280778885,
      "learning_rate": 0.0006851131541725601,
      "loss": 2.0897693634033203,
      "step": 5946
    },
    {
      "epoch": 0.6057612498154098,
      "grad_norm": 0.15823666751384735,
      "learning_rate": 0.0006847595473833099,
      "loss": 2.1809895038604736,
      "step": 5948
    },
    {
      "epoch": 0.6059649355080176,
      "grad_norm": 0.10821446031332016,
      "learning_rate": 0.0006844059405940595,
      "loss": 2.132140874862671,
      "step": 5950
    },
    {
      "epoch": 0.6061686212006253,
      "grad_norm": 0.1111520379781723,
      "learning_rate": 0.000684052333804809,
      "loss": 2.123976230621338,
      "step": 5952
    },
    {
      "epoch": 0.6063723068932331,
      "grad_norm": 0.12216538935899734,
      "learning_rate": 0.0006836987270155587,
      "loss": 2.2142128944396973,
      "step": 5954
    },
    {
      "epoch": 0.6065759925858408,
      "grad_norm": 0.11304319649934769,
      "learning_rate": 0.0006833451202263083,
      "loss": 2.1550352573394775,
      "step": 5956
    },
    {
      "epoch": 0.6067796782784485,
      "grad_norm": 0.12120669335126877,
      "learning_rate": 0.000682991513437058,
      "loss": 2.205153226852417,
      "step": 5958
    },
    {
      "epoch": 0.6069833639710562,
      "grad_norm": 0.15847378969192505,
      "learning_rate": 0.0006826379066478077,
      "loss": 2.2370009422302246,
      "step": 5960
    },
    {
      "epoch": 0.607187049663664,
      "grad_norm": 0.11909697949886322,
      "learning_rate": 0.0006822842998585574,
      "loss": 2.1927297115325928,
      "step": 5962
    },
    {
      "epoch": 0.6073907353562717,
      "grad_norm": 0.12275227159261703,
      "learning_rate": 0.000681930693069307,
      "loss": 2.1182925701141357,
      "step": 5964
    },
    {
      "epoch": 0.6075944210488795,
      "grad_norm": 0.1329088807106018,
      "learning_rate": 0.0006815770862800565,
      "loss": 2.243060350418091,
      "step": 5966
    },
    {
      "epoch": 0.6077981067414873,
      "grad_norm": 0.12506160140037537,
      "learning_rate": 0.0006812234794908062,
      "loss": 2.2525110244750977,
      "step": 5968
    },
    {
      "epoch": 0.6080017924340949,
      "grad_norm": 0.11776109784841537,
      "learning_rate": 0.0006808698727015558,
      "loss": 2.143366575241089,
      "step": 5970
    },
    {
      "epoch": 0.6082054781267027,
      "grad_norm": 0.13655391335487366,
      "learning_rate": 0.0006805162659123056,
      "loss": 2.2589454650878906,
      "step": 5972
    },
    {
      "epoch": 0.6084091638193104,
      "grad_norm": 0.12635111808776855,
      "learning_rate": 0.0006801626591230552,
      "loss": 2.260399580001831,
      "step": 5974
    },
    {
      "epoch": 0.6086128495119182,
      "grad_norm": 0.12260592728853226,
      "learning_rate": 0.0006798090523338049,
      "loss": 2.210442543029785,
      "step": 5976
    },
    {
      "epoch": 0.6088165352045259,
      "grad_norm": 0.1142953634262085,
      "learning_rate": 0.0006794554455445545,
      "loss": 2.1625008583068848,
      "step": 5978
    },
    {
      "epoch": 0.6090202208971336,
      "grad_norm": 0.10492588579654694,
      "learning_rate": 0.000679101838755304,
      "loss": 2.157905340194702,
      "step": 5980
    },
    {
      "epoch": 0.6092239065897413,
      "grad_norm": 0.10436489433050156,
      "learning_rate": 0.0006787482319660537,
      "loss": 2.1605606079101562,
      "step": 5982
    },
    {
      "epoch": 0.6094275922823491,
      "grad_norm": 0.11659426242113113,
      "learning_rate": 0.0006783946251768034,
      "loss": 2.173847198486328,
      "step": 5984
    },
    {
      "epoch": 0.6096312779749569,
      "grad_norm": 0.1403086632490158,
      "learning_rate": 0.0006780410183875531,
      "loss": 2.0930469036102295,
      "step": 5986
    },
    {
      "epoch": 0.6098349636675646,
      "grad_norm": 0.10844416171312332,
      "learning_rate": 0.0006776874115983027,
      "loss": 2.1215381622314453,
      "step": 5988
    },
    {
      "epoch": 0.6100386493601723,
      "grad_norm": 0.1294063776731491,
      "learning_rate": 0.0006773338048090524,
      "loss": 2.2616848945617676,
      "step": 5990
    },
    {
      "epoch": 0.61024233505278,
      "grad_norm": 0.1429341733455658,
      "learning_rate": 0.000676980198019802,
      "loss": 2.164083480834961,
      "step": 5992
    },
    {
      "epoch": 0.6104460207453878,
      "grad_norm": 0.11299315840005875,
      "learning_rate": 0.0006766265912305515,
      "loss": 2.200873374938965,
      "step": 5994
    },
    {
      "epoch": 0.6106497064379955,
      "grad_norm": 0.1586959809064865,
      "learning_rate": 0.0006762729844413013,
      "loss": 2.1433215141296387,
      "step": 5996
    },
    {
      "epoch": 0.6108533921306033,
      "grad_norm": 0.16817247867584229,
      "learning_rate": 0.0006759193776520509,
      "loss": 2.236090898513794,
      "step": 5998
    },
    {
      "epoch": 0.611057077823211,
      "grad_norm": 0.10995350033044815,
      "learning_rate": 0.0006755657708628006,
      "loss": 2.262341022491455,
      "step": 6000
    },
    {
      "epoch": 0.6112607635158187,
      "grad_norm": 0.1350107640028,
      "learning_rate": 0.0006752121640735502,
      "loss": 2.158784866333008,
      "step": 6002
    },
    {
      "epoch": 0.6114644492084265,
      "grad_norm": 0.12271104753017426,
      "learning_rate": 0.0006748585572843,
      "loss": 2.1531057357788086,
      "step": 6004
    },
    {
      "epoch": 0.6116681349010342,
      "grad_norm": 0.13481822609901428,
      "learning_rate": 0.0006745049504950496,
      "loss": 2.2012364864349365,
      "step": 6006
    },
    {
      "epoch": 0.611871820593642,
      "grad_norm": 0.12433091551065445,
      "learning_rate": 0.0006741513437057991,
      "loss": 2.147671699523926,
      "step": 6008
    },
    {
      "epoch": 0.6120755062862497,
      "grad_norm": 0.12741665542125702,
      "learning_rate": 0.0006737977369165488,
      "loss": 2.1015470027923584,
      "step": 6010
    },
    {
      "epoch": 0.6122791919788574,
      "grad_norm": 0.11500316113233566,
      "learning_rate": 0.0006734441301272984,
      "loss": 2.1692709922790527,
      "step": 6012
    },
    {
      "epoch": 0.6124828776714651,
      "grad_norm": 0.2202855795621872,
      "learning_rate": 0.0006730905233380481,
      "loss": 2.339313507080078,
      "step": 6014
    },
    {
      "epoch": 0.6126865633640729,
      "grad_norm": 0.12213063985109329,
      "learning_rate": 0.0006727369165487977,
      "loss": 2.2237448692321777,
      "step": 6016
    },
    {
      "epoch": 0.6128902490566807,
      "grad_norm": 0.11124632507562637,
      "learning_rate": 0.0006723833097595475,
      "loss": 2.347067356109619,
      "step": 6018
    },
    {
      "epoch": 0.6130939347492884,
      "grad_norm": 0.1313537210226059,
      "learning_rate": 0.0006720297029702971,
      "loss": 2.1775741577148438,
      "step": 6020
    },
    {
      "epoch": 0.6132976204418961,
      "grad_norm": 0.10514973849058151,
      "learning_rate": 0.0006716760961810466,
      "loss": 2.1862263679504395,
      "step": 6022
    },
    {
      "epoch": 0.6135013061345038,
      "grad_norm": 0.13106107711791992,
      "learning_rate": 0.0006713224893917963,
      "loss": 2.262377977371216,
      "step": 6024
    },
    {
      "epoch": 0.6137049918271116,
      "grad_norm": 0.12000846117734909,
      "learning_rate": 0.0006709688826025459,
      "loss": 2.1819043159484863,
      "step": 6026
    },
    {
      "epoch": 0.6139086775197193,
      "grad_norm": 0.11120782047510147,
      "learning_rate": 0.0006706152758132957,
      "loss": 2.2372324466705322,
      "step": 6028
    },
    {
      "epoch": 0.6141123632123271,
      "grad_norm": 0.1539832353591919,
      "learning_rate": 0.0006702616690240453,
      "loss": 2.1907992362976074,
      "step": 6030
    },
    {
      "epoch": 0.6143160489049347,
      "grad_norm": 0.10874677449464798,
      "learning_rate": 0.000669908062234795,
      "loss": 2.2678589820861816,
      "step": 6032
    },
    {
      "epoch": 0.6145197345975425,
      "grad_norm": 0.10985569655895233,
      "learning_rate": 0.0006695544554455446,
      "loss": 2.1475181579589844,
      "step": 6034
    },
    {
      "epoch": 0.6147234202901503,
      "grad_norm": 0.12657906115055084,
      "learning_rate": 0.0006692008486562942,
      "loss": 2.2552380561828613,
      "step": 6036
    },
    {
      "epoch": 0.614927105982758,
      "grad_norm": 0.1191960945725441,
      "learning_rate": 0.0006688472418670438,
      "loss": 2.2110211849212646,
      "step": 6038
    },
    {
      "epoch": 0.6151307916753658,
      "grad_norm": 0.14636239409446716,
      "learning_rate": 0.0006684936350777935,
      "loss": 2.2359893321990967,
      "step": 6040
    },
    {
      "epoch": 0.6153344773679735,
      "grad_norm": 0.11240778863430023,
      "learning_rate": 0.0006681400282885432,
      "loss": 2.2522647380828857,
      "step": 6042
    },
    {
      "epoch": 0.6155381630605812,
      "grad_norm": 0.13003510236740112,
      "learning_rate": 0.0006677864214992928,
      "loss": 2.258657455444336,
      "step": 6044
    },
    {
      "epoch": 0.6157418487531889,
      "grad_norm": 0.12382952868938446,
      "learning_rate": 0.0006674328147100425,
      "loss": 2.185727596282959,
      "step": 6046
    },
    {
      "epoch": 0.6159455344457967,
      "grad_norm": 0.12961411476135254,
      "learning_rate": 0.0006670792079207921,
      "loss": 2.284958839416504,
      "step": 6048
    },
    {
      "epoch": 0.6161492201384045,
      "grad_norm": 0.13513757288455963,
      "learning_rate": 0.0006667256011315417,
      "loss": 2.1242177486419678,
      "step": 6050
    },
    {
      "epoch": 0.6163529058310122,
      "grad_norm": 0.1475076675415039,
      "learning_rate": 0.0006663719943422914,
      "loss": 2.2287795543670654,
      "step": 6052
    },
    {
      "epoch": 0.6165565915236199,
      "grad_norm": 0.11314855515956879,
      "learning_rate": 0.000666018387553041,
      "loss": 2.163574695587158,
      "step": 6054
    },
    {
      "epoch": 0.6167602772162276,
      "grad_norm": 0.13114269077777863,
      "learning_rate": 0.0006656647807637907,
      "loss": 2.1446096897125244,
      "step": 6056
    },
    {
      "epoch": 0.6169639629088354,
      "grad_norm": 0.10525120049715042,
      "learning_rate": 0.0006653111739745403,
      "loss": 2.0769128799438477,
      "step": 6058
    },
    {
      "epoch": 0.6171676486014431,
      "grad_norm": 0.15548484027385712,
      "learning_rate": 0.00066495756718529,
      "loss": 2.1659669876098633,
      "step": 6060
    },
    {
      "epoch": 0.6173713342940509,
      "grad_norm": 0.12814033031463623,
      "learning_rate": 0.0006646039603960397,
      "loss": 2.125365734100342,
      "step": 6062
    },
    {
      "epoch": 0.6175750199866585,
      "grad_norm": 0.14213046431541443,
      "learning_rate": 0.0006642503536067893,
      "loss": 2.202376365661621,
      "step": 6064
    },
    {
      "epoch": 0.6177787056792663,
      "grad_norm": 0.12267129868268967,
      "learning_rate": 0.0006638967468175389,
      "loss": 2.192208766937256,
      "step": 6066
    },
    {
      "epoch": 0.617982391371874,
      "grad_norm": 0.11448483169078827,
      "learning_rate": 0.0006635431400282885,
      "loss": 2.2092103958129883,
      "step": 6068
    },
    {
      "epoch": 0.6181860770644818,
      "grad_norm": 0.15266162157058716,
      "learning_rate": 0.0006631895332390382,
      "loss": 2.3116114139556885,
      "step": 6070
    },
    {
      "epoch": 0.6183897627570896,
      "grad_norm": 0.11963768303394318,
      "learning_rate": 0.0006628359264497878,
      "loss": 2.1441726684570312,
      "step": 6072
    },
    {
      "epoch": 0.6185934484496973,
      "grad_norm": 0.11775033175945282,
      "learning_rate": 0.0006624823196605376,
      "loss": 2.181398868560791,
      "step": 6074
    },
    {
      "epoch": 0.618797134142305,
      "grad_norm": 0.16307245194911957,
      "learning_rate": 0.0006621287128712872,
      "loss": 2.2187881469726562,
      "step": 6076
    },
    {
      "epoch": 0.6190008198349127,
      "grad_norm": 0.18259114027023315,
      "learning_rate": 0.0006617751060820368,
      "loss": 2.317248821258545,
      "step": 6078
    },
    {
      "epoch": 0.6192045055275205,
      "grad_norm": 0.11615128815174103,
      "learning_rate": 0.0006614214992927864,
      "loss": 2.042428970336914,
      "step": 6080
    },
    {
      "epoch": 0.6194081912201282,
      "grad_norm": 0.10759855061769485,
      "learning_rate": 0.000661067892503536,
      "loss": 2.0464155673980713,
      "step": 6082
    },
    {
      "epoch": 0.619611876912736,
      "grad_norm": 0.14527547359466553,
      "learning_rate": 0.0006607142857142857,
      "loss": 2.116197109222412,
      "step": 6084
    },
    {
      "epoch": 0.6198155626053437,
      "grad_norm": 0.10908454656600952,
      "learning_rate": 0.0006603606789250354,
      "loss": 2.1556079387664795,
      "step": 6086
    },
    {
      "epoch": 0.6200192482979514,
      "grad_norm": 0.1309065818786621,
      "learning_rate": 0.0006600070721357851,
      "loss": 2.1859753131866455,
      "step": 6088
    },
    {
      "epoch": 0.6202229339905592,
      "grad_norm": 0.13339336216449738,
      "learning_rate": 0.0006596534653465347,
      "loss": 2.2213683128356934,
      "step": 6090
    },
    {
      "epoch": 0.6204266196831669,
      "grad_norm": 0.130723237991333,
      "learning_rate": 0.0006592998585572843,
      "loss": 2.283025026321411,
      "step": 6092
    },
    {
      "epoch": 0.6206303053757747,
      "grad_norm": 0.13604722917079926,
      "learning_rate": 0.0006589462517680339,
      "loss": 2.1760754585266113,
      "step": 6094
    },
    {
      "epoch": 0.6208339910683823,
      "grad_norm": 0.12190213799476624,
      "learning_rate": 0.0006585926449787835,
      "loss": 2.1973321437835693,
      "step": 6096
    },
    {
      "epoch": 0.6210376767609901,
      "grad_norm": 0.11351391673088074,
      "learning_rate": 0.0006582390381895333,
      "loss": 2.175511360168457,
      "step": 6098
    },
    {
      "epoch": 0.6212413624535978,
      "grad_norm": 0.12845958769321442,
      "learning_rate": 0.0006578854314002829,
      "loss": 2.186943531036377,
      "step": 6100
    },
    {
      "epoch": 0.6214450481462056,
      "grad_norm": 0.12718899548053741,
      "learning_rate": 0.0006575318246110326,
      "loss": 2.0796475410461426,
      "step": 6102
    },
    {
      "epoch": 0.6216487338388134,
      "grad_norm": 0.11504922807216644,
      "learning_rate": 0.0006571782178217822,
      "loss": 2.303457736968994,
      "step": 6104
    },
    {
      "epoch": 0.6218524195314211,
      "grad_norm": 0.11883866786956787,
      "learning_rate": 0.0006568246110325318,
      "loss": 2.173351764678955,
      "step": 6106
    },
    {
      "epoch": 0.6220561052240288,
      "grad_norm": 0.1289854198694229,
      "learning_rate": 0.0006564710042432815,
      "loss": 2.267148494720459,
      "step": 6108
    },
    {
      "epoch": 0.6222597909166365,
      "grad_norm": 0.18656279146671295,
      "learning_rate": 0.0006561173974540311,
      "loss": 2.3352484703063965,
      "step": 6110
    },
    {
      "epoch": 0.6224634766092443,
      "grad_norm": 0.1268947869539261,
      "learning_rate": 0.0006557637906647808,
      "loss": 2.1629884243011475,
      "step": 6112
    },
    {
      "epoch": 0.622667162301852,
      "grad_norm": 0.13543035089969635,
      "learning_rate": 0.0006554101838755304,
      "loss": 2.071101665496826,
      "step": 6114
    },
    {
      "epoch": 0.6228708479944598,
      "grad_norm": 0.15298506617546082,
      "learning_rate": 0.0006550565770862801,
      "loss": 2.239231586456299,
      "step": 6116
    },
    {
      "epoch": 0.6230745336870674,
      "grad_norm": 0.10753388702869415,
      "learning_rate": 0.0006547029702970297,
      "loss": 2.029538631439209,
      "step": 6118
    },
    {
      "epoch": 0.6232782193796752,
      "grad_norm": 0.13123008608818054,
      "learning_rate": 0.0006543493635077794,
      "loss": 2.179166078567505,
      "step": 6120
    },
    {
      "epoch": 0.623481905072283,
      "grad_norm": 0.14134056866168976,
      "learning_rate": 0.000653995756718529,
      "loss": 2.2137534618377686,
      "step": 6122
    },
    {
      "epoch": 0.6236855907648907,
      "grad_norm": 0.11275806277990341,
      "learning_rate": 0.0006536421499292786,
      "loss": 2.235398292541504,
      "step": 6124
    },
    {
      "epoch": 0.6238892764574985,
      "grad_norm": 0.11304907500743866,
      "learning_rate": 0.0006532885431400283,
      "loss": 2.231987237930298,
      "step": 6126
    },
    {
      "epoch": 0.6240929621501061,
      "grad_norm": 0.1115996465086937,
      "learning_rate": 0.0006529349363507779,
      "loss": 2.3585634231567383,
      "step": 6128
    },
    {
      "epoch": 0.6242966478427139,
      "grad_norm": 0.1282646805047989,
      "learning_rate": 0.0006525813295615276,
      "loss": 2.2568631172180176,
      "step": 6130
    },
    {
      "epoch": 0.6245003335353216,
      "grad_norm": 0.13580602407455444,
      "learning_rate": 0.0006522277227722773,
      "loss": 2.2572414875030518,
      "step": 6132
    },
    {
      "epoch": 0.6247040192279294,
      "grad_norm": 0.12228091806173325,
      "learning_rate": 0.0006518741159830269,
      "loss": 2.1419975757598877,
      "step": 6134
    },
    {
      "epoch": 0.6249077049205372,
      "grad_norm": 0.12068063765764236,
      "learning_rate": 0.0006515205091937766,
      "loss": 2.1627721786499023,
      "step": 6136
    },
    {
      "epoch": 0.6251113906131448,
      "grad_norm": 0.11344760656356812,
      "learning_rate": 0.0006511669024045261,
      "loss": 2.1179652214050293,
      "step": 6138
    },
    {
      "epoch": 0.6253150763057526,
      "grad_norm": 0.12166816741228104,
      "learning_rate": 0.0006508132956152758,
      "loss": 2.1018338203430176,
      "step": 6140
    },
    {
      "epoch": 0.6255187619983603,
      "grad_norm": 0.11945467442274094,
      "learning_rate": 0.0006504596888260254,
      "loss": 2.2756927013397217,
      "step": 6142
    },
    {
      "epoch": 0.6257224476909681,
      "grad_norm": 0.13658872246742249,
      "learning_rate": 0.0006501060820367752,
      "loss": 2.226067066192627,
      "step": 6144
    },
    {
      "epoch": 0.6259261333835758,
      "grad_norm": 0.13649815320968628,
      "learning_rate": 0.0006497524752475248,
      "loss": 2.182080030441284,
      "step": 6146
    },
    {
      "epoch": 0.6261298190761836,
      "grad_norm": 0.14311586320400238,
      "learning_rate": 0.0006493988684582744,
      "loss": 2.177544116973877,
      "step": 6148
    },
    {
      "epoch": 0.6263335047687912,
      "grad_norm": 0.11827678233385086,
      "learning_rate": 0.0006490452616690241,
      "loss": 2.1747031211853027,
      "step": 6150
    },
    {
      "epoch": 0.626537190461399,
      "grad_norm": 0.1305646449327469,
      "learning_rate": 0.0006486916548797736,
      "loss": 2.271838665008545,
      "step": 6152
    },
    {
      "epoch": 0.6267408761540068,
      "grad_norm": 0.13568271696567535,
      "learning_rate": 0.0006483380480905234,
      "loss": 2.16910982131958,
      "step": 6154
    },
    {
      "epoch": 0.6269445618466145,
      "grad_norm": 0.09940563887357712,
      "learning_rate": 0.000647984441301273,
      "loss": 2.149360179901123,
      "step": 6156
    },
    {
      "epoch": 0.6271482475392223,
      "grad_norm": 0.14230294525623322,
      "learning_rate": 0.0006476308345120227,
      "loss": 2.170987129211426,
      "step": 6158
    },
    {
      "epoch": 0.6273519332318299,
      "grad_norm": 0.12457246333360672,
      "learning_rate": 0.0006472772277227723,
      "loss": 2.218899726867676,
      "step": 6160
    },
    {
      "epoch": 0.6275556189244377,
      "grad_norm": 0.11063449084758759,
      "learning_rate": 0.0006469236209335219,
      "loss": 2.1820201873779297,
      "step": 6162
    },
    {
      "epoch": 0.6277593046170454,
      "grad_norm": 0.13509277999401093,
      "learning_rate": 0.0006465700141442716,
      "loss": 2.0119972229003906,
      "step": 6164
    },
    {
      "epoch": 0.6279629903096532,
      "grad_norm": 0.1170012429356575,
      "learning_rate": 0.0006462164073550212,
      "loss": 2.0899338722229004,
      "step": 6166
    },
    {
      "epoch": 0.628166676002261,
      "grad_norm": 0.14438816905021667,
      "learning_rate": 0.0006458628005657709,
      "loss": 2.18418550491333,
      "step": 6168
    },
    {
      "epoch": 0.6283703616948686,
      "grad_norm": 0.1475767195224762,
      "learning_rate": 0.0006455091937765205,
      "loss": 2.2060208320617676,
      "step": 6170
    },
    {
      "epoch": 0.6285740473874764,
      "grad_norm": 0.14569556713104248,
      "learning_rate": 0.0006451555869872702,
      "loss": 2.2189316749572754,
      "step": 6172
    },
    {
      "epoch": 0.6287777330800841,
      "grad_norm": 0.19569337368011475,
      "learning_rate": 0.0006448019801980198,
      "loss": 2.3008322715759277,
      "step": 6174
    },
    {
      "epoch": 0.6289814187726919,
      "grad_norm": 0.1300409883260727,
      "learning_rate": 0.0006444483734087694,
      "loss": 2.2209055423736572,
      "step": 6176
    },
    {
      "epoch": 0.6291851044652996,
      "grad_norm": 0.13190506398677826,
      "learning_rate": 0.0006440947666195192,
      "loss": 2.2814812660217285,
      "step": 6178
    },
    {
      "epoch": 0.6293887901579074,
      "grad_norm": 0.11865761876106262,
      "learning_rate": 0.0006437411598302687,
      "loss": 2.232536792755127,
      "step": 6180
    },
    {
      "epoch": 0.629592475850515,
      "grad_norm": 0.12957367300987244,
      "learning_rate": 0.0006433875530410184,
      "loss": 2.198939085006714,
      "step": 6182
    },
    {
      "epoch": 0.6297961615431228,
      "grad_norm": 0.12187027186155319,
      "learning_rate": 0.000643033946251768,
      "loss": 2.186192512512207,
      "step": 6184
    },
    {
      "epoch": 0.6299998472357305,
      "grad_norm": 0.1284366399049759,
      "learning_rate": 0.0006426803394625177,
      "loss": 2.2756242752075195,
      "step": 6186
    },
    {
      "epoch": 0.6302035329283383,
      "grad_norm": 0.10745688527822495,
      "learning_rate": 0.0006423267326732674,
      "loss": 2.231365919113159,
      "step": 6188
    },
    {
      "epoch": 0.6304072186209461,
      "grad_norm": 0.11868152767419815,
      "learning_rate": 0.000641973125884017,
      "loss": 2.1645612716674805,
      "step": 6190
    },
    {
      "epoch": 0.6306109043135537,
      "grad_norm": 0.13540369272232056,
      "learning_rate": 0.0006416195190947667,
      "loss": 2.2290327548980713,
      "step": 6192
    },
    {
      "epoch": 0.6308145900061615,
      "grad_norm": 0.14016500115394592,
      "learning_rate": 0.0006412659123055162,
      "loss": 2.0966062545776367,
      "step": 6194
    },
    {
      "epoch": 0.6310182756987692,
      "grad_norm": 0.14389309287071228,
      "learning_rate": 0.0006409123055162659,
      "loss": 2.4025676250457764,
      "step": 6196
    },
    {
      "epoch": 0.631221961391377,
      "grad_norm": 0.10881274193525314,
      "learning_rate": 0.0006405586987270155,
      "loss": 2.1253957748413086,
      "step": 6198
    },
    {
      "epoch": 0.6314256470839847,
      "grad_norm": 0.12776364386081696,
      "learning_rate": 0.0006402050919377653,
      "loss": 2.288860559463501,
      "step": 6200
    },
    {
      "epoch": 0.6316293327765924,
      "grad_norm": 0.11858262121677399,
      "learning_rate": 0.0006398514851485149,
      "loss": 2.321470260620117,
      "step": 6202
    },
    {
      "epoch": 0.6318330184692001,
      "grad_norm": 0.13187777996063232,
      "learning_rate": 0.0006394978783592645,
      "loss": 2.1928927898406982,
      "step": 6204
    },
    {
      "epoch": 0.6320367041618079,
      "grad_norm": 0.13595299422740936,
      "learning_rate": 0.0006391442715700142,
      "loss": 2.245807647705078,
      "step": 6206
    },
    {
      "epoch": 0.6322403898544157,
      "grad_norm": 0.1525571048259735,
      "learning_rate": 0.0006387906647807637,
      "loss": 2.1730074882507324,
      "step": 6208
    },
    {
      "epoch": 0.6324440755470234,
      "grad_norm": 0.2069864273071289,
      "learning_rate": 0.0006384370579915134,
      "loss": 2.1084887981414795,
      "step": 6210
    },
    {
      "epoch": 0.6326477612396312,
      "grad_norm": 0.1270495057106018,
      "learning_rate": 0.0006380834512022631,
      "loss": 2.306097984313965,
      "step": 6212
    },
    {
      "epoch": 0.6328514469322388,
      "grad_norm": 0.1272631138563156,
      "learning_rate": 0.0006377298444130128,
      "loss": 2.2024667263031006,
      "step": 6214
    },
    {
      "epoch": 0.6330551326248466,
      "grad_norm": 0.1326553374528885,
      "learning_rate": 0.0006373762376237624,
      "loss": 2.287569046020508,
      "step": 6216
    },
    {
      "epoch": 0.6332588183174543,
      "grad_norm": 0.13511191308498383,
      "learning_rate": 0.000637022630834512,
      "loss": 2.22867488861084,
      "step": 6218
    },
    {
      "epoch": 0.6334625040100621,
      "grad_norm": 0.1232408732175827,
      "learning_rate": 0.0006366690240452617,
      "loss": 2.3530139923095703,
      "step": 6220
    },
    {
      "epoch": 0.6336661897026699,
      "grad_norm": 0.12675848603248596,
      "learning_rate": 0.0006363154172560112,
      "loss": 2.1226301193237305,
      "step": 6222
    },
    {
      "epoch": 0.6338698753952775,
      "grad_norm": 0.1181819960474968,
      "learning_rate": 0.000635961810466761,
      "loss": 2.2860100269317627,
      "step": 6224
    },
    {
      "epoch": 0.6340735610878853,
      "grad_norm": 0.14696665108203888,
      "learning_rate": 0.0006356082036775106,
      "loss": 2.171440839767456,
      "step": 6226
    },
    {
      "epoch": 0.634277246780493,
      "grad_norm": 0.1299416571855545,
      "learning_rate": 0.0006352545968882603,
      "loss": 2.146697998046875,
      "step": 6228
    },
    {
      "epoch": 0.6344809324731008,
      "grad_norm": 0.13183563947677612,
      "learning_rate": 0.0006349009900990099,
      "loss": 2.2170350551605225,
      "step": 6230
    },
    {
      "epoch": 0.6346846181657085,
      "grad_norm": 0.11225762218236923,
      "learning_rate": 0.0006345473833097595,
      "loss": 2.224008083343506,
      "step": 6232
    },
    {
      "epoch": 0.6348883038583162,
      "grad_norm": 0.12826114892959595,
      "learning_rate": 0.0006341937765205093,
      "loss": 2.2194204330444336,
      "step": 6234
    },
    {
      "epoch": 0.6350919895509239,
      "grad_norm": 0.11958542466163635,
      "learning_rate": 0.0006338401697312589,
      "loss": 2.2406110763549805,
      "step": 6236
    },
    {
      "epoch": 0.6352956752435317,
      "grad_norm": 0.11838272958993912,
      "learning_rate": 0.0006334865629420085,
      "loss": 1.9759302139282227,
      "step": 6238
    },
    {
      "epoch": 0.6354993609361395,
      "grad_norm": 0.11560579389333725,
      "learning_rate": 0.0006331329561527581,
      "loss": 2.17691707611084,
      "step": 6240
    },
    {
      "epoch": 0.6357030466287472,
      "grad_norm": 0.12701982259750366,
      "learning_rate": 0.0006327793493635078,
      "loss": 2.1876869201660156,
      "step": 6242
    },
    {
      "epoch": 0.635906732321355,
      "grad_norm": 0.12560832500457764,
      "learning_rate": 0.0006324257425742574,
      "loss": 2.124069929122925,
      "step": 6244
    },
    {
      "epoch": 0.6361104180139626,
      "grad_norm": 0.10792442411184311,
      "learning_rate": 0.0006320721357850071,
      "loss": 2.0237839221954346,
      "step": 6246
    },
    {
      "epoch": 0.6363141037065704,
      "grad_norm": 0.14660875499248505,
      "learning_rate": 0.0006317185289957568,
      "loss": 2.3088369369506836,
      "step": 6248
    },
    {
      "epoch": 0.6365177893991781,
      "grad_norm": 0.11672408133745193,
      "learning_rate": 0.0006313649222065064,
      "loss": 2.232931613922119,
      "step": 6250
    },
    {
      "epoch": 0.6367214750917859,
      "grad_norm": 0.13354046642780304,
      "learning_rate": 0.000631011315417256,
      "loss": 2.0900802612304688,
      "step": 6252
    },
    {
      "epoch": 0.6369251607843937,
      "grad_norm": 0.11946399509906769,
      "learning_rate": 0.0006306577086280056,
      "loss": 2.2826590538024902,
      "step": 6254
    },
    {
      "epoch": 0.6371288464770013,
      "grad_norm": 0.1229013130068779,
      "learning_rate": 0.0006303041018387554,
      "loss": 2.0985138416290283,
      "step": 6256
    },
    {
      "epoch": 0.6373325321696091,
      "grad_norm": 0.1395273357629776,
      "learning_rate": 0.000629950495049505,
      "loss": 2.183420181274414,
      "step": 6258
    },
    {
      "epoch": 0.6375362178622168,
      "grad_norm": 0.11173433810472488,
      "learning_rate": 0.0006295968882602546,
      "loss": 2.193664073944092,
      "step": 6260
    },
    {
      "epoch": 0.6377399035548246,
      "grad_norm": 0.12663209438323975,
      "learning_rate": 0.0006292432814710043,
      "loss": 2.2583093643188477,
      "step": 6262
    },
    {
      "epoch": 0.6379435892474323,
      "grad_norm": 0.12875330448150635,
      "learning_rate": 0.0006288896746817539,
      "loss": 2.2027816772460938,
      "step": 6264
    },
    {
      "epoch": 0.63814727494004,
      "grad_norm": 0.11642230302095413,
      "learning_rate": 0.0006285360678925035,
      "loss": 2.1806936264038086,
      "step": 6266
    },
    {
      "epoch": 0.6383509606326477,
      "grad_norm": 0.10332033038139343,
      "learning_rate": 0.0006281824611032532,
      "loss": 2.1539623737335205,
      "step": 6268
    },
    {
      "epoch": 0.6385546463252555,
      "grad_norm": 0.1458934247493744,
      "learning_rate": 0.0006278288543140029,
      "loss": 2.3475983142852783,
      "step": 6270
    },
    {
      "epoch": 0.6387583320178633,
      "grad_norm": 0.18642187118530273,
      "learning_rate": 0.0006274752475247525,
      "loss": 2.1527843475341797,
      "step": 6272
    },
    {
      "epoch": 0.638962017710471,
      "grad_norm": 0.12385013699531555,
      "learning_rate": 0.0006271216407355021,
      "loss": 2.166287899017334,
      "step": 6274
    },
    {
      "epoch": 0.6391657034030787,
      "grad_norm": 0.12003276497125626,
      "learning_rate": 0.0006267680339462518,
      "loss": 2.2307095527648926,
      "step": 6276
    },
    {
      "epoch": 0.6393693890956864,
      "grad_norm": 0.1173548772931099,
      "learning_rate": 0.0006264144271570014,
      "loss": 2.2704288959503174,
      "step": 6278
    },
    {
      "epoch": 0.6395730747882942,
      "grad_norm": 0.11846067011356354,
      "learning_rate": 0.0006260608203677511,
      "loss": 2.265699863433838,
      "step": 6280
    },
    {
      "epoch": 0.6397767604809019,
      "grad_norm": 0.16386178135871887,
      "learning_rate": 0.0006257072135785007,
      "loss": 2.276601552963257,
      "step": 6282
    },
    {
      "epoch": 0.6399804461735097,
      "grad_norm": 0.11389186233282089,
      "learning_rate": 0.0006253536067892504,
      "loss": 2.160027027130127,
      "step": 6284
    },
    {
      "epoch": 0.6401841318661174,
      "grad_norm": 0.11156056821346283,
      "learning_rate": 0.000625,
      "loss": 2.1013312339782715,
      "step": 6286
    },
    {
      "epoch": 0.6403878175587251,
      "grad_norm": 0.12171629816293716,
      "learning_rate": 0.0006246463932107496,
      "loss": 2.134366512298584,
      "step": 6288
    },
    {
      "epoch": 0.6405915032513328,
      "grad_norm": 0.12361984699964523,
      "learning_rate": 0.0006242927864214994,
      "loss": 2.3297533988952637,
      "step": 6290
    },
    {
      "epoch": 0.6407951889439406,
      "grad_norm": 0.12072043120861053,
      "learning_rate": 0.000623939179632249,
      "loss": 2.133990526199341,
      "step": 6292
    },
    {
      "epoch": 0.6409988746365484,
      "grad_norm": 0.11242806911468506,
      "learning_rate": 0.0006235855728429986,
      "loss": 2.1084296703338623,
      "step": 6294
    },
    {
      "epoch": 0.6412025603291561,
      "grad_norm": 0.1372925192117691,
      "learning_rate": 0.0006232319660537482,
      "loss": 2.1608097553253174,
      "step": 6296
    },
    {
      "epoch": 0.6414062460217638,
      "grad_norm": 0.15135297179222107,
      "learning_rate": 0.0006228783592644979,
      "loss": 2.1530208587646484,
      "step": 6298
    },
    {
      "epoch": 0.6416099317143715,
      "grad_norm": 0.1209937259554863,
      "learning_rate": 0.0006225247524752475,
      "loss": 2.2466535568237305,
      "step": 6300
    },
    {
      "epoch": 0.6418136174069793,
      "grad_norm": 0.1445452868938446,
      "learning_rate": 0.0006221711456859972,
      "loss": 2.122600793838501,
      "step": 6302
    },
    {
      "epoch": 0.642017303099587,
      "grad_norm": 0.16167102754116058,
      "learning_rate": 0.0006218175388967469,
      "loss": 2.1701765060424805,
      "step": 6304
    },
    {
      "epoch": 0.6422209887921948,
      "grad_norm": 0.12057188898324966,
      "learning_rate": 0.0006214639321074965,
      "loss": 2.0993752479553223,
      "step": 6306
    },
    {
      "epoch": 0.6424246744848024,
      "grad_norm": 0.13527452945709229,
      "learning_rate": 0.0006211103253182461,
      "loss": 2.1016199588775635,
      "step": 6308
    },
    {
      "epoch": 0.6426283601774102,
      "grad_norm": 0.13119293749332428,
      "learning_rate": 0.0006207567185289957,
      "loss": 2.0926597118377686,
      "step": 6310
    },
    {
      "epoch": 0.642832045870018,
      "grad_norm": 0.11772466450929642,
      "learning_rate": 0.0006204031117397454,
      "loss": 2.2520034313201904,
      "step": 6312
    },
    {
      "epoch": 0.6430357315626257,
      "grad_norm": 0.1104651615023613,
      "learning_rate": 0.0006200495049504951,
      "loss": 2.196279525756836,
      "step": 6314
    },
    {
      "epoch": 0.6432394172552335,
      "grad_norm": 0.11810880154371262,
      "learning_rate": 0.0006196958981612447,
      "loss": 2.2100324630737305,
      "step": 6316
    },
    {
      "epoch": 0.6434431029478412,
      "grad_norm": 0.12361878901720047,
      "learning_rate": 0.0006193422913719944,
      "loss": 2.1428375244140625,
      "step": 6318
    },
    {
      "epoch": 0.6436467886404489,
      "grad_norm": 0.10758315026760101,
      "learning_rate": 0.000618988684582744,
      "loss": 2.1669249534606934,
      "step": 6320
    },
    {
      "epoch": 0.6438504743330566,
      "grad_norm": 0.1224251389503479,
      "learning_rate": 0.0006186350777934936,
      "loss": 2.128905773162842,
      "step": 6322
    },
    {
      "epoch": 0.6440541600256644,
      "grad_norm": 0.12061122804880142,
      "learning_rate": 0.0006182814710042432,
      "loss": 2.0873522758483887,
      "step": 6324
    },
    {
      "epoch": 0.6442578457182722,
      "grad_norm": 0.1419738531112671,
      "learning_rate": 0.000617927864214993,
      "loss": 2.263314962387085,
      "step": 6326
    },
    {
      "epoch": 0.6444615314108799,
      "grad_norm": 0.14759784936904907,
      "learning_rate": 0.0006175742574257426,
      "loss": 2.3812756538391113,
      "step": 6328
    },
    {
      "epoch": 0.6446652171034876,
      "grad_norm": 0.12879441678524017,
      "learning_rate": 0.0006172206506364922,
      "loss": 2.17915415763855,
      "step": 6330
    },
    {
      "epoch": 0.6448689027960953,
      "grad_norm": 0.12756220996379852,
      "learning_rate": 0.0006168670438472419,
      "loss": 2.163940906524658,
      "step": 6332
    },
    {
      "epoch": 0.6450725884887031,
      "grad_norm": 0.12183970957994461,
      "learning_rate": 0.0006165134370579915,
      "loss": 2.1834867000579834,
      "step": 6334
    },
    {
      "epoch": 0.6452762741813108,
      "grad_norm": 0.12366587668657303,
      "learning_rate": 0.0006161598302687411,
      "loss": 2.2444117069244385,
      "step": 6336
    },
    {
      "epoch": 0.6454799598739186,
      "grad_norm": 0.11047738045454025,
      "learning_rate": 0.0006158062234794908,
      "loss": 2.226478099822998,
      "step": 6338
    },
    {
      "epoch": 0.6456836455665262,
      "grad_norm": 0.15367628633975983,
      "learning_rate": 0.0006154526166902405,
      "loss": 2.2373239994049072,
      "step": 6340
    },
    {
      "epoch": 0.645887331259134,
      "grad_norm": 0.13164633512496948,
      "learning_rate": 0.0006150990099009901,
      "loss": 2.1757845878601074,
      "step": 6342
    },
    {
      "epoch": 0.6460910169517418,
      "grad_norm": 0.10945349186658859,
      "learning_rate": 0.0006147454031117397,
      "loss": 2.2596559524536133,
      "step": 6344
    },
    {
      "epoch": 0.6462947026443495,
      "grad_norm": 0.13242891430854797,
      "learning_rate": 0.0006143917963224894,
      "loss": 2.3594071865081787,
      "step": 6346
    },
    {
      "epoch": 0.6464983883369573,
      "grad_norm": 0.16045989096164703,
      "learning_rate": 0.0006140381895332391,
      "loss": 2.2350716590881348,
      "step": 6348
    },
    {
      "epoch": 0.646702074029565,
      "grad_norm": 0.17283880710601807,
      "learning_rate": 0.0006136845827439888,
      "loss": 2.228585958480835,
      "step": 6350
    },
    {
      "epoch": 0.6469057597221727,
      "grad_norm": 0.14210273325443268,
      "learning_rate": 0.0006133309759547383,
      "loss": 2.141378402709961,
      "step": 6352
    },
    {
      "epoch": 0.6471094454147804,
      "grad_norm": 0.1255907118320465,
      "learning_rate": 0.000612977369165488,
      "loss": 2.2373721599578857,
      "step": 6354
    },
    {
      "epoch": 0.6473131311073882,
      "grad_norm": 0.11482849717140198,
      "learning_rate": 0.0006126237623762376,
      "loss": 2.092116355895996,
      "step": 6356
    },
    {
      "epoch": 0.647516816799996,
      "grad_norm": 0.1410919427871704,
      "learning_rate": 0.0006122701555869872,
      "loss": 2.183887004852295,
      "step": 6358
    },
    {
      "epoch": 0.6477205024926037,
      "grad_norm": 0.12539854645729065,
      "learning_rate": 0.000611916548797737,
      "loss": 2.2171103954315186,
      "step": 6360
    },
    {
      "epoch": 0.6479241881852114,
      "grad_norm": 0.11664523929357529,
      "learning_rate": 0.0006115629420084866,
      "loss": 2.1765294075012207,
      "step": 6362
    },
    {
      "epoch": 0.6481278738778191,
      "grad_norm": 0.14048056304454803,
      "learning_rate": 0.0006112093352192363,
      "loss": 2.201698064804077,
      "step": 6364
    },
    {
      "epoch": 0.6483315595704269,
      "grad_norm": 0.18025995790958405,
      "learning_rate": 0.0006108557284299858,
      "loss": 2.2862625122070312,
      "step": 6366
    },
    {
      "epoch": 0.6485352452630346,
      "grad_norm": 0.13381128013134003,
      "learning_rate": 0.0006105021216407355,
      "loss": 2.1433138847351074,
      "step": 6368
    },
    {
      "epoch": 0.6487389309556424,
      "grad_norm": 0.11735348403453827,
      "learning_rate": 0.0006101485148514851,
      "loss": 2.1498143672943115,
      "step": 6370
    },
    {
      "epoch": 0.64894261664825,
      "grad_norm": 0.11966418474912643,
      "learning_rate": 0.0006097949080622348,
      "loss": 2.1217751502990723,
      "step": 6372
    },
    {
      "epoch": 0.6491463023408578,
      "grad_norm": 0.12271948158740997,
      "learning_rate": 0.0006094413012729845,
      "loss": 2.2053582668304443,
      "step": 6374
    },
    {
      "epoch": 0.6493499880334656,
      "grad_norm": 0.14635367691516876,
      "learning_rate": 0.0006090876944837341,
      "loss": 2.1338839530944824,
      "step": 6376
    },
    {
      "epoch": 0.6495536737260733,
      "grad_norm": 0.12828940153121948,
      "learning_rate": 0.0006087340876944838,
      "loss": 2.1675705909729004,
      "step": 6378
    },
    {
      "epoch": 0.6497573594186811,
      "grad_norm": 0.10813546925783157,
      "learning_rate": 0.0006083804809052333,
      "loss": 2.3040976524353027,
      "step": 6380
    },
    {
      "epoch": 0.6499610451112887,
      "grad_norm": 0.14479957520961761,
      "learning_rate": 0.0006080268741159831,
      "loss": 2.1643574237823486,
      "step": 6382
    },
    {
      "epoch": 0.6501647308038965,
      "grad_norm": 0.11798340082168579,
      "learning_rate": 0.0006076732673267327,
      "loss": 2.1976842880249023,
      "step": 6384
    },
    {
      "epoch": 0.6503684164965042,
      "grad_norm": 0.12092196196317673,
      "learning_rate": 0.0006073196605374823,
      "loss": 2.156385660171509,
      "step": 6386
    },
    {
      "epoch": 0.650572102189112,
      "grad_norm": 0.127323180437088,
      "learning_rate": 0.000606966053748232,
      "loss": 2.149869918823242,
      "step": 6388
    },
    {
      "epoch": 0.6507757878817197,
      "grad_norm": 0.12048517167568207,
      "learning_rate": 0.0006066124469589816,
      "loss": 2.218135118484497,
      "step": 6390
    },
    {
      "epoch": 0.6509794735743275,
      "grad_norm": 0.14185552299022675,
      "learning_rate": 0.0006062588401697313,
      "loss": 2.2942543029785156,
      "step": 6392
    },
    {
      "epoch": 0.6511831592669352,
      "grad_norm": 0.13854925334453583,
      "learning_rate": 0.0006059052333804809,
      "loss": 2.125215530395508,
      "step": 6394
    },
    {
      "epoch": 0.6513868449595429,
      "grad_norm": 0.14960628747940063,
      "learning_rate": 0.0006055516265912306,
      "loss": 2.1029367446899414,
      "step": 6396
    },
    {
      "epoch": 0.6515905306521507,
      "grad_norm": 0.18106921017169952,
      "learning_rate": 0.0006051980198019802,
      "loss": 2.220560073852539,
      "step": 6398
    },
    {
      "epoch": 0.6517942163447584,
      "grad_norm": 0.14215978980064392,
      "learning_rate": 0.0006048444130127298,
      "loss": 2.29769229888916,
      "step": 6400
    },
    {
      "epoch": 0.6519979020373662,
      "grad_norm": 0.12443889677524567,
      "learning_rate": 0.0006044908062234795,
      "loss": 2.2445082664489746,
      "step": 6402
    },
    {
      "epoch": 0.6522015877299738,
      "grad_norm": 0.12338683009147644,
      "learning_rate": 0.0006041371994342291,
      "loss": 2.255765676498413,
      "step": 6404
    },
    {
      "epoch": 0.6524052734225816,
      "grad_norm": 0.13481898605823517,
      "learning_rate": 0.0006037835926449789,
      "loss": 2.283073902130127,
      "step": 6406
    },
    {
      "epoch": 0.6526089591151893,
      "grad_norm": 0.11238790303468704,
      "learning_rate": 0.0006034299858557284,
      "loss": 2.2795767784118652,
      "step": 6408
    },
    {
      "epoch": 0.6528126448077971,
      "grad_norm": 0.1488676220178604,
      "learning_rate": 0.0006030763790664781,
      "loss": 2.238154649734497,
      "step": 6410
    },
    {
      "epoch": 0.6530163305004049,
      "grad_norm": 0.12391486763954163,
      "learning_rate": 0.0006027227722772277,
      "loss": 2.320791721343994,
      "step": 6412
    },
    {
      "epoch": 0.6532200161930125,
      "grad_norm": 0.1296369880437851,
      "learning_rate": 0.0006023691654879773,
      "loss": 2.2048072814941406,
      "step": 6414
    },
    {
      "epoch": 0.6534237018856203,
      "grad_norm": 0.12247279286384583,
      "learning_rate": 0.0006020155586987271,
      "loss": 2.247986078262329,
      "step": 6416
    },
    {
      "epoch": 0.653627387578228,
      "grad_norm": 0.14384430646896362,
      "learning_rate": 0.0006016619519094767,
      "loss": 2.246626377105713,
      "step": 6418
    },
    {
      "epoch": 0.6538310732708358,
      "grad_norm": 0.13940465450286865,
      "learning_rate": 0.0006013083451202264,
      "loss": 2.2653145790100098,
      "step": 6420
    },
    {
      "epoch": 0.6540347589634435,
      "grad_norm": 0.11411906778812408,
      "learning_rate": 0.0006009547383309759,
      "loss": 2.158053398132324,
      "step": 6422
    },
    {
      "epoch": 0.6542384446560513,
      "grad_norm": 0.11596006900072098,
      "learning_rate": 0.0006006011315417256,
      "loss": 2.1088225841522217,
      "step": 6424
    },
    {
      "epoch": 0.654442130348659,
      "grad_norm": 0.12322848290205002,
      "learning_rate": 0.0006002475247524752,
      "loss": 2.0411884784698486,
      "step": 6426
    },
    {
      "epoch": 0.6546458160412667,
      "grad_norm": 0.11348558217287064,
      "learning_rate": 0.0005998939179632249,
      "loss": 2.102726936340332,
      "step": 6428
    },
    {
      "epoch": 0.6548495017338745,
      "grad_norm": 0.11993367969989777,
      "learning_rate": 0.0005995403111739746,
      "loss": 2.176690101623535,
      "step": 6430
    },
    {
      "epoch": 0.6550531874264822,
      "grad_norm": 0.1274234652519226,
      "learning_rate": 0.0005991867043847242,
      "loss": 2.2784857749938965,
      "step": 6432
    },
    {
      "epoch": 0.65525687311909,
      "grad_norm": 0.12404479086399078,
      "learning_rate": 0.0005988330975954739,
      "loss": 2.286421775817871,
      "step": 6434
    },
    {
      "epoch": 0.6554605588116976,
      "grad_norm": 0.12470923364162445,
      "learning_rate": 0.0005984794908062234,
      "loss": 2.189089775085449,
      "step": 6436
    },
    {
      "epoch": 0.6556642445043054,
      "grad_norm": 0.15292224287986755,
      "learning_rate": 0.0005981258840169731,
      "loss": 2.225403308868408,
      "step": 6438
    },
    {
      "epoch": 0.6558679301969131,
      "grad_norm": 0.14307346940040588,
      "learning_rate": 0.0005977722772277228,
      "loss": 2.2401931285858154,
      "step": 6440
    },
    {
      "epoch": 0.6560716158895209,
      "grad_norm": 0.12224774062633514,
      "learning_rate": 0.0005974186704384724,
      "loss": 2.2023086547851562,
      "step": 6442
    },
    {
      "epoch": 0.6562753015821287,
      "grad_norm": 0.10780738294124603,
      "learning_rate": 0.0005970650636492221,
      "loss": 2.099853754043579,
      "step": 6444
    },
    {
      "epoch": 0.6564789872747363,
      "grad_norm": 0.12523357570171356,
      "learning_rate": 0.0005967114568599717,
      "loss": 2.3040812015533447,
      "step": 6446
    },
    {
      "epoch": 0.6566826729673441,
      "grad_norm": 0.12012991309165955,
      "learning_rate": 0.0005963578500707214,
      "loss": 2.0525214672088623,
      "step": 6448
    },
    {
      "epoch": 0.6568863586599518,
      "grad_norm": 0.1304541826248169,
      "learning_rate": 0.000596004243281471,
      "loss": 2.187079429626465,
      "step": 6450
    },
    {
      "epoch": 0.6570900443525596,
      "grad_norm": 0.14596222341060638,
      "learning_rate": 0.0005956506364922207,
      "loss": 2.231369733810425,
      "step": 6452
    },
    {
      "epoch": 0.6572937300451673,
      "grad_norm": 0.15714357793331146,
      "learning_rate": 0.0005952970297029703,
      "loss": 2.2782702445983887,
      "step": 6454
    },
    {
      "epoch": 0.6574974157377751,
      "grad_norm": 0.1273224800825119,
      "learning_rate": 0.0005949434229137199,
      "loss": 2.1936092376708984,
      "step": 6456
    },
    {
      "epoch": 0.6577011014303827,
      "grad_norm": 0.12941332161426544,
      "learning_rate": 0.0005945898161244696,
      "loss": 2.291776657104492,
      "step": 6458
    },
    {
      "epoch": 0.6579047871229905,
      "grad_norm": 0.10689939558506012,
      "learning_rate": 0.0005942362093352192,
      "loss": 2.2545504570007324,
      "step": 6460
    },
    {
      "epoch": 0.6581084728155983,
      "grad_norm": 0.13396690785884857,
      "learning_rate": 0.000593882602545969,
      "loss": 2.1225883960723877,
      "step": 6462
    },
    {
      "epoch": 0.658312158508206,
      "grad_norm": 0.12885591387748718,
      "learning_rate": 0.0005935289957567186,
      "loss": 2.238191843032837,
      "step": 6464
    },
    {
      "epoch": 0.6585158442008138,
      "grad_norm": 0.13396909832954407,
      "learning_rate": 0.0005931753889674682,
      "loss": 2.1699047088623047,
      "step": 6466
    },
    {
      "epoch": 0.6587195298934214,
      "grad_norm": 0.1269996613264084,
      "learning_rate": 0.0005928217821782178,
      "loss": 2.0954627990722656,
      "step": 6468
    },
    {
      "epoch": 0.6589232155860292,
      "grad_norm": 0.1470264196395874,
      "learning_rate": 0.0005924681753889674,
      "loss": 2.1755571365356445,
      "step": 6470
    },
    {
      "epoch": 0.6591269012786369,
      "grad_norm": 0.10489588230848312,
      "learning_rate": 0.0005921145685997171,
      "loss": 2.0854508876800537,
      "step": 6472
    },
    {
      "epoch": 0.6593305869712447,
      "grad_norm": 0.12009543180465698,
      "learning_rate": 0.0005917609618104668,
      "loss": 2.0961790084838867,
      "step": 6474
    },
    {
      "epoch": 0.6595342726638524,
      "grad_norm": 0.12135636061429977,
      "learning_rate": 0.0005914073550212165,
      "loss": 2.1420645713806152,
      "step": 6476
    },
    {
      "epoch": 0.6597379583564601,
      "grad_norm": 0.13952437043190002,
      "learning_rate": 0.0005910537482319661,
      "loss": 2.108983039855957,
      "step": 6478
    },
    {
      "epoch": 0.6599416440490679,
      "grad_norm": 0.11508040875196457,
      "learning_rate": 0.0005907001414427157,
      "loss": 2.215282440185547,
      "step": 6480
    },
    {
      "epoch": 0.6601453297416756,
      "grad_norm": 0.14740854501724243,
      "learning_rate": 0.0005903465346534653,
      "loss": 2.3571109771728516,
      "step": 6482
    },
    {
      "epoch": 0.6603490154342834,
      "grad_norm": 0.12140098214149475,
      "learning_rate": 0.0005899929278642149,
      "loss": 2.275908946990967,
      "step": 6484
    },
    {
      "epoch": 0.6605527011268911,
      "grad_norm": 0.11165975034236908,
      "learning_rate": 0.0005896393210749647,
      "loss": 2.03855037689209,
      "step": 6486
    },
    {
      "epoch": 0.6607563868194989,
      "grad_norm": 0.1395028829574585,
      "learning_rate": 0.0005892857142857143,
      "loss": 2.3523659706115723,
      "step": 6488
    },
    {
      "epoch": 0.6609600725121065,
      "grad_norm": 0.12047180533409119,
      "learning_rate": 0.000588932107496464,
      "loss": 2.188145875930786,
      "step": 6490
    },
    {
      "epoch": 0.6611637582047143,
      "grad_norm": 0.12289261817932129,
      "learning_rate": 0.0005885785007072136,
      "loss": 2.136874198913574,
      "step": 6492
    },
    {
      "epoch": 0.661367443897322,
      "grad_norm": 0.11287609487771988,
      "learning_rate": 0.0005882248939179632,
      "loss": 2.0606629848480225,
      "step": 6494
    },
    {
      "epoch": 0.6615711295899298,
      "grad_norm": 0.11279138922691345,
      "learning_rate": 0.0005878712871287129,
      "loss": 2.1829240322113037,
      "step": 6496
    },
    {
      "epoch": 0.6617748152825376,
      "grad_norm": 0.16172116994857788,
      "learning_rate": 0.0005875176803394625,
      "loss": 2.1559619903564453,
      "step": 6498
    },
    {
      "epoch": 0.6619785009751452,
      "grad_norm": 0.1525506228208542,
      "learning_rate": 0.0005871640735502122,
      "loss": 2.1542797088623047,
      "step": 6500
    },
    {
      "epoch": 0.662182186667753,
      "grad_norm": 0.12072119861841202,
      "learning_rate": 0.0005868104667609618,
      "loss": 2.1415228843688965,
      "step": 6502
    },
    {
      "epoch": 0.6623858723603607,
      "grad_norm": 0.1287134736776352,
      "learning_rate": 0.0005864568599717115,
      "loss": 2.221951723098755,
      "step": 6504
    },
    {
      "epoch": 0.6625895580529685,
      "grad_norm": 0.15478810667991638,
      "learning_rate": 0.0005861032531824611,
      "loss": 2.1871962547302246,
      "step": 6506
    },
    {
      "epoch": 0.6627932437455762,
      "grad_norm": 0.1158335879445076,
      "learning_rate": 0.0005857496463932108,
      "loss": 1.9978922605514526,
      "step": 6508
    },
    {
      "epoch": 0.6629969294381839,
      "grad_norm": 0.11401882022619247,
      "learning_rate": 0.0005853960396039604,
      "loss": 2.029592990875244,
      "step": 6510
    },
    {
      "epoch": 0.6632006151307916,
      "grad_norm": 0.174535870552063,
      "learning_rate": 0.00058504243281471,
      "loss": 2.204000949859619,
      "step": 6512
    },
    {
      "epoch": 0.6634043008233994,
      "grad_norm": 0.11549143493175507,
      "learning_rate": 0.0005846888260254597,
      "loss": 2.2037675380706787,
      "step": 6514
    },
    {
      "epoch": 0.6636079865160072,
      "grad_norm": 0.1382184773683548,
      "learning_rate": 0.0005843352192362093,
      "loss": 2.2127857208251953,
      "step": 6516
    },
    {
      "epoch": 0.6638116722086149,
      "grad_norm": 0.16223035752773285,
      "learning_rate": 0.000583981612446959,
      "loss": 2.2565813064575195,
      "step": 6518
    },
    {
      "epoch": 0.6640153579012226,
      "grad_norm": 0.11714797466993332,
      "learning_rate": 0.0005836280056577087,
      "loss": 2.273468017578125,
      "step": 6520
    },
    {
      "epoch": 0.6642190435938303,
      "grad_norm": 0.4852347671985626,
      "learning_rate": 0.0005832743988684583,
      "loss": 2.0566821098327637,
      "step": 6522
    },
    {
      "epoch": 0.6644227292864381,
      "grad_norm": 0.1535605937242508,
      "learning_rate": 0.0005829207920792079,
      "loss": 2.2503199577331543,
      "step": 6524
    },
    {
      "epoch": 0.6646264149790458,
      "grad_norm": 0.11933865398168564,
      "learning_rate": 0.0005825671852899575,
      "loss": 2.1197421550750732,
      "step": 6526
    },
    {
      "epoch": 0.6648301006716536,
      "grad_norm": 0.11563781648874283,
      "learning_rate": 0.0005822135785007072,
      "loss": 2.122138738632202,
      "step": 6528
    },
    {
      "epoch": 0.6650337863642614,
      "grad_norm": 0.12865518033504486,
      "learning_rate": 0.0005818599717114569,
      "loss": 2.1091251373291016,
      "step": 6530
    },
    {
      "epoch": 0.665237472056869,
      "grad_norm": 0.12114135921001434,
      "learning_rate": 0.0005815063649222066,
      "loss": 2.270144462585449,
      "step": 6532
    },
    {
      "epoch": 0.6654411577494768,
      "grad_norm": 0.14262057840824127,
      "learning_rate": 0.0005811527581329562,
      "loss": 2.119873046875,
      "step": 6534
    },
    {
      "epoch": 0.6656448434420845,
      "grad_norm": 0.10933762043714523,
      "learning_rate": 0.0005807991513437058,
      "loss": 2.1234047412872314,
      "step": 6536
    },
    {
      "epoch": 0.6658485291346923,
      "grad_norm": 0.12160074710845947,
      "learning_rate": 0.0005804455445544554,
      "loss": 2.1256518363952637,
      "step": 6538
    },
    {
      "epoch": 0.6660522148273,
      "grad_norm": 0.10659493505954742,
      "learning_rate": 0.000580091937765205,
      "loss": 2.206352710723877,
      "step": 6540
    },
    {
      "epoch": 0.6662559005199077,
      "grad_norm": 0.12685905396938324,
      "learning_rate": 0.0005797383309759548,
      "loss": 2.2779455184936523,
      "step": 6542
    },
    {
      "epoch": 0.6664595862125154,
      "grad_norm": 0.1141398698091507,
      "learning_rate": 0.0005793847241867044,
      "loss": 2.1719439029693604,
      "step": 6544
    },
    {
      "epoch": 0.6666632719051232,
      "grad_norm": 0.13413862884044647,
      "learning_rate": 0.0005790311173974541,
      "loss": 2.2580254077911377,
      "step": 6546
    },
    {
      "epoch": 0.666866957597731,
      "grad_norm": 0.1164148598909378,
      "learning_rate": 0.0005786775106082037,
      "loss": 2.181084156036377,
      "step": 6548
    },
    {
      "epoch": 0.6670706432903387,
      "grad_norm": 0.11869501322507858,
      "learning_rate": 0.0005783239038189534,
      "loss": 2.0936636924743652,
      "step": 6550
    },
    {
      "epoch": 0.6672743289829464,
      "grad_norm": 0.12283818423748016,
      "learning_rate": 0.0005779702970297029,
      "loss": 2.3596150875091553,
      "step": 6552
    },
    {
      "epoch": 0.6674780146755541,
      "grad_norm": 0.1308673769235611,
      "learning_rate": 0.0005776166902404526,
      "loss": 2.1611294746398926,
      "step": 6554
    },
    {
      "epoch": 0.6676817003681619,
      "grad_norm": 0.11719886213541031,
      "learning_rate": 0.0005772630834512023,
      "loss": 2.1986703872680664,
      "step": 6556
    },
    {
      "epoch": 0.6678853860607696,
      "grad_norm": 0.13262321054935455,
      "learning_rate": 0.0005769094766619519,
      "loss": 2.1901347637176514,
      "step": 6558
    },
    {
      "epoch": 0.6680890717533774,
      "grad_norm": 0.10917594283819199,
      "learning_rate": 0.0005765558698727016,
      "loss": 2.2495675086975098,
      "step": 6560
    },
    {
      "epoch": 0.6682927574459852,
      "grad_norm": 0.1289949119091034,
      "learning_rate": 0.0005762022630834512,
      "loss": 2.171152114868164,
      "step": 6562
    },
    {
      "epoch": 0.6684964431385928,
      "grad_norm": 0.12655465304851532,
      "learning_rate": 0.000575848656294201,
      "loss": 2.3222098350524902,
      "step": 6564
    },
    {
      "epoch": 0.6687001288312006,
      "grad_norm": 0.13164937496185303,
      "learning_rate": 0.0005754950495049505,
      "loss": 2.2147216796875,
      "step": 6566
    },
    {
      "epoch": 0.6689038145238083,
      "grad_norm": 0.14243540167808533,
      "learning_rate": 0.0005751414427157001,
      "loss": 2.0573272705078125,
      "step": 6568
    },
    {
      "epoch": 0.6691075002164161,
      "grad_norm": 0.11161145567893982,
      "learning_rate": 0.0005747878359264498,
      "loss": 2.1974034309387207,
      "step": 6570
    },
    {
      "epoch": 0.6693111859090238,
      "grad_norm": 0.16507673263549805,
      "learning_rate": 0.0005744342291371994,
      "loss": 2.227705955505371,
      "step": 6572
    },
    {
      "epoch": 0.6695148716016315,
      "grad_norm": 0.12704689800739288,
      "learning_rate": 0.0005740806223479491,
      "loss": 2.2133145332336426,
      "step": 6574
    },
    {
      "epoch": 0.6697185572942392,
      "grad_norm": 0.12253903597593307,
      "learning_rate": 0.0005737270155586988,
      "loss": 2.2653653621673584,
      "step": 6576
    },
    {
      "epoch": 0.669922242986847,
      "grad_norm": 0.1311369091272354,
      "learning_rate": 0.0005733734087694485,
      "loss": 2.234814167022705,
      "step": 6578
    },
    {
      "epoch": 0.6701259286794548,
      "grad_norm": 0.10396166890859604,
      "learning_rate": 0.000573019801980198,
      "loss": 2.08652400970459,
      "step": 6580
    },
    {
      "epoch": 0.6703296143720625,
      "grad_norm": 0.18537047505378723,
      "learning_rate": 0.0005726661951909476,
      "loss": 2.278042793273926,
      "step": 6582
    },
    {
      "epoch": 0.6705333000646702,
      "grad_norm": 0.11489458382129669,
      "learning_rate": 0.0005723125884016973,
      "loss": 2.303668975830078,
      "step": 6584
    },
    {
      "epoch": 0.6707369857572779,
      "grad_norm": 0.10374318808317184,
      "learning_rate": 0.0005719589816124469,
      "loss": 2.03580379486084,
      "step": 6586
    },
    {
      "epoch": 0.6709406714498857,
      "grad_norm": 0.15975044667720795,
      "learning_rate": 0.0005716053748231967,
      "loss": 2.226984977722168,
      "step": 6588
    },
    {
      "epoch": 0.6711443571424934,
      "grad_norm": 0.12064096331596375,
      "learning_rate": 0.0005712517680339463,
      "loss": 2.1514739990234375,
      "step": 6590
    },
    {
      "epoch": 0.6713480428351012,
      "grad_norm": 0.113404780626297,
      "learning_rate": 0.000570898161244696,
      "loss": 2.121080160140991,
      "step": 6592
    },
    {
      "epoch": 0.671551728527709,
      "grad_norm": 0.14040005207061768,
      "learning_rate": 0.0005705445544554455,
      "loss": 2.0831592082977295,
      "step": 6594
    },
    {
      "epoch": 0.6717554142203166,
      "grad_norm": 0.1487715244293213,
      "learning_rate": 0.0005701909476661951,
      "loss": 2.154799699783325,
      "step": 6596
    },
    {
      "epoch": 0.6719590999129244,
      "grad_norm": 0.12559741735458374,
      "learning_rate": 0.0005698373408769448,
      "loss": 2.196333408355713,
      "step": 6598
    },
    {
      "epoch": 0.6721627856055321,
      "grad_norm": 0.13625319302082062,
      "learning_rate": 0.0005694837340876945,
      "loss": 2.09314227104187,
      "step": 6600
    },
    {
      "epoch": 0.6723664712981399,
      "grad_norm": 0.1350695788860321,
      "learning_rate": 0.0005691301272984442,
      "loss": 2.243417501449585,
      "step": 6602
    },
    {
      "epoch": 0.6725701569907476,
      "grad_norm": 0.16215839982032776,
      "learning_rate": 0.0005687765205091938,
      "loss": 2.269503355026245,
      "step": 6604
    },
    {
      "epoch": 0.6727738426833553,
      "grad_norm": 0.14052601158618927,
      "learning_rate": 0.0005684229137199435,
      "loss": 2.0220835208892822,
      "step": 6606
    },
    {
      "epoch": 0.672977528375963,
      "grad_norm": 0.1260094940662384,
      "learning_rate": 0.000568069306930693,
      "loss": 2.1667070388793945,
      "step": 6608
    },
    {
      "epoch": 0.6731812140685708,
      "grad_norm": 0.1328732967376709,
      "learning_rate": 0.0005677157001414426,
      "loss": 2.1231536865234375,
      "step": 6610
    },
    {
      "epoch": 0.6733848997611785,
      "grad_norm": 0.14626561105251312,
      "learning_rate": 0.0005673620933521924,
      "loss": 2.093517303466797,
      "step": 6612
    },
    {
      "epoch": 0.6735885854537863,
      "grad_norm": 0.11670532822608948,
      "learning_rate": 0.000567008486562942,
      "loss": 2.2321834564208984,
      "step": 6614
    },
    {
      "epoch": 0.673792271146394,
      "grad_norm": 0.11348307132720947,
      "learning_rate": 0.0005666548797736917,
      "loss": 2.1260340213775635,
      "step": 6616
    },
    {
      "epoch": 0.6739959568390017,
      "grad_norm": 0.12412236630916595,
      "learning_rate": 0.0005663012729844413,
      "loss": 2.349534273147583,
      "step": 6618
    },
    {
      "epoch": 0.6741996425316095,
      "grad_norm": 0.15200240910053253,
      "learning_rate": 0.000565947666195191,
      "loss": 2.046051025390625,
      "step": 6620
    },
    {
      "epoch": 0.6744033282242172,
      "grad_norm": 0.13432519137859344,
      "learning_rate": 0.0005655940594059406,
      "loss": 2.144474506378174,
      "step": 6622
    },
    {
      "epoch": 0.674607013916825,
      "grad_norm": 0.1348087191581726,
      "learning_rate": 0.0005652404526166902,
      "loss": 2.0913615226745605,
      "step": 6624
    },
    {
      "epoch": 0.6748106996094326,
      "grad_norm": 0.11194302141666412,
      "learning_rate": 0.0005648868458274399,
      "loss": 2.2023816108703613,
      "step": 6626
    },
    {
      "epoch": 0.6750143853020404,
      "grad_norm": 0.1325799971818924,
      "learning_rate": 0.0005645332390381895,
      "loss": 2.054123640060425,
      "step": 6628
    },
    {
      "epoch": 0.6752180709946481,
      "grad_norm": 0.12262541800737381,
      "learning_rate": 0.0005641796322489392,
      "loss": 2.289379835128784,
      "step": 6630
    },
    {
      "epoch": 0.6754217566872559,
      "grad_norm": 0.10524732619524002,
      "learning_rate": 0.0005638260254596888,
      "loss": 2.09783935546875,
      "step": 6632
    },
    {
      "epoch": 0.6756254423798637,
      "grad_norm": 0.11849725246429443,
      "learning_rate": 0.0005634724186704386,
      "loss": 2.1308250427246094,
      "step": 6634
    },
    {
      "epoch": 0.6758291280724714,
      "grad_norm": 0.16306555271148682,
      "learning_rate": 0.0005631188118811881,
      "loss": 2.310553550720215,
      "step": 6636
    },
    {
      "epoch": 0.6760328137650791,
      "grad_norm": 0.14013512432575226,
      "learning_rate": 0.0005627652050919377,
      "loss": 2.247560977935791,
      "step": 6638
    },
    {
      "epoch": 0.6762364994576868,
      "grad_norm": 0.135228231549263,
      "learning_rate": 0.0005624115983026874,
      "loss": 2.227097988128662,
      "step": 6640
    },
    {
      "epoch": 0.6764401851502946,
      "grad_norm": 0.13446179032325745,
      "learning_rate": 0.000562057991513437,
      "loss": 2.208894729614258,
      "step": 6642
    },
    {
      "epoch": 0.6766438708429023,
      "grad_norm": 0.1198221817612648,
      "learning_rate": 0.0005617043847241868,
      "loss": 2.133399248123169,
      "step": 6644
    },
    {
      "epoch": 0.6768475565355101,
      "grad_norm": 0.1290067583322525,
      "learning_rate": 0.0005613507779349364,
      "loss": 2.093759536743164,
      "step": 6646
    },
    {
      "epoch": 0.6770512422281177,
      "grad_norm": 0.11179652810096741,
      "learning_rate": 0.0005609971711456861,
      "loss": 2.2451858520507812,
      "step": 6648
    },
    {
      "epoch": 0.6772549279207255,
      "grad_norm": 0.1475992053747177,
      "learning_rate": 0.0005606435643564357,
      "loss": 2.100156307220459,
      "step": 6650
    },
    {
      "epoch": 0.6774586136133333,
      "grad_norm": 0.1513018012046814,
      "learning_rate": 0.0005602899575671852,
      "loss": 2.1722660064697266,
      "step": 6652
    },
    {
      "epoch": 0.677662299305941,
      "grad_norm": 0.14956222474575043,
      "learning_rate": 0.0005599363507779349,
      "loss": 2.1785855293273926,
      "step": 6654
    },
    {
      "epoch": 0.6778659849985488,
      "grad_norm": 0.10434434562921524,
      "learning_rate": 0.0005595827439886846,
      "loss": 2.1852383613586426,
      "step": 6656
    },
    {
      "epoch": 0.6780696706911564,
      "grad_norm": 0.1263023316860199,
      "learning_rate": 0.0005592291371994343,
      "loss": 2.215329647064209,
      "step": 6658
    },
    {
      "epoch": 0.6782733563837642,
      "grad_norm": 0.10283675044775009,
      "learning_rate": 0.0005588755304101839,
      "loss": 2.1194419860839844,
      "step": 6660
    },
    {
      "epoch": 0.6784770420763719,
      "grad_norm": 0.11971983313560486,
      "learning_rate": 0.0005585219236209336,
      "loss": 2.196824550628662,
      "step": 6662
    },
    {
      "epoch": 0.6786807277689797,
      "grad_norm": 0.10413091629743576,
      "learning_rate": 0.0005581683168316832,
      "loss": 2.2233364582061768,
      "step": 6664
    },
    {
      "epoch": 0.6788844134615875,
      "grad_norm": 0.13281486928462982,
      "learning_rate": 0.0005578147100424327,
      "loss": 2.164261817932129,
      "step": 6666
    },
    {
      "epoch": 0.6790880991541952,
      "grad_norm": 0.13574522733688354,
      "learning_rate": 0.0005574611032531825,
      "loss": 2.1897754669189453,
      "step": 6668
    },
    {
      "epoch": 0.6792917848468029,
      "grad_norm": 0.1459997296333313,
      "learning_rate": 0.0005571074964639321,
      "loss": 2.2012763023376465,
      "step": 6670
    },
    {
      "epoch": 0.6794954705394106,
      "grad_norm": 0.1078740656375885,
      "learning_rate": 0.0005567538896746818,
      "loss": 2.158254384994507,
      "step": 6672
    },
    {
      "epoch": 0.6796991562320184,
      "grad_norm": 0.12663856148719788,
      "learning_rate": 0.0005564002828854314,
      "loss": 2.074051856994629,
      "step": 6674
    },
    {
      "epoch": 0.6799028419246261,
      "grad_norm": 0.156760111451149,
      "learning_rate": 0.0005560466760961811,
      "loss": 2.137676954269409,
      "step": 6676
    },
    {
      "epoch": 0.6801065276172339,
      "grad_norm": 0.1275133192539215,
      "learning_rate": 0.0005556930693069308,
      "loss": 2.251425266265869,
      "step": 6678
    },
    {
      "epoch": 0.6803102133098415,
      "grad_norm": 0.1379198580980301,
      "learning_rate": 0.0005553394625176803,
      "loss": 2.1907591819763184,
      "step": 6680
    },
    {
      "epoch": 0.6805138990024493,
      "grad_norm": 0.132883220911026,
      "learning_rate": 0.00055498585572843,
      "loss": 2.198934555053711,
      "step": 6682
    },
    {
      "epoch": 0.680717584695057,
      "grad_norm": 0.14394374191761017,
      "learning_rate": 0.0005546322489391796,
      "loss": 2.263066291809082,
      "step": 6684
    },
    {
      "epoch": 0.6809212703876648,
      "grad_norm": 0.10799068957567215,
      "learning_rate": 0.0005542786421499293,
      "loss": 2.100717782974243,
      "step": 6686
    },
    {
      "epoch": 0.6811249560802726,
      "grad_norm": 0.13743162155151367,
      "learning_rate": 0.0005539250353606789,
      "loss": 2.3152527809143066,
      "step": 6688
    },
    {
      "epoch": 0.6813286417728802,
      "grad_norm": 0.12321525067090988,
      "learning_rate": 0.0005535714285714287,
      "loss": 2.186284065246582,
      "step": 6690
    },
    {
      "epoch": 0.681532327465488,
      "grad_norm": 0.11392373591661453,
      "learning_rate": 0.0005532178217821783,
      "loss": 2.280444860458374,
      "step": 6692
    },
    {
      "epoch": 0.6817360131580957,
      "grad_norm": 0.16040953993797302,
      "learning_rate": 0.0005528642149929278,
      "loss": 2.215158462524414,
      "step": 6694
    },
    {
      "epoch": 0.6819396988507035,
      "grad_norm": 0.1149672418832779,
      "learning_rate": 0.0005525106082036775,
      "loss": 2.2340517044067383,
      "step": 6696
    },
    {
      "epoch": 0.6821433845433112,
      "grad_norm": 0.1260398030281067,
      "learning_rate": 0.0005521570014144271,
      "loss": 2.1147336959838867,
      "step": 6698
    },
    {
      "epoch": 0.682347070235919,
      "grad_norm": 0.12939248979091644,
      "learning_rate": 0.0005518033946251768,
      "loss": 2.08651065826416,
      "step": 6700
    },
    {
      "epoch": 0.6825507559285267,
      "grad_norm": 0.11006031930446625,
      "learning_rate": 0.0005514497878359265,
      "loss": 2.1740167140960693,
      "step": 6702
    },
    {
      "epoch": 0.6827544416211344,
      "grad_norm": 0.10850805789232254,
      "learning_rate": 0.0005510961810466762,
      "loss": 2.165470600128174,
      "step": 6704
    },
    {
      "epoch": 0.6829581273137422,
      "grad_norm": 0.12513421475887299,
      "learning_rate": 0.0005507425742574258,
      "loss": 2.2967214584350586,
      "step": 6706
    },
    {
      "epoch": 0.6831618130063499,
      "grad_norm": 0.14173394441604614,
      "learning_rate": 0.0005503889674681753,
      "loss": 2.222811698913574,
      "step": 6708
    },
    {
      "epoch": 0.6833654986989577,
      "grad_norm": 0.13794700801372528,
      "learning_rate": 0.000550035360678925,
      "loss": 2.265748977661133,
      "step": 6710
    },
    {
      "epoch": 0.6835691843915653,
      "grad_norm": 0.1817992478609085,
      "learning_rate": 0.0005496817538896746,
      "loss": 2.2276451587677,
      "step": 6712
    },
    {
      "epoch": 0.6837728700841731,
      "grad_norm": 0.12601616978645325,
      "learning_rate": 0.0005493281471004244,
      "loss": 2.303058624267578,
      "step": 6714
    },
    {
      "epoch": 0.6839765557767808,
      "grad_norm": 0.12855610251426697,
      "learning_rate": 0.000548974540311174,
      "loss": 2.2546474933624268,
      "step": 6716
    },
    {
      "epoch": 0.6841802414693886,
      "grad_norm": 0.10457655787467957,
      "learning_rate": 0.0005486209335219237,
      "loss": 2.2778005599975586,
      "step": 6718
    },
    {
      "epoch": 0.6843839271619964,
      "grad_norm": 0.11468460410833359,
      "learning_rate": 0.0005482673267326733,
      "loss": 2.166349411010742,
      "step": 6720
    },
    {
      "epoch": 0.684587612854604,
      "grad_norm": 0.1298896223306656,
      "learning_rate": 0.0005479137199434228,
      "loss": 2.145139694213867,
      "step": 6722
    },
    {
      "epoch": 0.6847912985472118,
      "grad_norm": 0.11265810579061508,
      "learning_rate": 0.0005475601131541726,
      "loss": 2.049501657485962,
      "step": 6724
    },
    {
      "epoch": 0.6849949842398195,
      "grad_norm": 0.11239831149578094,
      "learning_rate": 0.0005472065063649222,
      "loss": 2.2235913276672363,
      "step": 6726
    },
    {
      "epoch": 0.6851986699324273,
      "grad_norm": 0.11931256204843521,
      "learning_rate": 0.0005468528995756719,
      "loss": 2.15226411819458,
      "step": 6728
    },
    {
      "epoch": 0.685402355625035,
      "grad_norm": 0.12729452550411224,
      "learning_rate": 0.0005464992927864215,
      "loss": 2.222428321838379,
      "step": 6730
    },
    {
      "epoch": 0.6856060413176428,
      "grad_norm": 0.12783776223659515,
      "learning_rate": 0.0005461456859971712,
      "loss": 2.1212165355682373,
      "step": 6732
    },
    {
      "epoch": 0.6858097270102504,
      "grad_norm": 0.17432455718517303,
      "learning_rate": 0.0005457920792079208,
      "loss": 2.1301302909851074,
      "step": 6734
    },
    {
      "epoch": 0.6860134127028582,
      "grad_norm": 0.11008235812187195,
      "learning_rate": 0.0005454384724186703,
      "loss": 2.0820789337158203,
      "step": 6736
    },
    {
      "epoch": 0.686217098395466,
      "grad_norm": 0.11058143526315689,
      "learning_rate": 0.0005450848656294201,
      "loss": 2.2084619998931885,
      "step": 6738
    },
    {
      "epoch": 0.6864207840880737,
      "grad_norm": 0.13979071378707886,
      "learning_rate": 0.0005447312588401697,
      "loss": 2.144663095474243,
      "step": 6740
    },
    {
      "epoch": 0.6866244697806815,
      "grad_norm": 0.12404558807611465,
      "learning_rate": 0.0005443776520509194,
      "loss": 2.1864492893218994,
      "step": 6742
    },
    {
      "epoch": 0.6868281554732891,
      "grad_norm": 0.11094963550567627,
      "learning_rate": 0.000544024045261669,
      "loss": 2.0284128189086914,
      "step": 6744
    },
    {
      "epoch": 0.6870318411658969,
      "grad_norm": 0.13939952850341797,
      "learning_rate": 0.0005436704384724188,
      "loss": 2.2857885360717773,
      "step": 6746
    },
    {
      "epoch": 0.6872355268585046,
      "grad_norm": 0.1267445981502533,
      "learning_rate": 0.0005433168316831684,
      "loss": 2.2768096923828125,
      "step": 6748
    },
    {
      "epoch": 0.6874392125511124,
      "grad_norm": 0.12709426879882812,
      "learning_rate": 0.000542963224893918,
      "loss": 2.1641266345977783,
      "step": 6750
    },
    {
      "epoch": 0.6876428982437202,
      "grad_norm": 0.13848792016506195,
      "learning_rate": 0.0005426096181046676,
      "loss": 2.249333381652832,
      "step": 6752
    },
    {
      "epoch": 0.6878465839363278,
      "grad_norm": 0.12109556049108505,
      "learning_rate": 0.0005422560113154172,
      "loss": 2.0525009632110596,
      "step": 6754
    },
    {
      "epoch": 0.6880502696289356,
      "grad_norm": 0.1202191486954689,
      "learning_rate": 0.0005419024045261669,
      "loss": 2.0405631065368652,
      "step": 6756
    },
    {
      "epoch": 0.6882539553215433,
      "grad_norm": 0.10964738577604294,
      "learning_rate": 0.0005415487977369165,
      "loss": 2.0808825492858887,
      "step": 6758
    },
    {
      "epoch": 0.6884576410141511,
      "grad_norm": 0.11084166914224625,
      "learning_rate": 0.0005411951909476663,
      "loss": 2.147836208343506,
      "step": 6760
    },
    {
      "epoch": 0.6886613267067588,
      "grad_norm": 0.11232597380876541,
      "learning_rate": 0.0005408415841584159,
      "loss": 2.2039971351623535,
      "step": 6762
    },
    {
      "epoch": 0.6888650123993665,
      "grad_norm": 0.11466162651777267,
      "learning_rate": 0.0005404879773691655,
      "loss": 2.2470755577087402,
      "step": 6764
    },
    {
      "epoch": 0.6890686980919742,
      "grad_norm": 0.11326336115598679,
      "learning_rate": 0.0005401343705799151,
      "loss": 2.209963321685791,
      "step": 6766
    },
    {
      "epoch": 0.689272383784582,
      "grad_norm": 0.1510191708803177,
      "learning_rate": 0.0005397807637906647,
      "loss": 2.25813364982605,
      "step": 6768
    },
    {
      "epoch": 0.6894760694771898,
      "grad_norm": 0.11930445581674576,
      "learning_rate": 0.0005394271570014145,
      "loss": 2.30900239944458,
      "step": 6770
    },
    {
      "epoch": 0.6896797551697975,
      "grad_norm": 0.11531718820333481,
      "learning_rate": 0.0005390735502121641,
      "loss": 2.116684913635254,
      "step": 6772
    },
    {
      "epoch": 0.6898834408624053,
      "grad_norm": 0.11620393395423889,
      "learning_rate": 0.0005387199434229138,
      "loss": 2.312725067138672,
      "step": 6774
    },
    {
      "epoch": 0.6900871265550129,
      "grad_norm": 0.12310457229614258,
      "learning_rate": 0.0005383663366336634,
      "loss": 2.262577533721924,
      "step": 6776
    },
    {
      "epoch": 0.6902908122476207,
      "grad_norm": 0.10666100680828094,
      "learning_rate": 0.000538012729844413,
      "loss": 2.2063493728637695,
      "step": 6778
    },
    {
      "epoch": 0.6904944979402284,
      "grad_norm": 0.1439969837665558,
      "learning_rate": 0.0005376591230551626,
      "loss": 2.2763195037841797,
      "step": 6780
    },
    {
      "epoch": 0.6906981836328362,
      "grad_norm": 0.12671814858913422,
      "learning_rate": 0.0005373055162659123,
      "loss": 2.250988483428955,
      "step": 6782
    },
    {
      "epoch": 0.690901869325444,
      "grad_norm": 0.09972506761550903,
      "learning_rate": 0.000536951909476662,
      "loss": 1.820725679397583,
      "step": 6784
    },
    {
      "epoch": 0.6911055550180516,
      "grad_norm": 0.12202069163322449,
      "learning_rate": 0.0005365983026874116,
      "loss": 2.1470184326171875,
      "step": 6786
    },
    {
      "epoch": 0.6913092407106594,
      "grad_norm": 0.13062098622322083,
      "learning_rate": 0.0005362446958981613,
      "loss": 2.18206787109375,
      "step": 6788
    },
    {
      "epoch": 0.6915129264032671,
      "grad_norm": 0.13292749226093292,
      "learning_rate": 0.0005358910891089109,
      "loss": 2.2037277221679688,
      "step": 6790
    },
    {
      "epoch": 0.6917166120958749,
      "grad_norm": 0.12543079257011414,
      "learning_rate": 0.0005355374823196605,
      "loss": 2.078394889831543,
      "step": 6792
    },
    {
      "epoch": 0.6919202977884826,
      "grad_norm": 0.1503777951002121,
      "learning_rate": 0.0005351838755304102,
      "loss": 2.229627847671509,
      "step": 6794
    },
    {
      "epoch": 0.6921239834810903,
      "grad_norm": 0.12184897810220718,
      "learning_rate": 0.0005348302687411598,
      "loss": 2.27811861038208,
      "step": 6796
    },
    {
      "epoch": 0.692327669173698,
      "grad_norm": 0.1325872391462326,
      "learning_rate": 0.0005344766619519095,
      "loss": 2.281785011291504,
      "step": 6798
    },
    {
      "epoch": 0.6925313548663058,
      "grad_norm": 0.1461370885372162,
      "learning_rate": 0.0005341230551626591,
      "loss": 2.2570953369140625,
      "step": 6800
    },
    {
      "epoch": 0.6927350405589136,
      "grad_norm": 0.1253480762243271,
      "learning_rate": 0.0005337694483734088,
      "loss": 2.0168187618255615,
      "step": 6802
    },
    {
      "epoch": 0.6929387262515213,
      "grad_norm": 0.14210952818393707,
      "learning_rate": 0.0005334158415841585,
      "loss": 2.200817584991455,
      "step": 6804
    },
    {
      "epoch": 0.6931424119441291,
      "grad_norm": 0.10086695849895477,
      "learning_rate": 0.0005330622347949081,
      "loss": 2.1379098892211914,
      "step": 6806
    },
    {
      "epoch": 0.6933460976367367,
      "grad_norm": 0.09976791590452194,
      "learning_rate": 0.0005327086280056577,
      "loss": 2.224630832672119,
      "step": 6808
    },
    {
      "epoch": 0.6935497833293445,
      "grad_norm": 0.11090613156557083,
      "learning_rate": 0.0005323550212164073,
      "loss": 2.1224312782287598,
      "step": 6810
    },
    {
      "epoch": 0.6937534690219522,
      "grad_norm": 0.12469596415758133,
      "learning_rate": 0.000532001414427157,
      "loss": 2.207042694091797,
      "step": 6812
    },
    {
      "epoch": 0.69395715471456,
      "grad_norm": 0.10464311391115189,
      "learning_rate": 0.0005316478076379066,
      "loss": 2.14520001411438,
      "step": 6814
    },
    {
      "epoch": 0.6941608404071677,
      "grad_norm": 0.10965576022863388,
      "learning_rate": 0.0005312942008486564,
      "loss": 2.1586925983428955,
      "step": 6816
    },
    {
      "epoch": 0.6943645260997754,
      "grad_norm": 0.11370889097452164,
      "learning_rate": 0.000530940594059406,
      "loss": 2.3265748023986816,
      "step": 6818
    },
    {
      "epoch": 0.6945682117923832,
      "grad_norm": 0.11593182384967804,
      "learning_rate": 0.0005305869872701556,
      "loss": 2.2026352882385254,
      "step": 6820
    },
    {
      "epoch": 0.6947718974849909,
      "grad_norm": 0.14425961673259735,
      "learning_rate": 0.0005302333804809052,
      "loss": 2.1614785194396973,
      "step": 6822
    },
    {
      "epoch": 0.6949755831775987,
      "grad_norm": 0.11645737290382385,
      "learning_rate": 0.0005298797736916548,
      "loss": 2.2242069244384766,
      "step": 6824
    },
    {
      "epoch": 0.6951792688702064,
      "grad_norm": 0.15144312381744385,
      "learning_rate": 0.0005295261669024045,
      "loss": 2.1054575443267822,
      "step": 6826
    },
    {
      "epoch": 0.6953829545628141,
      "grad_norm": 0.14532288908958435,
      "learning_rate": 0.0005291725601131542,
      "loss": 2.2070953845977783,
      "step": 6828
    },
    {
      "epoch": 0.6955866402554218,
      "grad_norm": 0.11884139478206635,
      "learning_rate": 0.0005288189533239039,
      "loss": 2.2070672512054443,
      "step": 6830
    },
    {
      "epoch": 0.6957903259480296,
      "grad_norm": 0.15934845805168152,
      "learning_rate": 0.0005284653465346535,
      "loss": 2.096407413482666,
      "step": 6832
    },
    {
      "epoch": 0.6959940116406373,
      "grad_norm": 0.1483648121356964,
      "learning_rate": 0.0005281117397454031,
      "loss": 2.300933599472046,
      "step": 6834
    },
    {
      "epoch": 0.6961976973332451,
      "grad_norm": 0.13565392792224884,
      "learning_rate": 0.0005277581329561527,
      "loss": 2.224341869354248,
      "step": 6836
    },
    {
      "epoch": 0.6964013830258529,
      "grad_norm": 0.10340264439582825,
      "learning_rate": 0.0005274045261669023,
      "loss": 2.1174023151397705,
      "step": 6838
    },
    {
      "epoch": 0.6966050687184605,
      "grad_norm": 0.12272170186042786,
      "learning_rate": 0.0005270509193776521,
      "loss": 2.269014358520508,
      "step": 6840
    },
    {
      "epoch": 0.6968087544110683,
      "grad_norm": 0.1081172302365303,
      "learning_rate": 0.0005266973125884017,
      "loss": 2.221181869506836,
      "step": 6842
    },
    {
      "epoch": 0.697012440103676,
      "grad_norm": 0.13140040636062622,
      "learning_rate": 0.0005263437057991514,
      "loss": 2.2864701747894287,
      "step": 6844
    },
    {
      "epoch": 0.6972161257962838,
      "grad_norm": 0.11849227547645569,
      "learning_rate": 0.000525990099009901,
      "loss": 2.1338229179382324,
      "step": 6846
    },
    {
      "epoch": 0.6974198114888915,
      "grad_norm": 0.13458551466464996,
      "learning_rate": 0.0005256364922206506,
      "loss": 2.099022388458252,
      "step": 6848
    },
    {
      "epoch": 0.6976234971814992,
      "grad_norm": 0.1349409520626068,
      "learning_rate": 0.0005252828854314004,
      "loss": 2.15263032913208,
      "step": 6850
    },
    {
      "epoch": 0.6978271828741069,
      "grad_norm": 0.14674067497253418,
      "learning_rate": 0.0005249292786421499,
      "loss": 2.260681629180908,
      "step": 6852
    },
    {
      "epoch": 0.6980308685667147,
      "grad_norm": 0.1135706827044487,
      "learning_rate": 0.0005245756718528996,
      "loss": 2.156937599182129,
      "step": 6854
    },
    {
      "epoch": 0.6982345542593225,
      "grad_norm": 0.1222739964723587,
      "learning_rate": 0.0005242220650636492,
      "loss": 2.1285228729248047,
      "step": 6856
    },
    {
      "epoch": 0.6984382399519302,
      "grad_norm": 0.11722546815872192,
      "learning_rate": 0.0005238684582743989,
      "loss": 2.1365976333618164,
      "step": 6858
    },
    {
      "epoch": 0.6986419256445379,
      "grad_norm": 0.11624598503112793,
      "learning_rate": 0.0005235148514851485,
      "loss": 2.1651082038879395,
      "step": 6860
    },
    {
      "epoch": 0.6988456113371456,
      "grad_norm": 0.11327168345451355,
      "learning_rate": 0.0005231612446958982,
      "loss": 2.082371950149536,
      "step": 6862
    },
    {
      "epoch": 0.6990492970297534,
      "grad_norm": 0.1273626983165741,
      "learning_rate": 0.0005228076379066479,
      "loss": 2.1960766315460205,
      "step": 6864
    },
    {
      "epoch": 0.6992529827223611,
      "grad_norm": 0.10699215531349182,
      "learning_rate": 0.0005224540311173974,
      "loss": 2.0987491607666016,
      "step": 6866
    },
    {
      "epoch": 0.6994566684149689,
      "grad_norm": 0.1506192833185196,
      "learning_rate": 0.0005221004243281471,
      "loss": 2.146350860595703,
      "step": 6868
    },
    {
      "epoch": 0.6996603541075765,
      "grad_norm": 0.1318938136100769,
      "learning_rate": 0.0005217468175388967,
      "loss": 2.260868549346924,
      "step": 6870
    },
    {
      "epoch": 0.6998640398001843,
      "grad_norm": 0.13125275075435638,
      "learning_rate": 0.0005213932107496465,
      "loss": 2.1559906005859375,
      "step": 6872
    },
    {
      "epoch": 0.7000677254927921,
      "grad_norm": 0.11935018748044968,
      "learning_rate": 0.0005210396039603961,
      "loss": 2.1525304317474365,
      "step": 6874
    },
    {
      "epoch": 0.7002714111853998,
      "grad_norm": 0.10396468639373779,
      "learning_rate": 0.0005206859971711457,
      "loss": 2.1302080154418945,
      "step": 6876
    },
    {
      "epoch": 0.7004750968780076,
      "grad_norm": 0.10967627167701721,
      "learning_rate": 0.0005203323903818954,
      "loss": 2.2471625804901123,
      "step": 6878
    },
    {
      "epoch": 0.7006787825706153,
      "grad_norm": 0.13004520535469055,
      "learning_rate": 0.0005199787835926449,
      "loss": 2.201538324356079,
      "step": 6880
    },
    {
      "epoch": 0.700882468263223,
      "grad_norm": 0.11840129643678665,
      "learning_rate": 0.0005196251768033946,
      "loss": 2.200690269470215,
      "step": 6882
    },
    {
      "epoch": 0.7010861539558307,
      "grad_norm": 0.11906684935092926,
      "learning_rate": 0.0005192715700141443,
      "loss": 2.1921639442443848,
      "step": 6884
    },
    {
      "epoch": 0.7012898396484385,
      "grad_norm": 0.10488682985305786,
      "learning_rate": 0.000518917963224894,
      "loss": 2.286344289779663,
      "step": 6886
    },
    {
      "epoch": 0.7014935253410463,
      "grad_norm": 0.13285717368125916,
      "learning_rate": 0.0005185643564356436,
      "loss": 2.1521122455596924,
      "step": 6888
    },
    {
      "epoch": 0.701697211033654,
      "grad_norm": 0.13174422085285187,
      "learning_rate": 0.0005182107496463932,
      "loss": 2.196108818054199,
      "step": 6890
    },
    {
      "epoch": 0.7019008967262617,
      "grad_norm": 0.13051487505435944,
      "learning_rate": 0.0005178571428571429,
      "loss": 2.194432497024536,
      "step": 6892
    },
    {
      "epoch": 0.7021045824188694,
      "grad_norm": 0.12321194261312485,
      "learning_rate": 0.0005175035360678924,
      "loss": 2.3285269737243652,
      "step": 6894
    },
    {
      "epoch": 0.7023082681114772,
      "grad_norm": 0.12682370841503143,
      "learning_rate": 0.0005171499292786422,
      "loss": 2.0157248973846436,
      "step": 6896
    },
    {
      "epoch": 0.7025119538040849,
      "grad_norm": 0.10864172130823135,
      "learning_rate": 0.0005167963224893918,
      "loss": 2.182844638824463,
      "step": 6898
    },
    {
      "epoch": 0.7027156394966927,
      "grad_norm": 0.13734708726406097,
      "learning_rate": 0.0005164427157001415,
      "loss": 2.3467068672180176,
      "step": 6900
    },
    {
      "epoch": 0.7029193251893003,
      "grad_norm": 0.1108136996626854,
      "learning_rate": 0.0005160891089108911,
      "loss": 2.0998494625091553,
      "step": 6902
    },
    {
      "epoch": 0.7031230108819081,
      "grad_norm": 0.13328956067562103,
      "learning_rate": 0.0005157355021216407,
      "loss": 2.285036563873291,
      "step": 6904
    },
    {
      "epoch": 0.7033266965745159,
      "grad_norm": 0.13540716469287872,
      "learning_rate": 0.0005153818953323905,
      "loss": 2.2303614616394043,
      "step": 6906
    },
    {
      "epoch": 0.7035303822671236,
      "grad_norm": 0.11679026484489441,
      "learning_rate": 0.00051502828854314,
      "loss": 2.271360397338867,
      "step": 6908
    },
    {
      "epoch": 0.7037340679597314,
      "grad_norm": 0.12904059886932373,
      "learning_rate": 0.0005146746817538897,
      "loss": 2.180288791656494,
      "step": 6910
    },
    {
      "epoch": 0.7039377536523391,
      "grad_norm": 0.12192552536725998,
      "learning_rate": 0.0005143210749646393,
      "loss": 2.2354555130004883,
      "step": 6912
    },
    {
      "epoch": 0.7041414393449468,
      "grad_norm": 0.12568314373493195,
      "learning_rate": 0.000513967468175389,
      "loss": 2.1784815788269043,
      "step": 6914
    },
    {
      "epoch": 0.7043451250375545,
      "grad_norm": 0.11430572718381882,
      "learning_rate": 0.0005136138613861386,
      "loss": 2.2153565883636475,
      "step": 6916
    },
    {
      "epoch": 0.7045488107301623,
      "grad_norm": 0.11258040368556976,
      "learning_rate": 0.0005132602545968883,
      "loss": 2.158749580383301,
      "step": 6918
    },
    {
      "epoch": 0.70475249642277,
      "grad_norm": 0.11168136447668076,
      "learning_rate": 0.000512906647807638,
      "loss": 2.1535043716430664,
      "step": 6920
    },
    {
      "epoch": 0.7049561821153778,
      "grad_norm": 0.12356610596179962,
      "learning_rate": 0.0005125530410183875,
      "loss": 2.1780519485473633,
      "step": 6922
    },
    {
      "epoch": 0.7051598678079855,
      "grad_norm": 0.12887756526470184,
      "learning_rate": 0.0005121994342291372,
      "loss": 2.246436595916748,
      "step": 6924
    },
    {
      "epoch": 0.7053635535005932,
      "grad_norm": 0.1220330148935318,
      "learning_rate": 0.0005118458274398868,
      "loss": 2.117661476135254,
      "step": 6926
    },
    {
      "epoch": 0.705567239193201,
      "grad_norm": 0.1210666075348854,
      "learning_rate": 0.0005114922206506365,
      "loss": 2.1994590759277344,
      "step": 6928
    },
    {
      "epoch": 0.7057709248858087,
      "grad_norm": 0.11968952417373657,
      "learning_rate": 0.0005111386138613862,
      "loss": 2.279005527496338,
      "step": 6930
    },
    {
      "epoch": 0.7059746105784165,
      "grad_norm": 0.12168609350919724,
      "learning_rate": 0.0005107850070721358,
      "loss": 2.187859535217285,
      "step": 6932
    },
    {
      "epoch": 0.7061782962710241,
      "grad_norm": 0.1239394024014473,
      "learning_rate": 0.0005104314002828855,
      "loss": 2.2158749103546143,
      "step": 6934
    },
    {
      "epoch": 0.7063819819636319,
      "grad_norm": 0.14086398482322693,
      "learning_rate": 0.000510077793493635,
      "loss": 1.967200517654419,
      "step": 6936
    },
    {
      "epoch": 0.7065856676562396,
      "grad_norm": 0.13051286339759827,
      "learning_rate": 0.0005097241867043847,
      "loss": 2.17700457572937,
      "step": 6938
    },
    {
      "epoch": 0.7067893533488474,
      "grad_norm": 0.11586655676364899,
      "learning_rate": 0.0005093705799151343,
      "loss": 2.1640360355377197,
      "step": 6940
    },
    {
      "epoch": 0.7069930390414552,
      "grad_norm": 0.1330806016921997,
      "learning_rate": 0.0005090169731258841,
      "loss": 2.3092732429504395,
      "step": 6942
    },
    {
      "epoch": 0.7071967247340629,
      "grad_norm": 0.12130452692508698,
      "learning_rate": 0.0005086633663366337,
      "loss": 2.341068744659424,
      "step": 6944
    },
    {
      "epoch": 0.7074004104266706,
      "grad_norm": 0.14402107894420624,
      "learning_rate": 0.0005083097595473833,
      "loss": 2.077707290649414,
      "step": 6946
    },
    {
      "epoch": 0.7076040961192783,
      "grad_norm": 0.14220306277275085,
      "learning_rate": 0.000507956152758133,
      "loss": 2.104092597961426,
      "step": 6948
    },
    {
      "epoch": 0.7078077818118861,
      "grad_norm": 0.13854025304317474,
      "learning_rate": 0.0005076025459688826,
      "loss": 2.0620651245117188,
      "step": 6950
    },
    {
      "epoch": 0.7080114675044938,
      "grad_norm": 0.17001405358314514,
      "learning_rate": 0.0005072489391796323,
      "loss": 2.254617691040039,
      "step": 6952
    },
    {
      "epoch": 0.7082151531971016,
      "grad_norm": 0.11951429396867752,
      "learning_rate": 0.0005068953323903819,
      "loss": 2.1004600524902344,
      "step": 6954
    },
    {
      "epoch": 0.7084188388897092,
      "grad_norm": 0.14726273715496063,
      "learning_rate": 0.0005065417256011316,
      "loss": 2.2137393951416016,
      "step": 6956
    },
    {
      "epoch": 0.708622524582317,
      "grad_norm": 0.1123536229133606,
      "learning_rate": 0.0005061881188118812,
      "loss": 2.137648582458496,
      "step": 6958
    },
    {
      "epoch": 0.7088262102749248,
      "grad_norm": 0.14448675513267517,
      "learning_rate": 0.0005058345120226308,
      "loss": 2.0487186908721924,
      "step": 6960
    },
    {
      "epoch": 0.7090298959675325,
      "grad_norm": 0.11007431149482727,
      "learning_rate": 0.0005054809052333805,
      "loss": 2.102475166320801,
      "step": 6962
    },
    {
      "epoch": 0.7092335816601403,
      "grad_norm": 0.14165034890174866,
      "learning_rate": 0.0005051272984441302,
      "loss": 2.3335328102111816,
      "step": 6964
    },
    {
      "epoch": 0.7094372673527479,
      "grad_norm": 0.12197605520486832,
      "learning_rate": 0.0005047736916548798,
      "loss": 2.2052712440490723,
      "step": 6966
    },
    {
      "epoch": 0.7096409530453557,
      "grad_norm": 0.10166604071855545,
      "learning_rate": 0.0005044200848656294,
      "loss": 2.1714093685150146,
      "step": 6968
    },
    {
      "epoch": 0.7098446387379634,
      "grad_norm": 0.10901937633752823,
      "learning_rate": 0.0005040664780763791,
      "loss": 2.2511954307556152,
      "step": 6970
    },
    {
      "epoch": 0.7100483244305712,
      "grad_norm": 0.1339273601770401,
      "learning_rate": 0.0005037128712871287,
      "loss": 2.08305025100708,
      "step": 6972
    },
    {
      "epoch": 0.710252010123179,
      "grad_norm": 0.1623830795288086,
      "learning_rate": 0.0005033592644978783,
      "loss": 2.1462132930755615,
      "step": 6974
    },
    {
      "epoch": 0.7104556958157867,
      "grad_norm": 0.1284138560295105,
      "learning_rate": 0.0005030056577086281,
      "loss": 2.213456869125366,
      "step": 6976
    },
    {
      "epoch": 0.7106593815083944,
      "grad_norm": 0.10798266530036926,
      "learning_rate": 0.0005026520509193777,
      "loss": 2.2416625022888184,
      "step": 6978
    },
    {
      "epoch": 0.7108630672010021,
      "grad_norm": 0.13622315227985382,
      "learning_rate": 0.0005022984441301273,
      "loss": 2.2664809226989746,
      "step": 6980
    },
    {
      "epoch": 0.7110667528936099,
      "grad_norm": 0.11397131532430649,
      "learning_rate": 0.0005019448373408769,
      "loss": 2.212125301361084,
      "step": 6982
    },
    {
      "epoch": 0.7112704385862176,
      "grad_norm": 0.13327623903751373,
      "learning_rate": 0.0005015912305516266,
      "loss": 2.300661563873291,
      "step": 6984
    },
    {
      "epoch": 0.7114741242788254,
      "grad_norm": 0.12121587246656418,
      "learning_rate": 0.0005012376237623762,
      "loss": 2.154974937438965,
      "step": 6986
    },
    {
      "epoch": 0.711677809971433,
      "grad_norm": 0.1465798169374466,
      "learning_rate": 0.0005008840169731259,
      "loss": 2.24436616897583,
      "step": 6988
    },
    {
      "epoch": 0.7118814956640408,
      "grad_norm": 0.16294050216674805,
      "learning_rate": 0.0005005304101838756,
      "loss": 2.247267961502075,
      "step": 6990
    },
    {
      "epoch": 0.7120851813566486,
      "grad_norm": 0.1436827927827835,
      "learning_rate": 0.0005001768033946252,
      "loss": 2.0917468070983887,
      "step": 6992
    },
    {
      "epoch": 0.7122888670492563,
      "grad_norm": 0.11639072000980377,
      "learning_rate": 0.0004998231966053748,
      "loss": 2.1337637901306152,
      "step": 6994
    },
    {
      "epoch": 0.7124925527418641,
      "grad_norm": 0.17277725040912628,
      "learning_rate": 0.0004994695898161245,
      "loss": 2.135268211364746,
      "step": 6996
    },
    {
      "epoch": 0.7126962384344717,
      "grad_norm": 0.1544792652130127,
      "learning_rate": 0.0004991159830268742,
      "loss": 2.161101818084717,
      "step": 6998
    },
    {
      "epoch": 0.7128999241270795,
      "grad_norm": 0.10412278026342392,
      "learning_rate": 0.0004987623762376238,
      "loss": 2.2153496742248535,
      "step": 7000
    },
    {
      "epoch": 0.7131036098196872,
      "grad_norm": 0.1178407073020935,
      "learning_rate": 0.0004984087694483734,
      "loss": 2.2387804985046387,
      "step": 7002
    },
    {
      "epoch": 0.713307295512295,
      "grad_norm": 0.11781677603721619,
      "learning_rate": 0.000498055162659123,
      "loss": 2.341035842895508,
      "step": 7004
    },
    {
      "epoch": 0.7135109812049028,
      "grad_norm": 0.11440229415893555,
      "learning_rate": 0.0004977015558698727,
      "loss": 2.118098735809326,
      "step": 7006
    },
    {
      "epoch": 0.7137146668975104,
      "grad_norm": 0.14138607680797577,
      "learning_rate": 0.0004973479490806223,
      "loss": 2.1691908836364746,
      "step": 7008
    },
    {
      "epoch": 0.7139183525901182,
      "grad_norm": 0.12438711524009705,
      "learning_rate": 0.0004969943422913721,
      "loss": 2.3250412940979004,
      "step": 7010
    },
    {
      "epoch": 0.7141220382827259,
      "grad_norm": 0.12304583191871643,
      "learning_rate": 0.0004966407355021217,
      "loss": 2.22184419631958,
      "step": 7012
    },
    {
      "epoch": 0.7143257239753337,
      "grad_norm": 0.11633428931236267,
      "learning_rate": 0.0004962871287128713,
      "loss": 2.061891794204712,
      "step": 7014
    },
    {
      "epoch": 0.7145294096679414,
      "grad_norm": 0.12756654620170593,
      "learning_rate": 0.0004959335219236209,
      "loss": 2.1992828845977783,
      "step": 7016
    },
    {
      "epoch": 0.7147330953605492,
      "grad_norm": 0.10544373840093613,
      "learning_rate": 0.0004955799151343705,
      "loss": 2.1591296195983887,
      "step": 7018
    },
    {
      "epoch": 0.7149367810531568,
      "grad_norm": 0.11059703677892685,
      "learning_rate": 0.0004952263083451202,
      "loss": 2.17386531829834,
      "step": 7020
    },
    {
      "epoch": 0.7151404667457646,
      "grad_norm": 0.11344355344772339,
      "learning_rate": 0.0004948727015558699,
      "loss": 2.1254782676696777,
      "step": 7022
    },
    {
      "epoch": 0.7153441524383723,
      "grad_norm": 0.11303359270095825,
      "learning_rate": 0.0004945190947666196,
      "loss": 2.2117433547973633,
      "step": 7024
    },
    {
      "epoch": 0.7155478381309801,
      "grad_norm": 0.11579477787017822,
      "learning_rate": 0.0004941654879773692,
      "loss": 2.1964292526245117,
      "step": 7026
    },
    {
      "epoch": 0.7157515238235879,
      "grad_norm": 0.13907431066036224,
      "learning_rate": 0.0004938118811881188,
      "loss": 2.2176661491394043,
      "step": 7028
    },
    {
      "epoch": 0.7159552095161955,
      "grad_norm": 0.12066128104925156,
      "learning_rate": 0.0004934582743988684,
      "loss": 2.2213096618652344,
      "step": 7030
    },
    {
      "epoch": 0.7161588952088033,
      "grad_norm": 0.13106687366962433,
      "learning_rate": 0.0004931046676096182,
      "loss": 2.0997073650360107,
      "step": 7032
    },
    {
      "epoch": 0.716362580901411,
      "grad_norm": 0.11424015462398529,
      "learning_rate": 0.0004927510608203678,
      "loss": 2.193410873413086,
      "step": 7034
    },
    {
      "epoch": 0.7165662665940188,
      "grad_norm": 0.1493590921163559,
      "learning_rate": 0.0004923974540311174,
      "loss": 2.2846996784210205,
      "step": 7036
    },
    {
      "epoch": 0.7167699522866265,
      "grad_norm": 0.1178729459643364,
      "learning_rate": 0.0004920438472418671,
      "loss": 2.0571224689483643,
      "step": 7038
    },
    {
      "epoch": 0.7169736379792342,
      "grad_norm": 0.12649530172348022,
      "learning_rate": 0.0004916902404526167,
      "loss": 2.23508882522583,
      "step": 7040
    },
    {
      "epoch": 0.717177323671842,
      "grad_norm": 0.10769683867692947,
      "learning_rate": 0.0004913366336633663,
      "loss": 2.0882158279418945,
      "step": 7042
    },
    {
      "epoch": 0.7173810093644497,
      "grad_norm": 0.11791504174470901,
      "learning_rate": 0.000490983026874116,
      "loss": 2.132289409637451,
      "step": 7044
    },
    {
      "epoch": 0.7175846950570575,
      "grad_norm": 0.10153969377279282,
      "learning_rate": 0.0004906294200848657,
      "loss": 2.084827423095703,
      "step": 7046
    },
    {
      "epoch": 0.7177883807496652,
      "grad_norm": 0.12384435534477234,
      "learning_rate": 0.0004902758132956153,
      "loss": 2.2569081783294678,
      "step": 7048
    },
    {
      "epoch": 0.717992066442273,
      "grad_norm": 0.10688748210668564,
      "learning_rate": 0.0004899222065063649,
      "loss": 2.290553569793701,
      "step": 7050
    },
    {
      "epoch": 0.7181957521348806,
      "grad_norm": 0.10259644687175751,
      "learning_rate": 0.0004895685997171146,
      "loss": 2.1048638820648193,
      "step": 7052
    },
    {
      "epoch": 0.7183994378274884,
      "grad_norm": 0.10575117915868759,
      "learning_rate": 0.0004892149929278642,
      "loss": 2.086376667022705,
      "step": 7054
    },
    {
      "epoch": 0.7186031235200961,
      "grad_norm": 0.14850200712680817,
      "learning_rate": 0.0004888613861386139,
      "loss": 2.1660053730010986,
      "step": 7056
    },
    {
      "epoch": 0.7188068092127039,
      "grad_norm": 0.11399565637111664,
      "learning_rate": 0.0004885077793493635,
      "loss": 2.2702999114990234,
      "step": 7058
    },
    {
      "epoch": 0.7190104949053117,
      "grad_norm": 0.14110171794891357,
      "learning_rate": 0.0004881541725601132,
      "loss": 2.2832415103912354,
      "step": 7060
    },
    {
      "epoch": 0.7192141805979193,
      "grad_norm": 0.12176508456468582,
      "learning_rate": 0.0004878005657708628,
      "loss": 2.2140321731567383,
      "step": 7062
    },
    {
      "epoch": 0.7194178662905271,
      "grad_norm": 0.13676300644874573,
      "learning_rate": 0.00048744695898161243,
      "loss": 2.1871821880340576,
      "step": 7064
    },
    {
      "epoch": 0.7196215519831348,
      "grad_norm": 0.11305111646652222,
      "learning_rate": 0.0004870933521923621,
      "loss": 2.2487239837646484,
      "step": 7066
    },
    {
      "epoch": 0.7198252376757426,
      "grad_norm": 0.14864030480384827,
      "learning_rate": 0.0004867397454031117,
      "loss": 2.2602970600128174,
      "step": 7068
    },
    {
      "epoch": 0.7200289233683503,
      "grad_norm": 0.13264355063438416,
      "learning_rate": 0.0004863861386138614,
      "loss": 2.0703766345977783,
      "step": 7070
    },
    {
      "epoch": 0.720232609060958,
      "grad_norm": 0.13812260329723358,
      "learning_rate": 0.00048603253182461105,
      "loss": 2.1952476501464844,
      "step": 7072
    },
    {
      "epoch": 0.7204362947535657,
      "grad_norm": 0.13004131615161896,
      "learning_rate": 0.0004856789250353607,
      "loss": 2.0939459800720215,
      "step": 7074
    },
    {
      "epoch": 0.7206399804461735,
      "grad_norm": 0.12985344231128693,
      "learning_rate": 0.00048532531824611034,
      "loss": 2.278099536895752,
      "step": 7076
    },
    {
      "epoch": 0.7208436661387813,
      "grad_norm": 0.12438429147005081,
      "learning_rate": 0.00048497171145685995,
      "loss": 2.318068265914917,
      "step": 7078
    },
    {
      "epoch": 0.721047351831389,
      "grad_norm": 0.1187286451458931,
      "learning_rate": 0.0004846181046676096,
      "loss": 2.2220163345336914,
      "step": 7080
    },
    {
      "epoch": 0.7212510375239968,
      "grad_norm": 0.12190424650907516,
      "learning_rate": 0.0004842644978783593,
      "loss": 2.2009873390197754,
      "step": 7082
    },
    {
      "epoch": 0.7214547232166044,
      "grad_norm": 0.1090736836194992,
      "learning_rate": 0.0004839108910891089,
      "loss": 2.095428466796875,
      "step": 7084
    },
    {
      "epoch": 0.7216584089092122,
      "grad_norm": 0.1165740117430687,
      "learning_rate": 0.0004835572842998586,
      "loss": 2.1530308723449707,
      "step": 7086
    },
    {
      "epoch": 0.7218620946018199,
      "grad_norm": 0.13559509813785553,
      "learning_rate": 0.00048320367751060825,
      "loss": 2.1633596420288086,
      "step": 7088
    },
    {
      "epoch": 0.7220657802944277,
      "grad_norm": 0.11289870738983154,
      "learning_rate": 0.00048285007072135786,
      "loss": 2.127248764038086,
      "step": 7090
    },
    {
      "epoch": 0.7222694659870355,
      "grad_norm": 0.10785143822431564,
      "learning_rate": 0.0004824964639321075,
      "loss": 2.265580415725708,
      "step": 7092
    },
    {
      "epoch": 0.7224731516796431,
      "grad_norm": 0.13236449658870697,
      "learning_rate": 0.00048214285714285715,
      "loss": 2.3245160579681396,
      "step": 7094
    },
    {
      "epoch": 0.7226768373722509,
      "grad_norm": 0.12247749418020248,
      "learning_rate": 0.0004817892503536068,
      "loss": 2.2694199085235596,
      "step": 7096
    },
    {
      "epoch": 0.7228805230648586,
      "grad_norm": 0.14264729619026184,
      "learning_rate": 0.00048143564356435643,
      "loss": 2.2861008644104004,
      "step": 7098
    },
    {
      "epoch": 0.7230842087574664,
      "grad_norm": 0.1297949254512787,
      "learning_rate": 0.0004810820367751061,
      "loss": 2.161975622177124,
      "step": 7100
    },
    {
      "epoch": 0.7232878944500741,
      "grad_norm": 0.12222341448068619,
      "learning_rate": 0.00048072842998585577,
      "loss": 2.1518025398254395,
      "step": 7102
    },
    {
      "epoch": 0.7234915801426818,
      "grad_norm": 0.10791588574647903,
      "learning_rate": 0.0004803748231966054,
      "loss": 2.191044807434082,
      "step": 7104
    },
    {
      "epoch": 0.7236952658352895,
      "grad_norm": 0.1126575693488121,
      "learning_rate": 0.000480021216407355,
      "loss": 2.06609845161438,
      "step": 7106
    },
    {
      "epoch": 0.7238989515278973,
      "grad_norm": 0.11554937064647675,
      "learning_rate": 0.00047966760961810467,
      "loss": 2.13643217086792,
      "step": 7108
    },
    {
      "epoch": 0.724102637220505,
      "grad_norm": 0.10772731155157089,
      "learning_rate": 0.00047931400282885434,
      "loss": 2.1242775917053223,
      "step": 7110
    },
    {
      "epoch": 0.7243063229131128,
      "grad_norm": 0.12580707669258118,
      "learning_rate": 0.00047896039603960395,
      "loss": 2.1768360137939453,
      "step": 7112
    },
    {
      "epoch": 0.7245100086057206,
      "grad_norm": 0.12318222969770432,
      "learning_rate": 0.0004786067892503536,
      "loss": 2.1703805923461914,
      "step": 7114
    },
    {
      "epoch": 0.7247136942983282,
      "grad_norm": 0.10905760526657104,
      "learning_rate": 0.0004782531824611033,
      "loss": 2.2191622257232666,
      "step": 7116
    },
    {
      "epoch": 0.724917379990936,
      "grad_norm": 0.12346996366977692,
      "learning_rate": 0.0004778995756718529,
      "loss": 2.1143085956573486,
      "step": 7118
    },
    {
      "epoch": 0.7251210656835437,
      "grad_norm": 0.11048220098018646,
      "learning_rate": 0.0004775459688826025,
      "loss": 2.124051809310913,
      "step": 7120
    },
    {
      "epoch": 0.7253247513761515,
      "grad_norm": 0.13477946817874908,
      "learning_rate": 0.0004771923620933522,
      "loss": 2.049523115158081,
      "step": 7122
    },
    {
      "epoch": 0.7255284370687592,
      "grad_norm": 0.10880770534276962,
      "learning_rate": 0.00047683875530410186,
      "loss": 2.165774345397949,
      "step": 7124
    },
    {
      "epoch": 0.7257321227613669,
      "grad_norm": 0.13520145416259766,
      "learning_rate": 0.0004764851485148515,
      "loss": 2.243072032928467,
      "step": 7126
    },
    {
      "epoch": 0.7259358084539747,
      "grad_norm": 0.10832219570875168,
      "learning_rate": 0.00047613154172560114,
      "loss": 2.269648313522339,
      "step": 7128
    },
    {
      "epoch": 0.7261394941465824,
      "grad_norm": 0.13039006292819977,
      "learning_rate": 0.0004757779349363508,
      "loss": 2.1435000896453857,
      "step": 7130
    },
    {
      "epoch": 0.7263431798391902,
      "grad_norm": 0.1539987176656723,
      "learning_rate": 0.0004754243281471005,
      "loss": 2.2977187633514404,
      "step": 7132
    },
    {
      "epoch": 0.7265468655317979,
      "grad_norm": 0.14365899562835693,
      "learning_rate": 0.00047507072135785004,
      "loss": 2.1962013244628906,
      "step": 7134
    },
    {
      "epoch": 0.7267505512244056,
      "grad_norm": 0.12068725377321243,
      "learning_rate": 0.0004747171145685997,
      "loss": 2.265542984008789,
      "step": 7136
    },
    {
      "epoch": 0.7269542369170133,
      "grad_norm": 0.1352309286594391,
      "learning_rate": 0.0004743635077793494,
      "loss": 2.153000593185425,
      "step": 7138
    },
    {
      "epoch": 0.7271579226096211,
      "grad_norm": 0.10219433158636093,
      "learning_rate": 0.000474009900990099,
      "loss": 2.047924518585205,
      "step": 7140
    },
    {
      "epoch": 0.7273616083022288,
      "grad_norm": 0.11700855940580368,
      "learning_rate": 0.00047365629420084867,
      "loss": 2.188159942626953,
      "step": 7142
    },
    {
      "epoch": 0.7275652939948366,
      "grad_norm": 0.13516391813755035,
      "learning_rate": 0.00047330268741159834,
      "loss": 2.1916911602020264,
      "step": 7144
    },
    {
      "epoch": 0.7277689796874443,
      "grad_norm": 0.1262577474117279,
      "learning_rate": 0.000472949080622348,
      "loss": 2.184720993041992,
      "step": 7146
    },
    {
      "epoch": 0.727972665380052,
      "grad_norm": 0.10985517501831055,
      "learning_rate": 0.00047259547383309757,
      "loss": 2.222505569458008,
      "step": 7148
    },
    {
      "epoch": 0.7281763510726598,
      "grad_norm": 0.15139977633953094,
      "learning_rate": 0.00047224186704384724,
      "loss": 2.124080181121826,
      "step": 7150
    },
    {
      "epoch": 0.7283800367652675,
      "grad_norm": 0.15406298637390137,
      "learning_rate": 0.0004718882602545969,
      "loss": 2.158275604248047,
      "step": 7152
    },
    {
      "epoch": 0.7285837224578753,
      "grad_norm": 0.20447224378585815,
      "learning_rate": 0.0004715346534653465,
      "loss": 2.251479148864746,
      "step": 7154
    },
    {
      "epoch": 0.728787408150483,
      "grad_norm": 0.11569373309612274,
      "learning_rate": 0.0004711810466760962,
      "loss": 2.278141498565674,
      "step": 7156
    },
    {
      "epoch": 0.7289910938430907,
      "grad_norm": 0.10532794892787933,
      "learning_rate": 0.00047082743988684586,
      "loss": 2.0274252891540527,
      "step": 7158
    },
    {
      "epoch": 0.7291947795356984,
      "grad_norm": 0.1218039020895958,
      "learning_rate": 0.00047047383309759553,
      "loss": 2.183382511138916,
      "step": 7160
    },
    {
      "epoch": 0.7293984652283062,
      "grad_norm": 0.13361816108226776,
      "learning_rate": 0.0004701202263083451,
      "loss": 2.0631232261657715,
      "step": 7162
    },
    {
      "epoch": 0.729602150920914,
      "grad_norm": 0.1175391897559166,
      "learning_rate": 0.00046976661951909476,
      "loss": 2.2810473442077637,
      "step": 7164
    },
    {
      "epoch": 0.7298058366135217,
      "grad_norm": 0.11012241244316101,
      "learning_rate": 0.00046941301272984443,
      "loss": 2.227750539779663,
      "step": 7166
    },
    {
      "epoch": 0.7300095223061294,
      "grad_norm": 0.11332917213439941,
      "learning_rate": 0.00046905940594059404,
      "loss": 2.153705358505249,
      "step": 7168
    },
    {
      "epoch": 0.7302132079987371,
      "grad_norm": 0.10746357589960098,
      "learning_rate": 0.0004687057991513437,
      "loss": 2.1796658039093018,
      "step": 7170
    },
    {
      "epoch": 0.7304168936913449,
      "grad_norm": 0.12665142118930817,
      "learning_rate": 0.0004683521923620934,
      "loss": 2.338369369506836,
      "step": 7172
    },
    {
      "epoch": 0.7306205793839526,
      "grad_norm": 0.12800605595111847,
      "learning_rate": 0.00046799858557284305,
      "loss": 2.263645887374878,
      "step": 7174
    },
    {
      "epoch": 0.7308242650765604,
      "grad_norm": 0.11213421821594238,
      "learning_rate": 0.0004676449787835926,
      "loss": 2.124521017074585,
      "step": 7176
    },
    {
      "epoch": 0.731027950769168,
      "grad_norm": 0.11961778253316879,
      "learning_rate": 0.0004672913719943423,
      "loss": 2.1657209396362305,
      "step": 7178
    },
    {
      "epoch": 0.7312316364617758,
      "grad_norm": 0.14346137642860413,
      "learning_rate": 0.00046693776520509195,
      "loss": 2.3094279766082764,
      "step": 7180
    },
    {
      "epoch": 0.7314353221543836,
      "grad_norm": 0.1244010254740715,
      "learning_rate": 0.0004665841584158416,
      "loss": 2.2332985401153564,
      "step": 7182
    },
    {
      "epoch": 0.7316390078469913,
      "grad_norm": 0.11637738347053528,
      "learning_rate": 0.00046623055162659123,
      "loss": 2.174863338470459,
      "step": 7184
    },
    {
      "epoch": 0.7318426935395991,
      "grad_norm": 0.13326050341129303,
      "learning_rate": 0.0004658769448373409,
      "loss": 2.1995720863342285,
      "step": 7186
    },
    {
      "epoch": 0.7320463792322068,
      "grad_norm": 0.11370902508497238,
      "learning_rate": 0.00046552333804809057,
      "loss": 2.2718095779418945,
      "step": 7188
    },
    {
      "epoch": 0.7322500649248145,
      "grad_norm": 0.13393689692020416,
      "learning_rate": 0.00046516973125884013,
      "loss": 2.246312141418457,
      "step": 7190
    },
    {
      "epoch": 0.7324537506174222,
      "grad_norm": 0.11491590738296509,
      "learning_rate": 0.0004648161244695898,
      "loss": 2.137793779373169,
      "step": 7192
    },
    {
      "epoch": 0.73265743631003,
      "grad_norm": 0.15885479748249054,
      "learning_rate": 0.00046446251768033947,
      "loss": 2.1160168647766113,
      "step": 7194
    },
    {
      "epoch": 0.7328611220026378,
      "grad_norm": 0.10194899886846542,
      "learning_rate": 0.00046410891089108914,
      "loss": 2.0920462608337402,
      "step": 7196
    },
    {
      "epoch": 0.7330648076952455,
      "grad_norm": 0.11147967725992203,
      "learning_rate": 0.00046375530410183876,
      "loss": 2.178867816925049,
      "step": 7198
    },
    {
      "epoch": 0.7332684933878532,
      "grad_norm": 0.118509441614151,
      "learning_rate": 0.0004634016973125884,
      "loss": 2.0628676414489746,
      "step": 7200
    },
    {
      "epoch": 0.7334721790804609,
      "grad_norm": 0.1170046329498291,
      "learning_rate": 0.0004630480905233381,
      "loss": 2.1773734092712402,
      "step": 7202
    },
    {
      "epoch": 0.7336758647730687,
      "grad_norm": 0.17506229877471924,
      "learning_rate": 0.00046269448373408766,
      "loss": 2.2308316230773926,
      "step": 7204
    },
    {
      "epoch": 0.7338795504656764,
      "grad_norm": 0.165228009223938,
      "learning_rate": 0.0004623408769448373,
      "loss": 2.1990795135498047,
      "step": 7206
    },
    {
      "epoch": 0.7340832361582842,
      "grad_norm": 0.11517639458179474,
      "learning_rate": 0.000461987270155587,
      "loss": 2.1720337867736816,
      "step": 7208
    },
    {
      "epoch": 0.7342869218508918,
      "grad_norm": 0.16199757158756256,
      "learning_rate": 0.00046163366336633666,
      "loss": 2.1234359741210938,
      "step": 7210
    },
    {
      "epoch": 0.7344906075434996,
      "grad_norm": 0.1322079449892044,
      "learning_rate": 0.0004612800565770863,
      "loss": 2.068535804748535,
      "step": 7212
    },
    {
      "epoch": 0.7346942932361074,
      "grad_norm": 0.11777742207050323,
      "learning_rate": 0.00046092644978783595,
      "loss": 2.117947816848755,
      "step": 7214
    },
    {
      "epoch": 0.7348979789287151,
      "grad_norm": 0.12734408676624298,
      "learning_rate": 0.0004605728429985856,
      "loss": 2.0927577018737793,
      "step": 7216
    },
    {
      "epoch": 0.7351016646213229,
      "grad_norm": 0.10726716369390488,
      "learning_rate": 0.0004602192362093352,
      "loss": 2.2249088287353516,
      "step": 7218
    },
    {
      "epoch": 0.7353053503139306,
      "grad_norm": 0.12114494293928146,
      "learning_rate": 0.00045986562942008485,
      "loss": 2.3604941368103027,
      "step": 7220
    },
    {
      "epoch": 0.7355090360065383,
      "grad_norm": 0.1141815185546875,
      "learning_rate": 0.0004595120226308345,
      "loss": 2.291379928588867,
      "step": 7222
    },
    {
      "epoch": 0.735712721699146,
      "grad_norm": 0.12462089955806732,
      "learning_rate": 0.0004591584158415842,
      "loss": 2.1920008659362793,
      "step": 7224
    },
    {
      "epoch": 0.7359164073917538,
      "grad_norm": 0.1273108422756195,
      "learning_rate": 0.0004588048090523338,
      "loss": 2.329928398132324,
      "step": 7226
    },
    {
      "epoch": 0.7361200930843615,
      "grad_norm": 0.11966310441493988,
      "learning_rate": 0.00045845120226308347,
      "loss": 2.2020182609558105,
      "step": 7228
    },
    {
      "epoch": 0.7363237787769693,
      "grad_norm": 0.11796797066926956,
      "learning_rate": 0.00045809759547383314,
      "loss": 2.1978845596313477,
      "step": 7230
    },
    {
      "epoch": 0.736527464469577,
      "grad_norm": 0.1325305998325348,
      "learning_rate": 0.00045774398868458276,
      "loss": 2.0491514205932617,
      "step": 7232
    },
    {
      "epoch": 0.7367311501621847,
      "grad_norm": 0.11030648648738861,
      "learning_rate": 0.00045739038189533237,
      "loss": 2.146531820297241,
      "step": 7234
    },
    {
      "epoch": 0.7369348358547925,
      "grad_norm": 0.10756353288888931,
      "learning_rate": 0.00045703677510608204,
      "loss": 2.265913486480713,
      "step": 7236
    },
    {
      "epoch": 0.7371385215474002,
      "grad_norm": 0.14712779223918915,
      "learning_rate": 0.0004566831683168317,
      "loss": 2.2468481063842773,
      "step": 7238
    },
    {
      "epoch": 0.737342207240008,
      "grad_norm": 0.11845679581165314,
      "learning_rate": 0.0004563295615275813,
      "loss": 2.114072799682617,
      "step": 7240
    },
    {
      "epoch": 0.7375458929326156,
      "grad_norm": 0.12973470985889435,
      "learning_rate": 0.000455975954738331,
      "loss": 2.0625159740448,
      "step": 7242
    },
    {
      "epoch": 0.7377495786252234,
      "grad_norm": 0.1029026210308075,
      "learning_rate": 0.00045562234794908066,
      "loss": 2.0892293453216553,
      "step": 7244
    },
    {
      "epoch": 0.7379532643178311,
      "grad_norm": 0.11782599985599518,
      "learning_rate": 0.0004552687411598303,
      "loss": 2.0024781227111816,
      "step": 7246
    },
    {
      "epoch": 0.7381569500104389,
      "grad_norm": 0.11654394865036011,
      "learning_rate": 0.0004549151343705799,
      "loss": 2.2230732440948486,
      "step": 7248
    },
    {
      "epoch": 0.7383606357030467,
      "grad_norm": 0.12037082761526108,
      "learning_rate": 0.00045456152758132956,
      "loss": 2.2383041381835938,
      "step": 7250
    },
    {
      "epoch": 0.7385643213956543,
      "grad_norm": 0.10866861790418625,
      "learning_rate": 0.00045420792079207923,
      "loss": 2.2433395385742188,
      "step": 7252
    },
    {
      "epoch": 0.7387680070882621,
      "grad_norm": 0.11474991589784622,
      "learning_rate": 0.00045385431400282885,
      "loss": 2.164694309234619,
      "step": 7254
    },
    {
      "epoch": 0.7389716927808698,
      "grad_norm": 0.12148222327232361,
      "learning_rate": 0.0004535007072135785,
      "loss": 2.2003796100616455,
      "step": 7256
    },
    {
      "epoch": 0.7391753784734776,
      "grad_norm": 0.11337953805923462,
      "learning_rate": 0.0004531471004243282,
      "loss": 2.21478271484375,
      "step": 7258
    },
    {
      "epoch": 0.7393790641660853,
      "grad_norm": 0.10550428926944733,
      "learning_rate": 0.0004527934936350778,
      "loss": 2.2526681423187256,
      "step": 7260
    },
    {
      "epoch": 0.7395827498586931,
      "grad_norm": 0.12159799039363861,
      "learning_rate": 0.0004524398868458274,
      "loss": 2.0609195232391357,
      "step": 7262
    },
    {
      "epoch": 0.7397864355513007,
      "grad_norm": 0.13437095284461975,
      "learning_rate": 0.0004520862800565771,
      "loss": 2.2763326168060303,
      "step": 7264
    },
    {
      "epoch": 0.7399901212439085,
      "grad_norm": 0.11303786933422089,
      "learning_rate": 0.00045173267326732675,
      "loss": 2.2615270614624023,
      "step": 7266
    },
    {
      "epoch": 0.7401938069365163,
      "grad_norm": 0.14116591215133667,
      "learning_rate": 0.00045137906647807637,
      "loss": 2.208930015563965,
      "step": 7268
    },
    {
      "epoch": 0.740397492629124,
      "grad_norm": 0.10280372947454453,
      "learning_rate": 0.00045102545968882604,
      "loss": 2.0703845024108887,
      "step": 7270
    },
    {
      "epoch": 0.7406011783217318,
      "grad_norm": 0.12233562767505646,
      "learning_rate": 0.0004506718528995757,
      "loss": 2.2962920665740967,
      "step": 7272
    },
    {
      "epoch": 0.7408048640143394,
      "grad_norm": 0.11534818261861801,
      "learning_rate": 0.0004503182461103253,
      "loss": 2.077782154083252,
      "step": 7274
    },
    {
      "epoch": 0.7410085497069472,
      "grad_norm": 0.1281866878271103,
      "learning_rate": 0.00044996463932107494,
      "loss": 2.273240089416504,
      "step": 7276
    },
    {
      "epoch": 0.7412122353995549,
      "grad_norm": 0.13504695892333984,
      "learning_rate": 0.0004496110325318246,
      "loss": 2.1387414932250977,
      "step": 7278
    },
    {
      "epoch": 0.7414159210921627,
      "grad_norm": 0.12348013371229172,
      "learning_rate": 0.0004492574257425743,
      "loss": 2.11075758934021,
      "step": 7280
    },
    {
      "epoch": 0.7416196067847705,
      "grad_norm": 0.15741267800331116,
      "learning_rate": 0.00044890381895332395,
      "loss": 2.1937255859375,
      "step": 7282
    },
    {
      "epoch": 0.7418232924773781,
      "grad_norm": 0.15440931916236877,
      "learning_rate": 0.00044855021216407356,
      "loss": 2.0940775871276855,
      "step": 7284
    },
    {
      "epoch": 0.7420269781699859,
      "grad_norm": 0.11438805609941483,
      "learning_rate": 0.00044819660537482323,
      "loss": 2.256411552429199,
      "step": 7286
    },
    {
      "epoch": 0.7422306638625936,
      "grad_norm": 0.1551392525434494,
      "learning_rate": 0.00044784299858557285,
      "loss": 2.247454881668091,
      "step": 7288
    },
    {
      "epoch": 0.7424343495552014,
      "grad_norm": 0.11763888597488403,
      "learning_rate": 0.00044748939179632246,
      "loss": 2.0576279163360596,
      "step": 7290
    },
    {
      "epoch": 0.7426380352478091,
      "grad_norm": 0.1602763682603836,
      "learning_rate": 0.00044713578500707213,
      "loss": 2.209932327270508,
      "step": 7292
    },
    {
      "epoch": 0.7428417209404169,
      "grad_norm": 0.09371227771043777,
      "learning_rate": 0.0004467821782178218,
      "loss": 2.1280276775360107,
      "step": 7294
    },
    {
      "epoch": 0.7430454066330245,
      "grad_norm": 0.11199456453323364,
      "learning_rate": 0.00044642857142857147,
      "loss": 2.224663257598877,
      "step": 7296
    },
    {
      "epoch": 0.7432490923256323,
      "grad_norm": 0.13807111978530884,
      "learning_rate": 0.0004460749646393211,
      "loss": 2.091377019882202,
      "step": 7298
    },
    {
      "epoch": 0.7434527780182401,
      "grad_norm": 0.11320603638887405,
      "learning_rate": 0.00044572135785007075,
      "loss": 2.1508383750915527,
      "step": 7300
    },
    {
      "epoch": 0.7436564637108478,
      "grad_norm": 0.11144614964723587,
      "learning_rate": 0.00044536775106082037,
      "loss": 2.143433094024658,
      "step": 7302
    },
    {
      "epoch": 0.7438601494034556,
      "grad_norm": 0.11699941009283066,
      "learning_rate": 0.00044501414427157,
      "loss": 2.124083995819092,
      "step": 7304
    },
    {
      "epoch": 0.7440638350960632,
      "grad_norm": 0.10632345825433731,
      "learning_rate": 0.00044466053748231965,
      "loss": 2.0535824298858643,
      "step": 7306
    },
    {
      "epoch": 0.744267520788671,
      "grad_norm": 0.11475978046655655,
      "learning_rate": 0.0004443069306930693,
      "loss": 2.2596375942230225,
      "step": 7308
    },
    {
      "epoch": 0.7444712064812787,
      "grad_norm": 0.11019923537969589,
      "learning_rate": 0.000443953323903819,
      "loss": 2.0679681301116943,
      "step": 7310
    },
    {
      "epoch": 0.7446748921738865,
      "grad_norm": 0.1460941582918167,
      "learning_rate": 0.0004435997171145686,
      "loss": 2.166365146636963,
      "step": 7312
    },
    {
      "epoch": 0.7448785778664943,
      "grad_norm": 0.11354739964008331,
      "learning_rate": 0.0004432461103253183,
      "loss": 2.1577391624450684,
      "step": 7314
    },
    {
      "epoch": 0.7450822635591019,
      "grad_norm": 0.1457887589931488,
      "learning_rate": 0.0004428925035360679,
      "loss": 2.1072070598602295,
      "step": 7316
    },
    {
      "epoch": 0.7452859492517097,
      "grad_norm": 0.10286836326122284,
      "learning_rate": 0.0004425388967468175,
      "loss": 2.2254176139831543,
      "step": 7318
    },
    {
      "epoch": 0.7454896349443174,
      "grad_norm": 0.11904450505971909,
      "learning_rate": 0.0004421852899575672,
      "loss": 2.1901261806488037,
      "step": 7320
    },
    {
      "epoch": 0.7456933206369252,
      "grad_norm": 0.11921921372413635,
      "learning_rate": 0.00044183168316831684,
      "loss": 2.1913692951202393,
      "step": 7322
    },
    {
      "epoch": 0.7458970063295329,
      "grad_norm": 0.11820491403341293,
      "learning_rate": 0.0004414780763790665,
      "loss": 2.1933579444885254,
      "step": 7324
    },
    {
      "epoch": 0.7461006920221407,
      "grad_norm": 0.15072855353355408,
      "learning_rate": 0.00044112446958981613,
      "loss": 2.186699390411377,
      "step": 7326
    },
    {
      "epoch": 0.7463043777147483,
      "grad_norm": 0.1401892751455307,
      "learning_rate": 0.0004407708628005658,
      "loss": 2.302140235900879,
      "step": 7328
    },
    {
      "epoch": 0.7465080634073561,
      "grad_norm": 0.12423475086688995,
      "learning_rate": 0.0004404172560113154,
      "loss": 2.1404850482940674,
      "step": 7330
    },
    {
      "epoch": 0.7467117490999639,
      "grad_norm": 0.13450407981872559,
      "learning_rate": 0.0004400636492220651,
      "loss": 2.157552719116211,
      "step": 7332
    },
    {
      "epoch": 0.7469154347925716,
      "grad_norm": 0.13404616713523865,
      "learning_rate": 0.0004397100424328147,
      "loss": 2.123631000518799,
      "step": 7334
    },
    {
      "epoch": 0.7471191204851794,
      "grad_norm": 0.12365002930164337,
      "learning_rate": 0.00043935643564356437,
      "loss": 2.204796314239502,
      "step": 7336
    },
    {
      "epoch": 0.747322806177787,
      "grad_norm": 0.11499237269163132,
      "learning_rate": 0.00043900282885431404,
      "loss": 2.2845616340637207,
      "step": 7338
    },
    {
      "epoch": 0.7475264918703948,
      "grad_norm": 0.14364288747310638,
      "learning_rate": 0.00043864922206506365,
      "loss": 2.250135898590088,
      "step": 7340
    },
    {
      "epoch": 0.7477301775630025,
      "grad_norm": 0.11465809494256973,
      "learning_rate": 0.0004382956152758133,
      "loss": 2.146016836166382,
      "step": 7342
    },
    {
      "epoch": 0.7479338632556103,
      "grad_norm": 0.11732444167137146,
      "learning_rate": 0.00043794200848656294,
      "loss": 2.1014628410339355,
      "step": 7344
    },
    {
      "epoch": 0.748137548948218,
      "grad_norm": 0.11173725128173828,
      "learning_rate": 0.0004375884016973126,
      "loss": 2.085916042327881,
      "step": 7346
    },
    {
      "epoch": 0.7483412346408257,
      "grad_norm": 0.12824851274490356,
      "learning_rate": 0.0004372347949080622,
      "loss": 1.8141976594924927,
      "step": 7348
    },
    {
      "epoch": 0.7485449203334335,
      "grad_norm": 0.10794989764690399,
      "learning_rate": 0.0004368811881188119,
      "loss": 2.1259522438049316,
      "step": 7350
    },
    {
      "epoch": 0.7487486060260412,
      "grad_norm": 0.1125539094209671,
      "learning_rate": 0.00043652758132956156,
      "loss": 2.2020225524902344,
      "step": 7352
    },
    {
      "epoch": 0.748952291718649,
      "grad_norm": 0.1437269002199173,
      "learning_rate": 0.0004361739745403112,
      "loss": 2.3557262420654297,
      "step": 7354
    },
    {
      "epoch": 0.7491559774112567,
      "grad_norm": 0.1123492494225502,
      "learning_rate": 0.00043582036775106084,
      "loss": 2.241939067840576,
      "step": 7356
    },
    {
      "epoch": 0.7493596631038645,
      "grad_norm": 0.1291377991437912,
      "learning_rate": 0.00043546676096181046,
      "loss": 2.366107940673828,
      "step": 7358
    },
    {
      "epoch": 0.7495633487964721,
      "grad_norm": 0.12234491109848022,
      "learning_rate": 0.00043511315417256013,
      "loss": 2.08597993850708,
      "step": 7360
    },
    {
      "epoch": 0.7497670344890799,
      "grad_norm": 0.1162547618150711,
      "learning_rate": 0.00043475954738330974,
      "loss": 2.2991538047790527,
      "step": 7362
    },
    {
      "epoch": 0.7499707201816876,
      "grad_norm": 0.13435781002044678,
      "learning_rate": 0.0004344059405940594,
      "loss": 2.2634224891662598,
      "step": 7364
    },
    {
      "epoch": 0.7501744058742954,
      "grad_norm": 0.1013956144452095,
      "learning_rate": 0.0004340523338048091,
      "loss": 2.0876545906066895,
      "step": 7366
    },
    {
      "epoch": 0.7503780915669032,
      "grad_norm": 0.17289522290229797,
      "learning_rate": 0.0004336987270155587,
      "loss": 2.167109489440918,
      "step": 7368
    },
    {
      "epoch": 0.7505817772595108,
      "grad_norm": 0.13097891211509705,
      "learning_rate": 0.00043334512022630837,
      "loss": 2.21126389503479,
      "step": 7370
    },
    {
      "epoch": 0.7507854629521186,
      "grad_norm": 0.1172076165676117,
      "learning_rate": 0.000432991513437058,
      "loss": 2.162156820297241,
      "step": 7372
    },
    {
      "epoch": 0.7509891486447263,
      "grad_norm": 0.11858830600976944,
      "learning_rate": 0.00043263790664780765,
      "loss": 2.282296657562256,
      "step": 7374
    },
    {
      "epoch": 0.7511928343373341,
      "grad_norm": 0.13674502074718475,
      "learning_rate": 0.00043228429985855727,
      "loss": 2.2164413928985596,
      "step": 7376
    },
    {
      "epoch": 0.7513965200299418,
      "grad_norm": 0.11274705082178116,
      "learning_rate": 0.00043193069306930693,
      "loss": 2.149714946746826,
      "step": 7378
    },
    {
      "epoch": 0.7516002057225495,
      "grad_norm": 0.12953652441501617,
      "learning_rate": 0.0004315770862800566,
      "loss": 2.2392377853393555,
      "step": 7380
    },
    {
      "epoch": 0.7518038914151572,
      "grad_norm": 0.13338637351989746,
      "learning_rate": 0.0004312234794908063,
      "loss": 2.2033238410949707,
      "step": 7382
    },
    {
      "epoch": 0.752007577107765,
      "grad_norm": 0.11828739941120148,
      "learning_rate": 0.0004308698727015559,
      "loss": 2.257500171661377,
      "step": 7384
    },
    {
      "epoch": 0.7522112628003728,
      "grad_norm": 0.12033595889806747,
      "learning_rate": 0.0004305162659123055,
      "loss": 2.1677632331848145,
      "step": 7386
    },
    {
      "epoch": 0.7524149484929805,
      "grad_norm": 0.11717444658279419,
      "learning_rate": 0.0004301626591230552,
      "loss": 2.310243606567383,
      "step": 7388
    },
    {
      "epoch": 0.7526186341855882,
      "grad_norm": 0.12593597173690796,
      "learning_rate": 0.0004298090523338048,
      "loss": 2.1541004180908203,
      "step": 7390
    },
    {
      "epoch": 0.7528223198781959,
      "grad_norm": 0.11839443445205688,
      "learning_rate": 0.00042945544554455446,
      "loss": 2.105591058731079,
      "step": 7392
    },
    {
      "epoch": 0.7530260055708037,
      "grad_norm": 0.10842331498861313,
      "learning_rate": 0.0004291018387553041,
      "loss": 2.1172218322753906,
      "step": 7394
    },
    {
      "epoch": 0.7532296912634114,
      "grad_norm": 0.13087032735347748,
      "learning_rate": 0.0004287482319660538,
      "loss": 2.235990524291992,
      "step": 7396
    },
    {
      "epoch": 0.7534333769560192,
      "grad_norm": 0.1095251590013504,
      "learning_rate": 0.0004283946251768034,
      "loss": 2.2477762699127197,
      "step": 7398
    },
    {
      "epoch": 0.753637062648627,
      "grad_norm": 0.13663724064826965,
      "learning_rate": 0.000428041018387553,
      "loss": 2.1807003021240234,
      "step": 7400
    },
    {
      "epoch": 0.7538407483412346,
      "grad_norm": 0.11933542788028717,
      "learning_rate": 0.0004276874115983027,
      "loss": 2.210165500640869,
      "step": 7402
    },
    {
      "epoch": 0.7540444340338424,
      "grad_norm": 0.14587624371051788,
      "learning_rate": 0.0004273338048090523,
      "loss": 2.2777812480926514,
      "step": 7404
    },
    {
      "epoch": 0.7542481197264501,
      "grad_norm": 0.13167260587215424,
      "learning_rate": 0.000426980198019802,
      "loss": 2.27335786819458,
      "step": 7406
    },
    {
      "epoch": 0.7544518054190579,
      "grad_norm": 0.1576562374830246,
      "learning_rate": 0.00042662659123055165,
      "loss": 2.124171257019043,
      "step": 7408
    },
    {
      "epoch": 0.7546554911116656,
      "grad_norm": 0.1346999853849411,
      "learning_rate": 0.0004262729844413013,
      "loss": 2.184021234512329,
      "step": 7410
    },
    {
      "epoch": 0.7548591768042733,
      "grad_norm": 0.1283533275127411,
      "learning_rate": 0.00042591937765205093,
      "loss": 2.258882761001587,
      "step": 7412
    },
    {
      "epoch": 0.755062862496881,
      "grad_norm": 0.18324197828769684,
      "learning_rate": 0.00042556577086280055,
      "loss": 2.1795120239257812,
      "step": 7414
    },
    {
      "epoch": 0.7552665481894888,
      "grad_norm": 0.11130441725254059,
      "learning_rate": 0.0004252121640735502,
      "loss": 2.24591064453125,
      "step": 7416
    },
    {
      "epoch": 0.7554702338820966,
      "grad_norm": 0.1315958946943283,
      "learning_rate": 0.00042485855728429983,
      "loss": 2.2127785682678223,
      "step": 7418
    },
    {
      "epoch": 0.7556739195747043,
      "grad_norm": 0.15486687421798706,
      "learning_rate": 0.0004245049504950495,
      "loss": 2.3016796112060547,
      "step": 7420
    },
    {
      "epoch": 0.755877605267312,
      "grad_norm": 0.14735770225524902,
      "learning_rate": 0.00042415134370579917,
      "loss": 2.1790242195129395,
      "step": 7422
    },
    {
      "epoch": 0.7560812909599197,
      "grad_norm": 0.10496921837329865,
      "learning_rate": 0.00042379773691654884,
      "loss": 2.0984597206115723,
      "step": 7424
    },
    {
      "epoch": 0.7562849766525275,
      "grad_norm": 0.12590621411800385,
      "learning_rate": 0.00042344413012729846,
      "loss": 2.311577558517456,
      "step": 7426
    },
    {
      "epoch": 0.7564886623451352,
      "grad_norm": 0.1263175755739212,
      "learning_rate": 0.00042309052333804807,
      "loss": 2.3410801887512207,
      "step": 7428
    },
    {
      "epoch": 0.756692348037743,
      "grad_norm": 0.15581254661083221,
      "learning_rate": 0.00042273691654879774,
      "loss": 2.371112823486328,
      "step": 7430
    },
    {
      "epoch": 0.7568960337303507,
      "grad_norm": 0.14434748888015747,
      "learning_rate": 0.0004223833097595474,
      "loss": 2.0486397743225098,
      "step": 7432
    },
    {
      "epoch": 0.7570997194229584,
      "grad_norm": 0.15814325213432312,
      "learning_rate": 0.000422029702970297,
      "loss": 2.1823554039001465,
      "step": 7434
    },
    {
      "epoch": 0.7573034051155662,
      "grad_norm": 0.1430942565202713,
      "learning_rate": 0.0004216760961810467,
      "loss": 2.2210817337036133,
      "step": 7436
    },
    {
      "epoch": 0.7575070908081739,
      "grad_norm": 0.11624810844659805,
      "learning_rate": 0.00042132248939179636,
      "loss": 2.2832999229431152,
      "step": 7438
    },
    {
      "epoch": 0.7577107765007817,
      "grad_norm": 0.11466860771179199,
      "learning_rate": 0.000420968882602546,
      "loss": 2.197319984436035,
      "step": 7440
    },
    {
      "epoch": 0.7579144621933894,
      "grad_norm": 0.11587543785572052,
      "learning_rate": 0.0004206152758132956,
      "loss": 2.0645103454589844,
      "step": 7442
    },
    {
      "epoch": 0.7581181478859971,
      "grad_norm": 0.12006194144487381,
      "learning_rate": 0.00042026166902404526,
      "loss": 2.1752264499664307,
      "step": 7444
    },
    {
      "epoch": 0.7583218335786048,
      "grad_norm": 0.11728306114673615,
      "learning_rate": 0.00041990806223479493,
      "loss": 2.0015344619750977,
      "step": 7446
    },
    {
      "epoch": 0.7585255192712126,
      "grad_norm": 0.11294545978307724,
      "learning_rate": 0.00041955445544554455,
      "loss": 2.160616159439087,
      "step": 7448
    },
    {
      "epoch": 0.7587292049638203,
      "grad_norm": 0.11578211933374405,
      "learning_rate": 0.0004192008486562942,
      "loss": 2.2277212142944336,
      "step": 7450
    },
    {
      "epoch": 0.7589328906564281,
      "grad_norm": 0.16180458664894104,
      "learning_rate": 0.0004188472418670439,
      "loss": 2.0954928398132324,
      "step": 7452
    },
    {
      "epoch": 0.7591365763490358,
      "grad_norm": 0.1492527574300766,
      "learning_rate": 0.0004184936350777935,
      "loss": 2.3696627616882324,
      "step": 7454
    },
    {
      "epoch": 0.7593402620416435,
      "grad_norm": 0.11103566735982895,
      "learning_rate": 0.0004181400282885431,
      "loss": 2.0913469791412354,
      "step": 7456
    },
    {
      "epoch": 0.7595439477342513,
      "grad_norm": 0.11937806010246277,
      "learning_rate": 0.0004177864214992928,
      "loss": 2.321812152862549,
      "step": 7458
    },
    {
      "epoch": 0.759747633426859,
      "grad_norm": 0.12647084891796112,
      "learning_rate": 0.00041743281471004246,
      "loss": 2.1124353408813477,
      "step": 7460
    },
    {
      "epoch": 0.7599513191194668,
      "grad_norm": 0.19287514686584473,
      "learning_rate": 0.00041707920792079207,
      "loss": 2.172269344329834,
      "step": 7462
    },
    {
      "epoch": 0.7601550048120745,
      "grad_norm": 0.13867880403995514,
      "learning_rate": 0.00041672560113154174,
      "loss": 2.243438720703125,
      "step": 7464
    },
    {
      "epoch": 0.7603586905046822,
      "grad_norm": 0.09831921756267548,
      "learning_rate": 0.0004163719943422914,
      "loss": 2.13106632232666,
      "step": 7466
    },
    {
      "epoch": 0.76056237619729,
      "grad_norm": 0.11364050209522247,
      "learning_rate": 0.000416018387553041,
      "loss": 2.3317391872406006,
      "step": 7468
    },
    {
      "epoch": 0.7607660618898977,
      "grad_norm": 0.1759137213230133,
      "learning_rate": 0.00041566478076379064,
      "loss": 2.140282154083252,
      "step": 7470
    },
    {
      "epoch": 0.7609697475825055,
      "grad_norm": 0.1298263818025589,
      "learning_rate": 0.0004153111739745403,
      "loss": 2.1035356521606445,
      "step": 7472
    },
    {
      "epoch": 0.7611734332751132,
      "grad_norm": 0.1303761601448059,
      "learning_rate": 0.00041495756718529,
      "loss": 2.2078933715820312,
      "step": 7474
    },
    {
      "epoch": 0.7613771189677209,
      "grad_norm": 0.11295116692781448,
      "learning_rate": 0.0004146039603960396,
      "loss": 2.1742048263549805,
      "step": 7476
    },
    {
      "epoch": 0.7615808046603286,
      "grad_norm": 0.13440586626529694,
      "learning_rate": 0.00041425035360678926,
      "loss": 2.1521472930908203,
      "step": 7478
    },
    {
      "epoch": 0.7617844903529364,
      "grad_norm": 0.40169113874435425,
      "learning_rate": 0.00041389674681753893,
      "loss": 2.281069755554199,
      "step": 7480
    },
    {
      "epoch": 0.7619881760455441,
      "grad_norm": 0.1263132393360138,
      "learning_rate": 0.0004135431400282886,
      "loss": 2.115163803100586,
      "step": 7482
    },
    {
      "epoch": 0.7621918617381519,
      "grad_norm": 0.11750753968954086,
      "learning_rate": 0.00041318953323903816,
      "loss": 2.2711591720581055,
      "step": 7484
    },
    {
      "epoch": 0.7623955474307595,
      "grad_norm": 0.11017706990242004,
      "learning_rate": 0.00041283592644978783,
      "loss": 2.092867136001587,
      "step": 7486
    },
    {
      "epoch": 0.7625992331233673,
      "grad_norm": 0.13958165049552917,
      "learning_rate": 0.0004124823196605375,
      "loss": 2.3185930252075195,
      "step": 7488
    },
    {
      "epoch": 0.7628029188159751,
      "grad_norm": 0.1818593293428421,
      "learning_rate": 0.0004121287128712871,
      "loss": 2.2179160118103027,
      "step": 7490
    },
    {
      "epoch": 0.7630066045085828,
      "grad_norm": 0.11625786125659943,
      "learning_rate": 0.0004117751060820368,
      "loss": 2.2617874145507812,
      "step": 7492
    },
    {
      "epoch": 0.7632102902011906,
      "grad_norm": 0.11954747140407562,
      "learning_rate": 0.00041142149929278645,
      "loss": 2.2066941261291504,
      "step": 7494
    },
    {
      "epoch": 0.7634139758937982,
      "grad_norm": 0.139661505818367,
      "learning_rate": 0.0004110678925035361,
      "loss": 2.1448638439178467,
      "step": 7496
    },
    {
      "epoch": 0.763617661586406,
      "grad_norm": 0.11286605149507523,
      "learning_rate": 0.0004107142857142857,
      "loss": 2.1637625694274902,
      "step": 7498
    },
    {
      "epoch": 0.7638213472790137,
      "grad_norm": 0.12274090945720673,
      "learning_rate": 0.00041036067892503535,
      "loss": 2.078033447265625,
      "step": 7500
    },
    {
      "epoch": 0.7640250329716215,
      "grad_norm": 0.11895301938056946,
      "learning_rate": 0.000410007072135785,
      "loss": 2.2567403316497803,
      "step": 7502
    },
    {
      "epoch": 0.7642287186642293,
      "grad_norm": 0.12840381264686584,
      "learning_rate": 0.00040965346534653464,
      "loss": 2.234395980834961,
      "step": 7504
    },
    {
      "epoch": 0.764432404356837,
      "grad_norm": 0.11796200275421143,
      "learning_rate": 0.0004092998585572843,
      "loss": 2.234586477279663,
      "step": 7506
    },
    {
      "epoch": 0.7646360900494447,
      "grad_norm": 0.1231144368648529,
      "learning_rate": 0.000408946251768034,
      "loss": 2.3069896697998047,
      "step": 7508
    },
    {
      "epoch": 0.7648397757420524,
      "grad_norm": 0.12426431477069855,
      "learning_rate": 0.00040859264497878365,
      "loss": 2.11164927482605,
      "step": 7510
    },
    {
      "epoch": 0.7650434614346602,
      "grad_norm": 0.1193031519651413,
      "learning_rate": 0.0004082390381895332,
      "loss": 2.210097312927246,
      "step": 7512
    },
    {
      "epoch": 0.7652471471272679,
      "grad_norm": 0.10683692991733551,
      "learning_rate": 0.0004078854314002829,
      "loss": 2.0934340953826904,
      "step": 7514
    },
    {
      "epoch": 0.7654508328198757,
      "grad_norm": 0.10641656070947647,
      "learning_rate": 0.00040753182461103255,
      "loss": 1.9638471603393555,
      "step": 7516
    },
    {
      "epoch": 0.7656545185124833,
      "grad_norm": 0.13605116307735443,
      "learning_rate": 0.00040717821782178216,
      "loss": 2.0468478202819824,
      "step": 7518
    },
    {
      "epoch": 0.7658582042050911,
      "grad_norm": 0.15182848274707794,
      "learning_rate": 0.00040682461103253183,
      "loss": 2.2245593070983887,
      "step": 7520
    },
    {
      "epoch": 0.7660618898976989,
      "grad_norm": 0.1198265478014946,
      "learning_rate": 0.0004064710042432815,
      "loss": 2.124192237854004,
      "step": 7522
    },
    {
      "epoch": 0.7662655755903066,
      "grad_norm": 0.10609632730484009,
      "learning_rate": 0.00040611739745403117,
      "loss": 2.1620004177093506,
      "step": 7524
    },
    {
      "epoch": 0.7664692612829144,
      "grad_norm": 0.11195284873247147,
      "learning_rate": 0.00040576379066478073,
      "loss": 2.125565767288208,
      "step": 7526
    },
    {
      "epoch": 0.766672946975522,
      "grad_norm": 0.11904114484786987,
      "learning_rate": 0.0004054101838755304,
      "loss": 2.2705302238464355,
      "step": 7528
    },
    {
      "epoch": 0.7668766326681298,
      "grad_norm": 0.10706792771816254,
      "learning_rate": 0.00040505657708628007,
      "loss": 2.10801100730896,
      "step": 7530
    },
    {
      "epoch": 0.7670803183607375,
      "grad_norm": 0.10932281613349915,
      "learning_rate": 0.00040470297029702974,
      "loss": 2.244555950164795,
      "step": 7532
    },
    {
      "epoch": 0.7672840040533453,
      "grad_norm": 0.11411692202091217,
      "learning_rate": 0.00040434936350777935,
      "loss": 1.9592053890228271,
      "step": 7534
    },
    {
      "epoch": 0.767487689745953,
      "grad_norm": 0.10032481700181961,
      "learning_rate": 0.000403995756718529,
      "loss": 2.100128173828125,
      "step": 7536
    },
    {
      "epoch": 0.7676913754385608,
      "grad_norm": 0.10926875472068787,
      "learning_rate": 0.0004036421499292787,
      "loss": 2.1398258209228516,
      "step": 7538
    },
    {
      "epoch": 0.7678950611311685,
      "grad_norm": 0.1495082527399063,
      "learning_rate": 0.00040328854314002825,
      "loss": 2.1789891719818115,
      "step": 7540
    },
    {
      "epoch": 0.7680987468237762,
      "grad_norm": 0.15170566737651825,
      "learning_rate": 0.0004029349363507779,
      "loss": 2.2617626190185547,
      "step": 7542
    },
    {
      "epoch": 0.768302432516384,
      "grad_norm": 0.10486247390508652,
      "learning_rate": 0.0004025813295615276,
      "loss": 2.2156033515930176,
      "step": 7544
    },
    {
      "epoch": 0.7685061182089917,
      "grad_norm": 0.14316405355930328,
      "learning_rate": 0.00040222772277227726,
      "loss": 2.1725852489471436,
      "step": 7546
    },
    {
      "epoch": 0.7687098039015995,
      "grad_norm": 0.11964017152786255,
      "learning_rate": 0.0004018741159830269,
      "loss": 2.222818374633789,
      "step": 7548
    },
    {
      "epoch": 0.7689134895942071,
      "grad_norm": 0.12732727825641632,
      "learning_rate": 0.00040152050919377654,
      "loss": 2.157738447189331,
      "step": 7550
    },
    {
      "epoch": 0.7691171752868149,
      "grad_norm": 0.11193649470806122,
      "learning_rate": 0.0004011669024045262,
      "loss": 2.2767796516418457,
      "step": 7552
    },
    {
      "epoch": 0.7693208609794226,
      "grad_norm": 0.13614420592784882,
      "learning_rate": 0.0004008132956152758,
      "loss": 2.156355381011963,
      "step": 7554
    },
    {
      "epoch": 0.7695245466720304,
      "grad_norm": 0.11354421079158783,
      "learning_rate": 0.00040045968882602544,
      "loss": 2.1452484130859375,
      "step": 7556
    },
    {
      "epoch": 0.7697282323646382,
      "grad_norm": 0.09845248609781265,
      "learning_rate": 0.0004001060820367751,
      "loss": 2.110684871673584,
      "step": 7558
    },
    {
      "epoch": 0.7699319180572458,
      "grad_norm": 0.1163569837808609,
      "learning_rate": 0.0003997524752475248,
      "loss": 2.2935280799865723,
      "step": 7560
    },
    {
      "epoch": 0.7701356037498536,
      "grad_norm": 0.12381293624639511,
      "learning_rate": 0.0003993988684582744,
      "loss": 2.1601250171661377,
      "step": 7562
    },
    {
      "epoch": 0.7703392894424613,
      "grad_norm": 0.11196603626012802,
      "learning_rate": 0.00039904526166902407,
      "loss": 2.0937094688415527,
      "step": 7564
    },
    {
      "epoch": 0.7705429751350691,
      "grad_norm": 0.14257127046585083,
      "learning_rate": 0.00039869165487977374,
      "loss": 2.0962276458740234,
      "step": 7566
    },
    {
      "epoch": 0.7707466608276768,
      "grad_norm": 0.12125086784362793,
      "learning_rate": 0.0003983380480905233,
      "loss": 2.145346164703369,
      "step": 7568
    },
    {
      "epoch": 0.7709503465202846,
      "grad_norm": 0.1341014951467514,
      "learning_rate": 0.00039798444130127297,
      "loss": 2.1533329486846924,
      "step": 7570
    },
    {
      "epoch": 0.7711540322128922,
      "grad_norm": 0.22499701380729675,
      "learning_rate": 0.00039763083451202264,
      "loss": 2.273378372192383,
      "step": 7572
    },
    {
      "epoch": 0.7713577179055,
      "grad_norm": 0.14800283312797546,
      "learning_rate": 0.0003972772277227723,
      "loss": 2.0623230934143066,
      "step": 7574
    },
    {
      "epoch": 0.7715614035981078,
      "grad_norm": 0.11905153840780258,
      "learning_rate": 0.0003969236209335219,
      "loss": 2.1785905361175537,
      "step": 7576
    },
    {
      "epoch": 0.7717650892907155,
      "grad_norm": 0.11357702314853668,
      "learning_rate": 0.0003965700141442716,
      "loss": 2.1547703742980957,
      "step": 7578
    },
    {
      "epoch": 0.7719687749833233,
      "grad_norm": 0.11633023619651794,
      "learning_rate": 0.00039621640735502126,
      "loss": 2.049820899963379,
      "step": 7580
    },
    {
      "epoch": 0.7721724606759309,
      "grad_norm": 0.11940281838178635,
      "learning_rate": 0.0003958628005657709,
      "loss": 2.243802547454834,
      "step": 7582
    },
    {
      "epoch": 0.7723761463685387,
      "grad_norm": 0.14270330965518951,
      "learning_rate": 0.0003955091937765205,
      "loss": 2.2973577976226807,
      "step": 7584
    },
    {
      "epoch": 0.7725798320611464,
      "grad_norm": 0.1379833072423935,
      "learning_rate": 0.00039515558698727016,
      "loss": 2.0888309478759766,
      "step": 7586
    },
    {
      "epoch": 0.7727835177537542,
      "grad_norm": 0.13017885386943817,
      "learning_rate": 0.00039480198019801983,
      "loss": 2.304452896118164,
      "step": 7588
    },
    {
      "epoch": 0.772987203446362,
      "grad_norm": 0.10758107155561447,
      "learning_rate": 0.00039444837340876944,
      "loss": 2.1230616569519043,
      "step": 7590
    },
    {
      "epoch": 0.7731908891389696,
      "grad_norm": 0.10968238115310669,
      "learning_rate": 0.0003940947666195191,
      "loss": 2.2026686668395996,
      "step": 7592
    },
    {
      "epoch": 0.7733945748315774,
      "grad_norm": 0.12587088346481323,
      "learning_rate": 0.0003937411598302688,
      "loss": 2.1762261390686035,
      "step": 7594
    },
    {
      "epoch": 0.7735982605241851,
      "grad_norm": 0.131463885307312,
      "learning_rate": 0.0003933875530410184,
      "loss": 2.0876197814941406,
      "step": 7596
    },
    {
      "epoch": 0.7738019462167929,
      "grad_norm": 0.09080865234136581,
      "learning_rate": 0.000393033946251768,
      "loss": 2.036515951156616,
      "step": 7598
    },
    {
      "epoch": 0.7740056319094006,
      "grad_norm": 0.11709316819906235,
      "learning_rate": 0.0003926803394625177,
      "loss": 2.1828577518463135,
      "step": 7600
    },
    {
      "epoch": 0.7742093176020084,
      "grad_norm": 0.12605930864810944,
      "learning_rate": 0.00039232673267326735,
      "loss": 2.174643039703369,
      "step": 7602
    },
    {
      "epoch": 0.774413003294616,
      "grad_norm": 0.11395516991615295,
      "learning_rate": 0.00039197312588401697,
      "loss": 2.116788864135742,
      "step": 7604
    },
    {
      "epoch": 0.7746166889872238,
      "grad_norm": 0.14740853011608124,
      "learning_rate": 0.00039161951909476663,
      "loss": 2.213912010192871,
      "step": 7606
    },
    {
      "epoch": 0.7748203746798316,
      "grad_norm": 0.11909151822328568,
      "learning_rate": 0.0003912659123055163,
      "loss": 2.1161956787109375,
      "step": 7608
    },
    {
      "epoch": 0.7750240603724393,
      "grad_norm": 0.1315525472164154,
      "learning_rate": 0.0003909123055162659,
      "loss": 2.207735776901245,
      "step": 7610
    },
    {
      "epoch": 0.7752277460650471,
      "grad_norm": 0.11480554938316345,
      "learning_rate": 0.00039055869872701553,
      "loss": 2.233954906463623,
      "step": 7612
    },
    {
      "epoch": 0.7754314317576547,
      "grad_norm": 0.12707406282424927,
      "learning_rate": 0.0003902050919377652,
      "loss": 2.3344337940216064,
      "step": 7614
    },
    {
      "epoch": 0.7756351174502625,
      "grad_norm": 0.1290854811668396,
      "learning_rate": 0.00038985148514851487,
      "loss": 2.1170644760131836,
      "step": 7616
    },
    {
      "epoch": 0.7758388031428702,
      "grad_norm": 0.11787138879299164,
      "learning_rate": 0.0003894978783592645,
      "loss": 2.223832130432129,
      "step": 7618
    },
    {
      "epoch": 0.776042488835478,
      "grad_norm": 0.1293686032295227,
      "learning_rate": 0.00038914427157001416,
      "loss": 2.1878719329833984,
      "step": 7620
    },
    {
      "epoch": 0.7762461745280858,
      "grad_norm": 0.13177616894245148,
      "learning_rate": 0.0003887906647807638,
      "loss": 2.230968713760376,
      "step": 7622
    },
    {
      "epoch": 0.7764498602206934,
      "grad_norm": 0.12280303984880447,
      "learning_rate": 0.00038843705799151344,
      "loss": 2.2277979850769043,
      "step": 7624
    },
    {
      "epoch": 0.7766535459133012,
      "grad_norm": 0.13360098004341125,
      "learning_rate": 0.00038808345120226306,
      "loss": 2.207016944885254,
      "step": 7626
    },
    {
      "epoch": 0.7768572316059089,
      "grad_norm": 0.12297825515270233,
      "learning_rate": 0.0003877298444130127,
      "loss": 2.217522144317627,
      "step": 7628
    },
    {
      "epoch": 0.7770609172985167,
      "grad_norm": 0.12209200114011765,
      "learning_rate": 0.0003873762376237624,
      "loss": 2.3337690830230713,
      "step": 7630
    },
    {
      "epoch": 0.7772646029911244,
      "grad_norm": 0.12440410256385803,
      "learning_rate": 0.00038702263083451206,
      "loss": 2.1043553352355957,
      "step": 7632
    },
    {
      "epoch": 0.7774682886837321,
      "grad_norm": 0.12056566774845123,
      "learning_rate": 0.0003866690240452617,
      "loss": 2.191894292831421,
      "step": 7634
    },
    {
      "epoch": 0.7776719743763398,
      "grad_norm": 0.10340801626443863,
      "learning_rate": 0.00038631541725601135,
      "loss": 2.1667542457580566,
      "step": 7636
    },
    {
      "epoch": 0.7778756600689476,
      "grad_norm": 0.10759221762418747,
      "learning_rate": 0.00038596181046676096,
      "loss": 2.1982367038726807,
      "step": 7638
    },
    {
      "epoch": 0.7780793457615554,
      "grad_norm": 0.1852722465991974,
      "learning_rate": 0.0003856082036775106,
      "loss": 2.145864963531494,
      "step": 7640
    },
    {
      "epoch": 0.7782830314541631,
      "grad_norm": 0.11820481717586517,
      "learning_rate": 0.00038525459688826025,
      "loss": 2.211538314819336,
      "step": 7642
    },
    {
      "epoch": 0.7784867171467709,
      "grad_norm": 0.1548953801393509,
      "learning_rate": 0.0003849009900990099,
      "loss": 2.145458698272705,
      "step": 7644
    },
    {
      "epoch": 0.7786904028393785,
      "grad_norm": 0.16986437141895294,
      "learning_rate": 0.0003845473833097596,
      "loss": 2.301546573638916,
      "step": 7646
    },
    {
      "epoch": 0.7788940885319863,
      "grad_norm": 0.10694389045238495,
      "learning_rate": 0.0003841937765205092,
      "loss": 2.0890426635742188,
      "step": 7648
    },
    {
      "epoch": 0.779097774224594,
      "grad_norm": 0.16169637441635132,
      "learning_rate": 0.00038384016973125887,
      "loss": 2.208787441253662,
      "step": 7650
    },
    {
      "epoch": 0.7793014599172018,
      "grad_norm": 0.11806575208902359,
      "learning_rate": 0.0003834865629420085,
      "loss": 2.1338729858398438,
      "step": 7652
    },
    {
      "epoch": 0.7795051456098095,
      "grad_norm": 0.15782637894153595,
      "learning_rate": 0.0003831329561527581,
      "loss": 2.2428481578826904,
      "step": 7654
    },
    {
      "epoch": 0.7797088313024172,
      "grad_norm": 0.10875310748815536,
      "learning_rate": 0.00038277934936350777,
      "loss": 2.2264809608459473,
      "step": 7656
    },
    {
      "epoch": 0.779912516995025,
      "grad_norm": 0.11186566948890686,
      "learning_rate": 0.00038242574257425744,
      "loss": 2.180967092514038,
      "step": 7658
    },
    {
      "epoch": 0.7801162026876327,
      "grad_norm": 0.11339803040027618,
      "learning_rate": 0.0003820721357850071,
      "loss": 2.091559886932373,
      "step": 7660
    },
    {
      "epoch": 0.7803198883802405,
      "grad_norm": 0.11863519251346588,
      "learning_rate": 0.0003817185289957567,
      "loss": 2.2517690658569336,
      "step": 7662
    },
    {
      "epoch": 0.7805235740728482,
      "grad_norm": 0.12345506250858307,
      "learning_rate": 0.0003813649222065064,
      "loss": 2.176023483276367,
      "step": 7664
    },
    {
      "epoch": 0.7807272597654559,
      "grad_norm": 0.10906413197517395,
      "learning_rate": 0.000381011315417256,
      "loss": 2.165923833847046,
      "step": 7666
    },
    {
      "epoch": 0.7809309454580636,
      "grad_norm": 0.16013115644454956,
      "learning_rate": 0.0003806577086280056,
      "loss": 2.0865366458892822,
      "step": 7668
    },
    {
      "epoch": 0.7811346311506714,
      "grad_norm": 0.12339378148317337,
      "learning_rate": 0.0003803041018387553,
      "loss": 2.2233622074127197,
      "step": 7670
    },
    {
      "epoch": 0.7813383168432791,
      "grad_norm": 0.14184053242206573,
      "learning_rate": 0.00037995049504950496,
      "loss": 2.1619672775268555,
      "step": 7672
    },
    {
      "epoch": 0.7815420025358869,
      "grad_norm": 0.11955928802490234,
      "learning_rate": 0.00037959688826025463,
      "loss": 2.0848207473754883,
      "step": 7674
    },
    {
      "epoch": 0.7817456882284947,
      "grad_norm": 0.10260606557130814,
      "learning_rate": 0.00037924328147100425,
      "loss": 2.1751086711883545,
      "step": 7676
    },
    {
      "epoch": 0.7819493739211023,
      "grad_norm": 0.12498044967651367,
      "learning_rate": 0.0003788896746817539,
      "loss": 2.342471122741699,
      "step": 7678
    },
    {
      "epoch": 0.7821530596137101,
      "grad_norm": 0.11868558824062347,
      "learning_rate": 0.00037853606789250353,
      "loss": 2.1160783767700195,
      "step": 7680
    },
    {
      "epoch": 0.7823567453063178,
      "grad_norm": 0.1190086081624031,
      "learning_rate": 0.0003781824611032532,
      "loss": 2.2391655445098877,
      "step": 7682
    },
    {
      "epoch": 0.7825604309989256,
      "grad_norm": 0.12111752480268478,
      "learning_rate": 0.0003778288543140028,
      "loss": 2.216029167175293,
      "step": 7684
    },
    {
      "epoch": 0.7827641166915333,
      "grad_norm": 0.11023472994565964,
      "learning_rate": 0.0003774752475247525,
      "loss": 2.2736315727233887,
      "step": 7686
    },
    {
      "epoch": 0.782967802384141,
      "grad_norm": 0.12620225548744202,
      "learning_rate": 0.00037712164073550215,
      "loss": 2.101241111755371,
      "step": 7688
    },
    {
      "epoch": 0.7831714880767487,
      "grad_norm": 0.11720751225948334,
      "learning_rate": 0.00037676803394625177,
      "loss": 2.0816850662231445,
      "step": 7690
    },
    {
      "epoch": 0.7833751737693565,
      "grad_norm": 0.1140446737408638,
      "learning_rate": 0.00037641442715700144,
      "loss": 2.282790184020996,
      "step": 7692
    },
    {
      "epoch": 0.7835788594619643,
      "grad_norm": 0.11281349509954453,
      "learning_rate": 0.00037606082036775105,
      "loss": 2.2381839752197266,
      "step": 7694
    },
    {
      "epoch": 0.783782545154572,
      "grad_norm": 0.1395283192396164,
      "learning_rate": 0.0003757072135785007,
      "loss": 2.240267753601074,
      "step": 7696
    },
    {
      "epoch": 0.7839862308471797,
      "grad_norm": 0.12682893872261047,
      "learning_rate": 0.00037535360678925034,
      "loss": 2.1894359588623047,
      "step": 7698
    },
    {
      "epoch": 0.7841899165397874,
      "grad_norm": 0.12481030821800232,
      "learning_rate": 0.000375,
      "loss": 2.1927409172058105,
      "step": 7700
    },
    {
      "epoch": 0.7843936022323952,
      "grad_norm": 0.10465788841247559,
      "learning_rate": 0.0003746463932107497,
      "loss": 2.3093597888946533,
      "step": 7702
    },
    {
      "epoch": 0.7845972879250029,
      "grad_norm": 0.11697810143232346,
      "learning_rate": 0.0003742927864214993,
      "loss": 2.2218029499053955,
      "step": 7704
    },
    {
      "epoch": 0.7848009736176107,
      "grad_norm": 0.1168629378080368,
      "learning_rate": 0.00037393917963224896,
      "loss": 2.1052603721618652,
      "step": 7706
    },
    {
      "epoch": 0.7850046593102185,
      "grad_norm": 0.1322336345911026,
      "learning_rate": 0.0003735855728429986,
      "loss": 2.163619041442871,
      "step": 7708
    },
    {
      "epoch": 0.7852083450028261,
      "grad_norm": 0.14012421667575836,
      "learning_rate": 0.00037323196605374825,
      "loss": 2.2746999263763428,
      "step": 7710
    },
    {
      "epoch": 0.7854120306954339,
      "grad_norm": 0.14665111899375916,
      "learning_rate": 0.00037287835926449786,
      "loss": 2.356703281402588,
      "step": 7712
    },
    {
      "epoch": 0.7856157163880416,
      "grad_norm": 0.12099422514438629,
      "learning_rate": 0.00037252475247524753,
      "loss": 2.256875514984131,
      "step": 7714
    },
    {
      "epoch": 0.7858194020806494,
      "grad_norm": 0.14951945841312408,
      "learning_rate": 0.0003721711456859972,
      "loss": 2.2216458320617676,
      "step": 7716
    },
    {
      "epoch": 0.7860230877732571,
      "grad_norm": 0.118695929646492,
      "learning_rate": 0.0003718175388967468,
      "loss": 2.101557493209839,
      "step": 7718
    },
    {
      "epoch": 0.7862267734658648,
      "grad_norm": 0.1440853625535965,
      "learning_rate": 0.0003714639321074965,
      "loss": 2.2409732341766357,
      "step": 7720
    },
    {
      "epoch": 0.7864304591584725,
      "grad_norm": 0.11635091155767441,
      "learning_rate": 0.0003711103253182461,
      "loss": 2.1755027770996094,
      "step": 7722
    },
    {
      "epoch": 0.7866341448510803,
      "grad_norm": 0.12196182459592819,
      "learning_rate": 0.00037075671852899577,
      "loss": 2.2817277908325195,
      "step": 7724
    },
    {
      "epoch": 0.786837830543688,
      "grad_norm": 0.14254382252693176,
      "learning_rate": 0.0003704031117397454,
      "loss": 2.261566162109375,
      "step": 7726
    },
    {
      "epoch": 0.7870415162362958,
      "grad_norm": 0.12391708046197891,
      "learning_rate": 0.00037004950495049505,
      "loss": 2.162715435028076,
      "step": 7728
    },
    {
      "epoch": 0.7872452019289035,
      "grad_norm": 0.11442998051643372,
      "learning_rate": 0.0003696958981612447,
      "loss": 2.193514347076416,
      "step": 7730
    },
    {
      "epoch": 0.7874488876215112,
      "grad_norm": 0.12188281863927841,
      "learning_rate": 0.0003693422913719944,
      "loss": 2.178208589553833,
      "step": 7732
    },
    {
      "epoch": 0.787652573314119,
      "grad_norm": 0.11716936528682709,
      "learning_rate": 0.000368988684582744,
      "loss": 2.3788881301879883,
      "step": 7734
    },
    {
      "epoch": 0.7878562590067267,
      "grad_norm": 0.14356903731822968,
      "learning_rate": 0.0003686350777934936,
      "loss": 2.2849326133728027,
      "step": 7736
    },
    {
      "epoch": 0.7880599446993345,
      "grad_norm": 0.11025405675172806,
      "learning_rate": 0.0003682814710042433,
      "loss": 2.1291496753692627,
      "step": 7738
    },
    {
      "epoch": 0.7882636303919422,
      "grad_norm": 0.10432004928588867,
      "learning_rate": 0.0003679278642149929,
      "loss": 2.140965461730957,
      "step": 7740
    },
    {
      "epoch": 0.7884673160845499,
      "grad_norm": 0.17200422286987305,
      "learning_rate": 0.0003675742574257426,
      "loss": 2.290426254272461,
      "step": 7742
    },
    {
      "epoch": 0.7886710017771577,
      "grad_norm": 0.1137315109372139,
      "learning_rate": 0.00036722065063649224,
      "loss": 2.258269786834717,
      "step": 7744
    },
    {
      "epoch": 0.7888746874697654,
      "grad_norm": 0.12694571912288666,
      "learning_rate": 0.0003668670438472419,
      "loss": 2.2404730319976807,
      "step": 7746
    },
    {
      "epoch": 0.7890783731623732,
      "grad_norm": 0.11810439825057983,
      "learning_rate": 0.00036651343705799153,
      "loss": 2.0956907272338867,
      "step": 7748
    },
    {
      "epoch": 0.7892820588549809,
      "grad_norm": 0.13670235872268677,
      "learning_rate": 0.00036615983026874114,
      "loss": 2.224611759185791,
      "step": 7750
    },
    {
      "epoch": 0.7894857445475886,
      "grad_norm": 0.11581224948167801,
      "learning_rate": 0.0003658062234794908,
      "loss": 2.4227583408355713,
      "step": 7752
    },
    {
      "epoch": 0.7896894302401963,
      "grad_norm": 0.135577991604805,
      "learning_rate": 0.00036545261669024043,
      "loss": 2.21651029586792,
      "step": 7754
    },
    {
      "epoch": 0.7898931159328041,
      "grad_norm": 0.1313857138156891,
      "learning_rate": 0.0003650990099009901,
      "loss": 2.2684528827667236,
      "step": 7756
    },
    {
      "epoch": 0.7900968016254118,
      "grad_norm": 0.12197069823741913,
      "learning_rate": 0.00036474540311173977,
      "loss": 2.1991610527038574,
      "step": 7758
    },
    {
      "epoch": 0.7903004873180196,
      "grad_norm": 0.10910201817750931,
      "learning_rate": 0.00036439179632248944,
      "loss": 2.2714810371398926,
      "step": 7760
    },
    {
      "epoch": 0.7905041730106273,
      "grad_norm": 0.11678063869476318,
      "learning_rate": 0.00036403818953323905,
      "loss": 2.2665984630584717,
      "step": 7762
    },
    {
      "epoch": 0.790707858703235,
      "grad_norm": 0.1205020323395729,
      "learning_rate": 0.00036368458274398867,
      "loss": 2.1842050552368164,
      "step": 7764
    },
    {
      "epoch": 0.7909115443958428,
      "grad_norm": 0.11365606635808945,
      "learning_rate": 0.00036333097595473834,
      "loss": 2.1317765712738037,
      "step": 7766
    },
    {
      "epoch": 0.7911152300884505,
      "grad_norm": 0.11252666264772415,
      "learning_rate": 0.00036297736916548795,
      "loss": 2.181363105773926,
      "step": 7768
    },
    {
      "epoch": 0.7913189157810583,
      "grad_norm": 0.12137565016746521,
      "learning_rate": 0.0003626237623762376,
      "loss": 2.0766987800598145,
      "step": 7770
    },
    {
      "epoch": 0.7915226014736659,
      "grad_norm": 0.11630336195230484,
      "learning_rate": 0.0003622701555869873,
      "loss": 2.105882167816162,
      "step": 7772
    },
    {
      "epoch": 0.7917262871662737,
      "grad_norm": 0.12031614035367966,
      "learning_rate": 0.00036191654879773696,
      "loss": 2.061141014099121,
      "step": 7774
    },
    {
      "epoch": 0.7919299728588814,
      "grad_norm": 0.14484915137290955,
      "learning_rate": 0.0003615629420084866,
      "loss": 2.234410047531128,
      "step": 7776
    },
    {
      "epoch": 0.7921336585514892,
      "grad_norm": 0.1271427869796753,
      "learning_rate": 0.0003612093352192362,
      "loss": 2.184514284133911,
      "step": 7778
    },
    {
      "epoch": 0.792337344244097,
      "grad_norm": 0.14735957980155945,
      "learning_rate": 0.00036085572842998586,
      "loss": 2.166403293609619,
      "step": 7780
    },
    {
      "epoch": 0.7925410299367047,
      "grad_norm": 0.14090785384178162,
      "learning_rate": 0.00036050212164073553,
      "loss": 2.206090211868286,
      "step": 7782
    },
    {
      "epoch": 0.7927447156293124,
      "grad_norm": 0.14911113679409027,
      "learning_rate": 0.00036014851485148514,
      "loss": 2.1798465251922607,
      "step": 7784
    },
    {
      "epoch": 0.7929484013219201,
      "grad_norm": 0.10830841213464737,
      "learning_rate": 0.0003597949080622348,
      "loss": 2.180508613586426,
      "step": 7786
    },
    {
      "epoch": 0.7931520870145279,
      "grad_norm": 0.10368246585130692,
      "learning_rate": 0.0003594413012729845,
      "loss": 2.0911457538604736,
      "step": 7788
    },
    {
      "epoch": 0.7933557727071356,
      "grad_norm": 0.10041835159063339,
      "learning_rate": 0.0003590876944837341,
      "loss": 2.0827999114990234,
      "step": 7790
    },
    {
      "epoch": 0.7935594583997434,
      "grad_norm": 0.1357717365026474,
      "learning_rate": 0.0003587340876944837,
      "loss": 2.2814533710479736,
      "step": 7792
    },
    {
      "epoch": 0.793763144092351,
      "grad_norm": 0.1257532387971878,
      "learning_rate": 0.0003583804809052334,
      "loss": 2.2639904022216797,
      "step": 7794
    },
    {
      "epoch": 0.7939668297849588,
      "grad_norm": 0.1238815188407898,
      "learning_rate": 0.00035802687411598305,
      "loss": 2.2129359245300293,
      "step": 7796
    },
    {
      "epoch": 0.7941705154775666,
      "grad_norm": 0.1200995147228241,
      "learning_rate": 0.00035767326732673267,
      "loss": 2.3335886001586914,
      "step": 7798
    },
    {
      "epoch": 0.7943742011701743,
      "grad_norm": 0.11411391198635101,
      "learning_rate": 0.00035731966053748233,
      "loss": 2.184903144836426,
      "step": 7800
    },
    {
      "epoch": 0.7945778868627821,
      "grad_norm": 0.11360441893339157,
      "learning_rate": 0.000356966053748232,
      "loss": 2.0192551612854004,
      "step": 7802
    },
    {
      "epoch": 0.7947815725553897,
      "grad_norm": 0.1392737627029419,
      "learning_rate": 0.0003566124469589816,
      "loss": 2.1399433612823486,
      "step": 7804
    },
    {
      "epoch": 0.7949852582479975,
      "grad_norm": 0.11316756904125214,
      "learning_rate": 0.00035625884016973123,
      "loss": 2.1551513671875,
      "step": 7806
    },
    {
      "epoch": 0.7951889439406052,
      "grad_norm": 0.12153292447328568,
      "learning_rate": 0.0003559052333804809,
      "loss": 2.2388648986816406,
      "step": 7808
    },
    {
      "epoch": 0.795392629633213,
      "grad_norm": 0.10022859275341034,
      "learning_rate": 0.0003555516265912306,
      "loss": 2.0744667053222656,
      "step": 7810
    },
    {
      "epoch": 0.7955963153258208,
      "grad_norm": 0.13442184031009674,
      "learning_rate": 0.0003551980198019802,
      "loss": 2.1698596477508545,
      "step": 7812
    },
    {
      "epoch": 0.7958000010184285,
      "grad_norm": 0.13044272363185883,
      "learning_rate": 0.00035484441301272986,
      "loss": 2.1697113513946533,
      "step": 7814
    },
    {
      "epoch": 0.7960036867110362,
      "grad_norm": 0.11960259079933167,
      "learning_rate": 0.0003544908062234795,
      "loss": 2.226241111755371,
      "step": 7816
    },
    {
      "epoch": 0.7962073724036439,
      "grad_norm": 0.11224517226219177,
      "learning_rate": 0.00035413719943422914,
      "loss": 2.0630970001220703,
      "step": 7818
    },
    {
      "epoch": 0.7964110580962517,
      "grad_norm": 0.11859650164842606,
      "learning_rate": 0.00035378359264497876,
      "loss": 2.10903263092041,
      "step": 7820
    },
    {
      "epoch": 0.7966147437888594,
      "grad_norm": 0.12817636132240295,
      "learning_rate": 0.0003534299858557284,
      "loss": 2.1610026359558105,
      "step": 7822
    },
    {
      "epoch": 0.7968184294814672,
      "grad_norm": 0.12125498056411743,
      "learning_rate": 0.0003530763790664781,
      "loss": 2.1833319664001465,
      "step": 7824
    },
    {
      "epoch": 0.7970221151740748,
      "grad_norm": 0.11539766192436218,
      "learning_rate": 0.0003527227722772277,
      "loss": 2.3660316467285156,
      "step": 7826
    },
    {
      "epoch": 0.7972258008666826,
      "grad_norm": 0.107422836124897,
      "learning_rate": 0.0003523691654879774,
      "loss": 2.009080648422241,
      "step": 7828
    },
    {
      "epoch": 0.7974294865592904,
      "grad_norm": 0.13805031776428223,
      "learning_rate": 0.00035201555869872705,
      "loss": 2.132293224334717,
      "step": 7830
    },
    {
      "epoch": 0.7976331722518981,
      "grad_norm": 0.1033753752708435,
      "learning_rate": 0.0003516619519094767,
      "loss": 2.0949833393096924,
      "step": 7832
    },
    {
      "epoch": 0.7978368579445059,
      "grad_norm": 0.11357478052377701,
      "learning_rate": 0.0003513083451202263,
      "loss": 2.2085893154144287,
      "step": 7834
    },
    {
      "epoch": 0.7980405436371135,
      "grad_norm": 0.14497891068458557,
      "learning_rate": 0.00035095473833097595,
      "loss": 2.2840495109558105,
      "step": 7836
    },
    {
      "epoch": 0.7982442293297213,
      "grad_norm": 0.12452489882707596,
      "learning_rate": 0.0003506011315417256,
      "loss": 2.231752872467041,
      "step": 7838
    },
    {
      "epoch": 0.798447915022329,
      "grad_norm": 0.1325587034225464,
      "learning_rate": 0.00035024752475247523,
      "loss": 2.1613895893096924,
      "step": 7840
    },
    {
      "epoch": 0.7986516007149368,
      "grad_norm": 0.10710452497005463,
      "learning_rate": 0.0003498939179632249,
      "loss": 2.1133694648742676,
      "step": 7842
    },
    {
      "epoch": 0.7988552864075446,
      "grad_norm": 0.13017281889915466,
      "learning_rate": 0.00034954031117397457,
      "loss": 2.292755126953125,
      "step": 7844
    },
    {
      "epoch": 0.7990589721001523,
      "grad_norm": 0.11663831025362015,
      "learning_rate": 0.00034918670438472424,
      "loss": 2.089218854904175,
      "step": 7846
    },
    {
      "epoch": 0.79926265779276,
      "grad_norm": 0.1280330866575241,
      "learning_rate": 0.0003488330975954738,
      "loss": 2.33420467376709,
      "step": 7848
    },
    {
      "epoch": 0.7994663434853677,
      "grad_norm": 0.11932696402072906,
      "learning_rate": 0.00034847949080622347,
      "loss": 2.150791645050049,
      "step": 7850
    },
    {
      "epoch": 0.7996700291779755,
      "grad_norm": 0.11601509153842926,
      "learning_rate": 0.00034812588401697314,
      "loss": 2.045419216156006,
      "step": 7852
    },
    {
      "epoch": 0.7998737148705832,
      "grad_norm": 0.10614177584648132,
      "learning_rate": 0.00034777227722772276,
      "loss": 2.104888439178467,
      "step": 7854
    },
    {
      "epoch": 0.800077400563191,
      "grad_norm": 0.1283053606748581,
      "learning_rate": 0.0003474186704384724,
      "loss": 2.1801650524139404,
      "step": 7856
    },
    {
      "epoch": 0.8002810862557986,
      "grad_norm": 0.11199472844600677,
      "learning_rate": 0.0003470650636492221,
      "loss": 2.157182216644287,
      "step": 7858
    },
    {
      "epoch": 0.8004847719484064,
      "grad_norm": 0.12129276245832443,
      "learning_rate": 0.00034671145685997176,
      "loss": 2.124790668487549,
      "step": 7860
    },
    {
      "epoch": 0.8006884576410142,
      "grad_norm": 0.130443274974823,
      "learning_rate": 0.0003463578500707213,
      "loss": 2.2528934478759766,
      "step": 7862
    },
    {
      "epoch": 0.8008921433336219,
      "grad_norm": 0.10921832919120789,
      "learning_rate": 0.000346004243281471,
      "loss": 2.229609966278076,
      "step": 7864
    },
    {
      "epoch": 0.8010958290262297,
      "grad_norm": 0.13497889041900635,
      "learning_rate": 0.00034565063649222066,
      "loss": 2.2059149742126465,
      "step": 7866
    },
    {
      "epoch": 0.8012995147188373,
      "grad_norm": 0.1207050159573555,
      "learning_rate": 0.0003452970297029703,
      "loss": 2.205594778060913,
      "step": 7868
    },
    {
      "epoch": 0.8015032004114451,
      "grad_norm": 0.1399155706167221,
      "learning_rate": 0.00034494342291371995,
      "loss": 2.233565330505371,
      "step": 7870
    },
    {
      "epoch": 0.8017068861040528,
      "grad_norm": 0.16016075015068054,
      "learning_rate": 0.0003445898161244696,
      "loss": 2.1284096240997314,
      "step": 7872
    },
    {
      "epoch": 0.8019105717966606,
      "grad_norm": 0.10748561471700668,
      "learning_rate": 0.0003442362093352193,
      "loss": 2.127472400665283,
      "step": 7874
    },
    {
      "epoch": 0.8021142574892683,
      "grad_norm": 0.12328855693340302,
      "learning_rate": 0.00034388260254596885,
      "loss": 2.2378950119018555,
      "step": 7876
    },
    {
      "epoch": 0.802317943181876,
      "grad_norm": 0.11242914944887161,
      "learning_rate": 0.0003435289957567185,
      "loss": 2.031320571899414,
      "step": 7878
    },
    {
      "epoch": 0.8025216288744838,
      "grad_norm": 0.12081172317266464,
      "learning_rate": 0.0003431753889674682,
      "loss": 2.1255502700805664,
      "step": 7880
    },
    {
      "epoch": 0.8027253145670915,
      "grad_norm": 0.15110699832439423,
      "learning_rate": 0.00034282178217821786,
      "loss": 2.162748336791992,
      "step": 7882
    },
    {
      "epoch": 0.8029290002596993,
      "grad_norm": 0.10879082977771759,
      "learning_rate": 0.00034246817538896747,
      "loss": 2.149047374725342,
      "step": 7884
    },
    {
      "epoch": 0.803132685952307,
      "grad_norm": 0.12531301379203796,
      "learning_rate": 0.00034211456859971714,
      "loss": 2.146850347518921,
      "step": 7886
    },
    {
      "epoch": 0.8033363716449148,
      "grad_norm": 0.11774736642837524,
      "learning_rate": 0.0003417609618104668,
      "loss": 2.2381961345672607,
      "step": 7888
    },
    {
      "epoch": 0.8035400573375224,
      "grad_norm": 0.11990202218294144,
      "learning_rate": 0.00034140735502121637,
      "loss": 2.1238341331481934,
      "step": 7890
    },
    {
      "epoch": 0.8037437430301302,
      "grad_norm": 0.11240259557962418,
      "learning_rate": 0.00034105374823196604,
      "loss": 2.174626350402832,
      "step": 7892
    },
    {
      "epoch": 0.803947428722738,
      "grad_norm": 0.14520208537578583,
      "learning_rate": 0.0003407001414427157,
      "loss": 2.320854663848877,
      "step": 7894
    },
    {
      "epoch": 0.8041511144153457,
      "grad_norm": 0.13072127103805542,
      "learning_rate": 0.0003403465346534654,
      "loss": 2.1848936080932617,
      "step": 7896
    },
    {
      "epoch": 0.8043548001079535,
      "grad_norm": 0.11539946496486664,
      "learning_rate": 0.000339992927864215,
      "loss": 2.1903555393218994,
      "step": 7898
    },
    {
      "epoch": 0.8045584858005611,
      "grad_norm": 0.1254122257232666,
      "learning_rate": 0.00033963932107496466,
      "loss": 2.1296486854553223,
      "step": 7900
    },
    {
      "epoch": 0.8047621714931689,
      "grad_norm": 0.11741260439157486,
      "learning_rate": 0.00033928571428571433,
      "loss": 2.1216063499450684,
      "step": 7902
    },
    {
      "epoch": 0.8049658571857766,
      "grad_norm": 0.1271190643310547,
      "learning_rate": 0.0003389321074964639,
      "loss": 2.1028285026550293,
      "step": 7904
    },
    {
      "epoch": 0.8051695428783844,
      "grad_norm": 0.12315487116575241,
      "learning_rate": 0.00033857850070721356,
      "loss": 2.1565890312194824,
      "step": 7906
    },
    {
      "epoch": 0.8053732285709921,
      "grad_norm": 0.11878010630607605,
      "learning_rate": 0.00033822489391796323,
      "loss": 2.2738547325134277,
      "step": 7908
    },
    {
      "epoch": 0.8055769142635998,
      "grad_norm": 0.12007664144039154,
      "learning_rate": 0.0003378712871287129,
      "loss": 2.3603086471557617,
      "step": 7910
    },
    {
      "epoch": 0.8057805999562075,
      "grad_norm": 0.15576930344104767,
      "learning_rate": 0.0003375176803394625,
      "loss": 2.2025277614593506,
      "step": 7912
    },
    {
      "epoch": 0.8059842856488153,
      "grad_norm": 0.1333644986152649,
      "learning_rate": 0.0003371640735502122,
      "loss": 2.2077016830444336,
      "step": 7914
    },
    {
      "epoch": 0.8061879713414231,
      "grad_norm": 0.1180582344532013,
      "learning_rate": 0.00033681046676096185,
      "loss": 2.2466368675231934,
      "step": 7916
    },
    {
      "epoch": 0.8063916570340308,
      "grad_norm": 0.13653147220611572,
      "learning_rate": 0.0003364568599717114,
      "loss": 2.306959629058838,
      "step": 7918
    },
    {
      "epoch": 0.8065953427266386,
      "grad_norm": 0.10488647222518921,
      "learning_rate": 0.0003361032531824611,
      "loss": 2.1754517555236816,
      "step": 7920
    },
    {
      "epoch": 0.8067990284192462,
      "grad_norm": 0.11933264881372452,
      "learning_rate": 0.00033574964639321075,
      "loss": 2.1893322467803955,
      "step": 7922
    },
    {
      "epoch": 0.807002714111854,
      "grad_norm": 0.12756097316741943,
      "learning_rate": 0.0003353960396039604,
      "loss": 2.0425446033477783,
      "step": 7924
    },
    {
      "epoch": 0.8072063998044617,
      "grad_norm": 0.13093633949756622,
      "learning_rate": 0.00033504243281471004,
      "loss": 2.1561169624328613,
      "step": 7926
    },
    {
      "epoch": 0.8074100854970695,
      "grad_norm": 0.11788320541381836,
      "learning_rate": 0.0003346888260254597,
      "loss": 2.1200101375579834,
      "step": 7928
    },
    {
      "epoch": 0.8076137711896773,
      "grad_norm": 0.1213313490152359,
      "learning_rate": 0.0003343352192362094,
      "loss": 2.1425743103027344,
      "step": 7930
    },
    {
      "epoch": 0.8078174568822849,
      "grad_norm": 0.10630789399147034,
      "learning_rate": 0.000333981612446959,
      "loss": 2.1267776489257812,
      "step": 7932
    },
    {
      "epoch": 0.8080211425748927,
      "grad_norm": 0.11860230565071106,
      "learning_rate": 0.0003336280056577086,
      "loss": 2.2239789962768555,
      "step": 7934
    },
    {
      "epoch": 0.8082248282675004,
      "grad_norm": 0.14578630030155182,
      "learning_rate": 0.0003332743988684583,
      "loss": 2.1244964599609375,
      "step": 7936
    },
    {
      "epoch": 0.8084285139601082,
      "grad_norm": 0.09773004800081253,
      "learning_rate": 0.00033292079207920795,
      "loss": 2.099438190460205,
      "step": 7938
    },
    {
      "epoch": 0.8086321996527159,
      "grad_norm": 0.10993091762065887,
      "learning_rate": 0.00033256718528995756,
      "loss": 2.1105313301086426,
      "step": 7940
    },
    {
      "epoch": 0.8088358853453236,
      "grad_norm": 0.1085745319724083,
      "learning_rate": 0.00033221357850070723,
      "loss": 2.1075339317321777,
      "step": 7942
    },
    {
      "epoch": 0.8090395710379313,
      "grad_norm": 0.13988660275936127,
      "learning_rate": 0.0003318599717114569,
      "loss": 2.1321730613708496,
      "step": 7944
    },
    {
      "epoch": 0.8092432567305391,
      "grad_norm": 0.13297894597053528,
      "learning_rate": 0.0003315063649222065,
      "loss": 2.199112892150879,
      "step": 7946
    },
    {
      "epoch": 0.8094469424231469,
      "grad_norm": 0.13374896347522736,
      "learning_rate": 0.00033115275813295613,
      "loss": 2.236461639404297,
      "step": 7948
    },
    {
      "epoch": 0.8096506281157546,
      "grad_norm": 0.11001221835613251,
      "learning_rate": 0.0003307991513437058,
      "loss": 2.1736462116241455,
      "step": 7950
    },
    {
      "epoch": 0.8098543138083624,
      "grad_norm": 0.1128813698887825,
      "learning_rate": 0.00033044554455445547,
      "loss": 2.2416186332702637,
      "step": 7952
    },
    {
      "epoch": 0.81005799950097,
      "grad_norm": 0.13606499135494232,
      "learning_rate": 0.0003300919377652051,
      "loss": 2.3409762382507324,
      "step": 7954
    },
    {
      "epoch": 0.8102616851935778,
      "grad_norm": 0.1328086405992508,
      "learning_rate": 0.00032973833097595475,
      "loss": 1.9811738729476929,
      "step": 7956
    },
    {
      "epoch": 0.8104653708861855,
      "grad_norm": 0.13104325532913208,
      "learning_rate": 0.0003293847241867044,
      "loss": 2.11552357673645,
      "step": 7958
    },
    {
      "epoch": 0.8106690565787933,
      "grad_norm": 0.1382075995206833,
      "learning_rate": 0.00032903111739745404,
      "loss": 2.1411213874816895,
      "step": 7960
    },
    {
      "epoch": 0.810872742271401,
      "grad_norm": 0.12922942638397217,
      "learning_rate": 0.00032867751060820365,
      "loss": 2.3317108154296875,
      "step": 7962
    },
    {
      "epoch": 0.8110764279640087,
      "grad_norm": 0.09431646764278412,
      "learning_rate": 0.0003283239038189533,
      "loss": 2.1062374114990234,
      "step": 7964
    },
    {
      "epoch": 0.8112801136566165,
      "grad_norm": 0.11610108613967896,
      "learning_rate": 0.000327970297029703,
      "loss": 2.2143399715423584,
      "step": 7966
    },
    {
      "epoch": 0.8114837993492242,
      "grad_norm": 0.14476923644542694,
      "learning_rate": 0.0003276166902404526,
      "loss": 2.1183996200561523,
      "step": 7968
    },
    {
      "epoch": 0.811687485041832,
      "grad_norm": 0.11989686638116837,
      "learning_rate": 0.0003272630834512023,
      "loss": 2.304844379425049,
      "step": 7970
    },
    {
      "epoch": 0.8118911707344397,
      "grad_norm": 0.12282120436429977,
      "learning_rate": 0.00032690947666195194,
      "loss": 2.2896976470947266,
      "step": 7972
    },
    {
      "epoch": 0.8120948564270474,
      "grad_norm": 0.143123596906662,
      "learning_rate": 0.00032655586987270156,
      "loss": 2.1320688724517822,
      "step": 7974
    },
    {
      "epoch": 0.8122985421196551,
      "grad_norm": 0.15987257659435272,
      "learning_rate": 0.0003262022630834512,
      "loss": 2.2129688262939453,
      "step": 7976
    },
    {
      "epoch": 0.8125022278122629,
      "grad_norm": 0.10305670648813248,
      "learning_rate": 0.00032584865629420084,
      "loss": 2.2866668701171875,
      "step": 7978
    },
    {
      "epoch": 0.8127059135048706,
      "grad_norm": 0.11754024773836136,
      "learning_rate": 0.0003254950495049505,
      "loss": 2.1374356746673584,
      "step": 7980
    },
    {
      "epoch": 0.8129095991974784,
      "grad_norm": 0.12411908805370331,
      "learning_rate": 0.0003251414427157002,
      "loss": 2.2158761024475098,
      "step": 7982
    },
    {
      "epoch": 0.8131132848900862,
      "grad_norm": 0.13163192570209503,
      "learning_rate": 0.0003247878359264498,
      "loss": 2.364628791809082,
      "step": 7984
    },
    {
      "epoch": 0.8133169705826938,
      "grad_norm": 0.10726015269756317,
      "learning_rate": 0.00032443422913719947,
      "loss": 2.219970464706421,
      "step": 7986
    },
    {
      "epoch": 0.8135206562753016,
      "grad_norm": 0.1443360447883606,
      "learning_rate": 0.0003240806223479491,
      "loss": 2.1823906898498535,
      "step": 7988
    },
    {
      "epoch": 0.8137243419679093,
      "grad_norm": 0.15451686084270477,
      "learning_rate": 0.0003237270155586987,
      "loss": 2.2824175357818604,
      "step": 7990
    },
    {
      "epoch": 0.8139280276605171,
      "grad_norm": 0.1379615217447281,
      "learning_rate": 0.00032337340876944837,
      "loss": 2.2311177253723145,
      "step": 7992
    },
    {
      "epoch": 0.8141317133531248,
      "grad_norm": 0.1307673156261444,
      "learning_rate": 0.00032301980198019804,
      "loss": 2.17642879486084,
      "step": 7994
    },
    {
      "epoch": 0.8143353990457325,
      "grad_norm": 0.11893113702535629,
      "learning_rate": 0.0003226661951909477,
      "loss": 2.2852461338043213,
      "step": 7996
    },
    {
      "epoch": 0.8145390847383402,
      "grad_norm": 0.1239408552646637,
      "learning_rate": 0.0003223125884016973,
      "loss": 2.1812634468078613,
      "step": 7998
    },
    {
      "epoch": 0.814742770430948,
      "grad_norm": 0.12821362912654877,
      "learning_rate": 0.000321958981612447,
      "loss": 2.084040641784668,
      "step": 8000
    },
    {
      "epoch": 0.8149464561235558,
      "grad_norm": 0.13979463279247284,
      "learning_rate": 0.0003216053748231966,
      "loss": 2.2751717567443848,
      "step": 8002
    },
    {
      "epoch": 0.8151501418161635,
      "grad_norm": 0.10401247441768646,
      "learning_rate": 0.0003212517680339462,
      "loss": 2.17498779296875,
      "step": 8004
    },
    {
      "epoch": 0.8153538275087712,
      "grad_norm": 0.10649096220731735,
      "learning_rate": 0.0003208981612446959,
      "loss": 2.192232608795166,
      "step": 8006
    },
    {
      "epoch": 0.8155575132013789,
      "grad_norm": 0.14865191280841827,
      "learning_rate": 0.00032054455445544556,
      "loss": 2.2152342796325684,
      "step": 8008
    },
    {
      "epoch": 0.8157611988939867,
      "grad_norm": 0.1365330070257187,
      "learning_rate": 0.00032019094766619523,
      "loss": 2.2106075286865234,
      "step": 8010
    },
    {
      "epoch": 0.8159648845865944,
      "grad_norm": 0.16810238361358643,
      "learning_rate": 0.00031983734087694484,
      "loss": 2.194718599319458,
      "step": 8012
    },
    {
      "epoch": 0.8161685702792022,
      "grad_norm": 0.13432274758815765,
      "learning_rate": 0.0003194837340876945,
      "loss": 2.234405517578125,
      "step": 8014
    },
    {
      "epoch": 0.8163722559718098,
      "grad_norm": 0.13303619623184204,
      "learning_rate": 0.0003191301272984441,
      "loss": 2.1715807914733887,
      "step": 8016
    },
    {
      "epoch": 0.8165759416644176,
      "grad_norm": 0.12310335040092468,
      "learning_rate": 0.00031877652050919374,
      "loss": 2.1876044273376465,
      "step": 8018
    },
    {
      "epoch": 0.8167796273570254,
      "grad_norm": 0.13124507665634155,
      "learning_rate": 0.0003184229137199434,
      "loss": 2.1879422664642334,
      "step": 8020
    },
    {
      "epoch": 0.8169833130496331,
      "grad_norm": 0.1898803859949112,
      "learning_rate": 0.0003180693069306931,
      "loss": 2.380600929260254,
      "step": 8022
    },
    {
      "epoch": 0.8171869987422409,
      "grad_norm": 0.10968513786792755,
      "learning_rate": 0.00031771570014144275,
      "loss": 2.2040743827819824,
      "step": 8024
    },
    {
      "epoch": 0.8173906844348486,
      "grad_norm": 0.1593126654624939,
      "learning_rate": 0.00031736209335219237,
      "loss": 2.113182544708252,
      "step": 8026
    },
    {
      "epoch": 0.8175943701274563,
      "grad_norm": 0.13096506893634796,
      "learning_rate": 0.00031700848656294203,
      "loss": 2.3056507110595703,
      "step": 8028
    },
    {
      "epoch": 0.817798055820064,
      "grad_norm": 0.11455587297677994,
      "learning_rate": 0.00031665487977369165,
      "loss": 2.2828519344329834,
      "step": 8030
    },
    {
      "epoch": 0.8180017415126718,
      "grad_norm": 0.1341342031955719,
      "learning_rate": 0.0003163012729844413,
      "loss": 2.1049230098724365,
      "step": 8032
    },
    {
      "epoch": 0.8182054272052796,
      "grad_norm": 0.1107359379529953,
      "learning_rate": 0.00031594766619519093,
      "loss": 2.201828718185425,
      "step": 8034
    },
    {
      "epoch": 0.8184091128978873,
      "grad_norm": 0.12417546659708023,
      "learning_rate": 0.0003155940594059406,
      "loss": 2.1709675788879395,
      "step": 8036
    },
    {
      "epoch": 0.818612798590495,
      "grad_norm": 0.11010025441646576,
      "learning_rate": 0.00031524045261669027,
      "loss": 2.2036774158477783,
      "step": 8038
    },
    {
      "epoch": 0.8188164842831027,
      "grad_norm": 0.11170239001512527,
      "learning_rate": 0.0003148868458274399,
      "loss": 2.1654257774353027,
      "step": 8040
    },
    {
      "epoch": 0.8190201699757105,
      "grad_norm": 0.1297641098499298,
      "learning_rate": 0.00031453323903818956,
      "loss": 2.132801055908203,
      "step": 8042
    },
    {
      "epoch": 0.8192238556683182,
      "grad_norm": 0.11011791229248047,
      "learning_rate": 0.00031417963224893917,
      "loss": 2.092798948287964,
      "step": 8044
    },
    {
      "epoch": 0.819427541360926,
      "grad_norm": 0.10870732367038727,
      "learning_rate": 0.00031382602545968884,
      "loss": 2.265456199645996,
      "step": 8046
    },
    {
      "epoch": 0.8196312270535336,
      "grad_norm": 0.11831321567296982,
      "learning_rate": 0.00031347241867043846,
      "loss": 2.1301932334899902,
      "step": 8048
    },
    {
      "epoch": 0.8198349127461414,
      "grad_norm": 0.11510367691516876,
      "learning_rate": 0.0003131188118811881,
      "loss": 2.139995574951172,
      "step": 8050
    },
    {
      "epoch": 0.8200385984387492,
      "grad_norm": 0.13694868981838226,
      "learning_rate": 0.0003127652050919378,
      "loss": 2.1256093978881836,
      "step": 8052
    },
    {
      "epoch": 0.8202422841313569,
      "grad_norm": 0.10975997149944305,
      "learning_rate": 0.0003124115983026874,
      "loss": 2.1189522743225098,
      "step": 8054
    },
    {
      "epoch": 0.8204459698239647,
      "grad_norm": 0.14725029468536377,
      "learning_rate": 0.0003120579915134371,
      "loss": 2.1858015060424805,
      "step": 8056
    },
    {
      "epoch": 0.8206496555165724,
      "grad_norm": 0.1016884595155716,
      "learning_rate": 0.0003117043847241867,
      "loss": 2.0896153450012207,
      "step": 8058
    },
    {
      "epoch": 0.8208533412091801,
      "grad_norm": 0.11884958297014236,
      "learning_rate": 0.00031135077793493636,
      "loss": 2.1291792392730713,
      "step": 8060
    },
    {
      "epoch": 0.8210570269017878,
      "grad_norm": 0.12609048187732697,
      "learning_rate": 0.000310997171145686,
      "loss": 2.1893134117126465,
      "step": 8062
    },
    {
      "epoch": 0.8212607125943956,
      "grad_norm": 0.17347615957260132,
      "learning_rate": 0.00031064356435643565,
      "loss": 2.260197162628174,
      "step": 8064
    },
    {
      "epoch": 0.8214643982870034,
      "grad_norm": 0.13777856528759003,
      "learning_rate": 0.0003102899575671853,
      "loss": 2.1995537281036377,
      "step": 8066
    },
    {
      "epoch": 0.8216680839796111,
      "grad_norm": 0.1226726546883583,
      "learning_rate": 0.00030993635077793493,
      "loss": 2.075878381729126,
      "step": 8068
    },
    {
      "epoch": 0.8218717696722188,
      "grad_norm": 0.1311327964067459,
      "learning_rate": 0.0003095827439886846,
      "loss": 2.1719696521759033,
      "step": 8070
    },
    {
      "epoch": 0.8220754553648265,
      "grad_norm": 0.11353713274002075,
      "learning_rate": 0.0003092291371994342,
      "loss": 2.078310012817383,
      "step": 8072
    },
    {
      "epoch": 0.8222791410574343,
      "grad_norm": 0.11187119036912918,
      "learning_rate": 0.0003088755304101839,
      "loss": 1.971495270729065,
      "step": 8074
    },
    {
      "epoch": 0.822482826750042,
      "grad_norm": 0.13391372561454773,
      "learning_rate": 0.0003085219236209335,
      "loss": 2.265627861022949,
      "step": 8076
    },
    {
      "epoch": 0.8226865124426498,
      "grad_norm": 0.11978642642498016,
      "learning_rate": 0.00030816831683168317,
      "loss": 2.1827898025512695,
      "step": 8078
    },
    {
      "epoch": 0.8228901981352574,
      "grad_norm": 0.11301986128091812,
      "learning_rate": 0.00030781471004243284,
      "loss": 2.2173566818237305,
      "step": 8080
    },
    {
      "epoch": 0.8230938838278652,
      "grad_norm": 0.15681442618370056,
      "learning_rate": 0.0003074611032531825,
      "loss": 2.20855450630188,
      "step": 8082
    },
    {
      "epoch": 0.823297569520473,
      "grad_norm": 0.12136926501989365,
      "learning_rate": 0.0003071074964639321,
      "loss": 2.102145195007324,
      "step": 8084
    },
    {
      "epoch": 0.8235012552130807,
      "grad_norm": 0.12810854613780975,
      "learning_rate": 0.00030675388967468174,
      "loss": 2.146660089492798,
      "step": 8086
    },
    {
      "epoch": 0.8237049409056885,
      "grad_norm": 0.1437087059020996,
      "learning_rate": 0.0003064002828854314,
      "loss": 2.1165590286254883,
      "step": 8088
    },
    {
      "epoch": 0.8239086265982962,
      "grad_norm": 0.12178312987089157,
      "learning_rate": 0.000306046676096181,
      "loss": 2.099893093109131,
      "step": 8090
    },
    {
      "epoch": 0.8241123122909039,
      "grad_norm": 0.13698460161685944,
      "learning_rate": 0.0003056930693069307,
      "loss": 2.222064971923828,
      "step": 8092
    },
    {
      "epoch": 0.8243159979835116,
      "grad_norm": 0.11475176364183426,
      "learning_rate": 0.00030533946251768036,
      "loss": 2.220221757888794,
      "step": 8094
    },
    {
      "epoch": 0.8245196836761194,
      "grad_norm": 0.10099946707487106,
      "learning_rate": 0.00030498585572843003,
      "loss": 2.192148208618164,
      "step": 8096
    },
    {
      "epoch": 0.8247233693687271,
      "grad_norm": 0.13495968282222748,
      "learning_rate": 0.00030463224893917965,
      "loss": 2.1532557010650635,
      "step": 8098
    },
    {
      "epoch": 0.8249270550613349,
      "grad_norm": 0.121132493019104,
      "learning_rate": 0.00030427864214992926,
      "loss": 2.218238353729248,
      "step": 8100
    },
    {
      "epoch": 0.8251307407539425,
      "grad_norm": 0.11693278700113297,
      "learning_rate": 0.00030392503536067893,
      "loss": 2.103487968444824,
      "step": 8102
    },
    {
      "epoch": 0.8253344264465503,
      "grad_norm": 0.13688667118549347,
      "learning_rate": 0.00030357142857142855,
      "loss": 2.2046384811401367,
      "step": 8104
    },
    {
      "epoch": 0.8255381121391581,
      "grad_norm": 0.10592490434646606,
      "learning_rate": 0.0003032178217821782,
      "loss": 2.29494047164917,
      "step": 8106
    },
    {
      "epoch": 0.8257417978317658,
      "grad_norm": 0.1287696212530136,
      "learning_rate": 0.0003028642149929279,
      "loss": 2.1722428798675537,
      "step": 8108
    },
    {
      "epoch": 0.8259454835243736,
      "grad_norm": 0.12433689087629318,
      "learning_rate": 0.00030251060820367755,
      "loss": 2.2682852745056152,
      "step": 8110
    },
    {
      "epoch": 0.8261491692169812,
      "grad_norm": 0.12177790701389313,
      "learning_rate": 0.00030215700141442717,
      "loss": 2.1335549354553223,
      "step": 8112
    },
    {
      "epoch": 0.826352854909589,
      "grad_norm": 0.11575634777545929,
      "learning_rate": 0.0003018033946251768,
      "loss": 2.332488536834717,
      "step": 8114
    },
    {
      "epoch": 0.8265565406021967,
      "grad_norm": 0.10892880707979202,
      "learning_rate": 0.00030144978783592645,
      "loss": 2.054640293121338,
      "step": 8116
    },
    {
      "epoch": 0.8267602262948045,
      "grad_norm": 0.11481916159391403,
      "learning_rate": 0.00030109618104667607,
      "loss": 1.8858925104141235,
      "step": 8118
    },
    {
      "epoch": 0.8269639119874123,
      "grad_norm": 0.11296623945236206,
      "learning_rate": 0.00030074257425742574,
      "loss": 2.2703938484191895,
      "step": 8120
    },
    {
      "epoch": 0.8271675976800199,
      "grad_norm": 0.12593531608581543,
      "learning_rate": 0.0003003889674681754,
      "loss": 2.1617164611816406,
      "step": 8122
    },
    {
      "epoch": 0.8273712833726277,
      "grad_norm": 0.15464521944522858,
      "learning_rate": 0.0003000353606789251,
      "loss": 2.0424087047576904,
      "step": 8124
    },
    {
      "epoch": 0.8275749690652354,
      "grad_norm": 0.11228951811790466,
      "learning_rate": 0.0002996817538896747,
      "loss": 2.165393829345703,
      "step": 8126
    },
    {
      "epoch": 0.8277786547578432,
      "grad_norm": 0.12009846419095993,
      "learning_rate": 0.0002993281471004243,
      "loss": 2.2267022132873535,
      "step": 8128
    },
    {
      "epoch": 0.8279823404504509,
      "grad_norm": 0.12092482298612595,
      "learning_rate": 0.000298974540311174,
      "loss": 2.14404296875,
      "step": 8130
    },
    {
      "epoch": 0.8281860261430587,
      "grad_norm": 0.10342790186405182,
      "learning_rate": 0.00029862093352192365,
      "loss": 2.279728651046753,
      "step": 8132
    },
    {
      "epoch": 0.8283897118356663,
      "grad_norm": 0.1312059760093689,
      "learning_rate": 0.00029826732673267326,
      "loss": 2.3124895095825195,
      "step": 8134
    },
    {
      "epoch": 0.8285933975282741,
      "grad_norm": 0.1155388280749321,
      "learning_rate": 0.00029791371994342293,
      "loss": 2.1318321228027344,
      "step": 8136
    },
    {
      "epoch": 0.8287970832208819,
      "grad_norm": 0.13187064230442047,
      "learning_rate": 0.0002975601131541726,
      "loss": 2.1411540508270264,
      "step": 8138
    },
    {
      "epoch": 0.8290007689134896,
      "grad_norm": 0.12355957180261612,
      "learning_rate": 0.0002972065063649222,
      "loss": 2.227459669113159,
      "step": 8140
    },
    {
      "epoch": 0.8292044546060974,
      "grad_norm": 0.11651693284511566,
      "learning_rate": 0.00029685289957567183,
      "loss": 2.219754219055176,
      "step": 8142
    },
    {
      "epoch": 0.829408140298705,
      "grad_norm": 0.1323528289794922,
      "learning_rate": 0.0002964992927864215,
      "loss": 2.2001566886901855,
      "step": 8144
    },
    {
      "epoch": 0.8296118259913128,
      "grad_norm": 0.1294148862361908,
      "learning_rate": 0.00029614568599717117,
      "loss": 2.0082762241363525,
      "step": 8146
    },
    {
      "epoch": 0.8298155116839205,
      "grad_norm": 0.15559935569763184,
      "learning_rate": 0.0002957920792079208,
      "loss": 2.2428364753723145,
      "step": 8148
    },
    {
      "epoch": 0.8300191973765283,
      "grad_norm": 0.11053565889596939,
      "learning_rate": 0.00029543847241867045,
      "loss": 2.0708746910095215,
      "step": 8150
    },
    {
      "epoch": 0.830222883069136,
      "grad_norm": 0.10959014296531677,
      "learning_rate": 0.0002950848656294201,
      "loss": 2.2283148765563965,
      "step": 8152
    },
    {
      "epoch": 0.8304265687617437,
      "grad_norm": 0.1112988218665123,
      "learning_rate": 0.00029473125884016974,
      "loss": 2.0667216777801514,
      "step": 8154
    },
    {
      "epoch": 0.8306302544543515,
      "grad_norm": 0.1151966080069542,
      "learning_rate": 0.00029437765205091935,
      "loss": 2.156421184539795,
      "step": 8156
    },
    {
      "epoch": 0.8308339401469592,
      "grad_norm": 0.10428447276353836,
      "learning_rate": 0.000294024045261669,
      "loss": 2.137812614440918,
      "step": 8158
    },
    {
      "epoch": 0.831037625839567,
      "grad_norm": 0.11743973940610886,
      "learning_rate": 0.0002936704384724187,
      "loss": 2.2097039222717285,
      "step": 8160
    },
    {
      "epoch": 0.8312413115321747,
      "grad_norm": 0.13384787738323212,
      "learning_rate": 0.0002933168316831683,
      "loss": 2.1719961166381836,
      "step": 8162
    },
    {
      "epoch": 0.8314449972247825,
      "grad_norm": 0.13763411343097687,
      "learning_rate": 0.000292963224893918,
      "loss": 2.3455810546875,
      "step": 8164
    },
    {
      "epoch": 0.8316486829173901,
      "grad_norm": 0.1144433319568634,
      "learning_rate": 0.00029260961810466764,
      "loss": 2.2283341884613037,
      "step": 8166
    },
    {
      "epoch": 0.8318523686099979,
      "grad_norm": 0.1701311618089676,
      "learning_rate": 0.00029225601131541726,
      "loss": 2.039825201034546,
      "step": 8168
    },
    {
      "epoch": 0.8320560543026057,
      "grad_norm": 0.11572291702032089,
      "learning_rate": 0.0002919024045261669,
      "loss": 2.183323860168457,
      "step": 8170
    },
    {
      "epoch": 0.8322597399952134,
      "grad_norm": 0.12210246920585632,
      "learning_rate": 0.00029154879773691654,
      "loss": 2.1413285732269287,
      "step": 8172
    },
    {
      "epoch": 0.8324634256878212,
      "grad_norm": 0.11230811476707458,
      "learning_rate": 0.0002911951909476662,
      "loss": 2.191534996032715,
      "step": 8174
    },
    {
      "epoch": 0.8326671113804288,
      "grad_norm": 0.11386552453041077,
      "learning_rate": 0.00029084158415841583,
      "loss": 2.2909979820251465,
      "step": 8176
    },
    {
      "epoch": 0.8328707970730366,
      "grad_norm": 0.11456192284822464,
      "learning_rate": 0.0002904879773691655,
      "loss": 2.216874122619629,
      "step": 8178
    },
    {
      "epoch": 0.8330744827656443,
      "grad_norm": 0.10446296632289886,
      "learning_rate": 0.00029013437057991517,
      "loss": 2.1059224605560303,
      "step": 8180
    },
    {
      "epoch": 0.8332781684582521,
      "grad_norm": 0.12611356377601624,
      "learning_rate": 0.00028978076379066484,
      "loss": 2.232449531555176,
      "step": 8182
    },
    {
      "epoch": 0.8334818541508598,
      "grad_norm": 0.11137253046035767,
      "learning_rate": 0.0002894271570014144,
      "loss": 2.3332977294921875,
      "step": 8184
    },
    {
      "epoch": 0.8336855398434675,
      "grad_norm": 0.1446768045425415,
      "learning_rate": 0.00028907355021216407,
      "loss": 2.346497058868408,
      "step": 8186
    },
    {
      "epoch": 0.8338892255360753,
      "grad_norm": 0.120219886302948,
      "learning_rate": 0.00028871994342291374,
      "loss": 2.1025195121765137,
      "step": 8188
    },
    {
      "epoch": 0.834092911228683,
      "grad_norm": 0.1158127561211586,
      "learning_rate": 0.00028836633663366335,
      "loss": 2.1734864711761475,
      "step": 8190
    },
    {
      "epoch": 0.8342965969212908,
      "grad_norm": 0.11088211834430695,
      "learning_rate": 0.000288012729844413,
      "loss": 2.150383949279785,
      "step": 8192
    },
    {
      "epoch": 0.8345002826138985,
      "grad_norm": 0.11049996316432953,
      "learning_rate": 0.0002876591230551627,
      "loss": 2.2497141361236572,
      "step": 8194
    },
    {
      "epoch": 0.8347039683065063,
      "grad_norm": 0.11799979954957962,
      "learning_rate": 0.00028730551626591236,
      "loss": 2.054424285888672,
      "step": 8196
    },
    {
      "epoch": 0.8349076539991139,
      "grad_norm": 0.13123847544193268,
      "learning_rate": 0.0002869519094766619,
      "loss": 2.219024658203125,
      "step": 8198
    },
    {
      "epoch": 0.8351113396917217,
      "grad_norm": 0.11464262753725052,
      "learning_rate": 0.0002865983026874116,
      "loss": 2.2179956436157227,
      "step": 8200
    },
    {
      "epoch": 0.8353150253843294,
      "grad_norm": 0.14699681103229523,
      "learning_rate": 0.00028624469589816126,
      "loss": 2.199876308441162,
      "step": 8202
    },
    {
      "epoch": 0.8355187110769372,
      "grad_norm": 0.15808647871017456,
      "learning_rate": 0.0002858910891089109,
      "loss": 2.2072510719299316,
      "step": 8204
    },
    {
      "epoch": 0.835722396769545,
      "grad_norm": 0.11796831339597702,
      "learning_rate": 0.00028553748231966054,
      "loss": 2.159189224243164,
      "step": 8206
    },
    {
      "epoch": 0.8359260824621526,
      "grad_norm": 0.14400368928909302,
      "learning_rate": 0.0002851838755304102,
      "loss": 2.3204164505004883,
      "step": 8208
    },
    {
      "epoch": 0.8361297681547604,
      "grad_norm": 0.12601636350154877,
      "learning_rate": 0.0002848302687411599,
      "loss": 2.2370080947875977,
      "step": 8210
    },
    {
      "epoch": 0.8363334538473681,
      "grad_norm": 0.10966749489307404,
      "learning_rate": 0.00028447666195190944,
      "loss": 2.158388614654541,
      "step": 8212
    },
    {
      "epoch": 0.8365371395399759,
      "grad_norm": 0.12567226588726044,
      "learning_rate": 0.0002841230551626591,
      "loss": 2.3187856674194336,
      "step": 8214
    },
    {
      "epoch": 0.8367408252325836,
      "grad_norm": 0.10295213758945465,
      "learning_rate": 0.0002837694483734088,
      "loss": 2.1725988388061523,
      "step": 8216
    },
    {
      "epoch": 0.8369445109251913,
      "grad_norm": 0.11227910965681076,
      "learning_rate": 0.0002834158415841584,
      "loss": 2.0376136302948,
      "step": 8218
    },
    {
      "epoch": 0.837148196617799,
      "grad_norm": 0.12836399674415588,
      "learning_rate": 0.00028306223479490807,
      "loss": 2.1986751556396484,
      "step": 8220
    },
    {
      "epoch": 0.8373518823104068,
      "grad_norm": 0.12827168405056,
      "learning_rate": 0.00028270862800565774,
      "loss": 2.183912992477417,
      "step": 8222
    },
    {
      "epoch": 0.8375555680030146,
      "grad_norm": 0.1480412781238556,
      "learning_rate": 0.0002823550212164074,
      "loss": 2.2657904624938965,
      "step": 8224
    },
    {
      "epoch": 0.8377592536956223,
      "grad_norm": 0.09822309762239456,
      "learning_rate": 0.00028200141442715697,
      "loss": 2.1148505210876465,
      "step": 8226
    },
    {
      "epoch": 0.8379629393882301,
      "grad_norm": 0.13883230090141296,
      "learning_rate": 0.00028164780763790663,
      "loss": 2.3130335807800293,
      "step": 8228
    },
    {
      "epoch": 0.8381666250808377,
      "grad_norm": 0.10887761414051056,
      "learning_rate": 0.0002812942008486563,
      "loss": 2.1761960983276367,
      "step": 8230
    },
    {
      "epoch": 0.8383703107734455,
      "grad_norm": 0.10945425927639008,
      "learning_rate": 0.000280940594059406,
      "loss": 2.0355682373046875,
      "step": 8232
    },
    {
      "epoch": 0.8385739964660532,
      "grad_norm": 0.13266301155090332,
      "learning_rate": 0.0002805869872701556,
      "loss": 2.185271739959717,
      "step": 8234
    },
    {
      "epoch": 0.838777682158661,
      "grad_norm": 0.11107184737920761,
      "learning_rate": 0.00028023338048090526,
      "loss": 2.159743309020996,
      "step": 8236
    },
    {
      "epoch": 0.8389813678512688,
      "grad_norm": 0.1162906363606453,
      "learning_rate": 0.0002798797736916549,
      "loss": 2.1305041313171387,
      "step": 8238
    },
    {
      "epoch": 0.8391850535438764,
      "grad_norm": 0.11628192663192749,
      "learning_rate": 0.0002795261669024045,
      "loss": 2.1309189796447754,
      "step": 8240
    },
    {
      "epoch": 0.8393887392364842,
      "grad_norm": 0.12799851596355438,
      "learning_rate": 0.00027917256011315416,
      "loss": 2.1863956451416016,
      "step": 8242
    },
    {
      "epoch": 0.8395924249290919,
      "grad_norm": 0.1126675233244896,
      "learning_rate": 0.0002788189533239038,
      "loss": 2.243083953857422,
      "step": 8244
    },
    {
      "epoch": 0.8397961106216997,
      "grad_norm": 0.1194322258234024,
      "learning_rate": 0.0002784653465346535,
      "loss": 2.230428695678711,
      "step": 8246
    },
    {
      "epoch": 0.8399997963143074,
      "grad_norm": 0.12561801075935364,
      "learning_rate": 0.0002781117397454031,
      "loss": 2.2583255767822266,
      "step": 8248
    },
    {
      "epoch": 0.8402034820069151,
      "grad_norm": 0.13620789349079132,
      "learning_rate": 0.0002777581329561528,
      "loss": 2.211564779281616,
      "step": 8250
    },
    {
      "epoch": 0.8404071676995228,
      "grad_norm": 0.13178542256355286,
      "learning_rate": 0.00027740452616690245,
      "loss": 1.9948687553405762,
      "step": 8252
    },
    {
      "epoch": 0.8406108533921306,
      "grad_norm": 0.12301908433437347,
      "learning_rate": 0.000277050919377652,
      "loss": 2.093050479888916,
      "step": 8254
    },
    {
      "epoch": 0.8408145390847384,
      "grad_norm": 0.11671198159456253,
      "learning_rate": 0.0002766973125884017,
      "loss": 2.0945258140563965,
      "step": 8256
    },
    {
      "epoch": 0.8410182247773461,
      "grad_norm": 0.1272357702255249,
      "learning_rate": 0.00027634370579915135,
      "loss": 2.231387138366699,
      "step": 8258
    },
    {
      "epoch": 0.8412219104699538,
      "grad_norm": 0.13541650772094727,
      "learning_rate": 0.000275990099009901,
      "loss": 2.085049629211426,
      "step": 8260
    },
    {
      "epoch": 0.8414255961625615,
      "grad_norm": 0.13152475655078888,
      "learning_rate": 0.00027563649222065063,
      "loss": 2.0924694538116455,
      "step": 8262
    },
    {
      "epoch": 0.8416292818551693,
      "grad_norm": 0.11533188074827194,
      "learning_rate": 0.0002752828854314003,
      "loss": 2.2445003986358643,
      "step": 8264
    },
    {
      "epoch": 0.841832967547777,
      "grad_norm": 0.10579276829957962,
      "learning_rate": 0.00027492927864214997,
      "loss": 2.2019500732421875,
      "step": 8266
    },
    {
      "epoch": 0.8420366532403848,
      "grad_norm": 0.11459474265575409,
      "learning_rate": 0.00027457567185289953,
      "loss": 2.054989814758301,
      "step": 8268
    },
    {
      "epoch": 0.8422403389329925,
      "grad_norm": 0.12538108229637146,
      "learning_rate": 0.0002742220650636492,
      "loss": 2.2099571228027344,
      "step": 8270
    },
    {
      "epoch": 0.8424440246256002,
      "grad_norm": 0.11322376877069473,
      "learning_rate": 0.00027386845827439887,
      "loss": 2.172736167907715,
      "step": 8272
    },
    {
      "epoch": 0.842647710318208,
      "grad_norm": 0.10338419675827026,
      "learning_rate": 0.00027351485148514854,
      "loss": 2.133481502532959,
      "step": 8274
    },
    {
      "epoch": 0.8428513960108157,
      "grad_norm": 0.11346311122179031,
      "learning_rate": 0.00027316124469589816,
      "loss": 2.2061805725097656,
      "step": 8276
    },
    {
      "epoch": 0.8430550817034235,
      "grad_norm": 0.10214720666408539,
      "learning_rate": 0.0002728076379066478,
      "loss": 2.0279951095581055,
      "step": 8278
    },
    {
      "epoch": 0.8432587673960312,
      "grad_norm": 0.1339000165462494,
      "learning_rate": 0.0002724540311173975,
      "loss": 2.3271613121032715,
      "step": 8280
    },
    {
      "epoch": 0.8434624530886389,
      "grad_norm": 0.12192840874195099,
      "learning_rate": 0.0002721004243281471,
      "loss": 2.211956739425659,
      "step": 8282
    },
    {
      "epoch": 0.8436661387812466,
      "grad_norm": 0.1460392326116562,
      "learning_rate": 0.0002717468175388967,
      "loss": 2.1759257316589355,
      "step": 8284
    },
    {
      "epoch": 0.8438698244738544,
      "grad_norm": 0.11992810666561127,
      "learning_rate": 0.0002713932107496464,
      "loss": 2.283182382583618,
      "step": 8286
    },
    {
      "epoch": 0.8440735101664621,
      "grad_norm": 0.10964006930589676,
      "learning_rate": 0.00027103960396039606,
      "loss": 2.2879862785339355,
      "step": 8288
    },
    {
      "epoch": 0.8442771958590699,
      "grad_norm": 0.0993550717830658,
      "learning_rate": 0.0002706859971711457,
      "loss": 2.1982622146606445,
      "step": 8290
    },
    {
      "epoch": 0.8444808815516776,
      "grad_norm": 0.11887910217046738,
      "learning_rate": 0.00027033239038189535,
      "loss": 2.051905870437622,
      "step": 8292
    },
    {
      "epoch": 0.8446845672442853,
      "grad_norm": 0.11579185724258423,
      "learning_rate": 0.000269978783592645,
      "loss": 2.1379499435424805,
      "step": 8294
    },
    {
      "epoch": 0.8448882529368931,
      "grad_norm": 0.13413473963737488,
      "learning_rate": 0.00026962517680339463,
      "loss": 2.3044910430908203,
      "step": 8296
    },
    {
      "epoch": 0.8450919386295008,
      "grad_norm": 0.14118216931819916,
      "learning_rate": 0.00026927157001414425,
      "loss": 2.1587095260620117,
      "step": 8298
    },
    {
      "epoch": 0.8452956243221086,
      "grad_norm": 0.14223818480968475,
      "learning_rate": 0.0002689179632248939,
      "loss": 2.207369804382324,
      "step": 8300
    },
    {
      "epoch": 0.8454993100147163,
      "grad_norm": 0.10472667217254639,
      "learning_rate": 0.0002685643564356436,
      "loss": 2.2701778411865234,
      "step": 8302
    },
    {
      "epoch": 0.845702995707324,
      "grad_norm": 0.15801021456718445,
      "learning_rate": 0.0002682107496463932,
      "loss": 2.291372537612915,
      "step": 8304
    },
    {
      "epoch": 0.8459066813999317,
      "grad_norm": 0.11322212219238281,
      "learning_rate": 0.00026785714285714287,
      "loss": 2.1924514770507812,
      "step": 8306
    },
    {
      "epoch": 0.8461103670925395,
      "grad_norm": 0.1068267971277237,
      "learning_rate": 0.00026750353606789254,
      "loss": 2.131700038909912,
      "step": 8308
    },
    {
      "epoch": 0.8463140527851473,
      "grad_norm": 0.1100311204791069,
      "learning_rate": 0.00026714992927864215,
      "loss": 2.032182216644287,
      "step": 8310
    },
    {
      "epoch": 0.846517738477755,
      "grad_norm": 0.14688874781131744,
      "learning_rate": 0.00026679632248939177,
      "loss": 2.0810444355010986,
      "step": 8312
    },
    {
      "epoch": 0.8467214241703627,
      "grad_norm": 0.1112656444311142,
      "learning_rate": 0.00026644271570014144,
      "loss": 2.132812023162842,
      "step": 8314
    },
    {
      "epoch": 0.8469251098629704,
      "grad_norm": 0.11189007014036179,
      "learning_rate": 0.0002660891089108911,
      "loss": 2.1007399559020996,
      "step": 8316
    },
    {
      "epoch": 0.8471287955555782,
      "grad_norm": 0.12699809670448303,
      "learning_rate": 0.0002657355021216407,
      "loss": 2.0912351608276367,
      "step": 8318
    },
    {
      "epoch": 0.8473324812481859,
      "grad_norm": 0.11670741438865662,
      "learning_rate": 0.0002653818953323904,
      "loss": 2.0917186737060547,
      "step": 8320
    },
    {
      "epoch": 0.8475361669407937,
      "grad_norm": 0.1360200047492981,
      "learning_rate": 0.00026502828854314006,
      "loss": 2.2597107887268066,
      "step": 8322
    },
    {
      "epoch": 0.8477398526334013,
      "grad_norm": 0.1223158985376358,
      "learning_rate": 0.0002646746817538897,
      "loss": 2.169398307800293,
      "step": 8324
    },
    {
      "epoch": 0.8479435383260091,
      "grad_norm": 0.11410562694072723,
      "learning_rate": 0.0002643210749646393,
      "loss": 2.2131333351135254,
      "step": 8326
    },
    {
      "epoch": 0.8481472240186169,
      "grad_norm": 0.12003970891237259,
      "learning_rate": 0.00026396746817538896,
      "loss": 2.053128480911255,
      "step": 8328
    },
    {
      "epoch": 0.8483509097112246,
      "grad_norm": 0.11861306428909302,
      "learning_rate": 0.00026361386138613863,
      "loss": 2.3377788066864014,
      "step": 8330
    },
    {
      "epoch": 0.8485545954038324,
      "grad_norm": 0.12225455045700073,
      "learning_rate": 0.0002632602545968883,
      "loss": 2.1822712421417236,
      "step": 8332
    },
    {
      "epoch": 0.8487582810964401,
      "grad_norm": 0.1247875913977623,
      "learning_rate": 0.0002629066478076379,
      "loss": 2.2435860633850098,
      "step": 8334
    },
    {
      "epoch": 0.8489619667890478,
      "grad_norm": 0.13966140151023865,
      "learning_rate": 0.0002625530410183876,
      "loss": 2.1572821140289307,
      "step": 8336
    },
    {
      "epoch": 0.8491656524816555,
      "grad_norm": 0.122560515999794,
      "learning_rate": 0.0002621994342291372,
      "loss": 2.1824545860290527,
      "step": 8338
    },
    {
      "epoch": 0.8493693381742633,
      "grad_norm": 0.13850367069244385,
      "learning_rate": 0.0002618458274398868,
      "loss": 2.0465328693389893,
      "step": 8340
    },
    {
      "epoch": 0.8495730238668711,
      "grad_norm": 0.13734789192676544,
      "learning_rate": 0.0002614922206506365,
      "loss": 2.1171507835388184,
      "step": 8342
    },
    {
      "epoch": 0.8497767095594788,
      "grad_norm": 0.10661063343286514,
      "learning_rate": 0.00026113861386138615,
      "loss": 2.153157949447632,
      "step": 8344
    },
    {
      "epoch": 0.8499803952520865,
      "grad_norm": 0.11612168699502945,
      "learning_rate": 0.0002607850070721358,
      "loss": 2.22538161277771,
      "step": 8346
    },
    {
      "epoch": 0.8501840809446942,
      "grad_norm": 0.15666557848453522,
      "learning_rate": 0.00026043140028288544,
      "loss": 2.093848705291748,
      "step": 8348
    },
    {
      "epoch": 0.850387766637302,
      "grad_norm": 0.12051813304424286,
      "learning_rate": 0.0002600777934936351,
      "loss": 2.208604335784912,
      "step": 8350
    },
    {
      "epoch": 0.8505914523299097,
      "grad_norm": 0.12767551839351654,
      "learning_rate": 0.0002597241867043847,
      "loss": 2.268094539642334,
      "step": 8352
    },
    {
      "epoch": 0.8507951380225175,
      "grad_norm": 0.0986771434545517,
      "learning_rate": 0.00025937057991513434,
      "loss": 2.102081537246704,
      "step": 8354
    },
    {
      "epoch": 0.8509988237151251,
      "grad_norm": 0.11310850828886032,
      "learning_rate": 0.000259016973125884,
      "loss": 2.111302614212036,
      "step": 8356
    },
    {
      "epoch": 0.8512025094077329,
      "grad_norm": 0.1145244836807251,
      "learning_rate": 0.0002586633663366337,
      "loss": 2.2429986000061035,
      "step": 8358
    },
    {
      "epoch": 0.8514061951003407,
      "grad_norm": 0.12063159793615341,
      "learning_rate": 0.00025830975954738335,
      "loss": 2.2087011337280273,
      "step": 8360
    },
    {
      "epoch": 0.8516098807929484,
      "grad_norm": 0.1408926248550415,
      "learning_rate": 0.00025795615275813296,
      "loss": 2.1767945289611816,
      "step": 8362
    },
    {
      "epoch": 0.8518135664855562,
      "grad_norm": 0.12885715067386627,
      "learning_rate": 0.00025760254596888263,
      "loss": 2.175534725189209,
      "step": 8364
    },
    {
      "epoch": 0.8520172521781639,
      "grad_norm": 0.10878703743219376,
      "learning_rate": 0.00025724893917963225,
      "loss": 2.2103095054626465,
      "step": 8366
    },
    {
      "epoch": 0.8522209378707716,
      "grad_norm": 0.11623060703277588,
      "learning_rate": 0.00025689533239038186,
      "loss": 2.1428353786468506,
      "step": 8368
    },
    {
      "epoch": 0.8524246235633793,
      "grad_norm": 0.11249405890703201,
      "learning_rate": 0.00025654172560113153,
      "loss": 2.1615452766418457,
      "step": 8370
    },
    {
      "epoch": 0.8526283092559871,
      "grad_norm": 0.11520761251449585,
      "learning_rate": 0.0002561881188118812,
      "loss": 2.1071832180023193,
      "step": 8372
    },
    {
      "epoch": 0.8528319949485949,
      "grad_norm": 0.12151440232992172,
      "learning_rate": 0.00025583451202263087,
      "loss": 2.2071304321289062,
      "step": 8374
    },
    {
      "epoch": 0.8530356806412026,
      "grad_norm": 0.13034015893936157,
      "learning_rate": 0.0002554809052333805,
      "loss": 2.121131420135498,
      "step": 8376
    },
    {
      "epoch": 0.8532393663338103,
      "grad_norm": 0.14621728658676147,
      "learning_rate": 0.00025512729844413015,
      "loss": 2.2630138397216797,
      "step": 8378
    },
    {
      "epoch": 0.853443052026418,
      "grad_norm": 0.1391804963350296,
      "learning_rate": 0.00025477369165487977,
      "loss": 2.2119970321655273,
      "step": 8380
    },
    {
      "epoch": 0.8536467377190258,
      "grad_norm": 0.1220344603061676,
      "learning_rate": 0.00025442008486562944,
      "loss": 2.235598087310791,
      "step": 8382
    },
    {
      "epoch": 0.8538504234116335,
      "grad_norm": 0.12038014829158783,
      "learning_rate": 0.00025406647807637905,
      "loss": 2.319049596786499,
      "step": 8384
    },
    {
      "epoch": 0.8540541091042413,
      "grad_norm": 0.1351180225610733,
      "learning_rate": 0.0002537128712871287,
      "loss": 2.1155710220336914,
      "step": 8386
    },
    {
      "epoch": 0.8542577947968489,
      "grad_norm": 0.13637913763523102,
      "learning_rate": 0.0002533592644978784,
      "loss": 2.3453030586242676,
      "step": 8388
    },
    {
      "epoch": 0.8544614804894567,
      "grad_norm": 0.11665526777505875,
      "learning_rate": 0.000253005657708628,
      "loss": 2.2128748893737793,
      "step": 8390
    },
    {
      "epoch": 0.8546651661820645,
      "grad_norm": 0.11951042711734772,
      "learning_rate": 0.0002526520509193777,
      "loss": 2.207770347595215,
      "step": 8392
    },
    {
      "epoch": 0.8548688518746722,
      "grad_norm": 0.1752869188785553,
      "learning_rate": 0.0002522984441301273,
      "loss": 2.089493751525879,
      "step": 8394
    },
    {
      "epoch": 0.85507253756728,
      "grad_norm": 0.1205112412571907,
      "learning_rate": 0.00025194483734087696,
      "loss": 2.0937373638153076,
      "step": 8396
    },
    {
      "epoch": 0.8552762232598876,
      "grad_norm": 0.10692280530929565,
      "learning_rate": 0.0002515912305516266,
      "loss": 2.184696674346924,
      "step": 8398
    },
    {
      "epoch": 0.8554799089524954,
      "grad_norm": 0.11870463937520981,
      "learning_rate": 0.00025123762376237624,
      "loss": 2.2345457077026367,
      "step": 8400
    },
    {
      "epoch": 0.8556835946451031,
      "grad_norm": 0.11211320757865906,
      "learning_rate": 0.0002508840169731259,
      "loss": 2.2606935501098633,
      "step": 8402
    },
    {
      "epoch": 0.8558872803377109,
      "grad_norm": 0.12075284123420715,
      "learning_rate": 0.00025053041018387553,
      "loss": 2.2091240882873535,
      "step": 8404
    },
    {
      "epoch": 0.8560909660303186,
      "grad_norm": 0.117253877222538,
      "learning_rate": 0.0002501768033946252,
      "loss": 2.2250380516052246,
      "step": 8406
    },
    {
      "epoch": 0.8562946517229264,
      "grad_norm": 0.1285357028245926,
      "learning_rate": 0.0002498231966053748,
      "loss": 2.138946771621704,
      "step": 8408
    },
    {
      "epoch": 0.856498337415534,
      "grad_norm": 0.1256190687417984,
      "learning_rate": 0.0002494695898161245,
      "loss": 2.215114116668701,
      "step": 8410
    },
    {
      "epoch": 0.8567020231081418,
      "grad_norm": 0.1113329604268074,
      "learning_rate": 0.00024911598302687415,
      "loss": 2.1613717079162598,
      "step": 8412
    },
    {
      "epoch": 0.8569057088007496,
      "grad_norm": 0.12475811690092087,
      "learning_rate": 0.00024876237623762377,
      "loss": 2.212381362915039,
      "step": 8414
    },
    {
      "epoch": 0.8571093944933573,
      "grad_norm": 0.1165296658873558,
      "learning_rate": 0.0002484087694483734,
      "loss": 2.204674243927002,
      "step": 8416
    },
    {
      "epoch": 0.8573130801859651,
      "grad_norm": 0.11375758796930313,
      "learning_rate": 0.00024805516265912305,
      "loss": 2.286102771759033,
      "step": 8418
    },
    {
      "epoch": 0.8575167658785727,
      "grad_norm": 0.10971733927726746,
      "learning_rate": 0.0002477015558698727,
      "loss": 2.189349889755249,
      "step": 8420
    },
    {
      "epoch": 0.8577204515711805,
      "grad_norm": 0.11781566590070724,
      "learning_rate": 0.00024734794908062234,
      "loss": 2.2244186401367188,
      "step": 8422
    },
    {
      "epoch": 0.8579241372637882,
      "grad_norm": 0.11757552623748779,
      "learning_rate": 0.000246994342291372,
      "loss": 2.215592861175537,
      "step": 8424
    },
    {
      "epoch": 0.858127822956396,
      "grad_norm": 0.11571701616048813,
      "learning_rate": 0.0002466407355021217,
      "loss": 2.3399405479431152,
      "step": 8426
    },
    {
      "epoch": 0.8583315086490038,
      "grad_norm": 0.12327166646718979,
      "learning_rate": 0.0002462871287128713,
      "loss": 2.0645580291748047,
      "step": 8428
    },
    {
      "epoch": 0.8585351943416114,
      "grad_norm": 0.13052764534950256,
      "learning_rate": 0.00024593352192362096,
      "loss": 2.163888692855835,
      "step": 8430
    },
    {
      "epoch": 0.8587388800342192,
      "grad_norm": 0.1090330258011818,
      "learning_rate": 0.0002455799151343706,
      "loss": 2.1595637798309326,
      "step": 8432
    },
    {
      "epoch": 0.8589425657268269,
      "grad_norm": 0.13364097476005554,
      "learning_rate": 0.00024522630834512024,
      "loss": 2.29101300239563,
      "step": 8434
    },
    {
      "epoch": 0.8591462514194347,
      "grad_norm": 0.11617843061685562,
      "learning_rate": 0.00024487270155586986,
      "loss": 2.202939510345459,
      "step": 8436
    },
    {
      "epoch": 0.8593499371120424,
      "grad_norm": 0.1415223628282547,
      "learning_rate": 0.00024451909476661953,
      "loss": 2.1375296115875244,
      "step": 8438
    },
    {
      "epoch": 0.8595536228046502,
      "grad_norm": 0.12689262628555298,
      "learning_rate": 0.0002441654879773692,
      "loss": 2.2098135948181152,
      "step": 8440
    },
    {
      "epoch": 0.8597573084972578,
      "grad_norm": 0.11725840717554092,
      "learning_rate": 0.0002438118811881188,
      "loss": 2.2235193252563477,
      "step": 8442
    },
    {
      "epoch": 0.8599609941898656,
      "grad_norm": 0.13775768876075745,
      "learning_rate": 0.00024345827439886848,
      "loss": 2.184720516204834,
      "step": 8444
    },
    {
      "epoch": 0.8601646798824734,
      "grad_norm": 0.14161314070224762,
      "learning_rate": 0.0002431046676096181,
      "loss": 2.206542491912842,
      "step": 8446
    },
    {
      "epoch": 0.8603683655750811,
      "grad_norm": 0.10476462543010712,
      "learning_rate": 0.00024275106082036777,
      "loss": 2.2269973754882812,
      "step": 8448
    },
    {
      "epoch": 0.8605720512676889,
      "grad_norm": 0.12448402494192123,
      "learning_rate": 0.0002423974540311174,
      "loss": 2.1448426246643066,
      "step": 8450
    },
    {
      "epoch": 0.8607757369602965,
      "grad_norm": 0.11328263580799103,
      "learning_rate": 0.00024204384724186705,
      "loss": 2.1843338012695312,
      "step": 8452
    },
    {
      "epoch": 0.8609794226529043,
      "grad_norm": 0.1095777302980423,
      "learning_rate": 0.0002416902404526167,
      "loss": 2.105656623840332,
      "step": 8454
    },
    {
      "epoch": 0.861183108345512,
      "grad_norm": 0.1250767707824707,
      "learning_rate": 0.00024133663366336633,
      "loss": 2.1865129470825195,
      "step": 8456
    },
    {
      "epoch": 0.8613867940381198,
      "grad_norm": 0.13790586590766907,
      "learning_rate": 0.000240983026874116,
      "loss": 2.317653179168701,
      "step": 8458
    },
    {
      "epoch": 0.8615904797307276,
      "grad_norm": 0.12765635550022125,
      "learning_rate": 0.00024062942008486562,
      "loss": 2.127234697341919,
      "step": 8460
    },
    {
      "epoch": 0.8617941654233352,
      "grad_norm": 0.107661671936512,
      "learning_rate": 0.0002402758132956153,
      "loss": 2.0503127574920654,
      "step": 8462
    },
    {
      "epoch": 0.861997851115943,
      "grad_norm": 0.11359362304210663,
      "learning_rate": 0.00023992220650636493,
      "loss": 2.1298742294311523,
      "step": 8464
    },
    {
      "epoch": 0.8622015368085507,
      "grad_norm": 0.12243159860372543,
      "learning_rate": 0.00023956859971711457,
      "loss": 2.2245724201202393,
      "step": 8466
    },
    {
      "epoch": 0.8624052225011585,
      "grad_norm": 0.14036670327186584,
      "learning_rate": 0.00023921499292786421,
      "loss": 2.202219247817993,
      "step": 8468
    },
    {
      "epoch": 0.8626089081937662,
      "grad_norm": 0.11809424310922623,
      "learning_rate": 0.00023886138613861386,
      "loss": 2.049093008041382,
      "step": 8470
    },
    {
      "epoch": 0.862812593886374,
      "grad_norm": 0.11745058745145798,
      "learning_rate": 0.00023850777934936353,
      "loss": 2.0901360511779785,
      "step": 8472
    },
    {
      "epoch": 0.8630162795789816,
      "grad_norm": 0.1440880298614502,
      "learning_rate": 0.00023815417256011314,
      "loss": 2.2689685821533203,
      "step": 8474
    },
    {
      "epoch": 0.8632199652715894,
      "grad_norm": 0.10802867263555527,
      "learning_rate": 0.0002378005657708628,
      "loss": 2.195464611053467,
      "step": 8476
    },
    {
      "epoch": 0.8634236509641972,
      "grad_norm": 0.15347449481487274,
      "learning_rate": 0.00023744695898161245,
      "loss": 2.2079405784606934,
      "step": 8478
    },
    {
      "epoch": 0.8636273366568049,
      "grad_norm": 0.12634378671646118,
      "learning_rate": 0.00023709335219236212,
      "loss": 2.224449634552002,
      "step": 8480
    },
    {
      "epoch": 0.8638310223494127,
      "grad_norm": 0.14492622017860413,
      "learning_rate": 0.00023673974540311174,
      "loss": 2.2172038555145264,
      "step": 8482
    },
    {
      "epoch": 0.8640347080420203,
      "grad_norm": 0.11542665958404541,
      "learning_rate": 0.00023638613861386138,
      "loss": 2.1100192070007324,
      "step": 8484
    },
    {
      "epoch": 0.8642383937346281,
      "grad_norm": 0.1080252006649971,
      "learning_rate": 0.00023603253182461105,
      "loss": 2.1588706970214844,
      "step": 8486
    },
    {
      "epoch": 0.8644420794272358,
      "grad_norm": 0.13318414986133575,
      "learning_rate": 0.00023567892503536066,
      "loss": 2.3123390674591064,
      "step": 8488
    },
    {
      "epoch": 0.8646457651198436,
      "grad_norm": 0.12402119487524033,
      "learning_rate": 0.00023532531824611033,
      "loss": 2.244814872741699,
      "step": 8490
    },
    {
      "epoch": 0.8648494508124513,
      "grad_norm": 0.1350593864917755,
      "learning_rate": 0.00023497171145685998,
      "loss": 2.234158754348755,
      "step": 8492
    },
    {
      "epoch": 0.865053136505059,
      "grad_norm": 0.14904925227165222,
      "learning_rate": 0.00023461810466760964,
      "loss": 2.2257602214813232,
      "step": 8494
    },
    {
      "epoch": 0.8652568221976668,
      "grad_norm": 0.10348611325025558,
      "learning_rate": 0.00023426449787835926,
      "loss": 2.120408535003662,
      "step": 8496
    },
    {
      "epoch": 0.8654605078902745,
      "grad_norm": 0.14494602382183075,
      "learning_rate": 0.00023391089108910893,
      "loss": 2.198107957839966,
      "step": 8498
    },
    {
      "epoch": 0.8656641935828823,
      "grad_norm": 0.12150433659553528,
      "learning_rate": 0.00023355728429985857,
      "loss": 2.1607232093811035,
      "step": 8500
    },
    {
      "epoch": 0.86586787927549,
      "grad_norm": 0.10378558933734894,
      "learning_rate": 0.00023320367751060819,
      "loss": 2.0558764934539795,
      "step": 8502
    },
    {
      "epoch": 0.8660715649680977,
      "grad_norm": 0.11181873828172684,
      "learning_rate": 0.00023285007072135786,
      "loss": 2.1510961055755615,
      "step": 8504
    },
    {
      "epoch": 0.8662752506607054,
      "grad_norm": 0.11998780071735382,
      "learning_rate": 0.0002324964639321075,
      "loss": 2.096949577331543,
      "step": 8506
    },
    {
      "epoch": 0.8664789363533132,
      "grad_norm": 0.12660878896713257,
      "learning_rate": 0.00023214285714285717,
      "loss": 2.1523098945617676,
      "step": 8508
    },
    {
      "epoch": 0.866682622045921,
      "grad_norm": 0.14653544127941132,
      "learning_rate": 0.00023178925035360678,
      "loss": 2.231461524963379,
      "step": 8510
    },
    {
      "epoch": 0.8668863077385287,
      "grad_norm": 0.12010907381772995,
      "learning_rate": 0.00023143564356435645,
      "loss": 2.2240724563598633,
      "step": 8512
    },
    {
      "epoch": 0.8670899934311365,
      "grad_norm": 0.12040183693170547,
      "learning_rate": 0.0002310820367751061,
      "loss": 2.2657361030578613,
      "step": 8514
    },
    {
      "epoch": 0.8672936791237441,
      "grad_norm": 0.12833143770694733,
      "learning_rate": 0.0002307284299858557,
      "loss": 2.2078356742858887,
      "step": 8516
    },
    {
      "epoch": 0.8674973648163519,
      "grad_norm": 0.14156359434127808,
      "learning_rate": 0.00023037482319660538,
      "loss": 2.1271138191223145,
      "step": 8518
    },
    {
      "epoch": 0.8677010505089596,
      "grad_norm": 0.10809633135795593,
      "learning_rate": 0.00023002121640735502,
      "loss": 2.08061146736145,
      "step": 8520
    },
    {
      "epoch": 0.8679047362015674,
      "grad_norm": 0.15787367522716522,
      "learning_rate": 0.0002296676096181047,
      "loss": 2.252051830291748,
      "step": 8522
    },
    {
      "epoch": 0.8681084218941751,
      "grad_norm": 0.11143327504396439,
      "learning_rate": 0.0002293140028288543,
      "loss": 2.2866134643554688,
      "step": 8524
    },
    {
      "epoch": 0.8683121075867828,
      "grad_norm": 0.17430421710014343,
      "learning_rate": 0.00022896039603960397,
      "loss": 2.1505560874938965,
      "step": 8526
    },
    {
      "epoch": 0.8685157932793905,
      "grad_norm": 0.12307973206043243,
      "learning_rate": 0.00022860678925035362,
      "loss": 2.270566463470459,
      "step": 8528
    },
    {
      "epoch": 0.8687194789719983,
      "grad_norm": 0.11457561701536179,
      "learning_rate": 0.00022825318246110326,
      "loss": 2.187713146209717,
      "step": 8530
    },
    {
      "epoch": 0.8689231646646061,
      "grad_norm": 0.11708100885152817,
      "learning_rate": 0.0002278995756718529,
      "loss": 2.242441177368164,
      "step": 8532
    },
    {
      "epoch": 0.8691268503572138,
      "grad_norm": 0.1197577565908432,
      "learning_rate": 0.00022754596888260254,
      "loss": 2.1839301586151123,
      "step": 8534
    },
    {
      "epoch": 0.8693305360498215,
      "grad_norm": 0.1326673924922943,
      "learning_rate": 0.0002271923620933522,
      "loss": 2.1439943313598633,
      "step": 8536
    },
    {
      "epoch": 0.8695342217424292,
      "grad_norm": 0.13155077397823334,
      "learning_rate": 0.00022683875530410183,
      "loss": 2.1065292358398438,
      "step": 8538
    },
    {
      "epoch": 0.869737907435037,
      "grad_norm": 0.1408987045288086,
      "learning_rate": 0.0002264851485148515,
      "loss": 2.0899457931518555,
      "step": 8540
    },
    {
      "epoch": 0.8699415931276447,
      "grad_norm": 0.11619631946086884,
      "learning_rate": 0.00022613154172560114,
      "loss": 2.216660499572754,
      "step": 8542
    },
    {
      "epoch": 0.8701452788202525,
      "grad_norm": 0.11811387538909912,
      "learning_rate": 0.00022577793493635078,
      "loss": 2.1566035747528076,
      "step": 8544
    },
    {
      "epoch": 0.8703489645128603,
      "grad_norm": 0.144367054104805,
      "learning_rate": 0.00022542432814710042,
      "loss": 2.141432762145996,
      "step": 8546
    },
    {
      "epoch": 0.8705526502054679,
      "grad_norm": 0.13643218576908112,
      "learning_rate": 0.0002250707213578501,
      "loss": 2.2348856925964355,
      "step": 8548
    },
    {
      "epoch": 0.8707563358980757,
      "grad_norm": 0.12052277475595474,
      "learning_rate": 0.00022471711456859973,
      "loss": 2.1496362686157227,
      "step": 8550
    },
    {
      "epoch": 0.8709600215906834,
      "grad_norm": 0.12558241188526154,
      "learning_rate": 0.00022436350777934935,
      "loss": 2.294705629348755,
      "step": 8552
    },
    {
      "epoch": 0.8711637072832912,
      "grad_norm": 0.11195098608732224,
      "learning_rate": 0.00022400990099009902,
      "loss": 2.2785191535949707,
      "step": 8554
    },
    {
      "epoch": 0.8713673929758989,
      "grad_norm": 0.10720366984605789,
      "learning_rate": 0.00022365629420084866,
      "loss": 2.094865322113037,
      "step": 8556
    },
    {
      "epoch": 0.8715710786685066,
      "grad_norm": 0.1427599936723709,
      "learning_rate": 0.0002233026874115983,
      "loss": 2.2420663833618164,
      "step": 8558
    },
    {
      "epoch": 0.8717747643611143,
      "grad_norm": 0.11800504475831985,
      "learning_rate": 0.00022294908062234795,
      "loss": 2.1647143363952637,
      "step": 8560
    },
    {
      "epoch": 0.8719784500537221,
      "grad_norm": 0.12628668546676636,
      "learning_rate": 0.00022259547383309762,
      "loss": 2.2645249366760254,
      "step": 8562
    },
    {
      "epoch": 0.8721821357463299,
      "grad_norm": 0.13859009742736816,
      "learning_rate": 0.00022224186704384726,
      "loss": 2.266035795211792,
      "step": 8564
    },
    {
      "epoch": 0.8723858214389376,
      "grad_norm": 0.12860837578773499,
      "learning_rate": 0.00022188826025459687,
      "loss": 2.2051000595092773,
      "step": 8566
    },
    {
      "epoch": 0.8725895071315453,
      "grad_norm": 0.13687168061733246,
      "learning_rate": 0.00022153465346534654,
      "loss": 2.131985664367676,
      "step": 8568
    },
    {
      "epoch": 0.872793192824153,
      "grad_norm": 0.12343737483024597,
      "learning_rate": 0.00022118104667609618,
      "loss": 2.2190775871276855,
      "step": 8570
    },
    {
      "epoch": 0.8729968785167608,
      "grad_norm": 0.12059484422206879,
      "learning_rate": 0.00022082743988684583,
      "loss": 2.203526020050049,
      "step": 8572
    },
    {
      "epoch": 0.8732005642093685,
      "grad_norm": 0.13423702120780945,
      "learning_rate": 0.00022047383309759547,
      "loss": 2.323028802871704,
      "step": 8574
    },
    {
      "epoch": 0.8734042499019763,
      "grad_norm": 0.11954315751791,
      "learning_rate": 0.00022012022630834514,
      "loss": 2.2378478050231934,
      "step": 8576
    },
    {
      "epoch": 0.873607935594584,
      "grad_norm": 0.13080434501171112,
      "learning_rate": 0.00021976661951909478,
      "loss": 2.244542121887207,
      "step": 8578
    },
    {
      "epoch": 0.8738116212871917,
      "grad_norm": 0.11287268251180649,
      "learning_rate": 0.00021941301272984442,
      "loss": 2.0941715240478516,
      "step": 8580
    },
    {
      "epoch": 0.8740153069797995,
      "grad_norm": 0.13824176788330078,
      "learning_rate": 0.00021905940594059406,
      "loss": 2.094637870788574,
      "step": 8582
    },
    {
      "epoch": 0.8742189926724072,
      "grad_norm": 0.11493363231420517,
      "learning_rate": 0.0002187057991513437,
      "loss": 2.1497933864593506,
      "step": 8584
    },
    {
      "epoch": 0.874422678365015,
      "grad_norm": 0.15639877319335938,
      "learning_rate": 0.00021835219236209335,
      "loss": 2.190887451171875,
      "step": 8586
    },
    {
      "epoch": 0.8746263640576227,
      "grad_norm": 0.11394330114126205,
      "learning_rate": 0.000217998585572843,
      "loss": 2.2105135917663574,
      "step": 8588
    },
    {
      "epoch": 0.8748300497502304,
      "grad_norm": 0.11914867162704468,
      "learning_rate": 0.00021764497878359266,
      "loss": 2.283567428588867,
      "step": 8590
    },
    {
      "epoch": 0.8750337354428381,
      "grad_norm": 0.12108974903821945,
      "learning_rate": 0.0002172913719943423,
      "loss": 2.146623373031616,
      "step": 8592
    },
    {
      "epoch": 0.8752374211354459,
      "grad_norm": 0.12431849539279938,
      "learning_rate": 0.00021693776520509194,
      "loss": 2.1602001190185547,
      "step": 8594
    },
    {
      "epoch": 0.8754411068280537,
      "grad_norm": 0.11861249804496765,
      "learning_rate": 0.0002165841584158416,
      "loss": 2.2617430686950684,
      "step": 8596
    },
    {
      "epoch": 0.8756447925206614,
      "grad_norm": 0.10858278721570969,
      "learning_rate": 0.00021623055162659126,
      "loss": 2.170060634613037,
      "step": 8598
    },
    {
      "epoch": 0.875848478213269,
      "grad_norm": 0.1141049787402153,
      "learning_rate": 0.00021587694483734087,
      "loss": 2.1838669776916504,
      "step": 8600
    },
    {
      "epoch": 0.8760521639058768,
      "grad_norm": 0.14124970138072968,
      "learning_rate": 0.0002155233380480905,
      "loss": 2.224459171295166,
      "step": 8602
    },
    {
      "epoch": 0.8762558495984846,
      "grad_norm": 0.10870836675167084,
      "learning_rate": 0.00021516973125884018,
      "loss": 2.1437809467315674,
      "step": 8604
    },
    {
      "epoch": 0.8764595352910923,
      "grad_norm": 0.12143480777740479,
      "learning_rate": 0.00021481612446958982,
      "loss": 2.2585930824279785,
      "step": 8606
    },
    {
      "epoch": 0.8766632209837001,
      "grad_norm": 0.11944614350795746,
      "learning_rate": 0.00021446251768033947,
      "loss": 2.2984962463378906,
      "step": 8608
    },
    {
      "epoch": 0.8768669066763078,
      "grad_norm": 0.10895052552223206,
      "learning_rate": 0.0002141089108910891,
      "loss": 2.288649082183838,
      "step": 8610
    },
    {
      "epoch": 0.8770705923689155,
      "grad_norm": 0.1341923624277115,
      "learning_rate": 0.00021375530410183878,
      "loss": 2.170300245285034,
      "step": 8612
    },
    {
      "epoch": 0.8772742780615232,
      "grad_norm": 0.1229577511548996,
      "learning_rate": 0.0002134016973125884,
      "loss": 2.172239303588867,
      "step": 8614
    },
    {
      "epoch": 0.877477963754131,
      "grad_norm": 0.12646044790744781,
      "learning_rate": 0.00021304809052333804,
      "loss": 2.154872179031372,
      "step": 8616
    },
    {
      "epoch": 0.8776816494467388,
      "grad_norm": 0.11387313902378082,
      "learning_rate": 0.0002126944837340877,
      "loss": 2.188283920288086,
      "step": 8618
    },
    {
      "epoch": 0.8778853351393465,
      "grad_norm": 0.10949497669935226,
      "learning_rate": 0.00021234087694483735,
      "loss": 2.144660234451294,
      "step": 8620
    },
    {
      "epoch": 0.8780890208319542,
      "grad_norm": 0.11790304630994797,
      "learning_rate": 0.000211987270155587,
      "loss": 2.159522771835327,
      "step": 8622
    },
    {
      "epoch": 0.8782927065245619,
      "grad_norm": 0.12817120552062988,
      "learning_rate": 0.00021163366336633663,
      "loss": 2.1442184448242188,
      "step": 8624
    },
    {
      "epoch": 0.8784963922171697,
      "grad_norm": 0.11384691298007965,
      "learning_rate": 0.0002112800565770863,
      "loss": 2.129572868347168,
      "step": 8626
    },
    {
      "epoch": 0.8787000779097774,
      "grad_norm": 0.13049568235874176,
      "learning_rate": 0.00021092644978783592,
      "loss": 2.2553982734680176,
      "step": 8628
    },
    {
      "epoch": 0.8789037636023852,
      "grad_norm": 0.20881371200084686,
      "learning_rate": 0.00021057284299858559,
      "loss": 2.2394680976867676,
      "step": 8630
    },
    {
      "epoch": 0.8791074492949928,
      "grad_norm": 0.13170358538627625,
      "learning_rate": 0.00021021923620933523,
      "loss": 2.1932296752929688,
      "step": 8632
    },
    {
      "epoch": 0.8793111349876006,
      "grad_norm": 0.103530153632164,
      "learning_rate": 0.00020986562942008487,
      "loss": 2.0488381385803223,
      "step": 8634
    },
    {
      "epoch": 0.8795148206802084,
      "grad_norm": 0.1286340355873108,
      "learning_rate": 0.0002095120226308345,
      "loss": 2.1444809436798096,
      "step": 8636
    },
    {
      "epoch": 0.8797185063728161,
      "grad_norm": 0.16179896891117096,
      "learning_rate": 0.00020915841584158415,
      "loss": 2.040479898452759,
      "step": 8638
    },
    {
      "epoch": 0.8799221920654239,
      "grad_norm": 0.12930890917778015,
      "learning_rate": 0.00020880480905233382,
      "loss": 2.234626054763794,
      "step": 8640
    },
    {
      "epoch": 0.8801258777580315,
      "grad_norm": 0.11444754898548126,
      "learning_rate": 0.00020845120226308344,
      "loss": 2.23046875,
      "step": 8642
    },
    {
      "epoch": 0.8803295634506393,
      "grad_norm": 0.12268662452697754,
      "learning_rate": 0.0002080975954738331,
      "loss": 2.305014133453369,
      "step": 8644
    },
    {
      "epoch": 0.880533249143247,
      "grad_norm": 0.11240265518426895,
      "learning_rate": 0.00020774398868458275,
      "loss": 2.140080690383911,
      "step": 8646
    },
    {
      "epoch": 0.8807369348358548,
      "grad_norm": 0.11980671435594559,
      "learning_rate": 0.00020739038189533242,
      "loss": 2.1856260299682617,
      "step": 8648
    },
    {
      "epoch": 0.8809406205284626,
      "grad_norm": 0.12316741049289703,
      "learning_rate": 0.00020703677510608203,
      "loss": 2.248438835144043,
      "step": 8650
    },
    {
      "epoch": 0.8811443062210703,
      "grad_norm": 0.11319687962532043,
      "learning_rate": 0.00020668316831683168,
      "loss": 2.1249537467956543,
      "step": 8652
    },
    {
      "epoch": 0.881347991913678,
      "grad_norm": 0.11331547051668167,
      "learning_rate": 0.00020632956152758135,
      "loss": 2.186190366744995,
      "step": 8654
    },
    {
      "epoch": 0.8815516776062857,
      "grad_norm": 0.12187769263982773,
      "learning_rate": 0.00020597595473833096,
      "loss": 2.222234010696411,
      "step": 8656
    },
    {
      "epoch": 0.8817553632988935,
      "grad_norm": 0.11001472920179367,
      "learning_rate": 0.00020562234794908063,
      "loss": 2.163268566131592,
      "step": 8658
    },
    {
      "epoch": 0.8819590489915012,
      "grad_norm": 0.12073531001806259,
      "learning_rate": 0.00020526874115983027,
      "loss": 2.2579097747802734,
      "step": 8660
    },
    {
      "epoch": 0.882162734684109,
      "grad_norm": 0.11925148963928223,
      "learning_rate": 0.00020491513437057994,
      "loss": 2.13577938079834,
      "step": 8662
    },
    {
      "epoch": 0.8823664203767166,
      "grad_norm": 0.15424753725528717,
      "learning_rate": 0.00020456152758132956,
      "loss": 2.209768772125244,
      "step": 8664
    },
    {
      "epoch": 0.8825701060693244,
      "grad_norm": 0.11383488774299622,
      "learning_rate": 0.0002042079207920792,
      "loss": 2.2221741676330566,
      "step": 8666
    },
    {
      "epoch": 0.8827737917619322,
      "grad_norm": 0.11979332566261292,
      "learning_rate": 0.00020385431400282887,
      "loss": 2.312906265258789,
      "step": 8668
    },
    {
      "epoch": 0.8829774774545399,
      "grad_norm": 0.11553912609815598,
      "learning_rate": 0.00020350070721357848,
      "loss": 2.276459217071533,
      "step": 8670
    },
    {
      "epoch": 0.8831811631471477,
      "grad_norm": 0.10982818156480789,
      "learning_rate": 0.00020314710042432815,
      "loss": 2.1315226554870605,
      "step": 8672
    },
    {
      "epoch": 0.8833848488397553,
      "grad_norm": 0.11733166128396988,
      "learning_rate": 0.0002027934936350778,
      "loss": 2.062188148498535,
      "step": 8674
    },
    {
      "epoch": 0.8835885345323631,
      "grad_norm": 0.14271990954875946,
      "learning_rate": 0.00020243988684582746,
      "loss": 2.214179277420044,
      "step": 8676
    },
    {
      "epoch": 0.8837922202249708,
      "grad_norm": 0.13803835213184357,
      "learning_rate": 0.00020208628005657708,
      "loss": 2.292682647705078,
      "step": 8678
    },
    {
      "epoch": 0.8839959059175786,
      "grad_norm": 0.11644811183214188,
      "learning_rate": 0.00020173267326732675,
      "loss": 2.2054896354675293,
      "step": 8680
    },
    {
      "epoch": 0.8841995916101864,
      "grad_norm": 0.11151887476444244,
      "learning_rate": 0.0002013790664780764,
      "loss": 1.9073848724365234,
      "step": 8682
    },
    {
      "epoch": 0.8844032773027941,
      "grad_norm": 0.10937956720590591,
      "learning_rate": 0.000201025459688826,
      "loss": 2.213294744491577,
      "step": 8684
    },
    {
      "epoch": 0.8846069629954018,
      "grad_norm": 0.12395720928907394,
      "learning_rate": 0.00020067185289957568,
      "loss": 2.108579158782959,
      "step": 8686
    },
    {
      "epoch": 0.8848106486880095,
      "grad_norm": 0.12344598025083542,
      "learning_rate": 0.00020031824611032532,
      "loss": 2.085294246673584,
      "step": 8688
    },
    {
      "epoch": 0.8850143343806173,
      "grad_norm": 0.12511208653450012,
      "learning_rate": 0.000199964639321075,
      "loss": 2.1639671325683594,
      "step": 8690
    },
    {
      "epoch": 0.885218020073225,
      "grad_norm": 0.11997579783201218,
      "learning_rate": 0.0001996110325318246,
      "loss": 2.126923084259033,
      "step": 8692
    },
    {
      "epoch": 0.8854217057658328,
      "grad_norm": 0.11609110981225967,
      "learning_rate": 0.00019925742574257427,
      "loss": 2.1809425354003906,
      "step": 8694
    },
    {
      "epoch": 0.8856253914584404,
      "grad_norm": 0.1109338253736496,
      "learning_rate": 0.00019890381895332391,
      "loss": 2.1304118633270264,
      "step": 8696
    },
    {
      "epoch": 0.8858290771510482,
      "grad_norm": 0.12461572885513306,
      "learning_rate": 0.00019855021216407356,
      "loss": 2.218538284301758,
      "step": 8698
    },
    {
      "epoch": 0.886032762843656,
      "grad_norm": 0.11005209386348724,
      "learning_rate": 0.0001981966053748232,
      "loss": 2.145294189453125,
      "step": 8700
    },
    {
      "epoch": 0.8862364485362637,
      "grad_norm": 0.11893009394407272,
      "learning_rate": 0.00019784299858557284,
      "loss": 2.084887742996216,
      "step": 8702
    },
    {
      "epoch": 0.8864401342288715,
      "grad_norm": 0.1343289315700531,
      "learning_rate": 0.0001974893917963225,
      "loss": 2.1550183296203613,
      "step": 8704
    },
    {
      "epoch": 0.8866438199214791,
      "grad_norm": 0.1153336837887764,
      "learning_rate": 0.00019713578500707213,
      "loss": 2.2115297317504883,
      "step": 8706
    },
    {
      "epoch": 0.8868475056140869,
      "grad_norm": 0.11430255323648453,
      "learning_rate": 0.0001967821782178218,
      "loss": 2.2781476974487305,
      "step": 8708
    },
    {
      "epoch": 0.8870511913066946,
      "grad_norm": 0.1211911216378212,
      "learning_rate": 0.00019642857142857144,
      "loss": 2.157557487487793,
      "step": 8710
    },
    {
      "epoch": 0.8872548769993024,
      "grad_norm": 0.11213487386703491,
      "learning_rate": 0.00019607496463932108,
      "loss": 2.3043456077575684,
      "step": 8712
    },
    {
      "epoch": 0.8874585626919101,
      "grad_norm": 0.1559244990348816,
      "learning_rate": 0.00019572135785007072,
      "loss": 2.2327914237976074,
      "step": 8714
    },
    {
      "epoch": 0.8876622483845179,
      "grad_norm": 0.1277015209197998,
      "learning_rate": 0.00019536775106082036,
      "loss": 2.0877797603607178,
      "step": 8716
    },
    {
      "epoch": 0.8878659340771256,
      "grad_norm": 0.15825821459293365,
      "learning_rate": 0.00019501414427157003,
      "loss": 2.0249428749084473,
      "step": 8718
    },
    {
      "epoch": 0.8880696197697333,
      "grad_norm": 0.12276486307382584,
      "learning_rate": 0.00019466053748231965,
      "loss": 2.1304941177368164,
      "step": 8720
    },
    {
      "epoch": 0.8882733054623411,
      "grad_norm": 0.09890832006931305,
      "learning_rate": 0.00019430693069306932,
      "loss": 2.1875271797180176,
      "step": 8722
    },
    {
      "epoch": 0.8884769911549488,
      "grad_norm": 0.12316036224365234,
      "learning_rate": 0.00019395332390381896,
      "loss": 2.267254590988159,
      "step": 8724
    },
    {
      "epoch": 0.8886806768475566,
      "grad_norm": 0.11223430931568146,
      "learning_rate": 0.0001935997171145686,
      "loss": 2.2298150062561035,
      "step": 8726
    },
    {
      "epoch": 0.8888843625401642,
      "grad_norm": 0.12014144659042358,
      "learning_rate": 0.00019324611032531824,
      "loss": 2.2084856033325195,
      "step": 8728
    },
    {
      "epoch": 0.889088048232772,
      "grad_norm": 0.13105423748493195,
      "learning_rate": 0.0001928925035360679,
      "loss": 2.2206828594207764,
      "step": 8730
    },
    {
      "epoch": 0.8892917339253797,
      "grad_norm": 0.15459951758384705,
      "learning_rate": 0.00019253889674681756,
      "loss": 2.1972062587738037,
      "step": 8732
    },
    {
      "epoch": 0.8894954196179875,
      "grad_norm": 0.11419671028852463,
      "learning_rate": 0.00019218528995756717,
      "loss": 2.179816246032715,
      "step": 8734
    },
    {
      "epoch": 0.8896991053105953,
      "grad_norm": 0.12023697048425674,
      "learning_rate": 0.00019183168316831684,
      "loss": 2.242523193359375,
      "step": 8736
    },
    {
      "epoch": 0.8899027910032029,
      "grad_norm": 0.1227419301867485,
      "learning_rate": 0.00019147807637906648,
      "loss": 2.2264890670776367,
      "step": 8738
    },
    {
      "epoch": 0.8901064766958107,
      "grad_norm": 0.11458329111337662,
      "learning_rate": 0.00019112446958981612,
      "loss": 2.0837860107421875,
      "step": 8740
    },
    {
      "epoch": 0.8903101623884184,
      "grad_norm": 0.11515060067176819,
      "learning_rate": 0.00019077086280056577,
      "loss": 2.262943744659424,
      "step": 8742
    },
    {
      "epoch": 0.8905138480810262,
      "grad_norm": 0.13230116665363312,
      "learning_rate": 0.00019041725601131544,
      "loss": 2.2053327560424805,
      "step": 8744
    },
    {
      "epoch": 0.8907175337736339,
      "grad_norm": 0.10584226995706558,
      "learning_rate": 0.00019006364922206508,
      "loss": 2.2615833282470703,
      "step": 8746
    },
    {
      "epoch": 0.8909212194662416,
      "grad_norm": 0.13956141471862793,
      "learning_rate": 0.00018971004243281472,
      "loss": 2.088707685470581,
      "step": 8748
    },
    {
      "epoch": 0.8911249051588493,
      "grad_norm": 0.11561837792396545,
      "learning_rate": 0.00018935643564356436,
      "loss": 2.2327218055725098,
      "step": 8750
    },
    {
      "epoch": 0.8913285908514571,
      "grad_norm": 0.1369146853685379,
      "learning_rate": 0.000189002828854314,
      "loss": 2.3133413791656494,
      "step": 8752
    },
    {
      "epoch": 0.8915322765440649,
      "grad_norm": 0.12861958146095276,
      "learning_rate": 0.00018864922206506365,
      "loss": 2.146803617477417,
      "step": 8754
    },
    {
      "epoch": 0.8917359622366726,
      "grad_norm": 0.12289351969957352,
      "learning_rate": 0.0001882956152758133,
      "loss": 2.356349229812622,
      "step": 8756
    },
    {
      "epoch": 0.8919396479292804,
      "grad_norm": 0.13229168951511383,
      "learning_rate": 0.00018794200848656296,
      "loss": 2.1774399280548096,
      "step": 8758
    },
    {
      "epoch": 0.892143333621888,
      "grad_norm": 0.13941340148448944,
      "learning_rate": 0.0001875884016973126,
      "loss": 2.1579394340515137,
      "step": 8760
    },
    {
      "epoch": 0.8923470193144958,
      "grad_norm": 0.15535907447338104,
      "learning_rate": 0.00018723479490806224,
      "loss": 2.1134822368621826,
      "step": 8762
    },
    {
      "epoch": 0.8925507050071035,
      "grad_norm": 0.11805552244186401,
      "learning_rate": 0.00018688118811881188,
      "loss": 2.116799831390381,
      "step": 8764
    },
    {
      "epoch": 0.8927543906997113,
      "grad_norm": 0.11726606637239456,
      "learning_rate": 0.00018652758132956153,
      "loss": 2.187803030014038,
      "step": 8766
    },
    {
      "epoch": 0.8929580763923191,
      "grad_norm": 0.13380779325962067,
      "learning_rate": 0.00018617397454031117,
      "loss": 2.375462293624878,
      "step": 8768
    },
    {
      "epoch": 0.8931617620849267,
      "grad_norm": 0.13009732961654663,
      "learning_rate": 0.0001858203677510608,
      "loss": 2.252134323120117,
      "step": 8770
    },
    {
      "epoch": 0.8933654477775345,
      "grad_norm": 0.10745349526405334,
      "learning_rate": 0.00018546676096181048,
      "loss": 2.0997958183288574,
      "step": 8772
    },
    {
      "epoch": 0.8935691334701422,
      "grad_norm": 0.1031663566827774,
      "learning_rate": 0.00018511315417256012,
      "loss": 2.067868947982788,
      "step": 8774
    },
    {
      "epoch": 0.89377281916275,
      "grad_norm": 0.13304099440574646,
      "learning_rate": 0.00018475954738330976,
      "loss": 2.143285036087036,
      "step": 8776
    },
    {
      "epoch": 0.8939765048553577,
      "grad_norm": 0.11828266084194183,
      "learning_rate": 0.0001844059405940594,
      "loss": 2.2378745079040527,
      "step": 8778
    },
    {
      "epoch": 0.8941801905479654,
      "grad_norm": 0.10015836358070374,
      "learning_rate": 0.00018405233380480908,
      "loss": 2.125204086303711,
      "step": 8780
    },
    {
      "epoch": 0.8943838762405731,
      "grad_norm": 0.10107137262821198,
      "learning_rate": 0.0001836987270155587,
      "loss": 2.057300567626953,
      "step": 8782
    },
    {
      "epoch": 0.8945875619331809,
      "grad_norm": 0.11897963285446167,
      "learning_rate": 0.00018334512022630833,
      "loss": 2.2754015922546387,
      "step": 8784
    },
    {
      "epoch": 0.8947912476257887,
      "grad_norm": 0.11783910542726517,
      "learning_rate": 0.000182991513437058,
      "loss": 2.171006202697754,
      "step": 8786
    },
    {
      "epoch": 0.8949949333183964,
      "grad_norm": 0.11347558349370956,
      "learning_rate": 0.00018263790664780765,
      "loss": 2.0790390968322754,
      "step": 8788
    },
    {
      "epoch": 0.8951986190110042,
      "grad_norm": 0.13937658071517944,
      "learning_rate": 0.0001822842998585573,
      "loss": 2.2311654090881348,
      "step": 8790
    },
    {
      "epoch": 0.8954023047036118,
      "grad_norm": 0.12405553460121155,
      "learning_rate": 0.00018193069306930693,
      "loss": 2.087327480316162,
      "step": 8792
    },
    {
      "epoch": 0.8956059903962196,
      "grad_norm": 0.10281388461589813,
      "learning_rate": 0.0001815770862800566,
      "loss": 2.135857105255127,
      "step": 8794
    },
    {
      "epoch": 0.8958096760888273,
      "grad_norm": 0.13713425397872925,
      "learning_rate": 0.00018122347949080621,
      "loss": 2.212280511856079,
      "step": 8796
    },
    {
      "epoch": 0.8960133617814351,
      "grad_norm": 0.13629990816116333,
      "learning_rate": 0.00018086987270155586,
      "loss": 2.3162546157836914,
      "step": 8798
    },
    {
      "epoch": 0.8962170474740428,
      "grad_norm": 0.1022983193397522,
      "learning_rate": 0.00018051626591230553,
      "loss": 2.0116970539093018,
      "step": 8800
    },
    {
      "epoch": 0.8964207331666505,
      "grad_norm": 0.10931524634361267,
      "learning_rate": 0.00018016265912305517,
      "loss": 2.1432080268859863,
      "step": 8802
    },
    {
      "epoch": 0.8966244188592583,
      "grad_norm": 0.15693961083889008,
      "learning_rate": 0.0001798090523338048,
      "loss": 2.09150767326355,
      "step": 8804
    },
    {
      "epoch": 0.896828104551866,
      "grad_norm": 0.10967199504375458,
      "learning_rate": 0.00017945544554455445,
      "loss": 2.127716064453125,
      "step": 8806
    },
    {
      "epoch": 0.8970317902444738,
      "grad_norm": 0.1257874071598053,
      "learning_rate": 0.00017910183875530412,
      "loss": 2.2761945724487305,
      "step": 8808
    },
    {
      "epoch": 0.8972354759370815,
      "grad_norm": 0.13055185973644257,
      "learning_rate": 0.00017874823196605374,
      "loss": 2.226621627807617,
      "step": 8810
    },
    {
      "epoch": 0.8974391616296892,
      "grad_norm": 0.1228129118680954,
      "learning_rate": 0.0001783946251768034,
      "loss": 2.313627243041992,
      "step": 8812
    },
    {
      "epoch": 0.8976428473222969,
      "grad_norm": 0.10689161717891693,
      "learning_rate": 0.00017804101838755305,
      "loss": 2.2030715942382812,
      "step": 8814
    },
    {
      "epoch": 0.8978465330149047,
      "grad_norm": 0.16135095059871674,
      "learning_rate": 0.0001776874115983027,
      "loss": 2.2419281005859375,
      "step": 8816
    },
    {
      "epoch": 0.8980502187075124,
      "grad_norm": 0.1346459686756134,
      "learning_rate": 0.00017733380480905233,
      "loss": 2.207225799560547,
      "step": 8818
    },
    {
      "epoch": 0.8982539044001202,
      "grad_norm": 0.14773862063884735,
      "learning_rate": 0.00017698019801980197,
      "loss": 2.302567481994629,
      "step": 8820
    },
    {
      "epoch": 0.898457590092728,
      "grad_norm": 0.11851039528846741,
      "learning_rate": 0.00017662659123055164,
      "loss": 2.2524752616882324,
      "step": 8822
    },
    {
      "epoch": 0.8986612757853356,
      "grad_norm": 0.13583876192569733,
      "learning_rate": 0.00017627298444130126,
      "loss": 2.1906228065490723,
      "step": 8824
    },
    {
      "epoch": 0.8988649614779434,
      "grad_norm": 0.15085342526435852,
      "learning_rate": 0.00017591937765205093,
      "loss": 2.131749153137207,
      "step": 8826
    },
    {
      "epoch": 0.8990686471705511,
      "grad_norm": 0.12935005128383636,
      "learning_rate": 0.00017556577086280057,
      "loss": 2.2213480472564697,
      "step": 8828
    },
    {
      "epoch": 0.8992723328631589,
      "grad_norm": 0.1244739294052124,
      "learning_rate": 0.00017521216407355024,
      "loss": 2.192734718322754,
      "step": 8830
    },
    {
      "epoch": 0.8994760185557666,
      "grad_norm": 0.11433504521846771,
      "learning_rate": 0.00017485855728429986,
      "loss": 2.0381648540496826,
      "step": 8832
    },
    {
      "epoch": 0.8996797042483743,
      "grad_norm": 0.11442320048809052,
      "learning_rate": 0.0001745049504950495,
      "loss": 2.109994888305664,
      "step": 8834
    },
    {
      "epoch": 0.899883389940982,
      "grad_norm": 0.15000766515731812,
      "learning_rate": 0.00017415134370579917,
      "loss": 2.091477394104004,
      "step": 8836
    },
    {
      "epoch": 0.9000870756335898,
      "grad_norm": 0.1240738183259964,
      "learning_rate": 0.00017379773691654878,
      "loss": 2.2859935760498047,
      "step": 8838
    },
    {
      "epoch": 0.9002907613261976,
      "grad_norm": 0.11392781883478165,
      "learning_rate": 0.00017344413012729845,
      "loss": 2.2345194816589355,
      "step": 8840
    },
    {
      "epoch": 0.9004944470188053,
      "grad_norm": 0.11828096956014633,
      "learning_rate": 0.0001730905233380481,
      "loss": 2.2284059524536133,
      "step": 8842
    },
    {
      "epoch": 0.900698132711413,
      "grad_norm": 0.1320345401763916,
      "learning_rate": 0.00017273691654879776,
      "loss": 2.269773006439209,
      "step": 8844
    },
    {
      "epoch": 0.9009018184040207,
      "grad_norm": 0.12056339532136917,
      "learning_rate": 0.00017238330975954738,
      "loss": 2.2143449783325195,
      "step": 8846
    },
    {
      "epoch": 0.9011055040966285,
      "grad_norm": 0.11936300992965698,
      "learning_rate": 0.00017202970297029702,
      "loss": 2.165102481842041,
      "step": 8848
    },
    {
      "epoch": 0.9013091897892362,
      "grad_norm": 0.12048445641994476,
      "learning_rate": 0.0001716760961810467,
      "loss": 2.1098053455352783,
      "step": 8850
    },
    {
      "epoch": 0.901512875481844,
      "grad_norm": 0.12414637207984924,
      "learning_rate": 0.0001713224893917963,
      "loss": 2.3082380294799805,
      "step": 8852
    },
    {
      "epoch": 0.9017165611744518,
      "grad_norm": 0.11457009613513947,
      "learning_rate": 0.00017096888260254597,
      "loss": 2.329371213912964,
      "step": 8854
    },
    {
      "epoch": 0.9019202468670594,
      "grad_norm": 0.1334666907787323,
      "learning_rate": 0.00017061527581329562,
      "loss": 2.1836836338043213,
      "step": 8856
    },
    {
      "epoch": 0.9021239325596672,
      "grad_norm": 0.11559416353702545,
      "learning_rate": 0.00017026166902404529,
      "loss": 2.2307286262512207,
      "step": 8858
    },
    {
      "epoch": 0.9023276182522749,
      "grad_norm": 0.15669883787631989,
      "learning_rate": 0.0001699080622347949,
      "loss": 2.2046420574188232,
      "step": 8860
    },
    {
      "epoch": 0.9025313039448827,
      "grad_norm": 0.11745938658714294,
      "learning_rate": 0.00016955445544554457,
      "loss": 2.293147563934326,
      "step": 8862
    },
    {
      "epoch": 0.9027349896374904,
      "grad_norm": 0.13183525204658508,
      "learning_rate": 0.0001692008486562942,
      "loss": 2.1128296852111816,
      "step": 8864
    },
    {
      "epoch": 0.9029386753300981,
      "grad_norm": 0.10533929616212845,
      "learning_rate": 0.00016884724186704383,
      "loss": 2.0971641540527344,
      "step": 8866
    },
    {
      "epoch": 0.9031423610227058,
      "grad_norm": 0.11306287348270416,
      "learning_rate": 0.0001684936350777935,
      "loss": 2.197619676589966,
      "step": 8868
    },
    {
      "epoch": 0.9033460467153136,
      "grad_norm": 0.11756132543087006,
      "learning_rate": 0.00016814002828854314,
      "loss": 2.22320556640625,
      "step": 8870
    },
    {
      "epoch": 0.9035497324079214,
      "grad_norm": 0.1221703514456749,
      "learning_rate": 0.0001677864214992928,
      "loss": 2.1226420402526855,
      "step": 8872
    },
    {
      "epoch": 0.9037534181005291,
      "grad_norm": 0.13557785749435425,
      "learning_rate": 0.00016743281471004242,
      "loss": 2.0702154636383057,
      "step": 8874
    },
    {
      "epoch": 0.9039571037931368,
      "grad_norm": 0.09528390318155289,
      "learning_rate": 0.0001670792079207921,
      "loss": 2.169198989868164,
      "step": 8876
    },
    {
      "epoch": 0.9041607894857445,
      "grad_norm": 0.12492747604846954,
      "learning_rate": 0.00016672560113154173,
      "loss": 2.4683499336242676,
      "step": 8878
    },
    {
      "epoch": 0.9043644751783523,
      "grad_norm": 0.10677724331617355,
      "learning_rate": 0.00016637199434229138,
      "loss": 2.207334518432617,
      "step": 8880
    },
    {
      "epoch": 0.90456816087096,
      "grad_norm": 0.10995791852474213,
      "learning_rate": 0.00016601838755304102,
      "loss": 2.328812599182129,
      "step": 8882
    },
    {
      "epoch": 0.9047718465635678,
      "grad_norm": 0.13895897567272186,
      "learning_rate": 0.00016566478076379066,
      "loss": 2.2678914070129395,
      "step": 8884
    },
    {
      "epoch": 0.9049755322561754,
      "grad_norm": 0.11563248932361603,
      "learning_rate": 0.00016531117397454033,
      "loss": 2.14737606048584,
      "step": 8886
    },
    {
      "epoch": 0.9051792179487832,
      "grad_norm": 0.12580618262290955,
      "learning_rate": 0.00016495756718528995,
      "loss": 2.2339253425598145,
      "step": 8888
    },
    {
      "epoch": 0.905382903641391,
      "grad_norm": 0.10404669493436813,
      "learning_rate": 0.00016460396039603961,
      "loss": 2.114990711212158,
      "step": 8890
    },
    {
      "epoch": 0.9055865893339987,
      "grad_norm": 0.12889915704727173,
      "learning_rate": 0.00016425035360678926,
      "loss": 2.199639320373535,
      "step": 8892
    },
    {
      "epoch": 0.9057902750266065,
      "grad_norm": 0.13924485445022583,
      "learning_rate": 0.0001638967468175389,
      "loss": 2.2144699096679688,
      "step": 8894
    },
    {
      "epoch": 0.9059939607192142,
      "grad_norm": 0.13329054415225983,
      "learning_rate": 0.00016354314002828854,
      "loss": 2.257235050201416,
      "step": 8896
    },
    {
      "epoch": 0.9061976464118219,
      "grad_norm": 0.12625521421432495,
      "learning_rate": 0.00016318953323903818,
      "loss": 2.2226150035858154,
      "step": 8898
    },
    {
      "epoch": 0.9064013321044296,
      "grad_norm": 0.14877082407474518,
      "learning_rate": 0.00016283592644978785,
      "loss": 1.9926061630249023,
      "step": 8900
    },
    {
      "epoch": 0.9066050177970374,
      "grad_norm": 0.12353583425283432,
      "learning_rate": 0.00016248231966053747,
      "loss": 2.2316904067993164,
      "step": 8902
    },
    {
      "epoch": 0.9068087034896452,
      "grad_norm": 0.10882313549518585,
      "learning_rate": 0.00016212871287128714,
      "loss": 2.194981575012207,
      "step": 8904
    },
    {
      "epoch": 0.9070123891822529,
      "grad_norm": 0.11598377674818039,
      "learning_rate": 0.00016177510608203678,
      "loss": 2.0864343643188477,
      "step": 8906
    },
    {
      "epoch": 0.9072160748748606,
      "grad_norm": 0.10239724814891815,
      "learning_rate": 0.00016142149929278642,
      "loss": 2.1714963912963867,
      "step": 8908
    },
    {
      "epoch": 0.9074197605674683,
      "grad_norm": 0.1065215989947319,
      "learning_rate": 0.00016106789250353606,
      "loss": 2.0639092922210693,
      "step": 8910
    },
    {
      "epoch": 0.9076234462600761,
      "grad_norm": 0.1523260921239853,
      "learning_rate": 0.00016071428571428573,
      "loss": 2.226921796798706,
      "step": 8912
    },
    {
      "epoch": 0.9078271319526838,
      "grad_norm": 0.13321976363658905,
      "learning_rate": 0.00016036067892503538,
      "loss": 2.256279706954956,
      "step": 8914
    },
    {
      "epoch": 0.9080308176452916,
      "grad_norm": 0.10672559589147568,
      "learning_rate": 0.000160007072135785,
      "loss": 2.214966297149658,
      "step": 8916
    },
    {
      "epoch": 0.9082345033378992,
      "grad_norm": 0.11705255508422852,
      "learning_rate": 0.00015965346534653466,
      "loss": 2.238558530807495,
      "step": 8918
    },
    {
      "epoch": 0.908438189030507,
      "grad_norm": 0.11571221053600311,
      "learning_rate": 0.0001592998585572843,
      "loss": 2.2459261417388916,
      "step": 8920
    },
    {
      "epoch": 0.9086418747231148,
      "grad_norm": 0.10245706140995026,
      "learning_rate": 0.00015894625176803394,
      "loss": 2.1738944053649902,
      "step": 8922
    },
    {
      "epoch": 0.9088455604157225,
      "grad_norm": 0.12699541449546814,
      "learning_rate": 0.00015859264497878359,
      "loss": 2.196711540222168,
      "step": 8924
    },
    {
      "epoch": 0.9090492461083303,
      "grad_norm": 0.11207503825426102,
      "learning_rate": 0.00015823903818953326,
      "loss": 2.1669697761535645,
      "step": 8926
    },
    {
      "epoch": 0.909252931800938,
      "grad_norm": 0.1204100027680397,
      "learning_rate": 0.0001578854314002829,
      "loss": 2.264072895050049,
      "step": 8928
    },
    {
      "epoch": 0.9094566174935457,
      "grad_norm": 0.2036898285150528,
      "learning_rate": 0.00015753182461103254,
      "loss": 2.301436424255371,
      "step": 8930
    },
    {
      "epoch": 0.9096603031861534,
      "grad_norm": 0.141203373670578,
      "learning_rate": 0.00015717821782178218,
      "loss": 2.2779464721679688,
      "step": 8932
    },
    {
      "epoch": 0.9098639888787612,
      "grad_norm": 0.13236236572265625,
      "learning_rate": 0.00015682461103253182,
      "loss": 1.9855408668518066,
      "step": 8934
    },
    {
      "epoch": 0.910067674571369,
      "grad_norm": 0.13756974041461945,
      "learning_rate": 0.00015647100424328147,
      "loss": 2.1687393188476562,
      "step": 8936
    },
    {
      "epoch": 0.9102713602639767,
      "grad_norm": 0.11964678019285202,
      "learning_rate": 0.0001561173974540311,
      "loss": 2.2004313468933105,
      "step": 8938
    },
    {
      "epoch": 0.9104750459565844,
      "grad_norm": 0.13169556856155396,
      "learning_rate": 0.00015576379066478078,
      "loss": 2.222411632537842,
      "step": 8940
    },
    {
      "epoch": 0.9106787316491921,
      "grad_norm": 0.13725151121616364,
      "learning_rate": 0.00015541018387553042,
      "loss": 2.15348219871521,
      "step": 8942
    },
    {
      "epoch": 0.9108824173417999,
      "grad_norm": 0.1538146436214447,
      "learning_rate": 0.00015505657708628006,
      "loss": 2.249782085418701,
      "step": 8944
    },
    {
      "epoch": 0.9110861030344076,
      "grad_norm": 0.11282789707183838,
      "learning_rate": 0.0001547029702970297,
      "loss": 2.317354202270508,
      "step": 8946
    },
    {
      "epoch": 0.9112897887270154,
      "grad_norm": 0.11200334876775742,
      "learning_rate": 0.00015434936350777935,
      "loss": 2.0774712562561035,
      "step": 8948
    },
    {
      "epoch": 0.911493474419623,
      "grad_norm": 0.11471602320671082,
      "learning_rate": 0.000153995756718529,
      "loss": 2.1480939388275146,
      "step": 8950
    },
    {
      "epoch": 0.9116971601122308,
      "grad_norm": 0.10099772363901138,
      "learning_rate": 0.00015364214992927863,
      "loss": 2.0983986854553223,
      "step": 8952
    },
    {
      "epoch": 0.9119008458048385,
      "grad_norm": 0.12485602498054504,
      "learning_rate": 0.0001532885431400283,
      "loss": 2.3122477531433105,
      "step": 8954
    },
    {
      "epoch": 0.9121045314974463,
      "grad_norm": 0.10889948159456253,
      "learning_rate": 0.00015293493635077794,
      "loss": 2.118825912475586,
      "step": 8956
    },
    {
      "epoch": 0.9123082171900541,
      "grad_norm": 0.11204277724027634,
      "learning_rate": 0.00015258132956152759,
      "loss": 2.1865835189819336,
      "step": 8958
    },
    {
      "epoch": 0.9125119028826618,
      "grad_norm": 0.1006772443652153,
      "learning_rate": 0.00015222772277227723,
      "loss": 2.04062557220459,
      "step": 8960
    },
    {
      "epoch": 0.9127155885752695,
      "grad_norm": 0.11083020269870758,
      "learning_rate": 0.0001518741159830269,
      "loss": 2.1278209686279297,
      "step": 8962
    },
    {
      "epoch": 0.9129192742678772,
      "grad_norm": 0.12478788942098618,
      "learning_rate": 0.0001515205091937765,
      "loss": 2.124885082244873,
      "step": 8964
    },
    {
      "epoch": 0.913122959960485,
      "grad_norm": 0.10898878425359726,
      "learning_rate": 0.00015116690240452615,
      "loss": 2.050325393676758,
      "step": 8966
    },
    {
      "epoch": 0.9133266456530927,
      "grad_norm": 0.13455548882484436,
      "learning_rate": 0.00015081329561527582,
      "loss": 2.281848430633545,
      "step": 8968
    },
    {
      "epoch": 0.9135303313457005,
      "grad_norm": 0.16252915561199188,
      "learning_rate": 0.00015045968882602547,
      "loss": 2.1518914699554443,
      "step": 8970
    },
    {
      "epoch": 0.9137340170383081,
      "grad_norm": 0.11150234937667847,
      "learning_rate": 0.0001501060820367751,
      "loss": 2.2325332164764404,
      "step": 8972
    },
    {
      "epoch": 0.9139377027309159,
      "grad_norm": 0.1249770000576973,
      "learning_rate": 0.00014975247524752475,
      "loss": 2.1818580627441406,
      "step": 8974
    },
    {
      "epoch": 0.9141413884235237,
      "grad_norm": 0.11338584870100021,
      "learning_rate": 0.00014939886845827442,
      "loss": 2.235604763031006,
      "step": 8976
    },
    {
      "epoch": 0.9143450741161314,
      "grad_norm": 0.11494046449661255,
      "learning_rate": 0.00014904526166902403,
      "loss": 2.1389262676239014,
      "step": 8978
    },
    {
      "epoch": 0.9145487598087392,
      "grad_norm": 0.10616730898618698,
      "learning_rate": 0.0001486916548797737,
      "loss": 2.175537586212158,
      "step": 8980
    },
    {
      "epoch": 0.9147524455013468,
      "grad_norm": 0.11528082937002182,
      "learning_rate": 0.00014833804809052335,
      "loss": 2.3140759468078613,
      "step": 8982
    },
    {
      "epoch": 0.9149561311939546,
      "grad_norm": 0.10917354375123978,
      "learning_rate": 0.000147984441301273,
      "loss": 2.265113115310669,
      "step": 8984
    },
    {
      "epoch": 0.9151598168865623,
      "grad_norm": 0.1066412404179573,
      "learning_rate": 0.00014763083451202263,
      "loss": 2.040346384048462,
      "step": 8986
    },
    {
      "epoch": 0.9153635025791701,
      "grad_norm": 0.11359008401632309,
      "learning_rate": 0.00014727722772277227,
      "loss": 2.249979019165039,
      "step": 8988
    },
    {
      "epoch": 0.9155671882717779,
      "grad_norm": 0.11188136786222458,
      "learning_rate": 0.00014692362093352194,
      "loss": 2.0909860134124756,
      "step": 8990
    },
    {
      "epoch": 0.9157708739643855,
      "grad_norm": 0.132910817861557,
      "learning_rate": 0.00014657001414427156,
      "loss": 2.2820587158203125,
      "step": 8992
    },
    {
      "epoch": 0.9159745596569933,
      "grad_norm": 0.10948612540960312,
      "learning_rate": 0.00014621640735502123,
      "loss": 2.0870285034179688,
      "step": 8994
    },
    {
      "epoch": 0.916178245349601,
      "grad_norm": 0.10962677001953125,
      "learning_rate": 0.00014586280056577087,
      "loss": 2.2313077449798584,
      "step": 8996
    },
    {
      "epoch": 0.9163819310422088,
      "grad_norm": 0.12755367159843445,
      "learning_rate": 0.0001455091937765205,
      "loss": 2.2467494010925293,
      "step": 8998
    },
    {
      "epoch": 0.9165856167348165,
      "grad_norm": 0.14634546637535095,
      "learning_rate": 0.00014515558698727015,
      "loss": 2.128525733947754,
      "step": 9000
    },
    {
      "epoch": 0.9167893024274243,
      "grad_norm": 0.10384708642959595,
      "learning_rate": 0.0001448019801980198,
      "loss": 1.9265986680984497,
      "step": 9002
    },
    {
      "epoch": 0.9169929881200319,
      "grad_norm": 0.15008504688739777,
      "learning_rate": 0.00014444837340876946,
      "loss": 2.2665181159973145,
      "step": 9004
    },
    {
      "epoch": 0.9171966738126397,
      "grad_norm": 0.11339323222637177,
      "learning_rate": 0.00014409476661951908,
      "loss": 2.2663462162017822,
      "step": 9006
    },
    {
      "epoch": 0.9174003595052475,
      "grad_norm": 0.14448335766792297,
      "learning_rate": 0.00014374115983026875,
      "loss": 2.1150894165039062,
      "step": 9008
    },
    {
      "epoch": 0.9176040451978552,
      "grad_norm": 0.15524272620677948,
      "learning_rate": 0.0001433875530410184,
      "loss": 2.1795175075531006,
      "step": 9010
    },
    {
      "epoch": 0.917807730890463,
      "grad_norm": 0.14125113189220428,
      "learning_rate": 0.00014303394625176806,
      "loss": 2.1106529235839844,
      "step": 9012
    },
    {
      "epoch": 0.9180114165830706,
      "grad_norm": 0.11567751318216324,
      "learning_rate": 0.00014268033946251768,
      "loss": 2.2552590370178223,
      "step": 9014
    },
    {
      "epoch": 0.9182151022756784,
      "grad_norm": 0.10591063648462296,
      "learning_rate": 0.00014232673267326732,
      "loss": 2.2479562759399414,
      "step": 9016
    },
    {
      "epoch": 0.9184187879682861,
      "grad_norm": 0.1409359574317932,
      "learning_rate": 0.000141973125884017,
      "loss": 2.2173280715942383,
      "step": 9018
    },
    {
      "epoch": 0.9186224736608939,
      "grad_norm": 0.11282593756914139,
      "learning_rate": 0.0001416195190947666,
      "loss": 2.184047222137451,
      "step": 9020
    },
    {
      "epoch": 0.9188261593535016,
      "grad_norm": 0.11251885443925858,
      "learning_rate": 0.00014126591230551627,
      "loss": 2.29182767868042,
      "step": 9022
    },
    {
      "epoch": 0.9190298450461093,
      "grad_norm": 0.11577679961919785,
      "learning_rate": 0.0001409123055162659,
      "loss": 2.3339033126831055,
      "step": 9024
    },
    {
      "epoch": 0.919233530738717,
      "grad_norm": 0.11365494877099991,
      "learning_rate": 0.00014055869872701558,
      "loss": 2.1103248596191406,
      "step": 9026
    },
    {
      "epoch": 0.9194372164313248,
      "grad_norm": 0.11859577894210815,
      "learning_rate": 0.0001402050919377652,
      "loss": 2.100550651550293,
      "step": 9028
    },
    {
      "epoch": 0.9196409021239326,
      "grad_norm": 0.12535902857780457,
      "learning_rate": 0.00013985148514851487,
      "loss": 2.216801643371582,
      "step": 9030
    },
    {
      "epoch": 0.9198445878165403,
      "grad_norm": 0.18121246993541718,
      "learning_rate": 0.0001394978783592645,
      "loss": 2.1620142459869385,
      "step": 9032
    },
    {
      "epoch": 0.9200482735091481,
      "grad_norm": 0.12805581092834473,
      "learning_rate": 0.00013914427157001412,
      "loss": 2.1993041038513184,
      "step": 9034
    },
    {
      "epoch": 0.9202519592017557,
      "grad_norm": 0.14736780524253845,
      "learning_rate": 0.0001387906647807638,
      "loss": 2.1060566902160645,
      "step": 9036
    },
    {
      "epoch": 0.9204556448943635,
      "grad_norm": 0.12449242174625397,
      "learning_rate": 0.00013843705799151344,
      "loss": 2.113171100616455,
      "step": 9038
    },
    {
      "epoch": 0.9206593305869712,
      "grad_norm": 0.11784624308347702,
      "learning_rate": 0.0001380834512022631,
      "loss": 2.161017417907715,
      "step": 9040
    },
    {
      "epoch": 0.920863016279579,
      "grad_norm": 0.13825462758541107,
      "learning_rate": 0.00013772984441301272,
      "loss": 2.136171817779541,
      "step": 9042
    },
    {
      "epoch": 0.9210667019721868,
      "grad_norm": 0.12577888369560242,
      "learning_rate": 0.0001373762376237624,
      "loss": 2.218874931335449,
      "step": 9044
    },
    {
      "epoch": 0.9212703876647944,
      "grad_norm": 0.1216711699962616,
      "learning_rate": 0.00013702263083451203,
      "loss": 2.228893280029297,
      "step": 9046
    },
    {
      "epoch": 0.9214740733574022,
      "grad_norm": 0.1272527575492859,
      "learning_rate": 0.00013666902404526165,
      "loss": 2.2088754177093506,
      "step": 9048
    },
    {
      "epoch": 0.9216777590500099,
      "grad_norm": 0.11234042048454285,
      "learning_rate": 0.00013631541725601132,
      "loss": 2.1950128078460693,
      "step": 9050
    },
    {
      "epoch": 0.9218814447426177,
      "grad_norm": 0.12438163161277771,
      "learning_rate": 0.00013596181046676096,
      "loss": 2.2238359451293945,
      "step": 9052
    },
    {
      "epoch": 0.9220851304352254,
      "grad_norm": 0.11969199031591415,
      "learning_rate": 0.00013560820367751063,
      "loss": 2.1414575576782227,
      "step": 9054
    },
    {
      "epoch": 0.9222888161278331,
      "grad_norm": 0.10232295095920563,
      "learning_rate": 0.00013525459688826024,
      "loss": 2.128643274307251,
      "step": 9056
    },
    {
      "epoch": 0.9224925018204408,
      "grad_norm": 0.1271112859249115,
      "learning_rate": 0.0001349009900990099,
      "loss": 2.212498664855957,
      "step": 9058
    },
    {
      "epoch": 0.9226961875130486,
      "grad_norm": 0.12364928424358368,
      "learning_rate": 0.00013454738330975955,
      "loss": 2.336595058441162,
      "step": 9060
    },
    {
      "epoch": 0.9228998732056564,
      "grad_norm": 0.1287878453731537,
      "learning_rate": 0.0001341937765205092,
      "loss": 2.1608264446258545,
      "step": 9062
    },
    {
      "epoch": 0.9231035588982641,
      "grad_norm": 0.11372310668230057,
      "learning_rate": 0.00013384016973125884,
      "loss": 2.1642227172851562,
      "step": 9064
    },
    {
      "epoch": 0.9233072445908719,
      "grad_norm": 0.12102334201335907,
      "learning_rate": 0.00013348656294200848,
      "loss": 2.1749467849731445,
      "step": 9066
    },
    {
      "epoch": 0.9235109302834795,
      "grad_norm": 0.1364644467830658,
      "learning_rate": 0.00013313295615275815,
      "loss": 2.2044432163238525,
      "step": 9068
    },
    {
      "epoch": 0.9237146159760873,
      "grad_norm": 0.11966898292303085,
      "learning_rate": 0.00013277934936350777,
      "loss": 2.0846052169799805,
      "step": 9070
    },
    {
      "epoch": 0.923918301668695,
      "grad_norm": 0.1220482587814331,
      "learning_rate": 0.00013242574257425743,
      "loss": 2.217620849609375,
      "step": 9072
    },
    {
      "epoch": 0.9241219873613028,
      "grad_norm": 0.13672393560409546,
      "learning_rate": 0.00013207213578500708,
      "loss": 2.3254871368408203,
      "step": 9074
    },
    {
      "epoch": 0.9243256730539106,
      "grad_norm": 0.1155407577753067,
      "learning_rate": 0.00013171852899575672,
      "loss": 2.189077854156494,
      "step": 9076
    },
    {
      "epoch": 0.9245293587465182,
      "grad_norm": 0.12148711830377579,
      "learning_rate": 0.00013136492220650636,
      "loss": 2.1969215869903564,
      "step": 9078
    },
    {
      "epoch": 0.924733044439126,
      "grad_norm": 0.13652998208999634,
      "learning_rate": 0.00013101131541725603,
      "loss": 2.22811222076416,
      "step": 9080
    },
    {
      "epoch": 0.9249367301317337,
      "grad_norm": 0.12698987126350403,
      "learning_rate": 0.00013065770862800567,
      "loss": 2.331167697906494,
      "step": 9082
    },
    {
      "epoch": 0.9251404158243415,
      "grad_norm": 0.11161459237337112,
      "learning_rate": 0.0001303041018387553,
      "loss": 2.296201705932617,
      "step": 9084
    },
    {
      "epoch": 0.9253441015169492,
      "grad_norm": 0.12136905640363693,
      "learning_rate": 0.00012995049504950496,
      "loss": 2.3020119667053223,
      "step": 9086
    },
    {
      "epoch": 0.9255477872095569,
      "grad_norm": 0.10611245036125183,
      "learning_rate": 0.0001295968882602546,
      "loss": 2.2527971267700195,
      "step": 9088
    },
    {
      "epoch": 0.9257514729021646,
      "grad_norm": 0.10820122063159943,
      "learning_rate": 0.00012924328147100424,
      "loss": 2.049954414367676,
      "step": 9090
    },
    {
      "epoch": 0.9259551585947724,
      "grad_norm": 0.10664884746074677,
      "learning_rate": 0.00012888967468175388,
      "loss": 2.200984477996826,
      "step": 9092
    },
    {
      "epoch": 0.9261588442873802,
      "grad_norm": 0.11485740542411804,
      "learning_rate": 0.00012853606789250355,
      "loss": 2.2292580604553223,
      "step": 9094
    },
    {
      "epoch": 0.9263625299799879,
      "grad_norm": 0.11779829114675522,
      "learning_rate": 0.0001281824611032532,
      "loss": 2.183425188064575,
      "step": 9096
    },
    {
      "epoch": 0.9265662156725957,
      "grad_norm": 0.09601492434740067,
      "learning_rate": 0.0001278288543140028,
      "loss": 2.083014488220215,
      "step": 9098
    },
    {
      "epoch": 0.9267699013652033,
      "grad_norm": 0.13369454443454742,
      "learning_rate": 0.00012747524752475248,
      "loss": 2.3276000022888184,
      "step": 9100
    },
    {
      "epoch": 0.9269735870578111,
      "grad_norm": 0.10692059248685837,
      "learning_rate": 0.00012712164073550212,
      "loss": 2.3091630935668945,
      "step": 9102
    },
    {
      "epoch": 0.9271772727504188,
      "grad_norm": 0.12820445001125336,
      "learning_rate": 0.00012676803394625176,
      "loss": 2.2460110187530518,
      "step": 9104
    },
    {
      "epoch": 0.9273809584430266,
      "grad_norm": 0.11383721232414246,
      "learning_rate": 0.0001264144271570014,
      "loss": 2.1684036254882812,
      "step": 9106
    },
    {
      "epoch": 0.9275846441356344,
      "grad_norm": 0.1334749460220337,
      "learning_rate": 0.00012606082036775108,
      "loss": 2.162416934967041,
      "step": 9108
    },
    {
      "epoch": 0.927788329828242,
      "grad_norm": 0.11877596378326416,
      "learning_rate": 0.00012570721357850072,
      "loss": 2.2965049743652344,
      "step": 9110
    },
    {
      "epoch": 0.9279920155208498,
      "grad_norm": 0.12788313627243042,
      "learning_rate": 0.00012535360678925036,
      "loss": 2.3388123512268066,
      "step": 9112
    },
    {
      "epoch": 0.9281957012134575,
      "grad_norm": 0.10805898904800415,
      "learning_rate": 0.000125,
      "loss": 2.132335662841797,
      "step": 9114
    },
    {
      "epoch": 0.9283993869060653,
      "grad_norm": 0.09816355258226395,
      "learning_rate": 0.00012464639321074964,
      "loss": 2.1868371963500977,
      "step": 9116
    },
    {
      "epoch": 0.928603072598673,
      "grad_norm": 0.11228480190038681,
      "learning_rate": 0.0001242927864214993,
      "loss": 2.2351274490356445,
      "step": 9118
    },
    {
      "epoch": 0.9288067582912807,
      "grad_norm": 0.11491575837135315,
      "learning_rate": 0.00012393917963224896,
      "loss": 2.3226869106292725,
      "step": 9120
    },
    {
      "epoch": 0.9290104439838884,
      "grad_norm": 0.12003765255212784,
      "learning_rate": 0.0001235855728429986,
      "loss": 2.265395164489746,
      "step": 9122
    },
    {
      "epoch": 0.9292141296764962,
      "grad_norm": 0.13683675229549408,
      "learning_rate": 0.00012323196605374824,
      "loss": 2.2457499504089355,
      "step": 9124
    },
    {
      "epoch": 0.929417815369104,
      "grad_norm": 0.1251429319381714,
      "learning_rate": 0.00012287835926449788,
      "loss": 2.1820826530456543,
      "step": 9126
    },
    {
      "epoch": 0.9296215010617117,
      "grad_norm": 0.14274528622627258,
      "learning_rate": 0.00012252475247524753,
      "loss": 2.2386746406555176,
      "step": 9128
    },
    {
      "epoch": 0.9298251867543194,
      "grad_norm": 0.12531468272209167,
      "learning_rate": 0.00012217114568599717,
      "loss": 2.1713576316833496,
      "step": 9130
    },
    {
      "epoch": 0.9300288724469271,
      "grad_norm": 0.12717337906360626,
      "learning_rate": 0.00012181753889674682,
      "loss": 2.319697856903076,
      "step": 9132
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 0.12723161280155182,
      "learning_rate": 0.00012146393210749647,
      "loss": 2.154240608215332,
      "step": 9134
    },
    {
      "epoch": 0.9304362438321426,
      "grad_norm": 0.13798104226589203,
      "learning_rate": 0.00012111032531824612,
      "loss": 2.249922275543213,
      "step": 9136
    },
    {
      "epoch": 0.9306399295247504,
      "grad_norm": 0.11848773807287216,
      "learning_rate": 0.00012075671852899576,
      "loss": 2.2781968116760254,
      "step": 9138
    },
    {
      "epoch": 0.9308436152173581,
      "grad_norm": 0.11025025695562363,
      "learning_rate": 0.0001204031117397454,
      "loss": 2.3106565475463867,
      "step": 9140
    },
    {
      "epoch": 0.9310473009099658,
      "grad_norm": 0.10690717399120331,
      "learning_rate": 0.00012004950495049505,
      "loss": 2.0993638038635254,
      "step": 9142
    },
    {
      "epoch": 0.9312509866025735,
      "grad_norm": 0.13328315317630768,
      "learning_rate": 0.00011969589816124469,
      "loss": 2.1752381324768066,
      "step": 9144
    },
    {
      "epoch": 0.9314546722951813,
      "grad_norm": 0.10852692276239395,
      "learning_rate": 0.00011934229137199435,
      "loss": 2.363539695739746,
      "step": 9146
    },
    {
      "epoch": 0.9316583579877891,
      "grad_norm": 0.10953141748905182,
      "learning_rate": 0.00011898868458274399,
      "loss": 2.236314296722412,
      "step": 9148
    },
    {
      "epoch": 0.9318620436803968,
      "grad_norm": 0.12454991042613983,
      "learning_rate": 0.00011863507779349364,
      "loss": 2.1407134532928467,
      "step": 9150
    },
    {
      "epoch": 0.9320657293730045,
      "grad_norm": 0.11077451705932617,
      "learning_rate": 0.00011828147100424329,
      "loss": 2.2655487060546875,
      "step": 9152
    },
    {
      "epoch": 0.9322694150656122,
      "grad_norm": 0.15028011798858643,
      "learning_rate": 0.00011792786421499294,
      "loss": 2.1741437911987305,
      "step": 9154
    },
    {
      "epoch": 0.93247310075822,
      "grad_norm": 0.1051202267408371,
      "learning_rate": 0.00011757425742574257,
      "loss": 2.185573101043701,
      "step": 9156
    },
    {
      "epoch": 0.9326767864508277,
      "grad_norm": 0.12138890475034714,
      "learning_rate": 0.00011722065063649221,
      "loss": 2.254913568496704,
      "step": 9158
    },
    {
      "epoch": 0.9328804721434355,
      "grad_norm": 0.16427718102931976,
      "learning_rate": 0.00011686704384724187,
      "loss": 2.1415646076202393,
      "step": 9160
    },
    {
      "epoch": 0.9330841578360431,
      "grad_norm": 0.11628793179988861,
      "learning_rate": 0.00011651343705799151,
      "loss": 2.0478434562683105,
      "step": 9162
    },
    {
      "epoch": 0.9332878435286509,
      "grad_norm": 0.11973326653242111,
      "learning_rate": 0.00011615983026874117,
      "loss": 2.147087574005127,
      "step": 9164
    },
    {
      "epoch": 0.9334915292212587,
      "grad_norm": 0.11862340569496155,
      "learning_rate": 0.00011580622347949081,
      "loss": 2.2078781127929688,
      "step": 9166
    },
    {
      "epoch": 0.9336952149138664,
      "grad_norm": 0.11614961177110672,
      "learning_rate": 0.00011545261669024046,
      "loss": 2.2461538314819336,
      "step": 9168
    },
    {
      "epoch": 0.9338989006064742,
      "grad_norm": 0.10843377560377121,
      "learning_rate": 0.0001150990099009901,
      "loss": 2.086442708969116,
      "step": 9170
    },
    {
      "epoch": 0.9341025862990819,
      "grad_norm": 0.12939366698265076,
      "learning_rate": 0.00011474540311173975,
      "loss": 2.267122268676758,
      "step": 9172
    },
    {
      "epoch": 0.9343062719916896,
      "grad_norm": 0.11181405186653137,
      "learning_rate": 0.00011439179632248939,
      "loss": 1.9863405227661133,
      "step": 9174
    },
    {
      "epoch": 0.9345099576842973,
      "grad_norm": 0.13082173466682434,
      "learning_rate": 0.00011403818953323903,
      "loss": 2.2257165908813477,
      "step": 9176
    },
    {
      "epoch": 0.9347136433769051,
      "grad_norm": 0.102768175303936,
      "learning_rate": 0.00011368458274398869,
      "loss": 2.196721315383911,
      "step": 9178
    },
    {
      "epoch": 0.9349173290695129,
      "grad_norm": 0.11763963103294373,
      "learning_rate": 0.00011333097595473833,
      "loss": 2.1949877738952637,
      "step": 9180
    },
    {
      "epoch": 0.9351210147621206,
      "grad_norm": 0.11588013172149658,
      "learning_rate": 0.00011297736916548799,
      "loss": 2.191368579864502,
      "step": 9182
    },
    {
      "epoch": 0.9353247004547283,
      "grad_norm": 0.1093534603714943,
      "learning_rate": 0.00011262376237623763,
      "loss": 2.2454843521118164,
      "step": 9184
    },
    {
      "epoch": 0.935528386147336,
      "grad_norm": 0.13738007843494415,
      "learning_rate": 0.00011227015558698727,
      "loss": 2.2433559894561768,
      "step": 9186
    },
    {
      "epoch": 0.9357320718399438,
      "grad_norm": 0.11395280808210373,
      "learning_rate": 0.00011191654879773693,
      "loss": 2.0781924724578857,
      "step": 9188
    },
    {
      "epoch": 0.9359357575325515,
      "grad_norm": 0.10984939336776733,
      "learning_rate": 0.00011156294200848656,
      "loss": 2.1403844356536865,
      "step": 9190
    },
    {
      "epoch": 0.9361394432251593,
      "grad_norm": 0.14347213506698608,
      "learning_rate": 0.00011120933521923621,
      "loss": 2.1784892082214355,
      "step": 9192
    },
    {
      "epoch": 0.9363431289177669,
      "grad_norm": 0.12389325350522995,
      "learning_rate": 0.00011085572842998585,
      "loss": 2.1596412658691406,
      "step": 9194
    },
    {
      "epoch": 0.9365468146103747,
      "grad_norm": 0.13000668585300446,
      "learning_rate": 0.00011050212164073551,
      "loss": 2.1611685752868652,
      "step": 9196
    },
    {
      "epoch": 0.9367505003029825,
      "grad_norm": 0.12349081039428711,
      "learning_rate": 0.00011014851485148515,
      "loss": 2.1272740364074707,
      "step": 9198
    },
    {
      "epoch": 0.9369541859955902,
      "grad_norm": 0.10120643675327301,
      "learning_rate": 0.0001097949080622348,
      "loss": 2.132107734680176,
      "step": 9200
    },
    {
      "epoch": 0.937157871688198,
      "grad_norm": 0.10754270106554031,
      "learning_rate": 0.00010944130127298445,
      "loss": 2.265242576599121,
      "step": 9202
    },
    {
      "epoch": 0.9373615573808057,
      "grad_norm": 0.15750521421432495,
      "learning_rate": 0.00010908769448373409,
      "loss": 2.160656452178955,
      "step": 9204
    },
    {
      "epoch": 0.9375652430734134,
      "grad_norm": 0.15385378897190094,
      "learning_rate": 0.00010873408769448373,
      "loss": 2.218907356262207,
      "step": 9206
    },
    {
      "epoch": 0.9377689287660211,
      "grad_norm": 0.11380434036254883,
      "learning_rate": 0.00010838048090523338,
      "loss": 2.3030762672424316,
      "step": 9208
    },
    {
      "epoch": 0.9379726144586289,
      "grad_norm": 0.12990029156208038,
      "learning_rate": 0.00010802687411598303,
      "loss": 2.200498580932617,
      "step": 9210
    },
    {
      "epoch": 0.9381763001512367,
      "grad_norm": 0.10778291523456573,
      "learning_rate": 0.00010767326732673267,
      "loss": 2.1122517585754395,
      "step": 9212
    },
    {
      "epoch": 0.9383799858438444,
      "grad_norm": 0.14351433515548706,
      "learning_rate": 0.00010731966053748232,
      "loss": 2.305807590484619,
      "step": 9214
    },
    {
      "epoch": 0.9385836715364521,
      "grad_norm": 0.15836837887763977,
      "learning_rate": 0.00010696605374823197,
      "loss": 2.2919695377349854,
      "step": 9216
    },
    {
      "epoch": 0.9387873572290598,
      "grad_norm": 0.12802541255950928,
      "learning_rate": 0.00010661244695898161,
      "loss": 2.2272937297821045,
      "step": 9218
    },
    {
      "epoch": 0.9389910429216676,
      "grad_norm": 0.11549891531467438,
      "learning_rate": 0.00010625884016973127,
      "loss": 2.303804874420166,
      "step": 9220
    },
    {
      "epoch": 0.9391947286142753,
      "grad_norm": 0.10720356553792953,
      "learning_rate": 0.00010590523338048091,
      "loss": 2.293506622314453,
      "step": 9222
    },
    {
      "epoch": 0.9393984143068831,
      "grad_norm": 0.10646568238735199,
      "learning_rate": 0.00010555162659123055,
      "loss": 2.145371675491333,
      "step": 9224
    },
    {
      "epoch": 0.9396020999994907,
      "grad_norm": 0.10978415608406067,
      "learning_rate": 0.0001051980198019802,
      "loss": 2.2089126110076904,
      "step": 9226
    },
    {
      "epoch": 0.9398057856920985,
      "grad_norm": 0.11341012269258499,
      "learning_rate": 0.00010484441301272984,
      "loss": 2.1192691326141357,
      "step": 9228
    },
    {
      "epoch": 0.9400094713847063,
      "grad_norm": 0.10308238863945007,
      "learning_rate": 0.0001044908062234795,
      "loss": 2.077390193939209,
      "step": 9230
    },
    {
      "epoch": 0.940213157077314,
      "grad_norm": 0.11100722849369049,
      "learning_rate": 0.00010413719943422914,
      "loss": 2.372842788696289,
      "step": 9232
    },
    {
      "epoch": 0.9404168427699218,
      "grad_norm": 0.11067204177379608,
      "learning_rate": 0.00010378359264497879,
      "loss": 2.0601258277893066,
      "step": 9234
    },
    {
      "epoch": 0.9406205284625295,
      "grad_norm": 0.10486249625682831,
      "learning_rate": 0.00010342998585572843,
      "loss": 2.0915703773498535,
      "step": 9236
    },
    {
      "epoch": 0.9408242141551372,
      "grad_norm": 0.12225145846605301,
      "learning_rate": 0.00010307637906647809,
      "loss": 2.2491812705993652,
      "step": 9238
    },
    {
      "epoch": 0.9410278998477449,
      "grad_norm": 0.1642012894153595,
      "learning_rate": 0.00010272277227722772,
      "loss": 2.2399520874023438,
      "step": 9240
    },
    {
      "epoch": 0.9412315855403527,
      "grad_norm": 0.14131928980350494,
      "learning_rate": 0.00010236916548797736,
      "loss": 2.120420455932617,
      "step": 9242
    },
    {
      "epoch": 0.9414352712329604,
      "grad_norm": 0.10976595431566238,
      "learning_rate": 0.00010201555869872702,
      "loss": 2.124222755432129,
      "step": 9244
    },
    {
      "epoch": 0.9416389569255682,
      "grad_norm": 0.17113180458545685,
      "learning_rate": 0.00010166195190947666,
      "loss": 2.157519817352295,
      "step": 9246
    },
    {
      "epoch": 0.9418426426181759,
      "grad_norm": 0.10121867060661316,
      "learning_rate": 0.00010130834512022632,
      "loss": 2.1751132011413574,
      "step": 9248
    },
    {
      "epoch": 0.9420463283107836,
      "grad_norm": 0.12448973953723907,
      "learning_rate": 0.00010095473833097596,
      "loss": 2.1618363857269287,
      "step": 9250
    },
    {
      "epoch": 0.9422500140033914,
      "grad_norm": 0.10711071640253067,
      "learning_rate": 0.00010060113154172561,
      "loss": 2.209846019744873,
      "step": 9252
    },
    {
      "epoch": 0.9424536996959991,
      "grad_norm": 0.10651564598083496,
      "learning_rate": 0.00010024752475247526,
      "loss": 2.1102561950683594,
      "step": 9254
    },
    {
      "epoch": 0.9426573853886069,
      "grad_norm": 0.1262148767709732,
      "learning_rate": 9.989391796322488e-05,
      "loss": 2.270341396331787,
      "step": 9256
    },
    {
      "epoch": 0.9428610710812145,
      "grad_norm": 0.12098293751478195,
      "learning_rate": 9.954031117397454e-05,
      "loss": 2.1850128173828125,
      "step": 9258
    },
    {
      "epoch": 0.9430647567738223,
      "grad_norm": 0.1474771648645401,
      "learning_rate": 9.918670438472418e-05,
      "loss": 2.092100143432617,
      "step": 9260
    },
    {
      "epoch": 0.94326844246643,
      "grad_norm": 0.11741890013217926,
      "learning_rate": 9.883309759547384e-05,
      "loss": 2.2454986572265625,
      "step": 9262
    },
    {
      "epoch": 0.9434721281590378,
      "grad_norm": 0.1303734928369522,
      "learning_rate": 9.847949080622348e-05,
      "loss": 2.3352105617523193,
      "step": 9264
    },
    {
      "epoch": 0.9436758138516456,
      "grad_norm": 0.12090927362442017,
      "learning_rate": 9.812588401697314e-05,
      "loss": 2.212773323059082,
      "step": 9266
    },
    {
      "epoch": 0.9438794995442532,
      "grad_norm": 0.127187579870224,
      "learning_rate": 9.777227722772278e-05,
      "loss": 2.288011074066162,
      "step": 9268
    },
    {
      "epoch": 0.944083185236861,
      "grad_norm": 0.12787781655788422,
      "learning_rate": 9.741867043847242e-05,
      "loss": 2.221132516860962,
      "step": 9270
    },
    {
      "epoch": 0.9442868709294687,
      "grad_norm": 0.104576475918293,
      "learning_rate": 9.706506364922208e-05,
      "loss": 2.0811891555786133,
      "step": 9272
    },
    {
      "epoch": 0.9444905566220765,
      "grad_norm": 0.12026559561491013,
      "learning_rate": 9.67114568599717e-05,
      "loss": 2.132599115371704,
      "step": 9274
    },
    {
      "epoch": 0.9446942423146842,
      "grad_norm": 0.1199767217040062,
      "learning_rate": 9.635785007072136e-05,
      "loss": 2.250892162322998,
      "step": 9276
    },
    {
      "epoch": 0.944897928007292,
      "grad_norm": 0.11914829164743423,
      "learning_rate": 9.6004243281471e-05,
      "loss": 2.104621171951294,
      "step": 9278
    },
    {
      "epoch": 0.9451016136998996,
      "grad_norm": 0.1463329941034317,
      "learning_rate": 9.565063649222066e-05,
      "loss": 2.11488676071167,
      "step": 9280
    },
    {
      "epoch": 0.9453052993925074,
      "grad_norm": 0.11628095805644989,
      "learning_rate": 9.52970297029703e-05,
      "loss": 2.164492607116699,
      "step": 9282
    },
    {
      "epoch": 0.9455089850851152,
      "grad_norm": 0.10800041258335114,
      "learning_rate": 9.494342291371994e-05,
      "loss": 2.0224151611328125,
      "step": 9284
    },
    {
      "epoch": 0.9457126707777229,
      "grad_norm": 0.11747812479734421,
      "learning_rate": 9.45898161244696e-05,
      "loss": 2.2268574237823486,
      "step": 9286
    },
    {
      "epoch": 0.9459163564703307,
      "grad_norm": 0.10736073553562164,
      "learning_rate": 9.423620933521924e-05,
      "loss": 2.147866725921631,
      "step": 9288
    },
    {
      "epoch": 0.9461200421629383,
      "grad_norm": 0.11454814672470093,
      "learning_rate": 9.388260254596888e-05,
      "loss": 2.1760528087615967,
      "step": 9290
    },
    {
      "epoch": 0.9463237278555461,
      "grad_norm": 0.1253494918346405,
      "learning_rate": 9.352899575671852e-05,
      "loss": 2.14935564994812,
      "step": 9292
    },
    {
      "epoch": 0.9465274135481538,
      "grad_norm": 0.14049692451953888,
      "learning_rate": 9.317538896746818e-05,
      "loss": 2.130643367767334,
      "step": 9294
    },
    {
      "epoch": 0.9467310992407616,
      "grad_norm": 0.1123703196644783,
      "learning_rate": 9.282178217821782e-05,
      "loss": 2.084892511367798,
      "step": 9296
    },
    {
      "epoch": 0.9469347849333694,
      "grad_norm": 0.11608383059501648,
      "learning_rate": 9.246817538896747e-05,
      "loss": 2.3804268836975098,
      "step": 9298
    },
    {
      "epoch": 0.947138470625977,
      "grad_norm": 0.11187326908111572,
      "learning_rate": 9.211456859971712e-05,
      "loss": 2.230483055114746,
      "step": 9300
    },
    {
      "epoch": 0.9473421563185848,
      "grad_norm": 0.12833163142204285,
      "learning_rate": 9.176096181046676e-05,
      "loss": 2.2496094703674316,
      "step": 9302
    },
    {
      "epoch": 0.9475458420111925,
      "grad_norm": 0.11539774388074875,
      "learning_rate": 9.140735502121642e-05,
      "loss": 2.2908129692077637,
      "step": 9304
    },
    {
      "epoch": 0.9477495277038003,
      "grad_norm": 0.11081960797309875,
      "learning_rate": 9.105374823196605e-05,
      "loss": 2.1521410942077637,
      "step": 9306
    },
    {
      "epoch": 0.947953213396408,
      "grad_norm": 0.11019822955131531,
      "learning_rate": 9.07001414427157e-05,
      "loss": 2.138852119445801,
      "step": 9308
    },
    {
      "epoch": 0.9481568990890158,
      "grad_norm": 0.11895892024040222,
      "learning_rate": 9.034653465346535e-05,
      "loss": 2.238398551940918,
      "step": 9310
    },
    {
      "epoch": 0.9483605847816234,
      "grad_norm": 0.11299639195203781,
      "learning_rate": 8.999292786421499e-05,
      "loss": 2.127558708190918,
      "step": 9312
    },
    {
      "epoch": 0.9485642704742312,
      "grad_norm": 0.14211677014827728,
      "learning_rate": 8.963932107496464e-05,
      "loss": 2.1127636432647705,
      "step": 9314
    },
    {
      "epoch": 0.948767956166839,
      "grad_norm": 0.1361115425825119,
      "learning_rate": 8.928571428571429e-05,
      "loss": 2.1815731525421143,
      "step": 9316
    },
    {
      "epoch": 0.9489716418594467,
      "grad_norm": 0.11957821995019913,
      "learning_rate": 8.893210749646394e-05,
      "loss": 2.321918487548828,
      "step": 9318
    },
    {
      "epoch": 0.9491753275520545,
      "grad_norm": 0.12223505973815918,
      "learning_rate": 8.857850070721358e-05,
      "loss": 2.1873111724853516,
      "step": 9320
    },
    {
      "epoch": 0.9493790132446621,
      "grad_norm": 0.12203627079725266,
      "learning_rate": 8.822489391796323e-05,
      "loss": 2.189189910888672,
      "step": 9322
    },
    {
      "epoch": 0.9495826989372699,
      "grad_norm": 0.11903677135705948,
      "learning_rate": 8.787128712871287e-05,
      "loss": 2.180325508117676,
      "step": 9324
    },
    {
      "epoch": 0.9497863846298776,
      "grad_norm": 0.12294286489486694,
      "learning_rate": 8.751768033946251e-05,
      "loss": 2.218317985534668,
      "step": 9326
    },
    {
      "epoch": 0.9499900703224854,
      "grad_norm": 0.1259346306324005,
      "learning_rate": 8.716407355021217e-05,
      "loss": 2.1491622924804688,
      "step": 9328
    },
    {
      "epoch": 0.9501937560150931,
      "grad_norm": 0.10246121883392334,
      "learning_rate": 8.681046676096181e-05,
      "loss": 2.192349910736084,
      "step": 9330
    },
    {
      "epoch": 0.9503974417077008,
      "grad_norm": 0.13106542825698853,
      "learning_rate": 8.645685997171146e-05,
      "loss": 2.276650905609131,
      "step": 9332
    },
    {
      "epoch": 0.9506011274003086,
      "grad_norm": 0.13523755967617035,
      "learning_rate": 8.61032531824611e-05,
      "loss": 2.219226598739624,
      "step": 9334
    },
    {
      "epoch": 0.9508048130929163,
      "grad_norm": 0.1089225560426712,
      "learning_rate": 8.574964639321076e-05,
      "loss": 2.161916971206665,
      "step": 9336
    },
    {
      "epoch": 0.9510084987855241,
      "grad_norm": 0.12683171033859253,
      "learning_rate": 8.53960396039604e-05,
      "loss": 2.166428327560425,
      "step": 9338
    },
    {
      "epoch": 0.9512121844781318,
      "grad_norm": 0.11972793191671371,
      "learning_rate": 8.504243281471003e-05,
      "loss": 2.2006397247314453,
      "step": 9340
    },
    {
      "epoch": 0.9514158701707396,
      "grad_norm": 0.11917595565319061,
      "learning_rate": 8.468882602545969e-05,
      "loss": 2.2170767784118652,
      "step": 9342
    },
    {
      "epoch": 0.9516195558633472,
      "grad_norm": 0.1250493973493576,
      "learning_rate": 8.433521923620933e-05,
      "loss": 2.3466780185699463,
      "step": 9344
    },
    {
      "epoch": 0.951823241555955,
      "grad_norm": 0.12364175170660019,
      "learning_rate": 8.398161244695899e-05,
      "loss": 2.1524055004119873,
      "step": 9346
    },
    {
      "epoch": 0.9520269272485627,
      "grad_norm": 0.10736975818872452,
      "learning_rate": 8.362800565770863e-05,
      "loss": 2.241367816925049,
      "step": 9348
    },
    {
      "epoch": 0.9522306129411705,
      "grad_norm": 0.10855330526828766,
      "learning_rate": 8.327439886845828e-05,
      "loss": 2.1704888343811035,
      "step": 9350
    },
    {
      "epoch": 0.9524342986337783,
      "grad_norm": 0.12158513814210892,
      "learning_rate": 8.292079207920793e-05,
      "loss": 2.245576858520508,
      "step": 9352
    },
    {
      "epoch": 0.9526379843263859,
      "grad_norm": 0.11190270632505417,
      "learning_rate": 8.256718528995757e-05,
      "loss": 2.1433193683624268,
      "step": 9354
    },
    {
      "epoch": 0.9528416700189937,
      "grad_norm": 0.10676945000886917,
      "learning_rate": 8.221357850070721e-05,
      "loss": 2.1020150184631348,
      "step": 9356
    },
    {
      "epoch": 0.9530453557116014,
      "grad_norm": 0.11045260727405548,
      "learning_rate": 8.185997171145685e-05,
      "loss": 2.195345163345337,
      "step": 9358
    },
    {
      "epoch": 0.9532490414042092,
      "grad_norm": 0.10934927314519882,
      "learning_rate": 8.150636492220651e-05,
      "loss": 2.205892562866211,
      "step": 9360
    },
    {
      "epoch": 0.9534527270968169,
      "grad_norm": 0.12425072491168976,
      "learning_rate": 8.115275813295615e-05,
      "loss": 2.1409032344818115,
      "step": 9362
    },
    {
      "epoch": 0.9536564127894246,
      "grad_norm": 0.10545887798070908,
      "learning_rate": 8.079915134370581e-05,
      "loss": 2.1317555904388428,
      "step": 9364
    },
    {
      "epoch": 0.9538600984820323,
      "grad_norm": 0.13107380270957947,
      "learning_rate": 8.044554455445545e-05,
      "loss": 2.181759834289551,
      "step": 9366
    },
    {
      "epoch": 0.9540637841746401,
      "grad_norm": 0.13036632537841797,
      "learning_rate": 8.009193776520509e-05,
      "loss": 2.1651690006256104,
      "step": 9368
    },
    {
      "epoch": 0.9542674698672479,
      "grad_norm": 0.11655186116695404,
      "learning_rate": 7.973833097595475e-05,
      "loss": 2.182793378829956,
      "step": 9370
    },
    {
      "epoch": 0.9544711555598556,
      "grad_norm": 0.10612672567367554,
      "learning_rate": 7.938472418670438e-05,
      "loss": 2.0898447036743164,
      "step": 9372
    },
    {
      "epoch": 0.9546748412524633,
      "grad_norm": 0.1383945345878601,
      "learning_rate": 7.903111739745403e-05,
      "loss": 2.1392641067504883,
      "step": 9374
    },
    {
      "epoch": 0.954878526945071,
      "grad_norm": 0.12192133814096451,
      "learning_rate": 7.867751060820367e-05,
      "loss": 2.2283270359039307,
      "step": 9376
    },
    {
      "epoch": 0.9550822126376788,
      "grad_norm": 0.14710542559623718,
      "learning_rate": 7.832390381895333e-05,
      "loss": 2.198166847229004,
      "step": 9378
    },
    {
      "epoch": 0.9552858983302865,
      "grad_norm": 0.11825201660394669,
      "learning_rate": 7.797029702970297e-05,
      "loss": 2.0016307830810547,
      "step": 9380
    },
    {
      "epoch": 0.9554895840228943,
      "grad_norm": 0.1253078579902649,
      "learning_rate": 7.761669024045261e-05,
      "loss": 2.21748685836792,
      "step": 9382
    },
    {
      "epoch": 0.9556932697155021,
      "grad_norm": 0.11027050018310547,
      "learning_rate": 7.726308345120227e-05,
      "loss": 1.986274242401123,
      "step": 9384
    },
    {
      "epoch": 0.9558969554081097,
      "grad_norm": 0.11016513407230377,
      "learning_rate": 7.690947666195191e-05,
      "loss": 2.0804338455200195,
      "step": 9386
    },
    {
      "epoch": 0.9561006411007175,
      "grad_norm": 0.10916751623153687,
      "learning_rate": 7.655586987270157e-05,
      "loss": 2.02974271774292,
      "step": 9388
    },
    {
      "epoch": 0.9563043267933252,
      "grad_norm": 0.11007675528526306,
      "learning_rate": 7.62022630834512e-05,
      "loss": 2.122117042541504,
      "step": 9390
    },
    {
      "epoch": 0.956508012485933,
      "grad_norm": 0.125327467918396,
      "learning_rate": 7.584865629420085e-05,
      "loss": 2.2634007930755615,
      "step": 9392
    },
    {
      "epoch": 0.9567116981785407,
      "grad_norm": 0.11772444099187851,
      "learning_rate": 7.54950495049505e-05,
      "loss": 2.1846489906311035,
      "step": 9394
    },
    {
      "epoch": 0.9569153838711484,
      "grad_norm": 0.1031203642487526,
      "learning_rate": 7.514144271570014e-05,
      "loss": 2.279487133026123,
      "step": 9396
    },
    {
      "epoch": 0.9571190695637561,
      "grad_norm": 0.1199166476726532,
      "learning_rate": 7.478783592644979e-05,
      "loss": 2.063375473022461,
      "step": 9398
    },
    {
      "epoch": 0.9573227552563639,
      "grad_norm": 0.14151445031166077,
      "learning_rate": 7.443422913719943e-05,
      "loss": 2.1262831687927246,
      "step": 9400
    },
    {
      "epoch": 0.9575264409489717,
      "grad_norm": 0.1062505841255188,
      "learning_rate": 7.408062234794909e-05,
      "loss": 2.1298398971557617,
      "step": 9402
    },
    {
      "epoch": 0.9577301266415794,
      "grad_norm": 0.1329280287027359,
      "learning_rate": 7.372701555869873e-05,
      "loss": 2.227614402770996,
      "step": 9404
    },
    {
      "epoch": 0.9579338123341871,
      "grad_norm": 0.11898434907197952,
      "learning_rate": 7.337340876944837e-05,
      "loss": 2.162668228149414,
      "step": 9406
    },
    {
      "epoch": 0.9581374980267948,
      "grad_norm": 0.14432136714458466,
      "learning_rate": 7.301980198019802e-05,
      "loss": 2.1372737884521484,
      "step": 9408
    },
    {
      "epoch": 0.9583411837194026,
      "grad_norm": 0.12269629538059235,
      "learning_rate": 7.266619519094766e-05,
      "loss": 2.0643668174743652,
      "step": 9410
    },
    {
      "epoch": 0.9585448694120103,
      "grad_norm": 0.1369505077600479,
      "learning_rate": 7.231258840169731e-05,
      "loss": 2.2292754650115967,
      "step": 9412
    },
    {
      "epoch": 0.9587485551046181,
      "grad_norm": 0.126042902469635,
      "learning_rate": 7.195898161244696e-05,
      "loss": 2.1363258361816406,
      "step": 9414
    },
    {
      "epoch": 0.9589522407972259,
      "grad_norm": 0.11448273062705994,
      "learning_rate": 7.160537482319661e-05,
      "loss": 2.025153875350952,
      "step": 9416
    },
    {
      "epoch": 0.9591559264898335,
      "grad_norm": 0.12317217141389847,
      "learning_rate": 7.125176803394626e-05,
      "loss": 2.092733144760132,
      "step": 9418
    },
    {
      "epoch": 0.9593596121824413,
      "grad_norm": 0.11160948872566223,
      "learning_rate": 7.089816124469591e-05,
      "loss": 2.160646438598633,
      "step": 9420
    },
    {
      "epoch": 0.959563297875049,
      "grad_norm": 0.13286028802394867,
      "learning_rate": 7.054455445544554e-05,
      "loss": 2.199183464050293,
      "step": 9422
    },
    {
      "epoch": 0.9597669835676568,
      "grad_norm": 0.12729914486408234,
      "learning_rate": 7.019094766619518e-05,
      "loss": 2.234994888305664,
      "step": 9424
    },
    {
      "epoch": 0.9599706692602645,
      "grad_norm": 0.1365533024072647,
      "learning_rate": 6.983734087694484e-05,
      "loss": 2.137701988220215,
      "step": 9426
    },
    {
      "epoch": 0.9601743549528722,
      "grad_norm": 0.12780171632766724,
      "learning_rate": 6.948373408769448e-05,
      "loss": 2.3242692947387695,
      "step": 9428
    },
    {
      "epoch": 0.9603780406454799,
      "grad_norm": 0.11898702383041382,
      "learning_rate": 6.913012729844414e-05,
      "loss": 2.303419828414917,
      "step": 9430
    },
    {
      "epoch": 0.9605817263380877,
      "grad_norm": 0.11971818655729294,
      "learning_rate": 6.877652050919378e-05,
      "loss": 2.160088539123535,
      "step": 9432
    },
    {
      "epoch": 0.9607854120306955,
      "grad_norm": 0.12343229353427887,
      "learning_rate": 6.842291371994343e-05,
      "loss": 2.2596869468688965,
      "step": 9434
    },
    {
      "epoch": 0.9609890977233032,
      "grad_norm": 0.11701284348964691,
      "learning_rate": 6.806930693069308e-05,
      "loss": 2.292311668395996,
      "step": 9436
    },
    {
      "epoch": 0.9611927834159109,
      "grad_norm": 0.12352582067251205,
      "learning_rate": 6.771570014144272e-05,
      "loss": 2.228360891342163,
      "step": 9438
    },
    {
      "epoch": 0.9613964691085186,
      "grad_norm": 0.1383569985628128,
      "learning_rate": 6.736209335219236e-05,
      "loss": 2.183994770050049,
      "step": 9440
    },
    {
      "epoch": 0.9616001548011264,
      "grad_norm": 0.12198712676763535,
      "learning_rate": 6.7008486562942e-05,
      "loss": 2.2112557888031006,
      "step": 9442
    },
    {
      "epoch": 0.9618038404937341,
      "grad_norm": 0.11943031847476959,
      "learning_rate": 6.665487977369166e-05,
      "loss": 2.1269545555114746,
      "step": 9444
    },
    {
      "epoch": 0.9620075261863419,
      "grad_norm": 0.1132841408252716,
      "learning_rate": 6.63012729844413e-05,
      "loss": 2.0687255859375,
      "step": 9446
    },
    {
      "epoch": 0.9622112118789496,
      "grad_norm": 0.10773348063230515,
      "learning_rate": 6.594766619519096e-05,
      "loss": 2.167182683944702,
      "step": 9448
    },
    {
      "epoch": 0.9624148975715573,
      "grad_norm": 0.11921016871929169,
      "learning_rate": 6.55940594059406e-05,
      "loss": 2.2770628929138184,
      "step": 9450
    },
    {
      "epoch": 0.962618583264165,
      "grad_norm": 0.12770235538482666,
      "learning_rate": 6.524045261669024e-05,
      "loss": 2.1104214191436768,
      "step": 9452
    },
    {
      "epoch": 0.9628222689567728,
      "grad_norm": 0.13634364306926727,
      "learning_rate": 6.48868458274399e-05,
      "loss": 2.1756458282470703,
      "step": 9454
    },
    {
      "epoch": 0.9630259546493806,
      "grad_norm": 0.10495017468929291,
      "learning_rate": 6.453323903818952e-05,
      "loss": 2.2565627098083496,
      "step": 9456
    },
    {
      "epoch": 0.9632296403419883,
      "grad_norm": 0.10255300253629684,
      "learning_rate": 6.417963224893918e-05,
      "loss": 2.315197467803955,
      "step": 9458
    },
    {
      "epoch": 0.963433326034596,
      "grad_norm": 0.1457284837961197,
      "learning_rate": 6.382602545968882e-05,
      "loss": 2.3984270095825195,
      "step": 9460
    },
    {
      "epoch": 0.9636370117272037,
      "grad_norm": 0.10401006042957306,
      "learning_rate": 6.347241867043848e-05,
      "loss": 2.117025852203369,
      "step": 9462
    },
    {
      "epoch": 0.9638406974198115,
      "grad_norm": 0.12813422083854675,
      "learning_rate": 6.311881188118812e-05,
      "loss": 2.119793176651001,
      "step": 9464
    },
    {
      "epoch": 0.9640443831124192,
      "grad_norm": 0.10822150856256485,
      "learning_rate": 6.276520509193776e-05,
      "loss": 2.1817126274108887,
      "step": 9466
    },
    {
      "epoch": 0.964248068805027,
      "grad_norm": 0.12386409193277359,
      "learning_rate": 6.24115983026874e-05,
      "loss": 2.19362473487854,
      "step": 9468
    },
    {
      "epoch": 0.9644517544976347,
      "grad_norm": 0.12313814461231232,
      "learning_rate": 6.205799151343706e-05,
      "loss": 2.0600738525390625,
      "step": 9470
    },
    {
      "epoch": 0.9646554401902424,
      "grad_norm": 0.13068616390228271,
      "learning_rate": 6.17043847241867e-05,
      "loss": 2.2872676849365234,
      "step": 9472
    },
    {
      "epoch": 0.9648591258828502,
      "grad_norm": 0.11379009485244751,
      "learning_rate": 6.135077793493636e-05,
      "loss": 2.220521926879883,
      "step": 9474
    },
    {
      "epoch": 0.9650628115754579,
      "grad_norm": 0.11344245076179504,
      "learning_rate": 6.0997171145685994e-05,
      "loss": 2.1224608421325684,
      "step": 9476
    },
    {
      "epoch": 0.9652664972680657,
      "grad_norm": 0.1101764589548111,
      "learning_rate": 6.064356435643564e-05,
      "loss": 2.265286922454834,
      "step": 9478
    },
    {
      "epoch": 0.9654701829606734,
      "grad_norm": 0.1288902461528778,
      "learning_rate": 6.028995756718529e-05,
      "loss": 2.256955862045288,
      "step": 9480
    },
    {
      "epoch": 0.9656738686532811,
      "grad_norm": 0.12054713815450668,
      "learning_rate": 5.993635077793494e-05,
      "loss": 2.0165507793426514,
      "step": 9482
    },
    {
      "epoch": 0.9658775543458888,
      "grad_norm": 0.14531421661376953,
      "learning_rate": 5.958274398868459e-05,
      "loss": 2.1757190227508545,
      "step": 9484
    },
    {
      "epoch": 0.9660812400384966,
      "grad_norm": 0.11881791800260544,
      "learning_rate": 5.9229137199434226e-05,
      "loss": 2.1887636184692383,
      "step": 9486
    },
    {
      "epoch": 0.9662849257311044,
      "grad_norm": 0.1043429747223854,
      "learning_rate": 5.8875530410183875e-05,
      "loss": 2.13974666595459,
      "step": 9488
    },
    {
      "epoch": 0.9664886114237121,
      "grad_norm": 0.10344362258911133,
      "learning_rate": 5.8521923620933524e-05,
      "loss": 2.1253175735473633,
      "step": 9490
    },
    {
      "epoch": 0.9666922971163198,
      "grad_norm": 0.14983251690864563,
      "learning_rate": 5.816831683168317e-05,
      "loss": 2.1936001777648926,
      "step": 9492
    },
    {
      "epoch": 0.9668959828089275,
      "grad_norm": 0.10708626359701157,
      "learning_rate": 5.7814710042432815e-05,
      "loss": 2.322547435760498,
      "step": 9494
    },
    {
      "epoch": 0.9670996685015353,
      "grad_norm": 0.13726229965686798,
      "learning_rate": 5.7461103253182464e-05,
      "loss": 2.117645263671875,
      "step": 9496
    },
    {
      "epoch": 0.967303354194143,
      "grad_norm": 0.11894974857568741,
      "learning_rate": 5.710749646393211e-05,
      "loss": 2.110074520111084,
      "step": 9498
    },
    {
      "epoch": 0.9675070398867508,
      "grad_norm": 0.14057964086532593,
      "learning_rate": 5.6753889674681755e-05,
      "loss": 2.211768627166748,
      "step": 9500
    },
    {
      "epoch": 0.9677107255793584,
      "grad_norm": 0.1133980005979538,
      "learning_rate": 5.64002828854314e-05,
      "loss": 2.2346906661987305,
      "step": 9502
    },
    {
      "epoch": 0.9679144112719662,
      "grad_norm": 0.11173833906650543,
      "learning_rate": 5.6046676096181046e-05,
      "loss": 2.1793203353881836,
      "step": 9504
    },
    {
      "epoch": 0.968118096964574,
      "grad_norm": 0.11201004683971405,
      "learning_rate": 5.5693069306930695e-05,
      "loss": 2.344062328338623,
      "step": 9506
    },
    {
      "epoch": 0.9683217826571817,
      "grad_norm": 0.10441546142101288,
      "learning_rate": 5.5339462517680344e-05,
      "loss": 2.249624252319336,
      "step": 9508
    },
    {
      "epoch": 0.9685254683497895,
      "grad_norm": 0.10716905444860458,
      "learning_rate": 5.4985855728429986e-05,
      "loss": 2.1474807262420654,
      "step": 9510
    },
    {
      "epoch": 0.9687291540423971,
      "grad_norm": 0.12900523841381073,
      "learning_rate": 5.4632248939179635e-05,
      "loss": 2.2663862705230713,
      "step": 9512
    },
    {
      "epoch": 0.9689328397350049,
      "grad_norm": 0.12088318914175034,
      "learning_rate": 5.427864214992928e-05,
      "loss": 2.165233850479126,
      "step": 9514
    },
    {
      "epoch": 0.9691365254276126,
      "grad_norm": 0.10814917832612991,
      "learning_rate": 5.3925035360678926e-05,
      "loss": 2.2060327529907227,
      "step": 9516
    },
    {
      "epoch": 0.9693402111202204,
      "grad_norm": 0.11579562723636627,
      "learning_rate": 5.357142857142857e-05,
      "loss": 2.1695926189422607,
      "step": 9518
    },
    {
      "epoch": 0.9695438968128282,
      "grad_norm": 0.11138170957565308,
      "learning_rate": 5.321782178217822e-05,
      "loss": 2.0535714626312256,
      "step": 9520
    },
    {
      "epoch": 0.9697475825054359,
      "grad_norm": 0.12351871281862259,
      "learning_rate": 5.286421499292787e-05,
      "loss": 2.3228697776794434,
      "step": 9522
    },
    {
      "epoch": 0.9699512681980436,
      "grad_norm": 0.1224401667714119,
      "learning_rate": 5.2510608203677516e-05,
      "loss": 2.0469675064086914,
      "step": 9524
    },
    {
      "epoch": 0.9701549538906513,
      "grad_norm": 0.11771542578935623,
      "learning_rate": 5.215700141442716e-05,
      "loss": 2.0779523849487305,
      "step": 9526
    },
    {
      "epoch": 0.9703586395832591,
      "grad_norm": 0.11742597818374634,
      "learning_rate": 5.18033946251768e-05,
      "loss": 2.218693256378174,
      "step": 9528
    },
    {
      "epoch": 0.9705623252758668,
      "grad_norm": 0.1166633740067482,
      "learning_rate": 5.144978783592645e-05,
      "loss": 2.2231760025024414,
      "step": 9530
    },
    {
      "epoch": 0.9707660109684746,
      "grad_norm": 0.10180516541004181,
      "learning_rate": 5.10961810466761e-05,
      "loss": 2.2016797065734863,
      "step": 9532
    },
    {
      "epoch": 0.9709696966610822,
      "grad_norm": 0.10645721852779388,
      "learning_rate": 5.074257425742575e-05,
      "loss": 2.1183133125305176,
      "step": 9534
    },
    {
      "epoch": 0.97117338235369,
      "grad_norm": 0.11095879971981049,
      "learning_rate": 5.038896746817539e-05,
      "loss": 2.0676369667053223,
      "step": 9536
    },
    {
      "epoch": 0.9713770680462978,
      "grad_norm": 0.0977313295006752,
      "learning_rate": 5.003536067892504e-05,
      "loss": 2.075376033782959,
      "step": 9538
    },
    {
      "epoch": 0.9715807537389055,
      "grad_norm": 0.14062370359897614,
      "learning_rate": 4.968175388967469e-05,
      "loss": 2.071795701980591,
      "step": 9540
    },
    {
      "epoch": 0.9717844394315133,
      "grad_norm": 0.09673719108104706,
      "learning_rate": 4.932814710042433e-05,
      "loss": 2.10202956199646,
      "step": 9542
    },
    {
      "epoch": 0.9719881251241209,
      "grad_norm": 0.1278025358915329,
      "learning_rate": 4.897454031117397e-05,
      "loss": 2.198004722595215,
      "step": 9544
    },
    {
      "epoch": 0.9721918108167287,
      "grad_norm": 0.14980904757976532,
      "learning_rate": 4.862093352192362e-05,
      "loss": 2.2535157203674316,
      "step": 9546
    },
    {
      "epoch": 0.9723954965093364,
      "grad_norm": 0.12090009450912476,
      "learning_rate": 4.826732673267327e-05,
      "loss": 2.0773472785949707,
      "step": 9548
    },
    {
      "epoch": 0.9725991822019442,
      "grad_norm": 0.11658138036727905,
      "learning_rate": 4.791371994342292e-05,
      "loss": 2.160092830657959,
      "step": 9550
    },
    {
      "epoch": 0.972802867894552,
      "grad_norm": 0.12280880659818649,
      "learning_rate": 4.756011315417256e-05,
      "loss": 2.190091609954834,
      "step": 9552
    },
    {
      "epoch": 0.9730065535871597,
      "grad_norm": 0.11875976622104645,
      "learning_rate": 4.720650636492221e-05,
      "loss": 2.3168935775756836,
      "step": 9554
    },
    {
      "epoch": 0.9732102392797674,
      "grad_norm": 0.12153297662734985,
      "learning_rate": 4.685289957567185e-05,
      "loss": 2.111185312271118,
      "step": 9556
    },
    {
      "epoch": 0.9734139249723751,
      "grad_norm": 0.16477368772029877,
      "learning_rate": 4.64992927864215e-05,
      "loss": 2.2440977096557617,
      "step": 9558
    },
    {
      "epoch": 0.9736176106649829,
      "grad_norm": 0.12579382956027985,
      "learning_rate": 4.614568599717114e-05,
      "loss": 2.1934242248535156,
      "step": 9560
    },
    {
      "epoch": 0.9738212963575906,
      "grad_norm": 0.12224811315536499,
      "learning_rate": 4.579207920792079e-05,
      "loss": 2.1892356872558594,
      "step": 9562
    },
    {
      "epoch": 0.9740249820501984,
      "grad_norm": 0.14544080197811127,
      "learning_rate": 4.543847241867044e-05,
      "loss": 2.175766944885254,
      "step": 9564
    },
    {
      "epoch": 0.974228667742806,
      "grad_norm": 0.11315970122814178,
      "learning_rate": 4.508486562942009e-05,
      "loss": 2.2184996604919434,
      "step": 9566
    },
    {
      "epoch": 0.9744323534354138,
      "grad_norm": 0.11514592915773392,
      "learning_rate": 4.473125884016973e-05,
      "loss": 2.2325117588043213,
      "step": 9568
    },
    {
      "epoch": 0.9746360391280215,
      "grad_norm": 0.1155860647559166,
      "learning_rate": 4.4377652050919374e-05,
      "loss": 2.1235179901123047,
      "step": 9570
    },
    {
      "epoch": 0.9748397248206293,
      "grad_norm": 0.12861591577529907,
      "learning_rate": 4.4024045261669023e-05,
      "loss": 2.2270522117614746,
      "step": 9572
    },
    {
      "epoch": 0.9750434105132371,
      "grad_norm": 0.1381348967552185,
      "learning_rate": 4.367043847241867e-05,
      "loss": 2.140352249145508,
      "step": 9574
    },
    {
      "epoch": 0.9752470962058447,
      "grad_norm": 0.12470867484807968,
      "learning_rate": 4.3316831683168315e-05,
      "loss": 2.1918561458587646,
      "step": 9576
    },
    {
      "epoch": 0.9754507818984525,
      "grad_norm": 0.11691895872354507,
      "learning_rate": 4.2963224893917964e-05,
      "loss": 2.224324941635132,
      "step": 9578
    },
    {
      "epoch": 0.9756544675910602,
      "grad_norm": 0.11534804850816727,
      "learning_rate": 4.260961810466761e-05,
      "loss": 2.2702388763427734,
      "step": 9580
    },
    {
      "epoch": 0.975858153283668,
      "grad_norm": 0.11978413909673691,
      "learning_rate": 4.225601131541726e-05,
      "loss": 2.1585116386413574,
      "step": 9582
    },
    {
      "epoch": 0.9760618389762757,
      "grad_norm": 0.13044314086437225,
      "learning_rate": 4.19024045261669e-05,
      "loss": 2.1497554779052734,
      "step": 9584
    },
    {
      "epoch": 0.9762655246688835,
      "grad_norm": 0.11200161278247833,
      "learning_rate": 4.1548797736916546e-05,
      "loss": 2.141892433166504,
      "step": 9586
    },
    {
      "epoch": 0.9764692103614911,
      "grad_norm": 0.10528165847063065,
      "learning_rate": 4.1195190947666195e-05,
      "loss": 2.023982048034668,
      "step": 9588
    },
    {
      "epoch": 0.9766728960540989,
      "grad_norm": 0.10994535684585571,
      "learning_rate": 4.0841584158415844e-05,
      "loss": 2.1929001808166504,
      "step": 9590
    },
    {
      "epoch": 0.9768765817467067,
      "grad_norm": 0.1258140504360199,
      "learning_rate": 4.048797736916549e-05,
      "loss": 2.3627352714538574,
      "step": 9592
    },
    {
      "epoch": 0.9770802674393144,
      "grad_norm": 0.11771140247583389,
      "learning_rate": 4.0134370579915135e-05,
      "loss": 2.1675357818603516,
      "step": 9594
    },
    {
      "epoch": 0.9772839531319222,
      "grad_norm": 0.14429117739200592,
      "learning_rate": 3.9780763790664784e-05,
      "loss": 2.278122901916504,
      "step": 9596
    },
    {
      "epoch": 0.9774876388245298,
      "grad_norm": 0.1254463493824005,
      "learning_rate": 3.9427157001414426e-05,
      "loss": 2.2560315132141113,
      "step": 9598
    },
    {
      "epoch": 0.9776913245171376,
      "grad_norm": 0.13532914221286774,
      "learning_rate": 3.9073550212164075e-05,
      "loss": 2.2616963386535645,
      "step": 9600
    },
    {
      "epoch": 0.9778950102097453,
      "grad_norm": 0.12094525247812271,
      "learning_rate": 3.871994342291372e-05,
      "loss": 2.2684993743896484,
      "step": 9602
    },
    {
      "epoch": 0.9780986959023531,
      "grad_norm": 0.12271078675985336,
      "learning_rate": 3.8366336633663367e-05,
      "loss": 2.2275843620300293,
      "step": 9604
    },
    {
      "epoch": 0.9783023815949609,
      "grad_norm": 0.13218599557876587,
      "learning_rate": 3.8012729844413016e-05,
      "loss": 2.2804880142211914,
      "step": 9606
    },
    {
      "epoch": 0.9785060672875685,
      "grad_norm": 0.1097869724035263,
      "learning_rate": 3.7659123055162664e-05,
      "loss": 2.296509265899658,
      "step": 9608
    },
    {
      "epoch": 0.9787097529801763,
      "grad_norm": 0.15021127462387085,
      "learning_rate": 3.730551626591231e-05,
      "loss": 2.2287650108337402,
      "step": 9610
    },
    {
      "epoch": 0.978913438672784,
      "grad_norm": 0.123515285551548,
      "learning_rate": 3.695190947666195e-05,
      "loss": 2.101952075958252,
      "step": 9612
    },
    {
      "epoch": 0.9791171243653918,
      "grad_norm": 0.12042821198701859,
      "learning_rate": 3.65983026874116e-05,
      "loss": 2.1473236083984375,
      "step": 9614
    },
    {
      "epoch": 0.9793208100579995,
      "grad_norm": 0.17097489535808563,
      "learning_rate": 3.624469589816125e-05,
      "loss": 2.3172669410705566,
      "step": 9616
    },
    {
      "epoch": 0.9795244957506072,
      "grad_norm": 0.11563076823949814,
      "learning_rate": 3.589108910891089e-05,
      "loss": 2.186427116394043,
      "step": 9618
    },
    {
      "epoch": 0.9797281814432149,
      "grad_norm": 0.13045251369476318,
      "learning_rate": 3.553748231966054e-05,
      "loss": 2.138580560684204,
      "step": 9620
    },
    {
      "epoch": 0.9799318671358227,
      "grad_norm": 0.13359467685222626,
      "learning_rate": 3.518387553041019e-05,
      "loss": 2.2954893112182617,
      "step": 9622
    },
    {
      "epoch": 0.9801355528284305,
      "grad_norm": 0.11476019024848938,
      "learning_rate": 3.4830268741159836e-05,
      "loss": 2.1506845951080322,
      "step": 9624
    },
    {
      "epoch": 0.9803392385210382,
      "grad_norm": 0.1241152435541153,
      "learning_rate": 3.447666195190947e-05,
      "loss": 2.2127017974853516,
      "step": 9626
    },
    {
      "epoch": 0.980542924213646,
      "grad_norm": 0.10535994917154312,
      "learning_rate": 3.412305516265912e-05,
      "loss": 2.223968029022217,
      "step": 9628
    },
    {
      "epoch": 0.9807466099062536,
      "grad_norm": 0.1207202672958374,
      "learning_rate": 3.376944837340877e-05,
      "loss": 2.292062520980835,
      "step": 9630
    },
    {
      "epoch": 0.9809502955988614,
      "grad_norm": 0.11794230341911316,
      "learning_rate": 3.341584158415842e-05,
      "loss": 2.196157932281494,
      "step": 9632
    },
    {
      "epoch": 0.9811539812914691,
      "grad_norm": 0.1267300546169281,
      "learning_rate": 3.306223479490806e-05,
      "loss": 2.173892021179199,
      "step": 9634
    },
    {
      "epoch": 0.9813576669840769,
      "grad_norm": 0.12518678605556488,
      "learning_rate": 3.270862800565771e-05,
      "loss": 2.1752190589904785,
      "step": 9636
    },
    {
      "epoch": 0.9815613526766847,
      "grad_norm": 0.10903537273406982,
      "learning_rate": 3.235502121640736e-05,
      "loss": 2.022575855255127,
      "step": 9638
    },
    {
      "epoch": 0.9817650383692923,
      "grad_norm": 0.1316116452217102,
      "learning_rate": 3.2001414427157e-05,
      "loss": 2.199944496154785,
      "step": 9640
    },
    {
      "epoch": 0.9819687240619001,
      "grad_norm": 0.1285460889339447,
      "learning_rate": 3.164780763790665e-05,
      "loss": 2.2900710105895996,
      "step": 9642
    },
    {
      "epoch": 0.9821724097545078,
      "grad_norm": 0.1365768313407898,
      "learning_rate": 3.129420084865629e-05,
      "loss": 2.2441487312316895,
      "step": 9644
    },
    {
      "epoch": 0.9823760954471156,
      "grad_norm": 0.1222805604338646,
      "learning_rate": 3.094059405940594e-05,
      "loss": 2.1171083450317383,
      "step": 9646
    },
    {
      "epoch": 0.9825797811397233,
      "grad_norm": 0.11548007279634476,
      "learning_rate": 3.058698727015559e-05,
      "loss": 2.0737314224243164,
      "step": 9648
    },
    {
      "epoch": 0.982783466832331,
      "grad_norm": 0.13433538377285004,
      "learning_rate": 3.0233380480905232e-05,
      "loss": 1.9914588928222656,
      "step": 9650
    },
    {
      "epoch": 0.9829871525249387,
      "grad_norm": 0.10751388967037201,
      "learning_rate": 2.987977369165488e-05,
      "loss": 1.9814914464950562,
      "step": 9652
    },
    {
      "epoch": 0.9831908382175465,
      "grad_norm": 0.1207691952586174,
      "learning_rate": 2.9526166902404527e-05,
      "loss": 2.241441488265991,
      "step": 9654
    },
    {
      "epoch": 0.9833945239101543,
      "grad_norm": 0.11693581938743591,
      "learning_rate": 2.9172560113154176e-05,
      "loss": 2.136688232421875,
      "step": 9656
    },
    {
      "epoch": 0.983598209602762,
      "grad_norm": 0.10817644000053406,
      "learning_rate": 2.8818953323903818e-05,
      "loss": 1.9172933101654053,
      "step": 9658
    },
    {
      "epoch": 0.9838018952953698,
      "grad_norm": 0.12429602444171906,
      "learning_rate": 2.8465346534653467e-05,
      "loss": 2.0071120262145996,
      "step": 9660
    },
    {
      "epoch": 0.9840055809879774,
      "grad_norm": 0.13119038939476013,
      "learning_rate": 2.8111739745403113e-05,
      "loss": 2.1469383239746094,
      "step": 9662
    },
    {
      "epoch": 0.9842092666805852,
      "grad_norm": 0.11853819340467453,
      "learning_rate": 2.7758132956152758e-05,
      "loss": 2.060894012451172,
      "step": 9664
    },
    {
      "epoch": 0.9844129523731929,
      "grad_norm": 0.12018376588821411,
      "learning_rate": 2.7404526166902404e-05,
      "loss": 2.202319860458374,
      "step": 9666
    },
    {
      "epoch": 0.9846166380658007,
      "grad_norm": 0.1097240000963211,
      "learning_rate": 2.7050919377652053e-05,
      "loss": 2.1972334384918213,
      "step": 9668
    },
    {
      "epoch": 0.9848203237584084,
      "grad_norm": 0.12636548280715942,
      "learning_rate": 2.6697312588401698e-05,
      "loss": 2.0660274028778076,
      "step": 9670
    },
    {
      "epoch": 0.9850240094510161,
      "grad_norm": 0.16716332733631134,
      "learning_rate": 2.6343705799151344e-05,
      "loss": 2.2095346450805664,
      "step": 9672
    },
    {
      "epoch": 0.9852276951436238,
      "grad_norm": 0.13222898542881012,
      "learning_rate": 2.599009900990099e-05,
      "loss": 2.189622640609741,
      "step": 9674
    },
    {
      "epoch": 0.9854313808362316,
      "grad_norm": 0.10951217263936996,
      "learning_rate": 2.563649222065064e-05,
      "loss": 2.3005852699279785,
      "step": 9676
    },
    {
      "epoch": 0.9856350665288394,
      "grad_norm": 0.13000153005123138,
      "learning_rate": 2.528288543140028e-05,
      "loss": 2.250972270965576,
      "step": 9678
    },
    {
      "epoch": 0.9858387522214471,
      "grad_norm": 0.12066211551427841,
      "learning_rate": 2.492927864214993e-05,
      "loss": 2.2426559925079346,
      "step": 9680
    },
    {
      "epoch": 0.9860424379140548,
      "grad_norm": 0.10310465097427368,
      "learning_rate": 2.4575671852899575e-05,
      "loss": 2.101156711578369,
      "step": 9682
    },
    {
      "epoch": 0.9862461236066625,
      "grad_norm": 0.10478894412517548,
      "learning_rate": 2.4222065063649224e-05,
      "loss": 2.2042813301086426,
      "step": 9684
    },
    {
      "epoch": 0.9864498092992703,
      "grad_norm": 0.10848657041788101,
      "learning_rate": 2.3868458274398866e-05,
      "loss": 2.2650132179260254,
      "step": 9686
    },
    {
      "epoch": 0.986653494991878,
      "grad_norm": 0.11941387504339218,
      "learning_rate": 2.3514851485148515e-05,
      "loss": 2.181137800216675,
      "step": 9688
    },
    {
      "epoch": 0.9868571806844858,
      "grad_norm": 0.12692472338676453,
      "learning_rate": 2.316124469589816e-05,
      "loss": 2.191352367401123,
      "step": 9690
    },
    {
      "epoch": 0.9870608663770936,
      "grad_norm": 0.11461591720581055,
      "learning_rate": 2.2807637906647807e-05,
      "loss": 2.130669593811035,
      "step": 9692
    },
    {
      "epoch": 0.9872645520697012,
      "grad_norm": 0.11235085129737854,
      "learning_rate": 2.2454031117397456e-05,
      "loss": 2.194911479949951,
      "step": 9694
    },
    {
      "epoch": 0.987468237762309,
      "grad_norm": 0.11506561189889908,
      "learning_rate": 2.21004243281471e-05,
      "loss": 2.2360763549804688,
      "step": 9696
    },
    {
      "epoch": 0.9876719234549167,
      "grad_norm": 0.12685999274253845,
      "learning_rate": 2.174681753889675e-05,
      "loss": 2.3418896198272705,
      "step": 9698
    },
    {
      "epoch": 0.9878756091475245,
      "grad_norm": 0.11136867105960846,
      "learning_rate": 2.1393210749646392e-05,
      "loss": 2.1279220581054688,
      "step": 9700
    },
    {
      "epoch": 0.9880792948401322,
      "grad_norm": 0.11498718708753586,
      "learning_rate": 2.103960396039604e-05,
      "loss": 2.2023324966430664,
      "step": 9702
    },
    {
      "epoch": 0.9882829805327399,
      "grad_norm": 0.12807564437389374,
      "learning_rate": 2.0685997171145687e-05,
      "loss": 2.2242074012756348,
      "step": 9704
    },
    {
      "epoch": 0.9884866662253476,
      "grad_norm": 0.10460840910673141,
      "learning_rate": 2.0332390381895333e-05,
      "loss": 2.100865364074707,
      "step": 9706
    },
    {
      "epoch": 0.9886903519179554,
      "grad_norm": 0.10067147761583328,
      "learning_rate": 1.9978783592644978e-05,
      "loss": 2.1886072158813477,
      "step": 9708
    },
    {
      "epoch": 0.9888940376105632,
      "grad_norm": 0.12140684574842453,
      "learning_rate": 1.9625176803394627e-05,
      "loss": 2.2306129932403564,
      "step": 9710
    },
    {
      "epoch": 0.9890977233031709,
      "grad_norm": 0.11440601944923401,
      "learning_rate": 1.9271570014144273e-05,
      "loss": 2.1204004287719727,
      "step": 9712
    },
    {
      "epoch": 0.9893014089957786,
      "grad_norm": 0.11610212177038193,
      "learning_rate": 1.891796322489392e-05,
      "loss": 2.0451507568359375,
      "step": 9714
    },
    {
      "epoch": 0.9895050946883863,
      "grad_norm": 0.13186109066009521,
      "learning_rate": 1.8564356435643564e-05,
      "loss": 2.293870449066162,
      "step": 9716
    },
    {
      "epoch": 0.9897087803809941,
      "grad_norm": 0.1389576494693756,
      "learning_rate": 1.8210749646393213e-05,
      "loss": 2.256408214569092,
      "step": 9718
    },
    {
      "epoch": 0.9899124660736018,
      "grad_norm": 0.12352164089679718,
      "learning_rate": 1.7857142857142855e-05,
      "loss": 2.203220844268799,
      "step": 9720
    },
    {
      "epoch": 0.9901161517662096,
      "grad_norm": 0.12324763834476471,
      "learning_rate": 1.7503536067892504e-05,
      "loss": 2.2785794734954834,
      "step": 9722
    },
    {
      "epoch": 0.9903198374588174,
      "grad_norm": 0.11663363873958588,
      "learning_rate": 1.714992927864215e-05,
      "loss": 2.178011894226074,
      "step": 9724
    },
    {
      "epoch": 0.990523523151425,
      "grad_norm": 0.1087685078382492,
      "learning_rate": 1.67963224893918e-05,
      "loss": 2.0656275749206543,
      "step": 9726
    },
    {
      "epoch": 0.9907272088440328,
      "grad_norm": 0.13544976711273193,
      "learning_rate": 1.644271570014144e-05,
      "loss": 2.284142017364502,
      "step": 9728
    },
    {
      "epoch": 0.9909308945366405,
      "grad_norm": 0.10650062561035156,
      "learning_rate": 1.608910891089109e-05,
      "loss": 2.177692174911499,
      "step": 9730
    },
    {
      "epoch": 0.9911345802292483,
      "grad_norm": 0.14362026751041412,
      "learning_rate": 1.5735502121640735e-05,
      "loss": 2.258323907852173,
      "step": 9732
    },
    {
      "epoch": 0.991338265921856,
      "grad_norm": 0.1177920401096344,
      "learning_rate": 1.538189533239038e-05,
      "loss": 2.2031025886535645,
      "step": 9734
    },
    {
      "epoch": 0.9915419516144637,
      "grad_norm": 0.14983351528644562,
      "learning_rate": 1.5028288543140028e-05,
      "loss": 2.118962287902832,
      "step": 9736
    },
    {
      "epoch": 0.9917456373070714,
      "grad_norm": 0.09992138296365738,
      "learning_rate": 1.4674681753889676e-05,
      "loss": 2.0944111347198486,
      "step": 9738
    },
    {
      "epoch": 0.9919493229996792,
      "grad_norm": 0.12135495990514755,
      "learning_rate": 1.4321074964639321e-05,
      "loss": 2.240431785583496,
      "step": 9740
    },
    {
      "epoch": 0.992153008692287,
      "grad_norm": 0.11646091192960739,
      "learning_rate": 1.3967468175388968e-05,
      "loss": 2.2257771492004395,
      "step": 9742
    },
    {
      "epoch": 0.9923566943848947,
      "grad_norm": 0.12933209538459778,
      "learning_rate": 1.3613861386138614e-05,
      "loss": 2.1571578979492188,
      "step": 9744
    },
    {
      "epoch": 0.9925603800775024,
      "grad_norm": 0.13703583180904388,
      "learning_rate": 1.326025459688826e-05,
      "loss": 2.2181191444396973,
      "step": 9746
    },
    {
      "epoch": 0.9927640657701101,
      "grad_norm": 0.16202270984649658,
      "learning_rate": 1.2906647807637907e-05,
      "loss": 2.334251880645752,
      "step": 9748
    },
    {
      "epoch": 0.9929677514627179,
      "grad_norm": 0.11567133665084839,
      "learning_rate": 1.2553041018387553e-05,
      "loss": 2.182380199432373,
      "step": 9750
    },
    {
      "epoch": 0.9931714371553256,
      "grad_norm": 0.09885074943304062,
      "learning_rate": 1.21994342291372e-05,
      "loss": 2.1116645336151123,
      "step": 9752
    },
    {
      "epoch": 0.9933751228479334,
      "grad_norm": 0.10741697996854782,
      "learning_rate": 1.1845827439886845e-05,
      "loss": 2.24072265625,
      "step": 9754
    },
    {
      "epoch": 0.993578808540541,
      "grad_norm": 0.12508566677570343,
      "learning_rate": 1.1492220650636493e-05,
      "loss": 2.314736843109131,
      "step": 9756
    },
    {
      "epoch": 0.9937824942331488,
      "grad_norm": 0.12866725027561188,
      "learning_rate": 1.1138613861386138e-05,
      "loss": 2.3079721927642822,
      "step": 9758
    },
    {
      "epoch": 0.9939861799257566,
      "grad_norm": 0.13441456854343414,
      "learning_rate": 1.0785007072135784e-05,
      "loss": 2.2090160846710205,
      "step": 9760
    },
    {
      "epoch": 0.9941898656183643,
      "grad_norm": 0.1220303326845169,
      "learning_rate": 1.0431400282885431e-05,
      "loss": 2.3190228939056396,
      "step": 9762
    },
    {
      "epoch": 0.9943935513109721,
      "grad_norm": 0.14049310982227325,
      "learning_rate": 1.0077793493635077e-05,
      "loss": 2.1879591941833496,
      "step": 9764
    },
    {
      "epoch": 0.9945972370035798,
      "grad_norm": 0.11104830354452133,
      "learning_rate": 9.724186704384724e-06,
      "loss": 2.1529951095581055,
      "step": 9766
    },
    {
      "epoch": 0.9948009226961875,
      "grad_norm": 0.10130801051855087,
      "learning_rate": 9.37057991513437e-06,
      "loss": 2.223810911178589,
      "step": 9768
    },
    {
      "epoch": 0.9950046083887952,
      "grad_norm": 0.10204557329416275,
      "learning_rate": 9.016973125884019e-06,
      "loss": 2.1392641067504883,
      "step": 9770
    },
    {
      "epoch": 0.995208294081403,
      "grad_norm": 0.10794952511787415,
      "learning_rate": 8.663366336633664e-06,
      "loss": 2.192312717437744,
      "step": 9772
    },
    {
      "epoch": 0.9954119797740107,
      "grad_norm": 0.10588128864765167,
      "learning_rate": 8.30975954738331e-06,
      "loss": 2.3503150939941406,
      "step": 9774
    },
    {
      "epoch": 0.9956156654666185,
      "grad_norm": 0.11458659917116165,
      "learning_rate": 7.956152758132957e-06,
      "loss": 2.1300034523010254,
      "step": 9776
    },
    {
      "epoch": 0.9958193511592262,
      "grad_norm": 0.10543055832386017,
      "learning_rate": 7.602545968882603e-06,
      "loss": 1.9876337051391602,
      "step": 9778
    },
    {
      "epoch": 0.9960230368518339,
      "grad_norm": 0.12369866669178009,
      "learning_rate": 7.248939179632248e-06,
      "loss": 2.253140926361084,
      "step": 9780
    },
    {
      "epoch": 0.9962267225444417,
      "grad_norm": 0.12806814908981323,
      "learning_rate": 6.895332390381896e-06,
      "loss": 2.0085337162017822,
      "step": 9782
    },
    {
      "epoch": 0.9964304082370494,
      "grad_norm": 0.11297745257616043,
      "learning_rate": 6.541725601131542e-06,
      "loss": 2.2648372650146484,
      "step": 9784
    },
    {
      "epoch": 0.9966340939296572,
      "grad_norm": 0.11328990012407303,
      "learning_rate": 6.1881188118811885e-06,
      "loss": 2.054495096206665,
      "step": 9786
    },
    {
      "epoch": 0.9968377796222648,
      "grad_norm": 0.139146089553833,
      "learning_rate": 5.834512022630835e-06,
      "loss": 2.3090758323669434,
      "step": 9788
    },
    {
      "epoch": 0.9970414653148726,
      "grad_norm": 0.12158659845590591,
      "learning_rate": 5.480905233380481e-06,
      "loss": 2.1137328147888184,
      "step": 9790
    },
    {
      "epoch": 0.9972451510074803,
      "grad_norm": 0.12075679749250412,
      "learning_rate": 5.127298444130128e-06,
      "loss": 2.157978057861328,
      "step": 9792
    },
    {
      "epoch": 0.9974488367000881,
      "grad_norm": 0.11339784413576126,
      "learning_rate": 4.7736916548797735e-06,
      "loss": 2.195131301879883,
      "step": 9794
    },
    {
      "epoch": 0.9976525223926959,
      "grad_norm": 0.13893643021583557,
      "learning_rate": 4.42008486562942e-06,
      "loss": 2.171344518661499,
      "step": 9796
    },
    {
      "epoch": 0.9978562080853036,
      "grad_norm": 0.12680979073047638,
      "learning_rate": 4.066478076379066e-06,
      "loss": 2.1075198650360107,
      "step": 9798
    },
    {
      "epoch": 0.9980598937779113,
      "grad_norm": 0.14741966128349304,
      "learning_rate": 3.7128712871287128e-06,
      "loss": 2.3135738372802734,
      "step": 9800
    },
    {
      "epoch": 0.998263579470519,
      "grad_norm": 0.13972274959087372,
      "learning_rate": 3.3592644978783596e-06,
      "loss": 2.318100929260254,
      "step": 9802
    },
    {
      "epoch": 0.9984672651631268,
      "grad_norm": 0.1303679496049881,
      "learning_rate": 3.0056577086280057e-06,
      "loss": 2.198122978210449,
      "step": 9804
    },
    {
      "epoch": 0.9986709508557345,
      "grad_norm": 0.14517098665237427,
      "learning_rate": 2.652050919377652e-06,
      "loss": 2.1365013122558594,
      "step": 9806
    },
    {
      "epoch": 0.9988746365483423,
      "grad_norm": 0.10320740938186646,
      "learning_rate": 2.2984441301272985e-06,
      "loss": 2.350672483444214,
      "step": 9808
    },
    {
      "epoch": 0.99907832224095,
      "grad_norm": 0.1261180341243744,
      "learning_rate": 1.944837340876945e-06,
      "loss": 2.136641263961792,
      "step": 9810
    },
    {
      "epoch": 0.9992820079335577,
      "grad_norm": 0.1099807396531105,
      "learning_rate": 1.5912305516265914e-06,
      "loss": 2.222689151763916,
      "step": 9812
    },
    {
      "epoch": 0.9994856936261655,
      "grad_norm": 0.11894405633211136,
      "learning_rate": 1.2376237623762377e-06,
      "loss": 2.220721960067749,
      "step": 9814
    },
    {
      "epoch": 0.9996893793187732,
      "grad_norm": 0.11139527708292007,
      "learning_rate": 8.84016973125884e-07,
      "loss": 2.096712589263916,
      "step": 9816
    },
    {
      "epoch": 0.999893065011381,
      "grad_norm": 0.10845275968313217,
      "learning_rate": 5.304101838755304e-07,
      "loss": 2.1310038566589355,
      "step": 9818
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.45586344599723816,
      "learning_rate": 1.768033946251768e-07,
      "loss": 2.367912530899048,
      "step": 9820
    }
  ],
  "logging_steps": 2,
  "max_steps": 9820,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 982,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.0263528708500357e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}