{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "loss_align": 0.9942637085914612,
      "loss_contrastive": 0.07272572815418243,
      "loss_gen": 3.6496880054473877,
      "neg_sim_mean": -0.021537963300943375,
      "pos_sim_mean": 0.005736307241022587,
      "step": 0
    },
    {
      "epoch": 0.0002,
      "grad_norm": 21.918048858642578,
      "learning_rate": 0.0,
      "loss": 4.6527,
      "step": 1
    },
    {
      "epoch": 0.0002,
      "loss_align": 0.9702985882759094,
      "loss_contrastive": 0.02239394560456276,
      "loss_gen": 3.8620786666870117,
      "neg_sim_mean": -0.04790465906262398,
      "pos_sim_mean": 0.02970139868557453,
      "step": 1
    },
    {
      "epoch": 0.0004,
      "grad_norm": 29.39208221435547,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 4.8351,
      "step": 2
    },
    {
      "epoch": 0.0004,
      "loss_align": 0.9753555059432983,
      "loss_contrastive": 0.09612817317247391,
      "loss_gen": 3.9898574352264404,
      "neg_sim_mean": 0.020772660151124,
      "pos_sim_mean": 0.024644482880830765,
      "step": 2
    },
    {
      "epoch": 0.0006,
      "grad_norm": 26.888919830322266,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 4.9767,
      "step": 3
    },
    {
      "epoch": 0.0006,
      "loss_align": 0.9988664984703064,
      "loss_contrastive": 0.06666826456785202,
      "loss_gen": 2.931046962738037,
      "neg_sim_mean": -0.032198235392570496,
      "pos_sim_mean": 0.001133500598371029,
      "step": 3
    },
    {
      "epoch": 0.0008,
      "grad_norm": 17.074249267578125,
      "learning_rate": 6.000000000000001e-07,
      "loss": 3.9379,
      "step": 4
    },
    {
      "epoch": 0.0008,
      "loss_align": 1.007588505744934,
      "loss_contrastive": 0.11202633380889893,
      "loss_gen": 3.1525700092315674,
      "neg_sim_mean": 0.00443777721375227,
      "pos_sim_mean": -0.007588559295982122,
      "step": 4
    },
    {
      "epoch": 0.001,
      "grad_norm": 20.29343032836914,
      "learning_rate": 8.000000000000001e-07,
      "loss": 4.1736,
      "step": 5
    },
    {
      "epoch": 0.001,
      "loss_align": 1.0128737688064575,
      "loss_contrastive": 0.11190008372068405,
      "loss_gen": 3.9644124507904053,
      "neg_sim_mean": -0.0009736744686961174,
      "pos_sim_mean": -0.012873755767941475,
      "step": 5
    },
    {
      "epoch": 0.0012,
      "grad_norm": 22.427831649780273,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 4.9907,
      "step": 6
    },
    {
      "epoch": 0.0012,
      "loss_align": 1.0087602138519287,
      "loss_contrastive": 0.09481802582740784,
      "loss_gen": 2.773501396179199,
      "neg_sim_mean": -0.013942252844572067,
      "pos_sim_mean": -0.008760272525250912,
      "step": 6
    },
    {
      "epoch": 0.0014,
      "grad_norm": 15.676722526550293,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 3.7936,
      "step": 7
    },
    {
      "epoch": 0.0014,
      "loss_align": 1.020876407623291,
      "loss_contrastive": 0.12078294157981873,
      "loss_gen": 3.330004930496216,
      "neg_sim_mean": -9.347125887870789e-05,
      "pos_sim_mean": -0.02087641879916191,
      "step": 7
    },
    {
      "epoch": 0.0016,
      "grad_norm": 24.25899314880371,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 4.3654,
      "step": 8
    },
    {
      "epoch": 0.0016,
      "loss_align": 0.9810000658035278,
      "loss_contrastive": 0.10565060377120972,
      "loss_gen": 4.310175895690918,
      "neg_sim_mean": 0.024650512263178825,
      "pos_sim_mean": 0.018999911844730377,
      "step": 8
    },
    {
      "epoch": 0.0018,
      "grad_norm": 27.559354782104492,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 5.3039,
      "step": 9
    },
    {
      "epoch": 0.0018,
      "loss_align": 1.0098626613616943,
      "loss_contrastive": 0.11169713735580444,
      "loss_gen": 3.6010777950286865,
      "neg_sim_mean": 0.0018344121053814888,
      "pos_sim_mean": -0.009862720966339111,
      "step": 9
    },
    {
      "epoch": 0.002,
      "grad_norm": 25.55319595336914,
      "learning_rate": 1.8e-06,
      "loss": 4.6243,
      "step": 10
    },
    {
      "epoch": 0.002,
      "loss_align": 0.9795165657997131,
      "loss_contrastive": 0.04436018317937851,
      "loss_gen": 4.809248924255371,
      "neg_sim_mean": -0.03515639901161194,
      "pos_sim_mean": 0.020483417436480522,
      "step": 10
    },
    {
      "epoch": 0.0022,
      "grad_norm": 31.905231475830078,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 5.7941,
      "step": 11
    },
    {
      "epoch": 0.0022,
      "loss_align": 0.9736458659172058,
      "loss_contrastive": 0.09268258512020111,
      "loss_gen": 3.7080647945404053,
      "neg_sim_mean": 0.019036706537008286,
      "pos_sim_mean": 0.02635413035750389,
      "step": 11
    },
    {
      "epoch": 0.0024,
      "grad_norm": 19.600208282470703,
      "learning_rate": 2.2e-06,
      "loss": 4.6928,
      "step": 12
    },
    {
      "epoch": 0.0024,
      "loss_align": 0.957268476486206,
      "loss_contrastive": 0.10046496987342834,
      "loss_gen": 3.4305624961853027,
      "neg_sim_mean": 0.04319649562239647,
      "pos_sim_mean": 0.04273153096437454,
      "step": 12
    },
    {
      "epoch": 0.0026,
      "grad_norm": 20.176837921142578,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 4.3999,
      "step": 13
    },
    {
      "epoch": 0.0026,
      "loss_align": 0.9927799105644226,
      "loss_contrastive": 0.0679057165980339,
      "loss_gen": 4.5076584815979,
      "neg_sim_mean": -0.024874180555343628,
      "pos_sim_mean": 0.007220106199383736,
      "step": 13
    },
    {
      "epoch": 0.0028,
      "grad_norm": 33.02507019042969,
      "learning_rate": 2.6e-06,
      "loss": 5.5086,
      "step": 14
    },
    {
      "epoch": 0.0028,
      "loss_align": 0.9459438323974609,
      "loss_contrastive": 0.0871078297495842,
      "loss_gen": 3.2837975025177,
      "neg_sim_mean": 0.041164010763168335,
      "pos_sim_mean": 0.054056182503700256,
      "step": 14
    },
    {
      "epoch": 0.003,
      "grad_norm": 17.62779998779297,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 4.2402,
      "step": 15
    },
    {
      "epoch": 0.003,
      "loss_align": 0.9445565938949585,
      "loss_contrastive": 0.09578719735145569,
      "loss_gen": 3.5344018936157227,
      "neg_sim_mean": 0.05123058706521988,
      "pos_sim_mean": 0.05544339120388031,
      "step": 15
    },
    {
      "epoch": 0.0032,
      "grad_norm": 22.368989944458008,
      "learning_rate": 3e-06,
      "loss": 4.4905,
      "step": 16
    },
    {
      "epoch": 0.0032,
      "loss_align": 0.9622821807861328,
      "loss_contrastive": 0.10560566186904907,
      "loss_gen": 2.9163591861724854,
      "neg_sim_mean": 0.04332345724105835,
      "pos_sim_mean": 0.0377177931368351,
      "step": 16
    },
    {
      "epoch": 0.0034,
      "grad_norm": 16.32938003540039,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 3.8913,
      "step": 17
    },
    {
      "epoch": 0.0034,
      "loss_align": 0.9792921543121338,
      "loss_contrastive": 0.08709508925676346,
      "loss_gen": 3.6251840591430664,
      "neg_sim_mean": 0.007802926003932953,
      "pos_sim_mean": 0.020707841962575912,
      "step": 17
    },
    {
      "epoch": 0.0036,
      "grad_norm": 20.001964569091797,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 4.6149,
      "step": 18
    },
    {
      "epoch": 0.0036,
      "loss_align": 0.9524857401847839,
      "loss_contrastive": 0.056059330701828,
      "loss_gen": 3.60418963432312,
      "neg_sim_mean": 0.003573586232960224,
      "pos_sim_mean": 0.047514256089925766,
      "step": 18
    },
    {
      "epoch": 0.0038,
      "grad_norm": 18.6129207611084,
      "learning_rate": 3.6e-06,
      "loss": 4.5634,
      "step": 19
    },
    {
      "epoch": 0.0038,
      "loss_align": 0.9411982893943787,
      "loss_contrastive": 0.09804661571979523,
      "loss_gen": 3.5008902549743652,
      "neg_sim_mean": 0.056848328560590744,
      "pos_sim_mean": 0.058801718056201935,
      "step": 19
    },
    {
      "epoch": 0.004,
      "grad_norm": 20.449588775634766,
      "learning_rate": 3.8e-06,
      "loss": 4.4539,
      "step": 20
    },
    {
      "epoch": 0.004,
      "loss_align": 0.9662308692932129,
      "loss_contrastive": 0.14629852771759033,
      "loss_gen": 3.590441942214966,
      "neg_sim_mean": 0.08006766438484192,
      "pos_sim_mean": 0.03376913443207741,
      "step": 20
    },
    {
      "epoch": 0.0042,
      "grad_norm": 22.47841453552246,
      "learning_rate": 4.000000000000001e-06,
      "loss": 4.5742,
      "step": 21
    },
    {
      "epoch": 0.0042,
      "loss_align": 0.9371734261512756,
      "loss_contrastive": 0.043859224766492844,
      "loss_gen": 4.544889450073242,
      "neg_sim_mean": 0.00668581947684288,
      "pos_sim_mean": 0.06282659620046616,
      "step": 21
    },
    {
      "epoch": 0.0044,
      "grad_norm": 24.942380905151367,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 5.4873,
      "step": 22
    },
    {
      "epoch": 0.0044,
      "loss_align": 0.9201489090919495,
      "loss_contrastive": 0.03178812563419342,
      "loss_gen": 5.774998664855957,
      "neg_sim_mean": 0.011639206670224667,
      "pos_sim_mean": 0.07985108345746994,
      "step": 22
    },
    {
      "epoch": 0.0046,
      "grad_norm": 34.17338562011719,
      "learning_rate": 4.4e-06,
      "loss": 6.699,
      "step": 23
    },
    {
      "epoch": 0.0046,
      "loss_align": 0.9444453120231628,
      "loss_contrastive": 0.09761178493499756,
      "loss_gen": 3.171567916870117,
      "neg_sim_mean": 0.05316648259758949,
      "pos_sim_mean": 0.05555470287799835,
      "step": 23
    },
    {
      "epoch": 0.0048,
      "grad_norm": 16.437185287475586,
      "learning_rate": 4.6e-06,
      "loss": 4.1277,
      "step": 24
    },
    {
      "epoch": 0.0048,
      "loss_align": 0.9159988164901733,
      "loss_contrastive": 0.10857212543487549,
      "loss_gen": 3.7278759479522705,
      "neg_sim_mean": 0.09257330000400543,
      "pos_sim_mean": 0.08400117605924606,
      "step": 24
    },
    {
      "epoch": 0.005,
      "grad_norm": 15.854809761047363,
      "learning_rate": 4.800000000000001e-06,
      "loss": 4.6569,
      "step": 25
    },
    {
      "epoch": 0.005,
      "loss_align": 0.8865982294082642,
      "loss_contrastive": 0.09812586009502411,
      "loss_gen": 3.203885793685913,
      "neg_sim_mean": 0.11152764409780502,
      "pos_sim_mean": 0.11340178549289703,
      "step": 25
    },
    {
      "epoch": 0.0052,
      "grad_norm": 14.885098457336426,
      "learning_rate": 5e-06,
      "loss": 4.1023,
      "step": 26
    },
    {
      "epoch": 0.0052,
      "loss_align": 0.9219717979431152,
      "loss_contrastive": 0.1072048619389534,
      "loss_gen": 3.093815326690674,
      "neg_sim_mean": 0.08523304015398026,
      "pos_sim_mean": 0.07802817970514297,
      "step": 26
    },
    {
      "epoch": 0.0054,
      "grad_norm": 15.76525592803955,
      "learning_rate": 5.2e-06,
      "loss": 4.0287,
      "step": 27
    },
    {
      "epoch": 0.0054,
      "loss_align": 0.9056798219680786,
      "loss_contrastive": 0.04480614513158798,
      "loss_gen": 4.641787528991699,
      "neg_sim_mean": 0.03912629187107086,
      "pos_sim_mean": 0.094320148229599,
      "step": 27
    },
    {
      "epoch": 0.0056,
      "grad_norm": 29.079565048217773,
      "learning_rate": 5.4e-06,
      "loss": 5.5528,
      "step": 28
    },
    {
      "epoch": 0.0056,
      "loss_align": 0.9496100544929504,
      "loss_contrastive": 0.15820033848285675,
      "loss_gen": 3.668917655944824,
      "neg_sim_mean": 0.1085902750492096,
      "pos_sim_mean": 0.050389934331178665,
      "step": 28
    },
    {
      "epoch": 0.0058,
      "grad_norm": 17.85655975341797,
      "learning_rate": 5.600000000000001e-06,
      "loss": 4.6375,
      "step": 29
    },
    {
      "epoch": 0.0058,
      "loss_align": 0.9217426180839539,
      "loss_contrastive": 0.12320903688669205,
      "loss_gen": 3.4917659759521484,
      "neg_sim_mean": 0.10146643221378326,
      "pos_sim_mean": 0.07825739681720734,
      "step": 29
    },
    {
      "epoch": 0.006,
      "grad_norm": 15.457758903503418,
      "learning_rate": 5.8e-06,
      "loss": 4.4283,
      "step": 30
    },
    {
      "epoch": 0.006,
      "loss_align": 0.8883821964263916,
      "loss_contrastive": 0.09343116730451584,
      "loss_gen": 3.071113109588623,
      "neg_sim_mean": 0.10504893958568573,
      "pos_sim_mean": 0.11161777377128601,
      "step": 30
    },
    {
      "epoch": 0.0062,
      "grad_norm": 14.50932788848877,
      "learning_rate": 6e-06,
      "loss": 3.9707,
      "step": 31
    },
    {
      "epoch": 0.0062,
      "loss_align": 0.95461106300354,
      "loss_contrastive": 0.13441772758960724,
      "loss_gen": 4.001429557800293,
      "neg_sim_mean": 0.07980667054653168,
      "pos_sim_mean": 0.04538894072175026,
      "step": 31
    },
    {
      "epoch": 0.0064,
      "grad_norm": 19.05087661743164,
      "learning_rate": 6.2e-06,
      "loss": 4.9722,
      "step": 32
    },
    {
      "epoch": 0.0064,
      "loss_align": 0.826172411441803,
      "loss_contrastive": 0.05838645249605179,
      "loss_gen": 3.4629833698272705,
      "neg_sim_mean": 0.1322140395641327,
      "pos_sim_mean": 0.17382758855819702,
      "step": 32
    },
    {
      "epoch": 0.0066,
      "grad_norm": 15.777167320251465,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 4.2962,
      "step": 33
    },
    {
      "epoch": 0.0066,
      "loss_align": 0.9190036654472351,
      "loss_contrastive": 0.13078434765338898,
      "loss_gen": 3.4302785396575928,
      "neg_sim_mean": 0.11178069561719894,
      "pos_sim_mean": 0.08099634945392609,
      "step": 33
    },
    {
      "epoch": 0.0068,
      "grad_norm": 15.0590238571167,
      "learning_rate": 6.6e-06,
      "loss": 4.365,
      "step": 34
    },
    {
      "epoch": 0.0068,
      "loss_align": 0.9050677418708801,
      "loss_contrastive": 0.1397014558315277,
      "loss_gen": 3.696084499359131,
      "neg_sim_mean": 0.13463369011878967,
      "pos_sim_mean": 0.09493224322795868,
      "step": 34
    },
    {
      "epoch": 0.007,
      "grad_norm": 16.979101181030273,
      "learning_rate": 6.800000000000001e-06,
      "loss": 4.6179,
      "step": 35
    },
    {
      "epoch": 0.007,
      "loss_align": 0.8336895704269409,
      "loss_contrastive": 0.12123862653970718,
      "loss_gen": 2.495558023452759,
      "neg_sim_mean": 0.18754902482032776,
      "pos_sim_mean": 0.1663103997707367,
      "step": 35
    },
    {
      "epoch": 0.0072,
      "grad_norm": 12.087753295898438,
      "learning_rate": 7.000000000000001e-06,
      "loss": 3.3438,
      "step": 36
    },
    {
      "epoch": 0.0072,
      "loss_align": 0.8218684196472168,
      "loss_contrastive": 0.08065801113843918,
      "loss_gen": 2.7867674827575684,
      "neg_sim_mean": 0.15878956019878387,
      "pos_sim_mean": 0.17813155055046082,
      "step": 36
    },
    {
      "epoch": 0.0074,
      "grad_norm": 11.986642837524414,
      "learning_rate": 7.2e-06,
      "loss": 3.6183,
      "step": 37
    },
    {
      "epoch": 0.0074,
      "loss_align": 0.8215869665145874,
      "loss_contrastive": 0.10193813592195511,
      "loss_gen": 2.964951992034912,
      "neg_sim_mean": 0.18035119771957397,
      "pos_sim_mean": 0.17841306328773499,
      "step": 37
    },
    {
      "epoch": 0.0076,
      "grad_norm": 15.644676208496094,
      "learning_rate": 7.4e-06,
      "loss": 3.7988,
      "step": 38
    },
    {
      "epoch": 0.0076,
      "loss_align": 0.8186773061752319,
      "loss_contrastive": 0.047773249447345734,
      "loss_gen": 2.6558144092559814,
      "neg_sim_mean": 0.12909594178199768,
      "pos_sim_mean": 0.18132269382476807,
      "step": 38
    },
    {
      "epoch": 0.0078,
      "grad_norm": 12.507756233215332,
      "learning_rate": 7.6e-06,
      "loss": 3.4802,
      "step": 39
    },
    {
      "epoch": 0.0078,
      "loss_align": 0.8496583104133606,
      "loss_contrastive": 0.05744795501232147,
      "loss_gen": 2.832801342010498,
      "neg_sim_mean": 0.10778964310884476,
      "pos_sim_mean": 0.1503416895866394,
      "step": 39
    },
    {
      "epoch": 0.008,
      "grad_norm": 12.709632873535156,
      "learning_rate": 7.8e-06,
      "loss": 3.6894,
      "step": 40
    },
    {
      "epoch": 0.008,
      "loss_align": 0.8296085000038147,
      "loss_contrastive": 0.12013795226812363,
      "loss_gen": 3.493558883666992,
      "neg_sim_mean": 0.19052943587303162,
      "pos_sim_mean": 0.1703914850950241,
      "step": 40
    },
    {
      "epoch": 0.0082,
      "grad_norm": 15.279488563537598,
      "learning_rate": 8.000000000000001e-06,
      "loss": 4.3376,
      "step": 41
    },
    {
      "epoch": 0.0082,
      "loss_align": 0.7448031306266785,
      "loss_contrastive": 0.10309085249900818,
      "loss_gen": 2.3758950233459473,
      "neg_sim_mean": 0.2582877278327942,
      "pos_sim_mean": 0.25519686937332153,
      "step": 41
    },
    {
      "epoch": 0.0084,
      "grad_norm": 16.256059646606445,
      "learning_rate": 8.200000000000001e-06,
      "loss": 3.1331,
      "step": 42
    },
    {
      "epoch": 0.0084,
      "loss_align": 0.7382489442825317,
      "loss_contrastive": 0.0009646564722061157,
      "loss_gen": 2.9738128185272217,
      "neg_sim_mean": 0.16271574795246124,
      "pos_sim_mean": 0.26175108551979065,
      "step": 42
    },
    {
      "epoch": 0.0086,
      "grad_norm": 14.677474021911621,
      "learning_rate": 8.400000000000001e-06,
      "loss": 3.7122,
      "step": 43
    },
    {
      "epoch": 0.0086,
      "loss_align": 0.7861385345458984,
      "loss_contrastive": 0.09791464358568192,
      "loss_gen": 2.438361406326294,
      "neg_sim_mean": 0.21177607774734497,
      "pos_sim_mean": 0.21386143565177917,
      "step": 43
    },
    {
      "epoch": 0.0088,
      "grad_norm": 10.813572883605957,
      "learning_rate": 8.599999999999999e-06,
      "loss": 3.2362,
      "step": 44
    },
    {
      "epoch": 0.0088,
      "loss_align": 0.7317331433296204,
      "loss_contrastive": 0.1316801905632019,
      "loss_gen": 3.192765235900879,
      "neg_sim_mean": 0.299947053194046,
      "pos_sim_mean": 0.26826685667037964,
      "step": 44
    },
    {
      "epoch": 0.009,
      "grad_norm": 13.193706512451172,
      "learning_rate": 8.8e-06,
      "loss": 3.9403,
      "step": 45
    },
    {
      "epoch": 0.009,
      "loss_align": 0.7809799313545227,
      "loss_contrastive": 0.13233759999275208,
      "loss_gen": 2.7396602630615234,
      "neg_sim_mean": 0.25135767459869385,
      "pos_sim_mean": 0.2190200686454773,
      "step": 45
    },
    {
      "epoch": 0.0092,
      "grad_norm": 12.229599952697754,
      "learning_rate": 9e-06,
      "loss": 3.5365,
      "step": 46
    },
    {
      "epoch": 0.0092,
      "loss_align": 0.7956300377845764,
      "loss_contrastive": 0.12198884040117264,
      "loss_gen": 2.540943145751953,
      "neg_sim_mean": 0.2263588011264801,
      "pos_sim_mean": 0.20436996221542358,
      "step": 46
    },
    {
      "epoch": 0.0094,
      "grad_norm": 10.704208374023438,
      "learning_rate": 9.2e-06,
      "loss": 3.3512,
      "step": 47
    },
    {
      "epoch": 0.0094,
      "loss_align": 0.7346162796020508,
      "loss_contrastive": 0.07674460113048553,
      "loss_gen": 2.6254754066467285,
      "neg_sim_mean": 0.24212832748889923,
      "pos_sim_mean": 0.2653837203979492,
      "step": 47
    },
    {
      "epoch": 0.0096,
      "grad_norm": 11.647940635681152,
      "learning_rate": 9.4e-06,
      "loss": 3.3693,
      "step": 48
    },
    {
      "epoch": 0.0096,
      "loss_align": 0.7396198511123657,
      "loss_contrastive": 0.1206829845905304,
      "loss_gen": 2.2302074432373047,
      "neg_sim_mean": 0.28106313943862915,
      "pos_sim_mean": 0.2603801488876343,
      "step": 48
    },
    {
      "epoch": 0.0098,
      "grad_norm": 9.473556518554688,
      "learning_rate": 9.600000000000001e-06,
      "loss": 2.9843,
      "step": 49
    },
    {
      "epoch": 0.0098,
      "loss_align": 0.7412474751472473,
      "loss_contrastive": 0.07920639216899872,
      "loss_gen": 2.639563798904419,
      "neg_sim_mean": 0.23795892298221588,
      "pos_sim_mean": 0.2587525248527527,
      "step": 49
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.229619979858398,
      "learning_rate": 9.800000000000001e-06,
      "loss": 3.3903,
      "step": 50
    },
    {
      "epoch": 0.01,
      "loss_align": 0.6880612969398499,
      "loss_contrastive": 0.13164761662483215,
      "loss_gen": 2.177542209625244,
      "neg_sim_mean": 0.3435863256454468,
      "pos_sim_mean": 0.31193870306015015,
      "step": 50
    },
    {
      "epoch": 0.0102,
      "grad_norm": 9.512763023376465,
      "learning_rate": 1e-05,
      "loss": 2.8814,
      "step": 51
    },
    {
      "epoch": 0.0102,
      "loss_align": 0.6583797335624695,
      "loss_contrastive": 0.11556267738342285,
      "loss_gen": 2.449509620666504,
      "neg_sim_mean": 0.35718294978141785,
      "pos_sim_mean": 0.3416202664375305,
      "step": 51
    },
    {
      "epoch": 0.0104,
      "grad_norm": 10.614006996154785,
      "learning_rate": 1.02e-05,
      "loss": 3.1218,
      "step": 52
    },
    {
      "epoch": 0.0104,
      "loss_align": 0.6712257862091064,
      "loss_contrastive": 0.042567431926727295,
      "loss_gen": 3.004467010498047,
      "neg_sim_mean": 0.2713416516780853,
      "pos_sim_mean": 0.32877421379089355,
      "step": 52
    },
    {
      "epoch": 0.0106,
      "grad_norm": 14.341841697692871,
      "learning_rate": 1.04e-05,
      "loss": 3.6808,
      "step": 53
    },
    {
      "epoch": 0.0106,
      "loss_align": 0.6612468361854553,
      "loss_contrastive": 0.1345336139202118,
      "loss_gen": 2.251572608947754,
      "neg_sim_mean": 0.37328678369522095,
      "pos_sim_mean": 0.3387531638145447,
      "step": 53
    },
    {
      "epoch": 0.0108,
      "grad_norm": 9.561186790466309,
      "learning_rate": 1.06e-05,
      "loss": 2.929,
      "step": 54
    },
    {
      "epoch": 0.0108,
      "loss_align": 0.6905045509338379,
      "loss_contrastive": 0.14946249127388,
      "loss_gen": 2.071143865585327,
      "neg_sim_mean": 0.3589579463005066,
      "pos_sim_mean": 0.3094954490661621,
      "step": 54
    },
    {
      "epoch": 0.011,
      "grad_norm": 9.639748573303223,
      "learning_rate": 1.08e-05,
      "loss": 2.7796,
      "step": 55
    },
    {
      "epoch": 0.011,
      "loss_align": 0.6784470081329346,
      "loss_contrastive": 0.10852450132369995,
      "loss_gen": 2.3387811183929443,
      "neg_sim_mean": 0.33007752895355225,
      "pos_sim_mean": 0.3215530216693878,
      "step": 55
    },
    {
      "epoch": 0.0112,
      "grad_norm": 9.88353157043457,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 3.0303,
      "step": 56
    },
    {
      "epoch": 0.0112,
      "loss_align": 0.6140450835227966,
      "loss_contrastive": 0.02258327603340149,
      "loss_gen": 2.6231532096862793,
      "neg_sim_mean": 0.30853819847106934,
      "pos_sim_mean": 0.38595491647720337,
      "step": 56
    },
    {
      "epoch": 0.0114,
      "grad_norm": 12.226712226867676,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 3.2399,
      "step": 57
    },
    {
      "epoch": 0.0114,
      "loss_align": 0.6351897716522217,
      "loss_contrastive": 0.05079531669616699,
      "loss_gen": 2.964829921722412,
      "neg_sim_mean": 0.3156055212020874,
      "pos_sim_mean": 0.36481019854545593,
      "step": 57
    },
    {
      "epoch": 0.0116,
      "grad_norm": 13.603100776672363,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 3.6061,
      "step": 58
    },
    {
      "epoch": 0.0116,
      "loss_align": 0.7089629769325256,
      "loss_contrastive": 0.07569476962089539,
      "loss_gen": 2.5895583629608154,
      "neg_sim_mean": 0.26673179864883423,
      "pos_sim_mean": 0.29103702306747437,
      "step": 58
    },
    {
      "epoch": 0.0118,
      "grad_norm": 11.643229484558105,
      "learning_rate": 1.16e-05,
      "loss": 3.3076,
      "step": 59
    },
    {
      "epoch": 0.0118,
      "loss_align": 0.5301264524459839,
      "loss_contrastive": 0.0,
      "loss_gen": 2.386326313018799,
      "neg_sim_mean": 0.3325059413909912,
      "pos_sim_mean": 0.4698735475540161,
      "step": 59
    },
    {
      "epoch": 0.012,
      "grad_norm": 10.79203987121582,
      "learning_rate": 1.18e-05,
      "loss": 2.9165,
      "step": 60
    },
    {
      "epoch": 0.012,
      "loss_align": 0.5132398009300232,
      "loss_contrastive": 0.103852778673172,
      "loss_gen": 1.6955246925354004,
      "neg_sim_mean": 0.4906129837036133,
      "pos_sim_mean": 0.4867601990699768,
      "step": 60
    },
    {
      "epoch": 0.0122,
      "grad_norm": 7.665698051452637,
      "learning_rate": 1.2e-05,
      "loss": 2.2212,
      "step": 61
    },
    {
      "epoch": 0.0122,
      "loss_align": 0.6132921576499939,
      "loss_contrastive": 0.09708240628242493,
      "loss_gen": 1.9204792976379395,
      "neg_sim_mean": 0.3837902545928955,
      "pos_sim_mean": 0.3867078423500061,
      "step": 61
    },
    {
      "epoch": 0.0124,
      "grad_norm": 8.695037841796875,
      "learning_rate": 1.22e-05,
      "loss": 2.5454,
      "step": 62
    },
    {
      "epoch": 0.0124,
      "loss_align": 0.5966374278068542,
      "loss_contrastive": 0.16039970517158508,
      "loss_gen": 2.140732765197754,
      "neg_sim_mean": 0.4637622833251953,
      "pos_sim_mean": 0.40336257219314575,
      "step": 62
    },
    {
      "epoch": 0.0126,
      "grad_norm": 9.519328117370605,
      "learning_rate": 1.24e-05,
      "loss": 2.7566,
      "step": 63
    },
    {
      "epoch": 0.0126,
      "loss_align": 0.5495351552963257,
      "loss_contrastive": 0.1285141110420227,
      "loss_gen": 2.046060800552368,
      "neg_sim_mean": 0.4789789319038391,
      "pos_sim_mean": 0.45046481490135193,
      "step": 63
    },
    {
      "epoch": 0.0128,
      "grad_norm": 9.557918548583984,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 2.611,
      "step": 64
    },
    {
      "epoch": 0.0128,
      "loss_align": 0.5378457307815552,
      "loss_contrastive": 0.12752854824066162,
      "loss_gen": 2.1978530883789062,
      "neg_sim_mean": 0.48968279361724854,
      "pos_sim_mean": 0.46215423941612244,
      "step": 64
    },
    {
      "epoch": 0.013,
      "grad_norm": 9.978578567504883,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 2.751,
      "step": 65
    },
    {
      "epoch": 0.013,
      "loss_align": 0.5960726737976074,
      "loss_contrastive": 0.09827882051467896,
      "loss_gen": 1.5204423666000366,
      "neg_sim_mean": 0.4022061228752136,
      "pos_sim_mean": 0.4039272964000702,
      "step": 65
    },
    {
      "epoch": 0.0132,
      "grad_norm": 7.861701011657715,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 2.1283,
      "step": 66
    },
    {
      "epoch": 0.0132,
      "loss_align": 0.537016749382019,
      "loss_contrastive": 0.14187759160995483,
      "loss_gen": 2.261051654815674,
      "neg_sim_mean": 0.5048608183860779,
      "pos_sim_mean": 0.46298322081565857,
      "step": 66
    },
    {
      "epoch": 0.0134,
      "grad_norm": 10.714107513427734,
      "learning_rate": 1.32e-05,
      "loss": 2.8151,
      "step": 67
    },
    {
      "epoch": 0.0134,
      "loss_align": 0.5281546711921692,
      "loss_contrastive": 0.029558807611465454,
      "loss_gen": 2.1331241130828857,
      "neg_sim_mean": 0.40140414237976074,
      "pos_sim_mean": 0.4718453288078308,
      "step": 67
    },
    {
      "epoch": 0.0136,
      "grad_norm": 11.18114948272705,
      "learning_rate": 1.3400000000000002e-05,
      "loss": 2.6648,
      "step": 68
    },
    {
      "epoch": 0.0136,
      "loss_align": 0.5267449617385864,
      "loss_contrastive": 0.14218485355377197,
      "loss_gen": 1.7187196016311646,
      "neg_sim_mean": 0.5154398679733276,
      "pos_sim_mean": 0.4732550084590912,
      "step": 68
    },
    {
      "epoch": 0.0138,
      "grad_norm": 8.394133567810059,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 2.2625,
      "step": 69
    },
    {
      "epoch": 0.0138,
      "loss_align": 0.5136808156967163,
      "loss_contrastive": 0.05205422639846802,
      "loss_gen": 2.1263411045074463,
      "neg_sim_mean": 0.4383734166622162,
      "pos_sim_mean": 0.4863191843032837,
      "step": 69
    },
    {
      "epoch": 0.014,
      "grad_norm": 9.563084602355957,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 2.6463,
      "step": 70
    },
    {
      "epoch": 0.014,
      "loss_align": 0.4873695969581604,
      "loss_contrastive": 0.10363522171974182,
      "loss_gen": 1.849208950996399,
      "neg_sim_mean": 0.5162656307220459,
      "pos_sim_mean": 0.5126304030418396,
      "step": 70
    },
    {
      "epoch": 0.0142,
      "grad_norm": 8.68874740600586,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 2.349,
      "step": 71
    },
    {
      "epoch": 0.0142,
      "loss_align": 0.48186391592025757,
      "loss_contrastive": 0.09741541743278503,
      "loss_gen": 2.034801959991455,
      "neg_sim_mean": 0.5155515074729919,
      "pos_sim_mean": 0.5181360840797424,
      "step": 71
    },
    {
      "epoch": 0.0144,
      "grad_norm": 10.55654239654541,
      "learning_rate": 1.42e-05,
      "loss": 2.5284,
      "step": 72
    },
    {
      "epoch": 0.0144,
      "loss_align": 0.5059541463851929,
      "loss_contrastive": 0.08418717980384827,
      "loss_gen": 1.322789192199707,
      "neg_sim_mean": 0.4782330095767975,
      "pos_sim_mean": 0.49404582381248474,
      "step": 72
    },
    {
      "epoch": 0.0146,
      "grad_norm": 6.7354021072387695,
      "learning_rate": 1.44e-05,
      "loss": 1.8388,
      "step": 73
    },
    {
      "epoch": 0.0146,
      "loss_align": 0.4257744550704956,
      "loss_contrastive": 0.08087888360023499,
      "loss_gen": 1.518596887588501,
      "neg_sim_mean": 0.5551044344902039,
      "pos_sim_mean": 0.5742255449295044,
      "step": 73
    },
    {
      "epoch": 0.0148,
      "grad_norm": 8.009544372558594,
      "learning_rate": 1.4599999999999999e-05,
      "loss": 1.9541,
      "step": 74
    },
    {
      "epoch": 0.0148,
      "loss_align": 0.5321369767189026,
      "loss_contrastive": 0.06969842314720154,
      "loss_gen": 1.5258313417434692,
      "neg_sim_mean": 0.4375614523887634,
      "pos_sim_mean": 0.4678630232810974,
      "step": 74
    },
    {
      "epoch": 0.015,
      "grad_norm": 8.370376586914062,
      "learning_rate": 1.48e-05,
      "loss": 2.0663,
      "step": 75
    },
    {
      "epoch": 0.015,
      "loss_align": 0.4676152467727661,
      "loss_contrastive": 0.10522052645683289,
      "loss_gen": 1.8016488552093506,
      "neg_sim_mean": 0.5376052856445312,
      "pos_sim_mean": 0.5323847532272339,
      "step": 75
    },
    {
      "epoch": 0.0152,
      "grad_norm": 9.52915096282959,
      "learning_rate": 1.5e-05,
      "loss": 2.2819,
      "step": 76
    },
    {
      "epoch": 0.0152,
      "loss_align": 0.4664064645767212,
      "loss_contrastive": 0.09682747721672058,
      "loss_gen": 1.5262576341629028,
      "neg_sim_mean": 0.5304210186004639,
      "pos_sim_mean": 0.5335935354232788,
      "step": 76
    },
    {
      "epoch": 0.0154,
      "grad_norm": 8.199832916259766,
      "learning_rate": 1.52e-05,
      "loss": 2.0043,
      "step": 77
    },
    {
      "epoch": 0.0154,
      "loss_align": 0.5189933776855469,
      "loss_contrastive": 0.06978881359100342,
      "loss_gen": 1.481092929840088,
      "neg_sim_mean": 0.4507954716682434,
      "pos_sim_mean": 0.4810066521167755,
      "step": 77
    },
    {
      "epoch": 0.0156,
      "grad_norm": 8.740116119384766,
      "learning_rate": 1.54e-05,
      "loss": 2.0085,
      "step": 78
    },
    {
      "epoch": 0.0156,
      "loss_align": 0.5054870843887329,
      "loss_contrastive": 0.15398544073104858,
      "loss_gen": 1.331873893737793,
      "neg_sim_mean": 0.5484983921051025,
      "pos_sim_mean": 0.4945129454135895,
      "step": 78
    },
    {
      "epoch": 0.0158,
      "grad_norm": 7.922560214996338,
      "learning_rate": 1.56e-05,
      "loss": 1.8558,
      "step": 79
    },
    {
      "epoch": 0.0158,
      "loss_align": 0.3893936276435852,
      "loss_contrastive": 0.13433891534805298,
      "loss_gen": 1.2726539373397827,
      "neg_sim_mean": 0.6449452638626099,
      "pos_sim_mean": 0.6106063723564148,
      "step": 79
    },
    {
      "epoch": 0.016,
      "grad_norm": 7.739639759063721,
      "learning_rate": 1.58e-05,
      "loss": 1.6782,
      "step": 80
    },
    {
      "epoch": 0.016,
      "loss_align": 0.5253163576126099,
      "loss_contrastive": 0.17719313502311707,
      "loss_gen": 1.7089356184005737,
      "neg_sim_mean": 0.5518767833709717,
      "pos_sim_mean": 0.47468364238739014,
      "step": 80
    },
    {
      "epoch": 0.0162,
      "grad_norm": 9.99183177947998,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.2555,
      "step": 81
    },
    {
      "epoch": 0.0162,
      "loss_align": 0.5663001537322998,
      "loss_contrastive": 0.16061246395111084,
      "loss_gen": 1.5435893535614014,
      "neg_sim_mean": 0.4943122863769531,
      "pos_sim_mean": 0.4336998164653778,
      "step": 81
    },
    {
      "epoch": 0.0164,
      "grad_norm": 9.411971092224121,
      "learning_rate": 1.62e-05,
      "loss": 2.1292,
      "step": 82
    },
    {
      "epoch": 0.0164,
      "loss_align": 0.43331432342529297,
      "loss_contrastive": 0.050407201051712036,
      "loss_gen": 1.2708936929702759,
      "neg_sim_mean": 0.5170928835868835,
      "pos_sim_mean": 0.566685676574707,
      "step": 82
    },
    {
      "epoch": 0.0166,
      "grad_norm": 7.446043968200684,
      "learning_rate": 1.6400000000000002e-05,
      "loss": 1.7103,
      "step": 83
    },
    {
      "epoch": 0.0166,
      "loss_align": 0.46144676208496094,
      "loss_contrastive": 0.19877079129219055,
      "loss_gen": 1.3601137399673462,
      "neg_sim_mean": 0.6373240351676941,
      "pos_sim_mean": 0.5385532379150391,
      "step": 83
    },
    {
      "epoch": 0.0168,
      "grad_norm": 8.93761157989502,
      "learning_rate": 1.66e-05,
      "loss": 1.8454,
      "step": 84
    },
    {
      "epoch": 0.0168,
      "loss_align": 0.5782209038734436,
      "loss_contrastive": 0.17690232396125793,
      "loss_gen": 1.1786130666732788,
      "neg_sim_mean": 0.4986814260482788,
      "pos_sim_mean": 0.4217790961265564,
      "step": 84
    },
    {
      "epoch": 0.017,
      "grad_norm": 8.277304649353027,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 1.7781,
      "step": 85
    },
    {
      "epoch": 0.017,
      "loss_align": 0.4318523406982422,
      "loss_contrastive": 0.1358771026134491,
      "loss_gen": 1.0663363933563232,
      "neg_sim_mean": 0.6040247678756714,
      "pos_sim_mean": 0.5681476593017578,
      "step": 85
    },
    {
      "epoch": 0.0172,
      "grad_norm": 7.474723815917969,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 1.5145,
      "step": 86
    },
    {
      "epoch": 0.0172,
      "loss_align": 0.5235000848770142,
      "loss_contrastive": 0.112688809633255,
      "loss_gen": 1.4870156049728394,
      "neg_sim_mean": 0.4891887307167053,
      "pos_sim_mean": 0.47649991512298584,
      "step": 86
    },
    {
      "epoch": 0.0174,
      "grad_norm": 11.508679389953613,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 2.024,
      "step": 87
    },
    {
      "epoch": 0.0174,
      "loss_align": 0.3861648440361023,
      "loss_contrastive": 0.14061981439590454,
      "loss_gen": 0.9645462036132812,
      "neg_sim_mean": 0.6544549465179443,
      "pos_sim_mean": 0.6138351559638977,
      "step": 87
    },
    {
      "epoch": 0.0176,
      "grad_norm": 6.978907108306885,
      "learning_rate": 1.74e-05,
      "loss": 1.3676,
      "step": 88
    },
    {
      "epoch": 0.0176,
      "loss_align": 0.3800521492958069,
      "loss_contrastive": 0.009497463703155518,
      "loss_gen": 0.9382809400558472,
      "neg_sim_mean": 0.5294452905654907,
      "pos_sim_mean": 0.6199478507041931,
      "step": 88
    },
    {
      "epoch": 0.0178,
      "grad_norm": 7.171457767486572,
      "learning_rate": 1.76e-05,
      "loss": 1.3195,
      "step": 89
    },
    {
      "epoch": 0.0178,
      "loss_align": 0.5074197053909302,
      "loss_contrastive": 0.10188746452331543,
      "loss_gen": 1.4425121545791626,
      "neg_sim_mean": 0.49446773529052734,
      "pos_sim_mean": 0.49258026480674744,
      "step": 89
    },
    {
      "epoch": 0.018,
      "grad_norm": 11.864455223083496,
      "learning_rate": 1.78e-05,
      "loss": 1.9622,
      "step": 90
    },
    {
      "epoch": 0.018,
      "loss_align": 0.5051394104957581,
      "loss_contrastive": 0.08943858742713928,
      "loss_gen": 1.0757887363433838,
      "neg_sim_mean": 0.4842991828918457,
      "pos_sim_mean": 0.49486058950424194,
      "step": 90
    },
    {
      "epoch": 0.0182,
      "grad_norm": 8.178779602050781,
      "learning_rate": 1.8e-05,
      "loss": 1.5917,
      "step": 91
    },
    {
      "epoch": 0.0182,
      "loss_align": 0.422987163066864,
      "loss_contrastive": 0.0,
      "loss_gen": 1.2294102907180786,
      "neg_sim_mean": 0.4743393659591675,
      "pos_sim_mean": 0.577012836933136,
      "step": 91
    },
    {
      "epoch": 0.0184,
      "grad_norm": 9.131509780883789,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 1.6524,
      "step": 92
    },
    {
      "epoch": 0.0184,
      "loss_align": 0.4116121530532837,
      "loss_contrastive": 0.10387048125267029,
      "loss_gen": 0.8256558775901794,
      "neg_sim_mean": 0.5922583341598511,
      "pos_sim_mean": 0.5883878469467163,
      "step": 92
    },
    {
      "epoch": 0.0186,
      "grad_norm": 7.05711555480957,
      "learning_rate": 1.84e-05,
      "loss": 1.2497,
      "step": 93
    },
    {
      "epoch": 0.0186,
      "loss_align": 0.6160340905189514,
      "loss_contrastive": 0.09046682715415955,
      "loss_gen": 0.7694029808044434,
      "neg_sim_mean": 0.3744327425956726,
      "pos_sim_mean": 0.3839659094810486,
      "step": 93
    },
    {
      "epoch": 0.0188,
      "grad_norm": 6.543647766113281,
      "learning_rate": 1.86e-05,
      "loss": 1.3963,
      "step": 94
    },
    {
      "epoch": 0.0188,
      "loss_align": 0.5352718830108643,
      "loss_contrastive": 0.1408308446407318,
      "loss_gen": 0.7927210330963135,
      "neg_sim_mean": 0.505558967590332,
      "pos_sim_mean": 0.46472811698913574,
      "step": 94
    },
    {
      "epoch": 0.019,
      "grad_norm": 6.190351486206055,
      "learning_rate": 1.88e-05,
      "loss": 1.3449,
      "step": 95
    },
    {
      "epoch": 0.019,
      "loss_align": 0.475008487701416,
      "loss_contrastive": 0.07031485438346863,
      "loss_gen": 0.6501150727272034,
      "neg_sim_mean": 0.4953063726425171,
      "pos_sim_mean": 0.524991512298584,
      "step": 95
    },
    {
      "epoch": 0.0192,
      "grad_norm": 6.025124549865723,
      "learning_rate": 1.9e-05,
      "loss": 1.1336,
      "step": 96
    },
    {
      "epoch": 0.0192,
      "loss_align": 0.4548482298851013,
      "loss_contrastive": 0.12366250157356262,
      "loss_gen": 0.8180707097053528,
      "neg_sim_mean": 0.5688142776489258,
      "pos_sim_mean": 0.5451517701148987,
      "step": 96
    },
    {
      "epoch": 0.0194,
      "grad_norm": 7.409269332885742,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 1.2878,
      "step": 97
    },
    {
      "epoch": 0.0194,
      "loss_align": 0.480344295501709,
      "loss_contrastive": 0.19971153140068054,
      "loss_gen": 0.664563000202179,
      "neg_sim_mean": 0.619367241859436,
      "pos_sim_mean": 0.519655704498291,
      "step": 97
    },
    {
      "epoch": 0.0196,
      "grad_norm": 5.628553867340088,
      "learning_rate": 1.94e-05,
      "loss": 1.1689,
      "step": 98
    },
    {
      "epoch": 0.0196,
      "loss_align": 0.39690375328063965,
      "loss_contrastive": 0.07461345195770264,
      "loss_gen": 0.6062299013137817,
      "neg_sim_mean": 0.5777096748352051,
      "pos_sim_mean": 0.6030962467193604,
      "step": 98
    },
    {
      "epoch": 0.0198,
      "grad_norm": 5.665779113769531,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 1.0121,
      "step": 99
    },
    {
      "epoch": 0.0198,
      "loss_align": 0.46377670764923096,
      "loss_contrastive": 0.1109900176525116,
      "loss_gen": 0.6972107887268066,
      "neg_sim_mean": 0.5472133159637451,
      "pos_sim_mean": 0.536223292350769,
      "step": 99
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.834100246429443,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 1.1743,
      "step": 100
    },
    {
      "epoch": 0.02,
      "loss_align": 0.5219244360923767,
      "loss_contrastive": 0.16768130660057068,
      "loss_gen": 0.7385355830192566,
      "neg_sim_mean": 0.5457568764686584,
      "pos_sim_mean": 0.4780755639076233,
      "step": 100
    },
    {
      "epoch": 0.0202,
      "grad_norm": 7.200370788574219,
      "learning_rate": 2e-05,
      "loss": 1.2806,
      "step": 101
    },
    {
      "epoch": 0.0202,
      "loss_align": 0.3620588183403015,
      "loss_contrastive": 0.09243297576904297,
      "loss_gen": 0.8922696709632874,
      "neg_sim_mean": 0.6303741335868835,
      "pos_sim_mean": 0.6379411816596985,
      "step": 101
    },
    {
      "epoch": 0.0204,
      "grad_norm": 7.492624759674072,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 1.2654,
      "step": 102
    },
    {
      "epoch": 0.0204,
      "loss_align": 0.44170892238616943,
      "loss_contrastive": 0.08292636275291443,
      "loss_gen": 0.5384169816970825,
      "neg_sim_mean": 0.5412174463272095,
      "pos_sim_mean": 0.5582910776138306,
      "step": 102
    },
    {
      "epoch": 0.0206,
      "grad_norm": 5.349119186401367,
      "learning_rate": 2.04e-05,
      "loss": 0.9901,
      "step": 103
    },
    {
      "epoch": 0.0206,
      "loss_align": 0.3772209882736206,
      "loss_contrastive": 0.12209177017211914,
      "loss_gen": 0.5812702178955078,
      "neg_sim_mean": 0.6448707580566406,
      "pos_sim_mean": 0.6227790117263794,
      "step": 103
    },
    {
      "epoch": 0.0208,
      "grad_norm": 5.701079368591309,
      "learning_rate": 2.06e-05,
      "loss": 0.9731,
      "step": 104
    },
    {
      "epoch": 0.0208,
      "loss_align": 0.3319185972213745,
      "loss_contrastive": 0.0,
      "loss_gen": 0.6483643054962158,
      "neg_sim_mean": 0.5276594161987305,
      "pos_sim_mean": 0.6680814027786255,
      "step": 104
    },
    {
      "epoch": 0.021,
      "grad_norm": 5.976101398468018,
      "learning_rate": 2.08e-05,
      "loss": 0.9803,
      "step": 105
    },
    {
      "epoch": 0.021,
      "loss_align": 0.38477861881256104,
      "loss_contrastive": 0.054271399974823,
      "loss_gen": 0.6572605967521667,
      "neg_sim_mean": 0.569492757320404,
      "pos_sim_mean": 0.615221381187439,
      "step": 105
    },
    {
      "epoch": 0.0212,
      "grad_norm": 6.074682235717773,
      "learning_rate": 2.1e-05,
      "loss": 1.0486,
      "step": 106
    },
    {
      "epoch": 0.0212,
      "loss_align": 0.5165272951126099,
      "loss_contrastive": 0.20313546061515808,
      "loss_gen": 0.5726805925369263,
      "neg_sim_mean": 0.5866081714630127,
      "pos_sim_mean": 0.48347270488739014,
      "step": 106
    },
    {
      "epoch": 0.0214,
      "grad_norm": 5.787851333618164,
      "learning_rate": 2.12e-05,
      "loss": 1.1136,
      "step": 107
    },
    {
      "epoch": 0.0214,
      "loss_align": 0.5159496068954468,
      "loss_contrastive": 0.1658209264278412,
      "loss_gen": 1.0340982675552368,
      "neg_sim_mean": 0.5498713254928589,
      "pos_sim_mean": 0.4840503931045532,
      "step": 107
    },
    {
      "epoch": 0.0216,
      "grad_norm": 18.090478897094727,
      "learning_rate": 2.1400000000000002e-05,
      "loss": 1.5699,
      "step": 108
    },
    {
      "epoch": 0.0216,
      "loss_align": 0.4952784776687622,
      "loss_contrastive": 0.061788707971572876,
      "loss_gen": 0.6078722476959229,
      "neg_sim_mean": 0.46651023626327515,
      "pos_sim_mean": 0.5047215223312378,
      "step": 108
    },
    {
      "epoch": 0.0218,
      "grad_norm": 5.741234302520752,
      "learning_rate": 2.16e-05,
      "loss": 1.1106,
      "step": 109
    },
    {
      "epoch": 0.0218,
      "loss_align": 0.4238319396972656,
      "loss_contrastive": 0.06900975108146667,
      "loss_gen": 0.7976281046867371,
      "neg_sim_mean": 0.5451778173446655,
      "pos_sim_mean": 0.5761680603027344,
      "step": 109
    },
    {
      "epoch": 0.022,
      "grad_norm": 8.026098251342773,
      "learning_rate": 2.18e-05,
      "loss": 1.2297,
      "step": 110
    },
    {
      "epoch": 0.022,
      "loss_align": 0.44836366176605225,
      "loss_contrastive": 0.16959556937217712,
      "loss_gen": 0.735077440738678,
      "neg_sim_mean": 0.6212319135665894,
      "pos_sim_mean": 0.5516363382339478,
      "step": 110
    },
    {
      "epoch": 0.0222,
      "grad_norm": 7.907925128936768,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 1.2038,
      "step": 111
    },
    {
      "epoch": 0.0222,
      "loss_align": 0.45366400480270386,
      "loss_contrastive": 0.10640111565589905,
      "loss_gen": 0.45560717582702637,
      "neg_sim_mean": 0.5527371168136597,
      "pos_sim_mean": 0.5463359951972961,
      "step": 111
    },
    {
      "epoch": 0.0224,
      "grad_norm": 5.79366397857666,
      "learning_rate": 2.22e-05,
      "loss": 0.922,
      "step": 112
    },
    {
      "epoch": 0.0224,
      "loss_align": 0.4365161657333374,
      "loss_contrastive": 0.11140164732933044,
      "loss_gen": 0.5615648031234741,
      "neg_sim_mean": 0.5748854875564575,
      "pos_sim_mean": 0.5634838342666626,
      "step": 112
    },
    {
      "epoch": 0.0226,
      "grad_norm": 6.1144328117370605,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 1.0114,
      "step": 113
    },
    {
      "epoch": 0.0226,
      "loss_align": 0.5039534568786621,
      "loss_contrastive": 0.10762885212898254,
      "loss_gen": 0.6410218477249146,
      "neg_sim_mean": 0.5036754012107849,
      "pos_sim_mean": 0.4960465431213379,
      "step": 113
    },
    {
      "epoch": 0.0228,
      "grad_norm": 5.596155643463135,
      "learning_rate": 2.26e-05,
      "loss": 1.1579,
      "step": 114
    },
    {
      "epoch": 0.0228,
      "loss_align": 0.5725840926170349,
      "loss_contrastive": 0.15272587537765503,
      "loss_gen": 0.7210561633110046,
      "neg_sim_mean": 0.4801417887210846,
      "pos_sim_mean": 0.4274159073829651,
      "step": 114
    },
    {
      "epoch": 0.023,
      "grad_norm": 6.242228984832764,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 1.312,
      "step": 115
    },
    {
      "epoch": 0.023,
      "loss_align": 0.44844526052474976,
      "loss_contrastive": 0.017503947019577026,
      "loss_gen": 0.6860126256942749,
      "neg_sim_mean": 0.46905869245529175,
      "pos_sim_mean": 0.5515547394752502,
      "step": 115
    },
    {
      "epoch": 0.0232,
      "grad_norm": 7.527559280395508,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 1.1366,
      "step": 116
    },
    {
      "epoch": 0.0232,
      "loss_align": 0.5161222815513611,
      "loss_contrastive": 0.18133214116096497,
      "loss_gen": 0.6481294631958008,
      "neg_sim_mean": 0.5652098655700684,
      "pos_sim_mean": 0.4838777184486389,
      "step": 116
    },
    {
      "epoch": 0.0234,
      "grad_norm": 5.732221603393555,
      "learning_rate": 2.32e-05,
      "loss": 1.186,
      "step": 117
    },
    {
      "epoch": 0.0234,
      "loss_align": 0.5400981307029724,
      "loss_contrastive": 0.17835858464241028,
      "loss_gen": 0.5223103761672974,
      "neg_sim_mean": 0.5382604598999023,
      "pos_sim_mean": 0.4599018692970276,
      "step": 117
    },
    {
      "epoch": 0.0236,
      "grad_norm": 5.8995890617370605,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 1.0838,
      "step": 118
    },
    {
      "epoch": 0.0236,
      "loss_align": 0.43449240922927856,
      "loss_contrastive": 0.07260408997535706,
      "loss_gen": 0.5163992643356323,
      "neg_sim_mean": 0.538111686706543,
      "pos_sim_mean": 0.5655075907707214,
      "step": 118
    },
    {
      "epoch": 0.0238,
      "grad_norm": 5.593770503997803,
      "learning_rate": 2.36e-05,
      "loss": 0.9596,
      "step": 119
    },
    {
      "epoch": 0.0238,
      "loss_align": 0.35005801916122437,
      "loss_contrastive": 0.04423165321350098,
      "loss_gen": 0.891004204750061,
      "neg_sim_mean": 0.5941736102104187,
      "pos_sim_mean": 0.6499419808387756,
      "step": 119
    },
    {
      "epoch": 0.024,
      "grad_norm": 7.618281841278076,
      "learning_rate": 2.38e-05,
      "loss": 1.2464,
      "step": 120
    },
    {
      "epoch": 0.024,
      "loss_align": 0.3597596287727356,
      "loss_contrastive": 0.07968389987945557,
      "loss_gen": 0.7778334617614746,
      "neg_sim_mean": 0.6199242472648621,
      "pos_sim_mean": 0.6402403712272644,
      "step": 120
    },
    {
      "epoch": 0.0242,
      "grad_norm": 7.1683526039123535,
      "learning_rate": 2.4e-05,
      "loss": 1.1472,
      "step": 121
    },
    {
      "epoch": 0.0242,
      "loss_align": 0.4337579011917114,
      "loss_contrastive": 0.11505505442619324,
      "loss_gen": 0.40096521377563477,
      "neg_sim_mean": 0.5812971591949463,
      "pos_sim_mean": 0.5662420988082886,
      "step": 121
    },
    {
      "epoch": 0.0244,
      "grad_norm": 5.518289089202881,
      "learning_rate": 2.4200000000000002e-05,
      "loss": 0.8485,
      "step": 122
    },
    {
      "epoch": 0.0244,
      "loss_align": 0.5001288652420044,
      "loss_contrastive": 0.1420278549194336,
      "loss_gen": 0.4841098189353943,
      "neg_sim_mean": 0.5418990254402161,
      "pos_sim_mean": 0.499871164560318,
      "step": 122
    },
    {
      "epoch": 0.0246,
      "grad_norm": 5.579812526702881,
      "learning_rate": 2.44e-05,
      "loss": 1.0013,
      "step": 123
    },
    {
      "epoch": 0.0246,
      "loss_align": 0.4203141927719116,
      "loss_contrastive": 0.03033265471458435,
      "loss_gen": 0.6163560748100281,
      "neg_sim_mean": 0.5100184679031372,
      "pos_sim_mean": 0.5796858072280884,
      "step": 123
    },
    {
      "epoch": 0.0248,
      "grad_norm": 7.614130973815918,
      "learning_rate": 2.46e-05,
      "loss": 1.0403,
      "step": 124
    },
    {
      "epoch": 0.0248,
      "loss_align": 0.4647247791290283,
      "loss_contrastive": 0.07258716225624084,
      "loss_gen": 0.561968982219696,
      "neg_sim_mean": 0.507862389087677,
      "pos_sim_mean": 0.5352752208709717,
      "step": 124
    },
    {
      "epoch": 0.025,
      "grad_norm": 7.4538421630859375,
      "learning_rate": 2.48e-05,
      "loss": 1.0354,
      "step": 125
    },
    {
      "epoch": 0.025,
      "loss_align": 0.3536456823348999,
      "loss_contrastive": 0.0,
      "loss_gen": 0.46237608790397644,
      "neg_sim_mean": 0.47011834383010864,
      "pos_sim_mean": 0.6463543176651001,
      "step": 125
    },
    {
      "epoch": 0.0252,
      "grad_norm": 5.15724515914917,
      "learning_rate": 2.5e-05,
      "loss": 0.816,
      "step": 126
    },
    {
      "epoch": 0.0252,
      "loss_align": 0.4914821982383728,
      "loss_contrastive": 0.12179073691368103,
      "loss_gen": 0.4015607535839081,
      "neg_sim_mean": 0.5303085446357727,
      "pos_sim_mean": 0.5085178017616272,
      "step": 126
    },
    {
      "epoch": 0.0254,
      "grad_norm": 6.065829753875732,
      "learning_rate": 2.5200000000000003e-05,
      "loss": 0.9077,
      "step": 127
    },
    {
      "epoch": 0.0254,
      "loss_align": 0.4929812550544739,
      "loss_contrastive": 0.07658860087394714,
      "loss_gen": 0.46357086300849915,
      "neg_sim_mean": 0.48360735177993774,
      "pos_sim_mean": 0.5070187449455261,
      "step": 127
    },
    {
      "epoch": 0.0256,
      "grad_norm": 5.866573810577393,
      "learning_rate": 2.54e-05,
      "loss": 0.9657,
      "step": 128
    },
    {
      "epoch": 0.0256,
      "loss_align": 0.37694692611694336,
      "loss_contrastive": 0.10056078433990479,
      "loss_gen": 0.5241020917892456,
      "neg_sim_mean": 0.6236138343811035,
      "pos_sim_mean": 0.6230530738830566,
      "step": 128
    },
    {
      "epoch": 0.0258,
      "grad_norm": 7.725815296173096,
      "learning_rate": 2.5600000000000002e-05,
      "loss": 0.9131,
      "step": 129
    },
    {
      "epoch": 0.0258,
      "loss_align": 0.37931203842163086,
      "loss_contrastive": 0.08095955848693848,
      "loss_gen": 0.47829946875572205,
      "neg_sim_mean": 0.6016474962234497,
      "pos_sim_mean": 0.6206879615783691,
      "step": 129
    },
    {
      "epoch": 0.026,
      "grad_norm": 6.325316905975342,
      "learning_rate": 2.58e-05,
      "loss": 0.8673,
      "step": 130
    },
    {
      "epoch": 0.026,
      "loss_align": 0.501050591468811,
      "loss_contrastive": 0.16646218299865723,
      "loss_gen": 0.3322647213935852,
      "neg_sim_mean": 0.5654115676879883,
      "pos_sim_mean": 0.4989493787288666,
      "step": 130
    },
    {
      "epoch": 0.0262,
      "grad_norm": 5.124115943908691,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.8533,
      "step": 131
    },
    {
      "epoch": 0.0262,
      "loss_align": 0.3523743748664856,
      "loss_contrastive": 0.0,
      "loss_gen": 0.43963444232940674,
      "neg_sim_mean": 0.5298004150390625,
      "pos_sim_mean": 0.6476256251335144,
      "step": 131
    },
    {
      "epoch": 0.0264,
      "grad_norm": 4.9384446144104,
      "learning_rate": 2.6200000000000003e-05,
      "loss": 0.792,
      "step": 132
    },
    {
      "epoch": 0.0264,
      "loss_align": 0.42369329929351807,
      "loss_contrastive": 0.1639403998851776,
      "loss_gen": 0.4881097972393036,
      "neg_sim_mean": 0.640247106552124,
      "pos_sim_mean": 0.5763067007064819,
      "step": 132
    },
    {
      "epoch": 0.0266,
      "grad_norm": 5.532848358154297,
      "learning_rate": 2.64e-05,
      "loss": 0.9315,
      "step": 133
    },
    {
      "epoch": 0.0266,
      "loss_align": 0.5546563267707825,
      "loss_contrastive": 0.2616322338581085,
      "loss_gen": 0.4650026857852936,
      "neg_sim_mean": 0.6069759130477905,
      "pos_sim_mean": 0.44534367322921753,
      "step": 133
    },
    {
      "epoch": 0.0268,
      "grad_norm": 5.724526882171631,
      "learning_rate": 2.6600000000000003e-05,
      "loss": 1.0511,
      "step": 134
    },
    {
      "epoch": 0.0268,
      "loss_align": 0.5451817512512207,
      "loss_contrastive": 0.14517667889595032,
      "loss_gen": 0.4518257975578308,
      "neg_sim_mean": 0.4999949634075165,
      "pos_sim_mean": 0.4548182785511017,
      "step": 134
    },
    {
      "epoch": 0.027,
      "grad_norm": 5.622949123382568,
      "learning_rate": 2.6800000000000004e-05,
      "loss": 1.0144,
      "step": 135
    },
    {
      "epoch": 0.027,
      "loss_align": 0.5452142357826233,
      "loss_contrastive": 0.1324155032634735,
      "loss_gen": 0.4926644563674927,
      "neg_sim_mean": 0.4872012734413147,
      "pos_sim_mean": 0.4547857642173767,
      "step": 135
    },
    {
      "epoch": 0.0272,
      "grad_norm": 6.344630718231201,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 1.0538,
      "step": 136
    },
    {
      "epoch": 0.0272,
      "loss_align": 0.4895946979522705,
      "loss_contrastive": 0.0929425060749054,
      "loss_gen": 0.38136211037635803,
      "neg_sim_mean": 0.5033478140830994,
      "pos_sim_mean": 0.5104053020477295,
      "step": 136
    },
    {
      "epoch": 0.0274,
      "grad_norm": 5.762722492218018,
      "learning_rate": 2.7200000000000004e-05,
      "loss": 0.8821,
      "step": 137
    },
    {
      "epoch": 0.0274,
      "loss_align": 0.4660536050796509,
      "loss_contrastive": 0.11991056799888611,
      "loss_gen": 0.3930093050003052,
      "neg_sim_mean": 0.5538569688796997,
      "pos_sim_mean": 0.5339463949203491,
      "step": 137
    },
    {
      "epoch": 0.0276,
      "grad_norm": 4.595864772796631,
      "learning_rate": 2.7400000000000002e-05,
      "loss": 0.8735,
      "step": 138
    },
    {
      "epoch": 0.0276,
      "loss_align": 0.44260454177856445,
      "loss_contrastive": 0.10978779196739197,
      "loss_gen": 0.6118648648262024,
      "neg_sim_mean": 0.567183256149292,
      "pos_sim_mean": 0.5573954582214355,
      "step": 138
    },
    {
      "epoch": 0.0278,
      "grad_norm": 5.919750213623047,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 1.0676,
      "step": 139
    },
    {
      "epoch": 0.0278,
      "loss_align": 0.456429660320282,
      "loss_contrastive": 0.174020916223526,
      "loss_gen": 0.38002222776412964,
      "neg_sim_mean": 0.6175912618637085,
      "pos_sim_mean": 0.543570339679718,
      "step": 139
    },
    {
      "epoch": 0.028,
      "grad_norm": 4.8983612060546875,
      "learning_rate": 2.7800000000000005e-05,
      "loss": 0.8573,
      "step": 140
    },
    {
      "epoch": 0.028,
      "loss_align": 0.4696155786514282,
      "loss_contrastive": 0.11397907137870789,
      "loss_gen": 0.46198615431785583,
      "neg_sim_mean": 0.5443634986877441,
      "pos_sim_mean": 0.5303844213485718,
      "step": 140
    },
    {
      "epoch": 0.0282,
      "grad_norm": 5.105416297912598,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.9453,
      "step": 141
    },
    {
      "epoch": 0.0282,
      "loss_align": 0.3786832094192505,
      "loss_contrastive": 0.041602492332458496,
      "loss_gen": 0.3417510390281677,
      "neg_sim_mean": 0.5629192590713501,
      "pos_sim_mean": 0.6213167905807495,
      "step": 141
    },
    {
      "epoch": 0.0284,
      "grad_norm": 4.7344865798950195,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 0.7254,
      "step": 142
    },
    {
      "epoch": 0.0284,
      "loss_align": 0.5640031695365906,
      "loss_contrastive": 0.1420368254184723,
      "loss_gen": 0.5055482387542725,
      "neg_sim_mean": 0.4780336618423462,
      "pos_sim_mean": 0.4359968304634094,
      "step": 142
    },
    {
      "epoch": 0.0286,
      "grad_norm": 6.369258403778076,
      "learning_rate": 2.84e-05,
      "loss": 1.0866,
      "step": 143
    },
    {
      "epoch": 0.0286,
      "loss_align": 0.35473883152008057,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3792015314102173,
      "neg_sim_mean": 0.5122250318527222,
      "pos_sim_mean": 0.6452611684799194,
      "step": 143
    },
    {
      "epoch": 0.0288,
      "grad_norm": 6.1714091300964355,
      "learning_rate": 2.86e-05,
      "loss": 0.7339,
      "step": 144
    },
    {
      "epoch": 0.0288,
      "loss_align": 0.5478449463844299,
      "loss_contrastive": 0.16181877255439758,
      "loss_gen": 0.2154841125011444,
      "neg_sim_mean": 0.5139738321304321,
      "pos_sim_mean": 0.45215505361557007,
      "step": 144
    },
    {
      "epoch": 0.029,
      "grad_norm": 3.738401412963867,
      "learning_rate": 2.88e-05,
      "loss": 0.7827,
      "step": 145
    },
    {
      "epoch": 0.029,
      "loss_align": 0.48112308979034424,
      "loss_contrastive": 0.12663379311561584,
      "loss_gen": 0.24522559344768524,
      "neg_sim_mean": 0.5455107092857361,
      "pos_sim_mean": 0.5188769102096558,
      "step": 145
    },
    {
      "epoch": 0.0292,
      "grad_norm": 4.148242950439453,
      "learning_rate": 2.9e-05,
      "loss": 0.7415,
      "step": 146
    },
    {
      "epoch": 0.0292,
      "loss_align": 0.37362635135650635,
      "loss_contrastive": 0.03840160369873047,
      "loss_gen": 0.4379314184188843,
      "neg_sim_mean": 0.5647752285003662,
      "pos_sim_mean": 0.6263736486434937,
      "step": 146
    },
    {
      "epoch": 0.0294,
      "grad_norm": 6.075196266174316,
      "learning_rate": 2.9199999999999998e-05,
      "loss": 0.8162,
      "step": 147
    },
    {
      "epoch": 0.0294,
      "loss_align": 0.4628152847290039,
      "loss_contrastive": 0.0632070004940033,
      "loss_gen": 0.43002742528915405,
      "neg_sim_mean": 0.5003917217254639,
      "pos_sim_mean": 0.5371847152709961,
      "step": 147
    },
    {
      "epoch": 0.0296,
      "grad_norm": 6.197295188903809,
      "learning_rate": 2.94e-05,
      "loss": 0.9004,
      "step": 148
    },
    {
      "epoch": 0.0296,
      "loss_align": 0.34182465076446533,
      "loss_contrastive": 0.045772671699523926,
      "loss_gen": 0.6695640683174133,
      "neg_sim_mean": 0.6039479970932007,
      "pos_sim_mean": 0.6581753492355347,
      "step": 148
    },
    {
      "epoch": 0.0298,
      "grad_norm": 7.203254699707031,
      "learning_rate": 2.96e-05,
      "loss": 1.0169,
      "step": 149
    },
    {
      "epoch": 0.0298,
      "loss_align": 0.4264000654220581,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4289204180240631,
      "neg_sim_mean": 0.417951762676239,
      "pos_sim_mean": 0.5735999345779419,
      "step": 149
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.918241500854492,
      "learning_rate": 2.98e-05,
      "loss": 0.8553,
      "step": 150
    },
    {
      "epoch": 0.03,
      "loss_align": 0.5878881216049194,
      "loss_contrastive": 0.21409893035888672,
      "loss_gen": 0.24673619866371155,
      "neg_sim_mean": 0.5262108445167542,
      "pos_sim_mean": 0.41211190819740295,
      "step": 150
    },
    {
      "epoch": 0.0302,
      "grad_norm": 2.455801010131836,
      "learning_rate": 3e-05,
      "loss": 0.8603,
      "step": 151
    },
    {
      "epoch": 0.0302,
      "loss_align": 0.5454258322715759,
      "loss_contrastive": 0.13967391848564148,
      "loss_gen": 0.27479374408721924,
      "neg_sim_mean": 0.49424809217453003,
      "pos_sim_mean": 0.4545741677284241,
      "step": 151
    },
    {
      "epoch": 0.0304,
      "grad_norm": 4.048895359039307,
      "learning_rate": 3.02e-05,
      "loss": 0.837,
      "step": 152
    },
    {
      "epoch": 0.0304,
      "loss_align": 0.45105183124542236,
      "loss_contrastive": 0.10559436678886414,
      "loss_gen": 0.30397528409957886,
      "neg_sim_mean": 0.5545425415039062,
      "pos_sim_mean": 0.5489481687545776,
      "step": 152
    },
    {
      "epoch": 0.0306,
      "grad_norm": 4.6391825675964355,
      "learning_rate": 3.04e-05,
      "loss": 0.7677,
      "step": 153
    },
    {
      "epoch": 0.0306,
      "loss_align": 0.39970195293426514,
      "loss_contrastive": 0.03682434558868408,
      "loss_gen": 0.17031940817832947,
      "neg_sim_mean": 0.537122368812561,
      "pos_sim_mean": 0.6002980470657349,
      "step": 153
    },
    {
      "epoch": 0.0308,
      "grad_norm": 2.920287609100342,
      "learning_rate": 3.06e-05,
      "loss": 0.5744,
      "step": 154
    },
    {
      "epoch": 0.0308,
      "loss_align": 0.4240163564682007,
      "loss_contrastive": 0.15769335627555847,
      "loss_gen": 0.422208309173584,
      "neg_sim_mean": 0.6336770057678223,
      "pos_sim_mean": 0.5759836435317993,
      "step": 154
    },
    {
      "epoch": 0.031,
      "grad_norm": 5.949808120727539,
      "learning_rate": 3.08e-05,
      "loss": 0.8651,
      "step": 155
    },
    {
      "epoch": 0.031,
      "loss_align": 0.43330472707748413,
      "loss_contrastive": 0.1491251289844513,
      "loss_gen": 0.48682302236557007,
      "neg_sim_mean": 0.6158204078674316,
      "pos_sim_mean": 0.5666952729225159,
      "step": 155
    },
    {
      "epoch": 0.0312,
      "grad_norm": 5.893032073974609,
      "learning_rate": 3.1e-05,
      "loss": 0.938,
      "step": 156
    },
    {
      "epoch": 0.0312,
      "loss_align": 0.40632832050323486,
      "loss_contrastive": 0.12900522351264954,
      "loss_gen": 0.27475398778915405,
      "neg_sim_mean": 0.6226769089698792,
      "pos_sim_mean": 0.5936716794967651,
      "step": 156
    },
    {
      "epoch": 0.0314,
      "grad_norm": 3.9093079566955566,
      "learning_rate": 3.12e-05,
      "loss": 0.6966,
      "step": 157
    },
    {
      "epoch": 0.0314,
      "loss_align": 0.39049315452575684,
      "loss_contrastive": 0.0017915964126586914,
      "loss_gen": 0.3422077000141144,
      "neg_sim_mean": 0.511298418045044,
      "pos_sim_mean": 0.6095068454742432,
      "step": 157
    },
    {
      "epoch": 0.0316,
      "grad_norm": 4.825846195220947,
      "learning_rate": 3.1400000000000004e-05,
      "loss": 0.7329,
      "step": 158
    },
    {
      "epoch": 0.0316,
      "loss_align": 0.4338375926017761,
      "loss_contrastive": 0.08994874358177185,
      "loss_gen": 0.3600914478302002,
      "neg_sim_mean": 0.5561111569404602,
      "pos_sim_mean": 0.5661624073982239,
      "step": 158
    },
    {
      "epoch": 0.0318,
      "grad_norm": 5.602405071258545,
      "learning_rate": 3.16e-05,
      "loss": 0.8047,
      "step": 159
    },
    {
      "epoch": 0.0318,
      "loss_align": 0.5351468324661255,
      "loss_contrastive": 0.07343053817749023,
      "loss_gen": 0.4278753101825714,
      "neg_sim_mean": 0.43828368186950684,
      "pos_sim_mean": 0.4648531377315521,
      "step": 159
    },
    {
      "epoch": 0.032,
      "grad_norm": 5.235030174255371,
      "learning_rate": 3.18e-05,
      "loss": 0.9718,
      "step": 160
    },
    {
      "epoch": 0.032,
      "loss_align": 0.5241231322288513,
      "loss_contrastive": 0.0907234251499176,
      "loss_gen": 0.44842445850372314,
      "neg_sim_mean": 0.46660029888153076,
      "pos_sim_mean": 0.4758768677711487,
      "step": 160
    },
    {
      "epoch": 0.0322,
      "grad_norm": 5.759180545806885,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.9834,
      "step": 161
    },
    {
      "epoch": 0.0322,
      "loss_align": 0.5117963552474976,
      "loss_contrastive": 0.11687672138214111,
      "loss_gen": 0.36261340975761414,
      "neg_sim_mean": 0.5050803422927856,
      "pos_sim_mean": 0.48820361495018005,
      "step": 161
    },
    {
      "epoch": 0.0324,
      "grad_norm": 4.8927531242370605,
      "learning_rate": 3.2200000000000003e-05,
      "loss": 0.8884,
      "step": 162
    },
    {
      "epoch": 0.0324,
      "loss_align": 0.3956359028816223,
      "loss_contrastive": 0.02845442295074463,
      "loss_gen": 0.1593475043773651,
      "neg_sim_mean": 0.5328184962272644,
      "pos_sim_mean": 0.6043640971183777,
      "step": 162
    },
    {
      "epoch": 0.0326,
      "grad_norm": 3.532364845275879,
      "learning_rate": 3.24e-05,
      "loss": 0.5584,
      "step": 163
    },
    {
      "epoch": 0.0326,
      "loss_align": 0.45397186279296875,
      "loss_contrastive": 0.07152697443962097,
      "loss_gen": 0.23670007288455963,
      "neg_sim_mean": 0.5175551176071167,
      "pos_sim_mean": 0.5460281372070312,
      "step": 163
    },
    {
      "epoch": 0.0328,
      "grad_norm": 4.519861221313477,
      "learning_rate": 3.26e-05,
      "loss": 0.6993,
      "step": 164
    },
    {
      "epoch": 0.0328,
      "loss_align": 0.54201340675354,
      "loss_contrastive": 0.11592993140220642,
      "loss_gen": 0.3073796033859253,
      "neg_sim_mean": 0.47391656041145325,
      "pos_sim_mean": 0.45798662304878235,
      "step": 164
    },
    {
      "epoch": 0.033,
      "grad_norm": 4.596250057220459,
      "learning_rate": 3.2800000000000004e-05,
      "loss": 0.8633,
      "step": 165
    },
    {
      "epoch": 0.033,
      "loss_align": 0.4238414764404297,
      "loss_contrastive": 0.04615470767021179,
      "loss_gen": 0.26937904953956604,
      "neg_sim_mean": 0.5223132371902466,
      "pos_sim_mean": 0.5761585235595703,
      "step": 165
    },
    {
      "epoch": 0.0332,
      "grad_norm": 4.7662529945373535,
      "learning_rate": 3.3e-05,
      "loss": 0.6988,
      "step": 166
    },
    {
      "epoch": 0.0332,
      "loss_align": 0.4628787040710449,
      "loss_contrastive": 0.1318475306034088,
      "loss_gen": 0.4048099219799042,
      "neg_sim_mean": 0.5689688324928284,
      "pos_sim_mean": 0.5371212959289551,
      "step": 166
    },
    {
      "epoch": 0.0334,
      "grad_norm": 5.654176712036133,
      "learning_rate": 3.32e-05,
      "loss": 0.8835,
      "step": 167
    },
    {
      "epoch": 0.0334,
      "loss_align": 0.4189443588256836,
      "loss_contrastive": 0.08637967705726624,
      "loss_gen": 0.23229509592056274,
      "neg_sim_mean": 0.5674353241920471,
      "pos_sim_mean": 0.5810556411743164,
      "step": 167
    },
    {
      "epoch": 0.0336,
      "grad_norm": 4.541215896606445,
      "learning_rate": 3.3400000000000005e-05,
      "loss": 0.6616,
      "step": 168
    },
    {
      "epoch": 0.0336,
      "loss_align": 0.4642232060432434,
      "loss_contrastive": 0.0,
      "loss_gen": 0.38820165395736694,
      "neg_sim_mean": 0.4350593686103821,
      "pos_sim_mean": 0.5357767939567566,
      "step": 168
    },
    {
      "epoch": 0.0338,
      "grad_norm": 5.548286437988281,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 0.8524,
      "step": 169
    },
    {
      "epoch": 0.0338,
      "loss_align": 0.46382415294647217,
      "loss_contrastive": 0.07637140154838562,
      "loss_gen": 0.11571227759122849,
      "neg_sim_mean": 0.5125472545623779,
      "pos_sim_mean": 0.5361758470535278,
      "step": 169
    },
    {
      "epoch": 0.034,
      "grad_norm": 2.2568464279174805,
      "learning_rate": 3.38e-05,
      "loss": 0.5887,
      "step": 170
    },
    {
      "epoch": 0.034,
      "loss_align": 0.48151350021362305,
      "loss_contrastive": 0.14905700087547302,
      "loss_gen": 0.37822994589805603,
      "neg_sim_mean": 0.5675435066223145,
      "pos_sim_mean": 0.518486499786377,
      "step": 170
    },
    {
      "epoch": 0.0342,
      "grad_norm": 7.788582801818848,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.8776,
      "step": 171
    },
    {
      "epoch": 0.0342,
      "loss_align": 0.43759608268737793,
      "loss_contrastive": 0.09290966391563416,
      "loss_gen": 0.2575598359107971,
      "neg_sim_mean": 0.5553135871887207,
      "pos_sim_mean": 0.5624039173126221,
      "step": 171
    },
    {
      "epoch": 0.0344,
      "grad_norm": 4.4936676025390625,
      "learning_rate": 3.4200000000000005e-05,
      "loss": 0.7063,
      "step": 172
    },
    {
      "epoch": 0.0344,
      "loss_align": 0.4292588233947754,
      "loss_contrastive": 0.0031045377254486084,
      "loss_gen": 0.36139237880706787,
      "neg_sim_mean": 0.4738457202911377,
      "pos_sim_mean": 0.5707411766052246,
      "step": 172
    },
    {
      "epoch": 0.0346,
      "grad_norm": 5.737329959869385,
      "learning_rate": 3.4399999999999996e-05,
      "loss": 0.791,
      "step": 173
    },
    {
      "epoch": 0.0346,
      "loss_align": 0.36715030670166016,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12730902433395386,
      "neg_sim_mean": 0.5014106035232544,
      "pos_sim_mean": 0.6328496932983398,
      "step": 173
    },
    {
      "epoch": 0.0348,
      "grad_norm": 2.9097838401794434,
      "learning_rate": 3.46e-05,
      "loss": 0.4945,
      "step": 174
    },
    {
      "epoch": 0.0348,
      "loss_align": 0.466037392616272,
      "loss_contrastive": 0.054704517126083374,
      "loss_gen": 0.30287548899650574,
      "neg_sim_mean": 0.4886671304702759,
      "pos_sim_mean": 0.533962607383728,
      "step": 174
    },
    {
      "epoch": 0.035,
      "grad_norm": 6.570965766906738,
      "learning_rate": 3.48e-05,
      "loss": 0.7755,
      "step": 175
    },
    {
      "epoch": 0.035,
      "loss_align": 0.4086287021636963,
      "loss_contrastive": 0.021452337503433228,
      "loss_gen": 0.36525678634643555,
      "neg_sim_mean": 0.5128236413002014,
      "pos_sim_mean": 0.5913712978363037,
      "step": 175
    },
    {
      "epoch": 0.0352,
      "grad_norm": 5.993163108825684,
      "learning_rate": 3.5e-05,
      "loss": 0.7765,
      "step": 176
    },
    {
      "epoch": 0.0352,
      "loss_align": 0.45112860202789307,
      "loss_contrastive": 0.047946274280548096,
      "loss_gen": 0.09432340413331985,
      "neg_sim_mean": 0.4968176782131195,
      "pos_sim_mean": 0.5488713979721069,
      "step": 176
    },
    {
      "epoch": 0.0354,
      "grad_norm": 2.9195799827575684,
      "learning_rate": 3.52e-05,
      "loss": 0.5512,
      "step": 177
    },
    {
      "epoch": 0.0354,
      "loss_align": 0.3656498193740845,
      "loss_contrastive": 0.05842381715774536,
      "loss_gen": 0.40867263078689575,
      "neg_sim_mean": 0.592773973941803,
      "pos_sim_mean": 0.6343501806259155,
      "step": 177
    },
    {
      "epoch": 0.0356,
      "grad_norm": 4.96196174621582,
      "learning_rate": 3.54e-05,
      "loss": 0.7813,
      "step": 178
    },
    {
      "epoch": 0.0356,
      "loss_align": 0.4446887969970703,
      "loss_contrastive": 0.1458244025707245,
      "loss_gen": 0.24491609632968903,
      "neg_sim_mean": 0.6011356115341187,
      "pos_sim_mean": 0.5553112030029297,
      "step": 178
    },
    {
      "epoch": 0.0358,
      "grad_norm": 4.348042011260986,
      "learning_rate": 3.56e-05,
      "loss": 0.7071,
      "step": 179
    },
    {
      "epoch": 0.0358,
      "loss_align": 0.46612274646759033,
      "loss_contrastive": 0.1318652331829071,
      "loss_gen": 0.2709472179412842,
      "neg_sim_mean": 0.5657424926757812,
      "pos_sim_mean": 0.5338772535324097,
      "step": 179
    },
    {
      "epoch": 0.036,
      "grad_norm": 4.797234535217285,
      "learning_rate": 3.58e-05,
      "loss": 0.7529,
      "step": 180
    },
    {
      "epoch": 0.036,
      "loss_align": 0.35726314783096313,
      "loss_contrastive": 0.11358612775802612,
      "loss_gen": 0.18004070222377777,
      "neg_sim_mean": 0.6563229560852051,
      "pos_sim_mean": 0.6427368521690369,
      "step": 180
    },
    {
      "epoch": 0.0362,
      "grad_norm": 4.062804698944092,
      "learning_rate": 3.6e-05,
      "loss": 0.5509,
      "step": 181
    },
    {
      "epoch": 0.0362,
      "loss_align": 0.5258806943893433,
      "loss_contrastive": 0.17780962586402893,
      "loss_gen": 0.3028411865234375,
      "neg_sim_mean": 0.5519289374351501,
      "pos_sim_mean": 0.47411930561065674,
      "step": 181
    },
    {
      "epoch": 0.0364,
      "grad_norm": 5.753906726837158,
      "learning_rate": 3.62e-05,
      "loss": 0.8501,
      "step": 182
    },
    {
      "epoch": 0.0364,
      "loss_align": 0.46896159648895264,
      "loss_contrastive": 0.14960268139839172,
      "loss_gen": 0.30617743730545044,
      "neg_sim_mean": 0.5806410908699036,
      "pos_sim_mean": 0.5310384035110474,
      "step": 182
    },
    {
      "epoch": 0.0366,
      "grad_norm": 5.785995960235596,
      "learning_rate": 3.6400000000000004e-05,
      "loss": 0.7931,
      "step": 183
    },
    {
      "epoch": 0.0366,
      "loss_align": 0.46741658449172974,
      "loss_contrastive": 0.13203856348991394,
      "loss_gen": 0.6183087825775146,
      "neg_sim_mean": 0.5646219849586487,
      "pos_sim_mean": 0.5325834155082703,
      "step": 183
    },
    {
      "epoch": 0.0368,
      "grad_norm": 8.358132362365723,
      "learning_rate": 3.66e-05,
      "loss": 1.1016,
      "step": 184
    },
    {
      "epoch": 0.0368,
      "loss_align": 0.37738776206970215,
      "loss_contrastive": 0.1014525294303894,
      "loss_gen": 0.36192262172698975,
      "neg_sim_mean": 0.6240647435188293,
      "pos_sim_mean": 0.6226122379302979,
      "step": 184
    },
    {
      "epoch": 0.037,
      "grad_norm": 5.411758899688721,
      "learning_rate": 3.68e-05,
      "loss": 0.7515,
      "step": 185
    },
    {
      "epoch": 0.037,
      "loss_align": 0.47929519414901733,
      "loss_contrastive": 0.02412673830986023,
      "loss_gen": 0.3436315357685089,
      "neg_sim_mean": 0.4448315501213074,
      "pos_sim_mean": 0.5207048058509827,
      "step": 185
    },
    {
      "epoch": 0.0372,
      "grad_norm": 4.824592113494873,
      "learning_rate": 3.7e-05,
      "loss": 0.8258,
      "step": 186
    },
    {
      "epoch": 0.0372,
      "loss_align": 0.5040410161018372,
      "loss_contrastive": 0.22798308730125427,
      "loss_gen": 0.2898566722869873,
      "neg_sim_mean": 0.6239420771598816,
      "pos_sim_mean": 0.49595898389816284,
      "step": 186
    },
    {
      "epoch": 0.0374,
      "grad_norm": 5.337868690490723,
      "learning_rate": 3.72e-05,
      "loss": 0.8213,
      "step": 187
    },
    {
      "epoch": 0.0374,
      "loss_align": 0.523390531539917,
      "loss_contrastive": 0.12103748321533203,
      "loss_gen": 0.2579670548439026,
      "neg_sim_mean": 0.4976469576358795,
      "pos_sim_mean": 0.476609468460083,
      "step": 187
    },
    {
      "epoch": 0.0376,
      "grad_norm": 5.076371192932129,
      "learning_rate": 3.74e-05,
      "loss": 0.7959,
      "step": 188
    },
    {
      "epoch": 0.0376,
      "loss_align": 0.45778363943099976,
      "loss_contrastive": 0.17587032914161682,
      "loss_gen": 0.24829734861850739,
      "neg_sim_mean": 0.6180866956710815,
      "pos_sim_mean": 0.5422163605690002,
      "step": 188
    },
    {
      "epoch": 0.0378,
      "grad_norm": 4.108355522155762,
      "learning_rate": 3.76e-05,
      "loss": 0.7272,
      "step": 189
    },
    {
      "epoch": 0.0378,
      "loss_align": 0.4470306634902954,
      "loss_contrastive": 0.16562989354133606,
      "loss_gen": 0.5239307284355164,
      "neg_sim_mean": 0.6185992360115051,
      "pos_sim_mean": 0.5529693365097046,
      "step": 189
    },
    {
      "epoch": 0.038,
      "grad_norm": 6.678189754486084,
      "learning_rate": 3.7800000000000004e-05,
      "loss": 0.9908,
      "step": 190
    },
    {
      "epoch": 0.038,
      "loss_align": 0.3724656105041504,
      "loss_contrastive": 0.02852684259414673,
      "loss_gen": 0.2685311734676361,
      "neg_sim_mean": 0.5560612082481384,
      "pos_sim_mean": 0.6275343894958496,
      "step": 190
    },
    {
      "epoch": 0.0382,
      "grad_norm": 4.318970203399658,
      "learning_rate": 3.8e-05,
      "loss": 0.6444,
      "step": 191
    },
    {
      "epoch": 0.0382,
      "loss_align": 0.4535631537437439,
      "loss_contrastive": 0.1173078715801239,
      "loss_gen": 0.22960515320301056,
      "neg_sim_mean": 0.5637447237968445,
      "pos_sim_mean": 0.5464368462562561,
      "step": 191
    },
    {
      "epoch": 0.0384,
      "grad_norm": 5.094582557678223,
      "learning_rate": 3.82e-05,
      "loss": 0.6972,
      "step": 192
    },
    {
      "epoch": 0.0384,
      "loss_align": 0.3571040630340576,
      "loss_contrastive": 0.11535775661468506,
      "loss_gen": 0.3527502715587616,
      "neg_sim_mean": 0.6582536697387695,
      "pos_sim_mean": 0.6428959369659424,
      "step": 192
    },
    {
      "epoch": 0.0386,
      "grad_norm": 5.486941814422607,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 0.7237,
      "step": 193
    },
    {
      "epoch": 0.0386,
      "loss_align": 0.33489859104156494,
      "loss_contrastive": 0.11046695709228516,
      "loss_gen": 0.4050337076187134,
      "neg_sim_mean": 0.6755683422088623,
      "pos_sim_mean": 0.6651014089584351,
      "step": 193
    },
    {
      "epoch": 0.0388,
      "grad_norm": 5.438366889953613,
      "learning_rate": 3.86e-05,
      "loss": 0.7532,
      "step": 194
    },
    {
      "epoch": 0.0388,
      "loss_align": 0.3869221806526184,
      "loss_contrastive": 0.0,
      "loss_gen": 0.06825444847345352,
      "neg_sim_mean": 0.47894224524497986,
      "pos_sim_mean": 0.6130778193473816,
      "step": 194
    },
    {
      "epoch": 0.039,
      "grad_norm": 3.130199432373047,
      "learning_rate": 3.88e-05,
      "loss": 0.4552,
      "step": 195
    },
    {
      "epoch": 0.039,
      "loss_align": 0.38975560665130615,
      "loss_contrastive": 0.15466523170471191,
      "loss_gen": 0.12358077615499496,
      "neg_sim_mean": 0.6649096012115479,
      "pos_sim_mean": 0.6102443933486938,
      "step": 195
    },
    {
      "epoch": 0.0392,
      "grad_norm": 3.200873374938965,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.5319,
      "step": 196
    },
    {
      "epoch": 0.0392,
      "loss_align": 0.4974863529205322,
      "loss_contrastive": 0.09418979287147522,
      "loss_gen": 0.3449096083641052,
      "neg_sim_mean": 0.49670344591140747,
      "pos_sim_mean": 0.5025136470794678,
      "step": 196
    },
    {
      "epoch": 0.0394,
      "grad_norm": 5.417786598205566,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 0.8537,
      "step": 197
    },
    {
      "epoch": 0.0394,
      "loss_align": 0.4987875819206238,
      "loss_contrastive": 0.08878776431083679,
      "loss_gen": 0.2563721835613251,
      "neg_sim_mean": 0.4900001883506775,
      "pos_sim_mean": 0.5012124180793762,
      "step": 197
    },
    {
      "epoch": 0.0396,
      "grad_norm": 4.75882625579834,
      "learning_rate": 3.94e-05,
      "loss": 0.7658,
      "step": 198
    },
    {
      "epoch": 0.0396,
      "loss_align": 0.5195451974868774,
      "loss_contrastive": 0.10531061887741089,
      "loss_gen": 0.21138906478881836,
      "neg_sim_mean": 0.48576539754867554,
      "pos_sim_mean": 0.48045477271080017,
      "step": 198
    },
    {
      "epoch": 0.0398,
      "grad_norm": 4.9001874923706055,
      "learning_rate": 3.960000000000001e-05,
      "loss": 0.7436,
      "step": 199
    },
    {
      "epoch": 0.0398,
      "loss_align": 0.36207395792007446,
      "loss_contrastive": 0.13273674249649048,
      "loss_gen": 0.24235498905181885,
      "neg_sim_mean": 0.6706627607345581,
      "pos_sim_mean": 0.6379260420799255,
      "step": 199
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.139596939086914,
      "learning_rate": 3.9800000000000005e-05,
      "loss": 0.6204,
      "step": 200
    },
    {
      "epoch": 0.04,
      "loss_align": 0.4193331003189087,
      "loss_contrastive": 0.0015492141246795654,
      "loss_gen": 0.03515394777059555,
      "neg_sim_mean": 0.48221611976623535,
      "pos_sim_mean": 0.5806668996810913,
      "step": 200
    },
    {
      "epoch": 0.0402,
      "grad_norm": 2.470125436782837,
      "learning_rate": 4e-05,
      "loss": 0.4547,
      "step": 201
    },
    {
      "epoch": 0.0402,
      "loss_align": 0.4070572853088379,
      "loss_contrastive": 0.15103933215141296,
      "loss_gen": 0.2164098173379898,
      "neg_sim_mean": 0.6439820528030396,
      "pos_sim_mean": 0.5929427146911621,
      "step": 201
    },
    {
      "epoch": 0.0404,
      "grad_norm": 4.722174167633057,
      "learning_rate": 4.02e-05,
      "loss": 0.6416,
      "step": 202
    },
    {
      "epoch": 0.0404,
      "loss_align": 0.6185012459754944,
      "loss_contrastive": 0.13889071345329285,
      "loss_gen": 0.4773666262626648,
      "neg_sim_mean": 0.42038947343826294,
      "pos_sim_mean": 0.3814987540245056,
      "step": 202
    },
    {
      "epoch": 0.0406,
      "grad_norm": 6.516274452209473,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 1.1125,
      "step": 203
    },
    {
      "epoch": 0.0406,
      "loss_align": 0.47003817558288574,
      "loss_contrastive": 0.05200830101966858,
      "loss_gen": 0.03311925008893013,
      "neg_sim_mean": 0.4819701313972473,
      "pos_sim_mean": 0.5299618244171143,
      "step": 203
    },
    {
      "epoch": 0.0408,
      "grad_norm": 4.208193302154541,
      "learning_rate": 4.0600000000000004e-05,
      "loss": 0.5094,
      "step": 204
    },
    {
      "epoch": 0.0408,
      "loss_align": 0.5821070075035095,
      "loss_contrastive": 0.19321903586387634,
      "loss_gen": 0.21021775901317596,
      "neg_sim_mean": 0.5111120343208313,
      "pos_sim_mean": 0.4178929924964905,
      "step": 204
    },
    {
      "epoch": 0.041,
      "grad_norm": 4.678169250488281,
      "learning_rate": 4.08e-05,
      "loss": 0.8155,
      "step": 205
    },
    {
      "epoch": 0.041,
      "loss_align": 0.37039393186569214,
      "loss_contrastive": 0.0,
      "loss_gen": 0.30753833055496216,
      "neg_sim_mean": 0.4497188925743103,
      "pos_sim_mean": 0.6296060681343079,
      "step": 205
    },
    {
      "epoch": 0.0412,
      "grad_norm": 4.998572826385498,
      "learning_rate": 4.1e-05,
      "loss": 0.6779,
      "step": 206
    },
    {
      "epoch": 0.0412,
      "loss_align": 0.41097861528396606,
      "loss_contrastive": 0.0425589382648468,
      "loss_gen": 0.3987511694431305,
      "neg_sim_mean": 0.5315803289413452,
      "pos_sim_mean": 0.5890213847160339,
      "step": 206
    },
    {
      "epoch": 0.0414,
      "grad_norm": 5.542154788970947,
      "learning_rate": 4.12e-05,
      "loss": 0.8148,
      "step": 207
    },
    {
      "epoch": 0.0414,
      "loss_align": 0.3082458972930908,
      "loss_contrastive": 0.07713371515274048,
      "loss_gen": 0.25100159645080566,
      "neg_sim_mean": 0.6688877940177917,
      "pos_sim_mean": 0.6917541027069092,
      "step": 207
    },
    {
      "epoch": 0.0416,
      "grad_norm": 4.361125946044922,
      "learning_rate": 4.14e-05,
      "loss": 0.5685,
      "step": 208
    },
    {
      "epoch": 0.0416,
      "loss_align": 0.42902737855911255,
      "loss_contrastive": 0.06247606873512268,
      "loss_gen": 0.30839261412620544,
      "neg_sim_mean": 0.5334486961364746,
      "pos_sim_mean": 0.5709726214408875,
      "step": 208
    },
    {
      "epoch": 0.0418,
      "grad_norm": 5.7608723640441895,
      "learning_rate": 4.16e-05,
      "loss": 0.7449,
      "step": 209
    },
    {
      "epoch": 0.0418,
      "loss_align": 0.36819398403167725,
      "loss_contrastive": 0.0,
      "loss_gen": 0.022685201838612556,
      "neg_sim_mean": 0.48849114775657654,
      "pos_sim_mean": 0.6318060159683228,
      "step": 209
    },
    {
      "epoch": 0.042,
      "grad_norm": 1.9100204706192017,
      "learning_rate": 4.18e-05,
      "loss": 0.3909,
      "step": 210
    },
    {
      "epoch": 0.042,
      "loss_align": 0.4912400245666504,
      "loss_contrastive": 0.09620729088783264,
      "loss_gen": 0.3758351504802704,
      "neg_sim_mean": 0.5049672722816467,
      "pos_sim_mean": 0.5087599754333496,
      "step": 210
    },
    {
      "epoch": 0.0422,
      "grad_norm": 6.388913154602051,
      "learning_rate": 4.2e-05,
      "loss": 0.8786,
      "step": 211
    },
    {
      "epoch": 0.0422,
      "loss_align": 0.5224940180778503,
      "loss_contrastive": 0.11184930801391602,
      "loss_gen": 0.01331443339586258,
      "neg_sim_mean": 0.48935529589653015,
      "pos_sim_mean": 0.47750598192214966,
      "step": 211
    },
    {
      "epoch": 0.0424,
      "grad_norm": 1.382930874824524,
      "learning_rate": 4.22e-05,
      "loss": 0.5492,
      "step": 212
    },
    {
      "epoch": 0.0424,
      "loss_align": 0.33492934703826904,
      "loss_contrastive": 0.10174006223678589,
      "loss_gen": 0.4115355312824249,
      "neg_sim_mean": 0.6668106913566589,
      "pos_sim_mean": 0.665070652961731,
      "step": 212
    },
    {
      "epoch": 0.0426,
      "grad_norm": 6.874267101287842,
      "learning_rate": 4.24e-05,
      "loss": 0.7587,
      "step": 213
    },
    {
      "epoch": 0.0426,
      "loss_align": 0.5931655168533325,
      "loss_contrastive": 0.055337756872177124,
      "loss_gen": 0.40168601274490356,
      "neg_sim_mean": 0.36217227578163147,
      "pos_sim_mean": 0.40683451294898987,
      "step": 213
    },
    {
      "epoch": 0.0428,
      "grad_norm": 7.9680256843566895,
      "learning_rate": 4.26e-05,
      "loss": 1.0015,
      "step": 214
    },
    {
      "epoch": 0.0428,
      "loss_align": 0.27333712577819824,
      "loss_contrastive": 0.08907902240753174,
      "loss_gen": 0.26698532700538635,
      "neg_sim_mean": 0.7157418727874756,
      "pos_sim_mean": 0.7266628742218018,
      "step": 214
    },
    {
      "epoch": 0.043,
      "grad_norm": 4.72618293762207,
      "learning_rate": 4.2800000000000004e-05,
      "loss": 0.551,
      "step": 215
    },
    {
      "epoch": 0.043,
      "loss_align": 0.37628066539764404,
      "loss_contrastive": 0.07319754362106323,
      "loss_gen": 0.23546908795833588,
      "neg_sim_mean": 0.5969168543815613,
      "pos_sim_mean": 0.623719334602356,
      "step": 215
    },
    {
      "epoch": 0.0432,
      "grad_norm": 4.86954927444458,
      "learning_rate": 4.3e-05,
      "loss": 0.6205,
      "step": 216
    },
    {
      "epoch": 0.0432,
      "loss_align": 0.46930307149887085,
      "loss_contrastive": 0.09858909249305725,
      "loss_gen": 0.39767855405807495,
      "neg_sim_mean": 0.5292860269546509,
      "pos_sim_mean": 0.5306969285011292,
      "step": 216
    },
    {
      "epoch": 0.0434,
      "grad_norm": 5.8015360832214355,
      "learning_rate": 4.32e-05,
      "loss": 0.8788,
      "step": 217
    },
    {
      "epoch": 0.0434,
      "loss_align": 0.6396294832229614,
      "loss_contrastive": 0.12948283553123474,
      "loss_gen": 0.25968223810195923,
      "neg_sim_mean": 0.3898533582687378,
      "pos_sim_mean": 0.3603705167770386,
      "step": 217
    },
    {
      "epoch": 0.0436,
      "grad_norm": 4.880274772644043,
      "learning_rate": 4.3400000000000005e-05,
      "loss": 0.9148,
      "step": 218
    },
    {
      "epoch": 0.0436,
      "loss_align": 0.3388627767562866,
      "loss_contrastive": 0.12458407878875732,
      "loss_gen": 0.42343705892562866,
      "neg_sim_mean": 0.6857212781906128,
      "pos_sim_mean": 0.6611372232437134,
      "step": 218
    },
    {
      "epoch": 0.0438,
      "grad_norm": 6.799839496612549,
      "learning_rate": 4.36e-05,
      "loss": 0.7772,
      "step": 219
    },
    {
      "epoch": 0.0438,
      "loss_align": 0.5652725696563721,
      "loss_contrastive": 0.07376101613044739,
      "loss_gen": 0.19646260142326355,
      "neg_sim_mean": 0.4084884524345398,
      "pos_sim_mean": 0.43472743034362793,
      "step": 219
    },
    {
      "epoch": 0.044,
      "grad_norm": 3.526517629623413,
      "learning_rate": 4.38e-05,
      "loss": 0.7706,
      "step": 220
    },
    {
      "epoch": 0.044,
      "loss_align": 0.3360708951950073,
      "loss_contrastive": 0.045284271240234375,
      "loss_gen": 0.45388010144233704,
      "neg_sim_mean": 0.6092133522033691,
      "pos_sim_mean": 0.6639291048049927,
      "step": 220
    },
    {
      "epoch": 0.0442,
      "grad_norm": 5.848195552825928,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.7954,
      "step": 221
    },
    {
      "epoch": 0.0442,
      "loss_align": 0.35201072692871094,
      "loss_contrastive": 0.1155250072479248,
      "loss_gen": 0.29120421409606934,
      "neg_sim_mean": 0.663514256477356,
      "pos_sim_mean": 0.6479892730712891,
      "step": 221
    },
    {
      "epoch": 0.0444,
      "grad_norm": 5.063302516937256,
      "learning_rate": 4.4200000000000004e-05,
      "loss": 0.6571,
      "step": 222
    },
    {
      "epoch": 0.0444,
      "loss_align": 0.49108701944351196,
      "loss_contrastive": 0.20790830254554749,
      "loss_gen": 0.3838474750518799,
      "neg_sim_mean": 0.6168212890625,
      "pos_sim_mean": 0.508912980556488,
      "step": 222
    },
    {
      "epoch": 0.0446,
      "grad_norm": 5.3686628341674805,
      "learning_rate": 4.44e-05,
      "loss": 0.8999,
      "step": 223
    },
    {
      "epoch": 0.0446,
      "loss_align": 0.6108537912368774,
      "loss_contrastive": 0.21872830390930176,
      "loss_gen": 0.19756469130516052,
      "neg_sim_mean": 0.5078745484352112,
      "pos_sim_mean": 0.38914623856544495,
      "step": 223
    },
    {
      "epoch": 0.0448,
      "grad_norm": 5.142086029052734,
      "learning_rate": 4.46e-05,
      "loss": 0.8347,
      "step": 224
    },
    {
      "epoch": 0.0448,
      "loss_align": 0.3787654638290405,
      "loss_contrastive": 0.06892496347427368,
      "loss_gen": 0.32899048924446106,
      "neg_sim_mean": 0.5901594758033752,
      "pos_sim_mean": 0.6212345361709595,
      "step": 224
    },
    {
      "epoch": 0.045,
      "grad_norm": 6.467951774597168,
      "learning_rate": 4.4800000000000005e-05,
      "loss": 0.716,
      "step": 225
    },
    {
      "epoch": 0.045,
      "loss_align": 0.5989740490913391,
      "loss_contrastive": 0.2207367718219757,
      "loss_gen": 0.2234962284564972,
      "neg_sim_mean": 0.5217627286911011,
      "pos_sim_mean": 0.4010259509086609,
      "step": 225
    },
    {
      "epoch": 0.0452,
      "grad_norm": 5.462329387664795,
      "learning_rate": 4.5e-05,
      "loss": 0.849,
      "step": 226
    },
    {
      "epoch": 0.0452,
      "loss_align": 0.411739706993103,
      "loss_contrastive": 0.07298555970191956,
      "loss_gen": 0.26122745871543884,
      "neg_sim_mean": 0.561245858669281,
      "pos_sim_mean": 0.588260293006897,
      "step": 226
    },
    {
      "epoch": 0.0454,
      "grad_norm": 4.922479152679443,
      "learning_rate": 4.52e-05,
      "loss": 0.6817,
      "step": 227
    },
    {
      "epoch": 0.0454,
      "loss_align": 0.4499979019165039,
      "loss_contrastive": 0.019897550344467163,
      "loss_gen": 0.11976142972707748,
      "neg_sim_mean": 0.46989965438842773,
      "pos_sim_mean": 0.5500020980834961,
      "step": 227
    },
    {
      "epoch": 0.0456,
      "grad_norm": 3.930938959121704,
      "learning_rate": 4.5400000000000006e-05,
      "loss": 0.5721,
      "step": 228
    },
    {
      "epoch": 0.0456,
      "loss_align": 0.5086046457290649,
      "loss_contrastive": 0.08088609576225281,
      "loss_gen": 0.11951319128274918,
      "neg_sim_mean": 0.47228145599365234,
      "pos_sim_mean": 0.49139535427093506,
      "step": 228
    },
    {
      "epoch": 0.0458,
      "grad_norm": 3.379380702972412,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 0.6378,
      "step": 229
    },
    {
      "epoch": 0.0458,
      "loss_align": 0.4313257336616516,
      "loss_contrastive": 0.19576922059059143,
      "loss_gen": 0.5617209672927856,
      "neg_sim_mean": 0.6644434928894043,
      "pos_sim_mean": 0.5686742663383484,
      "step": 229
    },
    {
      "epoch": 0.046,
      "grad_norm": 6.384127616882324,
      "learning_rate": 4.58e-05,
      "loss": 1.0165,
      "step": 230
    },
    {
      "epoch": 0.046,
      "loss_align": 0.45461422204971313,
      "loss_contrastive": 0.07654014229774475,
      "loss_gen": 0.19116415083408356,
      "neg_sim_mean": 0.5219259262084961,
      "pos_sim_mean": 0.5453857779502869,
      "step": 230
    },
    {
      "epoch": 0.0462,
      "grad_norm": 4.232696533203125,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.655,
      "step": 231
    },
    {
      "epoch": 0.0462,
      "loss_align": 0.4119494557380676,
      "loss_contrastive": 0.13541820645332336,
      "loss_gen": 0.26615193486213684,
      "neg_sim_mean": 0.6234687566757202,
      "pos_sim_mean": 0.5880505442619324,
      "step": 231
    },
    {
      "epoch": 0.0464,
      "grad_norm": 4.3925909996032715,
      "learning_rate": 4.6200000000000005e-05,
      "loss": 0.6944,
      "step": 232
    },
    {
      "epoch": 0.0464,
      "loss_align": 0.4214956760406494,
      "loss_contrastive": 0.08769652247428894,
      "loss_gen": 0.2582911252975464,
      "neg_sim_mean": 0.566200852394104,
      "pos_sim_mean": 0.5785043239593506,
      "step": 232
    },
    {
      "epoch": 0.0466,
      "grad_norm": 4.265562057495117,
      "learning_rate": 4.64e-05,
      "loss": 0.6903,
      "step": 233
    },
    {
      "epoch": 0.0466,
      "loss_align": 0.6266130805015564,
      "loss_contrastive": 0.2686317265033722,
      "loss_gen": 0.2205888032913208,
      "neg_sim_mean": 0.5420186519622803,
      "pos_sim_mean": 0.3733869194984436,
      "step": 233
    },
    {
      "epoch": 0.0468,
      "grad_norm": 5.174211502075195,
      "learning_rate": 4.660000000000001e-05,
      "loss": 0.8794,
      "step": 234
    },
    {
      "epoch": 0.0468,
      "loss_align": 0.5774863958358765,
      "loss_contrastive": 0.1838034689426422,
      "loss_gen": 0.3201653063297272,
      "neg_sim_mean": 0.5063170790672302,
      "pos_sim_mean": 0.42251360416412354,
      "step": 234
    },
    {
      "epoch": 0.047,
      "grad_norm": 5.27044677734375,
      "learning_rate": 4.6800000000000006e-05,
      "loss": 0.9197,
      "step": 235
    },
    {
      "epoch": 0.047,
      "loss_align": 0.460491418838501,
      "loss_contrastive": 0.14295092225074768,
      "loss_gen": 0.5197129249572754,
      "neg_sim_mean": 0.5824595093727112,
      "pos_sim_mean": 0.539508581161499,
      "step": 235
    },
    {
      "epoch": 0.0472,
      "grad_norm": 6.835518836975098,
      "learning_rate": 4.7e-05,
      "loss": 0.9974,
      "step": 236
    },
    {
      "epoch": 0.0472,
      "loss_align": 0.4083653688430786,
      "loss_contrastive": 0.1630488932132721,
      "loss_gen": 0.2601219117641449,
      "neg_sim_mean": 0.654683530330658,
      "pos_sim_mean": 0.5916346311569214,
      "step": 236
    },
    {
      "epoch": 0.0474,
      "grad_norm": 4.645096778869629,
      "learning_rate": 4.72e-05,
      "loss": 0.6881,
      "step": 237
    },
    {
      "epoch": 0.0474,
      "loss_align": 0.46165943145751953,
      "loss_contrastive": 0.14866796135902405,
      "loss_gen": 0.31034043431282043,
      "neg_sim_mean": 0.587008535861969,
      "pos_sim_mean": 0.5383405685424805,
      "step": 237
    },
    {
      "epoch": 0.0476,
      "grad_norm": 4.073322772979736,
      "learning_rate": 4.74e-05,
      "loss": 0.7898,
      "step": 238
    },
    {
      "epoch": 0.0476,
      "loss_align": 0.36977505683898926,
      "loss_contrastive": 0.10008436441421509,
      "loss_gen": 0.3198358118534088,
      "neg_sim_mean": 0.6303092837333679,
      "pos_sim_mean": 0.6302249431610107,
      "step": 238
    },
    {
      "epoch": 0.0478,
      "grad_norm": 5.563520908355713,
      "learning_rate": 4.76e-05,
      "loss": 0.7016,
      "step": 239
    },
    {
      "epoch": 0.0478,
      "loss_align": 0.5424026250839233,
      "loss_contrastive": 0.1392132043838501,
      "loss_gen": 0.27561110258102417,
      "neg_sim_mean": 0.49681058526039124,
      "pos_sim_mean": 0.45759737491607666,
      "step": 239
    },
    {
      "epoch": 0.048,
      "grad_norm": 5.1413421630859375,
      "learning_rate": 4.78e-05,
      "loss": 0.8347,
      "step": 240
    },
    {
      "epoch": 0.048,
      "loss_align": 0.4093318581581116,
      "loss_contrastive": 0.09441617131233215,
      "loss_gen": 0.25190868973731995,
      "neg_sim_mean": 0.5850843191146851,
      "pos_sim_mean": 0.5906681418418884,
      "step": 240
    },
    {
      "epoch": 0.0482,
      "grad_norm": 4.511435508728027,
      "learning_rate": 4.8e-05,
      "loss": 0.6726,
      "step": 241
    },
    {
      "epoch": 0.0482,
      "loss_align": 0.6466504335403442,
      "loss_contrastive": 0.2366483211517334,
      "loss_gen": 0.3592701554298401,
      "neg_sim_mean": 0.489997923374176,
      "pos_sim_mean": 0.35334959626197815,
      "step": 241
    },
    {
      "epoch": 0.0484,
      "grad_norm": 5.573434352874756,
      "learning_rate": 4.82e-05,
      "loss": 1.0343,
      "step": 242
    },
    {
      "epoch": 0.0484,
      "loss_align": 0.5739877223968506,
      "loss_contrastive": 0.0822533369064331,
      "loss_gen": 0.24384821951389313,
      "neg_sim_mean": 0.408265620470047,
      "pos_sim_mean": 0.4260122776031494,
      "step": 242
    },
    {
      "epoch": 0.0486,
      "grad_norm": 5.54188871383667,
      "learning_rate": 4.8400000000000004e-05,
      "loss": 0.8277,
      "step": 243
    },
    {
      "epoch": 0.0486,
      "loss_align": 0.32498717308044434,
      "loss_contrastive": 0.12468171119689941,
      "loss_gen": 0.2890492379665375,
      "neg_sim_mean": 0.6996945142745972,
      "pos_sim_mean": 0.6750128269195557,
      "step": 243
    },
    {
      "epoch": 0.0488,
      "grad_norm": 4.54641580581665,
      "learning_rate": 4.86e-05,
      "loss": 0.629,
      "step": 244
    },
    {
      "epoch": 0.0488,
      "loss_align": 0.614560604095459,
      "loss_contrastive": 0.13533315062522888,
      "loss_gen": 0.2084585577249527,
      "neg_sim_mean": 0.4207725524902344,
      "pos_sim_mean": 0.385439395904541,
      "step": 244
    },
    {
      "epoch": 0.049,
      "grad_norm": 4.8028178215026855,
      "learning_rate": 4.88e-05,
      "loss": 0.8393,
      "step": 245
    },
    {
      "epoch": 0.049,
      "loss_align": 0.356687068939209,
      "loss_contrastive": 0.0,
      "loss_gen": 0.08213675767183304,
      "neg_sim_mean": 0.5180912017822266,
      "pos_sim_mean": 0.643312931060791,
      "step": 245
    },
    {
      "epoch": 0.0492,
      "grad_norm": 4.882893085479736,
      "learning_rate": 4.9e-05,
      "loss": 0.4388,
      "step": 246
    },
    {
      "epoch": 0.0492,
      "loss_align": 0.42453014850616455,
      "loss_contrastive": 0.04599890112876892,
      "loss_gen": 0.045828286558389664,
      "neg_sim_mean": 0.5214687585830688,
      "pos_sim_mean": 0.5754698514938354,
      "step": 246
    },
    {
      "epoch": 0.0494,
      "grad_norm": 3.4083759784698486,
      "learning_rate": 4.92e-05,
      "loss": 0.4759,
      "step": 247
    },
    {
      "epoch": 0.0494,
      "loss_align": 0.5257960557937622,
      "loss_contrastive": 0.202639639377594,
      "loss_gen": 0.2958846688270569,
      "neg_sim_mean": 0.5768435597419739,
      "pos_sim_mean": 0.4742039144039154,
      "step": 247
    },
    {
      "epoch": 0.0496,
      "grad_norm": 6.893433094024658,
      "learning_rate": 4.94e-05,
      "loss": 0.846,
      "step": 248
    },
    {
      "epoch": 0.0496,
      "loss_align": 0.42617738246917725,
      "loss_contrastive": 0.24319961667060852,
      "loss_gen": 0.2520601451396942,
      "neg_sim_mean": 0.7170222401618958,
      "pos_sim_mean": 0.5738226175308228,
      "step": 248
    },
    {
      "epoch": 0.0498,
      "grad_norm": 4.866425514221191,
      "learning_rate": 4.96e-05,
      "loss": 0.7074,
      "step": 249
    },
    {
      "epoch": 0.0498,
      "loss_align": 0.5471271276473999,
      "loss_contrastive": 0.22253307700157166,
      "loss_gen": 0.24347712099552155,
      "neg_sim_mean": 0.5754059553146362,
      "pos_sim_mean": 0.4528728723526001,
      "step": 249
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.737672805786133,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 0.8173,
      "step": 250
    },
    {
      "epoch": 0.05,
      "loss_align": 0.481001079082489,
      "loss_contrastive": 0.05573800206184387,
      "loss_gen": 0.32801729440689087,
      "neg_sim_mean": 0.47473692893981934,
      "pos_sim_mean": 0.518998920917511,
      "step": 250
    },
    {
      "epoch": 0.0502,
      "grad_norm": 4.703484535217285,
      "learning_rate": 5e-05,
      "loss": 0.8157,
      "step": 251
    },
    {
      "epoch": 0.0502,
      "loss_align": 0.4996238350868225,
      "loss_contrastive": 0.14242246747016907,
      "loss_gen": 0.41569769382476807,
      "neg_sim_mean": 0.542798638343811,
      "pos_sim_mean": 0.5003761649131775,
      "step": 251
    },
    {
      "epoch": 0.0504,
      "grad_norm": 5.284154891967773,
      "learning_rate": 5.02e-05,
      "loss": 0.9324,
      "step": 252
    },
    {
      "epoch": 0.0504,
      "loss_align": 0.5890151858329773,
      "loss_contrastive": 0.15849655866622925,
      "loss_gen": 0.30204111337661743,
      "neg_sim_mean": 0.46948137879371643,
      "pos_sim_mean": 0.4109848141670227,
      "step": 252
    },
    {
      "epoch": 0.0506,
      "grad_norm": 4.875269889831543,
      "learning_rate": 5.0400000000000005e-05,
      "loss": 0.9101,
      "step": 253
    },
    {
      "epoch": 0.0506,
      "loss_align": 0.5081731081008911,
      "loss_contrastive": 0.08616387844085693,
      "loss_gen": 0.23367026448249817,
      "neg_sim_mean": 0.4779908061027527,
      "pos_sim_mean": 0.4918269217014313,
      "step": 253
    },
    {
      "epoch": 0.0508,
      "grad_norm": 4.689417362213135,
      "learning_rate": 5.0600000000000003e-05,
      "loss": 0.7522,
      "step": 254
    },
    {
      "epoch": 0.0508,
      "loss_align": 0.32942134141921997,
      "loss_contrastive": 0.0,
      "loss_gen": 0.016742128878831863,
      "neg_sim_mean": 0.5276771783828735,
      "pos_sim_mean": 0.67057865858078,
      "step": 254
    },
    {
      "epoch": 0.051,
      "grad_norm": 1.8705847263336182,
      "learning_rate": 5.08e-05,
      "loss": 0.3462,
      "step": 255
    },
    {
      "epoch": 0.051,
      "loss_align": 0.5240673422813416,
      "loss_contrastive": 0.0840815007686615,
      "loss_gen": 0.18703612685203552,
      "neg_sim_mean": 0.4600141644477844,
      "pos_sim_mean": 0.47593265771865845,
      "step": 255
    },
    {
      "epoch": 0.0512,
      "grad_norm": 4.508375644683838,
      "learning_rate": 5.1000000000000006e-05,
      "loss": 0.7212,
      "step": 256
    },
    {
      "epoch": 0.0512,
      "loss_align": 0.42827099561691284,
      "loss_contrastive": 0.028737574815750122,
      "loss_gen": 0.35315632820129395,
      "neg_sim_mean": 0.5004665851593018,
      "pos_sim_mean": 0.5717290043830872,
      "step": 256
    },
    {
      "epoch": 0.0514,
      "grad_norm": 6.479555606842041,
      "learning_rate": 5.1200000000000004e-05,
      "loss": 0.7849,
      "step": 257
    },
    {
      "epoch": 0.0514,
      "loss_align": 0.39881056547164917,
      "loss_contrastive": 0.0030194520950317383,
      "loss_gen": 0.3644055724143982,
      "neg_sim_mean": 0.5042088627815247,
      "pos_sim_mean": 0.6011894345283508,
      "step": 257
    },
    {
      "epoch": 0.0516,
      "grad_norm": 4.6571044921875,
      "learning_rate": 5.14e-05,
      "loss": 0.7636,
      "step": 258
    },
    {
      "epoch": 0.0516,
      "loss_align": 0.39381474256515503,
      "loss_contrastive": 0.06924396753311157,
      "loss_gen": 0.3451721668243408,
      "neg_sim_mean": 0.5754292011260986,
      "pos_sim_mean": 0.606185257434845,
      "step": 258
    },
    {
      "epoch": 0.0518,
      "grad_norm": 5.305119037628174,
      "learning_rate": 5.16e-05,
      "loss": 0.7473,
      "step": 259
    },
    {
      "epoch": 0.0518,
      "loss_align": 0.4239400625228882,
      "loss_contrastive": 0.1292162835597992,
      "loss_gen": 0.22204053401947021,
      "neg_sim_mean": 0.6052762269973755,
      "pos_sim_mean": 0.5760599374771118,
      "step": 259
    },
    {
      "epoch": 0.052,
      "grad_norm": 5.2480363845825195,
      "learning_rate": 5.1800000000000005e-05,
      "loss": 0.6615,
      "step": 260
    },
    {
      "epoch": 0.052,
      "loss_align": 0.41434401273727417,
      "loss_contrastive": 0.03853300213813782,
      "loss_gen": 0.02386961132287979,
      "neg_sim_mean": 0.5241889953613281,
      "pos_sim_mean": 0.5856559872627258,
      "step": 260
    },
    {
      "epoch": 0.0522,
      "grad_norm": 2.6282272338867188,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 0.4428,
      "step": 261
    },
    {
      "epoch": 0.0522,
      "loss_align": 0.3494071960449219,
      "loss_contrastive": 0.10118973255157471,
      "loss_gen": 0.3735581934452057,
      "neg_sim_mean": 0.6517825126647949,
      "pos_sim_mean": 0.6505928039550781,
      "step": 261
    },
    {
      "epoch": 0.0524,
      "grad_norm": 5.018040180206299,
      "learning_rate": 5.22e-05,
      "loss": 0.7351,
      "step": 262
    },
    {
      "epoch": 0.0524,
      "loss_align": 0.4548985958099365,
      "loss_contrastive": 0.1486181914806366,
      "loss_gen": 0.29202476143836975,
      "neg_sim_mean": 0.5937196016311646,
      "pos_sim_mean": 0.5451014041900635,
      "step": 262
    },
    {
      "epoch": 0.0526,
      "grad_norm": 5.506120204925537,
      "learning_rate": 5.2400000000000007e-05,
      "loss": 0.7648,
      "step": 263
    },
    {
      "epoch": 0.0526,
      "loss_align": 0.33343183994293213,
      "loss_contrastive": 0.10613828897476196,
      "loss_gen": 0.23484091460704803,
      "neg_sim_mean": 0.6727064251899719,
      "pos_sim_mean": 0.6665681600570679,
      "step": 263
    },
    {
      "epoch": 0.0528,
      "grad_norm": 4.80199670791626,
      "learning_rate": 5.2600000000000005e-05,
      "loss": 0.581,
      "step": 264
    },
    {
      "epoch": 0.0528,
      "loss_align": 0.4052566885948181,
      "loss_contrastive": 0.1249859631061554,
      "loss_gen": 0.47244834899902344,
      "neg_sim_mean": 0.6197292804718018,
      "pos_sim_mean": 0.5947433114051819,
      "step": 264
    },
    {
      "epoch": 0.053,
      "grad_norm": 5.156983375549316,
      "learning_rate": 5.28e-05,
      "loss": 0.8927,
      "step": 265
    },
    {
      "epoch": 0.053,
      "loss_align": 0.46728968620300293,
      "loss_contrastive": 0.16301807761192322,
      "loss_gen": 0.3115480840206146,
      "neg_sim_mean": 0.5957283973693848,
      "pos_sim_mean": 0.5327103137969971,
      "step": 265
    },
    {
      "epoch": 0.0532,
      "grad_norm": 4.972663879394531,
      "learning_rate": 5.300000000000001e-05,
      "loss": 0.7984,
      "step": 266
    },
    {
      "epoch": 0.0532,
      "loss_align": 0.43015652894973755,
      "loss_contrastive": 0.04194048047065735,
      "loss_gen": 0.21069566905498505,
      "neg_sim_mean": 0.5117839574813843,
      "pos_sim_mean": 0.5698434710502625,
      "step": 266
    },
    {
      "epoch": 0.0534,
      "grad_norm": 3.990090847015381,
      "learning_rate": 5.3200000000000006e-05,
      "loss": 0.6459,
      "step": 267
    },
    {
      "epoch": 0.0534,
      "loss_align": 0.38078582286834717,
      "loss_contrastive": 0.09226131439208984,
      "loss_gen": 0.23020076751708984,
      "neg_sim_mean": 0.6114754676818848,
      "pos_sim_mean": 0.6192141771316528,
      "step": 267
    },
    {
      "epoch": 0.0536,
      "grad_norm": 3.867818593978882,
      "learning_rate": 5.3400000000000004e-05,
      "loss": 0.6221,
      "step": 268
    },
    {
      "epoch": 0.0536,
      "loss_align": 0.39416176080703735,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013133194297552109,
      "neg_sim_mean": 0.49985599517822266,
      "pos_sim_mean": 0.6058382391929626,
      "step": 268
    },
    {
      "epoch": 0.0538,
      "grad_norm": 1.6620198488235474,
      "learning_rate": 5.360000000000001e-05,
      "loss": 0.4073,
      "step": 269
    },
    {
      "epoch": 0.0538,
      "loss_align": 0.4961317777633667,
      "loss_contrastive": 0.10660484433174133,
      "loss_gen": 0.158174067735672,
      "neg_sim_mean": 0.5104730725288391,
      "pos_sim_mean": 0.5038682222366333,
      "step": 269
    },
    {
      "epoch": 0.054,
      "grad_norm": 4.6255998611450195,
      "learning_rate": 5.380000000000001e-05,
      "loss": 0.6671,
      "step": 270
    },
    {
      "epoch": 0.054,
      "loss_align": 0.469959020614624,
      "loss_contrastive": 0.08540597558021545,
      "loss_gen": 0.4723392426967621,
      "neg_sim_mean": 0.5154469609260559,
      "pos_sim_mean": 0.530040979385376,
      "step": 270
    },
    {
      "epoch": 0.0542,
      "grad_norm": 5.432751655578613,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 0.9525,
      "step": 271
    },
    {
      "epoch": 0.0542,
      "loss_align": 0.30330127477645874,
      "loss_contrastive": 0.059406816959381104,
      "loss_gen": 0.30023789405822754,
      "neg_sim_mean": 0.6561055183410645,
      "pos_sim_mean": 0.6966987252235413,
      "step": 271
    },
    {
      "epoch": 0.0544,
      "grad_norm": 4.852248668670654,
      "learning_rate": 5.420000000000001e-05,
      "loss": 0.6107,
      "step": 272
    },
    {
      "epoch": 0.0544,
      "loss_align": 0.4635313153266907,
      "loss_contrastive": 0.179774671792984,
      "loss_gen": 0.23081180453300476,
      "neg_sim_mean": 0.6162433624267578,
      "pos_sim_mean": 0.5364686846733093,
      "step": 272
    },
    {
      "epoch": 0.0546,
      "grad_norm": 4.976598739624023,
      "learning_rate": 5.440000000000001e-05,
      "loss": 0.7159,
      "step": 273
    },
    {
      "epoch": 0.0546,
      "loss_align": 0.4917494058609009,
      "loss_contrastive": 0.09121090173721313,
      "loss_gen": 0.009099382907152176,
      "neg_sim_mean": 0.49946150183677673,
      "pos_sim_mean": 0.5082505941390991,
      "step": 273
    },
    {
      "epoch": 0.0548,
      "grad_norm": 1.3662452697753906,
      "learning_rate": 5.4600000000000006e-05,
      "loss": 0.5118,
      "step": 274
    },
    {
      "epoch": 0.0548,
      "loss_align": 0.38475853204727173,
      "loss_contrastive": 0.03613311052322388,
      "loss_gen": 0.33630484342575073,
      "neg_sim_mean": 0.5513745546340942,
      "pos_sim_mean": 0.6152414679527283,
      "step": 274
    },
    {
      "epoch": 0.055,
      "grad_norm": 4.683475971221924,
      "learning_rate": 5.4800000000000004e-05,
      "loss": 0.7254,
      "step": 275
    },
    {
      "epoch": 0.055,
      "loss_align": 0.36632347106933594,
      "loss_contrastive": 0.06235063076019287,
      "loss_gen": 0.1921115517616272,
      "neg_sim_mean": 0.596027135848999,
      "pos_sim_mean": 0.6336765289306641,
      "step": 275
    },
    {
      "epoch": 0.0552,
      "grad_norm": 4.331364631652832,
      "learning_rate": 5.500000000000001e-05,
      "loss": 0.5659,
      "step": 276
    },
    {
      "epoch": 0.0552,
      "loss_align": 0.34376585483551025,
      "loss_contrastive": 0.013016819953918457,
      "loss_gen": 0.29203400015830994,
      "neg_sim_mean": 0.5692509412765503,
      "pos_sim_mean": 0.6562341451644897,
      "step": 276
    },
    {
      "epoch": 0.0554,
      "grad_norm": 4.276541233062744,
      "learning_rate": 5.520000000000001e-05,
      "loss": 0.6374,
      "step": 277
    },
    {
      "epoch": 0.0554,
      "loss_align": 0.3854435086250305,
      "loss_contrastive": 0.0,
      "loss_gen": 0.023721996694803238,
      "neg_sim_mean": 0.49419763684272766,
      "pos_sim_mean": 0.6145564913749695,
      "step": 277
    },
    {
      "epoch": 0.0556,
      "grad_norm": 2.3585269451141357,
      "learning_rate": 5.5400000000000005e-05,
      "loss": 0.4092,
      "step": 278
    },
    {
      "epoch": 0.0556,
      "loss_align": 0.3805563449859619,
      "loss_contrastive": 0.0368083119392395,
      "loss_gen": 0.4072188436985016,
      "neg_sim_mean": 0.5562519431114197,
      "pos_sim_mean": 0.6194436550140381,
      "step": 278
    },
    {
      "epoch": 0.0558,
      "grad_norm": 5.382613182067871,
      "learning_rate": 5.560000000000001e-05,
      "loss": 0.7922,
      "step": 279
    },
    {
      "epoch": 0.0558,
      "loss_align": 0.45411795377731323,
      "loss_contrastive": 0.048998624086380005,
      "loss_gen": 0.3856373131275177,
      "neg_sim_mean": 0.49488067626953125,
      "pos_sim_mean": 0.5458820462226868,
      "step": 279
    },
    {
      "epoch": 0.056,
      "grad_norm": 5.145069122314453,
      "learning_rate": 5.580000000000001e-05,
      "loss": 0.8456,
      "step": 280
    },
    {
      "epoch": 0.056,
      "loss_align": 0.46554064750671387,
      "loss_contrastive": 0.04531696438789368,
      "loss_gen": 0.4121096730232239,
      "neg_sim_mean": 0.4797763228416443,
      "pos_sim_mean": 0.5344593524932861,
      "step": 280
    },
    {
      "epoch": 0.0562,
      "grad_norm": 4.6745758056640625,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 0.8831,
      "step": 281
    },
    {
      "epoch": 0.0562,
      "loss_align": 0.3882499933242798,
      "loss_contrastive": 0.05017364025115967,
      "loss_gen": 0.43025121092796326,
      "neg_sim_mean": 0.561923623085022,
      "pos_sim_mean": 0.6117500066757202,
      "step": 281
    },
    {
      "epoch": 0.0564,
      "grad_norm": 5.610178470611572,
      "learning_rate": 5.620000000000001e-05,
      "loss": 0.8245,
      "step": 282
    },
    {
      "epoch": 0.0564,
      "loss_align": 0.309329628944397,
      "loss_contrastive": 0.05165362358093262,
      "loss_gen": 0.30018460750579834,
      "neg_sim_mean": 0.6423239707946777,
      "pos_sim_mean": 0.690670371055603,
      "step": 282
    },
    {
      "epoch": 0.0566,
      "grad_norm": 4.295063495635986,
      "learning_rate": 5.6399999999999995e-05,
      "loss": 0.6157,
      "step": 283
    },
    {
      "epoch": 0.0566,
      "loss_align": 0.45340174436569214,
      "loss_contrastive": 0.09421971440315247,
      "loss_gen": 0.22427253425121307,
      "neg_sim_mean": 0.5408179759979248,
      "pos_sim_mean": 0.5465982556343079,
      "step": 283
    },
    {
      "epoch": 0.0568,
      "grad_norm": 4.566743850708008,
      "learning_rate": 5.66e-05,
      "loss": 0.689,
      "step": 284
    },
    {
      "epoch": 0.0568,
      "loss_align": 0.47805851697921753,
      "loss_contrastive": 0.0690416693687439,
      "loss_gen": 0.00602889247238636,
      "neg_sim_mean": 0.49098315834999084,
      "pos_sim_mean": 0.5219414830207825,
      "step": 284
    },
    {
      "epoch": 0.057,
      "grad_norm": 1.1359729766845703,
      "learning_rate": 5.68e-05,
      "loss": 0.4924,
      "step": 285
    },
    {
      "epoch": 0.057,
      "loss_align": 0.31669890880584717,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010499703697860241,
      "neg_sim_mean": 0.49416470527648926,
      "pos_sim_mean": 0.6833010911941528,
      "step": 285
    },
    {
      "epoch": 0.0572,
      "grad_norm": 1.4718316793441772,
      "learning_rate": 5.6999999999999996e-05,
      "loss": 0.3272,
      "step": 286
    },
    {
      "epoch": 0.0572,
      "loss_align": 0.6288338303565979,
      "loss_contrastive": 0.20569661259651184,
      "loss_gen": 0.259816437959671,
      "neg_sim_mean": 0.4768627882003784,
      "pos_sim_mean": 0.3711661696434021,
      "step": 286
    },
    {
      "epoch": 0.0574,
      "grad_norm": 5.241859436035156,
      "learning_rate": 5.72e-05,
      "loss": 0.9133,
      "step": 287
    },
    {
      "epoch": 0.0574,
      "loss_align": 0.6209638714790344,
      "loss_contrastive": 0.17714637517929077,
      "loss_gen": 0.3145131766796112,
      "neg_sim_mean": 0.4561825096607208,
      "pos_sim_mean": 0.3790361285209656,
      "step": 287
    },
    {
      "epoch": 0.0576,
      "grad_norm": 5.219918727874756,
      "learning_rate": 5.74e-05,
      "loss": 0.9567,
      "step": 288
    },
    {
      "epoch": 0.0576,
      "loss_align": 0.3460838198661804,
      "loss_contrastive": 0.044139087200164795,
      "loss_gen": 0.21821516752243042,
      "neg_sim_mean": 0.5980552434921265,
      "pos_sim_mean": 0.6539161801338196,
      "step": 288
    },
    {
      "epoch": 0.0578,
      "grad_norm": 5.375040054321289,
      "learning_rate": 5.76e-05,
      "loss": 0.5696,
      "step": 289
    },
    {
      "epoch": 0.0578,
      "loss_align": 0.43071091175079346,
      "loss_contrastive": 0.16556110978126526,
      "loss_gen": 0.38559967279434204,
      "neg_sim_mean": 0.6348502039909363,
      "pos_sim_mean": 0.5692890882492065,
      "step": 289
    },
    {
      "epoch": 0.058,
      "grad_norm": 5.675469875335693,
      "learning_rate": 5.7799999999999995e-05,
      "loss": 0.8362,
      "step": 290
    },
    {
      "epoch": 0.058,
      "loss_align": 0.4360924959182739,
      "loss_contrastive": 0.03387758135795593,
      "loss_gen": 0.002976140007376671,
      "neg_sim_mean": 0.4977850914001465,
      "pos_sim_mean": 0.5639075040817261,
      "step": 290
    },
    {
      "epoch": 0.0582,
      "grad_norm": 0.9267215728759766,
      "learning_rate": 5.8e-05,
      "loss": 0.4431,
      "step": 291
    },
    {
      "epoch": 0.0582,
      "loss_align": 0.5341778993606567,
      "loss_contrastive": 0.16504326462745667,
      "loss_gen": 0.2934609651565552,
      "neg_sim_mean": 0.5308653712272644,
      "pos_sim_mean": 0.46582210063934326,
      "step": 291
    },
    {
      "epoch": 0.0584,
      "grad_norm": 5.148142337799072,
      "learning_rate": 5.82e-05,
      "loss": 0.8474,
      "step": 292
    },
    {
      "epoch": 0.0584,
      "loss_align": 0.3558098077774048,
      "loss_contrastive": 0.0,
      "loss_gen": 0.33001086115837097,
      "neg_sim_mean": 0.5323160290718079,
      "pos_sim_mean": 0.6441901922225952,
      "step": 292
    },
    {
      "epoch": 0.0586,
      "grad_norm": 5.427891254425049,
      "learning_rate": 5.8399999999999997e-05,
      "loss": 0.6858,
      "step": 293
    },
    {
      "epoch": 0.0586,
      "loss_align": 0.5044517517089844,
      "loss_contrastive": 0.11915528774261475,
      "loss_gen": 0.2957059144973755,
      "neg_sim_mean": 0.5147035121917725,
      "pos_sim_mean": 0.49554821848869324,
      "step": 293
    },
    {
      "epoch": 0.0588,
      "grad_norm": 5.341741561889648,
      "learning_rate": 5.86e-05,
      "loss": 0.8145,
      "step": 294
    },
    {
      "epoch": 0.0588,
      "loss_align": 0.3823840618133545,
      "loss_contrastive": 0.011786341667175293,
      "loss_gen": 0.3713425099849701,
      "neg_sim_mean": 0.5294022560119629,
      "pos_sim_mean": 0.6176159381866455,
      "step": 294
    },
    {
      "epoch": 0.059,
      "grad_norm": 6.125507831573486,
      "learning_rate": 5.88e-05,
      "loss": 0.7551,
      "step": 295
    },
    {
      "epoch": 0.059,
      "loss_align": 0.25980836153030396,
      "loss_contrastive": 0.06558889150619507,
      "loss_gen": 0.47227105498313904,
      "neg_sim_mean": 0.7057805061340332,
      "pos_sim_mean": 0.740191638469696,
      "step": 295
    },
    {
      "epoch": 0.0592,
      "grad_norm": 6.229730606079102,
      "learning_rate": 5.9e-05,
      "loss": 0.74,
      "step": 296
    },
    {
      "epoch": 0.0592,
      "loss_align": 0.5434862971305847,
      "loss_contrastive": 0.08262965083122253,
      "loss_gen": 0.450653076171875,
      "neg_sim_mean": 0.4391433596611023,
      "pos_sim_mean": 0.4565137028694153,
      "step": 296
    },
    {
      "epoch": 0.0594,
      "grad_norm": 7.525043487548828,
      "learning_rate": 5.92e-05,
      "loss": 1.0041,
      "step": 297
    },
    {
      "epoch": 0.0594,
      "loss_align": 0.4507449269294739,
      "loss_contrastive": 0.20027926564216614,
      "loss_gen": 0.2993481755256653,
      "neg_sim_mean": 0.6495343446731567,
      "pos_sim_mean": 0.5492550730705261,
      "step": 297
    },
    {
      "epoch": 0.0596,
      "grad_norm": 5.694822788238525,
      "learning_rate": 5.94e-05,
      "loss": 0.7741,
      "step": 298
    },
    {
      "epoch": 0.0596,
      "loss_align": 0.4338114261627197,
      "loss_contrastive": 0.09122392535209656,
      "loss_gen": 0.2834455966949463,
      "neg_sim_mean": 0.5574125051498413,
      "pos_sim_mean": 0.5661885738372803,
      "step": 298
    },
    {
      "epoch": 0.0598,
      "grad_norm": 4.5995635986328125,
      "learning_rate": 5.96e-05,
      "loss": 0.7282,
      "step": 299
    },
    {
      "epoch": 0.0598,
      "loss_align": 0.4437243938446045,
      "loss_contrastive": 0.14663168787956238,
      "loss_gen": 0.20131410658359528,
      "neg_sim_mean": 0.6029072999954224,
      "pos_sim_mean": 0.5562756061553955,
      "step": 299
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.595040798187256,
      "learning_rate": 5.9800000000000003e-05,
      "loss": 0.6626,
      "step": 300
    },
    {
      "epoch": 0.06,
      "loss_align": 0.530337929725647,
      "loss_contrastive": 0.10590624809265137,
      "loss_gen": 0.3063012957572937,
      "neg_sim_mean": 0.47556835412979126,
      "pos_sim_mean": 0.4696621000766754,
      "step": 300
    },
    {
      "epoch": 0.0602,
      "grad_norm": 4.424520015716553,
      "learning_rate": 6e-05,
      "loss": 0.8493,
      "step": 301
    },
    {
      "epoch": 0.0602,
      "loss_align": 0.27902674674987793,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03605237230658531,
      "neg_sim_mean": 0.47069212794303894,
      "pos_sim_mean": 0.7209732532501221,
      "step": 301
    },
    {
      "epoch": 0.0604,
      "grad_norm": 2.504791498184204,
      "learning_rate": 6.02e-05,
      "loss": 0.3151,
      "step": 302
    },
    {
      "epoch": 0.0604,
      "loss_align": 0.5610643029212952,
      "loss_contrastive": 0.20361486077308655,
      "loss_gen": 0.12760688364505768,
      "neg_sim_mean": 0.5425505638122559,
      "pos_sim_mean": 0.43893569707870483,
      "step": 302
    },
    {
      "epoch": 0.0606,
      "grad_norm": 3.626265525817871,
      "learning_rate": 6.04e-05,
      "loss": 0.7131,
      "step": 303
    },
    {
      "epoch": 0.0606,
      "loss_align": 0.46864646673202515,
      "loss_contrastive": 0.10907384753227234,
      "loss_gen": 0.22868120670318604,
      "neg_sim_mean": 0.5404273867607117,
      "pos_sim_mean": 0.5313535332679749,
      "step": 303
    },
    {
      "epoch": 0.0608,
      "grad_norm": 4.49381685256958,
      "learning_rate": 6.06e-05,
      "loss": 0.7104,
      "step": 304
    },
    {
      "epoch": 0.0608,
      "loss_align": 0.3932090997695923,
      "loss_contrastive": 0.10989546775817871,
      "loss_gen": 0.38171565532684326,
      "neg_sim_mean": 0.6166863441467285,
      "pos_sim_mean": 0.6067909002304077,
      "step": 304
    },
    {
      "epoch": 0.061,
      "grad_norm": 6.9596357345581055,
      "learning_rate": 6.08e-05,
      "loss": 0.7881,
      "step": 305
    },
    {
      "epoch": 0.061,
      "loss_align": 0.49088162183761597,
      "loss_contrastive": 0.08579429984092712,
      "loss_gen": 0.20831529796123505,
      "neg_sim_mean": 0.49491268396377563,
      "pos_sim_mean": 0.509118378162384,
      "step": 305
    },
    {
      "epoch": 0.0612,
      "grad_norm": 4.13602352142334,
      "learning_rate": 6.1e-05,
      "loss": 0.7095,
      "step": 306
    },
    {
      "epoch": 0.0612,
      "loss_align": 0.5644314289093018,
      "loss_contrastive": 0.08270826935768127,
      "loss_gen": 0.19467021524906158,
      "neg_sim_mean": 0.418276846408844,
      "pos_sim_mean": 0.43556857109069824,
      "step": 306
    },
    {
      "epoch": 0.0614,
      "grad_norm": 3.6161742210388184,
      "learning_rate": 6.12e-05,
      "loss": 0.769,
      "step": 307
    },
    {
      "epoch": 0.0614,
      "loss_align": 0.423122763633728,
      "loss_contrastive": 0.17348244786262512,
      "loss_gen": 0.30935657024383545,
      "neg_sim_mean": 0.6503596901893616,
      "pos_sim_mean": 0.576877236366272,
      "step": 307
    },
    {
      "epoch": 0.0616,
      "grad_norm": 4.900595664978027,
      "learning_rate": 6.14e-05,
      "loss": 0.7533,
      "step": 308
    },
    {
      "epoch": 0.0616,
      "loss_align": 0.3430370092391968,
      "loss_contrastive": 0.09688675403594971,
      "loss_gen": 0.39436399936676025,
      "neg_sim_mean": 0.653849720954895,
      "pos_sim_mean": 0.6569629907608032,
      "step": 308
    },
    {
      "epoch": 0.0618,
      "grad_norm": 4.536523818969727,
      "learning_rate": 6.16e-05,
      "loss": 0.749,
      "step": 309
    },
    {
      "epoch": 0.0618,
      "loss_align": 0.521238386631012,
      "loss_contrastive": 0.16906949877738953,
      "loss_gen": 0.22365009784698486,
      "neg_sim_mean": 0.547831118106842,
      "pos_sim_mean": 0.47876161336898804,
      "step": 309
    },
    {
      "epoch": 0.062,
      "grad_norm": 5.709066867828369,
      "learning_rate": 6.18e-05,
      "loss": 0.7652,
      "step": 310
    },
    {
      "epoch": 0.062,
      "loss_align": 0.600702702999115,
      "loss_contrastive": 0.14551496505737305,
      "loss_gen": 0.32096418738365173,
      "neg_sim_mean": 0.44481226801872253,
      "pos_sim_mean": 0.399297297000885,
      "step": 310
    },
    {
      "epoch": 0.0622,
      "grad_norm": 4.535719871520996,
      "learning_rate": 6.2e-05,
      "loss": 0.9391,
      "step": 311
    },
    {
      "epoch": 0.0622,
      "loss_align": 0.448344349861145,
      "loss_contrastive": 0.03282618522644043,
      "loss_gen": 0.035047952085733414,
      "neg_sim_mean": 0.4844818413257599,
      "pos_sim_mean": 0.551655650138855,
      "step": 311
    },
    {
      "epoch": 0.0624,
      "grad_norm": 3.2128710746765137,
      "learning_rate": 6.220000000000001e-05,
      "loss": 0.4873,
      "step": 312
    },
    {
      "epoch": 0.0624,
      "loss_align": 0.4253150224685669,
      "loss_contrastive": 0.05562189221382141,
      "loss_gen": 0.3526041507720947,
      "neg_sim_mean": 0.530306875705719,
      "pos_sim_mean": 0.5746849775314331,
      "step": 312
    },
    {
      "epoch": 0.0626,
      "grad_norm": 5.638452053070068,
      "learning_rate": 6.24e-05,
      "loss": 0.7846,
      "step": 313
    },
    {
      "epoch": 0.0626,
      "loss_align": 0.5848060846328735,
      "loss_contrastive": 0.10045161843299866,
      "loss_gen": 0.26459601521492004,
      "neg_sim_mean": 0.4156455397605896,
      "pos_sim_mean": 0.41519391536712646,
      "step": 313
    },
    {
      "epoch": 0.0628,
      "grad_norm": 5.103758335113525,
      "learning_rate": 6.26e-05,
      "loss": 0.8615,
      "step": 314
    },
    {
      "epoch": 0.0628,
      "loss_align": 0.42542803287506104,
      "loss_contrastive": 0.011058181524276733,
      "loss_gen": 0.029108887538313866,
      "neg_sim_mean": 0.4856301546096802,
      "pos_sim_mean": 0.574571967124939,
      "step": 314
    },
    {
      "epoch": 0.063,
      "grad_norm": 2.90736985206604,
      "learning_rate": 6.280000000000001e-05,
      "loss": 0.4559,
      "step": 315
    },
    {
      "epoch": 0.063,
      "loss_align": 0.3830270767211914,
      "loss_contrastive": 0.07869541645050049,
      "loss_gen": 0.19960369169712067,
      "neg_sim_mean": 0.5956683158874512,
      "pos_sim_mean": 0.6169729232788086,
      "step": 315
    },
    {
      "epoch": 0.0632,
      "grad_norm": 4.186357498168945,
      "learning_rate": 6.3e-05,
      "loss": 0.5921,
      "step": 316
    },
    {
      "epoch": 0.0632,
      "loss_align": 0.47197723388671875,
      "loss_contrastive": 0.052725523710250854,
      "loss_gen": 0.40882056951522827,
      "neg_sim_mean": 0.4807482957839966,
      "pos_sim_mean": 0.5280227661132812,
      "step": 316
    },
    {
      "epoch": 0.0634,
      "grad_norm": 5.690145015716553,
      "learning_rate": 6.32e-05,
      "loss": 0.8871,
      "step": 317
    },
    {
      "epoch": 0.0634,
      "loss_align": 0.43605685234069824,
      "loss_contrastive": 0.02707543969154358,
      "loss_gen": 0.31867504119873047,
      "neg_sim_mean": 0.4910185933113098,
      "pos_sim_mean": 0.5639431476593018,
      "step": 317
    },
    {
      "epoch": 0.0636,
      "grad_norm": 4.238920211791992,
      "learning_rate": 6.340000000000001e-05,
      "loss": 0.758,
      "step": 318
    },
    {
      "epoch": 0.0636,
      "loss_align": 0.33419734239578247,
      "loss_contrastive": 0.08587521314620972,
      "loss_gen": 0.6526414155960083,
      "neg_sim_mean": 0.6516778469085693,
      "pos_sim_mean": 0.6658026576042175,
      "step": 318
    },
    {
      "epoch": 0.0638,
      "grad_norm": 6.058183670043945,
      "learning_rate": 6.36e-05,
      "loss": 0.9971,
      "step": 319
    },
    {
      "epoch": 0.0638,
      "loss_align": 0.5518919825553894,
      "loss_contrastive": 0.11067241430282593,
      "loss_gen": 0.24513131380081177,
      "neg_sim_mean": 0.458780437707901,
      "pos_sim_mean": 0.4481080174446106,
      "step": 319
    },
    {
      "epoch": 0.064,
      "grad_norm": 6.029996395111084,
      "learning_rate": 6.38e-05,
      "loss": 0.8103,
      "step": 320
    },
    {
      "epoch": 0.064,
      "loss_align": 0.4657469391822815,
      "loss_contrastive": 0.041497111320495605,
      "loss_gen": 0.3455536663532257,
      "neg_sim_mean": 0.4757501780986786,
      "pos_sim_mean": 0.5342530608177185,
      "step": 320
    },
    {
      "epoch": 0.0642,
      "grad_norm": 4.66371488571167,
      "learning_rate": 6.400000000000001e-05,
      "loss": 0.8163,
      "step": 321
    },
    {
      "epoch": 0.0642,
      "loss_align": 0.5116283297538757,
      "loss_contrastive": 0.08968672156333923,
      "loss_gen": 0.28506144881248474,
      "neg_sim_mean": 0.478058397769928,
      "pos_sim_mean": 0.48837167024612427,
      "step": 321
    },
    {
      "epoch": 0.0644,
      "grad_norm": 4.74101448059082,
      "learning_rate": 6.42e-05,
      "loss": 0.8075,
      "step": 322
    },
    {
      "epoch": 0.0644,
      "loss_align": 0.548302948474884,
      "loss_contrastive": 0.0913148820400238,
      "loss_gen": 0.1963711827993393,
      "neg_sim_mean": 0.44301193952560425,
      "pos_sim_mean": 0.45169705152511597,
      "step": 322
    },
    {
      "epoch": 0.0646,
      "grad_norm": 3.5068557262420654,
      "learning_rate": 6.440000000000001e-05,
      "loss": 0.7556,
      "step": 323
    },
    {
      "epoch": 0.0646,
      "loss_align": 0.42675507068634033,
      "loss_contrastive": 0.11255773901939392,
      "loss_gen": 0.18465672433376312,
      "neg_sim_mean": 0.5858026742935181,
      "pos_sim_mean": 0.5732449293136597,
      "step": 323
    },
    {
      "epoch": 0.0648,
      "grad_norm": 3.966852903366089,
      "learning_rate": 6.460000000000001e-05,
      "loss": 0.6249,
      "step": 324
    },
    {
      "epoch": 0.0648,
      "loss_align": 0.47591614723205566,
      "loss_contrastive": 0.20744016766548157,
      "loss_gen": 0.21738098561763763,
      "neg_sim_mean": 0.6315240263938904,
      "pos_sim_mean": 0.5240838527679443,
      "step": 324
    },
    {
      "epoch": 0.065,
      "grad_norm": 4.314692497253418,
      "learning_rate": 6.48e-05,
      "loss": 0.7182,
      "step": 325
    },
    {
      "epoch": 0.065,
      "loss_align": 0.33260536193847656,
      "loss_contrastive": 0.04809468984603882,
      "loss_gen": 0.37737876176834106,
      "neg_sim_mean": 0.6154893040657043,
      "pos_sim_mean": 0.6673946380615234,
      "step": 325
    },
    {
      "epoch": 0.0652,
      "grad_norm": 5.907522201538086,
      "learning_rate": 6.500000000000001e-05,
      "loss": 0.7158,
      "step": 326
    },
    {
      "epoch": 0.0652,
      "loss_align": 0.4354217052459717,
      "loss_contrastive": 0.02397364377975464,
      "loss_gen": 0.21466128528118134,
      "neg_sim_mean": 0.48855194449424744,
      "pos_sim_mean": 0.5645782947540283,
      "step": 326
    },
    {
      "epoch": 0.0654,
      "grad_norm": 5.717843532562256,
      "learning_rate": 6.52e-05,
      "loss": 0.653,
      "step": 327
    },
    {
      "epoch": 0.0654,
      "loss_align": 0.5474451780319214,
      "loss_contrastive": 0.10887846350669861,
      "loss_gen": 0.2771158218383789,
      "neg_sim_mean": 0.4614332914352417,
      "pos_sim_mean": 0.4525548219680786,
      "step": 327
    },
    {
      "epoch": 0.0656,
      "grad_norm": 4.923692226409912,
      "learning_rate": 6.54e-05,
      "loss": 0.8376,
      "step": 328
    },
    {
      "epoch": 0.0656,
      "loss_align": 0.3660566806793213,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0251726396381855,
      "neg_sim_mean": 0.49154984951019287,
      "pos_sim_mean": 0.6339433193206787,
      "step": 328
    },
    {
      "epoch": 0.0658,
      "grad_norm": 2.3932039737701416,
      "learning_rate": 6.560000000000001e-05,
      "loss": 0.3912,
      "step": 329
    },
    {
      "epoch": 0.0658,
      "loss_align": 0.5232217311859131,
      "loss_contrastive": 0.11625421047210693,
      "loss_gen": 0.018825747072696686,
      "neg_sim_mean": 0.49303245544433594,
      "pos_sim_mean": 0.4767782390117645,
      "step": 329
    },
    {
      "epoch": 0.066,
      "grad_norm": 2.200507879257202,
      "learning_rate": 6.58e-05,
      "loss": 0.556,
      "step": 330
    },
    {
      "epoch": 0.066,
      "loss_align": 0.4295527935028076,
      "loss_contrastive": 0.029740601778030396,
      "loss_gen": 0.008928554132580757,
      "neg_sim_mean": 0.5001878142356873,
      "pos_sim_mean": 0.5704472064971924,
      "step": 330
    },
    {
      "epoch": 0.0662,
      "grad_norm": 1.386909008026123,
      "learning_rate": 6.6e-05,
      "loss": 0.4421,
      "step": 331
    },
    {
      "epoch": 0.0662,
      "loss_align": 0.402418851852417,
      "loss_contrastive": 0.005085974931716919,
      "loss_gen": 0.003735925070941448,
      "neg_sim_mean": 0.5026671290397644,
      "pos_sim_mean": 0.597581148147583,
      "step": 331
    },
    {
      "epoch": 0.0664,
      "grad_norm": 0.8450384736061096,
      "learning_rate": 6.620000000000001e-05,
      "loss": 0.4068,
      "step": 332
    },
    {
      "epoch": 0.0664,
      "loss_align": 0.4455946683883667,
      "loss_contrastive": 0.03537103533744812,
      "loss_gen": 0.14478816092014313,
      "neg_sim_mean": 0.4897763729095459,
      "pos_sim_mean": 0.5544053316116333,
      "step": 332
    },
    {
      "epoch": 0.0666,
      "grad_norm": 3.0666725635528564,
      "learning_rate": 6.64e-05,
      "loss": 0.5946,
      "step": 333
    },
    {
      "epoch": 0.0666,
      "loss_align": 0.3182886242866516,
      "loss_contrastive": 0.017264068126678467,
      "loss_gen": 0.3486067056655884,
      "neg_sim_mean": 0.598975419998169,
      "pos_sim_mean": 0.6817113757133484,
      "step": 333
    },
    {
      "epoch": 0.0668,
      "grad_norm": 4.80580472946167,
      "learning_rate": 6.66e-05,
      "loss": 0.669,
      "step": 334
    },
    {
      "epoch": 0.0668,
      "loss_align": 0.3852011561393738,
      "loss_contrastive": 0.11780893802642822,
      "loss_gen": 0.5941314697265625,
      "neg_sim_mean": 0.6326077580451965,
      "pos_sim_mean": 0.6147988438606262,
      "step": 334
    },
    {
      "epoch": 0.067,
      "grad_norm": 6.660198211669922,
      "learning_rate": 6.680000000000001e-05,
      "loss": 0.9935,
      "step": 335
    },
    {
      "epoch": 0.067,
      "loss_align": 0.49430376291275024,
      "loss_contrastive": 0.11848291754722595,
      "loss_gen": 0.2028721570968628,
      "neg_sim_mean": 0.5241791605949402,
      "pos_sim_mean": 0.5056962370872498,
      "step": 335
    },
    {
      "epoch": 0.0672,
      "grad_norm": 4.190873622894287,
      "learning_rate": 6.7e-05,
      "loss": 0.7114,
      "step": 336
    },
    {
      "epoch": 0.0672,
      "loss_align": 0.46882718801498413,
      "loss_contrastive": 0.10648354887962341,
      "loss_gen": 0.44794681668281555,
      "neg_sim_mean": 0.5376563668251038,
      "pos_sim_mean": 0.5311728119850159,
      "step": 336
    },
    {
      "epoch": 0.0674,
      "grad_norm": 5.682254791259766,
      "learning_rate": 6.720000000000001e-05,
      "loss": 0.9296,
      "step": 337
    },
    {
      "epoch": 0.0674,
      "loss_align": 0.37483179569244385,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0016936250030994415,
      "neg_sim_mean": 0.4922822117805481,
      "pos_sim_mean": 0.6251682043075562,
      "step": 337
    },
    {
      "epoch": 0.0676,
      "grad_norm": 0.6566380262374878,
      "learning_rate": 6.740000000000001e-05,
      "loss": 0.3765,
      "step": 338
    },
    {
      "epoch": 0.0676,
      "loss_align": 0.4472421407699585,
      "loss_contrastive": 0.08579584956169128,
      "loss_gen": 0.5782883167266846,
      "neg_sim_mean": 0.5385537147521973,
      "pos_sim_mean": 0.5527578592300415,
      "step": 338
    },
    {
      "epoch": 0.0678,
      "grad_norm": 7.548217296600342,
      "learning_rate": 6.76e-05,
      "loss": 1.0358,
      "step": 339
    },
    {
      "epoch": 0.0678,
      "loss_align": 0.5067899227142334,
      "loss_contrastive": 0.21325898170471191,
      "loss_gen": 0.3384286165237427,
      "neg_sim_mean": 0.6064690351486206,
      "pos_sim_mean": 0.4932100474834442,
      "step": 339
    },
    {
      "epoch": 0.068,
      "grad_norm": 4.788571834564209,
      "learning_rate": 6.780000000000001e-05,
      "loss": 0.8708,
      "step": 340
    },
    {
      "epoch": 0.068,
      "loss_align": 0.46385467052459717,
      "loss_contrastive": 0.05941453576087952,
      "loss_gen": 0.0013867092784494162,
      "neg_sim_mean": 0.4955598711967468,
      "pos_sim_mean": 0.5361453294754028,
      "step": 340
    },
    {
      "epoch": 0.0682,
      "grad_norm": 0.7629148960113525,
      "learning_rate": 6.800000000000001e-05,
      "loss": 0.4724,
      "step": 341
    },
    {
      "epoch": 0.0682,
      "loss_align": 0.39613914489746094,
      "loss_contrastive": 0.11171913146972656,
      "loss_gen": 0.42420467734336853,
      "neg_sim_mean": 0.6155799627304077,
      "pos_sim_mean": 0.6038608551025391,
      "step": 341
    },
    {
      "epoch": 0.0684,
      "grad_norm": 5.494924068450928,
      "learning_rate": 6.82e-05,
      "loss": 0.8338,
      "step": 342
    },
    {
      "epoch": 0.0684,
      "loss_align": 0.32429778575897217,
      "loss_contrastive": 0.10570299625396729,
      "loss_gen": 0.3368292450904846,
      "neg_sim_mean": 0.6814051866531372,
      "pos_sim_mean": 0.6757022142410278,
      "step": 342
    },
    {
      "epoch": 0.0686,
      "grad_norm": 4.522895336151123,
      "learning_rate": 6.840000000000001e-05,
      "loss": 0.6738,
      "step": 343
    },
    {
      "epoch": 0.0686,
      "loss_align": 0.43500447273254395,
      "loss_contrastive": 0.0821826159954071,
      "loss_gen": 0.3024328052997589,
      "neg_sim_mean": 0.5471781492233276,
      "pos_sim_mean": 0.564995527267456,
      "step": 343
    },
    {
      "epoch": 0.0688,
      "grad_norm": 4.159493446350098,
      "learning_rate": 6.860000000000001e-05,
      "loss": 0.7473,
      "step": 344
    },
    {
      "epoch": 0.0688,
      "loss_align": 0.44816458225250244,
      "loss_contrastive": 0.04841288924217224,
      "loss_gen": 0.003199824132025242,
      "neg_sim_mean": 0.5002483129501343,
      "pos_sim_mean": 0.5518354177474976,
      "step": 344
    },
    {
      "epoch": 0.069,
      "grad_norm": 0.8564828038215637,
      "learning_rate": 6.879999999999999e-05,
      "loss": 0.4572,
      "step": 345
    },
    {
      "epoch": 0.069,
      "loss_align": 0.671516478061676,
      "loss_contrastive": 0.26994651556015015,
      "loss_gen": 0.004107239190489054,
      "neg_sim_mean": 0.4984300434589386,
      "pos_sim_mean": 0.328483521938324,
      "step": 345
    },
    {
      "epoch": 0.0692,
      "grad_norm": 1.0143541097640991,
      "learning_rate": 6.9e-05,
      "loss": 0.708,
      "step": 346
    },
    {
      "epoch": 0.0692,
      "loss_align": 0.47809892892837524,
      "loss_contrastive": 0.08367440104484558,
      "loss_gen": 0.409329354763031,
      "neg_sim_mean": 0.5055754780769348,
      "pos_sim_mean": 0.5219010710716248,
      "step": 346
    },
    {
      "epoch": 0.0694,
      "grad_norm": 5.247970104217529,
      "learning_rate": 6.92e-05,
      "loss": 0.8975,
      "step": 347
    },
    {
      "epoch": 0.0694,
      "loss_align": 0.48336005210876465,
      "loss_contrastive": 0.07669851183891296,
      "loss_gen": 0.003727170405909419,
      "neg_sim_mean": 0.4933384656906128,
      "pos_sim_mean": 0.5166399478912354,
      "step": 347
    },
    {
      "epoch": 0.0696,
      "grad_norm": 0.8973610401153564,
      "learning_rate": 6.939999999999999e-05,
      "loss": 0.4963,
      "step": 348
    },
    {
      "epoch": 0.0696,
      "loss_align": 0.49082666635513306,
      "loss_contrastive": 0.052122920751571655,
      "loss_gen": 0.2702235281467438,
      "neg_sim_mean": 0.4612962603569031,
      "pos_sim_mean": 0.5091733336448669,
      "step": 348
    },
    {
      "epoch": 0.0698,
      "grad_norm": 4.402535915374756,
      "learning_rate": 6.96e-05,
      "loss": 0.7673,
      "step": 349
    },
    {
      "epoch": 0.0698,
      "loss_align": 0.44297003746032715,
      "loss_contrastive": 0.05240628123283386,
      "loss_gen": 0.19370132684707642,
      "neg_sim_mean": 0.5094362497329712,
      "pos_sim_mean": 0.5570299625396729,
      "step": 349
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.9894957542419434,
      "learning_rate": 6.98e-05,
      "loss": 0.643,
      "step": 350
    },
    {
      "epoch": 0.07,
      "loss_align": 0.4861633777618408,
      "loss_contrastive": 0.05250057578086853,
      "loss_gen": 0.4117458164691925,
      "neg_sim_mean": 0.4663372039794922,
      "pos_sim_mean": 0.5138366222381592,
      "step": 350
    },
    {
      "epoch": 0.0702,
      "grad_norm": 5.111454486846924,
      "learning_rate": 7e-05,
      "loss": 0.9042,
      "step": 351
    },
    {
      "epoch": 0.0702,
      "loss_align": 0.4951908588409424,
      "loss_contrastive": 0.06768515706062317,
      "loss_gen": 0.0017980093834921718,
      "neg_sim_mean": 0.47249430418014526,
      "pos_sim_mean": 0.5048091411590576,
      "step": 351
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.7478939294815063,
      "learning_rate": 7.02e-05,
      "loss": 0.5051,
      "step": 352
    },
    {
      "epoch": 0.0704,
      "loss_align": 0.39835822582244873,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003277980722486973,
      "neg_sim_mean": 0.4676506519317627,
      "pos_sim_mean": 0.6016417741775513,
      "step": 352
    },
    {
      "epoch": 0.0706,
      "grad_norm": 0.7532848715782166,
      "learning_rate": 7.04e-05,
      "loss": 0.4016,
      "step": 353
    },
    {
      "epoch": 0.0706,
      "loss_align": 0.4730045199394226,
      "loss_contrastive": 0.0031453967094421387,
      "loss_gen": 0.5186734795570374,
      "neg_sim_mean": 0.430140882730484,
      "pos_sim_mean": 0.5269954800605774,
      "step": 353
    },
    {
      "epoch": 0.0708,
      "grad_norm": 7.243414402008057,
      "learning_rate": 7.06e-05,
      "loss": 0.9921,
      "step": 354
    },
    {
      "epoch": 0.0708,
      "loss_align": 0.39293867349624634,
      "loss_contrastive": 0.09335893392562866,
      "loss_gen": 0.18366065621376038,
      "neg_sim_mean": 0.6004202365875244,
      "pos_sim_mean": 0.6070613265037537,
      "step": 354
    },
    {
      "epoch": 0.071,
      "grad_norm": 4.060479164123535,
      "learning_rate": 7.08e-05,
      "loss": 0.5878,
      "step": 355
    },
    {
      "epoch": 0.071,
      "loss_align": 0.4535837173461914,
      "loss_contrastive": 0.0014834702014923096,
      "loss_gen": 0.4248813986778259,
      "neg_sim_mean": 0.4478997588157654,
      "pos_sim_mean": 0.5464162826538086,
      "step": 355
    },
    {
      "epoch": 0.0712,
      "grad_norm": 5.706402778625488,
      "learning_rate": 7.1e-05,
      "loss": 0.8786,
      "step": 356
    },
    {
      "epoch": 0.0712,
      "loss_align": 0.6851861476898193,
      "loss_contrastive": 0.3088478744029999,
      "loss_gen": 0.4400654733181,
      "neg_sim_mean": 0.523661732673645,
      "pos_sim_mean": 0.31481385231018066,
      "step": 356
    },
    {
      "epoch": 0.0714,
      "grad_norm": 5.954972743988037,
      "learning_rate": 7.12e-05,
      "loss": 1.1623,
      "step": 357
    },
    {
      "epoch": 0.0714,
      "loss_align": 0.4807562828063965,
      "loss_contrastive": 0.03335386514663696,
      "loss_gen": 0.002248788019642234,
      "neg_sim_mean": 0.45259758830070496,
      "pos_sim_mean": 0.5192437171936035,
      "step": 357
    },
    {
      "epoch": 0.0716,
      "grad_norm": 0.7275072932243347,
      "learning_rate": 7.14e-05,
      "loss": 0.487,
      "step": 358
    },
    {
      "epoch": 0.0716,
      "loss_align": 0.4374760389328003,
      "loss_contrastive": 0.07890471816062927,
      "loss_gen": 0.4509928524494171,
      "neg_sim_mean": 0.5414286851882935,
      "pos_sim_mean": 0.5625239610671997,
      "step": 358
    },
    {
      "epoch": 0.0718,
      "grad_norm": 6.710362434387207,
      "learning_rate": 7.16e-05,
      "loss": 0.8979,
      "step": 359
    },
    {
      "epoch": 0.0718,
      "loss_align": 0.549680769443512,
      "loss_contrastive": 0.12249001860618591,
      "loss_gen": 0.21264636516571045,
      "neg_sim_mean": 0.4728092551231384,
      "pos_sim_mean": 0.45031923055648804,
      "step": 359
    },
    {
      "epoch": 0.072,
      "grad_norm": 4.085011959075928,
      "learning_rate": 7.18e-05,
      "loss": 0.777,
      "step": 360
    },
    {
      "epoch": 0.072,
      "loss_align": 0.4639604687690735,
      "loss_contrastive": 0.06582888960838318,
      "loss_gen": 0.43543028831481934,
      "neg_sim_mean": 0.5018684267997742,
      "pos_sim_mean": 0.5360395312309265,
      "step": 360
    },
    {
      "epoch": 0.0722,
      "grad_norm": 5.455506324768066,
      "learning_rate": 7.2e-05,
      "loss": 0.9073,
      "step": 361
    },
    {
      "epoch": 0.0722,
      "loss_align": 0.42825162410736084,
      "loss_contrastive": 0.09723839163780212,
      "loss_gen": 0.2533588707447052,
      "neg_sim_mean": 0.5689867734909058,
      "pos_sim_mean": 0.5717483758926392,
      "step": 361
    },
    {
      "epoch": 0.0724,
      "grad_norm": 3.782238006591797,
      "learning_rate": 7.22e-05,
      "loss": 0.6933,
      "step": 362
    },
    {
      "epoch": 0.0724,
      "loss_align": 0.47528672218322754,
      "loss_contrastive": 0.030077308416366577,
      "loss_gen": 0.02897115796804428,
      "neg_sim_mean": 0.4547905921936035,
      "pos_sim_mean": 0.5247132778167725,
      "step": 362
    },
    {
      "epoch": 0.0726,
      "grad_norm": 2.407893419265747,
      "learning_rate": 7.24e-05,
      "loss": 0.5079,
      "step": 363
    },
    {
      "epoch": 0.0726,
      "loss_align": 0.4003833532333374,
      "loss_contrastive": 0.1082359254360199,
      "loss_gen": 0.2508118450641632,
      "neg_sim_mean": 0.607852578163147,
      "pos_sim_mean": 0.5996166467666626,
      "step": 363
    },
    {
      "epoch": 0.0728,
      "grad_norm": 4.653700828552246,
      "learning_rate": 7.26e-05,
      "loss": 0.6642,
      "step": 364
    },
    {
      "epoch": 0.0728,
      "loss_align": 0.33570241928100586,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012177412398159504,
      "neg_sim_mean": 0.4564843773841858,
      "pos_sim_mean": 0.6642975807189941,
      "step": 364
    },
    {
      "epoch": 0.073,
      "grad_norm": 1.6301313638687134,
      "learning_rate": 7.280000000000001e-05,
      "loss": 0.3479,
      "step": 365
    },
    {
      "epoch": 0.073,
      "loss_align": 0.5490612983703613,
      "loss_contrastive": 0.16478702425956726,
      "loss_gen": 0.12675829231739044,
      "neg_sim_mean": 0.5157257318496704,
      "pos_sim_mean": 0.45093870162963867,
      "step": 365
    },
    {
      "epoch": 0.0732,
      "grad_norm": 3.1496853828430176,
      "learning_rate": 7.3e-05,
      "loss": 0.6956,
      "step": 366
    },
    {
      "epoch": 0.0732,
      "loss_align": 0.4794219732284546,
      "loss_contrastive": 0.12450727820396423,
      "loss_gen": 0.41048720479011536,
      "neg_sim_mean": 0.5450853109359741,
      "pos_sim_mean": 0.5205780267715454,
      "step": 366
    },
    {
      "epoch": 0.0734,
      "grad_norm": 5.219537258148193,
      "learning_rate": 7.32e-05,
      "loss": 0.9049,
      "step": 367
    },
    {
      "epoch": 0.0734,
      "loss_align": 0.5135320425033569,
      "loss_contrastive": 0.08167412877082825,
      "loss_gen": 0.4013638496398926,
      "neg_sim_mean": 0.4681420922279358,
      "pos_sim_mean": 0.48646795749664307,
      "step": 367
    },
    {
      "epoch": 0.0736,
      "grad_norm": 5.801942348480225,
      "learning_rate": 7.340000000000001e-05,
      "loss": 0.9247,
      "step": 368
    },
    {
      "epoch": 0.0736,
      "loss_align": 0.45559877157211304,
      "loss_contrastive": 0.051877349615097046,
      "loss_gen": 0.3646402657032013,
      "neg_sim_mean": 0.4962785840034485,
      "pos_sim_mean": 0.544401228427887,
      "step": 368
    },
    {
      "epoch": 0.0738,
      "grad_norm": 4.329456806182861,
      "learning_rate": 7.36e-05,
      "loss": 0.8265,
      "step": 369
    },
    {
      "epoch": 0.0738,
      "loss_align": 0.4546751379966736,
      "loss_contrastive": 0.027013152837753296,
      "loss_gen": 0.3936033844947815,
      "neg_sim_mean": 0.4723380208015442,
      "pos_sim_mean": 0.5453248620033264,
      "step": 369
    },
    {
      "epoch": 0.074,
      "grad_norm": 6.04612922668457,
      "learning_rate": 7.38e-05,
      "loss": 0.8515,
      "step": 370
    },
    {
      "epoch": 0.074,
      "loss_align": 0.5391757488250732,
      "loss_contrastive": 0.10950401425361633,
      "loss_gen": 0.0018375017680227757,
      "neg_sim_mean": 0.47032827138900757,
      "pos_sim_mean": 0.46082425117492676,
      "step": 370
    },
    {
      "epoch": 0.0742,
      "grad_norm": 0.7154759764671326,
      "learning_rate": 7.4e-05,
      "loss": 0.5542,
      "step": 371
    },
    {
      "epoch": 0.0742,
      "loss_align": 0.5057868361473083,
      "loss_contrastive": 0.07807061076164246,
      "loss_gen": 0.0029718857258558273,
      "neg_sim_mean": 0.4722837805747986,
      "pos_sim_mean": 0.49421316385269165,
      "step": 371
    },
    {
      "epoch": 0.0744,
      "grad_norm": 0.7874810099601746,
      "learning_rate": 7.42e-05,
      "loss": 0.5181,
      "step": 372
    },
    {
      "epoch": 0.0744,
      "loss_align": 0.5000601410865784,
      "loss_contrastive": 0.07916498184204102,
      "loss_gen": 0.4697619080543518,
      "neg_sim_mean": 0.4791048467159271,
      "pos_sim_mean": 0.49993985891342163,
      "step": 372
    },
    {
      "epoch": 0.0746,
      "grad_norm": 6.051591396331787,
      "learning_rate": 7.44e-05,
      "loss": 0.9793,
      "step": 373
    },
    {
      "epoch": 0.0746,
      "loss_align": 0.40419310331344604,
      "loss_contrastive": 0.03142932057380676,
      "loss_gen": 0.20410574972629547,
      "neg_sim_mean": 0.5272362232208252,
      "pos_sim_mean": 0.595806896686554,
      "step": 373
    },
    {
      "epoch": 0.0748,
      "grad_norm": 4.424078941345215,
      "learning_rate": 7.46e-05,
      "loss": 0.6121,
      "step": 374
    },
    {
      "epoch": 0.0748,
      "loss_align": 0.5178163051605225,
      "loss_contrastive": 0.026456505060195923,
      "loss_gen": 0.32603076100349426,
      "neg_sim_mean": 0.40864020586013794,
      "pos_sim_mean": 0.48218369483947754,
      "step": 374
    },
    {
      "epoch": 0.075,
      "grad_norm": 5.227490425109863,
      "learning_rate": 7.48e-05,
      "loss": 0.847,
      "step": 375
    },
    {
      "epoch": 0.075,
      "loss_align": 0.39743077754974365,
      "loss_contrastive": 0.1064075231552124,
      "loss_gen": 0.17494401335716248,
      "neg_sim_mean": 0.6089767217636108,
      "pos_sim_mean": 0.6025692224502563,
      "step": 375
    },
    {
      "epoch": 0.0752,
      "grad_norm": 3.591777801513672,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.5851,
      "step": 376
    },
    {
      "epoch": 0.0752,
      "loss_align": 0.45858097076416016,
      "loss_contrastive": 0.14738556742668152,
      "loss_gen": 0.27076849341392517,
      "neg_sim_mean": 0.5888046026229858,
      "pos_sim_mean": 0.5414190292358398,
      "step": 376
    },
    {
      "epoch": 0.0754,
      "grad_norm": 3.770490884780884,
      "learning_rate": 7.52e-05,
      "loss": 0.747,
      "step": 377
    },
    {
      "epoch": 0.0754,
      "loss_align": 0.6427107453346252,
      "loss_contrastive": 0.26429733633995056,
      "loss_gen": 0.38217344880104065,
      "neg_sim_mean": 0.5215865969657898,
      "pos_sim_mean": 0.35728925466537476,
      "step": 377
    },
    {
      "epoch": 0.0756,
      "grad_norm": 4.926638126373291,
      "learning_rate": 7.54e-05,
      "loss": 1.0566,
      "step": 378
    },
    {
      "epoch": 0.0756,
      "loss_align": 0.3975936770439148,
      "loss_contrastive": 0.07636982202529907,
      "loss_gen": 0.3701359033584595,
      "neg_sim_mean": 0.5787761211395264,
      "pos_sim_mean": 0.6024063229560852,
      "step": 378
    },
    {
      "epoch": 0.0758,
      "grad_norm": 4.9919023513793945,
      "learning_rate": 7.560000000000001e-05,
      "loss": 0.7769,
      "step": 379
    },
    {
      "epoch": 0.0758,
      "loss_align": 0.4141695499420166,
      "loss_contrastive": 0.14713796973228455,
      "loss_gen": 0.25238847732543945,
      "neg_sim_mean": 0.6329684257507324,
      "pos_sim_mean": 0.5858304500579834,
      "step": 379
    },
    {
      "epoch": 0.076,
      "grad_norm": 3.4124372005462646,
      "learning_rate": 7.58e-05,
      "loss": 0.6842,
      "step": 380
    },
    {
      "epoch": 0.076,
      "loss_align": 0.36626875400543213,
      "loss_contrastive": 0.0977480411529541,
      "loss_gen": 0.35784459114074707,
      "neg_sim_mean": 0.6314792633056641,
      "pos_sim_mean": 0.6337312459945679,
      "step": 380
    },
    {
      "epoch": 0.0762,
      "grad_norm": 4.660800933837891,
      "learning_rate": 7.6e-05,
      "loss": 0.7358,
      "step": 381
    },
    {
      "epoch": 0.0762,
      "loss_align": 0.3874434232711792,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00367049896158278,
      "neg_sim_mean": 0.48180997371673584,
      "pos_sim_mean": 0.6125565767288208,
      "step": 381
    },
    {
      "epoch": 0.0764,
      "grad_norm": 0.7436308860778809,
      "learning_rate": 7.620000000000001e-05,
      "loss": 0.3911,
      "step": 382
    },
    {
      "epoch": 0.0764,
      "loss_align": 0.51130211353302,
      "loss_contrastive": 0.11528649926185608,
      "loss_gen": 0.21482126414775848,
      "neg_sim_mean": 0.5039843916893005,
      "pos_sim_mean": 0.48869788646698,
      "step": 382
    },
    {
      "epoch": 0.0766,
      "grad_norm": 4.360066890716553,
      "learning_rate": 7.64e-05,
      "loss": 0.74,
      "step": 383
    },
    {
      "epoch": 0.0766,
      "loss_align": 0.380273699760437,
      "loss_contrastive": 0.12465494871139526,
      "loss_gen": 0.36605143547058105,
      "neg_sim_mean": 0.6443812251091003,
      "pos_sim_mean": 0.619726300239563,
      "step": 383
    },
    {
      "epoch": 0.0768,
      "grad_norm": 4.58119535446167,
      "learning_rate": 7.66e-05,
      "loss": 0.7613,
      "step": 384
    },
    {
      "epoch": 0.0768,
      "loss_align": 0.43524259328842163,
      "loss_contrastive": 0.11912170052528381,
      "loss_gen": 0.16371871531009674,
      "neg_sim_mean": 0.5838791131973267,
      "pos_sim_mean": 0.5647574067115784,
      "step": 384
    },
    {
      "epoch": 0.077,
      "grad_norm": 3.4283688068389893,
      "learning_rate": 7.680000000000001e-05,
      "loss": 0.6133,
      "step": 385
    },
    {
      "epoch": 0.077,
      "loss_align": 0.40881621837615967,
      "loss_contrastive": 0.1510504186153412,
      "loss_gen": 0.2407779097557068,
      "neg_sim_mean": 0.642234206199646,
      "pos_sim_mean": 0.5911837816238403,
      "step": 385
    },
    {
      "epoch": 0.0772,
      "grad_norm": 3.3193774223327637,
      "learning_rate": 7.7e-05,
      "loss": 0.6677,
      "step": 386
    },
    {
      "epoch": 0.0772,
      "loss_align": 0.5780316591262817,
      "loss_contrastive": 0.19671028852462769,
      "loss_gen": 0.17370890080928802,
      "neg_sim_mean": 0.518678605556488,
      "pos_sim_mean": 0.4219683110713959,
      "step": 386
    },
    {
      "epoch": 0.0774,
      "grad_norm": 3.4766311645507812,
      "learning_rate": 7.72e-05,
      "loss": 0.7753,
      "step": 387
    },
    {
      "epoch": 0.0774,
      "loss_align": 0.4521546959877014,
      "loss_contrastive": 0.09361997246742249,
      "loss_gen": 0.3103458881378174,
      "neg_sim_mean": 0.5414652824401855,
      "pos_sim_mean": 0.5478453040122986,
      "step": 387
    },
    {
      "epoch": 0.0776,
      "grad_norm": 4.773631572723389,
      "learning_rate": 7.740000000000001e-05,
      "loss": 0.7737,
      "step": 388
    },
    {
      "epoch": 0.0776,
      "loss_align": 0.4398806095123291,
      "loss_contrastive": 0.12958785891532898,
      "loss_gen": 0.24058128893375397,
      "neg_sim_mean": 0.5897072553634644,
      "pos_sim_mean": 0.5601193904876709,
      "step": 388
    },
    {
      "epoch": 0.0778,
      "grad_norm": 4.191360950469971,
      "learning_rate": 7.76e-05,
      "loss": 0.696,
      "step": 389
    },
    {
      "epoch": 0.0778,
      "loss_align": 0.5175379514694214,
      "loss_contrastive": 0.11020436882972717,
      "loss_gen": 0.041520338505506516,
      "neg_sim_mean": 0.4926663935184479,
      "pos_sim_mean": 0.4824620187282562,
      "step": 389
    },
    {
      "epoch": 0.078,
      "grad_norm": 3.3229169845581055,
      "learning_rate": 7.780000000000001e-05,
      "loss": 0.5723,
      "step": 390
    },
    {
      "epoch": 0.078,
      "loss_align": 0.43353891372680664,
      "loss_contrastive": 0.15381845831871033,
      "loss_gen": 0.08748160302639008,
      "neg_sim_mean": 0.6202795505523682,
      "pos_sim_mean": 0.5664610862731934,
      "step": 390
    },
    {
      "epoch": 0.0782,
      "grad_norm": 2.3095526695251465,
      "learning_rate": 7.800000000000001e-05,
      "loss": 0.5395,
      "step": 391
    },
    {
      "epoch": 0.0782,
      "loss_align": 0.3479191064834595,
      "loss_contrastive": 0.0672106146812439,
      "loss_gen": 0.1700022667646408,
      "neg_sim_mean": 0.6192914843559265,
      "pos_sim_mean": 0.6520808935165405,
      "step": 391
    },
    {
      "epoch": 0.0784,
      "grad_norm": 3.4490246772766113,
      "learning_rate": 7.82e-05,
      "loss": 0.526,
      "step": 392
    },
    {
      "epoch": 0.0784,
      "loss_align": 0.41754865646362305,
      "loss_contrastive": 0.19237861037254333,
      "loss_gen": 0.2551043927669525,
      "neg_sim_mean": 0.6748299598693848,
      "pos_sim_mean": 0.582451343536377,
      "step": 392
    },
    {
      "epoch": 0.0786,
      "grad_norm": 4.2545485496521,
      "learning_rate": 7.840000000000001e-05,
      "loss": 0.6957,
      "step": 393
    },
    {
      "epoch": 0.0786,
      "loss_align": 0.44602543115615845,
      "loss_contrastive": 0.14134958386421204,
      "loss_gen": 0.362663209438324,
      "neg_sim_mean": 0.5953241586685181,
      "pos_sim_mean": 0.5539745688438416,
      "step": 393
    },
    {
      "epoch": 0.0788,
      "grad_norm": 3.941119909286499,
      "learning_rate": 7.860000000000001e-05,
      "loss": 0.8257,
      "step": 394
    },
    {
      "epoch": 0.0788,
      "loss_align": 0.39573168754577637,
      "loss_contrastive": 0.0,
      "loss_gen": 0.020309949293732643,
      "neg_sim_mean": 0.4961617588996887,
      "pos_sim_mean": 0.6042683124542236,
      "step": 394
    },
    {
      "epoch": 0.079,
      "grad_norm": 2.1361331939697266,
      "learning_rate": 7.88e-05,
      "loss": 0.416,
      "step": 395
    },
    {
      "epoch": 0.079,
      "loss_align": 0.4075862169265747,
      "loss_contrastive": 0.08757564425468445,
      "loss_gen": 0.5752713680267334,
      "neg_sim_mean": 0.5799894332885742,
      "pos_sim_mean": 0.5924137830734253,
      "step": 395
    },
    {
      "epoch": 0.0792,
      "grad_norm": 5.682326316833496,
      "learning_rate": 7.900000000000001e-05,
      "loss": 0.9934,
      "step": 396
    },
    {
      "epoch": 0.0792,
      "loss_align": 0.6398273706436157,
      "loss_contrastive": 0.24922117590904236,
      "loss_gen": 0.5007286667823792,
      "neg_sim_mean": 0.5093938112258911,
      "pos_sim_mean": 0.3601726293563843,
      "step": 396
    },
    {
      "epoch": 0.0794,
      "grad_norm": 4.89358377456665,
      "learning_rate": 7.920000000000001e-05,
      "loss": 1.1705,
      "step": 397
    },
    {
      "epoch": 0.0794,
      "loss_align": 0.4421043395996094,
      "loss_contrastive": 0.04033669829368591,
      "loss_gen": 0.43837612867355347,
      "neg_sim_mean": 0.498232364654541,
      "pos_sim_mean": 0.5578956604003906,
      "step": 397
    },
    {
      "epoch": 0.0796,
      "grad_norm": 4.360952377319336,
      "learning_rate": 7.94e-05,
      "loss": 0.8853,
      "step": 398
    },
    {
      "epoch": 0.0796,
      "loss_align": 0.49768710136413574,
      "loss_contrastive": 0.16978952288627625,
      "loss_gen": 0.25004449486732483,
      "neg_sim_mean": 0.572102427482605,
      "pos_sim_mean": 0.5023128986358643,
      "step": 398
    },
    {
      "epoch": 0.0798,
      "grad_norm": 3.7914743423461914,
      "learning_rate": 7.960000000000001e-05,
      "loss": 0.7681,
      "step": 399
    },
    {
      "epoch": 0.0798,
      "loss_align": 0.3970690965652466,
      "loss_contrastive": 0.05822521448135376,
      "loss_gen": 0.5047485828399658,
      "neg_sim_mean": 0.5611560940742493,
      "pos_sim_mean": 0.6029309034347534,
      "step": 399
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.970183849334717,
      "learning_rate": 7.98e-05,
      "loss": 0.9088,
      "step": 400
    },
    {
      "epoch": 0.08,
      "loss_align": 0.3981393575668335,
      "loss_contrastive": 0.06308174133300781,
      "loss_gen": 0.3263634741306305,
      "neg_sim_mean": 0.5649423599243164,
      "pos_sim_mean": 0.6018606424331665,
      "step": 400
    },
    {
      "epoch": 0.0802,
      "grad_norm": 4.33458948135376,
      "learning_rate": 8e-05,
      "loss": 0.7321,
      "step": 401
    },
    {
      "epoch": 0.0802,
      "loss_align": 0.3366857171058655,
      "loss_contrastive": 0.04397827386856079,
      "loss_gen": 0.5081435441970825,
      "neg_sim_mean": 0.6072925329208374,
      "pos_sim_mean": 0.6633142828941345,
      "step": 401
    },
    {
      "epoch": 0.0804,
      "grad_norm": 4.862358570098877,
      "learning_rate": 8.020000000000001e-05,
      "loss": 0.8501,
      "step": 402
    },
    {
      "epoch": 0.0804,
      "loss_align": 0.36409759521484375,
      "loss_contrastive": 0.07265317440032959,
      "loss_gen": 0.15502485632896423,
      "neg_sim_mean": 0.6085555553436279,
      "pos_sim_mean": 0.6359024047851562,
      "step": 402
    },
    {
      "epoch": 0.0806,
      "grad_norm": 3.279250144958496,
      "learning_rate": 8.04e-05,
      "loss": 0.5278,
      "step": 403
    },
    {
      "epoch": 0.0806,
      "loss_align": 0.4097338318824768,
      "loss_contrastive": 0.14729705452919006,
      "loss_gen": 0.250876784324646,
      "neg_sim_mean": 0.6375632286071777,
      "pos_sim_mean": 0.5902661681175232,
      "step": 403
    },
    {
      "epoch": 0.0808,
      "grad_norm": 3.425713062286377,
      "learning_rate": 8.060000000000001e-05,
      "loss": 0.6783,
      "step": 404
    },
    {
      "epoch": 0.0808,
      "loss_align": 0.4179200530052185,
      "loss_contrastive": 0.022445887327194214,
      "loss_gen": 0.003522260347381234,
      "neg_sim_mean": 0.5045258402824402,
      "pos_sim_mean": 0.5820799469947815,
      "step": 404
    },
    {
      "epoch": 0.081,
      "grad_norm": 0.7722868919372559,
      "learning_rate": 8.080000000000001e-05,
      "loss": 0.4241,
      "step": 405
    },
    {
      "epoch": 0.081,
      "loss_align": 0.5613247156143188,
      "loss_contrastive": 0.17252695560455322,
      "loss_gen": 0.21971936523914337,
      "neg_sim_mean": 0.5112022161483765,
      "pos_sim_mean": 0.43867525458335876,
      "step": 405
    },
    {
      "epoch": 0.0812,
      "grad_norm": 4.052060604095459,
      "learning_rate": 8.1e-05,
      "loss": 0.8017,
      "step": 406
    },
    {
      "epoch": 0.0812,
      "loss_align": 0.4290711283683777,
      "loss_contrastive": 0.058231860399246216,
      "loss_gen": 0.415510892868042,
      "neg_sim_mean": 0.529160737991333,
      "pos_sim_mean": 0.5709288716316223,
      "step": 406
    },
    {
      "epoch": 0.0814,
      "grad_norm": 4.412705898284912,
      "learning_rate": 8.120000000000001e-05,
      "loss": 0.8516,
      "step": 407
    },
    {
      "epoch": 0.0814,
      "loss_align": 0.5203638076782227,
      "loss_contrastive": 0.1260955035686493,
      "loss_gen": 0.35942342877388,
      "neg_sim_mean": 0.5057317018508911,
      "pos_sim_mean": 0.47963619232177734,
      "step": 407
    },
    {
      "epoch": 0.0816,
      "grad_norm": 5.0423150062561035,
      "learning_rate": 8.14e-05,
      "loss": 0.8949,
      "step": 408
    },
    {
      "epoch": 0.0816,
      "loss_align": 0.5250416994094849,
      "loss_contrastive": 0.12609374523162842,
      "loss_gen": 0.03944796323776245,
      "neg_sim_mean": 0.5010520219802856,
      "pos_sim_mean": 0.47495827078819275,
      "step": 408
    },
    {
      "epoch": 0.0818,
      "grad_norm": 2.9710311889648438,
      "learning_rate": 8.16e-05,
      "loss": 0.5796,
      "step": 409
    },
    {
      "epoch": 0.0818,
      "loss_align": 0.4129505157470703,
      "loss_contrastive": 0.07929381728172302,
      "loss_gen": 0.3143634796142578,
      "neg_sim_mean": 0.5663433074951172,
      "pos_sim_mean": 0.5870494842529297,
      "step": 409
    },
    {
      "epoch": 0.082,
      "grad_norm": 3.8568522930145264,
      "learning_rate": 8.18e-05,
      "loss": 0.7368,
      "step": 410
    },
    {
      "epoch": 0.082,
      "loss_align": 0.43600720167160034,
      "loss_contrastive": 0.03647306561470032,
      "loss_gen": 0.01184909325093031,
      "neg_sim_mean": 0.5004658699035645,
      "pos_sim_mean": 0.5639927983283997,
      "step": 410
    },
    {
      "epoch": 0.0822,
      "grad_norm": 1.5112674236297607,
      "learning_rate": 8.2e-05,
      "loss": 0.4522,
      "step": 411
    },
    {
      "epoch": 0.0822,
      "loss_align": 0.3330690264701843,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4750140607357025,
      "neg_sim_mean": 0.5138184428215027,
      "pos_sim_mean": 0.6669309735298157,
      "step": 411
    },
    {
      "epoch": 0.0824,
      "grad_norm": 4.79976224899292,
      "learning_rate": 8.22e-05,
      "loss": 0.8081,
      "step": 412
    },
    {
      "epoch": 0.0824,
      "loss_align": 0.29279935359954834,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3194854259490967,
      "neg_sim_mean": 0.5838892459869385,
      "pos_sim_mean": 0.7072006464004517,
      "step": 412
    },
    {
      "epoch": 0.0826,
      "grad_norm": 4.3745527267456055,
      "learning_rate": 8.24e-05,
      "loss": 0.6123,
      "step": 413
    },
    {
      "epoch": 0.0826,
      "loss_align": 0.3482082486152649,
      "loss_contrastive": 0.10089689493179321,
      "loss_gen": 0.22501716017723083,
      "neg_sim_mean": 0.6526886224746704,
      "pos_sim_mean": 0.6517917513847351,
      "step": 413
    },
    {
      "epoch": 0.0828,
      "grad_norm": 3.780738592147827,
      "learning_rate": 8.26e-05,
      "loss": 0.5853,
      "step": 414
    },
    {
      "epoch": 0.0828,
      "loss_align": 0.39084064960479736,
      "loss_contrastive": 0.10150593519210815,
      "loss_gen": 0.3831116557121277,
      "neg_sim_mean": 0.6106652617454529,
      "pos_sim_mean": 0.6091593503952026,
      "step": 414
    },
    {
      "epoch": 0.083,
      "grad_norm": 4.435139179229736,
      "learning_rate": 8.28e-05,
      "loss": 0.7861,
      "step": 415
    },
    {
      "epoch": 0.083,
      "loss_align": 0.3516823649406433,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20437543094158173,
      "neg_sim_mean": 0.49983808398246765,
      "pos_sim_mean": 0.6483176350593567,
      "step": 415
    },
    {
      "epoch": 0.0832,
      "grad_norm": 3.804145574569702,
      "learning_rate": 8.3e-05,
      "loss": 0.5561,
      "step": 416
    },
    {
      "epoch": 0.0832,
      "loss_align": 0.48356837034225464,
      "loss_contrastive": 0.09367182850837708,
      "loss_gen": 0.35508206486701965,
      "neg_sim_mean": 0.5101034641265869,
      "pos_sim_mean": 0.5164316296577454,
      "step": 416
    },
    {
      "epoch": 0.0834,
      "grad_norm": 5.300083637237549,
      "learning_rate": 8.32e-05,
      "loss": 0.8499,
      "step": 417
    },
    {
      "epoch": 0.0834,
      "loss_align": 0.41884708404541016,
      "loss_contrastive": 0.010599732398986816,
      "loss_gen": 0.0014167225454002619,
      "neg_sim_mean": 0.49175265431404114,
      "pos_sim_mean": 0.5811529159545898,
      "step": 417
    },
    {
      "epoch": 0.0836,
      "grad_norm": 0.5809375643730164,
      "learning_rate": 8.34e-05,
      "loss": 0.4215,
      "step": 418
    },
    {
      "epoch": 0.0836,
      "loss_align": 0.4018477201461792,
      "loss_contrastive": 0.036357372999191284,
      "loss_gen": 0.19712625443935394,
      "neg_sim_mean": 0.5345096588134766,
      "pos_sim_mean": 0.5981522798538208,
      "step": 418
    },
    {
      "epoch": 0.0838,
      "grad_norm": 3.737154722213745,
      "learning_rate": 8.36e-05,
      "loss": 0.6033,
      "step": 419
    },
    {
      "epoch": 0.0838,
      "loss_align": 0.48143160343170166,
      "loss_contrastive": 0.05035760998725891,
      "loss_gen": 0.25971853733062744,
      "neg_sim_mean": 0.46892601251602173,
      "pos_sim_mean": 0.5185683965682983,
      "step": 419
    },
    {
      "epoch": 0.084,
      "grad_norm": 3.4813711643218994,
      "learning_rate": 8.38e-05,
      "loss": 0.7472,
      "step": 420
    },
    {
      "epoch": 0.084,
      "loss_align": 0.337583065032959,
      "loss_contrastive": 0.11215651035308838,
      "loss_gen": 0.45489996671676636,
      "neg_sim_mean": 0.6745734214782715,
      "pos_sim_mean": 0.662416934967041,
      "step": 420
    },
    {
      "epoch": 0.0842,
      "grad_norm": 6.716109752655029,
      "learning_rate": 8.4e-05,
      "loss": 0.8059,
      "step": 421
    },
    {
      "epoch": 0.0842,
      "loss_align": 0.3975580930709839,
      "loss_contrastive": 0.05564451217651367,
      "loss_gen": 0.17397984862327576,
      "neg_sim_mean": 0.5580863952636719,
      "pos_sim_mean": 0.6024419069290161,
      "step": 421
    },
    {
      "epoch": 0.0844,
      "grad_norm": 2.6638877391815186,
      "learning_rate": 8.42e-05,
      "loss": 0.5782,
      "step": 422
    },
    {
      "epoch": 0.0844,
      "loss_align": 0.3838083744049072,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0021254473831504583,
      "neg_sim_mean": 0.48517879843711853,
      "pos_sim_mean": 0.6161916255950928,
      "step": 422
    },
    {
      "epoch": 0.0846,
      "grad_norm": 0.5322901010513306,
      "learning_rate": 8.44e-05,
      "loss": 0.3859,
      "step": 423
    },
    {
      "epoch": 0.0846,
      "loss_align": 0.5175900459289551,
      "loss_contrastive": 0.10224759578704834,
      "loss_gen": 0.0050362879410386086,
      "neg_sim_mean": 0.4846575856208801,
      "pos_sim_mean": 0.4824099838733673,
      "step": 423
    },
    {
      "epoch": 0.0848,
      "grad_norm": 0.9296917915344238,
      "learning_rate": 8.46e-05,
      "loss": 0.5349,
      "step": 424
    },
    {
      "epoch": 0.0848,
      "loss_align": 0.51645427942276,
      "loss_contrastive": 0.031508803367614746,
      "loss_gen": 0.34910279512405396,
      "neg_sim_mean": 0.4150545299053192,
      "pos_sim_mean": 0.48354572057724,
      "step": 424
    },
    {
      "epoch": 0.085,
      "grad_norm": 4.581533432006836,
      "learning_rate": 8.48e-05,
      "loss": 0.8693,
      "step": 425
    },
    {
      "epoch": 0.085,
      "loss_align": 0.5146937370300293,
      "loss_contrastive": 0.10222062468528748,
      "loss_gen": 0.3287046253681183,
      "neg_sim_mean": 0.48752692341804504,
      "pos_sim_mean": 0.4853062927722931,
      "step": 425
    },
    {
      "epoch": 0.0852,
      "grad_norm": 3.842251777648926,
      "learning_rate": 8.5e-05,
      "loss": 0.8557,
      "step": 426
    },
    {
      "epoch": 0.0852,
      "loss_align": 0.5757423639297485,
      "loss_contrastive": 0.15635553002357483,
      "loss_gen": 0.02473149076104164,
      "neg_sim_mean": 0.48061317205429077,
      "pos_sim_mean": 0.42425763607025146,
      "step": 426
    },
    {
      "epoch": 0.0854,
      "grad_norm": 2.119562864303589,
      "learning_rate": 8.52e-05,
      "loss": 0.6192,
      "step": 427
    },
    {
      "epoch": 0.0854,
      "loss_align": 0.3455466628074646,
      "loss_contrastive": 0.10221070051193237,
      "loss_gen": 0.4673607051372528,
      "neg_sim_mean": 0.6566640138626099,
      "pos_sim_mean": 0.6544533371925354,
      "step": 427
    },
    {
      "epoch": 0.0856,
      "grad_norm": 4.731100559234619,
      "learning_rate": 8.54e-05,
      "loss": 0.8252,
      "step": 428
    },
    {
      "epoch": 0.0856,
      "loss_align": 0.43048369884490967,
      "loss_contrastive": 0.1466568410396576,
      "loss_gen": 0.18964815139770508,
      "neg_sim_mean": 0.6161731481552124,
      "pos_sim_mean": 0.5695163011550903,
      "step": 428
    },
    {
      "epoch": 0.0858,
      "grad_norm": 3.0134031772613525,
      "learning_rate": 8.560000000000001e-05,
      "loss": 0.6377,
      "step": 429
    },
    {
      "epoch": 0.0858,
      "loss_align": 0.404013991355896,
      "loss_contrastive": 0.0862489640712738,
      "loss_gen": 0.38108062744140625,
      "neg_sim_mean": 0.5822349786758423,
      "pos_sim_mean": 0.595986008644104,
      "step": 429
    },
    {
      "epoch": 0.086,
      "grad_norm": 4.471373081207275,
      "learning_rate": 8.58e-05,
      "loss": 0.7954,
      "step": 430
    },
    {
      "epoch": 0.086,
      "loss_align": 0.5015715956687927,
      "loss_contrastive": 0.11473885178565979,
      "loss_gen": 0.19762149453163147,
      "neg_sim_mean": 0.5131672620773315,
      "pos_sim_mean": 0.4984284043312073,
      "step": 430
    },
    {
      "epoch": 0.0862,
      "grad_norm": 3.225703716278076,
      "learning_rate": 8.6e-05,
      "loss": 0.713,
      "step": 431
    },
    {
      "epoch": 0.0862,
      "loss_align": 0.3461792469024658,
      "loss_contrastive": 0.062429845333099365,
      "loss_gen": 0.35574987530708313,
      "neg_sim_mean": 0.6162505745887756,
      "pos_sim_mean": 0.6538207530975342,
      "step": 431
    },
    {
      "epoch": 0.0864,
      "grad_norm": 4.347576141357422,
      "learning_rate": 8.620000000000001e-05,
      "loss": 0.7094,
      "step": 432
    },
    {
      "epoch": 0.0864,
      "loss_align": 0.5761409997940063,
      "loss_contrastive": 0.0924917459487915,
      "loss_gen": 0.1860271841287613,
      "neg_sim_mean": 0.41635075211524963,
      "pos_sim_mean": 0.42385900020599365,
      "step": 432
    },
    {
      "epoch": 0.0866,
      "grad_norm": 2.796299457550049,
      "learning_rate": 8.64e-05,
      "loss": 0.7733,
      "step": 433
    },
    {
      "epoch": 0.0866,
      "loss_align": 0.45419394969940186,
      "loss_contrastive": 0.09727498888969421,
      "loss_gen": 0.34615373611450195,
      "neg_sim_mean": 0.5430810451507568,
      "pos_sim_mean": 0.5458060503005981,
      "step": 433
    },
    {
      "epoch": 0.0868,
      "grad_norm": 4.942275047302246,
      "learning_rate": 8.66e-05,
      "loss": 0.812,
      "step": 434
    },
    {
      "epoch": 0.0868,
      "loss_align": 0.5397575497627258,
      "loss_contrastive": 0.07443892955780029,
      "loss_gen": 0.5285648107528687,
      "neg_sim_mean": 0.43468138575553894,
      "pos_sim_mean": 0.46024245023727417,
      "step": 434
    },
    {
      "epoch": 0.087,
      "grad_norm": 4.7968854904174805,
      "learning_rate": 8.680000000000001e-05,
      "loss": 1.0773,
      "step": 435
    },
    {
      "epoch": 0.087,
      "loss_align": 0.609548032283783,
      "loss_contrastive": 0.1967945098876953,
      "loss_gen": 0.001061509014107287,
      "neg_sim_mean": 0.48724648356437683,
      "pos_sim_mean": 0.39045196771621704,
      "step": 435
    },
    {
      "epoch": 0.0872,
      "grad_norm": 0.6225231289863586,
      "learning_rate": 8.7e-05,
      "loss": 0.6342,
      "step": 436
    },
    {
      "epoch": 0.0872,
      "loss_align": 0.40812259912490845,
      "loss_contrastive": 0.07902833819389343,
      "loss_gen": 0.4341534376144409,
      "neg_sim_mean": 0.5709057450294495,
      "pos_sim_mean": 0.5918774008750916,
      "step": 436
    },
    {
      "epoch": 0.0874,
      "grad_norm": 4.448585033416748,
      "learning_rate": 8.72e-05,
      "loss": 0.8518,
      "step": 437
    },
    {
      "epoch": 0.0874,
      "loss_align": 0.41063010692596436,
      "loss_contrastive": 0.12966647744178772,
      "loss_gen": 0.2924273908138275,
      "neg_sim_mean": 0.6190363764762878,
      "pos_sim_mean": 0.5893698930740356,
      "step": 437
    },
    {
      "epoch": 0.0876,
      "grad_norm": 3.7536420822143555,
      "learning_rate": 8.740000000000001e-05,
      "loss": 0.7186,
      "step": 438
    },
    {
      "epoch": 0.0876,
      "loss_align": 0.5961723923683167,
      "loss_contrastive": 0.15028443932533264,
      "loss_gen": 0.4348244369029999,
      "neg_sim_mean": 0.45411205291748047,
      "pos_sim_mean": 0.40382760763168335,
      "step": 438
    },
    {
      "epoch": 0.0878,
      "grad_norm": 4.556652069091797,
      "learning_rate": 8.76e-05,
      "loss": 1.049,
      "step": 439
    },
    {
      "epoch": 0.0878,
      "loss_align": 0.5016962885856628,
      "loss_contrastive": 0.040253251791000366,
      "loss_gen": 0.20099440217018127,
      "neg_sim_mean": 0.438556969165802,
      "pos_sim_mean": 0.49830371141433716,
      "step": 439
    },
    {
      "epoch": 0.088,
      "grad_norm": 3.290280818939209,
      "learning_rate": 8.78e-05,
      "loss": 0.7075,
      "step": 440
    },
    {
      "epoch": 0.088,
      "loss_align": 0.3871675133705139,
      "loss_contrastive": 0.10211366415023804,
      "loss_gen": 0.252108633518219,
      "neg_sim_mean": 0.6149461269378662,
      "pos_sim_mean": 0.6128324866294861,
      "step": 440
    },
    {
      "epoch": 0.0882,
      "grad_norm": 3.886051893234253,
      "learning_rate": 8.800000000000001e-05,
      "loss": 0.6515,
      "step": 441
    },
    {
      "epoch": 0.0882,
      "loss_align": 0.46764349937438965,
      "loss_contrastive": 0.13494965434074402,
      "loss_gen": 0.46818894147872925,
      "neg_sim_mean": 0.5673061609268188,
      "pos_sim_mean": 0.5323565006256104,
      "step": 441
    },
    {
      "epoch": 0.0884,
      "grad_norm": 4.945428371429443,
      "learning_rate": 8.82e-05,
      "loss": 0.952,
      "step": 442
    },
    {
      "epoch": 0.0884,
      "loss_align": 0.5334311723709106,
      "loss_contrastive": 0.1679806411266327,
      "loss_gen": 0.3521794378757477,
      "neg_sim_mean": 0.5345494747161865,
      "pos_sim_mean": 0.46656882762908936,
      "step": 442
    },
    {
      "epoch": 0.0886,
      "grad_norm": 4.946000576019287,
      "learning_rate": 8.840000000000001e-05,
      "loss": 0.9058,
      "step": 443
    },
    {
      "epoch": 0.0886,
      "loss_align": 0.38061273097991943,
      "loss_contrastive": 0.11504662036895752,
      "loss_gen": 0.4185822010040283,
      "neg_sim_mean": 0.6344338655471802,
      "pos_sim_mean": 0.6193872690200806,
      "step": 443
    },
    {
      "epoch": 0.0888,
      "grad_norm": 4.045678615570068,
      "learning_rate": 8.86e-05,
      "loss": 0.813,
      "step": 444
    },
    {
      "epoch": 0.0888,
      "loss_align": 0.3717274069786072,
      "loss_contrastive": 0.13050305843353271,
      "loss_gen": 0.3968946039676666,
      "neg_sim_mean": 0.6587756276130676,
      "pos_sim_mean": 0.6282725930213928,
      "step": 444
    },
    {
      "epoch": 0.089,
      "grad_norm": 4.264738082885742,
      "learning_rate": 8.88e-05,
      "loss": 0.7843,
      "step": 445
    },
    {
      "epoch": 0.089,
      "loss_align": 0.4742838740348816,
      "loss_contrastive": 0.14173713326454163,
      "loss_gen": 0.3077826499938965,
      "neg_sim_mean": 0.5674532651901245,
      "pos_sim_mean": 0.5257161259651184,
      "step": 445
    },
    {
      "epoch": 0.0892,
      "grad_norm": 4.667286396026611,
      "learning_rate": 8.900000000000001e-05,
      "loss": 0.7991,
      "step": 446
    },
    {
      "epoch": 0.0892,
      "loss_align": 0.34590429067611694,
      "loss_contrastive": 0.10349798202514648,
      "loss_gen": 0.2204710692167282,
      "neg_sim_mean": 0.6575936675071716,
      "pos_sim_mean": 0.6540957093238831,
      "step": 446
    },
    {
      "epoch": 0.0894,
      "grad_norm": 3.3027095794677734,
      "learning_rate": 8.92e-05,
      "loss": 0.5788,
      "step": 447
    },
    {
      "epoch": 0.0894,
      "loss_align": 0.3956815004348755,
      "loss_contrastive": 0.08602344989776611,
      "loss_gen": 0.30913570523262024,
      "neg_sim_mean": 0.5903419256210327,
      "pos_sim_mean": 0.6043184995651245,
      "step": 447
    },
    {
      "epoch": 0.0896,
      "grad_norm": 4.217752456665039,
      "learning_rate": 8.94e-05,
      "loss": 0.7151,
      "step": 448
    },
    {
      "epoch": 0.0896,
      "loss_align": 0.5181633234024048,
      "loss_contrastive": 0.10474377870559692,
      "loss_gen": 0.19797468185424805,
      "neg_sim_mean": 0.486580491065979,
      "pos_sim_mean": 0.4818367063999176,
      "step": 448
    },
    {
      "epoch": 0.0898,
      "grad_norm": 3.511166572570801,
      "learning_rate": 8.960000000000001e-05,
      "loss": 0.7287,
      "step": 449
    },
    {
      "epoch": 0.0898,
      "loss_align": 0.5723294019699097,
      "loss_contrastive": 0.2052057683467865,
      "loss_gen": 0.2068861424922943,
      "neg_sim_mean": 0.5328763723373413,
      "pos_sim_mean": 0.42767059803009033,
      "step": 449
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.545010805130005,
      "learning_rate": 8.98e-05,
      "loss": 0.8038,
      "step": 450
    },
    {
      "epoch": 0.09,
      "loss_align": 0.5471369028091431,
      "loss_contrastive": 0.07023802399635315,
      "loss_gen": 0.2634942829608917,
      "neg_sim_mean": 0.42310112714767456,
      "pos_sim_mean": 0.45286309719085693,
      "step": 450
    },
    {
      "epoch": 0.0902,
      "grad_norm": 4.038363456726074,
      "learning_rate": 9e-05,
      "loss": 0.8191,
      "step": 451
    },
    {
      "epoch": 0.0902,
      "loss_align": 0.32911181449890137,
      "loss_contrastive": 0.0815589427947998,
      "loss_gen": 0.24098335206508636,
      "neg_sim_mean": 0.6524471044540405,
      "pos_sim_mean": 0.6708881855010986,
      "step": 451
    },
    {
      "epoch": 0.0904,
      "grad_norm": 3.482308864593506,
      "learning_rate": 9.020000000000001e-05,
      "loss": 0.5799,
      "step": 452
    },
    {
      "epoch": 0.0904,
      "loss_align": 0.42986053228378296,
      "loss_contrastive": 0.16141310334205627,
      "loss_gen": 0.22208437323570251,
      "neg_sim_mean": 0.6315525770187378,
      "pos_sim_mean": 0.570139467716217,
      "step": 452
    },
    {
      "epoch": 0.0906,
      "grad_norm": 4.315195083618164,
      "learning_rate": 9.04e-05,
      "loss": 0.6713,
      "step": 453
    },
    {
      "epoch": 0.0906,
      "loss_align": 0.43917399644851685,
      "loss_contrastive": 0.038366734981536865,
      "loss_gen": 0.06906841695308685,
      "neg_sim_mean": 0.4991927444934845,
      "pos_sim_mean": 0.5608260035514832,
      "step": 453
    },
    {
      "epoch": 0.0908,
      "grad_norm": 2.9627504348754883,
      "learning_rate": 9.06e-05,
      "loss": 0.5128,
      "step": 454
    },
    {
      "epoch": 0.0908,
      "loss_align": 0.565840482711792,
      "loss_contrastive": 0.246699720621109,
      "loss_gen": 0.3146573007106781,
      "neg_sim_mean": 0.5808592438697815,
      "pos_sim_mean": 0.434159517288208,
      "step": 454
    },
    {
      "epoch": 0.091,
      "grad_norm": 5.454765319824219,
      "learning_rate": 9.080000000000001e-05,
      "loss": 0.9101,
      "step": 455
    },
    {
      "epoch": 0.091,
      "loss_align": 0.46151572465896606,
      "loss_contrastive": 0.018484175205230713,
      "loss_gen": 0.16123898327350616,
      "neg_sim_mean": 0.4569684565067291,
      "pos_sim_mean": 0.5384842753410339,
      "step": 455
    },
    {
      "epoch": 0.0912,
      "grad_norm": 3.0848774909973145,
      "learning_rate": 9.1e-05,
      "loss": 0.625,
      "step": 456
    },
    {
      "epoch": 0.0912,
      "loss_align": 0.4455358386039734,
      "loss_contrastive": 0.12866535782814026,
      "loss_gen": 0.21712112426757812,
      "neg_sim_mean": 0.5831295251846313,
      "pos_sim_mean": 0.5544641613960266,
      "step": 456
    },
    {
      "epoch": 0.0914,
      "grad_norm": 2.7614476680755615,
      "learning_rate": 9.120000000000001e-05,
      "loss": 0.6781,
      "step": 457
    },
    {
      "epoch": 0.0914,
      "loss_align": 0.38033998012542725,
      "loss_contrastive": 0.1267942190170288,
      "loss_gen": 0.207331120967865,
      "neg_sim_mean": 0.6464542150497437,
      "pos_sim_mean": 0.6196600198745728,
      "step": 457
    },
    {
      "epoch": 0.0916,
      "grad_norm": 3.6359410285949707,
      "learning_rate": 9.140000000000001e-05,
      "loss": 0.6029,
      "step": 458
    },
    {
      "epoch": 0.0916,
      "loss_align": 0.5437890291213989,
      "loss_contrastive": 0.1404792070388794,
      "loss_gen": 0.03774070739746094,
      "neg_sim_mean": 0.49669018387794495,
      "pos_sim_mean": 0.4562109708786011,
      "step": 458
    },
    {
      "epoch": 0.0918,
      "grad_norm": 2.4140336513519287,
      "learning_rate": 9.16e-05,
      "loss": 0.5984,
      "step": 459
    },
    {
      "epoch": 0.0918,
      "loss_align": 0.5039806365966797,
      "loss_contrastive": 0.11066555976867676,
      "loss_gen": 0.18695782124996185,
      "neg_sim_mean": 0.5066848993301392,
      "pos_sim_mean": 0.4960193336009979,
      "step": 459
    },
    {
      "epoch": 0.092,
      "grad_norm": 2.881483316421509,
      "learning_rate": 9.180000000000001e-05,
      "loss": 0.7042,
      "step": 460
    },
    {
      "epoch": 0.092,
      "loss_align": 0.4646780490875244,
      "loss_contrastive": 0.02260798215866089,
      "loss_gen": 0.21971358358860016,
      "neg_sim_mean": 0.45792993903160095,
      "pos_sim_mean": 0.5353219509124756,
      "step": 460
    },
    {
      "epoch": 0.0922,
      "grad_norm": 3.33453106880188,
      "learning_rate": 9.200000000000001e-05,
      "loss": 0.6871,
      "step": 461
    },
    {
      "epoch": 0.0922,
      "loss_align": 0.5182762145996094,
      "loss_contrastive": 0.12314975261688232,
      "loss_gen": 0.5454838275909424,
      "neg_sim_mean": 0.5048735737800598,
      "pos_sim_mean": 0.481723815202713,
      "step": 461
    },
    {
      "epoch": 0.0924,
      "grad_norm": 5.307323932647705,
      "learning_rate": 9.22e-05,
      "loss": 1.0785,
      "step": 462
    },
    {
      "epoch": 0.0924,
      "loss_align": 0.42607980966567993,
      "loss_contrastive": 0.0184670090675354,
      "loss_gen": 0.013788396492600441,
      "neg_sim_mean": 0.49238720536231995,
      "pos_sim_mean": 0.5739201903343201,
      "step": 462
    },
    {
      "epoch": 0.0926,
      "grad_norm": 1.483919382095337,
      "learning_rate": 9.240000000000001e-05,
      "loss": 0.4421,
      "step": 463
    },
    {
      "epoch": 0.0926,
      "loss_align": 0.3141027092933655,
      "loss_contrastive": 0.13370460271835327,
      "loss_gen": 0.2591756284236908,
      "neg_sim_mean": 0.7196018695831299,
      "pos_sim_mean": 0.6858972907066345,
      "step": 463
    },
    {
      "epoch": 0.0928,
      "grad_norm": 3.481818675994873,
      "learning_rate": 9.260000000000001e-05,
      "loss": 0.5893,
      "step": 464
    },
    {
      "epoch": 0.0928,
      "loss_align": 0.3924556374549866,
      "loss_contrastive": 0.0896804928779602,
      "loss_gen": 0.1888902336359024,
      "neg_sim_mean": 0.5972248315811157,
      "pos_sim_mean": 0.6075443625450134,
      "step": 464
    },
    {
      "epoch": 0.093,
      "grad_norm": 3.736905813217163,
      "learning_rate": 9.28e-05,
      "loss": 0.5921,
      "step": 465
    },
    {
      "epoch": 0.093,
      "loss_align": 0.2965332269668579,
      "loss_contrastive": 0.06022268533706665,
      "loss_gen": 0.41440653800964355,
      "neg_sim_mean": 0.6636894345283508,
      "pos_sim_mean": 0.7034667730331421,
      "step": 465
    },
    {
      "epoch": 0.0932,
      "grad_norm": 4.371766090393066,
      "learning_rate": 9.300000000000001e-05,
      "loss": 0.7182,
      "step": 466
    },
    {
      "epoch": 0.0932,
      "loss_align": 0.4635568857192993,
      "loss_contrastive": 0.07480117678642273,
      "loss_gen": 0.3021027147769928,
      "neg_sim_mean": 0.5112442970275879,
      "pos_sim_mean": 0.5364431142807007,
      "step": 466
    },
    {
      "epoch": 0.0934,
      "grad_norm": 3.3767004013061523,
      "learning_rate": 9.320000000000002e-05,
      "loss": 0.7746,
      "step": 467
    },
    {
      "epoch": 0.0934,
      "loss_align": 0.38195234537124634,
      "loss_contrastive": 0.10075265169143677,
      "loss_gen": 0.5963585376739502,
      "neg_sim_mean": 0.6188002824783325,
      "pos_sim_mean": 0.6180476546287537,
      "step": 467
    },
    {
      "epoch": 0.0936,
      "grad_norm": 4.772719383239746,
      "learning_rate": 9.340000000000001e-05,
      "loss": 0.9904,
      "step": 468
    },
    {
      "epoch": 0.0936,
      "loss_align": 0.569688081741333,
      "loss_contrastive": 0.17260253429412842,
      "loss_gen": 0.19415995478630066,
      "neg_sim_mean": 0.5029144287109375,
      "pos_sim_mean": 0.4303118884563446,
      "step": 468
    },
    {
      "epoch": 0.0938,
      "grad_norm": 3.20894718170166,
      "learning_rate": 9.360000000000001e-05,
      "loss": 0.7846,
      "step": 469
    },
    {
      "epoch": 0.0938,
      "loss_align": 0.5344507098197937,
      "loss_contrastive": 0.20410498976707458,
      "loss_gen": 0.1976403445005417,
      "neg_sim_mean": 0.5696542859077454,
      "pos_sim_mean": 0.4655492901802063,
      "step": 469
    },
    {
      "epoch": 0.094,
      "grad_norm": 3.2696621417999268,
      "learning_rate": 9.38e-05,
      "loss": 0.7566,
      "step": 470
    },
    {
      "epoch": 0.094,
      "loss_align": 0.3388940095901489,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005872477777302265,
      "neg_sim_mean": 0.491276353597641,
      "pos_sim_mean": 0.6611059904098511,
      "step": 470
    },
    {
      "epoch": 0.0942,
      "grad_norm": 0.7722316384315491,
      "learning_rate": 9.4e-05,
      "loss": 0.3448,
      "step": 471
    },
    {
      "epoch": 0.0942,
      "loss_align": 0.5227543711662292,
      "loss_contrastive": 0.25504520535469055,
      "loss_gen": 0.48536399006843567,
      "neg_sim_mean": 0.6322908401489258,
      "pos_sim_mean": 0.47724562883377075,
      "step": 471
    },
    {
      "epoch": 0.0944,
      "grad_norm": 4.05070686340332,
      "learning_rate": 9.42e-05,
      "loss": 1.0387,
      "step": 472
    },
    {
      "epoch": 0.0944,
      "loss_align": 0.38469940423965454,
      "loss_contrastive": 0.06871145963668823,
      "loss_gen": 0.13671478629112244,
      "neg_sim_mean": 0.5840120315551758,
      "pos_sim_mean": 0.6153005957603455,
      "step": 472
    },
    {
      "epoch": 0.0946,
      "grad_norm": 2.4582037925720215,
      "learning_rate": 9.44e-05,
      "loss": 0.5297,
      "step": 473
    },
    {
      "epoch": 0.0946,
      "loss_align": 0.39132893085479736,
      "loss_contrastive": 0.11676353216171265,
      "loss_gen": 0.44765704870224,
      "neg_sim_mean": 0.6254345774650574,
      "pos_sim_mean": 0.6086710691452026,
      "step": 473
    },
    {
      "epoch": 0.0948,
      "grad_norm": 4.8486433029174805,
      "learning_rate": 9.46e-05,
      "loss": 0.853,
      "step": 474
    },
    {
      "epoch": 0.0948,
      "loss_align": 0.3451981544494629,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0051374235190451145,
      "neg_sim_mean": 0.4971766471862793,
      "pos_sim_mean": 0.6548018455505371,
      "step": 474
    },
    {
      "epoch": 0.095,
      "grad_norm": 0.6763274669647217,
      "learning_rate": 9.48e-05,
      "loss": 0.3503,
      "step": 475
    },
    {
      "epoch": 0.095,
      "loss_align": 0.4478611350059509,
      "loss_contrastive": 0.2110518515110016,
      "loss_gen": 0.4418632388114929,
      "neg_sim_mean": 0.6631907224655151,
      "pos_sim_mean": 0.5521388649940491,
      "step": 475
    },
    {
      "epoch": 0.0952,
      "grad_norm": 3.757988452911377,
      "learning_rate": 9.5e-05,
      "loss": 0.9151,
      "step": 476
    },
    {
      "epoch": 0.0952,
      "loss_align": 0.3650110960006714,
      "loss_contrastive": 0.058017194271087646,
      "loss_gen": 0.3598990738391876,
      "neg_sim_mean": 0.5930060744285583,
      "pos_sim_mean": 0.6349889039993286,
      "step": 476
    },
    {
      "epoch": 0.0954,
      "grad_norm": 3.324298858642578,
      "learning_rate": 9.52e-05,
      "loss": 0.7319,
      "step": 477
    },
    {
      "epoch": 0.0954,
      "loss_align": 0.3855522871017456,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006627580150961876,
      "neg_sim_mean": 0.49676355719566345,
      "pos_sim_mean": 0.6144477128982544,
      "step": 477
    },
    {
      "epoch": 0.0956,
      "grad_norm": 0.8676964640617371,
      "learning_rate": 9.54e-05,
      "loss": 0.3922,
      "step": 478
    },
    {
      "epoch": 0.0956,
      "loss_align": 0.4780313968658447,
      "loss_contrastive": 0.10453823208808899,
      "loss_gen": 0.23325708508491516,
      "neg_sim_mean": 0.5265068411827087,
      "pos_sim_mean": 0.5219686031341553,
      "step": 478
    },
    {
      "epoch": 0.0958,
      "grad_norm": 2.5136210918426514,
      "learning_rate": 9.56e-05,
      "loss": 0.7238,
      "step": 479
    },
    {
      "epoch": 0.0958,
      "loss_align": 0.3759951591491699,
      "loss_contrastive": 0.1004229187965393,
      "loss_gen": 0.41034621000289917,
      "neg_sim_mean": 0.6244277358055115,
      "pos_sim_mean": 0.6240048408508301,
      "step": 479
    },
    {
      "epoch": 0.096,
      "grad_norm": 3.7067666053771973,
      "learning_rate": 9.58e-05,
      "loss": 0.7984,
      "step": 480
    },
    {
      "epoch": 0.096,
      "loss_align": 0.4519369602203369,
      "loss_contrastive": 0.06731036305427551,
      "loss_gen": 0.3692488670349121,
      "neg_sim_mean": 0.5153734087944031,
      "pos_sim_mean": 0.5480630397796631,
      "step": 480
    },
    {
      "epoch": 0.0962,
      "grad_norm": 4.446903228759766,
      "learning_rate": 9.6e-05,
      "loss": 0.8293,
      "step": 481
    },
    {
      "epoch": 0.0962,
      "loss_align": 0.2625422477722168,
      "loss_contrastive": 0.033325910568237305,
      "loss_gen": 0.32494333386421204,
      "neg_sim_mean": 0.6707836389541626,
      "pos_sim_mean": 0.7374577522277832,
      "step": 481
    },
    {
      "epoch": 0.0964,
      "grad_norm": 3.24831223487854,
      "learning_rate": 9.620000000000001e-05,
      "loss": 0.5915,
      "step": 482
    },
    {
      "epoch": 0.0964,
      "loss_align": 0.4706183671951294,
      "loss_contrastive": 0.07452866435050964,
      "loss_gen": 0.0011397392954677343,
      "neg_sim_mean": 0.5039103031158447,
      "pos_sim_mean": 0.5293816328048706,
      "step": 482
    },
    {
      "epoch": 0.0966,
      "grad_norm": 0.538504421710968,
      "learning_rate": 9.64e-05,
      "loss": 0.4807,
      "step": 483
    },
    {
      "epoch": 0.0966,
      "loss_align": 0.3782179355621338,
      "loss_contrastive": 0.12761640548706055,
      "loss_gen": 0.23315215110778809,
      "neg_sim_mean": 0.6493984460830688,
      "pos_sim_mean": 0.6217820644378662,
      "step": 483
    },
    {
      "epoch": 0.0968,
      "grad_norm": 2.6561083793640137,
      "learning_rate": 9.66e-05,
      "loss": 0.6267,
      "step": 484
    },
    {
      "epoch": 0.0968,
      "loss_align": 0.39888322353363037,
      "loss_contrastive": 0.1252487301826477,
      "loss_gen": 0.3898780643939972,
      "neg_sim_mean": 0.6263654828071594,
      "pos_sim_mean": 0.6011167764663696,
      "step": 484
    },
    {
      "epoch": 0.097,
      "grad_norm": 4.325795650482178,
      "learning_rate": 9.680000000000001e-05,
      "loss": 0.8038,
      "step": 485
    },
    {
      "epoch": 0.097,
      "loss_align": 0.45941162109375,
      "loss_contrastive": 0.276534765958786,
      "loss_gen": 0.18923167884349823,
      "neg_sim_mean": 0.7171231508255005,
      "pos_sim_mean": 0.54058837890625,
      "step": 485
    },
    {
      "epoch": 0.0972,
      "grad_norm": 2.976823568344116,
      "learning_rate": 9.7e-05,
      "loss": 0.6818,
      "step": 486
    },
    {
      "epoch": 0.0972,
      "loss_align": 0.40487009286880493,
      "loss_contrastive": 0.2043611705303192,
      "loss_gen": 0.527759850025177,
      "neg_sim_mean": 0.6994910836219788,
      "pos_sim_mean": 0.5951299071311951,
      "step": 486
    },
    {
      "epoch": 0.0974,
      "grad_norm": 4.372931957244873,
      "learning_rate": 9.72e-05,
      "loss": 0.9572,
      "step": 487
    },
    {
      "epoch": 0.0974,
      "loss_align": 0.48635345697402954,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1348273605108261,
      "neg_sim_mean": 0.36578047275543213,
      "pos_sim_mean": 0.5136465430259705,
      "step": 487
    },
    {
      "epoch": 0.0976,
      "grad_norm": 2.4796011447906494,
      "learning_rate": 9.74e-05,
      "loss": 0.6212,
      "step": 488
    },
    {
      "epoch": 0.0976,
      "loss_align": 0.3338127136230469,
      "loss_contrastive": 0.028707265853881836,
      "loss_gen": 0.26140281558036804,
      "neg_sim_mean": 0.594894528388977,
      "pos_sim_mean": 0.6661872863769531,
      "step": 488
    },
    {
      "epoch": 0.0978,
      "grad_norm": 2.785179376602173,
      "learning_rate": 9.76e-05,
      "loss": 0.5987,
      "step": 489
    },
    {
      "epoch": 0.0978,
      "loss_align": 0.5446986556053162,
      "loss_contrastive": 0.14168569445610046,
      "loss_gen": 0.31998562812805176,
      "neg_sim_mean": 0.4969870448112488,
      "pos_sim_mean": 0.45530134439468384,
      "step": 489
    },
    {
      "epoch": 0.098,
      "grad_norm": 2.9265170097351074,
      "learning_rate": 9.78e-05,
      "loss": 0.8817,
      "step": 490
    },
    {
      "epoch": 0.098,
      "loss_align": 0.42032361030578613,
      "loss_contrastive": 0.02114245295524597,
      "loss_gen": 0.03888625279068947,
      "neg_sim_mean": 0.5008188486099243,
      "pos_sim_mean": 0.5796763896942139,
      "step": 490
    },
    {
      "epoch": 0.0982,
      "grad_norm": 1.9794269800186157,
      "learning_rate": 9.8e-05,
      "loss": 0.4617,
      "step": 491
    },
    {
      "epoch": 0.0982,
      "loss_align": 0.4856947660446167,
      "loss_contrastive": 0.17035439610481262,
      "loss_gen": 0.2918778955936432,
      "neg_sim_mean": 0.5846596360206604,
      "pos_sim_mean": 0.5143052339553833,
      "step": 491
    },
    {
      "epoch": 0.0984,
      "grad_norm": 3.150728464126587,
      "learning_rate": 9.82e-05,
      "loss": 0.798,
      "step": 492
    },
    {
      "epoch": 0.0984,
      "loss_align": 0.48964905738830566,
      "loss_contrastive": 0.09003427624702454,
      "loss_gen": 0.021161606535315514,
      "neg_sim_mean": 0.5003852248191833,
      "pos_sim_mean": 0.5103509426116943,
      "step": 492
    },
    {
      "epoch": 0.0986,
      "grad_norm": 1.6929090023040771,
      "learning_rate": 9.84e-05,
      "loss": 0.5216,
      "step": 493
    },
    {
      "epoch": 0.0986,
      "loss_align": 0.33209019899368286,
      "loss_contrastive": 0.055351316928863525,
      "loss_gen": 0.3759758174419403,
      "neg_sim_mean": 0.6232610940933228,
      "pos_sim_mean": 0.6679098010063171,
      "step": 493
    },
    {
      "epoch": 0.0988,
      "grad_norm": 3.8076603412628174,
      "learning_rate": 9.86e-05,
      "loss": 0.7147,
      "step": 494
    },
    {
      "epoch": 0.0988,
      "loss_align": 0.3515878915786743,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3064098656177521,
      "neg_sim_mean": 0.5056571960449219,
      "pos_sim_mean": 0.6484121084213257,
      "step": 494
    },
    {
      "epoch": 0.099,
      "grad_norm": 3.2578043937683105,
      "learning_rate": 9.88e-05,
      "loss": 0.658,
      "step": 495
    },
    {
      "epoch": 0.099,
      "loss_align": 0.5017431974411011,
      "loss_contrastive": 0.2374567985534668,
      "loss_gen": 0.29842403531074524,
      "neg_sim_mean": 0.6357135772705078,
      "pos_sim_mean": 0.49825677275657654,
      "step": 495
    },
    {
      "epoch": 0.0992,
      "grad_norm": 3.167583465576172,
      "learning_rate": 9.900000000000001e-05,
      "loss": 0.8287,
      "step": 496
    },
    {
      "epoch": 0.0992,
      "loss_align": 0.4902908205986023,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25812309980392456,
      "neg_sim_mean": 0.40171271562576294,
      "pos_sim_mean": 0.5097091794013977,
      "step": 496
    },
    {
      "epoch": 0.0994,
      "grad_norm": 2.880500078201294,
      "learning_rate": 9.92e-05,
      "loss": 0.7484,
      "step": 497
    },
    {
      "epoch": 0.0994,
      "loss_align": 0.4597048759460449,
      "loss_contrastive": 0.03746393322944641,
      "loss_gen": 0.31600359082221985,
      "neg_sim_mean": 0.47775906324386597,
      "pos_sim_mean": 0.5402951240539551,
      "step": 497
    },
    {
      "epoch": 0.0996,
      "grad_norm": 3.0270421504974365,
      "learning_rate": 9.94e-05,
      "loss": 0.7802,
      "step": 498
    },
    {
      "epoch": 0.0996,
      "loss_align": 0.3365094065666199,
      "loss_contrastive": 0.0815577507019043,
      "loss_gen": 0.22673964500427246,
      "neg_sim_mean": 0.6450483202934265,
      "pos_sim_mean": 0.6634905934333801,
      "step": 498
    },
    {
      "epoch": 0.0998,
      "grad_norm": 2.405831813812256,
      "learning_rate": 9.960000000000001e-05,
      "loss": 0.573,
      "step": 499
    },
    {
      "epoch": 0.0998,
      "loss_align": 0.46667754650115967,
      "loss_contrastive": 0.09215858578681946,
      "loss_gen": 0.18818432092666626,
      "neg_sim_mean": 0.5254810452461243,
      "pos_sim_mean": 0.5333224534988403,
      "step": 499
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.743401050567627,
      "learning_rate": 9.98e-05,
      "loss": 0.6659,
      "step": 500
    },
    {
      "epoch": 0.1,
      "loss_align": 0.4887056350708008,
      "loss_contrastive": 0.0836648941040039,
      "loss_gen": 0.007132143713533878,
      "neg_sim_mean": 0.4949592649936676,
      "pos_sim_mean": 0.5112943649291992,
      "step": 500
    },
    {
      "epoch": 0.1002,
      "grad_norm": 0.970007598400116,
      "learning_rate": 0.0001,
      "loss": 0.5059,
      "step": 501
    },
    {
      "epoch": 0.1002,
      "loss_align": 0.47226619720458984,
      "loss_contrastive": 0.06677857041358948,
      "loss_gen": 0.016111945733428,
      "neg_sim_mean": 0.4945123791694641,
      "pos_sim_mean": 0.5277338027954102,
      "step": 501
    },
    {
      "epoch": 0.1004,
      "grad_norm": 1.6721558570861816,
      "learning_rate": 9.99999878153037e-05,
      "loss": 0.4964,
      "step": 502
    },
    {
      "epoch": 0.1004,
      "loss_align": 0.4609861969947815,
      "loss_contrastive": 0.10937485098838806,
      "loss_gen": 0.19558942317962646,
      "neg_sim_mean": 0.548388659954071,
      "pos_sim_mean": 0.5390138030052185,
      "step": 502
    },
    {
      "epoch": 0.1006,
      "grad_norm": 2.8448097705841064,
      "learning_rate": 9.999995126122076e-05,
      "loss": 0.6697,
      "step": 503
    },
    {
      "epoch": 0.1006,
      "loss_align": 0.3278157114982605,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13446250557899475,
      "neg_sim_mean": 0.532453179359436,
      "pos_sim_mean": 0.6721842885017395,
      "step": 503
    },
    {
      "epoch": 0.1008,
      "grad_norm": 1.9590929746627808,
      "learning_rate": 9.999989033776898e-05,
      "loss": 0.4623,
      "step": 504
    },
    {
      "epoch": 0.1008,
      "loss_align": 0.4914957880973816,
      "loss_contrastive": 0.08290764689445496,
      "loss_gen": 0.0020252459216862917,
      "neg_sim_mean": 0.49141186475753784,
      "pos_sim_mean": 0.5085042119026184,
      "step": 504
    },
    {
      "epoch": 0.101,
      "grad_norm": 0.5561604499816895,
      "learning_rate": 9.999980504497803e-05,
      "loss": 0.5035,
      "step": 505
    },
    {
      "epoch": 0.101,
      "loss_align": 0.4197920560836792,
      "loss_contrastive": 0.05464252829551697,
      "loss_gen": 0.2086109071969986,
      "neg_sim_mean": 0.5348504781723022,
      "pos_sim_mean": 0.5802079439163208,
      "step": 505
    },
    {
      "epoch": 0.1012,
      "grad_norm": 2.7892799377441406,
      "learning_rate": 9.999969538288952e-05,
      "loss": 0.635,
      "step": 506
    },
    {
      "epoch": 0.1012,
      "loss_align": 0.39227402210235596,
      "loss_contrastive": 0.1581520438194275,
      "loss_gen": 0.30985382199287415,
      "neg_sim_mean": 0.6658779978752136,
      "pos_sim_mean": 0.607725977897644,
      "step": 506
    },
    {
      "epoch": 0.1014,
      "grad_norm": 2.960606575012207,
      "learning_rate": 9.999956135155687e-05,
      "loss": 0.7211,
      "step": 507
    },
    {
      "epoch": 0.1014,
      "loss_align": 0.5874917507171631,
      "loss_contrastive": 0.21548137068748474,
      "loss_gen": 0.299978107213974,
      "neg_sim_mean": 0.5279896259307861,
      "pos_sim_mean": 0.4125082492828369,
      "step": 507
    },
    {
      "epoch": 0.1016,
      "grad_norm": 3.33560848236084,
      "learning_rate": 9.999940295104544e-05,
      "loss": 0.9133,
      "step": 508
    },
    {
      "epoch": 0.1016,
      "loss_align": 0.5498975515365601,
      "loss_contrastive": 0.07746121287345886,
      "loss_gen": 0.23971131443977356,
      "neg_sim_mean": 0.42756369709968567,
      "pos_sim_mean": 0.45010247826576233,
      "step": 508
    },
    {
      "epoch": 0.1018,
      "grad_norm": 3.2006425857543945,
      "learning_rate": 9.999922018143242e-05,
      "loss": 0.7989,
      "step": 509
    },
    {
      "epoch": 0.1018,
      "loss_align": 0.5626869797706604,
      "loss_contrastive": 0.14349675178527832,
      "loss_gen": 0.008863704279065132,
      "neg_sim_mean": 0.4808097779750824,
      "pos_sim_mean": 0.4373130202293396,
      "step": 509
    },
    {
      "epoch": 0.102,
      "grad_norm": 1.0685726404190063,
      "learning_rate": 9.999901304280685e-05,
      "loss": 0.5888,
      "step": 510
    },
    {
      "epoch": 0.102,
      "loss_align": 0.45979154109954834,
      "loss_contrastive": 0.13974520564079285,
      "loss_gen": 0.28818628191947937,
      "neg_sim_mean": 0.579953670501709,
      "pos_sim_mean": 0.5402084589004517,
      "step": 510
    },
    {
      "epoch": 0.1022,
      "grad_norm": 2.9796035289764404,
      "learning_rate": 9.999878153526974e-05,
      "loss": 0.7647,
      "step": 511
    },
    {
      "epoch": 0.1022,
      "loss_align": 0.4668494462966919,
      "loss_contrastive": 0.015423506498336792,
      "loss_gen": 0.3880884349346161,
      "neg_sim_mean": 0.4485740661621094,
      "pos_sim_mean": 0.5331505537033081,
      "step": 511
    },
    {
      "epoch": 0.1024,
      "grad_norm": 3.597560405731201,
      "learning_rate": 9.99985256589339e-05,
      "loss": 0.8568,
      "step": 512
    },
    {
      "epoch": 0.1024,
      "loss_align": 0.4835778474807739,
      "loss_contrastive": 0.06183534860610962,
      "loss_gen": 0.0038468053098767996,
      "neg_sim_mean": 0.47825750708580017,
      "pos_sim_mean": 0.5164221525192261,
      "step": 512
    },
    {
      "epoch": 0.1026,
      "grad_norm": 0.6889423727989197,
      "learning_rate": 9.999824541392405e-05,
      "loss": 0.4948,
      "step": 513
    },
    {
      "epoch": 0.1026,
      "loss_align": 0.51387619972229,
      "loss_contrastive": 0.09265050292015076,
      "loss_gen": 0.0018236003816127777,
      "neg_sim_mean": 0.4787743091583252,
      "pos_sim_mean": 0.48612380027770996,
      "step": 513
    },
    {
      "epoch": 0.1028,
      "grad_norm": 0.5397586822509766,
      "learning_rate": 9.999794080037674e-05,
      "loss": 0.5268,
      "step": 514
    },
    {
      "epoch": 0.1028,
      "loss_align": 0.5212321877479553,
      "loss_contrastive": 0.15049925446510315,
      "loss_gen": 0.5150099992752075,
      "neg_sim_mean": 0.5292670726776123,
      "pos_sim_mean": 0.4787678122520447,
      "step": 514
    },
    {
      "epoch": 0.103,
      "grad_norm": 4.021456241607666,
      "learning_rate": 9.99976118184405e-05,
      "loss": 1.0543,
      "step": 515
    },
    {
      "epoch": 0.103,
      "loss_align": 0.4157896041870117,
      "loss_contrastive": 0.06657883524894714,
      "loss_gen": 0.33566585183143616,
      "neg_sim_mean": 0.5507892370223999,
      "pos_sim_mean": 0.5842103958129883,
      "step": 515
    },
    {
      "epoch": 0.1032,
      "grad_norm": 3.724581480026245,
      "learning_rate": 9.999725846827562e-05,
      "loss": 0.7594,
      "step": 516
    },
    {
      "epoch": 0.1032,
      "loss_align": 0.46773993968963623,
      "loss_contrastive": 0.06634357571601868,
      "loss_gen": 0.4768458604812622,
      "neg_sim_mean": 0.4986036419868469,
      "pos_sim_mean": 0.5322600603103638,
      "step": 516
    },
    {
      "epoch": 0.1034,
      "grad_norm": 3.910008668899536,
      "learning_rate": 9.999688075005433e-05,
      "loss": 0.9525,
      "step": 517
    },
    {
      "epoch": 0.1034,
      "loss_align": 0.4952429533004761,
      "loss_contrastive": 0.07131204009056091,
      "loss_gen": 0.002322487998753786,
      "neg_sim_mean": 0.4760690927505493,
      "pos_sim_mean": 0.5047570466995239,
      "step": 517
    },
    {
      "epoch": 0.1036,
      "grad_norm": 0.5586229562759399,
      "learning_rate": 9.999647866396073e-05,
      "loss": 0.5061,
      "step": 518
    },
    {
      "epoch": 0.1036,
      "loss_align": 0.39570140838623047,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0040769679471850395,
      "neg_sim_mean": 0.47331807017326355,
      "pos_sim_mean": 0.6042985916137695,
      "step": 518
    },
    {
      "epoch": 0.1038,
      "grad_norm": 0.6186424493789673,
      "learning_rate": 9.999605221019081e-05,
      "loss": 0.3998,
      "step": 519
    },
    {
      "epoch": 0.1038,
      "loss_align": 0.3948557376861572,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004842238966375589,
      "neg_sim_mean": 0.4708152413368225,
      "pos_sim_mean": 0.6051442623138428,
      "step": 519
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.6747104525566101,
      "learning_rate": 9.999560138895238e-05,
      "loss": 0.3997,
      "step": 520
    },
    {
      "epoch": 0.104,
      "loss_align": 0.44048595428466797,
      "loss_contrastive": 0.009174555540084839,
      "loss_gen": 0.0015654755989089608,
      "neg_sim_mean": 0.46868860721588135,
      "pos_sim_mean": 0.559514045715332,
      "step": 520
    },
    {
      "epoch": 0.1042,
      "grad_norm": 0.4921855926513672,
      "learning_rate": 9.999512620046522e-05,
      "loss": 0.4432,
      "step": 521
    },
    {
      "epoch": 0.1042,
      "loss_align": 0.6582093238830566,
      "loss_contrastive": 0.17487016320228577,
      "loss_gen": 0.5442762970924377,
      "neg_sim_mean": 0.4166608452796936,
      "pos_sim_mean": 0.34179067611694336,
      "step": 521
    },
    {
      "epoch": 0.1044,
      "grad_norm": 3.8642821311950684,
      "learning_rate": 9.999462664496088e-05,
      "loss": 1.2235,
      "step": 522
    },
    {
      "epoch": 0.1044,
      "loss_align": 0.443666934967041,
      "loss_contrastive": 0.12409207224845886,
      "loss_gen": 0.22674313187599182,
      "neg_sim_mean": 0.5804251432418823,
      "pos_sim_mean": 0.556333065032959,
      "step": 522
    },
    {
      "epoch": 0.1046,
      "grad_norm": 3.5880911350250244,
      "learning_rate": 9.999410272268285e-05,
      "loss": 0.6853,
      "step": 523
    },
    {
      "epoch": 0.1046,
      "loss_align": 0.6524568796157837,
      "loss_contrastive": 0.23690807819366455,
      "loss_gen": 0.27083703875541687,
      "neg_sim_mean": 0.48445120453834534,
      "pos_sim_mean": 0.3475431203842163,
      "step": 523
    },
    {
      "epoch": 0.1048,
      "grad_norm": 3.415421724319458,
      "learning_rate": 9.99935544338865e-05,
      "loss": 0.9517,
      "step": 524
    },
    {
      "epoch": 0.1048,
      "loss_align": 0.4039626717567444,
      "loss_contrastive": 0.0909741222858429,
      "loss_gen": 0.2557157278060913,
      "neg_sim_mean": 0.587011456489563,
      "pos_sim_mean": 0.5960373282432556,
      "step": 524
    },
    {
      "epoch": 0.105,
      "grad_norm": 2.706165075302124,
      "learning_rate": 9.999298177883903e-05,
      "loss": 0.6706,
      "step": 525
    },
    {
      "epoch": 0.105,
      "loss_align": 0.3269914984703064,
      "loss_contrastive": 0.07710474729537964,
      "loss_gen": 0.3164523243904114,
      "neg_sim_mean": 0.6501132249832153,
      "pos_sim_mean": 0.6730085015296936,
      "step": 525
    },
    {
      "epoch": 0.1052,
      "grad_norm": 2.8242194652557373,
      "learning_rate": 9.999238475781957e-05,
      "loss": 0.6527,
      "step": 526
    },
    {
      "epoch": 0.1052,
      "loss_align": 0.4367765784263611,
      "loss_contrastive": 0.07211491465568542,
      "loss_gen": 0.18719786405563354,
      "neg_sim_mean": 0.5353383421897888,
      "pos_sim_mean": 0.5632234215736389,
      "step": 526
    },
    {
      "epoch": 0.1054,
      "grad_norm": 2.7242627143859863,
      "learning_rate": 9.999176337111908e-05,
      "loss": 0.6326,
      "step": 527
    },
    {
      "epoch": 0.1054,
      "loss_align": 0.5569044947624207,
      "loss_contrastive": 0.12914639711380005,
      "loss_gen": 0.16006748378276825,
      "neg_sim_mean": 0.47224190831184387,
      "pos_sim_mean": 0.44309550523757935,
      "step": 527
    },
    {
      "epoch": 0.1056,
      "grad_norm": 2.3128318786621094,
      "learning_rate": 9.999111761904046e-05,
      "loss": 0.7325,
      "step": 528
    },
    {
      "epoch": 0.1056,
      "loss_align": 0.3989017605781555,
      "loss_contrastive": 0.10118627548217773,
      "loss_gen": 0.16920778155326843,
      "neg_sim_mean": 0.6022844910621643,
      "pos_sim_mean": 0.6010982394218445,
      "step": 528
    },
    {
      "epoch": 0.1058,
      "grad_norm": 2.400599956512451,
      "learning_rate": 9.999044750189838e-05,
      "loss": 0.5803,
      "step": 529
    },
    {
      "epoch": 0.1058,
      "loss_align": 0.5287190675735474,
      "loss_contrastive": 0.10080942511558533,
      "loss_gen": 0.2899731695652008,
      "neg_sim_mean": 0.47209036350250244,
      "pos_sim_mean": 0.47128093242645264,
      "step": 529
    },
    {
      "epoch": 0.106,
      "grad_norm": 3.2884511947631836,
      "learning_rate": 9.99897530200195e-05,
      "loss": 0.8308,
      "step": 530
    },
    {
      "epoch": 0.106,
      "loss_align": 0.5057154297828674,
      "loss_contrastive": 0.03153336048126221,
      "loss_gen": 0.3079286217689514,
      "neg_sim_mean": 0.42581793665885925,
      "pos_sim_mean": 0.49428457021713257,
      "step": 530
    },
    {
      "epoch": 0.1062,
      "grad_norm": 3.8651719093322754,
      "learning_rate": 9.998903417374228e-05,
      "loss": 0.8174,
      "step": 531
    },
    {
      "epoch": 0.1062,
      "loss_align": 0.4316657781600952,
      "loss_contrastive": 0.07338151335716248,
      "loss_gen": 0.5200848579406738,
      "neg_sim_mean": 0.5417157411575317,
      "pos_sim_mean": 0.5683342218399048,
      "step": 531
    },
    {
      "epoch": 0.1064,
      "grad_norm": 4.500788688659668,
      "learning_rate": 9.998829096341706e-05,
      "loss": 0.9606,
      "step": 532
    },
    {
      "epoch": 0.1064,
      "loss_align": 0.3878391981124878,
      "loss_contrastive": 0.05528676509857178,
      "loss_gen": 0.20077082514762878,
      "neg_sim_mean": 0.5674475431442261,
      "pos_sim_mean": 0.6121608018875122,
      "step": 532
    },
    {
      "epoch": 0.1066,
      "grad_norm": 2.937908411026001,
      "learning_rate": 9.998752338940612e-05,
      "loss": 0.5952,
      "step": 533
    },
    {
      "epoch": 0.1066,
      "loss_align": 0.3350294232368469,
      "loss_contrastive": 0.06865674257278442,
      "loss_gen": 0.3180646300315857,
      "neg_sim_mean": 0.6336272954940796,
      "pos_sim_mean": 0.6649705767631531,
      "step": 533
    },
    {
      "epoch": 0.1068,
      "grad_norm": 3.14916729927063,
      "learning_rate": 9.99867314520835e-05,
      "loss": 0.6613,
      "step": 534
    },
    {
      "epoch": 0.1068,
      "loss_align": 0.446502685546875,
      "loss_contrastive": 0.07159408926963806,
      "loss_gen": 0.2958413064479828,
      "neg_sim_mean": 0.5250914096832275,
      "pos_sim_mean": 0.553497314453125,
      "step": 534
    },
    {
      "epoch": 0.107,
      "grad_norm": 2.806112766265869,
      "learning_rate": 9.998591515183524e-05,
      "loss": 0.7509,
      "step": 535
    },
    {
      "epoch": 0.107,
      "loss_align": 0.4319499731063843,
      "loss_contrastive": 0.053706616163253784,
      "loss_gen": 0.28488072752952576,
      "neg_sim_mean": 0.521756649017334,
      "pos_sim_mean": 0.5680500268936157,
      "step": 535
    },
    {
      "epoch": 0.1072,
      "grad_norm": 3.3860392570495605,
      "learning_rate": 9.998507448905917e-05,
      "loss": 0.7233,
      "step": 536
    },
    {
      "epoch": 0.1072,
      "loss_align": 0.5064229965209961,
      "loss_contrastive": 0.21473941206932068,
      "loss_gen": 0.3445166051387787,
      "neg_sim_mean": 0.6083164215087891,
      "pos_sim_mean": 0.4935770034790039,
      "step": 536
    },
    {
      "epoch": 0.1074,
      "grad_norm": 3.06276273727417,
      "learning_rate": 9.9984209464165e-05,
      "loss": 0.8767,
      "step": 537
    },
    {
      "epoch": 0.1074,
      "loss_align": 0.43135297298431396,
      "loss_contrastive": 0.12252512574195862,
      "loss_gen": 0.286087304353714,
      "neg_sim_mean": 0.5911721587181091,
      "pos_sim_mean": 0.568647027015686,
      "step": 537
    },
    {
      "epoch": 0.1076,
      "grad_norm": 3.072375774383545,
      "learning_rate": 9.998332007757436e-05,
      "loss": 0.7321,
      "step": 538
    },
    {
      "epoch": 0.1076,
      "loss_align": 0.5642942190170288,
      "loss_contrastive": 0.1330878734588623,
      "loss_gen": 0.3787286579608917,
      "neg_sim_mean": 0.4687936305999756,
      "pos_sim_mean": 0.4357057511806488,
      "step": 538
    },
    {
      "epoch": 0.1078,
      "grad_norm": 4.038261890411377,
      "learning_rate": 9.998240632972073e-05,
      "loss": 0.959,
      "step": 539
    },
    {
      "epoch": 0.1078,
      "loss_align": 0.5260153412818909,
      "loss_contrastive": 0.11309781670570374,
      "loss_gen": 0.037278082221746445,
      "neg_sim_mean": 0.48708248138427734,
      "pos_sim_mean": 0.47398465871810913,
      "step": 539
    },
    {
      "epoch": 0.108,
      "grad_norm": 2.018883466720581,
      "learning_rate": 9.998146822104943e-05,
      "loss": 0.5769,
      "step": 540
    },
    {
      "epoch": 0.108,
      "loss_align": 0.4074360132217407,
      "loss_contrastive": 0.13019075989723206,
      "loss_gen": 0.4851318597793579,
      "neg_sim_mean": 0.6227547526359558,
      "pos_sim_mean": 0.5925639867782593,
      "step": 540
    },
    {
      "epoch": 0.1082,
      "grad_norm": 3.217215061187744,
      "learning_rate": 9.998050575201771e-05,
      "loss": 0.9082,
      "step": 541
    },
    {
      "epoch": 0.1082,
      "loss_align": 0.38387733697891235,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11464329808950424,
      "neg_sim_mean": 0.49110880494117737,
      "pos_sim_mean": 0.6161226630210876,
      "step": 541
    },
    {
      "epoch": 0.1084,
      "grad_norm": 2.941823720932007,
      "learning_rate": 9.997951892309468e-05,
      "loss": 0.4985,
      "step": 542
    },
    {
      "epoch": 0.1084,
      "loss_align": 0.411771297454834,
      "loss_contrastive": 0.13736459612846375,
      "loss_gen": 0.15727528929710388,
      "neg_sim_mean": 0.6255933046340942,
      "pos_sim_mean": 0.588228702545166,
      "step": 542
    },
    {
      "epoch": 0.1086,
      "grad_norm": 2.4264392852783203,
      "learning_rate": 9.997850773476126e-05,
      "loss": 0.5855,
      "step": 543
    },
    {
      "epoch": 0.1086,
      "loss_align": 0.45948725938796997,
      "loss_contrastive": 0.054339081048965454,
      "loss_gen": 0.05464860796928406,
      "neg_sim_mean": 0.49485182762145996,
      "pos_sim_mean": 0.54051274061203,
      "step": 543
    },
    {
      "epoch": 0.1088,
      "grad_norm": 2.1435513496398926,
      "learning_rate": 9.997747218751032e-05,
      "loss": 0.5207,
      "step": 544
    },
    {
      "epoch": 0.1088,
      "loss_align": 0.3934832811355591,
      "loss_contrastive": 0.07364034652709961,
      "loss_gen": 0.3741568922996521,
      "neg_sim_mean": 0.5801570415496826,
      "pos_sim_mean": 0.6065167188644409,
      "step": 544
    },
    {
      "epoch": 0.109,
      "grad_norm": 3.6743171215057373,
      "learning_rate": 9.997641228184655e-05,
      "loss": 0.7765,
      "step": 545
    },
    {
      "epoch": 0.109,
      "loss_align": 0.37418919801712036,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01183298509567976,
      "neg_sim_mean": 0.49789753556251526,
      "pos_sim_mean": 0.6258108019828796,
      "step": 545
    },
    {
      "epoch": 0.1092,
      "grad_norm": 1.1464004516601562,
      "learning_rate": 9.997532801828658e-05,
      "loss": 0.386,
      "step": 546
    },
    {
      "epoch": 0.1092,
      "loss_align": 0.39144933223724365,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009903489612042904,
      "neg_sim_mean": 0.4985063672065735,
      "pos_sim_mean": 0.6085506677627563,
      "step": 546
    },
    {
      "epoch": 0.1094,
      "grad_norm": 1.045884370803833,
      "learning_rate": 9.997421939735884e-05,
      "loss": 0.4014,
      "step": 547
    },
    {
      "epoch": 0.1094,
      "loss_align": 0.29664015769958496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010159339755773544,
      "neg_sim_mean": 0.4992067217826843,
      "pos_sim_mean": 0.703359842300415,
      "step": 547
    },
    {
      "epoch": 0.1096,
      "grad_norm": 1.1010795831680298,
      "learning_rate": 9.997308641960365e-05,
      "loss": 0.3068,
      "step": 548
    },
    {
      "epoch": 0.1096,
      "loss_align": 0.36793482303619385,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0023836807813495398,
      "neg_sim_mean": 0.4986743628978729,
      "pos_sim_mean": 0.6320651769638062,
      "step": 548
    },
    {
      "epoch": 0.1098,
      "grad_norm": 0.4622403383255005,
      "learning_rate": 9.997192908557323e-05,
      "loss": 0.3703,
      "step": 549
    },
    {
      "epoch": 0.1098,
      "loss_align": 0.4001031517982483,
      "loss_contrastive": 0.11103692650794983,
      "loss_gen": 0.23482680320739746,
      "neg_sim_mean": 0.610933780670166,
      "pos_sim_mean": 0.5998968482017517,
      "step": 549
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.6626534461975098,
      "learning_rate": 9.997074739583162e-05,
      "loss": 0.6483,
      "step": 550
    },
    {
      "epoch": 0.11,
      "loss_align": 0.4511202573776245,
      "loss_contrastive": 0.09007224440574646,
      "loss_gen": 0.23838414251804352,
      "neg_sim_mean": 0.5389519929885864,
      "pos_sim_mean": 0.5488797426223755,
      "step": 550
    },
    {
      "epoch": 0.1102,
      "grad_norm": 2.6960275173187256,
      "learning_rate": 9.99695413509548e-05,
      "loss": 0.7003,
      "step": 551
    },
    {
      "epoch": 0.1102,
      "loss_align": 0.47223716974258423,
      "loss_contrastive": 0.0221843421459198,
      "loss_gen": 0.34729084372520447,
      "neg_sim_mean": 0.44994717836380005,
      "pos_sim_mean": 0.5277628302574158,
      "step": 551
    },
    {
      "epoch": 0.1104,
      "grad_norm": 3.233841896057129,
      "learning_rate": 9.996831095153055e-05,
      "loss": 0.8222,
      "step": 552
    },
    {
      "epoch": 0.1104,
      "loss_align": 0.3795865774154663,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0006066289497539401,
      "neg_sim_mean": 0.49771952629089355,
      "pos_sim_mean": 0.6204134225845337,
      "step": 552
    },
    {
      "epoch": 0.1106,
      "grad_norm": 0.3713701665401459,
      "learning_rate": 9.996705619815855e-05,
      "loss": 0.3802,
      "step": 553
    },
    {
      "epoch": 0.1106,
      "loss_align": 0.4921460747718811,
      "loss_contrastive": 0.1149873435497284,
      "loss_gen": 0.4220338463783264,
      "neg_sim_mean": 0.5228412747383118,
      "pos_sim_mean": 0.5078539252281189,
      "step": 553
    },
    {
      "epoch": 0.1108,
      "grad_norm": 3.5696845054626465,
      "learning_rate": 9.99657770914504e-05,
      "loss": 0.928,
      "step": 554
    },
    {
      "epoch": 0.1108,
      "loss_align": 0.3234391212463379,
      "loss_contrastive": 0.06135845184326172,
      "loss_gen": 0.44023647904396057,
      "neg_sim_mean": 0.6379193067550659,
      "pos_sim_mean": 0.6765608787536621,
      "step": 554
    },
    {
      "epoch": 0.111,
      "grad_norm": 4.113605499267578,
      "learning_rate": 9.996447363202946e-05,
      "loss": 0.771,
      "step": 555
    },
    {
      "epoch": 0.111,
      "loss_align": 0.3189834952354431,
      "loss_contrastive": 0.02861100435256958,
      "loss_gen": 0.14383572340011597,
      "neg_sim_mean": 0.6096274852752686,
      "pos_sim_mean": 0.6810165047645569,
      "step": 555
    },
    {
      "epoch": 0.1112,
      "grad_norm": 1.7632378339767456,
      "learning_rate": 9.996314582053106e-05,
      "loss": 0.4663,
      "step": 556
    },
    {
      "epoch": 0.1112,
      "loss_align": 0.4782789945602417,
      "loss_contrastive": 0.08583769202232361,
      "loss_gen": 0.3610673248767853,
      "neg_sim_mean": 0.5075587034225464,
      "pos_sim_mean": 0.5217210054397583,
      "step": 556
    },
    {
      "epoch": 0.1114,
      "grad_norm": 3.5256285667419434,
      "learning_rate": 9.996179365760234e-05,
      "loss": 0.8496,
      "step": 557
    },
    {
      "epoch": 0.1114,
      "loss_align": 0.4461480379104614,
      "loss_contrastive": 0.11632207036018372,
      "loss_gen": 0.3422077000141144,
      "neg_sim_mean": 0.5701740384101868,
      "pos_sim_mean": 0.5538519620895386,
      "step": 557
    },
    {
      "epoch": 0.1116,
      "grad_norm": 2.9104201793670654,
      "learning_rate": 9.996041714390235e-05,
      "loss": 0.8023,
      "step": 558
    },
    {
      "epoch": 0.1116,
      "loss_align": 0.4056745767593384,
      "loss_contrastive": 0.11734244227409363,
      "loss_gen": 0.40818560123443604,
      "neg_sim_mean": 0.6116678714752197,
      "pos_sim_mean": 0.5943254232406616,
      "step": 558
    },
    {
      "epoch": 0.1118,
      "grad_norm": 3.2366340160369873,
      "learning_rate": 9.995901628010196e-05,
      "loss": 0.8279,
      "step": 559
    },
    {
      "epoch": 0.1118,
      "loss_align": 0.3900188207626343,
      "loss_contrastive": 0.13773059844970703,
      "loss_gen": 0.18023064732551575,
      "neg_sim_mean": 0.6477117538452148,
      "pos_sim_mean": 0.6099811792373657,
      "step": 559
    },
    {
      "epoch": 0.112,
      "grad_norm": 2.790740728378296,
      "learning_rate": 9.995759106688393e-05,
      "loss": 0.5868,
      "step": 560
    },
    {
      "epoch": 0.112,
      "loss_align": 0.33132117986679077,
      "loss_contrastive": 0.06728321313858032,
      "loss_gen": 0.4401678144931793,
      "neg_sim_mean": 0.6359620094299316,
      "pos_sim_mean": 0.6686788201332092,
      "step": 560
    },
    {
      "epoch": 0.1122,
      "grad_norm": 3.5018374919891357,
      "learning_rate": 9.995614150494293e-05,
      "loss": 0.7796,
      "step": 561
    },
    {
      "epoch": 0.1122,
      "loss_align": 0.36388254165649414,
      "loss_contrastive": 0.1587809920310974,
      "loss_gen": 0.34163492918014526,
      "neg_sim_mean": 0.6948984265327454,
      "pos_sim_mean": 0.6361174583435059,
      "step": 561
    },
    {
      "epoch": 0.1124,
      "grad_norm": 2.9726345539093018,
      "learning_rate": 9.995466759498542e-05,
      "loss": 0.7246,
      "step": 562
    },
    {
      "epoch": 0.1124,
      "loss_align": 0.3800450563430786,
      "loss_contrastive": 0.08307790756225586,
      "loss_gen": 0.34884870052337646,
      "neg_sim_mean": 0.6030328273773193,
      "pos_sim_mean": 0.6199549436569214,
      "step": 562
    },
    {
      "epoch": 0.1126,
      "grad_norm": 3.3977279663085938,
      "learning_rate": 9.995316933772978e-05,
      "loss": 0.7389,
      "step": 563
    },
    {
      "epoch": 0.1126,
      "loss_align": 0.3776051998138428,
      "loss_contrastive": 0.08995020389556885,
      "loss_gen": 0.37404412031173706,
      "neg_sim_mean": 0.6123449802398682,
      "pos_sim_mean": 0.6223948001861572,
      "step": 563
    },
    {
      "epoch": 0.1128,
      "grad_norm": 3.1040971279144287,
      "learning_rate": 9.995164673390625e-05,
      "loss": 0.7624,
      "step": 564
    },
    {
      "epoch": 0.1128,
      "loss_align": 0.3711404800415039,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007538940757513046,
      "neg_sim_mean": 0.4978642761707306,
      "pos_sim_mean": 0.6288595199584961,
      "step": 564
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.7858462929725647,
      "learning_rate": 9.995009978425691e-05,
      "loss": 0.3787,
      "step": 565
    },
    {
      "epoch": 0.113,
      "loss_align": 0.5092572569847107,
      "loss_contrastive": 0.048275142908096313,
      "loss_gen": 0.42117002606391907,
      "neg_sim_mean": 0.4390178918838501,
      "pos_sim_mean": 0.4907427430152893,
      "step": 565
    },
    {
      "epoch": 0.1132,
      "grad_norm": 3.1914124488830566,
      "learning_rate": 9.994852848953574e-05,
      "loss": 0.9362,
      "step": 566
    },
    {
      "epoch": 0.1132,
      "loss_align": 0.40450966358184814,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0031890999525785446,
      "neg_sim_mean": 0.4937420189380646,
      "pos_sim_mean": 0.5954903364181519,
      "step": 566
    },
    {
      "epoch": 0.1134,
      "grad_norm": 0.4928176701068878,
      "learning_rate": 9.994693285050857e-05,
      "loss": 0.4077,
      "step": 567
    },
    {
      "epoch": 0.1134,
      "loss_align": 0.46821773052215576,
      "loss_contrastive": 0.06051450967788696,
      "loss_gen": 0.00345805287361145,
      "neg_sim_mean": 0.4922967851161957,
      "pos_sim_mean": 0.5317822694778442,
      "step": 567
    },
    {
      "epoch": 0.1136,
      "grad_norm": 0.5637243986129761,
      "learning_rate": 9.994531286795308e-05,
      "loss": 0.4789,
      "step": 568
    },
    {
      "epoch": 0.1136,
      "loss_align": 0.44876736402511597,
      "loss_contrastive": 0.1020989716053009,
      "loss_gen": 0.4178033173084259,
      "neg_sim_mean": 0.5533316135406494,
      "pos_sim_mean": 0.551232635974884,
      "step": 568
    },
    {
      "epoch": 0.1138,
      "grad_norm": 3.990514039993286,
      "learning_rate": 9.994366854265886e-05,
      "loss": 0.8788,
      "step": 569
    },
    {
      "epoch": 0.1138,
      "loss_align": 0.38842666149139404,
      "loss_contrastive": 0.049339890480041504,
      "loss_gen": 0.2518976628780365,
      "neg_sim_mean": 0.5609132051467896,
      "pos_sim_mean": 0.611573338508606,
      "step": 569
    },
    {
      "epoch": 0.114,
      "grad_norm": 2.464989423751831,
      "learning_rate": 9.99419998754273e-05,
      "loss": 0.6462,
      "step": 570
    },
    {
      "epoch": 0.114,
      "loss_align": 0.48297256231307983,
      "loss_contrastive": 0.04769745469093323,
      "loss_gen": 0.19450142979621887,
      "neg_sim_mean": 0.46472489833831787,
      "pos_sim_mean": 0.5170274376869202,
      "step": 570
    },
    {
      "epoch": 0.1142,
      "grad_norm": 2.5457005500793457,
      "learning_rate": 9.99403068670717e-05,
      "loss": 0.6832,
      "step": 571
    },
    {
      "epoch": 0.1142,
      "loss_align": 0.42469263076782227,
      "loss_contrastive": 0.023895829916000366,
      "loss_gen": 0.3043645918369293,
      "neg_sim_mean": 0.4992032051086426,
      "pos_sim_mean": 0.5753073692321777,
      "step": 571
    },
    {
      "epoch": 0.1144,
      "grad_norm": 3.201894521713257,
      "learning_rate": 9.993858951841724e-05,
      "loss": 0.7319,
      "step": 572
    },
    {
      "epoch": 0.1144,
      "loss_align": 0.5414120554924011,
      "loss_contrastive": 0.13686731457710266,
      "loss_gen": 0.24077308177947998,
      "neg_sim_mean": 0.495455265045166,
      "pos_sim_mean": 0.4585879445075989,
      "step": 572
    },
    {
      "epoch": 0.1146,
      "grad_norm": 3.106646776199341,
      "learning_rate": 9.993684783030088e-05,
      "loss": 0.7986,
      "step": 573
    },
    {
      "epoch": 0.1146,
      "loss_align": 0.3638627529144287,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2242559939622879,
      "neg_sim_mean": 0.5289855003356934,
      "pos_sim_mean": 0.6361372470855713,
      "step": 573
    },
    {
      "epoch": 0.1148,
      "grad_norm": 3.2109596729278564,
      "learning_rate": 9.993508180357154e-05,
      "loss": 0.5881,
      "step": 574
    },
    {
      "epoch": 0.1148,
      "loss_align": 0.35389482975006104,
      "loss_contrastive": 0.17258191108703613,
      "loss_gen": 0.2443963587284088,
      "neg_sim_mean": 0.7186870574951172,
      "pos_sim_mean": 0.646105170249939,
      "step": 574
    },
    {
      "epoch": 0.115,
      "grad_norm": 2.616030693054199,
      "learning_rate": 9.993329143908994e-05,
      "loss": 0.619,
      "step": 575
    },
    {
      "epoch": 0.115,
      "loss_align": 0.3865119218826294,
      "loss_contrastive": 0.062493860721588135,
      "loss_gen": 0.44901785254478455,
      "neg_sim_mean": 0.5759819149971008,
      "pos_sim_mean": 0.6134880781173706,
      "step": 575
    },
    {
      "epoch": 0.1152,
      "grad_norm": 3.5999913215637207,
      "learning_rate": 9.99314767377287e-05,
      "loss": 0.843,
      "step": 576
    },
    {
      "epoch": 0.1152,
      "loss_align": 0.317419171333313,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4241628646850586,
      "neg_sim_mean": 0.5496351718902588,
      "pos_sim_mean": 0.682580828666687,
      "step": 576
    },
    {
      "epoch": 0.1154,
      "grad_norm": 3.263704776763916,
      "learning_rate": 9.992963770037227e-05,
      "loss": 0.7416,
      "step": 577
    },
    {
      "epoch": 0.1154,
      "loss_align": 0.3023151159286499,
      "loss_contrastive": 0.06276166439056396,
      "loss_gen": 0.2972114086151123,
      "neg_sim_mean": 0.6604465246200562,
      "pos_sim_mean": 0.6976848840713501,
      "step": 577
    },
    {
      "epoch": 0.1156,
      "grad_norm": 3.7628371715545654,
      "learning_rate": 9.992777432791697e-05,
      "loss": 0.6071,
      "step": 578
    },
    {
      "epoch": 0.1156,
      "loss_align": 0.3792600631713867,
      "loss_contrastive": 0.12913578748703003,
      "loss_gen": 0.3453991413116455,
      "neg_sim_mean": 0.6498757004737854,
      "pos_sim_mean": 0.6207399368286133,
      "step": 578
    },
    {
      "epoch": 0.1158,
      "grad_norm": 2.5107851028442383,
      "learning_rate": 9.9925886621271e-05,
      "loss": 0.7402,
      "step": 579
    },
    {
      "epoch": 0.1158,
      "loss_align": 0.4531214237213135,
      "loss_contrastive": 0.14489862322807312,
      "loss_gen": 0.36563554406166077,
      "neg_sim_mean": 0.5917772054672241,
      "pos_sim_mean": 0.5468785762786865,
      "step": 579
    },
    {
      "epoch": 0.116,
      "grad_norm": 2.7850372791290283,
      "learning_rate": 9.992397458135438e-05,
      "loss": 0.8361,
      "step": 580
    },
    {
      "epoch": 0.116,
      "loss_align": 0.45180100202560425,
      "loss_contrastive": 0.14573052525520325,
      "loss_gen": 0.24534624814987183,
      "neg_sim_mean": 0.5939295291900635,
      "pos_sim_mean": 0.5481989979743958,
      "step": 580
    },
    {
      "epoch": 0.1162,
      "grad_norm": 2.695176362991333,
      "learning_rate": 9.992203820909906e-05,
      "loss": 0.7146,
      "step": 581
    },
    {
      "epoch": 0.1162,
      "loss_align": 0.5101398229598999,
      "loss_contrastive": 0.10674965381622314,
      "loss_gen": 0.34081658720970154,
      "neg_sim_mean": 0.49660980701446533,
      "pos_sim_mean": 0.4898601472377777,
      "step": 581
    },
    {
      "epoch": 0.1164,
      "grad_norm": 2.9453821182250977,
      "learning_rate": 9.992007750544876e-05,
      "loss": 0.8638,
      "step": 582
    },
    {
      "epoch": 0.1164,
      "loss_align": 0.6353825926780701,
      "loss_contrastive": 0.21926483511924744,
      "loss_gen": 0.03290059044957161,
      "neg_sim_mean": 0.48388224840164185,
      "pos_sim_mean": 0.36461740732192993,
      "step": 582
    },
    {
      "epoch": 0.1166,
      "grad_norm": 1.8507673740386963,
      "learning_rate": 9.991809247135913e-05,
      "loss": 0.6946,
      "step": 583
    },
    {
      "epoch": 0.1166,
      "loss_align": 0.5740150213241577,
      "loss_contrastive": 0.09673097729682922,
      "loss_gen": 0.13634264469146729,
      "neg_sim_mean": 0.422715961933136,
      "pos_sim_mean": 0.4259849786758423,
      "step": 583
    },
    {
      "epoch": 0.1168,
      "grad_norm": 2.1636366844177246,
      "learning_rate": 9.991608310779762e-05,
      "loss": 0.722,
      "step": 584
    },
    {
      "epoch": 0.1168,
      "loss_align": 0.427034854888916,
      "loss_contrastive": 0.022582322359085083,
      "loss_gen": 0.17714759707450867,
      "neg_sim_mean": 0.49554747343063354,
      "pos_sim_mean": 0.572965145111084,
      "step": 584
    },
    {
      "epoch": 0.117,
      "grad_norm": 2.4814882278442383,
      "learning_rate": 9.991404941574361e-05,
      "loss": 0.6069,
      "step": 585
    },
    {
      "epoch": 0.117,
      "loss_align": 0.43061351776123047,
      "loss_contrastive": 0.047666996717453,
      "loss_gen": 0.19700144231319427,
      "neg_sim_mean": 0.517053484916687,
      "pos_sim_mean": 0.5693864822387695,
      "step": 585
    },
    {
      "epoch": 0.1172,
      "grad_norm": 2.599050283432007,
      "learning_rate": 9.991199139618827e-05,
      "loss": 0.6333,
      "step": 586
    },
    {
      "epoch": 0.1172,
      "loss_align": 0.4164583683013916,
      "loss_contrastive": 0.03767392039299011,
      "loss_gen": 0.2524871826171875,
      "neg_sim_mean": 0.521215558052063,
      "pos_sim_mean": 0.5835416316986084,
      "step": 586
    },
    {
      "epoch": 0.1174,
      "grad_norm": 2.620216131210327,
      "learning_rate": 9.990990905013467e-05,
      "loss": 0.6735,
      "step": 587
    },
    {
      "epoch": 0.1174,
      "loss_align": 0.35335034132003784,
      "loss_contrastive": 0.12480872869491577,
      "loss_gen": 0.33394554257392883,
      "neg_sim_mean": 0.67145836353302,
      "pos_sim_mean": 0.6466496586799622,
      "step": 587
    },
    {
      "epoch": 0.1176,
      "grad_norm": 2.8795106410980225,
      "learning_rate": 9.990780237859769e-05,
      "loss": 0.7023,
      "step": 588
    },
    {
      "epoch": 0.1176,
      "loss_align": 0.3496718406677246,
      "loss_contrastive": 0.09820139408111572,
      "loss_gen": 0.19857442378997803,
      "neg_sim_mean": 0.6485295295715332,
      "pos_sim_mean": 0.6503281593322754,
      "step": 588
    },
    {
      "epoch": 0.1178,
      "grad_norm": 2.369922637939453,
      "learning_rate": 9.990567138260413e-05,
      "loss": 0.56,
      "step": 589
    },
    {
      "epoch": 0.1178,
      "loss_align": 0.5756129026412964,
      "loss_contrastive": 0.2126084268093109,
      "loss_gen": 0.5919660329818726,
      "neg_sim_mean": 0.536995530128479,
      "pos_sim_mean": 0.4243870973587036,
      "step": 589
    },
    {
      "epoch": 0.118,
      "grad_norm": 3.7364869117736816,
      "learning_rate": 9.990351606319261e-05,
      "loss": 1.1931,
      "step": 590
    },
    {
      "epoch": 0.118,
      "loss_align": 0.33744144439697266,
      "loss_contrastive": 0.0729442834854126,
      "loss_gen": 0.39674094319343567,
      "neg_sim_mean": 0.635502815246582,
      "pos_sim_mean": 0.6625585556030273,
      "step": 590
    },
    {
      "epoch": 0.1182,
      "grad_norm": 2.8960201740264893,
      "learning_rate": 9.990133642141359e-05,
      "loss": 0.7429,
      "step": 591
    },
    {
      "epoch": 0.1182,
      "loss_align": 0.6182481050491333,
      "loss_contrastive": 0.21870732307434082,
      "loss_gen": 0.40291380882263184,
      "neg_sim_mean": 0.5004591941833496,
      "pos_sim_mean": 0.3817518651485443,
      "step": 591
    },
    {
      "epoch": 0.1184,
      "grad_norm": 3.206916093826294,
      "learning_rate": 9.989913245832941e-05,
      "loss": 1.0474,
      "step": 592
    },
    {
      "epoch": 0.1184,
      "loss_align": 0.6317662596702576,
      "loss_contrastive": 0.17779746651649475,
      "loss_gen": 0.2567140758037567,
      "neg_sim_mean": 0.44603121280670166,
      "pos_sim_mean": 0.36823374032974243,
      "step": 592
    },
    {
      "epoch": 0.1186,
      "grad_norm": 2.5757386684417725,
      "learning_rate": 9.989690417501423e-05,
      "loss": 0.9098,
      "step": 593
    },
    {
      "epoch": 0.1186,
      "loss_align": 0.3678758144378662,
      "loss_contrastive": 0.008341014385223389,
      "loss_gen": 0.08543526381254196,
      "neg_sim_mean": 0.5404651761054993,
      "pos_sim_mean": 0.6321241855621338,
      "step": 593
    },
    {
      "epoch": 0.1188,
      "grad_norm": 1.7249510288238525,
      "learning_rate": 9.989465157255412e-05,
      "loss": 0.4543,
      "step": 594
    },
    {
      "epoch": 0.1188,
      "loss_align": 0.5131674408912659,
      "loss_contrastive": 0.11487582325935364,
      "loss_gen": 0.019541533663868904,
      "neg_sim_mean": 0.5017083883285522,
      "pos_sim_mean": 0.48683255910873413,
      "step": 594
    },
    {
      "epoch": 0.119,
      "grad_norm": 1.3134962320327759,
      "learning_rate": 9.989237465204697e-05,
      "loss": 0.5465,
      "step": 595
    },
    {
      "epoch": 0.119,
      "loss_align": 0.3829871416091919,
      "loss_contrastive": 0.11817717552185059,
      "loss_gen": 0.29755571484565735,
      "neg_sim_mean": 0.6351900100708008,
      "pos_sim_mean": 0.6170128583908081,
      "step": 595
    },
    {
      "epoch": 0.1192,
      "grad_norm": 2.8205604553222656,
      "learning_rate": 9.98900734146025e-05,
      "loss": 0.6947,
      "step": 596
    },
    {
      "epoch": 0.1192,
      "loss_align": 0.40916377305984497,
      "loss_contrastive": 0.08384951949119568,
      "loss_gen": 0.22556596994400024,
      "neg_sim_mean": 0.5746857523918152,
      "pos_sim_mean": 0.590836226940155,
      "step": 596
    },
    {
      "epoch": 0.1194,
      "grad_norm": 2.5816900730133057,
      "learning_rate": 9.988774786134234e-05,
      "loss": 0.6448,
      "step": 597
    },
    {
      "epoch": 0.1194,
      "loss_align": 0.4883003234863281,
      "loss_contrastive": 0.07848384976387024,
      "loss_gen": 0.3707527220249176,
      "neg_sim_mean": 0.4901835322380066,
      "pos_sim_mean": 0.5116996765136719,
      "step": 597
    },
    {
      "epoch": 0.1196,
      "grad_norm": 2.9190709590911865,
      "learning_rate": 9.988539799339988e-05,
      "loss": 0.8685,
      "step": 598
    },
    {
      "epoch": 0.1196,
      "loss_align": 0.4161292314529419,
      "loss_contrastive": 0.03446361422538757,
      "loss_gen": 0.24348974227905273,
      "neg_sim_mean": 0.5183343887329102,
      "pos_sim_mean": 0.5838707685470581,
      "step": 598
    },
    {
      "epoch": 0.1198,
      "grad_norm": 2.7049262523651123,
      "learning_rate": 9.98830238119205e-05,
      "loss": 0.6638,
      "step": 599
    },
    {
      "epoch": 0.1198,
      "loss_align": 0.5179650783538818,
      "loss_contrastive": 0.11712220311164856,
      "loss_gen": 0.28989219665527344,
      "neg_sim_mean": 0.4991571605205536,
      "pos_sim_mean": 0.48203495144844055,
      "step": 599
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.832916498184204,
      "learning_rate": 9.988062531806126e-05,
      "loss": 0.8219,
      "step": 600
    },
    {
      "epoch": 0.12,
      "loss_align": 0.49974721670150757,
      "loss_contrastive": 0.061229825019836426,
      "loss_gen": 0.5145480036735535,
      "neg_sim_mean": 0.46148261427879333,
      "pos_sim_mean": 0.5002527832984924,
      "step": 600
    },
    {
      "epoch": 0.1202,
      "grad_norm": 4.001679420471191,
      "learning_rate": 9.987820251299122e-05,
      "loss": 1.0216,
      "step": 601
    },
    {
      "epoch": 0.1202,
      "loss_align": 0.4821522831916809,
      "loss_contrastive": 0.09116974472999573,
      "loss_gen": 0.0034530875273048878,
      "neg_sim_mean": 0.5090174674987793,
      "pos_sim_mean": 0.5178477168083191,
      "step": 601
    },
    {
      "epoch": 0.1204,
      "grad_norm": 0.5520172715187073,
      "learning_rate": 9.987575539789119e-05,
      "loss": 0.4965,
      "step": 602
    },
    {
      "epoch": 0.1204,
      "loss_align": 0.5227758884429932,
      "loss_contrastive": 0.2642989456653595,
      "loss_gen": 0.1639774590730667,
      "neg_sim_mean": 0.6415230631828308,
      "pos_sim_mean": 0.47722411155700684,
      "step": 602
    },
    {
      "epoch": 0.1206,
      "grad_norm": 2.045729875564575,
      "learning_rate": 9.987328397395387e-05,
      "loss": 0.7185,
      "step": 603
    },
    {
      "epoch": 0.1206,
      "loss_align": 0.5455121994018555,
      "loss_contrastive": 0.13425713777542114,
      "loss_gen": 0.23653972148895264,
      "neg_sim_mean": 0.48874494433403015,
      "pos_sim_mean": 0.45448780059814453,
      "step": 603
    },
    {
      "epoch": 0.1208,
      "grad_norm": 2.714970588684082,
      "learning_rate": 9.987078824238383e-05,
      "loss": 0.7982,
      "step": 604
    },
    {
      "epoch": 0.1208,
      "loss_align": 0.4205460548400879,
      "loss_contrastive": 0.050825685262680054,
      "loss_gen": 0.5020553469657898,
      "neg_sim_mean": 0.5302796363830566,
      "pos_sim_mean": 0.5794539451599121,
      "step": 604
    },
    {
      "epoch": 0.121,
      "grad_norm": 4.01848840713501,
      "learning_rate": 9.986826820439743e-05,
      "loss": 0.9287,
      "step": 605
    },
    {
      "epoch": 0.121,
      "loss_align": 0.5337874889373779,
      "loss_contrastive": 0.10799264907836914,
      "loss_gen": 0.434208482503891,
      "neg_sim_mean": 0.4742051661014557,
      "pos_sim_mean": 0.46621251106262207,
      "step": 605
    },
    {
      "epoch": 0.1212,
      "grad_norm": 2.804670810699463,
      "learning_rate": 9.986572386122291e-05,
      "loss": 0.981,
      "step": 606
    },
    {
      "epoch": 0.1212,
      "loss_align": 0.44201135635375977,
      "loss_contrastive": 0.17685887217521667,
      "loss_gen": 0.530718982219696,
      "neg_sim_mean": 0.6348475217819214,
      "pos_sim_mean": 0.5579886436462402,
      "step": 606
    },
    {
      "epoch": 0.1214,
      "grad_norm": 3.112023115158081,
      "learning_rate": 9.986315521410035e-05,
      "loss": 0.994,
      "step": 607
    },
    {
      "epoch": 0.1214,
      "loss_align": 0.34437090158462524,
      "loss_contrastive": 0.031733810901641846,
      "loss_gen": 0.5530574321746826,
      "neg_sim_mean": 0.5873628854751587,
      "pos_sim_mean": 0.6556290984153748,
      "step": 607
    },
    {
      "epoch": 0.1216,
      "grad_norm": 3.4744718074798584,
      "learning_rate": 9.986056226428169e-05,
      "loss": 0.9012,
      "step": 608
    },
    {
      "epoch": 0.1216,
      "loss_align": 0.4644779562950134,
      "loss_contrastive": 0.1643942892551422,
      "loss_gen": 0.49761778116226196,
      "neg_sim_mean": 0.5999163389205933,
      "pos_sim_mean": 0.5355220437049866,
      "step": 608
    },
    {
      "epoch": 0.1218,
      "grad_norm": 3.1859657764434814,
      "learning_rate": 9.98579450130307e-05,
      "loss": 0.9818,
      "step": 609
    },
    {
      "epoch": 0.1218,
      "loss_align": 0.48091065883636475,
      "loss_contrastive": 0.10315218567848206,
      "loss_gen": 0.24048474431037903,
      "neg_sim_mean": 0.5222415328025818,
      "pos_sim_mean": 0.5190893411636353,
      "step": 609
    },
    {
      "epoch": 0.122,
      "grad_norm": 3.014065742492676,
      "learning_rate": 9.985530346162299e-05,
      "loss": 0.7338,
      "step": 610
    },
    {
      "epoch": 0.122,
      "loss_align": 0.3838212490081787,
      "loss_contrastive": 0.10227930545806885,
      "loss_gen": 0.4565899670124054,
      "neg_sim_mean": 0.6184580326080322,
      "pos_sim_mean": 0.6161787509918213,
      "step": 610
    },
    {
      "epoch": 0.1222,
      "grad_norm": 3.804196357727051,
      "learning_rate": 9.985263761134602e-05,
      "loss": 0.8527,
      "step": 611
    },
    {
      "epoch": 0.1222,
      "loss_align": 0.45828115940093994,
      "loss_contrastive": 0.08374831080436707,
      "loss_gen": 0.008712650276720524,
      "neg_sim_mean": 0.5254671573638916,
      "pos_sim_mean": 0.5417188405990601,
      "step": 611
    },
    {
      "epoch": 0.1224,
      "grad_norm": 0.843600869178772,
      "learning_rate": 9.98499474634991e-05,
      "loss": 0.477,
      "step": 612
    },
    {
      "epoch": 0.1224,
      "loss_align": 0.47944170236587524,
      "loss_contrastive": 0.16975608468055725,
      "loss_gen": 0.27314847707748413,
      "neg_sim_mean": 0.5903143882751465,
      "pos_sim_mean": 0.5205582976341248,
      "step": 612
    },
    {
      "epoch": 0.1226,
      "grad_norm": 2.8927550315856934,
      "learning_rate": 9.984723301939337e-05,
      "loss": 0.773,
      "step": 613
    },
    {
      "epoch": 0.1226,
      "loss_align": 0.3782106041908264,
      "loss_contrastive": 0.10155618190765381,
      "loss_gen": 0.3989322781562805,
      "neg_sim_mean": 0.6233455538749695,
      "pos_sim_mean": 0.6217893958091736,
      "step": 613
    },
    {
      "epoch": 0.1228,
      "grad_norm": 3.395869493484497,
      "learning_rate": 9.98444942803518e-05,
      "loss": 0.7893,
      "step": 614
    },
    {
      "epoch": 0.1228,
      "loss_align": 0.3891267776489258,
      "loss_contrastive": 0.12352323532104492,
      "loss_gen": 0.3973917067050934,
      "neg_sim_mean": 0.6343964338302612,
      "pos_sim_mean": 0.6108732223510742,
      "step": 614
    },
    {
      "epoch": 0.123,
      "grad_norm": 2.641756534576416,
      "learning_rate": 9.984173124770923e-05,
      "loss": 0.8013,
      "step": 615
    },
    {
      "epoch": 0.123,
      "loss_align": 0.4932379722595215,
      "loss_contrastive": 0.11891612410545349,
      "loss_gen": 0.005908751394599676,
      "neg_sim_mean": 0.5256781578063965,
      "pos_sim_mean": 0.5067620277404785,
      "step": 615
    },
    {
      "epoch": 0.1232,
      "grad_norm": 0.725159227848053,
      "learning_rate": 9.983894392281237e-05,
      "loss": 0.5134,
      "step": 616
    },
    {
      "epoch": 0.1232,
      "loss_align": 0.47067177295684814,
      "loss_contrastive": 0.11715158820152283,
      "loss_gen": 0.22832204401493073,
      "neg_sim_mean": 0.5464798212051392,
      "pos_sim_mean": 0.5293282270431519,
      "step": 616
    },
    {
      "epoch": 0.1234,
      "grad_norm": 2.3342761993408203,
      "learning_rate": 9.983613230701967e-05,
      "loss": 0.7131,
      "step": 617
    },
    {
      "epoch": 0.1234,
      "loss_align": 0.5356869101524353,
      "loss_contrastive": 0.03664162755012512,
      "loss_gen": 0.3012901246547699,
      "neg_sim_mean": 0.4009547233581543,
      "pos_sim_mean": 0.4643130898475647,
      "step": 617
    },
    {
      "epoch": 0.1236,
      "grad_norm": 2.42525577545166,
      "learning_rate": 9.983329640170149e-05,
      "loss": 0.8414,
      "step": 618
    },
    {
      "epoch": 0.1236,
      "loss_align": 0.48973459005355835,
      "loss_contrastive": 0.11158332228660583,
      "loss_gen": 0.4099193811416626,
      "neg_sim_mean": 0.521848738193512,
      "pos_sim_mean": 0.5102654099464417,
      "step": 618
    },
    {
      "epoch": 0.1238,
      "grad_norm": 3.069512128829956,
      "learning_rate": 9.983043620824005e-05,
      "loss": 0.913,
      "step": 619
    },
    {
      "epoch": 0.1238,
      "loss_align": 0.5287526249885559,
      "loss_contrastive": 0.1816929280757904,
      "loss_gen": 0.2327118217945099,
      "neg_sim_mean": 0.552940309047699,
      "pos_sim_mean": 0.4712473750114441,
      "step": 619
    },
    {
      "epoch": 0.124,
      "grad_norm": 2.465766429901123,
      "learning_rate": 9.982755172802933e-05,
      "loss": 0.7833,
      "step": 620
    },
    {
      "epoch": 0.124,
      "loss_align": 0.5737291574478149,
      "loss_contrastive": 0.19139662384986877,
      "loss_gen": 0.03475527465343475,
      "neg_sim_mean": 0.5176674723625183,
      "pos_sim_mean": 0.42627084255218506,
      "step": 620
    },
    {
      "epoch": 0.1242,
      "grad_norm": 1.7092374563217163,
      "learning_rate": 9.982464296247522e-05,
      "loss": 0.6315,
      "step": 621
    },
    {
      "epoch": 0.1242,
      "loss_align": 0.341070294380188,
      "loss_contrastive": 0.09257102012634277,
      "loss_gen": 0.2625069320201874,
      "neg_sim_mean": 0.6515007019042969,
      "pos_sim_mean": 0.658929705619812,
      "step": 621
    },
    {
      "epoch": 0.1244,
      "grad_norm": 2.4740452766418457,
      "learning_rate": 9.982170991299539e-05,
      "loss": 0.6147,
      "step": 622
    },
    {
      "epoch": 0.1244,
      "loss_align": 0.47737324237823486,
      "loss_contrastive": 0.14087948203086853,
      "loss_gen": 0.4068867862224579,
      "neg_sim_mean": 0.5635062456130981,
      "pos_sim_mean": 0.5226267576217651,
      "step": 622
    },
    {
      "epoch": 0.1246,
      "grad_norm": 2.76275897026062,
      "learning_rate": 9.981875258101942e-05,
      "loss": 0.9012,
      "step": 623
    },
    {
      "epoch": 0.1246,
      "loss_align": 0.5514424443244934,
      "loss_contrastive": 0.14643976092338562,
      "loss_gen": 0.2647116184234619,
      "neg_sim_mean": 0.4949973225593567,
      "pos_sim_mean": 0.4485575556755066,
      "step": 623
    },
    {
      "epoch": 0.1248,
      "grad_norm": 2.6022865772247314,
      "learning_rate": 9.981577096798863e-05,
      "loss": 0.8337,
      "step": 624
    },
    {
      "epoch": 0.1248,
      "loss_align": 0.439399778842926,
      "loss_contrastive": 0.115488201379776,
      "loss_gen": 0.17854219675064087,
      "neg_sim_mean": 0.5760884284973145,
      "pos_sim_mean": 0.560600221157074,
      "step": 624
    },
    {
      "epoch": 0.125,
      "grad_norm": 2.1827542781829834,
      "learning_rate": 9.981276507535624e-05,
      "loss": 0.6318,
      "step": 625
    },
    {
      "epoch": 0.125,
      "loss_align": 0.4499041438102722,
      "loss_contrastive": 0.11219367384910583,
      "loss_gen": 0.20486760139465332,
      "neg_sim_mean": 0.5622895359992981,
      "pos_sim_mean": 0.5500958561897278,
      "step": 625
    },
    {
      "epoch": 0.1252,
      "grad_norm": 2.6703808307647705,
      "learning_rate": 9.980973490458728e-05,
      "loss": 0.6682,
      "step": 626
    },
    {
      "epoch": 0.1252,
      "loss_align": 0.31220686435699463,
      "loss_contrastive": 0.01869732141494751,
      "loss_gen": 0.19906507432460785,
      "neg_sim_mean": 0.606490433216095,
      "pos_sim_mean": 0.6877931356430054,
      "step": 626
    },
    {
      "epoch": 0.1254,
      "grad_norm": 1.9722697734832764,
      "learning_rate": 9.980668045715864e-05,
      "loss": 0.5135,
      "step": 627
    },
    {
      "epoch": 0.1254,
      "loss_align": 0.45462483167648315,
      "loss_contrastive": 0.1290539801120758,
      "loss_gen": 0.3903633952140808,
      "neg_sim_mean": 0.5744291543960571,
      "pos_sim_mean": 0.5453751683235168,
      "step": 627
    },
    {
      "epoch": 0.1256,
      "grad_norm": 3.0606024265289307,
      "learning_rate": 9.980360173455898e-05,
      "loss": 0.8605,
      "step": 628
    },
    {
      "epoch": 0.1256,
      "loss_align": 0.45687198638916016,
      "loss_contrastive": 0.012387841939926147,
      "loss_gen": 0.25300589203834534,
      "neg_sim_mean": 0.45551586151123047,
      "pos_sim_mean": 0.5431280136108398,
      "step": 628
    },
    {
      "epoch": 0.1258,
      "grad_norm": 2.297520875930786,
      "learning_rate": 9.980049873828887e-05,
      "loss": 0.7114,
      "step": 629
    },
    {
      "epoch": 0.1258,
      "loss_align": 0.5042926669120789,
      "loss_contrastive": 0.11505106091499329,
      "loss_gen": 0.28062695264816284,
      "neg_sim_mean": 0.5107583999633789,
      "pos_sim_mean": 0.49570733308792114,
      "step": 629
    },
    {
      "epoch": 0.126,
      "grad_norm": 2.756640672683716,
      "learning_rate": 9.979737146986064e-05,
      "loss": 0.7987,
      "step": 630
    },
    {
      "epoch": 0.126,
      "loss_align": 0.41748881340026855,
      "loss_contrastive": 0.049887508153915405,
      "loss_gen": 0.3544781804084778,
      "neg_sim_mean": 0.5323987007141113,
      "pos_sim_mean": 0.5825111865997314,
      "step": 630
    },
    {
      "epoch": 0.1262,
      "grad_norm": 3.392361879348755,
      "learning_rate": 9.979421993079852e-05,
      "loss": 0.778,
      "step": 631
    },
    {
      "epoch": 0.1262,
      "loss_align": 0.41441309452056885,
      "loss_contrastive": 0.0296458899974823,
      "loss_gen": 0.015546969138085842,
      "neg_sim_mean": 0.5152328014373779,
      "pos_sim_mean": 0.5855869054794312,
      "step": 631
    },
    {
      "epoch": 0.1264,
      "grad_norm": 1.1845502853393555,
      "learning_rate": 9.97910441226385e-05,
      "loss": 0.4335,
      "step": 632
    },
    {
      "epoch": 0.1264,
      "loss_align": 0.45448440313339233,
      "loss_contrastive": 0.11613216996192932,
      "loss_gen": 0.220089852809906,
      "neg_sim_mean": 0.5616477727890015,
      "pos_sim_mean": 0.5455155968666077,
      "step": 632
    },
    {
      "epoch": 0.1266,
      "grad_norm": 2.493333578109741,
      "learning_rate": 9.978784404692847e-05,
      "loss": 0.6885,
      "step": 633
    },
    {
      "epoch": 0.1266,
      "loss_align": 0.4491947889328003,
      "loss_contrastive": 0.0643969476222992,
      "loss_gen": 0.0010434705764055252,
      "neg_sim_mean": 0.5152021646499634,
      "pos_sim_mean": 0.5508052110671997,
      "step": 633
    },
    {
      "epoch": 0.1268,
      "grad_norm": 0.4091656804084778,
      "learning_rate": 9.978461970522806e-05,
      "loss": 0.458,
      "step": 634
    },
    {
      "epoch": 0.1268,
      "loss_align": 0.4666574001312256,
      "loss_contrastive": 0.05577939748764038,
      "loss_gen": 0.046691376715898514,
      "neg_sim_mean": 0.4891220033168793,
      "pos_sim_mean": 0.5333425998687744,
      "step": 634
    },
    {
      "epoch": 0.127,
      "grad_norm": 1.355676293373108,
      "learning_rate": 9.978137109910878e-05,
      "loss": 0.52,
      "step": 635
    },
    {
      "epoch": 0.127,
      "loss_align": 0.4564405679702759,
      "loss_contrastive": 0.09864768385887146,
      "loss_gen": 0.38492631912231445,
      "neg_sim_mean": 0.5422071218490601,
      "pos_sim_mean": 0.5435594320297241,
      "step": 635
    },
    {
      "epoch": 0.1272,
      "grad_norm": 2.8207526206970215,
      "learning_rate": 9.977809823015401e-05,
      "loss": 0.8532,
      "step": 636
    },
    {
      "epoch": 0.1272,
      "loss_align": 0.35607731342315674,
      "loss_contrastive": 0.1050952672958374,
      "loss_gen": 0.3598295748233795,
      "neg_sim_mean": 0.6490179300308228,
      "pos_sim_mean": 0.6439226865768433,
      "step": 636
    },
    {
      "epoch": 0.1274,
      "grad_norm": 3.4598560333251953,
      "learning_rate": 9.977480109995885e-05,
      "loss": 0.7285,
      "step": 637
    },
    {
      "epoch": 0.1274,
      "loss_align": 0.3730594515800476,
      "loss_contrastive": 0.028466403484344482,
      "loss_gen": 0.17375624179840088,
      "neg_sim_mean": 0.555406928062439,
      "pos_sim_mean": 0.6269405484199524,
      "step": 637
    },
    {
      "epoch": 0.1276,
      "grad_norm": 2.502551555633545,
      "learning_rate": 9.977147971013032e-05,
      "loss": 0.5502,
      "step": 638
    },
    {
      "epoch": 0.1276,
      "loss_align": 0.41736435890197754,
      "loss_contrastive": 0.10160800814628601,
      "loss_gen": 0.35120779275894165,
      "neg_sim_mean": 0.584243655204773,
      "pos_sim_mean": 0.5826356410980225,
      "step": 638
    },
    {
      "epoch": 0.1278,
      "grad_norm": 3.2628989219665527,
      "learning_rate": 9.97681340622872e-05,
      "loss": 0.7808,
      "step": 639
    },
    {
      "epoch": 0.1278,
      "loss_align": 0.3668009042739868,
      "loss_contrastive": 0.08004623651504517,
      "loss_gen": 0.26964882016181946,
      "neg_sim_mean": 0.6132453083992004,
      "pos_sim_mean": 0.6331990957260132,
      "step": 639
    },
    {
      "epoch": 0.128,
      "grad_norm": 2.643354892730713,
      "learning_rate": 9.976476415806013e-05,
      "loss": 0.6461,
      "step": 640
    },
    {
      "epoch": 0.128,
      "loss_align": 0.47347378730773926,
      "loss_contrastive": 0.0541422963142395,
      "loss_gen": 0.3826301097869873,
      "neg_sim_mean": 0.4806685149669647,
      "pos_sim_mean": 0.5265262126922607,
      "step": 640
    },
    {
      "epoch": 0.1282,
      "grad_norm": 3.4593253135681152,
      "learning_rate": 9.976136999909156e-05,
      "loss": 0.8626,
      "step": 641
    },
    {
      "epoch": 0.1282,
      "loss_align": 0.3931915760040283,
      "loss_contrastive": 0.11886155605316162,
      "loss_gen": 0.31854158639907837,
      "neg_sim_mean": 0.6256699562072754,
      "pos_sim_mean": 0.6068084239959717,
      "step": 641
    },
    {
      "epoch": 0.1284,
      "grad_norm": 2.4703211784362793,
      "learning_rate": 9.975795158703576e-05,
      "loss": 0.726,
      "step": 642
    },
    {
      "epoch": 0.1284,
      "loss_align": 0.43370944261550903,
      "loss_contrastive": 0.056544095277786255,
      "loss_gen": 0.01571236364543438,
      "neg_sim_mean": 0.5228346586227417,
      "pos_sim_mean": 0.566290557384491,
      "step": 642
    },
    {
      "epoch": 0.1286,
      "grad_norm": 1.1352488994598389,
      "learning_rate": 9.975450892355882e-05,
      "loss": 0.4562,
      "step": 643
    },
    {
      "epoch": 0.1286,
      "loss_align": 0.46920979022979736,
      "loss_contrastive": 0.08353695273399353,
      "loss_gen": 0.367667019367218,
      "neg_sim_mean": 0.5143271684646606,
      "pos_sim_mean": 0.5307902097702026,
      "step": 643
    },
    {
      "epoch": 0.1288,
      "grad_norm": 2.9066431522369385,
      "learning_rate": 9.975104201033866e-05,
      "loss": 0.8469,
      "step": 644
    },
    {
      "epoch": 0.1288,
      "loss_align": 0.43527770042419434,
      "loss_contrastive": 0.09244367480278015,
      "loss_gen": 0.2883821725845337,
      "neg_sim_mean": 0.5571659803390503,
      "pos_sim_mean": 0.5647222995758057,
      "step": 644
    },
    {
      "epoch": 0.129,
      "grad_norm": 2.5104451179504395,
      "learning_rate": 9.974755084906502e-05,
      "loss": 0.7348,
      "step": 645
    },
    {
      "epoch": 0.129,
      "loss_align": 0.4874076843261719,
      "loss_contrastive": 0.17182496190071106,
      "loss_gen": 0.2079632580280304,
      "neg_sim_mean": 0.5844172835350037,
      "pos_sim_mean": 0.5125923156738281,
      "step": 645
    },
    {
      "epoch": 0.1292,
      "grad_norm": 1.9525834321975708,
      "learning_rate": 9.974403544143941e-05,
      "loss": 0.716,
      "step": 646
    },
    {
      "epoch": 0.1292,
      "loss_align": 0.4950123429298401,
      "loss_contrastive": 0.1029137670993805,
      "loss_gen": 0.31720903515815735,
      "neg_sim_mean": 0.5079014301300049,
      "pos_sim_mean": 0.5049876570701599,
      "step": 646
    },
    {
      "epoch": 0.1294,
      "grad_norm": 2.9301509857177734,
      "learning_rate": 9.974049578917523e-05,
      "loss": 0.8246,
      "step": 647
    },
    {
      "epoch": 0.1294,
      "loss_align": 0.5189793705940247,
      "loss_contrastive": 0.015231996774673462,
      "loss_gen": 0.2434999942779541,
      "neg_sim_mean": 0.3962526321411133,
      "pos_sim_mean": 0.48102062940597534,
      "step": 647
    },
    {
      "epoch": 0.1296,
      "grad_norm": 2.665950298309326,
      "learning_rate": 9.973693189399766e-05,
      "loss": 0.7643,
      "step": 648
    },
    {
      "epoch": 0.1296,
      "loss_align": 0.4452759027481079,
      "loss_contrastive": 0.1214357316493988,
      "loss_gen": 0.31816062331199646,
      "neg_sim_mean": 0.5761598348617554,
      "pos_sim_mean": 0.5547240972518921,
      "step": 648
    },
    {
      "epoch": 0.1298,
      "grad_norm": 2.5087292194366455,
      "learning_rate": 9.973334375764372e-05,
      "loss": 0.778,
      "step": 649
    },
    {
      "epoch": 0.1298,
      "loss_align": 0.3509063720703125,
      "loss_contrastive": 0.04387855529785156,
      "loss_gen": 0.39939531683921814,
      "neg_sim_mean": 0.5929721593856812,
      "pos_sim_mean": 0.6490936279296875,
      "step": 649
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.758498191833496,
      "learning_rate": 9.972973138186217e-05,
      "loss": 0.7556,
      "step": 650
    },
    {
      "epoch": 0.13,
      "loss_align": 0.4307512640953064,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3415634334087372,
      "neg_sim_mean": 0.46576353907585144,
      "pos_sim_mean": 0.5692487359046936,
      "step": 650
    },
    {
      "epoch": 0.1302,
      "grad_norm": 2.272902011871338,
      "learning_rate": 9.972609476841367e-05,
      "loss": 0.7723,
      "step": 651
    },
    {
      "epoch": 0.1302,
      "loss_align": 0.3824067711830139,
      "loss_contrastive": 0.0013112425804138184,
      "loss_gen": 0.01367025077342987,
      "neg_sim_mean": 0.518904447555542,
      "pos_sim_mean": 0.6175932288169861,
      "step": 651
    },
    {
      "epoch": 0.1304,
      "grad_norm": 1.0996509790420532,
      "learning_rate": 9.972243391907067e-05,
      "loss": 0.3962,
      "step": 652
    },
    {
      "epoch": 0.1304,
      "loss_align": 0.5128887891769409,
      "loss_contrastive": 0.21758240461349487,
      "loss_gen": 0.22346378862857819,
      "neg_sim_mean": 0.6046936511993408,
      "pos_sim_mean": 0.48711124062538147,
      "step": 652
    },
    {
      "epoch": 0.1306,
      "grad_norm": 2.3901896476745605,
      "learning_rate": 9.97187488356174e-05,
      "loss": 0.7625,
      "step": 653
    },
    {
      "epoch": 0.1306,
      "loss_align": 0.4846712350845337,
      "loss_contrastive": 0.10320433974266052,
      "loss_gen": 0.3110545575618744,
      "neg_sim_mean": 0.5185331106185913,
      "pos_sim_mean": 0.5153287649154663,
      "step": 653
    },
    {
      "epoch": 0.1308,
      "grad_norm": 2.877790927886963,
      "learning_rate": 9.971503951984995e-05,
      "loss": 0.8081,
      "step": 654
    },
    {
      "epoch": 0.1308,
      "loss_align": 0.3886021375656128,
      "loss_contrastive": 0.06376063823699951,
      "loss_gen": 0.3146986961364746,
      "neg_sim_mean": 0.5751584768295288,
      "pos_sim_mean": 0.6113978624343872,
      "step": 654
    },
    {
      "epoch": 0.131,
      "grad_norm": 2.454122543334961,
      "learning_rate": 9.971130597357618e-05,
      "loss": 0.711,
      "step": 655
    },
    {
      "epoch": 0.131,
      "loss_align": 0.41102713346481323,
      "loss_contrastive": 0.014243513345718384,
      "loss_gen": 0.4079318940639496,
      "neg_sim_mean": 0.5032163858413696,
      "pos_sim_mean": 0.5889728665351868,
      "step": 655
    },
    {
      "epoch": 0.1312,
      "grad_norm": 3.6248137950897217,
      "learning_rate": 9.970754819861577e-05,
      "loss": 0.8207,
      "step": 656
    },
    {
      "epoch": 0.1312,
      "loss_align": 0.4220503568649292,
      "loss_contrastive": 0.11359784007072449,
      "loss_gen": 0.27095746994018555,
      "neg_sim_mean": 0.5915474891662598,
      "pos_sim_mean": 0.5779496431350708,
      "step": 656
    },
    {
      "epoch": 0.1314,
      "grad_norm": 2.4745254516601562,
      "learning_rate": 9.970376619680024e-05,
      "loss": 0.7066,
      "step": 657
    },
    {
      "epoch": 0.1314,
      "loss_align": 0.43664276599884033,
      "loss_contrastive": 0.07271149754524231,
      "loss_gen": 0.3818773627281189,
      "neg_sim_mean": 0.5360687375068665,
      "pos_sim_mean": 0.5633572340011597,
      "step": 657
    },
    {
      "epoch": 0.1316,
      "grad_norm": 3.0864031314849854,
      "learning_rate": 9.969995996997285e-05,
      "loss": 0.8272,
      "step": 658
    },
    {
      "epoch": 0.1316,
      "loss_align": 0.3458050489425659,
      "loss_contrastive": 0.09073680639266968,
      "loss_gen": 0.28791695833206177,
      "neg_sim_mean": 0.6449317336082458,
      "pos_sim_mean": 0.6541949510574341,
      "step": 658
    },
    {
      "epoch": 0.1318,
      "grad_norm": 2.3080570697784424,
      "learning_rate": 9.969612951998874e-05,
      "loss": 0.6446,
      "step": 659
    },
    {
      "epoch": 0.1318,
      "loss_align": 0.3709260821342468,
      "loss_contrastive": 0.10785108804702759,
      "loss_gen": 0.2565406262874603,
      "neg_sim_mean": 0.6369249820709229,
      "pos_sim_mean": 0.6290739178657532,
      "step": 659
    },
    {
      "epoch": 0.132,
      "grad_norm": 2.457520008087158,
      "learning_rate": 9.969227484871484e-05,
      "loss": 0.6404,
      "step": 660
    },
    {
      "epoch": 0.132,
      "loss_align": 0.4239267110824585,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2749828100204468,
      "neg_sim_mean": 0.4714524745941162,
      "pos_sim_mean": 0.5760732889175415,
      "step": 660
    },
    {
      "epoch": 0.1322,
      "grad_norm": 2.3392879962921143,
      "learning_rate": 9.968839595802982e-05,
      "loss": 0.6989,
      "step": 661
    },
    {
      "epoch": 0.1322,
      "loss_align": 0.4574732184410095,
      "loss_contrastive": 0.14524772763252258,
      "loss_gen": 0.3437926471233368,
      "neg_sim_mean": 0.5877745151519775,
      "pos_sim_mean": 0.5425267815589905,
      "step": 661
    },
    {
      "epoch": 0.1324,
      "grad_norm": 2.636033296585083,
      "learning_rate": 9.968449284982424e-05,
      "loss": 0.8187,
      "step": 662
    },
    {
      "epoch": 0.1324,
      "loss_align": 0.3840606212615967,
      "loss_contrastive": 0.10563695430755615,
      "loss_gen": 0.42052778601646423,
      "neg_sim_mean": 0.6215763092041016,
      "pos_sim_mean": 0.6159393787384033,
      "step": 662
    },
    {
      "epoch": 0.1326,
      "grad_norm": 3.284907817840576,
      "learning_rate": 9.968056552600043e-05,
      "loss": 0.8173,
      "step": 663
    },
    {
      "epoch": 0.1326,
      "loss_align": 0.46797382831573486,
      "loss_contrastive": 0.1425069272518158,
      "loss_gen": 0.3805599510669708,
      "neg_sim_mean": 0.5745331048965454,
      "pos_sim_mean": 0.5320261716842651,
      "step": 663
    },
    {
      "epoch": 0.1328,
      "grad_norm": 2.956063985824585,
      "learning_rate": 9.96766139884725e-05,
      "loss": 0.8656,
      "step": 664
    },
    {
      "epoch": 0.1328,
      "loss_align": 0.43896961212158203,
      "loss_contrastive": 0.0936954915523529,
      "loss_gen": 0.45466214418411255,
      "neg_sim_mean": 0.5547258853912354,
      "pos_sim_mean": 0.561030387878418,
      "step": 664
    },
    {
      "epoch": 0.133,
      "grad_norm": 3.4581685066223145,
      "learning_rate": 9.967263823916638e-05,
      "loss": 0.9049,
      "step": 665
    },
    {
      "epoch": 0.133,
      "loss_align": 0.47321224212646484,
      "loss_contrastive": 0.15897151827812195,
      "loss_gen": 0.6599470973014832,
      "neg_sim_mean": 0.5857592821121216,
      "pos_sim_mean": 0.5267877578735352,
      "step": 665
    },
    {
      "epoch": 0.1332,
      "grad_norm": 3.335344076156616,
      "learning_rate": 9.966863828001982e-05,
      "loss": 1.1522,
      "step": 666
    },
    {
      "epoch": 0.1332,
      "loss_align": 0.44522225856781006,
      "loss_contrastive": 0.06754431128501892,
      "loss_gen": 0.28176119923591614,
      "neg_sim_mean": 0.5223220586776733,
      "pos_sim_mean": 0.5547777414321899,
      "step": 666
    },
    {
      "epoch": 0.1334,
      "grad_norm": 2.237610340118408,
      "learning_rate": 9.966461411298236e-05,
      "loss": 0.7351,
      "step": 667
    },
    {
      "epoch": 0.1334,
      "loss_align": 0.42418158054351807,
      "loss_contrastive": 0.04666987061500549,
      "loss_gen": 0.2221786379814148,
      "neg_sim_mean": 0.5224882960319519,
      "pos_sim_mean": 0.5758184194564819,
      "step": 667
    },
    {
      "epoch": 0.1336,
      "grad_norm": 2.276703357696533,
      "learning_rate": 9.966056574001529e-05,
      "loss": 0.652,
      "step": 668
    },
    {
      "epoch": 0.1336,
      "loss_align": 0.41812968254089355,
      "loss_contrastive": 0.10030171275138855,
      "loss_gen": 0.29450300335884094,
      "neg_sim_mean": 0.5821720361709595,
      "pos_sim_mean": 0.5818703174591064,
      "step": 668
    },
    {
      "epoch": 0.1338,
      "grad_norm": 2.0652527809143066,
      "learning_rate": 9.965649316309178e-05,
      "loss": 0.7247,
      "step": 669
    },
    {
      "epoch": 0.1338,
      "loss_align": 0.5434994697570801,
      "loss_contrastive": 0.2838391661643982,
      "loss_gen": 0.32598385214805603,
      "neg_sim_mean": 0.640339732170105,
      "pos_sim_mean": 0.4565005600452423,
      "step": 669
    },
    {
      "epoch": 0.134,
      "grad_norm": 2.4615535736083984,
      "learning_rate": 9.965239638419672e-05,
      "loss": 0.9035,
      "step": 670
    },
    {
      "epoch": 0.134,
      "loss_align": 0.31239593029022217,
      "loss_contrastive": 0.05319976806640625,
      "loss_gen": 0.3357142210006714,
      "neg_sim_mean": 0.6408038139343262,
      "pos_sim_mean": 0.6876040697097778,
      "step": 670
    },
    {
      "epoch": 0.1342,
      "grad_norm": 2.237487554550171,
      "learning_rate": 9.964827540532685e-05,
      "loss": 0.6545,
      "step": 671
    },
    {
      "epoch": 0.1342,
      "loss_align": 0.36964648962020874,
      "loss_contrastive": 0.10284245014190674,
      "loss_gen": 0.22607816755771637,
      "neg_sim_mean": 0.6331959366798401,
      "pos_sim_mean": 0.6303535103797913,
      "step": 671
    },
    {
      "epoch": 0.1344,
      "grad_norm": 2.1537892818450928,
      "learning_rate": 9.964413022849068e-05,
      "loss": 0.6081,
      "step": 672
    },
    {
      "epoch": 0.1344,
      "loss_align": 0.5187164545059204,
      "loss_contrastive": 0.05987226963043213,
      "loss_gen": 0.25949394702911377,
      "neg_sim_mean": 0.4411557912826538,
      "pos_sim_mean": 0.4812835156917572,
      "step": 672
    },
    {
      "epoch": 0.1346,
      "grad_norm": 2.1216933727264404,
      "learning_rate": 9.963996085570854e-05,
      "loss": 0.7854,
      "step": 673
    },
    {
      "epoch": 0.1346,
      "loss_align": 0.3968659043312073,
      "loss_contrastive": 0.0940735936164856,
      "loss_gen": 0.30690035223960876,
      "neg_sim_mean": 0.5972076654434204,
      "pos_sim_mean": 0.6031340956687927,
      "step": 673
    },
    {
      "epoch": 0.1348,
      "grad_norm": 3.2947347164154053,
      "learning_rate": 9.96357672890125e-05,
      "loss": 0.7151,
      "step": 674
    },
    {
      "epoch": 0.1348,
      "loss_align": 0.500877857208252,
      "loss_contrastive": 0.16699323058128357,
      "loss_gen": 0.2790600061416626,
      "neg_sim_mean": 0.5661153793334961,
      "pos_sim_mean": 0.49912214279174805,
      "step": 674
    },
    {
      "epoch": 0.135,
      "grad_norm": 2.2711880207061768,
      "learning_rate": 9.963154953044645e-05,
      "loss": 0.8,
      "step": 675
    },
    {
      "epoch": 0.135,
      "loss_align": 0.3515005111694336,
      "loss_contrastive": 0.06745004653930664,
      "loss_gen": 0.19055530428886414,
      "neg_sim_mean": 0.6159495115280151,
      "pos_sim_mean": 0.6484994888305664,
      "step": 675
    },
    {
      "epoch": 0.1352,
      "grad_norm": 1.6157941818237305,
      "learning_rate": 9.962730758206611e-05,
      "loss": 0.5501,
      "step": 676
    },
    {
      "epoch": 0.1352,
      "loss_align": 0.39032912254333496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010426471941173077,
      "neg_sim_mean": 0.5096209645271301,
      "pos_sim_mean": 0.609670877456665,
      "step": 676
    },
    {
      "epoch": 0.1354,
      "grad_norm": 0.9279466271400452,
      "learning_rate": 9.962304144593893e-05,
      "loss": 0.4008,
      "step": 677
    },
    {
      "epoch": 0.1354,
      "loss_align": 0.4997749924659729,
      "loss_contrastive": 0.165770024061203,
      "loss_gen": 0.240028977394104,
      "neg_sim_mean": 0.5659950375556946,
      "pos_sim_mean": 0.5002250075340271,
      "step": 677
    },
    {
      "epoch": 0.1356,
      "grad_norm": 2.7873423099517822,
      "learning_rate": 9.961875112414416e-05,
      "loss": 0.7597,
      "step": 678
    },
    {
      "epoch": 0.1356,
      "loss_align": 0.45183539390563965,
      "loss_contrastive": 0.11293467879295349,
      "loss_gen": 0.41346409916877747,
      "neg_sim_mean": 0.5610992908477783,
      "pos_sim_mean": 0.5481646060943604,
      "step": 678
    },
    {
      "epoch": 0.1358,
      "grad_norm": 2.893907070159912,
      "learning_rate": 9.961443661877289e-05,
      "loss": 0.8789,
      "step": 679
    },
    {
      "epoch": 0.1358,
      "loss_align": 0.35681354999542236,
      "loss_contrastive": 0.12887835502624512,
      "loss_gen": 0.31898918747901917,
      "neg_sim_mean": 0.6720647811889648,
      "pos_sim_mean": 0.6431864500045776,
      "step": 679
    },
    {
      "epoch": 0.136,
      "grad_norm": 2.4175913333892822,
      "learning_rate": 9.961009793192792e-05,
      "loss": 0.6913,
      "step": 680
    },
    {
      "epoch": 0.136,
      "loss_align": 0.31748348474502563,
      "loss_contrastive": 0.017237961292266846,
      "loss_gen": 0.343657523393631,
      "neg_sim_mean": 0.5997544527053833,
      "pos_sim_mean": 0.6825165152549744,
      "step": 680
    },
    {
      "epoch": 0.1362,
      "grad_norm": 2.477168083190918,
      "learning_rate": 9.96057350657239e-05,
      "loss": 0.6632,
      "step": 681
    },
    {
      "epoch": 0.1362,
      "loss_align": 0.39929473400115967,
      "loss_contrastive": 0.10324752330780029,
      "loss_gen": 0.2090003788471222,
      "neg_sim_mean": 0.6039527654647827,
      "pos_sim_mean": 0.6007052659988403,
      "step": 681
    },
    {
      "epoch": 0.1364,
      "grad_norm": 2.0785181522369385,
      "learning_rate": 9.960134802228723e-05,
      "loss": 0.6207,
      "step": 682
    },
    {
      "epoch": 0.1364,
      "loss_align": 0.501621663570404,
      "loss_contrastive": 0.11153319478034973,
      "loss_gen": 0.227961465716362,
      "neg_sim_mean": 0.5099115371704102,
      "pos_sim_mean": 0.49837833642959595,
      "step": 682
    },
    {
      "epoch": 0.1366,
      "grad_norm": 2.1123433113098145,
      "learning_rate": 9.959693680375608e-05,
      "loss": 0.743,
      "step": 683
    },
    {
      "epoch": 0.1366,
      "loss_align": 0.39233994483947754,
      "loss_contrastive": 0.1361764669418335,
      "loss_gen": 0.296735554933548,
      "neg_sim_mean": 0.643836498260498,
      "pos_sim_mean": 0.6076600551605225,
      "step": 683
    },
    {
      "epoch": 0.1368,
      "grad_norm": 2.14851713180542,
      "learning_rate": 9.959250141228045e-05,
      "loss": 0.7054,
      "step": 684
    },
    {
      "epoch": 0.1368,
      "loss_align": 0.42448174953460693,
      "loss_contrastive": 0.09684213995933533,
      "loss_gen": 0.30011963844299316,
      "neg_sim_mean": 0.5723603963851929,
      "pos_sim_mean": 0.5755182504653931,
      "step": 684
    },
    {
      "epoch": 0.137,
      "grad_norm": 2.1397409439086914,
      "learning_rate": 9.958804185002209e-05,
      "loss": 0.7362,
      "step": 685
    },
    {
      "epoch": 0.137,
      "loss_align": 0.5019726157188416,
      "loss_contrastive": 0.13389745354652405,
      "loss_gen": 0.1614999622106552,
      "neg_sim_mean": 0.531924843788147,
      "pos_sim_mean": 0.49802738428115845,
      "step": 685
    },
    {
      "epoch": 0.1372,
      "grad_norm": 1.9195445775985718,
      "learning_rate": 9.958355811915451e-05,
      "loss": 0.6795,
      "step": 686
    },
    {
      "epoch": 0.1372,
      "loss_align": 0.4181784987449646,
      "loss_contrastive": 0.08404532074928284,
      "loss_gen": 0.26578888297080994,
      "neg_sim_mean": 0.5658668279647827,
      "pos_sim_mean": 0.5818215012550354,
      "step": 686
    },
    {
      "epoch": 0.1374,
      "grad_norm": 2.27827525138855,
      "learning_rate": 9.957905022186309e-05,
      "loss": 0.6941,
      "step": 687
    },
    {
      "epoch": 0.1374,
      "loss_align": 0.3166170120239258,
      "loss_contrastive": 0.009017109870910645,
      "loss_gen": 0.32517826557159424,
      "neg_sim_mean": 0.592400074005127,
      "pos_sim_mean": 0.6833829879760742,
      "step": 687
    },
    {
      "epoch": 0.1376,
      "grad_norm": 2.75704026222229,
      "learning_rate": 9.957451816034486e-05,
      "loss": 0.6429,
      "step": 688
    },
    {
      "epoch": 0.1376,
      "loss_align": 0.4121260643005371,
      "loss_contrastive": 0.14406761527061462,
      "loss_gen": 0.22016161680221558,
      "neg_sim_mean": 0.631941556930542,
      "pos_sim_mean": 0.5878739356994629,
      "step": 688
    },
    {
      "epoch": 0.1378,
      "grad_norm": 2.180234670639038,
      "learning_rate": 9.956996193680873e-05,
      "loss": 0.6496,
      "step": 689
    },
    {
      "epoch": 0.1378,
      "loss_align": 0.52382493019104,
      "loss_contrastive": 0.13456851243972778,
      "loss_gen": 0.3233279287815094,
      "neg_sim_mean": 0.5107436180114746,
      "pos_sim_mean": 0.47617509961128235,
      "step": 689
    },
    {
      "epoch": 0.138,
      "grad_norm": 2.3199055194854736,
      "learning_rate": 9.956538155347534e-05,
      "loss": 0.8633,
      "step": 690
    },
    {
      "epoch": 0.138,
      "loss_align": 0.43185579776763916,
      "loss_contrastive": 0.052414149045944214,
      "loss_gen": 0.05538291484117508,
      "neg_sim_mean": 0.5205583572387695,
      "pos_sim_mean": 0.5681442022323608,
      "step": 690
    },
    {
      "epoch": 0.1382,
      "grad_norm": 1.7194679975509644,
      "learning_rate": 9.956077701257709e-05,
      "loss": 0.4935,
      "step": 691
    },
    {
      "epoch": 0.1382,
      "loss_align": 0.5883260369300842,
      "loss_contrastive": 0.12330275774002075,
      "loss_gen": 0.18423904478549957,
      "neg_sim_mean": 0.434976726770401,
      "pos_sim_mean": 0.41167396306991577,
      "step": 691
    },
    {
      "epoch": 0.1384,
      "grad_norm": 1.779068946838379,
      "learning_rate": 9.95561483163582e-05,
      "loss": 0.7874,
      "step": 692
    },
    {
      "epoch": 0.1384,
      "loss_align": 0.436049222946167,
      "loss_contrastive": 0.14041879773139954,
      "loss_gen": 0.5418235063552856,
      "neg_sim_mean": 0.604369580745697,
      "pos_sim_mean": 0.563950777053833,
      "step": 692
    },
    {
      "epoch": 0.1386,
      "grad_norm": 3.432802677154541,
      "learning_rate": 9.955149546707465e-05,
      "loss": 0.9947,
      "step": 693
    },
    {
      "epoch": 0.1386,
      "loss_align": 0.4725421667098999,
      "loss_contrastive": 0.11274239420890808,
      "loss_gen": 0.20196859538555145,
      "neg_sim_mean": 0.5402002334594727,
      "pos_sim_mean": 0.5274578332901001,
      "step": 693
    },
    {
      "epoch": 0.1388,
      "grad_norm": 2.1908202171325684,
      "learning_rate": 9.954681846699414e-05,
      "loss": 0.688,
      "step": 694
    },
    {
      "epoch": 0.1388,
      "loss_align": 0.5271130204200745,
      "loss_contrastive": 0.11961814761161804,
      "loss_gen": 0.6099897027015686,
      "neg_sim_mean": 0.49250513315200806,
      "pos_sim_mean": 0.47288697957992554,
      "step": 694
    },
    {
      "epoch": 0.139,
      "grad_norm": 3.5581071376800537,
      "learning_rate": 9.954211731839624e-05,
      "loss": 1.1515,
      "step": 695
    },
    {
      "epoch": 0.139,
      "loss_align": 0.4489672780036926,
      "loss_contrastive": 0.006399482488632202,
      "loss_gen": 0.3674847185611725,
      "neg_sim_mean": 0.45743221044540405,
      "pos_sim_mean": 0.5510327219963074,
      "step": 695
    },
    {
      "epoch": 0.1392,
      "grad_norm": 2.3628642559051514,
      "learning_rate": 9.953739202357218e-05,
      "loss": 0.8172,
      "step": 696
    },
    {
      "epoch": 0.1392,
      "loss_align": 0.4769051671028137,
      "loss_contrastive": 0.14357700943946838,
      "loss_gen": 0.28602102398872375,
      "neg_sim_mean": 0.5666718482971191,
      "pos_sim_mean": 0.5230948328971863,
      "step": 696
    },
    {
      "epoch": 0.1394,
      "grad_norm": 2.835441827774048,
      "learning_rate": 9.953264258482504e-05,
      "loss": 0.7802,
      "step": 697
    },
    {
      "epoch": 0.1394,
      "loss_align": 0.6348860263824463,
      "loss_contrastive": 0.15244394540786743,
      "loss_gen": 0.28940239548683167,
      "neg_sim_mean": 0.41755789518356323,
      "pos_sim_mean": 0.3651139438152313,
      "step": 697
    },
    {
      "epoch": 0.1396,
      "grad_norm": 2.405369281768799,
      "learning_rate": 9.952786900446965e-05,
      "loss": 0.9426,
      "step": 698
    },
    {
      "epoch": 0.1396,
      "loss_align": 0.462355375289917,
      "loss_contrastive": 0.14571437239646912,
      "loss_gen": 0.33251234889030457,
      "neg_sim_mean": 0.5833590030670166,
      "pos_sim_mean": 0.537644624710083,
      "step": 698
    },
    {
      "epoch": 0.1398,
      "grad_norm": 2.5895442962646484,
      "learning_rate": 9.952307128483256e-05,
      "loss": 0.8124,
      "step": 699
    },
    {
      "epoch": 0.1398,
      "loss_align": 0.45530128479003906,
      "loss_contrastive": 0.12113246321678162,
      "loss_gen": 0.29038506746292114,
      "neg_sim_mean": 0.565831184387207,
      "pos_sim_mean": 0.5446987152099609,
      "step": 699
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.6017985343933105,
      "learning_rate": 9.951824942825215e-05,
      "loss": 0.7602,
      "step": 700
    },
    {
      "epoch": 0.14,
      "loss_align": 0.4584038257598877,
      "loss_contrastive": 0.13340917229652405,
      "loss_gen": 0.2697322368621826,
      "neg_sim_mean": 0.5750053524971008,
      "pos_sim_mean": 0.5415961742401123,
      "step": 700
    },
    {
      "epoch": 0.1402,
      "grad_norm": 2.311006784439087,
      "learning_rate": 9.951340343707852e-05,
      "loss": 0.7441,
      "step": 701
    },
    {
      "epoch": 0.1402,
      "loss_align": 0.40069717168807983,
      "loss_contrastive": 0.0709172785282135,
      "loss_gen": 0.3570605516433716,
      "neg_sim_mean": 0.5702201128005981,
      "pos_sim_mean": 0.5993028283119202,
      "step": 701
    },
    {
      "epoch": 0.1404,
      "grad_norm": 2.3849124908447266,
      "learning_rate": 9.950853331367356e-05,
      "loss": 0.7663,
      "step": 702
    },
    {
      "epoch": 0.1404,
      "loss_align": 0.47093796730041504,
      "loss_contrastive": 0.12612232565879822,
      "loss_gen": 0.3631810247898102,
      "neg_sim_mean": 0.5551843643188477,
      "pos_sim_mean": 0.529062032699585,
      "step": 702
    },
    {
      "epoch": 0.1406,
      "grad_norm": 2.5819623470306396,
      "learning_rate": 9.950363906041089e-05,
      "loss": 0.8493,
      "step": 703
    },
    {
      "epoch": 0.1406,
      "loss_align": 0.3210163116455078,
      "loss_contrastive": 0.08343452215194702,
      "loss_gen": 0.30793631076812744,
      "neg_sim_mean": 0.6624181866645813,
      "pos_sim_mean": 0.6789836883544922,
      "step": 703
    },
    {
      "epoch": 0.1408,
      "grad_norm": 2.48396372795105,
      "learning_rate": 9.94987206796759e-05,
      "loss": 0.639,
      "step": 704
    },
    {
      "epoch": 0.1408,
      "loss_align": 0.374326229095459,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028152069076895714,
      "neg_sim_mean": 0.522883951663971,
      "pos_sim_mean": 0.625673770904541,
      "step": 704
    },
    {
      "epoch": 0.141,
      "grad_norm": 1.4620615243911743,
      "learning_rate": 9.949377817386579e-05,
      "loss": 0.4025,
      "step": 705
    },
    {
      "epoch": 0.141,
      "loss_align": 0.4519764184951782,
      "loss_contrastive": 0.08468332886695862,
      "loss_gen": 0.09139774739742279,
      "neg_sim_mean": 0.5327069163322449,
      "pos_sim_mean": 0.5480235815048218,
      "step": 705
    },
    {
      "epoch": 0.1412,
      "grad_norm": 1.7072526216506958,
      "learning_rate": 9.948881154538945e-05,
      "loss": 0.5535,
      "step": 706
    },
    {
      "epoch": 0.1412,
      "loss_align": 0.5052671432495117,
      "loss_contrastive": 0.12975165247917175,
      "loss_gen": 0.02655801735818386,
      "neg_sim_mean": 0.5244845151901245,
      "pos_sim_mean": 0.4947328567504883,
      "step": 706
    },
    {
      "epoch": 0.1414,
      "grad_norm": 1.329419732093811,
      "learning_rate": 9.948382079666755e-05,
      "loss": 0.5474,
      "step": 707
    },
    {
      "epoch": 0.1414,
      "loss_align": 0.43411123752593994,
      "loss_contrastive": 0.12069258093833923,
      "loss_gen": 0.18003328144550323,
      "neg_sim_mean": 0.5865813493728638,
      "pos_sim_mean": 0.5658887624740601,
      "step": 707
    },
    {
      "epoch": 0.1416,
      "grad_norm": 1.9298638105392456,
      "learning_rate": 9.947880593013255e-05,
      "loss": 0.6286,
      "step": 708
    },
    {
      "epoch": 0.1416,
      "loss_align": 0.44584059715270996,
      "loss_contrastive": 0.08387133479118347,
      "loss_gen": 0.3308878242969513,
      "neg_sim_mean": 0.538030743598938,
      "pos_sim_mean": 0.55415940284729,
      "step": 708
    },
    {
      "epoch": 0.1418,
      "grad_norm": 2.360288143157959,
      "learning_rate": 9.947376694822861e-05,
      "loss": 0.7868,
      "step": 709
    },
    {
      "epoch": 0.1418,
      "loss_align": 0.3534266948699951,
      "loss_contrastive": 0.12423312664031982,
      "loss_gen": 0.46589016914367676,
      "neg_sim_mean": 0.6708064079284668,
      "pos_sim_mean": 0.6465733051300049,
      "step": 709
    },
    {
      "epoch": 0.142,
      "grad_norm": 2.927459716796875,
      "learning_rate": 9.946870385341167e-05,
      "loss": 0.8342,
      "step": 710
    },
    {
      "epoch": 0.142,
      "loss_align": 0.3785419464111328,
      "loss_contrastive": 0.0024924278259277344,
      "loss_gen": 0.009839199483394623,
      "neg_sim_mean": 0.523950457572937,
      "pos_sim_mean": 0.6214580535888672,
      "step": 710
    },
    {
      "epoch": 0.1422,
      "grad_norm": 0.9294435977935791,
      "learning_rate": 9.946361664814943e-05,
      "loss": 0.3887,
      "step": 711
    },
    {
      "epoch": 0.1422,
      "loss_align": 0.5016694664955139,
      "loss_contrastive": 0.20197775959968567,
      "loss_gen": 0.41502514481544495,
      "neg_sim_mean": 0.6003082990646362,
      "pos_sim_mean": 0.4983305335044861,
      "step": 711
    },
    {
      "epoch": 0.1424,
      "grad_norm": 2.4292855262756348,
      "learning_rate": 9.945850533492131e-05,
      "loss": 0.9409,
      "step": 712
    },
    {
      "epoch": 0.1424,
      "loss_align": 0.3367766737937927,
      "loss_contrastive": 0.04128235578536987,
      "loss_gen": 0.19935746490955353,
      "neg_sim_mean": 0.6045056581497192,
      "pos_sim_mean": 0.6632233262062073,
      "step": 712
    },
    {
      "epoch": 0.1426,
      "grad_norm": 2.0345633029937744,
      "learning_rate": 9.945336991621853e-05,
      "loss": 0.5411,
      "step": 713
    },
    {
      "epoch": 0.1426,
      "loss_align": 0.4539632797241211,
      "loss_contrastive": 0.11293813586235046,
      "loss_gen": 0.22663459181785583,
      "neg_sim_mean": 0.5589748620986938,
      "pos_sim_mean": 0.5460367202758789,
      "step": 713
    },
    {
      "epoch": 0.1428,
      "grad_norm": 2.3816347122192383,
      "learning_rate": 9.944821039454402e-05,
      "loss": 0.6942,
      "step": 714
    },
    {
      "epoch": 0.1428,
      "loss_align": 0.3749389052391052,
      "loss_contrastive": 0.09628134965896606,
      "loss_gen": 0.185208261013031,
      "neg_sim_mean": 0.6213424205780029,
      "pos_sim_mean": 0.6250610947608948,
      "step": 714
    },
    {
      "epoch": 0.143,
      "grad_norm": 2.0237174034118652,
      "learning_rate": 9.944302677241246e-05,
      "loss": 0.5717,
      "step": 715
    },
    {
      "epoch": 0.143,
      "loss_align": 0.46234601736068726,
      "loss_contrastive": 0.18054941296577454,
      "loss_gen": 0.2267376333475113,
      "neg_sim_mean": 0.6182034015655518,
      "pos_sim_mean": 0.5376539826393127,
      "step": 715
    },
    {
      "epoch": 0.1432,
      "grad_norm": 2.044534206390381,
      "learning_rate": 9.94378190523503e-05,
      "loss": 0.7107,
      "step": 716
    },
    {
      "epoch": 0.1432,
      "loss_align": 0.5528774261474609,
      "loss_contrastive": 0.11239442229270935,
      "loss_gen": 0.17740535736083984,
      "neg_sim_mean": 0.4595169723033905,
      "pos_sim_mean": 0.4471225440502167,
      "step": 716
    },
    {
      "epoch": 0.1434,
      "grad_norm": 1.878922462463379,
      "learning_rate": 9.94325872368957e-05,
      "loss": 0.7438,
      "step": 717
    },
    {
      "epoch": 0.1434,
      "loss_align": 0.44703882932662964,
      "loss_contrastive": 0.08952471613883972,
      "loss_gen": 0.3992731273174286,
      "neg_sim_mean": 0.5424858927726746,
      "pos_sim_mean": 0.5529611706733704,
      "step": 717
    },
    {
      "epoch": 0.1436,
      "grad_norm": 2.7987654209136963,
      "learning_rate": 9.94273313285986e-05,
      "loss": 0.8571,
      "step": 718
    },
    {
      "epoch": 0.1436,
      "loss_align": 0.520528256893158,
      "loss_contrastive": 0.19234302639961243,
      "loss_gen": 0.21613827347755432,
      "neg_sim_mean": 0.571814775466919,
      "pos_sim_mean": 0.47947174310684204,
      "step": 718
    },
    {
      "epoch": 0.1438,
      "grad_norm": 2.2411224842071533,
      "learning_rate": 9.942205133002068e-05,
      "loss": 0.7597,
      "step": 719
    },
    {
      "epoch": 0.1438,
      "loss_align": 0.5254244804382324,
      "loss_contrastive": 0.12818634510040283,
      "loss_gen": 0.21798397600650787,
      "neg_sim_mean": 0.5027618408203125,
      "pos_sim_mean": 0.4745754897594452,
      "step": 719
    },
    {
      "epoch": 0.144,
      "grad_norm": 2.308933973312378,
      "learning_rate": 9.941674724373531e-05,
      "loss": 0.7588,
      "step": 720
    },
    {
      "epoch": 0.144,
      "loss_align": 0.37178534269332886,
      "loss_contrastive": 0.11361998319625854,
      "loss_gen": 0.3708633780479431,
      "neg_sim_mean": 0.6418346166610718,
      "pos_sim_mean": 0.6282146573066711,
      "step": 720
    },
    {
      "epoch": 0.1442,
      "grad_norm": 2.4599595069885254,
      "learning_rate": 9.941141907232765e-05,
      "loss": 0.7563,
      "step": 721
    },
    {
      "epoch": 0.1442,
      "loss_align": 0.49610722064971924,
      "loss_contrastive": 0.0500011146068573,
      "loss_gen": 0.13659334182739258,
      "neg_sim_mean": 0.45389389991760254,
      "pos_sim_mean": 0.5038927793502808,
      "step": 721
    },
    {
      "epoch": 0.1444,
      "grad_norm": 1.757448434829712,
      "learning_rate": 9.94060668183946e-05,
      "loss": 0.6387,
      "step": 722
    },
    {
      "epoch": 0.1444,
      "loss_align": 0.44384056329727173,
      "loss_contrastive": 0.07499995827674866,
      "loss_gen": 0.523643970489502,
      "neg_sim_mean": 0.5311594009399414,
      "pos_sim_mean": 0.5561594367027283,
      "step": 722
    },
    {
      "epoch": 0.1446,
      "grad_norm": 2.504560947418213,
      "learning_rate": 9.940069048454476e-05,
      "loss": 0.9765,
      "step": 723
    },
    {
      "epoch": 0.1446,
      "loss_align": 0.3588891625404358,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010805698111653328,
      "neg_sim_mean": 0.5350767374038696,
      "pos_sim_mean": 0.6411108374595642,
      "step": 723
    },
    {
      "epoch": 0.1448,
      "grad_norm": 0.8151720762252808,
      "learning_rate": 9.939529007339852e-05,
      "loss": 0.3697,
      "step": 724
    },
    {
      "epoch": 0.1448,
      "loss_align": 0.43168479204177856,
      "loss_contrastive": 0.06899937987327576,
      "loss_gen": 0.0018304495606571436,
      "neg_sim_mean": 0.5373145937919617,
      "pos_sim_mean": 0.5683152079582214,
      "step": 724
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.41593337059020996,
      "learning_rate": 9.938986558758795e-05,
      "loss": 0.4418,
      "step": 725
    },
    {
      "epoch": 0.145,
      "loss_align": 0.36704492568969727,
      "loss_contrastive": 0.12342393398284912,
      "loss_gen": 0.1978827863931656,
      "neg_sim_mean": 0.656378984451294,
      "pos_sim_mean": 0.6329550743103027,
      "step": 725
    },
    {
      "epoch": 0.1452,
      "grad_norm": 2.116053819656372,
      "learning_rate": 9.938441702975689e-05,
      "loss": 0.5797,
      "step": 726
    },
    {
      "epoch": 0.1452,
      "loss_align": 0.33593326807022095,
      "loss_contrastive": 0.0,
      "loss_gen": 0.34085866808891296,
      "neg_sim_mean": 0.5520090460777283,
      "pos_sim_mean": 0.664066731929779,
      "step": 726
    },
    {
      "epoch": 0.1454,
      "grad_norm": 2.509838342666626,
      "learning_rate": 9.93789444025609e-05,
      "loss": 0.6768,
      "step": 727
    },
    {
      "epoch": 0.1454,
      "loss_align": 0.3733600378036499,
      "loss_contrastive": 0.007187783718109131,
      "loss_gen": 0.024905076250433922,
      "neg_sim_mean": 0.5338277220726013,
      "pos_sim_mean": 0.6266399621963501,
      "step": 727
    },
    {
      "epoch": 0.1456,
      "grad_norm": 1.4935091733932495,
      "learning_rate": 9.937344770866728e-05,
      "loss": 0.3991,
      "step": 728
    },
    {
      "epoch": 0.1456,
      "loss_align": 0.3747902512550354,
      "loss_contrastive": 0.12742453813552856,
      "loss_gen": 0.5052085518836975,
      "neg_sim_mean": 0.6526342630386353,
      "pos_sim_mean": 0.6252097487449646,
      "step": 728
    },
    {
      "epoch": 0.1458,
      "grad_norm": 3.4471099376678467,
      "learning_rate": 9.936792695075502e-05,
      "loss": 0.8953,
      "step": 729
    },
    {
      "epoch": 0.1458,
      "loss_align": 0.44896769523620605,
      "loss_contrastive": 0.016644150018692017,
      "loss_gen": 0.18632568418979645,
      "neg_sim_mean": 0.46767646074295044,
      "pos_sim_mean": 0.551032304763794,
      "step": 729
    },
    {
      "epoch": 0.146,
      "grad_norm": 1.883104681968689,
      "learning_rate": 9.936238213151491e-05,
      "loss": 0.6373,
      "step": 730
    },
    {
      "epoch": 0.146,
      "loss_align": 0.27310264110565186,
      "loss_contrastive": 0.037860214710235596,
      "loss_gen": 0.32719215750694275,
      "neg_sim_mean": 0.6647575497627258,
      "pos_sim_mean": 0.7268973588943481,
      "step": 730
    },
    {
      "epoch": 0.1462,
      "grad_norm": 2.795058012008667,
      "learning_rate": 9.93568132536494e-05,
      "loss": 0.6048,
      "step": 731
    },
    {
      "epoch": 0.1462,
      "loss_align": 0.3448220491409302,
      "loss_contrastive": 0.12680411338806152,
      "loss_gen": 0.42179208993911743,
      "neg_sim_mean": 0.6819820404052734,
      "pos_sim_mean": 0.6551779508590698,
      "step": 731
    },
    {
      "epoch": 0.1464,
      "grad_norm": 3.1829211711883545,
      "learning_rate": 9.93512203198727e-05,
      "loss": 0.7818,
      "step": 732
    },
    {
      "epoch": 0.1464,
      "loss_align": 0.45181727409362793,
      "loss_contrastive": 0.07925352454185486,
      "loss_gen": 0.005832582712173462,
      "neg_sim_mean": 0.5274362564086914,
      "pos_sim_mean": 0.5481827259063721,
      "step": 732
    },
    {
      "epoch": 0.1466,
      "grad_norm": 0.6099985837936401,
      "learning_rate": 9.934560333291076e-05,
      "loss": 0.4672,
      "step": 733
    },
    {
      "epoch": 0.1466,
      "loss_align": 0.4701445698738098,
      "loss_contrastive": 0.22964146733283997,
      "loss_gen": 0.34968873858451843,
      "neg_sim_mean": 0.6594969034194946,
      "pos_sim_mean": 0.5298554301261902,
      "step": 733
    },
    {
      "epoch": 0.1468,
      "grad_norm": 2.5029618740081787,
      "learning_rate": 9.93399622955012e-05,
      "loss": 0.8474,
      "step": 734
    },
    {
      "epoch": 0.1468,
      "loss_align": 0.4742974042892456,
      "loss_contrastive": 0.1966380774974823,
      "loss_gen": 0.22393207252025604,
      "neg_sim_mean": 0.6223406791687012,
      "pos_sim_mean": 0.5257025957107544,
      "step": 734
    },
    {
      "epoch": 0.147,
      "grad_norm": 1.9645209312438965,
      "learning_rate": 9.93342972103934e-05,
      "loss": 0.7218,
      "step": 735
    },
    {
      "epoch": 0.147,
      "loss_align": 0.542301595211029,
      "loss_contrastive": 0.15581998229026794,
      "loss_gen": 0.31992483139038086,
      "neg_sim_mean": 0.5135183930397034,
      "pos_sim_mean": 0.45769840478897095,
      "step": 735
    },
    {
      "epoch": 0.1472,
      "grad_norm": 2.634200096130371,
      "learning_rate": 9.932860808034848e-05,
      "loss": 0.8809,
      "step": 736
    },
    {
      "epoch": 0.1472,
      "loss_align": 0.4448798894882202,
      "loss_contrastive": 0.07890406250953674,
      "loss_gen": 0.000735304958652705,
      "neg_sim_mean": 0.534024178981781,
      "pos_sim_mean": 0.5551201105117798,
      "step": 736
    },
    {
      "epoch": 0.1474,
      "grad_norm": 0.3980177640914917,
      "learning_rate": 9.932289490813922e-05,
      "loss": 0.4551,
      "step": 737
    },
    {
      "epoch": 0.1474,
      "loss_align": 0.4128260016441345,
      "loss_contrastive": 0.13626894354820251,
      "loss_gen": 0.4034982919692993,
      "neg_sim_mean": 0.6234429478645325,
      "pos_sim_mean": 0.5871739983558655,
      "step": 737
    },
    {
      "epoch": 0.1476,
      "grad_norm": 2.3286006450653076,
      "learning_rate": 9.931715769655015e-05,
      "loss": 0.8327,
      "step": 738
    },
    {
      "epoch": 0.1476,
      "loss_align": 0.4548969864845276,
      "loss_contrastive": 0.14649203419685364,
      "loss_gen": 0.3754461109638214,
      "neg_sim_mean": 0.5915950536727905,
      "pos_sim_mean": 0.5451030135154724,
      "step": 738
    },
    {
      "epoch": 0.1478,
      "grad_norm": 2.7763872146606445,
      "learning_rate": 9.931139644837754e-05,
      "loss": 0.8479,
      "step": 739
    },
    {
      "epoch": 0.1478,
      "loss_align": 0.325594425201416,
      "loss_contrastive": 0.12856680154800415,
      "loss_gen": 0.2409670203924179,
      "neg_sim_mean": 0.7029723525047302,
      "pos_sim_mean": 0.674405574798584,
      "step": 739
    },
    {
      "epoch": 0.148,
      "grad_norm": 2.5653879642486572,
      "learning_rate": 9.930561116642935e-05,
      "loss": 0.582,
      "step": 740
    },
    {
      "epoch": 0.148,
      "loss_align": 0.43642544746398926,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23055587708950043,
      "neg_sim_mean": 0.4558696448802948,
      "pos_sim_mean": 0.5635745525360107,
      "step": 740
    },
    {
      "epoch": 0.1482,
      "grad_norm": 2.5987682342529297,
      "learning_rate": 9.929980185352526e-05,
      "loss": 0.667,
      "step": 741
    },
    {
      "epoch": 0.1482,
      "loss_align": 0.4547067880630493,
      "loss_contrastive": 0.12100300192832947,
      "loss_gen": 0.30187562108039856,
      "neg_sim_mean": 0.5662962198257446,
      "pos_sim_mean": 0.5452932119369507,
      "step": 741
    },
    {
      "epoch": 0.1484,
      "grad_norm": 2.0060837268829346,
      "learning_rate": 9.929396851249661e-05,
      "loss": 0.7711,
      "step": 742
    },
    {
      "epoch": 0.1484,
      "loss_align": 0.4326239824295044,
      "loss_contrastive": 0.07422217726707458,
      "loss_gen": 0.23178425431251526,
      "neg_sim_mean": 0.5415982007980347,
      "pos_sim_mean": 0.5673760175704956,
      "step": 742
    },
    {
      "epoch": 0.1486,
      "grad_norm": 2.4758667945861816,
      "learning_rate": 9.928811114618656e-05,
      "loss": 0.6733,
      "step": 743
    },
    {
      "epoch": 0.1486,
      "loss_align": 0.424052357673645,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3773021101951599,
      "neg_sim_mean": 0.45137646794319153,
      "pos_sim_mean": 0.575947642326355,
      "step": 743
    },
    {
      "epoch": 0.1488,
      "grad_norm": 2.3427772521972656,
      "learning_rate": 9.928222975744991e-05,
      "loss": 0.8014,
      "step": 744
    },
    {
      "epoch": 0.1488,
      "loss_align": 0.3839113712310791,
      "loss_contrastive": 0.009579658508300781,
      "loss_gen": 0.018019631505012512,
      "neg_sim_mean": 0.5256682634353638,
      "pos_sim_mean": 0.6160886287689209,
      "step": 744
    },
    {
      "epoch": 0.149,
      "grad_norm": 1.216425895690918,
      "learning_rate": 9.927632434915314e-05,
      "loss": 0.4031,
      "step": 745
    },
    {
      "epoch": 0.149,
      "loss_align": 0.407598614692688,
      "loss_contrastive": 0.1438896358013153,
      "loss_gen": 0.27671557664871216,
      "neg_sim_mean": 0.6362910270690918,
      "pos_sim_mean": 0.592401385307312,
      "step": 745
    },
    {
      "epoch": 0.1492,
      "grad_norm": 2.750720262527466,
      "learning_rate": 9.927039492417452e-05,
      "loss": 0.7016,
      "step": 746
    },
    {
      "epoch": 0.1492,
      "loss_align": 0.4437780976295471,
      "loss_contrastive": 0.06999489665031433,
      "loss_gen": 0.002466969657689333,
      "neg_sim_mean": 0.5262168049812317,
      "pos_sim_mean": 0.5562219023704529,
      "step": 746
    },
    {
      "epoch": 0.1494,
      "grad_norm": 0.4465416967868805,
      "learning_rate": 9.926444148540393e-05,
      "loss": 0.4546,
      "step": 747
    },
    {
      "epoch": 0.1494,
      "loss_align": 0.48387813568115234,
      "loss_contrastive": 0.11977896094322205,
      "loss_gen": 0.23137451708316803,
      "neg_sim_mean": 0.5359008312225342,
      "pos_sim_mean": 0.5161218643188477,
      "step": 747
    },
    {
      "epoch": 0.1496,
      "grad_norm": 2.163264036178589,
      "learning_rate": 9.925846403574305e-05,
      "loss": 0.7296,
      "step": 748
    },
    {
      "epoch": 0.1496,
      "loss_align": 0.4034906029701233,
      "loss_contrastive": 0.07693985104560852,
      "loss_gen": 0.13921363651752472,
      "neg_sim_mean": 0.5734492540359497,
      "pos_sim_mean": 0.5965093970298767,
      "step": 748
    },
    {
      "epoch": 0.1498,
      "grad_norm": 2.0369250774383545,
      "learning_rate": 9.925246257810519e-05,
      "loss": 0.5519,
      "step": 749
    },
    {
      "epoch": 0.1498,
      "loss_align": 0.3626716732978821,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0028627540450543165,
      "neg_sim_mean": 0.5190661549568176,
      "pos_sim_mean": 0.6373283267021179,
      "step": 749
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4237993657588959,
      "learning_rate": 9.924643711541539e-05,
      "loss": 0.3655,
      "step": 750
    },
    {
      "epoch": 0.15,
      "loss_align": 0.5935620665550232,
      "loss_contrastive": 0.13282832503318787,
      "loss_gen": 0.25477099418640137,
      "neg_sim_mean": 0.43926626443862915,
      "pos_sim_mean": 0.4064379334449768,
      "step": 750
    },
    {
      "epoch": 0.1502,
      "grad_norm": 2.050766944885254,
      "learning_rate": 9.924038765061042e-05,
      "loss": 0.8643,
      "step": 751
    },
    {
      "epoch": 0.1502,
      "loss_align": 0.5311694741249084,
      "loss_contrastive": 0.18268552422523499,
      "loss_gen": 0.1622648388147354,
      "neg_sim_mean": 0.551516056060791,
      "pos_sim_mean": 0.46883052587509155,
      "step": 751
    },
    {
      "epoch": 0.1504,
      "grad_norm": 1.8545762300491333,
      "learning_rate": 9.923431418663865e-05,
      "loss": 0.7154,
      "step": 752
    },
    {
      "epoch": 0.1504,
      "loss_align": 0.47957563400268555,
      "loss_contrastive": 0.09667512774467468,
      "loss_gen": 0.20537713170051575,
      "neg_sim_mean": 0.5170994997024536,
      "pos_sim_mean": 0.5204243659973145,
      "step": 752
    },
    {
      "epoch": 0.1506,
      "grad_norm": 2.8152430057525635,
      "learning_rate": 9.922821672646027e-05,
      "loss": 0.6966,
      "step": 753
    },
    {
      "epoch": 0.1506,
      "loss_align": 0.45585471391677856,
      "loss_contrastive": 0.09585198760032654,
      "loss_gen": 0.29928794503211975,
      "neg_sim_mean": 0.5399972796440125,
      "pos_sim_mean": 0.5441452860832214,
      "step": 753
    },
    {
      "epoch": 0.1508,
      "grad_norm": 2.009305715560913,
      "learning_rate": 9.92220952730471e-05,
      "loss": 0.7666,
      "step": 754
    },
    {
      "epoch": 0.1508,
      "loss_align": 0.449588418006897,
      "loss_contrastive": 0.09247830510139465,
      "loss_gen": 0.19881723821163177,
      "neg_sim_mean": 0.5428898930549622,
      "pos_sim_mean": 0.550411581993103,
      "step": 754
    },
    {
      "epoch": 0.151,
      "grad_norm": 1.8571348190307617,
      "learning_rate": 9.921594982938263e-05,
      "loss": 0.6595,
      "step": 755
    },
    {
      "epoch": 0.151,
      "loss_align": 0.47200852632522583,
      "loss_contrastive": 0.10370543599128723,
      "loss_gen": 0.3316049873828888,
      "neg_sim_mean": 0.5316969156265259,
      "pos_sim_mean": 0.5279914736747742,
      "step": 755
    },
    {
      "epoch": 0.1512,
      "grad_norm": 2.4131031036376953,
      "learning_rate": 9.92097803984621e-05,
      "loss": 0.8161,
      "step": 756
    },
    {
      "epoch": 0.1512,
      "loss_align": 0.3648728132247925,
      "loss_contrastive": 0.03060305118560791,
      "loss_gen": 0.21366052329540253,
      "neg_sim_mean": 0.5657302141189575,
      "pos_sim_mean": 0.6351271867752075,
      "step": 756
    },
    {
      "epoch": 0.1514,
      "grad_norm": 2.141334295272827,
      "learning_rate": 9.920358698329242e-05,
      "loss": 0.5822,
      "step": 757
    },
    {
      "epoch": 0.1514,
      "loss_align": 0.332699716091156,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2801547050476074,
      "neg_sim_mean": 0.5282911658287048,
      "pos_sim_mean": 0.667300283908844,
      "step": 757
    },
    {
      "epoch": 0.1516,
      "grad_norm": 2.3398799896240234,
      "learning_rate": 9.919736958689217e-05,
      "loss": 0.6129,
      "step": 758
    },
    {
      "epoch": 0.1516,
      "loss_align": 0.4308595657348633,
      "loss_contrastive": 0.13897892832756042,
      "loss_gen": 0.30252715945243835,
      "neg_sim_mean": 0.6081193685531616,
      "pos_sim_mean": 0.5691404342651367,
      "step": 758
    },
    {
      "epoch": 0.1518,
      "grad_norm": 2.2776012420654297,
      "learning_rate": 9.919112821229163e-05,
      "loss": 0.7501,
      "step": 759
    },
    {
      "epoch": 0.1518,
      "loss_align": 0.357499361038208,
      "loss_contrastive": 0.07262694835662842,
      "loss_gen": 0.26388877630233765,
      "neg_sim_mean": 0.6151275634765625,
      "pos_sim_mean": 0.642500638961792,
      "step": 759
    },
    {
      "epoch": 0.152,
      "grad_norm": 2.520012378692627,
      "learning_rate": 9.918486286253279e-05,
      "loss": 0.6301,
      "step": 760
    },
    {
      "epoch": 0.152,
      "loss_align": 0.4497736692428589,
      "loss_contrastive": 0.1864217221736908,
      "loss_gen": 0.24137093126773834,
      "neg_sim_mean": 0.6366480588912964,
      "pos_sim_mean": 0.5502263307571411,
      "step": 760
    },
    {
      "epoch": 0.1522,
      "grad_norm": 2.2072904109954834,
      "learning_rate": 9.917857354066931e-05,
      "loss": 0.7135,
      "step": 761
    },
    {
      "epoch": 0.1522,
      "loss_align": 0.5963643789291382,
      "loss_contrastive": 0.215447336435318,
      "loss_gen": 0.4339088499546051,
      "neg_sim_mean": 0.5190829634666443,
      "pos_sim_mean": 0.4036356210708618,
      "step": 761
    },
    {
      "epoch": 0.1524,
      "grad_norm": 2.7324728965759277,
      "learning_rate": 9.917226024976649e-05,
      "loss": 1.0561,
      "step": 762
    },
    {
      "epoch": 0.1524,
      "loss_align": 0.5438852310180664,
      "loss_contrastive": 0.036769360303878784,
      "loss_gen": 0.33893921971321106,
      "neg_sim_mean": 0.39288410544395447,
      "pos_sim_mean": 0.4561147391796112,
      "step": 762
    },
    {
      "epoch": 0.1526,
      "grad_norm": 2.323103189468384,
      "learning_rate": 9.916592299290138e-05,
      "loss": 0.8872,
      "step": 763
    },
    {
      "epoch": 0.1526,
      "loss_align": 0.5227675437927246,
      "loss_contrastive": 0.09588384628295898,
      "loss_gen": 0.2518293559551239,
      "neg_sim_mean": 0.47311627864837646,
      "pos_sim_mean": 0.477232426404953,
      "step": 763
    },
    {
      "epoch": 0.1528,
      "grad_norm": 2.4027631282806396,
      "learning_rate": 9.915956177316267e-05,
      "loss": 0.7861,
      "step": 764
    },
    {
      "epoch": 0.1528,
      "loss_align": 0.4228851795196533,
      "loss_contrastive": 0.1491898000240326,
      "loss_gen": 0.4053054749965668,
      "neg_sim_mean": 0.6263046264648438,
      "pos_sim_mean": 0.5771148204803467,
      "step": 764
    },
    {
      "epoch": 0.153,
      "grad_norm": 2.765937089920044,
      "learning_rate": 9.915317659365077e-05,
      "loss": 0.8461,
      "step": 765
    },
    {
      "epoch": 0.153,
      "loss_align": 0.4392566680908203,
      "loss_contrastive": 0.18206587433815002,
      "loss_gen": 0.1321311891078949,
      "neg_sim_mean": 0.6428092122077942,
      "pos_sim_mean": 0.5607433319091797,
      "step": 765
    },
    {
      "epoch": 0.1532,
      "grad_norm": 1.3994423151016235,
      "learning_rate": 9.914676745747772e-05,
      "loss": 0.5932,
      "step": 766
    },
    {
      "epoch": 0.1532,
      "loss_align": 0.4152982234954834,
      "loss_contrastive": 0.08993253111839294,
      "loss_gen": 0.639436662197113,
      "neg_sim_mean": 0.574634313583374,
      "pos_sim_mean": 0.5847017765045166,
      "step": 766
    },
    {
      "epoch": 0.1534,
      "grad_norm": 4.107356548309326,
      "learning_rate": 9.914033436776724e-05,
      "loss": 1.0655,
      "step": 767
    },
    {
      "epoch": 0.1534,
      "loss_align": 0.5846061706542969,
      "loss_contrastive": 0.21355822682380676,
      "loss_gen": 0.3602442443370819,
      "neg_sim_mean": 0.5289520621299744,
      "pos_sim_mean": 0.4153938293457031,
      "step": 767
    },
    {
      "epoch": 0.1536,
      "grad_norm": 2.7683725357055664,
      "learning_rate": 9.913387732765475e-05,
      "loss": 0.9705,
      "step": 768
    },
    {
      "epoch": 0.1536,
      "loss_align": 0.4024617075920105,
      "loss_contrastive": 0.11832728981971741,
      "loss_gen": 0.3304308354854584,
      "neg_sim_mean": 0.6158655881881714,
      "pos_sim_mean": 0.5975382924079895,
      "step": 768
    },
    {
      "epoch": 0.1538,
      "grad_norm": 2.9039812088012695,
      "learning_rate": 9.912739634028733e-05,
      "loss": 0.7471,
      "step": 769
    },
    {
      "epoch": 0.1538,
      "loss_align": 0.30548906326293945,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18500696122646332,
      "neg_sim_mean": 0.582070529460907,
      "pos_sim_mean": 0.6945109367370605,
      "step": 769
    },
    {
      "epoch": 0.154,
      "grad_norm": 2.1774327754974365,
      "learning_rate": 9.912089140882376e-05,
      "loss": 0.4905,
      "step": 770
    },
    {
      "epoch": 0.154,
      "loss_align": 0.36628973484039307,
      "loss_contrastive": 0.10801684856414795,
      "loss_gen": 0.16923213005065918,
      "neg_sim_mean": 0.641727089881897,
      "pos_sim_mean": 0.6337102651596069,
      "step": 770
    },
    {
      "epoch": 0.1542,
      "grad_norm": 2.0456345081329346,
      "learning_rate": 9.911436253643445e-05,
      "loss": 0.5485,
      "step": 771
    },
    {
      "epoch": 0.1542,
      "loss_align": 0.4210021495819092,
      "loss_contrastive": 0.06671962141990662,
      "loss_gen": 0.04774986952543259,
      "neg_sim_mean": 0.5457174777984619,
      "pos_sim_mean": 0.5789978504180908,
      "step": 771
    },
    {
      "epoch": 0.1544,
      "grad_norm": 1.7869011163711548,
      "learning_rate": 9.910780972630146e-05,
      "loss": 0.4768,
      "step": 772
    },
    {
      "epoch": 0.1544,
      "loss_align": 0.49806779623031616,
      "loss_contrastive": 0.1341109573841095,
      "loss_gen": 0.35776597261428833,
      "neg_sim_mean": 0.5360431671142578,
      "pos_sim_mean": 0.5019322037696838,
      "step": 772
    },
    {
      "epoch": 0.1546,
      "grad_norm": 2.4380040168762207,
      "learning_rate": 9.91012329816186e-05,
      "loss": 0.8719,
      "step": 773
    },
    {
      "epoch": 0.1546,
      "loss_align": 0.4081364870071411,
      "loss_contrastive": 0.2250429093837738,
      "loss_gen": 0.264457106590271,
      "neg_sim_mean": 0.7169064283370972,
      "pos_sim_mean": 0.5918635129928589,
      "step": 773
    },
    {
      "epoch": 0.1548,
      "grad_norm": 2.140528440475464,
      "learning_rate": 9.909463230559127e-05,
      "loss": 0.6996,
      "step": 774
    },
    {
      "epoch": 0.1548,
      "loss_align": 0.4638570547103882,
      "loss_contrastive": 0.012655287981033325,
      "loss_gen": 0.2280658334493637,
      "neg_sim_mean": 0.4487982392311096,
      "pos_sim_mean": 0.5361429452896118,
      "step": 774
    },
    {
      "epoch": 0.155,
      "grad_norm": 2.432093620300293,
      "learning_rate": 9.908800770143653e-05,
      "loss": 0.6934,
      "step": 775
    },
    {
      "epoch": 0.155,
      "loss_align": 0.5155020952224731,
      "loss_contrastive": 0.15497690439224243,
      "loss_gen": 0.020322151482105255,
      "neg_sim_mean": 0.5394748449325562,
      "pos_sim_mean": 0.48449793457984924,
      "step": 775
    },
    {
      "epoch": 0.1552,
      "grad_norm": 1.5067108869552612,
      "learning_rate": 9.908135917238321e-05,
      "loss": 0.5544,
      "step": 776
    },
    {
      "epoch": 0.1552,
      "loss_align": 0.4388543367385864,
      "loss_contrastive": 0.15122851729393005,
      "loss_gen": 0.29084083437919617,
      "neg_sim_mean": 0.6123741865158081,
      "pos_sim_mean": 0.5611456632614136,
      "step": 776
    },
    {
      "epoch": 0.1554,
      "grad_norm": 2.3264598846435547,
      "learning_rate": 9.907468672167165e-05,
      "loss": 0.7478,
      "step": 777
    },
    {
      "epoch": 0.1554,
      "loss_align": 0.4251037836074829,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2788790762424469,
      "neg_sim_mean": 0.47281157970428467,
      "pos_sim_mean": 0.5748962163925171,
      "step": 777
    },
    {
      "epoch": 0.1556,
      "grad_norm": 2.5961573123931885,
      "learning_rate": 9.906799035255396e-05,
      "loss": 0.704,
      "step": 778
    },
    {
      "epoch": 0.1556,
      "loss_align": 0.4611303210258484,
      "loss_contrastive": 0.10465672612190247,
      "loss_gen": 0.20402924716472626,
      "neg_sim_mean": 0.5435264110565186,
      "pos_sim_mean": 0.5388696789741516,
      "step": 778
    },
    {
      "epoch": 0.1558,
      "grad_norm": 2.24774432182312,
      "learning_rate": 9.906127006829384e-05,
      "loss": 0.6777,
      "step": 779
    },
    {
      "epoch": 0.1558,
      "loss_align": 0.29459303617477417,
      "loss_contrastive": 0.08195251226425171,
      "loss_gen": 0.3272571563720703,
      "neg_sim_mean": 0.6873594522476196,
      "pos_sim_mean": 0.7054069638252258,
      "step": 779
    },
    {
      "epoch": 0.156,
      "grad_norm": 2.933387517929077,
      "learning_rate": 9.90545258721667e-05,
      "loss": 0.6317,
      "step": 780
    },
    {
      "epoch": 0.156,
      "loss_align": 0.44617176055908203,
      "loss_contrastive": 0.06306609511375427,
      "loss_gen": 0.23909765481948853,
      "neg_sim_mean": 0.5168943405151367,
      "pos_sim_mean": 0.553828239440918,
      "step": 780
    },
    {
      "epoch": 0.1562,
      "grad_norm": 2.23122501373291,
      "learning_rate": 9.904775776745958e-05,
      "loss": 0.6928,
      "step": 781
    },
    {
      "epoch": 0.1562,
      "loss_align": 0.48330509662628174,
      "loss_contrastive": 0.17298665642738342,
      "loss_gen": 0.26402953267097473,
      "neg_sim_mean": 0.5896815657615662,
      "pos_sim_mean": 0.5166949033737183,
      "step": 781
    },
    {
      "epoch": 0.1564,
      "grad_norm": 2.0882744789123535,
      "learning_rate": 9.904096575747116e-05,
      "loss": 0.7681,
      "step": 782
    },
    {
      "epoch": 0.1564,
      "loss_align": 0.3507314920425415,
      "loss_contrastive": 0.17604851722717285,
      "loss_gen": 0.3118339478969574,
      "neg_sim_mean": 0.7253170013427734,
      "pos_sim_mean": 0.6492685079574585,
      "step": 782
    },
    {
      "epoch": 0.1566,
      "grad_norm": 2.2415409088134766,
      "learning_rate": 9.903414984551179e-05,
      "loss": 0.6837,
      "step": 783
    },
    {
      "epoch": 0.1566,
      "loss_align": 0.527445912361145,
      "loss_contrastive": 0.16190561652183533,
      "loss_gen": 0.3045770525932312,
      "neg_sim_mean": 0.5344597101211548,
      "pos_sim_mean": 0.472554087638855,
      "step": 783
    },
    {
      "epoch": 0.1568,
      "grad_norm": 2.426459550857544,
      "learning_rate": 9.902731003490344e-05,
      "loss": 0.8515,
      "step": 784
    },
    {
      "epoch": 0.1568,
      "loss_align": 0.3384361267089844,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011295333504676819,
      "neg_sim_mean": 0.4543708264827728,
      "pos_sim_mean": 0.6615638732910156,
      "step": 784
    },
    {
      "epoch": 0.157,
      "grad_norm": 1.0525271892547607,
      "learning_rate": 9.90204463289798e-05,
      "loss": 0.3497,
      "step": 785
    },
    {
      "epoch": 0.157,
      "loss_align": 0.4182763695716858,
      "loss_contrastive": 0.08982005715370178,
      "loss_gen": 0.40641483664512634,
      "neg_sim_mean": 0.5715436935424805,
      "pos_sim_mean": 0.5817236304283142,
      "step": 785
    },
    {
      "epoch": 0.1572,
      "grad_norm": 2.8277697563171387,
      "learning_rate": 9.901355873108609e-05,
      "loss": 0.8355,
      "step": 786
    },
    {
      "epoch": 0.1572,
      "loss_align": 0.25851333141326904,
      "loss_contrastive": 0.038597822189331055,
      "loss_gen": 0.214501291513443,
      "neg_sim_mean": 0.6800844669342041,
      "pos_sim_mean": 0.741486668586731,
      "step": 786
    },
    {
      "epoch": 0.1574,
      "grad_norm": 2.4349851608276367,
      "learning_rate": 9.900664724457931e-05,
      "loss": 0.4776,
      "step": 787
    },
    {
      "epoch": 0.1574,
      "loss_align": 0.5734924077987671,
      "loss_contrastive": 0.24323174357414246,
      "loss_gen": 0.4371512234210968,
      "neg_sim_mean": 0.5697393417358398,
      "pos_sim_mean": 0.4265075922012329,
      "step": 787
    },
    {
      "epoch": 0.1576,
      "grad_norm": 3.3129329681396484,
      "learning_rate": 9.899971187282798e-05,
      "loss": 1.0398,
      "step": 788
    },
    {
      "epoch": 0.1576,
      "loss_align": 0.3153911828994751,
      "loss_contrastive": 0.008826017379760742,
      "loss_gen": 0.6011728644371033,
      "neg_sim_mean": 0.5934348106384277,
      "pos_sim_mean": 0.6846088171005249,
      "step": 788
    },
    {
      "epoch": 0.1578,
      "grad_norm": 3.727766513824463,
      "learning_rate": 9.899275261921234e-05,
      "loss": 0.9176,
      "step": 789
    },
    {
      "epoch": 0.1578,
      "loss_align": 0.438180148601532,
      "loss_contrastive": 0.1617087423801422,
      "loss_gen": 0.05909745395183563,
      "neg_sim_mean": 0.6235285997390747,
      "pos_sim_mean": 0.561819851398468,
      "step": 789
    },
    {
      "epoch": 0.158,
      "grad_norm": 1.1831594705581665,
      "learning_rate": 9.898576948712426e-05,
      "loss": 0.5167,
      "step": 790
    },
    {
      "epoch": 0.158,
      "loss_align": 0.3691306710243225,
      "loss_contrastive": 0.0,
      "loss_gen": 0.28370097279548645,
      "neg_sim_mean": 0.5120148658752441,
      "pos_sim_mean": 0.6308693289756775,
      "step": 790
    },
    {
      "epoch": 0.1582,
      "grad_norm": 2.497265100479126,
      "learning_rate": 9.89787624799672e-05,
      "loss": 0.6528,
      "step": 791
    },
    {
      "epoch": 0.1582,
      "loss_align": 0.5263890027999878,
      "loss_contrastive": 0.16958889365196228,
      "loss_gen": 0.004163224715739489,
      "neg_sim_mean": 0.543199896812439,
      "pos_sim_mean": 0.4736109972000122,
      "step": 791
    },
    {
      "epoch": 0.1584,
      "grad_norm": 0.8759443163871765,
      "learning_rate": 9.897173160115632e-05,
      "loss": 0.5509,
      "step": 792
    },
    {
      "epoch": 0.1584,
      "loss_align": 0.4222955107688904,
      "loss_contrastive": 0.13077428936958313,
      "loss_gen": 0.11335790902376175,
      "neg_sim_mean": 0.6084787845611572,
      "pos_sim_mean": 0.5777044892311096,
      "step": 792
    },
    {
      "epoch": 0.1586,
      "grad_norm": 1.460977554321289,
      "learning_rate": 9.896467685411838e-05,
      "loss": 0.5513,
      "step": 793
    },
    {
      "epoch": 0.1586,
      "loss_align": 0.4402117133140564,
      "loss_contrastive": 0.11679127812385559,
      "loss_gen": 0.23831139504909515,
      "neg_sim_mean": 0.5765795707702637,
      "pos_sim_mean": 0.5597882866859436,
      "step": 793
    },
    {
      "epoch": 0.1588,
      "grad_norm": 1.951909065246582,
      "learning_rate": 9.895759824229176e-05,
      "loss": 0.6925,
      "step": 794
    },
    {
      "epoch": 0.1588,
      "loss_align": 0.31030136346817017,
      "loss_contrastive": 0.0,
      "loss_gen": 0.37756338715553284,
      "neg_sim_mean": 0.5681697130203247,
      "pos_sim_mean": 0.6896986365318298,
      "step": 794
    },
    {
      "epoch": 0.159,
      "grad_norm": 3.2657532691955566,
      "learning_rate": 9.89504957691265e-05,
      "loss": 0.6879,
      "step": 795
    },
    {
      "epoch": 0.159,
      "loss_align": 0.435843825340271,
      "loss_contrastive": 0.04061153531074524,
      "loss_gen": 0.20808514952659607,
      "neg_sim_mean": 0.5047677159309387,
      "pos_sim_mean": 0.564156174659729,
      "step": 795
    },
    {
      "epoch": 0.1592,
      "grad_norm": 6.022551536560059,
      "learning_rate": 9.894336943808426e-05,
      "loss": 0.6488,
      "step": 796
    },
    {
      "epoch": 0.1592,
      "loss_align": 0.47713154554367065,
      "loss_contrastive": 0.13647696375846863,
      "loss_gen": 0.30393359065055847,
      "neg_sim_mean": 0.5593454241752625,
      "pos_sim_mean": 0.5228684544563293,
      "step": 796
    },
    {
      "epoch": 0.1594,
      "grad_norm": 2.8155243396759033,
      "learning_rate": 9.893621925263832e-05,
      "loss": 0.7974,
      "step": 797
    },
    {
      "epoch": 0.1594,
      "loss_align": 0.34248095750808716,
      "loss_contrastive": 0.027743637561798096,
      "loss_gen": 0.3196139931678772,
      "neg_sim_mean": 0.585262656211853,
      "pos_sim_mean": 0.6575190424919128,
      "step": 797
    },
    {
      "epoch": 0.1596,
      "grad_norm": 2.5005154609680176,
      "learning_rate": 9.892904521627361e-05,
      "loss": 0.6654,
      "step": 798
    },
    {
      "epoch": 0.1596,
      "loss_align": 0.5149094462394714,
      "loss_contrastive": 0.17910125851631165,
      "loss_gen": 0.2533152103424072,
      "neg_sim_mean": 0.5641918182373047,
      "pos_sim_mean": 0.48509055376052856,
      "step": 798
    },
    {
      "epoch": 0.1598,
      "grad_norm": 3.138225555419922,
      "learning_rate": 9.892184733248666e-05,
      "loss": 0.7897,
      "step": 799
    },
    {
      "epoch": 0.1598,
      "loss_align": 0.37094563245773315,
      "loss_contrastive": 0.12143629789352417,
      "loss_gen": 0.3896617889404297,
      "neg_sim_mean": 0.6504906415939331,
      "pos_sim_mean": 0.6290543675422668,
      "step": 799
    },
    {
      "epoch": 0.16,
      "grad_norm": 3.4220082759857178,
      "learning_rate": 9.891462560478562e-05,
      "loss": 0.7752,
      "step": 800
    },
    {
      "epoch": 0.16,
      "loss_align": 0.44443243741989136,
      "loss_contrastive": 0.11696842312812805,
      "loss_gen": 0.30391380190849304,
      "neg_sim_mean": 0.5725359916687012,
      "pos_sim_mean": 0.5555675625801086,
      "step": 800
    },
    {
      "epoch": 0.1602,
      "grad_norm": 2.012058734893799,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.7624,
      "step": 801
    },
    {
      "epoch": 0.1602,
      "loss_align": 0.43887120485305786,
      "loss_contrastive": 0.08446136116981506,
      "loss_gen": 0.19373854994773865,
      "neg_sim_mean": 0.5455901622772217,
      "pos_sim_mean": 0.5611287951469421,
      "step": 801
    },
    {
      "epoch": 0.1604,
      "grad_norm": 2.4058761596679688,
      "learning_rate": 9.890011063173205e-05,
      "loss": 0.6427,
      "step": 802
    },
    {
      "epoch": 0.1604,
      "loss_align": 0.33138561248779297,
      "loss_contrastive": 0.06650221347808838,
      "loss_gen": 0.4905419945716858,
      "neg_sim_mean": 0.6351165771484375,
      "pos_sim_mean": 0.668614387512207,
      "step": 802
    },
    {
      "epoch": 0.1606,
      "grad_norm": 3.5331215858459473,
      "learning_rate": 9.889281739345395e-05,
      "loss": 0.8299,
      "step": 803
    },
    {
      "epoch": 0.1606,
      "loss_align": 0.406832218170166,
      "loss_contrastive": 0.08777841925621033,
      "loss_gen": 0.2861657738685608,
      "neg_sim_mean": 0.5809462070465088,
      "pos_sim_mean": 0.593167781829834,
      "step": 803
    },
    {
      "epoch": 0.1608,
      "grad_norm": 2.8923392295837402,
      "learning_rate": 9.888550032541059e-05,
      "loss": 0.7035,
      "step": 804
    },
    {
      "epoch": 0.1608,
      "loss_align": 0.2863469123840332,
      "loss_contrastive": 0.08445560932159424,
      "loss_gen": 0.17403370141983032,
      "neg_sim_mean": 0.6981086730957031,
      "pos_sim_mean": 0.7136530876159668,
      "step": 804
    },
    {
      "epoch": 0.161,
      "grad_norm": 2.677778959274292,
      "learning_rate": 9.887815943116827e-05,
      "loss": 0.4705,
      "step": 805
    },
    {
      "epoch": 0.161,
      "loss_align": 0.3899669051170349,
      "loss_contrastive": 0.00968468189239502,
      "loss_gen": 0.20376881957054138,
      "neg_sim_mean": 0.5197177529335022,
      "pos_sim_mean": 0.6100330948829651,
      "step": 805
    },
    {
      "epoch": 0.1612,
      "grad_norm": 2.2850182056427,
      "learning_rate": 9.88707947143048e-05,
      "loss": 0.5949,
      "step": 806
    },
    {
      "epoch": 0.1612,
      "loss_align": 0.4181969165802002,
      "loss_contrastive": 0.0360548198223114,
      "loss_gen": 0.30929693579673767,
      "neg_sim_mean": 0.5178579092025757,
      "pos_sim_mean": 0.5818030834197998,
      "step": 806
    },
    {
      "epoch": 0.1614,
      "grad_norm": 2.4515910148620605,
      "learning_rate": 9.886340617840968e-05,
      "loss": 0.7318,
      "step": 807
    },
    {
      "epoch": 0.1614,
      "loss_align": 0.549542248249054,
      "loss_contrastive": 0.13975000381469727,
      "loss_gen": 0.47213587164878845,
      "neg_sim_mean": 0.4902077615261078,
      "pos_sim_mean": 0.45045775175094604,
      "step": 807
    },
    {
      "epoch": 0.1616,
      "grad_norm": 3.651691198348999,
      "learning_rate": 9.8855993827084e-05,
      "loss": 1.0384,
      "step": 808
    },
    {
      "epoch": 0.1616,
      "loss_align": 0.3831942081451416,
      "loss_contrastive": 0.12207382917404175,
      "loss_gen": 0.2012890726327896,
      "neg_sim_mean": 0.6388795971870422,
      "pos_sim_mean": 0.6168057918548584,
      "step": 808
    },
    {
      "epoch": 0.1618,
      "grad_norm": 2.406679630279541,
      "learning_rate": 9.884855766394042e-05,
      "loss": 0.5991,
      "step": 809
    },
    {
      "epoch": 0.1618,
      "loss_align": 0.4039510488510132,
      "loss_contrastive": 0.036712855100631714,
      "loss_gen": 0.3603123128414154,
      "neg_sim_mean": 0.532761812210083,
      "pos_sim_mean": 0.5960489511489868,
      "step": 809
    },
    {
      "epoch": 0.162,
      "grad_norm": 2.976299524307251,
      "learning_rate": 9.884109769260325e-05,
      "loss": 0.7687,
      "step": 810
    },
    {
      "epoch": 0.162,
      "loss_align": 0.39007455110549927,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005227214191108942,
      "neg_sim_mean": 0.5014854669570923,
      "pos_sim_mean": 0.6099254488945007,
      "step": 810
    },
    {
      "epoch": 0.1622,
      "grad_norm": 0.5771586894989014,
      "learning_rate": 9.88336139167084e-05,
      "loss": 0.3953,
      "step": 811
    },
    {
      "epoch": 0.1622,
      "loss_align": 0.47823071479797363,
      "loss_contrastive": 0.0646263062953949,
      "loss_gen": 0.0067055365070700645,
      "neg_sim_mean": 0.48639559745788574,
      "pos_sim_mean": 0.5217692852020264,
      "step": 811
    },
    {
      "epoch": 0.1624,
      "grad_norm": 0.7441064715385437,
      "learning_rate": 9.882610633990337e-05,
      "loss": 0.4927,
      "step": 812
    },
    {
      "epoch": 0.1624,
      "loss_align": 0.39392101764678955,
      "loss_contrastive": 0.03919947147369385,
      "loss_gen": 0.3053688406944275,
      "neg_sim_mean": 0.5452784299850464,
      "pos_sim_mean": 0.6060789823532104,
      "step": 812
    },
    {
      "epoch": 0.1626,
      "grad_norm": 2.7479610443115234,
      "learning_rate": 9.881857496584726e-05,
      "loss": 0.704,
      "step": 813
    },
    {
      "epoch": 0.1626,
      "loss_align": 0.37530815601348877,
      "loss_contrastive": 0.0,
      "loss_gen": 0.31160464882850647,
      "neg_sim_mean": 0.4444255232810974,
      "pos_sim_mean": 0.6246918439865112,
      "step": 813
    },
    {
      "epoch": 0.1628,
      "grad_norm": 2.4568450450897217,
      "learning_rate": 9.881101979821075e-05,
      "loss": 0.6869,
      "step": 814
    },
    {
      "epoch": 0.1628,
      "loss_align": 0.41810721158981323,
      "loss_contrastive": 0.10447958111763,
      "loss_gen": 0.22292543947696686,
      "neg_sim_mean": 0.5863723754882812,
      "pos_sim_mean": 0.5818927884101868,
      "step": 814
    },
    {
      "epoch": 0.163,
      "grad_norm": 2.0443825721740723,
      "learning_rate": 9.880344084067616e-05,
      "loss": 0.6536,
      "step": 815
    },
    {
      "epoch": 0.163,
      "loss_align": 0.4627094268798828,
      "loss_contrastive": 0.16875538229942322,
      "loss_gen": 0.19237300753593445,
      "neg_sim_mean": 0.6060459613800049,
      "pos_sim_mean": 0.5372905731201172,
      "step": 815
    },
    {
      "epoch": 0.1632,
      "grad_norm": 2.2425527572631836,
      "learning_rate": 9.879583809693738e-05,
      "loss": 0.6753,
      "step": 816
    },
    {
      "epoch": 0.1632,
      "loss_align": 0.3655756115913391,
      "loss_contrastive": 0.11353588104248047,
      "loss_gen": 0.7659692764282227,
      "neg_sim_mean": 0.6479602456092834,
      "pos_sim_mean": 0.6344243884086609,
      "step": 816
    },
    {
      "epoch": 0.1634,
      "grad_norm": 3.7879796028137207,
      "learning_rate": 9.878821157069989e-05,
      "loss": 1.1452,
      "step": 817
    },
    {
      "epoch": 0.1634,
      "loss_align": 0.44912809133529663,
      "loss_contrastive": 0.16511979699134827,
      "loss_gen": 0.5062975287437439,
      "neg_sim_mean": 0.6159917116165161,
      "pos_sim_mean": 0.5508719086647034,
      "step": 817
    },
    {
      "epoch": 0.1636,
      "grad_norm": 3.040212631225586,
      "learning_rate": 9.878056126568075e-05,
      "loss": 0.9752,
      "step": 818
    },
    {
      "epoch": 0.1636,
      "loss_align": 0.463875949382782,
      "loss_contrastive": 0.22690477967262268,
      "loss_gen": 0.2911381423473358,
      "neg_sim_mean": 0.6630288362503052,
      "pos_sim_mean": 0.536124050617218,
      "step": 818
    },
    {
      "epoch": 0.1638,
      "grad_norm": 3.6137759685516357,
      "learning_rate": 9.877288718560866e-05,
      "loss": 0.7822,
      "step": 819
    },
    {
      "epoch": 0.1638,
      "loss_align": 0.3474893569946289,
      "loss_contrastive": 0.1768243908882141,
      "loss_gen": 0.2894814908504486,
      "neg_sim_mean": 0.7293350100517273,
      "pos_sim_mean": 0.6525106430053711,
      "step": 819
    },
    {
      "epoch": 0.164,
      "grad_norm": 2.4010701179504395,
      "learning_rate": 9.876518933422386e-05,
      "loss": 0.6582,
      "step": 820
    },
    {
      "epoch": 0.164,
      "loss_align": 0.3427044153213501,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09892360121011734,
      "neg_sim_mean": 0.5280126929283142,
      "pos_sim_mean": 0.6572955846786499,
      "step": 820
    },
    {
      "epoch": 0.1642,
      "grad_norm": 1.4443258047103882,
      "learning_rate": 9.875746771527816e-05,
      "loss": 0.4416,
      "step": 821
    },
    {
      "epoch": 0.1642,
      "loss_align": 0.4802544116973877,
      "loss_contrastive": 0.21780642867088318,
      "loss_gen": 0.20414860546588898,
      "neg_sim_mean": 0.63755202293396,
      "pos_sim_mean": 0.5197455883026123,
      "step": 821
    },
    {
      "epoch": 0.1644,
      "grad_norm": 1.9687540531158447,
      "learning_rate": 9.874972233253504e-05,
      "loss": 0.7105,
      "step": 822
    },
    {
      "epoch": 0.1644,
      "loss_align": 0.4491797089576721,
      "loss_contrastive": 0.09492632746696472,
      "loss_gen": 0.19375625252723694,
      "neg_sim_mean": 0.5457466244697571,
      "pos_sim_mean": 0.5508202910423279,
      "step": 822
    },
    {
      "epoch": 0.1646,
      "grad_norm": 2.0333588123321533,
      "learning_rate": 9.874195318976945e-05,
      "loss": 0.6543,
      "step": 823
    },
    {
      "epoch": 0.1646,
      "loss_align": 0.38312196731567383,
      "loss_contrastive": 0.07241320610046387,
      "loss_gen": 0.2594773471355438,
      "neg_sim_mean": 0.5892912149429321,
      "pos_sim_mean": 0.6168780326843262,
      "step": 823
    },
    {
      "epoch": 0.1648,
      "grad_norm": 2.2985525131225586,
      "learning_rate": 9.873416029076801e-05,
      "loss": 0.6513,
      "step": 824
    },
    {
      "epoch": 0.1648,
      "loss_align": 0.3277777433395386,
      "loss_contrastive": 0.10337281227111816,
      "loss_gen": 0.4628726840019226,
      "neg_sim_mean": 0.6755950450897217,
      "pos_sim_mean": 0.6722222566604614,
      "step": 824
    },
    {
      "epoch": 0.165,
      "grad_norm": 3.0914697647094727,
      "learning_rate": 9.872634363932887e-05,
      "loss": 0.8031,
      "step": 825
    },
    {
      "epoch": 0.165,
      "loss_align": 0.34866517782211304,
      "loss_contrastive": 0.14177244901657104,
      "loss_gen": 0.29038840532302856,
      "neg_sim_mean": 0.6931072473526001,
      "pos_sim_mean": 0.651334822177887,
      "step": 825
    },
    {
      "epoch": 0.1652,
      "grad_norm": 2.37093448638916,
      "learning_rate": 9.871850323926177e-05,
      "loss": 0.6561,
      "step": 826
    },
    {
      "epoch": 0.1652,
      "loss_align": 0.2670246362686157,
      "loss_contrastive": 0.0,
      "loss_gen": 0.034035298973321915,
      "neg_sim_mean": 0.455016553401947,
      "pos_sim_mean": 0.7329753637313843,
      "step": 826
    },
    {
      "epoch": 0.1654,
      "grad_norm": 2.735124111175537,
      "learning_rate": 9.871063909438803e-05,
      "loss": 0.3011,
      "step": 827
    },
    {
      "epoch": 0.1654,
      "loss_align": 0.29649269580841064,
      "loss_contrastive": 0.03614389896392822,
      "loss_gen": 0.32688841223716736,
      "neg_sim_mean": 0.6396511793136597,
      "pos_sim_mean": 0.7035073041915894,
      "step": 827
    },
    {
      "epoch": 0.1656,
      "grad_norm": 2.518077850341797,
      "learning_rate": 9.870275120854054e-05,
      "loss": 0.6277,
      "step": 828
    },
    {
      "epoch": 0.1656,
      "loss_align": 0.3844172954559326,
      "loss_contrastive": 0.1423642635345459,
      "loss_gen": 0.34708768129348755,
      "neg_sim_mean": 0.6579469442367554,
      "pos_sim_mean": 0.6155827045440674,
      "step": 828
    },
    {
      "epoch": 0.1658,
      "grad_norm": 2.7247345447540283,
      "learning_rate": 9.869483958556375e-05,
      "loss": 0.7486,
      "step": 829
    },
    {
      "epoch": 0.1658,
      "loss_align": 0.23447465896606445,
      "loss_contrastive": 0.0,
      "loss_gen": 0.28793230652809143,
      "neg_sim_mean": 0.5054068565368652,
      "pos_sim_mean": 0.7655253410339355,
      "step": 829
    },
    {
      "epoch": 0.166,
      "grad_norm": 2.490105152130127,
      "learning_rate": 9.868690422931372e-05,
      "loss": 0.5224,
      "step": 830
    },
    {
      "epoch": 0.166,
      "loss_align": 0.49662649631500244,
      "loss_contrastive": 0.15894952416419983,
      "loss_gen": 0.2845914363861084,
      "neg_sim_mean": 0.5623230338096619,
      "pos_sim_mean": 0.5033735036849976,
      "step": 830
    },
    {
      "epoch": 0.1662,
      "grad_norm": 2.553978443145752,
      "learning_rate": 9.867894514365802e-05,
      "loss": 0.8003,
      "step": 831
    },
    {
      "epoch": 0.1662,
      "loss_align": 0.2922865152359009,
      "loss_contrastive": 0.09366554021835327,
      "loss_gen": 0.43181923031806946,
      "neg_sim_mean": 0.7013790011405945,
      "pos_sim_mean": 0.7077134847640991,
      "step": 831
    },
    {
      "epoch": 0.1664,
      "grad_norm": 2.3542160987854004,
      "learning_rate": 9.867096233247581e-05,
      "loss": 0.7353,
      "step": 832
    },
    {
      "epoch": 0.1664,
      "loss_align": 0.3811153173446655,
      "loss_contrastive": 0.09897422790527344,
      "loss_gen": 0.40567654371261597,
      "neg_sim_mean": 0.61785888671875,
      "pos_sim_mean": 0.6188846826553345,
      "step": 832
    },
    {
      "epoch": 0.1666,
      "grad_norm": 2.6590354442596436,
      "learning_rate": 9.86629557996578e-05,
      "loss": 0.7987,
      "step": 833
    },
    {
      "epoch": 0.1666,
      "loss_align": 0.34258973598480225,
      "loss_contrastive": 0.11785531044006348,
      "loss_gen": 0.4814872145652771,
      "neg_sim_mean": 0.6752655506134033,
      "pos_sim_mean": 0.6574102640151978,
      "step": 833
    },
    {
      "epoch": 0.1668,
      "grad_norm": 3.162956476211548,
      "learning_rate": 9.865492554910633e-05,
      "loss": 0.8382,
      "step": 834
    },
    {
      "epoch": 0.1668,
      "loss_align": 0.5517908930778503,
      "loss_contrastive": 0.17808184027671814,
      "loss_gen": 0.31114083528518677,
      "neg_sim_mean": 0.5262909531593323,
      "pos_sim_mean": 0.44820910692214966,
      "step": 834
    },
    {
      "epoch": 0.167,
      "grad_norm": 2.6080715656280518,
      "learning_rate": 9.86468715847352e-05,
      "loss": 0.8843,
      "step": 835
    },
    {
      "epoch": 0.167,
      "loss_align": 0.30058664083480835,
      "loss_contrastive": 0.09718281030654907,
      "loss_gen": 0.3534819185733795,
      "neg_sim_mean": 0.6965961456298828,
      "pos_sim_mean": 0.6994133591651917,
      "step": 835
    },
    {
      "epoch": 0.1672,
      "grad_norm": 2.54770827293396,
      "learning_rate": 9.863879391046984e-05,
      "loss": 0.6657,
      "step": 836
    },
    {
      "epoch": 0.1672,
      "loss_align": 0.5094277858734131,
      "loss_contrastive": 0.15219327807426453,
      "loss_gen": 0.3128429353237152,
      "neg_sim_mean": 0.5427654981613159,
      "pos_sim_mean": 0.4905722141265869,
      "step": 836
    },
    {
      "epoch": 0.1674,
      "grad_norm": 3.01906418800354,
      "learning_rate": 9.863069253024719e-05,
      "loss": 0.8405,
      "step": 837
    },
    {
      "epoch": 0.1674,
      "loss_align": 0.47175222635269165,
      "loss_contrastive": 0.17883697152137756,
      "loss_gen": 0.1419340968132019,
      "neg_sim_mean": 0.6070847511291504,
      "pos_sim_mean": 0.5282477736473083,
      "step": 837
    },
    {
      "epoch": 0.1676,
      "grad_norm": 1.720219612121582,
      "learning_rate": 9.862256744801577e-05,
      "loss": 0.6351,
      "step": 838
    },
    {
      "epoch": 0.1676,
      "loss_align": 0.36182039976119995,
      "loss_contrastive": 0.11841511726379395,
      "loss_gen": 0.3302779197692871,
      "neg_sim_mean": 0.6565946936607361,
      "pos_sim_mean": 0.6381796002388,
      "step": 838
    },
    {
      "epoch": 0.1678,
      "grad_norm": 2.2059526443481445,
      "learning_rate": 9.861441866773564e-05,
      "loss": 0.7063,
      "step": 839
    },
    {
      "epoch": 0.1678,
      "loss_align": 0.4156886339187622,
      "loss_contrastive": 0.16144809126853943,
      "loss_gen": 0.1200157105922699,
      "neg_sim_mean": 0.6457594633102417,
      "pos_sim_mean": 0.5843113660812378,
      "step": 839
    },
    {
      "epoch": 0.168,
      "grad_norm": 1.6624672412872314,
      "learning_rate": 9.860624619337844e-05,
      "loss": 0.5551,
      "step": 840
    },
    {
      "epoch": 0.168,
      "loss_align": 0.4611319303512573,
      "loss_contrastive": 0.12479332089424133,
      "loss_gen": 0.19707030057907104,
      "neg_sim_mean": 0.5636613965034485,
      "pos_sim_mean": 0.5388680696487427,
      "step": 840
    },
    {
      "epoch": 0.1682,
      "grad_norm": 2.087273597717285,
      "learning_rate": 9.859805002892732e-05,
      "loss": 0.6732,
      "step": 841
    },
    {
      "epoch": 0.1682,
      "loss_align": 0.46054673194885254,
      "loss_contrastive": 0.14204886555671692,
      "loss_gen": 0.028002994135022163,
      "neg_sim_mean": 0.5815021395683289,
      "pos_sim_mean": 0.5394532680511475,
      "step": 841
    },
    {
      "epoch": 0.1684,
      "grad_norm": 1.4682905673980713,
      "learning_rate": 9.858983017837698e-05,
      "loss": 0.5056,
      "step": 842
    },
    {
      "epoch": 0.1684,
      "loss_align": 0.3190884590148926,
      "loss_contrastive": 0.08817136287689209,
      "loss_gen": 0.48293745517730713,
      "neg_sim_mean": 0.6690828800201416,
      "pos_sim_mean": 0.6809115409851074,
      "step": 842
    },
    {
      "epoch": 0.1686,
      "grad_norm": 3.026524543762207,
      "learning_rate": 9.85815866457337e-05,
      "loss": 0.8126,
      "step": 843
    },
    {
      "epoch": 0.1686,
      "loss_align": 0.3829152584075928,
      "loss_contrastive": 0.07767373323440552,
      "loss_gen": 0.5328832864761353,
      "neg_sim_mean": 0.5947584509849548,
      "pos_sim_mean": 0.6170847415924072,
      "step": 843
    },
    {
      "epoch": 0.1688,
      "grad_norm": 3.2954282760620117,
      "learning_rate": 9.857331943501527e-05,
      "loss": 0.9251,
      "step": 844
    },
    {
      "epoch": 0.1688,
      "loss_align": 0.27915048599243164,
      "loss_contrastive": 0.09325659275054932,
      "loss_gen": 0.33398565649986267,
      "neg_sim_mean": 0.7141060829162598,
      "pos_sim_mean": 0.7208495140075684,
      "step": 844
    },
    {
      "epoch": 0.169,
      "grad_norm": 2.5486180782318115,
      "learning_rate": 9.8565028550251e-05,
      "loss": 0.6243,
      "step": 845
    },
    {
      "epoch": 0.169,
      "loss_align": 0.4287601113319397,
      "loss_contrastive": 0.1708090603351593,
      "loss_gen": 0.32080313563346863,
      "neg_sim_mean": 0.6420489549636841,
      "pos_sim_mean": 0.5712398886680603,
      "step": 845
    },
    {
      "epoch": 0.1692,
      "grad_norm": 2.649864912033081,
      "learning_rate": 9.855671399548181e-05,
      "loss": 0.7701,
      "step": 846
    },
    {
      "epoch": 0.1692,
      "loss_align": 0.4099959135055542,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028157783672213554,
      "neg_sim_mean": 0.46989160776138306,
      "pos_sim_mean": 0.5900040864944458,
      "step": 846
    },
    {
      "epoch": 0.1694,
      "grad_norm": 1.5234758853912354,
      "learning_rate": 9.854837577476008e-05,
      "loss": 0.4382,
      "step": 847
    },
    {
      "epoch": 0.1694,
      "loss_align": 0.3867309093475342,
      "loss_contrastive": 0.12363505363464355,
      "loss_gen": 0.3208252191543579,
      "neg_sim_mean": 0.6369041204452515,
      "pos_sim_mean": 0.6132690906524658,
      "step": 847
    },
    {
      "epoch": 0.1696,
      "grad_norm": 2.650357246398926,
      "learning_rate": 9.854001389214978e-05,
      "loss": 0.7224,
      "step": 848
    },
    {
      "epoch": 0.1696,
      "loss_align": 0.26128649711608887,
      "loss_contrastive": 0.05150938034057617,
      "loss_gen": 0.11377652734518051,
      "neg_sim_mean": 0.6902228593826294,
      "pos_sim_mean": 0.7387135028839111,
      "step": 848
    },
    {
      "epoch": 0.1698,
      "grad_norm": 1.5707221031188965,
      "learning_rate": 9.853162835172637e-05,
      "loss": 0.3812,
      "step": 849
    },
    {
      "epoch": 0.1698,
      "loss_align": 0.36958467960357666,
      "loss_contrastive": 0.16450893878936768,
      "loss_gen": 0.4088340401649475,
      "neg_sim_mean": 0.6949242353439331,
      "pos_sim_mean": 0.6304153203964233,
      "step": 849
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.080021619796753,
      "learning_rate": 9.852321915757687e-05,
      "loss": 0.7982,
      "step": 850
    },
    {
      "epoch": 0.17,
      "loss_align": 0.3476414680480957,
      "loss_contrastive": 0.02833002805709839,
      "loss_gen": 0.014311186969280243,
      "neg_sim_mean": 0.5806885361671448,
      "pos_sim_mean": 0.6523585319519043,
      "step": 850
    },
    {
      "epoch": 0.1702,
      "grad_norm": 1.0317028760910034,
      "learning_rate": 9.851478631379982e-05,
      "loss": 0.3654,
      "step": 851
    },
    {
      "epoch": 0.1702,
      "loss_align": 0.3878794312477112,
      "loss_contrastive": 0.08775204420089722,
      "loss_gen": 0.4560433030128479,
      "neg_sim_mean": 0.5998725891113281,
      "pos_sim_mean": 0.6121205687522888,
      "step": 851
    },
    {
      "epoch": 0.1704,
      "grad_norm": 3.0134506225585938,
      "learning_rate": 9.85063298245053e-05,
      "loss": 0.8545,
      "step": 852
    },
    {
      "epoch": 0.1704,
      "loss_align": 0.4978073835372925,
      "loss_contrastive": 0.15978160500526428,
      "loss_gen": 0.4042084813117981,
      "neg_sim_mean": 0.5619742274284363,
      "pos_sim_mean": 0.5021926164627075,
      "step": 852
    },
    {
      "epoch": 0.1706,
      "grad_norm": 2.6582531929016113,
      "learning_rate": 9.849784969381486e-05,
      "loss": 0.9212,
      "step": 853
    },
    {
      "epoch": 0.1706,
      "loss_align": 0.3876449465751648,
      "loss_contrastive": 0.08704715967178345,
      "loss_gen": 0.232362300157547,
      "neg_sim_mean": 0.5994021892547607,
      "pos_sim_mean": 0.6123550534248352,
      "step": 853
    },
    {
      "epoch": 0.1708,
      "grad_norm": 2.343400716781616,
      "learning_rate": 9.848934592586166e-05,
      "loss": 0.6305,
      "step": 854
    },
    {
      "epoch": 0.1708,
      "loss_align": 0.23570466041564941,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0031812412198632956,
      "neg_sim_mean": 0.48332396149635315,
      "pos_sim_mean": 0.7642953395843506,
      "step": 854
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.4383050203323364,
      "learning_rate": 9.84808185247903e-05,
      "loss": 0.2389,
      "step": 855
    },
    {
      "epoch": 0.171,
      "loss_align": 0.35934311151504517,
      "loss_contrastive": 0.10859721899032593,
      "loss_gen": 0.21660107374191284,
      "neg_sim_mean": 0.6492540836334229,
      "pos_sim_mean": 0.6406568884849548,
      "step": 855
    },
    {
      "epoch": 0.1712,
      "grad_norm": 2.147675037384033,
      "learning_rate": 9.847226749475695e-05,
      "loss": 0.589,
      "step": 856
    },
    {
      "epoch": 0.1712,
      "loss_align": 0.32376301288604736,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0005795969627797604,
      "neg_sim_mean": 0.5056392550468445,
      "pos_sim_mean": 0.6762369871139526,
      "step": 856
    },
    {
      "epoch": 0.1714,
      "grad_norm": 0.49917125701904297,
      "learning_rate": 9.846369283992926e-05,
      "loss": 0.3243,
      "step": 857
    },
    {
      "epoch": 0.1714,
      "loss_align": 0.31288278102874756,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20459803938865662,
      "neg_sim_mean": 0.583777129650116,
      "pos_sim_mean": 0.6871172189712524,
      "step": 857
    },
    {
      "epoch": 0.1716,
      "grad_norm": 2.19319748878479,
      "learning_rate": 9.845509456448643e-05,
      "loss": 0.5175,
      "step": 858
    },
    {
      "epoch": 0.1716,
      "loss_align": 0.44105279445648193,
      "loss_contrastive": 0.20558640360832214,
      "loss_gen": 0.344039648771286,
      "neg_sim_mean": 0.6645336151123047,
      "pos_sim_mean": 0.5589472055435181,
      "step": 858
    },
    {
      "epoch": 0.1718,
      "grad_norm": 3.001392364501953,
      "learning_rate": 9.844647267261916e-05,
      "loss": 0.8098,
      "step": 859
    },
    {
      "epoch": 0.1718,
      "loss_align": 0.2642136812210083,
      "loss_contrastive": 0.09820526838302612,
      "loss_gen": 0.3033656179904938,
      "neg_sim_mean": 0.7339915633201599,
      "pos_sim_mean": 0.7357863187789917,
      "step": 859
    },
    {
      "epoch": 0.172,
      "grad_norm": 2.6809115409851074,
      "learning_rate": 9.843782716852963e-05,
      "loss": 0.5794,
      "step": 860
    },
    {
      "epoch": 0.172,
      "loss_align": 0.3598838448524475,
      "loss_contrastive": 0.10100698471069336,
      "loss_gen": 0.0017086411826312542,
      "neg_sim_mean": 0.6411231160163879,
      "pos_sim_mean": 0.6401161551475525,
      "step": 860
    },
    {
      "epoch": 0.1722,
      "grad_norm": 1.0038012266159058,
      "learning_rate": 9.842915805643155e-05,
      "loss": 0.3737,
      "step": 861
    },
    {
      "epoch": 0.1722,
      "loss_align": 0.48355770111083984,
      "loss_contrastive": 0.2221379578113556,
      "loss_gen": 0.26313316822052,
      "neg_sim_mean": 0.6385802626609802,
      "pos_sim_mean": 0.5164422988891602,
      "step": 861
    },
    {
      "epoch": 0.1724,
      "grad_norm": 2.8418591022491455,
      "learning_rate": 9.842046534055019e-05,
      "loss": 0.7733,
      "step": 862
    },
    {
      "epoch": 0.1724,
      "loss_align": 0.37855273485183716,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012235555797815323,
      "neg_sim_mean": 0.49055546522140503,
      "pos_sim_mean": 0.6214472651481628,
      "step": 862
    },
    {
      "epoch": 0.1726,
      "grad_norm": 13.53811264038086,
      "learning_rate": 9.841174902512223e-05,
      "loss": 0.3908,
      "step": 863
    },
    {
      "epoch": 0.1726,
      "loss_align": 0.28533506393432617,
      "loss_contrastive": 0.09218257665634155,
      "loss_gen": 0.4484202563762665,
      "neg_sim_mean": 0.7068474888801575,
      "pos_sim_mean": 0.7146649360656738,
      "step": 863
    },
    {
      "epoch": 0.1728,
      "grad_norm": 3.2706565856933594,
      "learning_rate": 9.840300911439591e-05,
      "loss": 0.7448,
      "step": 864
    },
    {
      "epoch": 0.1728,
      "loss_align": 0.3126174211502075,
      "loss_contrastive": 0.0789041519165039,
      "loss_gen": 0.3500964939594269,
      "neg_sim_mean": 0.6662867069244385,
      "pos_sim_mean": 0.6873825788497925,
      "step": 864
    },
    {
      "epoch": 0.173,
      "grad_norm": 2.325822353363037,
      "learning_rate": 9.839424561263093e-05,
      "loss": 0.6722,
      "step": 865
    },
    {
      "epoch": 0.173,
      "loss_align": 0.290835440158844,
      "loss_contrastive": 0.07518899440765381,
      "loss_gen": 0.4357629120349884,
      "neg_sim_mean": 0.6843535304069519,
      "pos_sim_mean": 0.709164559841156,
      "step": 865
    },
    {
      "epoch": 0.1732,
      "grad_norm": 3.1243209838867188,
      "learning_rate": 9.838545852409857e-05,
      "loss": 0.7356,
      "step": 866
    },
    {
      "epoch": 0.1732,
      "loss_align": 0.3169596195220947,
      "loss_contrastive": 0.0693972110748291,
      "loss_gen": 0.361904501914978,
      "neg_sim_mean": 0.6524375677108765,
      "pos_sim_mean": 0.6830403804779053,
      "step": 866
    },
    {
      "epoch": 0.1734,
      "grad_norm": 2.5544135570526123,
      "learning_rate": 9.837664785308149e-05,
      "loss": 0.6872,
      "step": 867
    },
    {
      "epoch": 0.1734,
      "loss_align": 0.4538049101829529,
      "loss_contrastive": 0.17246904969215393,
      "loss_gen": 0.4453934133052826,
      "neg_sim_mean": 0.6186641454696655,
      "pos_sim_mean": 0.5461950898170471,
      "step": 867
    },
    {
      "epoch": 0.1736,
      "grad_norm": 2.8866782188415527,
      "learning_rate": 9.836781360387396e-05,
      "loss": 0.9199,
      "step": 868
    },
    {
      "epoch": 0.1736,
      "loss_align": 0.4676128625869751,
      "loss_contrastive": 0.20998749136924744,
      "loss_gen": 0.3990058898925781,
      "neg_sim_mean": 0.6423746347427368,
      "pos_sim_mean": 0.5323871374130249,
      "step": 868
    },
    {
      "epoch": 0.1738,
      "grad_norm": 3.1298060417175293,
      "learning_rate": 9.835895578078165e-05,
      "loss": 0.8918,
      "step": 869
    },
    {
      "epoch": 0.1738,
      "loss_align": 0.36798346042633057,
      "loss_contrastive": 0.08292877674102783,
      "loss_gen": 0.3428002893924713,
      "neg_sim_mean": 0.6149452924728394,
      "pos_sim_mean": 0.6320165395736694,
      "step": 869
    },
    {
      "epoch": 0.174,
      "grad_norm": 2.50455641746521,
      "learning_rate": 9.835007438812177e-05,
      "loss": 0.7207,
      "step": 870
    },
    {
      "epoch": 0.174,
      "loss_align": 0.2815623879432678,
      "loss_contrastive": 0.10570985078811646,
      "loss_gen": 0.517719030380249,
      "neg_sim_mean": 0.7241474390029907,
      "pos_sim_mean": 0.7184376120567322,
      "step": 870
    },
    {
      "epoch": 0.1742,
      "grad_norm": 3.300144910812378,
      "learning_rate": 9.834116943022298e-05,
      "loss": 0.812,
      "step": 871
    },
    {
      "epoch": 0.1742,
      "loss_align": 0.41577398777008057,
      "loss_contrastive": 0.12489119172096252,
      "loss_gen": 0.00850186962634325,
      "neg_sim_mean": 0.6091172099113464,
      "pos_sim_mean": 0.5842260122299194,
      "step": 871
    },
    {
      "epoch": 0.1744,
      "grad_norm": 1.4420585632324219,
      "learning_rate": 9.833224091142547e-05,
      "loss": 0.4393,
      "step": 872
    },
    {
      "epoch": 0.1744,
      "loss_align": 0.355690062046051,
      "loss_contrastive": 0.07834804058074951,
      "loss_gen": 0.3223697543144226,
      "neg_sim_mean": 0.6226579546928406,
      "pos_sim_mean": 0.644309937953949,
      "step": 872
    },
    {
      "epoch": 0.1746,
      "grad_norm": 2.4603495597839355,
      "learning_rate": 9.832328883608088e-05,
      "loss": 0.6875,
      "step": 873
    },
    {
      "epoch": 0.1746,
      "loss_align": 0.3077791929244995,
      "loss_contrastive": 0.08233559131622314,
      "loss_gen": 0.783178448677063,
      "neg_sim_mean": 0.6745563745498657,
      "pos_sim_mean": 0.6922208070755005,
      "step": 873
    },
    {
      "epoch": 0.1748,
      "grad_norm": 3.8271431922912598,
      "learning_rate": 9.831431320855235e-05,
      "loss": 1.1008,
      "step": 874
    },
    {
      "epoch": 0.1748,
      "loss_align": 0.2872183322906494,
      "loss_contrastive": 0.15570837259292603,
      "loss_gen": 0.2734012305736542,
      "neg_sim_mean": 0.7684900164604187,
      "pos_sim_mean": 0.7127816677093506,
      "step": 874
    },
    {
      "epoch": 0.175,
      "grad_norm": 2.2056493759155273,
      "learning_rate": 9.830531403321451e-05,
      "loss": 0.5793,
      "step": 875
    },
    {
      "epoch": 0.175,
      "loss_align": 0.3659363389015198,
      "loss_contrastive": 0.18256878852844238,
      "loss_gen": 0.19181594252586365,
      "neg_sim_mean": 0.7166324257850647,
      "pos_sim_mean": 0.6340636610984802,
      "step": 875
    },
    {
      "epoch": 0.1752,
      "grad_norm": 1.983808159828186,
      "learning_rate": 9.829629131445342e-05,
      "loss": 0.5797,
      "step": 876
    },
    {
      "epoch": 0.1752,
      "loss_align": 0.6493775844573975,
      "loss_contrastive": 0.5032268166542053,
      "loss_gen": 0.3167959451675415,
      "neg_sim_mean": 0.7538492679595947,
      "pos_sim_mean": 0.3506224453449249,
      "step": 876
    },
    {
      "epoch": 0.1754,
      "grad_norm": 2.5115318298339844,
      "learning_rate": 9.828724505666664e-05,
      "loss": 1.0266,
      "step": 877
    },
    {
      "epoch": 0.1754,
      "loss_align": 0.4385397434234619,
      "loss_contrastive": 0.10207626223564148,
      "loss_gen": 0.030258357524871826,
      "neg_sim_mean": 0.563536524772644,
      "pos_sim_mean": 0.5614602565765381,
      "step": 877
    },
    {
      "epoch": 0.1756,
      "grad_norm": 1.6320137977600098,
      "learning_rate": 9.827817526426324e-05,
      "loss": 0.481,
      "step": 878
    },
    {
      "epoch": 0.1756,
      "loss_align": 0.26978766918182373,
      "loss_contrastive": 0.02463388442993164,
      "loss_gen": 0.18136896193027496,
      "neg_sim_mean": 0.65484619140625,
      "pos_sim_mean": 0.7302123308181763,
      "step": 878
    },
    {
      "epoch": 0.1758,
      "grad_norm": 2.0254127979278564,
      "learning_rate": 9.82690819416637e-05,
      "loss": 0.4541,
      "step": 879
    },
    {
      "epoch": 0.1758,
      "loss_align": 0.49751198291778564,
      "loss_contrastive": 0.19614467024803162,
      "loss_gen": 0.46209976077079773,
      "neg_sim_mean": 0.5986326932907104,
      "pos_sim_mean": 0.5024880170822144,
      "step": 879
    },
    {
      "epoch": 0.176,
      "grad_norm": 3.03190279006958,
      "learning_rate": 9.82599650933e-05,
      "loss": 0.9831,
      "step": 880
    },
    {
      "epoch": 0.176,
      "loss_align": 0.3926035165786743,
      "loss_contrastive": 0.13189023733139038,
      "loss_gen": 0.2599489092826843,
      "neg_sim_mean": 0.6392866969108582,
      "pos_sim_mean": 0.6073964834213257,
      "step": 880
    },
    {
      "epoch": 0.1762,
      "grad_norm": 2.3398234844207764,
      "learning_rate": 9.825082472361557e-05,
      "loss": 0.6684,
      "step": 881
    },
    {
      "epoch": 0.1762,
      "loss_align": 0.3936949372291565,
      "loss_contrastive": 0.08434885740280151,
      "loss_gen": 0.3523317873477936,
      "neg_sim_mean": 0.5906538963317871,
      "pos_sim_mean": 0.6063050627708435,
      "step": 881
    },
    {
      "epoch": 0.1764,
      "grad_norm": 2.285114288330078,
      "learning_rate": 9.824166083706534e-05,
      "loss": 0.7561,
      "step": 882
    },
    {
      "epoch": 0.1764,
      "loss_align": 0.4063529372215271,
      "loss_contrastive": 0.06082412600517273,
      "loss_gen": 0.3308386206626892,
      "neg_sim_mean": 0.5544711947441101,
      "pos_sim_mean": 0.5936470627784729,
      "step": 882
    },
    {
      "epoch": 0.1766,
      "grad_norm": 2.3753702640533447,
      "learning_rate": 9.823247343811567e-05,
      "loss": 0.7445,
      "step": 883
    },
    {
      "epoch": 0.1766,
      "loss_align": 0.29404252767562866,
      "loss_contrastive": 0.11815804243087769,
      "loss_gen": 0.43304160237312317,
      "neg_sim_mean": 0.7241154909133911,
      "pos_sim_mean": 0.7059574723243713,
      "step": 883
    },
    {
      "epoch": 0.1768,
      "grad_norm": 2.6221280097961426,
      "learning_rate": 9.822326253124437e-05,
      "loss": 0.7413,
      "step": 884
    },
    {
      "epoch": 0.1768,
      "loss_align": 0.5493403673171997,
      "loss_contrastive": 0.15649205446243286,
      "loss_gen": 0.2945738434791565,
      "neg_sim_mean": 0.50715172290802,
      "pos_sim_mean": 0.4506596624851227,
      "step": 884
    },
    {
      "epoch": 0.177,
      "grad_norm": 2.7851836681365967,
      "learning_rate": 9.821402812094073e-05,
      "loss": 0.8627,
      "step": 885
    },
    {
      "epoch": 0.177,
      "loss_align": 0.22825807332992554,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04513034224510193,
      "neg_sim_mean": 0.49370887875556946,
      "pos_sim_mean": 0.7717419266700745,
      "step": 885
    },
    {
      "epoch": 0.1772,
      "grad_norm": 1.9039539098739624,
      "learning_rate": 9.820477021170551e-05,
      "loss": 0.2734,
      "step": 886
    },
    {
      "epoch": 0.1772,
      "loss_align": 0.48779040575027466,
      "loss_contrastive": 0.1814335286617279,
      "loss_gen": 0.34839749336242676,
      "neg_sim_mean": 0.5936431288719177,
      "pos_sim_mean": 0.5122095942497253,
      "step": 886
    },
    {
      "epoch": 0.1774,
      "grad_norm": 2.9575300216674805,
      "learning_rate": 9.819548880805087e-05,
      "loss": 0.858,
      "step": 887
    },
    {
      "epoch": 0.1774,
      "loss_align": 0.27611392736434937,
      "loss_contrastive": 0.07088738679885864,
      "loss_gen": 0.26274964213371277,
      "neg_sim_mean": 0.6947734355926514,
      "pos_sim_mean": 0.7238860726356506,
      "step": 887
    },
    {
      "epoch": 0.1776,
      "grad_norm": 2.877164840698242,
      "learning_rate": 9.81861839145005e-05,
      "loss": 0.5474,
      "step": 888
    },
    {
      "epoch": 0.1776,
      "loss_align": 0.47611701488494873,
      "loss_contrastive": 0.08200231194496155,
      "loss_gen": 0.21953025460243225,
      "neg_sim_mean": 0.5058853030204773,
      "pos_sim_mean": 0.5238829851150513,
      "step": 888
    },
    {
      "epoch": 0.1778,
      "grad_norm": 2.6176538467407227,
      "learning_rate": 9.817685553558944e-05,
      "loss": 0.7055,
      "step": 889
    },
    {
      "epoch": 0.1778,
      "loss_align": 0.19503676891326904,
      "loss_contrastive": 0.05021703243255615,
      "loss_gen": 0.31482040882110596,
      "neg_sim_mean": 0.7551802396774292,
      "pos_sim_mean": 0.804963231086731,
      "step": 889
    },
    {
      "epoch": 0.178,
      "grad_norm": 2.589047431945801,
      "learning_rate": 9.816750367586425e-05,
      "loss": 0.5159,
      "step": 890
    },
    {
      "epoch": 0.178,
      "loss_align": 0.2509123682975769,
      "loss_contrastive": 0.057305097579956055,
      "loss_gen": 0.4348580241203308,
      "neg_sim_mean": 0.7063927054405212,
      "pos_sim_mean": 0.7490876317024231,
      "step": 890
    },
    {
      "epoch": 0.1782,
      "grad_norm": 2.4927542209625244,
      "learning_rate": 9.815812833988291e-05,
      "loss": 0.6926,
      "step": 891
    },
    {
      "epoch": 0.1782,
      "loss_align": 0.19079256057739258,
      "loss_contrastive": 0.05030930042266846,
      "loss_gen": 0.16693630814552307,
      "neg_sim_mean": 0.759516716003418,
      "pos_sim_mean": 0.8092074394226074,
      "step": 891
    },
    {
      "epoch": 0.1784,
      "grad_norm": 1.8996330499649048,
      "learning_rate": 9.814872953221485e-05,
      "loss": 0.3638,
      "step": 892
    },
    {
      "epoch": 0.1784,
      "loss_align": 0.395041823387146,
      "loss_contrastive": 0.021591782569885254,
      "loss_gen": 0.013273191638290882,
      "neg_sim_mean": 0.5265499353408813,
      "pos_sim_mean": 0.604958176612854,
      "step": 892
    },
    {
      "epoch": 0.1786,
      "grad_norm": 1.0170646905899048,
      "learning_rate": 9.813930725744094e-05,
      "loss": 0.4109,
      "step": 893
    },
    {
      "epoch": 0.1786,
      "loss_align": 0.35099148750305176,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14654067158699036,
      "neg_sim_mean": 0.5237900018692017,
      "pos_sim_mean": 0.6490085124969482,
      "step": 893
    },
    {
      "epoch": 0.1788,
      "grad_norm": 2.166337013244629,
      "learning_rate": 9.812986152015348e-05,
      "loss": 0.4975,
      "step": 894
    },
    {
      "epoch": 0.1788,
      "loss_align": 0.19735312461853027,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2054246962070465,
      "neg_sim_mean": 0.6215919256210327,
      "pos_sim_mean": 0.8026468753814697,
      "step": 894
    },
    {
      "epoch": 0.179,
      "grad_norm": 1.6263539791107178,
      "learning_rate": 9.81203923249562e-05,
      "loss": 0.4028,
      "step": 895
    },
    {
      "epoch": 0.179,
      "loss_align": 0.3196835517883301,
      "loss_contrastive": 0.005352020263671875,
      "loss_gen": 0.008241917937994003,
      "neg_sim_mean": 0.5856684446334839,
      "pos_sim_mean": 0.6803164482116699,
      "step": 895
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.9726157188415527,
      "learning_rate": 9.811089967646428e-05,
      "loss": 0.3286,
      "step": 896
    },
    {
      "epoch": 0.1792,
      "loss_align": 0.2018193006515503,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3399876654148102,
      "neg_sim_mean": 0.6411837339401245,
      "pos_sim_mean": 0.7981806993484497,
      "step": 896
    },
    {
      "epoch": 0.1794,
      "grad_norm": 2.498446226119995,
      "learning_rate": 9.81013835793043e-05,
      "loss": 0.5418,
      "step": 897
    },
    {
      "epoch": 0.1794,
      "loss_align": 0.3441987633705139,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007585222367197275,
      "neg_sim_mean": 0.44325748085975647,
      "pos_sim_mean": 0.6558012366294861,
      "step": 897
    },
    {
      "epoch": 0.1796,
      "grad_norm": 0.7972870469093323,
      "learning_rate": 9.809184403811431e-05,
      "loss": 0.3518,
      "step": 898
    },
    {
      "epoch": 0.1796,
      "loss_align": 0.45301353931427,
      "loss_contrastive": 0.14455685019493103,
      "loss_gen": 0.4382624626159668,
      "neg_sim_mean": 0.5915433168411255,
      "pos_sim_mean": 0.54698646068573,
      "step": 898
    },
    {
      "epoch": 0.1798,
      "grad_norm": 2.378143787384033,
      "learning_rate": 9.808228105754376e-05,
      "loss": 0.9086,
      "step": 899
    },
    {
      "epoch": 0.1798,
      "loss_align": 0.4257020950317383,
      "loss_contrastive": 0.07584795355796814,
      "loss_gen": 0.3278009295463562,
      "neg_sim_mean": 0.5501458644866943,
      "pos_sim_mean": 0.5742979049682617,
      "step": 899
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.5055713653564453,
      "learning_rate": 9.807269464225355e-05,
      "loss": 0.7626,
      "step": 900
    },
    {
      "epoch": 0.18,
      "loss_align": 0.34358155727386475,
      "loss_contrastive": 0.13785183429718018,
      "loss_gen": 0.6247925162315369,
      "neg_sim_mean": 0.6942702531814575,
      "pos_sim_mean": 0.6564184427261353,
      "step": 900
    },
    {
      "epoch": 0.1802,
      "grad_norm": 3.915189027786255,
      "learning_rate": 9.806308479691595e-05,
      "loss": 0.9849,
      "step": 901
    },
    {
      "epoch": 0.1802,
      "loss_align": 0.3366107940673828,
      "loss_contrastive": 0.08275604248046875,
      "loss_gen": 0.3307775855064392,
      "neg_sim_mean": 0.646145224571228,
      "pos_sim_mean": 0.6633892059326172,
      "step": 901
    },
    {
      "epoch": 0.1804,
      "grad_norm": 2.1551833152770996,
      "learning_rate": 9.80534515262147e-05,
      "loss": 0.6773,
      "step": 902
    },
    {
      "epoch": 0.1804,
      "loss_align": 0.3545852303504944,
      "loss_contrastive": 0.049369215965270996,
      "loss_gen": 0.27083781361579895,
      "neg_sim_mean": 0.5947839617729187,
      "pos_sim_mean": 0.6454147696495056,
      "step": 902
    },
    {
      "epoch": 0.1806,
      "grad_norm": 2.249073028564453,
      "learning_rate": 9.804379483484494e-05,
      "loss": 0.6313,
      "step": 903
    },
    {
      "epoch": 0.1806,
      "loss_align": 0.4511266350746155,
      "loss_contrastive": 0.06027248501777649,
      "loss_gen": 0.0037720161490142345,
      "neg_sim_mean": 0.5091458559036255,
      "pos_sim_mean": 0.5488733649253845,
      "step": 903
    },
    {
      "epoch": 0.1808,
      "grad_norm": 0.6310865879058838,
      "learning_rate": 9.80341147275132e-05,
      "loss": 0.4621,
      "step": 904
    },
    {
      "epoch": 0.1808,
      "loss_align": 0.253578782081604,
      "loss_contrastive": 0.061604440212249756,
      "loss_gen": 0.24747395515441895,
      "neg_sim_mean": 0.7080256342887878,
      "pos_sim_mean": 0.746421217918396,
      "step": 904
    },
    {
      "epoch": 0.181,
      "grad_norm": 2.1150248050689697,
      "learning_rate": 9.802441120893749e-05,
      "loss": 0.5084,
      "step": 905
    },
    {
      "epoch": 0.181,
      "loss_align": 0.5340636372566223,
      "loss_contrastive": 0.22442492842674255,
      "loss_gen": 0.21011123061180115,
      "neg_sim_mean": 0.5903612971305847,
      "pos_sim_mean": 0.4659363627433777,
      "step": 905
    },
    {
      "epoch": 0.1812,
      "grad_norm": 2.1786911487579346,
      "learning_rate": 9.801468428384716e-05,
      "loss": 0.7711,
      "step": 906
    },
    {
      "epoch": 0.1812,
      "loss_align": 0.38482165336608887,
      "loss_contrastive": 0.11249947547912598,
      "loss_gen": 0.32430464029312134,
      "neg_sim_mean": 0.6276777982711792,
      "pos_sim_mean": 0.6151783466339111,
      "step": 906
    },
    {
      "epoch": 0.1814,
      "grad_norm": 2.8118796348571777,
      "learning_rate": 9.8004933956983e-05,
      "loss": 0.7226,
      "step": 907
    },
    {
      "epoch": 0.1814,
      "loss_align": 0.38334691524505615,
      "loss_contrastive": 0.12477624416351318,
      "loss_gen": 0.15813155472278595,
      "neg_sim_mean": 0.6414293050765991,
      "pos_sim_mean": 0.6166530847549438,
      "step": 907
    },
    {
      "epoch": 0.1816,
      "grad_norm": 1.7415874004364014,
      "learning_rate": 9.799516023309719e-05,
      "loss": 0.5565,
      "step": 908
    },
    {
      "epoch": 0.1816,
      "loss_align": 0.3383120894432068,
      "loss_contrastive": 0.09810984134674072,
      "loss_gen": 0.2957366406917572,
      "neg_sim_mean": 0.659797728061676,
      "pos_sim_mean": 0.6616879105567932,
      "step": 908
    },
    {
      "epoch": 0.1818,
      "grad_norm": 2.4961020946502686,
      "learning_rate": 9.798536311695334e-05,
      "loss": 0.6458,
      "step": 909
    },
    {
      "epoch": 0.1818,
      "loss_align": 0.35832154750823975,
      "loss_contrastive": 0.0,
      "loss_gen": 0.35753926634788513,
      "neg_sim_mean": 0.5336954593658447,
      "pos_sim_mean": 0.6416784524917603,
      "step": 909
    },
    {
      "epoch": 0.182,
      "grad_norm": 2.5443859100341797,
      "learning_rate": 9.797554261332645e-05,
      "loss": 0.7159,
      "step": 910
    },
    {
      "epoch": 0.182,
      "loss_align": 0.3173929452896118,
      "loss_contrastive": 0.09695184230804443,
      "loss_gen": 0.21693715453147888,
      "neg_sim_mean": 0.6795588731765747,
      "pos_sim_mean": 0.6826070547103882,
      "step": 910
    },
    {
      "epoch": 0.1822,
      "grad_norm": 1.875941514968872,
      "learning_rate": 9.796569872700288e-05,
      "loss": 0.546,
      "step": 911
    },
    {
      "epoch": 0.1822,
      "loss_align": 0.33427155017852783,
      "loss_contrastive": 0.17595374584197998,
      "loss_gen": 0.5228172540664673,
      "neg_sim_mean": 0.7416821718215942,
      "pos_sim_mean": 0.6657284498214722,
      "step": 911
    },
    {
      "epoch": 0.1824,
      "grad_norm": 4.642751693725586,
      "learning_rate": 9.795583146278046e-05,
      "loss": 0.8782,
      "step": 912
    },
    {
      "epoch": 0.1824,
      "loss_align": 0.4755624532699585,
      "loss_contrastive": 0.10904762148857117,
      "loss_gen": 0.3407049775123596,
      "neg_sim_mean": 0.5334851741790771,
      "pos_sim_mean": 0.5244375467300415,
      "step": 912
    },
    {
      "epoch": 0.1826,
      "grad_norm": 2.76220440864563,
      "learning_rate": 9.794594082546835e-05,
      "loss": 0.8294,
      "step": 913
    },
    {
      "epoch": 0.1826,
      "loss_align": 0.32643330097198486,
      "loss_contrastive": 0.1401458978652954,
      "loss_gen": 0.315660685300827,
      "neg_sim_mean": 0.7137125730514526,
      "pos_sim_mean": 0.6735666990280151,
      "step": 913
    },
    {
      "epoch": 0.1828,
      "grad_norm": 2.4877822399139404,
      "learning_rate": 9.793602681988714e-05,
      "loss": 0.6589,
      "step": 914
    },
    {
      "epoch": 0.1828,
      "loss_align": 0.2783929109573364,
      "loss_contrastive": 0.0751839280128479,
      "loss_gen": 0.3504835069179535,
      "neg_sim_mean": 0.6967909932136536,
      "pos_sim_mean": 0.7216070890426636,
      "step": 914
    },
    {
      "epoch": 0.183,
      "grad_norm": 2.3919429779052734,
      "learning_rate": 9.79260894508688e-05,
      "loss": 0.6379,
      "step": 915
    },
    {
      "epoch": 0.183,
      "loss_align": 0.486777126789093,
      "loss_contrastive": 0.1712615191936493,
      "loss_gen": 0.09535960853099823,
      "neg_sim_mean": 0.5844843983650208,
      "pos_sim_mean": 0.513222873210907,
      "step": 915
    },
    {
      "epoch": 0.1832,
      "grad_norm": 1.325232982635498,
      "learning_rate": 9.791612872325667e-05,
      "loss": 0.6027,
      "step": 916
    },
    {
      "epoch": 0.1832,
      "loss_align": 0.3565281629562378,
      "loss_contrastive": 0.024611234664916992,
      "loss_gen": 0.24705570936203003,
      "neg_sim_mean": 0.5680830478668213,
      "pos_sim_mean": 0.6434718370437622,
      "step": 916
    },
    {
      "epoch": 0.1834,
      "grad_norm": 2.269813299179077,
      "learning_rate": 9.790614464190549e-05,
      "loss": 0.6065,
      "step": 917
    },
    {
      "epoch": 0.1834,
      "loss_align": 0.32899582386016846,
      "loss_contrastive": 0.09926092624664307,
      "loss_gen": 0.35992875695228577,
      "neg_sim_mean": 0.6702650785446167,
      "pos_sim_mean": 0.6710041761398315,
      "step": 917
    },
    {
      "epoch": 0.1836,
      "grad_norm": 2.5376737117767334,
      "learning_rate": 9.789613721168139e-05,
      "loss": 0.7008,
      "step": 918
    },
    {
      "epoch": 0.1836,
      "loss_align": 0.3670032024383545,
      "loss_contrastive": 0.06503903865814209,
      "loss_gen": 0.2939162254333496,
      "neg_sim_mean": 0.5980358123779297,
      "pos_sim_mean": 0.6329967975616455,
      "step": 918
    },
    {
      "epoch": 0.1838,
      "grad_norm": 2.8354151248931885,
      "learning_rate": 9.788610643746184e-05,
      "loss": 0.6687,
      "step": 919
    },
    {
      "epoch": 0.1838,
      "loss_align": 0.26310765743255615,
      "loss_contrastive": 0.08138275146484375,
      "loss_gen": 0.14690029621124268,
      "neg_sim_mean": 0.7182750701904297,
      "pos_sim_mean": 0.7368923425674438,
      "step": 919
    },
    {
      "epoch": 0.184,
      "grad_norm": 1.7721166610717773,
      "learning_rate": 9.787605232413574e-05,
      "loss": 0.4198,
      "step": 920
    },
    {
      "epoch": 0.184,
      "loss_align": 0.31534063816070557,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0317830853164196,
      "neg_sim_mean": 0.5370048880577087,
      "pos_sim_mean": 0.6846593618392944,
      "step": 920
    },
    {
      "epoch": 0.1842,
      "grad_norm": 1.44808828830719,
      "learning_rate": 9.786597487660337e-05,
      "loss": 0.3471,
      "step": 921
    },
    {
      "epoch": 0.1842,
      "loss_align": 0.2281663417816162,
      "loss_contrastive": 0.09142845869064331,
      "loss_gen": 0.24596330523490906,
      "neg_sim_mean": 0.7632620930671692,
      "pos_sim_mean": 0.7718336582183838,
      "step": 921
    },
    {
      "epoch": 0.1844,
      "grad_norm": 2.0251834392547607,
      "learning_rate": 9.785587409977632e-05,
      "loss": 0.4851,
      "step": 922
    },
    {
      "epoch": 0.1844,
      "loss_align": 0.3323215842247009,
      "loss_contrastive": 0.11734646558761597,
      "loss_gen": 0.31623101234436035,
      "neg_sim_mean": 0.6850248575210571,
      "pos_sim_mean": 0.6676784157752991,
      "step": 922
    },
    {
      "epoch": 0.1846,
      "grad_norm": 2.3379528522491455,
      "learning_rate": 9.784574999857757e-05,
      "loss": 0.6626,
      "step": 923
    },
    {
      "epoch": 0.1846,
      "loss_align": 0.3401716947555542,
      "loss_contrastive": 0.12492382526397705,
      "loss_gen": 0.497518926858902,
      "neg_sim_mean": 0.6847521066665649,
      "pos_sim_mean": 0.6598283052444458,
      "step": 923
    },
    {
      "epoch": 0.1848,
      "grad_norm": 3.244202136993408,
      "learning_rate": 9.783560257794154e-05,
      "loss": 0.8527,
      "step": 924
    },
    {
      "epoch": 0.1848,
      "loss_align": 0.2814016342163086,
      "loss_contrastive": 0.09088289737701416,
      "loss_gen": 0.33393189311027527,
      "neg_sim_mean": 0.7094812393188477,
      "pos_sim_mean": 0.7185983657836914,
      "step": 924
    },
    {
      "epoch": 0.185,
      "grad_norm": 2.5662364959716797,
      "learning_rate": 9.78254318428139e-05,
      "loss": 0.6262,
      "step": 925
    },
    {
      "epoch": 0.185,
      "loss_align": 0.34187477827072144,
      "loss_contrastive": 0.20478862524032593,
      "loss_gen": 0.27333521842956543,
      "neg_sim_mean": 0.7629138231277466,
      "pos_sim_mean": 0.6581252217292786,
      "step": 925
    },
    {
      "epoch": 0.1852,
      "grad_norm": 2.4246950149536133,
      "learning_rate": 9.781523779815179e-05,
      "loss": 0.6398,
      "step": 926
    },
    {
      "epoch": 0.1852,
      "loss_align": 0.22614681720733643,
      "loss_contrastive": 0.11010372638702393,
      "loss_gen": 0.5355212092399597,
      "neg_sim_mean": 0.7839568853378296,
      "pos_sim_mean": 0.7738531827926636,
      "step": 926
    },
    {
      "epoch": 0.1854,
      "grad_norm": 3.254915475845337,
      "learning_rate": 9.780502044892362e-05,
      "loss": 0.7749,
      "step": 927
    },
    {
      "epoch": 0.1854,
      "loss_align": 0.40454524755477905,
      "loss_contrastive": 0.1889292299747467,
      "loss_gen": 0.39104217290878296,
      "neg_sim_mean": 0.6843839883804321,
      "pos_sim_mean": 0.595454752445221,
      "step": 927
    },
    {
      "epoch": 0.1856,
      "grad_norm": 3.350266218185425,
      "learning_rate": 9.779477980010924e-05,
      "loss": 0.8183,
      "step": 928
    },
    {
      "epoch": 0.1856,
      "loss_align": 0.141282856464386,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2560200095176697,
      "neg_sim_mean": 0.7112950086593628,
      "pos_sim_mean": 0.858717143535614,
      "step": 928
    },
    {
      "epoch": 0.1858,
      "grad_norm": 2.0439627170562744,
      "learning_rate": 9.778451585669982e-05,
      "loss": 0.3973,
      "step": 929
    },
    {
      "epoch": 0.1858,
      "loss_align": 0.2689533233642578,
      "loss_contrastive": 0.07333743572235107,
      "loss_gen": 0.3634451925754547,
      "neg_sim_mean": 0.7043840885162354,
      "pos_sim_mean": 0.7310466766357422,
      "step": 929
    },
    {
      "epoch": 0.186,
      "grad_norm": 2.381758213043213,
      "learning_rate": 9.777422862369783e-05,
      "loss": 0.6412,
      "step": 930
    },
    {
      "epoch": 0.186,
      "loss_align": 0.28113359212875366,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03197580203413963,
      "neg_sim_mean": 0.43901216983795166,
      "pos_sim_mean": 0.7188664078712463,
      "step": 930
    },
    {
      "epoch": 0.1862,
      "grad_norm": 1.381705403327942,
      "learning_rate": 9.776391810611718e-05,
      "loss": 0.3131,
      "step": 931
    },
    {
      "epoch": 0.1862,
      "loss_align": 0.39008206129074097,
      "loss_contrastive": 0.12547969818115234,
      "loss_gen": 0.23152700066566467,
      "neg_sim_mean": 0.6353976130485535,
      "pos_sim_mean": 0.609917938709259,
      "step": 931
    },
    {
      "epoch": 0.1864,
      "grad_norm": 2.3064138889312744,
      "learning_rate": 9.775358430898311e-05,
      "loss": 0.6367,
      "step": 932
    },
    {
      "epoch": 0.1864,
      "loss_align": 0.2655559778213501,
      "loss_contrastive": 0.041671693325042725,
      "loss_gen": 0.4152446985244751,
      "neg_sim_mean": 0.6761156916618347,
      "pos_sim_mean": 0.7344440221786499,
      "step": 932
    },
    {
      "epoch": 0.1866,
      "grad_norm": 3.16756272315979,
      "learning_rate": 9.774322723733216e-05,
      "loss": 0.6858,
      "step": 933
    },
    {
      "epoch": 0.1866,
      "loss_align": 0.26114344596862793,
      "loss_contrastive": 0.06506204605102539,
      "loss_gen": 0.24087616801261902,
      "neg_sim_mean": 0.7039185762405396,
      "pos_sim_mean": 0.7388565540313721,
      "step": 933
    },
    {
      "epoch": 0.1868,
      "grad_norm": 1.983178734779358,
      "learning_rate": 9.773284689621222e-05,
      "loss": 0.5098,
      "step": 934
    },
    {
      "epoch": 0.1868,
      "loss_align": 0.3978651762008667,
      "loss_contrastive": 0.0,
      "loss_gen": 0.027086615562438965,
      "neg_sim_mean": 0.40627896785736084,
      "pos_sim_mean": 0.6021348237991333,
      "step": 934
    },
    {
      "epoch": 0.187,
      "grad_norm": 1.6771299839019775,
      "learning_rate": 9.772244329068261e-05,
      "loss": 0.425,
      "step": 935
    },
    {
      "epoch": 0.187,
      "loss_align": 0.27740931510925293,
      "loss_contrastive": 0.170038640499115,
      "loss_gen": 0.238499715924263,
      "neg_sim_mean": 0.7926293015480042,
      "pos_sim_mean": 0.7225906848907471,
      "step": 935
    },
    {
      "epoch": 0.1872,
      "grad_norm": 2.193420648574829,
      "learning_rate": 9.771201642581385e-05,
      "loss": 0.5363,
      "step": 936
    },
    {
      "epoch": 0.1872,
      "loss_align": 0.4913354516029358,
      "loss_contrastive": 0.19941309094429016,
      "loss_gen": 0.33544084429740906,
      "neg_sim_mean": 0.6080776453018188,
      "pos_sim_mean": 0.5086645483970642,
      "step": 936
    },
    {
      "epoch": 0.1874,
      "grad_norm": 2.704489231109619,
      "learning_rate": 9.77015663066879e-05,
      "loss": 0.8507,
      "step": 937
    },
    {
      "epoch": 0.1874,
      "loss_align": 0.30457818508148193,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27154913544654846,
      "neg_sim_mean": 0.5952380895614624,
      "pos_sim_mean": 0.6954218149185181,
      "step": 937
    },
    {
      "epoch": 0.1876,
      "grad_norm": 2.627809762954712,
      "learning_rate": 9.769109293839802e-05,
      "loss": 0.5761,
      "step": 938
    },
    {
      "epoch": 0.1876,
      "loss_align": 0.26739561557769775,
      "loss_contrastive": 0.10147953033447266,
      "loss_gen": 0.23336432874202728,
      "neg_sim_mean": 0.734083890914917,
      "pos_sim_mean": 0.7326043844223022,
      "step": 938
    },
    {
      "epoch": 0.1878,
      "grad_norm": 2.0571978092193604,
      "learning_rate": 9.76805963260488e-05,
      "loss": 0.5129,
      "step": 939
    },
    {
      "epoch": 0.1878,
      "loss_align": 0.2812669277191162,
      "loss_contrastive": 0.11841696500778198,
      "loss_gen": 0.20785576105117798,
      "neg_sim_mean": 0.7371500134468079,
      "pos_sim_mean": 0.7187330722808838,
      "step": 939
    },
    {
      "epoch": 0.188,
      "grad_norm": 2.525144577026367,
      "learning_rate": 9.767007647475617e-05,
      "loss": 0.5033,
      "step": 940
    },
    {
      "epoch": 0.188,
      "loss_align": 0.43451887369155884,
      "loss_contrastive": 0.1706579029560089,
      "loss_gen": 0.2093353271484375,
      "neg_sim_mean": 0.6361390352249146,
      "pos_sim_mean": 0.5654811263084412,
      "step": 940
    },
    {
      "epoch": 0.1882,
      "grad_norm": 2.019552707672119,
      "learning_rate": 9.765953338964735e-05,
      "loss": 0.6643,
      "step": 941
    },
    {
      "epoch": 0.1882,
      "loss_align": 0.32118475437164307,
      "loss_contrastive": 0.023932039737701416,
      "loss_gen": 0.31543073058128357,
      "neg_sim_mean": 0.6027472615242004,
      "pos_sim_mean": 0.6788152456283569,
      "step": 941
    },
    {
      "epoch": 0.1884,
      "grad_norm": 2.095071315765381,
      "learning_rate": 9.764896707586096e-05,
      "loss": 0.6395,
      "step": 942
    },
    {
      "epoch": 0.1884,
      "loss_align": 0.25607240200042725,
      "loss_contrastive": 0.1814512014389038,
      "loss_gen": 0.32966339588165283,
      "neg_sim_mean": 0.8253787755966187,
      "pos_sim_mean": 0.7439275979995728,
      "step": 942
    },
    {
      "epoch": 0.1886,
      "grad_norm": 2.068887948989868,
      "learning_rate": 9.763837753854683e-05,
      "loss": 0.6075,
      "step": 943
    },
    {
      "epoch": 0.1886,
      "loss_align": 0.4612598419189453,
      "loss_contrastive": 0.16613826155662537,
      "loss_gen": 0.2956877052783966,
      "neg_sim_mean": 0.6048784255981445,
      "pos_sim_mean": 0.5387401580810547,
      "step": 943
    },
    {
      "epoch": 0.1888,
      "grad_norm": 2.2713797092437744,
      "learning_rate": 9.762776478286622e-05,
      "loss": 0.7769,
      "step": 944
    },
    {
      "epoch": 0.1888,
      "loss_align": 0.44079047441482544,
      "loss_contrastive": 0.1507945954799652,
      "loss_gen": 0.32157406210899353,
      "neg_sim_mean": 0.6100041270256042,
      "pos_sim_mean": 0.5592095255851746,
      "step": 944
    },
    {
      "epoch": 0.189,
      "grad_norm": 2.8855643272399902,
      "learning_rate": 9.761712881399164e-05,
      "loss": 0.7805,
      "step": 945
    },
    {
      "epoch": 0.189,
      "loss_align": 0.3488141894340515,
      "loss_contrastive": 0.123776376247406,
      "loss_gen": 0.5525150299072266,
      "neg_sim_mean": 0.6749621629714966,
      "pos_sim_mean": 0.6511858105659485,
      "step": 945
    },
    {
      "epoch": 0.1892,
      "grad_norm": 2.9072415828704834,
      "learning_rate": 9.760646963710694e-05,
      "loss": 0.9162,
      "step": 946
    },
    {
      "epoch": 0.1892,
      "loss_align": 0.3523114323616028,
      "loss_contrastive": 0.19569438695907593,
      "loss_gen": 0.22032620012760162,
      "neg_sim_mean": 0.7433829307556152,
      "pos_sim_mean": 0.6476885676383972,
      "step": 946
    },
    {
      "epoch": 0.1894,
      "grad_norm": 1.8336920738220215,
      "learning_rate": 9.759578725740727e-05,
      "loss": 0.5961,
      "step": 947
    },
    {
      "epoch": 0.1894,
      "loss_align": 0.36985641717910767,
      "loss_contrastive": 0.03734630346298218,
      "loss_gen": 0.31564861536026,
      "neg_sim_mean": 0.5674898624420166,
      "pos_sim_mean": 0.6301435828208923,
      "step": 947
    },
    {
      "epoch": 0.1896,
      "grad_norm": 2.126828193664551,
      "learning_rate": 9.75850816800991e-05,
      "loss": 0.69,
      "step": 948
    },
    {
      "epoch": 0.1896,
      "loss_align": 0.3257027268409729,
      "loss_contrastive": 0.14391475915908813,
      "loss_gen": 0.46968355774879456,
      "neg_sim_mean": 0.7182120084762573,
      "pos_sim_mean": 0.6742972731590271,
      "step": 948
    },
    {
      "epoch": 0.1898,
      "grad_norm": 2.8265979290008545,
      "learning_rate": 9.757435291040016e-05,
      "loss": 0.8127,
      "step": 949
    },
    {
      "epoch": 0.1898,
      "loss_align": 0.30205631256103516,
      "loss_contrastive": 0.10405242443084717,
      "loss_gen": 0.2506158649921417,
      "neg_sim_mean": 0.7019960880279541,
      "pos_sim_mean": 0.6979436874389648,
      "step": 949
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.8618170022964478,
      "learning_rate": 9.756360095353956e-05,
      "loss": 0.5652,
      "step": 950
    },
    {
      "epoch": 0.19,
      "loss_align": 0.4024907350540161,
      "loss_contrastive": 0.18640843033790588,
      "loss_gen": 0.3312056064605713,
      "neg_sim_mean": 0.6839177012443542,
      "pos_sim_mean": 0.5975092649459839,
      "step": 950
    },
    {
      "epoch": 0.1902,
      "grad_norm": 2.744610071182251,
      "learning_rate": 9.755282581475769e-05,
      "loss": 0.7561,
      "step": 951
    },
    {
      "epoch": 0.1902,
      "loss_align": 0.4318023920059204,
      "loss_contrastive": 0.10793808102607727,
      "loss_gen": 0.38174545764923096,
      "neg_sim_mean": 0.5761356949806213,
      "pos_sim_mean": 0.5681976079940796,
      "step": 951
    },
    {
      "epoch": 0.1904,
      "grad_norm": 2.7758727073669434,
      "learning_rate": 9.754202749930616e-05,
      "loss": 0.8265,
      "step": 952
    },
    {
      "epoch": 0.1904,
      "loss_align": 0.4340564012527466,
      "loss_contrastive": 0.09396454691886902,
      "loss_gen": 0.2778797149658203,
      "neg_sim_mean": 0.5599081516265869,
      "pos_sim_mean": 0.5659435987472534,
      "step": 952
    },
    {
      "epoch": 0.1906,
      "grad_norm": 2.3001210689544678,
      "learning_rate": 9.753120601244801e-05,
      "loss": 0.7232,
      "step": 953
    },
    {
      "epoch": 0.1906,
      "loss_align": 0.4326409101486206,
      "loss_contrastive": 0.24644574522972107,
      "loss_gen": 0.3735186755657196,
      "neg_sim_mean": 0.7138048410415649,
      "pos_sim_mean": 0.5673590898513794,
      "step": 953
    },
    {
      "epoch": 0.1908,
      "grad_norm": 3.07806396484375,
      "learning_rate": 9.752036135945744e-05,
      "loss": 0.8357,
      "step": 954
    },
    {
      "epoch": 0.1908,
      "loss_align": 0.37265896797180176,
      "loss_contrastive": 0.0765988826751709,
      "loss_gen": 0.2499469518661499,
      "neg_sim_mean": 0.6039398908615112,
      "pos_sim_mean": 0.6273410320281982,
      "step": 954
    },
    {
      "epoch": 0.191,
      "grad_norm": 2.365572452545166,
      "learning_rate": 9.750949354562004e-05,
      "loss": 0.6318,
      "step": 955
    },
    {
      "epoch": 0.191,
      "loss_align": 0.25524359941482544,
      "loss_contrastive": 0.0055925846099853516,
      "loss_gen": 0.22718645632266998,
      "neg_sim_mean": 0.650348961353302,
      "pos_sim_mean": 0.7447564005851746,
      "step": 955
    },
    {
      "epoch": 0.1912,
      "grad_norm": 2.013307809829712,
      "learning_rate": 9.749860257623263e-05,
      "loss": 0.4831,
      "step": 956
    },
    {
      "epoch": 0.1912,
      "loss_align": 0.4423692226409912,
      "loss_contrastive": 0.16556581854820251,
      "loss_gen": 0.436479777097702,
      "neg_sim_mean": 0.6231966018676758,
      "pos_sim_mean": 0.5576307773590088,
      "step": 956
    },
    {
      "epoch": 0.1914,
      "grad_norm": 2.3088772296905518,
      "learning_rate": 9.748768845660334e-05,
      "loss": 0.8987,
      "step": 957
    },
    {
      "epoch": 0.1914,
      "loss_align": 0.34037983417510986,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01707773655653,
      "neg_sim_mean": 0.42119282484054565,
      "pos_sim_mean": 0.6596201658248901,
      "step": 957
    },
    {
      "epoch": 0.1916,
      "grad_norm": 1.0668606758117676,
      "learning_rate": 9.74767511920516e-05,
      "loss": 0.3575,
      "step": 958
    },
    {
      "epoch": 0.1916,
      "loss_align": 0.3266538977622986,
      "loss_contrastive": 0.1242331862449646,
      "loss_gen": 0.3649652600288391,
      "neg_sim_mean": 0.6975792646408081,
      "pos_sim_mean": 0.6733461022377014,
      "step": 958
    },
    {
      "epoch": 0.1918,
      "grad_norm": 2.2648584842681885,
      "learning_rate": 9.746579078790807e-05,
      "loss": 0.7065,
      "step": 959
    },
    {
      "epoch": 0.1918,
      "loss_align": 0.28766006231307983,
      "loss_contrastive": 0.08289813995361328,
      "loss_gen": 0.3435540795326233,
      "neg_sim_mean": 0.6952380537986755,
      "pos_sim_mean": 0.7123399376869202,
      "step": 959
    },
    {
      "epoch": 0.192,
      "grad_norm": 2.0672383308410645,
      "learning_rate": 9.745480724951473e-05,
      "loss": 0.6412,
      "step": 960
    },
    {
      "epoch": 0.192,
      "loss_align": 0.4580222964286804,
      "loss_contrastive": 0.11023136973381042,
      "loss_gen": 0.18820537626743317,
      "neg_sim_mean": 0.5522090792655945,
      "pos_sim_mean": 0.5419777035713196,
      "step": 960
    },
    {
      "epoch": 0.1922,
      "grad_norm": 2.085536241531372,
      "learning_rate": 9.744380058222483e-05,
      "loss": 0.6595,
      "step": 961
    },
    {
      "epoch": 0.1922,
      "loss_align": 0.44460636377334595,
      "loss_contrastive": 0.27228978276252747,
      "loss_gen": 0.529231607913971,
      "neg_sim_mean": 0.727683424949646,
      "pos_sim_mean": 0.555393636226654,
      "step": 961
    },
    {
      "epoch": 0.1924,
      "grad_norm": 2.588758945465088,
      "learning_rate": 9.743277079140287e-05,
      "loss": 1.0065,
      "step": 962
    },
    {
      "epoch": 0.1924,
      "loss_align": 0.3027612566947937,
      "loss_contrastive": 0.09636563062667847,
      "loss_gen": 0.35818690061569214,
      "neg_sim_mean": 0.6936043500900269,
      "pos_sim_mean": 0.6972387433052063,
      "step": 962
    },
    {
      "epoch": 0.1926,
      "grad_norm": 2.3563311100006104,
      "learning_rate": 9.742171788242466e-05,
      "loss": 0.6725,
      "step": 963
    },
    {
      "epoch": 0.1926,
      "loss_align": 0.42063313722610474,
      "loss_contrastive": 0.07897254824638367,
      "loss_gen": 0.24519877135753632,
      "neg_sim_mean": 0.5583394169807434,
      "pos_sim_mean": 0.5793668627738953,
      "step": 963
    },
    {
      "epoch": 0.1928,
      "grad_norm": 2.241769790649414,
      "learning_rate": 9.741064186067723e-05,
      "loss": 0.6753,
      "step": 964
    },
    {
      "epoch": 0.1928,
      "loss_align": 0.42649388313293457,
      "loss_contrastive": 0.17303547263145447,
      "loss_gen": 0.4465709328651428,
      "neg_sim_mean": 0.6465415954589844,
      "pos_sim_mean": 0.5735061168670654,
      "step": 964
    },
    {
      "epoch": 0.193,
      "grad_norm": 2.3848464488983154,
      "learning_rate": 9.739954273155891e-05,
      "loss": 0.8938,
      "step": 965
    },
    {
      "epoch": 0.193,
      "loss_align": 0.3941345810890198,
      "loss_contrastive": 0.06841105222702026,
      "loss_gen": 0.029474487528204918,
      "neg_sim_mean": 0.5742764472961426,
      "pos_sim_mean": 0.6058654189109802,
      "step": 965
    },
    {
      "epoch": 0.1932,
      "grad_norm": 1.2948051691055298,
      "learning_rate": 9.73884205004793e-05,
      "loss": 0.4318,
      "step": 966
    },
    {
      "epoch": 0.1932,
      "loss_align": 0.3789750337600708,
      "loss_contrastive": 0.11516362428665161,
      "loss_gen": 0.18467938899993896,
      "neg_sim_mean": 0.6361885666847229,
      "pos_sim_mean": 0.6210249662399292,
      "step": 966
    },
    {
      "epoch": 0.1934,
      "grad_norm": 1.6665513515472412,
      "learning_rate": 9.737727517285919e-05,
      "loss": 0.5775,
      "step": 967
    },
    {
      "epoch": 0.1934,
      "loss_align": 0.38765913248062134,
      "loss_contrastive": 0.09731811285018921,
      "loss_gen": 0.5179067850112915,
      "neg_sim_mean": 0.60965895652771,
      "pos_sim_mean": 0.6123408675193787,
      "step": 967
    },
    {
      "epoch": 0.1936,
      "grad_norm": 2.682548999786377,
      "learning_rate": 9.736610675413071e-05,
      "loss": 0.9172,
      "step": 968
    },
    {
      "epoch": 0.1936,
      "loss_align": 0.371778666973114,
      "loss_contrastive": 0.21680611371994019,
      "loss_gen": 0.2764454782009125,
      "neg_sim_mean": 0.7450274229049683,
      "pos_sim_mean": 0.628221333026886,
      "step": 968
    },
    {
      "epoch": 0.1938,
      "grad_norm": 2.577968120574951,
      "learning_rate": 9.735491524973722e-05,
      "loss": 0.6742,
      "step": 969
    },
    {
      "epoch": 0.1938,
      "loss_align": 0.3358204960823059,
      "loss_contrastive": 0.05780684947967529,
      "loss_gen": 0.33331555128097534,
      "neg_sim_mean": 0.6219863295555115,
      "pos_sim_mean": 0.6641795039176941,
      "step": 969
    },
    {
      "epoch": 0.194,
      "grad_norm": 1.9925694465637207,
      "learning_rate": 9.73437006651333e-05,
      "loss": 0.6761,
      "step": 970
    },
    {
      "epoch": 0.194,
      "loss_align": 0.4457436203956604,
      "loss_contrastive": 0.13556018471717834,
      "loss_gen": 0.33875319361686707,
      "neg_sim_mean": 0.5898165702819824,
      "pos_sim_mean": 0.5542563796043396,
      "step": 970
    },
    {
      "epoch": 0.1942,
      "grad_norm": 2.6791770458221436,
      "learning_rate": 9.733246300578483e-05,
      "loss": 0.8008,
      "step": 971
    },
    {
      "epoch": 0.1942,
      "loss_align": 0.25693535804748535,
      "loss_contrastive": 0.06941819190979004,
      "loss_gen": 0.21036751568317413,
      "neg_sim_mean": 0.7124828100204468,
      "pos_sim_mean": 0.7430646419525146,
      "step": 971
    },
    {
      "epoch": 0.1944,
      "grad_norm": 1.852287769317627,
      "learning_rate": 9.732120227716888e-05,
      "loss": 0.4756,
      "step": 972
    },
    {
      "epoch": 0.1944,
      "loss_align": 0.35374462604522705,
      "loss_contrastive": 0.09059059619903564,
      "loss_gen": 0.25844407081604004,
      "neg_sim_mean": 0.6368459463119507,
      "pos_sim_mean": 0.646255373954773,
      "step": 972
    },
    {
      "epoch": 0.1946,
      "grad_norm": 2.1529593467712402,
      "learning_rate": 9.730991848477379e-05,
      "loss": 0.6231,
      "step": 973
    },
    {
      "epoch": 0.1946,
      "loss_align": 0.3853747248649597,
      "loss_contrastive": 0.0777772068977356,
      "loss_gen": 0.3228794038295746,
      "neg_sim_mean": 0.592402458190918,
      "pos_sim_mean": 0.6146252751350403,
      "step": 973
    },
    {
      "epoch": 0.1948,
      "grad_norm": 2.6168107986450195,
      "learning_rate": 9.729861163409919e-05,
      "loss": 0.7176,
      "step": 974
    },
    {
      "epoch": 0.1948,
      "loss_align": 0.35084402561187744,
      "loss_contrastive": 0.11471617221832275,
      "loss_gen": 0.21726511418819427,
      "neg_sim_mean": 0.6638721227645874,
      "pos_sim_mean": 0.6491559743881226,
      "step": 974
    },
    {
      "epoch": 0.195,
      "grad_norm": 2.341388463973999,
      "learning_rate": 9.728728173065585e-05,
      "loss": 0.5819,
      "step": 975
    },
    {
      "epoch": 0.195,
      "loss_align": 0.3091347813606262,
      "loss_contrastive": 0.044569432735443115,
      "loss_gen": 0.2972387969493866,
      "neg_sim_mean": 0.635434627532959,
      "pos_sim_mean": 0.6908652186393738,
      "step": 975
    },
    {
      "epoch": 0.1952,
      "grad_norm": 2.2907960414886475,
      "learning_rate": 9.727592877996585e-05,
      "loss": 0.6117,
      "step": 976
    },
    {
      "epoch": 0.1952,
      "loss_align": 0.22896987199783325,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0038066108245402575,
      "neg_sim_mean": 0.40857306122779846,
      "pos_sim_mean": 0.7710301280021667,
      "step": 976
    },
    {
      "epoch": 0.1954,
      "grad_norm": 0.46951302886009216,
      "learning_rate": 9.726455278756248e-05,
      "loss": 0.2328,
      "step": 977
    },
    {
      "epoch": 0.1954,
      "loss_align": 0.38410484790802,
      "loss_contrastive": 0.01327204704284668,
      "loss_gen": 0.29705268144607544,
      "neg_sim_mean": 0.5291671752929688,
      "pos_sim_mean": 0.61589515209198,
      "step": 977
    },
    {
      "epoch": 0.1956,
      "grad_norm": 2.0849945545196533,
      "learning_rate": 9.725315375899024e-05,
      "loss": 0.6828,
      "step": 978
    },
    {
      "epoch": 0.1956,
      "loss_align": 0.2595188617706299,
      "loss_contrastive": 0.12114298343658447,
      "loss_gen": 0.4535013735294342,
      "neg_sim_mean": 0.7616240978240967,
      "pos_sim_mean": 0.7404811382293701,
      "step": 978
    },
    {
      "epoch": 0.1958,
      "grad_norm": 3.385451555252075,
      "learning_rate": 9.724173169980491e-05,
      "loss": 0.7276,
      "step": 979
    },
    {
      "epoch": 0.1958,
      "loss_align": 0.33725595474243164,
      "loss_contrastive": 0.06909346580505371,
      "loss_gen": 0.610077440738678,
      "neg_sim_mean": 0.6318374872207642,
      "pos_sim_mean": 0.6627440452575684,
      "step": 979
    },
    {
      "epoch": 0.196,
      "grad_norm": 2.693323850631714,
      "learning_rate": 9.723028661557345e-05,
      "loss": 0.9556,
      "step": 980
    },
    {
      "epoch": 0.196,
      "loss_align": 0.43730342388153076,
      "loss_contrastive": 0.15536817908287048,
      "loss_gen": 0.3097597360610962,
      "neg_sim_mean": 0.6180647611618042,
      "pos_sim_mean": 0.5626965761184692,
      "step": 980
    },
    {
      "epoch": 0.1962,
      "grad_norm": 2.5746943950653076,
      "learning_rate": 9.721881851187406e-05,
      "loss": 0.7657,
      "step": 981
    },
    {
      "epoch": 0.1962,
      "loss_align": 0.29538649320602417,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005041954573243856,
      "neg_sim_mean": 0.45655524730682373,
      "pos_sim_mean": 0.7046135067939758,
      "step": 981
    },
    {
      "epoch": 0.1964,
      "grad_norm": 0.49245360493659973,
      "learning_rate": 9.720732739429614e-05,
      "loss": 0.3004,
      "step": 982
    },
    {
      "epoch": 0.1964,
      "loss_align": 0.3770679235458374,
      "loss_contrastive": 0.1233665943145752,
      "loss_gen": 0.36263954639434814,
      "neg_sim_mean": 0.6462986469268799,
      "pos_sim_mean": 0.6229320764541626,
      "step": 982
    },
    {
      "epoch": 0.1966,
      "grad_norm": 2.865204334259033,
      "learning_rate": 9.719581326844032e-05,
      "loss": 0.7545,
      "step": 983
    },
    {
      "epoch": 0.1966,
      "loss_align": 0.5920677185058594,
      "loss_contrastive": 0.4437357783317566,
      "loss_gen": 0.2400614470243454,
      "neg_sim_mean": 0.7516680955886841,
      "pos_sim_mean": 0.407932311296463,
      "step": 983
    },
    {
      "epoch": 0.1968,
      "grad_norm": 2.577491044998169,
      "learning_rate": 9.718427613991848e-05,
      "loss": 0.8854,
      "step": 984
    },
    {
      "epoch": 0.1968,
      "loss_align": 0.2520977258682251,
      "loss_contrastive": 0.10878181457519531,
      "loss_gen": 0.1734030395746231,
      "neg_sim_mean": 0.7566840648651123,
      "pos_sim_mean": 0.7479022741317749,
      "step": 984
    },
    {
      "epoch": 0.197,
      "grad_norm": 2.147550344467163,
      "learning_rate": 9.717271601435363e-05,
      "loss": 0.4386,
      "step": 985
    },
    {
      "epoch": 0.197,
      "loss_align": 0.3854717016220093,
      "loss_contrastive": 0.18840360641479492,
      "loss_gen": 0.3716434836387634,
      "neg_sim_mean": 0.7029318809509277,
      "pos_sim_mean": 0.6145282983779907,
      "step": 985
    },
    {
      "epoch": 0.1972,
      "grad_norm": 2.5639853477478027,
      "learning_rate": 9.716113289738004e-05,
      "loss": 0.7797,
      "step": 986
    },
    {
      "epoch": 0.1972,
      "loss_align": 0.37148112058639526,
      "loss_contrastive": 0.0025527477264404297,
      "loss_gen": 0.5035820007324219,
      "neg_sim_mean": 0.5310716032981873,
      "pos_sim_mean": 0.6285188794136047,
      "step": 986
    },
    {
      "epoch": 0.1974,
      "grad_norm": 3.5728931427001953,
      "learning_rate": 9.714952679464323e-05,
      "loss": 0.8754,
      "step": 987
    },
    {
      "epoch": 0.1974,
      "loss_align": 0.4234936237335205,
      "loss_contrastive": 0.22928908467292786,
      "loss_gen": 0.2713203430175781,
      "neg_sim_mean": 0.7057954668998718,
      "pos_sim_mean": 0.5765063762664795,
      "step": 987
    },
    {
      "epoch": 0.1976,
      "grad_norm": 2.4472289085388184,
      "learning_rate": 9.713789771179982e-05,
      "loss": 0.7223,
      "step": 988
    },
    {
      "epoch": 0.1976,
      "loss_align": 0.18994319438934326,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19575950503349304,
      "neg_sim_mean": 0.7022315263748169,
      "pos_sim_mean": 0.8100568056106567,
      "step": 988
    },
    {
      "epoch": 0.1978,
      "grad_norm": 2.1228489875793457,
      "learning_rate": 9.712624565451772e-05,
      "loss": 0.3857,
      "step": 989
    },
    {
      "epoch": 0.1978,
      "loss_align": 0.434795618057251,
      "loss_contrastive": 0.09299060702323914,
      "loss_gen": 0.009949731640517712,
      "neg_sim_mean": 0.5581949949264526,
      "pos_sim_mean": 0.565204381942749,
      "step": 989
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.8263043761253357,
      "learning_rate": 9.711457062847595e-05,
      "loss": 0.4559,
      "step": 990
    },
    {
      "epoch": 0.198,
      "loss_align": 0.26272785663604736,
      "loss_contrastive": 0.06714564561843872,
      "loss_gen": 0.34275779128074646,
      "neg_sim_mean": 0.7044177651405334,
      "pos_sim_mean": 0.7372721433639526,
      "step": 990
    },
    {
      "epoch": 0.1982,
      "grad_norm": 2.4446394443511963,
      "learning_rate": 9.710287263936484e-05,
      "loss": 0.6135,
      "step": 991
    },
    {
      "epoch": 0.1982,
      "loss_align": 0.47349631786346436,
      "loss_contrastive": 0.2597522437572479,
      "loss_gen": 0.19275173544883728,
      "neg_sim_mean": 0.686255931854248,
      "pos_sim_mean": 0.5265036821365356,
      "step": 991
    },
    {
      "epoch": 0.1984,
      "grad_norm": 1.8861078023910522,
      "learning_rate": 9.70911516928858e-05,
      "loss": 0.6974,
      "step": 992
    },
    {
      "epoch": 0.1984,
      "loss_align": 0.2384873628616333,
      "loss_contrastive": 0.0,
      "loss_gen": 0.014009028673171997,
      "neg_sim_mean": 0.38629332184791565,
      "pos_sim_mean": 0.7615126371383667,
      "step": 992
    },
    {
      "epoch": 0.1986,
      "grad_norm": 0.9785428643226624,
      "learning_rate": 9.707940779475151e-05,
      "loss": 0.2525,
      "step": 993
    },
    {
      "epoch": 0.1986,
      "loss_align": 0.35813701152801514,
      "loss_contrastive": 0.19900858402252197,
      "loss_gen": 0.26224201917648315,
      "neg_sim_mean": 0.7408715486526489,
      "pos_sim_mean": 0.6418629884719849,
      "step": 993
    },
    {
      "epoch": 0.1988,
      "grad_norm": 2.0030405521392822,
      "learning_rate": 9.70676409506858e-05,
      "loss": 0.6443,
      "step": 994
    },
    {
      "epoch": 0.1988,
      "loss_align": 0.44233936071395874,
      "loss_contrastive": 0.16140881180763245,
      "loss_gen": 0.16641199588775635,
      "neg_sim_mean": 0.6190694570541382,
      "pos_sim_mean": 0.5576606392860413,
      "step": 994
    },
    {
      "epoch": 0.199,
      "grad_norm": 2.0080292224884033,
      "learning_rate": 9.705585116642364e-05,
      "loss": 0.6281,
      "step": 995
    },
    {
      "epoch": 0.199,
      "loss_align": 0.3044671416282654,
      "loss_contrastive": 0.0814366340637207,
      "loss_gen": 0.4599587917327881,
      "neg_sim_mean": 0.6769694685935974,
      "pos_sim_mean": 0.6955328583717346,
      "step": 995
    },
    {
      "epoch": 0.1992,
      "grad_norm": 2.4589385986328125,
      "learning_rate": 9.704403844771128e-05,
      "loss": 0.7742,
      "step": 996
    },
    {
      "epoch": 0.1992,
      "loss_align": 0.4135902523994446,
      "loss_contrastive": 0.1987568438053131,
      "loss_gen": 0.3260107934474945,
      "neg_sim_mean": 0.685166597366333,
      "pos_sim_mean": 0.5864097476005554,
      "step": 996
    },
    {
      "epoch": 0.1994,
      "grad_norm": 2.0860798358917236,
      "learning_rate": 9.703220280030608e-05,
      "loss": 0.7635,
      "step": 997
    },
    {
      "epoch": 0.1994,
      "loss_align": 0.312113881111145,
      "loss_contrastive": 0.05463826656341553,
      "loss_gen": 0.2935503423213959,
      "neg_sim_mean": 0.6425243616104126,
      "pos_sim_mean": 0.687886118888855,
      "step": 997
    },
    {
      "epoch": 0.1996,
      "grad_norm": 2.187670946121216,
      "learning_rate": 9.702034422997658e-05,
      "loss": 0.6122,
      "step": 998
    },
    {
      "epoch": 0.1996,
      "loss_align": 0.4404866695404053,
      "loss_contrastive": 0.12057831883430481,
      "loss_gen": 0.17638134956359863,
      "neg_sim_mean": 0.580091655254364,
      "pos_sim_mean": 0.5595133304595947,
      "step": 998
    },
    {
      "epoch": 0.1998,
      "grad_norm": 1.794672966003418,
      "learning_rate": 9.700846274250251e-05,
      "loss": 0.6313,
      "step": 999
    },
    {
      "epoch": 0.1998,
      "loss_align": 0.207061767578125,
      "loss_contrastive": 0.0,
      "loss_gen": 0.023862561210989952,
      "neg_sim_mean": 0.4322751462459564,
      "pos_sim_mean": 0.792938232421875,
      "step": 999
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.6777244806289673,
      "learning_rate": 9.699655834367478e-05,
      "loss": 0.2309,
      "step": 1000
    },
    {
      "epoch": 0.2,
      "loss_align": 0.3526079058647156,
      "loss_contrastive": 0.10055691003799438,
      "loss_gen": 0.4167080819606781,
      "neg_sim_mean": 0.6479489803314209,
      "pos_sim_mean": 0.6473920941352844,
      "step": 1000
    },
    {
      "epoch": 0.2002,
      "grad_norm": 3.1165049076080322,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.7814,
      "step": 1001
    },
    {
      "epoch": 0.2002,
      "loss_align": 0.35312604904174805,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007865036837756634,
      "neg_sim_mean": 0.45290786027908325,
      "pos_sim_mean": 0.646873950958252,
      "step": 1001
    },
    {
      "epoch": 0.2004,
      "grad_norm": 1.012306809425354,
      "learning_rate": 9.697268083517767e-05,
      "loss": 0.361,
      "step": 1002
    },
    {
      "epoch": 0.2004,
      "loss_align": 0.254131019115448,
      "loss_contrastive": 0.10842019319534302,
      "loss_gen": 0.28392717242240906,
      "neg_sim_mean": 0.7542891502380371,
      "pos_sim_mean": 0.745868980884552,
      "step": 1002
    },
    {
      "epoch": 0.2006,
      "grad_norm": 2.06250262260437,
      "learning_rate": 9.696070773714591e-05,
      "loss": 0.5511,
      "step": 1003
    },
    {
      "epoch": 0.2006,
      "loss_align": 0.3478192687034607,
      "loss_contrastive": 0.00990229845046997,
      "loss_gen": 0.010447260923683643,
      "neg_sim_mean": 0.5620830059051514,
      "pos_sim_mean": 0.6521807312965393,
      "step": 1003
    },
    {
      "epoch": 0.2008,
      "grad_norm": 0.8418737053871155,
      "learning_rate": 9.694871175103571e-05,
      "loss": 0.3595,
      "step": 1004
    },
    {
      "epoch": 0.2008,
      "loss_align": 0.5549522042274475,
      "loss_contrastive": 0.25689902901649475,
      "loss_gen": 0.2758694291114807,
      "neg_sim_mean": 0.6019468307495117,
      "pos_sim_mean": 0.4450477957725525,
      "step": 1004
    },
    {
      "epoch": 0.201,
      "grad_norm": 2.801445722579956,
      "learning_rate": 9.693669288269372e-05,
      "loss": 0.8616,
      "step": 1005
    },
    {
      "epoch": 0.201,
      "loss_align": 0.4665229916572571,
      "loss_contrastive": 0.12559804320335388,
      "loss_gen": 0.39291828870773315,
      "neg_sim_mean": 0.5590750575065613,
      "pos_sim_mean": 0.5334770083427429,
      "step": 1005
    },
    {
      "epoch": 0.2012,
      "grad_norm": 2.6170899868011475,
      "learning_rate": 9.69246511379778e-05,
      "loss": 0.8745,
      "step": 1006
    },
    {
      "epoch": 0.2012,
      "loss_align": 0.3117350935935974,
      "loss_contrastive": 0.08808839321136475,
      "loss_gen": 0.2832641005516052,
      "neg_sim_mean": 0.6763532757759094,
      "pos_sim_mean": 0.6882649064064026,
      "step": 1006
    },
    {
      "epoch": 0.2014,
      "grad_norm": 2.2779555320739746,
      "learning_rate": 9.691258652275698e-05,
      "loss": 0.6056,
      "step": 1007
    },
    {
      "epoch": 0.2014,
      "loss_align": 0.3938157558441162,
      "loss_contrastive": 0.044244229793548584,
      "loss_gen": 0.12226498872041702,
      "neg_sim_mean": 0.5504284501075745,
      "pos_sim_mean": 0.6061842441558838,
      "step": 1007
    },
    {
      "epoch": 0.2016,
      "grad_norm": 1.9422143697738647,
      "learning_rate": 9.690049904291139e-05,
      "loss": 0.5214,
      "step": 1008
    },
    {
      "epoch": 0.2016,
      "loss_align": 0.3785039186477661,
      "loss_contrastive": 0.11667740345001221,
      "loss_gen": 0.3112224042415619,
      "neg_sim_mean": 0.6381734609603882,
      "pos_sim_mean": 0.6214960813522339,
      "step": 1008
    },
    {
      "epoch": 0.2018,
      "grad_norm": 2.2658421993255615,
      "learning_rate": 9.68883887043323e-05,
      "loss": 0.7037,
      "step": 1009
    },
    {
      "epoch": 0.2018,
      "loss_align": 0.2004004716873169,
      "loss_contrastive": 0.10145890712738037,
      "loss_gen": 0.3561411201953888,
      "neg_sim_mean": 0.8010584115982056,
      "pos_sim_mean": 0.7995995283126831,
      "step": 1009
    },
    {
      "epoch": 0.202,
      "grad_norm": 2.557936191558838,
      "learning_rate": 9.68762555129222e-05,
      "loss": 0.5687,
      "step": 1010
    },
    {
      "epoch": 0.202,
      "loss_align": 0.25573229789733887,
      "loss_contrastive": 0.03856503963470459,
      "loss_gen": 0.26963940262794495,
      "neg_sim_mean": 0.6828327178955078,
      "pos_sim_mean": 0.7442677021026611,
      "step": 1010
    },
    {
      "epoch": 0.2022,
      "grad_norm": 2.709620714187622,
      "learning_rate": 9.686409947459458e-05,
      "loss": 0.53,
      "step": 1011
    },
    {
      "epoch": 0.2022,
      "loss_align": 0.3522747755050659,
      "loss_contrastive": 0.0801076889038086,
      "loss_gen": 0.4078730046749115,
      "neg_sim_mean": 0.6278328895568848,
      "pos_sim_mean": 0.6477252244949341,
      "step": 1011
    },
    {
      "epoch": 0.2024,
      "grad_norm": 2.3988943099975586,
      "learning_rate": 9.68519205952742e-05,
      "loss": 0.7698,
      "step": 1012
    },
    {
      "epoch": 0.2024,
      "loss_align": 0.36263787746429443,
      "loss_contrastive": 0.09227573871612549,
      "loss_gen": 0.27774107456207275,
      "neg_sim_mean": 0.6296378374099731,
      "pos_sim_mean": 0.6373621225357056,
      "step": 1012
    },
    {
      "epoch": 0.2026,
      "grad_norm": 2.2644529342651367,
      "learning_rate": 9.68397188808969e-05,
      "loss": 0.6515,
      "step": 1013
    },
    {
      "epoch": 0.2026,
      "loss_align": 0.28464192152023315,
      "loss_contrastive": 0.09893804788589478,
      "loss_gen": 0.1702079176902771,
      "neg_sim_mean": 0.7142961025238037,
      "pos_sim_mean": 0.7153580784797668,
      "step": 1013
    },
    {
      "epoch": 0.2028,
      "grad_norm": 1.8027820587158203,
      "learning_rate": 9.682749433740962e-05,
      "loss": 0.4667,
      "step": 1014
    },
    {
      "epoch": 0.2028,
      "loss_align": 0.26866209506988525,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001650901511311531,
      "neg_sim_mean": 0.513590395450592,
      "pos_sim_mean": 0.7313379049301147,
      "step": 1014
    },
    {
      "epoch": 0.203,
      "grad_norm": 0.4449164569377899,
      "learning_rate": 9.681524697077047e-05,
      "loss": 0.2703,
      "step": 1015
    },
    {
      "epoch": 0.203,
      "loss_align": 0.2737584114074707,
      "loss_contrastive": 0.10649925470352173,
      "loss_gen": 0.2833368182182312,
      "neg_sim_mean": 0.7327408194541931,
      "pos_sim_mean": 0.7262415885925293,
      "step": 1015
    },
    {
      "epoch": 0.2032,
      "grad_norm": 2.425530433654785,
      "learning_rate": 9.680297678694867e-05,
      "loss": 0.5699,
      "step": 1016
    },
    {
      "epoch": 0.2032,
      "loss_align": 0.41917192935943604,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003610544139519334,
      "neg_sim_mean": 0.40780168771743774,
      "pos_sim_mean": 0.580828070640564,
      "step": 1016
    },
    {
      "epoch": 0.2034,
      "grad_norm": 0.6126342415809631,
      "learning_rate": 9.679068379192456e-05,
      "loss": 0.4228,
      "step": 1017
    },
    {
      "epoch": 0.2034,
      "loss_align": 0.3007829189300537,
      "loss_contrastive": 0.1282089352607727,
      "loss_gen": 0.3717288076877594,
      "neg_sim_mean": 0.7274259924888611,
      "pos_sim_mean": 0.6992170810699463,
      "step": 1017
    },
    {
      "epoch": 0.2036,
      "grad_norm": 2.6180312633514404,
      "learning_rate": 9.677836799168957e-05,
      "loss": 0.6879,
      "step": 1018
    },
    {
      "epoch": 0.2036,
      "loss_align": 0.2644498348236084,
      "loss_contrastive": 0.10520470142364502,
      "loss_gen": 0.2350032925605774,
      "neg_sim_mean": 0.7407548427581787,
      "pos_sim_mean": 0.7355501651763916,
      "step": 1018
    },
    {
      "epoch": 0.2038,
      "grad_norm": 1.7784169912338257,
      "learning_rate": 9.676602939224629e-05,
      "loss": 0.5121,
      "step": 1019
    },
    {
      "epoch": 0.2038,
      "loss_align": 0.40569090843200684,
      "loss_contrastive": 0.15786483883857727,
      "loss_gen": 0.1971205323934555,
      "neg_sim_mean": 0.6521739363670349,
      "pos_sim_mean": 0.5943090915679932,
      "step": 1019
    },
    {
      "epoch": 0.204,
      "grad_norm": 2.034613847732544,
      "learning_rate": 9.675366799960841e-05,
      "loss": 0.6218,
      "step": 1020
    },
    {
      "epoch": 0.204,
      "loss_align": 0.18722283840179443,
      "loss_contrastive": 0.04269707202911377,
      "loss_gen": 0.23803327977657318,
      "neg_sim_mean": 0.7554742097854614,
      "pos_sim_mean": 0.8127771615982056,
      "step": 1020
    },
    {
      "epoch": 0.2042,
      "grad_norm": 1.9408036470413208,
      "learning_rate": 9.674128381980072e-05,
      "loss": 0.4304,
      "step": 1021
    },
    {
      "epoch": 0.2042,
      "loss_align": 0.329174280166626,
      "loss_contrastive": 0.08305442333221436,
      "loss_gen": 0.1479014903306961,
      "neg_sim_mean": 0.6538801193237305,
      "pos_sim_mean": 0.670825719833374,
      "step": 1021
    },
    {
      "epoch": 0.2044,
      "grad_norm": 1.891326904296875,
      "learning_rate": 9.672887685885912e-05,
      "loss": 0.487,
      "step": 1022
    },
    {
      "epoch": 0.2044,
      "loss_align": 0.156660258769989,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17051269114017487,
      "neg_sim_mean": 0.6458012461662292,
      "pos_sim_mean": 0.843339741230011,
      "step": 1022
    },
    {
      "epoch": 0.2046,
      "grad_norm": 1.7118597030639648,
      "learning_rate": 9.671644712283061e-05,
      "loss": 0.3272,
      "step": 1023
    },
    {
      "epoch": 0.2046,
      "loss_align": 0.20798885822296143,
      "loss_contrastive": 0.07792377471923828,
      "loss_gen": 0.34543001651763916,
      "neg_sim_mean": 0.769934892654419,
      "pos_sim_mean": 0.7920111417770386,
      "step": 1023
    },
    {
      "epoch": 0.2048,
      "grad_norm": 2.387714385986328,
      "learning_rate": 9.670399461777329e-05,
      "loss": 0.5628,
      "step": 1024
    },
    {
      "epoch": 0.2048,
      "loss_align": 0.3255469799041748,
      "loss_contrastive": 0.08598637580871582,
      "loss_gen": 0.23557992279529572,
      "neg_sim_mean": 0.6604393720626831,
      "pos_sim_mean": 0.6744530200958252,
      "step": 1024
    },
    {
      "epoch": 0.205,
      "grad_norm": 2.040832281112671,
      "learning_rate": 9.669151934975634e-05,
      "loss": 0.5714,
      "step": 1025
    },
    {
      "epoch": 0.205,
      "loss_align": 0.33581793308258057,
      "loss_contrastive": 0.05579805374145508,
      "loss_gen": 0.2726467251777649,
      "neg_sim_mean": 0.6199800968170166,
      "pos_sim_mean": 0.6641820669174194,
      "step": 1025
    },
    {
      "epoch": 0.2052,
      "grad_norm": 2.4192914962768555,
      "learning_rate": 9.667902132486009e-05,
      "loss": 0.6152,
      "step": 1026
    },
    {
      "epoch": 0.2052,
      "loss_align": 0.2213795781135559,
      "loss_contrastive": 0.12030136585235596,
      "loss_gen": 0.24241787195205688,
      "neg_sim_mean": 0.7989217638969421,
      "pos_sim_mean": 0.7786204218864441,
      "step": 1026
    },
    {
      "epoch": 0.2054,
      "grad_norm": 1.781551480293274,
      "learning_rate": 9.66665005491759e-05,
      "loss": 0.4782,
      "step": 1027
    },
    {
      "epoch": 0.2054,
      "loss_align": 0.2671889066696167,
      "loss_contrastive": 0.002659916877746582,
      "loss_gen": 0.3531951904296875,
      "neg_sim_mean": 0.635470986366272,
      "pos_sim_mean": 0.7328110933303833,
      "step": 1027
    },
    {
      "epoch": 0.2056,
      "grad_norm": 2.9044113159179688,
      "learning_rate": 9.665395702880626e-05,
      "loss": 0.6207,
      "step": 1028
    },
    {
      "epoch": 0.2056,
      "loss_align": 0.18659639358520508,
      "loss_contrastive": 0.11483055353164673,
      "loss_gen": 0.39395684003829956,
      "neg_sim_mean": 0.8282341361045837,
      "pos_sim_mean": 0.8134036064147949,
      "step": 1028
    },
    {
      "epoch": 0.2058,
      "grad_norm": 2.6046195030212402,
      "learning_rate": 9.664139076986473e-05,
      "loss": 0.5943,
      "step": 1029
    },
    {
      "epoch": 0.2058,
      "loss_align": 0.214128315448761,
      "loss_contrastive": 0.0,
      "loss_gen": 0.018536241725087166,
      "neg_sim_mean": 0.6006737947463989,
      "pos_sim_mean": 0.785871684551239,
      "step": 1029
    },
    {
      "epoch": 0.206,
      "grad_norm": 1.0643248558044434,
      "learning_rate": 9.662880177847594e-05,
      "loss": 0.2327,
      "step": 1030
    },
    {
      "epoch": 0.206,
      "loss_align": 0.3251906633377075,
      "loss_contrastive": 0.09603631496429443,
      "loss_gen": 0.24769191443920135,
      "neg_sim_mean": 0.670845627784729,
      "pos_sim_mean": 0.6748093366622925,
      "step": 1030
    },
    {
      "epoch": 0.2062,
      "grad_norm": 2.5122711658477783,
      "learning_rate": 9.661619006077561e-05,
      "loss": 0.5844,
      "step": 1031
    },
    {
      "epoch": 0.2062,
      "loss_align": 0.18520903587341309,
      "loss_contrastive": 0.11105173826217651,
      "loss_gen": 0.23463979363441467,
      "neg_sim_mean": 0.8258426785469055,
      "pos_sim_mean": 0.8147909641265869,
      "step": 1031
    },
    {
      "epoch": 0.2064,
      "grad_norm": 2.1040186882019043,
      "learning_rate": 9.660355562291055e-05,
      "loss": 0.4332,
      "step": 1032
    },
    {
      "epoch": 0.2064,
      "loss_align": 0.44084179401397705,
      "loss_contrastive": 0.038310885429382324,
      "loss_gen": 0.011204729788005352,
      "neg_sim_mean": 0.49746909737586975,
      "pos_sim_mean": 0.559158205986023,
      "step": 1032
    },
    {
      "epoch": 0.2066,
      "grad_norm": 1.0772347450256348,
      "learning_rate": 9.659089847103863e-05,
      "loss": 0.4566,
      "step": 1033
    },
    {
      "epoch": 0.2066,
      "loss_align": 0.4828331470489502,
      "loss_contrastive": 0.06979942321777344,
      "loss_gen": 0.32502102851867676,
      "neg_sim_mean": 0.4869662821292877,
      "pos_sim_mean": 0.5171668529510498,
      "step": 1033
    },
    {
      "epoch": 0.2068,
      "grad_norm": 2.116213083267212,
      "learning_rate": 9.657821861132878e-05,
      "loss": 0.8162,
      "step": 1034
    },
    {
      "epoch": 0.2068,
      "loss_align": 0.4904736876487732,
      "loss_contrastive": 0.14525440335273743,
      "loss_gen": 0.2620701789855957,
      "neg_sim_mean": 0.5547807216644287,
      "pos_sim_mean": 0.5095263123512268,
      "step": 1034
    },
    {
      "epoch": 0.207,
      "grad_norm": 2.046550989151001,
      "learning_rate": 9.656551604996102e-05,
      "loss": 0.77,
      "step": 1035
    },
    {
      "epoch": 0.207,
      "loss_align": 0.36411118507385254,
      "loss_contrastive": 0.12653100490570068,
      "loss_gen": 0.011901898309588432,
      "neg_sim_mean": 0.6624197959899902,
      "pos_sim_mean": 0.6358888149261475,
      "step": 1035
    },
    {
      "epoch": 0.2072,
      "grad_norm": 0.920789361000061,
      "learning_rate": 9.655279079312642e-05,
      "loss": 0.3912,
      "step": 1036
    },
    {
      "epoch": 0.2072,
      "loss_align": 0.39312970638275146,
      "loss_contrastive": 0.1723315715789795,
      "loss_gen": 0.1896347552537918,
      "neg_sim_mean": 0.6792018413543701,
      "pos_sim_mean": 0.6068702936172485,
      "step": 1036
    },
    {
      "epoch": 0.2074,
      "grad_norm": 2.0649731159210205,
      "learning_rate": 9.654004284702711e-05,
      "loss": 0.6034,
      "step": 1037
    },
    {
      "epoch": 0.2074,
      "loss_align": 0.2897202968597412,
      "loss_contrastive": 0.12033504247665405,
      "loss_gen": 0.5414879322052002,
      "neg_sim_mean": 0.7306147217750549,
      "pos_sim_mean": 0.7102797031402588,
      "step": 1037
    },
    {
      "epoch": 0.2076,
      "grad_norm": 2.6561973094940186,
      "learning_rate": 9.652727221787631e-05,
      "loss": 0.8456,
      "step": 1038
    },
    {
      "epoch": 0.2076,
      "loss_align": 0.15744715929031372,
      "loss_contrastive": 0.08692377805709839,
      "loss_gen": 0.3160199224948883,
      "neg_sim_mean": 0.8294765949249268,
      "pos_sim_mean": 0.8425528407096863,
      "step": 1038
    },
    {
      "epoch": 0.2078,
      "grad_norm": 2.1871137619018555,
      "learning_rate": 9.651447891189825e-05,
      "loss": 0.4839,
      "step": 1039
    },
    {
      "epoch": 0.2078,
      "loss_align": 0.22375750541687012,
      "loss_contrastive": 0.18349969387054443,
      "loss_gen": 0.4470013678073883,
      "neg_sim_mean": 0.8597421646118164,
      "pos_sim_mean": 0.7762424945831299,
      "step": 1039
    },
    {
      "epoch": 0.208,
      "grad_norm": 2.912187099456787,
      "learning_rate": 9.65016629353282e-05,
      "loss": 0.6928,
      "step": 1040
    },
    {
      "epoch": 0.208,
      "loss_align": 0.32632768154144287,
      "loss_contrastive": 0.030666589736938477,
      "loss_gen": 0.07574692368507385,
      "neg_sim_mean": 0.6043388843536377,
      "pos_sim_mean": 0.6736723184585571,
      "step": 1040
    },
    {
      "epoch": 0.2082,
      "grad_norm": 2.1337103843688965,
      "learning_rate": 9.648882429441257e-05,
      "loss": 0.4058,
      "step": 1041
    },
    {
      "epoch": 0.2082,
      "loss_align": 0.4693363904953003,
      "loss_contrastive": 0.16020354628562927,
      "loss_gen": 0.16776591539382935,
      "neg_sim_mean": 0.5908671617507935,
      "pos_sim_mean": 0.5306636095046997,
      "step": 1041
    },
    {
      "epoch": 0.2084,
      "grad_norm": 1.7334511280059814,
      "learning_rate": 9.647596299540873e-05,
      "loss": 0.6563,
      "step": 1042
    },
    {
      "epoch": 0.2084,
      "loss_align": 0.26979005336761475,
      "loss_contrastive": 0.06552112102508545,
      "loss_gen": 0.42864713072776794,
      "neg_sim_mean": 0.6957310438156128,
      "pos_sim_mean": 0.7302099466323853,
      "step": 1042
    },
    {
      "epoch": 0.2086,
      "grad_norm": 3.9302260875701904,
      "learning_rate": 9.646307904458513e-05,
      "loss": 0.7063,
      "step": 1043
    },
    {
      "epoch": 0.2086,
      "loss_align": 0.44693470001220703,
      "loss_contrastive": 0.1288590133190155,
      "loss_gen": 0.26597121357917786,
      "neg_sim_mean": 0.581924319267273,
      "pos_sim_mean": 0.553065299987793,
      "step": 1043
    },
    {
      "epoch": 0.2088,
      "grad_norm": 2.314216375350952,
      "learning_rate": 9.645017244822123e-05,
      "loss": 0.7284,
      "step": 1044
    },
    {
      "epoch": 0.2088,
      "loss_align": 0.3139948844909668,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009304980747401714,
      "neg_sim_mean": 0.4688768982887268,
      "pos_sim_mean": 0.6860051155090332,
      "step": 1044
    },
    {
      "epoch": 0.209,
      "grad_norm": 0.8284803628921509,
      "learning_rate": 9.643724321260756e-05,
      "loss": 0.3233,
      "step": 1045
    },
    {
      "epoch": 0.209,
      "loss_align": 0.19359451532363892,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2624931335449219,
      "neg_sim_mean": 0.6958110928535461,
      "pos_sim_mean": 0.8064054846763611,
      "step": 1045
    },
    {
      "epoch": 0.2092,
      "grad_norm": 2.6075408458709717,
      "learning_rate": 9.642429134404569e-05,
      "loss": 0.4561,
      "step": 1046
    },
    {
      "epoch": 0.2092,
      "loss_align": 0.1960902214050293,
      "loss_contrastive": 0.10420346260070801,
      "loss_gen": 0.21704518795013428,
      "neg_sim_mean": 0.8081132173538208,
      "pos_sim_mean": 0.8039097785949707,
      "step": 1046
    },
    {
      "epoch": 0.2094,
      "grad_norm": 2.3291015625,
      "learning_rate": 9.641131684884817e-05,
      "loss": 0.4256,
      "step": 1047
    },
    {
      "epoch": 0.2094,
      "loss_align": 0.3546334505081177,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004068049602210522,
      "neg_sim_mean": 0.37332671880722046,
      "pos_sim_mean": 0.6453665494918823,
      "step": 1047
    },
    {
      "epoch": 0.2096,
      "grad_norm": 0.5811882019042969,
      "learning_rate": 9.639831973333864e-05,
      "loss": 0.3587,
      "step": 1048
    },
    {
      "epoch": 0.2096,
      "loss_align": 0.29559624195098877,
      "loss_contrastive": 0.06510573625564575,
      "loss_gen": 0.42187270522117615,
      "neg_sim_mean": 0.6695094704627991,
      "pos_sim_mean": 0.7044037580490112,
      "step": 1048
    },
    {
      "epoch": 0.2098,
      "grad_norm": 2.6563632488250732,
      "learning_rate": 9.638530000385171e-05,
      "loss": 0.7253,
      "step": 1049
    },
    {
      "epoch": 0.2098,
      "loss_align": 0.24937379360198975,
      "loss_contrastive": 0.03411662578582764,
      "loss_gen": 0.15779007971286774,
      "neg_sim_mean": 0.68474280834198,
      "pos_sim_mean": 0.7506262063980103,
      "step": 1049
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.7565451860427856,
      "learning_rate": 9.637225766673307e-05,
      "loss": 0.4113,
      "step": 1050
    },
    {
      "epoch": 0.21,
      "loss_align": 0.3798903226852417,
      "loss_contrastive": 0.20976614952087402,
      "loss_gen": 0.4176521301269531,
      "neg_sim_mean": 0.7298758029937744,
      "pos_sim_mean": 0.6201096773147583,
      "step": 1050
    },
    {
      "epoch": 0.2102,
      "grad_norm": 2.8262128829956055,
      "learning_rate": 9.635919272833938e-05,
      "loss": 0.8227,
      "step": 1051
    },
    {
      "epoch": 0.2102,
      "loss_align": 0.3525412082672119,
      "loss_contrastive": 0.10602402687072754,
      "loss_gen": 0.4794984459877014,
      "neg_sim_mean": 0.6534827947616577,
      "pos_sim_mean": 0.6474587917327881,
      "step": 1051
    },
    {
      "epoch": 0.2104,
      "grad_norm": 2.6513869762420654,
      "learning_rate": 9.634610519503832e-05,
      "loss": 0.8448,
      "step": 1052
    },
    {
      "epoch": 0.2104,
      "loss_align": 0.38436973094940186,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00256079132668674,
      "neg_sim_mean": 0.4509815573692322,
      "pos_sim_mean": 0.6156302690505981,
      "step": 1052
    },
    {
      "epoch": 0.2106,
      "grad_norm": 0.6859422922134399,
      "learning_rate": 9.63329950732086e-05,
      "loss": 0.3869,
      "step": 1053
    },
    {
      "epoch": 0.2106,
      "loss_align": 0.22665345668792725,
      "loss_contrastive": 0.06332004070281982,
      "loss_gen": 0.1902395337820053,
      "neg_sim_mean": 0.7366665601730347,
      "pos_sim_mean": 0.7733465433120728,
      "step": 1053
    },
    {
      "epoch": 0.2108,
      "grad_norm": 1.8724268674850464,
      "learning_rate": 9.631986236923997e-05,
      "loss": 0.4245,
      "step": 1054
    },
    {
      "epoch": 0.2108,
      "loss_align": 0.2397196888923645,
      "loss_contrastive": 0.1236909031867981,
      "loss_gen": 0.2738282084465027,
      "neg_sim_mean": 0.7839711904525757,
      "pos_sim_mean": 0.7602803111076355,
      "step": 1054
    },
    {
      "epoch": 0.211,
      "grad_norm": 2.0902230739593506,
      "learning_rate": 9.630670708953311e-05,
      "loss": 0.5284,
      "step": 1055
    },
    {
      "epoch": 0.211,
      "loss_align": 0.44423454999923706,
      "loss_contrastive": 0.07926526665687561,
      "loss_gen": 0.4806457459926605,
      "neg_sim_mean": 0.535030722618103,
      "pos_sim_mean": 0.5557654500007629,
      "step": 1055
    },
    {
      "epoch": 0.2112,
      "grad_norm": 2.907822370529175,
      "learning_rate": 9.629352924049975e-05,
      "loss": 0.9344,
      "step": 1056
    },
    {
      "epoch": 0.2112,
      "loss_align": 0.20570993423461914,
      "loss_contrastive": 0.1152108907699585,
      "loss_gen": 0.3366546332836151,
      "neg_sim_mean": 0.8095009326934814,
      "pos_sim_mean": 0.7942900657653809,
      "step": 1056
    },
    {
      "epoch": 0.2114,
      "grad_norm": 1.9949431419372559,
      "learning_rate": 9.628032882856261e-05,
      "loss": 0.5562,
      "step": 1057
    },
    {
      "epoch": 0.2114,
      "loss_align": 0.3168456554412842,
      "loss_contrastive": 0.05317455530166626,
      "loss_gen": 0.2511458992958069,
      "neg_sim_mean": 0.6363288760185242,
      "pos_sim_mean": 0.6831543445587158,
      "step": 1057
    },
    {
      "epoch": 0.2116,
      "grad_norm": 2.0866594314575195,
      "learning_rate": 9.626710586015543e-05,
      "loss": 0.5744,
      "step": 1058
    },
    {
      "epoch": 0.2116,
      "loss_align": 0.20267105102539062,
      "loss_contrastive": 0.045923054218292236,
      "loss_gen": 0.29658764600753784,
      "neg_sim_mean": 0.7432519793510437,
      "pos_sim_mean": 0.7973289489746094,
      "step": 1058
    },
    {
      "epoch": 0.2118,
      "grad_norm": 2.190643548965454,
      "learning_rate": 9.62538603417229e-05,
      "loss": 0.5048,
      "step": 1059
    },
    {
      "epoch": 0.2118,
      "loss_align": 0.28530311584472656,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0014514016220346093,
      "neg_sim_mean": 0.5583482980728149,
      "pos_sim_mean": 0.7146968841552734,
      "step": 1059
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.46491318941116333,
      "learning_rate": 9.624059227972075e-05,
      "loss": 0.2868,
      "step": 1060
    },
    {
      "epoch": 0.212,
      "loss_align": 0.28251010179519653,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02167319692671299,
      "neg_sim_mean": 0.32693779468536377,
      "pos_sim_mean": 0.7174898982048035,
      "step": 1060
    },
    {
      "epoch": 0.2122,
      "grad_norm": 1.3290477991104126,
      "learning_rate": 9.622730168061567e-05,
      "loss": 0.3042,
      "step": 1061
    },
    {
      "epoch": 0.2122,
      "loss_align": 0.22073078155517578,
      "loss_contrastive": 0.12613749504089355,
      "loss_gen": 0.3277617394924164,
      "neg_sim_mean": 0.8054066896438599,
      "pos_sim_mean": 0.7792692184448242,
      "step": 1061
    },
    {
      "epoch": 0.2124,
      "grad_norm": 2.4157307147979736,
      "learning_rate": 9.62139885508853e-05,
      "loss": 0.5636,
      "step": 1062
    },
    {
      "epoch": 0.2124,
      "loss_align": 0.3976665735244751,
      "loss_contrastive": 0.0856509804725647,
      "loss_gen": 0.304877907037735,
      "neg_sim_mean": 0.5879843831062317,
      "pos_sim_mean": 0.6023334264755249,
      "step": 1062
    },
    {
      "epoch": 0.2126,
      "grad_norm": 2.363758087158203,
      "learning_rate": 9.620065289701834e-05,
      "loss": 0.7128,
      "step": 1063
    },
    {
      "epoch": 0.2126,
      "loss_align": 0.3180077075958252,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003590845502912998,
      "neg_sim_mean": 0.4307221472263336,
      "pos_sim_mean": 0.6819922924041748,
      "step": 1063
    },
    {
      "epoch": 0.2128,
      "grad_norm": 0.521578848361969,
      "learning_rate": 9.618729472551441e-05,
      "loss": 0.3216,
      "step": 1064
    },
    {
      "epoch": 0.2128,
      "loss_align": 0.366178035736084,
      "loss_contrastive": 0.21303439140319824,
      "loss_gen": 0.3182355761528015,
      "neg_sim_mean": 0.7468563318252563,
      "pos_sim_mean": 0.633821964263916,
      "step": 1064
    },
    {
      "epoch": 0.213,
      "grad_norm": 2.2587027549743652,
      "learning_rate": 9.617391404288412e-05,
      "loss": 0.71,
      "step": 1065
    },
    {
      "epoch": 0.213,
      "loss_align": 0.5274367332458496,
      "loss_contrastive": 0.11701595783233643,
      "loss_gen": 0.5794643759727478,
      "neg_sim_mean": 0.4895792007446289,
      "pos_sim_mean": 0.472563236951828,
      "step": 1065
    },
    {
      "epoch": 0.2132,
      "grad_norm": 3.3940765857696533,
      "learning_rate": 9.616051085564906e-05,
      "loss": 1.1209,
      "step": 1066
    },
    {
      "epoch": 0.2132,
      "loss_align": 0.27590906620025635,
      "loss_contrastive": 0.004156708717346191,
      "loss_gen": 0.22587087750434875,
      "neg_sim_mean": 0.6282476186752319,
      "pos_sim_mean": 0.7240909337997437,
      "step": 1066
    },
    {
      "epoch": 0.2134,
      "grad_norm": 2.054948329925537,
      "learning_rate": 9.614708517034175e-05,
      "loss": 0.5023,
      "step": 1067
    },
    {
      "epoch": 0.2134,
      "loss_align": 0.35981643199920654,
      "loss_contrastive": 0.19194108247756958,
      "loss_gen": 0.3043254613876343,
      "neg_sim_mean": 0.7321246266365051,
      "pos_sim_mean": 0.6401835680007935,
      "step": 1067
    },
    {
      "epoch": 0.2136,
      "grad_norm": 2.2820889949798584,
      "learning_rate": 9.613363699350575e-05,
      "loss": 0.6872,
      "step": 1068
    },
    {
      "epoch": 0.2136,
      "loss_align": 0.254061758518219,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013573172502219677,
      "neg_sim_mean": 0.5057094097137451,
      "pos_sim_mean": 0.745938241481781,
      "step": 1068
    },
    {
      "epoch": 0.2138,
      "grad_norm": 1.2299586534500122,
      "learning_rate": 9.612016633169551e-05,
      "loss": 0.2676,
      "step": 1069
    },
    {
      "epoch": 0.2138,
      "loss_align": 0.3506375551223755,
      "loss_contrastive": 0.13944804668426514,
      "loss_gen": 0.27729636430740356,
      "neg_sim_mean": 0.6888104677200317,
      "pos_sim_mean": 0.6493624448776245,
      "step": 1069
    },
    {
      "epoch": 0.214,
      "grad_norm": 2.122001886367798,
      "learning_rate": 9.610667319147647e-05,
      "loss": 0.6447,
      "step": 1070
    },
    {
      "epoch": 0.214,
      "loss_align": 0.23557448387145996,
      "loss_contrastive": 0.07315051555633545,
      "loss_gen": 0.33054178953170776,
      "neg_sim_mean": 0.7375760078430176,
      "pos_sim_mean": 0.76442551612854,
      "step": 1070
    },
    {
      "epoch": 0.2142,
      "grad_norm": 2.3280255794525146,
      "learning_rate": 9.609315757942503e-05,
      "loss": 0.5749,
      "step": 1071
    },
    {
      "epoch": 0.2142,
      "loss_align": 0.2611730098724365,
      "loss_contrastive": 0.08044219017028809,
      "loss_gen": 0.2324356585741043,
      "neg_sim_mean": 0.7192691564559937,
      "pos_sim_mean": 0.7388269901275635,
      "step": 1071
    },
    {
      "epoch": 0.2144,
      "grad_norm": 2.018324851989746,
      "learning_rate": 9.607961950212854e-05,
      "loss": 0.5033,
      "step": 1072
    },
    {
      "epoch": 0.2144,
      "loss_align": 0.3348270654678345,
      "loss_contrastive": 0.14221715927124023,
      "loss_gen": 0.3255847096443176,
      "neg_sim_mean": 0.7073900699615479,
      "pos_sim_mean": 0.6651729345321655,
      "step": 1072
    },
    {
      "epoch": 0.2146,
      "grad_norm": 2.3424785137176514,
      "learning_rate": 9.606605896618529e-05,
      "loss": 0.6775,
      "step": 1073
    },
    {
      "epoch": 0.2146,
      "loss_align": 0.38542139530181885,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0010178667725995183,
      "neg_sim_mean": 0.4599305987358093,
      "pos_sim_mean": 0.6145786046981812,
      "step": 1073
    },
    {
      "epoch": 0.2148,
      "grad_norm": 0.5929842591285706,
      "learning_rate": 9.605247597820448e-05,
      "loss": 0.3864,
      "step": 1074
    },
    {
      "epoch": 0.2148,
      "loss_align": 0.37225306034088135,
      "loss_contrastive": 0.20379936695098877,
      "loss_gen": 0.23859217762947083,
      "neg_sim_mean": 0.7315462827682495,
      "pos_sim_mean": 0.6277469396591187,
      "step": 1074
    },
    {
      "epoch": 0.215,
      "grad_norm": 2.0289885997772217,
      "learning_rate": 9.603887054480636e-05,
      "loss": 0.6353,
      "step": 1075
    },
    {
      "epoch": 0.215,
      "loss_align": 0.2545706033706665,
      "loss_contrastive": 0.0678337812423706,
      "loss_gen": 0.22854962944984436,
      "neg_sim_mean": 0.7132631540298462,
      "pos_sim_mean": 0.7454293966293335,
      "step": 1075
    },
    {
      "epoch": 0.2152,
      "grad_norm": 1.6416313648223877,
      "learning_rate": 9.602524267262203e-05,
      "loss": 0.4913,
      "step": 1076
    },
    {
      "epoch": 0.2152,
      "loss_align": 0.23613160848617554,
      "loss_contrastive": 0.12879669666290283,
      "loss_gen": 0.6144083738327026,
      "neg_sim_mean": 0.7926650643348694,
      "pos_sim_mean": 0.7638683915138245,
      "step": 1076
    },
    {
      "epoch": 0.2154,
      "grad_norm": 2.8977856636047363,
      "learning_rate": 9.601159236829352e-05,
      "loss": 0.866,
      "step": 1077
    },
    {
      "epoch": 0.2154,
      "loss_align": 0.255160927772522,
      "loss_contrastive": 0.10749894380569458,
      "loss_gen": 0.28725889325141907,
      "neg_sim_mean": 0.7523379921913147,
      "pos_sim_mean": 0.744839072227478,
      "step": 1077
    },
    {
      "epoch": 0.2156,
      "grad_norm": 2.3649916648864746,
      "learning_rate": 9.599791963847386e-05,
      "loss": 0.5553,
      "step": 1078
    },
    {
      "epoch": 0.2156,
      "loss_align": 0.11738592386245728,
      "loss_contrastive": 0.035703837871551514,
      "loss_gen": 0.23252791166305542,
      "neg_sim_mean": 0.8183178901672363,
      "pos_sim_mean": 0.8826140761375427,
      "step": 1078
    },
    {
      "epoch": 0.2158,
      "grad_norm": 2.1585893630981445,
      "learning_rate": 9.598422448982696e-05,
      "loss": 0.3542,
      "step": 1079
    },
    {
      "epoch": 0.2158,
      "loss_align": 0.3529309034347534,
      "loss_contrastive": 0.26258039474487305,
      "loss_gen": 0.3194299042224884,
      "neg_sim_mean": 0.8096494674682617,
      "pos_sim_mean": 0.6470690965652466,
      "step": 1079
    },
    {
      "epoch": 0.216,
      "grad_norm": 2.7934160232543945,
      "learning_rate": 9.597050692902765e-05,
      "loss": 0.7039,
      "step": 1080
    },
    {
      "epoch": 0.216,
      "loss_align": 0.3175346851348877,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02053104154765606,
      "neg_sim_mean": 0.5728607177734375,
      "pos_sim_mean": 0.6824653148651123,
      "step": 1080
    },
    {
      "epoch": 0.2162,
      "grad_norm": 1.3019870519638062,
      "learning_rate": 9.595676696276172e-05,
      "loss": 0.3381,
      "step": 1081
    },
    {
      "epoch": 0.2162,
      "loss_align": 0.40987592935562134,
      "loss_contrastive": 0.025392264127731323,
      "loss_gen": 0.33328333497047424,
      "neg_sim_mean": 0.5155163407325745,
      "pos_sim_mean": 0.5901240706443787,
      "step": 1081
    },
    {
      "epoch": 0.2164,
      "grad_norm": 2.359510898590088,
      "learning_rate": 9.594300459772587e-05,
      "loss": 0.7462,
      "step": 1082
    },
    {
      "epoch": 0.2164,
      "loss_align": 0.32487595081329346,
      "loss_contrastive": 0.14645922183990479,
      "loss_gen": 0.1729775369167328,
      "neg_sim_mean": 0.7215832471847534,
      "pos_sim_mean": 0.6751240491867065,
      "step": 1082
    },
    {
      "epoch": 0.2166,
      "grad_norm": 1.6041500568389893,
      "learning_rate": 9.59292198406277e-05,
      "loss": 0.5154,
      "step": 1083
    },
    {
      "epoch": 0.2166,
      "loss_align": 0.3130437135696411,
      "loss_contrastive": 0.060896456241607666,
      "loss_gen": 0.15222850441932678,
      "neg_sim_mean": 0.6478527188301086,
      "pos_sim_mean": 0.6869562864303589,
      "step": 1083
    },
    {
      "epoch": 0.2168,
      "grad_norm": 1.6583309173583984,
      "learning_rate": 9.591541269818573e-05,
      "loss": 0.4726,
      "step": 1084
    },
    {
      "epoch": 0.2168,
      "loss_align": 0.3673211336135864,
      "loss_contrastive": 0.09050363302230835,
      "loss_gen": 0.28036147356033325,
      "neg_sim_mean": 0.623182475566864,
      "pos_sim_mean": 0.6326788663864136,
      "step": 1084
    },
    {
      "epoch": 0.217,
      "grad_norm": 2.0852036476135254,
      "learning_rate": 9.59015831771294e-05,
      "loss": 0.6585,
      "step": 1085
    },
    {
      "epoch": 0.217,
      "loss_align": 0.24300521612167358,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015467623015865684,
      "neg_sim_mean": 0.3186497688293457,
      "pos_sim_mean": 0.7569947838783264,
      "step": 1085
    },
    {
      "epoch": 0.2172,
      "grad_norm": 0.5619054436683655,
      "learning_rate": 9.588773128419906e-05,
      "loss": 0.2446,
      "step": 1086
    },
    {
      "epoch": 0.2172,
      "loss_align": 0.4485796093940735,
      "loss_contrastive": 0.03240075707435608,
      "loss_gen": 0.42208123207092285,
      "neg_sim_mean": 0.48382115364074707,
      "pos_sim_mean": 0.5514203906059265,
      "step": 1086
    },
    {
      "epoch": 0.2174,
      "grad_norm": 2.3381378650665283,
      "learning_rate": 9.587385702614593e-05,
      "loss": 0.8745,
      "step": 1087
    },
    {
      "epoch": 0.2174,
      "loss_align": 0.22568905353546143,
      "loss_contrastive": 0.07014966011047363,
      "loss_gen": 0.5037911534309387,
      "neg_sim_mean": 0.7444605827331543,
      "pos_sim_mean": 0.7743109464645386,
      "step": 1087
    },
    {
      "epoch": 0.2176,
      "grad_norm": 2.620889186859131,
      "learning_rate": 9.585996040973218e-05,
      "loss": 0.7379,
      "step": 1088
    },
    {
      "epoch": 0.2176,
      "loss_align": 0.35170185565948486,
      "loss_contrastive": 0.009810924530029297,
      "loss_gen": 0.016467031091451645,
      "neg_sim_mean": 0.5581090450286865,
      "pos_sim_mean": 0.6482981443405151,
      "step": 1088
    },
    {
      "epoch": 0.2178,
      "grad_norm": 0.9596202969551086,
      "learning_rate": 9.584604144173083e-05,
      "loss": 0.3693,
      "step": 1089
    },
    {
      "epoch": 0.2178,
      "loss_align": 0.35682952404022217,
      "loss_contrastive": 0.058743178844451904,
      "loss_gen": 0.5376690626144409,
      "neg_sim_mean": 0.6019136309623718,
      "pos_sim_mean": 0.6431704759597778,
      "step": 1089
    },
    {
      "epoch": 0.218,
      "grad_norm": 2.5957493782043457,
      "learning_rate": 9.583210012892582e-05,
      "loss": 0.9015,
      "step": 1090
    },
    {
      "epoch": 0.218,
      "loss_align": 0.37900203466415405,
      "loss_contrastive": 0.16850298643112183,
      "loss_gen": 0.26629507541656494,
      "neg_sim_mean": 0.6895009279251099,
      "pos_sim_mean": 0.620997965335846,
      "step": 1090
    },
    {
      "epoch": 0.2182,
      "grad_norm": 3.0195658206939697,
      "learning_rate": 9.581813647811198e-05,
      "loss": 0.6655,
      "step": 1091
    },
    {
      "epoch": 0.2182,
      "loss_align": 0.2647925019264221,
      "loss_contrastive": 0.13398802280426025,
      "loss_gen": 0.22638073563575745,
      "neg_sim_mean": 0.7691954970359802,
      "pos_sim_mean": 0.7352074980735779,
      "step": 1091
    },
    {
      "epoch": 0.2184,
      "grad_norm": 2.114933967590332,
      "learning_rate": 9.580415049609503e-05,
      "loss": 0.5073,
      "step": 1092
    },
    {
      "epoch": 0.2184,
      "loss_align": 0.44217348098754883,
      "loss_contrastive": 0.18428859114646912,
      "loss_gen": 0.18602876365184784,
      "neg_sim_mean": 0.6421151161193848,
      "pos_sim_mean": 0.5578265190124512,
      "step": 1092
    },
    {
      "epoch": 0.2186,
      "grad_norm": 1.969165325164795,
      "learning_rate": 9.579014218969158e-05,
      "loss": 0.6503,
      "step": 1093
    },
    {
      "epoch": 0.2186,
      "loss_align": 0.2896972894668579,
      "loss_contrastive": 0.12541288137435913,
      "loss_gen": 0.3108312487602234,
      "neg_sim_mean": 0.7357155680656433,
      "pos_sim_mean": 0.7103027105331421,
      "step": 1093
    },
    {
      "epoch": 0.2188,
      "grad_norm": 1.9449167251586914,
      "learning_rate": 9.577611156572908e-05,
      "loss": 0.6156,
      "step": 1094
    },
    {
      "epoch": 0.2188,
      "loss_align": 0.3861861824989319,
      "loss_contrastive": 0.04692584276199341,
      "loss_gen": 0.46311715245246887,
      "neg_sim_mean": 0.5607396364212036,
      "pos_sim_mean": 0.6138138175010681,
      "step": 1094
    },
    {
      "epoch": 0.219,
      "grad_norm": 2.5143158435821533,
      "learning_rate": 9.576205863104588e-05,
      "loss": 0.8549,
      "step": 1095
    },
    {
      "epoch": 0.219,
      "loss_align": 0.3604586124420166,
      "loss_contrastive": 0.17082834243774414,
      "loss_gen": 0.15555796027183533,
      "neg_sim_mean": 0.7103697061538696,
      "pos_sim_mean": 0.6395413875579834,
      "step": 1095
    },
    {
      "epoch": 0.2192,
      "grad_norm": 1.591081976890564,
      "learning_rate": 9.574798339249125e-05,
      "loss": 0.5365,
      "step": 1096
    },
    {
      "epoch": 0.2192,
      "loss_align": 0.35802769660949707,
      "loss_contrastive": 0.11050057411193848,
      "loss_gen": 0.29082340002059937,
      "neg_sim_mean": 0.6524728536605835,
      "pos_sim_mean": 0.6419723033905029,
      "step": 1096
    },
    {
      "epoch": 0.2194,
      "grad_norm": 2.699857473373413,
      "learning_rate": 9.573388585692525e-05,
      "loss": 0.6621,
      "step": 1097
    },
    {
      "epoch": 0.2194,
      "loss_align": 0.1157982349395752,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007044696249067783,
      "neg_sim_mean": 0.40579503774642944,
      "pos_sim_mean": 0.8842017650604248,
      "step": 1097
    },
    {
      "epoch": 0.2196,
      "grad_norm": 0.6110581159591675,
      "learning_rate": 9.571976603121888e-05,
      "loss": 0.1228,
      "step": 1098
    },
    {
      "epoch": 0.2196,
      "loss_align": 0.17077994346618652,
      "loss_contrastive": 0.11601150035858154,
      "loss_gen": 0.22213536500930786,
      "neg_sim_mean": 0.8452315330505371,
      "pos_sim_mean": 0.8292200565338135,
      "step": 1098
    },
    {
      "epoch": 0.2198,
      "grad_norm": 1.6836833953857422,
      "learning_rate": 9.570562392225396e-05,
      "loss": 0.4068,
      "step": 1099
    },
    {
      "epoch": 0.2198,
      "loss_align": 0.28661489486694336,
      "loss_contrastive": 0.0364454984664917,
      "loss_gen": 0.31671226024627686,
      "neg_sim_mean": 0.6498305797576904,
      "pos_sim_mean": 0.7133851051330566,
      "step": 1099
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.9558165073394775,
      "learning_rate": 9.569145953692315e-05,
      "loss": 0.6077,
      "step": 1100
    },
    {
      "epoch": 0.22,
      "loss_align": 0.3405250906944275,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01813654974102974,
      "neg_sim_mean": 0.4767199456691742,
      "pos_sim_mean": 0.6594749093055725,
      "step": 1100
    },
    {
      "epoch": 0.2202,
      "grad_norm": 1.1400787830352783,
      "learning_rate": 9.567727288213005e-05,
      "loss": 0.3587,
      "step": 1101
    },
    {
      "epoch": 0.2202,
      "loss_align": 0.2963863015174866,
      "loss_contrastive": 0.2064303755760193,
      "loss_gen": 0.3723633289337158,
      "neg_sim_mean": 0.8100440502166748,
      "pos_sim_mean": 0.7036136984825134,
      "step": 1101
    },
    {
      "epoch": 0.2204,
      "grad_norm": 2.5038650035858154,
      "learning_rate": 9.566306396478904e-05,
      "loss": 0.6935,
      "step": 1102
    },
    {
      "epoch": 0.2204,
      "loss_align": 0.2897545099258423,
      "loss_contrastive": 0.1461331844329834,
      "loss_gen": 0.24904853105545044,
      "neg_sim_mean": 0.7563786506652832,
      "pos_sim_mean": 0.7102454900741577,
      "step": 1102
    },
    {
      "epoch": 0.2206,
      "grad_norm": 1.926424264907837,
      "learning_rate": 9.564883279182537e-05,
      "loss": 0.5563,
      "step": 1103
    },
    {
      "epoch": 0.2206,
      "loss_align": 0.22597289085388184,
      "loss_contrastive": 0.09109997749328613,
      "loss_gen": 0.35222458839416504,
      "neg_sim_mean": 0.7651270627975464,
      "pos_sim_mean": 0.7740271091461182,
      "step": 1103
    },
    {
      "epoch": 0.2208,
      "grad_norm": 2.39408540725708,
      "learning_rate": 9.563457937017515e-05,
      "loss": 0.5891,
      "step": 1104
    },
    {
      "epoch": 0.2208,
      "loss_align": 0.3609501123428345,
      "loss_contrastive": 0.09019553661346436,
      "loss_gen": 0.29265913367271423,
      "neg_sim_mean": 0.629245400428772,
      "pos_sim_mean": 0.6390498876571655,
      "step": 1104
    },
    {
      "epoch": 0.221,
      "grad_norm": 2.4007091522216797,
      "learning_rate": 9.562030370678532e-05,
      "loss": 0.6644,
      "step": 1105
    },
    {
      "epoch": 0.221,
      "loss_align": 0.2966240644454956,
      "loss_contrastive": 0.006657063961029053,
      "loss_gen": 0.35671642422676086,
      "neg_sim_mean": 0.6100329756736755,
      "pos_sim_mean": 0.7033759355545044,
      "step": 1105
    },
    {
      "epoch": 0.2212,
      "grad_norm": 2.7044708728790283,
      "learning_rate": 9.560600580861365e-05,
      "loss": 0.6541,
      "step": 1106
    },
    {
      "epoch": 0.2212,
      "loss_align": 0.19875967502593994,
      "loss_contrastive": 0.07854419946670532,
      "loss_gen": 0.34353023767471313,
      "neg_sim_mean": 0.7797845005989075,
      "pos_sim_mean": 0.8012403249740601,
      "step": 1106
    },
    {
      "epoch": 0.2214,
      "grad_norm": 2.2235240936279297,
      "learning_rate": 9.55916856826288e-05,
      "loss": 0.5517,
      "step": 1107
    },
    {
      "epoch": 0.2214,
      "loss_align": 0.41086649894714355,
      "loss_contrastive": 0.11010906100273132,
      "loss_gen": 0.23271793127059937,
      "neg_sim_mean": 0.5992425680160522,
      "pos_sim_mean": 0.5891335010528564,
      "step": 1107
    },
    {
      "epoch": 0.2216,
      "grad_norm": 2.3982646465301514,
      "learning_rate": 9.557734333581019e-05,
      "loss": 0.6568,
      "step": 1108
    },
    {
      "epoch": 0.2216,
      "loss_align": 0.44944101572036743,
      "loss_contrastive": 0.07320168614387512,
      "loss_gen": 0.32721975445747375,
      "neg_sim_mean": 0.5237606763839722,
      "pos_sim_mean": 0.5505589842796326,
      "step": 1108
    },
    {
      "epoch": 0.2218,
      "grad_norm": 2.519359588623047,
      "learning_rate": 9.556297877514812e-05,
      "loss": 0.7854,
      "step": 1109
    },
    {
      "epoch": 0.2218,
      "loss_align": 0.4618258476257324,
      "loss_contrastive": 0.16634711623191833,
      "loss_gen": 0.18996231257915497,
      "neg_sim_mean": 0.6045212745666504,
      "pos_sim_mean": 0.5381741523742676,
      "step": 1109
    },
    {
      "epoch": 0.222,
      "grad_norm": 1.8484441041946411,
      "learning_rate": 9.55485920076437e-05,
      "loss": 0.6717,
      "step": 1110
    },
    {
      "epoch": 0.222,
      "loss_align": 0.3014535903930664,
      "loss_contrastive": 0.16961455345153809,
      "loss_gen": 0.3392043113708496,
      "neg_sim_mean": 0.7681609392166138,
      "pos_sim_mean": 0.6985464096069336,
      "step": 1110
    },
    {
      "epoch": 0.2222,
      "grad_norm": 2.342106819152832,
      "learning_rate": 9.553418304030886e-05,
      "loss": 0.661,
      "step": 1111
    },
    {
      "epoch": 0.2222,
      "loss_align": 0.2840961217880249,
      "loss_contrastive": 0.0,
      "loss_gen": 0.018718842417001724,
      "neg_sim_mean": 0.4885326027870178,
      "pos_sim_mean": 0.7159038782119751,
      "step": 1111
    },
    {
      "epoch": 0.2224,
      "grad_norm": 1.130345106124878,
      "learning_rate": 9.551975188016638e-05,
      "loss": 0.3028,
      "step": 1112
    },
    {
      "epoch": 0.2224,
      "loss_align": 0.1403583288192749,
      "loss_contrastive": 0.14349830150604248,
      "loss_gen": 0.34318986535072327,
      "neg_sim_mean": 0.9031399488449097,
      "pos_sim_mean": 0.8596416711807251,
      "step": 1112
    },
    {
      "epoch": 0.2226,
      "grad_norm": 2.193359136581421,
      "learning_rate": 9.550529853424979e-05,
      "loss": 0.5008,
      "step": 1113
    },
    {
      "epoch": 0.2226,
      "loss_align": 0.25312352180480957,
      "loss_contrastive": 0.0,
      "loss_gen": 0.020682917907834053,
      "neg_sim_mean": 0.39680811762809753,
      "pos_sim_mean": 0.7468764781951904,
      "step": 1113
    },
    {
      "epoch": 0.2228,
      "grad_norm": 1.4232479333877563,
      "learning_rate": 9.54908230096035e-05,
      "loss": 0.2738,
      "step": 1114
    },
    {
      "epoch": 0.2228,
      "loss_align": 0.3164339065551758,
      "loss_contrastive": 0.13131266832351685,
      "loss_gen": 0.2940122187137604,
      "neg_sim_mean": 0.7148787379264832,
      "pos_sim_mean": 0.6835660934448242,
      "step": 1114
    },
    {
      "epoch": 0.223,
      "grad_norm": 2.266876220703125,
      "learning_rate": 9.547632531328272e-05,
      "loss": 0.6262,
      "step": 1115
    },
    {
      "epoch": 0.223,
      "loss_align": 0.3088490962982178,
      "loss_contrastive": 0.07625710964202881,
      "loss_gen": 0.19441567361354828,
      "neg_sim_mean": 0.6674079895019531,
      "pos_sim_mean": 0.6911509037017822,
      "step": 1115
    },
    {
      "epoch": 0.2232,
      "grad_norm": 2.1525580883026123,
      "learning_rate": 9.546180545235344e-05,
      "loss": 0.5124,
      "step": 1116
    },
    {
      "epoch": 0.2232,
      "loss_align": 0.2161860466003418,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01812225766479969,
      "neg_sim_mean": 0.4098963141441345,
      "pos_sim_mean": 0.7838139533996582,
      "step": 1116
    },
    {
      "epoch": 0.2234,
      "grad_norm": 1.2812988758087158,
      "learning_rate": 9.544726343389244e-05,
      "loss": 0.2343,
      "step": 1117
    },
    {
      "epoch": 0.2234,
      "loss_align": 0.2877234220504761,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005195552948862314,
      "neg_sim_mean": 0.4309488534927368,
      "pos_sim_mean": 0.7122765779495239,
      "step": 1117
    },
    {
      "epoch": 0.2236,
      "grad_norm": 0.6406148076057434,
      "learning_rate": 9.543269926498735e-05,
      "loss": 0.2929,
      "step": 1118
    },
    {
      "epoch": 0.2236,
      "loss_align": 0.30653810501098633,
      "loss_contrastive": 0.1893274188041687,
      "loss_gen": 0.21794550120830536,
      "neg_sim_mean": 0.7827892899513245,
      "pos_sim_mean": 0.6934618949890137,
      "step": 1118
    },
    {
      "epoch": 0.2238,
      "grad_norm": 1.940883755683899,
      "learning_rate": 9.541811295273656e-05,
      "loss": 0.5472,
      "step": 1119
    },
    {
      "epoch": 0.2238,
      "loss_align": 0.3094162344932556,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007184531539678574,
      "neg_sim_mean": 0.386791855096817,
      "pos_sim_mean": 0.6905837655067444,
      "step": 1119
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.8452802300453186,
      "learning_rate": 9.540350450424926e-05,
      "loss": 0.3166,
      "step": 1120
    },
    {
      "epoch": 0.224,
      "loss_align": 0.373420774936676,
      "loss_contrastive": 0.15550166368484497,
      "loss_gen": 0.2681884169578552,
      "neg_sim_mean": 0.682080864906311,
      "pos_sim_mean": 0.626579225063324,
      "step": 1120
    },
    {
      "epoch": 0.2242,
      "grad_norm": 2.310701370239258,
      "learning_rate": 9.538887392664544e-05,
      "loss": 0.6603,
      "step": 1121
    },
    {
      "epoch": 0.2242,
      "loss_align": 0.19037950038909912,
      "loss_contrastive": 0.10232782363891602,
      "loss_gen": 0.710658609867096,
      "neg_sim_mean": 0.811948299407959,
      "pos_sim_mean": 0.8096204996109009,
      "step": 1121
    },
    {
      "epoch": 0.2244,
      "grad_norm": 3.023073434829712,
      "learning_rate": 9.537422122705585e-05,
      "loss": 0.9133,
      "step": 1122
    },
    {
      "epoch": 0.2244,
      "loss_align": 0.23874413967132568,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0013460644986480474,
      "neg_sim_mean": 0.37928974628448486,
      "pos_sim_mean": 0.7612558603286743,
      "step": 1122
    },
    {
      "epoch": 0.2246,
      "grad_norm": 0.44369325041770935,
      "learning_rate": 9.535954641262206e-05,
      "loss": 0.2401,
      "step": 1123
    },
    {
      "epoch": 0.2246,
      "loss_align": 0.3022177219390869,
      "loss_contrastive": 0.12120628356933594,
      "loss_gen": 0.21770328283309937,
      "neg_sim_mean": 0.7189885377883911,
      "pos_sim_mean": 0.6977822780609131,
      "step": 1123
    },
    {
      "epoch": 0.2248,
      "grad_norm": 2.0480573177337646,
      "learning_rate": 9.534484949049635e-05,
      "loss": 0.5345,
      "step": 1124
    },
    {
      "epoch": 0.2248,
      "loss_align": 0.2939174771308899,
      "loss_contrastive": 0.10850274562835693,
      "loss_gen": 0.31275463104248047,
      "neg_sim_mean": 0.7145852446556091,
      "pos_sim_mean": 0.7060825228691101,
      "step": 1124
    },
    {
      "epoch": 0.225,
      "grad_norm": 2.1360409259796143,
      "learning_rate": 9.533013046784189e-05,
      "loss": 0.6197,
      "step": 1125
    },
    {
      "epoch": 0.225,
      "loss_align": 0.34289681911468506,
      "loss_contrastive": 0.24783861637115479,
      "loss_gen": 0.49986526370048523,
      "neg_sim_mean": 0.8049417734146118,
      "pos_sim_mean": 0.6571031808853149,
      "step": 1125
    },
    {
      "epoch": 0.2252,
      "grad_norm": 2.682520866394043,
      "learning_rate": 9.53153893518325e-05,
      "loss": 0.8725,
      "step": 1126
    },
    {
      "epoch": 0.2252,
      "loss_align": 0.39569389820098877,
      "loss_contrastive": 0.07023334503173828,
      "loss_gen": 0.6172829270362854,
      "neg_sim_mean": 0.5745394229888916,
      "pos_sim_mean": 0.6043061017990112,
      "step": 1126
    },
    {
      "epoch": 0.2254,
      "grad_norm": 3.3143258094787598,
      "learning_rate": 9.530062614965285e-05,
      "loss": 1.0214,
      "step": 1127
    },
    {
      "epoch": 0.2254,
      "loss_align": 0.2529408931732178,
      "loss_contrastive": 0.1315205693244934,
      "loss_gen": 0.3941288888454437,
      "neg_sim_mean": 0.7785796523094177,
      "pos_sim_mean": 0.7470591068267822,
      "step": 1127
    },
    {
      "epoch": 0.2256,
      "grad_norm": 2.5374326705932617,
      "learning_rate": 9.528584086849832e-05,
      "loss": 0.6629,
      "step": 1128
    },
    {
      "epoch": 0.2256,
      "loss_align": 0.15549767017364502,
      "loss_contrastive": 0.0756826400756836,
      "loss_gen": 0.15928034484386444,
      "neg_sim_mean": 0.8201849460601807,
      "pos_sim_mean": 0.844502329826355,
      "step": 1128
    },
    {
      "epoch": 0.2258,
      "grad_norm": 1.4746872186660767,
      "learning_rate": 9.52710335155751e-05,
      "loss": 0.3239,
      "step": 1129
    },
    {
      "epoch": 0.2258,
      "loss_align": 0.36364448070526123,
      "loss_contrastive": 0.05269557237625122,
      "loss_gen": 0.3008453845977783,
      "neg_sim_mean": 0.5890510678291321,
      "pos_sim_mean": 0.6363555192947388,
      "step": 1129
    },
    {
      "epoch": 0.226,
      "grad_norm": 2.1354713439941406,
      "learning_rate": 9.525620409810007e-05,
      "loss": 0.6708,
      "step": 1130
    },
    {
      "epoch": 0.226,
      "loss_align": 0.22498679161071777,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2277303785085678,
      "neg_sim_mean": 0.6327342987060547,
      "pos_sim_mean": 0.7750132083892822,
      "step": 1130
    },
    {
      "epoch": 0.2262,
      "grad_norm": 2.3612427711486816,
      "learning_rate": 9.524135262330098e-05,
      "loss": 0.4527,
      "step": 1131
    },
    {
      "epoch": 0.2262,
      "loss_align": 0.1871001124382019,
      "loss_contrastive": 0.09802192449569702,
      "loss_gen": 0.2511994242668152,
      "neg_sim_mean": 0.8109217882156372,
      "pos_sim_mean": 0.8128998875617981,
      "step": 1131
    },
    {
      "epoch": 0.2264,
      "grad_norm": 2.3108668327331543,
      "learning_rate": 9.522647909841621e-05,
      "loss": 0.4501,
      "step": 1132
    },
    {
      "epoch": 0.2264,
      "loss_align": 0.2446819543838501,
      "loss_contrastive": 0.04303157329559326,
      "loss_gen": 0.37716245651245117,
      "neg_sim_mean": 0.6983495950698853,
      "pos_sim_mean": 0.7553180456161499,
      "step": 1132
    },
    {
      "epoch": 0.2266,
      "grad_norm": 2.3631699085235596,
      "learning_rate": 9.521158353069494e-05,
      "loss": 0.627,
      "step": 1133
    },
    {
      "epoch": 0.2266,
      "loss_align": 0.35988450050354004,
      "loss_contrastive": 0.047985196113586426,
      "loss_gen": 0.013651831075549126,
      "neg_sim_mean": 0.5881006717681885,
      "pos_sim_mean": 0.64011549949646,
      "step": 1133
    },
    {
      "epoch": 0.2268,
      "grad_norm": 1.0662858486175537,
      "learning_rate": 9.519666592739709e-05,
      "loss": 0.3793,
      "step": 1134
    },
    {
      "epoch": 0.2268,
      "loss_align": 0.23093414306640625,
      "loss_contrastive": 0.05857229232788086,
      "loss_gen": 0.36206498742103577,
      "neg_sim_mean": 0.7276381254196167,
      "pos_sim_mean": 0.7690658569335938,
      "step": 1134
    },
    {
      "epoch": 0.227,
      "grad_norm": 2.1638333797454834,
      "learning_rate": 9.518172629579333e-05,
      "loss": 0.6,
      "step": 1135
    },
    {
      "epoch": 0.227,
      "loss_align": 0.2699422836303711,
      "loss_contrastive": 0.11299985647201538,
      "loss_gen": 0.11226208508014679,
      "neg_sim_mean": 0.7430575489997864,
      "pos_sim_mean": 0.7300577163696289,
      "step": 1135
    },
    {
      "epoch": 0.2272,
      "grad_norm": 1.4529064893722534,
      "learning_rate": 9.516676464316505e-05,
      "loss": 0.3958,
      "step": 1136
    },
    {
      "epoch": 0.2272,
      "loss_align": 0.26556968688964844,
      "loss_contrastive": 0.09352320432662964,
      "loss_gen": 0.29824018478393555,
      "neg_sim_mean": 0.7279534935951233,
      "pos_sim_mean": 0.7344303131103516,
      "step": 1136
    },
    {
      "epoch": 0.2274,
      "grad_norm": 2.0982608795166016,
      "learning_rate": 9.515178097680437e-05,
      "loss": 0.575,
      "step": 1137
    },
    {
      "epoch": 0.2274,
      "loss_align": 0.27660417556762695,
      "loss_contrastive": 0.13063925504684448,
      "loss_gen": 0.32095691561698914,
      "neg_sim_mean": 0.7540350556373596,
      "pos_sim_mean": 0.723395824432373,
      "step": 1137
    },
    {
      "epoch": 0.2276,
      "grad_norm": 2.196917772293091,
      "learning_rate": 9.513677530401415e-05,
      "loss": 0.6132,
      "step": 1138
    },
    {
      "epoch": 0.2276,
      "loss_align": 0.16334617137908936,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003559691831469536,
      "neg_sim_mean": 0.4366157054901123,
      "pos_sim_mean": 0.8366538286209106,
      "step": 1138
    },
    {
      "epoch": 0.2278,
      "grad_norm": 0.5155234932899475,
      "learning_rate": 9.512174763210797e-05,
      "loss": 0.1669,
      "step": 1139
    },
    {
      "epoch": 0.2278,
      "loss_align": 0.3890874981880188,
      "loss_contrastive": 0.15033107995986938,
      "loss_gen": 0.2987753748893738,
      "neg_sim_mean": 0.6612435579299927,
      "pos_sim_mean": 0.6109125018119812,
      "step": 1139
    },
    {
      "epoch": 0.228,
      "grad_norm": 2.054464340209961,
      "learning_rate": 9.510669796841014e-05,
      "loss": 0.7059,
      "step": 1140
    },
    {
      "epoch": 0.228,
      "loss_align": 0.27735447883605957,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2358180582523346,
      "neg_sim_mean": 0.5952112674713135,
      "pos_sim_mean": 0.7226455211639404,
      "step": 1140
    },
    {
      "epoch": 0.2282,
      "grad_norm": 2.0276989936828613,
      "learning_rate": 9.50916263202557e-05,
      "loss": 0.5132,
      "step": 1141
    },
    {
      "epoch": 0.2282,
      "loss_align": 0.24394738674163818,
      "loss_contrastive": 0.07354491949081421,
      "loss_gen": 0.20468825101852417,
      "neg_sim_mean": 0.7295975089073181,
      "pos_sim_mean": 0.7560526132583618,
      "step": 1141
    },
    {
      "epoch": 0.2284,
      "grad_norm": 1.7448235750198364,
      "learning_rate": 9.507653269499035e-05,
      "loss": 0.4575,
      "step": 1142
    },
    {
      "epoch": 0.2284,
      "loss_align": 0.22382187843322754,
      "loss_contrastive": 0.05060982704162598,
      "loss_gen": 0.33077800273895264,
      "neg_sim_mean": 0.7267879247665405,
      "pos_sim_mean": 0.7761781215667725,
      "step": 1142
    },
    {
      "epoch": 0.2286,
      "grad_norm": 2.162210464477539,
      "learning_rate": 9.506141709997057e-05,
      "loss": 0.5607,
      "step": 1143
    },
    {
      "epoch": 0.2286,
      "loss_align": 0.10386621952056885,
      "loss_contrastive": 0.0,
      "loss_gen": 0.297100305557251,
      "neg_sim_mean": 0.7695891261100769,
      "pos_sim_mean": 0.8961337804794312,
      "step": 1143
    },
    {
      "epoch": 0.2288,
      "grad_norm": 2.1028995513916016,
      "learning_rate": 9.50462795425635e-05,
      "loss": 0.401,
      "step": 1144
    },
    {
      "epoch": 0.2288,
      "loss_align": 0.20910561084747314,
      "loss_contrastive": 0.07208293676376343,
      "loss_gen": 0.3687742352485657,
      "neg_sim_mean": 0.7629773020744324,
      "pos_sim_mean": 0.7908943891525269,
      "step": 1144
    },
    {
      "epoch": 0.229,
      "grad_norm": 2.630213737487793,
      "learning_rate": 9.503112003014703e-05,
      "loss": 0.5865,
      "step": 1145
    },
    {
      "epoch": 0.229,
      "loss_align": 0.15776342153549194,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2215508371591568,
      "neg_sim_mean": 0.7304009199142456,
      "pos_sim_mean": 0.8422365784645081,
      "step": 1145
    },
    {
      "epoch": 0.2292,
      "grad_norm": 1.4599227905273438,
      "learning_rate": 9.501593857010969e-05,
      "loss": 0.3793,
      "step": 1146
    },
    {
      "epoch": 0.2292,
      "loss_align": 0.3356708288192749,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0062903487123548985,
      "neg_sim_mean": 0.4330170750617981,
      "pos_sim_mean": 0.6643291711807251,
      "step": 1146
    },
    {
      "epoch": 0.2294,
      "grad_norm": 0.7295498251914978,
      "learning_rate": 9.500073516985074e-05,
      "loss": 0.342,
      "step": 1147
    },
    {
      "epoch": 0.2294,
      "loss_align": 0.21786445379257202,
      "loss_contrastive": 0.03386193513870239,
      "loss_gen": 0.3327307105064392,
      "neg_sim_mean": 0.7159974575042725,
      "pos_sim_mean": 0.782135546207428,
      "step": 1147
    },
    {
      "epoch": 0.2296,
      "grad_norm": 2.8367974758148193,
      "learning_rate": 9.498550983678015e-05,
      "loss": 0.5547,
      "step": 1148
    },
    {
      "epoch": 0.2296,
      "loss_align": 0.29904842376708984,
      "loss_contrastive": 0.04703116416931152,
      "loss_gen": 0.47675833106040955,
      "neg_sim_mean": 0.6479827165603638,
      "pos_sim_mean": 0.7009515762329102,
      "step": 1148
    },
    {
      "epoch": 0.2298,
      "grad_norm": 2.84675669670105,
      "learning_rate": 9.497026257831855e-05,
      "loss": 0.7815,
      "step": 1149
    },
    {
      "epoch": 0.2298,
      "loss_align": 0.4072531461715698,
      "loss_contrastive": 0.20633432269096375,
      "loss_gen": 0.31079256534576416,
      "neg_sim_mean": 0.6990811824798584,
      "pos_sim_mean": 0.5927468538284302,
      "step": 1149
    },
    {
      "epoch": 0.23,
      "grad_norm": 11.042607307434082,
      "learning_rate": 9.495499340189728e-05,
      "loss": 0.7428,
      "step": 1150
    },
    {
      "epoch": 0.23,
      "loss_align": 0.3924606442451477,
      "loss_contrastive": 0.20005923509597778,
      "loss_gen": 0.29700392484664917,
      "neg_sim_mean": 0.7075985670089722,
      "pos_sim_mean": 0.6075393557548523,
      "step": 1150
    },
    {
      "epoch": 0.2302,
      "grad_norm": 3.186835289001465,
      "learning_rate": 9.493970231495835e-05,
      "loss": 0.7135,
      "step": 1151
    },
    {
      "epoch": 0.2302,
      "loss_align": 0.3275444507598877,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03863540291786194,
      "neg_sim_mean": 0.4087068736553192,
      "pos_sim_mean": 0.6724555492401123,
      "step": 1151
    },
    {
      "epoch": 0.2304,
      "grad_norm": 1.7321935892105103,
      "learning_rate": 9.492438932495444e-05,
      "loss": 0.3662,
      "step": 1152
    },
    {
      "epoch": 0.2304,
      "loss_align": 0.2955175042152405,
      "loss_contrastive": 0.0,
      "loss_gen": 0.037279121577739716,
      "neg_sim_mean": 0.31440916657447815,
      "pos_sim_mean": 0.7044824957847595,
      "step": 1152
    },
    {
      "epoch": 0.2306,
      "grad_norm": 2.321319103240967,
      "learning_rate": 9.490905443934893e-05,
      "loss": 0.3328,
      "step": 1153
    },
    {
      "epoch": 0.2306,
      "loss_align": 0.40972113609313965,
      "loss_contrastive": 0.23533371090888977,
      "loss_gen": 0.16131019592285156,
      "neg_sim_mean": 0.7256125807762146,
      "pos_sim_mean": 0.5902788639068604,
      "step": 1153
    },
    {
      "epoch": 0.2308,
      "grad_norm": 1.9649556875228882,
      "learning_rate": 9.489369766561584e-05,
      "loss": 0.5993,
      "step": 1154
    },
    {
      "epoch": 0.2308,
      "loss_align": 0.2232953906059265,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26692044734954834,
      "neg_sim_mean": 0.6355262398719788,
      "pos_sim_mean": 0.7767046093940735,
      "step": 1154
    },
    {
      "epoch": 0.231,
      "grad_norm": 2.189997434616089,
      "learning_rate": 9.487831901123988e-05,
      "loss": 0.4902,
      "step": 1155
    },
    {
      "epoch": 0.231,
      "loss_align": 0.3039062023162842,
      "loss_contrastive": 0.20059841871261597,
      "loss_gen": 0.24697129428386688,
      "neg_sim_mean": 0.7966921925544739,
      "pos_sim_mean": 0.6960937976837158,
      "step": 1155
    },
    {
      "epoch": 0.2312,
      "grad_norm": 2.130819797515869,
      "learning_rate": 9.486291848371643e-05,
      "loss": 0.5749,
      "step": 1156
    },
    {
      "epoch": 0.2312,
      "loss_align": 0.32861846685409546,
      "loss_contrastive": 0.19906002283096313,
      "loss_gen": 0.3205695152282715,
      "neg_sim_mean": 0.7704415321350098,
      "pos_sim_mean": 0.6713815331459045,
      "step": 1156
    },
    {
      "epoch": 0.2314,
      "grad_norm": 2.449655771255493,
      "learning_rate": 9.48474960905515e-05,
      "loss": 0.6731,
      "step": 1157
    },
    {
      "epoch": 0.2314,
      "loss_align": 0.27029740810394287,
      "loss_contrastive": 0.01679295301437378,
      "loss_gen": 0.4034092426300049,
      "neg_sim_mean": 0.646495521068573,
      "pos_sim_mean": 0.7297025918960571,
      "step": 1157
    },
    {
      "epoch": 0.2316,
      "grad_norm": 2.680406332015991,
      "learning_rate": 9.483205183926181e-05,
      "loss": 0.6757,
      "step": 1158
    },
    {
      "epoch": 0.2316,
      "loss_align": 0.4375948905944824,
      "loss_contrastive": 0.018341928720474243,
      "loss_gen": 0.01226150430738926,
      "neg_sim_mean": 0.4807470440864563,
      "pos_sim_mean": 0.5624051094055176,
      "step": 1158
    },
    {
      "epoch": 0.2318,
      "grad_norm": 1.0025745630264282,
      "learning_rate": 9.481658573737465e-05,
      "loss": 0.4521,
      "step": 1159
    },
    {
      "epoch": 0.2318,
      "loss_align": 0.34415721893310547,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012810802087187767,
      "neg_sim_mean": 0.376905620098114,
      "pos_sim_mean": 0.6558427810668945,
      "step": 1159
    },
    {
      "epoch": 0.232,
      "grad_norm": 1.4682114124298096,
      "learning_rate": 9.480109779242805e-05,
      "loss": 0.357,
      "step": 1160
    },
    {
      "epoch": 0.232,
      "loss_align": 0.4745975732803345,
      "loss_contrastive": 0.04933050274848938,
      "loss_gen": 0.0017793813021853566,
      "neg_sim_mean": 0.4747329354286194,
      "pos_sim_mean": 0.5254024267196655,
      "step": 1160
    },
    {
      "epoch": 0.2322,
      "grad_norm": 0.7924551963806152,
      "learning_rate": 9.478558801197065e-05,
      "loss": 0.4823,
      "step": 1161
    },
    {
      "epoch": 0.2322,
      "loss_align": 0.22933846712112427,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2656950354576111,
      "neg_sim_mean": 0.6706569790840149,
      "pos_sim_mean": 0.7706615328788757,
      "step": 1161
    },
    {
      "epoch": 0.2324,
      "grad_norm": 2.010047435760498,
      "learning_rate": 9.47700564035617e-05,
      "loss": 0.495,
      "step": 1162
    },
    {
      "epoch": 0.2324,
      "loss_align": 0.36367809772491455,
      "loss_contrastive": 0.07969427108764648,
      "loss_gen": 0.32467564940452576,
      "neg_sim_mean": 0.616016149520874,
      "pos_sim_mean": 0.6363219022750854,
      "step": 1162
    },
    {
      "epoch": 0.2326,
      "grad_norm": 2.357328414916992,
      "learning_rate": 9.475450297477113e-05,
      "loss": 0.6979,
      "step": 1163
    },
    {
      "epoch": 0.2326,
      "loss_align": 0.31500232219696045,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3466358780860901,
      "neg_sim_mean": 0.5525594353675842,
      "pos_sim_mean": 0.6849976778030396,
      "step": 1163
    },
    {
      "epoch": 0.2328,
      "grad_norm": 2.467040777206421,
      "learning_rate": 9.473892773317952e-05,
      "loss": 0.6616,
      "step": 1164
    },
    {
      "epoch": 0.2328,
      "loss_align": 0.16306602954864502,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00622619828209281,
      "neg_sim_mean": 0.36307603120803833,
      "pos_sim_mean": 0.836933970451355,
      "step": 1164
    },
    {
      "epoch": 0.233,
      "grad_norm": 0.6646591424942017,
      "learning_rate": 9.4723330686378e-05,
      "loss": 0.1693,
      "step": 1165
    },
    {
      "epoch": 0.233,
      "loss_align": 0.26219117641448975,
      "loss_contrastive": 0.09840065240859985,
      "loss_gen": 0.20388810336589813,
      "neg_sim_mean": 0.7362094521522522,
      "pos_sim_mean": 0.7378088235855103,
      "step": 1165
    },
    {
      "epoch": 0.2332,
      "grad_norm": 1.7097710371017456,
      "learning_rate": 9.47077118419684e-05,
      "loss": 0.4779,
      "step": 1166
    },
    {
      "epoch": 0.2332,
      "loss_align": 0.2566859722137451,
      "loss_contrastive": 0.10826992988586426,
      "loss_gen": 0.24714629352092743,
      "neg_sim_mean": 0.7515839338302612,
      "pos_sim_mean": 0.7433140277862549,
      "step": 1166
    },
    {
      "epoch": 0.2334,
      "grad_norm": 2.068209171295166,
      "learning_rate": 9.46920712075632e-05,
      "loss": 0.5168,
      "step": 1167
    },
    {
      "epoch": 0.2334,
      "loss_align": 0.3527873158454895,
      "loss_contrastive": 0.05697554349899292,
      "loss_gen": 0.36801835894584656,
      "neg_sim_mean": 0.6041882038116455,
      "pos_sim_mean": 0.6472126841545105,
      "step": 1167
    },
    {
      "epoch": 0.2336,
      "grad_norm": 2.1226260662078857,
      "learning_rate": 9.467640879078539e-05,
      "loss": 0.7276,
      "step": 1168
    },
    {
      "epoch": 0.2336,
      "loss_align": 0.29767417907714844,
      "loss_contrastive": 0.23660582304000854,
      "loss_gen": 0.27196818590164185,
      "neg_sim_mean": 0.8389316201210022,
      "pos_sim_mean": 0.7023258209228516,
      "step": 1168
    },
    {
      "epoch": 0.2338,
      "grad_norm": 2.249723196029663,
      "learning_rate": 9.46607245992687e-05,
      "loss": 0.598,
      "step": 1169
    },
    {
      "epoch": 0.2338,
      "loss_align": 0.27133089303970337,
      "loss_contrastive": 0.11497330665588379,
      "loss_gen": 0.1601642668247223,
      "neg_sim_mean": 0.7436423897743225,
      "pos_sim_mean": 0.7286691069602966,
      "step": 1169
    },
    {
      "epoch": 0.234,
      "grad_norm": 2.356693983078003,
      "learning_rate": 9.464501864065735e-05,
      "loss": 0.4453,
      "step": 1170
    },
    {
      "epoch": 0.234,
      "loss_align": 0.25449979305267334,
      "loss_contrastive": 0.15797960758209229,
      "loss_gen": 0.19966112077236176,
      "neg_sim_mean": 0.803479790687561,
      "pos_sim_mean": 0.7455002069473267,
      "step": 1170
    },
    {
      "epoch": 0.2342,
      "grad_norm": 1.7434662580490112,
      "learning_rate": 9.462929092260628e-05,
      "loss": 0.4731,
      "step": 1171
    },
    {
      "epoch": 0.2342,
      "loss_align": 0.3822232484817505,
      "loss_contrastive": 0.0890728235244751,
      "loss_gen": 0.2582884132862091,
      "neg_sim_mean": 0.6068495512008667,
      "pos_sim_mean": 0.6177767515182495,
      "step": 1171
    },
    {
      "epoch": 0.2344,
      "grad_norm": 1.7987017631530762,
      "learning_rate": 9.461354145278099e-05,
      "loss": 0.6512,
      "step": 1172
    },
    {
      "epoch": 0.2344,
      "loss_align": 0.3659181594848633,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00497911823913455,
      "neg_sim_mean": 0.4030311703681946,
      "pos_sim_mean": 0.6340818405151367,
      "step": 1172
    },
    {
      "epoch": 0.2346,
      "grad_norm": 0.6648158431053162,
      "learning_rate": 9.459777023885755e-05,
      "loss": 0.3709,
      "step": 1173
    },
    {
      "epoch": 0.2346,
      "loss_align": 0.21590590476989746,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0028843667823821306,
      "neg_sim_mean": 0.26021987199783325,
      "pos_sim_mean": 0.7840940952301025,
      "step": 1173
    },
    {
      "epoch": 0.2348,
      "grad_norm": 0.48294171690940857,
      "learning_rate": 9.458197728852267e-05,
      "loss": 0.2188,
      "step": 1174
    },
    {
      "epoch": 0.2348,
      "loss_align": 0.2851905822753906,
      "loss_contrastive": 0.07581484317779541,
      "loss_gen": 0.4094771444797516,
      "neg_sim_mean": 0.6906242370605469,
      "pos_sim_mean": 0.7148094177246094,
      "step": 1174
    },
    {
      "epoch": 0.235,
      "grad_norm": 2.4139318466186523,
      "learning_rate": 9.456616260947366e-05,
      "loss": 0.7038,
      "step": 1175
    },
    {
      "epoch": 0.235,
      "loss_align": 0.15625643730163574,
      "loss_contrastive": 0.022620975971221924,
      "loss_gen": 0.2380080670118332,
      "neg_sim_mean": 0.7663645148277283,
      "pos_sim_mean": 0.8437435626983643,
      "step": 1175
    },
    {
      "epoch": 0.2352,
      "grad_norm": 2.058068037033081,
      "learning_rate": 9.45503262094184e-05,
      "loss": 0.397,
      "step": 1176
    },
    {
      "epoch": 0.2352,
      "loss_align": 0.3072502613067627,
      "loss_contrastive": 0.1633002758026123,
      "loss_gen": 0.23858053982257843,
      "neg_sim_mean": 0.7560499906539917,
      "pos_sim_mean": 0.6927497386932373,
      "step": 1176
    },
    {
      "epoch": 0.2354,
      "grad_norm": 2.40171217918396,
      "learning_rate": 9.453446809607533e-05,
      "loss": 0.5654,
      "step": 1177
    },
    {
      "epoch": 0.2354,
      "loss_align": 0.20530426502227783,
      "loss_contrastive": 0.07371193170547485,
      "loss_gen": 0.1650068759918213,
      "neg_sim_mean": 0.7684076428413391,
      "pos_sim_mean": 0.7946957349777222,
      "step": 1177
    },
    {
      "epoch": 0.2356,
      "grad_norm": 1.939557671546936,
      "learning_rate": 9.451858827717353e-05,
      "loss": 0.3792,
      "step": 1178
    },
    {
      "epoch": 0.2356,
      "loss_align": 0.1500687599182129,
      "loss_contrastive": 0.13292992115020752,
      "loss_gen": 0.49196943640708923,
      "neg_sim_mean": 0.8828611373901367,
      "pos_sim_mean": 0.8499312400817871,
      "step": 1178
    },
    {
      "epoch": 0.2358,
      "grad_norm": 2.431389331817627,
      "learning_rate": 9.450268676045262e-05,
      "loss": 0.658,
      "step": 1179
    },
    {
      "epoch": 0.2358,
      "loss_align": 0.26070868968963623,
      "loss_contrastive": 0.10268831253051758,
      "loss_gen": 0.30102473497390747,
      "neg_sim_mean": 0.7419795989990234,
      "pos_sim_mean": 0.7392913103103638,
      "step": 1179
    },
    {
      "epoch": 0.236,
      "grad_norm": 2.260953187942505,
      "learning_rate": 9.448676355366282e-05,
      "loss": 0.5741,
      "step": 1180
    },
    {
      "epoch": 0.236,
      "loss_align": 0.2462371587753296,
      "loss_contrastive": 0.15167534351348877,
      "loss_gen": 0.3037232756614685,
      "neg_sim_mean": 0.8054381608963013,
      "pos_sim_mean": 0.7537628412246704,
      "step": 1180
    },
    {
      "epoch": 0.2362,
      "grad_norm": 2.0495269298553467,
      "learning_rate": 9.447081866456489e-05,
      "loss": 0.5682,
      "step": 1181
    },
    {
      "epoch": 0.2362,
      "loss_align": 0.42268961668014526,
      "loss_contrastive": 0.009555190801620483,
      "loss_gen": 0.27707812190055847,
      "neg_sim_mean": 0.4868655800819397,
      "pos_sim_mean": 0.5773103833198547,
      "step": 1181
    },
    {
      "epoch": 0.2364,
      "grad_norm": 2.1811439990997314,
      "learning_rate": 9.445485210093017e-05,
      "loss": 0.7009,
      "step": 1182
    },
    {
      "epoch": 0.2364,
      "loss_align": 0.30276238918304443,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002984747989103198,
      "neg_sim_mean": 0.37942177057266235,
      "pos_sim_mean": 0.6972376108169556,
      "step": 1182
    },
    {
      "epoch": 0.2366,
      "grad_norm": 0.8576866388320923,
      "learning_rate": 9.443886387054058e-05,
      "loss": 0.3057,
      "step": 1183
    },
    {
      "epoch": 0.2366,
      "loss_align": 0.40227067470550537,
      "loss_contrastive": 0.2546598017215729,
      "loss_gen": 0.20998020470142365,
      "neg_sim_mean": 0.752389132976532,
      "pos_sim_mean": 0.5977293252944946,
      "step": 1183
    },
    {
      "epoch": 0.2368,
      "grad_norm": 2.319758653640747,
      "learning_rate": 9.442285398118859e-05,
      "loss": 0.6428,
      "step": 1184
    },
    {
      "epoch": 0.2368,
      "loss_align": 0.18990755081176758,
      "loss_contrastive": 0.06318902969360352,
      "loss_gen": 0.366312712430954,
      "neg_sim_mean": 0.773281455039978,
      "pos_sim_mean": 0.8100924491882324,
      "step": 1184
    },
    {
      "epoch": 0.237,
      "grad_norm": 2.6827149391174316,
      "learning_rate": 9.440682244067724e-05,
      "loss": 0.5638,
      "step": 1185
    },
    {
      "epoch": 0.237,
      "loss_align": 0.40951716899871826,
      "loss_contrastive": 0.13865545392036438,
      "loss_gen": 0.18016991019248962,
      "neg_sim_mean": 0.6291382908821106,
      "pos_sim_mean": 0.5904828310012817,
      "step": 1185
    },
    {
      "epoch": 0.2372,
      "grad_norm": 1.8026264905929565,
      "learning_rate": 9.439076925682006e-05,
      "loss": 0.6063,
      "step": 1186
    },
    {
      "epoch": 0.2372,
      "loss_align": 0.1740201711654663,
      "loss_contrastive": 0.019981741905212402,
      "loss_gen": 0.17839393019676208,
      "neg_sim_mean": 0.7459615468978882,
      "pos_sim_mean": 0.8259798288345337,
      "step": 1186
    },
    {
      "epoch": 0.2374,
      "grad_norm": 2.086522340774536,
      "learning_rate": 9.437469443744124e-05,
      "loss": 0.3548,
      "step": 1187
    },
    {
      "epoch": 0.2374,
      "loss_align": 0.3173861503601074,
      "loss_contrastive": 0.15474039316177368,
      "loss_gen": 0.23541101813316345,
      "neg_sim_mean": 0.7373542189598083,
      "pos_sim_mean": 0.6826138496398926,
      "step": 1187
    },
    {
      "epoch": 0.2376,
      "grad_norm": 1.9684669971466064,
      "learning_rate": 9.43585979903754e-05,
      "loss": 0.5714,
      "step": 1188
    },
    {
      "epoch": 0.2376,
      "loss_align": 0.31821560859680176,
      "loss_contrastive": 0.048602283000946045,
      "loss_gen": 0.20445775985717773,
      "neg_sim_mean": 0.6303866505622864,
      "pos_sim_mean": 0.6817843914031982,
      "step": 1188
    },
    {
      "epoch": 0.2378,
      "grad_norm": 1.9286285638809204,
      "learning_rate": 9.43424799234678e-05,
      "loss": 0.5285,
      "step": 1189
    },
    {
      "epoch": 0.2378,
      "loss_align": 0.18153929710388184,
      "loss_contrastive": 0.11084878444671631,
      "loss_gen": 0.1583927720785141,
      "neg_sim_mean": 0.8293094635009766,
      "pos_sim_mean": 0.8184607028961182,
      "step": 1189
    },
    {
      "epoch": 0.238,
      "grad_norm": 1.9079320430755615,
      "learning_rate": 9.432634024457414e-05,
      "loss": 0.3532,
      "step": 1190
    },
    {
      "epoch": 0.238,
      "loss_align": 0.3492215871810913,
      "loss_contrastive": 0.10577881336212158,
      "loss_gen": 0.11137869954109192,
      "neg_sim_mean": 0.6565572023391724,
      "pos_sim_mean": 0.6507784128189087,
      "step": 1190
    },
    {
      "epoch": 0.2382,
      "grad_norm": 1.9723923206329346,
      "learning_rate": 9.431017896156074e-05,
      "loss": 0.4733,
      "step": 1191
    },
    {
      "epoch": 0.2382,
      "loss_align": 0.25648874044418335,
      "loss_contrastive": 0.03811460733413696,
      "loss_gen": 0.22695890069007874,
      "neg_sim_mean": 0.6816258430480957,
      "pos_sim_mean": 0.7435112595558167,
      "step": 1191
    },
    {
      "epoch": 0.2384,
      "grad_norm": 2.1276535987854004,
      "learning_rate": 9.429399608230441e-05,
      "loss": 0.488,
      "step": 1192
    },
    {
      "epoch": 0.2384,
      "loss_align": 0.3426218032836914,
      "loss_contrastive": 0.18270504474639893,
      "loss_gen": 0.25395986437797546,
      "neg_sim_mean": 0.7400832176208496,
      "pos_sim_mean": 0.6573781967163086,
      "step": 1192
    },
    {
      "epoch": 0.2386,
      "grad_norm": 2.1733551025390625,
      "learning_rate": 9.427779161469246e-05,
      "loss": 0.6185,
      "step": 1193
    },
    {
      "epoch": 0.2386,
      "loss_align": 0.15832734107971191,
      "loss_contrastive": 0.10473167896270752,
      "loss_gen": 0.22762960195541382,
      "neg_sim_mean": 0.8464043140411377,
      "pos_sim_mean": 0.8416726589202881,
      "step": 1193
    },
    {
      "epoch": 0.2388,
      "grad_norm": 2.0783498287200928,
      "learning_rate": 9.426156556662276e-05,
      "loss": 0.3985,
      "step": 1194
    },
    {
      "epoch": 0.2388,
      "loss_align": 0.23831629753112793,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02988409623503685,
      "neg_sim_mean": 0.4672381281852722,
      "pos_sim_mean": 0.7616837024688721,
      "step": 1194
    },
    {
      "epoch": 0.239,
      "grad_norm": 1.313643217086792,
      "learning_rate": 9.424531794600372e-05,
      "loss": 0.2682,
      "step": 1195
    },
    {
      "epoch": 0.239,
      "loss_align": 0.2886682152748108,
      "loss_contrastive": 0.1353958249092102,
      "loss_gen": 0.3005959093570709,
      "neg_sim_mean": 0.7467275857925415,
      "pos_sim_mean": 0.7113317847251892,
      "step": 1195
    },
    {
      "epoch": 0.2392,
      "grad_norm": 2.8091330528259277,
      "learning_rate": 9.42290487607542e-05,
      "loss": 0.6055,
      "step": 1196
    },
    {
      "epoch": 0.2392,
      "loss_align": 0.258292555809021,
      "loss_contrastive": 0.11897432804107666,
      "loss_gen": 0.1752372533082962,
      "neg_sim_mean": 0.7606817483901978,
      "pos_sim_mean": 0.741707444190979,
      "step": 1196
    },
    {
      "epoch": 0.2394,
      "grad_norm": 1.6500059366226196,
      "learning_rate": 9.421275801880362e-05,
      "loss": 0.4478,
      "step": 1197
    },
    {
      "epoch": 0.2394,
      "loss_align": 0.2901543378829956,
      "loss_contrastive": 0.08106666803359985,
      "loss_gen": 0.3005558252334595,
      "neg_sim_mean": 0.6909123063087463,
      "pos_sim_mean": 0.7098456621170044,
      "step": 1197
    },
    {
      "epoch": 0.2396,
      "grad_norm": 2.3786463737487793,
      "learning_rate": 9.419644572809189e-05,
      "loss": 0.6004,
      "step": 1198
    },
    {
      "epoch": 0.2396,
      "loss_align": 0.22598761320114136,
      "loss_contrastive": 0.09181803464889526,
      "loss_gen": 0.2790994644165039,
      "neg_sim_mean": 0.765830397605896,
      "pos_sim_mean": 0.7740123867988586,
      "step": 1198
    },
    {
      "epoch": 0.2398,
      "grad_norm": 1.757612943649292,
      "learning_rate": 9.418011189656941e-05,
      "loss": 0.5161,
      "step": 1199
    },
    {
      "epoch": 0.2398,
      "loss_align": 0.3369225263595581,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03913811594247818,
      "neg_sim_mean": 0.35048729181289673,
      "pos_sim_mean": 0.6630774736404419,
      "step": 1199
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7052149772644043,
      "learning_rate": 9.416375653219709e-05,
      "loss": 0.3761,
      "step": 1200
    },
    {
      "epoch": 0.24,
      "loss_align": 0.311755895614624,
      "loss_contrastive": 0.09700000286102295,
      "loss_gen": 0.38943660259246826,
      "neg_sim_mean": 0.685244083404541,
      "pos_sim_mean": 0.688244104385376,
      "step": 1200
    },
    {
      "epoch": 0.2402,
      "grad_norm": 2.5513429641723633,
      "learning_rate": 9.414737964294636e-05,
      "loss": 0.7128,
      "step": 1201
    },
    {
      "epoch": 0.2402,
      "loss_align": 0.40048402547836304,
      "loss_contrastive": 0.3169701397418976,
      "loss_gen": 0.27118951082229614,
      "neg_sim_mean": 0.816486120223999,
      "pos_sim_mean": 0.599515974521637,
      "step": 1201
    },
    {
      "epoch": 0.2404,
      "grad_norm": 2.0705506801605225,
      "learning_rate": 9.413098123679908e-05,
      "loss": 0.7097,
      "step": 1202
    },
    {
      "epoch": 0.2404,
      "loss_align": 0.350716769695282,
      "loss_contrastive": 0.15003269910812378,
      "loss_gen": 0.24604938924312592,
      "neg_sim_mean": 0.6993159055709839,
      "pos_sim_mean": 0.649283230304718,
      "step": 1202
    },
    {
      "epoch": 0.2406,
      "grad_norm": 1.9969526529312134,
      "learning_rate": 9.411456132174767e-05,
      "loss": 0.6148,
      "step": 1203
    },
    {
      "epoch": 0.2406,
      "loss_align": 0.23030364513397217,
      "loss_contrastive": 0.08333081007003784,
      "loss_gen": 0.2703934609889984,
      "neg_sim_mean": 0.7530271410942078,
      "pos_sim_mean": 0.7696963548660278,
      "step": 1203
    },
    {
      "epoch": 0.2408,
      "grad_norm": 2.309121608734131,
      "learning_rate": 9.409811990579498e-05,
      "loss": 0.5107,
      "step": 1204
    },
    {
      "epoch": 0.2408,
      "loss_align": 0.35059064626693726,
      "loss_contrastive": 0.09798198938369751,
      "loss_gen": 0.24977828562259674,
      "neg_sim_mean": 0.6473913192749023,
      "pos_sim_mean": 0.6494093537330627,
      "step": 1204
    },
    {
      "epoch": 0.241,
      "grad_norm": 2.0936102867126465,
      "learning_rate": 9.408165699695435e-05,
      "loss": 0.6121,
      "step": 1205
    },
    {
      "epoch": 0.241,
      "loss_align": 0.25156646966934204,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009427805431187153,
      "neg_sim_mean": 0.36947154998779297,
      "pos_sim_mean": 0.748433530330658,
      "step": 1205
    },
    {
      "epoch": 0.2412,
      "grad_norm": 1.3324685096740723,
      "learning_rate": 9.40651726032496e-05,
      "loss": 0.261,
      "step": 1206
    },
    {
      "epoch": 0.2412,
      "loss_align": 0.19474709033966064,
      "loss_contrastive": 0.0479469895362854,
      "loss_gen": 0.372905969619751,
      "neg_sim_mean": 0.7531998753547668,
      "pos_sim_mean": 0.8052529096603394,
      "step": 1206
    },
    {
      "epoch": 0.2414,
      "grad_norm": 2.192831516265869,
      "learning_rate": 9.404866673271506e-05,
      "loss": 0.5734,
      "step": 1207
    },
    {
      "epoch": 0.2414,
      "loss_align": 0.3767378330230713,
      "loss_contrastive": 0.09070682525634766,
      "loss_gen": 0.012409725226461887,
      "neg_sim_mean": 0.6139689683914185,
      "pos_sim_mean": 0.6232621669769287,
      "step": 1207
    },
    {
      "epoch": 0.2416,
      "grad_norm": 1.0204719305038452,
      "learning_rate": 9.403213939339546e-05,
      "loss": 0.4,
      "step": 1208
    },
    {
      "epoch": 0.2416,
      "loss_align": 0.24745118618011475,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0039870827458798885,
      "neg_sim_mean": 0.5361372828483582,
      "pos_sim_mean": 0.7525488138198853,
      "step": 1208
    },
    {
      "epoch": 0.2418,
      "grad_norm": 0.7153217792510986,
      "learning_rate": 9.401559059334601e-05,
      "loss": 0.2514,
      "step": 1209
    },
    {
      "epoch": 0.2418,
      "loss_align": 0.28648966550827026,
      "loss_contrastive": 0.12263661623001099,
      "loss_gen": 0.1841481477022171,
      "neg_sim_mean": 0.7361469268798828,
      "pos_sim_mean": 0.7135103344917297,
      "step": 1209
    },
    {
      "epoch": 0.242,
      "grad_norm": 1.9094318151474,
      "learning_rate": 9.399902034063242e-05,
      "loss": 0.4854,
      "step": 1210
    },
    {
      "epoch": 0.242,
      "loss_align": 0.20650523900985718,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0060438611544668674,
      "neg_sim_mean": 0.44064855575561523,
      "pos_sim_mean": 0.7934947609901428,
      "step": 1210
    },
    {
      "epoch": 0.2422,
      "grad_norm": 0.6406786441802979,
      "learning_rate": 9.398242864333083e-05,
      "loss": 0.2125,
      "step": 1211
    },
    {
      "epoch": 0.2422,
      "loss_align": 0.2307220697402954,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01848321408033371,
      "neg_sim_mean": 0.48757365345954895,
      "pos_sim_mean": 0.7692779302597046,
      "step": 1211
    },
    {
      "epoch": 0.2424,
      "grad_norm": 1.2071224451065063,
      "learning_rate": 9.396581550952781e-05,
      "loss": 0.2492,
      "step": 1212
    },
    {
      "epoch": 0.2424,
      "loss_align": 0.24709182977676392,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3427548110485077,
      "neg_sim_mean": 0.6476860046386719,
      "pos_sim_mean": 0.7529081702232361,
      "step": 1212
    },
    {
      "epoch": 0.2426,
      "grad_norm": 2.4928438663482666,
      "learning_rate": 9.394918094732043e-05,
      "loss": 0.5898,
      "step": 1213
    },
    {
      "epoch": 0.2426,
      "loss_align": 0.3733471632003784,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00318972859531641,
      "neg_sim_mean": 0.4983869194984436,
      "pos_sim_mean": 0.6266528367996216,
      "step": 1213
    },
    {
      "epoch": 0.2428,
      "grad_norm": 0.5778024196624756,
      "learning_rate": 9.393252496481615e-05,
      "loss": 0.3765,
      "step": 1214
    },
    {
      "epoch": 0.2428,
      "loss_align": 0.23147451877593994,
      "loss_contrastive": 0.1156454086303711,
      "loss_gen": 0.1773955076932907,
      "neg_sim_mean": 0.7841708660125732,
      "pos_sim_mean": 0.7685254812240601,
      "step": 1214
    },
    {
      "epoch": 0.243,
      "grad_norm": 1.7107897996902466,
      "learning_rate": 9.391584757013289e-05,
      "loss": 0.4227,
      "step": 1215
    },
    {
      "epoch": 0.243,
      "loss_align": 0.29593586921691895,
      "loss_contrastive": 0.13108175992965698,
      "loss_gen": 0.26824986934661865,
      "neg_sim_mean": 0.7351458668708801,
      "pos_sim_mean": 0.704064130783081,
      "step": 1215
    },
    {
      "epoch": 0.2432,
      "grad_norm": 2.0354156494140625,
      "learning_rate": 9.389914877139903e-05,
      "loss": 0.5799,
      "step": 1216
    },
    {
      "epoch": 0.2432,
      "loss_align": 0.377766489982605,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0006794839282520115,
      "neg_sim_mean": 0.4578549861907959,
      "pos_sim_mean": 0.622233510017395,
      "step": 1216
    },
    {
      "epoch": 0.2434,
      "grad_norm": 0.6505597233772278,
      "learning_rate": 9.388242857675336e-05,
      "loss": 0.3784,
      "step": 1217
    },
    {
      "epoch": 0.2434,
      "loss_align": 0.2728097438812256,
      "loss_contrastive": 0.08589208126068115,
      "loss_gen": 0.536338210105896,
      "neg_sim_mean": 0.7130823135375977,
      "pos_sim_mean": 0.7271902561187744,
      "step": 1217
    },
    {
      "epoch": 0.2436,
      "grad_norm": 3.10520339012146,
      "learning_rate": 9.386568699434508e-05,
      "loss": 0.8195,
      "step": 1218
    },
    {
      "epoch": 0.2436,
      "loss_align": 0.23050594329833984,
      "loss_contrastive": 0.1067667007446289,
      "loss_gen": 0.31332406401634216,
      "neg_sim_mean": 0.7762607336044312,
      "pos_sim_mean": 0.7694940567016602,
      "step": 1218
    },
    {
      "epoch": 0.2438,
      "grad_norm": 1.996089220046997,
      "learning_rate": 9.384892403233384e-05,
      "loss": 0.5566,
      "step": 1219
    },
    {
      "epoch": 0.2438,
      "loss_align": 0.434739351272583,
      "loss_contrastive": 0.31246694922447205,
      "loss_gen": 0.3165879547595978,
      "neg_sim_mean": 0.7777276039123535,
      "pos_sim_mean": 0.565260648727417,
      "step": 1219
    },
    {
      "epoch": 0.244,
      "grad_norm": 2.0730466842651367,
      "learning_rate": 9.383213969888972e-05,
      "loss": 0.7888,
      "step": 1220
    },
    {
      "epoch": 0.244,
      "loss_align": 0.2902311086654663,
      "loss_contrastive": 0.0,
      "loss_gen": 0.32377055287361145,
      "neg_sim_mean": 0.4957343637943268,
      "pos_sim_mean": 0.7097688913345337,
      "step": 1220
    },
    {
      "epoch": 0.2442,
      "grad_norm": 2.1924359798431396,
      "learning_rate": 9.381533400219318e-05,
      "loss": 0.614,
      "step": 1221
    },
    {
      "epoch": 0.2442,
      "loss_align": 0.19643795490264893,
      "loss_contrastive": 0.1451268196105957,
      "loss_gen": 0.12798310816287994,
      "neg_sim_mean": 0.8486888408660889,
      "pos_sim_mean": 0.8035620450973511,
      "step": 1221
    },
    {
      "epoch": 0.2444,
      "grad_norm": 1.728240728378296,
      "learning_rate": 9.379850695043513e-05,
      "loss": 0.3418,
      "step": 1222
    },
    {
      "epoch": 0.2444,
      "loss_align": 0.31128567457199097,
      "loss_contrastive": 0.0171431303024292,
      "loss_gen": 0.17661628127098083,
      "neg_sim_mean": 0.6058574318885803,
      "pos_sim_mean": 0.688714325428009,
      "step": 1222
    },
    {
      "epoch": 0.2446,
      "grad_norm": 2.270468235015869,
      "learning_rate": 9.378165855181687e-05,
      "loss": 0.49,
      "step": 1223
    },
    {
      "epoch": 0.2446,
      "loss_align": 0.26592040061950684,
      "loss_contrastive": 0.09089851379394531,
      "loss_gen": 0.26160958409309387,
      "neg_sim_mean": 0.7249780893325806,
      "pos_sim_mean": 0.7340795993804932,
      "step": 1223
    },
    {
      "epoch": 0.2448,
      "grad_norm": 2.451890230178833,
      "learning_rate": 9.376478881455009e-05,
      "loss": 0.5384,
      "step": 1224
    },
    {
      "epoch": 0.2448,
      "loss_align": 0.2564541697502136,
      "loss_contrastive": 0.013835251331329346,
      "loss_gen": 0.28580084443092346,
      "neg_sim_mean": 0.6573810577392578,
      "pos_sim_mean": 0.7435458302497864,
      "step": 1224
    },
    {
      "epoch": 0.245,
      "grad_norm": 2.2567241191864014,
      "learning_rate": 9.37478977468569e-05,
      "loss": 0.5439,
      "step": 1225
    },
    {
      "epoch": 0.245,
      "loss_align": 0.37337183952331543,
      "loss_contrastive": 0.2559323310852051,
      "loss_gen": 0.3293715715408325,
      "neg_sim_mean": 0.7825604677200317,
      "pos_sim_mean": 0.6266281604766846,
      "step": 1225
    },
    {
      "epoch": 0.2452,
      "grad_norm": 2.2694413661956787,
      "learning_rate": 9.373098535696979e-05,
      "loss": 0.7335,
      "step": 1226
    },
    {
      "epoch": 0.2452,
      "loss_align": 0.24034655094146729,
      "loss_contrastive": 0.05932331085205078,
      "loss_gen": 0.6096598505973816,
      "neg_sim_mean": 0.7189767360687256,
      "pos_sim_mean": 0.7596534490585327,
      "step": 1226
    },
    {
      "epoch": 0.2454,
      "grad_norm": 2.7561697959899902,
      "learning_rate": 9.371405165313169e-05,
      "loss": 0.8571,
      "step": 1227
    },
    {
      "epoch": 0.2454,
      "loss_align": 0.10189616680145264,
      "loss_contrastive": 0.07655000686645508,
      "loss_gen": 0.6772276759147644,
      "neg_sim_mean": 0.8746538162231445,
      "pos_sim_mean": 0.8981038331985474,
      "step": 1227
    },
    {
      "epoch": 0.2456,
      "grad_norm": 3.6922507286071777,
      "learning_rate": 9.369709664359584e-05,
      "loss": 0.7883,
      "step": 1228
    },
    {
      "epoch": 0.2456,
      "loss_align": 0.22652924060821533,
      "loss_contrastive": 0.11779904365539551,
      "loss_gen": 0.339558482170105,
      "neg_sim_mean": 0.7912697792053223,
      "pos_sim_mean": 0.7734707593917847,
      "step": 1228
    },
    {
      "epoch": 0.2458,
      "grad_norm": 1.861728310585022,
      "learning_rate": 9.368012033662594e-05,
      "loss": 0.5802,
      "step": 1229
    },
    {
      "epoch": 0.2458,
      "loss_align": 0.37740612030029297,
      "loss_contrastive": 0.13865000009536743,
      "loss_gen": 0.286308228969574,
      "neg_sim_mean": 0.6612438559532166,
      "pos_sim_mean": 0.622593879699707,
      "step": 1229
    },
    {
      "epoch": 0.246,
      "grad_norm": 2.211129665374756,
      "learning_rate": 9.366312274049602e-05,
      "loss": 0.6804,
      "step": 1230
    },
    {
      "epoch": 0.246,
      "loss_align": 0.22741711139678955,
      "loss_contrastive": 0.0845038890838623,
      "loss_gen": 0.44066110253334045,
      "neg_sim_mean": 0.7570867538452148,
      "pos_sim_mean": 0.7725828886032104,
      "step": 1230
    },
    {
      "epoch": 0.2462,
      "grad_norm": 2.7397427558898926,
      "learning_rate": 9.364610386349049e-05,
      "loss": 0.6782,
      "step": 1231
    },
    {
      "epoch": 0.2462,
      "loss_align": 0.20772230625152588,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010869028978049755,
      "neg_sim_mean": 0.28853631019592285,
      "pos_sim_mean": 0.7922776937484741,
      "step": 1231
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.82407546043396,
      "learning_rate": 9.362906371390417e-05,
      "loss": 0.2186,
      "step": 1232
    },
    {
      "epoch": 0.2464,
      "loss_align": 0.407336950302124,
      "loss_contrastive": 0.08117875456809998,
      "loss_gen": 0.014563889242708683,
      "neg_sim_mean": 0.5738418102264404,
      "pos_sim_mean": 0.592663049697876,
      "step": 1232
    },
    {
      "epoch": 0.2466,
      "grad_norm": 1.3124635219573975,
      "learning_rate": 9.36120023000422e-05,
      "loss": 0.4316,
      "step": 1233
    },
    {
      "epoch": 0.2466,
      "loss_align": 0.11091840267181396,
      "loss_contrastive": 0.09821999073028564,
      "loss_gen": 0.195564404129982,
      "neg_sim_mean": 0.8873015642166138,
      "pos_sim_mean": 0.889081597328186,
      "step": 1233
    },
    {
      "epoch": 0.2468,
      "grad_norm": 1.4486119747161865,
      "learning_rate": 9.35949196302201e-05,
      "loss": 0.3183,
      "step": 1234
    },
    {
      "epoch": 0.2468,
      "loss_align": 0.27768564224243164,
      "loss_contrastive": 0.18080031871795654,
      "loss_gen": 0.4337530732154846,
      "neg_sim_mean": 0.803114652633667,
      "pos_sim_mean": 0.7223143577575684,
      "step": 1234
    },
    {
      "epoch": 0.247,
      "grad_norm": 2.694352149963379,
      "learning_rate": 9.357781571276379e-05,
      "loss": 0.7331,
      "step": 1235
    },
    {
      "epoch": 0.247,
      "loss_align": 0.3204280138015747,
      "loss_contrastive": 0.06563758850097656,
      "loss_gen": 0.012106997892260551,
      "neg_sim_mean": 0.645209550857544,
      "pos_sim_mean": 0.6795719861984253,
      "step": 1235
    },
    {
      "epoch": 0.2472,
      "grad_norm": 0.8973890542984009,
      "learning_rate": 9.356069055600948e-05,
      "loss": 0.3404,
      "step": 1236
    },
    {
      "epoch": 0.2472,
      "loss_align": 0.3120776414871216,
      "loss_contrastive": 0.09766507148742676,
      "loss_gen": 0.4597516357898712,
      "neg_sim_mean": 0.6855874061584473,
      "pos_sim_mean": 0.6879223585128784,
      "step": 1236
    },
    {
      "epoch": 0.2474,
      "grad_norm": 2.8420345783233643,
      "learning_rate": 9.354354416830377e-05,
      "loss": 0.7835,
      "step": 1237
    },
    {
      "epoch": 0.2474,
      "loss_align": 0.3301992416381836,
      "loss_contrastive": 0.07291454076766968,
      "loss_gen": 0.3310156464576721,
      "neg_sim_mean": 0.6427152752876282,
      "pos_sim_mean": 0.6698007583618164,
      "step": 1237
    },
    {
      "epoch": 0.2476,
      "grad_norm": 2.089721918106079,
      "learning_rate": 9.352637655800361e-05,
      "loss": 0.67,
      "step": 1238
    },
    {
      "epoch": 0.2476,
      "loss_align": 0.21007555723190308,
      "loss_contrastive": 0.11375969648361206,
      "loss_gen": 0.21680302917957306,
      "neg_sim_mean": 0.8036841154098511,
      "pos_sim_mean": 0.7899244427680969,
      "step": 1238
    },
    {
      "epoch": 0.2478,
      "grad_norm": 1.644802212715149,
      "learning_rate": 9.35091877334763e-05,
      "loss": 0.4405,
      "step": 1239
    },
    {
      "epoch": 0.2478,
      "loss_align": 0.39656734466552734,
      "loss_contrastive": 0.2042868733406067,
      "loss_gen": 0.5792660117149353,
      "neg_sim_mean": 0.7077195048332214,
      "pos_sim_mean": 0.6034326553344727,
      "step": 1239
    },
    {
      "epoch": 0.248,
      "grad_norm": 2.891129493713379,
      "learning_rate": 9.34919777030994e-05,
      "loss": 1.0003,
      "step": 1240
    },
    {
      "epoch": 0.248,
      "loss_align": 0.19918251037597656,
      "loss_contrastive": 0.10142475366592407,
      "loss_gen": 0.2340894341468811,
      "neg_sim_mean": 0.8022422194480896,
      "pos_sim_mean": 0.8008174896240234,
      "step": 1240
    },
    {
      "epoch": 0.2482,
      "grad_norm": 1.7860790491104126,
      "learning_rate": 9.347474647526095e-05,
      "loss": 0.4454,
      "step": 1241
    },
    {
      "epoch": 0.2482,
      "loss_align": 0.22997766733169556,
      "loss_contrastive": 0.04658228158950806,
      "loss_gen": 0.4798816442489624,
      "neg_sim_mean": 0.7166045904159546,
      "pos_sim_mean": 0.7700223326683044,
      "step": 1241
    },
    {
      "epoch": 0.2484,
      "grad_norm": 2.6772241592407227,
      "learning_rate": 9.34574940583592e-05,
      "loss": 0.7154,
      "step": 1242
    },
    {
      "epoch": 0.2484,
      "loss_align": 0.3823992609977722,
      "loss_contrastive": 0.20650702714920044,
      "loss_gen": 0.356012761592865,
      "neg_sim_mean": 0.7241077423095703,
      "pos_sim_mean": 0.6176007390022278,
      "step": 1242
    },
    {
      "epoch": 0.2486,
      "grad_norm": 2.702120780944824,
      "learning_rate": 9.344022046080276e-05,
      "loss": 0.7632,
      "step": 1243
    },
    {
      "epoch": 0.2486,
      "loss_align": 0.32700681686401367,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00442476524040103,
      "neg_sim_mean": 0.5080528259277344,
      "pos_sim_mean": 0.6729931831359863,
      "step": 1243
    },
    {
      "epoch": 0.2488,
      "grad_norm": 0.6037419438362122,
      "learning_rate": 9.342292569101061e-05,
      "loss": 0.3314,
      "step": 1244
    },
    {
      "epoch": 0.2488,
      "loss_align": 0.2635784149169922,
      "loss_contrastive": 0.11616212129592896,
      "loss_gen": 0.2491292506456375,
      "neg_sim_mean": 0.7525836825370789,
      "pos_sim_mean": 0.7364215850830078,
      "step": 1244
    },
    {
      "epoch": 0.249,
      "grad_norm": 1.9875189065933228,
      "learning_rate": 9.340560975741197e-05,
      "loss": 0.5266,
      "step": 1245
    },
    {
      "epoch": 0.249,
      "loss_align": 0.280948281288147,
      "loss_contrastive": 0.06663846969604492,
      "loss_gen": 0.3550574481487274,
      "neg_sim_mean": 0.68569016456604,
      "pos_sim_mean": 0.719051718711853,
      "step": 1245
    },
    {
      "epoch": 0.2492,
      "grad_norm": 2.5873725414276123,
      "learning_rate": 9.338827266844644e-05,
      "loss": 0.644,
      "step": 1246
    },
    {
      "epoch": 0.2492,
      "loss_align": 0.37227779626846313,
      "loss_contrastive": 0.09581542015075684,
      "loss_gen": 0.31981587409973145,
      "neg_sim_mean": 0.6235376000404358,
      "pos_sim_mean": 0.6277222037315369,
      "step": 1246
    },
    {
      "epoch": 0.2494,
      "grad_norm": 2.03666353225708,
      "learning_rate": 9.337091443256387e-05,
      "loss": 0.7036,
      "step": 1247
    },
    {
      "epoch": 0.2494,
      "loss_align": 0.3780101537704468,
      "loss_contrastive": 0.24116146564483643,
      "loss_gen": 0.42417383193969727,
      "neg_sim_mean": 0.7631512880325317,
      "pos_sim_mean": 0.6219898462295532,
      "step": 1247
    },
    {
      "epoch": 0.2496,
      "grad_norm": 2.181248903274536,
      "learning_rate": 9.33535350582245e-05,
      "loss": 0.8311,
      "step": 1248
    },
    {
      "epoch": 0.2496,
      "loss_align": 0.25765061378479004,
      "loss_contrastive": 0.12218612432479858,
      "loss_gen": 0.21611107885837555,
      "neg_sim_mean": 0.7645354866981506,
      "pos_sim_mean": 0.74234938621521,
      "step": 1248
    },
    {
      "epoch": 0.2498,
      "grad_norm": 1.6283421516418457,
      "learning_rate": 9.333613455389883e-05,
      "loss": 0.4884,
      "step": 1249
    },
    {
      "epoch": 0.2498,
      "loss_align": 0.4156665802001953,
      "loss_contrastive": 0.18061277270317078,
      "loss_gen": 0.24507617950439453,
      "neg_sim_mean": 0.6649461984634399,
      "pos_sim_mean": 0.5843334197998047,
      "step": 1249
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.2593438625335693,
      "learning_rate": 9.331871292806759e-05,
      "loss": 0.6824,
      "step": 1250
    },
    {
      "epoch": 0.25,
      "loss_align": 0.29210877418518066,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01281752996146679,
      "neg_sim_mean": 0.48201799392700195,
      "pos_sim_mean": 0.7078912258148193,
      "step": 1250
    },
    {
      "epoch": 0.2502,
      "grad_norm": 4.48463249206543,
      "learning_rate": 9.330127018922194e-05,
      "loss": 0.3049,
      "step": 1251
    },
    {
      "epoch": 0.2502,
      "loss_align": 0.18525290489196777,
      "loss_contrastive": 0.0725550651550293,
      "loss_gen": 0.390463650226593,
      "neg_sim_mean": 0.7873021364212036,
      "pos_sim_mean": 0.8147470951080322,
      "step": 1251
    },
    {
      "epoch": 0.2504,
      "grad_norm": 2.3656563758850098,
      "learning_rate": 9.328380634586322e-05,
      "loss": 0.5844,
      "step": 1252
    },
    {
      "epoch": 0.2504,
      "loss_align": 0.30493974685668945,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01344984583556652,
      "neg_sim_mean": 0.4493821859359741,
      "pos_sim_mean": 0.6950602531433105,
      "step": 1252
    },
    {
      "epoch": 0.2506,
      "grad_norm": 0.9167845249176025,
      "learning_rate": 9.326632140650311e-05,
      "loss": 0.3184,
      "step": 1253
    },
    {
      "epoch": 0.2506,
      "loss_align": 0.10741639137268066,
      "loss_contrastive": 0.0,
      "loss_gen": 0.024006815627217293,
      "neg_sim_mean": 0.4125003218650818,
      "pos_sim_mean": 0.8925836086273193,
      "step": 1253
    },
    {
      "epoch": 0.2508,
      "grad_norm": 7.383155345916748,
      "learning_rate": 9.324881537966354e-05,
      "loss": 0.1314,
      "step": 1254
    },
    {
      "epoch": 0.2508,
      "loss_align": 0.2020127773284912,
      "loss_contrastive": 0.11325693130493164,
      "loss_gen": 0.19578704237937927,
      "neg_sim_mean": 0.8112441301345825,
      "pos_sim_mean": 0.7979872226715088,
      "step": 1254
    },
    {
      "epoch": 0.251,
      "grad_norm": 1.787388801574707,
      "learning_rate": 9.323128827387675e-05,
      "loss": 0.4114,
      "step": 1255
    },
    {
      "epoch": 0.251,
      "loss_align": 0.4002242684364319,
      "loss_contrastive": 0.21508684754371643,
      "loss_gen": 0.14813232421875,
      "neg_sim_mean": 0.714862585067749,
      "pos_sim_mean": 0.5997757315635681,
      "step": 1255
    },
    {
      "epoch": 0.2512,
      "grad_norm": 1.6798369884490967,
      "learning_rate": 9.321374009768525e-05,
      "loss": 0.5742,
      "step": 1256
    },
    {
      "epoch": 0.2512,
      "loss_align": 0.4689469337463379,
      "loss_contrastive": 0.12131544947624207,
      "loss_gen": 0.2953067421913147,
      "neg_sim_mean": 0.5523685216903687,
      "pos_sim_mean": 0.5310530662536621,
      "step": 1256
    },
    {
      "epoch": 0.2514,
      "grad_norm": 1.7345199584960938,
      "learning_rate": 9.319617085964176e-05,
      "loss": 0.7788,
      "step": 1257
    },
    {
      "epoch": 0.2514,
      "loss_align": 0.364406943321228,
      "loss_contrastive": 0.17596834897994995,
      "loss_gen": 0.35252609848976135,
      "neg_sim_mean": 0.711561381816864,
      "pos_sim_mean": 0.635593056678772,
      "step": 1257
    },
    {
      "epoch": 0.2516,
      "grad_norm": 2.2111728191375732,
      "learning_rate": 9.317858056830936e-05,
      "loss": 0.738,
      "step": 1258
    },
    {
      "epoch": 0.2516,
      "loss_align": 0.4202669858932495,
      "loss_contrastive": 0.1396406590938568,
      "loss_gen": 0.3220921754837036,
      "neg_sim_mean": 0.6193736791610718,
      "pos_sim_mean": 0.5797330141067505,
      "step": 1258
    },
    {
      "epoch": 0.2518,
      "grad_norm": 2.091569423675537,
      "learning_rate": 9.316096923226135e-05,
      "loss": 0.7591,
      "step": 1259
    },
    {
      "epoch": 0.2518,
      "loss_align": 0.22614091634750366,
      "loss_contrastive": 0.14061814546585083,
      "loss_gen": 0.10891328752040863,
      "neg_sim_mean": 0.8144772052764893,
      "pos_sim_mean": 0.7738590836524963,
      "step": 1259
    },
    {
      "epoch": 0.252,
      "grad_norm": 1.6002318859100342,
      "learning_rate": 9.314333686008125e-05,
      "loss": 0.3519,
      "step": 1260
    },
    {
      "epoch": 0.252,
      "loss_align": 0.29824644327163696,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01621987856924534,
      "neg_sim_mean": 0.49122801423072815,
      "pos_sim_mean": 0.701753556728363,
      "step": 1260
    },
    {
      "epoch": 0.2522,
      "grad_norm": 7.0283074378967285,
      "learning_rate": 9.312568346036288e-05,
      "loss": 0.3145,
      "step": 1261
    },
    {
      "epoch": 0.2522,
      "loss_align": 0.23107123374938965,
      "loss_contrastive": 0.14568060636520386,
      "loss_gen": 0.401224821805954,
      "neg_sim_mean": 0.8146093487739563,
      "pos_sim_mean": 0.7689287662506104,
      "step": 1261
    },
    {
      "epoch": 0.2524,
      "grad_norm": 2.122560501098633,
      "learning_rate": 9.310800904171028e-05,
      "loss": 0.6498,
      "step": 1262
    },
    {
      "epoch": 0.2524,
      "loss_align": 0.4096873998641968,
      "loss_contrastive": 0.09382250905036926,
      "loss_gen": 0.4873107373714447,
      "neg_sim_mean": 0.584135115146637,
      "pos_sim_mean": 0.5903126001358032,
      "step": 1262
    },
    {
      "epoch": 0.2526,
      "grad_norm": 3.3667707443237305,
      "learning_rate": 9.309031361273775e-05,
      "loss": 0.9083,
      "step": 1263
    },
    {
      "epoch": 0.2526,
      "loss_align": 0.3126370906829834,
      "loss_contrastive": 0.0641704797744751,
      "loss_gen": 0.25284111499786377,
      "neg_sim_mean": 0.6515333652496338,
      "pos_sim_mean": 0.6873629093170166,
      "step": 1263
    },
    {
      "epoch": 0.2528,
      "grad_norm": 1.812436819076538,
      "learning_rate": 9.307259718206985e-05,
      "loss": 0.5732,
      "step": 1264
    },
    {
      "epoch": 0.2528,
      "loss_align": 0.23192012310028076,
      "loss_contrastive": 0.10503578186035156,
      "loss_gen": 0.4157458245754242,
      "neg_sim_mean": 0.7731156349182129,
      "pos_sim_mean": 0.7680798768997192,
      "step": 1264
    },
    {
      "epoch": 0.253,
      "grad_norm": 2.567600965499878,
      "learning_rate": 9.305485975834132e-05,
      "loss": 0.6603,
      "step": 1265
    },
    {
      "epoch": 0.253,
      "loss_align": 0.4040473699569702,
      "loss_contrastive": 0.1920364797115326,
      "loss_gen": 0.3179645836353302,
      "neg_sim_mean": 0.6879891157150269,
      "pos_sim_mean": 0.5959526300430298,
      "step": 1265
    },
    {
      "epoch": 0.2532,
      "grad_norm": 2.3827567100524902,
      "learning_rate": 9.30371013501972e-05,
      "loss": 0.7451,
      "step": 1266
    },
    {
      "epoch": 0.2532,
      "loss_align": 0.23102867603302002,
      "loss_contrastive": 0.12490880489349365,
      "loss_gen": 0.28408271074295044,
      "neg_sim_mean": 0.7938801050186157,
      "pos_sim_mean": 0.76897132396698,
      "step": 1266
    },
    {
      "epoch": 0.2534,
      "grad_norm": 2.1013476848602295,
      "learning_rate": 9.301932196629267e-05,
      "loss": 0.5301,
      "step": 1267
    },
    {
      "epoch": 0.2534,
      "loss_align": 0.3493042588233948,
      "loss_contrastive": 0.12447339296340942,
      "loss_gen": 0.22624707221984863,
      "neg_sim_mean": 0.6751691102981567,
      "pos_sim_mean": 0.6506957411766052,
      "step": 1267
    },
    {
      "epoch": 0.2536,
      "grad_norm": 1.8669161796569824,
      "learning_rate": 9.300152161529324e-05,
      "loss": 0.5905,
      "step": 1268
    },
    {
      "epoch": 0.2536,
      "loss_align": 0.3649252653121948,
      "loss_contrastive": 0.04110628366470337,
      "loss_gen": 0.04260864108800888,
      "neg_sim_mean": 0.5761809945106506,
      "pos_sim_mean": 0.6350747346878052,
      "step": 1268
    },
    {
      "epoch": 0.2538,
      "grad_norm": 1.9209027290344238,
      "learning_rate": 9.298370030587456e-05,
      "loss": 0.4125,
      "step": 1269
    },
    {
      "epoch": 0.2538,
      "loss_align": 0.22874176502227783,
      "loss_contrastive": 0.16407030820846558,
      "loss_gen": 0.29381096363067627,
      "neg_sim_mean": 0.8353285193443298,
      "pos_sim_mean": 0.7712582349777222,
      "step": 1269
    },
    {
      "epoch": 0.254,
      "grad_norm": 2.6844983100891113,
      "learning_rate": 9.296585804672252e-05,
      "loss": 0.5422,
      "step": 1270
    },
    {
      "epoch": 0.254,
      "loss_align": 0.3245280981063843,
      "loss_contrastive": 0.04907786846160889,
      "loss_gen": 0.2537236213684082,
      "neg_sim_mean": 0.6245497465133667,
      "pos_sim_mean": 0.6754719018936157,
      "step": 1270
    },
    {
      "epoch": 0.2542,
      "grad_norm": 1.99187433719635,
      "learning_rate": 9.294799484653323e-05,
      "loss": 0.5841,
      "step": 1271
    },
    {
      "epoch": 0.2542,
      "loss_align": 0.21155911684036255,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17345315217971802,
      "neg_sim_mean": 0.6245101094245911,
      "pos_sim_mean": 0.7884408831596375,
      "step": 1271
    },
    {
      "epoch": 0.2544,
      "grad_norm": 1.843809962272644,
      "learning_rate": 9.293011071401298e-05,
      "loss": 0.385,
      "step": 1272
    },
    {
      "epoch": 0.2544,
      "loss_align": 0.2459803819656372,
      "loss_contrastive": 0.2064971923828125,
      "loss_gen": 0.22471249103546143,
      "neg_sim_mean": 0.8605167865753174,
      "pos_sim_mean": 0.7540196180343628,
      "step": 1272
    },
    {
      "epoch": 0.2546,
      "grad_norm": 2.068045139312744,
      "learning_rate": 9.29122056578783e-05,
      "loss": 0.4955,
      "step": 1273
    },
    {
      "epoch": 0.2546,
      "loss_align": 0.30137741565704346,
      "loss_contrastive": 0.11101412773132324,
      "loss_gen": 0.25254660844802856,
      "neg_sim_mean": 0.7096366882324219,
      "pos_sim_mean": 0.6986225843429565,
      "step": 1273
    },
    {
      "epoch": 0.2548,
      "grad_norm": 1.9055217504501343,
      "learning_rate": 9.289427968685586e-05,
      "loss": 0.5672,
      "step": 1274
    },
    {
      "epoch": 0.2548,
      "loss_align": 0.18784809112548828,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3765106201171875,
      "neg_sim_mean": 0.5848128199577332,
      "pos_sim_mean": 0.8121519088745117,
      "step": 1274
    },
    {
      "epoch": 0.255,
      "grad_norm": 2.3721413612365723,
      "learning_rate": 9.287633280968261e-05,
      "loss": 0.5644,
      "step": 1275
    },
    {
      "epoch": 0.255,
      "loss_align": 0.20891988277435303,
      "loss_contrastive": 0.13234734535217285,
      "loss_gen": 0.292208194732666,
      "neg_sim_mean": 0.8234274387359619,
      "pos_sim_mean": 0.791080117225647,
      "step": 1275
    },
    {
      "epoch": 0.2552,
      "grad_norm": 2.2292680740356445,
      "learning_rate": 9.285836503510562e-05,
      "loss": 0.517,
      "step": 1276
    },
    {
      "epoch": 0.2552,
      "loss_align": 0.43555498123168945,
      "loss_contrastive": 0.14445963501930237,
      "loss_gen": 0.6124423742294312,
      "neg_sim_mean": 0.6089046597480774,
      "pos_sim_mean": 0.5644450187683105,
      "step": 1276
    },
    {
      "epoch": 0.2554,
      "grad_norm": 3.304922342300415,
      "learning_rate": 9.284037637188215e-05,
      "loss": 1.0653,
      "step": 1277
    },
    {
      "epoch": 0.2554,
      "loss_align": 0.21736550331115723,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03590136766433716,
      "neg_sim_mean": 0.3321831226348877,
      "pos_sim_mean": 0.7826344966888428,
      "step": 1277
    },
    {
      "epoch": 0.2556,
      "grad_norm": 1.4375592470169067,
      "learning_rate": 9.282236682877967e-05,
      "loss": 0.2533,
      "step": 1278
    },
    {
      "epoch": 0.2556,
      "loss_align": 0.2332613468170166,
      "loss_contrastive": 0.12214934825897217,
      "loss_gen": 0.3880118727684021,
      "neg_sim_mean": 0.7888879776000977,
      "pos_sim_mean": 0.7667386531829834,
      "step": 1278
    },
    {
      "epoch": 0.2558,
      "grad_norm": 2.7730488777160645,
      "learning_rate": 9.28043364145758e-05,
      "loss": 0.6359,
      "step": 1279
    },
    {
      "epoch": 0.2558,
      "loss_align": 0.31983691453933716,
      "loss_contrastive": 0.12024420499801636,
      "loss_gen": 0.6247671842575073,
      "neg_sim_mean": 0.7004072666168213,
      "pos_sim_mean": 0.6801630854606628,
      "step": 1279
    },
    {
      "epoch": 0.256,
      "grad_norm": 2.555514335632324,
      "learning_rate": 9.278628513805838e-05,
      "loss": 0.959,
      "step": 1280
    },
    {
      "epoch": 0.256,
      "loss_align": 0.25770580768585205,
      "loss_contrastive": 0.07822263240814209,
      "loss_gen": 0.34783995151519775,
      "neg_sim_mean": 0.7205168008804321,
      "pos_sim_mean": 0.742294192314148,
      "step": 1280
    },
    {
      "epoch": 0.2562,
      "grad_norm": 2.78027606010437,
      "learning_rate": 9.276821300802534e-05,
      "loss": 0.6149,
      "step": 1281
    },
    {
      "epoch": 0.2562,
      "loss_align": 0.38991016149520874,
      "loss_contrastive": 0.12158375978469849,
      "loss_gen": 0.2991960048675537,
      "neg_sim_mean": 0.6316735744476318,
      "pos_sim_mean": 0.6100898385047913,
      "step": 1281
    },
    {
      "epoch": 0.2564,
      "grad_norm": 2.062201738357544,
      "learning_rate": 9.275012003328483e-05,
      "loss": 0.7037,
      "step": 1282
    },
    {
      "epoch": 0.2564,
      "loss_align": 0.20952141284942627,
      "loss_contrastive": 0.08885502815246582,
      "loss_gen": 0.20445577800273895,
      "neg_sim_mean": 0.7793335914611816,
      "pos_sim_mean": 0.7904785871505737,
      "step": 1282
    },
    {
      "epoch": 0.2566,
      "grad_norm": 1.8667752742767334,
      "learning_rate": 9.273200622265516e-05,
      "loss": 0.4246,
      "step": 1283
    },
    {
      "epoch": 0.2566,
      "loss_align": 0.16909360885620117,
      "loss_contrastive": 0.05981320142745972,
      "loss_gen": 0.38175541162490845,
      "neg_sim_mean": 0.7907195687294006,
      "pos_sim_mean": 0.8309063911437988,
      "step": 1283
    },
    {
      "epoch": 0.2568,
      "grad_norm": 2.192629814147949,
      "learning_rate": 9.271387158496476e-05,
      "loss": 0.558,
      "step": 1284
    },
    {
      "epoch": 0.2568,
      "loss_align": 0.29637134075164795,
      "loss_contrastive": 0.17137587070465088,
      "loss_gen": 0.2045455425977707,
      "neg_sim_mean": 0.775004506111145,
      "pos_sim_mean": 0.703628659248352,
      "step": 1284
    },
    {
      "epoch": 0.257,
      "grad_norm": 1.8889775276184082,
      "learning_rate": 9.269571612905225e-05,
      "loss": 0.5215,
      "step": 1285
    },
    {
      "epoch": 0.257,
      "loss_align": 0.14285200834274292,
      "loss_contrastive": 0.03139251470565796,
      "loss_gen": 0.2021632343530655,
      "neg_sim_mean": 0.7885404825210571,
      "pos_sim_mean": 0.8571479916572571,
      "step": 1285
    },
    {
      "epoch": 0.2572,
      "grad_norm": 2.144747734069824,
      "learning_rate": 9.267753986376637e-05,
      "loss": 0.3488,
      "step": 1286
    },
    {
      "epoch": 0.2572,
      "loss_align": 0.2813771963119507,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01774011366069317,
      "neg_sim_mean": 0.37021759152412415,
      "pos_sim_mean": 0.7186228036880493,
      "step": 1286
    },
    {
      "epoch": 0.2574,
      "grad_norm": 0.9997801184654236,
      "learning_rate": 9.265934279796602e-05,
      "loss": 0.2991,
      "step": 1287
    },
    {
      "epoch": 0.2574,
      "loss_align": 0.13398516178131104,
      "loss_contrastive": 0.11539965867996216,
      "loss_gen": 0.27945756912231445,
      "neg_sim_mean": 0.8814144730567932,
      "pos_sim_mean": 0.866014838218689,
      "step": 1287
    },
    {
      "epoch": 0.2576,
      "grad_norm": 1.9779858589172363,
      "learning_rate": 9.264112494052022e-05,
      "loss": 0.4273,
      "step": 1288
    },
    {
      "epoch": 0.2576,
      "loss_align": 0.2893192172050476,
      "loss_contrastive": 0.003668487071990967,
      "loss_gen": 0.37599319219589233,
      "neg_sim_mean": 0.6143492460250854,
      "pos_sim_mean": 0.7106807827949524,
      "step": 1288
    },
    {
      "epoch": 0.2578,
      "grad_norm": 2.639309883117676,
      "learning_rate": 9.262288630030814e-05,
      "loss": 0.6658,
      "step": 1289
    },
    {
      "epoch": 0.2578,
      "loss_align": 0.17856299877166748,
      "loss_contrastive": 0.010178089141845703,
      "loss_gen": 0.24686089158058167,
      "neg_sim_mean": 0.7316150665283203,
      "pos_sim_mean": 0.8214370012283325,
      "step": 1289
    },
    {
      "epoch": 0.258,
      "grad_norm": 2.07429575920105,
      "learning_rate": 9.260462688621905e-05,
      "loss": 0.4266,
      "step": 1290
    },
    {
      "epoch": 0.258,
      "loss_align": 0.16511934995651245,
      "loss_contrastive": 0.08826535940170288,
      "loss_gen": 0.23642002046108246,
      "neg_sim_mean": 0.8231459856033325,
      "pos_sim_mean": 0.8348806500434875,
      "step": 1290
    },
    {
      "epoch": 0.2582,
      "grad_norm": 1.780773639678955,
      "learning_rate": 9.258634670715238e-05,
      "loss": 0.4121,
      "step": 1291
    },
    {
      "epoch": 0.2582,
      "loss_align": 0.3985900282859802,
      "loss_contrastive": 0.16404056549072266,
      "loss_gen": 0.42754846811294556,
      "neg_sim_mean": 0.6654505133628845,
      "pos_sim_mean": 0.6014099717140198,
      "step": 1291
    },
    {
      "epoch": 0.2584,
      "grad_norm": 2.614017963409424,
      "learning_rate": 9.256804577201767e-05,
      "loss": 0.8458,
      "step": 1292
    },
    {
      "epoch": 0.2584,
      "loss_align": 0.18572580814361572,
      "loss_contrastive": 0.08220267295837402,
      "loss_gen": 0.4810852110385895,
      "neg_sim_mean": 0.7964768409729004,
      "pos_sim_mean": 0.8142741918563843,
      "step": 1292
    },
    {
      "epoch": 0.2586,
      "grad_norm": 2.8001906871795654,
      "learning_rate": 9.254972408973461e-05,
      "loss": 0.6767,
      "step": 1293
    },
    {
      "epoch": 0.2586,
      "loss_align": 0.2351241111755371,
      "loss_contrastive": 0.08244222402572632,
      "loss_gen": 0.3667842745780945,
      "neg_sim_mean": 0.7473180890083313,
      "pos_sim_mean": 0.7648758888244629,
      "step": 1293
    },
    {
      "epoch": 0.2588,
      "grad_norm": 2.1363601684570312,
      "learning_rate": 9.253138166923289e-05,
      "loss": 0.6118,
      "step": 1294
    },
    {
      "epoch": 0.2588,
      "loss_align": 0.25564754009246826,
      "loss_contrastive": 0.0,
      "loss_gen": 0.49542906880378723,
      "neg_sim_mean": 0.6262807250022888,
      "pos_sim_mean": 0.7443524599075317,
      "step": 1294
    },
    {
      "epoch": 0.259,
      "grad_norm": 2.838226318359375,
      "learning_rate": 9.251301851945244e-05,
      "loss": 0.7511,
      "step": 1295
    },
    {
      "epoch": 0.259,
      "loss_align": 0.35115665197372437,
      "loss_contrastive": 0.18275290727615356,
      "loss_gen": 0.2324966937303543,
      "neg_sim_mean": 0.7315962314605713,
      "pos_sim_mean": 0.6488433480262756,
      "step": 1295
    },
    {
      "epoch": 0.2592,
      "grad_norm": 2.2880122661590576,
      "learning_rate": 9.249463464934321e-05,
      "loss": 0.6056,
      "step": 1296
    },
    {
      "epoch": 0.2592,
      "loss_align": 0.3295711874961853,
      "loss_contrastive": 0.20207899808883667,
      "loss_gen": 0.5659460425376892,
      "neg_sim_mean": 0.7725077867507935,
      "pos_sim_mean": 0.6704288125038147,
      "step": 1296
    },
    {
      "epoch": 0.2594,
      "grad_norm": 2.5717694759368896,
      "learning_rate": 9.247623006786527e-05,
      "loss": 0.9198,
      "step": 1297
    },
    {
      "epoch": 0.2594,
      "loss_align": 0.2331656813621521,
      "loss_contrastive": 0.16422754526138306,
      "loss_gen": 0.3678591549396515,
      "neg_sim_mean": 0.831061840057373,
      "pos_sim_mean": 0.7668343186378479,
      "step": 1297
    },
    {
      "epoch": 0.2596,
      "grad_norm": 2.233236789703369,
      "learning_rate": 9.245780478398883e-05,
      "loss": 0.6207,
      "step": 1298
    },
    {
      "epoch": 0.2596,
      "loss_align": 0.13675785064697266,
      "loss_contrastive": 0.11048692464828491,
      "loss_gen": 0.30702653527259827,
      "neg_sim_mean": 0.8737290501594543,
      "pos_sim_mean": 0.8632421493530273,
      "step": 1298
    },
    {
      "epoch": 0.2598,
      "grad_norm": 2.346881628036499,
      "learning_rate": 9.24393588066941e-05,
      "loss": 0.457,
      "step": 1299
    },
    {
      "epoch": 0.2598,
      "loss_align": 0.15720272064208984,
      "loss_contrastive": 0.08198028802871704,
      "loss_gen": 0.13734106719493866,
      "neg_sim_mean": 0.8247775435447693,
      "pos_sim_mean": 0.8427972793579102,
      "step": 1299
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.6675939559936523,
      "learning_rate": 9.242089214497145e-05,
      "loss": 0.3044,
      "step": 1300
    },
    {
      "epoch": 0.26,
      "loss_align": 0.26881635189056396,
      "loss_contrastive": 0.11887532472610474,
      "loss_gen": 0.2154977023601532,
      "neg_sim_mean": 0.7500589489936829,
      "pos_sim_mean": 0.731183648109436,
      "step": 1300
    },
    {
      "epoch": 0.2602,
      "grad_norm": 1.8143163919448853,
      "learning_rate": 9.24024048078213e-05,
      "loss": 0.4986,
      "step": 1301
    },
    {
      "epoch": 0.2602,
      "loss_align": 0.3072141408920288,
      "loss_contrastive": 0.08609247207641602,
      "loss_gen": 0.260934442281723,
      "neg_sim_mean": 0.6788783073425293,
      "pos_sim_mean": 0.6927858591079712,
      "step": 1301
    },
    {
      "epoch": 0.2604,
      "grad_norm": 2.18945050239563,
      "learning_rate": 9.238389680425416e-05,
      "loss": 0.5785,
      "step": 1302
    },
    {
      "epoch": 0.2604,
      "loss_align": 0.2505430579185486,
      "loss_contrastive": 0.0,
      "loss_gen": 0.022286580875515938,
      "neg_sim_mean": 0.4215013086795807,
      "pos_sim_mean": 0.7494569420814514,
      "step": 1302
    },
    {
      "epoch": 0.2606,
      "grad_norm": 0.9868737459182739,
      "learning_rate": 9.23653681432906e-05,
      "loss": 0.2728,
      "step": 1303
    },
    {
      "epoch": 0.2606,
      "loss_align": 0.21791404485702515,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006395143456757069,
      "neg_sim_mean": 0.5983943939208984,
      "pos_sim_mean": 0.7820859551429749,
      "step": 1303
    },
    {
      "epoch": 0.2608,
      "grad_norm": 0.5582861304283142,
      "learning_rate": 9.234681883396127e-05,
      "loss": 0.2243,
      "step": 1304
    },
    {
      "epoch": 0.2608,
      "loss_align": 0.14687657356262207,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2921047508716583,
      "neg_sim_mean": 0.673572301864624,
      "pos_sim_mean": 0.8531234264373779,
      "step": 1304
    },
    {
      "epoch": 0.261,
      "grad_norm": 2.261320114135742,
      "learning_rate": 9.23282488853069e-05,
      "loss": 0.439,
      "step": 1305
    },
    {
      "epoch": 0.261,
      "loss_align": 0.2602382302284241,
      "loss_contrastive": 0.15985256433486938,
      "loss_gen": 0.5248255133628845,
      "neg_sim_mean": 0.7996143102645874,
      "pos_sim_mean": 0.7397617697715759,
      "step": 1305
    },
    {
      "epoch": 0.2612,
      "grad_norm": 3.261416435241699,
      "learning_rate": 9.230965830637821e-05,
      "loss": 0.8042,
      "step": 1306
    },
    {
      "epoch": 0.2612,
      "loss_align": 0.3629630208015442,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008518035523593426,
      "neg_sim_mean": 0.4574142098426819,
      "pos_sim_mean": 0.6370369791984558,
      "step": 1306
    },
    {
      "epoch": 0.2614,
      "grad_norm": 0.7779417037963867,
      "learning_rate": 9.229104710623604e-05,
      "loss": 0.3715,
      "step": 1307
    },
    {
      "epoch": 0.2614,
      "loss_align": 0.20705807209014893,
      "loss_contrastive": 0.10498440265655518,
      "loss_gen": 0.21546941995620728,
      "neg_sim_mean": 0.7979263067245483,
      "pos_sim_mean": 0.7929419279098511,
      "step": 1307
    },
    {
      "epoch": 0.2616,
      "grad_norm": 2.0313453674316406,
      "learning_rate": 9.227241529395127e-05,
      "loss": 0.4351,
      "step": 1308
    },
    {
      "epoch": 0.2616,
      "loss_align": 0.20962107181549072,
      "loss_contrastive": 0.1171349287033081,
      "loss_gen": 0.3690911829471588,
      "neg_sim_mean": 0.8075138330459595,
      "pos_sim_mean": 0.7903789281845093,
      "step": 1308
    },
    {
      "epoch": 0.2618,
      "grad_norm": 2.243697166442871,
      "learning_rate": 9.225376287860484e-05,
      "loss": 0.5928,
      "step": 1309
    },
    {
      "epoch": 0.2618,
      "loss_align": 0.2203463315963745,
      "loss_contrastive": 0.11801522970199585,
      "loss_gen": 0.24735452234745026,
      "neg_sim_mean": 0.7976688742637634,
      "pos_sim_mean": 0.7796536684036255,
      "step": 1309
    },
    {
      "epoch": 0.262,
      "grad_norm": 2.1308021545410156,
      "learning_rate": 9.223508986928766e-05,
      "loss": 0.4819,
      "step": 1310
    },
    {
      "epoch": 0.262,
      "loss_align": 0.2801944613456726,
      "loss_contrastive": 0.03984469175338745,
      "loss_gen": 0.18601515889167786,
      "neg_sim_mean": 0.6596502065658569,
      "pos_sim_mean": 0.7198055386543274,
      "step": 1310
    },
    {
      "epoch": 0.2622,
      "grad_norm": 1.740089774131775,
      "learning_rate": 9.221639627510076e-05,
      "loss": 0.471,
      "step": 1311
    },
    {
      "epoch": 0.2622,
      "loss_align": 0.20832562446594238,
      "loss_contrastive": 0.007924556732177734,
      "loss_gen": 0.24530500173568726,
      "neg_sim_mean": 0.6995989084243774,
      "pos_sim_mean": 0.7916743755340576,
      "step": 1311
    },
    {
      "epoch": 0.2624,
      "grad_norm": 2.023608684539795,
      "learning_rate": 9.219768210515517e-05,
      "loss": 0.4546,
      "step": 1312
    },
    {
      "epoch": 0.2624,
      "loss_align": 0.34953272342681885,
      "loss_contrastive": 0.05954104661941528,
      "loss_gen": 0.16680893301963806,
      "neg_sim_mean": 0.6100082993507385,
      "pos_sim_mean": 0.6504672765731812,
      "step": 1312
    },
    {
      "epoch": 0.2626,
      "grad_norm": 1.8263267278671265,
      "learning_rate": 9.217894736857195e-05,
      "loss": 0.5235,
      "step": 1313
    },
    {
      "epoch": 0.2626,
      "loss_align": 0.45515984296798706,
      "loss_contrastive": 0.049823105335235596,
      "loss_gen": 0.3870081901550293,
      "neg_sim_mean": 0.494663268327713,
      "pos_sim_mean": 0.5448401570320129,
      "step": 1313
    },
    {
      "epoch": 0.2628,
      "grad_norm": 2.796184539794922,
      "learning_rate": 9.216019207448217e-05,
      "loss": 0.8481,
      "step": 1314
    },
    {
      "epoch": 0.2628,
      "loss_align": 0.20741033554077148,
      "loss_contrastive": 0.05202144384384155,
      "loss_gen": 0.25826239585876465,
      "neg_sim_mean": 0.7446110844612122,
      "pos_sim_mean": 0.7925896644592285,
      "step": 1314
    },
    {
      "epoch": 0.263,
      "grad_norm": 1.8034113645553589,
      "learning_rate": 9.214141623202695e-05,
      "loss": 0.4719,
      "step": 1315
    },
    {
      "epoch": 0.263,
      "loss_align": 0.27157723903656006,
      "loss_contrastive": 0.0,
      "loss_gen": 0.019704818725585938,
      "neg_sim_mean": 0.4676637053489685,
      "pos_sim_mean": 0.7284227609634399,
      "step": 1315
    },
    {
      "epoch": 0.2632,
      "grad_norm": 1.0254275798797607,
      "learning_rate": 9.212261985035739e-05,
      "loss": 0.2913,
      "step": 1316
    },
    {
      "epoch": 0.2632,
      "loss_align": 0.2569814920425415,
      "loss_contrastive": 0.10469043254852295,
      "loss_gen": 0.2776658833026886,
      "neg_sim_mean": 0.7477089166641235,
      "pos_sim_mean": 0.7430185079574585,
      "step": 1316
    },
    {
      "epoch": 0.2634,
      "grad_norm": 2.054260730743408,
      "learning_rate": 9.210380293863462e-05,
      "loss": 0.5472,
      "step": 1317
    },
    {
      "epoch": 0.2634,
      "loss_align": 0.38085365295410156,
      "loss_contrastive": 0.23690664768218994,
      "loss_gen": 0.5070936679840088,
      "neg_sim_mean": 0.7560529708862305,
      "pos_sim_mean": 0.6191463470458984,
      "step": 1317
    },
    {
      "epoch": 0.2636,
      "grad_norm": 3.013972759246826,
      "learning_rate": 9.208496550602979e-05,
      "loss": 0.9164,
      "step": 1318
    },
    {
      "epoch": 0.2636,
      "loss_align": 0.3081780672073364,
      "loss_contrastive": 0.13366270065307617,
      "loss_gen": 0.31400614976882935,
      "neg_sim_mean": 0.7254846096038818,
      "pos_sim_mean": 0.6918219327926636,
      "step": 1318
    },
    {
      "epoch": 0.2638,
      "grad_norm": 2.4555723667144775,
      "learning_rate": 9.206610756172402e-05,
      "loss": 0.6382,
      "step": 1319
    },
    {
      "epoch": 0.2638,
      "loss_align": 0.2323685884475708,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009867814369499683,
      "neg_sim_mean": 0.5529646873474121,
      "pos_sim_mean": 0.7676314115524292,
      "step": 1319
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.7206518054008484,
      "learning_rate": 9.204722911490846e-05,
      "loss": 0.2422,
      "step": 1320
    },
    {
      "epoch": 0.264,
      "loss_align": 0.23372888565063477,
      "loss_contrastive": 0.096282958984375,
      "loss_gen": 0.21526867151260376,
      "neg_sim_mean": 0.7625540494918823,
      "pos_sim_mean": 0.7662711143493652,
      "step": 1320
    },
    {
      "epoch": 0.2642,
      "grad_norm": 1.8241842985153198,
      "learning_rate": 9.202833017478422e-05,
      "loss": 0.4606,
      "step": 1321
    },
    {
      "epoch": 0.2642,
      "loss_align": 0.37491416931152344,
      "loss_contrastive": 0.16329425573349,
      "loss_gen": 0.19360241293907166,
      "neg_sim_mean": 0.6883800625801086,
      "pos_sim_mean": 0.6250858306884766,
      "step": 1321
    },
    {
      "epoch": 0.2644,
      "grad_norm": 2.002584218978882,
      "learning_rate": 9.200941075056242e-05,
      "loss": 0.5881,
      "step": 1322
    },
    {
      "epoch": 0.2644,
      "loss_align": 0.2687833905220032,
      "loss_contrastive": 0.0696602463722229,
      "loss_gen": 0.2829129993915558,
      "neg_sim_mean": 0.7008768320083618,
      "pos_sim_mean": 0.7312166094779968,
      "step": 1322
    },
    {
      "epoch": 0.2646,
      "grad_norm": 2.429079294204712,
      "learning_rate": 9.199047085146415e-05,
      "loss": 0.5601,
      "step": 1323
    },
    {
      "epoch": 0.2646,
      "loss_align": 0.2440088391304016,
      "loss_contrastive": 0.12291127443313599,
      "loss_gen": 0.3188379406929016,
      "neg_sim_mean": 0.7789024114608765,
      "pos_sim_mean": 0.7559911608695984,
      "step": 1323
    },
    {
      "epoch": 0.2648,
      "grad_norm": 2.4084129333496094,
      "learning_rate": 9.19715104867205e-05,
      "loss": 0.5776,
      "step": 1324
    },
    {
      "epoch": 0.2648,
      "loss_align": 0.27101242542266846,
      "loss_contrastive": 0.1097412109375,
      "loss_gen": 0.3540312349796295,
      "neg_sim_mean": 0.7387287616729736,
      "pos_sim_mean": 0.7289875745773315,
      "step": 1324
    },
    {
      "epoch": 0.265,
      "grad_norm": 2.3926467895507812,
      "learning_rate": 9.19525296655725e-05,
      "loss": 0.6382,
      "step": 1325
    },
    {
      "epoch": 0.265,
      "loss_align": 0.3404175639152527,
      "loss_contrastive": 0.09910649061203003,
      "loss_gen": 0.170003280043602,
      "neg_sim_mean": 0.6586889028549194,
      "pos_sim_mean": 0.6595824360847473,
      "step": 1325
    },
    {
      "epoch": 0.2652,
      "grad_norm": 1.882193922996521,
      "learning_rate": 9.193352839727121e-05,
      "loss": 0.5223,
      "step": 1326
    },
    {
      "epoch": 0.2652,
      "loss_align": 0.24964040517807007,
      "loss_contrastive": 0.0972413420677185,
      "loss_gen": 0.33613598346710205,
      "neg_sim_mean": 0.7476009130477905,
      "pos_sim_mean": 0.7503595948219299,
      "step": 1326
    },
    {
      "epoch": 0.2654,
      "grad_norm": 2.1362557411193848,
      "learning_rate": 9.191450669107758e-05,
      "loss": 0.5974,
      "step": 1327
    },
    {
      "epoch": 0.2654,
      "loss_align": 0.20015156269073486,
      "loss_contrastive": 0.0,
      "loss_gen": 0.023133782669901848,
      "neg_sim_mean": 0.456468790769577,
      "pos_sim_mean": 0.7998484373092651,
      "step": 1327
    },
    {
      "epoch": 0.2656,
      "grad_norm": 1.2658374309539795,
      "learning_rate": 9.189546455626257e-05,
      "loss": 0.2233,
      "step": 1328
    },
    {
      "epoch": 0.2656,
      "loss_align": 0.37411773204803467,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012971335090696812,
      "neg_sim_mean": 0.45471829175949097,
      "pos_sim_mean": 0.6258822679519653,
      "step": 1328
    },
    {
      "epoch": 0.2658,
      "grad_norm": 0.90057772397995,
      "learning_rate": 9.18764020021071e-05,
      "loss": 0.3871,
      "step": 1329
    },
    {
      "epoch": 0.2658,
      "loss_align": 0.3975430727005005,
      "loss_contrastive": 0.13096141815185547,
      "loss_gen": 0.2770216763019562,
      "neg_sim_mean": 0.6334183216094971,
      "pos_sim_mean": 0.6024569272994995,
      "step": 1329
    },
    {
      "epoch": 0.266,
      "grad_norm": 2.277409076690674,
      "learning_rate": 9.1857319037902e-05,
      "loss": 0.6903,
      "step": 1330
    },
    {
      "epoch": 0.266,
      "loss_align": 0.24730265140533447,
      "loss_contrastive": 0.08887189626693726,
      "loss_gen": 0.19677554070949554,
      "neg_sim_mean": 0.7415692210197449,
      "pos_sim_mean": 0.7526973485946655,
      "step": 1330
    },
    {
      "epoch": 0.2662,
      "grad_norm": 1.8715251684188843,
      "learning_rate": 9.183821567294809e-05,
      "loss": 0.4547,
      "step": 1331
    },
    {
      "epoch": 0.2662,
      "loss_align": 0.2963854670524597,
      "loss_contrastive": 0.1422775387763977,
      "loss_gen": 0.33362141251564026,
      "neg_sim_mean": 0.7458920478820801,
      "pos_sim_mean": 0.7036145329475403,
      "step": 1331
    },
    {
      "epoch": 0.2664,
      "grad_norm": 2.3847317695617676,
      "learning_rate": 9.181909191655612e-05,
      "loss": 0.6471,
      "step": 1332
    },
    {
      "epoch": 0.2664,
      "loss_align": 0.3248337507247925,
      "loss_contrastive": 0.060408055782318115,
      "loss_gen": 0.445643812417984,
      "neg_sim_mean": 0.6355742812156677,
      "pos_sim_mean": 0.6751662492752075,
      "step": 1332
    },
    {
      "epoch": 0.2666,
      "grad_norm": 2.455421209335327,
      "learning_rate": 9.179994777804677e-05,
      "loss": 0.7777,
      "step": 1333
    },
    {
      "epoch": 0.2666,
      "loss_align": 0.23950856924057007,
      "loss_contrastive": 0.15010637044906616,
      "loss_gen": 0.22765140235424042,
      "neg_sim_mean": 0.8105977773666382,
      "pos_sim_mean": 0.7604914307594299,
      "step": 1333
    },
    {
      "epoch": 0.2668,
      "grad_norm": 2.0401384830474854,
      "learning_rate": 9.178078326675069e-05,
      "loss": 0.4852,
      "step": 1334
    },
    {
      "epoch": 0.2668,
      "loss_align": 0.1718130111694336,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003519507357850671,
      "neg_sim_mean": 0.36353784799575806,
      "pos_sim_mean": 0.8281869888305664,
      "step": 1334
    },
    {
      "epoch": 0.267,
      "grad_norm": 0.5028029680252075,
      "learning_rate": 9.176159839200838e-05,
      "loss": 0.1753,
      "step": 1335
    },
    {
      "epoch": 0.267,
      "loss_align": 0.33581477403640747,
      "loss_contrastive": 0.11328864097595215,
      "loss_gen": 0.15912272036075592,
      "neg_sim_mean": 0.6774738430976868,
      "pos_sim_mean": 0.6641852259635925,
      "step": 1335
    },
    {
      "epoch": 0.2672,
      "grad_norm": 1.6243637800216675,
      "learning_rate": 9.174239316317033e-05,
      "loss": 0.5085,
      "step": 1336
    },
    {
      "epoch": 0.2672,
      "loss_align": 0.1477797031402588,
      "loss_contrastive": 0.06912600994110107,
      "loss_gen": 0.20308007299900055,
      "neg_sim_mean": 0.8213462829589844,
      "pos_sim_mean": 0.8522202968597412,
      "step": 1336
    },
    {
      "epoch": 0.2674,
      "grad_norm": 2.387382745742798,
      "learning_rate": 9.172316758959696e-05,
      "loss": 0.3592,
      "step": 1337
    },
    {
      "epoch": 0.2674,
      "loss_align": 0.1750137209892273,
      "loss_contrastive": 0.08772021532058716,
      "loss_gen": 0.1625591218471527,
      "neg_sim_mean": 0.812706470489502,
      "pos_sim_mean": 0.8249862790107727,
      "step": 1337
    },
    {
      "epoch": 0.2676,
      "grad_norm": 1.805914044380188,
      "learning_rate": 9.170392168065857e-05,
      "loss": 0.3481,
      "step": 1338
    },
    {
      "epoch": 0.2676,
      "loss_align": 0.27316874265670776,
      "loss_contrastive": 0.12033063173294067,
      "loss_gen": 0.1956794708967209,
      "neg_sim_mean": 0.747161865234375,
      "pos_sim_mean": 0.7268312573432922,
      "step": 1338
    },
    {
      "epoch": 0.2678,
      "grad_norm": 2.038041830062866,
      "learning_rate": 9.168465544573536e-05,
      "loss": 0.4833,
      "step": 1339
    },
    {
      "epoch": 0.2678,
      "loss_align": 0.27739179134368896,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008275218307971954,
      "neg_sim_mean": 0.39916718006134033,
      "pos_sim_mean": 0.722608208656311,
      "step": 1339
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.6704334020614624,
      "learning_rate": 9.16653688942175e-05,
      "loss": 0.2857,
      "step": 1340
    },
    {
      "epoch": 0.268,
      "loss_align": 0.19497883319854736,
      "loss_contrastive": 0.047883033752441406,
      "loss_gen": 0.347038596868515,
      "neg_sim_mean": 0.7529041767120361,
      "pos_sim_mean": 0.8050211668014526,
      "step": 1340
    },
    {
      "epoch": 0.2682,
      "grad_norm": 2.503345489501953,
      "learning_rate": 9.164606203550497e-05,
      "loss": 0.5478,
      "step": 1341
    },
    {
      "epoch": 0.2682,
      "loss_align": 0.357852041721344,
      "loss_contrastive": 0.2246529459953308,
      "loss_gen": 0.22455401718616486,
      "neg_sim_mean": 0.7668008804321289,
      "pos_sim_mean": 0.642147958278656,
      "step": 1341
    },
    {
      "epoch": 0.2684,
      "grad_norm": 2.1124491691589355,
      "learning_rate": 9.162673487900775e-05,
      "loss": 0.6094,
      "step": 1342
    },
    {
      "epoch": 0.2684,
      "loss_align": 0.47596240043640137,
      "loss_contrastive": 0.06151825189590454,
      "loss_gen": 0.12437551468610764,
      "neg_sim_mean": 0.48555585741996765,
      "pos_sim_mean": 0.5240375995635986,
      "step": 1342
    },
    {
      "epoch": 0.2686,
      "grad_norm": 1.7151422500610352,
      "learning_rate": 9.160738743414563e-05,
      "loss": 0.6077,
      "step": 1343
    },
    {
      "epoch": 0.2686,
      "loss_align": 0.32797765731811523,
      "loss_contrastive": 0.10349303483963013,
      "loss_gen": 0.2992999851703644,
      "neg_sim_mean": 0.675515353679657,
      "pos_sim_mean": 0.6720223426818848,
      "step": 1343
    },
    {
      "epoch": 0.2688,
      "grad_norm": 2.88981294631958,
      "learning_rate": 9.158801971034832e-05,
      "loss": 0.6397,
      "step": 1344
    },
    {
      "epoch": 0.2688,
      "loss_align": 0.2943572402000427,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008504788391292095,
      "neg_sim_mean": 0.5875091552734375,
      "pos_sim_mean": 0.7056427597999573,
      "step": 1344
    },
    {
      "epoch": 0.269,
      "grad_norm": 1.4243589639663696,
      "learning_rate": 9.156863171705543e-05,
      "loss": 0.3029,
      "step": 1345
    },
    {
      "epoch": 0.269,
      "loss_align": 0.20030343532562256,
      "loss_contrastive": 0.12032943964004517,
      "loss_gen": 0.1563563346862793,
      "neg_sim_mean": 0.8200259804725647,
      "pos_sim_mean": 0.7996965646743774,
      "step": 1345
    },
    {
      "epoch": 0.2692,
      "grad_norm": 1.8472926616668701,
      "learning_rate": 9.154922346371642e-05,
      "loss": 0.3711,
      "step": 1346
    },
    {
      "epoch": 0.2692,
      "loss_align": 0.2556288242340088,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00928543321788311,
      "neg_sim_mean": 0.3700709342956543,
      "pos_sim_mean": 0.7443711757659912,
      "step": 1346
    },
    {
      "epoch": 0.2694,
      "grad_norm": 0.8127310872077942,
      "learning_rate": 9.152979495979063e-05,
      "loss": 0.2649,
      "step": 1347
    },
    {
      "epoch": 0.2694,
      "loss_align": 0.4502585530281067,
      "loss_contrastive": 0.2412167489528656,
      "loss_gen": 0.24877019226551056,
      "neg_sim_mean": 0.6909582018852234,
      "pos_sim_mean": 0.5497414469718933,
      "step": 1347
    },
    {
      "epoch": 0.2696,
      "grad_norm": 2.2501561641693115,
      "learning_rate": 9.15103462147473e-05,
      "loss": 0.728,
      "step": 1348
    },
    {
      "epoch": 0.2696,
      "loss_align": 0.2998809218406677,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14922769367694855,
      "neg_sim_mean": 0.5738180875778198,
      "pos_sim_mean": 0.7001190781593323,
      "step": 1348
    },
    {
      "epoch": 0.2698,
      "grad_norm": 2.340876817703247,
      "learning_rate": 9.149087723806548e-05,
      "loss": 0.4491,
      "step": 1349
    },
    {
      "epoch": 0.2698,
      "loss_align": 0.40574848651885986,
      "loss_contrastive": 0.04055163264274597,
      "loss_gen": 0.30652496218681335,
      "neg_sim_mean": 0.5348031520843506,
      "pos_sim_mean": 0.5942515134811401,
      "step": 1349
    },
    {
      "epoch": 0.27,
      "grad_norm": 2.3425369262695312,
      "learning_rate": 9.147138803923416e-05,
      "loss": 0.7171,
      "step": 1350
    },
    {
      "epoch": 0.27,
      "loss_align": 0.22468841075897217,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006484671961516142,
      "neg_sim_mean": 0.3944063186645508,
      "pos_sim_mean": 0.7753115892410278,
      "step": 1350
    },
    {
      "epoch": 0.2702,
      "grad_norm": 0.8129034638404846,
      "learning_rate": 9.145187862775209e-05,
      "loss": 0.2312,
      "step": 1351
    },
    {
      "epoch": 0.2702,
      "loss_align": 0.44663476943969727,
      "loss_contrastive": 0.15514907240867615,
      "loss_gen": 0.06889398396015167,
      "neg_sim_mean": 0.6085143089294434,
      "pos_sim_mean": 0.5533652305603027,
      "step": 1351
    },
    {
      "epoch": 0.2704,
      "grad_norm": 2.064791679382324,
      "learning_rate": 9.143234901312794e-05,
      "loss": 0.5341,
      "step": 1352
    },
    {
      "epoch": 0.2704,
      "loss_align": 0.3161507844924927,
      "loss_contrastive": 0.21350812911987305,
      "loss_gen": 0.33947521448135376,
      "neg_sim_mean": 0.7973573207855225,
      "pos_sim_mean": 0.6838492155075073,
      "step": 1352
    },
    {
      "epoch": 0.2706,
      "grad_norm": 2.038731098175049,
      "learning_rate": 9.141279920488021e-05,
      "loss": 0.6812,
      "step": 1353
    },
    {
      "epoch": 0.2706,
      "loss_align": 0.3133190870285034,
      "loss_contrastive": 0.14788216352462769,
      "loss_gen": 0.3217576742172241,
      "neg_sim_mean": 0.7345630526542664,
      "pos_sim_mean": 0.6866809129714966,
      "step": 1353
    },
    {
      "epoch": 0.2708,
      "grad_norm": 2.3445472717285156,
      "learning_rate": 9.139322921253724e-05,
      "loss": 0.6528,
      "step": 1354
    },
    {
      "epoch": 0.2708,
      "loss_align": 0.2253882884979248,
      "loss_contrastive": 0.1230049729347229,
      "loss_gen": 0.22309310734272003,
      "neg_sim_mean": 0.7976166605949402,
      "pos_sim_mean": 0.7746117115020752,
      "step": 1354
    },
    {
      "epoch": 0.271,
      "grad_norm": 1.806708574295044,
      "learning_rate": 9.137363904563719e-05,
      "loss": 0.4632,
      "step": 1355
    },
    {
      "epoch": 0.271,
      "loss_align": 0.14690470695495605,
      "loss_contrastive": 0.05643749237060547,
      "loss_gen": 0.22685973346233368,
      "neg_sim_mean": 0.8095327615737915,
      "pos_sim_mean": 0.853095293045044,
      "step": 1355
    },
    {
      "epoch": 0.2712,
      "grad_norm": 1.8590829372406006,
      "learning_rate": 9.135402871372808e-05,
      "loss": 0.3805,
      "step": 1356
    },
    {
      "epoch": 0.2712,
      "loss_align": 0.2591225504875183,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16609449684619904,
      "neg_sim_mean": 0.5484834909439087,
      "pos_sim_mean": 0.7408774495124817,
      "step": 1356
    },
    {
      "epoch": 0.2714,
      "grad_norm": 1.8866857290267944,
      "learning_rate": 9.133439822636778e-05,
      "loss": 0.4252,
      "step": 1357
    },
    {
      "epoch": 0.2714,
      "loss_align": 0.5453770160675049,
      "loss_contrastive": 0.28364983201026917,
      "loss_gen": 0.30761075019836426,
      "neg_sim_mean": 0.6382728219032288,
      "pos_sim_mean": 0.4546229839324951,
      "step": 1357
    },
    {
      "epoch": 0.2716,
      "grad_norm": 2.6622281074523926,
      "learning_rate": 9.13147475931239e-05,
      "loss": 0.887,
      "step": 1358
    },
    {
      "epoch": 0.2716,
      "loss_align": 0.2348954677581787,
      "loss_contrastive": 0.03216832876205444,
      "loss_gen": 0.10203266888856888,
      "neg_sim_mean": 0.6972728371620178,
      "pos_sim_mean": 0.7651045322418213,
      "step": 1358
    },
    {
      "epoch": 0.2718,
      "grad_norm": 1.4859105348587036,
      "learning_rate": 9.129507682357394e-05,
      "loss": 0.3408,
      "step": 1359
    },
    {
      "epoch": 0.2718,
      "loss_align": 0.29762399196624756,
      "loss_contrastive": 0.0,
      "loss_gen": 0.014072110876441002,
      "neg_sim_mean": 0.39349377155303955,
      "pos_sim_mean": 0.7023760080337524,
      "step": 1359
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.9298372864723206,
      "learning_rate": 9.127538592730519e-05,
      "loss": 0.3117,
      "step": 1360
    },
    {
      "epoch": 0.272,
      "loss_align": 0.19658935070037842,
      "loss_contrastive": 0.06812429428100586,
      "loss_gen": 0.28110143542289734,
      "neg_sim_mean": 0.7715349197387695,
      "pos_sim_mean": 0.8034106492996216,
      "step": 1360
    },
    {
      "epoch": 0.2722,
      "grad_norm": 2.4342544078826904,
      "learning_rate": 9.125567491391476e-05,
      "loss": 0.4859,
      "step": 1361
    },
    {
      "epoch": 0.2722,
      "loss_align": 0.3582209348678589,
      "loss_contrastive": 0.056731343269348145,
      "loss_gen": 0.2751466631889343,
      "neg_sim_mean": 0.5985103845596313,
      "pos_sim_mean": 0.6417790651321411,
      "step": 1361
    },
    {
      "epoch": 0.2724,
      "grad_norm": 2.0433945655822754,
      "learning_rate": 9.123594379300955e-05,
      "loss": 0.6402,
      "step": 1362
    },
    {
      "epoch": 0.2724,
      "loss_align": 0.21271508932113647,
      "loss_contrastive": 0.0,
      "loss_gen": 0.36129868030548096,
      "neg_sim_mean": 0.645726203918457,
      "pos_sim_mean": 0.7872849106788635,
      "step": 1362
    },
    {
      "epoch": 0.2726,
      "grad_norm": 2.3452517986297607,
      "learning_rate": 9.12161925742063e-05,
      "loss": 0.574,
      "step": 1363
    },
    {
      "epoch": 0.2726,
      "loss_align": 0.2535989284515381,
      "loss_contrastive": 0.09952712059020996,
      "loss_gen": 0.3306367099285126,
      "neg_sim_mean": 0.745928168296814,
      "pos_sim_mean": 0.7464010715484619,
      "step": 1363
    },
    {
      "epoch": 0.2728,
      "grad_norm": 2.3246097564697266,
      "learning_rate": 9.119642126713147e-05,
      "loss": 0.5962,
      "step": 1364
    },
    {
      "epoch": 0.2728,
      "loss_align": 0.24910670518875122,
      "loss_contrastive": 0.1372736692428589,
      "loss_gen": 0.235764741897583,
      "neg_sim_mean": 0.7881669402122498,
      "pos_sim_mean": 0.7508932948112488,
      "step": 1364
    },
    {
      "epoch": 0.273,
      "grad_norm": 1.915942668914795,
      "learning_rate": 9.117662988142138e-05,
      "loss": 0.5013,
      "step": 1365
    },
    {
      "epoch": 0.273,
      "loss_align": 0.29374969005584717,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009945734404027462,
      "neg_sim_mean": 0.4114963412284851,
      "pos_sim_mean": 0.7062503099441528,
      "step": 1365
    },
    {
      "epoch": 0.2732,
      "grad_norm": 0.8947063684463501,
      "learning_rate": 9.11568184267221e-05,
      "loss": 0.3037,
      "step": 1366
    },
    {
      "epoch": 0.2732,
      "loss_align": 0.2548496723175049,
      "loss_contrastive": 0.06238365173339844,
      "loss_gen": 0.17962509393692017,
      "neg_sim_mean": 0.7075339555740356,
      "pos_sim_mean": 0.7451503276824951,
      "step": 1366
    },
    {
      "epoch": 0.2734,
      "grad_norm": 1.6536005735397339,
      "learning_rate": 9.11369869126895e-05,
      "loss": 0.442,
      "step": 1367
    },
    {
      "epoch": 0.2734,
      "loss_align": 0.15309631824493408,
      "loss_contrastive": 0.11620104312896729,
      "loss_gen": 0.457759290933609,
      "neg_sim_mean": 0.8631047010421753,
      "pos_sim_mean": 0.8469036817550659,
      "step": 1367
    },
    {
      "epoch": 0.2736,
      "grad_norm": 2.7010555267333984,
      "learning_rate": 9.111713534898922e-05,
      "loss": 0.6248,
      "step": 1368
    },
    {
      "epoch": 0.2736,
      "loss_align": 0.27798378467559814,
      "loss_contrastive": 0.18826544284820557,
      "loss_gen": 0.36092787981033325,
      "neg_sim_mean": 0.8102816343307495,
      "pos_sim_mean": 0.7220162153244019,
      "step": 1368
    },
    {
      "epoch": 0.2738,
      "grad_norm": 2.426992177963257,
      "learning_rate": 9.109726374529665e-05,
      "loss": 0.6615,
      "step": 1369
    },
    {
      "epoch": 0.2738,
      "loss_align": 0.2725485563278198,
      "loss_contrastive": 0.023225843906402588,
      "loss_gen": 0.24363453686237335,
      "neg_sim_mean": 0.6506772637367249,
      "pos_sim_mean": 0.7274514436721802,
      "step": 1369
    },
    {
      "epoch": 0.274,
      "grad_norm": 1.7859058380126953,
      "learning_rate": 9.107737211129701e-05,
      "loss": 0.519,
      "step": 1370
    },
    {
      "epoch": 0.274,
      "loss_align": 0.2820996046066284,
      "loss_contrastive": 0.14117836952209473,
      "loss_gen": 0.34391844272613525,
      "neg_sim_mean": 0.7590787410736084,
      "pos_sim_mean": 0.7179003953933716,
      "step": 1370
    },
    {
      "epoch": 0.2742,
      "grad_norm": 2.509997844696045,
      "learning_rate": 9.105746045668521e-05,
      "loss": 0.643,
      "step": 1371
    },
    {
      "epoch": 0.2742,
      "loss_align": 0.3669711947441101,
      "loss_contrastive": 0.1114853024482727,
      "loss_gen": 0.22471728920936584,
      "neg_sim_mean": 0.6445140838623047,
      "pos_sim_mean": 0.6330288052558899,
      "step": 1371
    },
    {
      "epoch": 0.2744,
      "grad_norm": 2.256462335586548,
      "learning_rate": 9.103752879116595e-05,
      "loss": 0.6051,
      "step": 1372
    },
    {
      "epoch": 0.2744,
      "loss_align": 0.22798001766204834,
      "loss_contrastive": 0.10219866037368774,
      "loss_gen": 0.34789493680000305,
      "neg_sim_mean": 0.7742186188697815,
      "pos_sim_mean": 0.7720199823379517,
      "step": 1372
    },
    {
      "epoch": 0.2746,
      "grad_norm": 2.62957501411438,
      "learning_rate": 9.10175771244537e-05,
      "loss": 0.5881,
      "step": 1373
    },
    {
      "epoch": 0.2746,
      "loss_align": 0.2546055316925049,
      "loss_contrastive": 0.12288767099380493,
      "loss_gen": 0.3918096721172333,
      "neg_sim_mean": 0.7682821154594421,
      "pos_sim_mean": 0.7453944683074951,
      "step": 1373
    },
    {
      "epoch": 0.2748,
      "grad_norm": 2.5758814811706543,
      "learning_rate": 9.099760546627261e-05,
      "loss": 0.6612,
      "step": 1374
    },
    {
      "epoch": 0.2748,
      "loss_align": 0.1335161328315735,
      "loss_contrastive": 0.06538766622543335,
      "loss_gen": 0.3551633954048157,
      "neg_sim_mean": 0.831871509552002,
      "pos_sim_mean": 0.8664838671684265,
      "step": 1374
    },
    {
      "epoch": 0.275,
      "grad_norm": 2.482902765274048,
      "learning_rate": 9.09776138263567e-05,
      "loss": 0.4965,
      "step": 1375
    },
    {
      "epoch": 0.275,
      "loss_align": 0.2982722520828247,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008501337841153145,
      "neg_sim_mean": 0.45723551511764526,
      "pos_sim_mean": 0.7017277479171753,
      "step": 1375
    },
    {
      "epoch": 0.2752,
      "grad_norm": 0.6887929439544678,
      "learning_rate": 9.09576022144496e-05,
      "loss": 0.3068,
      "step": 1376
    },
    {
      "epoch": 0.2752,
      "loss_align": 0.39744365215301514,
      "loss_contrastive": 0.09092545509338379,
      "loss_gen": 0.15464115142822266,
      "neg_sim_mean": 0.5934817790985107,
      "pos_sim_mean": 0.6025563478469849,
      "step": 1376
    },
    {
      "epoch": 0.2754,
      "grad_norm": 1.7988122701644897,
      "learning_rate": 9.093757064030473e-05,
      "loss": 0.563,
      "step": 1377
    },
    {
      "epoch": 0.2754,
      "loss_align": 0.13721394538879395,
      "loss_contrastive": 0.02467125654220581,
      "loss_gen": 0.2969443202018738,
      "neg_sim_mean": 0.787457287311554,
      "pos_sim_mean": 0.862786054611206,
      "step": 1377
    },
    {
      "epoch": 0.2756,
      "grad_norm": 2.193944215774536,
      "learning_rate": 9.091751911368525e-05,
      "loss": 0.4371,
      "step": 1378
    },
    {
      "epoch": 0.2756,
      "loss_align": 0.10618454217910767,
      "loss_contrastive": 0.07147711515426636,
      "loss_gen": 0.5486994981765747,
      "neg_sim_mean": 0.8652925491333008,
      "pos_sim_mean": 0.8938154578208923,
      "step": 1378
    },
    {
      "epoch": 0.2758,
      "grad_norm": 3.0467021465301514,
      "learning_rate": 9.089744764436403e-05,
      "loss": 0.6635,
      "step": 1379
    },
    {
      "epoch": 0.2758,
      "loss_align": 0.3463495969772339,
      "loss_contrastive": 0.1829148530960083,
      "loss_gen": 0.21213865280151367,
      "neg_sim_mean": 0.7365652322769165,
      "pos_sim_mean": 0.6536504030227661,
      "step": 1379
    },
    {
      "epoch": 0.276,
      "grad_norm": 2.08199405670166,
      "learning_rate": 9.087735624212365e-05,
      "loss": 0.5804,
      "step": 1380
    },
    {
      "epoch": 0.276,
      "loss_align": 0.2620844841003418,
      "loss_contrastive": 0.12810778617858887,
      "loss_gen": 0.4131752848625183,
      "neg_sim_mean": 0.7660232782363892,
      "pos_sim_mean": 0.7379155158996582,
      "step": 1380
    },
    {
      "epoch": 0.2762,
      "grad_norm": 2.3555164337158203,
      "learning_rate": 9.085724491675642e-05,
      "loss": 0.6906,
      "step": 1381
    },
    {
      "epoch": 0.2762,
      "loss_align": 0.17655229568481445,
      "loss_contrastive": 0.004427552223205566,
      "loss_gen": 0.22576788067817688,
      "neg_sim_mean": 0.7278752326965332,
      "pos_sim_mean": 0.8234477043151855,
      "step": 1381
    },
    {
      "epoch": 0.2764,
      "grad_norm": 1.5583312511444092,
      "learning_rate": 9.083711367806438e-05,
      "loss": 0.4029,
      "step": 1382
    },
    {
      "epoch": 0.2764,
      "loss_align": 0.25177139043807983,
      "loss_contrastive": 0.1341632604598999,
      "loss_gen": 0.37658339738845825,
      "neg_sim_mean": 0.7823918461799622,
      "pos_sim_mean": 0.7482286095619202,
      "step": 1382
    },
    {
      "epoch": 0.2766,
      "grad_norm": 3.761183977127075,
      "learning_rate": 9.081696253585921e-05,
      "loss": 0.6445,
      "step": 1383
    },
    {
      "epoch": 0.2766,
      "loss_align": 0.2544978857040405,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006158818956464529,
      "neg_sim_mean": 0.43900319933891296,
      "pos_sim_mean": 0.7455021142959595,
      "step": 1383
    },
    {
      "epoch": 0.2768,
      "grad_norm": 0.5604974031448364,
      "learning_rate": 9.079679149996234e-05,
      "loss": 0.2607,
      "step": 1384
    },
    {
      "epoch": 0.2768,
      "loss_align": 0.18246042728424072,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02948681451380253,
      "neg_sim_mean": 0.3463250994682312,
      "pos_sim_mean": 0.8175395727157593,
      "step": 1384
    },
    {
      "epoch": 0.277,
      "grad_norm": 1.2237426042556763,
      "learning_rate": 9.077660058020491e-05,
      "loss": 0.2119,
      "step": 1385
    },
    {
      "epoch": 0.277,
      "loss_align": 0.32738959789276123,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006747976411134005,
      "neg_sim_mean": 0.5410257577896118,
      "pos_sim_mean": 0.6726104021072388,
      "step": 1385
    },
    {
      "epoch": 0.2772,
      "grad_norm": 0.6796273589134216,
      "learning_rate": 9.075638978642771e-05,
      "loss": 0.3341,
      "step": 1386
    },
    {
      "epoch": 0.2772,
      "loss_align": 0.26803040504455566,
      "loss_contrastive": 0.1121014952659607,
      "loss_gen": 0.12568938732147217,
      "neg_sim_mean": 0.7440710663795471,
      "pos_sim_mean": 0.7319695949554443,
      "step": 1386
    },
    {
      "epoch": 0.2774,
      "grad_norm": 1.5228427648544312,
      "learning_rate": 9.073615912848126e-05,
      "loss": 0.4072,
      "step": 1387
    },
    {
      "epoch": 0.2774,
      "loss_align": 0.16599225997924805,
      "loss_contrastive": 0.0864112377166748,
      "loss_gen": 0.36283576488494873,
      "neg_sim_mean": 0.8204189538955688,
      "pos_sim_mean": 0.834007740020752,
      "step": 1387
    },
    {
      "epoch": 0.2776,
      "grad_norm": 2.279644727706909,
      "learning_rate": 9.07159086162257e-05,
      "loss": 0.5392,
      "step": 1388
    },
    {
      "epoch": 0.2776,
      "loss_align": 0.15400242805480957,
      "loss_contrastive": 0.0,
      "loss_gen": 0.014393056742846966,
      "neg_sim_mean": 0.20127275586128235,
      "pos_sim_mean": 0.8459975719451904,
      "step": 1388
    },
    {
      "epoch": 0.2778,
      "grad_norm": 0.9075749516487122,
      "learning_rate": 9.069563825953092e-05,
      "loss": 0.1684,
      "step": 1389
    },
    {
      "epoch": 0.2778,
      "loss_align": 0.24837255477905273,
      "loss_contrastive": 0.14171981811523438,
      "loss_gen": 0.2534845471382141,
      "neg_sim_mean": 0.7933472394943237,
      "pos_sim_mean": 0.7516274452209473,
      "step": 1389
    },
    {
      "epoch": 0.278,
      "grad_norm": 1.8667951822280884,
      "learning_rate": 9.06753480682764e-05,
      "loss": 0.5189,
      "step": 1390
    },
    {
      "epoch": 0.278,
      "loss_align": 0.2555427551269531,
      "loss_contrastive": 0.18773174285888672,
      "loss_gen": 0.4250108003616333,
      "neg_sim_mean": 0.8321889638900757,
      "pos_sim_mean": 0.7444572448730469,
      "step": 1390
    },
    {
      "epoch": 0.2782,
      "grad_norm": 2.4618003368377686,
      "learning_rate": 9.065503805235138e-05,
      "loss": 0.7031,
      "step": 1391
    },
    {
      "epoch": 0.2782,
      "loss_align": 0.2930319905281067,
      "loss_contrastive": 0.08460813760757446,
      "loss_gen": 0.2627771496772766,
      "neg_sim_mean": 0.6915761232376099,
      "pos_sim_mean": 0.7069680094718933,
      "step": 1391
    },
    {
      "epoch": 0.2784,
      "grad_norm": 2.1307549476623535,
      "learning_rate": 9.06347082216547e-05,
      "loss": 0.566,
      "step": 1392
    },
    {
      "epoch": 0.2784,
      "loss_align": 0.21739786863327026,
      "loss_contrastive": 0.0,
      "loss_gen": 0.30634236335754395,
      "neg_sim_mean": 0.4507838487625122,
      "pos_sim_mean": 0.7826021313667297,
      "step": 1392
    },
    {
      "epoch": 0.2786,
      "grad_norm": 2.1888227462768555,
      "learning_rate": 9.061435858609486e-05,
      "loss": 0.5237,
      "step": 1393
    },
    {
      "epoch": 0.2786,
      "loss_align": 0.15226471424102783,
      "loss_contrastive": 0.08973181247711182,
      "loss_gen": 0.24666400253772736,
      "neg_sim_mean": 0.8374670743942261,
      "pos_sim_mean": 0.8477352857589722,
      "step": 1393
    },
    {
      "epoch": 0.2788,
      "grad_norm": 1.7737308740615845,
      "learning_rate": 9.059398915559006e-05,
      "loss": 0.4097,
      "step": 1394
    },
    {
      "epoch": 0.2788,
      "loss_align": 0.30170583724975586,
      "loss_contrastive": 0.06829166412353516,
      "loss_gen": 0.23237748444080353,
      "neg_sim_mean": 0.6665858030319214,
      "pos_sim_mean": 0.6982941627502441,
      "step": 1394
    },
    {
      "epoch": 0.279,
      "grad_norm": 1.7427688837051392,
      "learning_rate": 9.057359994006806e-05,
      "loss": 0.5423,
      "step": 1395
    },
    {
      "epoch": 0.279,
      "loss_align": 0.08408576250076294,
      "loss_contrastive": 0.05349379777908325,
      "loss_gen": 0.2201918065547943,
      "neg_sim_mean": 0.8694080114364624,
      "pos_sim_mean": 0.9159142374992371,
      "step": 1395
    },
    {
      "epoch": 0.2792,
      "grad_norm": 1.8103233575820923,
      "learning_rate": 9.055319094946633e-05,
      "loss": 0.3107,
      "step": 1396
    },
    {
      "epoch": 0.2792,
      "loss_align": 0.18065297603607178,
      "loss_contrastive": 0.07707607746124268,
      "loss_gen": 0.23177510499954224,
      "neg_sim_mean": 0.796423077583313,
      "pos_sim_mean": 0.8193470239639282,
      "step": 1396
    },
    {
      "epoch": 0.2794,
      "grad_norm": 2.0538713932037354,
      "learning_rate": 9.0532762193732e-05,
      "loss": 0.4217,
      "step": 1397
    },
    {
      "epoch": 0.2794,
      "loss_align": 0.21574938297271729,
      "loss_contrastive": 0.1422652006149292,
      "loss_gen": 0.27878597378730774,
      "neg_sim_mean": 0.826515793800354,
      "pos_sim_mean": 0.7842506170272827,
      "step": 1397
    },
    {
      "epoch": 0.2796,
      "grad_norm": 1.8693212270736694,
      "learning_rate": 9.051231368282177e-05,
      "loss": 0.5116,
      "step": 1398
    },
    {
      "epoch": 0.2796,
      "loss_align": 0.1640310287475586,
      "loss_contrastive": 0.15140008926391602,
      "loss_gen": 0.18134461343288422,
      "neg_sim_mean": 0.8873690366744995,
      "pos_sim_mean": 0.8359689712524414,
      "step": 1398
    },
    {
      "epoch": 0.2798,
      "grad_norm": 1.7646918296813965,
      "learning_rate": 9.049184542670199e-05,
      "loss": 0.3635,
      "step": 1399
    },
    {
      "epoch": 0.2798,
      "loss_align": 0.22936546802520752,
      "loss_contrastive": 0.05300164222717285,
      "loss_gen": 0.20751889050006866,
      "neg_sim_mean": 0.7236361503601074,
      "pos_sim_mean": 0.7706345319747925,
      "step": 1399
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.7226871252059937,
      "learning_rate": 9.047135743534868e-05,
      "loss": 0.4432,
      "step": 1400
    },
    {
      "epoch": 0.28,
      "loss_align": 0.18093568086624146,
      "loss_contrastive": 0.0,
      "loss_gen": 0.017381587997078896,
      "neg_sim_mean": 0.35522836446762085,
      "pos_sim_mean": 0.8190643191337585,
      "step": 1400
    },
    {
      "epoch": 0.2802,
      "grad_norm": 0.9217726588249207,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.1983,
      "step": 1401
    },
    {
      "epoch": 0.2802,
      "loss_align": 0.13054126501083374,
      "loss_contrastive": 0.10484111309051514,
      "loss_gen": 0.21294093132019043,
      "neg_sim_mean": 0.8742998242378235,
      "pos_sim_mean": 0.8694587349891663,
      "step": 1401
    },
    {
      "epoch": 0.2804,
      "grad_norm": 2.0623738765716553,
      "learning_rate": 9.043032228689333e-05,
      "loss": 0.3561,
      "step": 1402
    },
    {
      "epoch": 0.2804,
      "loss_align": 0.1445755958557129,
      "loss_contrastive": 0.06248629093170166,
      "loss_gen": 0.20937085151672363,
      "neg_sim_mean": 0.8179106712341309,
      "pos_sim_mean": 0.8554244041442871,
      "step": 1402
    },
    {
      "epoch": 0.2806,
      "grad_norm": 2.121826410293579,
      "learning_rate": 9.040977514979137e-05,
      "loss": 0.3614,
      "step": 1403
    },
    {
      "epoch": 0.2806,
      "loss_align": 0.24184393882751465,
      "loss_contrastive": 0.07185286283493042,
      "loss_gen": 0.2654222249984741,
      "neg_sim_mean": 0.7300089001655579,
      "pos_sim_mean": 0.7581560611724854,
      "step": 1403
    },
    {
      "epoch": 0.2808,
      "grad_norm": 1.9822912216186523,
      "learning_rate": 9.038920831745587e-05,
      "loss": 0.5159,
      "step": 1404
    },
    {
      "epoch": 0.2808,
      "loss_align": 0.15431737899780273,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29245224595069885,
      "neg_sim_mean": 0.5084807276725769,
      "pos_sim_mean": 0.8456826210021973,
      "step": 1404
    },
    {
      "epoch": 0.281,
      "grad_norm": 2.167086601257324,
      "learning_rate": 9.036862179991093e-05,
      "loss": 0.4468,
      "step": 1405
    },
    {
      "epoch": 0.281,
      "loss_align": 0.22941523790359497,
      "loss_contrastive": 0.10215622186660767,
      "loss_gen": 0.2882685363292694,
      "neg_sim_mean": 0.7727409601211548,
      "pos_sim_mean": 0.770584762096405,
      "step": 1405
    },
    {
      "epoch": 0.2812,
      "grad_norm": 2.295203447341919,
      "learning_rate": 9.034801560719011e-05,
      "loss": 0.5299,
      "step": 1406
    },
    {
      "epoch": 0.2812,
      "loss_align": 0.2734227776527405,
      "loss_contrastive": 0.14796727895736694,
      "loss_gen": 0.3943270444869995,
      "neg_sim_mean": 0.7745444774627686,
      "pos_sim_mean": 0.7265772223472595,
      "step": 1406
    },
    {
      "epoch": 0.2814,
      "grad_norm": 2.6024391651153564,
      "learning_rate": 9.032738974933664e-05,
      "loss": 0.6855,
      "step": 1407
    },
    {
      "epoch": 0.2814,
      "loss_align": 0.4577629566192627,
      "loss_contrastive": 0.2557184398174286,
      "loss_gen": 0.21992933750152588,
      "neg_sim_mean": 0.6979554891586304,
      "pos_sim_mean": 0.5422370433807373,
      "step": 1407
    },
    {
      "epoch": 0.2816,
      "grad_norm": 1.6634706258773804,
      "learning_rate": 9.03067442364033e-05,
      "loss": 0.7084,
      "step": 1408
    },
    {
      "epoch": 0.2816,
      "loss_align": 0.25419920682907104,
      "loss_contrastive": 0.08404666185379028,
      "loss_gen": 0.43996623158454895,
      "neg_sim_mean": 0.7298474311828613,
      "pos_sim_mean": 0.745800793170929,
      "step": 1408
    },
    {
      "epoch": 0.2818,
      "grad_norm": 2.378164052963257,
      "learning_rate": 9.028607907845247e-05,
      "loss": 0.7043,
      "step": 1409
    },
    {
      "epoch": 0.2818,
      "loss_align": 0.22740519046783447,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02312346175312996,
      "neg_sim_mean": 0.4495953917503357,
      "pos_sim_mean": 0.7725948095321655,
      "step": 1409
    },
    {
      "epoch": 0.282,
      "grad_norm": 1.3587907552719116,
      "learning_rate": 9.02653942855561e-05,
      "loss": 0.2505,
      "step": 1410
    },
    {
      "epoch": 0.282,
      "loss_align": 0.3997762203216553,
      "loss_contrastive": 0.10957074165344238,
      "loss_gen": 0.33939382433891296,
      "neg_sim_mean": 0.6097944974899292,
      "pos_sim_mean": 0.6002237796783447,
      "step": 1410
    },
    {
      "epoch": 0.2822,
      "grad_norm": 1.932094693183899,
      "learning_rate": 9.02446898677957e-05,
      "loss": 0.7523,
      "step": 1411
    },
    {
      "epoch": 0.2822,
      "loss_align": 0.40971577167510986,
      "loss_contrastive": 0.2249208390712738,
      "loss_gen": 0.3311426341533661,
      "neg_sim_mean": 0.7152050733566284,
      "pos_sim_mean": 0.5902842283248901,
      "step": 1411
    },
    {
      "epoch": 0.2824,
      "grad_norm": 2.6926169395446777,
      "learning_rate": 9.022396583526239e-05,
      "loss": 0.7678,
      "step": 1412
    },
    {
      "epoch": 0.2824,
      "loss_align": 0.32267481088638306,
      "loss_contrastive": 0.0,
      "loss_gen": 0.026233036071062088,
      "neg_sim_mean": 0.38846200704574585,
      "pos_sim_mean": 0.6773251891136169,
      "step": 1412
    },
    {
      "epoch": 0.2826,
      "grad_norm": 1.2724858522415161,
      "learning_rate": 9.020322219805674e-05,
      "loss": 0.3489,
      "step": 1413
    },
    {
      "epoch": 0.2826,
      "loss_align": 0.3445572257041931,
      "loss_contrastive": 0.10109305381774902,
      "loss_gen": 0.19939151406288147,
      "neg_sim_mean": 0.656535804271698,
      "pos_sim_mean": 0.6554427742958069,
      "step": 1413
    },
    {
      "epoch": 0.2828,
      "grad_norm": 1.5931241512298584,
      "learning_rate": 9.018245896628899e-05,
      "loss": 0.5561,
      "step": 1414
    },
    {
      "epoch": 0.2828,
      "loss_align": 0.33739233016967773,
      "loss_contrastive": 0.05374199151992798,
      "loss_gen": 0.3271501958370209,
      "neg_sim_mean": 0.6163496375083923,
      "pos_sim_mean": 0.6626076698303223,
      "step": 1414
    },
    {
      "epoch": 0.283,
      "grad_norm": 1.9016315937042236,
      "learning_rate": 9.01616761500789e-05,
      "loss": 0.671,
      "step": 1415
    },
    {
      "epoch": 0.283,
      "loss_align": 0.3637472987174988,
      "loss_contrastive": 0.2301056981086731,
      "loss_gen": 0.2764038145542145,
      "neg_sim_mean": 0.7663583755493164,
      "pos_sim_mean": 0.6362527012825012,
      "step": 1415
    },
    {
      "epoch": 0.2832,
      "grad_norm": 2.1906840801239014,
      "learning_rate": 9.014087375955573e-05,
      "loss": 0.6678,
      "step": 1416
    },
    {
      "epoch": 0.2832,
      "loss_align": 0.14883625507354736,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011988038197159767,
      "neg_sim_mean": 0.4538556933403015,
      "pos_sim_mean": 0.8511637449264526,
      "step": 1416
    },
    {
      "epoch": 0.2834,
      "grad_norm": 0.7023838758468628,
      "learning_rate": 9.012005180485835e-05,
      "loss": 0.1608,
      "step": 1417
    },
    {
      "epoch": 0.2834,
      "loss_align": 0.1499783992767334,
      "loss_contrastive": 0.09437680244445801,
      "loss_gen": 0.23518984019756317,
      "neg_sim_mean": 0.8443983793258667,
      "pos_sim_mean": 0.8500216007232666,
      "step": 1417
    },
    {
      "epoch": 0.2836,
      "grad_norm": 1.6199089288711548,
      "learning_rate": 9.009921029613506e-05,
      "loss": 0.3965,
      "step": 1418
    },
    {
      "epoch": 0.2836,
      "loss_align": 0.24784386157989502,
      "loss_contrastive": 0.08666396141052246,
      "loss_gen": 0.2712171673774719,
      "neg_sim_mean": 0.7388200759887695,
      "pos_sim_mean": 0.752156138420105,
      "step": 1418
    },
    {
      "epoch": 0.2838,
      "grad_norm": 2.4452760219573975,
      "learning_rate": 9.007834924354383e-05,
      "loss": 0.5295,
      "step": 1419
    },
    {
      "epoch": 0.2838,
      "loss_align": 0.2844933271408081,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011218626983463764,
      "neg_sim_mean": 0.4327086806297302,
      "pos_sim_mean": 0.7155066728591919,
      "step": 1419
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.9373879432678223,
      "learning_rate": 9.005746865725207e-05,
      "loss": 0.2957,
      "step": 1420
    },
    {
      "epoch": 0.284,
      "loss_align": 0.20457732677459717,
      "loss_contrastive": 0.1634446382522583,
      "loss_gen": 0.2277696579694748,
      "neg_sim_mean": 0.8588672876358032,
      "pos_sim_mean": 0.7954226732254028,
      "step": 1420
    },
    {
      "epoch": 0.2842,
      "grad_norm": 2.1984176635742188,
      "learning_rate": 9.003656854743667e-05,
      "loss": 0.452,
      "step": 1421
    },
    {
      "epoch": 0.2842,
      "loss_align": 0.2871590256690979,
      "loss_contrastive": 0.10722178220748901,
      "loss_gen": 0.19786758720874786,
      "neg_sim_mean": 0.7200627326965332,
      "pos_sim_mean": 0.7128409743309021,
      "step": 1421
    },
    {
      "epoch": 0.2844,
      "grad_norm": 1.8509567975997925,
      "learning_rate": 9.001564892428415e-05,
      "loss": 0.4979,
      "step": 1422
    },
    {
      "epoch": 0.2844,
      "loss_align": 0.3664965033531189,
      "loss_contrastive": 0.13859015703201294,
      "loss_gen": 0.16326040029525757,
      "neg_sim_mean": 0.6720936298370361,
      "pos_sim_mean": 0.6335034966468811,
      "step": 1422
    },
    {
      "epoch": 0.2846,
      "grad_norm": 1.5954328775405884,
      "learning_rate": 8.999470979799048e-05,
      "loss": 0.5464,
      "step": 1423
    },
    {
      "epoch": 0.2846,
      "loss_align": 0.401095449924469,
      "loss_contrastive": 0.20699408650398254,
      "loss_gen": 0.167668879032135,
      "neg_sim_mean": 0.705898642539978,
      "pos_sim_mean": 0.598904550075531,
      "step": 1423
    },
    {
      "epoch": 0.2848,
      "grad_norm": 1.7198703289031982,
      "learning_rate": 8.99737511787611e-05,
      "loss": 0.5936,
      "step": 1424
    },
    {
      "epoch": 0.2848,
      "loss_align": 0.28810787200927734,
      "loss_contrastive": 0.21701455116271973,
      "loss_gen": 0.21585430204868317,
      "neg_sim_mean": 0.8289066553115845,
      "pos_sim_mean": 0.7118921279907227,
      "step": 1424
    },
    {
      "epoch": 0.285,
      "grad_norm": 2.480196714401245,
      "learning_rate": 8.995277307681099e-05,
      "loss": 0.53,
      "step": 1425
    },
    {
      "epoch": 0.285,
      "loss_align": 0.30144643783569336,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011317221447825432,
      "neg_sim_mean": 0.46199458837509155,
      "pos_sim_mean": 0.6985535621643066,
      "step": 1425
    },
    {
      "epoch": 0.2852,
      "grad_norm": 0.8206113576889038,
      "learning_rate": 8.993177550236464e-05,
      "loss": 0.3128,
      "step": 1426
    },
    {
      "epoch": 0.2852,
      "loss_align": 0.27764707803726196,
      "loss_contrastive": 0.0,
      "loss_gen": 0.015357396565377712,
      "neg_sim_mean": 0.5526482462882996,
      "pos_sim_mean": 0.722352921962738,
      "step": 1426
    },
    {
      "epoch": 0.2854,
      "grad_norm": 0.9145209193229675,
      "learning_rate": 8.991075846565602e-05,
      "loss": 0.293,
      "step": 1427
    },
    {
      "epoch": 0.2854,
      "loss_align": 0.30827295780181885,
      "loss_contrastive": 0.16944479942321777,
      "loss_gen": 0.1816585808992386,
      "neg_sim_mean": 0.761171817779541,
      "pos_sim_mean": 0.6917270421981812,
      "step": 1427
    },
    {
      "epoch": 0.2856,
      "grad_norm": 2.147658586502075,
      "learning_rate": 8.988972197692855e-05,
      "loss": 0.5103,
      "step": 1428
    },
    {
      "epoch": 0.2856,
      "loss_align": 0.4007452130317688,
      "loss_contrastive": 0.027881771326065063,
      "loss_gen": 0.007695936132222414,
      "neg_sim_mean": 0.5271365642547607,
      "pos_sim_mean": 0.5992547869682312,
      "step": 1428
    },
    {
      "epoch": 0.2858,
      "grad_norm": 1.2155873775482178,
      "learning_rate": 8.986866604643518e-05,
      "loss": 0.4118,
      "step": 1429
    },
    {
      "epoch": 0.2858,
      "loss_align": 0.11034542322158813,
      "loss_contrastive": 0.040368854999542236,
      "loss_gen": 0.12356262654066086,
      "neg_sim_mean": 0.8300234079360962,
      "pos_sim_mean": 0.8896545767784119,
      "step": 1429
    },
    {
      "epoch": 0.286,
      "grad_norm": 1.3839068412780762,
      "learning_rate": 8.984759068443831e-05,
      "loss": 0.2388,
      "step": 1430
    },
    {
      "epoch": 0.286,
      "loss_align": 0.2962380647659302,
      "loss_contrastive": 0.12424564361572266,
      "loss_gen": 0.3033168911933899,
      "neg_sim_mean": 0.7280075550079346,
      "pos_sim_mean": 0.7037619352340698,
      "step": 1430
    },
    {
      "epoch": 0.2862,
      "grad_norm": 2.1695735454559326,
      "learning_rate": 8.982649590120982e-05,
      "loss": 0.6145,
      "step": 1431
    },
    {
      "epoch": 0.2862,
      "loss_align": 0.23086941242218018,
      "loss_contrastive": 0.0957176685333252,
      "loss_gen": 0.3177970051765442,
      "neg_sim_mean": 0.7648482322692871,
      "pos_sim_mean": 0.7691305875778198,
      "step": 1431
    },
    {
      "epoch": 0.2864,
      "grad_norm": 2.1540942192077637,
      "learning_rate": 8.980538170703104e-05,
      "loss": 0.5602,
      "step": 1432
    },
    {
      "epoch": 0.2864,
      "loss_align": 0.17501616477966309,
      "loss_contrastive": 0.08780944347381592,
      "loss_gen": 0.34413257241249084,
      "neg_sim_mean": 0.8127932548522949,
      "pos_sim_mean": 0.8249838352203369,
      "step": 1432
    },
    {
      "epoch": 0.2866,
      "grad_norm": 2.359220266342163,
      "learning_rate": 8.978424811219277e-05,
      "loss": 0.5297,
      "step": 1433
    },
    {
      "epoch": 0.2866,
      "loss_align": 0.3071504831314087,
      "loss_contrastive": 0.16877615451812744,
      "loss_gen": 0.5479201674461365,
      "neg_sim_mean": 0.7616256475448608,
      "pos_sim_mean": 0.6928495168685913,
      "step": 1433
    },
    {
      "epoch": 0.2868,
      "grad_norm": 2.8743298053741455,
      "learning_rate": 8.976309512699528e-05,
      "loss": 0.8753,
      "step": 1434
    },
    {
      "epoch": 0.2868,
      "loss_align": 0.28773558139801025,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0019271643832325935,
      "neg_sim_mean": 0.37796300649642944,
      "pos_sim_mean": 0.7122644186019897,
      "step": 1434
    },
    {
      "epoch": 0.287,
      "grad_norm": 0.4790099859237671,
      "learning_rate": 8.974192276174829e-05,
      "loss": 0.2897,
      "step": 1435
    },
    {
      "epoch": 0.287,
      "loss_align": 0.5384359359741211,
      "loss_contrastive": 0.1165294349193573,
      "loss_gen": 0.11073250323534012,
      "neg_sim_mean": 0.4780935049057007,
      "pos_sim_mean": 0.4615640640258789,
      "step": 1435
    },
    {
      "epoch": 0.2872,
      "grad_norm": 2.058934211730957,
      "learning_rate": 8.972073102677091e-05,
      "loss": 0.6632,
      "step": 1436
    },
    {
      "epoch": 0.2872,
      "loss_align": 0.33629363775253296,
      "loss_contrastive": 0.2178615927696228,
      "loss_gen": 0.434631884098053,
      "neg_sim_mean": 0.7815679311752319,
      "pos_sim_mean": 0.663706362247467,
      "step": 1436
    },
    {
      "epoch": 0.2874,
      "grad_norm": 2.5669338703155518,
      "learning_rate": 8.969951993239177e-05,
      "loss": 0.7971,
      "step": 1437
    },
    {
      "epoch": 0.2874,
      "loss_align": 0.39539241790771484,
      "loss_contrastive": 0.16788733005523682,
      "loss_gen": 0.3685917258262634,
      "neg_sim_mean": 0.6724948883056641,
      "pos_sim_mean": 0.6046075820922852,
      "step": 1437
    },
    {
      "epoch": 0.2876,
      "grad_norm": 2.1533124446868896,
      "learning_rate": 8.96782894889489e-05,
      "loss": 0.7841,
      "step": 1438
    },
    {
      "epoch": 0.2876,
      "loss_align": 0.3284207582473755,
      "loss_contrastive": 0.24523329734802246,
      "loss_gen": 0.22332698106765747,
      "neg_sim_mean": 0.8168125152587891,
      "pos_sim_mean": 0.6715792417526245,
      "step": 1438
    },
    {
      "epoch": 0.2878,
      "grad_norm": 1.967281699180603,
      "learning_rate": 8.965703970678974e-05,
      "loss": 0.5812,
      "step": 1439
    },
    {
      "epoch": 0.2878,
      "loss_align": 0.3105289936065674,
      "loss_contrastive": 0.11394107341766357,
      "loss_gen": 0.13814544677734375,
      "neg_sim_mean": 0.7034120559692383,
      "pos_sim_mean": 0.6894710063934326,
      "step": 1439
    },
    {
      "epoch": 0.288,
      "grad_norm": 1.4834585189819336,
      "learning_rate": 8.963577059627118e-05,
      "loss": 0.4623,
      "step": 1440
    },
    {
      "epoch": 0.288,
      "loss_align": 0.209122896194458,
      "loss_contrastive": 0.0985291600227356,
      "loss_gen": 0.21384559571743011,
      "neg_sim_mean": 0.7894062399864197,
      "pos_sim_mean": 0.790877103805542,
      "step": 1440
    },
    {
      "epoch": 0.2882,
      "grad_norm": 2.1185193061828613,
      "learning_rate": 8.961448216775954e-05,
      "loss": 0.4348,
      "step": 1441
    },
    {
      "epoch": 0.2882,
      "loss_align": 0.10068678855895996,
      "loss_contrastive": 0.03593087196350098,
      "loss_gen": 0.1079675629734993,
      "neg_sim_mean": 0.8352440595626831,
      "pos_sim_mean": 0.89931321144104,
      "step": 1441
    },
    {
      "epoch": 0.2884,
      "grad_norm": 1.2510671615600586,
      "learning_rate": 8.959317443163054e-05,
      "loss": 0.213,
      "step": 1442
    },
    {
      "epoch": 0.2884,
      "loss_align": 0.15001589059829712,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2688310146331787,
      "neg_sim_mean": 0.7222028970718384,
      "pos_sim_mean": 0.8499841094017029,
      "step": 1442
    },
    {
      "epoch": 0.2886,
      "grad_norm": 2.2506535053253174,
      "learning_rate": 8.95718473982693e-05,
      "loss": 0.4188,
      "step": 1443
    },
    {
      "epoch": 0.2886,
      "loss_align": 0.20398396253585815,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004635953810065985,
      "neg_sim_mean": 0.6835967302322388,
      "pos_sim_mean": 0.7960160374641418,
      "step": 1443
    },
    {
      "epoch": 0.2888,
      "grad_norm": 0.48049455881118774,
      "learning_rate": 8.955050107807035e-05,
      "loss": 0.2086,
      "step": 1444
    },
    {
      "epoch": 0.2888,
      "loss_align": 0.2917017936706543,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006076182704418898,
      "neg_sim_mean": 0.47051650285720825,
      "pos_sim_mean": 0.7082982063293457,
      "step": 1444
    },
    {
      "epoch": 0.289,
      "grad_norm": 0.6968604922294617,
      "learning_rate": 8.952913548143765e-05,
      "loss": 0.2978,
      "step": 1445
    },
    {
      "epoch": 0.289,
      "loss_align": 0.3553581237792969,
      "loss_contrastive": 0.12444216012954712,
      "loss_gen": 0.5024534463882446,
      "neg_sim_mean": 0.6690840125083923,
      "pos_sim_mean": 0.6446418762207031,
      "step": 1445
    },
    {
      "epoch": 0.2892,
      "grad_norm": 3.2336642742156982,
      "learning_rate": 8.950775061878453e-05,
      "loss": 0.8727,
      "step": 1446
    },
    {
      "epoch": 0.2892,
      "loss_align": 0.5704221725463867,
      "loss_contrastive": 0.31942325830459595,
      "loss_gen": 0.38977959752082825,
      "neg_sim_mean": 0.6490010619163513,
      "pos_sim_mean": 0.4295777976512909,
      "step": 1446
    },
    {
      "epoch": 0.2894,
      "grad_norm": 2.755042314529419,
      "learning_rate": 8.948634650053369e-05,
      "loss": 0.9985,
      "step": 1447
    },
    {
      "epoch": 0.2894,
      "loss_align": 0.2762088179588318,
      "loss_contrastive": 0.08997434377670288,
      "loss_gen": 0.295868843793869,
      "neg_sim_mean": 0.7137655019760132,
      "pos_sim_mean": 0.7237911820411682,
      "step": 1447
    },
    {
      "epoch": 0.2896,
      "grad_norm": 1.8808990716934204,
      "learning_rate": 8.946492313711725e-05,
      "loss": 0.5829,
      "step": 1448
    },
    {
      "epoch": 0.2896,
      "loss_align": 0.30892276763916016,
      "loss_contrastive": 0.11276113986968994,
      "loss_gen": 0.26731863617897034,
      "neg_sim_mean": 0.7038383483886719,
      "pos_sim_mean": 0.6910772323608398,
      "step": 1448
    },
    {
      "epoch": 0.2898,
      "grad_norm": 1.7101833820343018,
      "learning_rate": 8.944348053897671e-05,
      "loss": 0.5898,
      "step": 1449
    },
    {
      "epoch": 0.2898,
      "loss_align": 0.31922799348831177,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011557026766240597,
      "neg_sim_mean": 0.46052825450897217,
      "pos_sim_mean": 0.6807720065116882,
      "step": 1449
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.8712276220321655,
      "learning_rate": 8.942201871656291e-05,
      "loss": 0.3308,
      "step": 1450
    },
    {
      "epoch": 0.29,
      "loss_align": 0.29656851291656494,
      "loss_contrastive": 0.06748497486114502,
      "loss_gen": 0.31061139702796936,
      "neg_sim_mean": 0.6709164381027222,
      "pos_sim_mean": 0.7034314870834351,
      "step": 1450
    },
    {
      "epoch": 0.2902,
      "grad_norm": 2.1035022735595703,
      "learning_rate": 8.940053768033609e-05,
      "loss": 0.6153,
      "step": 1451
    },
    {
      "epoch": 0.2902,
      "loss_align": 0.3106120824813843,
      "loss_contrastive": 0.11262929439544678,
      "loss_gen": 0.28892022371292114,
      "neg_sim_mean": 0.7020171880722046,
      "pos_sim_mean": 0.6893879175186157,
      "step": 1451
    },
    {
      "epoch": 0.2904,
      "grad_norm": 1.9860485792160034,
      "learning_rate": 8.937903744076587e-05,
      "loss": 0.613,
      "step": 1452
    },
    {
      "epoch": 0.2904,
      "loss_align": 0.19714659452438354,
      "loss_contrastive": 0.122067391872406,
      "loss_gen": 0.21955740451812744,
      "neg_sim_mean": 0.8249207735061646,
      "pos_sim_mean": 0.8028534054756165,
      "step": 1452
    },
    {
      "epoch": 0.2906,
      "grad_norm": 1.9676228761672974,
      "learning_rate": 8.935751800833117e-05,
      "loss": 0.4314,
      "step": 1453
    },
    {
      "epoch": 0.2906,
      "loss_align": 0.29690802097320557,
      "loss_contrastive": 0.08422195911407471,
      "loss_gen": 0.3600965738296509,
      "neg_sim_mean": 0.6873139142990112,
      "pos_sim_mean": 0.7030919790267944,
      "step": 1453
    },
    {
      "epoch": 0.2908,
      "grad_norm": 2.1677286624908447,
      "learning_rate": 8.93359793935203e-05,
      "loss": 0.6671,
      "step": 1454
    },
    {
      "epoch": 0.2908,
      "loss_align": 0.32639849185943604,
      "loss_contrastive": 0.11321306228637695,
      "loss_gen": 0.2742483913898468,
      "neg_sim_mean": 0.686814546585083,
      "pos_sim_mean": 0.673601508140564,
      "step": 1454
    },
    {
      "epoch": 0.291,
      "grad_norm": 1.826849102973938,
      "learning_rate": 8.931442160683094e-05,
      "loss": 0.6142,
      "step": 1455
    },
    {
      "epoch": 0.291,
      "loss_align": 0.21572202444076538,
      "loss_contrastive": 0.021386325359344482,
      "loss_gen": 0.42863166332244873,
      "neg_sim_mean": 0.7056642770767212,
      "pos_sim_mean": 0.7842779755592346,
      "step": 1455
    },
    {
      "epoch": 0.2912,
      "grad_norm": 1.9926718473434448,
      "learning_rate": 8.92928446587701e-05,
      "loss": 0.6469,
      "step": 1456
    },
    {
      "epoch": 0.2912,
      "loss_align": 0.2804328203201294,
      "loss_contrastive": 0.09396445751190186,
      "loss_gen": 0.22632232308387756,
      "neg_sim_mean": 0.7135316133499146,
      "pos_sim_mean": 0.7195671796798706,
      "step": 1456
    },
    {
      "epoch": 0.2914,
      "grad_norm": 2.2447075843811035,
      "learning_rate": 8.927124855985409e-05,
      "loss": 0.518,
      "step": 1457
    },
    {
      "epoch": 0.2914,
      "loss_align": 0.16492724418640137,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013348313979804516,
      "neg_sim_mean": 0.4307405352592468,
      "pos_sim_mean": 0.8350727558135986,
      "step": 1457
    },
    {
      "epoch": 0.2916,
      "grad_norm": 0.7867599725723267,
      "learning_rate": 8.924963332060863e-05,
      "loss": 0.1783,
      "step": 1458
    },
    {
      "epoch": 0.2916,
      "loss_align": 0.2261180281639099,
      "loss_contrastive": 0.00503164529800415,
      "loss_gen": 0.2322242110967636,
      "neg_sim_mean": 0.6789135932922363,
      "pos_sim_mean": 0.7738819718360901,
      "step": 1458
    },
    {
      "epoch": 0.2918,
      "grad_norm": 2.2326207160949707,
      "learning_rate": 8.922799895156867e-05,
      "loss": 0.4589,
      "step": 1459
    },
    {
      "epoch": 0.2918,
      "loss_align": 0.21687966585159302,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005854619201272726,
      "neg_sim_mean": 0.3465959429740906,
      "pos_sim_mean": 0.783120334148407,
      "step": 1459
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.5761947631835938,
      "learning_rate": 8.920634546327858e-05,
      "loss": 0.2227,
      "step": 1460
    },
    {
      "epoch": 0.292,
      "loss_align": 0.3310343027114868,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01042480580508709,
      "neg_sim_mean": 0.46878764033317566,
      "pos_sim_mean": 0.6689656972885132,
      "step": 1460
    },
    {
      "epoch": 0.2922,
      "grad_norm": 0.7198240756988525,
      "learning_rate": 8.9184672866292e-05,
      "loss": 0.3415,
      "step": 1461
    },
    {
      "epoch": 0.2922,
      "loss_align": 0.3227522373199463,
      "loss_contrastive": 0.15443897247314453,
      "loss_gen": 0.3886658549308777,
      "neg_sim_mean": 0.7316867113113403,
      "pos_sim_mean": 0.6772477626800537,
      "step": 1461
    },
    {
      "epoch": 0.2924,
      "grad_norm": 2.666471242904663,
      "learning_rate": 8.916298117117188e-05,
      "loss": 0.73,
      "step": 1462
    },
    {
      "epoch": 0.2924,
      "loss_align": 0.2840142250061035,
      "loss_contrastive": 0.0,
      "loss_gen": 0.023505449295043945,
      "neg_sim_mean": 0.5338003635406494,
      "pos_sim_mean": 0.7159857749938965,
      "step": 1462
    },
    {
      "epoch": 0.2926,
      "grad_norm": 1.1485737562179565,
      "learning_rate": 8.91412703884905e-05,
      "loss": 0.3075,
      "step": 1463
    },
    {
      "epoch": 0.2926,
      "loss_align": 0.3280975818634033,
      "loss_contrastive": 0.07539618015289307,
      "loss_gen": 0.3722975254058838,
      "neg_sim_mean": 0.6472985744476318,
      "pos_sim_mean": 0.6719024181365967,
      "step": 1463
    },
    {
      "epoch": 0.2928,
      "grad_norm": 2.2500672340393066,
      "learning_rate": 8.91195405288294e-05,
      "loss": 0.7094,
      "step": 1464
    },
    {
      "epoch": 0.2928,
      "loss_align": 0.12291949987411499,
      "loss_contrastive": 0.049380600452423096,
      "loss_gen": 0.25217103958129883,
      "neg_sim_mean": 0.8264610767364502,
      "pos_sim_mean": 0.877080500125885,
      "step": 1464
    },
    {
      "epoch": 0.293,
      "grad_norm": 1.6347031593322754,
      "learning_rate": 8.90977916027795e-05,
      "loss": 0.381,
      "step": 1465
    },
    {
      "epoch": 0.293,
      "loss_align": 0.22984296083450317,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009904852136969566,
      "neg_sim_mean": 0.46516233682632446,
      "pos_sim_mean": 0.7701570391654968,
      "step": 1465
    },
    {
      "epoch": 0.2932,
      "grad_norm": 0.7105045318603516,
      "learning_rate": 8.907602362094094e-05,
      "loss": 0.2397,
      "step": 1466
    },
    {
      "epoch": 0.2932,
      "loss_align": 0.12274175882339478,
      "loss_contrastive": 0.11114251613616943,
      "loss_gen": 0.25924623012542725,
      "neg_sim_mean": 0.8884007334709167,
      "pos_sim_mean": 0.8772582411766052,
      "step": 1466
    },
    {
      "epoch": 0.2934,
      "grad_norm": 1.9878697395324707,
      "learning_rate": 8.905423659392316e-05,
      "loss": 0.3953,
      "step": 1467
    },
    {
      "epoch": 0.2934,
      "loss_align": 0.2197188138961792,
      "loss_contrastive": 0.0033119916915893555,
      "loss_gen": 0.01285362709313631,
      "neg_sim_mean": 0.6835931539535522,
      "pos_sim_mean": 0.7802811861038208,
      "step": 1467
    },
    {
      "epoch": 0.2936,
      "grad_norm": 0.8544709086418152,
      "learning_rate": 8.903243053234491e-05,
      "loss": 0.233,
      "step": 1468
    },
    {
      "epoch": 0.2936,
      "loss_align": 0.5106626152992249,
      "loss_contrastive": 0.21479812264442444,
      "loss_gen": 0.29842615127563477,
      "neg_sim_mean": 0.6041355133056641,
      "pos_sim_mean": 0.48933738470077515,
      "step": 1468
    },
    {
      "epoch": 0.2938,
      "grad_norm": 1.9903188943862915,
      "learning_rate": 8.901060544683418e-05,
      "loss": 0.8349,
      "step": 1469
    },
    {
      "epoch": 0.2938,
      "loss_align": 0.2745400667190552,
      "loss_contrastive": 0.13312041759490967,
      "loss_gen": 0.39048004150390625,
      "neg_sim_mean": 0.7585803270339966,
      "pos_sim_mean": 0.7254599332809448,
      "step": 1469
    },
    {
      "epoch": 0.294,
      "grad_norm": 2.0494558811187744,
      "learning_rate": 8.898876134802826e-05,
      "loss": 0.681,
      "step": 1470
    },
    {
      "epoch": 0.294,
      "loss_align": 0.45338624715805054,
      "loss_contrastive": 0.20349302887916565,
      "loss_gen": 0.21110834181308746,
      "neg_sim_mean": 0.6501067876815796,
      "pos_sim_mean": 0.5466137528419495,
      "step": 1470
    },
    {
      "epoch": 0.2942,
      "grad_norm": 1.9778872728347778,
      "learning_rate": 8.896689824657372e-05,
      "loss": 0.6889,
      "step": 1471
    },
    {
      "epoch": 0.2942,
      "loss_align": 0.20965713262557983,
      "loss_contrastive": 0.14755481481552124,
      "loss_gen": 0.3894383907318115,
      "neg_sim_mean": 0.8378976583480835,
      "pos_sim_mean": 0.7903428673744202,
      "step": 1471
    },
    {
      "epoch": 0.2944,
      "grad_norm": 2.2378690242767334,
      "learning_rate": 8.894501615312633e-05,
      "loss": 0.6168,
      "step": 1472
    },
    {
      "epoch": 0.2944,
      "loss_align": 0.3241976499557495,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0013761556474491954,
      "neg_sim_mean": 0.4546814262866974,
      "pos_sim_mean": 0.6758023500442505,
      "step": 1472
    },
    {
      "epoch": 0.2946,
      "grad_norm": 0.4476165175437927,
      "learning_rate": 8.892311507835119e-05,
      "loss": 0.3256,
      "step": 1473
    },
    {
      "epoch": 0.2946,
      "loss_align": 0.37578046321868896,
      "loss_contrastive": 0.08160364627838135,
      "loss_gen": 0.19838257133960724,
      "neg_sim_mean": 0.6058231592178345,
      "pos_sim_mean": 0.624219536781311,
      "step": 1473
    },
    {
      "epoch": 0.2948,
      "grad_norm": 2.091373920440674,
      "learning_rate": 8.890119503292257e-05,
      "loss": 0.584,
      "step": 1474
    },
    {
      "epoch": 0.2948,
      "loss_align": 0.251267671585083,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26245367527008057,
      "neg_sim_mean": 0.6027848720550537,
      "pos_sim_mean": 0.748732328414917,
      "step": 1474
    },
    {
      "epoch": 0.295,
      "grad_norm": 2.379565477371216,
      "learning_rate": 8.88792560275241e-05,
      "loss": 0.5137,
      "step": 1475
    },
    {
      "epoch": 0.295,
      "loss_align": 0.4401298761367798,
      "loss_contrastive": 0.08872899413108826,
      "loss_gen": 0.28955578804016113,
      "neg_sim_mean": 0.548599123954773,
      "pos_sim_mean": 0.5598701238632202,
      "step": 1475
    },
    {
      "epoch": 0.2952,
      "grad_norm": 2.220806360244751,
      "learning_rate": 8.885729807284856e-05,
      "loss": 0.7403,
      "step": 1476
    },
    {
      "epoch": 0.2952,
      "loss_align": 0.1986856460571289,
      "loss_contrastive": 0.07344865798950195,
      "loss_gen": 0.28680768609046936,
      "neg_sim_mean": 0.7747629880905151,
      "pos_sim_mean": 0.8013143539428711,
      "step": 1476
    },
    {
      "epoch": 0.2954,
      "grad_norm": 2.0523202419281006,
      "learning_rate": 8.883532117959796e-05,
      "loss": 0.4943,
      "step": 1477
    },
    {
      "epoch": 0.2954,
      "loss_align": 0.2558424472808838,
      "loss_contrastive": 0.22203266620635986,
      "loss_gen": 0.2488250732421875,
      "neg_sim_mean": 0.8661901950836182,
      "pos_sim_mean": 0.7441575527191162,
      "step": 1477
    },
    {
      "epoch": 0.2956,
      "grad_norm": 2.1628074645996094,
      "learning_rate": 8.88133253584836e-05,
      "loss": 0.5313,
      "step": 1478
    },
    {
      "epoch": 0.2956,
      "loss_align": 0.3186863660812378,
      "loss_contrastive": 0.1421370506286621,
      "loss_gen": 0.26458096504211426,
      "neg_sim_mean": 0.7234506607055664,
      "pos_sim_mean": 0.6813136339187622,
      "step": 1478
    },
    {
      "epoch": 0.2958,
      "grad_norm": 2.313030481338501,
      "learning_rate": 8.879131062022598e-05,
      "loss": 0.6003,
      "step": 1479
    },
    {
      "epoch": 0.2958,
      "loss_align": 0.3433513045310974,
      "loss_contrastive": 0.18650978803634644,
      "loss_gen": 0.1802826076745987,
      "neg_sim_mean": 0.7431584596633911,
      "pos_sim_mean": 0.6566486954689026,
      "step": 1479
    },
    {
      "epoch": 0.296,
      "grad_norm": 2.0523765087127686,
      "learning_rate": 8.87692769755548e-05,
      "loss": 0.546,
      "step": 1480
    },
    {
      "epoch": 0.296,
      "loss_align": 0.42673540115356445,
      "loss_contrastive": 0.26771995425224304,
      "loss_gen": 0.2597218155860901,
      "neg_sim_mean": 0.7409845590591431,
      "pos_sim_mean": 0.5732645988464355,
      "step": 1480
    },
    {
      "epoch": 0.2962,
      "grad_norm": 2.551443338394165,
      "learning_rate": 8.874722443520899e-05,
      "loss": 0.7186,
      "step": 1481
    },
    {
      "epoch": 0.2962,
      "loss_align": 0.24025118350982666,
      "loss_contrastive": 0.07692420482635498,
      "loss_gen": 0.22797730565071106,
      "neg_sim_mean": 0.7366729974746704,
      "pos_sim_mean": 0.7597488164901733,
      "step": 1481
    },
    {
      "epoch": 0.2964,
      "grad_norm": 1.7645748853683472,
      "learning_rate": 8.872515300993669e-05,
      "loss": 0.4775,
      "step": 1482
    },
    {
      "epoch": 0.2964,
      "loss_align": 0.22858506441116333,
      "loss_contrastive": 0.11199086904525757,
      "loss_gen": 0.18720504641532898,
      "neg_sim_mean": 0.7834057807922363,
      "pos_sim_mean": 0.7714149355888367,
      "step": 1482
    },
    {
      "epoch": 0.2966,
      "grad_norm": 1.9162673950195312,
      "learning_rate": 8.870306271049527e-05,
      "loss": 0.4292,
      "step": 1483
    },
    {
      "epoch": 0.2966,
      "loss_align": 0.25117921829223633,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24685414135456085,
      "neg_sim_mean": 0.4612608551979065,
      "pos_sim_mean": 0.7488207817077637,
      "step": 1483
    },
    {
      "epoch": 0.2968,
      "grad_norm": 2.5428128242492676,
      "learning_rate": 8.868095354765125e-05,
      "loss": 0.498,
      "step": 1484
    },
    {
      "epoch": 0.2968,
      "loss_align": 0.2468661069869995,
      "loss_contrastive": 0.023668289184570312,
      "loss_gen": 0.4175589978694916,
      "neg_sim_mean": 0.6768021583557129,
      "pos_sim_mean": 0.7531338930130005,
      "step": 1484
    },
    {
      "epoch": 0.297,
      "grad_norm": 2.5392463207244873,
      "learning_rate": 8.865882553218037e-05,
      "loss": 0.6673,
      "step": 1485
    },
    {
      "epoch": 0.297,
      "loss_align": 0.25222349166870117,
      "loss_contrastive": 0.14982300996780396,
      "loss_gen": 0.34849104285240173,
      "neg_sim_mean": 0.7975994944572449,
      "pos_sim_mean": 0.7477765083312988,
      "step": 1485
    },
    {
      "epoch": 0.2972,
      "grad_norm": 2.2081568241119385,
      "learning_rate": 8.863667867486756e-05,
      "loss": 0.6187,
      "step": 1486
    },
    {
      "epoch": 0.2972,
      "loss_align": 0.1816920042037964,
      "loss_contrastive": 0.11663228273391724,
      "loss_gen": 0.1386813223361969,
      "neg_sim_mean": 0.8349402546882629,
      "pos_sim_mean": 0.8183079957962036,
      "step": 1486
    },
    {
      "epoch": 0.2974,
      "grad_norm": 1.631062388420105,
      "learning_rate": 8.861451298650691e-05,
      "loss": 0.3344,
      "step": 1487
    },
    {
      "epoch": 0.2974,
      "loss_align": 0.16603761911392212,
      "loss_contrastive": 0.0478779673576355,
      "loss_gen": 0.31122535467147827,
      "neg_sim_mean": 0.7818403244018555,
      "pos_sim_mean": 0.8339623808860779,
      "step": 1487
    },
    {
      "epoch": 0.2976,
      "grad_norm": 1.7382988929748535,
      "learning_rate": 8.859232847790175e-05,
      "loss": 0.483,
      "step": 1488
    },
    {
      "epoch": 0.2976,
      "loss_align": 0.13519418239593506,
      "loss_contrastive": 0.07374316453933716,
      "loss_gen": 0.22035521268844604,
      "neg_sim_mean": 0.8385489583015442,
      "pos_sim_mean": 0.8648058176040649,
      "step": 1488
    },
    {
      "epoch": 0.2978,
      "grad_norm": 1.8145588636398315,
      "learning_rate": 8.857012515986452e-05,
      "loss": 0.3644,
      "step": 1489
    },
    {
      "epoch": 0.2978,
      "loss_align": 0.24420708417892456,
      "loss_contrastive": 0.14091485738754272,
      "loss_gen": 0.3175278604030609,
      "neg_sim_mean": 0.7967077493667603,
      "pos_sim_mean": 0.7557929158210754,
      "step": 1489
    },
    {
      "epoch": 0.298,
      "grad_norm": 3.6949942111968994,
      "learning_rate": 8.854790304321681e-05,
      "loss": 0.5786,
      "step": 1490
    },
    {
      "epoch": 0.298,
      "loss_align": 0.2046680450439453,
      "loss_contrastive": 0.07876455783843994,
      "loss_gen": 0.15353012084960938,
      "neg_sim_mean": 0.7740964889526367,
      "pos_sim_mean": 0.7953319549560547,
      "step": 1490
    },
    {
      "epoch": 0.2982,
      "grad_norm": 1.4848413467407227,
      "learning_rate": 8.852566213878947e-05,
      "loss": 0.3676,
      "step": 1491
    },
    {
      "epoch": 0.2982,
      "loss_align": 0.270203173160553,
      "loss_contrastive": 0.0997469425201416,
      "loss_gen": 0.2018955647945404,
      "neg_sim_mean": 0.7295437455177307,
      "pos_sim_mean": 0.729796826839447,
      "step": 1491
    },
    {
      "epoch": 0.2984,
      "grad_norm": 1.8822436332702637,
      "learning_rate": 8.85034024574224e-05,
      "loss": 0.4841,
      "step": 1492
    },
    {
      "epoch": 0.2984,
      "loss_align": 0.11359310150146484,
      "loss_contrastive": 0.0,
      "loss_gen": 0.023633474484086037,
      "neg_sim_mean": 0.32063788175582886,
      "pos_sim_mean": 0.8864068984985352,
      "step": 1492
    },
    {
      "epoch": 0.2986,
      "grad_norm": 1.0519107580184937,
      "learning_rate": 8.848112400996473e-05,
      "loss": 0.1372,
      "step": 1493
    },
    {
      "epoch": 0.2986,
      "loss_align": 0.23084211349487305,
      "loss_contrastive": 0.1428452730178833,
      "loss_gen": 0.20913763344287872,
      "neg_sim_mean": 0.8120031356811523,
      "pos_sim_mean": 0.769157886505127,
      "step": 1493
    },
    {
      "epoch": 0.2988,
      "grad_norm": 1.7282675504684448,
      "learning_rate": 8.845882680727469e-05,
      "loss": 0.4571,
      "step": 1494
    },
    {
      "epoch": 0.2988,
      "loss_align": 0.33076733350753784,
      "loss_contrastive": 0.14708930253982544,
      "loss_gen": 0.4777136743068695,
      "neg_sim_mean": 0.7163219451904297,
      "pos_sim_mean": 0.6692326664924622,
      "step": 1494
    },
    {
      "epoch": 0.299,
      "grad_norm": 2.656099796295166,
      "learning_rate": 8.843651086021966e-05,
      "loss": 0.8261,
      "step": 1495
    },
    {
      "epoch": 0.299,
      "loss_align": 0.14484214782714844,
      "loss_contrastive": 0.10303604602813721,
      "loss_gen": 0.22669339179992676,
      "neg_sim_mean": 0.8581938743591309,
      "pos_sim_mean": 0.8551578521728516,
      "step": 1495
    },
    {
      "epoch": 0.2992,
      "grad_norm": 1.9143263101577759,
      "learning_rate": 8.841417617967618e-05,
      "loss": 0.3839,
      "step": 1496
    },
    {
      "epoch": 0.2992,
      "loss_align": 0.23084896802902222,
      "loss_contrastive": 0.11706805229187012,
      "loss_gen": 0.3344027101993561,
      "neg_sim_mean": 0.78621906042099,
      "pos_sim_mean": 0.7691510319709778,
      "step": 1496
    },
    {
      "epoch": 0.2994,
      "grad_norm": 2.6890463829040527,
      "learning_rate": 8.839182277652989e-05,
      "loss": 0.5793,
      "step": 1497
    },
    {
      "epoch": 0.2994,
      "loss_align": 0.23451757431030273,
      "loss_contrastive": 0.028398990631103516,
      "loss_gen": 0.1886952817440033,
      "neg_sim_mean": 0.6938813924789429,
      "pos_sim_mean": 0.7654824256896973,
      "step": 1497
    },
    {
      "epoch": 0.2996,
      "grad_norm": 1.6903541088104248,
      "learning_rate": 8.836945066167555e-05,
      "loss": 0.4266,
      "step": 1498
    },
    {
      "epoch": 0.2996,
      "loss_align": 0.23152828216552734,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011968708597123623,
      "neg_sim_mean": 0.3493209183216095,
      "pos_sim_mean": 0.7684717178344727,
      "step": 1498
    },
    {
      "epoch": 0.2998,
      "grad_norm": 0.7715897560119629,
      "learning_rate": 8.834705984601708e-05,
      "loss": 0.2435,
      "step": 1499
    },
    {
      "epoch": 0.2998,
      "loss_align": 0.24249398708343506,
      "loss_contrastive": 0.10891222953796387,
      "loss_gen": 0.34990963339805603,
      "neg_sim_mean": 0.7664182186126709,
      "pos_sim_mean": 0.7575060129165649,
      "step": 1499
    },
    {
      "epoch": 0.3,
      "grad_norm": 2.1670007705688477,
      "learning_rate": 8.832465034046749e-05,
      "loss": 0.6055,
      "step": 1500
    },
    {
      "epoch": 0.3,
      "loss_align": 0.15134769678115845,
      "loss_contrastive": 0.07039666175842285,
      "loss_gen": 0.35602182149887085,
      "neg_sim_mean": 0.8190489411354065,
      "pos_sim_mean": 0.8486523032188416,
      "step": 1500
    },
    {
      "epoch": 0.3002,
      "grad_norm": 2.1437876224517822,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.5158,
      "step": 1501
    },
    {
      "epoch": 0.3002,
      "loss_align": 0.25259220600128174,
      "loss_contrastive": 0.05707055330276489,
      "loss_gen": 0.07858426868915558,
      "neg_sim_mean": 0.7044783234596252,
      "pos_sim_mean": 0.7474077939987183,
      "step": 1501
    },
    {
      "epoch": 0.3004,
      "grad_norm": 1.1888149976730347,
      "learning_rate": 8.827977530339253e-05,
      "loss": 0.338,
      "step": 1502
    },
    {
      "epoch": 0.3004,
      "loss_align": 0.13917911052703857,
      "loss_contrastive": 0.022490620613098145,
      "loss_gen": 0.2237219661474228,
      "neg_sim_mean": 0.7833114862442017,
      "pos_sim_mean": 0.8608208894729614,
      "step": 1502
    },
    {
      "epoch": 0.3006,
      "grad_norm": 2.0592081546783447,
      "learning_rate": 8.825730979373872e-05,
      "loss": 0.3656,
      "step": 1503
    },
    {
      "epoch": 0.3006,
      "loss_align": 0.19211024045944214,
      "loss_contrastive": 0.1277438998222351,
      "loss_gen": 0.31213173270225525,
      "neg_sim_mean": 0.8356336355209351,
      "pos_sim_mean": 0.8078897595405579,
      "step": 1503
    },
    {
      "epoch": 0.3008,
      "grad_norm": 2.2221062183380127,
      "learning_rate": 8.823482563793685e-05,
      "loss": 0.5196,
      "step": 1504
    },
    {
      "epoch": 0.3008,
      "loss_align": 0.32017678022384644,
      "loss_contrastive": 0.0712742805480957,
      "loss_gen": 0.19248728454113007,
      "neg_sim_mean": 0.6510974764823914,
      "pos_sim_mean": 0.6798232197761536,
      "step": 1504
    },
    {
      "epoch": 0.301,
      "grad_norm": 2.260063648223877,
      "learning_rate": 8.821232284694545e-05,
      "loss": 0.5212,
      "step": 1505
    },
    {
      "epoch": 0.301,
      "loss_align": 0.30882948637008667,
      "loss_contrastive": 0.11721181869506836,
      "loss_gen": 0.37907901406288147,
      "neg_sim_mean": 0.7083823084831238,
      "pos_sim_mean": 0.6911705136299133,
      "step": 1505
    },
    {
      "epoch": 0.3012,
      "grad_norm": 2.2334587574005127,
      "learning_rate": 8.818980143173213e-05,
      "loss": 0.702,
      "step": 1506
    },
    {
      "epoch": 0.3012,
      "loss_align": 0.12563419342041016,
      "loss_contrastive": 0.13215744495391846,
      "loss_gen": 0.19683687388896942,
      "neg_sim_mean": 0.9065232276916504,
      "pos_sim_mean": 0.8743658065795898,
      "step": 1506
    },
    {
      "epoch": 0.3014,
      "grad_norm": 1.6267951726913452,
      "learning_rate": 8.81672614032735e-05,
      "loss": 0.3383,
      "step": 1507
    },
    {
      "epoch": 0.3014,
      "loss_align": 0.24356549978256226,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02698330208659172,
      "neg_sim_mean": 0.32699090242385864,
      "pos_sim_mean": 0.7564345002174377,
      "step": 1507
    },
    {
      "epoch": 0.3016,
      "grad_norm": 1.1694793701171875,
      "learning_rate": 8.814470277255531e-05,
      "loss": 0.2705,
      "step": 1508
    },
    {
      "epoch": 0.3016,
      "loss_align": 0.2618027925491333,
      "loss_contrastive": 0.059068918228149414,
      "loss_gen": 0.3237552046775818,
      "neg_sim_mean": 0.6972661018371582,
      "pos_sim_mean": 0.7381972074508667,
      "step": 1508
    },
    {
      "epoch": 0.3018,
      "grad_norm": 2.513019323348999,
      "learning_rate": 8.81221255505724e-05,
      "loss": 0.5926,
      "step": 1509
    },
    {
      "epoch": 0.3018,
      "loss_align": 0.2720465660095215,
      "loss_contrastive": 0.0713890790939331,
      "loss_gen": 0.1693231612443924,
      "neg_sim_mean": 0.6993424892425537,
      "pos_sim_mean": 0.7279534339904785,
      "step": 1509
    },
    {
      "epoch": 0.302,
      "grad_norm": 2.1938750743865967,
      "learning_rate": 8.80995297483286e-05,
      "loss": 0.4499,
      "step": 1510
    },
    {
      "epoch": 0.302,
      "loss_align": 0.2589590549468994,
      "loss_contrastive": 0.045897841453552246,
      "loss_gen": 0.2814682722091675,
      "neg_sim_mean": 0.6869387626647949,
      "pos_sim_mean": 0.7410409450531006,
      "step": 1510
    },
    {
      "epoch": 0.3022,
      "grad_norm": 1.9817708730697632,
      "learning_rate": 8.807691537683685e-05,
      "loss": 0.5459,
      "step": 1511
    },
    {
      "epoch": 0.3022,
      "loss_align": 0.19536668062210083,
      "loss_contrastive": 0.06366753578186035,
      "loss_gen": 0.2513737082481384,
      "neg_sim_mean": 0.7683008313179016,
      "pos_sim_mean": 0.8046333193778992,
      "step": 1511
    },
    {
      "epoch": 0.3024,
      "grad_norm": 2.396624803543091,
      "learning_rate": 8.80542824471191e-05,
      "loss": 0.4544,
      "step": 1512
    },
    {
      "epoch": 0.3024,
      "loss_align": 0.16401278972625732,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1717263013124466,
      "neg_sim_mean": 0.7027065753936768,
      "pos_sim_mean": 0.8359872102737427,
      "step": 1512
    },
    {
      "epoch": 0.3026,
      "grad_norm": 10.708481788635254,
      "learning_rate": 8.803163097020637e-05,
      "loss": 0.3357,
      "step": 1513
    },
    {
      "epoch": 0.3026,
      "loss_align": 0.40050768852233887,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010962392203509808,
      "neg_sim_mean": 0.36321595311164856,
      "pos_sim_mean": 0.5994923114776611,
      "step": 1513
    },
    {
      "epoch": 0.3028,
      "grad_norm": 1.4416900873184204,
      "learning_rate": 8.800896095713871e-05,
      "loss": 0.4115,
      "step": 1514
    },
    {
      "epoch": 0.3028,
      "loss_align": 0.3982371687889099,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009782282635569572,
      "neg_sim_mean": 0.36358967423439026,
      "pos_sim_mean": 0.6017628312110901,
      "step": 1514
    },
    {
      "epoch": 0.303,
      "grad_norm": 0.994678258895874,
      "learning_rate": 8.798627241896524e-05,
      "loss": 0.408,
      "step": 1515
    },
    {
      "epoch": 0.303,
      "loss_align": 0.23044466972351074,
      "loss_contrastive": 0.09326910972595215,
      "loss_gen": 0.23760604858398438,
      "neg_sim_mean": 0.7628244161605835,
      "pos_sim_mean": 0.7695553302764893,
      "step": 1515
    },
    {
      "epoch": 0.3032,
      "grad_norm": 1.9759114980697632,
      "learning_rate": 8.796356536674403e-05,
      "loss": 0.4792,
      "step": 1516
    },
    {
      "epoch": 0.3032,
      "loss_align": 0.2713346481323242,
      "loss_contrastive": 0.052096664905548096,
      "loss_gen": 0.06604522466659546,
      "neg_sim_mean": 0.680761992931366,
      "pos_sim_mean": 0.7286653518676758,
      "step": 1516
    },
    {
      "epoch": 0.3034,
      "grad_norm": 1.1079784631729126,
      "learning_rate": 8.794083981154229e-05,
      "loss": 0.3436,
      "step": 1517
    },
    {
      "epoch": 0.3034,
      "loss_align": 0.29703426361083984,
      "loss_contrastive": 0.08996772766113281,
      "loss_gen": 0.25752267241477966,
      "neg_sim_mean": 0.6929334402084351,
      "pos_sim_mean": 0.7029657363891602,
      "step": 1517
    },
    {
      "epoch": 0.3036,
      "grad_norm": 1.9603111743927002,
      "learning_rate": 8.79180957644361e-05,
      "loss": 0.5654,
      "step": 1518
    },
    {
      "epoch": 0.3036,
      "loss_align": 0.1565021276473999,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1832302063703537,
      "neg_sim_mean": 0.7329328060150146,
      "pos_sim_mean": 0.8434978723526001,
      "step": 1518
    },
    {
      "epoch": 0.3038,
      "grad_norm": 1.6222319602966309,
      "learning_rate": 8.789533323651066e-05,
      "loss": 0.3397,
      "step": 1519
    },
    {
      "epoch": 0.3038,
      "loss_align": 0.22053766250610352,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3062755763530731,
      "neg_sim_mean": 0.6579056978225708,
      "pos_sim_mean": 0.7794623374938965,
      "step": 1519
    },
    {
      "epoch": 0.304,
      "grad_norm": 1.7999722957611084,
      "learning_rate": 8.787255223886019e-05,
      "loss": 0.5268,
      "step": 1520
    },
    {
      "epoch": 0.304,
      "loss_align": 0.14849621057510376,
      "loss_contrastive": 0.10642200708389282,
      "loss_gen": 0.23602932691574097,
      "neg_sim_mean": 0.8579257726669312,
      "pos_sim_mean": 0.8515037894248962,
      "step": 1520
    },
    {
      "epoch": 0.3042,
      "grad_norm": 2.038374662399292,
      "learning_rate": 8.784975278258783e-05,
      "loss": 0.3973,
      "step": 1521
    },
    {
      "epoch": 0.3042,
      "loss_align": 0.2817467451095581,
      "loss_contrastive": 0.10119092464447021,
      "loss_gen": 0.3574848771095276,
      "neg_sim_mean": 0.7194441556930542,
      "pos_sim_mean": 0.7182532548904419,
      "step": 1521
    },
    {
      "epoch": 0.3044,
      "grad_norm": 2.735379219055176,
      "learning_rate": 8.782693487880575e-05,
      "loss": 0.6514,
      "step": 1522
    },
    {
      "epoch": 0.3044,
      "loss_align": 0.43016427755355835,
      "loss_contrastive": 0.13608774542808533,
      "loss_gen": 0.21711285412311554,
      "neg_sim_mean": 0.6059234738349915,
      "pos_sim_mean": 0.5698357224464417,
      "step": 1522
    },
    {
      "epoch": 0.3046,
      "grad_norm": 2.0322587490081787,
      "learning_rate": 8.780409853863517e-05,
      "loss": 0.6636,
      "step": 1523
    },
    {
      "epoch": 0.3046,
      "loss_align": 0.09467470645904541,
      "loss_contrastive": 0.11984115839004517,
      "loss_gen": 0.4720384180545807,
      "neg_sim_mean": 0.9251664280891418,
      "pos_sim_mean": 0.9053252935409546,
      "step": 1523
    },
    {
      "epoch": 0.3048,
      "grad_norm": 2.7267093658447266,
      "learning_rate": 8.778124377320618e-05,
      "loss": 0.5811,
      "step": 1524
    },
    {
      "epoch": 0.3048,
      "loss_align": 0.17104864120483398,
      "loss_contrastive": 0.08911740779876709,
      "loss_gen": 0.20625530183315277,
      "neg_sim_mean": 0.8180687427520752,
      "pos_sim_mean": 0.828951358795166,
      "step": 1524
    },
    {
      "epoch": 0.305,
      "grad_norm": 1.778530240058899,
      "learning_rate": 8.775837059365796e-05,
      "loss": 0.388,
      "step": 1525
    },
    {
      "epoch": 0.305,
      "loss_align": 0.28164219856262207,
      "loss_contrastive": 0.15282809734344482,
      "loss_gen": 0.2785595953464508,
      "neg_sim_mean": 0.7711858749389648,
      "pos_sim_mean": 0.7183578014373779,
      "step": 1525
    },
    {
      "epoch": 0.3052,
      "grad_norm": 2.0840184688568115,
      "learning_rate": 8.773547901113862e-05,
      "loss": 0.5785,
      "step": 1526
    },
    {
      "epoch": 0.3052,
      "loss_align": 0.3401613235473633,
      "loss_contrastive": 0.03512454032897949,
      "loss_gen": 0.01866324059665203,
      "neg_sim_mean": 0.5949631929397583,
      "pos_sim_mean": 0.6598386764526367,
      "step": 1526
    },
    {
      "epoch": 0.3054,
      "grad_norm": 1.2992994785308838,
      "learning_rate": 8.771256903680519e-05,
      "loss": 0.363,
      "step": 1527
    },
    {
      "epoch": 0.3054,
      "loss_align": 0.2849690914154053,
      "loss_contrastive": 0.15339726209640503,
      "loss_gen": 0.15921983122825623,
      "neg_sim_mean": 0.7684281468391418,
      "pos_sim_mean": 0.7150309085845947,
      "step": 1527
    },
    {
      "epoch": 0.3056,
      "grad_norm": 1.7344920635223389,
      "learning_rate": 8.768964068182378e-05,
      "loss": 0.4626,
      "step": 1528
    },
    {
      "epoch": 0.3056,
      "loss_align": 0.40516090393066406,
      "loss_contrastive": 0.18240824341773987,
      "loss_gen": 0.19021430611610413,
      "neg_sim_mean": 0.6772473454475403,
      "pos_sim_mean": 0.5948390960693359,
      "step": 1528
    },
    {
      "epoch": 0.3058,
      "grad_norm": 1.8974099159240723,
      "learning_rate": 8.766669395736935e-05,
      "loss": 0.6173,
      "step": 1529
    },
    {
      "epoch": 0.3058,
      "loss_align": 0.2605234384536743,
      "loss_contrastive": 0.09718525409698486,
      "loss_gen": 0.24509942531585693,
      "neg_sim_mean": 0.7366617918014526,
      "pos_sim_mean": 0.7394765615463257,
      "step": 1529
    },
    {
      "epoch": 0.306,
      "grad_norm": 2.3550424575805664,
      "learning_rate": 8.764372887462586e-05,
      "loss": 0.5173,
      "step": 1530
    },
    {
      "epoch": 0.306,
      "loss_align": 0.2154827117919922,
      "loss_contrastive": 0.08381712436676025,
      "loss_gen": 0.21817433834075928,
      "neg_sim_mean": 0.7683343887329102,
      "pos_sim_mean": 0.7845172882080078,
      "step": 1530
    },
    {
      "epoch": 0.3062,
      "grad_norm": 1.7915605306625366,
      "learning_rate": 8.762074544478623e-05,
      "loss": 0.4437,
      "step": 1531
    },
    {
      "epoch": 0.3062,
      "loss_align": 0.24197757244110107,
      "loss_contrastive": 0.15366804599761963,
      "loss_gen": 0.19465099275112152,
      "neg_sim_mean": 0.8116904497146606,
      "pos_sim_mean": 0.7580224275588989,
      "step": 1531
    },
    {
      "epoch": 0.3064,
      "grad_norm": 2.5179436206817627,
      "learning_rate": 8.759774367905228e-05,
      "loss": 0.4551,
      "step": 1532
    },
    {
      "epoch": 0.3064,
      "loss_align": 0.37399327754974365,
      "loss_contrastive": 0.07708591222763062,
      "loss_gen": 0.13054853677749634,
      "neg_sim_mean": 0.603092610836029,
      "pos_sim_mean": 0.6260067224502563,
      "step": 1532
    },
    {
      "epoch": 0.3066,
      "grad_norm": 1.4931204319000244,
      "learning_rate": 8.757472358863481e-05,
      "loss": 0.5138,
      "step": 1533
    },
    {
      "epoch": 0.3066,
      "loss_align": 0.11387777328491211,
      "loss_contrastive": 0.06977880001068115,
      "loss_gen": 0.28161367774009705,
      "neg_sim_mean": 0.8559010028839111,
      "pos_sim_mean": 0.8861222267150879,
      "step": 1533
    },
    {
      "epoch": 0.3068,
      "grad_norm": 2.2673287391662598,
      "learning_rate": 8.755168518475351e-05,
      "loss": 0.4039,
      "step": 1534
    },
    {
      "epoch": 0.3068,
      "loss_align": 0.24921506643295288,
      "loss_contrastive": 0.051178038120269775,
      "loss_gen": 0.19649752974510193,
      "neg_sim_mean": 0.701962947845459,
      "pos_sim_mean": 0.7507849335670471,
      "step": 1534
    },
    {
      "epoch": 0.307,
      "grad_norm": 1.7925400733947754,
      "learning_rate": 8.752862847863706e-05,
      "loss": 0.4519,
      "step": 1535
    },
    {
      "epoch": 0.307,
      "loss_align": 0.15922558307647705,
      "loss_contrastive": 0.0610315203666687,
      "loss_gen": 0.2751845717430115,
      "neg_sim_mean": 0.8018059134483337,
      "pos_sim_mean": 0.840774416923523,
      "step": 1535
    },
    {
      "epoch": 0.3072,
      "grad_norm": 1.7531975507736206,
      "learning_rate": 8.750555348152298e-05,
      "loss": 0.4417,
      "step": 1536
    },
    {
      "epoch": 0.3072,
      "loss_align": 0.14032894372940063,
      "loss_contrastive": 0.06991750001907349,
      "loss_gen": 0.3176782727241516,
      "neg_sim_mean": 0.8295885324478149,
      "pos_sim_mean": 0.8596710562705994,
      "step": 1536
    },
    {
      "epoch": 0.3074,
      "grad_norm": 2.0382728576660156,
      "learning_rate": 8.748246020465776e-05,
      "loss": 0.4664,
      "step": 1537
    },
    {
      "epoch": 0.3074,
      "loss_align": 0.31172001361846924,
      "loss_contrastive": 0.15820330381393433,
      "loss_gen": 0.2896787226200104,
      "neg_sim_mean": 0.7464832663536072,
      "pos_sim_mean": 0.6882799863815308,
      "step": 1537
    },
    {
      "epoch": 0.3076,
      "grad_norm": 2.8107004165649414,
      "learning_rate": 8.745934865929676e-05,
      "loss": 0.6204,
      "step": 1538
    },
    {
      "epoch": 0.3076,
      "loss_align": 0.17991596460342407,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00728801591321826,
      "neg_sim_mean": 0.34984129667282104,
      "pos_sim_mean": 0.8200840353965759,
      "step": 1538
    },
    {
      "epoch": 0.3078,
      "grad_norm": 0.5605076551437378,
      "learning_rate": 8.74362188567043e-05,
      "loss": 0.1872,
      "step": 1539
    },
    {
      "epoch": 0.3078,
      "loss_align": 0.32820188999176025,
      "loss_contrastive": 0.20481228828430176,
      "loss_gen": 0.19082951545715332,
      "neg_sim_mean": 0.7766103744506836,
      "pos_sim_mean": 0.6717981100082397,
      "step": 1539
    },
    {
      "epoch": 0.308,
      "grad_norm": 2.0195839405059814,
      "learning_rate": 8.741307080815357e-05,
      "loss": 0.5436,
      "step": 1540
    },
    {
      "epoch": 0.308,
      "loss_align": 0.11156797409057617,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010248063132166862,
      "neg_sim_mean": 0.6894248723983765,
      "pos_sim_mean": 0.8884320259094238,
      "step": 1540
    },
    {
      "epoch": 0.3082,
      "grad_norm": 0.6757187247276306,
      "learning_rate": 8.73899045249266e-05,
      "loss": 0.1218,
      "step": 1541
    },
    {
      "epoch": 0.3082,
      "loss_align": 0.21818625926971436,
      "loss_contrastive": 0.06602632999420166,
      "loss_gen": 0.26447421312332153,
      "neg_sim_mean": 0.7478400468826294,
      "pos_sim_mean": 0.7818137407302856,
      "step": 1541
    },
    {
      "epoch": 0.3084,
      "grad_norm": 1.8295320272445679,
      "learning_rate": 8.736672001831438e-05,
      "loss": 0.4906,
      "step": 1542
    },
    {
      "epoch": 0.3084,
      "loss_align": 0.2064826488494873,
      "loss_contrastive": 0.08221447467803955,
      "loss_gen": 0.21723099052906036,
      "neg_sim_mean": 0.7757318019866943,
      "pos_sim_mean": 0.7935173511505127,
      "step": 1542
    },
    {
      "epoch": 0.3086,
      "grad_norm": 1.894877314567566,
      "learning_rate": 8.734351729961678e-05,
      "loss": 0.4336,
      "step": 1543
    },
    {
      "epoch": 0.3086,
      "loss_align": 0.10822999477386475,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011361949145793915,
      "neg_sim_mean": 0.40275198221206665,
      "pos_sim_mean": 0.8917700052261353,
      "step": 1543
    },
    {
      "epoch": 0.3088,
      "grad_norm": 0.7483651041984558,
      "learning_rate": 8.732029638014249e-05,
      "loss": 0.1196,
      "step": 1544
    },
    {
      "epoch": 0.3088,
      "loss_align": 0.19633299112319946,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2435959130525589,
      "neg_sim_mean": 0.7013993263244629,
      "pos_sim_mean": 0.8036670088768005,
      "step": 1544
    },
    {
      "epoch": 0.309,
      "grad_norm": 2.088340997695923,
      "learning_rate": 8.729705727120911e-05,
      "loss": 0.4399,
      "step": 1545
    },
    {
      "epoch": 0.309,
      "loss_align": 0.19246500730514526,
      "loss_contrastive": 0.07680052518844604,
      "loss_gen": 0.2016880214214325,
      "neg_sim_mean": 0.7843354940414429,
      "pos_sim_mean": 0.8075349926948547,
      "step": 1545
    },
    {
      "epoch": 0.3092,
      "grad_norm": 1.4629321098327637,
      "learning_rate": 8.727379998414311e-05,
      "loss": 0.4034,
      "step": 1546
    },
    {
      "epoch": 0.3092,
      "loss_align": 0.21524953842163086,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00830573309212923,
      "neg_sim_mean": 0.2958757281303406,
      "pos_sim_mean": 0.7847504615783691,
      "step": 1546
    },
    {
      "epoch": 0.3094,
      "grad_norm": 0.6814629435539246,
      "learning_rate": 8.725052453027981e-05,
      "loss": 0.2236,
      "step": 1547
    },
    {
      "epoch": 0.3094,
      "loss_align": 0.3584336042404175,
      "loss_contrastive": 0.1873512864112854,
      "loss_gen": 0.6575900912284851,
      "neg_sim_mean": 0.72891765832901,
      "pos_sim_mean": 0.6415663957595825,
      "step": 1547
    },
    {
      "epoch": 0.3096,
      "grad_norm": 3.186229944229126,
      "learning_rate": 8.722723092096338e-05,
      "loss": 1.0385,
      "step": 1548
    },
    {
      "epoch": 0.3096,
      "loss_align": 0.2639063596725464,
      "loss_contrastive": 0.1623629927635193,
      "loss_gen": 0.2678842544555664,
      "neg_sim_mean": 0.798456609249115,
      "pos_sim_mean": 0.7360936403274536,
      "step": 1548
    },
    {
      "epoch": 0.3098,
      "grad_norm": 2.1023285388946533,
      "learning_rate": 8.720391916754683e-05,
      "loss": 0.5513,
      "step": 1549
    },
    {
      "epoch": 0.3098,
      "loss_align": 0.24771958589553833,
      "loss_contrastive": 0.11090224981307983,
      "loss_gen": 0.1736689954996109,
      "neg_sim_mean": 0.7631826400756836,
      "pos_sim_mean": 0.7522804141044617,
      "step": 1549
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.79746413230896,
      "learning_rate": 8.718058928139205e-05,
      "loss": 0.4347,
      "step": 1550
    },
    {
      "epoch": 0.31,
      "loss_align": 0.35740745067596436,
      "loss_contrastive": 0.01895129680633545,
      "loss_gen": 0.24271194636821747,
      "neg_sim_mean": 0.5615438222885132,
      "pos_sim_mean": 0.6425925493240356,
      "step": 1550
    },
    {
      "epoch": 0.3102,
      "grad_norm": 1.9069777727127075,
      "learning_rate": 8.715724127386972e-05,
      "loss": 0.6024,
      "step": 1551
    },
    {
      "epoch": 0.3102,
      "loss_align": 0.3083360195159912,
      "loss_contrastive": 0.12384450435638428,
      "loss_gen": 0.21938936412334442,
      "neg_sim_mean": 0.7155084609985352,
      "pos_sim_mean": 0.6916639804840088,
      "step": 1551
    },
    {
      "epoch": 0.3104,
      "grad_norm": 1.7667264938354492,
      "learning_rate": 8.713387515635938e-05,
      "loss": 0.5426,
      "step": 1552
    },
    {
      "epoch": 0.3104,
      "loss_align": 0.20558947324752808,
      "loss_contrastive": 0.1997097134590149,
      "loss_gen": 0.2142987996339798,
      "neg_sim_mean": 0.8941202163696289,
      "pos_sim_mean": 0.7944105267524719,
      "step": 1552
    },
    {
      "epoch": 0.3106,
      "grad_norm": 1.7276784181594849,
      "learning_rate": 8.711049094024942e-05,
      "loss": 0.4439,
      "step": 1553
    },
    {
      "epoch": 0.3106,
      "loss_align": 0.34405088424682617,
      "loss_contrastive": 0.12536245584487915,
      "loss_gen": 0.2301127314567566,
      "neg_sim_mean": 0.6813115477561951,
      "pos_sim_mean": 0.6559491157531738,
      "step": 1553
    },
    {
      "epoch": 0.3108,
      "grad_norm": 1.8499741554260254,
      "learning_rate": 8.708708863693697e-05,
      "loss": 0.5892,
      "step": 1554
    },
    {
      "epoch": 0.3108,
      "loss_align": 0.2095937728881836,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03614102303981781,
      "neg_sim_mean": 0.31344401836395264,
      "pos_sim_mean": 0.7904062271118164,
      "step": 1554
    },
    {
      "epoch": 0.311,
      "grad_norm": 1.3696777820587158,
      "learning_rate": 8.706366825782805e-05,
      "loss": 0.2457,
      "step": 1555
    },
    {
      "epoch": 0.311,
      "loss_align": 0.2547401189804077,
      "loss_contrastive": 0.0529862642288208,
      "loss_gen": 0.26975929737091064,
      "neg_sim_mean": 0.6982461214065552,
      "pos_sim_mean": 0.7452598810195923,
      "step": 1555
    },
    {
      "epoch": 0.3112,
      "grad_norm": 1.935594916343689,
      "learning_rate": 8.70402298143375e-05,
      "loss": 0.5309,
      "step": 1556
    },
    {
      "epoch": 0.3112,
      "loss_align": 0.1896006464958191,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011485384311527014,
      "neg_sim_mean": 0.4009777307510376,
      "pos_sim_mean": 0.8103993535041809,
      "step": 1556
    },
    {
      "epoch": 0.3114,
      "grad_norm": 0.42798569798469543,
      "learning_rate": 8.701677331788891e-05,
      "loss": 0.1907,
      "step": 1557
    },
    {
      "epoch": 0.3114,
      "loss_align": 0.1837339997291565,
      "loss_contrastive": 0.11425596475601196,
      "loss_gen": 0.2659924030303955,
      "neg_sim_mean": 0.8305219411849976,
      "pos_sim_mean": 0.8162660002708435,
      "step": 1557
    },
    {
      "epoch": 0.3116,
      "grad_norm": 1.9076263904571533,
      "learning_rate": 8.699329877991468e-05,
      "loss": 0.4634,
      "step": 1558
    },
    {
      "epoch": 0.3116,
      "loss_align": 0.2830333113670349,
      "loss_contrastive": 0.0,
      "loss_gen": 0.000986670609563589,
      "neg_sim_mean": 0.5788758397102356,
      "pos_sim_mean": 0.7169666886329651,
      "step": 1558
    },
    {
      "epoch": 0.3118,
      "grad_norm": 0.5588852167129517,
      "learning_rate": 8.696980621185602e-05,
      "loss": 0.284,
      "step": 1559
    },
    {
      "epoch": 0.3118,
      "loss_align": 0.24624300003051758,
      "loss_contrastive": 0.11292314529418945,
      "loss_gen": 0.11548107117414474,
      "neg_sim_mean": 0.766680121421814,
      "pos_sim_mean": 0.7537569999694824,
      "step": 1559
    },
    {
      "epoch": 0.312,
      "grad_norm": 1.6392719745635986,
      "learning_rate": 8.694629562516294e-05,
      "loss": 0.3753,
      "step": 1560
    },
    {
      "epoch": 0.312,
      "loss_align": 0.20194941759109497,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009252568706870079,
      "neg_sim_mean": 0.420762836933136,
      "pos_sim_mean": 0.798050582408905,
      "step": 1560
    },
    {
      "epoch": 0.3122,
      "grad_norm": 1.2589263916015625,
      "learning_rate": 8.692276703129421e-05,
      "loss": 0.2112,
      "step": 1561
    },
    {
      "epoch": 0.3122,
      "loss_align": 0.34169232845306396,
      "loss_contrastive": 0.18058836460113525,
      "loss_gen": 0.24790574610233307,
      "neg_sim_mean": 0.7388960123062134,
      "pos_sim_mean": 0.658307671546936,
      "step": 1561
    },
    {
      "epoch": 0.3124,
      "grad_norm": 1.965075969696045,
      "learning_rate": 8.689922044171734e-05,
      "loss": 0.6113,
      "step": 1562
    },
    {
      "epoch": 0.3124,
      "loss_align": 0.21890974044799805,
      "loss_contrastive": 0.08064186573028564,
      "loss_gen": 0.3751385807991028,
      "neg_sim_mean": 0.7617321014404297,
      "pos_sim_mean": 0.781090259552002,
      "step": 1562
    },
    {
      "epoch": 0.3126,
      "grad_norm": 2.0798089504241943,
      "learning_rate": 8.68756558679087e-05,
      "loss": 0.6037,
      "step": 1563
    },
    {
      "epoch": 0.3126,
      "loss_align": 0.32034832239151,
      "loss_contrastive": 0.09453779458999634,
      "loss_gen": 0.2059248983860016,
      "neg_sim_mean": 0.6741894483566284,
      "pos_sim_mean": 0.67965167760849,
      "step": 1563
    },
    {
      "epoch": 0.3128,
      "grad_norm": 2.0469865798950195,
      "learning_rate": 8.685207332135337e-05,
      "loss": 0.5376,
      "step": 1564
    },
    {
      "epoch": 0.3128,
      "loss_align": 0.22028982639312744,
      "loss_contrastive": 0.024841725826263428,
      "loss_gen": 0.2855398654937744,
      "neg_sim_mean": 0.7045518755912781,
      "pos_sim_mean": 0.7797101736068726,
      "step": 1564
    },
    {
      "epoch": 0.313,
      "grad_norm": 2.2333178520202637,
      "learning_rate": 8.682847281354517e-05,
      "loss": 0.5088,
      "step": 1565
    },
    {
      "epoch": 0.313,
      "loss_align": 0.2288196086883545,
      "loss_contrastive": 0.07746076583862305,
      "loss_gen": 0.4140479862689972,
      "neg_sim_mean": 0.7486411333084106,
      "pos_sim_mean": 0.7711803913116455,
      "step": 1565
    },
    {
      "epoch": 0.3132,
      "grad_norm": 2.414372205734253,
      "learning_rate": 8.680485435598673e-05,
      "loss": 0.6522,
      "step": 1566
    },
    {
      "epoch": 0.3132,
      "loss_align": 0.21129769086837769,
      "loss_contrastive": 0.13771885633468628,
      "loss_gen": 0.5293526649475098,
      "neg_sim_mean": 0.8264211416244507,
      "pos_sim_mean": 0.7887023091316223,
      "step": 1566
    },
    {
      "epoch": 0.3134,
      "grad_norm": 2.3901896476745605,
      "learning_rate": 8.678121796018937e-05,
      "loss": 0.7572,
      "step": 1567
    },
    {
      "epoch": 0.3134,
      "loss_align": 0.19834601879119873,
      "loss_contrastive": 0.10620105266571045,
      "loss_gen": 0.163115993142128,
      "neg_sim_mean": 0.8078550100326538,
      "pos_sim_mean": 0.8016539812088013,
      "step": 1567
    },
    {
      "epoch": 0.3136,
      "grad_norm": 1.8948463201522827,
      "learning_rate": 8.675756363767322e-05,
      "loss": 0.3742,
      "step": 1568
    },
    {
      "epoch": 0.3136,
      "loss_align": 0.2547808885574341,
      "loss_contrastive": 0.10082042217254639,
      "loss_gen": 0.16147445142269135,
      "neg_sim_mean": 0.7460395097732544,
      "pos_sim_mean": 0.7452191114425659,
      "step": 1568
    },
    {
      "epoch": 0.3138,
      "grad_norm": 1.8810151815414429,
      "learning_rate": 8.673389139996708e-05,
      "loss": 0.4284,
      "step": 1569
    },
    {
      "epoch": 0.3138,
      "loss_align": 0.2716518044471741,
      "loss_contrastive": 0.161113440990448,
      "loss_gen": 0.34863683581352234,
      "neg_sim_mean": 0.789461612701416,
      "pos_sim_mean": 0.7283481955528259,
      "step": 1569
    },
    {
      "epoch": 0.314,
      "grad_norm": 2.2451295852661133,
      "learning_rate": 8.671020125860851e-05,
      "loss": 0.6396,
      "step": 1570
    },
    {
      "epoch": 0.314,
      "loss_align": 0.26322340965270996,
      "loss_contrastive": 0.008543848991394043,
      "loss_gen": 0.26331281661987305,
      "neg_sim_mean": 0.6453204154968262,
      "pos_sim_mean": 0.73677659034729,
      "step": 1570
    },
    {
      "epoch": 0.3142,
      "grad_norm": 2.175018548965454,
      "learning_rate": 8.668649322514382e-05,
      "loss": 0.5276,
      "step": 1571
    },
    {
      "epoch": 0.3142,
      "loss_align": 0.0987699031829834,
      "loss_contrastive": 0.01890397071838379,
      "loss_gen": 0.5171607136726379,
      "neg_sim_mean": 0.8201340436935425,
      "pos_sim_mean": 0.9012300968170166,
      "step": 1571
    },
    {
      "epoch": 0.3144,
      "grad_norm": 2.513479471206665,
      "learning_rate": 8.666276731112801e-05,
      "loss": 0.6182,
      "step": 1572
    },
    {
      "epoch": 0.3144,
      "loss_align": 0.1925448179244995,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011281137354671955,
      "neg_sim_mean": 0.38017261028289795,
      "pos_sim_mean": 0.8074551820755005,
      "step": 1572
    },
    {
      "epoch": 0.3146,
      "grad_norm": 0.8962131142616272,
      "learning_rate": 8.663902352812478e-05,
      "loss": 0.2038,
      "step": 1573
    },
    {
      "epoch": 0.3146,
      "loss_align": 0.3145377039909363,
      "loss_contrastive": 0.1295042634010315,
      "loss_gen": 0.35310372710227966,
      "neg_sim_mean": 0.7149665355682373,
      "pos_sim_mean": 0.6854622960090637,
      "step": 1573
    },
    {
      "epoch": 0.3148,
      "grad_norm": 1.9357103109359741,
      "learning_rate": 8.661526188770659e-05,
      "loss": 0.6832,
      "step": 1574
    },
    {
      "epoch": 0.3148,
      "loss_align": 0.3141772150993347,
      "loss_contrastive": 0.0997423529624939,
      "loss_gen": 0.19243045151233673,
      "neg_sim_mean": 0.6855651140213013,
      "pos_sim_mean": 0.6858227849006653,
      "step": 1574
    },
    {
      "epoch": 0.315,
      "grad_norm": 1.403950810432434,
      "learning_rate": 8.659148240145456e-05,
      "loss": 0.5186,
      "step": 1575
    },
    {
      "epoch": 0.315,
      "loss_align": 0.19730150699615479,
      "loss_contrastive": 0.0012171268463134766,
      "loss_gen": 0.3188645541667938,
      "neg_sim_mean": 0.7039155960083008,
      "pos_sim_mean": 0.8026984930038452,
      "step": 1575
    },
    {
      "epoch": 0.3152,
      "grad_norm": 2.4639298915863037,
      "learning_rate": 8.656768508095853e-05,
      "loss": 0.5163,
      "step": 1576
    },
    {
      "epoch": 0.3152,
      "loss_align": 0.047405362129211426,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12129034847021103,
      "neg_sim_mean": 0.6513180732727051,
      "pos_sim_mean": 0.9525946378707886,
      "step": 1576
    },
    {
      "epoch": 0.3154,
      "grad_norm": 1.3044209480285645,
      "learning_rate": 8.654386993781703e-05,
      "loss": 0.1687,
      "step": 1577
    },
    {
      "epoch": 0.3154,
      "loss_align": 0.1006474494934082,
      "loss_contrastive": 0.07429838180541992,
      "loss_gen": 0.25501102209091187,
      "neg_sim_mean": 0.8736509084701538,
      "pos_sim_mean": 0.8993525505065918,
      "step": 1577
    },
    {
      "epoch": 0.3156,
      "grad_norm": 1.792632818222046,
      "learning_rate": 8.652003698363724e-05,
      "loss": 0.3646,
      "step": 1578
    },
    {
      "epoch": 0.3156,
      "loss_align": 0.23076671361923218,
      "loss_contrastive": 0.08027935028076172,
      "loss_gen": 0.3172222971916199,
      "neg_sim_mean": 0.7495126128196716,
      "pos_sim_mean": 0.7692332863807678,
      "step": 1578
    },
    {
      "epoch": 0.3158,
      "grad_norm": 2.405543088912964,
      "learning_rate": 8.649618623003508e-05,
      "loss": 0.5576,
      "step": 1579
    },
    {
      "epoch": 0.3158,
      "loss_align": 0.142197847366333,
      "loss_contrastive": 0.10506975650787354,
      "loss_gen": 0.30215415358543396,
      "neg_sim_mean": 0.8628718852996826,
      "pos_sim_mean": 0.857802152633667,
      "step": 1579
    },
    {
      "epoch": 0.316,
      "grad_norm": 1.8695098161697388,
      "learning_rate": 8.647231768863512e-05,
      "loss": 0.457,
      "step": 1580
    },
    {
      "epoch": 0.316,
      "loss_align": 0.26313889026641846,
      "loss_contrastive": 0.04573988914489746,
      "loss_gen": 0.24220211803913116,
      "neg_sim_mean": 0.6826009750366211,
      "pos_sim_mean": 0.7368611097335815,
      "step": 1580
    },
    {
      "epoch": 0.3162,
      "grad_norm": 1.9144333600997925,
      "learning_rate": 8.644843137107059e-05,
      "loss": 0.5108,
      "step": 1581
    },
    {
      "epoch": 0.3162,
      "loss_align": 0.06505858898162842,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27139392495155334,
      "neg_sim_mean": 0.8175430297851562,
      "pos_sim_mean": 0.9349414110183716,
      "step": 1581
    },
    {
      "epoch": 0.3164,
      "grad_norm": 1.9313966035842896,
      "learning_rate": 8.642452728898338e-05,
      "loss": 0.3365,
      "step": 1582
    },
    {
      "epoch": 0.3164,
      "loss_align": 0.1912279725074768,
      "loss_contrastive": 0.04056459665298462,
      "loss_gen": 0.28319793939590454,
      "neg_sim_mean": 0.7493366003036499,
      "pos_sim_mean": 0.8087720274925232,
      "step": 1582
    },
    {
      "epoch": 0.3166,
      "grad_norm": 1.8969430923461914,
      "learning_rate": 8.640060545402406e-05,
      "loss": 0.4793,
      "step": 1583
    },
    {
      "epoch": 0.3166,
      "loss_align": 0.17148357629776,
      "loss_contrastive": 0.12281054258346558,
      "loss_gen": 0.35135617852211,
      "neg_sim_mean": 0.8513269424438477,
      "pos_sim_mean": 0.82851642370224,
      "step": 1583
    },
    {
      "epoch": 0.3168,
      "grad_norm": 2.145258903503418,
      "learning_rate": 8.637666587785184e-05,
      "loss": 0.5376,
      "step": 1584
    },
    {
      "epoch": 0.3168,
      "loss_align": 0.3145660161972046,
      "loss_contrastive": 0.0963207483291626,
      "loss_gen": 0.23936492204666138,
      "neg_sim_mean": 0.6817547082901001,
      "pos_sim_mean": 0.6854339838027954,
      "step": 1584
    },
    {
      "epoch": 0.317,
      "grad_norm": 1.7997231483459473,
      "learning_rate": 8.63527085721346e-05,
      "loss": 0.5655,
      "step": 1585
    },
    {
      "epoch": 0.317,
      "loss_align": 0.22509461641311646,
      "loss_contrastive": 0.07190018892288208,
      "loss_gen": 0.2632492482662201,
      "neg_sim_mean": 0.7468055486679077,
      "pos_sim_mean": 0.7749053835868835,
      "step": 1585
    },
    {
      "epoch": 0.3172,
      "grad_norm": 2.16416072845459,
      "learning_rate": 8.63287335485488e-05,
      "loss": 0.497,
      "step": 1586
    },
    {
      "epoch": 0.3172,
      "loss_align": 0.21520715951919556,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0068838708102703094,
      "neg_sim_mean": 0.5562286376953125,
      "pos_sim_mean": 0.7847928404808044,
      "step": 1586
    },
    {
      "epoch": 0.3174,
      "grad_norm": 0.6778222918510437,
      "learning_rate": 8.630474081877959e-05,
      "loss": 0.2221,
      "step": 1587
    },
    {
      "epoch": 0.3174,
      "loss_align": 0.3618183135986328,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04434290900826454,
      "neg_sim_mean": 0.3807797133922577,
      "pos_sim_mean": 0.6381816864013672,
      "step": 1587
    },
    {
      "epoch": 0.3176,
      "grad_norm": 1.5714203119277954,
      "learning_rate": 8.628073039452076e-05,
      "loss": 0.4062,
      "step": 1588
    },
    {
      "epoch": 0.3176,
      "loss_align": 0.23697912693023682,
      "loss_contrastive": 0.1380121111869812,
      "loss_gen": 0.37344178557395935,
      "neg_sim_mean": 0.8010329604148865,
      "pos_sim_mean": 0.7630208730697632,
      "step": 1588
    },
    {
      "epoch": 0.3178,
      "grad_norm": 2.4488000869750977,
      "learning_rate": 8.625670228747467e-05,
      "loss": 0.627,
      "step": 1589
    },
    {
      "epoch": 0.3178,
      "loss_align": 0.34898078441619873,
      "loss_contrastive": 0.09124815464019775,
      "loss_gen": 0.28812679648399353,
      "neg_sim_mean": 0.6422673463821411,
      "pos_sim_mean": 0.6510192155838013,
      "step": 1589
    },
    {
      "epoch": 0.318,
      "grad_norm": 2.122382402420044,
      "learning_rate": 8.623265650935234e-05,
      "loss": 0.6481,
      "step": 1590
    },
    {
      "epoch": 0.318,
      "loss_align": 0.1760956048965454,
      "loss_contrastive": 0.13144612312316895,
      "loss_gen": 0.2668725550174713,
      "neg_sim_mean": 0.8553504943847656,
      "pos_sim_mean": 0.8239043951034546,
      "step": 1590
    },
    {
      "epoch": 0.3182,
      "grad_norm": 1.7500731945037842,
      "learning_rate": 8.620859307187339e-05,
      "loss": 0.4587,
      "step": 1591
    },
    {
      "epoch": 0.3182,
      "loss_align": 0.2540210485458374,
      "loss_contrastive": 0.09146618843078613,
      "loss_gen": 0.4187440574169159,
      "neg_sim_mean": 0.7374451160430908,
      "pos_sim_mean": 0.7459789514541626,
      "step": 1591
    },
    {
      "epoch": 0.3184,
      "grad_norm": 2.1706504821777344,
      "learning_rate": 8.618451198676603e-05,
      "loss": 0.6837,
      "step": 1592
    },
    {
      "epoch": 0.3184,
      "loss_align": 0.20220530033111572,
      "loss_contrastive": 0.09461557865142822,
      "loss_gen": 0.23253530263900757,
      "neg_sim_mean": 0.7924102544784546,
      "pos_sim_mean": 0.7977946996688843,
      "step": 1592
    },
    {
      "epoch": 0.3186,
      "grad_norm": 2.3432769775390625,
      "learning_rate": 8.616041326576711e-05,
      "loss": 0.4461,
      "step": 1593
    },
    {
      "epoch": 0.3186,
      "loss_align": 0.1380530595779419,
      "loss_contrastive": 0.08289653062820435,
      "loss_gen": 0.2398836761713028,
      "neg_sim_mean": 0.8448434472084045,
      "pos_sim_mean": 0.8619469404220581,
      "step": 1593
    },
    {
      "epoch": 0.3188,
      "grad_norm": 2.04162859916687,
      "learning_rate": 8.613629692062204e-05,
      "loss": 0.3879,
      "step": 1594
    },
    {
      "epoch": 0.3188,
      "loss_align": 0.14726710319519043,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0031944122165441513,
      "neg_sim_mean": 0.3448134660720825,
      "pos_sim_mean": 0.8527328968048096,
      "step": 1594
    },
    {
      "epoch": 0.319,
      "grad_norm": 0.5095478892326355,
      "learning_rate": 8.611216296308484e-05,
      "loss": 0.1505,
      "step": 1595
    },
    {
      "epoch": 0.319,
      "loss_align": 0.20788860321044922,
      "loss_contrastive": 0.09173864126205444,
      "loss_gen": 0.23673021793365479,
      "neg_sim_mean": 0.7838500142097473,
      "pos_sim_mean": 0.7921113967895508,
      "step": 1595
    },
    {
      "epoch": 0.3192,
      "grad_norm": 1.6984565258026123,
      "learning_rate": 8.608801140491811e-05,
      "loss": 0.4556,
      "step": 1596
    },
    {
      "epoch": 0.3192,
      "loss_align": 0.2488335371017456,
      "loss_contrastive": 0.04696756601333618,
      "loss_gen": 0.26893898844718933,
      "neg_sim_mean": 0.6981340050697327,
      "pos_sim_mean": 0.7511664628982544,
      "step": 1596
    },
    {
      "epoch": 0.3194,
      "grad_norm": 1.8592195510864258,
      "learning_rate": 8.606384225789303e-05,
      "loss": 0.5234,
      "step": 1597
    },
    {
      "epoch": 0.3194,
      "loss_align": 0.2308884859085083,
      "loss_contrastive": 0.0,
      "loss_gen": 0.027362069115042686,
      "neg_sim_mean": 0.29357054829597473,
      "pos_sim_mean": 0.7691115140914917,
      "step": 1597
    },
    {
      "epoch": 0.3196,
      "grad_norm": 1.1135114431381226,
      "learning_rate": 8.603965553378934e-05,
      "loss": 0.2583,
      "step": 1598
    },
    {
      "epoch": 0.3196,
      "loss_align": 0.178810715675354,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3061872720718384,
      "neg_sim_mean": 0.5765891671180725,
      "pos_sim_mean": 0.821189284324646,
      "step": 1598
    },
    {
      "epoch": 0.3198,
      "grad_norm": 2.4170656204223633,
      "learning_rate": 8.601545124439535e-05,
      "loss": 0.485,
      "step": 1599
    },
    {
      "epoch": 0.3198,
      "loss_align": 0.2091461420059204,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002967755775898695,
      "neg_sim_mean": 0.4194120168685913,
      "pos_sim_mean": 0.7908538579940796,
      "step": 1599
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4766463041305542,
      "learning_rate": 8.599122940150795e-05,
      "loss": 0.2121,
      "step": 1600
    },
    {
      "epoch": 0.32,
      "loss_align": 0.22248733043670654,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3237919211387634,
      "neg_sim_mean": 0.6518687605857849,
      "pos_sim_mean": 0.7775126695632935,
      "step": 1600
    },
    {
      "epoch": 0.3202,
      "grad_norm": 2.18330717086792,
      "learning_rate": 8.596699001693255e-05,
      "loss": 0.5463,
      "step": 1601
    },
    {
      "epoch": 0.3202,
      "loss_align": 0.14771270751953125,
      "loss_contrastive": 0.14612513780593872,
      "loss_gen": 0.288551390171051,
      "neg_sim_mean": 0.8984124064445496,
      "pos_sim_mean": 0.8522872924804688,
      "step": 1601
    },
    {
      "epoch": 0.3204,
      "grad_norm": 1.842478632926941,
      "learning_rate": 8.594273310248318e-05,
      "loss": 0.4538,
      "step": 1602
    },
    {
      "epoch": 0.3204,
      "loss_align": 0.41831719875335693,
      "loss_contrastive": 0.1779555380344391,
      "loss_gen": 0.36539164185523987,
      "neg_sim_mean": 0.6596383452415466,
      "pos_sim_mean": 0.5816828012466431,
      "step": 1602
    },
    {
      "epoch": 0.3206,
      "grad_norm": 2.5013036727905273,
      "learning_rate": 8.591845866998231e-05,
      "loss": 0.8051,
      "step": 1603
    },
    {
      "epoch": 0.3206,
      "loss_align": 0.1999218463897705,
      "loss_contrastive": 0.09820675849914551,
      "loss_gen": 0.2144298404455185,
      "neg_sim_mean": 0.7982848882675171,
      "pos_sim_mean": 0.8000781536102295,
      "step": 1603
    },
    {
      "epoch": 0.3208,
      "grad_norm": 1.9187188148498535,
      "learning_rate": 8.589416673126103e-05,
      "loss": 0.4261,
      "step": 1604
    },
    {
      "epoch": 0.3208,
      "loss_align": 0.1748523712158203,
      "loss_contrastive": 0.07790917158126831,
      "loss_gen": 0.2766612768173218,
      "neg_sim_mean": 0.8030567765235901,
      "pos_sim_mean": 0.8251476287841797,
      "step": 1604
    },
    {
      "epoch": 0.321,
      "grad_norm": 2.2230794429779053,
      "learning_rate": 8.586985729815894e-05,
      "loss": 0.4609,
      "step": 1605
    },
    {
      "epoch": 0.321,
      "loss_align": 0.12975263595581055,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2781105637550354,
      "neg_sim_mean": 0.7688134908676147,
      "pos_sim_mean": 0.8702473640441895,
      "step": 1605
    },
    {
      "epoch": 0.3212,
      "grad_norm": 1.8134979009628296,
      "learning_rate": 8.584553038252414e-05,
      "loss": 0.4079,
      "step": 1606
    },
    {
      "epoch": 0.3212,
      "loss_align": 0.36653733253479004,
      "loss_contrastive": 0.16836023330688477,
      "loss_gen": 0.08334110677242279,
      "neg_sim_mean": 0.7018228769302368,
      "pos_sim_mean": 0.63346266746521,
      "step": 1606
    },
    {
      "epoch": 0.3214,
      "grad_norm": 1.3504393100738525,
      "learning_rate": 8.58211859962133e-05,
      "loss": 0.4701,
      "step": 1607
    },
    {
      "epoch": 0.3214,
      "loss_align": 0.21653878688812256,
      "loss_contrastive": 0.06306231021881104,
      "loss_gen": 0.3386731445789337,
      "neg_sim_mean": 0.7465234994888306,
      "pos_sim_mean": 0.7834612131118774,
      "step": 1607
    },
    {
      "epoch": 0.3216,
      "grad_norm": 2.30208158493042,
      "learning_rate": 8.579682415109156e-05,
      "loss": 0.5628,
      "step": 1608
    },
    {
      "epoch": 0.3216,
      "loss_align": 0.3266794681549072,
      "loss_contrastive": 0.24094462394714355,
      "loss_gen": 0.2480890452861786,
      "neg_sim_mean": 0.8142651319503784,
      "pos_sim_mean": 0.6733205318450928,
      "step": 1608
    },
    {
      "epoch": 0.3218,
      "grad_norm": 2.076960802078247,
      "learning_rate": 8.577244485903259e-05,
      "loss": 0.6037,
      "step": 1609
    },
    {
      "epoch": 0.3218,
      "loss_align": 0.3614336848258972,
      "loss_contrastive": 0.1273404359817505,
      "loss_gen": 0.21646329760551453,
      "neg_sim_mean": 0.6659067273139954,
      "pos_sim_mean": 0.6385663151741028,
      "step": 1609
    },
    {
      "epoch": 0.322,
      "grad_norm": 1.851291298866272,
      "learning_rate": 8.574804813191858e-05,
      "loss": 0.5932,
      "step": 1610
    },
    {
      "epoch": 0.322,
      "loss_align": 0.20102381706237793,
      "loss_contrastive": 0.10700386762619019,
      "loss_gen": 0.20767278969287872,
      "neg_sim_mean": 0.8059800267219543,
      "pos_sim_mean": 0.7989761829376221,
      "step": 1610
    },
    {
      "epoch": 0.3222,
      "grad_norm": 1.5276837348937988,
      "learning_rate": 8.572363398164017e-05,
      "loss": 0.4215,
      "step": 1611
    },
    {
      "epoch": 0.3222,
      "loss_align": 0.16188335418701172,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011106421239674091,
      "neg_sim_mean": 0.4027406871318817,
      "pos_sim_mean": 0.8381166458129883,
      "step": 1611
    },
    {
      "epoch": 0.3224,
      "grad_norm": 0.7909839153289795,
      "learning_rate": 8.569920242009655e-05,
      "loss": 0.173,
      "step": 1612
    },
    {
      "epoch": 0.3224,
      "loss_align": 0.29924535751342773,
      "loss_contrastive": 0.182589590549469,
      "loss_gen": 0.10634540766477585,
      "neg_sim_mean": 0.7833442091941833,
      "pos_sim_mean": 0.7007546424865723,
      "step": 1612
    },
    {
      "epoch": 0.3226,
      "grad_norm": 1.6779345273971558,
      "learning_rate": 8.567475345919532e-05,
      "loss": 0.4275,
      "step": 1613
    },
    {
      "epoch": 0.3226,
      "loss_align": 0.33067625761032104,
      "loss_contrastive": 0.039773762226104736,
      "loss_gen": 0.3521304130554199,
      "neg_sim_mean": 0.6090974807739258,
      "pos_sim_mean": 0.669323742389679,
      "step": 1613
    },
    {
      "epoch": 0.3228,
      "grad_norm": 2.1589510440826416,
      "learning_rate": 8.565028711085265e-05,
      "loss": 0.6876,
      "step": 1614
    },
    {
      "epoch": 0.3228,
      "loss_align": 0.16329145431518555,
      "loss_contrastive": 0.1219131350517273,
      "loss_gen": 0.17021729052066803,
      "neg_sim_mean": 0.8586216568946838,
      "pos_sim_mean": 0.8367085456848145,
      "step": 1614
    },
    {
      "epoch": 0.323,
      "grad_norm": 1.7246286869049072,
      "learning_rate": 8.562580338699313e-05,
      "loss": 0.3481,
      "step": 1615
    },
    {
      "epoch": 0.323,
      "loss_align": 0.18139296770095825,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1641116738319397,
      "neg_sim_mean": 0.7139067649841309,
      "pos_sim_mean": 0.8186070322990417,
      "step": 1615
    },
    {
      "epoch": 0.3232,
      "grad_norm": 1.6989614963531494,
      "learning_rate": 8.560130229954984e-05,
      "loss": 0.3455,
      "step": 1616
    },
    {
      "epoch": 0.3232,
      "loss_align": 0.33448731899261475,
      "loss_contrastive": 0.1309235692024231,
      "loss_gen": 0.40350231528282166,
      "neg_sim_mean": 0.6964362263679504,
      "pos_sim_mean": 0.6655126810073853,
      "step": 1616
    },
    {
      "epoch": 0.3234,
      "grad_norm": 3.134629011154175,
      "learning_rate": 8.557678386046428e-05,
      "loss": 0.7537,
      "step": 1617
    },
    {
      "epoch": 0.3234,
      "loss_align": 0.19029009342193604,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011904857121407986,
      "neg_sim_mean": 0.3734191954135895,
      "pos_sim_mean": 0.809709906578064,
      "step": 1617
    },
    {
      "epoch": 0.3236,
      "grad_norm": 1.0695385932922363,
      "learning_rate": 8.555224808168644e-05,
      "loss": 0.2022,
      "step": 1618
    },
    {
      "epoch": 0.3236,
      "loss_align": 0.22943878173828125,
      "loss_contrastive": 0.1959991455078125,
      "loss_gen": 0.2369905263185501,
      "neg_sim_mean": 0.8665603399276733,
      "pos_sim_mean": 0.7705612182617188,
      "step": 1618
    },
    {
      "epoch": 0.3238,
      "grad_norm": 2.3348934650421143,
      "learning_rate": 8.552769497517482e-05,
      "loss": 0.4899,
      "step": 1619
    },
    {
      "epoch": 0.3238,
      "loss_align": 0.2537076473236084,
      "loss_contrastive": 0.2106422781944275,
      "loss_gen": 0.24054056406021118,
      "neg_sim_mean": 0.8569346070289612,
      "pos_sim_mean": 0.7462923526763916,
      "step": 1619
    },
    {
      "epoch": 0.324,
      "grad_norm": 1.7915163040161133,
      "learning_rate": 8.550312455289625e-05,
      "loss": 0.5195,
      "step": 1620
    },
    {
      "epoch": 0.324,
      "loss_align": 0.1360476016998291,
      "loss_contrastive": 0.07374519109725952,
      "loss_gen": 0.4828230142593384,
      "neg_sim_mean": 0.8376975655555725,
      "pos_sim_mean": 0.8639523983001709,
      "step": 1620
    },
    {
      "epoch": 0.3242,
      "grad_norm": 2.2264811992645264,
      "learning_rate": 8.547853682682604e-05,
      "loss": 0.6277,
      "step": 1621
    },
    {
      "epoch": 0.3242,
      "loss_align": 0.29661309719085693,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2790378928184509,
      "neg_sim_mean": 0.5634996294975281,
      "pos_sim_mean": 0.7033869028091431,
      "step": 1621
    },
    {
      "epoch": 0.3244,
      "grad_norm": 2.005655527114868,
      "learning_rate": 8.5453931808948e-05,
      "loss": 0.5757,
      "step": 1622
    },
    {
      "epoch": 0.3244,
      "loss_align": 0.5115126371383667,
      "loss_contrastive": 0.2086820900440216,
      "loss_gen": 0.25595152378082275,
      "neg_sim_mean": 0.5971694588661194,
      "pos_sim_mean": 0.4884873628616333,
      "step": 1622
    },
    {
      "epoch": 0.3246,
      "grad_norm": 2.4153647422790527,
      "learning_rate": 8.542930951125432e-05,
      "loss": 0.7925,
      "step": 1623
    },
    {
      "epoch": 0.3246,
      "loss_align": 0.27687084674835205,
      "loss_contrastive": 0.04262375831604004,
      "loss_gen": 0.39235812425613403,
      "neg_sim_mean": 0.6657528877258301,
      "pos_sim_mean": 0.723129153251648,
      "step": 1623
    },
    {
      "epoch": 0.3248,
      "grad_norm": 2.1411964893341064,
      "learning_rate": 8.540466994574555e-05,
      "loss": 0.6743,
      "step": 1624
    },
    {
      "epoch": 0.3248,
      "loss_align": 0.18437892198562622,
      "loss_contrastive": 0.1213299036026001,
      "loss_gen": 0.27881211042404175,
      "neg_sim_mean": 0.836950957775116,
      "pos_sim_mean": 0.8156210780143738,
      "step": 1624
    },
    {
      "epoch": 0.325,
      "grad_norm": 1.864424467086792,
      "learning_rate": 8.538001312443077e-05,
      "loss": 0.4778,
      "step": 1625
    },
    {
      "epoch": 0.325,
      "loss_align": 0.18128955364227295,
      "loss_contrastive": 0.05713081359863281,
      "loss_gen": 0.1829233467578888,
      "neg_sim_mean": 0.775841236114502,
      "pos_sim_mean": 0.818710446357727,
      "step": 1625
    },
    {
      "epoch": 0.3252,
      "grad_norm": 1.7990089654922485,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.3711,
      "step": 1626
    },
    {
      "epoch": 0.3252,
      "loss_align": 0.21756184101104736,
      "loss_contrastive": 0.13678443431854248,
      "loss_gen": 0.33032241463661194,
      "neg_sim_mean": 0.8192225694656372,
      "pos_sim_mean": 0.7824381589889526,
      "step": 1626
    },
    {
      "epoch": 0.3254,
      "grad_norm": 2.2877774238586426,
      "learning_rate": 8.533064776246125e-05,
      "loss": 0.5643,
      "step": 1627
    },
    {
      "epoch": 0.3254,
      "loss_align": 0.11136829853057861,
      "loss_contrastive": 0.10496973991394043,
      "loss_gen": 0.19163087010383606,
      "neg_sim_mean": 0.8936014175415039,
      "pos_sim_mean": 0.8886317014694214,
      "step": 1627
    },
    {
      "epoch": 0.3256,
      "grad_norm": 1.6115758419036865,
      "learning_rate": 8.530593924586659e-05,
      "loss": 0.3156,
      "step": 1628
    },
    {
      "epoch": 0.3256,
      "loss_align": 0.22084903717041016,
      "loss_contrastive": 0.11905479431152344,
      "loss_gen": 0.26042935252189636,
      "neg_sim_mean": 0.7982057332992554,
      "pos_sim_mean": 0.7791509628295898,
      "step": 1628
    },
    {
      "epoch": 0.3258,
      "grad_norm": 1.994745135307312,
      "learning_rate": 8.528121352158604e-05,
      "loss": 0.4956,
      "step": 1629
    },
    {
      "epoch": 0.3258,
      "loss_align": 0.38280820846557617,
      "loss_contrastive": 0.19372928142547607,
      "loss_gen": 0.22801516950130463,
      "neg_sim_mean": 0.710921049118042,
      "pos_sim_mean": 0.6171917915344238,
      "step": 1629
    },
    {
      "epoch": 0.326,
      "grad_norm": 1.7952793836593628,
      "learning_rate": 8.525647060167063e-05,
      "loss": 0.6341,
      "step": 1630
    },
    {
      "epoch": 0.326,
      "loss_align": 0.2437680959701538,
      "loss_contrastive": 0.08922135829925537,
      "loss_gen": 0.21268302202224731,
      "neg_sim_mean": 0.7454532384872437,
      "pos_sim_mean": 0.7562319040298462,
      "step": 1630
    },
    {
      "epoch": 0.3262,
      "grad_norm": 2.055982828140259,
      "learning_rate": 8.523171049817974e-05,
      "loss": 0.4672,
      "step": 1631
    },
    {
      "epoch": 0.3262,
      "loss_align": 0.17679989337921143,
      "loss_contrastive": 0.06418812274932861,
      "loss_gen": 0.27990105748176575,
      "neg_sim_mean": 0.7873882055282593,
      "pos_sim_mean": 0.8232001066207886,
      "step": 1631
    },
    {
      "epoch": 0.3264,
      "grad_norm": 2.041987895965576,
      "learning_rate": 8.520693322318116e-05,
      "loss": 0.4644,
      "step": 1632
    },
    {
      "epoch": 0.3264,
      "loss_align": 0.21672534942626953,
      "loss_contrastive": 0.05484288930892944,
      "loss_gen": 0.1554149091243744,
      "neg_sim_mean": 0.738117516040802,
      "pos_sim_mean": 0.7832746505737305,
      "step": 1632
    },
    {
      "epoch": 0.3266,
      "grad_norm": 1.5111712217330933,
      "learning_rate": 8.518213878875103e-05,
      "loss": 0.3787,
      "step": 1633
    },
    {
      "epoch": 0.3266,
      "loss_align": 0.14580696821212769,
      "loss_contrastive": 0.06964010000228882,
      "loss_gen": 0.3664219379425049,
      "neg_sim_mean": 0.8238331079483032,
      "pos_sim_mean": 0.8541930317878723,
      "step": 1633
    },
    {
      "epoch": 0.3268,
      "grad_norm": 2.2300355434417725,
      "learning_rate": 8.515732720697383e-05,
      "loss": 0.5206,
      "step": 1634
    },
    {
      "epoch": 0.3268,
      "loss_align": 0.2040344476699829,
      "loss_contrastive": 0.0,
      "loss_gen": 0.197567418217659,
      "neg_sim_mean": 0.6810770034790039,
      "pos_sim_mean": 0.7959655523300171,
      "step": 1634
    },
    {
      "epoch": 0.327,
      "grad_norm": 1.7171162366867065,
      "learning_rate": 8.513249848994246e-05,
      "loss": 0.4016,
      "step": 1635
    },
    {
      "epoch": 0.327,
      "loss_align": 0.22026383876800537,
      "loss_contrastive": 0.012300848960876465,
      "loss_gen": 0.37010160088539124,
      "neg_sim_mean": 0.6920369863510132,
      "pos_sim_mean": 0.7797361612319946,
      "step": 1635
    },
    {
      "epoch": 0.3272,
      "grad_norm": 1.9224302768707275,
      "learning_rate": 8.510765264975813e-05,
      "loss": 0.5918,
      "step": 1636
    },
    {
      "epoch": 0.3272,
      "loss_align": 0.2757059931755066,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2719745337963104,
      "neg_sim_mean": 0.5008509159088135,
      "pos_sim_mean": 0.7242940068244934,
      "step": 1636
    },
    {
      "epoch": 0.3274,
      "grad_norm": 2.0164737701416016,
      "learning_rate": 8.508278969853037e-05,
      "loss": 0.5477,
      "step": 1637
    },
    {
      "epoch": 0.3274,
      "loss_align": 0.1218869686126709,
      "loss_contrastive": 0.04655790328979492,
      "loss_gen": 0.39183342456817627,
      "neg_sim_mean": 0.8246709108352661,
      "pos_sim_mean": 0.8781130313873291,
      "step": 1637
    },
    {
      "epoch": 0.3276,
      "grad_norm": 2.065077781677246,
      "learning_rate": 8.505790964837713e-05,
      "loss": 0.5193,
      "step": 1638
    },
    {
      "epoch": 0.3276,
      "loss_align": 0.15152227878570557,
      "loss_contrastive": 0.0,
      "loss_gen": 0.027528198435902596,
      "neg_sim_mean": 0.47256842255592346,
      "pos_sim_mean": 0.8484777212142944,
      "step": 1638
    },
    {
      "epoch": 0.3278,
      "grad_norm": 1.2420003414154053,
      "learning_rate": 8.503301251142459e-05,
      "loss": 0.1791,
      "step": 1639
    },
    {
      "epoch": 0.3278,
      "loss_align": 0.122519850730896,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03050048090517521,
      "neg_sim_mean": 0.3445688784122467,
      "pos_sim_mean": 0.877480149269104,
      "step": 1639
    },
    {
      "epoch": 0.328,
      "grad_norm": 1.0704612731933594,
      "learning_rate": 8.500809829980734e-05,
      "loss": 0.153,
      "step": 1640
    },
    {
      "epoch": 0.328,
      "loss_align": 0.21040701866149902,
      "loss_contrastive": 0.056934475898742676,
      "loss_gen": 0.15275968611240387,
      "neg_sim_mean": 0.7465274333953857,
      "pos_sim_mean": 0.789592981338501,
      "step": 1640
    },
    {
      "epoch": 0.3282,
      "grad_norm": 1.7507376670837402,
      "learning_rate": 8.498316702566828e-05,
      "loss": 0.37,
      "step": 1641
    },
    {
      "epoch": 0.3282,
      "loss_align": 0.12992030382156372,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02399665303528309,
      "neg_sim_mean": 0.3615715205669403,
      "pos_sim_mean": 0.8700796961784363,
      "step": 1641
    },
    {
      "epoch": 0.3284,
      "grad_norm": 1.230289340019226,
      "learning_rate": 8.495821870115858e-05,
      "loss": 0.1539,
      "step": 1642
    },
    {
      "epoch": 0.3284,
      "loss_align": 0.16294056177139282,
      "loss_contrastive": 0.038127601146698,
      "loss_gen": 0.3588908612728119,
      "neg_sim_mean": 0.7751870155334473,
      "pos_sim_mean": 0.8370594382286072,
      "step": 1642
    },
    {
      "epoch": 0.3286,
      "grad_norm": 2.0645875930786133,
      "learning_rate": 8.493325333843776e-05,
      "loss": 0.5264,
      "step": 1643
    },
    {
      "epoch": 0.3286,
      "loss_align": 0.2761056423187256,
      "loss_contrastive": 0.11443924903869629,
      "loss_gen": 0.13785724341869354,
      "neg_sim_mean": 0.7383335828781128,
      "pos_sim_mean": 0.7238943576812744,
      "step": 1643
    },
    {
      "epoch": 0.3288,
      "grad_norm": 1.5824426412582397,
      "learning_rate": 8.490827094967363e-05,
      "loss": 0.4277,
      "step": 1644
    },
    {
      "epoch": 0.3288,
      "loss_align": 0.20356571674346924,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003159794257953763,
      "neg_sim_mean": 0.4745522141456604,
      "pos_sim_mean": 0.7964342832565308,
      "step": 1644
    },
    {
      "epoch": 0.329,
      "grad_norm": 0.43506038188934326,
      "learning_rate": 8.488327154704232e-05,
      "loss": 0.2067,
      "step": 1645
    },
    {
      "epoch": 0.329,
      "loss_align": 0.2491869330406189,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2099556177854538,
      "neg_sim_mean": 0.5665215253829956,
      "pos_sim_mean": 0.7508130669593811,
      "step": 1645
    },
    {
      "epoch": 0.3292,
      "grad_norm": 1.6675069332122803,
      "learning_rate": 8.485825514272824e-05,
      "loss": 0.4591,
      "step": 1646
    },
    {
      "epoch": 0.3292,
      "loss_align": 0.16352808475494385,
      "loss_contrastive": 0.028002262115478516,
      "loss_gen": 0.3044705092906952,
      "neg_sim_mean": 0.7644741535186768,
      "pos_sim_mean": 0.8364719152450562,
      "step": 1646
    },
    {
      "epoch": 0.3294,
      "grad_norm": 1.9389381408691406,
      "learning_rate": 8.483322174892404e-05,
      "loss": 0.4714,
      "step": 1647
    },
    {
      "epoch": 0.3294,
      "loss_align": 0.2786813974380493,
      "loss_contrastive": 0.005554795265197754,
      "loss_gen": 0.006678341422230005,
      "neg_sim_mean": 0.6268733739852905,
      "pos_sim_mean": 0.7213186025619507,
      "step": 1647
    },
    {
      "epoch": 0.3296,
      "grad_norm": 0.8492620587348938,
      "learning_rate": 8.480817137783073e-05,
      "loss": 0.286,
      "step": 1648
    },
    {
      "epoch": 0.3296,
      "loss_align": 0.13058888912200928,
      "loss_contrastive": 0.11874067783355713,
      "loss_gen": 0.2766217291355133,
      "neg_sim_mean": 0.8881517648696899,
      "pos_sim_mean": 0.8694111108779907,
      "step": 1648
    },
    {
      "epoch": 0.3298,
      "grad_norm": 2.0500478744506836,
      "learning_rate": 8.478310404165755e-05,
      "loss": 0.4215,
      "step": 1649
    },
    {
      "epoch": 0.3298,
      "loss_align": 0.39958643913269043,
      "loss_contrastive": 0.16359621286392212,
      "loss_gen": 0.32820847630500793,
      "neg_sim_mean": 0.6640097498893738,
      "pos_sim_mean": 0.6004135608673096,
      "step": 1649
    },
    {
      "epoch": 0.33,
      "grad_norm": 2.257153272628784,
      "learning_rate": 8.4758019752622e-05,
      "loss": 0.7474,
      "step": 1650
    },
    {
      "epoch": 0.33,
      "loss_align": 0.32995784282684326,
      "loss_contrastive": 0.17099136114120483,
      "loss_gen": 0.2708996534347534,
      "neg_sim_mean": 0.7410334944725037,
      "pos_sim_mean": 0.6700421571731567,
      "step": 1650
    },
    {
      "epoch": 0.3302,
      "grad_norm": 1.9612929821014404,
      "learning_rate": 8.473291852294987e-05,
      "loss": 0.6214,
      "step": 1651
    },
    {
      "epoch": 0.3302,
      "loss_align": 0.2247520089149475,
      "loss_contrastive": 0.14327579736709595,
      "loss_gen": 0.42733341455459595,
      "neg_sim_mean": 0.8185237646102905,
      "pos_sim_mean": 0.7752479910850525,
      "step": 1651
    },
    {
      "epoch": 0.3304,
      "grad_norm": 2.6147501468658447,
      "learning_rate": 8.47078003648752e-05,
      "loss": 0.6693,
      "step": 1652
    },
    {
      "epoch": 0.3304,
      "loss_align": 0.2876942753791809,
      "loss_contrastive": 0.05283695459365845,
      "loss_gen": 0.1460365355014801,
      "neg_sim_mean": 0.6651426553726196,
      "pos_sim_mean": 0.7123057246208191,
      "step": 1652
    },
    {
      "epoch": 0.3306,
      "grad_norm": 1.9987832307815552,
      "learning_rate": 8.468266529064025e-05,
      "loss": 0.4401,
      "step": 1653
    },
    {
      "epoch": 0.3306,
      "loss_align": 0.1307600736618042,
      "loss_contrastive": 0.04594719409942627,
      "loss_gen": 0.2172991782426834,
      "neg_sim_mean": 0.8151870965957642,
      "pos_sim_mean": 0.8692399263381958,
      "step": 1653
    },
    {
      "epoch": 0.3308,
      "grad_norm": 1.6770305633544922,
      "learning_rate": 8.465751331249559e-05,
      "loss": 0.3536,
      "step": 1654
    },
    {
      "epoch": 0.3308,
      "loss_align": 0.19249093532562256,
      "loss_contrastive": 0.05298346281051636,
      "loss_gen": 0.4813081920146942,
      "neg_sim_mean": 0.7604925036430359,
      "pos_sim_mean": 0.8075090646743774,
      "step": 1654
    },
    {
      "epoch": 0.331,
      "grad_norm": 2.59965443611145,
      "learning_rate": 8.463234444269995e-05,
      "loss": 0.6802,
      "step": 1655
    },
    {
      "epoch": 0.331,
      "loss_align": 0.41527748107910156,
      "loss_contrastive": 0.09818795323371887,
      "loss_gen": 0.2909018099308014,
      "neg_sim_mean": 0.5829104781150818,
      "pos_sim_mean": 0.5847225189208984,
      "step": 1655
    },
    {
      "epoch": 0.3312,
      "grad_norm": 2.398383617401123,
      "learning_rate": 8.460715869352035e-05,
      "loss": 0.718,
      "step": 1656
    },
    {
      "epoch": 0.3312,
      "loss_align": 0.2583826780319214,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1834942102432251,
      "neg_sim_mean": 0.6294806599617004,
      "pos_sim_mean": 0.7416173219680786,
      "step": 1656
    },
    {
      "epoch": 0.3314,
      "grad_norm": 1.5962258577346802,
      "learning_rate": 8.4581956077232e-05,
      "loss": 0.4419,
      "step": 1657
    },
    {
      "epoch": 0.3314,
      "loss_align": 0.1720505952835083,
      "loss_contrastive": 0.0,
      "loss_gen": 0.35594117641448975,
      "neg_sim_mean": 0.607498288154602,
      "pos_sim_mean": 0.8279494047164917,
      "step": 1657
    },
    {
      "epoch": 0.3316,
      "grad_norm": 2.383251190185547,
      "learning_rate": 8.45567366061184e-05,
      "loss": 0.528,
      "step": 1658
    },
    {
      "epoch": 0.3316,
      "loss_align": 0.19832241535186768,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0009093114640563726,
      "neg_sim_mean": 0.4061540365219116,
      "pos_sim_mean": 0.8016775846481323,
      "step": 1658
    },
    {
      "epoch": 0.3318,
      "grad_norm": 0.46155795454978943,
      "learning_rate": 8.453150029247114e-05,
      "loss": 0.1992,
      "step": 1659
    },
    {
      "epoch": 0.3318,
      "loss_align": 0.2397478222846985,
      "loss_contrastive": 0.1393294334411621,
      "loss_gen": 0.12091085314750671,
      "neg_sim_mean": 0.7995815873146057,
      "pos_sim_mean": 0.7602521777153015,
      "step": 1659
    },
    {
      "epoch": 0.332,
      "grad_norm": 1.521432638168335,
      "learning_rate": 8.450624714859016e-05,
      "loss": 0.3774,
      "step": 1660
    },
    {
      "epoch": 0.332,
      "loss_align": 0.3816855549812317,
      "loss_contrastive": 0.06922781467437744,
      "loss_gen": 0.23690813779830933,
      "neg_sim_mean": 0.5875422358512878,
      "pos_sim_mean": 0.6183144450187683,
      "step": 1660
    },
    {
      "epoch": 0.3322,
      "grad_norm": 1.9098093509674072,
      "learning_rate": 8.44809771867835e-05,
      "loss": 0.6269,
      "step": 1661
    },
    {
      "epoch": 0.3322,
      "loss_align": 0.11900520324707031,
      "loss_contrastive": 0.05348312854766846,
      "loss_gen": 0.30470043420791626,
      "neg_sim_mean": 0.8344779014587402,
      "pos_sim_mean": 0.8809947967529297,
      "step": 1661
    },
    {
      "epoch": 0.3324,
      "grad_norm": 2.6977643966674805,
      "learning_rate": 8.445569041936743e-05,
      "loss": 0.4301,
      "step": 1662
    },
    {
      "epoch": 0.3324,
      "loss_align": 0.1969916820526123,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29490232467651367,
      "neg_sim_mean": 0.4409315288066864,
      "pos_sim_mean": 0.8030083179473877,
      "step": 1662
    },
    {
      "epoch": 0.3326,
      "grad_norm": 1.9559448957443237,
      "learning_rate": 8.443038685866642e-05,
      "loss": 0.4919,
      "step": 1663
    },
    {
      "epoch": 0.3326,
      "loss_align": 0.13635492324829102,
      "loss_contrastive": 0.10174429416656494,
      "loss_gen": 0.3586962819099426,
      "neg_sim_mean": 0.865389347076416,
      "pos_sim_mean": 0.863645076751709,
      "step": 1663
    },
    {
      "epoch": 0.3328,
      "grad_norm": 2.1306092739105225,
      "learning_rate": 8.440506651701315e-05,
      "loss": 0.5073,
      "step": 1664
    },
    {
      "epoch": 0.3328,
      "loss_align": 0.215509831905365,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00771568575873971,
      "neg_sim_mean": 0.4687831401824951,
      "pos_sim_mean": 0.784490168094635,
      "step": 1664
    },
    {
      "epoch": 0.333,
      "grad_norm": 0.6451334953308105,
      "learning_rate": 8.437972940674838e-05,
      "loss": 0.2232,
      "step": 1665
    },
    {
      "epoch": 0.333,
      "loss_align": 0.1342748999595642,
      "loss_contrastive": 0.07902032136917114,
      "loss_gen": 0.18597589433193207,
      "neg_sim_mean": 0.844745397567749,
      "pos_sim_mean": 0.8657251000404358,
      "step": 1665
    },
    {
      "epoch": 0.3332,
      "grad_norm": 1.537246823310852,
      "learning_rate": 8.435437554022115e-05,
      "loss": 0.3297,
      "step": 1666
    },
    {
      "epoch": 0.3332,
      "loss_align": 0.08943164348602295,
      "loss_contrastive": 0.07150024175643921,
      "loss_gen": 0.3562995195388794,
      "neg_sim_mean": 0.8820685744285583,
      "pos_sim_mean": 0.910568356513977,
      "step": 1666
    },
    {
      "epoch": 0.3334,
      "grad_norm": 2.4733216762542725,
      "learning_rate": 8.432900492978863e-05,
      "loss": 0.4543,
      "step": 1667
    },
    {
      "epoch": 0.3334,
      "loss_align": 0.18308991193771362,
      "loss_contrastive": 0.028934240341186523,
      "loss_gen": 0.5033236742019653,
      "neg_sim_mean": 0.745844304561615,
      "pos_sim_mean": 0.8169100880622864,
      "step": 1667
    },
    {
      "epoch": 0.3336,
      "grad_norm": 3.0894010066986084,
      "learning_rate": 8.430361758781616e-05,
      "loss": 0.6899,
      "step": 1668
    },
    {
      "epoch": 0.3336,
      "loss_align": 0.25697118043899536,
      "loss_contrastive": 0.12225842475891113,
      "loss_gen": 0.20941261947155,
      "neg_sim_mean": 0.7652872204780579,
      "pos_sim_mean": 0.7430288195610046,
      "step": 1668
    },
    {
      "epoch": 0.3338,
      "grad_norm": 2.2813966274261475,
      "learning_rate": 8.427821352667718e-05,
      "loss": 0.4811,
      "step": 1669
    },
    {
      "epoch": 0.3338,
      "loss_align": 0.12843090295791626,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19348666071891785,
      "neg_sim_mean": 0.7185935974121094,
      "pos_sim_mean": 0.8715690970420837,
      "step": 1669
    },
    {
      "epoch": 0.334,
      "grad_norm": 1.949415683746338,
      "learning_rate": 8.425279275875336e-05,
      "loss": 0.3219,
      "step": 1670
    },
    {
      "epoch": 0.334,
      "loss_align": 0.23430514335632324,
      "loss_contrastive": 0.1592092514038086,
      "loss_gen": 0.1719057261943817,
      "neg_sim_mean": 0.8249040842056274,
      "pos_sim_mean": 0.7656948566436768,
      "step": 1670
    },
    {
      "epoch": 0.3342,
      "grad_norm": 2.004873752593994,
      "learning_rate": 8.422735529643444e-05,
      "loss": 0.4253,
      "step": 1671
    },
    {
      "epoch": 0.3342,
      "loss_align": 0.3791787624359131,
      "loss_contrastive": 0.03860628604888916,
      "loss_gen": 0.06211487203836441,
      "neg_sim_mean": 0.5594274997711182,
      "pos_sim_mean": 0.6208212375640869,
      "step": 1671
    },
    {
      "epoch": 0.3344,
      "grad_norm": 1.7855333089828491,
      "learning_rate": 8.420190115211834e-05,
      "loss": 0.4459,
      "step": 1672
    },
    {
      "epoch": 0.3344,
      "loss_align": 0.17416101694107056,
      "loss_contrastive": 0.11130714416503906,
      "loss_gen": 0.1687707006931305,
      "neg_sim_mean": 0.8371461033821106,
      "pos_sim_mean": 0.8258389830589294,
      "step": 1672
    },
    {
      "epoch": 0.3346,
      "grad_norm": 1.4235848188400269,
      "learning_rate": 8.417643033821114e-05,
      "loss": 0.3563,
      "step": 1673
    },
    {
      "epoch": 0.3346,
      "loss_align": 0.23435884714126587,
      "loss_contrastive": 0.13849979639053345,
      "loss_gen": 0.22864314913749695,
      "neg_sim_mean": 0.8041409254074097,
      "pos_sim_mean": 0.7656411528587341,
      "step": 1673
    },
    {
      "epoch": 0.3348,
      "grad_norm": 1.9832888841629028,
      "learning_rate": 8.415094286712694e-05,
      "loss": 0.4796,
      "step": 1674
    },
    {
      "epoch": 0.3348,
      "loss_align": 0.15757369995117188,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04453305900096893,
      "neg_sim_mean": 0.3641779124736786,
      "pos_sim_mean": 0.8424263000488281,
      "step": 1674
    },
    {
      "epoch": 0.335,
      "grad_norm": 1.358351707458496,
      "learning_rate": 8.41254387512881e-05,
      "loss": 0.2021,
      "step": 1675
    },
    {
      "epoch": 0.335,
      "loss_align": 0.23282462358474731,
      "loss_contrastive": 0.11222583055496216,
      "loss_gen": 0.18443350493907928,
      "neg_sim_mean": 0.7794011831283569,
      "pos_sim_mean": 0.7671753764152527,
      "step": 1675
    },
    {
      "epoch": 0.3352,
      "grad_norm": 1.4677391052246094,
      "learning_rate": 8.409991800312493e-05,
      "loss": 0.4307,
      "step": 1676
    },
    {
      "epoch": 0.3352,
      "loss_align": 0.33193516731262207,
      "loss_contrastive": 0.09704601764678955,
      "loss_gen": 0.3608814775943756,
      "neg_sim_mean": 0.6651108264923096,
      "pos_sim_mean": 0.6680648326873779,
      "step": 1676
    },
    {
      "epoch": 0.3354,
      "grad_norm": 3.505178451538086,
      "learning_rate": 8.4074380635076e-05,
      "loss": 0.7045,
      "step": 1677
    },
    {
      "epoch": 0.3354,
      "loss_align": 0.14366543292999268,
      "loss_contrastive": 0.048953890800476074,
      "loss_gen": 0.43494129180908203,
      "neg_sim_mean": 0.8052884340286255,
      "pos_sim_mean": 0.8563345670700073,
      "step": 1677
    },
    {
      "epoch": 0.3356,
      "grad_norm": 2.170562505722046,
      "learning_rate": 8.404882665958787e-05,
      "loss": 0.5845,
      "step": 1678
    },
    {
      "epoch": 0.3356,
      "loss_align": 0.36562126874923706,
      "loss_contrastive": 0.106747567653656,
      "loss_gen": 0.2575227618217468,
      "neg_sim_mean": 0.641126275062561,
      "pos_sim_mean": 0.6343787312507629,
      "step": 1678
    },
    {
      "epoch": 0.3358,
      "grad_norm": 2.897366523742676,
      "learning_rate": 8.402325608911526e-05,
      "loss": 0.636,
      "step": 1679
    },
    {
      "epoch": 0.3358,
      "loss_align": 0.47097843885421753,
      "loss_contrastive": 0.23098114132881165,
      "loss_gen": 0.36641108989715576,
      "neg_sim_mean": 0.6600027084350586,
      "pos_sim_mean": 0.5290215611457825,
      "step": 1679
    },
    {
      "epoch": 0.336,
      "grad_norm": 2.505586862564087,
      "learning_rate": 8.399766893612096e-05,
      "loss": 0.8651,
      "step": 1680
    },
    {
      "epoch": 0.336,
      "loss_align": 0.15132761001586914,
      "loss_contrastive": 0.0912124514579773,
      "loss_gen": 0.1510869860649109,
      "neg_sim_mean": 0.8398848176002502,
      "pos_sim_mean": 0.8486723899841309,
      "step": 1680
    },
    {
      "epoch": 0.3362,
      "grad_norm": 1.7556222677230835,
      "learning_rate": 8.397206521307584e-05,
      "loss": 0.3134,
      "step": 1681
    },
    {
      "epoch": 0.3362,
      "loss_align": 0.19791090488433838,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013612918555736542,
      "neg_sim_mean": 0.3008832335472107,
      "pos_sim_mean": 0.8020890951156616,
      "step": 1681
    },
    {
      "epoch": 0.3364,
      "grad_norm": 0.787927508354187,
      "learning_rate": 8.394644493245881e-05,
      "loss": 0.2115,
      "step": 1682
    },
    {
      "epoch": 0.3364,
      "loss_align": 0.17495006322860718,
      "loss_contrastive": 0.08607584238052368,
      "loss_gen": 0.3675861954689026,
      "neg_sim_mean": 0.8111257553100586,
      "pos_sim_mean": 0.8250499367713928,
      "step": 1682
    },
    {
      "epoch": 0.3366,
      "grad_norm": 2.4180986881256104,
      "learning_rate": 8.392080810675691e-05,
      "loss": 0.5529,
      "step": 1683
    },
    {
      "epoch": 0.3366,
      "loss_align": 0.4062808156013489,
      "loss_contrastive": 0.18988260626792908,
      "loss_gen": 0.34694012999534607,
      "neg_sim_mean": 0.6836017966270447,
      "pos_sim_mean": 0.5937191843986511,
      "step": 1683
    },
    {
      "epoch": 0.3368,
      "grad_norm": 2.3738722801208496,
      "learning_rate": 8.389515474846522e-05,
      "loss": 0.776,
      "step": 1684
    },
    {
      "epoch": 0.3368,
      "loss_align": 0.28470706939697266,
      "loss_contrastive": 0.05041545629501343,
      "loss_gen": 0.1786472052335739,
      "neg_sim_mean": 0.6657083630561829,
      "pos_sim_mean": 0.7152929306030273,
      "step": 1684
    },
    {
      "epoch": 0.337,
      "grad_norm": 1.4233825206756592,
      "learning_rate": 8.386948487008686e-05,
      "loss": 0.4694,
      "step": 1685
    },
    {
      "epoch": 0.337,
      "loss_align": 0.17332780361175537,
      "loss_contrastive": 0.12844330072402954,
      "loss_gen": 0.3025255501270294,
      "neg_sim_mean": 0.8551154732704163,
      "pos_sim_mean": 0.8266721963882446,
      "step": 1685
    },
    {
      "epoch": 0.3372,
      "grad_norm": 2.0291922092437744,
      "learning_rate": 8.384379848413304e-05,
      "loss": 0.4913,
      "step": 1686
    },
    {
      "epoch": 0.3372,
      "loss_align": 0.20490562915802002,
      "loss_contrastive": 0.10635745525360107,
      "loss_gen": 0.30865779519081116,
      "neg_sim_mean": 0.8014518022537231,
      "pos_sim_mean": 0.79509437084198,
      "step": 1686
    },
    {
      "epoch": 0.3374,
      "grad_norm": 2.344273090362549,
      "learning_rate": 8.381809560312297e-05,
      "loss": 0.5263,
      "step": 1687
    },
    {
      "epoch": 0.3374,
      "loss_align": 0.1480022668838501,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26907142996788025,
      "neg_sim_mean": 0.5665907859802246,
      "pos_sim_mean": 0.8519977331161499,
      "step": 1687
    },
    {
      "epoch": 0.3376,
      "grad_norm": 1.7228893041610718,
      "learning_rate": 8.379237623958394e-05,
      "loss": 0.4171,
      "step": 1688
    },
    {
      "epoch": 0.3376,
      "loss_align": 0.21797782182693481,
      "loss_contrastive": 0.12403565645217896,
      "loss_gen": 0.5293768644332886,
      "neg_sim_mean": 0.8060578107833862,
      "pos_sim_mean": 0.7820221781730652,
      "step": 1688
    },
    {
      "epoch": 0.3378,
      "grad_norm": 2.8748090267181396,
      "learning_rate": 8.376664040605122e-05,
      "loss": 0.7622,
      "step": 1689
    },
    {
      "epoch": 0.3378,
      "loss_align": 0.16195225715637207,
      "loss_contrastive": 0.03308504819869995,
      "loss_gen": 0.3217257559299469,
      "neg_sim_mean": 0.77113276720047,
      "pos_sim_mean": 0.8380477428436279,
      "step": 1689
    },
    {
      "epoch": 0.338,
      "grad_norm": 1.819718837738037,
      "learning_rate": 8.374088811506819e-05,
      "loss": 0.4876,
      "step": 1690
    },
    {
      "epoch": 0.338,
      "loss_align": 0.20320212841033936,
      "loss_contrastive": 0.0652322769165039,
      "loss_gen": 0.20692062377929688,
      "neg_sim_mean": 0.7620301246643066,
      "pos_sim_mean": 0.7967978715896606,
      "step": 1690
    },
    {
      "epoch": 0.3382,
      "grad_norm": 2.207883358001709,
      "learning_rate": 8.371511937918616e-05,
      "loss": 0.418,
      "step": 1691
    },
    {
      "epoch": 0.3382,
      "loss_align": 0.23110324144363403,
      "loss_contrastive": 0.0,
      "loss_gen": 0.027234317734837532,
      "neg_sim_mean": 0.3318008482456207,
      "pos_sim_mean": 0.768896758556366,
      "step": 1691
    },
    {
      "epoch": 0.3384,
      "grad_norm": 1.2985212802886963,
      "learning_rate": 8.368933421096454e-05,
      "loss": 0.2583,
      "step": 1692
    },
    {
      "epoch": 0.3384,
      "loss_align": 0.2524617910385132,
      "loss_contrastive": 0.020935535430908203,
      "loss_gen": 0.32377004623413086,
      "neg_sim_mean": 0.6684737205505371,
      "pos_sim_mean": 0.7475382089614868,
      "step": 1692
    },
    {
      "epoch": 0.3386,
      "grad_norm": 2.4240875244140625,
      "learning_rate": 8.366353262297069e-05,
      "loss": 0.5787,
      "step": 1693
    },
    {
      "epoch": 0.3386,
      "loss_align": 0.23286211490631104,
      "loss_contrastive": 0.06903409957885742,
      "loss_gen": 0.39688968658447266,
      "neg_sim_mean": 0.7361719608306885,
      "pos_sim_mean": 0.767137885093689,
      "step": 1693
    },
    {
      "epoch": 0.3388,
      "grad_norm": 2.322878360748291,
      "learning_rate": 8.363771462777998e-05,
      "loss": 0.638,
      "step": 1694
    },
    {
      "epoch": 0.3388,
      "loss_align": 0.22201716899871826,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009280172176659107,
      "neg_sim_mean": 0.37466633319854736,
      "pos_sim_mean": 0.7779828310012817,
      "step": 1694
    },
    {
      "epoch": 0.339,
      "grad_norm": 0.7655966877937317,
      "learning_rate": 8.361188023797582e-05,
      "loss": 0.2313,
      "step": 1695
    },
    {
      "epoch": 0.339,
      "loss_align": 0.29700732231140137,
      "loss_contrastive": 0.1397719383239746,
      "loss_gen": 0.16570407152175903,
      "neg_sim_mean": 0.7427645921707153,
      "pos_sim_mean": 0.7029926776885986,
      "step": 1695
    },
    {
      "epoch": 0.3392,
      "grad_norm": 1.5709035396575928,
      "learning_rate": 8.358602946614951e-05,
      "loss": 0.4795,
      "step": 1696
    },
    {
      "epoch": 0.3392,
      "loss_align": 0.15546774864196777,
      "loss_contrastive": 0.08377963304519653,
      "loss_gen": 0.26709264516830444,
      "neg_sim_mean": 0.8283118605613708,
      "pos_sim_mean": 0.8445322513580322,
      "step": 1696
    },
    {
      "epoch": 0.3394,
      "grad_norm": 1.9281715154647827,
      "learning_rate": 8.356016232490048e-05,
      "loss": 0.4326,
      "step": 1697
    },
    {
      "epoch": 0.3394,
      "loss_align": 0.3205884099006653,
      "loss_contrastive": 0.1517435908317566,
      "loss_gen": 0.2843138873577118,
      "neg_sim_mean": 0.7311551570892334,
      "pos_sim_mean": 0.6794115900993347,
      "step": 1697
    },
    {
      "epoch": 0.3396,
      "grad_norm": 2.1055173873901367,
      "learning_rate": 8.3534278826836e-05,
      "loss": 0.6231,
      "step": 1698
    },
    {
      "epoch": 0.3396,
      "loss_align": 0.08356094360351562,
      "loss_contrastive": 0.0,
      "loss_gen": 0.35693320631980896,
      "neg_sim_mean": 0.7930681705474854,
      "pos_sim_mean": 0.9164390563964844,
      "step": 1698
    },
    {
      "epoch": 0.3398,
      "grad_norm": 2.316502094268799,
      "learning_rate": 8.350837898457143e-05,
      "loss": 0.4405,
      "step": 1699
    },
    {
      "epoch": 0.3398,
      "loss_align": 0.300518274307251,
      "loss_contrastive": 0.1279754638671875,
      "loss_gen": 0.3309910297393799,
      "neg_sim_mean": 0.7274571657180786,
      "pos_sim_mean": 0.699481725692749,
      "step": 1699
    },
    {
      "epoch": 0.34,
      "grad_norm": 2.059562921524048,
      "learning_rate": 8.348246281072998e-05,
      "loss": 0.6469,
      "step": 1700
    },
    {
      "epoch": 0.34,
      "loss_align": 0.3128893971443176,
      "loss_contrastive": 0.030765235424041748,
      "loss_gen": 0.2683083415031433,
      "neg_sim_mean": 0.6178758144378662,
      "pos_sim_mean": 0.6871106028556824,
      "step": 1700
    },
    {
      "epoch": 0.3402,
      "grad_norm": 2.0488905906677246,
      "learning_rate": 8.345653031794292e-05,
      "loss": 0.5849,
      "step": 1701
    },
    {
      "epoch": 0.3402,
      "loss_align": 0.21120047569274902,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001540881348773837,
      "neg_sim_mean": 0.4159407913684845,
      "pos_sim_mean": 0.788799524307251,
      "step": 1701
    },
    {
      "epoch": 0.3404,
      "grad_norm": 0.4234159290790558,
      "learning_rate": 8.343058151884941e-05,
      "loss": 0.2127,
      "step": 1702
    },
    {
      "epoch": 0.3404,
      "loss_align": 0.1864413022994995,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00832067895680666,
      "neg_sim_mean": 0.5331451892852783,
      "pos_sim_mean": 0.8135586977005005,
      "step": 1702
    },
    {
      "epoch": 0.3406,
      "grad_norm": 0.6919605731964111,
      "learning_rate": 8.340461642609658e-05,
      "loss": 0.1948,
      "step": 1703
    },
    {
      "epoch": 0.3406,
      "loss_align": 0.17806464433670044,
      "loss_contrastive": 0.01813817024230957,
      "loss_gen": 0.15858516097068787,
      "neg_sim_mean": 0.7400735020637512,
      "pos_sim_mean": 0.8219353556632996,
      "step": 1703
    },
    {
      "epoch": 0.3408,
      "grad_norm": 1.5325477123260498,
      "learning_rate": 8.337863505233953e-05,
      "loss": 0.3388,
      "step": 1704
    },
    {
      "epoch": 0.3408,
      "loss_align": 0.23164403438568115,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005819633137434721,
      "neg_sim_mean": 0.574273943901062,
      "pos_sim_mean": 0.7683559656143188,
      "step": 1704
    },
    {
      "epoch": 0.341,
      "grad_norm": 0.7408190369606018,
      "learning_rate": 8.335263741024123e-05,
      "loss": 0.2375,
      "step": 1705
    },
    {
      "epoch": 0.341,
      "loss_align": 0.28358030319213867,
      "loss_contrastive": 0.029064834117889404,
      "loss_gen": 0.4382578134536743,
      "neg_sim_mean": 0.6454845070838928,
      "pos_sim_mean": 0.7164196968078613,
      "step": 1705
    },
    {
      "epoch": 0.3412,
      "grad_norm": 2.6875369548797607,
      "learning_rate": 8.332662351247262e-05,
      "loss": 0.7253,
      "step": 1706
    },
    {
      "epoch": 0.3412,
      "loss_align": 0.1837902069091797,
      "loss_contrastive": 0.1139177680015564,
      "loss_gen": 0.5075876116752625,
      "neg_sim_mean": 0.8301275372505188,
      "pos_sim_mean": 0.8162097930908203,
      "step": 1706
    },
    {
      "epoch": 0.3414,
      "grad_norm": 2.5496599674224854,
      "learning_rate": 8.330059337171258e-05,
      "loss": 0.705,
      "step": 1707
    },
    {
      "epoch": 0.3414,
      "loss_align": 0.26378440856933594,
      "loss_contrastive": 0.19351810216903687,
      "loss_gen": 0.24141483008861542,
      "neg_sim_mean": 0.829733669757843,
      "pos_sim_mean": 0.7362155914306641,
      "step": 1707
    },
    {
      "epoch": 0.3416,
      "grad_norm": 1.7498637437820435,
      "learning_rate": 8.327454700064788e-05,
      "loss": 0.5284,
      "step": 1708
    },
    {
      "epoch": 0.3416,
      "loss_align": 0.36614930629730225,
      "loss_contrastive": 0.13649141788482666,
      "loss_gen": 0.3053963780403137,
      "neg_sim_mean": 0.6703420877456665,
      "pos_sim_mean": 0.6338506937026978,
      "step": 1708
    },
    {
      "epoch": 0.3418,
      "grad_norm": 2.06306791305542,
      "learning_rate": 8.324848441197317e-05,
      "loss": 0.6879,
      "step": 1709
    },
    {
      "epoch": 0.3418,
      "loss_align": 0.2194165587425232,
      "loss_contrastive": 0.0686231255531311,
      "loss_gen": 0.5644789934158325,
      "neg_sim_mean": 0.74920654296875,
      "pos_sim_mean": 0.7805834412574768,
      "step": 1709
    },
    {
      "epoch": 0.342,
      "grad_norm": 2.652616262435913,
      "learning_rate": 8.322240561839109e-05,
      "loss": 0.7921,
      "step": 1710
    },
    {
      "epoch": 0.342,
      "loss_align": 0.23415321111679077,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22451449930667877,
      "neg_sim_mean": 0.6261646747589111,
      "pos_sim_mean": 0.7658467888832092,
      "step": 1710
    },
    {
      "epoch": 0.3422,
      "grad_norm": 1.8892347812652588,
      "learning_rate": 8.319631063261209e-05,
      "loss": 0.4587,
      "step": 1711
    },
    {
      "epoch": 0.3422,
      "loss_align": 0.2397049069404602,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007290488108992577,
      "neg_sim_mean": 0.28105711936950684,
      "pos_sim_mean": 0.7602950930595398,
      "step": 1711
    },
    {
      "epoch": 0.3424,
      "grad_norm": 0.5921566486358643,
      "learning_rate": 8.317019946735456e-05,
      "loss": 0.247,
      "step": 1712
    },
    {
      "epoch": 0.3424,
      "loss_align": 0.07420212030410767,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00356791028752923,
      "neg_sim_mean": 0.3844091594219208,
      "pos_sim_mean": 0.9257978796958923,
      "step": 1712
    },
    {
      "epoch": 0.3426,
      "grad_norm": 0.3202042281627655,
      "learning_rate": 8.314407213534476e-05,
      "loss": 0.0778,
      "step": 1713
    },
    {
      "epoch": 0.3426,
      "loss_align": 0.3260929584503174,
      "loss_contrastive": 0.08419209718704224,
      "loss_gen": 0.11170049756765366,
      "neg_sim_mean": 0.6580991148948669,
      "pos_sim_mean": 0.6739070415496826,
      "step": 1713
    },
    {
      "epoch": 0.3428,
      "grad_norm": 1.4560970067977905,
      "learning_rate": 8.311792864931684e-05,
      "loss": 0.4479,
      "step": 1714
    },
    {
      "epoch": 0.3428,
      "loss_align": 0.3363359570503235,
      "loss_contrastive": 0.17744213342666626,
      "loss_gen": 0.3225090503692627,
      "neg_sim_mean": 0.7411061525344849,
      "pos_sim_mean": 0.6636640429496765,
      "step": 1714
    },
    {
      "epoch": 0.343,
      "grad_norm": 2.2290709018707275,
      "learning_rate": 8.309176902201284e-05,
      "loss": 0.6801,
      "step": 1715
    },
    {
      "epoch": 0.343,
      "loss_align": 0.13210725784301758,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01361076533794403,
      "neg_sim_mean": 0.3266260623931885,
      "pos_sim_mean": 0.8678927421569824,
      "step": 1715
    },
    {
      "epoch": 0.3432,
      "grad_norm": 0.8566368818283081,
      "learning_rate": 8.306559326618259e-05,
      "loss": 0.1457,
      "step": 1716
    },
    {
      "epoch": 0.3432,
      "loss_align": 0.10304915904998779,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19094467163085938,
      "neg_sim_mean": 0.7579123973846436,
      "pos_sim_mean": 0.8969508409500122,
      "step": 1716
    },
    {
      "epoch": 0.3434,
      "grad_norm": 1.7353371381759644,
      "learning_rate": 8.303940139458389e-05,
      "loss": 0.294,
      "step": 1717
    },
    {
      "epoch": 0.3434,
      "loss_align": 0.15244781970977783,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22907663881778717,
      "neg_sim_mean": 0.6614803075790405,
      "pos_sim_mean": 0.8475521802902222,
      "step": 1717
    },
    {
      "epoch": 0.3436,
      "grad_norm": 1.7220642566680908,
      "learning_rate": 8.301319341998231e-05,
      "loss": 0.3815,
      "step": 1718
    },
    {
      "epoch": 0.3436,
      "loss_align": 0.21955764293670654,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0076186563819646835,
      "neg_sim_mean": 0.3581264615058899,
      "pos_sim_mean": 0.7804423570632935,
      "step": 1718
    },
    {
      "epoch": 0.3438,
      "grad_norm": 0.671852707862854,
      "learning_rate": 8.298696935515132e-05,
      "loss": 0.2272,
      "step": 1719
    },
    {
      "epoch": 0.3438,
      "loss_align": 0.13739442825317383,
      "loss_contrastive": 0.05342578887939453,
      "loss_gen": 0.11818448454141617,
      "neg_sim_mean": 0.8160313367843628,
      "pos_sim_mean": 0.8626055717468262,
      "step": 1719
    },
    {
      "epoch": 0.344,
      "grad_norm": 1.4108328819274902,
      "learning_rate": 8.296072921287217e-05,
      "loss": 0.262,
      "step": 1720
    },
    {
      "epoch": 0.344,
      "loss_align": 0.2877662181854248,
      "loss_contrastive": 0.06994736194610596,
      "loss_gen": 0.25378233194351196,
      "neg_sim_mean": 0.6821811199188232,
      "pos_sim_mean": 0.7122337818145752,
      "step": 1720
    },
    {
      "epoch": 0.3442,
      "grad_norm": 2.0601184368133545,
      "learning_rate": 8.293447300593402e-05,
      "loss": 0.5499,
      "step": 1721
    },
    {
      "epoch": 0.3442,
      "loss_align": 0.15753424167633057,
      "loss_contrastive": 0.06622493267059326,
      "loss_gen": 0.2506241798400879,
      "neg_sim_mean": 0.8086906671524048,
      "pos_sim_mean": 0.8424657583236694,
      "step": 1721
    },
    {
      "epoch": 0.3444,
      "grad_norm": 1.7420716285705566,
      "learning_rate": 8.290820074713384e-05,
      "loss": 0.4161,
      "step": 1722
    },
    {
      "epoch": 0.3444,
      "loss_align": 0.2730889320373535,
      "loss_contrastive": 0.0,
      "loss_gen": 0.019403710961341858,
      "neg_sim_mean": 0.5326687693595886,
      "pos_sim_mean": 0.7269110679626465,
      "step": 1722
    },
    {
      "epoch": 0.3446,
      "grad_norm": 1.5533586740493774,
      "learning_rate": 8.288191244927637e-05,
      "loss": 0.2925,
      "step": 1723
    },
    {
      "epoch": 0.3446,
      "loss_align": 0.21051651239395142,
      "loss_contrastive": 0.1232488751411438,
      "loss_gen": 0.38328343629837036,
      "neg_sim_mean": 0.8127323389053345,
      "pos_sim_mean": 0.7894834876060486,
      "step": 1723
    },
    {
      "epoch": 0.3448,
      "grad_norm": 2.312868595123291,
      "learning_rate": 8.285560812517423e-05,
      "loss": 0.6086,
      "step": 1724
    },
    {
      "epoch": 0.3448,
      "loss_align": 0.3996846079826355,
      "loss_contrastive": 0.17054146528244019,
      "loss_gen": 0.17383411526679993,
      "neg_sim_mean": 0.6708568334579468,
      "pos_sim_mean": 0.6003153920173645,
      "step": 1724
    },
    {
      "epoch": 0.345,
      "grad_norm": 1.8350250720977783,
      "learning_rate": 8.282928778764783e-05,
      "loss": 0.594,
      "step": 1725
    },
    {
      "epoch": 0.345,
      "loss_align": 0.18427777290344238,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0017995184753090143,
      "neg_sim_mean": 0.4569210410118103,
      "pos_sim_mean": 0.8157222270965576,
      "step": 1725
    },
    {
      "epoch": 0.3452,
      "grad_norm": 0.5038022994995117,
      "learning_rate": 8.280295144952536e-05,
      "loss": 0.1861,
      "step": 1726
    },
    {
      "epoch": 0.3452,
      "loss_align": 0.2540156841278076,
      "loss_contrastive": 0.13994866609573364,
      "loss_gen": 0.173132061958313,
      "neg_sim_mean": 0.7859329581260681,
      "pos_sim_mean": 0.7459843158721924,
      "step": 1726
    },
    {
      "epoch": 0.3454,
      "grad_norm": 1.7030261754989624,
      "learning_rate": 8.277659912364288e-05,
      "loss": 0.4439,
      "step": 1727
    },
    {
      "epoch": 0.3454,
      "loss_align": 0.32836246490478516,
      "loss_contrastive": 0.19141751527786255,
      "loss_gen": 0.33386197686195374,
      "neg_sim_mean": 0.7630550265312195,
      "pos_sim_mean": 0.6716375350952148,
      "step": 1727
    },
    {
      "epoch": 0.3456,
      "grad_norm": 2.4772846698760986,
      "learning_rate": 8.275023082284413e-05,
      "loss": 0.6852,
      "step": 1728
    },
    {
      "epoch": 0.3456,
      "loss_align": 0.21009236574172974,
      "loss_contrastive": 0.11418485641479492,
      "loss_gen": 0.2844017446041107,
      "neg_sim_mean": 0.8040924668312073,
      "pos_sim_mean": 0.7899076342582703,
      "step": 1728
    },
    {
      "epoch": 0.3458,
      "grad_norm": 1.9413261413574219,
      "learning_rate": 8.272384655998074e-05,
      "loss": 0.5082,
      "step": 1729
    },
    {
      "epoch": 0.3458,
      "loss_align": 0.29530370235443115,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006616082973778248,
      "neg_sim_mean": 0.34991151094436646,
      "pos_sim_mean": 0.7046962976455688,
      "step": 1729
    },
    {
      "epoch": 0.346,
      "grad_norm": 0.679625928401947,
      "learning_rate": 8.269744634791208e-05,
      "loss": 0.3019,
      "step": 1730
    },
    {
      "epoch": 0.346,
      "loss_align": 0.07145845890045166,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003840592224150896,
      "neg_sim_mean": 0.362863153219223,
      "pos_sim_mean": 0.9285415410995483,
      "step": 1730
    },
    {
      "epoch": 0.3462,
      "grad_norm": 0.34265878796577454,
      "learning_rate": 8.267103019950529e-05,
      "loss": 0.0753,
      "step": 1731
    },
    {
      "epoch": 0.3462,
      "loss_align": 0.4025266766548157,
      "loss_contrastive": 0.13778463006019592,
      "loss_gen": 0.3131106197834015,
      "neg_sim_mean": 0.6352579593658447,
      "pos_sim_mean": 0.5974733233451843,
      "step": 1731
    },
    {
      "epoch": 0.3464,
      "grad_norm": 2.166626453399658,
      "learning_rate": 8.264459812763525e-05,
      "loss": 0.7322,
      "step": 1732
    },
    {
      "epoch": 0.3464,
      "loss_align": 0.2027323842048645,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0017190705984830856,
      "neg_sim_mean": 0.39789414405822754,
      "pos_sim_mean": 0.7972676157951355,
      "step": 1732
    },
    {
      "epoch": 0.3466,
      "grad_norm": 0.42046767473220825,
      "learning_rate": 8.261815014518466e-05,
      "loss": 0.2045,
      "step": 1733
    },
    {
      "epoch": 0.3466,
      "loss_align": 0.17956042289733887,
      "loss_contrastive": 0.0,
      "loss_gen": 0.014836400747299194,
      "neg_sim_mean": 0.30781832337379456,
      "pos_sim_mean": 0.8204395771026611,
      "step": 1733
    },
    {
      "epoch": 0.3468,
      "grad_norm": 0.8871029019355774,
      "learning_rate": 8.259168626504395e-05,
      "loss": 0.1944,
      "step": 1734
    },
    {
      "epoch": 0.3468,
      "loss_align": 0.2716383934020996,
      "loss_contrastive": 0.10273027420043945,
      "loss_gen": 0.32234644889831543,
      "neg_sim_mean": 0.7310918569564819,
      "pos_sim_mean": 0.7283616065979004,
      "step": 1734
    },
    {
      "epoch": 0.347,
      "grad_norm": 2.416919231414795,
      "learning_rate": 8.256520650011127e-05,
      "loss": 0.6063,
      "step": 1735
    },
    {
      "epoch": 0.347,
      "loss_align": 0.21061837673187256,
      "loss_contrastive": 0.1171189546585083,
      "loss_gen": 0.34973084926605225,
      "neg_sim_mean": 0.8065005540847778,
      "pos_sim_mean": 0.7893816232681274,
      "step": 1735
    },
    {
      "epoch": 0.3472,
      "grad_norm": 2.603844404220581,
      "learning_rate": 8.253871086329255e-05,
      "loss": 0.5744,
      "step": 1736
    },
    {
      "epoch": 0.3472,
      "loss_align": 0.2972477674484253,
      "loss_contrastive": 0.1320110559463501,
      "loss_gen": 0.18364977836608887,
      "neg_sim_mean": 0.7347632646560669,
      "pos_sim_mean": 0.7027522325515747,
      "step": 1736
    },
    {
      "epoch": 0.3474,
      "grad_norm": 1.7044867277145386,
      "learning_rate": 8.251219936750144e-05,
      "loss": 0.4967,
      "step": 1737
    },
    {
      "epoch": 0.3474,
      "loss_align": 0.12882041931152344,
      "loss_contrastive": 0.013788104057312012,
      "loss_gen": 0.1753273904323578,
      "neg_sim_mean": 0.7849676609039307,
      "pos_sim_mean": 0.8711795806884766,
      "step": 1737
    },
    {
      "epoch": 0.3476,
      "grad_norm": 1.4525471925735474,
      "learning_rate": 8.248567202565933e-05,
      "loss": 0.3058,
      "step": 1738
    },
    {
      "epoch": 0.3476,
      "loss_align": 0.36730343103408813,
      "loss_contrastive": 0.08827811479568481,
      "loss_gen": 0.1733429878950119,
      "neg_sim_mean": 0.6209746599197388,
      "pos_sim_mean": 0.6326965689659119,
      "step": 1738
    },
    {
      "epoch": 0.3478,
      "grad_norm": 1.8194156885147095,
      "learning_rate": 8.245912885069531e-05,
      "loss": 0.5512,
      "step": 1739
    },
    {
      "epoch": 0.3478,
      "loss_align": 0.33159971237182617,
      "loss_contrastive": 0.0814676284790039,
      "loss_gen": 0.29396843910217285,
      "neg_sim_mean": 0.6498678922653198,
      "pos_sim_mean": 0.6684002876281738,
      "step": 1739
    },
    {
      "epoch": 0.348,
      "grad_norm": 2.403895139694214,
      "learning_rate": 8.243256985554621e-05,
      "loss": 0.6353,
      "step": 1740
    },
    {
      "epoch": 0.348,
      "loss_align": 0.23057186603546143,
      "loss_contrastive": 0.08601391315460205,
      "loss_gen": 0.27147820591926575,
      "neg_sim_mean": 0.7554420232772827,
      "pos_sim_mean": 0.7694281339645386,
      "step": 1740
    },
    {
      "epoch": 0.3482,
      "grad_norm": 2.231764078140259,
      "learning_rate": 8.240599505315655e-05,
      "loss": 0.5124,
      "step": 1741
    },
    {
      "epoch": 0.3482,
      "loss_align": 0.07844662666320801,
      "loss_contrastive": 0.06537675857543945,
      "loss_gen": 0.24244952201843262,
      "neg_sim_mean": 0.8869301080703735,
      "pos_sim_mean": 0.921553373336792,
      "step": 1741
    },
    {
      "epoch": 0.3484,
      "grad_norm": 2.041598320007324,
      "learning_rate": 8.237940445647858e-05,
      "loss": 0.3287,
      "step": 1742
    },
    {
      "epoch": 0.3484,
      "loss_align": 0.2646671533584595,
      "loss_contrastive": 0.15261781215667725,
      "loss_gen": 0.15673980116844177,
      "neg_sim_mean": 0.7879506349563599,
      "pos_sim_mean": 0.7353328466415405,
      "step": 1742
    },
    {
      "epoch": 0.3486,
      "grad_norm": 1.4071375131607056,
      "learning_rate": 8.235279807847223e-05,
      "loss": 0.4397,
      "step": 1743
    },
    {
      "epoch": 0.3486,
      "loss_align": 0.17269957065582275,
      "loss_contrastive": 0.06412434577941895,
      "loss_gen": 0.2875841557979584,
      "neg_sim_mean": 0.7914247512817383,
      "pos_sim_mean": 0.8273004293441772,
      "step": 1743
    },
    {
      "epoch": 0.3488,
      "grad_norm": 1.848241925239563,
      "learning_rate": 8.23261759321051e-05,
      "loss": 0.468,
      "step": 1744
    },
    {
      "epoch": 0.3488,
      "loss_align": 0.1331080198287964,
      "loss_contrastive": 0.03799450397491455,
      "loss_gen": 0.41274651885032654,
      "neg_sim_mean": 0.8048864603042603,
      "pos_sim_mean": 0.8668919801712036,
      "step": 1744
    },
    {
      "epoch": 0.349,
      "grad_norm": 2.331432580947876,
      "learning_rate": 8.229953803035255e-05,
      "loss": 0.5504,
      "step": 1745
    },
    {
      "epoch": 0.349,
      "loss_align": 0.23299318552017212,
      "loss_contrastive": 0.016846179962158203,
      "loss_gen": 0.2584165334701538,
      "neg_sim_mean": 0.6838529706001282,
      "pos_sim_mean": 0.7670068144798279,
      "step": 1745
    },
    {
      "epoch": 0.3492,
      "grad_norm": 2.129589080810547,
      "learning_rate": 8.227288438619754e-05,
      "loss": 0.4934,
      "step": 1746
    },
    {
      "epoch": 0.3492,
      "loss_align": 0.18663561344146729,
      "loss_contrastive": 0.11576998233795166,
      "loss_gen": 0.2045169174671173,
      "neg_sim_mean": 0.8291343450546265,
      "pos_sim_mean": 0.8133643865585327,
      "step": 1746
    },
    {
      "epoch": 0.3494,
      "grad_norm": 1.9781272411346436,
      "learning_rate": 8.224621501263072e-05,
      "loss": 0.405,
      "step": 1747
    },
    {
      "epoch": 0.3494,
      "loss_align": 0.2297821044921875,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27456340193748474,
      "neg_sim_mean": 0.4578413963317871,
      "pos_sim_mean": 0.7702178955078125,
      "step": 1747
    },
    {
      "epoch": 0.3496,
      "grad_norm": 2.2503058910369873,
      "learning_rate": 8.221952992265045e-05,
      "loss": 0.5043,
      "step": 1748
    },
    {
      "epoch": 0.3496,
      "loss_align": 0.21953845024108887,
      "loss_contrastive": 0.14461749792099,
      "loss_gen": 0.6478904485702515,
      "neg_sim_mean": 0.8250790238380432,
      "pos_sim_mean": 0.7804615497589111,
      "step": 1748
    },
    {
      "epoch": 0.3498,
      "grad_norm": 2.8464322090148926,
      "learning_rate": 8.21928291292627e-05,
      "loss": 0.8848,
      "step": 1749
    },
    {
      "epoch": 0.3498,
      "loss_align": 0.3300018310546875,
      "loss_contrastive": 0.09601724147796631,
      "loss_gen": 0.1577930599451065,
      "neg_sim_mean": 0.6660153865814209,
      "pos_sim_mean": 0.6699981689453125,
      "step": 1749
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.5639296770095825,
      "learning_rate": 8.21661126454811e-05,
      "loss": 0.4993,
      "step": 1750
    },
    {
      "epoch": 0.35,
      "loss_align": 0.18308734893798828,
      "loss_contrastive": 0.06000465154647827,
      "loss_gen": 0.26323458552360535,
      "neg_sim_mean": 0.7769172787666321,
      "pos_sim_mean": 0.8169126510620117,
      "step": 1750
    },
    {
      "epoch": 0.3502,
      "grad_norm": 2.0695297718048096,
      "learning_rate": 8.213938048432697e-05,
      "loss": 0.4535,
      "step": 1751
    },
    {
      "epoch": 0.3502,
      "loss_align": 0.14673089981079102,
      "loss_contrastive": 0.11737620830535889,
      "loss_gen": 0.4133201539516449,
      "neg_sim_mean": 0.87064528465271,
      "pos_sim_mean": 0.853269100189209,
      "step": 1751
    },
    {
      "epoch": 0.3504,
      "grad_norm": 2.233980894088745,
      "learning_rate": 8.211263265882923e-05,
      "loss": 0.5741,
      "step": 1752
    },
    {
      "epoch": 0.3504,
      "loss_align": 0.36981451511383057,
      "loss_contrastive": 0.1329372525215149,
      "loss_gen": 0.2544061541557312,
      "neg_sim_mean": 0.6631227135658264,
      "pos_sim_mean": 0.6301854848861694,
      "step": 1752
    },
    {
      "epoch": 0.3506,
      "grad_norm": 1.9609172344207764,
      "learning_rate": 8.208586918202443e-05,
      "loss": 0.6402,
      "step": 1753
    },
    {
      "epoch": 0.3506,
      "loss_align": 0.19207000732421875,
      "loss_contrastive": 0.12220346927642822,
      "loss_gen": 0.3782358467578888,
      "neg_sim_mean": 0.8301334381103516,
      "pos_sim_mean": 0.8079299926757812,
      "step": 1753
    },
    {
      "epoch": 0.3508,
      "grad_norm": 2.17154860496521,
      "learning_rate": 8.205909006695678e-05,
      "loss": 0.585,
      "step": 1754
    },
    {
      "epoch": 0.3508,
      "loss_align": 0.30465930700302124,
      "loss_contrastive": 0.0,
      "loss_gen": 0.33363375067710876,
      "neg_sim_mean": 0.5717660188674927,
      "pos_sim_mean": 0.6953406929969788,
      "step": 1754
    },
    {
      "epoch": 0.351,
      "grad_norm": 2.132749557495117,
      "learning_rate": 8.203229532667807e-05,
      "loss": 0.6383,
      "step": 1755
    },
    {
      "epoch": 0.351,
      "loss_align": 0.273978590965271,
      "loss_contrastive": 0.0470120906829834,
      "loss_gen": 0.23622433841228485,
      "neg_sim_mean": 0.6730334758758545,
      "pos_sim_mean": 0.726021409034729,
      "step": 1755
    },
    {
      "epoch": 0.3512,
      "grad_norm": 1.9350337982177734,
      "learning_rate": 8.200548497424778e-05,
      "loss": 0.5158,
      "step": 1756
    },
    {
      "epoch": 0.3512,
      "loss_align": 0.283454954624176,
      "loss_contrastive": 0.02218639850616455,
      "loss_gen": 0.1298452615737915,
      "neg_sim_mean": 0.6387314200401306,
      "pos_sim_mean": 0.716545045375824,
      "step": 1756
    },
    {
      "epoch": 0.3514,
      "grad_norm": 1.6291335821151733,
      "learning_rate": 8.19786590227329e-05,
      "loss": 0.416,
      "step": 1757
    },
    {
      "epoch": 0.3514,
      "loss_align": 0.15823894739151,
      "loss_contrastive": 0.07914656400680542,
      "loss_gen": 0.2809307873249054,
      "neg_sim_mean": 0.8209075927734375,
      "pos_sim_mean": 0.84176105260849,
      "step": 1757
    },
    {
      "epoch": 0.3516,
      "grad_norm": 2.0440523624420166,
      "learning_rate": 8.195181748520811e-05,
      "loss": 0.4487,
      "step": 1758
    },
    {
      "epoch": 0.3516,
      "loss_align": 0.23777639865875244,
      "loss_contrastive": 0.028033196926116943,
      "loss_gen": 0.14863324165344238,
      "neg_sim_mean": 0.6902567744255066,
      "pos_sim_mean": 0.7622236013412476,
      "step": 1758
    },
    {
      "epoch": 0.3518,
      "grad_norm": 1.754736304283142,
      "learning_rate": 8.192496037475562e-05,
      "loss": 0.3898,
      "step": 1759
    },
    {
      "epoch": 0.3518,
      "loss_align": 0.17399299144744873,
      "loss_contrastive": 0.077292799949646,
      "loss_gen": 0.18937809765338898,
      "neg_sim_mean": 0.8032997846603394,
      "pos_sim_mean": 0.8260070085525513,
      "step": 1759
    },
    {
      "epoch": 0.352,
      "grad_norm": 1.5587314367294312,
      "learning_rate": 8.189808770446528e-05,
      "loss": 0.3726,
      "step": 1760
    },
    {
      "epoch": 0.352,
      "loss_align": 0.2624548673629761,
      "loss_contrastive": 0.10527420043945312,
      "loss_gen": 0.28261932730674744,
      "neg_sim_mean": 0.7428193092346191,
      "pos_sim_mean": 0.7375451326370239,
      "step": 1760
    },
    {
      "epoch": 0.3522,
      "grad_norm": 1.914603352546692,
      "learning_rate": 8.18711994874345e-05,
      "loss": 0.5577,
      "step": 1761
    },
    {
      "epoch": 0.3522,
      "loss_align": 0.26837700605392456,
      "loss_contrastive": 0.12499147653579712,
      "loss_gen": 0.2639131546020508,
      "neg_sim_mean": 0.7566144466400146,
      "pos_sim_mean": 0.7316229939460754,
      "step": 1761
    },
    {
      "epoch": 0.3524,
      "grad_norm": 1.7445164918899536,
      "learning_rate": 8.184429573676826e-05,
      "loss": 0.5473,
      "step": 1762
    },
    {
      "epoch": 0.3524,
      "loss_align": 0.20751571655273438,
      "loss_contrastive": 0.09292656183242798,
      "loss_gen": 0.11559439450502396,
      "neg_sim_mean": 0.7854108214378357,
      "pos_sim_mean": 0.7924842834472656,
      "step": 1762
    },
    {
      "epoch": 0.3526,
      "grad_norm": 1.3454012870788574,
      "learning_rate": 8.181737646557912e-05,
      "loss": 0.3343,
      "step": 1763
    },
    {
      "epoch": 0.3526,
      "loss_align": 0.23908227682113647,
      "loss_contrastive": 0.12240660190582275,
      "loss_gen": 0.5020692944526672,
      "neg_sim_mean": 0.7833243012428284,
      "pos_sim_mean": 0.7609177231788635,
      "step": 1763
    },
    {
      "epoch": 0.3528,
      "grad_norm": 3.1825993061065674,
      "learning_rate": 8.179044168698721e-05,
      "loss": 0.7558,
      "step": 1764
    },
    {
      "epoch": 0.3528,
      "loss_align": 0.3288170099258423,
      "loss_contrastive": 0.10538214445114136,
      "loss_gen": 0.13170969486236572,
      "neg_sim_mean": 0.6765651106834412,
      "pos_sim_mean": 0.6711829900741577,
      "step": 1764
    },
    {
      "epoch": 0.353,
      "grad_norm": 1.6186189651489258,
      "learning_rate": 8.176349141412022e-05,
      "loss": 0.4732,
      "step": 1765
    },
    {
      "epoch": 0.353,
      "loss_align": 0.06025135517120361,
      "loss_contrastive": 0.0,
      "loss_gen": 0.06451407819986343,
      "neg_sim_mean": 0.7595001459121704,
      "pos_sim_mean": 0.9397486448287964,
      "step": 1765
    },
    {
      "epoch": 0.3532,
      "grad_norm": 1.1423856019973755,
      "learning_rate": 8.173652566011338e-05,
      "loss": 0.1248,
      "step": 1766
    },
    {
      "epoch": 0.3532,
      "loss_align": 0.18375694751739502,
      "loss_contrastive": 0.0,
      "loss_gen": 0.025312598794698715,
      "neg_sim_mean": 0.38678693771362305,
      "pos_sim_mean": 0.816243052482605,
      "step": 1766
    },
    {
      "epoch": 0.3534,
      "grad_norm": 1.1043701171875,
      "learning_rate": 8.170954443810948e-05,
      "loss": 0.2091,
      "step": 1767
    },
    {
      "epoch": 0.3534,
      "loss_align": 0.2738657593727112,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03127475082874298,
      "neg_sim_mean": 0.3644741177558899,
      "pos_sim_mean": 0.7261342406272888,
      "step": 1767
    },
    {
      "epoch": 0.3536,
      "grad_norm": 1.161710262298584,
      "learning_rate": 8.168254776125883e-05,
      "loss": 0.3051,
      "step": 1768
    },
    {
      "epoch": 0.3536,
      "loss_align": 0.39743709564208984,
      "loss_contrastive": 0.0,
      "loss_gen": 0.015065683983266354,
      "neg_sim_mean": 0.34741973876953125,
      "pos_sim_mean": 0.6025629043579102,
      "step": 1768
    },
    {
      "epoch": 0.3538,
      "grad_norm": 0.9787336587905884,
      "learning_rate": 8.165553564271928e-05,
      "loss": 0.4125,
      "step": 1769
    },
    {
      "epoch": 0.3538,
      "loss_align": 0.30498337745666504,
      "loss_contrastive": 0.14960944652557373,
      "loss_gen": 0.1880829781293869,
      "neg_sim_mean": 0.7446260452270508,
      "pos_sim_mean": 0.695016622543335,
      "step": 1769
    },
    {
      "epoch": 0.354,
      "grad_norm": 1.6065661907196045,
      "learning_rate": 8.162850809565623e-05,
      "loss": 0.511,
      "step": 1770
    },
    {
      "epoch": 0.354,
      "loss_align": 0.12383103370666504,
      "loss_contrastive": 0.11838537454605103,
      "loss_gen": 0.540890097618103,
      "neg_sim_mean": 0.8945543169975281,
      "pos_sim_mean": 0.876168966293335,
      "step": 1770
    },
    {
      "epoch": 0.3542,
      "grad_norm": 2.657676935195923,
      "learning_rate": 8.160146513324254e-05,
      "loss": 0.6789,
      "step": 1771
    },
    {
      "epoch": 0.3542,
      "loss_align": 0.14214611053466797,
      "loss_contrastive": 0.0885046124458313,
      "loss_gen": 0.40459930896759033,
      "neg_sim_mean": 0.8463584780693054,
      "pos_sim_mean": 0.857853889465332,
      "step": 1771
    },
    {
      "epoch": 0.3544,
      "grad_norm": 2.4735653400421143,
      "learning_rate": 8.157440676865866e-05,
      "loss": 0.5574,
      "step": 1772
    },
    {
      "epoch": 0.3544,
      "loss_align": 0.33302903175354004,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02314986102283001,
      "neg_sim_mean": 0.348158061504364,
      "pos_sim_mean": 0.66697096824646,
      "step": 1772
    },
    {
      "epoch": 0.3546,
      "grad_norm": 1.0350677967071533,
      "learning_rate": 8.154733301509248e-05,
      "loss": 0.3562,
      "step": 1773
    },
    {
      "epoch": 0.3546,
      "loss_align": 0.22188979387283325,
      "loss_contrastive": 0.0821332335472107,
      "loss_gen": 0.29797089099884033,
      "neg_sim_mean": 0.7602434158325195,
      "pos_sim_mean": 0.7781102061271667,
      "step": 1773
    },
    {
      "epoch": 0.3548,
      "grad_norm": 2.4215190410614014,
      "learning_rate": 8.152024388573945e-05,
      "loss": 0.5297,
      "step": 1774
    },
    {
      "epoch": 0.3548,
      "loss_align": 0.3101069927215576,
      "loss_contrastive": 0.014979243278503418,
      "loss_gen": 0.21017929911613464,
      "neg_sim_mean": 0.6048722267150879,
      "pos_sim_mean": 0.6898930072784424,
      "step": 1774
    },
    {
      "epoch": 0.355,
      "grad_norm": 1.9645711183547974,
      "learning_rate": 8.149313939380244e-05,
      "loss": 0.5221,
      "step": 1775
    },
    {
      "epoch": 0.355,
      "loss_align": 0.2828715443611145,
      "loss_contrastive": 0.14127439260482788,
      "loss_gen": 0.3027631342411041,
      "neg_sim_mean": 0.7584028244018555,
      "pos_sim_mean": 0.7171284556388855,
      "step": 1775
    },
    {
      "epoch": 0.3552,
      "grad_norm": 1.6803889274597168,
      "learning_rate": 8.146601955249188e-05,
      "loss": 0.6026,
      "step": 1776
    },
    {
      "epoch": 0.3552,
      "loss_align": 0.1508021354675293,
      "loss_contrastive": 0.06080448627471924,
      "loss_gen": 0.24160388112068176,
      "neg_sim_mean": 0.810002326965332,
      "pos_sim_mean": 0.8491978645324707,
      "step": 1776
    },
    {
      "epoch": 0.3554,
      "grad_norm": 1.7708957195281982,
      "learning_rate": 8.143888437502565e-05,
      "loss": 0.3997,
      "step": 1777
    },
    {
      "epoch": 0.3554,
      "loss_align": 0.14629638195037842,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0062069459818303585,
      "neg_sim_mean": 0.31984758377075195,
      "pos_sim_mean": 0.8537036180496216,
      "step": 1777
    },
    {
      "epoch": 0.3556,
      "grad_norm": 0.49134939908981323,
      "learning_rate": 8.141173387462909e-05,
      "loss": 0.1525,
      "step": 1778
    },
    {
      "epoch": 0.3556,
      "loss_align": 0.13076835870742798,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3453720211982727,
      "neg_sim_mean": 0.631640613079071,
      "pos_sim_mean": 0.869231641292572,
      "step": 1778
    },
    {
      "epoch": 0.3558,
      "grad_norm": 2.0291502475738525,
      "learning_rate": 8.138456806453503e-05,
      "loss": 0.4761,
      "step": 1779
    },
    {
      "epoch": 0.3558,
      "loss_align": 0.18484628200531006,
      "loss_contrastive": 0.08046197891235352,
      "loss_gen": 0.387012779712677,
      "neg_sim_mean": 0.7956156730651855,
      "pos_sim_mean": 0.8151537179946899,
      "step": 1779
    },
    {
      "epoch": 0.356,
      "grad_norm": 1.934741497039795,
      "learning_rate": 8.135738695798376e-05,
      "loss": 0.5815,
      "step": 1780
    },
    {
      "epoch": 0.356,
      "loss_align": 0.16852259635925293,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23013107478618622,
      "neg_sim_mean": 0.6968276500701904,
      "pos_sim_mean": 0.8314774036407471,
      "step": 1780
    },
    {
      "epoch": 0.3562,
      "grad_norm": 1.7883273363113403,
      "learning_rate": 8.133019056822304e-05,
      "loss": 0.3987,
      "step": 1781
    },
    {
      "epoch": 0.3562,
      "loss_align": 0.24567335844039917,
      "loss_contrastive": 0.1726018786430359,
      "loss_gen": 0.37214395403862,
      "neg_sim_mean": 0.8269284963607788,
      "pos_sim_mean": 0.7543266415596008,
      "step": 1781
    },
    {
      "epoch": 0.3564,
      "grad_norm": 2.237586498260498,
      "learning_rate": 8.130297890850802e-05,
      "loss": 0.6385,
      "step": 1782
    },
    {
      "epoch": 0.3564,
      "loss_align": 0.06624460220336914,
      "loss_contrastive": 0.043422698974609375,
      "loss_gen": 0.42095476388931274,
      "neg_sim_mean": 0.8771780729293823,
      "pos_sim_mean": 0.9337553977966309,
      "step": 1782
    },
    {
      "epoch": 0.3566,
      "grad_norm": 2.243381977081299,
      "learning_rate": 8.127575199210136e-05,
      "loss": 0.4924,
      "step": 1783
    },
    {
      "epoch": 0.3566,
      "loss_align": 0.3195497393608093,
      "loss_contrastive": 0.11314815282821655,
      "loss_gen": 0.39460107684135437,
      "neg_sim_mean": 0.6935983896255493,
      "pos_sim_mean": 0.6804502606391907,
      "step": 1783
    },
    {
      "epoch": 0.3568,
      "grad_norm": 2.225114345550537,
      "learning_rate": 8.124850983227312e-05,
      "loss": 0.7277,
      "step": 1784
    },
    {
      "epoch": 0.3568,
      "loss_align": 0.1965298056602478,
      "loss_contrastive": 0.19914519786834717,
      "loss_gen": 0.20681189000606537,
      "neg_sim_mean": 0.9026153683662415,
      "pos_sim_mean": 0.8034701943397522,
      "step": 1784
    },
    {
      "epoch": 0.357,
      "grad_norm": 1.7209655046463013,
      "learning_rate": 8.122125244230079e-05,
      "loss": 0.4272,
      "step": 1785
    },
    {
      "epoch": 0.357,
      "loss_align": 0.24064040184020996,
      "loss_contrastive": 0.04433012008666992,
      "loss_gen": 0.3415552079677582,
      "neg_sim_mean": 0.703689694404602,
      "pos_sim_mean": 0.75935959815979,
      "step": 1785
    },
    {
      "epoch": 0.3572,
      "grad_norm": 2.504072904586792,
      "learning_rate": 8.119397983546932e-05,
      "loss": 0.5875,
      "step": 1786
    },
    {
      "epoch": 0.3572,
      "loss_align": 0.1734742522239685,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19213569164276123,
      "neg_sim_mean": 0.7152763605117798,
      "pos_sim_mean": 0.8265257477760315,
      "step": 1786
    },
    {
      "epoch": 0.3574,
      "grad_norm": 1.3739327192306519,
      "learning_rate": 8.116669202507101e-05,
      "loss": 0.3656,
      "step": 1787
    },
    {
      "epoch": 0.3574,
      "loss_align": 0.1843264102935791,
      "loss_contrastive": 0.0,
      "loss_gen": 0.015302074141800404,
      "neg_sim_mean": 0.4118478000164032,
      "pos_sim_mean": 0.8156735897064209,
      "step": 1787
    },
    {
      "epoch": 0.3576,
      "grad_norm": 0.8792819380760193,
      "learning_rate": 8.113938902440564e-05,
      "loss": 0.1996,
      "step": 1788
    },
    {
      "epoch": 0.3576,
      "loss_align": 0.34333884716033936,
      "loss_contrastive": 0.100513756275177,
      "loss_gen": 0.2155410647392273,
      "neg_sim_mean": 0.6571748852729797,
      "pos_sim_mean": 0.6566611528396606,
      "step": 1788
    },
    {
      "epoch": 0.3578,
      "grad_norm": 1.9429576396942139,
      "learning_rate": 8.111207084678032e-05,
      "loss": 0.5709,
      "step": 1789
    },
    {
      "epoch": 0.3578,
      "loss_align": 0.20173394680023193,
      "loss_contrastive": 0.07901382446289062,
      "loss_gen": 0.1326597034931183,
      "neg_sim_mean": 0.7772798538208008,
      "pos_sim_mean": 0.7982660531997681,
      "step": 1789
    },
    {
      "epoch": 0.358,
      "grad_norm": 1.6384128332138062,
      "learning_rate": 8.108473750550965e-05,
      "loss": 0.3439,
      "step": 1790
    },
    {
      "epoch": 0.358,
      "loss_align": 0.15818411111831665,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003528531175106764,
      "neg_sim_mean": 0.26604145765304565,
      "pos_sim_mean": 0.8418158888816833,
      "step": 1790
    },
    {
      "epoch": 0.3582,
      "grad_norm": 0.4548089802265167,
      "learning_rate": 8.105738901391552e-05,
      "loss": 0.1617,
      "step": 1791
    },
    {
      "epoch": 0.3582,
      "loss_align": 0.13946306705474854,
      "loss_contrastive": 0.08114266395568848,
      "loss_gen": 0.17646688222885132,
      "neg_sim_mean": 0.841679573059082,
      "pos_sim_mean": 0.8605369329452515,
      "step": 1791
    },
    {
      "epoch": 0.3584,
      "grad_norm": 1.484968900680542,
      "learning_rate": 8.103002538532729e-05,
      "loss": 0.3257,
      "step": 1792
    },
    {
      "epoch": 0.3584,
      "loss_align": 0.18676257133483887,
      "loss_contrastive": 0.08023697137832642,
      "loss_gen": 0.3594473600387573,
      "neg_sim_mean": 0.7934743762016296,
      "pos_sim_mean": 0.8132374286651611,
      "step": 1792
    },
    {
      "epoch": 0.3586,
      "grad_norm": 2.0794923305511475,
      "learning_rate": 8.100264663308164e-05,
      "loss": 0.5558,
      "step": 1793
    },
    {
      "epoch": 0.3586,
      "loss_align": 0.18821048736572266,
      "loss_contrastive": 0.11496210098266602,
      "loss_gen": 0.41960060596466064,
      "neg_sim_mean": 0.8267515897750854,
      "pos_sim_mean": 0.8117895126342773,
      "step": 1793
    },
    {
      "epoch": 0.3588,
      "grad_norm": 2.1278982162475586,
      "learning_rate": 8.097525277052264e-05,
      "loss": 0.6216,
      "step": 1794
    },
    {
      "epoch": 0.3588,
      "loss_align": 0.33248019218444824,
      "loss_contrastive": 0.13641905784606934,
      "loss_gen": 0.40709811449050903,
      "neg_sim_mean": 0.7039388418197632,
      "pos_sim_mean": 0.6675198078155518,
      "step": 1794
    },
    {
      "epoch": 0.359,
      "grad_norm": 3.1140248775482178,
      "learning_rate": 8.094784381100174e-05,
      "loss": 0.7559,
      "step": 1795
    },
    {
      "epoch": 0.359,
      "loss_align": 0.1621614694595337,
      "loss_contrastive": 0.06296074390411377,
      "loss_gen": 0.34600499272346497,
      "neg_sim_mean": 0.8007992506027222,
      "pos_sim_mean": 0.8378385305404663,
      "step": 1795
    },
    {
      "epoch": 0.3592,
      "grad_norm": 2.1016693115234375,
      "learning_rate": 8.09204197678777e-05,
      "loss": 0.5157,
      "step": 1796
    },
    {
      "epoch": 0.3592,
      "loss_align": 0.21162641048431396,
      "loss_contrastive": 0.12664562463760376,
      "loss_gen": 0.18360859155654907,
      "neg_sim_mean": 0.8150191903114319,
      "pos_sim_mean": 0.788373589515686,
      "step": 1796
    },
    {
      "epoch": 0.3594,
      "grad_norm": 1.9693211317062378,
      "learning_rate": 8.089298065451672e-05,
      "loss": 0.4104,
      "step": 1797
    },
    {
      "epoch": 0.3594,
      "loss_align": 0.17712187767028809,
      "loss_contrastive": 0.0764855146408081,
      "loss_gen": 0.16978217661380768,
      "neg_sim_mean": 0.7993636131286621,
      "pos_sim_mean": 0.8228781223297119,
      "step": 1797
    },
    {
      "epoch": 0.3596,
      "grad_norm": 1.451027750968933,
      "learning_rate": 8.086552648429226e-05,
      "loss": 0.3561,
      "step": 1798
    },
    {
      "epoch": 0.3596,
      "loss_align": 0.1915876269340515,
      "loss_contrastive": 0.10588544607162476,
      "loss_gen": 0.3091868460178375,
      "neg_sim_mean": 0.8142977952957153,
      "pos_sim_mean": 0.8084123730659485,
      "step": 1798
    },
    {
      "epoch": 0.3598,
      "grad_norm": 2.02791428565979,
      "learning_rate": 8.083805727058513e-05,
      "loss": 0.5135,
      "step": 1799
    },
    {
      "epoch": 0.3598,
      "loss_align": 0.15590769052505493,
      "loss_contrastive": 0.09898561239242554,
      "loss_gen": 0.3201773464679718,
      "neg_sim_mean": 0.8430778980255127,
      "pos_sim_mean": 0.8440923094749451,
      "step": 1799
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.231729507446289,
      "learning_rate": 8.081057302678352e-05,
      "loss": 0.488,
      "step": 1800
    },
    {
      "epoch": 0.36,
      "loss_align": 0.2644039988517761,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012419015169143677,
      "neg_sim_mean": 0.2922181785106659,
      "pos_sim_mean": 0.7355960011482239,
      "step": 1800
    },
    {
      "epoch": 0.3602,
      "grad_norm": 0.7295103073120117,
      "learning_rate": 8.07830737662829e-05,
      "loss": 0.2768,
      "step": 1801
    },
    {
      "epoch": 0.3602,
      "loss_align": 0.20040512084960938,
      "loss_contrastive": 0.11885535717010498,
      "loss_gen": 0.46911150217056274,
      "neg_sim_mean": 0.8184502124786377,
      "pos_sim_mean": 0.7995948791503906,
      "step": 1801
    },
    {
      "epoch": 0.3604,
      "grad_norm": 2.4105560779571533,
      "learning_rate": 8.075555950248613e-05,
      "loss": 0.6838,
      "step": 1802
    },
    {
      "epoch": 0.3604,
      "loss_align": 0.1704857349395752,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003164211055263877,
      "neg_sim_mean": 0.33613166213035583,
      "pos_sim_mean": 0.8295142650604248,
      "step": 1802
    },
    {
      "epoch": 0.3606,
      "grad_norm": 0.44518500566482544,
      "learning_rate": 8.072803024880322e-05,
      "loss": 0.1736,
      "step": 1803
    },
    {
      "epoch": 0.3606,
      "loss_align": 0.1341233253479004,
      "loss_contrastive": 0.07704490423202515,
      "loss_gen": 0.2527095675468445,
      "neg_sim_mean": 0.8429215550422668,
      "pos_sim_mean": 0.8658766746520996,
      "step": 1803
    },
    {
      "epoch": 0.3608,
      "grad_norm": 2.1424357891082764,
      "learning_rate": 8.070048601865169e-05,
      "loss": 0.3961,
      "step": 1804
    },
    {
      "epoch": 0.3608,
      "loss_align": 0.24765557050704956,
      "loss_contrastive": 0.0035979747772216797,
      "loss_gen": 0.1925777643918991,
      "neg_sim_mean": 0.6559423804283142,
      "pos_sim_mean": 0.7523444294929504,
      "step": 1804
    },
    {
      "epoch": 0.361,
      "grad_norm": 1.7864577770233154,
      "learning_rate": 8.067292682545622e-05,
      "loss": 0.4407,
      "step": 1805
    },
    {
      "epoch": 0.361,
      "loss_align": 0.1523228883743286,
      "loss_contrastive": 0.1364859938621521,
      "loss_gen": 0.300616979598999,
      "neg_sim_mean": 0.8841630816459656,
      "pos_sim_mean": 0.8476771116256714,
      "step": 1805
    },
    {
      "epoch": 0.3612,
      "grad_norm": 1.7877801656723022,
      "learning_rate": 8.064535268264883e-05,
      "loss": 0.4693,
      "step": 1806
    },
    {
      "epoch": 0.3612,
      "loss_align": 0.23596704006195068,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005516884848475456,
      "neg_sim_mean": 0.5294023752212524,
      "pos_sim_mean": 0.7640329599380493,
      "step": 1806
    },
    {
      "epoch": 0.3614,
      "grad_norm": 0.6472501158714294,
      "learning_rate": 8.061776360366883e-05,
      "loss": 0.2415,
      "step": 1807
    },
    {
      "epoch": 0.3614,
      "loss_align": 0.1602182388305664,
      "loss_contrastive": 0.055982232093811035,
      "loss_gen": 0.26615020632743835,
      "neg_sim_mean": 0.7957639694213867,
      "pos_sim_mean": 0.8397817611694336,
      "step": 1807
    },
    {
      "epoch": 0.3616,
      "grad_norm": 1.8607956171035767,
      "learning_rate": 8.059015960196279e-05,
      "loss": 0.4331,
      "step": 1808
    },
    {
      "epoch": 0.3616,
      "loss_align": 0.17319899797439575,
      "loss_contrastive": 0.13738220930099487,
      "loss_gen": 0.2115083932876587,
      "neg_sim_mean": 0.8641831874847412,
      "pos_sim_mean": 0.8268010020256042,
      "step": 1808
    },
    {
      "epoch": 0.3618,
      "grad_norm": 1.630345106124878,
      "learning_rate": 8.056254069098459e-05,
      "loss": 0.4012,
      "step": 1809
    },
    {
      "epoch": 0.3618,
      "loss_align": 0.10761690139770508,
      "loss_contrastive": 0.100269615650177,
      "loss_gen": 0.34492871165275574,
      "neg_sim_mean": 0.892652690410614,
      "pos_sim_mean": 0.8923830986022949,
      "step": 1809
    },
    {
      "epoch": 0.362,
      "grad_norm": 2.1058223247528076,
      "learning_rate": 8.053490688419533e-05,
      "loss": 0.4646,
      "step": 1810
    },
    {
      "epoch": 0.362,
      "loss_align": 0.2335224747657776,
      "loss_contrastive": 0.06327027082443237,
      "loss_gen": 0.27398183941841125,
      "neg_sim_mean": 0.7297477722167969,
      "pos_sim_mean": 0.7664775252342224,
      "step": 1810
    },
    {
      "epoch": 0.3622,
      "grad_norm": 1.9473196268081665,
      "learning_rate": 8.05072581950634e-05,
      "loss": 0.5151,
      "step": 1811
    },
    {
      "epoch": 0.3622,
      "loss_align": 0.24052172899246216,
      "loss_contrastive": 0.11112558841705322,
      "loss_gen": 0.33335819840431213,
      "neg_sim_mean": 0.7706038355827332,
      "pos_sim_mean": 0.7594782710075378,
      "step": 1811
    },
    {
      "epoch": 0.3624,
      "grad_norm": 1.9583733081817627,
      "learning_rate": 8.04795946370644e-05,
      "loss": 0.5872,
      "step": 1812
    },
    {
      "epoch": 0.3624,
      "loss_align": 0.20934653282165527,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008859981782734394,
      "neg_sim_mean": 0.30010899901390076,
      "pos_sim_mean": 0.7906534671783447,
      "step": 1812
    },
    {
      "epoch": 0.3626,
      "grad_norm": 0.7181097865104675,
      "learning_rate": 8.045191622368128e-05,
      "loss": 0.2182,
      "step": 1813
    },
    {
      "epoch": 0.3626,
      "loss_align": 0.3014615774154663,
      "loss_contrastive": 0.07601523399353027,
      "loss_gen": 0.27460411190986633,
      "neg_sim_mean": 0.674553632736206,
      "pos_sim_mean": 0.6985384225845337,
      "step": 1813
    },
    {
      "epoch": 0.3628,
      "grad_norm": 1.806028962135315,
      "learning_rate": 8.04242229684041e-05,
      "loss": 0.5852,
      "step": 1814
    },
    {
      "epoch": 0.3628,
      "loss_align": 0.17279988527297974,
      "loss_contrastive": 0.11357855796813965,
      "loss_gen": 0.4088062047958374,
      "neg_sim_mean": 0.840778648853302,
      "pos_sim_mean": 0.8272001147270203,
      "step": 1814
    },
    {
      "epoch": 0.363,
      "grad_norm": 2.4305779933929443,
      "learning_rate": 8.039651488473028e-05,
      "loss": 0.5952,
      "step": 1815
    },
    {
      "epoch": 0.363,
      "loss_align": 0.13649511337280273,
      "loss_contrastive": 0.09021645784378052,
      "loss_gen": 0.16794373095035553,
      "neg_sim_mean": 0.8537213206291199,
      "pos_sim_mean": 0.8635048866271973,
      "step": 1815
    },
    {
      "epoch": 0.3632,
      "grad_norm": 1.6370786428451538,
      "learning_rate": 8.036879198616434e-05,
      "loss": 0.3153,
      "step": 1816
    },
    {
      "epoch": 0.3632,
      "loss_align": 0.2246069312095642,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1878943145275116,
      "neg_sim_mean": 0.6498917937278748,
      "pos_sim_mean": 0.7753930687904358,
      "step": 1816
    },
    {
      "epoch": 0.3634,
      "grad_norm": 1.8756835460662842,
      "learning_rate": 8.034105428621812e-05,
      "loss": 0.4125,
      "step": 1817
    },
    {
      "epoch": 0.3634,
      "loss_align": 0.28697001934051514,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3770064115524292,
      "neg_sim_mean": 0.5756097435951233,
      "pos_sim_mean": 0.7130299806594849,
      "step": 1817
    },
    {
      "epoch": 0.3636,
      "grad_norm": 2.398925304412842,
      "learning_rate": 8.031330179841062e-05,
      "loss": 0.664,
      "step": 1818
    },
    {
      "epoch": 0.3636,
      "loss_align": 0.22885137796401978,
      "loss_contrastive": 0.11152464151382446,
      "loss_gen": 0.20735882222652435,
      "neg_sim_mean": 0.7826732397079468,
      "pos_sim_mean": 0.7711486220359802,
      "step": 1818
    },
    {
      "epoch": 0.3638,
      "grad_norm": 1.6668294668197632,
      "learning_rate": 8.028553453626808e-05,
      "loss": 0.4496,
      "step": 1819
    },
    {
      "epoch": 0.3638,
      "loss_align": 0.15320098400115967,
      "loss_contrastive": 0.09374785423278809,
      "loss_gen": 0.25362297892570496,
      "neg_sim_mean": 0.8405468463897705,
      "pos_sim_mean": 0.8467990159988403,
      "step": 1819
    },
    {
      "epoch": 0.364,
      "grad_norm": 1.9346349239349365,
      "learning_rate": 8.02577525133239e-05,
      "loss": 0.4181,
      "step": 1820
    },
    {
      "epoch": 0.364,
      "loss_align": 0.3106347322463989,
      "loss_contrastive": 0.1757228970527649,
      "loss_gen": 0.40505534410476685,
      "neg_sim_mean": 0.7650881409645081,
      "pos_sim_mean": 0.6893652677536011,
      "step": 1820
    },
    {
      "epoch": 0.3642,
      "grad_norm": 2.235163450241089,
      "learning_rate": 8.022995574311876e-05,
      "loss": 0.7368,
      "step": 1821
    },
    {
      "epoch": 0.3642,
      "loss_align": 0.12452399730682373,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0023688438814133406,
      "neg_sim_mean": 0.3847596347332001,
      "pos_sim_mean": 0.8754760026931763,
      "step": 1821
    },
    {
      "epoch": 0.3644,
      "grad_norm": 0.35517168045043945,
      "learning_rate": 8.020214423920039e-05,
      "loss": 0.1269,
      "step": 1822
    },
    {
      "epoch": 0.3644,
      "loss_align": 0.200897216796875,
      "loss_contrastive": 0.06371045112609863,
      "loss_gen": 0.22302335500717163,
      "neg_sim_mean": 0.7628132104873657,
      "pos_sim_mean": 0.799102783203125,
      "step": 1822
    },
    {
      "epoch": 0.3646,
      "grad_norm": 1.9429899454116821,
      "learning_rate": 8.017431801512383e-05,
      "loss": 0.4316,
      "step": 1823
    },
    {
      "epoch": 0.3646,
      "loss_align": 0.10090434551239014,
      "loss_contrastive": 0.0061157941818237305,
      "loss_gen": 0.4181966185569763,
      "neg_sim_mean": 0.8052114248275757,
      "pos_sim_mean": 0.8990956544876099,
      "step": 1823
    },
    {
      "epoch": 0.3648,
      "grad_norm": 2.5033211708068848,
      "learning_rate": 8.014647708445124e-05,
      "loss": 0.5198,
      "step": 1824
    },
    {
      "epoch": 0.3648,
      "loss_align": 0.1629558801651001,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028269318863749504,
      "neg_sim_mean": 0.3398517966270447,
      "pos_sim_mean": 0.8370441198348999,
      "step": 1824
    },
    {
      "epoch": 0.365,
      "grad_norm": 1.3642574548721313,
      "learning_rate": 8.011862146075193e-05,
      "loss": 0.1912,
      "step": 1825
    },
    {
      "epoch": 0.365,
      "loss_align": 0.12739253044128418,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25351211428642273,
      "neg_sim_mean": 0.7651383280754089,
      "pos_sim_mean": 0.8726074695587158,
      "step": 1825
    },
    {
      "epoch": 0.3652,
      "grad_norm": 1.8415533304214478,
      "learning_rate": 8.009075115760243e-05,
      "loss": 0.3809,
      "step": 1826
    },
    {
      "epoch": 0.3652,
      "loss_align": 0.20329558849334717,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1512848436832428,
      "neg_sim_mean": 0.6928471922874451,
      "pos_sim_mean": 0.7967044115066528,
      "step": 1826
    },
    {
      "epoch": 0.3654,
      "grad_norm": 1.541783094406128,
      "learning_rate": 8.006286618858635e-05,
      "loss": 0.3546,
      "step": 1827
    },
    {
      "epoch": 0.3654,
      "loss_align": 0.3110211491584778,
      "loss_contrastive": 0.17813211679458618,
      "loss_gen": 0.3439876139163971,
      "neg_sim_mean": 0.7671109437942505,
      "pos_sim_mean": 0.6889788508415222,
      "step": 1827
    },
    {
      "epoch": 0.3656,
      "grad_norm": 2.2691824436187744,
      "learning_rate": 8.003496656729448e-05,
      "loss": 0.6764,
      "step": 1828
    },
    {
      "epoch": 0.3656,
      "loss_align": 0.12400609254837036,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003164751920849085,
      "neg_sim_mean": 0.36083459854125977,
      "pos_sim_mean": 0.8759939074516296,
      "step": 1828
    },
    {
      "epoch": 0.3658,
      "grad_norm": 0.3971754312515259,
      "learning_rate": 8.000705230732478e-05,
      "loss": 0.1272,
      "step": 1829
    },
    {
      "epoch": 0.3658,
      "loss_align": 0.0873226523399353,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1907568722963333,
      "neg_sim_mean": 0.7649950385093689,
      "pos_sim_mean": 0.9126773476600647,
      "step": 1829
    },
    {
      "epoch": 0.366,
      "grad_norm": 1.69815194606781,
      "learning_rate": 7.997912342228232e-05,
      "loss": 0.2781,
      "step": 1830
    },
    {
      "epoch": 0.366,
      "loss_align": 0.18595755100250244,
      "loss_contrastive": 0.0571213960647583,
      "loss_gen": 0.31931254267692566,
      "neg_sim_mean": 0.771163821220398,
      "pos_sim_mean": 0.8140424489974976,
      "step": 1830
    },
    {
      "epoch": 0.3662,
      "grad_norm": 1.790196180343628,
      "learning_rate": 7.99511799257793e-05,
      "loss": 0.5121,
      "step": 1831
    },
    {
      "epoch": 0.3662,
      "loss_align": 0.4218190908432007,
      "loss_contrastive": 0.11764046549797058,
      "loss_gen": 0.26084399223327637,
      "neg_sim_mean": 0.5958213806152344,
      "pos_sim_mean": 0.5781809091567993,
      "step": 1831
    },
    {
      "epoch": 0.3664,
      "grad_norm": 1.7131441831588745,
      "learning_rate": 7.992322183143503e-05,
      "loss": 0.6968,
      "step": 1832
    },
    {
      "epoch": 0.3664,
      "loss_align": 0.16123926639556885,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0021779353264719248,
      "neg_sim_mean": 0.6856765151023865,
      "pos_sim_mean": 0.8387607336044312,
      "step": 1832
    },
    {
      "epoch": 0.3666,
      "grad_norm": 0.3973623514175415,
      "learning_rate": 7.989524915287595e-05,
      "loss": 0.1634,
      "step": 1833
    },
    {
      "epoch": 0.3666,
      "loss_align": 0.25122249126434326,
      "loss_contrastive": 0.150529146194458,
      "loss_gen": 0.3379417955875397,
      "neg_sim_mean": 0.7993066310882568,
      "pos_sim_mean": 0.7487775087356567,
      "step": 1833
    },
    {
      "epoch": 0.3668,
      "grad_norm": 2.136350631713867,
      "learning_rate": 7.986726190373561e-05,
      "loss": 0.6072,
      "step": 1834
    },
    {
      "epoch": 0.3668,
      "loss_align": 0.18446898460388184,
      "loss_contrastive": 0.03328812122344971,
      "loss_gen": 0.18565109372138977,
      "neg_sim_mean": 0.74881911277771,
      "pos_sim_mean": 0.8155310153961182,
      "step": 1834
    },
    {
      "epoch": 0.367,
      "grad_norm": 1.4612773656845093,
      "learning_rate": 7.983926009765463e-05,
      "loss": 0.3741,
      "step": 1835
    },
    {
      "epoch": 0.367,
      "loss_align": 0.20811140537261963,
      "loss_contrastive": 0.145003080368042,
      "loss_gen": 0.3586626350879669,
      "neg_sim_mean": 0.8368916511535645,
      "pos_sim_mean": 0.7918885946273804,
      "step": 1835
    },
    {
      "epoch": 0.3672,
      "grad_norm": 2.101956367492676,
      "learning_rate": 7.98112437482808e-05,
      "loss": 0.5842,
      "step": 1836
    },
    {
      "epoch": 0.3672,
      "loss_align": 0.18938922882080078,
      "loss_contrastive": 0.14116966724395752,
      "loss_gen": 0.5285218358039856,
      "neg_sim_mean": 0.8517804145812988,
      "pos_sim_mean": 0.8106107711791992,
      "step": 1836
    },
    {
      "epoch": 0.3674,
      "grad_norm": 3.1028997898101807,
      "learning_rate": 7.97832128692689e-05,
      "loss": 0.7349,
      "step": 1837
    },
    {
      "epoch": 0.3674,
      "loss_align": 0.23401093482971191,
      "loss_contrastive": 0.13776147365570068,
      "loss_gen": 0.2574929893016815,
      "neg_sim_mean": 0.8037505149841309,
      "pos_sim_mean": 0.7659890651702881,
      "step": 1837
    },
    {
      "epoch": 0.3676,
      "grad_norm": 1.857134222984314,
      "learning_rate": 7.975516747428087e-05,
      "loss": 0.508,
      "step": 1838
    },
    {
      "epoch": 0.3676,
      "loss_align": 0.2024136185646057,
      "loss_contrastive": 0.07822084426879883,
      "loss_gen": 0.43044623732566833,
      "neg_sim_mean": 0.7758072018623352,
      "pos_sim_mean": 0.7975863814353943,
      "step": 1838
    },
    {
      "epoch": 0.3678,
      "grad_norm": 2.613476037979126,
      "learning_rate": 7.972710757698567e-05,
      "loss": 0.6422,
      "step": 1839
    },
    {
      "epoch": 0.3678,
      "loss_align": 0.30122125148773193,
      "loss_contrastive": 0.08695805072784424,
      "loss_gen": 0.32417795062065125,
      "neg_sim_mean": 0.6857367753982544,
      "pos_sim_mean": 0.6987787485122681,
      "step": 1839
    },
    {
      "epoch": 0.368,
      "grad_norm": 2.2702910900115967,
      "learning_rate": 7.969903319105935e-05,
      "loss": 0.6358,
      "step": 1840
    },
    {
      "epoch": 0.368,
      "loss_align": 0.2200758457183838,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19914710521697998,
      "neg_sim_mean": 0.6725616455078125,
      "pos_sim_mean": 0.7799241542816162,
      "step": 1840
    },
    {
      "epoch": 0.3682,
      "grad_norm": 1.8157877922058105,
      "learning_rate": 7.967094433018508e-05,
      "loss": 0.4192,
      "step": 1841
    },
    {
      "epoch": 0.3682,
      "loss_align": 0.20965421199798584,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009211298078298569,
      "neg_sim_mean": 0.346078485250473,
      "pos_sim_mean": 0.7903457880020142,
      "step": 1841
    },
    {
      "epoch": 0.3684,
      "grad_norm": 0.6682487726211548,
      "learning_rate": 7.964284100805297e-05,
      "loss": 0.2189,
      "step": 1842
    },
    {
      "epoch": 0.3684,
      "loss_align": 0.14037126302719116,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008198815397918224,
      "neg_sim_mean": 0.39117375016212463,
      "pos_sim_mean": 0.8596287369728088,
      "step": 1842
    },
    {
      "epoch": 0.3686,
      "grad_norm": 0.6525157690048218,
      "learning_rate": 7.961472323836026e-05,
      "loss": 0.1486,
      "step": 1843
    },
    {
      "epoch": 0.3686,
      "loss_align": 0.18305671215057373,
      "loss_contrastive": 0.14718300104141235,
      "loss_gen": 0.16816604137420654,
      "neg_sim_mean": 0.8641262650489807,
      "pos_sim_mean": 0.8169432878494263,
      "step": 1843
    },
    {
      "epoch": 0.3688,
      "grad_norm": 1.418088436126709,
      "learning_rate": 7.95865910348112e-05,
      "loss": 0.3689,
      "step": 1844
    },
    {
      "epoch": 0.3688,
      "loss_align": 0.24206113815307617,
      "loss_contrastive": 0.004186034202575684,
      "loss_gen": 0.005966844968497753,
      "neg_sim_mean": 0.6621248722076416,
      "pos_sim_mean": 0.7579388618469238,
      "step": 1844
    },
    {
      "epoch": 0.369,
      "grad_norm": 0.5636115670204163,
      "learning_rate": 7.95584444111171e-05,
      "loss": 0.2485,
      "step": 1845
    },
    {
      "epoch": 0.369,
      "loss_align": 0.12766844034194946,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25043898820877075,
      "neg_sim_mean": 0.7204475998878479,
      "pos_sim_mean": 0.8723315596580505,
      "step": 1845
    },
    {
      "epoch": 0.3692,
      "grad_norm": 1.8122678995132446,
      "learning_rate": 7.953028338099627e-05,
      "loss": 0.3781,
      "step": 1846
    },
    {
      "epoch": 0.3692,
      "loss_align": 0.1752585768699646,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22692139446735382,
      "neg_sim_mean": 0.6596486568450928,
      "pos_sim_mean": 0.8247414231300354,
      "step": 1846
    },
    {
      "epoch": 0.3694,
      "grad_norm": 1.9309754371643066,
      "learning_rate": 7.950210795817407e-05,
      "loss": 0.4022,
      "step": 1847
    },
    {
      "epoch": 0.3694,
      "loss_align": 0.22703897953033447,
      "loss_contrastive": 0.14805829524993896,
      "loss_gen": 0.6274051070213318,
      "neg_sim_mean": 0.8210192918777466,
      "pos_sim_mean": 0.7729610204696655,
      "step": 1847
    },
    {
      "epoch": 0.3696,
      "grad_norm": 2.597676992416382,
      "learning_rate": 7.947391815638284e-05,
      "loss": 0.8722,
      "step": 1848
    },
    {
      "epoch": 0.3696,
      "loss_align": 0.20254909992218018,
      "loss_contrastive": 0.07974410057067871,
      "loss_gen": 0.16778336465358734,
      "neg_sim_mean": 0.7771949768066406,
      "pos_sim_mean": 0.7974509000778198,
      "step": 1848
    },
    {
      "epoch": 0.3698,
      "grad_norm": 1.6177244186401367,
      "learning_rate": 7.944571398936193e-05,
      "loss": 0.3799,
      "step": 1849
    },
    {
      "epoch": 0.3698,
      "loss_align": 0.12472653388977051,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011762740323320031,
      "neg_sim_mean": 0.3434268832206726,
      "pos_sim_mean": 0.8752734661102295,
      "step": 1849
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.32949304580688477,
      "learning_rate": 7.941749547085777e-05,
      "loss": 0.1259,
      "step": 1850
    },
    {
      "epoch": 0.37,
      "loss_align": 0.1484203338623047,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0013553227763623,
      "neg_sim_mean": 0.4883388876914978,
      "pos_sim_mean": 0.8515796661376953,
      "step": 1850
    },
    {
      "epoch": 0.3702,
      "grad_norm": 0.5065698623657227,
      "learning_rate": 7.938926261462366e-05,
      "loss": 0.1498,
      "step": 1851
    },
    {
      "epoch": 0.3702,
      "loss_align": 0.29562121629714966,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0016609852900728583,
      "neg_sim_mean": 0.4397144317626953,
      "pos_sim_mean": 0.7043787837028503,
      "step": 1851
    },
    {
      "epoch": 0.3704,
      "grad_norm": 0.6247354745864868,
      "learning_rate": 7.936101543441998e-05,
      "loss": 0.2973,
      "step": 1852
    },
    {
      "epoch": 0.3704,
      "loss_align": 0.32700520753860474,
      "loss_contrastive": 0.13186508417129517,
      "loss_gen": 0.2190369963645935,
      "neg_sim_mean": 0.7048598527908325,
      "pos_sim_mean": 0.6729947924613953,
      "step": 1852
    },
    {
      "epoch": 0.3706,
      "grad_norm": 1.78740656375885,
      "learning_rate": 7.933275394401406e-05,
      "loss": 0.5619,
      "step": 1853
    },
    {
      "epoch": 0.3706,
      "loss_align": 0.1982986330986023,
      "loss_contrastive": 0.018704712390899658,
      "loss_gen": 0.059819817543029785,
      "neg_sim_mean": 0.7204060554504395,
      "pos_sim_mean": 0.8017013669013977,
      "step": 1853
    },
    {
      "epoch": 0.3708,
      "grad_norm": 1.1011886596679688,
      "learning_rate": 7.930447815718022e-05,
      "loss": 0.2604,
      "step": 1854
    },
    {
      "epoch": 0.3708,
      "loss_align": 0.20570117235183716,
      "loss_contrastive": 0.08435690402984619,
      "loss_gen": 0.27837204933166504,
      "neg_sim_mean": 0.7786557078361511,
      "pos_sim_mean": 0.7942988276481628,
      "step": 1854
    },
    {
      "epoch": 0.371,
      "grad_norm": 2.4270641803741455,
      "learning_rate": 7.92761880876997e-05,
      "loss": 0.4942,
      "step": 1855
    },
    {
      "epoch": 0.371,
      "loss_align": 0.2763513922691345,
      "loss_contrastive": 0.06001240015029907,
      "loss_gen": 0.25867265462875366,
      "neg_sim_mean": 0.6836609840393066,
      "pos_sim_mean": 0.7236486077308655,
      "step": 1855
    },
    {
      "epoch": 0.3712,
      "grad_norm": 2.0190529823303223,
      "learning_rate": 7.924788374936078e-05,
      "loss": 0.5422,
      "step": 1856
    },
    {
      "epoch": 0.3712,
      "loss_align": 0.12851840257644653,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002272552577778697,
      "neg_sim_mean": 0.30801519751548767,
      "pos_sim_mean": 0.8714815974235535,
      "step": 1856
    },
    {
      "epoch": 0.3714,
      "grad_norm": 0.4014641344547272,
      "learning_rate": 7.921956515595861e-05,
      "loss": 0.1308,
      "step": 1857
    },
    {
      "epoch": 0.3714,
      "loss_align": 0.13568425178527832,
      "loss_contrastive": 0.097057044506073,
      "loss_gen": 0.2597246468067169,
      "neg_sim_mean": 0.8613727688789368,
      "pos_sim_mean": 0.8643157482147217,
      "step": 1857
    },
    {
      "epoch": 0.3716,
      "grad_norm": 1.6439262628555298,
      "learning_rate": 7.919123232129535e-05,
      "loss": 0.4071,
      "step": 1858
    },
    {
      "epoch": 0.3716,
      "loss_align": 0.17509770393371582,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012452024035155773,
      "neg_sim_mean": 0.390115886926651,
      "pos_sim_mean": 0.8249022960662842,
      "step": 1858
    },
    {
      "epoch": 0.3718,
      "grad_norm": 0.8659335970878601,
      "learning_rate": 7.916288525918007e-05,
      "loss": 0.1875,
      "step": 1859
    },
    {
      "epoch": 0.3718,
      "loss_align": 0.2381419539451599,
      "loss_contrastive": 0.04796653985977173,
      "loss_gen": 0.1877664178609848,
      "neg_sim_mean": 0.7098245620727539,
      "pos_sim_mean": 0.7618580460548401,
      "step": 1859
    },
    {
      "epoch": 0.372,
      "grad_norm": 1.7026824951171875,
      "learning_rate": 7.913452398342881e-05,
      "loss": 0.4317,
      "step": 1860
    },
    {
      "epoch": 0.372,
      "loss_align": 0.2202235460281372,
      "loss_contrastive": 0.10222333669662476,
      "loss_gen": 0.32948803901672363,
      "neg_sim_mean": 0.7819997668266296,
      "pos_sim_mean": 0.7797764539718628,
      "step": 1860
    },
    {
      "epoch": 0.3722,
      "grad_norm": 2.124629259109497,
      "learning_rate": 7.910614850786448e-05,
      "loss": 0.562,
      "step": 1861
    },
    {
      "epoch": 0.3722,
      "loss_align": 0.1744556427001953,
      "loss_contrastive": 0.011456489562988281,
      "loss_gen": 0.22421248257160187,
      "neg_sim_mean": 0.7370008230209351,
      "pos_sim_mean": 0.8255443572998047,
      "step": 1861
    },
    {
      "epoch": 0.3724,
      "grad_norm": 2.1396327018737793,
      "learning_rate": 7.907775884631694e-05,
      "loss": 0.4,
      "step": 1862
    },
    {
      "epoch": 0.3724,
      "loss_align": 0.36889588832855225,
      "loss_contrastive": 0.029813647270202637,
      "loss_gen": 0.2697761654853821,
      "neg_sim_mean": 0.5609177350997925,
      "pos_sim_mean": 0.6311041116714478,
      "step": 1862
    },
    {
      "epoch": 0.3726,
      "grad_norm": 1.9800419807434082,
      "learning_rate": 7.9049355012623e-05,
      "loss": 0.6422,
      "step": 1863
    },
    {
      "epoch": 0.3726,
      "loss_align": 0.20432031154632568,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03201077878475189,
      "neg_sim_mean": 0.45672667026519775,
      "pos_sim_mean": 0.7956796884536743,
      "step": 1863
    },
    {
      "epoch": 0.3728,
      "grad_norm": 1.5240612030029297,
      "learning_rate": 7.90209370206263e-05,
      "loss": 0.2363,
      "step": 1864
    },
    {
      "epoch": 0.3728,
      "loss_align": 0.10921818017959595,
      "loss_contrastive": 0.11792159080505371,
      "loss_gen": 0.26519596576690674,
      "neg_sim_mean": 0.9087033867835999,
      "pos_sim_mean": 0.890781819820404,
      "step": 1864
    },
    {
      "epoch": 0.373,
      "grad_norm": 1.8661314249038696,
      "learning_rate": 7.899250488417747e-05,
      "loss": 0.3886,
      "step": 1865
    },
    {
      "epoch": 0.373,
      "loss_align": 0.19857442378997803,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0012856435496360064,
      "neg_sim_mean": 0.6441717147827148,
      "pos_sim_mean": 0.801425576210022,
      "step": 1865
    },
    {
      "epoch": 0.3732,
      "grad_norm": 0.5422137975692749,
      "learning_rate": 7.896405861713394e-05,
      "loss": 0.1999,
      "step": 1866
    },
    {
      "epoch": 0.3732,
      "loss_align": 0.2817167043685913,
      "loss_contrastive": 0.09177041053771973,
      "loss_gen": 0.25107330083847046,
      "neg_sim_mean": 0.7100536823272705,
      "pos_sim_mean": 0.7182832956314087,
      "step": 1866
    },
    {
      "epoch": 0.3734,
      "grad_norm": 2.252591133117676,
      "learning_rate": 7.893559823336012e-05,
      "loss": 0.5438,
      "step": 1867
    },
    {
      "epoch": 0.3734,
      "loss_align": 0.23144572973251343,
      "loss_contrastive": 0.01931864023208618,
      "loss_gen": 0.38982951641082764,
      "neg_sim_mean": 0.6878728866577148,
      "pos_sim_mean": 0.7685542702674866,
      "step": 1867
    },
    {
      "epoch": 0.3736,
      "grad_norm": 2.4216628074645996,
      "learning_rate": 7.890712374672723e-05,
      "loss": 0.6236,
      "step": 1868
    },
    {
      "epoch": 0.3736,
      "loss_align": 0.12280237674713135,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21371424198150635,
      "neg_sim_mean": 0.7297568321228027,
      "pos_sim_mean": 0.8771976232528687,
      "step": 1868
    },
    {
      "epoch": 0.3738,
      "grad_norm": 1.6154987812042236,
      "learning_rate": 7.887863517111338e-05,
      "loss": 0.3365,
      "step": 1869
    },
    {
      "epoch": 0.3738,
      "loss_align": 0.1834418773651123,
      "loss_contrastive": 0.0,
      "loss_gen": 0.307756245136261,
      "neg_sim_mean": 0.6981692910194397,
      "pos_sim_mean": 0.8165581226348877,
      "step": 1869
    },
    {
      "epoch": 0.374,
      "grad_norm": 1.9130148887634277,
      "learning_rate": 7.885013252040359e-05,
      "loss": 0.4912,
      "step": 1870
    },
    {
      "epoch": 0.374,
      "loss_align": 0.3333784341812134,
      "loss_contrastive": 0.1424289345741272,
      "loss_gen": 0.564798891544342,
      "neg_sim_mean": 0.7090504765510559,
      "pos_sim_mean": 0.6666215658187866,
      "step": 1870
    },
    {
      "epoch": 0.3742,
      "grad_norm": 3.362436056137085,
      "learning_rate": 7.882161580848967e-05,
      "loss": 0.9153,
      "step": 1871
    },
    {
      "epoch": 0.3742,
      "loss_align": 0.11314034461975098,
      "loss_contrastive": 0.0743173360824585,
      "loss_gen": 0.3899879455566406,
      "neg_sim_mean": 0.8611769676208496,
      "pos_sim_mean": 0.886859655380249,
      "step": 1871
    },
    {
      "epoch": 0.3744,
      "grad_norm": 2.442035675048828,
      "learning_rate": 7.879308504927035e-05,
      "loss": 0.512,
      "step": 1872
    },
    {
      "epoch": 0.3744,
      "loss_align": 0.1858581304550171,
      "loss_contrastive": 0.0629192590713501,
      "loss_gen": 0.2955109179019928,
      "neg_sim_mean": 0.7770611047744751,
      "pos_sim_mean": 0.8141418695449829,
      "step": 1872
    },
    {
      "epoch": 0.3746,
      "grad_norm": 1.7403088808059692,
      "learning_rate": 7.876454025665114e-05,
      "loss": 0.4889,
      "step": 1873
    },
    {
      "epoch": 0.3746,
      "loss_align": 0.12244975566864014,
      "loss_contrastive": 0.013656854629516602,
      "loss_gen": 0.40981489419937134,
      "neg_sim_mean": 0.7912070751190186,
      "pos_sim_mean": 0.8775502443313599,
      "step": 1873
    },
    {
      "epoch": 0.3748,
      "grad_norm": 2.659356117248535,
      "learning_rate": 7.873598144454443e-05,
      "loss": 0.5339,
      "step": 1874
    },
    {
      "epoch": 0.3748,
      "loss_align": 0.23197603225708008,
      "loss_contrastive": 0.16134512424468994,
      "loss_gen": 0.19896265864372253,
      "neg_sim_mean": 0.829369068145752,
      "pos_sim_mean": 0.7680239677429199,
      "step": 1874
    },
    {
      "epoch": 0.375,
      "grad_norm": 1.5686118602752686,
      "learning_rate": 7.87074086268695e-05,
      "loss": 0.4503,
      "step": 1875
    },
    {
      "epoch": 0.375,
      "loss_align": 0.1849367618560791,
      "loss_contrastive": 0.09032833576202393,
      "loss_gen": 0.22469505667686462,
      "neg_sim_mean": 0.8053915500640869,
      "pos_sim_mean": 0.8150632381439209,
      "step": 1875
    },
    {
      "epoch": 0.3752,
      "grad_norm": 1.7232731580734253,
      "learning_rate": 7.86788218175523e-05,
      "loss": 0.4205,
      "step": 1876
    },
    {
      "epoch": 0.3752,
      "loss_align": 0.24809235334396362,
      "loss_contrastive": 0.15517836809158325,
      "loss_gen": 0.20933464169502258,
      "neg_sim_mean": 0.8070859909057617,
      "pos_sim_mean": 0.7519076466560364,
      "step": 1876
    },
    {
      "epoch": 0.3754,
      "grad_norm": 1.9466642141342163,
      "learning_rate": 7.865022103052578e-05,
      "loss": 0.476,
      "step": 1877
    },
    {
      "epoch": 0.3754,
      "loss_align": 0.18454623222351074,
      "loss_contrastive": 0.04102742671966553,
      "loss_gen": 0.2650861144065857,
      "neg_sim_mean": 0.7564811706542969,
      "pos_sim_mean": 0.8154537677764893,
      "step": 1877
    },
    {
      "epoch": 0.3756,
      "grad_norm": 1.5432415008544922,
      "learning_rate": 7.862160627972955e-05,
      "loss": 0.4546,
      "step": 1878
    },
    {
      "epoch": 0.3756,
      "loss_align": 0.23968547582626343,
      "loss_contrastive": 0.08839744329452515,
      "loss_gen": 0.5473530888557434,
      "neg_sim_mean": 0.7487119436264038,
      "pos_sim_mean": 0.7603145241737366,
      "step": 1878
    },
    {
      "epoch": 0.3758,
      "grad_norm": 2.8103814125061035,
      "learning_rate": 7.859297757911013e-05,
      "loss": 0.7976,
      "step": 1879
    },
    {
      "epoch": 0.3758,
      "loss_align": 0.2127537727355957,
      "loss_contrastive": 0.07263356447219849,
      "loss_gen": 0.23446010053157806,
      "neg_sim_mean": 0.7598797678947449,
      "pos_sim_mean": 0.7872462272644043,
      "step": 1879
    },
    {
      "epoch": 0.376,
      "grad_norm": 1.6405730247497559,
      "learning_rate": 7.856433494262077e-05,
      "loss": 0.4559,
      "step": 1880
    },
    {
      "epoch": 0.376,
      "loss_align": 0.2992474436759949,
      "loss_contrastive": 0.0960661768913269,
      "loss_gen": 0.16625460982322693,
      "neg_sim_mean": 0.6968187093734741,
      "pos_sim_mean": 0.7007525563240051,
      "step": 1880
    },
    {
      "epoch": 0.3762,
      "grad_norm": 1.6213834285736084,
      "learning_rate": 7.85356783842216e-05,
      "loss": 0.477,
      "step": 1881
    },
    {
      "epoch": 0.3762,
      "loss_align": 0.19057965278625488,
      "loss_contrastive": 0.08774381875991821,
      "loss_gen": 0.22700831294059753,
      "neg_sim_mean": 0.7971641421318054,
      "pos_sim_mean": 0.8094203472137451,
      "step": 1881
    },
    {
      "epoch": 0.3764,
      "grad_norm": 1.9662343263626099,
      "learning_rate": 7.850700791787942e-05,
      "loss": 0.4281,
      "step": 1882
    },
    {
      "epoch": 0.3764,
      "loss_align": 0.12941104173660278,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2895298898220062,
      "neg_sim_mean": 0.7521596550941467,
      "pos_sim_mean": 0.8705889582633972,
      "step": 1882
    },
    {
      "epoch": 0.3766,
      "grad_norm": 2.3819656372070312,
      "learning_rate": 7.847832355756788e-05,
      "loss": 0.4189,
      "step": 1883
    },
    {
      "epoch": 0.3766,
      "loss_align": 0.12991547584533691,
      "loss_contrastive": 0.09166431427001953,
      "loss_gen": 0.35670432448387146,
      "neg_sim_mean": 0.8617488145828247,
      "pos_sim_mean": 0.8700845241546631,
      "step": 1883
    },
    {
      "epoch": 0.3768,
      "grad_norm": 2.1308107376098633,
      "learning_rate": 7.84496253172674e-05,
      "loss": 0.4976,
      "step": 1884
    },
    {
      "epoch": 0.3768,
      "loss_align": 0.15269804000854492,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008810088969767094,
      "neg_sim_mean": 0.45639461278915405,
      "pos_sim_mean": 0.8473019599914551,
      "step": 1884
    },
    {
      "epoch": 0.377,
      "grad_norm": 0.6147511601448059,
      "learning_rate": 7.842091321096515e-05,
      "loss": 0.1615,
      "step": 1885
    },
    {
      "epoch": 0.377,
      "loss_align": 0.2381880283355713,
      "loss_contrastive": 0.050730109214782715,
      "loss_gen": 0.3152073621749878,
      "neg_sim_mean": 0.7125420570373535,
      "pos_sim_mean": 0.7618119716644287,
      "step": 1885
    },
    {
      "epoch": 0.3772,
      "grad_norm": 1.760503888130188,
      "learning_rate": 7.839218725265506e-05,
      "loss": 0.5595,
      "step": 1886
    },
    {
      "epoch": 0.3772,
      "loss_align": 0.12148946523666382,
      "loss_contrastive": 0.00873565673828125,
      "loss_gen": 0.2714894115924835,
      "neg_sim_mean": 0.7872461676597595,
      "pos_sim_mean": 0.8785105347633362,
      "step": 1886
    },
    {
      "epoch": 0.3774,
      "grad_norm": 1.7079325914382935,
      "learning_rate": 7.836344745633783e-05,
      "loss": 0.394,
      "step": 1887
    },
    {
      "epoch": 0.3774,
      "loss_align": 0.20516204833984375,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3129279315471649,
      "neg_sim_mean": 0.44851332902908325,
      "pos_sim_mean": 0.7948379516601562,
      "step": 1887
    },
    {
      "epoch": 0.3776,
      "grad_norm": 1.9335650205612183,
      "learning_rate": 7.833469383602085e-05,
      "loss": 0.5181,
      "step": 1888
    },
    {
      "epoch": 0.3776,
      "loss_align": 0.37586504220962524,
      "loss_contrastive": 0.04737931489944458,
      "loss_gen": 0.28182724118232727,
      "neg_sim_mean": 0.5715142488479614,
      "pos_sim_mean": 0.6241349577903748,
      "step": 1888
    },
    {
      "epoch": 0.3778,
      "grad_norm": 2.2675468921661377,
      "learning_rate": 7.830592640571834e-05,
      "loss": 0.6634,
      "step": 1889
    },
    {
      "epoch": 0.3778,
      "loss_align": 0.26071029901504517,
      "loss_contrastive": 0.07771730422973633,
      "loss_gen": 0.38536337018013,
      "neg_sim_mean": 0.7170069813728333,
      "pos_sim_mean": 0.7392897009849548,
      "step": 1889
    },
    {
      "epoch": 0.378,
      "grad_norm": 1.9701366424560547,
      "learning_rate": 7.827714517945115e-05,
      "loss": 0.6554,
      "step": 1890
    },
    {
      "epoch": 0.378,
      "loss_align": 0.0949588418006897,
      "loss_contrastive": 0.06633585691452026,
      "loss_gen": 0.11641428619623184,
      "neg_sim_mean": 0.8713769912719727,
      "pos_sim_mean": 0.9050411581993103,
      "step": 1890
    },
    {
      "epoch": 0.3782,
      "grad_norm": 1.2701895236968994,
      "learning_rate": 7.82483501712469e-05,
      "loss": 0.2193,
      "step": 1891
    },
    {
      "epoch": 0.3782,
      "loss_align": 0.24406611919403076,
      "loss_contrastive": 0.11711740493774414,
      "loss_gen": 0.21709144115447998,
      "neg_sim_mean": 0.7730512619018555,
      "pos_sim_mean": 0.7559338808059692,
      "step": 1891
    },
    {
      "epoch": 0.3784,
      "grad_norm": 1.7330454587936401,
      "learning_rate": 7.821954139513996e-05,
      "loss": 0.4752,
      "step": 1892
    },
    {
      "epoch": 0.3784,
      "loss_align": 0.2552921772003174,
      "loss_contrastive": 0.13923048973083496,
      "loss_gen": 0.30524930357933044,
      "neg_sim_mean": 0.7839382886886597,
      "pos_sim_mean": 0.7447078227996826,
      "step": 1892
    },
    {
      "epoch": 0.3786,
      "grad_norm": 2.2533633708953857,
      "learning_rate": 7.819071886517134e-05,
      "loss": 0.5772,
      "step": 1893
    },
    {
      "epoch": 0.3786,
      "loss_align": 0.2590596079826355,
      "loss_contrastive": 0.009371459484100342,
      "loss_gen": 0.31793129444122314,
      "neg_sim_mean": 0.6503118276596069,
      "pos_sim_mean": 0.7409403920173645,
      "step": 1893
    },
    {
      "epoch": 0.3788,
      "grad_norm": 1.8223270177841187,
      "learning_rate": 7.816188259538884e-05,
      "loss": 0.5781,
      "step": 1894
    },
    {
      "epoch": 0.3788,
      "loss_align": 0.21356070041656494,
      "loss_contrastive": 0.05628407001495361,
      "loss_gen": 0.28916749358177185,
      "neg_sim_mean": 0.7427233457565308,
      "pos_sim_mean": 0.7864392995834351,
      "step": 1894
    },
    {
      "epoch": 0.379,
      "grad_norm": 1.7988765239715576,
      "learning_rate": 7.813303259984685e-05,
      "loss": 0.5095,
      "step": 1895
    },
    {
      "epoch": 0.379,
      "loss_align": 0.23995757102966309,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3435003459453583,
      "neg_sim_mean": 0.5873886346817017,
      "pos_sim_mean": 0.7600424289703369,
      "step": 1895
    },
    {
      "epoch": 0.3792,
      "grad_norm": 1.7275415658950806,
      "learning_rate": 7.810416889260653e-05,
      "loss": 0.5835,
      "step": 1896
    },
    {
      "epoch": 0.3792,
      "loss_align": 0.13030296564102173,
      "loss_contrastive": 0.10115557909011841,
      "loss_gen": 0.5096735954284668,
      "neg_sim_mean": 0.8708525896072388,
      "pos_sim_mean": 0.8696970343589783,
      "step": 1896
    },
    {
      "epoch": 0.3794,
      "grad_norm": 2.3637313842773438,
      "learning_rate": 7.807529148773572e-05,
      "loss": 0.6521,
      "step": 1897
    },
    {
      "epoch": 0.3794,
      "loss_align": 0.23865509033203125,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006251928396522999,
      "neg_sim_mean": 0.42369839549064636,
      "pos_sim_mean": 0.7613449096679688,
      "step": 1897
    },
    {
      "epoch": 0.3796,
      "grad_norm": 0.5395580530166626,
      "learning_rate": 7.80464003993089e-05,
      "loss": 0.2449,
      "step": 1898
    },
    {
      "epoch": 0.3796,
      "loss_align": 0.27788662910461426,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29689228534698486,
      "neg_sim_mean": 0.6131507158279419,
      "pos_sim_mean": 0.7221133708953857,
      "step": 1898
    },
    {
      "epoch": 0.3798,
      "grad_norm": 2.090304374694824,
      "learning_rate": 7.801749564140724e-05,
      "loss": 0.5748,
      "step": 1899
    },
    {
      "epoch": 0.3798,
      "loss_align": 0.28411728143692017,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14957937598228455,
      "neg_sim_mean": 0.5995557308197021,
      "pos_sim_mean": 0.7158827185630798,
      "step": 1899
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.6965235471725464,
      "learning_rate": 7.798857722811856e-05,
      "loss": 0.4337,
      "step": 1900
    },
    {
      "epoch": 0.38,
      "loss_align": 0.20893794298171997,
      "loss_contrastive": 0.060563623905181885,
      "loss_gen": 0.3406183123588562,
      "neg_sim_mean": 0.751625657081604,
      "pos_sim_mean": 0.79106205701828,
      "step": 1900
    },
    {
      "epoch": 0.3802,
      "grad_norm": 1.9941967725753784,
      "learning_rate": 7.795964517353735e-05,
      "loss": 0.5568,
      "step": 1901
    },
    {
      "epoch": 0.3802,
      "loss_align": 0.42124617099761963,
      "loss_contrastive": 0.10530087351799011,
      "loss_gen": 0.05328839272260666,
      "neg_sim_mean": 0.584054708480835,
      "pos_sim_mean": 0.5787538290023804,
      "step": 1901
    },
    {
      "epoch": 0.3804,
      "grad_norm": 1.488142967224121,
      "learning_rate": 7.793069949176473e-05,
      "loss": 0.4872,
      "step": 1902
    },
    {
      "epoch": 0.3804,
      "loss_align": 0.2484511137008667,
      "loss_contrastive": 0.0919460654258728,
      "loss_gen": 0.17326584458351135,
      "neg_sim_mean": 0.7434949278831482,
      "pos_sim_mean": 0.7515488862991333,
      "step": 1902
    },
    {
      "epoch": 0.3806,
      "grad_norm": 1.4694474935531616,
      "learning_rate": 7.79017401969085e-05,
      "loss": 0.4328,
      "step": 1903
    },
    {
      "epoch": 0.3806,
      "loss_align": 0.2145094871520996,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2008388787508011,
      "neg_sim_mean": 0.5744110941886902,
      "pos_sim_mean": 0.7854905128479004,
      "step": 1903
    },
    {
      "epoch": 0.3808,
      "grad_norm": 1.683977484703064,
      "learning_rate": 7.787276730308304e-05,
      "loss": 0.4153,
      "step": 1904
    },
    {
      "epoch": 0.3808,
      "loss_align": 0.2736114263534546,
      "loss_contrastive": 0.10227620601654053,
      "loss_gen": 0.23070214688777924,
      "neg_sim_mean": 0.728664755821228,
      "pos_sim_mean": 0.7263885736465454,
      "step": 1904
    },
    {
      "epoch": 0.381,
      "grad_norm": 1.9514930248260498,
      "learning_rate": 7.784378082440941e-05,
      "loss": 0.5166,
      "step": 1905
    },
    {
      "epoch": 0.381,
      "loss_align": 0.2806234359741211,
      "loss_contrastive": 0.2002490758895874,
      "loss_gen": 0.3011503517627716,
      "neg_sim_mean": 0.8196256160736084,
      "pos_sim_mean": 0.7193765640258789,
      "step": 1905
    },
    {
      "epoch": 0.3812,
      "grad_norm": 2.3959083557128906,
      "learning_rate": 7.781478077501525e-05,
      "loss": 0.6058,
      "step": 1906
    },
    {
      "epoch": 0.3812,
      "loss_align": 0.17583686113357544,
      "loss_contrastive": 0.07691454887390137,
      "loss_gen": 0.18645718693733215,
      "neg_sim_mean": 0.801077663898468,
      "pos_sim_mean": 0.8241631388664246,
      "step": 1906
    },
    {
      "epoch": 0.3814,
      "grad_norm": 1.5471899509429932,
      "learning_rate": 7.778576716903483e-05,
      "loss": 0.3715,
      "step": 1907
    },
    {
      "epoch": 0.3814,
      "loss_align": 0.2834278345108032,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0027901032008230686,
      "neg_sim_mean": 0.39952701330184937,
      "pos_sim_mean": 0.7165721654891968,
      "step": 1907
    },
    {
      "epoch": 0.3816,
      "grad_norm": 0.5461950898170471,
      "learning_rate": 7.775674002060905e-05,
      "loss": 0.2862,
      "step": 1908
    },
    {
      "epoch": 0.3816,
      "loss_align": 0.13815104961395264,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0019523618975654244,
      "neg_sim_mean": 0.5026402473449707,
      "pos_sim_mean": 0.8618489503860474,
      "step": 1908
    },
    {
      "epoch": 0.3818,
      "grad_norm": 0.48860886693000793,
      "learning_rate": 7.772769934388537e-05,
      "loss": 0.1401,
      "step": 1909
    },
    {
      "epoch": 0.3818,
      "loss_align": 0.35951632261276245,
      "loss_contrastive": 0.0868341326713562,
      "loss_gen": 0.23245957493782043,
      "neg_sim_mean": 0.6273177862167358,
      "pos_sim_mean": 0.6404836773872375,
      "step": 1909
    },
    {
      "epoch": 0.382,
      "grad_norm": 1.702452540397644,
      "learning_rate": 7.769864515301787e-05,
      "loss": 0.6024,
      "step": 1910
    },
    {
      "epoch": 0.382,
      "loss_align": 0.23456788063049316,
      "loss_contrastive": 0.05826282501220703,
      "loss_gen": 0.20076307654380798,
      "neg_sim_mean": 0.723694920539856,
      "pos_sim_mean": 0.7654321193695068,
      "step": 1910
    },
    {
      "epoch": 0.3822,
      "grad_norm": 1.7810797691345215,
      "learning_rate": 7.766957746216721e-05,
      "loss": 0.4423,
      "step": 1911
    },
    {
      "epoch": 0.3822,
      "loss_align": 0.17906224727630615,
      "loss_contrastive": 0.1662774682044983,
      "loss_gen": 0.2220892608165741,
      "neg_sim_mean": 0.8872151970863342,
      "pos_sim_mean": 0.8209377527236938,
      "step": 1911
    },
    {
      "epoch": 0.3824,
      "grad_norm": 1.7975573539733887,
      "learning_rate": 7.764049628550063e-05,
      "loss": 0.4211,
      "step": 1912
    },
    {
      "epoch": 0.3824,
      "loss_align": 0.17398971319198608,
      "loss_contrastive": 0.024602174758911133,
      "loss_gen": 0.2613906264305115,
      "neg_sim_mean": 0.7506124377250671,
      "pos_sim_mean": 0.8260102868080139,
      "step": 1912
    },
    {
      "epoch": 0.3826,
      "grad_norm": 1.7969367504119873,
      "learning_rate": 7.761140163719194e-05,
      "loss": 0.4383,
      "step": 1913
    },
    {
      "epoch": 0.3826,
      "loss_align": 0.26718032360076904,
      "loss_contrastive": 0.02607107162475586,
      "loss_gen": 0.323693186044693,
      "neg_sim_mean": 0.6588907241821289,
      "pos_sim_mean": 0.732819676399231,
      "step": 1913
    },
    {
      "epoch": 0.3828,
      "grad_norm": 2.2983345985412598,
      "learning_rate": 7.758229353142152e-05,
      "loss": 0.594,
      "step": 1914
    },
    {
      "epoch": 0.3828,
      "loss_align": 0.22237813472747803,
      "loss_contrastive": 0.13150066137313843,
      "loss_gen": 0.3095700442790985,
      "neg_sim_mean": 0.8091225028038025,
      "pos_sim_mean": 0.777621865272522,
      "step": 1914
    },
    {
      "epoch": 0.383,
      "grad_norm": 2.132195234298706,
      "learning_rate": 7.75531719823763e-05,
      "loss": 0.5477,
      "step": 1915
    },
    {
      "epoch": 0.383,
      "loss_align": 0.17031383514404297,
      "loss_contrastive": 0.0,
      "loss_gen": 0.33587127923965454,
      "neg_sim_mean": 0.674385666847229,
      "pos_sim_mean": 0.829686164855957,
      "step": 1915
    },
    {
      "epoch": 0.3832,
      "grad_norm": 1.7726424932479858,
      "learning_rate": 7.752403700424979e-05,
      "loss": 0.5062,
      "step": 1916
    },
    {
      "epoch": 0.3832,
      "loss_align": 0.31083881855010986,
      "loss_contrastive": 0.11398929357528687,
      "loss_gen": 0.4814998507499695,
      "neg_sim_mean": 0.7031504511833191,
      "pos_sim_mean": 0.6891611814498901,
      "step": 1916
    },
    {
      "epoch": 0.3834,
      "grad_norm": 2.163426637649536,
      "learning_rate": 7.7494888611242e-05,
      "loss": 0.806,
      "step": 1917
    },
    {
      "epoch": 0.3834,
      "loss_align": 0.14813649654388428,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01137454528361559,
      "neg_sim_mean": 0.34965699911117554,
      "pos_sim_mean": 0.8518635034561157,
      "step": 1917
    },
    {
      "epoch": 0.3836,
      "grad_norm": 0.7593345642089844,
      "learning_rate": 7.74657268175595e-05,
      "loss": 0.1595,
      "step": 1918
    },
    {
      "epoch": 0.3836,
      "loss_align": 0.22310912609100342,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007199198938906193,
      "neg_sim_mean": 0.21695174276828766,
      "pos_sim_mean": 0.7768908739089966,
      "step": 1918
    },
    {
      "epoch": 0.3838,
      "grad_norm": 0.6891863346099854,
      "learning_rate": 7.743655163741543e-05,
      "loss": 0.2303,
      "step": 1919
    },
    {
      "epoch": 0.3838,
      "loss_align": 0.10740005970001221,
      "loss_contrastive": 0.014147341251373291,
      "loss_gen": 0.30178770422935486,
      "neg_sim_mean": 0.8067472577095032,
      "pos_sim_mean": 0.8925999402999878,
      "step": 1919
    },
    {
      "epoch": 0.384,
      "grad_norm": 1.8222596645355225,
      "learning_rate": 7.740736308502938e-05,
      "loss": 0.4109,
      "step": 1920
    },
    {
      "epoch": 0.384,
      "loss_align": 0.18603920936584473,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015929180663079023,
      "neg_sim_mean": 0.2304525226354599,
      "pos_sim_mean": 0.8139607906341553,
      "step": 1920
    },
    {
      "epoch": 0.3842,
      "grad_norm": 0.4727036952972412,
      "learning_rate": 7.737816117462752e-05,
      "loss": 0.1876,
      "step": 1921
    },
    {
      "epoch": 0.3842,
      "loss_align": 0.18779516220092773,
      "loss_contrastive": 0.051236510276794434,
      "loss_gen": 0.2607313096523285,
      "neg_sim_mean": 0.7634413242340088,
      "pos_sim_mean": 0.8122048377990723,
      "step": 1921
    },
    {
      "epoch": 0.3844,
      "grad_norm": 1.6885530948638916,
      "learning_rate": 7.734894592044248e-05,
      "loss": 0.4547,
      "step": 1922
    },
    {
      "epoch": 0.3844,
      "loss_align": 0.22028577327728271,
      "loss_contrastive": 0.08386063575744629,
      "loss_gen": 0.12838178873062134,
      "neg_sim_mean": 0.7635748386383057,
      "pos_sim_mean": 0.7797142267227173,
      "step": 1922
    },
    {
      "epoch": 0.3846,
      "grad_norm": 1.5976665019989014,
      "learning_rate": 7.731971733671346e-05,
      "loss": 0.3587,
      "step": 1923
    },
    {
      "epoch": 0.3846,
      "loss_align": 0.1816871166229248,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2432197481393814,
      "neg_sim_mean": 0.6288161873817444,
      "pos_sim_mean": 0.8183128833770752,
      "step": 1923
    },
    {
      "epoch": 0.3848,
      "grad_norm": 1.8140321969985962,
      "learning_rate": 7.729047543768609e-05,
      "loss": 0.4249,
      "step": 1924
    },
    {
      "epoch": 0.3848,
      "loss_align": 0.25584208965301514,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20843203365802765,
      "neg_sim_mean": 0.4793843626976013,
      "pos_sim_mean": 0.7441579103469849,
      "step": 1924
    },
    {
      "epoch": 0.385,
      "grad_norm": 2.759136199951172,
      "learning_rate": 7.726122023761251e-05,
      "loss": 0.4643,
      "step": 1925
    },
    {
      "epoch": 0.385,
      "loss_align": 0.148645281791687,
      "loss_contrastive": 0.03780066967010498,
      "loss_gen": 0.3220762312412262,
      "neg_sim_mean": 0.7891553640365601,
      "pos_sim_mean": 0.851354718208313,
      "step": 1925
    },
    {
      "epoch": 0.3852,
      "grad_norm": 2.169638156890869,
      "learning_rate": 7.723195175075136e-05,
      "loss": 0.4753,
      "step": 1926
    },
    {
      "epoch": 0.3852,
      "loss_align": 0.29480159282684326,
      "loss_contrastive": 0.16716182231903076,
      "loss_gen": 0.27841717004776,
      "neg_sim_mean": 0.7723602056503296,
      "pos_sim_mean": 0.7051984071731567,
      "step": 1926
    },
    {
      "epoch": 0.3854,
      "grad_norm": 2.0056233406066895,
      "learning_rate": 7.720266999136773e-05,
      "loss": 0.5933,
      "step": 1927
    },
    {
      "epoch": 0.3854,
      "loss_align": 0.3606605529785156,
      "loss_contrastive": 0.2367156744003296,
      "loss_gen": 0.3321572542190552,
      "neg_sim_mean": 0.776055097579956,
      "pos_sim_mean": 0.6393394470214844,
      "step": 1927
    },
    {
      "epoch": 0.3856,
      "grad_norm": 2.102766752243042,
      "learning_rate": 7.717337497373324e-05,
      "loss": 0.7212,
      "step": 1928
    },
    {
      "epoch": 0.3856,
      "loss_align": 0.12658607959747314,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21886353194713593,
      "neg_sim_mean": 0.7661798000335693,
      "pos_sim_mean": 0.8734139204025269,
      "step": 1928
    },
    {
      "epoch": 0.3858,
      "grad_norm": 1.525866150856018,
      "learning_rate": 7.714406671212589e-05,
      "loss": 0.3454,
      "step": 1929
    },
    {
      "epoch": 0.3858,
      "loss_align": 0.1012732982635498,
      "loss_contrastive": 0.06173980236053467,
      "loss_gen": 0.2064107209444046,
      "neg_sim_mean": 0.860466480255127,
      "pos_sim_mean": 0.8987267017364502,
      "step": 1929
    },
    {
      "epoch": 0.386,
      "grad_norm": 1.6149846315383911,
      "learning_rate": 7.711474522083015e-05,
      "loss": 0.3151,
      "step": 1930
    },
    {
      "epoch": 0.386,
      "loss_align": 0.21210813522338867,
      "loss_contrastive": 0.09665179252624512,
      "loss_gen": 0.19725492596626282,
      "neg_sim_mean": 0.7845436334609985,
      "pos_sim_mean": 0.7878918647766113,
      "step": 1930
    },
    {
      "epoch": 0.3862,
      "grad_norm": 1.5396692752838135,
      "learning_rate": 7.7085410514137e-05,
      "loss": 0.421,
      "step": 1931
    },
    {
      "epoch": 0.3862,
      "loss_align": 0.12054789066314697,
      "loss_contrastive": 0.11244940757751465,
      "loss_gen": 0.19220581650733948,
      "neg_sim_mean": 0.8919014930725098,
      "pos_sim_mean": 0.879452109336853,
      "step": 1931
    },
    {
      "epoch": 0.3864,
      "grad_norm": 1.7944321632385254,
      "learning_rate": 7.705606260634379e-05,
      "loss": 0.3262,
      "step": 1932
    },
    {
      "epoch": 0.3864,
      "loss_align": 0.16417860984802246,
      "loss_contrastive": 0.0,
      "loss_gen": 0.027415726333856583,
      "neg_sim_mean": 0.4835830330848694,
      "pos_sim_mean": 0.8358213901519775,
      "step": 1932
    },
    {
      "epoch": 0.3866,
      "grad_norm": 1.1194868087768555,
      "learning_rate": 7.702670151175436e-05,
      "loss": 0.1916,
      "step": 1933
    },
    {
      "epoch": 0.3866,
      "loss_align": 0.23967456817626953,
      "loss_contrastive": 0.07515907287597656,
      "loss_gen": 0.2710065245628357,
      "neg_sim_mean": 0.7354844808578491,
      "pos_sim_mean": 0.7603254318237305,
      "step": 1933
    },
    {
      "epoch": 0.3868,
      "grad_norm": 1.9063690900802612,
      "learning_rate": 7.699732724467894e-05,
      "loss": 0.5197,
      "step": 1934
    },
    {
      "epoch": 0.3868,
      "loss_align": 0.1694786548614502,
      "loss_contrastive": 0.007709085941314697,
      "loss_gen": 0.2260298877954483,
      "neg_sim_mean": 0.7382304072380066,
      "pos_sim_mean": 0.8305213451385498,
      "step": 1934
    },
    {
      "epoch": 0.387,
      "grad_norm": 1.9449974298477173,
      "learning_rate": 7.696793981943417e-05,
      "loss": 0.3964,
      "step": 1935
    },
    {
      "epoch": 0.387,
      "loss_align": 0.13448834419250488,
      "loss_contrastive": 0.10729622840881348,
      "loss_gen": 0.19610708951950073,
      "neg_sim_mean": 0.8728078603744507,
      "pos_sim_mean": 0.8655116558074951,
      "step": 1935
    },
    {
      "epoch": 0.3872,
      "grad_norm": 1.72786283493042,
      "learning_rate": 7.693853925034315e-05,
      "loss": 0.3435,
      "step": 1936
    },
    {
      "epoch": 0.3872,
      "loss_align": 0.13551008701324463,
      "loss_contrastive": 0.04073810577392578,
      "loss_gen": 0.2581043541431427,
      "neg_sim_mean": 0.8052279949188232,
      "pos_sim_mean": 0.8644899129867554,
      "step": 1936
    },
    {
      "epoch": 0.3874,
      "grad_norm": 1.923495888710022,
      "learning_rate": 7.690912555173537e-05,
      "loss": 0.3985,
      "step": 1937
    },
    {
      "epoch": 0.3874,
      "loss_align": 0.2684285044670105,
      "loss_contrastive": 0.2198934555053711,
      "loss_gen": 0.5527350306510925,
      "neg_sim_mean": 0.8514649271965027,
      "pos_sim_mean": 0.7315714955329895,
      "step": 1937
    },
    {
      "epoch": 0.3876,
      "grad_norm": 2.6491799354553223,
      "learning_rate": 7.687969873794667e-05,
      "loss": 0.8476,
      "step": 1938
    },
    {
      "epoch": 0.3876,
      "loss_align": 0.26981091499328613,
      "loss_contrastive": 0.07513082027435303,
      "loss_gen": 0.3953418433666229,
      "neg_sim_mean": 0.705319881439209,
      "pos_sim_mean": 0.7301890850067139,
      "step": 1938
    },
    {
      "epoch": 0.3878,
      "grad_norm": 2.4529471397399902,
      "learning_rate": 7.685025882331936e-05,
      "loss": 0.6742,
      "step": 1939
    },
    {
      "epoch": 0.3878,
      "loss_align": 0.17812734842300415,
      "loss_contrastive": 0.15459716320037842,
      "loss_gen": 0.348511278629303,
      "neg_sim_mean": 0.8764697909355164,
      "pos_sim_mean": 0.8218726515769958,
      "step": 1939
    },
    {
      "epoch": 0.388,
      "grad_norm": 2.505741596221924,
      "learning_rate": 7.682080582220206e-05,
      "loss": 0.5452,
      "step": 1940
    },
    {
      "epoch": 0.388,
      "loss_align": 0.1980201005935669,
      "loss_contrastive": 0.1267595887184143,
      "loss_gen": 0.41487905383110046,
      "neg_sim_mean": 0.8287394642829895,
      "pos_sim_mean": 0.8019798994064331,
      "step": 1940
    },
    {
      "epoch": 0.3882,
      "grad_norm": 2.312455654144287,
      "learning_rate": 7.679133974894983e-05,
      "loss": 0.6281,
      "step": 1941
    },
    {
      "epoch": 0.3882,
      "loss_align": 0.12859821319580078,
      "loss_contrastive": 0.10853254795074463,
      "loss_gen": 0.166835218667984,
      "neg_sim_mean": 0.8799343109130859,
      "pos_sim_mean": 0.8714017868041992,
      "step": 1941
    },
    {
      "epoch": 0.3884,
      "grad_norm": 1.6724467277526855,
      "learning_rate": 7.676186061792407e-05,
      "loss": 0.3085,
      "step": 1942
    },
    {
      "epoch": 0.3884,
      "loss_align": 0.25904786586761475,
      "loss_contrastive": 0.09672772884368896,
      "loss_gen": 0.3730085790157318,
      "neg_sim_mean": 0.7376798391342163,
      "pos_sim_mean": 0.7409521341323853,
      "step": 1942
    },
    {
      "epoch": 0.3886,
      "grad_norm": 2.2003579139709473,
      "learning_rate": 7.673236844349256e-05,
      "loss": 0.6437,
      "step": 1943
    },
    {
      "epoch": 0.3886,
      "loss_align": 0.17449069023132324,
      "loss_contrastive": 0.05837714672088623,
      "loss_gen": 0.1426829844713211,
      "neg_sim_mean": 0.7838864326477051,
      "pos_sim_mean": 0.8255093097686768,
      "step": 1943
    },
    {
      "epoch": 0.3888,
      "grad_norm": 1.640709638595581,
      "learning_rate": 7.670286324002944e-05,
      "loss": 0.3242,
      "step": 1944
    },
    {
      "epoch": 0.3888,
      "loss_align": 0.19154983758926392,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23456747829914093,
      "neg_sim_mean": 0.6749454140663147,
      "pos_sim_mean": 0.8084501624107361,
      "step": 1944
    },
    {
      "epoch": 0.389,
      "grad_norm": 1.682354211807251,
      "learning_rate": 7.667334502191515e-05,
      "loss": 0.4261,
      "step": 1945
    },
    {
      "epoch": 0.389,
      "loss_align": 0.18249547481536865,
      "loss_contrastive": 0.07553184032440186,
      "loss_gen": 0.3923287093639374,
      "neg_sim_mean": 0.7930363416671753,
      "pos_sim_mean": 0.8175045251846313,
      "step": 1945
    },
    {
      "epoch": 0.3892,
      "grad_norm": 2.4480578899383545,
      "learning_rate": 7.66438138035365e-05,
      "loss": 0.5839,
      "step": 1946
    },
    {
      "epoch": 0.3892,
      "loss_align": 0.1939377784729004,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002715855138376355,
      "neg_sim_mean": 0.3324718773365021,
      "pos_sim_mean": 0.8060622215270996,
      "step": 1946
    },
    {
      "epoch": 0.3894,
      "grad_norm": 0.5107614398002625,
      "learning_rate": 7.66142695992867e-05,
      "loss": 0.1967,
      "step": 1947
    },
    {
      "epoch": 0.3894,
      "loss_align": 0.2730295658111572,
      "loss_contrastive": 0.16410565376281738,
      "loss_gen": 0.344460129737854,
      "neg_sim_mean": 0.7910760641098022,
      "pos_sim_mean": 0.7269704341888428,
      "step": 1947
    },
    {
      "epoch": 0.3896,
      "grad_norm": 2.4740912914276123,
      "learning_rate": 7.65847124235652e-05,
      "loss": 0.6372,
      "step": 1948
    },
    {
      "epoch": 0.3896,
      "loss_align": 0.33721303939819336,
      "loss_contrastive": 0.2916000485420227,
      "loss_gen": 0.45865872502326965,
      "neg_sim_mean": 0.8543869853019714,
      "pos_sim_mean": 0.6627869606018066,
      "step": 1948
    },
    {
      "epoch": 0.3898,
      "grad_norm": 2.460341215133667,
      "learning_rate": 7.655514229077784e-05,
      "loss": 0.8309,
      "step": 1949
    },
    {
      "epoch": 0.3898,
      "loss_align": 0.1439495086669922,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00838866364210844,
      "neg_sim_mean": 0.40929532051086426,
      "pos_sim_mean": 0.8560504913330078,
      "step": 1949
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.7622932195663452,
      "learning_rate": 7.65255592153367e-05,
      "loss": 0.1523,
      "step": 1950
    },
    {
      "epoch": 0.39,
      "loss_align": 0.20010173320770264,
      "loss_contrastive": 0.08840936422348022,
      "loss_gen": 0.22865073382854462,
      "neg_sim_mean": 0.7883076071739197,
      "pos_sim_mean": 0.7998982667922974,
      "step": 1950
    },
    {
      "epoch": 0.3902,
      "grad_norm": 1.7744957208633423,
      "learning_rate": 7.649596321166024e-05,
      "loss": 0.4394,
      "step": 1951
    },
    {
      "epoch": 0.3902,
      "loss_align": 0.12692618370056152,
      "loss_contrastive": 0.12910348176956177,
      "loss_gen": 0.4868432283401489,
      "neg_sim_mean": 0.9021772742271423,
      "pos_sim_mean": 0.8730738162994385,
      "step": 1951
    },
    {
      "epoch": 0.3904,
      "grad_norm": 2.6477246284484863,
      "learning_rate": 7.646635429417322e-05,
      "loss": 0.6293,
      "step": 1952
    },
    {
      "epoch": 0.3904,
      "loss_align": 0.36610305309295654,
      "loss_contrastive": 0.20766615867614746,
      "loss_gen": 0.1836560219526291,
      "neg_sim_mean": 0.741563081741333,
      "pos_sim_mean": 0.6338969469070435,
      "step": 1952
    },
    {
      "epoch": 0.3906,
      "grad_norm": 2.044857978820801,
      "learning_rate": 7.643673247730658e-05,
      "loss": 0.5747,
      "step": 1953
    },
    {
      "epoch": 0.3906,
      "loss_align": 0.1660398244857788,
      "loss_contrastive": 0.0,
      "loss_gen": 0.022852318361401558,
      "neg_sim_mean": 0.36719852685928345,
      "pos_sim_mean": 0.8339601755142212,
      "step": 1953
    },
    {
      "epoch": 0.3908,
      "grad_norm": 1.3344712257385254,
      "learning_rate": 7.640709777549773e-05,
      "loss": 0.1889,
      "step": 1954
    },
    {
      "epoch": 0.3908,
      "loss_align": 0.27609121799468994,
      "loss_contrastive": 0.005199253559112549,
      "loss_gen": 0.07574676722288132,
      "neg_sim_mean": 0.6291080117225647,
      "pos_sim_mean": 0.7239087820053101,
      "step": 1954
    },
    {
      "epoch": 0.391,
      "grad_norm": 1.1971125602722168,
      "learning_rate": 7.637745020319019e-05,
      "loss": 0.3525,
      "step": 1955
    },
    {
      "epoch": 0.391,
      "loss_align": 0.1915132999420166,
      "loss_contrastive": 0.09764289855957031,
      "loss_gen": 0.6810727119445801,
      "neg_sim_mean": 0.8061295747756958,
      "pos_sim_mean": 0.8084867000579834,
      "step": 1955
    },
    {
      "epoch": 0.3912,
      "grad_norm": 3.0455105304718018,
      "learning_rate": 7.634778977483389e-05,
      "loss": 0.8843,
      "step": 1956
    },
    {
      "epoch": 0.3912,
      "loss_align": 0.20455002784729004,
      "loss_contrastive": 0.11453014612197876,
      "loss_gen": 0.4531581997871399,
      "neg_sim_mean": 0.8099800944328308,
      "pos_sim_mean": 0.79544997215271,
      "step": 1956
    },
    {
      "epoch": 0.3914,
      "grad_norm": 2.431882381439209,
      "learning_rate": 7.631811650488489e-05,
      "loss": 0.6715,
      "step": 1957
    },
    {
      "epoch": 0.3914,
      "loss_align": 0.17170369625091553,
      "loss_contrastive": 0.13357830047607422,
      "loss_gen": 0.44372668862342834,
      "neg_sim_mean": 0.8618745803833008,
      "pos_sim_mean": 0.8282963037490845,
      "step": 1957
    },
    {
      "epoch": 0.3916,
      "grad_norm": 2.3147501945495605,
      "learning_rate": 7.628843040780567e-05,
      "loss": 0.6315,
      "step": 1958
    },
    {
      "epoch": 0.3916,
      "loss_align": 0.24316036701202393,
      "loss_contrastive": 0.030064821243286133,
      "loss_gen": 0.15641698241233826,
      "neg_sim_mean": 0.6869044303894043,
      "pos_sim_mean": 0.7568396329879761,
      "step": 1958
    },
    {
      "epoch": 0.3918,
      "grad_norm": 1.744675636291504,
      "learning_rate": 7.62587314980648e-05,
      "loss": 0.4032,
      "step": 1959
    },
    {
      "epoch": 0.3918,
      "loss_align": 0.2367544174194336,
      "loss_contrastive": 0.0710555911064148,
      "loss_gen": 0.3428444266319275,
      "neg_sim_mean": 0.7343011498451233,
      "pos_sim_mean": 0.7632455825805664,
      "step": 1959
    },
    {
      "epoch": 0.392,
      "grad_norm": 2.289303779602051,
      "learning_rate": 7.622901979013717e-05,
      "loss": 0.5881,
      "step": 1960
    },
    {
      "epoch": 0.392,
      "loss_align": 0.11670160293579102,
      "loss_contrastive": 0.006627142429351807,
      "loss_gen": 0.2755291163921356,
      "neg_sim_mean": 0.7899255156517029,
      "pos_sim_mean": 0.883298397064209,
      "step": 1960
    },
    {
      "epoch": 0.3922,
      "grad_norm": 2.4091877937316895,
      "learning_rate": 7.619929529850397e-05,
      "loss": 0.393,
      "step": 1961
    },
    {
      "epoch": 0.3922,
      "loss_align": 0.24082183837890625,
      "loss_contrastive": 0.14989042282104492,
      "loss_gen": 0.3156662583351135,
      "neg_sim_mean": 0.8090685606002808,
      "pos_sim_mean": 0.7591781616210938,
      "step": 1961
    },
    {
      "epoch": 0.3924,
      "grad_norm": 2.1315205097198486,
      "learning_rate": 7.616955803765249e-05,
      "loss": 0.5745,
      "step": 1962
    },
    {
      "epoch": 0.3924,
      "loss_align": 0.28809654712677,
      "loss_contrastive": 0.05529356002807617,
      "loss_gen": 0.24255892634391785,
      "neg_sim_mean": 0.6671969890594482,
      "pos_sim_mean": 0.71190345287323,
      "step": 1962
    },
    {
      "epoch": 0.3926,
      "grad_norm": 2.0127193927764893,
      "learning_rate": 7.613980802207633e-05,
      "loss": 0.5373,
      "step": 1963
    },
    {
      "epoch": 0.3926,
      "loss_align": 0.10008949041366577,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0012043805327266455,
      "neg_sim_mean": 0.32476937770843506,
      "pos_sim_mean": 0.8999105095863342,
      "step": 1963
    },
    {
      "epoch": 0.3928,
      "grad_norm": 0.2958042621612549,
      "learning_rate": 7.61100452662753e-05,
      "loss": 0.1013,
      "step": 1964
    },
    {
      "epoch": 0.3928,
      "loss_align": 0.20468294620513916,
      "loss_contrastive": 0.006834149360656738,
      "loss_gen": 0.2599714398384094,
      "neg_sim_mean": 0.7021511793136597,
      "pos_sim_mean": 0.7953170537948608,
      "step": 1964
    },
    {
      "epoch": 0.393,
      "grad_norm": 1.9616754055023193,
      "learning_rate": 7.60802697847554e-05,
      "loss": 0.4655,
      "step": 1965
    },
    {
      "epoch": 0.393,
      "loss_align": 0.2142575979232788,
      "loss_contrastive": 0.05661153793334961,
      "loss_gen": 0.3282117545604706,
      "neg_sim_mean": 0.7423539161682129,
      "pos_sim_mean": 0.7857424020767212,
      "step": 1965
    },
    {
      "epoch": 0.3932,
      "grad_norm": 2.208322763442993,
      "learning_rate": 7.605048159202883e-05,
      "loss": 0.5493,
      "step": 1966
    },
    {
      "epoch": 0.3932,
      "loss_align": 0.08977007865905762,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0018419971456751227,
      "neg_sim_mean": 0.335318922996521,
      "pos_sim_mean": 0.9102299213409424,
      "step": 1966
    },
    {
      "epoch": 0.3934,
      "grad_norm": 0.3231896758079529,
      "learning_rate": 7.602068070261399e-05,
      "loss": 0.0916,
      "step": 1967
    },
    {
      "epoch": 0.3934,
      "loss_align": 0.13586020469665527,
      "loss_contrastive": 0.046034038066864014,
      "loss_gen": 0.2342475801706314,
      "neg_sim_mean": 0.8101738095283508,
      "pos_sim_mean": 0.8641397953033447,
      "step": 1967
    },
    {
      "epoch": 0.3936,
      "grad_norm": 1.714193344116211,
      "learning_rate": 7.599086713103547e-05,
      "loss": 0.3756,
      "step": 1968
    },
    {
      "epoch": 0.3936,
      "loss_align": 0.31088340282440186,
      "loss_contrastive": 0.09264427423477173,
      "loss_gen": 0.13112926483154297,
      "neg_sim_mean": 0.681760847568512,
      "pos_sim_mean": 0.6891165971755981,
      "step": 1968
    },
    {
      "epoch": 0.3938,
      "grad_norm": 1.7190730571746826,
      "learning_rate": 7.596104089182408e-05,
      "loss": 0.4531,
      "step": 1969
    },
    {
      "epoch": 0.3938,
      "loss_align": 0.07760381698608398,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3114003837108612,
      "neg_sim_mean": 0.7490525841712952,
      "pos_sim_mean": 0.922396183013916,
      "step": 1969
    },
    {
      "epoch": 0.394,
      "grad_norm": 1.749712347984314,
      "learning_rate": 7.593120199951671e-05,
      "loss": 0.389,
      "step": 1970
    },
    {
      "epoch": 0.394,
      "loss_align": 0.15517258644104004,
      "loss_contrastive": 0.09891253709793091,
      "loss_gen": 0.21130450069904327,
      "neg_sim_mean": 0.843739926815033,
      "pos_sim_mean": 0.84482741355896,
      "step": 1970
    },
    {
      "epoch": 0.3942,
      "grad_norm": 1.7526087760925293,
      "learning_rate": 7.590135046865651e-05,
      "loss": 0.3783,
      "step": 1971
    },
    {
      "epoch": 0.3942,
      "loss_align": 0.04379183053970337,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2541551887989044,
      "neg_sim_mean": 0.8215116262435913,
      "pos_sim_mean": 0.9562081694602966,
      "step": 1971
    },
    {
      "epoch": 0.3944,
      "grad_norm": 1.9904065132141113,
      "learning_rate": 7.587148631379275e-05,
      "loss": 0.2979,
      "step": 1972
    },
    {
      "epoch": 0.3944,
      "loss_align": 0.1997596025466919,
      "loss_contrastive": 0.10558044910430908,
      "loss_gen": 0.3842889964580536,
      "neg_sim_mean": 0.8058208227157593,
      "pos_sim_mean": 0.8002403974533081,
      "step": 1972
    },
    {
      "epoch": 0.3946,
      "grad_norm": 2.0823845863342285,
      "learning_rate": 7.584160954948086e-05,
      "loss": 0.5967,
      "step": 1973
    },
    {
      "epoch": 0.3946,
      "loss_align": 0.19210600852966309,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2892533540725708,
      "neg_sim_mean": 0.6834781765937805,
      "pos_sim_mean": 0.8078939914703369,
      "step": 1973
    },
    {
      "epoch": 0.3948,
      "grad_norm": 1.899411916732788,
      "learning_rate": 7.581172019028238e-05,
      "loss": 0.4814,
      "step": 1974
    },
    {
      "epoch": 0.3948,
      "loss_align": 0.08392548561096191,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3189861476421356,
      "neg_sim_mean": 0.6234904527664185,
      "pos_sim_mean": 0.9160745143890381,
      "step": 1974
    },
    {
      "epoch": 0.395,
      "grad_norm": 2.145831823348999,
      "learning_rate": 7.578181825076506e-05,
      "loss": 0.4029,
      "step": 1975
    },
    {
      "epoch": 0.395,
      "loss_align": 0.1311037540435791,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0685625672340393,
      "neg_sim_mean": 0.370002418756485,
      "pos_sim_mean": 0.8688962459564209,
      "step": 1975
    },
    {
      "epoch": 0.3952,
      "grad_norm": 2.1235733032226562,
      "learning_rate": 7.575190374550272e-05,
      "loss": 0.1997,
      "step": 1976
    },
    {
      "epoch": 0.3952,
      "loss_align": 0.18236184120178223,
      "loss_contrastive": 0.005743861198425293,
      "loss_gen": 0.23311111330986023,
      "neg_sim_mean": 0.7233819961547852,
      "pos_sim_mean": 0.8176381587982178,
      "step": 1976
    },
    {
      "epoch": 0.3954,
      "grad_norm": 1.7679396867752075,
      "learning_rate": 7.572197668907532e-05,
      "loss": 0.4162,
      "step": 1977
    },
    {
      "epoch": 0.3954,
      "loss_align": 0.1979914903640747,
      "loss_contrastive": 0.08851301670074463,
      "loss_gen": 0.30784872174263,
      "neg_sim_mean": 0.790521502494812,
      "pos_sim_mean": 0.8020085096359253,
      "step": 1977
    },
    {
      "epoch": 0.3956,
      "grad_norm": 1.7535675764083862,
      "learning_rate": 7.569203709606898e-05,
      "loss": 0.5165,
      "step": 1978
    },
    {
      "epoch": 0.3956,
      "loss_align": 0.22267788648605347,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16176250576972961,
      "neg_sim_mean": 0.6152085065841675,
      "pos_sim_mean": 0.7773221135139465,
      "step": 1978
    },
    {
      "epoch": 0.3958,
      "grad_norm": 1.607372760772705,
      "learning_rate": 7.566208498107585e-05,
      "loss": 0.3844,
      "step": 1979
    },
    {
      "epoch": 0.3958,
      "loss_align": 0.17644500732421875,
      "loss_contrastive": 0.14470887184143066,
      "loss_gen": 0.1938781887292862,
      "neg_sim_mean": 0.868263840675354,
      "pos_sim_mean": 0.8235549926757812,
      "step": 1979
    },
    {
      "epoch": 0.396,
      "grad_norm": 1.5333034992218018,
      "learning_rate": 7.563212035869425e-05,
      "loss": 0.3877,
      "step": 1980
    },
    {
      "epoch": 0.396,
      "loss_align": 0.24040865898132324,
      "loss_contrastive": 0.1336759328842163,
      "loss_gen": 0.1407010555267334,
      "neg_sim_mean": 0.7932672500610352,
      "pos_sim_mean": 0.7595913410186768,
      "step": 1980
    },
    {
      "epoch": 0.3962,
      "grad_norm": 1.662361979484558,
      "learning_rate": 7.560214324352858e-05,
      "loss": 0.3972,
      "step": 1981
    },
    {
      "epoch": 0.3962,
      "loss_align": 0.17749303579330444,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2504741847515106,
      "neg_sim_mean": 0.6933271884918213,
      "pos_sim_mean": 0.8225069642066956,
      "step": 1981
    },
    {
      "epoch": 0.3964,
      "grad_norm": 2.7531938552856445,
      "learning_rate": 7.55721536501893e-05,
      "loss": 0.428,
      "step": 1982
    },
    {
      "epoch": 0.3964,
      "loss_align": 0.1838669776916504,
      "loss_contrastive": 0.05391061305999756,
      "loss_gen": 0.3521873354911804,
      "neg_sim_mean": 0.7700436115264893,
      "pos_sim_mean": 0.8161330223083496,
      "step": 1982
    },
    {
      "epoch": 0.3966,
      "grad_norm": 2.1472535133361816,
      "learning_rate": 7.5542151593293e-05,
      "loss": 0.5425,
      "step": 1983
    },
    {
      "epoch": 0.3966,
      "loss_align": 0.18234598636627197,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22941365838050842,
      "neg_sim_mean": 0.48916763067245483,
      "pos_sim_mean": 0.817654013633728,
      "step": 1983
    },
    {
      "epoch": 0.3968,
      "grad_norm": 1.6480425596237183,
      "learning_rate": 7.55121370874623e-05,
      "loss": 0.4118,
      "step": 1984
    },
    {
      "epoch": 0.3968,
      "loss_align": 0.16862916946411133,
      "loss_contrastive": 0.10882318019866943,
      "loss_gen": 0.22420687973499298,
      "neg_sim_mean": 0.8401939868927002,
      "pos_sim_mean": 0.8313708305358887,
      "step": 1984
    },
    {
      "epoch": 0.397,
      "grad_norm": 2.222815752029419,
      "learning_rate": 7.548211014732589e-05,
      "loss": 0.4059,
      "step": 1985
    },
    {
      "epoch": 0.397,
      "loss_align": 0.23432695865631104,
      "loss_contrastive": 0.055068016052246094,
      "loss_gen": 0.29449471831321716,
      "neg_sim_mean": 0.7207410335540771,
      "pos_sim_mean": 0.765673041343689,
      "step": 1985
    },
    {
      "epoch": 0.3972,
      "grad_norm": 2.1673262119293213,
      "learning_rate": 7.545207078751857e-05,
      "loss": 0.5354,
      "step": 1986
    },
    {
      "epoch": 0.3972,
      "loss_align": 0.07714641094207764,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2644027769565582,
      "neg_sim_mean": 0.761644721031189,
      "pos_sim_mean": 0.9228535890579224,
      "step": 1986
    },
    {
      "epoch": 0.3974,
      "grad_norm": 1.872213363647461,
      "learning_rate": 7.542201902268115e-05,
      "loss": 0.3415,
      "step": 1987
    },
    {
      "epoch": 0.3974,
      "loss_align": 0.27539652585983276,
      "loss_contrastive": 0.028582990169525146,
      "loss_gen": 0.4993652105331421,
      "neg_sim_mean": 0.6531864404678345,
      "pos_sim_mean": 0.7246034741401672,
      "step": 1987
    },
    {
      "epoch": 0.3976,
      "grad_norm": 2.5051259994506836,
      "learning_rate": 7.539195486746046e-05,
      "loss": 0.7782,
      "step": 1988
    },
    {
      "epoch": 0.3976,
      "loss_align": 0.2684319019317627,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0013553000753745437,
      "neg_sim_mean": 0.374701589345932,
      "pos_sim_mean": 0.7315680980682373,
      "step": 1988
    },
    {
      "epoch": 0.3978,
      "grad_norm": 0.5352356433868408,
      "learning_rate": 7.536187833650947e-05,
      "loss": 0.2698,
      "step": 1989
    },
    {
      "epoch": 0.3978,
      "loss_align": 0.10327130556106567,
      "loss_contrastive": 0.10724902153015137,
      "loss_gen": 0.5774789452552795,
      "neg_sim_mean": 0.9039776921272278,
      "pos_sim_mean": 0.8967286944389343,
      "step": 1989
    },
    {
      "epoch": 0.398,
      "grad_norm": 3.584965705871582,
      "learning_rate": 7.533178944448705e-05,
      "loss": 0.6936,
      "step": 1990
    },
    {
      "epoch": 0.398,
      "loss_align": 0.3096752166748047,
      "loss_contrastive": 0.15256130695343018,
      "loss_gen": 0.22524495422840118,
      "neg_sim_mean": 0.7428860664367676,
      "pos_sim_mean": 0.6903247833251953,
      "step": 1990
    },
    {
      "epoch": 0.3982,
      "grad_norm": 1.9175461530685425,
      "learning_rate": 7.530168820605818e-05,
      "loss": 0.5532,
      "step": 1991
    },
    {
      "epoch": 0.3982,
      "loss_align": 0.21465635299682617,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004261751659214497,
      "neg_sim_mean": 0.4668494164943695,
      "pos_sim_mean": 0.7853436470031738,
      "step": 1991
    },
    {
      "epoch": 0.3984,
      "grad_norm": 0.5126521587371826,
      "learning_rate": 7.52715746358939e-05,
      "loss": 0.2189,
      "step": 1992
    },
    {
      "epoch": 0.3984,
      "loss_align": 0.14769196510314941,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02361033484339714,
      "neg_sim_mean": 0.3580585718154907,
      "pos_sim_mean": 0.8523080348968506,
      "step": 1992
    },
    {
      "epoch": 0.3986,
      "grad_norm": 1.200824499130249,
      "learning_rate": 7.52414487486711e-05,
      "loss": 0.1713,
      "step": 1993
    },
    {
      "epoch": 0.3986,
      "loss_align": 0.15903550386428833,
      "loss_contrastive": 0.11340874433517456,
      "loss_gen": 0.16405433416366577,
      "neg_sim_mean": 0.8543732166290283,
      "pos_sim_mean": 0.8409644961357117,
      "step": 1993
    },
    {
      "epoch": 0.3988,
      "grad_norm": 1.4205758571624756,
      "learning_rate": 7.521131055907282e-05,
      "loss": 0.3367,
      "step": 1994
    },
    {
      "epoch": 0.3988,
      "loss_align": 0.1790705919265747,
      "loss_contrastive": 0.019113481044769287,
      "loss_gen": 0.27947694063186646,
      "neg_sim_mean": 0.7400428652763367,
      "pos_sim_mean": 0.8209294080734253,
      "step": 1994
    },
    {
      "epoch": 0.399,
      "grad_norm": 1.7014658451080322,
      "learning_rate": 7.518116008178805e-05,
      "loss": 0.4608,
      "step": 1995
    },
    {
      "epoch": 0.399,
      "loss_align": 0.1745915412902832,
      "loss_contrastive": 0.13393455743789673,
      "loss_gen": 0.22718556225299835,
      "neg_sim_mean": 0.8593429923057556,
      "pos_sim_mean": 0.8254084587097168,
      "step": 1995
    },
    {
      "epoch": 0.3992,
      "grad_norm": 1.613709807395935,
      "learning_rate": 7.515099733151177e-05,
      "loss": 0.4178,
      "step": 1996
    },
    {
      "epoch": 0.3992,
      "loss_align": 0.12805384397506714,
      "loss_contrastive": 0.1159813404083252,
      "loss_gen": 0.43555566668510437,
      "neg_sim_mean": 0.8879274725914001,
      "pos_sim_mean": 0.8719461560249329,
      "step": 1996
    },
    {
      "epoch": 0.3994,
      "grad_norm": 1.8596140146255493,
      "learning_rate": 7.512082232294491e-05,
      "loss": 0.5775,
      "step": 1997
    },
    {
      "epoch": 0.3994,
      "loss_align": 0.18767857551574707,
      "loss_contrastive": 0.058997273445129395,
      "loss_gen": 0.4190036952495575,
      "neg_sim_mean": 0.7713186740875244,
      "pos_sim_mean": 0.8123214244842529,
      "step": 1997
    },
    {
      "epoch": 0.3996,
      "grad_norm": 2.207597017288208,
      "learning_rate": 7.509063507079443e-05,
      "loss": 0.6138,
      "step": 1998
    },
    {
      "epoch": 0.3996,
      "loss_align": 0.08746039867401123,
      "loss_contrastive": 0.0,
      "loss_gen": 0.037678059190511703,
      "neg_sim_mean": 0.36588671803474426,
      "pos_sim_mean": 0.9125396013259888,
      "step": 1998
    },
    {
      "epoch": 0.3998,
      "grad_norm": 1.1072145700454712,
      "learning_rate": 7.506043558977321e-05,
      "loss": 0.1251,
      "step": 1999
    },
    {
      "epoch": 0.3998,
      "loss_align": 0.12444442510604858,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24666066467761993,
      "neg_sim_mean": 0.7721338272094727,
      "pos_sim_mean": 0.8755555748939514,
      "step": 1999
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.7002286911010742,
      "learning_rate": 7.503022389460014e-05,
      "loss": 0.3711,
      "step": 2000
    },
    {
      "epoch": 0.4,
      "loss_align": 0.17717498540878296,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04672468081116676,
      "neg_sim_mean": 0.3338260054588318,
      "pos_sim_mean": 0.822825014591217,
      "step": 2000
    },
    {
      "epoch": 0.4002,
      "grad_norm": 1.5470269918441772,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.2239,
      "step": 2001
    },
    {
      "epoch": 0.4002,
      "loss_align": 0.19012337923049927,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12186755239963531,
      "neg_sim_mean": 0.5708239078521729,
      "pos_sim_mean": 0.8098766207695007,
      "step": 2001
    },
    {
      "epoch": 0.4004,
      "grad_norm": 1.5428857803344727,
      "learning_rate": 7.496976392070358e-05,
      "loss": 0.312,
      "step": 2002
    },
    {
      "epoch": 0.4004,
      "loss_align": 0.25558483600616455,
      "loss_contrastive": 0.07055801153182983,
      "loss_gen": 0.30209243297576904,
      "neg_sim_mean": 0.7149731516838074,
      "pos_sim_mean": 0.7444151639938354,
      "step": 2002
    },
    {
      "epoch": 0.4006,
      "grad_norm": 1.723357081413269,
      "learning_rate": 7.493951567144754e-05,
      "loss": 0.5661,
      "step": 2003
    },
    {
      "epoch": 0.4006,
      "loss_align": 0.2384876012802124,
      "loss_contrastive": 0.06579697132110596,
      "loss_gen": 0.1807260811328888,
      "neg_sim_mean": 0.7273093461990356,
      "pos_sim_mean": 0.7615123987197876,
      "step": 2003
    },
    {
      "epoch": 0.4008,
      "grad_norm": 1.546653151512146,
      "learning_rate": 7.490925526697455e-05,
      "loss": 0.4271,
      "step": 2004
    },
    {
      "epoch": 0.4008,
      "loss_align": 0.1895134449005127,
      "loss_contrastive": 0.021123766899108887,
      "loss_gen": 0.20611122250556946,
      "neg_sim_mean": 0.7316102981567383,
      "pos_sim_mean": 0.8104865550994873,
      "step": 2004
    },
    {
      "epoch": 0.401,
      "grad_norm": 1.65488862991333,
      "learning_rate": 7.487898272203313e-05,
      "loss": 0.3982,
      "step": 2005
    },
    {
      "epoch": 0.401,
      "loss_align": 0.2052081823348999,
      "loss_contrastive": 0.09625053405761719,
      "loss_gen": 0.3897444009780884,
      "neg_sim_mean": 0.7910423278808594,
      "pos_sim_mean": 0.7947918176651001,
      "step": 2005
    },
    {
      "epoch": 0.4012,
      "grad_norm": 2.6042723655700684,
      "learning_rate": 7.484869805137778e-05,
      "loss": 0.6065,
      "step": 2006
    },
    {
      "epoch": 0.4012,
      "loss_align": 0.16580671072006226,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1453365534543991,
      "neg_sim_mean": 0.5938717126846313,
      "pos_sim_mean": 0.8341932892799377,
      "step": 2006
    },
    {
      "epoch": 0.4014,
      "grad_norm": 1.3507872819900513,
      "learning_rate": 7.481840126976885e-05,
      "loss": 0.3111,
      "step": 2007
    },
    {
      "epoch": 0.4014,
      "loss_align": 0.2248002290725708,
      "loss_contrastive": 0.1686055064201355,
      "loss_gen": 0.1430954486131668,
      "neg_sim_mean": 0.8438052535057068,
      "pos_sim_mean": 0.7751997709274292,
      "step": 2007
    },
    {
      "epoch": 0.4016,
      "grad_norm": 1.3002324104309082,
      "learning_rate": 7.478809239197264e-05,
      "loss": 0.3881,
      "step": 2008
    },
    {
      "epoch": 0.4016,
      "loss_align": 0.367959201335907,
      "loss_contrastive": 0.18590503931045532,
      "loss_gen": 0.23350322246551514,
      "neg_sim_mean": 0.7179458141326904,
      "pos_sim_mean": 0.632040798664093,
      "step": 2008
    },
    {
      "epoch": 0.4018,
      "grad_norm": 1.7921454906463623,
      "learning_rate": 7.475777143276133e-05,
      "loss": 0.6238,
      "step": 2009
    },
    {
      "epoch": 0.4018,
      "loss_align": 0.14811253547668457,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005549708381295204,
      "neg_sim_mean": 0.2450752556324005,
      "pos_sim_mean": 0.8518874645233154,
      "step": 2009
    },
    {
      "epoch": 0.402,
      "grad_norm": 0.6264021992683411,
      "learning_rate": 7.472743840691299e-05,
      "loss": 0.1537,
      "step": 2010
    },
    {
      "epoch": 0.402,
      "loss_align": 0.2166668176651001,
      "loss_contrastive": 0.12510162591934204,
      "loss_gen": 0.21792268753051758,
      "neg_sim_mean": 0.808434784412384,
      "pos_sim_mean": 0.7833331823348999,
      "step": 2010
    },
    {
      "epoch": 0.4022,
      "grad_norm": 1.8179057836532593,
      "learning_rate": 7.469709332921155e-05,
      "loss": 0.4496,
      "step": 2011
    },
    {
      "epoch": 0.4022,
      "loss_align": 0.31867367029190063,
      "loss_contrastive": 0.11680757999420166,
      "loss_gen": 0.3241109251976013,
      "neg_sim_mean": 0.6981338858604431,
      "pos_sim_mean": 0.6813263297080994,
      "step": 2011
    },
    {
      "epoch": 0.4024,
      "grad_norm": 2.400477647781372,
      "learning_rate": 7.466673621444685e-05,
      "loss": 0.6568,
      "step": 2012
    },
    {
      "epoch": 0.4024,
      "loss_align": 0.18789678812026978,
      "loss_contrastive": 0.0,
      "loss_gen": 0.024554524570703506,
      "neg_sim_mean": 0.32123851776123047,
      "pos_sim_mean": 0.8121032118797302,
      "step": 2012
    },
    {
      "epoch": 0.4026,
      "grad_norm": 1.2042416334152222,
      "learning_rate": 7.463636707741458e-05,
      "loss": 0.2125,
      "step": 2013
    },
    {
      "epoch": 0.4026,
      "loss_align": 0.2501784563064575,
      "loss_contrastive": 0.04998290538787842,
      "loss_gen": 0.31991854310035706,
      "neg_sim_mean": 0.699804425239563,
      "pos_sim_mean": 0.7498215436935425,
      "step": 2013
    },
    {
      "epoch": 0.4028,
      "grad_norm": 2.278700351715088,
      "learning_rate": 7.460598593291628e-05,
      "loss": 0.5761,
      "step": 2014
    },
    {
      "epoch": 0.4028,
      "loss_align": 0.25992488861083984,
      "loss_contrastive": 0.24221408367156982,
      "loss_gen": 0.5172268748283386,
      "neg_sim_mean": 0.8822891712188721,
      "pos_sim_mean": 0.7400751113891602,
      "step": 2014
    },
    {
      "epoch": 0.403,
      "grad_norm": 2.47257399559021,
      "learning_rate": 7.457559279575935e-05,
      "loss": 0.8062,
      "step": 2015
    },
    {
      "epoch": 0.403,
      "loss_align": 0.055655717849731445,
      "loss_contrastive": 0.00026476383209228516,
      "loss_gen": 0.2659248113632202,
      "neg_sim_mean": 0.8446090221405029,
      "pos_sim_mean": 0.9443442821502686,
      "step": 2015
    },
    {
      "epoch": 0.4032,
      "grad_norm": 1.7533198595046997,
      "learning_rate": 7.454518768075704e-05,
      "loss": 0.3216,
      "step": 2016
    },
    {
      "epoch": 0.4032,
      "loss_align": 0.1376122236251831,
      "loss_contrastive": 0.035822153091430664,
      "loss_gen": 0.27129802107810974,
      "neg_sim_mean": 0.7982099056243896,
      "pos_sim_mean": 0.8623877763748169,
      "step": 2016
    },
    {
      "epoch": 0.4034,
      "grad_norm": 1.7689992189407349,
      "learning_rate": 7.451477060272844e-05,
      "loss": 0.4132,
      "step": 2017
    },
    {
      "epoch": 0.4034,
      "loss_align": 0.14962661266326904,
      "loss_contrastive": 0.03863823413848877,
      "loss_gen": 0.35003992915153503,
      "neg_sim_mean": 0.7890115976333618,
      "pos_sim_mean": 0.850373387336731,
      "step": 2017
    },
    {
      "epoch": 0.4036,
      "grad_norm": 1.9655271768569946,
      "learning_rate": 7.448434157649846e-05,
      "loss": 0.5043,
      "step": 2018
    },
    {
      "epoch": 0.4036,
      "loss_align": 0.2461690902709961,
      "loss_contrastive": 0.14583063125610352,
      "loss_gen": 0.1350121796131134,
      "neg_sim_mean": 0.7996615171432495,
      "pos_sim_mean": 0.7538309097290039,
      "step": 2018
    },
    {
      "epoch": 0.4038,
      "grad_norm": 1.4442776441574097,
      "learning_rate": 7.445390061689782e-05,
      "loss": 0.3987,
      "step": 2019
    },
    {
      "epoch": 0.4038,
      "loss_align": 0.3855922222137451,
      "loss_contrastive": 0.12361061573028564,
      "loss_gen": 0.2771291434764862,
      "neg_sim_mean": 0.6380183696746826,
      "pos_sim_mean": 0.6144077777862549,
      "step": 2019
    },
    {
      "epoch": 0.404,
      "grad_norm": 2.1193759441375732,
      "learning_rate": 7.44234477387631e-05,
      "loss": 0.6776,
      "step": 2020
    },
    {
      "epoch": 0.404,
      "loss_align": 0.24878370761871338,
      "loss_contrastive": 0.12751376628875732,
      "loss_gen": 0.35628706216812134,
      "neg_sim_mean": 0.778730034828186,
      "pos_sim_mean": 0.7512162923812866,
      "step": 2020
    },
    {
      "epoch": 0.4042,
      "grad_norm": 2.279064655303955,
      "learning_rate": 7.439298295693665e-05,
      "loss": 0.6204,
      "step": 2021
    },
    {
      "epoch": 0.4042,
      "loss_align": 0.20253342390060425,
      "loss_contrastive": 0.049638450145721436,
      "loss_gen": 0.17973344027996063,
      "neg_sim_mean": 0.7471050024032593,
      "pos_sim_mean": 0.7974665760993958,
      "step": 2021
    },
    {
      "epoch": 0.4044,
      "grad_norm": 1.37368643283844,
      "learning_rate": 7.436250628626662e-05,
      "loss": 0.3882,
      "step": 2022
    },
    {
      "epoch": 0.4044,
      "loss_align": 0.19671082496643066,
      "loss_contrastive": 0.0,
      "loss_gen": 0.425650030374527,
      "neg_sim_mean": 0.610601007938385,
      "pos_sim_mean": 0.8032891750335693,
      "step": 2022
    },
    {
      "epoch": 0.4046,
      "grad_norm": 2.572145938873291,
      "learning_rate": 7.4332017741607e-05,
      "loss": 0.6224,
      "step": 2023
    },
    {
      "epoch": 0.4046,
      "loss_align": 0.1800529956817627,
      "loss_contrastive": 0.029246807098388672,
      "loss_gen": 0.28964683413505554,
      "neg_sim_mean": 0.7491937875747681,
      "pos_sim_mean": 0.8199470043182373,
      "step": 2023
    },
    {
      "epoch": 0.4048,
      "grad_norm": 1.6182286739349365,
      "learning_rate": 7.430151733781752e-05,
      "loss": 0.4732,
      "step": 2024
    },
    {
      "epoch": 0.4048,
      "loss_align": 0.1940145492553711,
      "loss_contrastive": 0.11228346824645996,
      "loss_gen": 0.16571882367134094,
      "neg_sim_mean": 0.818268895149231,
      "pos_sim_mean": 0.8059854507446289,
      "step": 2024
    },
    {
      "epoch": 0.405,
      "grad_norm": 1.5683213472366333,
      "learning_rate": 7.42710050897637e-05,
      "loss": 0.3732,
      "step": 2025
    },
    {
      "epoch": 0.405,
      "loss_align": 0.28487682342529297,
      "loss_contrastive": 0.06755006313323975,
      "loss_gen": 0.18720726668834686,
      "neg_sim_mean": 0.6826732158660889,
      "pos_sim_mean": 0.715123176574707,
      "step": 2025
    },
    {
      "epoch": 0.4052,
      "grad_norm": 1.7959885597229004,
      "learning_rate": 7.424048101231686e-05,
      "loss": 0.4802,
      "step": 2026
    },
    {
      "epoch": 0.4052,
      "loss_align": 0.10180222988128662,
      "loss_contrastive": 0.036789119243621826,
      "loss_gen": 0.23981447517871857,
      "neg_sim_mean": 0.8349868655204773,
      "pos_sim_mean": 0.8981977701187134,
      "step": 2026
    },
    {
      "epoch": 0.4054,
      "grad_norm": 1.9448013305664062,
      "learning_rate": 7.420994512035404e-05,
      "loss": 0.346,
      "step": 2027
    },
    {
      "epoch": 0.4054,
      "loss_align": 0.30962467193603516,
      "loss_contrastive": 0.029788196086883545,
      "loss_gen": 0.396129310131073,
      "neg_sim_mean": 0.6201635003089905,
      "pos_sim_mean": 0.6903753280639648,
      "step": 2027
    },
    {
      "epoch": 0.4056,
      "grad_norm": 3.1061387062072754,
      "learning_rate": 7.417939742875808e-05,
      "loss": 0.7093,
      "step": 2028
    },
    {
      "epoch": 0.4056,
      "loss_align": 0.23557090759277344,
      "loss_contrastive": 0.11013948917388916,
      "loss_gen": 0.22796297073364258,
      "neg_sim_mean": 0.7745685577392578,
      "pos_sim_mean": 0.7644290924072266,
      "step": 2028
    },
    {
      "epoch": 0.4058,
      "grad_norm": 1.7367475032806396,
      "learning_rate": 7.414883795241753e-05,
      "loss": 0.4768,
      "step": 2029
    },
    {
      "epoch": 0.4058,
      "loss_align": 0.2527047395706177,
      "loss_contrastive": 0.18151307106018066,
      "loss_gen": 0.31990760564804077,
      "neg_sim_mean": 0.8288083076477051,
      "pos_sim_mean": 0.7472952604293823,
      "step": 2029
    },
    {
      "epoch": 0.406,
      "grad_norm": 2.244982957839966,
      "learning_rate": 7.411826670622676e-05,
      "loss": 0.5944,
      "step": 2030
    },
    {
      "epoch": 0.406,
      "loss_align": 0.11291587352752686,
      "loss_contrastive": 0.03751581907272339,
      "loss_gen": 0.328532338142395,
      "neg_sim_mean": 0.8245999217033386,
      "pos_sim_mean": 0.8870841264724731,
      "step": 2030
    },
    {
      "epoch": 0.4062,
      "grad_norm": 2.249072551727295,
      "learning_rate": 7.408768370508576e-05,
      "loss": 0.446,
      "step": 2031
    },
    {
      "epoch": 0.4062,
      "loss_align": 0.14019536972045898,
      "loss_contrastive": 0.012912511825561523,
      "loss_gen": 0.34094423055648804,
      "neg_sim_mean": 0.7727171182632446,
      "pos_sim_mean": 0.859804630279541,
      "step": 2031
    },
    {
      "epoch": 0.4064,
      "grad_norm": 1.9236087799072266,
      "learning_rate": 7.405708896390037e-05,
      "loss": 0.4827,
      "step": 2032
    },
    {
      "epoch": 0.4064,
      "loss_align": 0.19195592403411865,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2043496072292328,
      "neg_sim_mean": 0.6913944482803345,
      "pos_sim_mean": 0.8080440759658813,
      "step": 2032
    },
    {
      "epoch": 0.4066,
      "grad_norm": 1.6222596168518066,
      "learning_rate": 7.402648249758203e-05,
      "loss": 0.3963,
      "step": 2033
    },
    {
      "epoch": 0.4066,
      "loss_align": 0.2508026361465454,
      "loss_contrastive": 0.09977608919143677,
      "loss_gen": 0.5617833137512207,
      "neg_sim_mean": 0.7489734292030334,
      "pos_sim_mean": 0.7491973638534546,
      "step": 2033
    },
    {
      "epoch": 0.4068,
      "grad_norm": 2.6227526664733887,
      "learning_rate": 7.399586432104804e-05,
      "loss": 0.8246,
      "step": 2034
    },
    {
      "epoch": 0.4068,
      "loss_align": 0.16309499740600586,
      "loss_contrastive": 0.08461415767669678,
      "loss_gen": 0.269067645072937,
      "neg_sim_mean": 0.821519136428833,
      "pos_sim_mean": 0.8369050025939941,
      "step": 2034
    },
    {
      "epoch": 0.407,
      "grad_norm": 2.2521002292633057,
      "learning_rate": 7.396523444922125e-05,
      "loss": 0.4423,
      "step": 2035
    },
    {
      "epoch": 0.407,
      "loss_align": 0.183823823928833,
      "loss_contrastive": 0.0899268388748169,
      "loss_gen": 0.21658295392990112,
      "neg_sim_mean": 0.806102991104126,
      "pos_sim_mean": 0.816176176071167,
      "step": 2035
    },
    {
      "epoch": 0.4072,
      "grad_norm": 1.900836706161499,
      "learning_rate": 7.393459289703035e-05,
      "loss": 0.4112,
      "step": 2036
    },
    {
      "epoch": 0.4072,
      "loss_align": 0.601197361946106,
      "loss_contrastive": 0.5169672966003418,
      "loss_gen": 0.11040909588336945,
      "neg_sim_mean": 0.8157699108123779,
      "pos_sim_mean": 0.39880263805389404,
      "step": 2036
    },
    {
      "epoch": 0.4074,
      "grad_norm": 1.4035462141036987,
      "learning_rate": 7.390393967940962e-05,
      "loss": 0.7736,
      "step": 2037
    },
    {
      "epoch": 0.4074,
      "loss_align": 0.15970420837402344,
      "loss_contrastive": 0.1316462755203247,
      "loss_gen": 0.2593839764595032,
      "neg_sim_mean": 0.8719420433044434,
      "pos_sim_mean": 0.8402957916259766,
      "step": 2037
    },
    {
      "epoch": 0.4076,
      "grad_norm": 1.8356211185455322,
      "learning_rate": 7.387327481129906e-05,
      "loss": 0.4349,
      "step": 2038
    },
    {
      "epoch": 0.4076,
      "loss_align": 0.2879534363746643,
      "loss_contrastive": 0.02689957618713379,
      "loss_gen": 0.24794313311576843,
      "neg_sim_mean": 0.6389461159706116,
      "pos_sim_mean": 0.7120465636253357,
      "step": 2038
    },
    {
      "epoch": 0.4078,
      "grad_norm": 2.1949105262756348,
      "learning_rate": 7.38425983076444e-05,
      "loss": 0.5391,
      "step": 2039
    },
    {
      "epoch": 0.4078,
      "loss_align": 0.37580257654190063,
      "loss_contrastive": 0.10821419954299927,
      "loss_gen": 0.1564950793981552,
      "neg_sim_mean": 0.6324115991592407,
      "pos_sim_mean": 0.6241974234580994,
      "step": 2039
    },
    {
      "epoch": 0.408,
      "grad_norm": 1.5352975130081177,
      "learning_rate": 7.381191018339696e-05,
      "loss": 0.5453,
      "step": 2040
    },
    {
      "epoch": 0.408,
      "loss_align": 0.182431161403656,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02892584539949894,
      "neg_sim_mean": 0.5044515132904053,
      "pos_sim_mean": 0.817568838596344,
      "step": 2040
    },
    {
      "epoch": 0.4082,
      "grad_norm": 1.293729543685913,
      "learning_rate": 7.378121045351378e-05,
      "loss": 0.2114,
      "step": 2041
    },
    {
      "epoch": 0.4082,
      "loss_align": 0.23132216930389404,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15643900632858276,
      "neg_sim_mean": 0.5571514368057251,
      "pos_sim_mean": 0.768677830696106,
      "step": 2041
    },
    {
      "epoch": 0.4084,
      "grad_norm": 1.5518512725830078,
      "learning_rate": 7.375049913295749e-05,
      "loss": 0.3878,
      "step": 2042
    },
    {
      "epoch": 0.4084,
      "loss_align": 0.24155861139297485,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02232290990650654,
      "neg_sim_mean": 0.5428922176361084,
      "pos_sim_mean": 0.7584413886070251,
      "step": 2042
    },
    {
      "epoch": 0.4086,
      "grad_norm": 1.05941903591156,
      "learning_rate": 7.371977623669647e-05,
      "loss": 0.2639,
      "step": 2043
    },
    {
      "epoch": 0.4086,
      "loss_align": 0.2224544882774353,
      "loss_contrastive": 0.0,
      "loss_gen": 0.017633937299251556,
      "neg_sim_mean": 0.31187349557876587,
      "pos_sim_mean": 0.7775455117225647,
      "step": 2043
    },
    {
      "epoch": 0.4088,
      "grad_norm": 0.9023754596710205,
      "learning_rate": 7.368904177970466e-05,
      "loss": 0.2401,
      "step": 2044
    },
    {
      "epoch": 0.4088,
      "loss_align": 0.12422597408294678,
      "loss_contrastive": 0.04071849584579468,
      "loss_gen": 0.11861910670995712,
      "neg_sim_mean": 0.81649249792099,
      "pos_sim_mean": 0.8757740259170532,
      "step": 2044
    },
    {
      "epoch": 0.409,
      "grad_norm": 1.294272541999817,
      "learning_rate": 7.365829577696166e-05,
      "loss": 0.2477,
      "step": 2045
    },
    {
      "epoch": 0.409,
      "loss_align": 0.0685083270072937,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20516380667686462,
      "neg_sim_mean": 0.8019022345542908,
      "pos_sim_mean": 0.9314916729927063,
      "step": 2045
    },
    {
      "epoch": 0.4092,
      "grad_norm": 1.6437513828277588,
      "learning_rate": 7.362753824345272e-05,
      "loss": 0.2737,
      "step": 2046
    },
    {
      "epoch": 0.4092,
      "loss_align": 0.21961331367492676,
      "loss_contrastive": 0.19136959314346313,
      "loss_gen": 0.1602460891008377,
      "neg_sim_mean": 0.8717562556266785,
      "pos_sim_mean": 0.7803866863250732,
      "step": 2046
    },
    {
      "epoch": 0.4094,
      "grad_norm": 1.6467819213867188,
      "learning_rate": 7.359676919416865e-05,
      "loss": 0.4028,
      "step": 2047
    },
    {
      "epoch": 0.4094,
      "loss_align": 0.13869190216064453,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01708219014108181,
      "neg_sim_mean": 0.4176746606826782,
      "pos_sim_mean": 0.8613080978393555,
      "step": 2047
    },
    {
      "epoch": 0.4096,
      "grad_norm": 0.8714888095855713,
      "learning_rate": 7.356598864410592e-05,
      "loss": 0.1558,
      "step": 2048
    },
    {
      "epoch": 0.4096,
      "loss_align": 0.1497829556465149,
      "loss_contrastive": 0.06743472814559937,
      "loss_gen": 0.3045084774494171,
      "neg_sim_mean": 0.8176517486572266,
      "pos_sim_mean": 0.8502170443534851,
      "step": 2048
    },
    {
      "epoch": 0.4098,
      "grad_norm": 2.1127872467041016,
      "learning_rate": 7.353519660826665e-05,
      "loss": 0.4624,
      "step": 2049
    },
    {
      "epoch": 0.4098,
      "loss_align": 0.14520466327667236,
      "loss_contrastive": 0.010053634643554688,
      "loss_gen": 0.2675189971923828,
      "neg_sim_mean": 0.7648489475250244,
      "pos_sim_mean": 0.8547953367233276,
      "step": 2049
    },
    {
      "epoch": 0.41,
      "grad_norm": 2.167250633239746,
      "learning_rate": 7.350439310165842e-05,
      "loss": 0.4139,
      "step": 2050
    },
    {
      "epoch": 0.41,
      "loss_align": 0.2529042959213257,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29182323813438416,
      "neg_sim_mean": 0.5444790124893188,
      "pos_sim_mean": 0.7470957040786743,
      "step": 2050
    },
    {
      "epoch": 0.4102,
      "grad_norm": 1.806875228881836,
      "learning_rate": 7.347357813929454e-05,
      "loss": 0.5447,
      "step": 2051
    },
    {
      "epoch": 0.4102,
      "loss_align": 0.1260671615600586,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14703799784183502,
      "neg_sim_mean": 0.6564871072769165,
      "pos_sim_mean": 0.8739328384399414,
      "step": 2051
    },
    {
      "epoch": 0.4104,
      "grad_norm": 1.4239555597305298,
      "learning_rate": 7.344275173619385e-05,
      "loss": 0.2731,
      "step": 2052
    },
    {
      "epoch": 0.4104,
      "loss_align": 0.16040325164794922,
      "loss_contrastive": 0.03477597236633301,
      "loss_gen": 0.15074491500854492,
      "neg_sim_mean": 0.7743726968765259,
      "pos_sim_mean": 0.8395967483520508,
      "step": 2052
    },
    {
      "epoch": 0.4106,
      "grad_norm": 1.5230730772018433,
      "learning_rate": 7.341191390738073e-05,
      "loss": 0.3153,
      "step": 2053
    },
    {
      "epoch": 0.4106,
      "loss_align": 0.3053598403930664,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004837876185774803,
      "neg_sim_mean": 0.4497148394584656,
      "pos_sim_mean": 0.6946401596069336,
      "step": 2053
    },
    {
      "epoch": 0.4108,
      "grad_norm": 0.98356032371521,
      "learning_rate": 7.33810646678852e-05,
      "loss": 0.3102,
      "step": 2054
    },
    {
      "epoch": 0.4108,
      "loss_align": 0.14515739679336548,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010823630727827549,
      "neg_sim_mean": 0.33820027112960815,
      "pos_sim_mean": 0.8548426032066345,
      "step": 2054
    },
    {
      "epoch": 0.411,
      "grad_norm": 0.753108561038971,
      "learning_rate": 7.335020403274278e-05,
      "loss": 0.156,
      "step": 2055
    },
    {
      "epoch": 0.411,
      "loss_align": 0.11185848712921143,
      "loss_contrastive": 0.06293272972106934,
      "loss_gen": 0.18968245387077332,
      "neg_sim_mean": 0.85107421875,
      "pos_sim_mean": 0.8881415128707886,
      "step": 2055
    },
    {
      "epoch": 0.4112,
      "grad_norm": 2.013014554977417,
      "learning_rate": 7.331933201699457e-05,
      "loss": 0.3091,
      "step": 2056
    },
    {
      "epoch": 0.4112,
      "loss_align": 0.09555846452713013,
      "loss_contrastive": 0.0,
      "loss_gen": 0.6742463707923889,
      "neg_sim_mean": 0.7879362106323242,
      "pos_sim_mean": 0.9044415354728699,
      "step": 2056
    },
    {
      "epoch": 0.4114,
      "grad_norm": 2.847902774810791,
      "learning_rate": 7.328844863568721e-05,
      "loss": 0.7698,
      "step": 2057
    },
    {
      "epoch": 0.4114,
      "loss_align": 0.13641095161437988,
      "loss_contrastive": 0.01643294095993042,
      "loss_gen": 0.10610145330429077,
      "neg_sim_mean": 0.7800219655036926,
      "pos_sim_mean": 0.8635890483856201,
      "step": 2057
    },
    {
      "epoch": 0.4116,
      "grad_norm": 1.3035128116607666,
      "learning_rate": 7.325755390387292e-05,
      "loss": 0.2445,
      "step": 2058
    },
    {
      "epoch": 0.4116,
      "loss_align": 0.1215277910232544,
      "loss_contrastive": 0.09308141469955444,
      "loss_gen": 0.32656213641166687,
      "neg_sim_mean": 0.8715535998344421,
      "pos_sim_mean": 0.8784722089767456,
      "step": 2058
    },
    {
      "epoch": 0.4118,
      "grad_norm": 1.8969401121139526,
      "learning_rate": 7.32266478366094e-05,
      "loss": 0.4593,
      "step": 2059
    },
    {
      "epoch": 0.4118,
      "loss_align": 0.11563408374786377,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0582822784781456,
      "neg_sim_mean": 0.6737194657325745,
      "pos_sim_mean": 0.8843659162521362,
      "step": 2059
    },
    {
      "epoch": 0.412,
      "grad_norm": 1.1610666513442993,
      "learning_rate": 7.319573044895986e-05,
      "loss": 0.1739,
      "step": 2060
    },
    {
      "epoch": 0.412,
      "loss_align": 0.3232001066207886,
      "loss_contrastive": 0.09392118453979492,
      "loss_gen": 0.2618580758571625,
      "neg_sim_mean": 0.6707210540771484,
      "pos_sim_mean": 0.6767998933792114,
      "step": 2060
    },
    {
      "epoch": 0.4122,
      "grad_norm": 1.9452685117721558,
      "learning_rate": 7.316480175599309e-05,
      "loss": 0.5963,
      "step": 2061
    },
    {
      "epoch": 0.4122,
      "loss_align": 0.16534864902496338,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23099201917648315,
      "neg_sim_mean": 0.7246769666671753,
      "pos_sim_mean": 0.8346513509750366,
      "step": 2061
    },
    {
      "epoch": 0.4124,
      "grad_norm": 1.8325951099395752,
      "learning_rate": 7.313386177278335e-05,
      "loss": 0.3963,
      "step": 2062
    },
    {
      "epoch": 0.4124,
      "loss_align": 0.15588140487670898,
      "loss_contrastive": 0.09170496463775635,
      "loss_gen": 0.3616030812263489,
      "neg_sim_mean": 0.8358235359191895,
      "pos_sim_mean": 0.844118595123291,
      "step": 2062
    },
    {
      "epoch": 0.4126,
      "grad_norm": 2.068814754486084,
      "learning_rate": 7.310291051441044e-05,
      "loss": 0.5285,
      "step": 2063
    },
    {
      "epoch": 0.4126,
      "loss_align": 0.1029043197631836,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12510846555233002,
      "neg_sim_mean": 0.6986340880393982,
      "pos_sim_mean": 0.8970956802368164,
      "step": 2063
    },
    {
      "epoch": 0.4128,
      "grad_norm": 1.439439058303833,
      "learning_rate": 7.307194799595958e-05,
      "loss": 0.228,
      "step": 2064
    },
    {
      "epoch": 0.4128,
      "loss_align": 0.2964736819267273,
      "loss_contrastive": 0.11725753545761108,
      "loss_gen": 0.27903082966804504,
      "neg_sim_mean": 0.7207838296890259,
      "pos_sim_mean": 0.7035263180732727,
      "step": 2064
    },
    {
      "epoch": 0.413,
      "grad_norm": 2.0296545028686523,
      "learning_rate": 7.304097423252155e-05,
      "loss": 0.5896,
      "step": 2065
    },
    {
      "epoch": 0.413,
      "loss_align": 0.04584014415740967,
      "loss_contrastive": 0.0,
      "loss_gen": 0.07040242105722427,
      "neg_sim_mean": 0.7062023878097534,
      "pos_sim_mean": 0.9541598558425903,
      "step": 2065
    },
    {
      "epoch": 0.4132,
      "grad_norm": 0.9114355444908142,
      "learning_rate": 7.300998923919259e-05,
      "loss": 0.1162,
      "step": 2066
    },
    {
      "epoch": 0.4132,
      "loss_align": 0.38339757919311523,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02353348769247532,
      "neg_sim_mean": 0.4311971068382263,
      "pos_sim_mean": 0.6166024208068848,
      "step": 2066
    },
    {
      "epoch": 0.4134,
      "grad_norm": 1.41655695438385,
      "learning_rate": 7.297899303107441e-05,
      "loss": 0.4069,
      "step": 2067
    },
    {
      "epoch": 0.4134,
      "loss_align": 0.1901242733001709,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0046746269799768925,
      "neg_sim_mean": 0.4291023015975952,
      "pos_sim_mean": 0.8098757266998291,
      "step": 2067
    },
    {
      "epoch": 0.4136,
      "grad_norm": 0.6626297235488892,
      "learning_rate": 7.294798562327416e-05,
      "loss": 0.1948,
      "step": 2068
    },
    {
      "epoch": 0.4136,
      "loss_align": 0.14095860719680786,
      "loss_contrastive": 0.08689814805984497,
      "loss_gen": 0.12604403495788574,
      "neg_sim_mean": 0.8459395170211792,
      "pos_sim_mean": 0.8590413928031921,
      "step": 2068
    },
    {
      "epoch": 0.4138,
      "grad_norm": 1.8599516153335571,
      "learning_rate": 7.29169670309045e-05,
      "loss": 0.2774,
      "step": 2069
    },
    {
      "epoch": 0.4138,
      "loss_align": 0.16852307319641113,
      "loss_contrastive": 0.05686831474304199,
      "loss_gen": 0.20425580441951752,
      "neg_sim_mean": 0.788345217704773,
      "pos_sim_mean": 0.8314769268035889,
      "step": 2069
    },
    {
      "epoch": 0.414,
      "grad_norm": 1.6913368701934814,
      "learning_rate": 7.28859372690835e-05,
      "loss": 0.3796,
      "step": 2070
    },
    {
      "epoch": 0.414,
      "loss_align": 0.2542457580566406,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013866978697478771,
      "neg_sim_mean": 0.5253069400787354,
      "pos_sim_mean": 0.7457542419433594,
      "step": 2070
    },
    {
      "epoch": 0.4142,
      "grad_norm": 0.9877485632896423,
      "learning_rate": 7.285489635293472e-05,
      "loss": 0.2681,
      "step": 2071
    },
    {
      "epoch": 0.4142,
      "loss_align": 0.23619574308395386,
      "loss_contrastive": 0.1500837206840515,
      "loss_gen": 0.3261416554450989,
      "neg_sim_mean": 0.8138879537582397,
      "pos_sim_mean": 0.7638042569160461,
      "step": 2071
    },
    {
      "epoch": 0.4144,
      "grad_norm": 2.2167129516601562,
      "learning_rate": 7.282384429758709e-05,
      "loss": 0.5803,
      "step": 2072
    },
    {
      "epoch": 0.4144,
      "loss_align": 0.28832507133483887,
      "loss_contrastive": 0.08792877197265625,
      "loss_gen": 0.17621566355228424,
      "neg_sim_mean": 0.6996036767959595,
      "pos_sim_mean": 0.7116749286651611,
      "step": 2072
    },
    {
      "epoch": 0.4146,
      "grad_norm": 1.8121660947799683,
      "learning_rate": 7.279278111817501e-05,
      "loss": 0.4751,
      "step": 2073
    },
    {
      "epoch": 0.4146,
      "loss_align": 0.23736703395843506,
      "loss_contrastive": 0.11431288719177246,
      "loss_gen": 0.2678946852684021,
      "neg_sim_mean": 0.7769458293914795,
      "pos_sim_mean": 0.7626329660415649,
      "step": 2073
    },
    {
      "epoch": 0.4148,
      "grad_norm": 2.156306028366089,
      "learning_rate": 7.276170682983829e-05,
      "loss": 0.519,
      "step": 2074
    },
    {
      "epoch": 0.4148,
      "loss_align": 0.13851165771484375,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0017213295213878155,
      "neg_sim_mean": 0.4530198574066162,
      "pos_sim_mean": 0.8614883422851562,
      "step": 2074
    },
    {
      "epoch": 0.415,
      "grad_norm": 0.6685593724250793,
      "learning_rate": 7.27306214477222e-05,
      "loss": 0.1402,
      "step": 2075
    },
    {
      "epoch": 0.415,
      "loss_align": 0.1631184220314026,
      "loss_contrastive": 0.033608853816986084,
      "loss_gen": 0.25002601742744446,
      "neg_sim_mean": 0.7704904079437256,
      "pos_sim_mean": 0.8368815779685974,
      "step": 2075
    },
    {
      "epoch": 0.4152,
      "grad_norm": 2.094970226287842,
      "learning_rate": 7.269952498697734e-05,
      "loss": 0.4172,
      "step": 2076
    },
    {
      "epoch": 0.4152,
      "loss_align": 0.10008323192596436,
      "loss_contrastive": 0.0,
      "loss_gen": 0.308709055185318,
      "neg_sim_mean": 0.7642055749893188,
      "pos_sim_mean": 0.8999167680740356,
      "step": 2076
    },
    {
      "epoch": 0.4154,
      "grad_norm": 2.094520330429077,
      "learning_rate": 7.266841746275976e-05,
      "loss": 0.4088,
      "step": 2077
    },
    {
      "epoch": 0.4154,
      "loss_align": 0.3473818302154541,
      "loss_contrastive": 0.08556950092315674,
      "loss_gen": 0.19279921054840088,
      "neg_sim_mean": 0.6381876468658447,
      "pos_sim_mean": 0.6526181697845459,
      "step": 2077
    },
    {
      "epoch": 0.4156,
      "grad_norm": 2.0617527961730957,
      "learning_rate": 7.26372988902309e-05,
      "loss": 0.5504,
      "step": 2078
    },
    {
      "epoch": 0.4156,
      "loss_align": 0.1475188136100769,
      "loss_contrastive": 0.1061200499534607,
      "loss_gen": 0.15994176268577576,
      "neg_sim_mean": 0.8586012125015259,
      "pos_sim_mean": 0.8524811863899231,
      "step": 2078
    },
    {
      "epoch": 0.4158,
      "grad_norm": 1.9351742267608643,
      "learning_rate": 7.260616928455754e-05,
      "loss": 0.3202,
      "step": 2079
    },
    {
      "epoch": 0.4158,
      "loss_align": 0.16250813007354736,
      "loss_contrastive": 0.0,
      "loss_gen": 0.020243056118488312,
      "neg_sim_mean": 0.2910839021205902,
      "pos_sim_mean": 0.8374918699264526,
      "step": 2079
    },
    {
      "epoch": 0.416,
      "grad_norm": 2.393876314163208,
      "learning_rate": 7.257502866091192e-05,
      "loss": 0.1828,
      "step": 2080
    },
    {
      "epoch": 0.416,
      "loss_align": 0.23745226860046387,
      "loss_contrastive": 0.08727020025253296,
      "loss_gen": 0.30953162908554077,
      "neg_sim_mean": 0.7498179078102112,
      "pos_sim_mean": 0.7625477313995361,
      "step": 2080
    },
    {
      "epoch": 0.4162,
      "grad_norm": 2.1483242511749268,
      "learning_rate": 7.254387703447154e-05,
      "loss": 0.5575,
      "step": 2081
    },
    {
      "epoch": 0.4162,
      "loss_align": 0.20886999368667603,
      "loss_contrastive": 0.10217827558517456,
      "loss_gen": 0.1983262002468109,
      "neg_sim_mean": 0.7933082580566406,
      "pos_sim_mean": 0.791130006313324,
      "step": 2081
    },
    {
      "epoch": 0.4164,
      "grad_norm": 1.7259869575500488,
      "learning_rate": 7.251271442041938e-05,
      "loss": 0.4195,
      "step": 2082
    },
    {
      "epoch": 0.4164,
      "loss_align": 0.21838688850402832,
      "loss_contrastive": 0.13629776239395142,
      "loss_gen": 0.1672925055027008,
      "neg_sim_mean": 0.8179108500480652,
      "pos_sim_mean": 0.7816131114959717,
      "step": 2082
    },
    {
      "epoch": 0.4166,
      "grad_norm": 1.7343480587005615,
      "learning_rate": 7.24815408339437e-05,
      "loss": 0.402,
      "step": 2083
    },
    {
      "epoch": 0.4166,
      "loss_align": 0.47691065073013306,
      "loss_contrastive": 0.31623831391334534,
      "loss_gen": 0.20754727721214294,
      "neg_sim_mean": 0.7393276691436768,
      "pos_sim_mean": 0.5230893492698669,
      "step": 2083
    },
    {
      "epoch": 0.4168,
      "grad_norm": 2.2007696628570557,
      "learning_rate": 7.245035629023812e-05,
      "loss": 0.7224,
      "step": 2084
    },
    {
      "epoch": 0.4168,
      "loss_align": 0.15443789958953857,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0033550849184393883,
      "neg_sim_mean": 0.37083443999290466,
      "pos_sim_mean": 0.8455621004104614,
      "step": 2084
    },
    {
      "epoch": 0.417,
      "grad_norm": 0.4824541211128235,
      "learning_rate": 7.241916080450163e-05,
      "loss": 0.1578,
      "step": 2085
    },
    {
      "epoch": 0.417,
      "loss_align": 0.14479857683181763,
      "loss_contrastive": 0.11811918020248413,
      "loss_gen": 0.4592745900154114,
      "neg_sim_mean": 0.8733205795288086,
      "pos_sim_mean": 0.8552014231681824,
      "step": 2085
    },
    {
      "epoch": 0.4172,
      "grad_norm": 2.579618453979492,
      "learning_rate": 7.238795439193848e-05,
      "loss": 0.6182,
      "step": 2086
    },
    {
      "epoch": 0.4172,
      "loss_align": 0.17345160245895386,
      "loss_contrastive": 0.05080002546310425,
      "loss_gen": 0.17514044046401978,
      "neg_sim_mean": 0.7773483991622925,
      "pos_sim_mean": 0.8265483975410461,
      "step": 2086
    },
    {
      "epoch": 0.4174,
      "grad_norm": 1.7651699781417847,
      "learning_rate": 7.235673706775836e-05,
      "loss": 0.3547,
      "step": 2087
    },
    {
      "epoch": 0.4174,
      "loss_align": 0.3832141160964966,
      "loss_contrastive": 0.10762250423431396,
      "loss_gen": 0.26371726393699646,
      "neg_sim_mean": 0.6244083642959595,
      "pos_sim_mean": 0.6167858839035034,
      "step": 2087
    },
    {
      "epoch": 0.4176,
      "grad_norm": 2.5085575580596924,
      "learning_rate": 7.232550884717617e-05,
      "loss": 0.6598,
      "step": 2088
    },
    {
      "epoch": 0.4176,
      "loss_align": 0.21994656324386597,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2701839804649353,
      "neg_sim_mean": 0.6161128282546997,
      "pos_sim_mean": 0.780053436756134,
      "step": 2088
    },
    {
      "epoch": 0.4178,
      "grad_norm": 1.9468051195144653,
      "learning_rate": 7.229426974541221e-05,
      "loss": 0.4901,
      "step": 2089
    },
    {
      "epoch": 0.4178,
      "loss_align": 0.2900393009185791,
      "loss_contrastive": 0.038982093334198,
      "loss_gen": 0.37119928002357483,
      "neg_sim_mean": 0.648942768573761,
      "pos_sim_mean": 0.7099606990814209,
      "step": 2089
    },
    {
      "epoch": 0.418,
      "grad_norm": 2.3871524333953857,
      "learning_rate": 7.226301977769199e-05,
      "loss": 0.6659,
      "step": 2090
    },
    {
      "epoch": 0.418,
      "loss_align": 0.1105278730392456,
      "loss_contrastive": 0.07581007480621338,
      "loss_gen": 0.27871376276016235,
      "neg_sim_mean": 0.8652821779251099,
      "pos_sim_mean": 0.8894721269607544,
      "step": 2090
    },
    {
      "epoch": 0.4182,
      "grad_norm": 2.152406930923462,
      "learning_rate": 7.223175895924638e-05,
      "loss": 0.3983,
      "step": 2091
    },
    {
      "epoch": 0.4182,
      "loss_align": 0.13353270292282104,
      "loss_contrastive": 0.08067411184310913,
      "loss_gen": 0.3305568993091583,
      "neg_sim_mean": 0.8471413850784302,
      "pos_sim_mean": 0.866467297077179,
      "step": 2091
    },
    {
      "epoch": 0.4184,
      "grad_norm": 2.156653881072998,
      "learning_rate": 7.220048730531153e-05,
      "loss": 0.4738,
      "step": 2092
    },
    {
      "epoch": 0.4184,
      "loss_align": 0.22468727827072144,
      "loss_contrastive": 0.0699341893196106,
      "loss_gen": 0.30194148421287537,
      "neg_sim_mean": 0.7452468872070312,
      "pos_sim_mean": 0.7753127217292786,
      "step": 2092
    },
    {
      "epoch": 0.4186,
      "grad_norm": 2.136679172515869,
      "learning_rate": 7.216920483112886e-05,
      "loss": 0.535,
      "step": 2093
    },
    {
      "epoch": 0.4186,
      "loss_align": 0.17442822456359863,
      "loss_contrastive": 0.0,
      "loss_gen": 0.029994655400514603,
      "neg_sim_mean": 0.4748098850250244,
      "pos_sim_mean": 0.8255717754364014,
      "step": 2093
    },
    {
      "epoch": 0.4188,
      "grad_norm": 1.0540275573730469,
      "learning_rate": 7.21379115519451e-05,
      "loss": 0.2044,
      "step": 2094
    },
    {
      "epoch": 0.4188,
      "loss_align": 0.1879870891571045,
      "loss_contrastive": 0.03827786445617676,
      "loss_gen": 0.1446606069803238,
      "neg_sim_mean": 0.7502907514572144,
      "pos_sim_mean": 0.8120129108428955,
      "step": 2094
    },
    {
      "epoch": 0.419,
      "grad_norm": 1.7868541479110718,
      "learning_rate": 7.210660748301214e-05,
      "loss": 0.3372,
      "step": 2095
    },
    {
      "epoch": 0.419,
      "loss_align": 0.15086591243743896,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09664762020111084,
      "neg_sim_mean": 0.6534902453422546,
      "pos_sim_mean": 0.849134087562561,
      "step": 2095
    },
    {
      "epoch": 0.4192,
      "grad_norm": 1.6261086463928223,
      "learning_rate": 7.207529263958726e-05,
      "loss": 0.2475,
      "step": 2096
    },
    {
      "epoch": 0.4192,
      "loss_align": 0.15964215993881226,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20245598256587982,
      "neg_sim_mean": 0.7212827205657959,
      "pos_sim_mean": 0.8403578400611877,
      "step": 2096
    },
    {
      "epoch": 0.4194,
      "grad_norm": 1.6100685596466064,
      "learning_rate": 7.204396703693294e-05,
      "loss": 0.3621,
      "step": 2097
    },
    {
      "epoch": 0.4194,
      "loss_align": 0.23140692710876465,
      "loss_contrastive": 0.048235535621643066,
      "loss_gen": 0.2240334451198578,
      "neg_sim_mean": 0.7168285846710205,
      "pos_sim_mean": 0.7685930728912354,
      "step": 2097
    },
    {
      "epoch": 0.4196,
      "grad_norm": 1.71221923828125,
      "learning_rate": 7.201263069031687e-05,
      "loss": 0.4612,
      "step": 2098
    },
    {
      "epoch": 0.4196,
      "loss_align": 0.17485332489013672,
      "loss_contrastive": 0.09993064403533936,
      "loss_gen": 0.35561662912368774,
      "neg_sim_mean": 0.8250772953033447,
      "pos_sim_mean": 0.8251466751098633,
      "step": 2098
    },
    {
      "epoch": 0.4198,
      "grad_norm": 2.0732369422912598,
      "learning_rate": 7.1981283615012e-05,
      "loss": 0.5425,
      "step": 2099
    },
    {
      "epoch": 0.4198,
      "loss_align": 0.12343418598175049,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1426393687725067,
      "neg_sim_mean": 0.625442385673523,
      "pos_sim_mean": 0.8765658140182495,
      "step": 2099
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.3517767190933228,
      "learning_rate": 7.194992582629654e-05,
      "loss": 0.2661,
      "step": 2100
    },
    {
      "epoch": 0.42,
      "loss_align": 0.17020189762115479,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21176773309707642,
      "neg_sim_mean": 0.6162791848182678,
      "pos_sim_mean": 0.8297981023788452,
      "step": 2100
    },
    {
      "epoch": 0.4202,
      "grad_norm": 1.6312851905822754,
      "learning_rate": 7.191855733945387e-05,
      "loss": 0.382,
      "step": 2101
    },
    {
      "epoch": 0.4202,
      "loss_align": 0.14241564273834229,
      "loss_contrastive": 0.10794854164123535,
      "loss_gen": 0.180076465010643,
      "neg_sim_mean": 0.8655328750610352,
      "pos_sim_mean": 0.8575843572616577,
      "step": 2101
    },
    {
      "epoch": 0.4204,
      "grad_norm": 1.7823710441589355,
      "learning_rate": 7.188717816977263e-05,
      "loss": 0.3354,
      "step": 2102
    },
    {
      "epoch": 0.4204,
      "loss_align": 0.20839250087738037,
      "loss_contrastive": 0.07234448194503784,
      "loss_gen": 0.25085076689720154,
      "neg_sim_mean": 0.7639519572257996,
      "pos_sim_mean": 0.7916074991226196,
      "step": 2102
    },
    {
      "epoch": 0.4206,
      "grad_norm": 2.29349946975708,
      "learning_rate": 7.185578833254664e-05,
      "loss": 0.4679,
      "step": 2103
    },
    {
      "epoch": 0.4206,
      "loss_align": 0.1512974500656128,
      "loss_contrastive": 0.07312685251235962,
      "loss_gen": 0.22008933126926422,
      "neg_sim_mean": 0.8218293786048889,
      "pos_sim_mean": 0.8487025499343872,
      "step": 2103
    },
    {
      "epoch": 0.4208,
      "grad_norm": 1.5095957517623901,
      "learning_rate": 7.182438784307495e-05,
      "loss": 0.3802,
      "step": 2104
    },
    {
      "epoch": 0.4208,
      "loss_align": 0.2761845588684082,
      "loss_contrastive": 0.2156360149383545,
      "loss_gen": 0.32674306631088257,
      "neg_sim_mean": 0.8394514322280884,
      "pos_sim_mean": 0.7238154411315918,
      "step": 2104
    },
    {
      "epoch": 0.421,
      "grad_norm": 2.4904141426086426,
      "learning_rate": 7.179297671666171e-05,
      "loss": 0.6288,
      "step": 2105
    },
    {
      "epoch": 0.421,
      "loss_align": 0.1608412265777588,
      "loss_contrastive": 0.07881742715835571,
      "loss_gen": 0.12972332537174225,
      "neg_sim_mean": 0.817976176738739,
      "pos_sim_mean": 0.8391587734222412,
      "step": 2105
    },
    {
      "epoch": 0.4212,
      "grad_norm": 1.3987497091293335,
      "learning_rate": 7.176155496861638e-05,
      "loss": 0.3,
      "step": 2106
    },
    {
      "epoch": 0.4212,
      "loss_align": 0.25199437141418457,
      "loss_contrastive": 0.1715325117111206,
      "loss_gen": 0.295246958732605,
      "neg_sim_mean": 0.8195381164550781,
      "pos_sim_mean": 0.7480056285858154,
      "step": 2106
    },
    {
      "epoch": 0.4214,
      "grad_norm": 1.9999120235443115,
      "learning_rate": 7.173012261425351e-05,
      "loss": 0.5678,
      "step": 2107
    },
    {
      "epoch": 0.4214,
      "loss_align": 0.10279077291488647,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2578642964363098,
      "neg_sim_mean": 0.7547852993011475,
      "pos_sim_mean": 0.8972092270851135,
      "step": 2107
    },
    {
      "epoch": 0.4216,
      "grad_norm": 2.0230939388275146,
      "learning_rate": 7.169867966889288e-05,
      "loss": 0.3607,
      "step": 2108
    },
    {
      "epoch": 0.4216,
      "loss_align": 0.19566011428833008,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0044865054078400135,
      "neg_sim_mean": 0.4473987817764282,
      "pos_sim_mean": 0.8043398857116699,
      "step": 2108
    },
    {
      "epoch": 0.4218,
      "grad_norm": 0.5207483172416687,
      "learning_rate": 7.166722614785937e-05,
      "loss": 0.2001,
      "step": 2109
    },
    {
      "epoch": 0.4218,
      "loss_align": 0.1545041799545288,
      "loss_contrastive": 0.09110379219055176,
      "loss_gen": 0.12488548457622528,
      "neg_sim_mean": 0.836599588394165,
      "pos_sim_mean": 0.8454958200454712,
      "step": 2109
    },
    {
      "epoch": 0.422,
      "grad_norm": 1.3995410203933716,
      "learning_rate": 7.163576206648304e-05,
      "loss": 0.2903,
      "step": 2110
    },
    {
      "epoch": 0.422,
      "loss_align": 0.28106260299682617,
      "loss_contrastive": 0.14394307136535645,
      "loss_gen": 0.22725367546081543,
      "neg_sim_mean": 0.7628804445266724,
      "pos_sim_mean": 0.7189373970031738,
      "step": 2110
    },
    {
      "epoch": 0.4222,
      "grad_norm": 1.7914843559265137,
      "learning_rate": 7.160428744009912e-05,
      "loss": 0.5256,
      "step": 2111
    },
    {
      "epoch": 0.4222,
      "loss_align": 0.04599207639694214,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2429395318031311,
      "neg_sim_mean": 0.7380727529525757,
      "pos_sim_mean": 0.9540079236030579,
      "step": 2111
    },
    {
      "epoch": 0.4224,
      "grad_norm": 1.9387234449386597,
      "learning_rate": 7.157280228404795e-05,
      "loss": 0.2889,
      "step": 2112
    },
    {
      "epoch": 0.4224,
      "loss_align": 0.2904067635536194,
      "loss_contrastive": 0.06120496988296509,
      "loss_gen": 0.30741044878959656,
      "neg_sim_mean": 0.6707981824874878,
      "pos_sim_mean": 0.7095932364463806,
      "step": 2112
    },
    {
      "epoch": 0.4226,
      "grad_norm": 2.1284873485565186,
      "learning_rate": 7.154130661367503e-05,
      "loss": 0.6052,
      "step": 2113
    },
    {
      "epoch": 0.4226,
      "loss_align": 0.15802961587905884,
      "loss_contrastive": 0.0,
      "loss_gen": 0.017027953639626503,
      "neg_sim_mean": 0.42348262667655945,
      "pos_sim_mean": 0.8419703841209412,
      "step": 2113
    },
    {
      "epoch": 0.4228,
      "grad_norm": 0.8822039365768433,
      "learning_rate": 7.150980044433093e-05,
      "loss": 0.1751,
      "step": 2114
    },
    {
      "epoch": 0.4228,
      "loss_align": 0.1795218586921692,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002928805770352483,
      "neg_sim_mean": 0.3944292366504669,
      "pos_sim_mean": 0.8204781413078308,
      "step": 2114
    },
    {
      "epoch": 0.423,
      "grad_norm": 0.5936741828918457,
      "learning_rate": 7.147828379137142e-05,
      "loss": 0.1825,
      "step": 2115
    },
    {
      "epoch": 0.423,
      "loss_align": 0.1577562689781189,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4545894265174866,
      "neg_sim_mean": 0.5220574736595154,
      "pos_sim_mean": 0.8422437310218811,
      "step": 2115
    },
    {
      "epoch": 0.4232,
      "grad_norm": 2.717639446258545,
      "learning_rate": 7.14467566701573e-05,
      "loss": 0.6123,
      "step": 2116
    },
    {
      "epoch": 0.4232,
      "loss_align": 0.19049286842346191,
      "loss_contrastive": 0.069682776927948,
      "loss_gen": 0.39113548398017883,
      "neg_sim_mean": 0.7791898846626282,
      "pos_sim_mean": 0.8095071315765381,
      "step": 2116
    },
    {
      "epoch": 0.4234,
      "grad_norm": 2.374199390411377,
      "learning_rate": 7.141521909605452e-05,
      "loss": 0.59,
      "step": 2117
    },
    {
      "epoch": 0.4234,
      "loss_align": 0.09178006649017334,
      "loss_contrastive": 0.08297479152679443,
      "loss_gen": 0.16821295022964478,
      "neg_sim_mean": 0.8911947011947632,
      "pos_sim_mean": 0.9082199335098267,
      "step": 2117
    },
    {
      "epoch": 0.4236,
      "grad_norm": 1.6806339025497437,
      "learning_rate": 7.138367108443411e-05,
      "loss": 0.27,
      "step": 2118
    },
    {
      "epoch": 0.4236,
      "loss_align": 0.0945885181427002,
      "loss_contrastive": 0.11536169052124023,
      "loss_gen": 0.3097189664840698,
      "neg_sim_mean": 0.9207731485366821,
      "pos_sim_mean": 0.9054114818572998,
      "step": 2118
    },
    {
      "epoch": 0.4238,
      "grad_norm": 1.9793360233306885,
      "learning_rate": 7.135211265067216e-05,
      "loss": 0.4182,
      "step": 2119
    },
    {
      "epoch": 0.4238,
      "loss_align": 0.24550890922546387,
      "loss_contrastive": 0.1558130979537964,
      "loss_gen": 0.19482764601707458,
      "neg_sim_mean": 0.8103041648864746,
      "pos_sim_mean": 0.7544910907745361,
      "step": 2119
    },
    {
      "epoch": 0.424,
      "grad_norm": 1.9688751697540283,
      "learning_rate": 7.132054381014995e-05,
      "loss": 0.459,
      "step": 2120
    },
    {
      "epoch": 0.424,
      "loss_align": 0.23544514179229736,
      "loss_contrastive": 0.03714883327484131,
      "loss_gen": 0.26492589712142944,
      "neg_sim_mean": 0.701703667640686,
      "pos_sim_mean": 0.7645548582077026,
      "step": 2120
    },
    {
      "epoch": 0.4242,
      "grad_norm": 2.5430495738983154,
      "learning_rate": 7.128896457825364e-05,
      "loss": 0.5048,
      "step": 2121
    },
    {
      "epoch": 0.4242,
      "loss_align": 0.15438032150268555,
      "loss_contrastive": 0.09666812419891357,
      "loss_gen": 0.32108184695243835,
      "neg_sim_mean": 0.8422877788543701,
      "pos_sim_mean": 0.8456196784973145,
      "step": 2121
    },
    {
      "epoch": 0.4244,
      "grad_norm": 1.6863951683044434,
      "learning_rate": 7.125737497037464e-05,
      "loss": 0.4871,
      "step": 2122
    },
    {
      "epoch": 0.4244,
      "loss_align": 0.2920827865600586,
      "loss_contrastive": 0.19431668519973755,
      "loss_gen": 0.20527446269989014,
      "neg_sim_mean": 0.802233874797821,
      "pos_sim_mean": 0.7079172134399414,
      "step": 2122
    },
    {
      "epoch": 0.4246,
      "grad_norm": 1.709509015083313,
      "learning_rate": 7.12257750019093e-05,
      "loss": 0.5207,
      "step": 2123
    },
    {
      "epoch": 0.4246,
      "loss_align": 0.17697155475616455,
      "loss_contrastive": 0.0,
      "loss_gen": 0.020728230476379395,
      "neg_sim_mean": 0.46083831787109375,
      "pos_sim_mean": 0.8230284452438354,
      "step": 2123
    },
    {
      "epoch": 0.4248,
      "grad_norm": 0.9947166442871094,
      "learning_rate": 7.119416468825908e-05,
      "loss": 0.1977,
      "step": 2124
    },
    {
      "epoch": 0.4248,
      "loss_align": 0.24385511875152588,
      "loss_contrastive": 0.11575597524642944,
      "loss_gen": 0.08222466707229614,
      "neg_sim_mean": 0.7719008326530457,
      "pos_sim_mean": 0.7561448812484741,
      "step": 2124
    },
    {
      "epoch": 0.425,
      "grad_norm": 1.340091347694397,
      "learning_rate": 7.116254404483049e-05,
      "loss": 0.34,
      "step": 2125
    },
    {
      "epoch": 0.425,
      "loss_align": 0.10474425554275513,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3202812969684601,
      "neg_sim_mean": 0.7715502977371216,
      "pos_sim_mean": 0.8952557444572449,
      "step": 2125
    },
    {
      "epoch": 0.4252,
      "grad_norm": 2.222264289855957,
      "learning_rate": 7.113091308703498e-05,
      "loss": 0.425,
      "step": 2126
    },
    {
      "epoch": 0.4252,
      "loss_align": 0.11267721652984619,
      "loss_contrastive": 0.020804524421691895,
      "loss_gen": 0.2371288686990738,
      "neg_sim_mean": 0.8081272840499878,
      "pos_sim_mean": 0.8873227834701538,
      "step": 2126
    },
    {
      "epoch": 0.4254,
      "grad_norm": 2.022841453552246,
      "learning_rate": 7.109927183028914e-05,
      "loss": 0.3523,
      "step": 2127
    },
    {
      "epoch": 0.4254,
      "loss_align": 0.36682021617889404,
      "loss_contrastive": 0.08093523979187012,
      "loss_gen": 0.31031113862991333,
      "neg_sim_mean": 0.6141149997711182,
      "pos_sim_mean": 0.633179783821106,
      "step": 2127
    },
    {
      "epoch": 0.4256,
      "grad_norm": 2.7195441722869873,
      "learning_rate": 7.106762029001455e-05,
      "loss": 0.6868,
      "step": 2128
    },
    {
      "epoch": 0.4256,
      "loss_align": 0.24300360679626465,
      "loss_contrastive": 0.20535290241241455,
      "loss_gen": 0.3044440746307373,
      "neg_sim_mean": 0.862349271774292,
      "pos_sim_mean": 0.7569963932037354,
      "step": 2128
    },
    {
      "epoch": 0.4258,
      "grad_norm": 2.2925024032592773,
      "learning_rate": 7.103595848163774e-05,
      "loss": 0.5721,
      "step": 2129
    },
    {
      "epoch": 0.4258,
      "loss_align": 0.24382156133651733,
      "loss_contrastive": 0.16181451082229614,
      "loss_gen": 0.2838786840438843,
      "neg_sim_mean": 0.8179929256439209,
      "pos_sim_mean": 0.7561784386634827,
      "step": 2129
    },
    {
      "epoch": 0.426,
      "grad_norm": 1.9964178800582886,
      "learning_rate": 7.100428642059033e-05,
      "loss": 0.5471,
      "step": 2130
    },
    {
      "epoch": 0.426,
      "loss_align": 0.22755730152130127,
      "loss_contrastive": 0.056823015213012695,
      "loss_gen": 0.2614210247993469,
      "neg_sim_mean": 0.7292656898498535,
      "pos_sim_mean": 0.7724426984786987,
      "step": 2130
    },
    {
      "epoch": 0.4262,
      "grad_norm": 1.7291045188903809,
      "learning_rate": 7.097260412230886e-05,
      "loss": 0.4958,
      "step": 2131
    },
    {
      "epoch": 0.4262,
      "loss_align": 0.25719964504241943,
      "loss_contrastive": 0.05824059247970581,
      "loss_gen": 0.11234169453382492,
      "neg_sim_mean": 0.7010409235954285,
      "pos_sim_mean": 0.7428003549575806,
      "step": 2131
    },
    {
      "epoch": 0.4264,
      "grad_norm": 1.4497325420379639,
      "learning_rate": 7.094091160223492e-05,
      "loss": 0.3765,
      "step": 2132
    },
    {
      "epoch": 0.4264,
      "loss_align": 0.15347087383270264,
      "loss_contrastive": 0.05452537536621094,
      "loss_gen": 0.34534093737602234,
      "neg_sim_mean": 0.8010544776916504,
      "pos_sim_mean": 0.8465291261672974,
      "step": 2132
    },
    {
      "epoch": 0.4266,
      "grad_norm": 2.053804874420166,
      "learning_rate": 7.090920887581506e-05,
      "loss": 0.5054,
      "step": 2133
    },
    {
      "epoch": 0.4266,
      "loss_align": 0.23183035850524902,
      "loss_contrastive": 0.10157668590545654,
      "loss_gen": 0.5522376298904419,
      "neg_sim_mean": 0.7697463035583496,
      "pos_sim_mean": 0.768169641494751,
      "step": 2133
    },
    {
      "epoch": 0.4268,
      "grad_norm": 2.776538610458374,
      "learning_rate": 7.087749595850084e-05,
      "loss": 0.7963,
      "step": 2134
    },
    {
      "epoch": 0.4268,
      "loss_align": 0.135817289352417,
      "loss_contrastive": 0.008385181427001953,
      "loss_gen": 0.45464199781417847,
      "neg_sim_mean": 0.772567868232727,
      "pos_sim_mean": 0.864182710647583,
      "step": 2134
    },
    {
      "epoch": 0.427,
      "grad_norm": 2.2041189670562744,
      "learning_rate": 7.084577286574868e-05,
      "loss": 0.5915,
      "step": 2135
    },
    {
      "epoch": 0.427,
      "loss_align": 0.07319211959838867,
      "loss_contrastive": 0.10373067855834961,
      "loss_gen": 0.14449170231819153,
      "neg_sim_mean": 0.930538535118103,
      "pos_sim_mean": 0.9268078804016113,
      "step": 2135
    },
    {
      "epoch": 0.4272,
      "grad_norm": 1.4762275218963623,
      "learning_rate": 7.081403961302006e-05,
      "loss": 0.2301,
      "step": 2136
    },
    {
      "epoch": 0.4272,
      "loss_align": 0.1176302433013916,
      "loss_contrastive": 0.09740573167800903,
      "loss_gen": 0.6765592098236084,
      "neg_sim_mean": 0.8797754645347595,
      "pos_sim_mean": 0.8823697566986084,
      "step": 2136
    },
    {
      "epoch": 0.4274,
      "grad_norm": 3.9643330574035645,
      "learning_rate": 7.07822962157814e-05,
      "loss": 0.8059,
      "step": 2137
    },
    {
      "epoch": 0.4274,
      "loss_align": 0.15501338243484497,
      "loss_contrastive": 0.11845594644546509,
      "loss_gen": 0.17647823691368103,
      "neg_sim_mean": 0.8634425401687622,
      "pos_sim_mean": 0.844986617565155,
      "step": 2137
    },
    {
      "epoch": 0.4276,
      "grad_norm": 1.4125028848648071,
      "learning_rate": 7.075054268950402e-05,
      "loss": 0.3457,
      "step": 2138
    },
    {
      "epoch": 0.4276,
      "loss_align": 0.3045130968093872,
      "loss_contrastive": 0.12524282932281494,
      "loss_gen": 0.21794776618480682,
      "neg_sim_mean": 0.7207297086715698,
      "pos_sim_mean": 0.6954869031906128,
      "step": 2138
    },
    {
      "epoch": 0.4278,
      "grad_norm": 1.5560311079025269,
      "learning_rate": 7.071877904966423e-05,
      "loss": 0.5375,
      "step": 2139
    },
    {
      "epoch": 0.4278,
      "loss_align": 0.26852357387542725,
      "loss_contrastive": 0.044254422187805176,
      "loss_gen": 0.35009080171585083,
      "neg_sim_mean": 0.67573082447052,
      "pos_sim_mean": 0.7314764261245728,
      "step": 2139
    },
    {
      "epoch": 0.428,
      "grad_norm": 2.8860552310943604,
      "learning_rate": 7.068700531174319e-05,
      "loss": 0.6239,
      "step": 2140
    },
    {
      "epoch": 0.428,
      "loss_align": 0.32259035110473633,
      "loss_contrastive": 0.1325892210006714,
      "loss_gen": 0.11002293229103088,
      "neg_sim_mean": 0.7099988460540771,
      "pos_sim_mean": 0.6774096488952637,
      "step": 2140
    },
    {
      "epoch": 0.4282,
      "grad_norm": 1.4977648258209229,
      "learning_rate": 7.06552214912271e-05,
      "loss": 0.4485,
      "step": 2141
    },
    {
      "epoch": 0.4282,
      "loss_align": 0.41304081678390503,
      "loss_contrastive": 0.1794341504573822,
      "loss_gen": 0.3864297568798065,
      "neg_sim_mean": 0.6663933396339417,
      "pos_sim_mean": 0.586959183216095,
      "step": 2141
    },
    {
      "epoch": 0.4284,
      "grad_norm": 2.6874325275421143,
      "learning_rate": 7.062342760360696e-05,
      "loss": 0.821,
      "step": 2142
    },
    {
      "epoch": 0.4284,
      "loss_align": 0.029469728469848633,
      "loss_contrastive": 0.047281086444854736,
      "loss_gen": 0.10127267986536026,
      "neg_sim_mean": 0.9178113341331482,
      "pos_sim_mean": 0.9705302715301514,
      "step": 2142
    },
    {
      "epoch": 0.4286,
      "grad_norm": 0.9689324498176575,
      "learning_rate": 7.059162366437874e-05,
      "loss": 0.1364,
      "step": 2143
    },
    {
      "epoch": 0.4286,
      "loss_align": 0.1266920566558838,
      "loss_contrastive": 0.1098601222038269,
      "loss_gen": 0.360585480928421,
      "neg_sim_mean": 0.8831680417060852,
      "pos_sim_mean": 0.8733079433441162,
      "step": 2143
    },
    {
      "epoch": 0.4288,
      "grad_norm": 2.219106912612915,
      "learning_rate": 7.055980968904332e-05,
      "loss": 0.5005,
      "step": 2144
    },
    {
      "epoch": 0.4288,
      "loss_align": 0.10006475448608398,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4013645052909851,
      "neg_sim_mean": 0.7769601345062256,
      "pos_sim_mean": 0.899935245513916,
      "step": 2144
    },
    {
      "epoch": 0.429,
      "grad_norm": 1.9609031677246094,
      "learning_rate": 7.05279856931064e-05,
      "loss": 0.5014,
      "step": 2145
    },
    {
      "epoch": 0.429,
      "loss_align": 0.10685694217681885,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22721339762210846,
      "neg_sim_mean": 0.7909392714500427,
      "pos_sim_mean": 0.8931430578231812,
      "step": 2145
    },
    {
      "epoch": 0.4292,
      "grad_norm": 1.871880292892456,
      "learning_rate": 7.049615169207864e-05,
      "loss": 0.3341,
      "step": 2146
    },
    {
      "epoch": 0.4292,
      "loss_align": 0.10457372665405273,
      "loss_contrastive": 0.10885828733444214,
      "loss_gen": 0.49846866726875305,
      "neg_sim_mean": 0.9042845368385315,
      "pos_sim_mean": 0.8954262733459473,
      "step": 2146
    },
    {
      "epoch": 0.4294,
      "grad_norm": 2.508683204650879,
      "learning_rate": 7.046430770147554e-05,
      "loss": 0.6161,
      "step": 2147
    },
    {
      "epoch": 0.4294,
      "loss_align": 0.29455578327178955,
      "loss_contrastive": 0.08523005247116089,
      "loss_gen": 0.35905879735946655,
      "neg_sim_mean": 0.6906742453575134,
      "pos_sim_mean": 0.7054442167282104,
      "step": 2147
    },
    {
      "epoch": 0.4296,
      "grad_norm": 2.063446521759033,
      "learning_rate": 7.043245373681747e-05,
      "loss": 0.6638,
      "step": 2148
    },
    {
      "epoch": 0.4296,
      "loss_align": 0.19634974002838135,
      "loss_contrastive": 0.09805083274841309,
      "loss_gen": 0.2311389446258545,
      "neg_sim_mean": 0.8017010688781738,
      "pos_sim_mean": 0.8036502599716187,
      "step": 2148
    },
    {
      "epoch": 0.4298,
      "grad_norm": 2.0596773624420166,
      "learning_rate": 7.040058981362965e-05,
      "loss": 0.4393,
      "step": 2149
    },
    {
      "epoch": 0.4298,
      "loss_align": 0.329082190990448,
      "loss_contrastive": 0.08166778087615967,
      "loss_gen": 0.24083299934864044,
      "neg_sim_mean": 0.6525855660438538,
      "pos_sim_mean": 0.670917809009552,
      "step": 2149
    },
    {
      "epoch": 0.43,
      "grad_norm": 2.0668933391571045,
      "learning_rate": 7.036871594744217e-05,
      "loss": 0.5797,
      "step": 2150
    },
    {
      "epoch": 0.43,
      "loss_align": 0.14524221420288086,
      "loss_contrastive": 0.0913550853729248,
      "loss_gen": 0.2428223043680191,
      "neg_sim_mean": 0.846112847328186,
      "pos_sim_mean": 0.8547577857971191,
      "step": 2150
    },
    {
      "epoch": 0.4302,
      "grad_norm": 1.711635708808899,
      "learning_rate": 7.033683215379002e-05,
      "loss": 0.399,
      "step": 2151
    },
    {
      "epoch": 0.4302,
      "loss_align": 0.17388659715652466,
      "loss_contrastive": 0.057031452655792236,
      "loss_gen": 0.2140377014875412,
      "neg_sim_mean": 0.7831448316574097,
      "pos_sim_mean": 0.8261134028434753,
      "step": 2151
    },
    {
      "epoch": 0.4304,
      "grad_norm": 1.9528337717056274,
      "learning_rate": 7.030493844821291e-05,
      "loss": 0.3948,
      "step": 2152
    },
    {
      "epoch": 0.4304,
      "loss_align": 0.2120925784111023,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0342068076133728,
      "neg_sim_mean": 0.4686771631240845,
      "pos_sim_mean": 0.7879074215888977,
      "step": 2152
    },
    {
      "epoch": 0.4306,
      "grad_norm": 1.2985972166061401,
      "learning_rate": 7.027303484625548e-05,
      "loss": 0.2463,
      "step": 2153
    },
    {
      "epoch": 0.4306,
      "loss_align": 0.14117419719696045,
      "loss_contrastive": 0.12621629238128662,
      "loss_gen": 0.23464035987854004,
      "neg_sim_mean": 0.8850420713424683,
      "pos_sim_mean": 0.8588258028030396,
      "step": 2153
    },
    {
      "epoch": 0.4308,
      "grad_norm": 1.5850825309753418,
      "learning_rate": 7.024112136346712e-05,
      "loss": 0.391,
      "step": 2154
    },
    {
      "epoch": 0.4308,
      "loss_align": 0.21988117694854736,
      "loss_contrastive": 0.08120584487915039,
      "loss_gen": 0.2520335018634796,
      "neg_sim_mean": 0.7613246440887451,
      "pos_sim_mean": 0.7801188230514526,
      "step": 2154
    },
    {
      "epoch": 0.431,
      "grad_norm": 2.0797085762023926,
      "learning_rate": 7.020919801540211e-05,
      "loss": 0.4817,
      "step": 2155
    },
    {
      "epoch": 0.431,
      "loss_align": 0.13826417922973633,
      "loss_contrastive": 0.061521947383880615,
      "loss_gen": 0.17665410041809082,
      "neg_sim_mean": 0.8232577443122864,
      "pos_sim_mean": 0.8617358207702637,
      "step": 2155
    },
    {
      "epoch": 0.4312,
      "grad_norm": 1.979451298713684,
      "learning_rate": 7.017726481761951e-05,
      "loss": 0.3223,
      "step": 2156
    },
    {
      "epoch": 0.4312,
      "loss_align": 0.2070326805114746,
      "loss_contrastive": 0.12738895416259766,
      "loss_gen": 0.10262210667133331,
      "neg_sim_mean": 0.8203562498092651,
      "pos_sim_mean": 0.7929673194885254,
      "step": 2156
    },
    {
      "epoch": 0.4314,
      "grad_norm": 1.3850096464157104,
      "learning_rate": 7.014532178568314e-05,
      "loss": 0.3249,
      "step": 2157
    },
    {
      "epoch": 0.4314,
      "loss_align": 0.3055253028869629,
      "loss_contrastive": 0.07909870147705078,
      "loss_gen": 0.21740902960300446,
      "neg_sim_mean": 0.67357337474823,
      "pos_sim_mean": 0.6944746971130371,
      "step": 2157
    },
    {
      "epoch": 0.4316,
      "grad_norm": 1.9645124673843384,
      "learning_rate": 7.011336893516166e-05,
      "loss": 0.5324,
      "step": 2158
    },
    {
      "epoch": 0.4316,
      "loss_align": 0.297155499458313,
      "loss_contrastive": 0.011895298957824707,
      "loss_gen": 0.12895944714546204,
      "neg_sim_mean": 0.6147397756576538,
      "pos_sim_mean": 0.702844500541687,
      "step": 2158
    },
    {
      "epoch": 0.4318,
      "grad_norm": 1.6978875398635864,
      "learning_rate": 7.00814062816285e-05,
      "loss": 0.4275,
      "step": 2159
    },
    {
      "epoch": 0.4318,
      "loss_align": 0.13196396827697754,
      "loss_contrastive": 0.01042085886001587,
      "loss_gen": 0.156470388174057,
      "neg_sim_mean": 0.7784568667411804,
      "pos_sim_mean": 0.8680360317230225,
      "step": 2159
    },
    {
      "epoch": 0.432,
      "grad_norm": 1.5463413000106812,
      "learning_rate": 7.004943384066187e-05,
      "loss": 0.2897,
      "step": 2160
    },
    {
      "epoch": 0.432,
      "loss_align": 0.3907848000526428,
      "loss_contrastive": 0.22019439935684204,
      "loss_gen": 0.3753284215927124,
      "neg_sim_mean": 0.7294095754623413,
      "pos_sim_mean": 0.6092151999473572,
      "step": 2160
    },
    {
      "epoch": 0.4322,
      "grad_norm": 2.339604377746582,
      "learning_rate": 7.001745162784477e-05,
      "loss": 0.7925,
      "step": 2161
    },
    {
      "epoch": 0.4322,
      "loss_align": 0.1920936107635498,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21074844896793365,
      "neg_sim_mean": 0.6020471453666687,
      "pos_sim_mean": 0.8079063892364502,
      "step": 2161
    },
    {
      "epoch": 0.4324,
      "grad_norm": 1.9074183702468872,
      "learning_rate": 6.998545965876488e-05,
      "loss": 0.4028,
      "step": 2162
    },
    {
      "epoch": 0.4324,
      "loss_align": 0.18263518810272217,
      "loss_contrastive": 0.09429651498794556,
      "loss_gen": 0.10664471983909607,
      "neg_sim_mean": 0.8116613030433655,
      "pos_sim_mean": 0.8173648118972778,
      "step": 2162
    },
    {
      "epoch": 0.4326,
      "grad_norm": 1.3473790884017944,
      "learning_rate": 6.995345794901477e-05,
      "loss": 0.3006,
      "step": 2163
    },
    {
      "epoch": 0.4326,
      "loss_align": 0.1780625581741333,
      "loss_contrastive": 0.07163989543914795,
      "loss_gen": 0.22628211975097656,
      "neg_sim_mean": 0.7935773134231567,
      "pos_sim_mean": 0.8219374418258667,
      "step": 2163
    },
    {
      "epoch": 0.4328,
      "grad_norm": 1.9184489250183105,
      "learning_rate": 6.992144651419164e-05,
      "loss": 0.4129,
      "step": 2164
    },
    {
      "epoch": 0.4328,
      "loss_align": 0.1476476788520813,
      "loss_contrastive": 0.0486336350440979,
      "loss_gen": 0.2066306471824646,
      "neg_sim_mean": 0.8009859323501587,
      "pos_sim_mean": 0.8523523211479187,
      "step": 2164
    },
    {
      "epoch": 0.433,
      "grad_norm": 1.8587520122528076,
      "learning_rate": 6.98894253698975e-05,
      "loss": 0.3601,
      "step": 2165
    },
    {
      "epoch": 0.433,
      "loss_align": 0.1948709487915039,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18347640335559845,
      "neg_sim_mean": 0.6852232217788696,
      "pos_sim_mean": 0.8051290512084961,
      "step": 2165
    },
    {
      "epoch": 0.4332,
      "grad_norm": 1.539846658706665,
      "learning_rate": 6.985739453173903e-05,
      "loss": 0.3783,
      "step": 2166
    },
    {
      "epoch": 0.4332,
      "loss_align": 0.2385188341140747,
      "loss_contrastive": 0.0,
      "loss_gen": 0.018620606511831284,
      "neg_sim_mean": 0.38843315839767456,
      "pos_sim_mean": 0.7614811658859253,
      "step": 2166
    },
    {
      "epoch": 0.4334,
      "grad_norm": 1.0618066787719727,
      "learning_rate": 6.982535401532772e-05,
      "loss": 0.2571,
      "step": 2167
    },
    {
      "epoch": 0.4334,
      "loss_align": 0.27840548753738403,
      "loss_contrastive": 0.12486821413040161,
      "loss_gen": 0.29931676387786865,
      "neg_sim_mean": 0.7464627027511597,
      "pos_sim_mean": 0.721594512462616,
      "step": 2167
    },
    {
      "epoch": 0.4336,
      "grad_norm": 2.2148516178131104,
      "learning_rate": 6.979330383627969e-05,
      "loss": 0.5927,
      "step": 2168
    },
    {
      "epoch": 0.4336,
      "loss_align": 0.22632884979248047,
      "loss_contrastive": 0.01316744089126587,
      "loss_gen": 0.1633223593235016,
      "neg_sim_mean": 0.6868385672569275,
      "pos_sim_mean": 0.7736711502075195,
      "step": 2168
    },
    {
      "epoch": 0.4338,
      "grad_norm": 1.6689903736114502,
      "learning_rate": 6.976124401021583e-05,
      "loss": 0.3912,
      "step": 2169
    },
    {
      "epoch": 0.4338,
      "loss_align": 0.17244470119476318,
      "loss_contrastive": 0.09937947988510132,
      "loss_gen": 0.28394362330436707,
      "neg_sim_mean": 0.8269347548484802,
      "pos_sim_mean": 0.8275552988052368,
      "step": 2169
    },
    {
      "epoch": 0.434,
      "grad_norm": 1.919156551361084,
      "learning_rate": 6.97291745527617e-05,
      "loss": 0.4683,
      "step": 2170
    },
    {
      "epoch": 0.434,
      "loss_align": 0.18979966640472412,
      "loss_contrastive": 0.11769223213195801,
      "loss_gen": 0.15109997987747192,
      "neg_sim_mean": 0.827892541885376,
      "pos_sim_mean": 0.8102003335952759,
      "step": 2170
    },
    {
      "epoch": 0.4342,
      "grad_norm": 1.3999724388122559,
      "learning_rate": 6.969709547954756e-05,
      "loss": 0.355,
      "step": 2171
    },
    {
      "epoch": 0.4342,
      "loss_align": 0.11479759216308594,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02231595665216446,
      "neg_sim_mean": 0.32332631945610046,
      "pos_sim_mean": 0.8852024078369141,
      "step": 2171
    },
    {
      "epoch": 0.4344,
      "grad_norm": 1.1006112098693848,
      "learning_rate": 6.966500680620837e-05,
      "loss": 0.1371,
      "step": 2172
    },
    {
      "epoch": 0.4344,
      "loss_align": 0.16753876209259033,
      "loss_contrastive": 0.07174146175384521,
      "loss_gen": 0.17732250690460205,
      "neg_sim_mean": 0.804202675819397,
      "pos_sim_mean": 0.8324612379074097,
      "step": 2172
    },
    {
      "epoch": 0.4346,
      "grad_norm": 1.595425009727478,
      "learning_rate": 6.963290854838376e-05,
      "loss": 0.3535,
      "step": 2173
    },
    {
      "epoch": 0.4346,
      "loss_align": 0.09928274154663086,
      "loss_contrastive": 0.09844428300857544,
      "loss_gen": 0.1857995092868805,
      "neg_sim_mean": 0.8991615176200867,
      "pos_sim_mean": 0.9007172584533691,
      "step": 2173
    },
    {
      "epoch": 0.4348,
      "grad_norm": 2.076289176940918,
      "learning_rate": 6.960080072171801e-05,
      "loss": 0.2969,
      "step": 2174
    },
    {
      "epoch": 0.4348,
      "loss_align": 0.14541500806808472,
      "loss_contrastive": 0.008583605289459229,
      "loss_gen": 0.28486213088035583,
      "neg_sim_mean": 0.7631685733795166,
      "pos_sim_mean": 0.8545849919319153,
      "step": 2174
    },
    {
      "epoch": 0.435,
      "grad_norm": 2.078909158706665,
      "learning_rate": 6.956868334186013e-05,
      "loss": 0.4313,
      "step": 2175
    },
    {
      "epoch": 0.435,
      "loss_align": 0.16180455684661865,
      "loss_contrastive": 0.021030545234680176,
      "loss_gen": 0.45607930421829224,
      "neg_sim_mean": 0.7592259645462036,
      "pos_sim_mean": 0.8381954431533813,
      "step": 2175
    },
    {
      "epoch": 0.4352,
      "grad_norm": 2.0485236644744873,
      "learning_rate": 6.953655642446368e-05,
      "loss": 0.6204,
      "step": 2176
    },
    {
      "epoch": 0.4352,
      "loss_align": 0.18910813331604004,
      "loss_contrastive": 0.14862239360809326,
      "loss_gen": 0.12937913835048676,
      "neg_sim_mean": 0.8595142364501953,
      "pos_sim_mean": 0.81089186668396,
      "step": 2176
    },
    {
      "epoch": 0.4354,
      "grad_norm": 1.5026315450668335,
      "learning_rate": 6.950441998518699e-05,
      "loss": 0.3363,
      "step": 2177
    },
    {
      "epoch": 0.4354,
      "loss_align": 0.2980530261993408,
      "loss_contrastive": 0.16637110710144043,
      "loss_gen": 0.20110124349594116,
      "neg_sim_mean": 0.7683180570602417,
      "pos_sim_mean": 0.7019469738006592,
      "step": 2177
    },
    {
      "epoch": 0.4356,
      "grad_norm": 1.8175064325332642,
      "learning_rate": 6.947227403969293e-05,
      "loss": 0.5191,
      "step": 2178
    },
    {
      "epoch": 0.4356,
      "loss_align": 0.22124183177947998,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3401600122451782,
      "neg_sim_mean": 0.5696365833282471,
      "pos_sim_mean": 0.77875816822052,
      "step": 2178
    },
    {
      "epoch": 0.4358,
      "grad_norm": 2.0661821365356445,
      "learning_rate": 6.944011860364905e-05,
      "loss": 0.5614,
      "step": 2179
    },
    {
      "epoch": 0.4358,
      "loss_align": 0.1474781632423401,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015519142616540194,
      "neg_sim_mean": 0.37338489294052124,
      "pos_sim_mean": 0.8525218367576599,
      "step": 2179
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.4714229702949524,
      "learning_rate": 6.940795369272754e-05,
      "loss": 0.149,
      "step": 2180
    },
    {
      "epoch": 0.436,
      "loss_align": 0.17988324165344238,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1172671690583229,
      "neg_sim_mean": 0.7026374340057373,
      "pos_sim_mean": 0.8201167583465576,
      "step": 2180
    },
    {
      "epoch": 0.4362,
      "grad_norm": 1.4900611639022827,
      "learning_rate": 6.937577932260515e-05,
      "loss": 0.2972,
      "step": 2181
    },
    {
      "epoch": 0.4362,
      "loss_align": 0.14485567808151245,
      "loss_contrastive": 0.017676234245300293,
      "loss_gen": 0.1652126908302307,
      "neg_sim_mean": 0.7728205323219299,
      "pos_sim_mean": 0.8551443219184875,
      "step": 2181
    },
    {
      "epoch": 0.4364,
      "grad_norm": 1.8583996295928955,
      "learning_rate": 6.934359550896333e-05,
      "loss": 0.3122,
      "step": 2182
    },
    {
      "epoch": 0.4364,
      "loss_align": 0.30119121074676514,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0025149849243462086,
      "neg_sim_mean": 0.31938111782073975,
      "pos_sim_mean": 0.6988087892532349,
      "step": 2182
    },
    {
      "epoch": 0.4366,
      "grad_norm": 0.8889106512069702,
      "learning_rate": 6.9311402267488e-05,
      "loss": 0.3037,
      "step": 2183
    },
    {
      "epoch": 0.4366,
      "loss_align": 0.3165954351425171,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008007718250155449,
      "neg_sim_mean": 0.4618818163871765,
      "pos_sim_mean": 0.6834045648574829,
      "step": 2183
    },
    {
      "epoch": 0.4368,
      "grad_norm": 0.7735477685928345,
      "learning_rate": 6.927919961386984e-05,
      "loss": 0.3246,
      "step": 2184
    },
    {
      "epoch": 0.4368,
      "loss_align": 0.11644965410232544,
      "loss_contrastive": 0.022299230098724365,
      "loss_gen": 0.1861712634563446,
      "neg_sim_mean": 0.805849552154541,
      "pos_sim_mean": 0.8835503458976746,
      "step": 2184
    },
    {
      "epoch": 0.437,
      "grad_norm": 1.618161678314209,
      "learning_rate": 6.924698756380398e-05,
      "loss": 0.3053,
      "step": 2185
    },
    {
      "epoch": 0.437,
      "loss_align": 0.20317673683166504,
      "loss_contrastive": 0.036180734634399414,
      "loss_gen": 0.11311482638120651,
      "neg_sim_mean": 0.7330039739608765,
      "pos_sim_mean": 0.796823263168335,
      "step": 2185
    },
    {
      "epoch": 0.4372,
      "grad_norm": 1.5004485845565796,
      "learning_rate": 6.921476613299018e-05,
      "loss": 0.3206,
      "step": 2186
    },
    {
      "epoch": 0.4372,
      "loss_align": 0.17814183235168457,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0314660482108593,
      "neg_sim_mean": 0.34703508019447327,
      "pos_sim_mean": 0.8218581676483154,
      "step": 2186
    },
    {
      "epoch": 0.4374,
      "grad_norm": 1.370407223701477,
      "learning_rate": 6.918253533713282e-05,
      "loss": 0.2096,
      "step": 2187
    },
    {
      "epoch": 0.4374,
      "loss_align": 0.14581936597824097,
      "loss_contrastive": 0.055828988552093506,
      "loss_gen": 0.06942329555749893,
      "neg_sim_mean": 0.8100095987319946,
      "pos_sim_mean": 0.854180634021759,
      "step": 2187
    },
    {
      "epoch": 0.4376,
      "grad_norm": 1.2299271821975708,
      "learning_rate": 6.915029519194076e-05,
      "loss": 0.2219,
      "step": 2188
    },
    {
      "epoch": 0.4376,
      "loss_align": 0.05673342943191528,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2838110327720642,
      "neg_sim_mean": 0.8272584676742554,
      "pos_sim_mean": 0.9432665705680847,
      "step": 2188
    },
    {
      "epoch": 0.4378,
      "grad_norm": 1.960166335105896,
      "learning_rate": 6.911804571312745e-05,
      "loss": 0.3405,
      "step": 2189
    },
    {
      "epoch": 0.4378,
      "loss_align": 0.1469118595123291,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14360716938972473,
      "neg_sim_mean": 0.5816497802734375,
      "pos_sim_mean": 0.8530881404876709,
      "step": 2189
    },
    {
      "epoch": 0.438,
      "grad_norm": 1.3525549173355103,
      "learning_rate": 6.908578691641092e-05,
      "loss": 0.2905,
      "step": 2190
    },
    {
      "epoch": 0.438,
      "loss_align": 0.1296539306640625,
      "loss_contrastive": 0.10276293754577637,
      "loss_gen": 0.25535452365875244,
      "neg_sim_mean": 0.873108983039856,
      "pos_sim_mean": 0.8703460693359375,
      "step": 2190
    },
    {
      "epoch": 0.4382,
      "grad_norm": 1.8698995113372803,
      "learning_rate": 6.905351881751372e-05,
      "loss": 0.3973,
      "step": 2191
    },
    {
      "epoch": 0.4382,
      "loss_align": 0.13085073232650757,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001505793072283268,
      "neg_sim_mean": 0.36131811141967773,
      "pos_sim_mean": 0.8691492676734924,
      "step": 2191
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.35210493206977844,
      "learning_rate": 6.902124143216289e-05,
      "loss": 0.1324,
      "step": 2192
    },
    {
      "epoch": 0.4384,
      "loss_align": 0.17376422882080078,
      "loss_contrastive": 0.0715489387512207,
      "loss_gen": 0.2796093225479126,
      "neg_sim_mean": 0.797784686088562,
      "pos_sim_mean": 0.8262357711791992,
      "step": 2192
    },
    {
      "epoch": 0.4386,
      "grad_norm": 1.6080970764160156,
      "learning_rate": 6.898895477609007e-05,
      "loss": 0.462,
      "step": 2193
    },
    {
      "epoch": 0.4386,
      "loss_align": 0.08695453405380249,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002003619447350502,
      "neg_sim_mean": 0.4026820659637451,
      "pos_sim_mean": 0.9130454659461975,
      "step": 2193
    },
    {
      "epoch": 0.4388,
      "grad_norm": 0.31985267996788025,
      "learning_rate": 6.895665886503135e-05,
      "loss": 0.089,
      "step": 2194
    },
    {
      "epoch": 0.4388,
      "loss_align": 0.11941730976104736,
      "loss_contrastive": 0.011517643928527832,
      "loss_gen": 0.3528590500354767,
      "neg_sim_mean": 0.7921003103256226,
      "pos_sim_mean": 0.8805826902389526,
      "step": 2194
    },
    {
      "epoch": 0.439,
      "grad_norm": 2.764840841293335,
      "learning_rate": 6.892435371472741e-05,
      "loss": 0.4737,
      "step": 2195
    },
    {
      "epoch": 0.439,
      "loss_align": 0.1655929684638977,
      "loss_contrastive": 9.638071060180664e-05,
      "loss_gen": 0.37712618708610535,
      "neg_sim_mean": 0.7345033884048462,
      "pos_sim_mean": 0.8344070315361023,
      "step": 2195
    },
    {
      "epoch": 0.4392,
      "grad_norm": 2.1508429050445557,
      "learning_rate": 6.889203934092336e-05,
      "loss": 0.5427,
      "step": 2196
    },
    {
      "epoch": 0.4392,
      "loss_align": 0.17556250095367432,
      "loss_contrastive": 0.10561919212341309,
      "loss_gen": 0.23273468017578125,
      "neg_sim_mean": 0.8300566673278809,
      "pos_sim_mean": 0.8244374990463257,
      "step": 2196
    },
    {
      "epoch": 0.4394,
      "grad_norm": 1.6243067979812622,
      "learning_rate": 6.885971575936883e-05,
      "loss": 0.421,
      "step": 2197
    },
    {
      "epoch": 0.4394,
      "loss_align": 0.11543524265289307,
      "loss_contrastive": 0.06474113464355469,
      "loss_gen": 0.33969971537590027,
      "neg_sim_mean": 0.8493058681488037,
      "pos_sim_mean": 0.8845647573471069,
      "step": 2197
    },
    {
      "epoch": 0.4396,
      "grad_norm": 2.3643500804901123,
      "learning_rate": 6.882738298581797e-05,
      "loss": 0.4629,
      "step": 2198
    },
    {
      "epoch": 0.4396,
      "loss_align": 0.20312058925628662,
      "loss_contrastive": 0.11831837892532349,
      "loss_gen": 0.31977587938308716,
      "neg_sim_mean": 0.815197765827179,
      "pos_sim_mean": 0.7968794107437134,
      "step": 2198
    },
    {
      "epoch": 0.4398,
      "grad_norm": 2.1543080806732178,
      "learning_rate": 6.879504103602935e-05,
      "loss": 0.5371,
      "step": 2199
    },
    {
      "epoch": 0.4398,
      "loss_align": 0.17992275953292847,
      "loss_contrastive": 0.011200368404388428,
      "loss_gen": 0.38758134841918945,
      "neg_sim_mean": 0.731277585029602,
      "pos_sim_mean": 0.8200772404670715,
      "step": 2199
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.5104470252990723,
      "learning_rate": 6.876268992576604e-05,
      "loss": 0.5688,
      "step": 2200
    },
    {
      "epoch": 0.44,
      "loss_align": 0.1170356273651123,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0049535613507032394,
      "neg_sim_mean": 0.34457117319107056,
      "pos_sim_mean": 0.8829643726348877,
      "step": 2200
    },
    {
      "epoch": 0.4402,
      "grad_norm": 0.5156263709068298,
      "learning_rate": 6.873032967079561e-05,
      "loss": 0.122,
      "step": 2201
    },
    {
      "epoch": 0.4402,
      "loss_align": 0.1200023889541626,
      "loss_contrastive": 0.07501095533370972,
      "loss_gen": 0.2312062382698059,
      "neg_sim_mean": 0.8550085425376892,
      "pos_sim_mean": 0.8799976110458374,
      "step": 2201
    },
    {
      "epoch": 0.4404,
      "grad_norm": 1.7920804023742676,
      "learning_rate": 6.869796028689001e-05,
      "loss": 0.3602,
      "step": 2202
    },
    {
      "epoch": 0.4404,
      "loss_align": 0.08599257469177246,
      "loss_contrastive": 0.0626528263092041,
      "loss_gen": 0.13834773004055023,
      "neg_sim_mean": 0.8766602277755737,
      "pos_sim_mean": 0.9140074253082275,
      "step": 2202
    },
    {
      "epoch": 0.4406,
      "grad_norm": 1.1095269918441772,
      "learning_rate": 6.866558178982575e-05,
      "loss": 0.2319,
      "step": 2203
    },
    {
      "epoch": 0.4406,
      "loss_align": 0.20543092489242554,
      "loss_contrastive": 0.05375128984451294,
      "loss_gen": 0.1898338496685028,
      "neg_sim_mean": 0.7483203411102295,
      "pos_sim_mean": 0.7945690751075745,
      "step": 2203
    },
    {
      "epoch": 0.4408,
      "grad_norm": 1.8744406700134277,
      "learning_rate": 6.863319419538366e-05,
      "loss": 0.4017,
      "step": 2204
    },
    {
      "epoch": 0.4408,
      "loss_align": 0.176863431930542,
      "loss_contrastive": 0.14004695415496826,
      "loss_gen": 0.46070438623428345,
      "neg_sim_mean": 0.8631834983825684,
      "pos_sim_mean": 0.823136568069458,
      "step": 2204
    },
    {
      "epoch": 0.441,
      "grad_norm": 2.484020233154297,
      "learning_rate": 6.860079751934908e-05,
      "loss": 0.6544,
      "step": 2205
    },
    {
      "epoch": 0.441,
      "loss_align": 0.1727602481842041,
      "loss_contrastive": 0.07858341932296753,
      "loss_gen": 0.4298420250415802,
      "neg_sim_mean": 0.8058231472969055,
      "pos_sim_mean": 0.8272397518157959,
      "step": 2205
    },
    {
      "epoch": 0.4412,
      "grad_norm": 2.433109760284424,
      "learning_rate": 6.856839177751176e-05,
      "loss": 0.612,
      "step": 2206
    },
    {
      "epoch": 0.4412,
      "loss_align": 0.19377100467681885,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0014660042943432927,
      "neg_sim_mean": 0.38525986671447754,
      "pos_sim_mean": 0.8062289953231812,
      "step": 2206
    },
    {
      "epoch": 0.4414,
      "grad_norm": 0.5614091753959656,
      "learning_rate": 6.853597698566582e-05,
      "loss": 0.1952,
      "step": 2207
    },
    {
      "epoch": 0.4414,
      "loss_align": 0.15984630584716797,
      "loss_contrastive": 0.010330557823181152,
      "loss_gen": 0.2610233724117279,
      "neg_sim_mean": 0.7504842281341553,
      "pos_sim_mean": 0.840153694152832,
      "step": 2207
    },
    {
      "epoch": 0.4416,
      "grad_norm": 1.9848085641860962,
      "learning_rate": 6.850355315960992e-05,
      "loss": 0.4221,
      "step": 2208
    },
    {
      "epoch": 0.4416,
      "loss_align": 0.19179797172546387,
      "loss_contrastive": 0.12496238946914673,
      "loss_gen": 0.1761186271905899,
      "neg_sim_mean": 0.833164393901825,
      "pos_sim_mean": 0.8082020282745361,
      "step": 2208
    },
    {
      "epoch": 0.4418,
      "grad_norm": 2.077502489089966,
      "learning_rate": 6.847112031514698e-05,
      "loss": 0.3829,
      "step": 2209
    },
    {
      "epoch": 0.4418,
      "loss_align": 0.1724950075149536,
      "loss_contrastive": 0.057200729846954346,
      "loss_gen": 0.37842458486557007,
      "neg_sim_mean": 0.7847056984901428,
      "pos_sim_mean": 0.8275049924850464,
      "step": 2209
    },
    {
      "epoch": 0.442,
      "grad_norm": 2.540062427520752,
      "learning_rate": 6.843867846808438e-05,
      "loss": 0.5578,
      "step": 2210
    },
    {
      "epoch": 0.442,
      "loss_align": 0.10372304916381836,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01413307897746563,
      "neg_sim_mean": 0.2790358066558838,
      "pos_sim_mean": 0.8962769508361816,
      "step": 2210
    },
    {
      "epoch": 0.4422,
      "grad_norm": 0.822120189666748,
      "learning_rate": 6.840622763423391e-05,
      "loss": 0.1179,
      "step": 2211
    },
    {
      "epoch": 0.4422,
      "loss_align": 0.17911440134048462,
      "loss_contrastive": 0.1362789273262024,
      "loss_gen": 0.1505652815103531,
      "neg_sim_mean": 0.8571645021438599,
      "pos_sim_mean": 0.8208855986595154,
      "step": 2211
    },
    {
      "epoch": 0.4424,
      "grad_norm": 1.6323866844177246,
      "learning_rate": 6.837376782941167e-05,
      "loss": 0.346,
      "step": 2212
    },
    {
      "epoch": 0.4424,
      "loss_align": 0.13769274950027466,
      "loss_contrastive": 0.10648459196090698,
      "loss_gen": 0.2423212081193924,
      "neg_sim_mean": 0.8687918186187744,
      "pos_sim_mean": 0.8623072504997253,
      "step": 2212
    },
    {
      "epoch": 0.4426,
      "grad_norm": 2.127504348754883,
      "learning_rate": 6.834129906943822e-05,
      "loss": 0.3928,
      "step": 2213
    },
    {
      "epoch": 0.4426,
      "loss_align": 0.2519788146018982,
      "loss_contrastive": 0.1356261968612671,
      "loss_gen": 0.21061882376670837,
      "neg_sim_mean": 0.783647358417511,
      "pos_sim_mean": 0.7480211853981018,
      "step": 2213
    },
    {
      "epoch": 0.4428,
      "grad_norm": 2.0667343139648438,
      "learning_rate": 6.830882137013839e-05,
      "loss": 0.4789,
      "step": 2214
    },
    {
      "epoch": 0.4428,
      "loss_align": 0.07662707567214966,
      "loss_contrastive": 0.07518249750137329,
      "loss_gen": 0.15659168362617493,
      "neg_sim_mean": 0.8985553979873657,
      "pos_sim_mean": 0.9233729243278503,
      "step": 2214
    },
    {
      "epoch": 0.443,
      "grad_norm": 1.225982427597046,
      "learning_rate": 6.827633474734145e-05,
      "loss": 0.2422,
      "step": 2215
    },
    {
      "epoch": 0.443,
      "loss_align": 0.5005133152008057,
      "loss_contrastive": 0.4494773745536804,
      "loss_gen": 0.15332219004631042,
      "neg_sim_mean": 0.8489640951156616,
      "pos_sim_mean": 0.4994867146015167,
      "step": 2215
    },
    {
      "epoch": 0.4432,
      "grad_norm": 1.631399154663086,
      "learning_rate": 6.824383921688098e-05,
      "loss": 0.7078,
      "step": 2216
    },
    {
      "epoch": 0.4432,
      "loss_align": 0.11724942922592163,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1477842777967453,
      "neg_sim_mean": 0.7371784448623657,
      "pos_sim_mean": 0.8827505707740784,
      "step": 2216
    },
    {
      "epoch": 0.4434,
      "grad_norm": 1.2603380680084229,
      "learning_rate": 6.821133479459492e-05,
      "loss": 0.265,
      "step": 2217
    },
    {
      "epoch": 0.4434,
      "loss_align": 0.09208309650421143,
      "loss_contrastive": 0.013748645782470703,
      "loss_gen": 0.08800091594457626,
      "neg_sim_mean": 0.8216655254364014,
      "pos_sim_mean": 0.9079169034957886,
      "step": 2217
    },
    {
      "epoch": 0.4436,
      "grad_norm": 0.9080158472061157,
      "learning_rate": 6.81788214963255e-05,
      "loss": 0.1817,
      "step": 2218
    },
    {
      "epoch": 0.4436,
      "loss_align": 0.13198637962341309,
      "loss_contrastive": 0.07257115840911865,
      "loss_gen": 0.27075260877609253,
      "neg_sim_mean": 0.8405847549438477,
      "pos_sim_mean": 0.8680136203765869,
      "step": 2218
    },
    {
      "epoch": 0.4438,
      "grad_norm": 2.3053505420684814,
      "learning_rate": 6.814629933791931e-05,
      "loss": 0.4114,
      "step": 2219
    },
    {
      "epoch": 0.4438,
      "loss_align": 0.2903352975845337,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0021231756545603275,
      "neg_sim_mean": 0.2355707287788391,
      "pos_sim_mean": 0.7096647024154663,
      "step": 2219
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.603298008441925,
      "learning_rate": 6.811376833522729e-05,
      "loss": 0.2925,
      "step": 2220
    },
    {
      "epoch": 0.444,
      "loss_align": 0.1681535243988037,
      "loss_contrastive": 0.09465682506561279,
      "loss_gen": 0.2755167782306671,
      "neg_sim_mean": 0.8265032768249512,
      "pos_sim_mean": 0.8318464756011963,
      "step": 2220
    },
    {
      "epoch": 0.4442,
      "grad_norm": 1.8293484449386597,
      "learning_rate": 6.808122850410461e-05,
      "loss": 0.455,
      "step": 2221
    },
    {
      "epoch": 0.4442,
      "loss_align": 0.13043129444122314,
      "loss_contrastive": 0.016726016998291016,
      "loss_gen": 0.345681756734848,
      "neg_sim_mean": 0.78629469871521,
      "pos_sim_mean": 0.8695687055587769,
      "step": 2221
    },
    {
      "epoch": 0.4444,
      "grad_norm": 2.2203755378723145,
      "learning_rate": 6.804867986041083e-05,
      "loss": 0.4781,
      "step": 2222
    },
    {
      "epoch": 0.4444,
      "loss_align": 0.15916776657104492,
      "loss_contrastive": 0.03253030776977539,
      "loss_gen": 0.1523626744747162,
      "neg_sim_mean": 0.7733625173568726,
      "pos_sim_mean": 0.8408322334289551,
      "step": 2222
    },
    {
      "epoch": 0.4446,
      "grad_norm": 1.5417938232421875,
      "learning_rate": 6.801612242000974e-05,
      "loss": 0.3154,
      "step": 2223
    },
    {
      "epoch": 0.4446,
      "loss_align": 0.15414828062057495,
      "loss_contrastive": 0.033282577991485596,
      "loss_gen": 0.22908732295036316,
      "neg_sim_mean": 0.7791342735290527,
      "pos_sim_mean": 0.845851719379425,
      "step": 2223
    },
    {
      "epoch": 0.4448,
      "grad_norm": 1.8901317119598389,
      "learning_rate": 6.798355619876944e-05,
      "loss": 0.3872,
      "step": 2224
    },
    {
      "epoch": 0.4448,
      "loss_align": 0.15679121017456055,
      "loss_contrastive": 0.04090893268585205,
      "loss_gen": 0.12480853497982025,
      "neg_sim_mean": 0.7841176986694336,
      "pos_sim_mean": 0.8432087898254395,
      "step": 2224
    },
    {
      "epoch": 0.445,
      "grad_norm": 1.1696051359176636,
      "learning_rate": 6.795098121256232e-05,
      "loss": 0.2865,
      "step": 2225
    },
    {
      "epoch": 0.445,
      "loss_align": 0.13401341438293457,
      "loss_contrastive": 0.057895541191101074,
      "loss_gen": 0.1796015202999115,
      "neg_sim_mean": 0.8238821029663086,
      "pos_sim_mean": 0.8659865856170654,
      "step": 2225
    },
    {
      "epoch": 0.4452,
      "grad_norm": 1.3724044561386108,
      "learning_rate": 6.7918397477265e-05,
      "loss": 0.3206,
      "step": 2226
    },
    {
      "epoch": 0.4452,
      "loss_align": 0.04672598838806152,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19636408984661102,
      "neg_sim_mean": 0.7277234196662903,
      "pos_sim_mean": 0.9532740116119385,
      "step": 2226
    },
    {
      "epoch": 0.4454,
      "grad_norm": 1.384148359298706,
      "learning_rate": 6.788580500875847e-05,
      "loss": 0.2431,
      "step": 2227
    },
    {
      "epoch": 0.4454,
      "loss_align": 0.19968384504318237,
      "loss_contrastive": 0.04287254810333252,
      "loss_gen": 0.2829383611679077,
      "neg_sim_mean": 0.7431886792182922,
      "pos_sim_mean": 0.8003161549568176,
      "step": 2227
    },
    {
      "epoch": 0.4456,
      "grad_norm": 1.758322834968567,
      "learning_rate": 6.785320382292783e-05,
      "loss": 0.4878,
      "step": 2228
    },
    {
      "epoch": 0.4456,
      "loss_align": 0.20272237062454224,
      "loss_contrastive": 0.04807180166244507,
      "loss_gen": 0.09772723913192749,
      "neg_sim_mean": 0.7453494071960449,
      "pos_sim_mean": 0.7972776293754578,
      "step": 2228
    },
    {
      "epoch": 0.4458,
      "grad_norm": 1.385088562965393,
      "learning_rate": 6.782059393566253e-05,
      "loss": 0.3062,
      "step": 2229
    },
    {
      "epoch": 0.4458,
      "loss_align": 0.1434476375579834,
      "loss_contrastive": 0.0416523814201355,
      "loss_gen": 0.16449673473834991,
      "neg_sim_mean": 0.7982047200202942,
      "pos_sim_mean": 0.8565523624420166,
      "step": 2229
    },
    {
      "epoch": 0.446,
      "grad_norm": 1.5853562355041504,
      "learning_rate": 6.778797536285624e-05,
      "loss": 0.3129,
      "step": 2230
    },
    {
      "epoch": 0.446,
      "loss_align": 0.08767497539520264,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3001021444797516,
      "neg_sim_mean": 0.6334004402160645,
      "pos_sim_mean": 0.9123250246047974,
      "step": 2230
    },
    {
      "epoch": 0.4462,
      "grad_norm": 2.022346019744873,
      "learning_rate": 6.775534812040685e-05,
      "loss": 0.3878,
      "step": 2231
    },
    {
      "epoch": 0.4462,
      "loss_align": 0.07699108123779297,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1459621638059616,
      "neg_sim_mean": 0.7435793876647949,
      "pos_sim_mean": 0.923008918762207,
      "step": 2231
    },
    {
      "epoch": 0.4464,
      "grad_norm": 1.334755778312683,
      "learning_rate": 6.772271222421649e-05,
      "loss": 0.223,
      "step": 2232
    },
    {
      "epoch": 0.4464,
      "loss_align": 0.18299221992492676,
      "loss_contrastive": 0.09801274538040161,
      "loss_gen": 0.26869532465934753,
      "neg_sim_mean": 0.8150205016136169,
      "pos_sim_mean": 0.8170077800750732,
      "step": 2232
    },
    {
      "epoch": 0.4466,
      "grad_norm": 2.0372748374938965,
      "learning_rate": 6.769006769019148e-05,
      "loss": 0.4634,
      "step": 2233
    },
    {
      "epoch": 0.4466,
      "loss_align": 0.22805935144424438,
      "loss_contrastive": 0.0446968674659729,
      "loss_gen": 0.3028155565261841,
      "neg_sim_mean": 0.7166374921798706,
      "pos_sim_mean": 0.7719406485557556,
      "step": 2233
    },
    {
      "epoch": 0.4468,
      "grad_norm": 2.4321446418762207,
      "learning_rate": 6.765741453424237e-05,
      "loss": 0.5362,
      "step": 2234
    },
    {
      "epoch": 0.4468,
      "loss_align": 0.16284435987472534,
      "loss_contrastive": 0.0,
      "loss_gen": 0.019580738618969917,
      "neg_sim_mean": 0.3115098178386688,
      "pos_sim_mean": 0.8371556401252747,
      "step": 2234
    },
    {
      "epoch": 0.447,
      "grad_norm": 0.984735369682312,
      "learning_rate": 6.762475277228392e-05,
      "loss": 0.1824,
      "step": 2235
    },
    {
      "epoch": 0.447,
      "loss_align": 0.14092206954956055,
      "loss_contrastive": 0.0,
      "loss_gen": 0.33890411257743835,
      "neg_sim_mean": 0.4679194688796997,
      "pos_sim_mean": 0.8590779304504395,
      "step": 2235
    },
    {
      "epoch": 0.4472,
      "grad_norm": 2.357485055923462,
      "learning_rate": 6.759208242023509e-05,
      "loss": 0.4798,
      "step": 2236
    },
    {
      "epoch": 0.4472,
      "loss_align": 0.17355185747146606,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01480785571038723,
      "neg_sim_mean": 0.41277605295181274,
      "pos_sim_mean": 0.8264481425285339,
      "step": 2236
    },
    {
      "epoch": 0.4474,
      "grad_norm": 0.9100253582000732,
      "learning_rate": 6.755940349401901e-05,
      "loss": 0.1884,
      "step": 2237
    },
    {
      "epoch": 0.4474,
      "loss_align": 0.09639275074005127,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24258163571357727,
      "neg_sim_mean": 0.7876144647598267,
      "pos_sim_mean": 0.9036072492599487,
      "step": 2237
    },
    {
      "epoch": 0.4476,
      "grad_norm": 1.7989354133605957,
      "learning_rate": 6.752671600956295e-05,
      "loss": 0.339,
      "step": 2238
    },
    {
      "epoch": 0.4476,
      "loss_align": 0.31070423126220703,
      "loss_contrastive": 0.059461891651153564,
      "loss_gen": 0.1589120775461197,
      "neg_sim_mean": 0.6487576365470886,
      "pos_sim_mean": 0.689295768737793,
      "step": 2238
    },
    {
      "epoch": 0.4478,
      "grad_norm": 1.7323817014694214,
      "learning_rate": 6.749401998279846e-05,
      "loss": 0.4768,
      "step": 2239
    },
    {
      "epoch": 0.4478,
      "loss_align": 0.13148772716522217,
      "loss_contrastive": 0.10264134407043457,
      "loss_gen": 0.10796165466308594,
      "neg_sim_mean": 0.8711535930633545,
      "pos_sim_mean": 0.8685122728347778,
      "step": 2239
    },
    {
      "epoch": 0.448,
      "grad_norm": 1.4479176998138428,
      "learning_rate": 6.746131542966112e-05,
      "loss": 0.2518,
      "step": 2240
    },
    {
      "epoch": 0.448,
      "loss_align": 0.11734241247177124,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1943877637386322,
      "neg_sim_mean": 0.6520842909812927,
      "pos_sim_mean": 0.8826575875282288,
      "step": 2240
    },
    {
      "epoch": 0.4482,
      "grad_norm": 1.3892337083816528,
      "learning_rate": 6.742860236609077e-05,
      "loss": 0.3117,
      "step": 2241
    },
    {
      "epoch": 0.4482,
      "loss_align": 0.18097907304763794,
      "loss_contrastive": 0.0951196551322937,
      "loss_gen": 0.42216548323631287,
      "neg_sim_mean": 0.8141405582427979,
      "pos_sim_mean": 0.8190209269523621,
      "step": 2241
    },
    {
      "epoch": 0.4484,
      "grad_norm": 2.7358038425445557,
      "learning_rate": 6.739588080803134e-05,
      "loss": 0.6146,
      "step": 2242
    },
    {
      "epoch": 0.4484,
      "loss_align": 0.16340118646621704,
      "loss_contrastive": 0.15685147047042847,
      "loss_gen": 0.1947091817855835,
      "neg_sim_mean": 0.8934502601623535,
      "pos_sim_mean": 0.836598813533783,
      "step": 2242
    },
    {
      "epoch": 0.4486,
      "grad_norm": 1.5946593284606934,
      "learning_rate": 6.736315077143096e-05,
      "loss": 0.3769,
      "step": 2243
    },
    {
      "epoch": 0.4486,
      "loss_align": 0.18479180335998535,
      "loss_contrastive": 0.010824263095855713,
      "loss_gen": 0.2738840579986572,
      "neg_sim_mean": 0.7260324358940125,
      "pos_sim_mean": 0.8152081966400146,
      "step": 2243
    },
    {
      "epoch": 0.4488,
      "grad_norm": 2.132218599319458,
      "learning_rate": 6.733041227224181e-05,
      "loss": 0.46,
      "step": 2244
    },
    {
      "epoch": 0.4488,
      "loss_align": 0.0835229754447937,
      "loss_contrastive": 0.09716743230819702,
      "loss_gen": 0.19108907878398895,
      "neg_sim_mean": 0.9136444330215454,
      "pos_sim_mean": 0.9164770245552063,
      "step": 2244
    },
    {
      "epoch": 0.449,
      "grad_norm": 1.8356037139892578,
      "learning_rate": 6.729766532642025e-05,
      "loss": 0.2863,
      "step": 2245
    },
    {
      "epoch": 0.449,
      "loss_align": 0.15842074155807495,
      "loss_contrastive": 0.0026915669441223145,
      "loss_gen": 0.24804840981960297,
      "neg_sim_mean": 0.7442708015441895,
      "pos_sim_mean": 0.841579258441925,
      "step": 2245
    },
    {
      "epoch": 0.4492,
      "grad_norm": 1.9761096239089966,
      "learning_rate": 6.726490994992674e-05,
      "loss": 0.4068,
      "step": 2246
    },
    {
      "epoch": 0.4492,
      "loss_align": 0.17639625072479248,
      "loss_contrastive": 0.0,
      "loss_gen": 0.05185306444764137,
      "neg_sim_mean": 0.4080280065536499,
      "pos_sim_mean": 0.8236037492752075,
      "step": 2246
    },
    {
      "epoch": 0.4494,
      "grad_norm": 1.7436901330947876,
      "learning_rate": 6.723214615872585e-05,
      "loss": 0.2282,
      "step": 2247
    },
    {
      "epoch": 0.4494,
      "loss_align": 0.1430705189704895,
      "loss_contrastive": 0.11408478021621704,
      "loss_gen": 0.275645911693573,
      "neg_sim_mean": 0.8710142374038696,
      "pos_sim_mean": 0.8569294810295105,
      "step": 2247
    },
    {
      "epoch": 0.4496,
      "grad_norm": 2.049276351928711,
      "learning_rate": 6.719937396878627e-05,
      "loss": 0.4324,
      "step": 2248
    },
    {
      "epoch": 0.4496,
      "loss_align": 0.10627084970474243,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03750435635447502,
      "neg_sim_mean": 0.3954889476299286,
      "pos_sim_mean": 0.8937291502952576,
      "step": 2248
    },
    {
      "epoch": 0.4498,
      "grad_norm": 1.3231263160705566,
      "learning_rate": 6.716659339608077e-05,
      "loss": 0.1438,
      "step": 2249
    },
    {
      "epoch": 0.4498,
      "loss_align": 0.12559998035430908,
      "loss_contrastive": 0.05488461256027222,
      "loss_gen": 0.307547926902771,
      "neg_sim_mean": 0.8292846083641052,
      "pos_sim_mean": 0.8744000196456909,
      "step": 2249
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.1235225200653076,
      "learning_rate": 6.713380445658618e-05,
      "loss": 0.4397,
      "step": 2250
    },
    {
      "epoch": 0.45,
      "loss_align": 0.21797943115234375,
      "loss_contrastive": 0.18816351890563965,
      "loss_gen": 0.27530670166015625,
      "neg_sim_mean": 0.870184063911438,
      "pos_sim_mean": 0.7820205688476562,
      "step": 2250
    },
    {
      "epoch": 0.4502,
      "grad_norm": 1.8068394660949707,
      "learning_rate": 6.710100716628344e-05,
      "loss": 0.5159,
      "step": 2251
    },
    {
      "epoch": 0.4502,
      "loss_align": 0.1823328733444214,
      "loss_contrastive": 0.0,
      "loss_gen": 0.5131062865257263,
      "neg_sim_mean": 0.7172564268112183,
      "pos_sim_mean": 0.8176671266555786,
      "step": 2251
    },
    {
      "epoch": 0.4504,
      "grad_norm": 2.46465802192688,
      "learning_rate": 6.706820154115754e-05,
      "loss": 0.6954,
      "step": 2252
    },
    {
      "epoch": 0.4504,
      "loss_align": 0.26522254943847656,
      "loss_contrastive": 0.14660245180130005,
      "loss_gen": 0.5010817050933838,
      "neg_sim_mean": 0.7813798785209656,
      "pos_sim_mean": 0.7347774505615234,
      "step": 2252
    },
    {
      "epoch": 0.4506,
      "grad_norm": 2.6422955989837646,
      "learning_rate": 6.70353875971976e-05,
      "loss": 0.7839,
      "step": 2253
    },
    {
      "epoch": 0.4506,
      "loss_align": 0.1850736141204834,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23622404038906097,
      "neg_sim_mean": 0.41092759370803833,
      "pos_sim_mean": 0.8149263858795166,
      "step": 2253
    },
    {
      "epoch": 0.4508,
      "grad_norm": 1.7629750967025757,
      "learning_rate": 6.700256535039664e-05,
      "loss": 0.4213,
      "step": 2254
    },
    {
      "epoch": 0.4508,
      "loss_align": 0.14547914266586304,
      "loss_contrastive": 0.04187649488449097,
      "loss_gen": 0.14317815005779266,
      "neg_sim_mean": 0.79639732837677,
      "pos_sim_mean": 0.854520857334137,
      "step": 2254
    },
    {
      "epoch": 0.451,
      "grad_norm": 1.4790070056915283,
      "learning_rate": 6.696973481675191e-05,
      "loss": 0.2937,
      "step": 2255
    },
    {
      "epoch": 0.451,
      "loss_align": 0.26650673151016235,
      "loss_contrastive": 0.02219754457473755,
      "loss_gen": 0.3577496409416199,
      "neg_sim_mean": 0.6556907892227173,
      "pos_sim_mean": 0.7334932684898376,
      "step": 2255
    },
    {
      "epoch": 0.4512,
      "grad_norm": 2.5671184062957764,
      "learning_rate": 6.693689601226458e-05,
      "loss": 0.6269,
      "step": 2256
    },
    {
      "epoch": 0.4512,
      "loss_align": 0.09806203842163086,
      "loss_contrastive": 0.08630907535552979,
      "loss_gen": 0.17046697437763214,
      "neg_sim_mean": 0.888247013092041,
      "pos_sim_mean": 0.9019379615783691,
      "step": 2256
    },
    {
      "epoch": 0.4514,
      "grad_norm": 1.4092618227005005,
      "learning_rate": 6.690404895293987e-05,
      "loss": 0.2789,
      "step": 2257
    },
    {
      "epoch": 0.4514,
      "loss_align": 0.1708289384841919,
      "loss_contrastive": 0.15697133541107178,
      "loss_gen": 0.1440281718969345,
      "neg_sim_mean": 0.886142373085022,
      "pos_sim_mean": 0.8291710615158081,
      "step": 2257
    },
    {
      "epoch": 0.4516,
      "grad_norm": 1.4736077785491943,
      "learning_rate": 6.687119365478706e-05,
      "loss": 0.3337,
      "step": 2258
    },
    {
      "epoch": 0.4516,
      "loss_align": 0.1580866575241089,
      "loss_contrastive": 0.10229265689849854,
      "loss_gen": 0.21078352630138397,
      "neg_sim_mean": 0.8442059755325317,
      "pos_sim_mean": 0.8419133424758911,
      "step": 2258
    },
    {
      "epoch": 0.4518,
      "grad_norm": 2.0600340366363525,
      "learning_rate": 6.683833013381941e-05,
      "loss": 0.3811,
      "step": 2259
    },
    {
      "epoch": 0.4518,
      "loss_align": 0.26116734743118286,
      "loss_contrastive": 0.1535550355911255,
      "loss_gen": 0.1739492565393448,
      "neg_sim_mean": 0.7923876643180847,
      "pos_sim_mean": 0.7388326525688171,
      "step": 2259
    },
    {
      "epoch": 0.452,
      "grad_norm": 1.730128288269043,
      "learning_rate": 6.680545840605423e-05,
      "loss": 0.4535,
      "step": 2260
    },
    {
      "epoch": 0.452,
      "loss_align": 0.11300486326217651,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002799590118229389,
      "neg_sim_mean": 0.33888959884643555,
      "pos_sim_mean": 0.8869951367378235,
      "step": 2260
    },
    {
      "epoch": 0.4522,
      "grad_norm": 0.45011085271835327,
      "learning_rate": 6.677257848751277e-05,
      "loss": 0.1158,
      "step": 2261
    },
    {
      "epoch": 0.4522,
      "loss_align": 0.2404012680053711,
      "loss_contrastive": 0.09152430295944214,
      "loss_gen": 0.35636261105537415,
      "neg_sim_mean": 0.7511230111122131,
      "pos_sim_mean": 0.7595987319946289,
      "step": 2261
    },
    {
      "epoch": 0.4524,
      "grad_norm": 2.1769096851348877,
      "learning_rate": 6.67396903942203e-05,
      "loss": 0.6077,
      "step": 2262
    },
    {
      "epoch": 0.4524,
      "loss_align": 0.30029022693634033,
      "loss_contrastive": 0.13631188869476318,
      "loss_gen": 0.19286876916885376,
      "neg_sim_mean": 0.7360216379165649,
      "pos_sim_mean": 0.6997097730636597,
      "step": 2262
    },
    {
      "epoch": 0.4526,
      "grad_norm": 2.4949686527252197,
      "learning_rate": 6.670679414220609e-05,
      "loss": 0.5095,
      "step": 2263
    },
    {
      "epoch": 0.4526,
      "loss_align": 0.09223997592926025,
      "loss_contrastive": 0.011083364486694336,
      "loss_gen": 0.19208188354969025,
      "neg_sim_mean": 0.8188433647155762,
      "pos_sim_mean": 0.9077600240707397,
      "step": 2263
    },
    {
      "epoch": 0.4528,
      "grad_norm": 1.7856425046920776,
      "learning_rate": 6.667388974750337e-05,
      "loss": 0.2857,
      "step": 2264
    },
    {
      "epoch": 0.4528,
      "loss_align": 0.1389039158821106,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19559207558631897,
      "neg_sim_mean": 0.688734769821167,
      "pos_sim_mean": 0.8610960841178894,
      "step": 2264
    },
    {
      "epoch": 0.453,
      "grad_norm": 1.4567476511001587,
      "learning_rate": 6.664097722614934e-05,
      "loss": 0.3345,
      "step": 2265
    },
    {
      "epoch": 0.453,
      "loss_align": 0.2337815761566162,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2138025015592575,
      "neg_sim_mean": 0.6645012497901917,
      "pos_sim_mean": 0.7662184238433838,
      "step": 2265
    },
    {
      "epoch": 0.4532,
      "grad_norm": 1.7045652866363525,
      "learning_rate": 6.660805659418516e-05,
      "loss": 0.4476,
      "step": 2266
    },
    {
      "epoch": 0.4532,
      "loss_align": 0.17453795671463013,
      "loss_contrastive": 0.034910619258880615,
      "loss_gen": 0.22574174404144287,
      "neg_sim_mean": 0.7603726387023926,
      "pos_sim_mean": 0.8254620432853699,
      "step": 2266
    },
    {
      "epoch": 0.4534,
      "grad_norm": 1.7604933977127075,
      "learning_rate": 6.657512786765599e-05,
      "loss": 0.4045,
      "step": 2267
    },
    {
      "epoch": 0.4534,
      "loss_align": 0.11282885074615479,
      "loss_contrastive": 0.13427412509918213,
      "loss_gen": 0.35138317942619324,
      "neg_sim_mean": 0.9214452505111694,
      "pos_sim_mean": 0.8871711492538452,
      "step": 2267
    },
    {
      "epoch": 0.4536,
      "grad_norm": 2.033834934234619,
      "learning_rate": 6.654219106261082e-05,
      "loss": 0.4803,
      "step": 2268
    },
    {
      "epoch": 0.4536,
      "loss_align": 0.17962145805358887,
      "loss_contrastive": 0.14430880546569824,
      "loss_gen": 0.18000547587871552,
      "neg_sim_mean": 0.8646873235702515,
      "pos_sim_mean": 0.8203785419464111,
      "step": 2268
    },
    {
      "epoch": 0.4538,
      "grad_norm": 1.8173903226852417,
      "learning_rate": 6.650924619510269e-05,
      "loss": 0.3769,
      "step": 2269
    },
    {
      "epoch": 0.4538,
      "loss_align": 0.10578042268753052,
      "loss_contrastive": 0.10580557584762573,
      "loss_gen": 0.2990017533302307,
      "neg_sim_mean": 0.9000251293182373,
      "pos_sim_mean": 0.8942195773124695,
      "step": 2269
    },
    {
      "epoch": 0.454,
      "grad_norm": 2.0856235027313232,
      "learning_rate": 6.647629328118851e-05,
      "loss": 0.4175,
      "step": 2270
    },
    {
      "epoch": 0.454,
      "loss_align": 0.21356737613677979,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0042802984826266766,
      "neg_sim_mean": 0.49842438101768494,
      "pos_sim_mean": 0.7864326238632202,
      "step": 2270
    },
    {
      "epoch": 0.4542,
      "grad_norm": 0.5760250091552734,
      "learning_rate": 6.644333233692916e-05,
      "loss": 0.2178,
      "step": 2271
    },
    {
      "epoch": 0.4542,
      "loss_align": 0.12488925457000732,
      "loss_contrastive": 0.056632161140441895,
      "loss_gen": 0.21484129130840302,
      "neg_sim_mean": 0.8317428827285767,
      "pos_sim_mean": 0.8751107454299927,
      "step": 2271
    },
    {
      "epoch": 0.4544,
      "grad_norm": 1.741165041923523,
      "learning_rate": 6.64103633783894e-05,
      "loss": 0.3465,
      "step": 2272
    },
    {
      "epoch": 0.4544,
      "loss_align": 0.11611014604568481,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2072378695011139,
      "neg_sim_mean": 0.6029267907142639,
      "pos_sim_mean": 0.8838898539543152,
      "step": 2272
    },
    {
      "epoch": 0.4546,
      "grad_norm": 2.0745744705200195,
      "learning_rate": 6.637738642163784e-05,
      "loss": 0.3233,
      "step": 2273
    },
    {
      "epoch": 0.4546,
      "loss_align": 0.22856903076171875,
      "loss_contrastive": 0.16067051887512207,
      "loss_gen": 0.2541951537132263,
      "neg_sim_mean": 0.8321014642715454,
      "pos_sim_mean": 0.7714309692382812,
      "step": 2273
    },
    {
      "epoch": 0.4548,
      "grad_norm": 2.2025818824768066,
      "learning_rate": 6.634440148274713e-05,
      "loss": 0.502,
      "step": 2274
    },
    {
      "epoch": 0.4548,
      "loss_align": 0.1221245527267456,
      "loss_contrastive": 0.07801604270935059,
      "loss_gen": 0.22235602140426636,
      "neg_sim_mean": 0.8558914661407471,
      "pos_sim_mean": 0.8778754472732544,
      "step": 2274
    },
    {
      "epoch": 0.455,
      "grad_norm": 1.6634503602981567,
      "learning_rate": 6.631140857779368e-05,
      "loss": 0.3538,
      "step": 2275
    },
    {
      "epoch": 0.455,
      "loss_align": 0.12349224090576172,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007484428118914366,
      "neg_sim_mean": 0.32378169894218445,
      "pos_sim_mean": 0.8765077590942383,
      "step": 2275
    },
    {
      "epoch": 0.4552,
      "grad_norm": 0.6097941994667053,
      "learning_rate": 6.627840772285784e-05,
      "loss": 0.131,
      "step": 2276
    },
    {
      "epoch": 0.4552,
      "loss_align": 0.08983272314071655,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14371098577976227,
      "neg_sim_mean": 0.7277461290359497,
      "pos_sim_mean": 0.9101672768592834,
      "step": 2276
    },
    {
      "epoch": 0.4554,
      "grad_norm": 1.45820152759552,
      "learning_rate": 6.624539893402382e-05,
      "loss": 0.2335,
      "step": 2277
    },
    {
      "epoch": 0.4554,
      "loss_align": 0.11878204345703125,
      "loss_contrastive": 0.05333399772644043,
      "loss_gen": 0.14302489161491394,
      "neg_sim_mean": 0.8345519304275513,
      "pos_sim_mean": 0.8812179565429688,
      "step": 2277
    },
    {
      "epoch": 0.4556,
      "grad_norm": 1.6794253587722778,
      "learning_rate": 6.621238222737972e-05,
      "loss": 0.2682,
      "step": 2278
    },
    {
      "epoch": 0.4556,
      "loss_align": 0.18660622835159302,
      "loss_contrastive": 0.06761366128921509,
      "loss_gen": 0.2454586774110794,
      "neg_sim_mean": 0.7810074090957642,
      "pos_sim_mean": 0.813393771648407,
      "step": 2278
    },
    {
      "epoch": 0.4558,
      "grad_norm": 1.6824582815170288,
      "learning_rate": 6.617935761901748e-05,
      "loss": 0.4402,
      "step": 2279
    },
    {
      "epoch": 0.4558,
      "loss_align": 0.09675371646881104,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2958611845970154,
      "neg_sim_mean": 0.7790403366088867,
      "pos_sim_mean": 0.903246283531189,
      "step": 2279
    },
    {
      "epoch": 0.456,
      "grad_norm": 1.6198521852493286,
      "learning_rate": 6.614632512503288e-05,
      "loss": 0.3926,
      "step": 2280
    },
    {
      "epoch": 0.456,
      "loss_align": 0.2922738790512085,
      "loss_contrastive": 0.08135223388671875,
      "loss_gen": 0.22095943987369537,
      "neg_sim_mean": 0.6890783309936523,
      "pos_sim_mean": 0.7077261209487915,
      "step": 2280
    },
    {
      "epoch": 0.4562,
      "grad_norm": 2.0568153858184814,
      "learning_rate": 6.611328476152557e-05,
      "loss": 0.523,
      "step": 2281
    },
    {
      "epoch": 0.4562,
      "loss_align": 0.3424842953681946,
      "loss_contrastive": 0.02396923303604126,
      "loss_gen": 0.15490801632404327,
      "neg_sim_mean": 0.5814849138259888,
      "pos_sim_mean": 0.6575157046318054,
      "step": 2281
    },
    {
      "epoch": 0.4564,
      "grad_norm": 1.6307977437973022,
      "learning_rate": 6.608023654459899e-05,
      "loss": 0.5003,
      "step": 2282
    },
    {
      "epoch": 0.4564,
      "loss_align": 0.27722418308258057,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29278287291526794,
      "neg_sim_mean": 0.5950148105621338,
      "pos_sim_mean": 0.7227758169174194,
      "step": 2282
    },
    {
      "epoch": 0.4566,
      "grad_norm": 1.789478063583374,
      "learning_rate": 6.604718049036048e-05,
      "loss": 0.57,
      "step": 2283
    },
    {
      "epoch": 0.4566,
      "loss_align": 0.2190706729888916,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04566919431090355,
      "neg_sim_mean": 0.5058499574661255,
      "pos_sim_mean": 0.7809293270111084,
      "step": 2283
    },
    {
      "epoch": 0.4568,
      "grad_norm": 1.4236148595809937,
      "learning_rate": 6.601411661492114e-05,
      "loss": 0.2647,
      "step": 2284
    },
    {
      "epoch": 0.4568,
      "loss_align": 0.15915709733963013,
      "loss_contrastive": 0.10507398843765259,
      "loss_gen": 0.2010304182767868,
      "neg_sim_mean": 0.8459168672561646,
      "pos_sim_mean": 0.8408429026603699,
      "step": 2284
    },
    {
      "epoch": 0.457,
      "grad_norm": 1.7498221397399902,
      "learning_rate": 6.59810449343959e-05,
      "loss": 0.3728,
      "step": 2285
    },
    {
      "epoch": 0.457,
      "loss_align": 0.09023207426071167,
      "loss_contrastive": 0.02464240789413452,
      "loss_gen": 0.232479989528656,
      "neg_sim_mean": 0.8344103097915649,
      "pos_sim_mean": 0.9097679257392883,
      "step": 2285
    },
    {
      "epoch": 0.4572,
      "grad_norm": 2.1970834732055664,
      "learning_rate": 6.594796546490351e-05,
      "loss": 0.3257,
      "step": 2286
    },
    {
      "epoch": 0.4572,
      "loss_align": 0.1943349838256836,
      "loss_contrastive": 0.008937656879425049,
      "loss_gen": 0.26500603556632996,
      "neg_sim_mean": 0.7146026492118835,
      "pos_sim_mean": 0.8056650161743164,
      "step": 2286
    },
    {
      "epoch": 0.4574,
      "grad_norm": 1.821365475654602,
      "learning_rate": 6.591487822256648e-05,
      "loss": 0.4604,
      "step": 2287
    },
    {
      "epoch": 0.4574,
      "loss_align": 0.09314197301864624,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19951367378234863,
      "neg_sim_mean": 0.7153511047363281,
      "pos_sim_mean": 0.9068580269813538,
      "step": 2287
    },
    {
      "epoch": 0.4576,
      "grad_norm": 1.6495128870010376,
      "learning_rate": 6.588178322351113e-05,
      "loss": 0.2927,
      "step": 2288
    },
    {
      "epoch": 0.4576,
      "loss_align": 0.11563968658447266,
      "loss_contrastive": 0.0,
      "loss_gen": 0.017717143520712852,
      "neg_sim_mean": 0.3570265769958496,
      "pos_sim_mean": 0.8843603134155273,
      "step": 2288
    },
    {
      "epoch": 0.4578,
      "grad_norm": 0.7618797421455383,
      "learning_rate": 6.58486804838676e-05,
      "loss": 0.1334,
      "step": 2289
    },
    {
      "epoch": 0.4578,
      "loss_align": 0.20962589979171753,
      "loss_contrastive": 0.03473562002182007,
      "loss_gen": 0.13996219635009766,
      "neg_sim_mean": 0.7251096963882446,
      "pos_sim_mean": 0.7903741002082825,
      "step": 2289
    },
    {
      "epoch": 0.458,
      "grad_norm": 1.2916641235351562,
      "learning_rate": 6.58155700197697e-05,
      "loss": 0.3538,
      "step": 2290
    },
    {
      "epoch": 0.458,
      "loss_align": 0.2769460082054138,
      "loss_contrastive": 0.0723537802696228,
      "loss_gen": 0.4372478127479553,
      "neg_sim_mean": 0.6954077482223511,
      "pos_sim_mean": 0.7230539917945862,
      "step": 2290
    },
    {
      "epoch": 0.4582,
      "grad_norm": 2.7800192832946777,
      "learning_rate": 6.578245184735513e-05,
      "loss": 0.7229,
      "step": 2291
    },
    {
      "epoch": 0.4582,
      "loss_align": 0.18936288356781006,
      "loss_contrastive": 0.0858469009399414,
      "loss_gen": 0.1772390455007553,
      "neg_sim_mean": 0.7964839935302734,
      "pos_sim_mean": 0.8106371164321899,
      "step": 2291
    },
    {
      "epoch": 0.4584,
      "grad_norm": 1.6011402606964111,
      "learning_rate": 6.574932598276525e-05,
      "loss": 0.3769,
      "step": 2292
    },
    {
      "epoch": 0.4584,
      "loss_align": 0.19840645790100098,
      "loss_contrastive": 0.17972254753112793,
      "loss_gen": 0.2785719931125641,
      "neg_sim_mean": 0.881316065788269,
      "pos_sim_mean": 0.801593542098999,
      "step": 2292
    },
    {
      "epoch": 0.4586,
      "grad_norm": 1.8636200428009033,
      "learning_rate": 6.57161924421452e-05,
      "loss": 0.4985,
      "step": 2293
    },
    {
      "epoch": 0.4586,
      "loss_align": 0.11124330759048462,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03988935798406601,
      "neg_sim_mean": 0.33241361379623413,
      "pos_sim_mean": 0.8887566924095154,
      "step": 2293
    },
    {
      "epoch": 0.4588,
      "grad_norm": 1.506708025932312,
      "learning_rate": 6.568305124164389e-05,
      "loss": 0.1511,
      "step": 2294
    },
    {
      "epoch": 0.4588,
      "loss_align": 0.13148713111877441,
      "loss_contrastive": 0.08141028881072998,
      "loss_gen": 0.1736574023962021,
      "neg_sim_mean": 0.8499231338500977,
      "pos_sim_mean": 0.8685128688812256,
      "step": 2294
    },
    {
      "epoch": 0.459,
      "grad_norm": 1.6229259967803955,
      "learning_rate": 6.564990239741391e-05,
      "loss": 0.3149,
      "step": 2295
    },
    {
      "epoch": 0.459,
      "loss_align": 0.2220064401626587,
      "loss_contrastive": 0.07531559467315674,
      "loss_gen": 0.17801178991794586,
      "neg_sim_mean": 0.7533091306686401,
      "pos_sim_mean": 0.7779935598373413,
      "step": 2295
    },
    {
      "epoch": 0.4592,
      "grad_norm": 1.6528067588806152,
      "learning_rate": 6.561674592561163e-05,
      "loss": 0.4091,
      "step": 2296
    },
    {
      "epoch": 0.4592,
      "loss_align": 0.12481081485748291,
      "loss_contrastive": 0.07848000526428223,
      "loss_gen": 0.25321996212005615,
      "neg_sim_mean": 0.8536691665649414,
      "pos_sim_mean": 0.8751891851425171,
      "step": 2296
    },
    {
      "epoch": 0.4594,
      "grad_norm": 1.864292860031128,
      "learning_rate": 6.55835818423971e-05,
      "loss": 0.3874,
      "step": 2297
    },
    {
      "epoch": 0.4594,
      "loss_align": 0.1714293360710144,
      "loss_contrastive": 0.09539669752120972,
      "loss_gen": 0.26592525839805603,
      "neg_sim_mean": 0.8239673376083374,
      "pos_sim_mean": 0.8285706639289856,
      "step": 2297
    },
    {
      "epoch": 0.4596,
      "grad_norm": 1.6924936771392822,
      "learning_rate": 6.55504101639341e-05,
      "loss": 0.4488,
      "step": 2298
    },
    {
      "epoch": 0.4596,
      "loss_align": 0.24037694931030273,
      "loss_contrastive": 0.0561138391494751,
      "loss_gen": 0.34461894631385803,
      "neg_sim_mean": 0.7157368659973145,
      "pos_sim_mean": 0.7596230506896973,
      "step": 2298
    },
    {
      "epoch": 0.4598,
      "grad_norm": 1.8472057580947876,
      "learning_rate": 6.551723090639007e-05,
      "loss": 0.5917,
      "step": 2299
    },
    {
      "epoch": 0.4598,
      "loss_align": 0.1913354992866516,
      "loss_contrastive": 0.06988316774368286,
      "loss_gen": 0.21980027854442596,
      "neg_sim_mean": 0.7785476446151733,
      "pos_sim_mean": 0.8086645007133484,
      "step": 2299
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.937591791152954,
      "learning_rate": 6.548404408593621e-05,
      "loss": 0.4195,
      "step": 2300
    },
    {
      "epoch": 0.46,
      "loss_align": 0.142555832862854,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16759547591209412,
      "neg_sim_mean": 0.633228063583374,
      "pos_sim_mean": 0.857444167137146,
      "step": 2300
    },
    {
      "epoch": 0.4602,
      "grad_norm": 1.8826621770858765,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.3102,
      "step": 2301
    },
    {
      "epoch": 0.4602,
      "loss_align": 0.38236773014068604,
      "loss_contrastive": 0.07909983396530151,
      "loss_gen": 0.3401808440685272,
      "neg_sim_mean": 0.5967320799827576,
      "pos_sim_mean": 0.617632269859314,
      "step": 2301
    },
    {
      "epoch": 0.4604,
      "grad_norm": 2.258342981338501,
      "learning_rate": 6.541764782100208e-05,
      "loss": 0.732,
      "step": 2302
    },
    {
      "epoch": 0.4604,
      "loss_align": 0.1895458698272705,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004370367154479027,
      "neg_sim_mean": 0.34253984689712524,
      "pos_sim_mean": 0.8104541301727295,
      "step": 2302
    },
    {
      "epoch": 0.4606,
      "grad_norm": 1.0339899063110352,
      "learning_rate": 6.538443840888254e-05,
      "loss": 0.1939,
      "step": 2303
    },
    {
      "epoch": 0.4606,
      "loss_align": 0.13134312629699707,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2723951041698456,
      "neg_sim_mean": 0.7683295011520386,
      "pos_sim_mean": 0.8686568737030029,
      "step": 2303
    },
    {
      "epoch": 0.4608,
      "grad_norm": 1.8680111169815063,
      "learning_rate": 6.53512214985746e-05,
      "loss": 0.4037,
      "step": 2304
    },
    {
      "epoch": 0.4608,
      "loss_align": 0.09926998615264893,
      "loss_contrastive": 0.010869622230529785,
      "loss_gen": 0.4110755920410156,
      "neg_sim_mean": 0.811599612236023,
      "pos_sim_mean": 0.9007300138473511,
      "step": 2304
    },
    {
      "epoch": 0.461,
      "grad_norm": 2.2880189418792725,
      "learning_rate": 6.531799710626779e-05,
      "loss": 0.5116,
      "step": 2305
    },
    {
      "epoch": 0.461,
      "loss_align": 0.11341798305511475,
      "loss_contrastive": 0.07823324203491211,
      "loss_gen": 0.41904962062835693,
      "neg_sim_mean": 0.8648152351379395,
      "pos_sim_mean": 0.8865820169448853,
      "step": 2305
    },
    {
      "epoch": 0.4612,
      "grad_norm": 1.9528107643127441,
      "learning_rate": 6.528476524815528e-05,
      "loss": 0.5419,
      "step": 2306
    },
    {
      "epoch": 0.4612,
      "loss_align": 0.22412192821502686,
      "loss_contrastive": 0.0637812614440918,
      "loss_gen": 0.13326503336429596,
      "neg_sim_mean": 0.739659309387207,
      "pos_sim_mean": 0.7758780717849731,
      "step": 2306
    },
    {
      "epoch": 0.4614,
      "grad_norm": 1.4552149772644043,
      "learning_rate": 6.52515259404339e-05,
      "loss": 0.365,
      "step": 2307
    },
    {
      "epoch": 0.4614,
      "loss_align": 0.26390594244003296,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013270487077534199,
      "neg_sim_mean": 0.44086626172065735,
      "pos_sim_mean": 0.736094057559967,
      "step": 2307
    },
    {
      "epoch": 0.4616,
      "grad_norm": 0.8601897358894348,
      "learning_rate": 6.521827919930401e-05,
      "loss": 0.2772,
      "step": 2308
    },
    {
      "epoch": 0.4616,
      "loss_align": 0.06824111938476562,
      "loss_contrastive": 0.0,
      "loss_gen": 0.017747078090906143,
      "neg_sim_mean": 0.4527308940887451,
      "pos_sim_mean": 0.9317588806152344,
      "step": 2308
    },
    {
      "epoch": 0.4618,
      "grad_norm": 0.7133414149284363,
      "learning_rate": 6.518502504096971e-05,
      "loss": 0.086,
      "step": 2309
    },
    {
      "epoch": 0.4618,
      "loss_align": 0.16849184036254883,
      "loss_contrastive": 0.022507429122924805,
      "loss_gen": 0.3499961793422699,
      "neg_sim_mean": 0.7540155649185181,
      "pos_sim_mean": 0.8315081596374512,
      "step": 2309
    },
    {
      "epoch": 0.462,
      "grad_norm": 2.2747843265533447,
      "learning_rate": 6.515176348163871e-05,
      "loss": 0.5212,
      "step": 2310
    },
    {
      "epoch": 0.462,
      "loss_align": 0.45164287090301514,
      "loss_contrastive": 0.18767854571342468,
      "loss_gen": 0.26827722787857056,
      "neg_sim_mean": 0.636035680770874,
      "pos_sim_mean": 0.5483571290969849,
      "step": 2310
    },
    {
      "epoch": 0.4622,
      "grad_norm": 3.3391971588134766,
      "learning_rate": 6.511849453752223e-05,
      "loss": 0.7424,
      "step": 2311
    },
    {
      "epoch": 0.4622,
      "loss_align": 0.13231396675109863,
      "loss_contrastive": 0.0858079195022583,
      "loss_gen": 0.16140733659267426,
      "neg_sim_mean": 0.8534939289093018,
      "pos_sim_mean": 0.8676860332489014,
      "step": 2311
    },
    {
      "epoch": 0.4624,
      "grad_norm": 1.5898280143737793,
      "learning_rate": 6.508521822483518e-05,
      "loss": 0.304,
      "step": 2312
    },
    {
      "epoch": 0.4624,
      "loss_align": 0.19425451755523682,
      "loss_contrastive": 0.008315324783325195,
      "loss_gen": 0.36796730756759644,
      "neg_sim_mean": 0.7140607833862305,
      "pos_sim_mean": 0.8057454824447632,
      "step": 2312
    },
    {
      "epoch": 0.4626,
      "grad_norm": 2.2518463134765625,
      "learning_rate": 6.505193455979603e-05,
      "loss": 0.5632,
      "step": 2313
    },
    {
      "epoch": 0.4626,
      "loss_align": 0.16761112213134766,
      "loss_contrastive": 0.139748215675354,
      "loss_gen": 0.21957647800445557,
      "neg_sim_mean": 0.8721370697021484,
      "pos_sim_mean": 0.8323888778686523,
      "step": 2313
    },
    {
      "epoch": 0.4628,
      "grad_norm": 1.5640982389450073,
      "learning_rate": 6.50186435586268e-05,
      "loss": 0.404,
      "step": 2314
    },
    {
      "epoch": 0.4628,
      "loss_align": 0.13613605499267578,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008176885545253754,
      "neg_sim_mean": 0.3940000534057617,
      "pos_sim_mean": 0.8638639450073242,
      "step": 2314
    },
    {
      "epoch": 0.463,
      "grad_norm": 0.6077010631561279,
      "learning_rate": 6.49853452375532e-05,
      "loss": 0.1443,
      "step": 2315
    },
    {
      "epoch": 0.463,
      "loss_align": 0.22798216342926025,
      "loss_contrastive": 0.10587549209594727,
      "loss_gen": 0.31058406829833984,
      "neg_sim_mean": 0.7778933048248291,
      "pos_sim_mean": 0.7720178365707397,
      "step": 2315
    },
    {
      "epoch": 0.4632,
      "grad_norm": 1.9109503030776978,
      "learning_rate": 6.495203961280434e-05,
      "loss": 0.5513,
      "step": 2316
    },
    {
      "epoch": 0.4632,
      "loss_align": 0.1885143518447876,
      "loss_contrastive": 0.18176651000976562,
      "loss_gen": 0.13525129854679108,
      "neg_sim_mean": 0.8932521343231201,
      "pos_sim_mean": 0.8114856481552124,
      "step": 2316
    },
    {
      "epoch": 0.4634,
      "grad_norm": 1.5993915796279907,
      "learning_rate": 6.491872670061302e-05,
      "loss": 0.3456,
      "step": 2317
    },
    {
      "epoch": 0.4634,
      "loss_align": 0.2920701503753662,
      "loss_contrastive": 0.1218792200088501,
      "loss_gen": 0.22304970026016235,
      "neg_sim_mean": 0.729809045791626,
      "pos_sim_mean": 0.7079298496246338,
      "step": 2317
    },
    {
      "epoch": 0.4636,
      "grad_norm": 1.999691367149353,
      "learning_rate": 6.488540651721553e-05,
      "loss": 0.5297,
      "step": 2318
    },
    {
      "epoch": 0.4636,
      "loss_align": 0.11499500274658203,
      "loss_contrastive": 0.10300731658935547,
      "loss_gen": 0.2042105346918106,
      "neg_sim_mean": 0.8880122900009155,
      "pos_sim_mean": 0.885004997253418,
      "step": 2318
    },
    {
      "epoch": 0.4638,
      "grad_norm": 1.3541475534439087,
      "learning_rate": 6.485207907885175e-05,
      "loss": 0.3316,
      "step": 2319
    },
    {
      "epoch": 0.4638,
      "loss_align": 0.09056752920150757,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3138047754764557,
      "neg_sim_mean": 0.7985481023788452,
      "pos_sim_mean": 0.9094324707984924,
      "step": 2319
    },
    {
      "epoch": 0.464,
      "grad_norm": 1.8659923076629639,
      "learning_rate": 6.481874440176506e-05,
      "loss": 0.4044,
      "step": 2320
    },
    {
      "epoch": 0.464,
      "loss_align": 0.11155366897583008,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3853341341018677,
      "neg_sim_mean": 0.6786266565322876,
      "pos_sim_mean": 0.8884463310241699,
      "step": 2320
    },
    {
      "epoch": 0.4642,
      "grad_norm": 2.1732897758483887,
      "learning_rate": 6.478540250220234e-05,
      "loss": 0.4969,
      "step": 2321
    },
    {
      "epoch": 0.4642,
      "loss_align": 0.27880263328552246,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1465936154127121,
      "neg_sim_mean": 0.5977414846420288,
      "pos_sim_mean": 0.7211973667144775,
      "step": 2321
    },
    {
      "epoch": 0.4644,
      "grad_norm": 1.672307014465332,
      "learning_rate": 6.475205339641407e-05,
      "loss": 0.4254,
      "step": 2322
    },
    {
      "epoch": 0.4644,
      "loss_align": 0.22304505109786987,
      "loss_contrastive": 0.09298169612884521,
      "loss_gen": 0.2218811810016632,
      "neg_sim_mean": 0.7699366211891174,
      "pos_sim_mean": 0.7769549489021301,
      "step": 2322
    },
    {
      "epoch": 0.4646,
      "grad_norm": 1.871933937072754,
      "learning_rate": 6.471869710065418e-05,
      "loss": 0.4561,
      "step": 2323
    },
    {
      "epoch": 0.4646,
      "loss_align": 0.2799278497695923,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3391662538051605,
      "neg_sim_mean": 0.5402880907058716,
      "pos_sim_mean": 0.7200721502304077,
      "step": 2323
    },
    {
      "epoch": 0.4648,
      "grad_norm": 2.4035768508911133,
      "learning_rate": 6.468533363118015e-05,
      "loss": 0.6191,
      "step": 2324
    },
    {
      "epoch": 0.4648,
      "loss_align": 0.13868820667266846,
      "loss_contrastive": 0.07769453525543213,
      "loss_gen": 0.18617461621761322,
      "neg_sim_mean": 0.8390063047409058,
      "pos_sim_mean": 0.8613117933273315,
      "step": 2324
    },
    {
      "epoch": 0.465,
      "grad_norm": 1.5245448350906372,
      "learning_rate": 6.465196300425287e-05,
      "loss": 0.3342,
      "step": 2325
    },
    {
      "epoch": 0.465,
      "loss_align": 0.08051681518554688,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15625718235969543,
      "neg_sim_mean": 0.7099982500076294,
      "pos_sim_mean": 0.9194831848144531,
      "step": 2325
    },
    {
      "epoch": 0.4652,
      "grad_norm": 1.199634075164795,
      "learning_rate": 6.461858523613684e-05,
      "loss": 0.2368,
      "step": 2326
    },
    {
      "epoch": 0.4652,
      "loss_align": 0.291090190410614,
      "loss_contrastive": 0.2592635750770569,
      "loss_gen": 0.17457464337348938,
      "neg_sim_mean": 0.868173360824585,
      "pos_sim_mean": 0.708909809589386,
      "step": 2326
    },
    {
      "epoch": 0.4654,
      "grad_norm": 3.593463659286499,
      "learning_rate": 6.458520034309994e-05,
      "loss": 0.4968,
      "step": 2327
    },
    {
      "epoch": 0.4654,
      "loss_align": 0.10810482501983643,
      "loss_contrastive": 0.04706460237503052,
      "loss_gen": 0.2887856662273407,
      "neg_sim_mean": 0.8389597535133362,
      "pos_sim_mean": 0.8918951749801636,
      "step": 2327
    },
    {
      "epoch": 0.4656,
      "grad_norm": 2.5231056213378906,
      "learning_rate": 6.455180834141359e-05,
      "loss": 0.4025,
      "step": 2328
    },
    {
      "epoch": 0.4656,
      "loss_align": 0.0862734317779541,
      "loss_contrastive": 0.05041229724884033,
      "loss_gen": 0.6214362978935242,
      "neg_sim_mean": 0.8641388416290283,
      "pos_sim_mean": 0.9137265682220459,
      "step": 2328
    },
    {
      "epoch": 0.4658,
      "grad_norm": 2.876417398452759,
      "learning_rate": 6.451840924735263e-05,
      "loss": 0.7138,
      "step": 2329
    },
    {
      "epoch": 0.4658,
      "loss_align": 0.21365761756896973,
      "loss_contrastive": 0.010442256927490234,
      "loss_gen": 0.38541287183761597,
      "neg_sim_mean": 0.6967846155166626,
      "pos_sim_mean": 0.7863423824310303,
      "step": 2329
    },
    {
      "epoch": 0.466,
      "grad_norm": 2.491999864578247,
      "learning_rate": 6.448500307719538e-05,
      "loss": 0.6003,
      "step": 2330
    },
    {
      "epoch": 0.466,
      "loss_align": 0.07478857040405273,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24549981951713562,
      "neg_sim_mean": 0.8242542743682861,
      "pos_sim_mean": 0.9252114295959473,
      "step": 2330
    },
    {
      "epoch": 0.4662,
      "grad_norm": 1.728295087814331,
      "learning_rate": 6.445158984722358e-05,
      "loss": 0.3203,
      "step": 2331
    },
    {
      "epoch": 0.4662,
      "loss_align": 0.19545143842697144,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24600426852703094,
      "neg_sim_mean": 0.7009051442146301,
      "pos_sim_mean": 0.8045485615730286,
      "step": 2331
    },
    {
      "epoch": 0.4664,
      "grad_norm": 1.9680100679397583,
      "learning_rate": 6.441816957372246e-05,
      "loss": 0.4415,
      "step": 2332
    },
    {
      "epoch": 0.4664,
      "loss_align": 0.0821220874786377,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20217657089233398,
      "neg_sim_mean": 0.568456768989563,
      "pos_sim_mean": 0.9178779125213623,
      "step": 2332
    },
    {
      "epoch": 0.4666,
      "grad_norm": 1.5864074230194092,
      "learning_rate": 6.438474227298064e-05,
      "loss": 0.2843,
      "step": 2333
    },
    {
      "epoch": 0.4666,
      "loss_align": 0.2652028799057007,
      "loss_contrastive": 0.10233032703399658,
      "loss_gen": 0.4129754602909088,
      "neg_sim_mean": 0.737127423286438,
      "pos_sim_mean": 0.7347971200942993,
      "step": 2333
    },
    {
      "epoch": 0.4668,
      "grad_norm": 2.4486351013183594,
      "learning_rate": 6.435130796129018e-05,
      "loss": 0.6905,
      "step": 2334
    },
    {
      "epoch": 0.4668,
      "loss_align": 0.22096127271652222,
      "loss_contrastive": 0.10160428285598755,
      "loss_gen": 0.3621286451816559,
      "neg_sim_mean": 0.7806429862976074,
      "pos_sim_mean": 0.7790387272834778,
      "step": 2334
    },
    {
      "epoch": 0.467,
      "grad_norm": 2.146153450012207,
      "learning_rate": 6.431786665494657e-05,
      "loss": 0.5953,
      "step": 2335
    },
    {
      "epoch": 0.467,
      "loss_align": 0.018312156200408936,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09223566949367523,
      "neg_sim_mean": 0.8087579011917114,
      "pos_sim_mean": 0.9816878437995911,
      "step": 2335
    },
    {
      "epoch": 0.4672,
      "grad_norm": 0.9000042676925659,
      "learning_rate": 6.428441837024868e-05,
      "loss": 0.1105,
      "step": 2336
    },
    {
      "epoch": 0.4672,
      "loss_align": 0.16498780250549316,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006656304467469454,
      "neg_sim_mean": 0.5173325538635254,
      "pos_sim_mean": 0.8350121974945068,
      "step": 2336
    },
    {
      "epoch": 0.4674,
      "grad_norm": 0.8560518622398376,
      "learning_rate": 6.42509631234988e-05,
      "loss": 0.1716,
      "step": 2337
    },
    {
      "epoch": 0.4674,
      "loss_align": 0.09723222255706787,
      "loss_contrastive": 0.04715728759765625,
      "loss_gen": 0.521767795085907,
      "neg_sim_mean": 0.8499250411987305,
      "pos_sim_mean": 0.9027677774429321,
      "step": 2337
    },
    {
      "epoch": 0.4676,
      "grad_norm": 2.396294593811035,
      "learning_rate": 6.421750093100264e-05,
      "loss": 0.6247,
      "step": 2338
    },
    {
      "epoch": 0.4676,
      "loss_align": 0.3924804925918579,
      "loss_contrastive": 0.12931573390960693,
      "loss_gen": 0.39431121945381165,
      "neg_sim_mean": 0.6368352174758911,
      "pos_sim_mean": 0.6075195074081421,
      "step": 2338
    },
    {
      "epoch": 0.4678,
      "grad_norm": 2.839203357696533,
      "learning_rate": 6.418403180906922e-05,
      "loss": 0.8023,
      "step": 2339
    },
    {
      "epoch": 0.4678,
      "loss_align": 0.13666057586669922,
      "loss_contrastive": 0.009107232093811035,
      "loss_gen": 0.12713439762592316,
      "neg_sim_mean": 0.7724466323852539,
      "pos_sim_mean": 0.8633394241333008,
      "step": 2339
    },
    {
      "epoch": 0.468,
      "grad_norm": 1.4184645414352417,
      "learning_rate": 6.415055577401102e-05,
      "loss": 0.2649,
      "step": 2340
    },
    {
      "epoch": 0.468,
      "loss_align": 0.2166258692741394,
      "loss_contrastive": 0.08311229944229126,
      "loss_gen": 0.2677643597126007,
      "neg_sim_mean": 0.766486406326294,
      "pos_sim_mean": 0.7833741307258606,
      "step": 2340
    },
    {
      "epoch": 0.4682,
      "grad_norm": 2.2356672286987305,
      "learning_rate": 6.411707284214384e-05,
      "loss": 0.4944,
      "step": 2341
    },
    {
      "epoch": 0.4682,
      "loss_align": 0.08021879196166992,
      "loss_contrastive": 0.04020041227340698,
      "loss_gen": 0.31863558292388916,
      "neg_sim_mean": 0.8599815964698792,
      "pos_sim_mean": 0.9197812080383301,
      "step": 2341
    },
    {
      "epoch": 0.4684,
      "grad_norm": 2.041862726211548,
      "learning_rate": 6.408358302978683e-05,
      "loss": 0.4037,
      "step": 2342
    },
    {
      "epoch": 0.4684,
      "loss_align": 0.12757349014282227,
      "loss_contrastive": 0.03269553184509277,
      "loss_gen": 0.11510208994150162,
      "neg_sim_mean": 0.8051220178604126,
      "pos_sim_mean": 0.8724265098571777,
      "step": 2342
    },
    {
      "epoch": 0.4686,
      "grad_norm": 1.3061941862106323,
      "learning_rate": 6.405008635326257e-05,
      "loss": 0.2466,
      "step": 2343
    },
    {
      "epoch": 0.4686,
      "loss_align": 0.2071324586868286,
      "loss_contrastive": 0.07528805732727051,
      "loss_gen": 0.3660692572593689,
      "neg_sim_mean": 0.768155574798584,
      "pos_sim_mean": 0.7928675413131714,
      "step": 2343
    },
    {
      "epoch": 0.4688,
      "grad_norm": 2.7577567100524902,
      "learning_rate": 6.401658282889688e-05,
      "loss": 0.5822,
      "step": 2344
    },
    {
      "epoch": 0.4688,
      "loss_align": 0.13901615142822266,
      "loss_contrastive": 0.14540815353393555,
      "loss_gen": 0.19546544551849365,
      "neg_sim_mean": 0.906391978263855,
      "pos_sim_mean": 0.8609838485717773,
      "step": 2344
    },
    {
      "epoch": 0.469,
      "grad_norm": 1.4503607749938965,
      "learning_rate": 6.3983072473019e-05,
      "loss": 0.3519,
      "step": 2345
    },
    {
      "epoch": 0.469,
      "loss_align": 0.14986228942871094,
      "loss_contrastive": 0.004733085632324219,
      "loss_gen": 0.16808722913265228,
      "neg_sim_mean": 0.7548707723617554,
      "pos_sim_mean": 0.8501377105712891,
      "step": 2345
    },
    {
      "epoch": 0.4692,
      "grad_norm": 1.1957608461380005,
      "learning_rate": 6.394955530196147e-05,
      "loss": 0.3185,
      "step": 2346
    },
    {
      "epoch": 0.4692,
      "loss_align": 0.14029335975646973,
      "loss_contrastive": 0.007009267807006836,
      "loss_gen": 0.17079126834869385,
      "neg_sim_mean": 0.7667158842086792,
      "pos_sim_mean": 0.8597066402435303,
      "step": 2346
    },
    {
      "epoch": 0.4694,
      "grad_norm": 1.661421298980713,
      "learning_rate": 6.391603133206015e-05,
      "loss": 0.3119,
      "step": 2347
    },
    {
      "epoch": 0.4694,
      "loss_align": 0.21696341037750244,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3232189416885376,
      "neg_sim_mean": 0.6567867398262024,
      "pos_sim_mean": 0.7830365896224976,
      "step": 2347
    },
    {
      "epoch": 0.4696,
      "grad_norm": 1.8256702423095703,
      "learning_rate": 6.388250057965421e-05,
      "loss": 0.5402,
      "step": 2348
    },
    {
      "epoch": 0.4696,
      "loss_align": 0.09102606773376465,
      "loss_contrastive": 0.0,
      "loss_gen": 0.253428190946579,
      "neg_sim_mean": 0.7820791006088257,
      "pos_sim_mean": 0.9089739322662354,
      "step": 2348
    },
    {
      "epoch": 0.4698,
      "grad_norm": 1.8196055889129639,
      "learning_rate": 6.384896306108612e-05,
      "loss": 0.3445,
      "step": 2349
    },
    {
      "epoch": 0.4698,
      "loss_align": 0.1335393190383911,
      "loss_contrastive": 0.10110199451446533,
      "loss_gen": 0.19440534710884094,
      "neg_sim_mean": 0.8675626516342163,
      "pos_sim_mean": 0.8664606809616089,
      "step": 2349
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.6749160289764404,
      "learning_rate": 6.381541879270169e-05,
      "loss": 0.3401,
      "step": 2350
    },
    {
      "epoch": 0.47,
      "loss_align": 0.132132887840271,
      "loss_contrastive": 0.060011088848114014,
      "loss_gen": 0.23372627794742584,
      "neg_sim_mean": 0.8278781771659851,
      "pos_sim_mean": 0.867867112159729,
      "step": 2350
    },
    {
      "epoch": 0.4702,
      "grad_norm": 1.564494252204895,
      "learning_rate": 6.378186779084995e-05,
      "loss": 0.3731,
      "step": 2351
    },
    {
      "epoch": 0.4702,
      "loss_align": 0.2873269319534302,
      "loss_contrastive": 0.11551141738891602,
      "loss_gen": 0.3050936758518219,
      "neg_sim_mean": 0.7281844615936279,
      "pos_sim_mean": 0.7126730680465698,
      "step": 2351
    },
    {
      "epoch": 0.4704,
      "grad_norm": 2.2825045585632324,
      "learning_rate": 6.374831007188332e-05,
      "loss": 0.6063,
      "step": 2352
    },
    {
      "epoch": 0.4704,
      "loss_align": 0.13800597190856934,
      "loss_contrastive": 0.06534647941589355,
      "loss_gen": 0.4963991641998291,
      "neg_sim_mean": 0.8273404836654663,
      "pos_sim_mean": 0.8619940280914307,
      "step": 2352
    },
    {
      "epoch": 0.4706,
      "grad_norm": 2.5759265422821045,
      "learning_rate": 6.371474565215734e-05,
      "loss": 0.6422,
      "step": 2353
    },
    {
      "epoch": 0.4706,
      "loss_align": 0.09208989143371582,
      "loss_contrastive": 0.045285820960998535,
      "loss_gen": 0.2534239888191223,
      "neg_sim_mean": 0.8531959056854248,
      "pos_sim_mean": 0.9079101085662842,
      "step": 2353
    },
    {
      "epoch": 0.4708,
      "grad_norm": 1.6646076440811157,
      "learning_rate": 6.368117454803093e-05,
      "loss": 0.3509,
      "step": 2354
    },
    {
      "epoch": 0.4708,
      "loss_align": 0.037186264991760254,
      "loss_contrastive": 0.08259493112564087,
      "loss_gen": 0.15838736295700073,
      "neg_sim_mean": 0.9454086422920227,
      "pos_sim_mean": 0.9628137350082397,
      "step": 2354
    },
    {
      "epoch": 0.471,
      "grad_norm": 1.2155897617340088,
      "learning_rate": 6.364759677586627e-05,
      "loss": 0.2055,
      "step": 2355
    },
    {
      "epoch": 0.471,
      "loss_align": 0.3172827959060669,
      "loss_contrastive": 0.12601667642593384,
      "loss_gen": 0.3580584228038788,
      "neg_sim_mean": 0.708733856678009,
      "pos_sim_mean": 0.6827172040939331,
      "step": 2355
    },
    {
      "epoch": 0.4712,
      "grad_norm": 2.3233768939971924,
      "learning_rate": 6.361401235202872e-05,
      "loss": 0.6905,
      "step": 2356
    },
    {
      "epoch": 0.4712,
      "loss_align": 0.29970669746398926,
      "loss_contrastive": 0.06734544038772583,
      "loss_gen": 0.21912173926830292,
      "neg_sim_mean": 0.6676387190818787,
      "pos_sim_mean": 0.7002933025360107,
      "step": 2356
    },
    {
      "epoch": 0.4714,
      "grad_norm": 2.295267343521118,
      "learning_rate": 6.358042129288694e-05,
      "loss": 0.5269,
      "step": 2357
    },
    {
      "epoch": 0.4714,
      "loss_align": 0.19143670797348022,
      "loss_contrastive": 0.10282927751541138,
      "loss_gen": 0.38024765253067017,
      "neg_sim_mean": 0.8113925457000732,
      "pos_sim_mean": 0.8085632920265198,
      "step": 2357
    },
    {
      "epoch": 0.4716,
      "grad_norm": 2.1619434356689453,
      "learning_rate": 6.35468236148128e-05,
      "loss": 0.584,
      "step": 2358
    },
    {
      "epoch": 0.4716,
      "loss_align": 0.0638226866722107,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23786833882331848,
      "neg_sim_mean": 0.8034989833831787,
      "pos_sim_mean": 0.9361773133277893,
      "step": 2358
    },
    {
      "epoch": 0.4718,
      "grad_norm": 1.7424460649490356,
      "learning_rate": 6.351321933418139e-05,
      "loss": 0.3017,
      "step": 2359
    },
    {
      "epoch": 0.4718,
      "loss_align": 0.21183645725250244,
      "loss_contrastive": 0.13896870613098145,
      "loss_gen": 0.21729010343551636,
      "neg_sim_mean": 0.8271322250366211,
      "pos_sim_mean": 0.7881635427474976,
      "step": 2359
    },
    {
      "epoch": 0.472,
      "grad_norm": 1.4682468175888062,
      "learning_rate": 6.347960846737106e-05,
      "loss": 0.4458,
      "step": 2360
    },
    {
      "epoch": 0.472,
      "loss_align": 0.15507495403289795,
      "loss_contrastive": 0.064475417137146,
      "loss_gen": 0.23968176543712616,
      "neg_sim_mean": 0.8094004392623901,
      "pos_sim_mean": 0.844925045967102,
      "step": 2360
    },
    {
      "epoch": 0.4722,
      "grad_norm": 2.024820327758789,
      "learning_rate": 6.344599103076329e-05,
      "loss": 0.4025,
      "step": 2361
    },
    {
      "epoch": 0.4722,
      "loss_align": 0.10709893703460693,
      "loss_contrastive": 0.09248948097229004,
      "loss_gen": 0.46519482135772705,
      "neg_sim_mean": 0.8853905200958252,
      "pos_sim_mean": 0.8929010629653931,
      "step": 2361
    },
    {
      "epoch": 0.4724,
      "grad_norm": 2.036245346069336,
      "learning_rate": 6.341236704074285e-05,
      "loss": 0.5834,
      "step": 2362
    },
    {
      "epoch": 0.4724,
      "loss_align": 0.1476823091506958,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04207290709018707,
      "neg_sim_mean": 0.4562041759490967,
      "pos_sim_mean": 0.8523176908493042,
      "step": 2362
    },
    {
      "epoch": 0.4726,
      "grad_norm": 1.4990911483764648,
      "learning_rate": 6.337873651369763e-05,
      "loss": 0.1898,
      "step": 2363
    },
    {
      "epoch": 0.4726,
      "loss_align": 0.11035174131393433,
      "loss_contrastive": 0.040060341358184814,
      "loss_gen": 0.21365651488304138,
      "neg_sim_mean": 0.8297085762023926,
      "pos_sim_mean": 0.8896482586860657,
      "step": 2363
    },
    {
      "epoch": 0.4728,
      "grad_norm": 1.7019065618515015,
      "learning_rate": 6.334509946601879e-05,
      "loss": 0.3288,
      "step": 2364
    },
    {
      "epoch": 0.4728,
      "loss_align": 0.31986093521118164,
      "loss_contrastive": 0.2379772663116455,
      "loss_gen": 0.1714615523815155,
      "neg_sim_mean": 0.818116307258606,
      "pos_sim_mean": 0.6801390647888184,
      "step": 2364
    },
    {
      "epoch": 0.473,
      "grad_norm": 1.8073550462722778,
      "learning_rate": 6.331145591410057e-05,
      "loss": 0.5199,
      "step": 2365
    },
    {
      "epoch": 0.473,
      "loss_align": 0.10771560668945312,
      "loss_contrastive": 0.0,
      "loss_gen": 0.273588091135025,
      "neg_sim_mean": 0.7916051745414734,
      "pos_sim_mean": 0.8922843933105469,
      "step": 2365
    },
    {
      "epoch": 0.4732,
      "grad_norm": 1.9492825269699097,
      "learning_rate": 6.327780587434044e-05,
      "loss": 0.3813,
      "step": 2366
    },
    {
      "epoch": 0.4732,
      "loss_align": 0.12334930896759033,
      "loss_contrastive": 0.0,
      "loss_gen": 0.021220693364739418,
      "neg_sim_mean": 0.33526918292045593,
      "pos_sim_mean": 0.8766506910324097,
      "step": 2366
    },
    {
      "epoch": 0.4734,
      "grad_norm": 0.9094032049179077,
      "learning_rate": 6.324414936313904e-05,
      "loss": 0.1446,
      "step": 2367
    },
    {
      "epoch": 0.4734,
      "loss_align": 0.12495380640029907,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03400842845439911,
      "neg_sim_mean": 0.46628373861312866,
      "pos_sim_mean": 0.8750461935997009,
      "step": 2367
    },
    {
      "epoch": 0.4736,
      "grad_norm": 1.0911766290664673,
      "learning_rate": 6.321048639690013e-05,
      "loss": 0.159,
      "step": 2368
    },
    {
      "epoch": 0.4736,
      "loss_align": 0.12662363052368164,
      "loss_contrastive": 0.0,
      "loss_gen": 0.040703367441892624,
      "neg_sim_mean": 0.28022271394729614,
      "pos_sim_mean": 0.8733763694763184,
      "step": 2368
    },
    {
      "epoch": 0.4738,
      "grad_norm": 1.3032602071762085,
      "learning_rate": 6.317681699203063e-05,
      "loss": 0.1673,
      "step": 2369
    },
    {
      "epoch": 0.4738,
      "loss_align": 0.13256335258483887,
      "loss_contrastive": 0.06175804138183594,
      "loss_gen": 0.15708743035793304,
      "neg_sim_mean": 0.8291946649551392,
      "pos_sim_mean": 0.8674366474151611,
      "step": 2369
    },
    {
      "epoch": 0.474,
      "grad_norm": 1.481279730796814,
      "learning_rate": 6.31431411649406e-05,
      "loss": 0.2971,
      "step": 2370
    },
    {
      "epoch": 0.474,
      "loss_align": 0.12317240238189697,
      "loss_contrastive": 0.0,
      "loss_gen": 0.30912938714027405,
      "neg_sim_mean": 0.7214937806129456,
      "pos_sim_mean": 0.876827597618103,
      "step": 2370
    },
    {
      "epoch": 0.4742,
      "grad_norm": 2.0501022338867188,
      "learning_rate": 6.310945893204324e-05,
      "loss": 0.4323,
      "step": 2371
    },
    {
      "epoch": 0.4742,
      "loss_align": 0.22655558586120605,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009984628297388554,
      "neg_sim_mean": 0.3704131841659546,
      "pos_sim_mean": 0.773444414138794,
      "step": 2371
    },
    {
      "epoch": 0.4744,
      "grad_norm": 0.8135990500450134,
      "learning_rate": 6.307577030975484e-05,
      "loss": 0.2365,
      "step": 2372
    },
    {
      "epoch": 0.4744,
      "loss_align": 0.09606355428695679,
      "loss_contrastive": 0.09887689352035522,
      "loss_gen": 0.10395557433366776,
      "neg_sim_mean": 0.9028133153915405,
      "pos_sim_mean": 0.9039364457130432,
      "step": 2372
    },
    {
      "epoch": 0.4746,
      "grad_norm": 1.1216830015182495,
      "learning_rate": 6.304207531449486e-05,
      "loss": 0.2119,
      "step": 2373
    },
    {
      "epoch": 0.4746,
      "loss_align": 0.1421893835067749,
      "loss_contrastive": 0.08971649408340454,
      "loss_gen": 0.15795738995075226,
      "neg_sim_mean": 0.8475270867347717,
      "pos_sim_mean": 0.8578106164932251,
      "step": 2373
    },
    {
      "epoch": 0.4748,
      "grad_norm": 1.3644349575042725,
      "learning_rate": 6.300837396268579e-05,
      "loss": 0.3109,
      "step": 2374
    },
    {
      "epoch": 0.4748,
      "loss_align": 0.0608675479888916,
      "loss_contrastive": 0.057125210762023926,
      "loss_gen": 0.2437417060136795,
      "neg_sim_mean": 0.8962576389312744,
      "pos_sim_mean": 0.9391324520111084,
      "step": 2374
    },
    {
      "epoch": 0.475,
      "grad_norm": 1.5784125328063965,
      "learning_rate": 6.297466627075327e-05,
      "loss": 0.3115,
      "step": 2375
    },
    {
      "epoch": 0.475,
      "loss_align": 0.09952032566070557,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010134653188288212,
      "neg_sim_mean": 0.40234270691871643,
      "pos_sim_mean": 0.9004796743392944,
      "step": 2375
    },
    {
      "epoch": 0.4752,
      "grad_norm": 0.6518652439117432,
      "learning_rate": 6.294095225512603e-05,
      "loss": 0.1097,
      "step": 2376
    },
    {
      "epoch": 0.4752,
      "loss_align": 0.15797603130340576,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005412626080214977,
      "neg_sim_mean": 0.34385570883750916,
      "pos_sim_mean": 0.8420239686965942,
      "step": 2376
    },
    {
      "epoch": 0.4754,
      "grad_norm": 0.5483437776565552,
      "learning_rate": 6.29072319322359e-05,
      "loss": 0.1634,
      "step": 2377
    },
    {
      "epoch": 0.4754,
      "loss_align": 0.12877333164215088,
      "loss_contrastive": 0.101898193359375,
      "loss_gen": 0.1426021307706833,
      "neg_sim_mean": 0.8731248378753662,
      "pos_sim_mean": 0.8712266683578491,
      "step": 2377
    },
    {
      "epoch": 0.4756,
      "grad_norm": 1.4216089248657227,
      "learning_rate": 6.28735053185177e-05,
      "loss": 0.2836,
      "step": 2378
    },
    {
      "epoch": 0.4756,
      "loss_align": 0.20358562469482422,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18768373131752014,
      "neg_sim_mean": 0.5897002816200256,
      "pos_sim_mean": 0.7964143753051758,
      "step": 2378
    },
    {
      "epoch": 0.4758,
      "grad_norm": 1.3922169208526611,
      "learning_rate": 6.283977243040939e-05,
      "loss": 0.3913,
      "step": 2379
    },
    {
      "epoch": 0.4758,
      "loss_align": 0.1621183156967163,
      "loss_contrastive": 0.08287882804870605,
      "loss_gen": 0.10861716419458389,
      "neg_sim_mean": 0.8207604885101318,
      "pos_sim_mean": 0.8378816843032837,
      "step": 2379
    },
    {
      "epoch": 0.476,
      "grad_norm": 1.3728240728378296,
      "learning_rate": 6.280603328435198e-05,
      "loss": 0.2807,
      "step": 2380
    },
    {
      "epoch": 0.476,
      "loss_align": 0.20830631256103516,
      "loss_contrastive": 0.055755674839019775,
      "loss_gen": 0.10892197489738464,
      "neg_sim_mean": 0.7474493384361267,
      "pos_sim_mean": 0.7916936874389648,
      "step": 2380
    },
    {
      "epoch": 0.4762,
      "grad_norm": 1.5188755989074707,
      "learning_rate": 6.277228789678953e-05,
      "loss": 0.3239,
      "step": 2381
    },
    {
      "epoch": 0.4762,
      "loss_align": 0.07861858606338501,
      "loss_contrastive": 0.06258159875869751,
      "loss_gen": 0.22213807702064514,
      "neg_sim_mean": 0.8839629888534546,
      "pos_sim_mean": 0.921381413936615,
      "step": 2381
    },
    {
      "epoch": 0.4764,
      "grad_norm": 2.0142056941986084,
      "learning_rate": 6.273853628416911e-05,
      "loss": 0.3083,
      "step": 2382
    },
    {
      "epoch": 0.4764,
      "loss_align": 0.17548781633377075,
      "loss_contrastive": 0.13002341985702515,
      "loss_gen": 0.23008650541305542,
      "neg_sim_mean": 0.8545355796813965,
      "pos_sim_mean": 0.8245121836662292,
      "step": 2382
    },
    {
      "epoch": 0.4766,
      "grad_norm": 2.363600492477417,
      "learning_rate": 6.270477846294087e-05,
      "loss": 0.4212,
      "step": 2383
    },
    {
      "epoch": 0.4766,
      "loss_align": 0.13919532299041748,
      "loss_contrastive": 0.08207237720489502,
      "loss_gen": 0.24721460044384003,
      "neg_sim_mean": 0.8428770303726196,
      "pos_sim_mean": 0.8608046770095825,
      "step": 2383
    },
    {
      "epoch": 0.4768,
      "grad_norm": 1.8422706127166748,
      "learning_rate": 6.267101444955792e-05,
      "loss": 0.3963,
      "step": 2384
    },
    {
      "epoch": 0.4768,
      "loss_align": 0.15381860733032227,
      "loss_contrastive": 0.016640543937683105,
      "loss_gen": 0.5136318206787109,
      "neg_sim_mean": 0.7628219127655029,
      "pos_sim_mean": 0.8461813926696777,
      "step": 2384
    },
    {
      "epoch": 0.477,
      "grad_norm": 2.5672011375427246,
      "learning_rate": 6.263724426047647e-05,
      "loss": 0.6694,
      "step": 2385
    },
    {
      "epoch": 0.477,
      "loss_align": 0.2288520336151123,
      "loss_contrastive": 0.14358556270599365,
      "loss_gen": 0.26603055000305176,
      "neg_sim_mean": 0.8147335052490234,
      "pos_sim_mean": 0.7711479663848877,
      "step": 2385
    },
    {
      "epoch": 0.4772,
      "grad_norm": 1.6791019439697266,
      "learning_rate": 6.26034679121557e-05,
      "loss": 0.5121,
      "step": 2386
    },
    {
      "epoch": 0.4772,
      "loss_align": 0.09591597318649292,
      "loss_contrastive": 0.057784855365753174,
      "loss_gen": 0.41938844323158264,
      "neg_sim_mean": 0.8618688583374023,
      "pos_sim_mean": 0.9040840268135071,
      "step": 2386
    },
    {
      "epoch": 0.4774,
      "grad_norm": 2.0246262550354004,
      "learning_rate": 6.256968542105775e-05,
      "loss": 0.5222,
      "step": 2387
    },
    {
      "epoch": 0.4774,
      "loss_align": 0.13527226448059082,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15767699480056763,
      "neg_sim_mean": 0.7279738783836365,
      "pos_sim_mean": 0.8647277355194092,
      "step": 2387
    },
    {
      "epoch": 0.4776,
      "grad_norm": 1.6194554567337036,
      "learning_rate": 6.253589680364785e-05,
      "loss": 0.2929,
      "step": 2388
    },
    {
      "epoch": 0.4776,
      "loss_align": 0.11998504400253296,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005607863422483206,
      "neg_sim_mean": 0.3746967911720276,
      "pos_sim_mean": 0.880014955997467,
      "step": 2388
    },
    {
      "epoch": 0.4778,
      "grad_norm": 0.49439406394958496,
      "learning_rate": 6.250210207639411e-05,
      "loss": 0.1256,
      "step": 2389
    },
    {
      "epoch": 0.4778,
      "loss_align": 0.2545599937438965,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006771197076886892,
      "neg_sim_mean": 0.4313652217388153,
      "pos_sim_mean": 0.7454400062561035,
      "step": 2389
    },
    {
      "epoch": 0.478,
      "grad_norm": 0.7447043657302856,
      "learning_rate": 6.24683012557677e-05,
      "loss": 0.2613,
      "step": 2390
    },
    {
      "epoch": 0.478,
      "loss_align": 0.1035999059677124,
      "loss_contrastive": 0.06651538610458374,
      "loss_gen": 0.14991848170757294,
      "neg_sim_mean": 0.8629154562950134,
      "pos_sim_mean": 0.8964000940322876,
      "step": 2390
    },
    {
      "epoch": 0.4782,
      "grad_norm": 1.6617175340652466,
      "learning_rate": 6.243449435824276e-05,
      "loss": 0.2615,
      "step": 2391
    },
    {
      "epoch": 0.4782,
      "loss_align": 0.2525235414505005,
      "loss_contrastive": 0.06610101461410522,
      "loss_gen": 0.2812207043170929,
      "neg_sim_mean": 0.7135774493217468,
      "pos_sim_mean": 0.7474764585494995,
      "step": 2391
    },
    {
      "epoch": 0.4784,
      "grad_norm": 1.6357816457748413,
      "learning_rate": 6.240068140029628e-05,
      "loss": 0.5417,
      "step": 2392
    },
    {
      "epoch": 0.4784,
      "loss_align": 0.17129725217819214,
      "loss_contrastive": 0.0329357385635376,
      "loss_gen": 0.20068854093551636,
      "neg_sim_mean": 0.7616384625434875,
      "pos_sim_mean": 0.8287027478218079,
      "step": 2392
    },
    {
      "epoch": 0.4786,
      "grad_norm": 2.078484296798706,
      "learning_rate": 6.236686239840836e-05,
      "loss": 0.3759,
      "step": 2393
    },
    {
      "epoch": 0.4786,
      "loss_align": 0.1885596513748169,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19440610706806183,
      "neg_sim_mean": 0.7079446315765381,
      "pos_sim_mean": 0.8114403486251831,
      "step": 2393
    },
    {
      "epoch": 0.4788,
      "grad_norm": 1.7937631607055664,
      "learning_rate": 6.233303736906193e-05,
      "loss": 0.383,
      "step": 2394
    },
    {
      "epoch": 0.4788,
      "loss_align": 0.07515740394592285,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0023901353124529123,
      "neg_sim_mean": 0.324931263923645,
      "pos_sim_mean": 0.9248425960540771,
      "step": 2394
    },
    {
      "epoch": 0.479,
      "grad_norm": 0.2949922978878021,
      "learning_rate": 6.22992063287429e-05,
      "loss": 0.0775,
      "step": 2395
    },
    {
      "epoch": 0.479,
      "loss_align": 0.11315637826919556,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2567247450351715,
      "neg_sim_mean": 0.7729756236076355,
      "pos_sim_mean": 0.8868436217308044,
      "step": 2395
    },
    {
      "epoch": 0.4792,
      "grad_norm": 1.550940990447998,
      "learning_rate": 6.226536929394013e-05,
      "loss": 0.3699,
      "step": 2396
    },
    {
      "epoch": 0.4792,
      "loss_align": 0.126501202583313,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23460914194583893,
      "neg_sim_mean": 0.5280914902687073,
      "pos_sim_mean": 0.873498797416687,
      "step": 2396
    },
    {
      "epoch": 0.4794,
      "grad_norm": 1.4500032663345337,
      "learning_rate": 6.223152628114537e-05,
      "loss": 0.3611,
      "step": 2397
    },
    {
      "epoch": 0.4794,
      "loss_align": 0.10650861263275146,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16395559906959534,
      "neg_sim_mean": 0.7754433155059814,
      "pos_sim_mean": 0.8934913873672485,
      "step": 2397
    },
    {
      "epoch": 0.4796,
      "grad_norm": 1.4914159774780273,
      "learning_rate": 6.219767730685329e-05,
      "loss": 0.2705,
      "step": 2398
    },
    {
      "epoch": 0.4796,
      "loss_align": 0.09479361772537231,
      "loss_contrastive": 0.0,
      "loss_gen": 0.019878452643752098,
      "neg_sim_mean": 0.23268882930278778,
      "pos_sim_mean": 0.9052063822746277,
      "step": 2398
    },
    {
      "epoch": 0.4798,
      "grad_norm": 2.440772771835327,
      "learning_rate": 6.216382238756146e-05,
      "loss": 0.1147,
      "step": 2399
    },
    {
      "epoch": 0.4798,
      "loss_align": 0.1928694248199463,
      "loss_contrastive": 0.02435898780822754,
      "loss_gen": 0.1904105544090271,
      "neg_sim_mean": 0.7314895391464233,
      "pos_sim_mean": 0.8071305751800537,
      "step": 2399
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.7153244018554688,
      "learning_rate": 6.212996153977037e-05,
      "loss": 0.3862,
      "step": 2400
    },
    {
      "epoch": 0.48,
      "loss_align": 0.17370522022247314,
      "loss_contrastive": 0.06993973255157471,
      "loss_gen": 0.12213020771741867,
      "neg_sim_mean": 0.7962344884872437,
      "pos_sim_mean": 0.8262947797775269,
      "step": 2400
    },
    {
      "epoch": 0.4802,
      "grad_norm": 1.3116973638534546,
      "learning_rate": 6.209609477998338e-05,
      "loss": 0.3042,
      "step": 2401
    },
    {
      "epoch": 0.4802,
      "loss_align": 0.45992910861968994,
      "loss_contrastive": 0.19752535223960876,
      "loss_gen": 0.37991225719451904,
      "neg_sim_mean": 0.6375962495803833,
      "pos_sim_mean": 0.5400708913803101,
      "step": 2401
    },
    {
      "epoch": 0.4804,
      "grad_norm": 3.5771801471710205,
      "learning_rate": 6.206222212470675e-05,
      "loss": 0.8635,
      "step": 2402
    },
    {
      "epoch": 0.4804,
      "loss_align": 0.11783826351165771,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12969167530536652,
      "neg_sim_mean": 0.7622181177139282,
      "pos_sim_mean": 0.8821617364883423,
      "step": 2402
    },
    {
      "epoch": 0.4806,
      "grad_norm": 1.4626002311706543,
      "learning_rate": 6.202834359044959e-05,
      "loss": 0.2475,
      "step": 2403
    },
    {
      "epoch": 0.4806,
      "loss_align": 0.27465397119522095,
      "loss_contrastive": 0.05619150400161743,
      "loss_gen": 0.1478358656167984,
      "neg_sim_mean": 0.6815375089645386,
      "pos_sim_mean": 0.725346028804779,
      "step": 2403
    },
    {
      "epoch": 0.4808,
      "grad_norm": 1.7530391216278076,
      "learning_rate": 6.199445919372389e-05,
      "loss": 0.4292,
      "step": 2404
    },
    {
      "epoch": 0.4808,
      "loss_align": 0.11807554960250854,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3010479509830475,
      "neg_sim_mean": 0.672987699508667,
      "pos_sim_mean": 0.8819244503974915,
      "step": 2404
    },
    {
      "epoch": 0.481,
      "grad_norm": 1.830617070198059,
      "learning_rate": 6.196056895104448e-05,
      "loss": 0.4191,
      "step": 2405
    },
    {
      "epoch": 0.481,
      "loss_align": 0.2703315019607544,
      "loss_contrastive": 0.006308317184448242,
      "loss_gen": 0.18889981508255005,
      "neg_sim_mean": 0.6359767913818359,
      "pos_sim_mean": 0.7296684980392456,
      "step": 2405
    },
    {
      "epoch": 0.4812,
      "grad_norm": 1.818926453590393,
      "learning_rate": 6.192667287892905e-05,
      "loss": 0.46,
      "step": 2406
    },
    {
      "epoch": 0.4812,
      "loss_align": 0.3476495146751404,
      "loss_contrastive": 0.07742148637771606,
      "loss_gen": 0.22742554545402527,
      "neg_sim_mean": 0.6297719478607178,
      "pos_sim_mean": 0.6523504853248596,
      "step": 2406
    },
    {
      "epoch": 0.4814,
      "grad_norm": 1.9733856916427612,
      "learning_rate": 6.189277099389816e-05,
      "loss": 0.5844,
      "step": 2407
    },
    {
      "epoch": 0.4814,
      "loss_align": 0.16057002544403076,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13876768946647644,
      "neg_sim_mean": 0.6729337573051453,
      "pos_sim_mean": 0.8394299745559692,
      "step": 2407
    },
    {
      "epoch": 0.4816,
      "grad_norm": 1.758565068244934,
      "learning_rate": 6.185886331247515e-05,
      "loss": 0.2993,
      "step": 2408
    },
    {
      "epoch": 0.4816,
      "loss_align": 0.12349796295166016,
      "loss_contrastive": 0.11435079574584961,
      "loss_gen": 0.35489627718925476,
      "neg_sim_mean": 0.8908528089523315,
      "pos_sim_mean": 0.8765020370483398,
      "step": 2408
    },
    {
      "epoch": 0.4818,
      "grad_norm": 2.2588748931884766,
      "learning_rate": 6.182494985118624e-05,
      "loss": 0.4921,
      "step": 2409
    },
    {
      "epoch": 0.4818,
      "loss_align": 0.15680217742919922,
      "loss_contrastive": 0.0154494047164917,
      "loss_gen": 0.1933307647705078,
      "neg_sim_mean": 0.7586472034454346,
      "pos_sim_mean": 0.8431978225708008,
      "step": 2409
    },
    {
      "epoch": 0.482,
      "grad_norm": 1.7589409351348877,
      "learning_rate": 6.179103062656042e-05,
      "loss": 0.352,
      "step": 2410
    },
    {
      "epoch": 0.482,
      "loss_align": 0.35599541664123535,
      "loss_contrastive": 0.08342629671096802,
      "loss_gen": 0.22926795482635498,
      "neg_sim_mean": 0.6274308562278748,
      "pos_sim_mean": 0.6440045833587646,
      "step": 2410
    },
    {
      "epoch": 0.4822,
      "grad_norm": 2.3892455101013184,
      "learning_rate": 6.17571056551295e-05,
      "loss": 0.5953,
      "step": 2411
    },
    {
      "epoch": 0.4822,
      "loss_align": 0.1962885856628418,
      "loss_contrastive": 0.0,
      "loss_gen": 0.07035704702138901,
      "neg_sim_mean": 0.6960432529449463,
      "pos_sim_mean": 0.8037114143371582,
      "step": 2411
    },
    {
      "epoch": 0.4824,
      "grad_norm": 1.063927412033081,
      "learning_rate": 6.172317495342812e-05,
      "loss": 0.2666,
      "step": 2412
    },
    {
      "epoch": 0.4824,
      "loss_align": 0.17516767978668213,
      "loss_contrastive": 0.09335470199584961,
      "loss_gen": 0.1764436960220337,
      "neg_sim_mean": 0.8181869983673096,
      "pos_sim_mean": 0.8248323202133179,
      "step": 2412
    },
    {
      "epoch": 0.4826,
      "grad_norm": 1.83170747756958,
      "learning_rate": 6.168923853799369e-05,
      "loss": 0.3628,
      "step": 2413
    },
    {
      "epoch": 0.4826,
      "loss_align": 0.13317126035690308,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007352376822382212,
      "neg_sim_mean": 0.425650417804718,
      "pos_sim_mean": 0.8668287396430969,
      "step": 2413
    },
    {
      "epoch": 0.4828,
      "grad_norm": 0.5395276546478271,
      "learning_rate": 6.16552964253664e-05,
      "loss": 0.1405,
      "step": 2414
    },
    {
      "epoch": 0.4828,
      "loss_align": 0.11304426193237305,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028134722262620926,
      "neg_sim_mean": 0.30781131982803345,
      "pos_sim_mean": 0.886955738067627,
      "step": 2414
    },
    {
      "epoch": 0.483,
      "grad_norm": 0.93443363904953,
      "learning_rate": 6.16213486320892e-05,
      "loss": 0.1412,
      "step": 2415
    },
    {
      "epoch": 0.483,
      "loss_align": 0.19230204820632935,
      "loss_contrastive": 0.10115092992782593,
      "loss_gen": 0.17788183689117432,
      "neg_sim_mean": 0.8088488578796387,
      "pos_sim_mean": 0.8076979517936707,
      "step": 2415
    },
    {
      "epoch": 0.4832,
      "grad_norm": 1.862151861190796,
      "learning_rate": 6.158739517470786e-05,
      "loss": 0.3823,
      "step": 2416
    },
    {
      "epoch": 0.4832,
      "loss_align": 0.1706451177597046,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1934128999710083,
      "neg_sim_mean": 0.6824910640716553,
      "pos_sim_mean": 0.8293548822402954,
      "step": 2416
    },
    {
      "epoch": 0.4834,
      "grad_norm": 1.6793296337127686,
      "learning_rate": 6.155343606977091e-05,
      "loss": 0.3641,
      "step": 2417
    },
    {
      "epoch": 0.4834,
      "loss_align": 0.14133459329605103,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00898527167737484,
      "neg_sim_mean": 0.48995155096054077,
      "pos_sim_mean": 0.858665406703949,
      "step": 2417
    },
    {
      "epoch": 0.4836,
      "grad_norm": 0.5911709666252136,
      "learning_rate": 6.151947133382954e-05,
      "loss": 0.1503,
      "step": 2418
    },
    {
      "epoch": 0.4836,
      "loss_align": 0.17937743663787842,
      "loss_contrastive": 0.18822240829467773,
      "loss_gen": 0.27787911891937256,
      "neg_sim_mean": 0.9088449478149414,
      "pos_sim_mean": 0.8206225633621216,
      "step": 2418
    },
    {
      "epoch": 0.4838,
      "grad_norm": 2.3434388637542725,
      "learning_rate": 6.148550098343778e-05,
      "loss": 0.4798,
      "step": 2419
    },
    {
      "epoch": 0.4838,
      "loss_align": 0.2988746166229248,
      "loss_contrastive": 0.15139782428741455,
      "loss_gen": 0.2593882977962494,
      "neg_sim_mean": 0.7525231838226318,
      "pos_sim_mean": 0.7011253833770752,
      "step": 2419
    },
    {
      "epoch": 0.484,
      "grad_norm": 2.2904579639434814,
      "learning_rate": 6.145152503515239e-05,
      "loss": 0.5764,
      "step": 2420
    },
    {
      "epoch": 0.484,
      "loss_align": 0.10446488857269287,
      "loss_contrastive": 0.07763409614562988,
      "loss_gen": 0.3328511416912079,
      "neg_sim_mean": 0.8731691837310791,
      "pos_sim_mean": 0.8955351114273071,
      "step": 2420
    },
    {
      "epoch": 0.4842,
      "grad_norm": 2.6232168674468994,
      "learning_rate": 6.141754350553279e-05,
      "loss": 0.4466,
      "step": 2421
    },
    {
      "epoch": 0.4842,
      "loss_align": 0.254810631275177,
      "loss_contrastive": 0.08811992406845093,
      "loss_gen": 0.40138518810272217,
      "neg_sim_mean": 0.733309268951416,
      "pos_sim_mean": 0.745189368724823,
      "step": 2421
    },
    {
      "epoch": 0.4844,
      "grad_norm": 2.4087653160095215,
      "learning_rate": 6.13835564111412e-05,
      "loss": 0.6668,
      "step": 2422
    },
    {
      "epoch": 0.4844,
      "loss_align": 0.08383524417877197,
      "loss_contrastive": 0.03045254945755005,
      "loss_gen": 0.29043394327163696,
      "neg_sim_mean": 0.8466172814369202,
      "pos_sim_mean": 0.916164755821228,
      "step": 2422
    },
    {
      "epoch": 0.4846,
      "grad_norm": 1.9050933122634888,
      "learning_rate": 6.134956376854251e-05,
      "loss": 0.3779,
      "step": 2423
    },
    {
      "epoch": 0.4846,
      "loss_align": 0.20442914962768555,
      "loss_contrastive": 0.025556862354278564,
      "loss_gen": 0.2882797122001648,
      "neg_sim_mean": 0.7211276888847351,
      "pos_sim_mean": 0.7955708503723145,
      "step": 2423
    },
    {
      "epoch": 0.4848,
      "grad_norm": 2.113996744155884,
      "learning_rate": 6.13155655943043e-05,
      "loss": 0.4958,
      "step": 2424
    },
    {
      "epoch": 0.4848,
      "loss_align": 0.09996926784515381,
      "loss_contrastive": 0.0,
      "loss_gen": 0.027018779888749123,
      "neg_sim_mean": 0.32323935627937317,
      "pos_sim_mean": 0.9000307321548462,
      "step": 2424
    },
    {
      "epoch": 0.485,
      "grad_norm": 1.2159343957901,
      "learning_rate": 6.128156190499688e-05,
      "loss": 0.127,
      "step": 2425
    },
    {
      "epoch": 0.485,
      "loss_align": 0.3010757565498352,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007075665518641472,
      "neg_sim_mean": 0.4776611030101776,
      "pos_sim_mean": 0.6989242434501648,
      "step": 2425
    },
    {
      "epoch": 0.4852,
      "grad_norm": 1.042484998703003,
      "learning_rate": 6.124755271719325e-05,
      "loss": 0.3082,
      "step": 2426
    },
    {
      "epoch": 0.4852,
      "loss_align": 0.08554720878601074,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006083989515900612,
      "neg_sim_mean": 0.28027501702308655,
      "pos_sim_mean": 0.9144527912139893,
      "step": 2426
    },
    {
      "epoch": 0.4854,
      "grad_norm": 0.4679703712463379,
      "learning_rate": 6.121353804746907e-05,
      "loss": 0.0916,
      "step": 2427
    },
    {
      "epoch": 0.4854,
      "loss_align": 0.07106095552444458,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21878497302532196,
      "neg_sim_mean": 0.7884617447853088,
      "pos_sim_mean": 0.9289390444755554,
      "step": 2427
    },
    {
      "epoch": 0.4856,
      "grad_norm": 1.7353689670562744,
      "learning_rate": 6.117951791240265e-05,
      "loss": 0.2898,
      "step": 2428
    },
    {
      "epoch": 0.4856,
      "loss_align": 0.10296082496643066,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007960851304233074,
      "neg_sim_mean": 0.40989193320274353,
      "pos_sim_mean": 0.8970391750335693,
      "step": 2428
    },
    {
      "epoch": 0.4858,
      "grad_norm": 0.5935572385787964,
      "learning_rate": 6.114549232857503e-05,
      "loss": 0.1109,
      "step": 2429
    },
    {
      "epoch": 0.4858,
      "loss_align": 0.1499091386795044,
      "loss_contrastive": 0.018706202507019043,
      "loss_gen": 0.1816893368959427,
      "neg_sim_mean": 0.7687970399856567,
      "pos_sim_mean": 0.8500908613204956,
      "step": 2429
    },
    {
      "epoch": 0.486,
      "grad_norm": 1.6714009046554565,
      "learning_rate": 6.111146131256983e-05,
      "loss": 0.3338,
      "step": 2430
    },
    {
      "epoch": 0.486,
      "loss_align": 0.28416121006011963,
      "loss_contrastive": 0.1687474250793457,
      "loss_gen": 0.26126739382743835,
      "neg_sim_mean": 0.7845861911773682,
      "pos_sim_mean": 0.7158387899398804,
      "step": 2430
    },
    {
      "epoch": 0.4862,
      "grad_norm": 2.3125059604644775,
      "learning_rate": 6.107742488097338e-05,
      "loss": 0.5657,
      "step": 2431
    },
    {
      "epoch": 0.4862,
      "loss_align": 0.40190744400024414,
      "loss_contrastive": 0.05006185173988342,
      "loss_gen": 0.25415492057800293,
      "neg_sim_mean": 0.5481544137001038,
      "pos_sim_mean": 0.5980925559997559,
      "step": 2431
    },
    {
      "epoch": 0.4864,
      "grad_norm": 2.546661138534546,
      "learning_rate": 6.104338305037459e-05,
      "loss": 0.6621,
      "step": 2432
    },
    {
      "epoch": 0.4864,
      "loss_align": 0.1894097924232483,
      "loss_contrastive": 0.1074400544166565,
      "loss_gen": 0.21073926985263824,
      "neg_sim_mean": 0.8180302381515503,
      "pos_sim_mean": 0.8105902075767517,
      "step": 2432
    },
    {
      "epoch": 0.4866,
      "grad_norm": 1.7458546161651611,
      "learning_rate": 6.100933583736508e-05,
      "loss": 0.413,
      "step": 2433
    },
    {
      "epoch": 0.4866,
      "loss_align": 0.1353616714477539,
      "loss_contrastive": 0.11283469200134277,
      "loss_gen": 0.3046617805957794,
      "neg_sim_mean": 0.877472996711731,
      "pos_sim_mean": 0.8646383285522461,
      "step": 2433
    },
    {
      "epoch": 0.4868,
      "grad_norm": 2.397552967071533,
      "learning_rate": 6.097528325853903e-05,
      "loss": 0.4536,
      "step": 2434
    },
    {
      "epoch": 0.4868,
      "loss_align": 0.1271604299545288,
      "loss_contrastive": 0.08730661869049072,
      "loss_gen": 0.18654805421829224,
      "neg_sim_mean": 0.860146164894104,
      "pos_sim_mean": 0.8728395700454712,
      "step": 2434
    },
    {
      "epoch": 0.487,
      "grad_norm": 1.950222373008728,
      "learning_rate": 6.0941225330493235e-05,
      "loss": 0.3242,
      "step": 2435
    },
    {
      "epoch": 0.487,
      "loss_align": 0.24348723888397217,
      "loss_contrastive": 0.14392197132110596,
      "loss_gen": 0.09899929910898209,
      "neg_sim_mean": 0.8004347085952759,
      "pos_sim_mean": 0.7565127611160278,
      "step": 2435
    },
    {
      "epoch": 0.4872,
      "grad_norm": 1.5003820657730103,
      "learning_rate": 6.090716206982714e-05,
      "loss": 0.3598,
      "step": 2436
    },
    {
      "epoch": 0.4872,
      "loss_align": 0.3062177896499634,
      "loss_contrastive": 0.14053940773010254,
      "loss_gen": 0.3532106876373291,
      "neg_sim_mean": 0.7343215942382812,
      "pos_sim_mean": 0.6937822103500366,
      "step": 2436
    },
    {
      "epoch": 0.4874,
      "grad_norm": 2.177974224090576,
      "learning_rate": 6.0873093493142744e-05,
      "loss": 0.6763,
      "step": 2437
    },
    {
      "epoch": 0.4874,
      "loss_align": 0.08234524726867676,
      "loss_contrastive": 0.05273759365081787,
      "loss_gen": 0.10057535022497177,
      "neg_sim_mean": 0.8703923225402832,
      "pos_sim_mean": 0.9176547527313232,
      "step": 2437
    },
    {
      "epoch": 0.4876,
      "grad_norm": 1.362637996673584,
      "learning_rate": 6.083901961704467e-05,
      "loss": 0.1892,
      "step": 2438
    },
    {
      "epoch": 0.4876,
      "loss_align": 0.1185045838356018,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00257893861271441,
      "neg_sim_mean": 0.31406253576278687,
      "pos_sim_mean": 0.8814954161643982,
      "step": 2438
    },
    {
      "epoch": 0.4878,
      "grad_norm": 0.4269247353076935,
      "learning_rate": 6.080494045814011e-05,
      "loss": 0.1211,
      "step": 2439
    },
    {
      "epoch": 0.4878,
      "loss_align": 0.22211074829101562,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3245435357093811,
      "neg_sim_mean": 0.41596341133117676,
      "pos_sim_mean": 0.7778892517089844,
      "step": 2439
    },
    {
      "epoch": 0.488,
      "grad_norm": 2.481224298477173,
      "learning_rate": 6.077085603303882e-05,
      "loss": 0.5467,
      "step": 2440
    },
    {
      "epoch": 0.488,
      "loss_align": 0.2091681957244873,
      "loss_contrastive": 0.15344762802124023,
      "loss_gen": 0.22298656404018402,
      "neg_sim_mean": 0.844279408454895,
      "pos_sim_mean": 0.7908318042755127,
      "step": 2440
    },
    {
      "epoch": 0.4882,
      "grad_norm": 1.6635725498199463,
      "learning_rate": 6.073676635835317e-05,
      "loss": 0.4506,
      "step": 2441
    },
    {
      "epoch": 0.4882,
      "loss_align": 0.14349102973937988,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007107590790838003,
      "neg_sim_mean": 0.5744126439094543,
      "pos_sim_mean": 0.8565089702606201,
      "step": 2441
    },
    {
      "epoch": 0.4884,
      "grad_norm": 0.6136196255683899,
      "learning_rate": 6.0702671450698e-05,
      "loss": 0.1506,
      "step": 2442
    },
    {
      "epoch": 0.4884,
      "loss_align": 0.22103726863861084,
      "loss_contrastive": 0.16350144147872925,
      "loss_gen": 0.18049360811710358,
      "neg_sim_mean": 0.8424641489982605,
      "pos_sim_mean": 0.7789627313613892,
      "step": 2442
    },
    {
      "epoch": 0.4886,
      "grad_norm": 1.6764458417892456,
      "learning_rate": 6.066857132669081e-05,
      "loss": 0.4212,
      "step": 2443
    },
    {
      "epoch": 0.4886,
      "loss_align": 0.09796100854873657,
      "loss_contrastive": 0.09940427541732788,
      "loss_gen": 0.21869659423828125,
      "neg_sim_mean": 0.9014432430267334,
      "pos_sim_mean": 0.9020389914512634,
      "step": 2443
    },
    {
      "epoch": 0.4888,
      "grad_norm": 1.8590798377990723,
      "learning_rate": 6.063446600295154e-05,
      "loss": 0.3286,
      "step": 2444
    },
    {
      "epoch": 0.4888,
      "loss_align": 0.08276635408401489,
      "loss_contrastive": 0.016704320907592773,
      "loss_gen": 0.18724459409713745,
      "neg_sim_mean": 0.83393794298172,
      "pos_sim_mean": 0.9172336459159851,
      "step": 2444
    },
    {
      "epoch": 0.489,
      "grad_norm": 1.9458783864974976,
      "learning_rate": 6.0600355496102745e-05,
      "loss": 0.272,
      "step": 2445
    },
    {
      "epoch": 0.489,
      "loss_align": 0.23094499111175537,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2107240855693817,
      "neg_sim_mean": 0.6442298293113708,
      "pos_sim_mean": 0.7690550088882446,
      "step": 2445
    },
    {
      "epoch": 0.4892,
      "grad_norm": 2.451303243637085,
      "learning_rate": 6.056623982276944e-05,
      "loss": 0.4417,
      "step": 2446
    },
    {
      "epoch": 0.4892,
      "loss_align": 0.12064504623413086,
      "loss_contrastive": 0.08216977119445801,
      "loss_gen": 0.21638531982898712,
      "neg_sim_mean": 0.8615247011184692,
      "pos_sim_mean": 0.8793549537658691,
      "step": 2446
    },
    {
      "epoch": 0.4894,
      "grad_norm": 1.5667043924331665,
      "learning_rate": 6.05321189995792e-05,
      "loss": 0.3469,
      "step": 2447
    },
    {
      "epoch": 0.4894,
      "loss_align": 0.01022028923034668,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0024036874528974295,
      "neg_sim_mean": 0.33229076862335205,
      "pos_sim_mean": 0.9897797107696533,
      "step": 2447
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.1978950798511505,
      "learning_rate": 6.049799304316214e-05,
      "loss": 0.0126,
      "step": 2448
    },
    {
      "epoch": 0.4896,
      "loss_align": 0.15884774923324585,
      "loss_contrastive": 0.06414180994033813,
      "loss_gen": 0.4998503029346466,
      "neg_sim_mean": 0.8052940368652344,
      "pos_sim_mean": 0.8411522507667542,
      "step": 2448
    },
    {
      "epoch": 0.4898,
      "grad_norm": 2.4465911388397217,
      "learning_rate": 6.046386197015076e-05,
      "loss": 0.6664,
      "step": 2449
    },
    {
      "epoch": 0.4898,
      "loss_align": 0.25181007385253906,
      "loss_contrastive": 0.05108189582824707,
      "loss_gen": 0.20622499287128448,
      "neg_sim_mean": 0.6992717981338501,
      "pos_sim_mean": 0.7481899261474609,
      "step": 2449
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.968401312828064,
      "learning_rate": 6.042972579718019e-05,
      "loss": 0.4642,
      "step": 2450
    },
    {
      "epoch": 0.49,
      "loss_align": 0.13531279563903809,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0036374793853610754,
      "neg_sim_mean": 0.4707900285720825,
      "pos_sim_mean": 0.8646872043609619,
      "step": 2450
    },
    {
      "epoch": 0.4902,
      "grad_norm": 0.45416906476020813,
      "learning_rate": 6.0395584540887963e-05,
      "loss": 0.139,
      "step": 2451
    },
    {
      "epoch": 0.4902,
      "loss_align": 0.19531762599945068,
      "loss_contrastive": 0.07376891374588013,
      "loss_gen": 0.19877657294273376,
      "neg_sim_mean": 0.7784512639045715,
      "pos_sim_mean": 0.8046823740005493,
      "step": 2451
    },
    {
      "epoch": 0.4904,
      "grad_norm": 1.719552993774414,
      "learning_rate": 6.0361438217914125e-05,
      "loss": 0.4029,
      "step": 2452
    },
    {
      "epoch": 0.4904,
      "loss_align": 0.27777963876724243,
      "loss_contrastive": 0.0,
      "loss_gen": 0.32792791724205017,
      "neg_sim_mean": 0.5631883144378662,
      "pos_sim_mean": 0.7222203612327576,
      "step": 2452
    },
    {
      "epoch": 0.4906,
      "grad_norm": 2.128969430923462,
      "learning_rate": 6.032728684490118e-05,
      "loss": 0.6057,
      "step": 2453
    },
    {
      "epoch": 0.4906,
      "loss_align": 0.17465734481811523,
      "loss_contrastive": 0.06755590438842773,
      "loss_gen": 0.49890175461769104,
      "neg_sim_mean": 0.7928985357284546,
      "pos_sim_mean": 0.8253426551818848,
      "step": 2453
    },
    {
      "epoch": 0.4908,
      "grad_norm": 2.379319667816162,
      "learning_rate": 6.029313043849407e-05,
      "loss": 0.6817,
      "step": 2454
    },
    {
      "epoch": 0.4908,
      "loss_align": 0.13007527589797974,
      "loss_contrastive": 0.00026702880859375,
      "loss_gen": 0.364562064409256,
      "neg_sim_mean": 0.7701917290687561,
      "pos_sim_mean": 0.8699247241020203,
      "step": 2454
    },
    {
      "epoch": 0.491,
      "grad_norm": 2.0249266624450684,
      "learning_rate": 6.025896901534023e-05,
      "loss": 0.4947,
      "step": 2455
    },
    {
      "epoch": 0.491,
      "loss_align": 0.12225198745727539,
      "loss_contrastive": 0.0,
      "loss_gen": 0.218730166554451,
      "neg_sim_mean": 0.7735191583633423,
      "pos_sim_mean": 0.8777480125427246,
      "step": 2455
    },
    {
      "epoch": 0.4912,
      "grad_norm": 1.7546199560165405,
      "learning_rate": 6.0224802592089513e-05,
      "loss": 0.341,
      "step": 2456
    },
    {
      "epoch": 0.4912,
      "loss_align": 0.25051069259643555,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09234799444675446,
      "neg_sim_mean": 0.5547239780426025,
      "pos_sim_mean": 0.7494893074035645,
      "step": 2456
    },
    {
      "epoch": 0.4914,
      "grad_norm": 1.4527864456176758,
      "learning_rate": 6.019063118539425e-05,
      "loss": 0.3429,
      "step": 2457
    },
    {
      "epoch": 0.4914,
      "loss_align": 0.09389233589172363,
      "loss_contrastive": 0.0692136287689209,
      "loss_gen": 0.09248391538858414,
      "neg_sim_mean": 0.8753212690353394,
      "pos_sim_mean": 0.9061076641082764,
      "step": 2457
    },
    {
      "epoch": 0.4916,
      "grad_norm": 1.2266850471496582,
      "learning_rate": 6.015645481190911e-05,
      "loss": 0.1947,
      "step": 2458
    },
    {
      "epoch": 0.4916,
      "loss_align": 0.2936882972717285,
      "loss_contrastive": 0.19431358575820923,
      "loss_gen": 0.21941612660884857,
      "neg_sim_mean": 0.8006252646446228,
      "pos_sim_mean": 0.7063117027282715,
      "step": 2458
    },
    {
      "epoch": 0.4918,
      "grad_norm": 1.9528381824493408,
      "learning_rate": 6.0122273488291304e-05,
      "loss": 0.5364,
      "step": 2459
    },
    {
      "epoch": 0.4918,
      "loss_align": 0.24071168899536133,
      "loss_contrastive": 0.0,
      "loss_gen": 0.41855135560035706,
      "neg_sim_mean": 0.5546934604644775,
      "pos_sim_mean": 0.7592883110046387,
      "step": 2459
    },
    {
      "epoch": 0.492,
      "grad_norm": 2.184838056564331,
      "learning_rate": 6.008808723120035e-05,
      "loss": 0.6593,
      "step": 2460
    },
    {
      "epoch": 0.492,
      "loss_align": 0.14864790439605713,
      "loss_contrastive": 0.15403372049331665,
      "loss_gen": 0.184220090508461,
      "neg_sim_mean": 0.9053857922554016,
      "pos_sim_mean": 0.8513520956039429,
      "step": 2460
    },
    {
      "epoch": 0.4922,
      "grad_norm": 1.661287784576416,
      "learning_rate": 6.005389605729824e-05,
      "loss": 0.3514,
      "step": 2461
    },
    {
      "epoch": 0.4922,
      "loss_align": 0.26371538639068604,
      "loss_contrastive": 0.04058825969696045,
      "loss_gen": 0.3204244077205658,
      "neg_sim_mean": 0.6768728494644165,
      "pos_sim_mean": 0.736284613609314,
      "step": 2461
    },
    {
      "epoch": 0.4924,
      "grad_norm": 2.3246383666992188,
      "learning_rate": 6.0019699983249324e-05,
      "loss": 0.589,
      "step": 2462
    },
    {
      "epoch": 0.4924,
      "loss_align": 0.11137014627456665,
      "loss_contrastive": 0.050541043281555176,
      "loss_gen": 0.4408610463142395,
      "neg_sim_mean": 0.8391708731651306,
      "pos_sim_mean": 0.8886298537254333,
      "step": 2462
    },
    {
      "epoch": 0.4926,
      "grad_norm": 2.079242467880249,
      "learning_rate": 5.9985499025720346e-05,
      "loss": 0.5583,
      "step": 2463
    },
    {
      "epoch": 0.4926,
      "loss_align": 0.11774003505706787,
      "loss_contrastive": 0.03979825973510742,
      "loss_gen": 0.20728273689746857,
      "neg_sim_mean": 0.8220582008361816,
      "pos_sim_mean": 0.8822599649429321,
      "step": 2463
    },
    {
      "epoch": 0.4928,
      "grad_norm": 1.7176631689071655,
      "learning_rate": 5.995129320138047e-05,
      "loss": 0.3298,
      "step": 2464
    },
    {
      "epoch": 0.4928,
      "loss_align": 0.1798609495162964,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003256036899983883,
      "neg_sim_mean": 0.37726688385009766,
      "pos_sim_mean": 0.8201390504837036,
      "step": 2464
    },
    {
      "epoch": 0.493,
      "grad_norm": 0.516295850276947,
      "learning_rate": 5.991708252690117e-05,
      "loss": 0.1831,
      "step": 2465
    },
    {
      "epoch": 0.493,
      "loss_align": 0.09518563747406006,
      "loss_contrastive": 0.06087374687194824,
      "loss_gen": 0.3385506570339203,
      "neg_sim_mean": 0.8656880855560303,
      "pos_sim_mean": 0.9048143625259399,
      "step": 2465
    },
    {
      "epoch": 0.4932,
      "grad_norm": 2.615292549133301,
      "learning_rate": 5.988286701895631e-05,
      "loss": 0.441,
      "step": 2466
    },
    {
      "epoch": 0.4932,
      "loss_align": 0.22914326190948486,
      "loss_contrastive": 0.1405467987060547,
      "loss_gen": 0.15099181234836578,
      "neg_sim_mean": 0.8114035129547119,
      "pos_sim_mean": 0.7708567380905151,
      "step": 2466
    },
    {
      "epoch": 0.4934,
      "grad_norm": 1.4290339946746826,
      "learning_rate": 5.9848646694222146e-05,
      "loss": 0.397,
      "step": 2467
    },
    {
      "epoch": 0.4934,
      "loss_align": 0.19010871648788452,
      "loss_contrastive": 0.09282469749450684,
      "loss_gen": 0.14935767650604248,
      "neg_sim_mean": 0.8027159571647644,
      "pos_sim_mean": 0.8098912835121155,
      "step": 2467
    },
    {
      "epoch": 0.4936,
      "grad_norm": 1.6746543645858765,
      "learning_rate": 5.98144215693772e-05,
      "loss": 0.3506,
      "step": 2468
    },
    {
      "epoch": 0.4936,
      "loss_align": 0.26337510347366333,
      "loss_contrastive": 0.042704999446868896,
      "loss_gen": 0.29277658462524414,
      "neg_sim_mean": 0.6793298721313477,
      "pos_sim_mean": 0.7366248965263367,
      "step": 2468
    },
    {
      "epoch": 0.4938,
      "grad_norm": 2.343899726867676,
      "learning_rate": 5.9780191661102415e-05,
      "loss": 0.5613,
      "step": 2469
    },
    {
      "epoch": 0.4938,
      "loss_align": 0.1186600923538208,
      "loss_contrastive": 0.09221100807189941,
      "loss_gen": 0.12948472797870636,
      "neg_sim_mean": 0.8735508918762207,
      "pos_sim_mean": 0.8813399076461792,
      "step": 2469
    },
    {
      "epoch": 0.494,
      "grad_norm": 1.5046190023422241,
      "learning_rate": 5.974595698608103e-05,
      "loss": 0.2592,
      "step": 2470
    },
    {
      "epoch": 0.494,
      "loss_align": 0.186650812625885,
      "loss_contrastive": 0.06717091798782349,
      "loss_gen": 0.1938575804233551,
      "neg_sim_mean": 0.7805200815200806,
      "pos_sim_mean": 0.813349187374115,
      "step": 2470
    },
    {
      "epoch": 0.4942,
      "grad_norm": 2.15801739692688,
      "learning_rate": 5.97117175609986e-05,
      "loss": 0.3886,
      "step": 2471
    },
    {
      "epoch": 0.4942,
      "loss_align": 0.24341940879821777,
      "loss_contrastive": 0.10826116800308228,
      "loss_gen": 0.1948927342891693,
      "neg_sim_mean": 0.7648417353630066,
      "pos_sim_mean": 0.7565805912017822,
      "step": 2471
    },
    {
      "epoch": 0.4944,
      "grad_norm": 1.8046643733978271,
      "learning_rate": 5.967747340254303e-05,
      "loss": 0.4513,
      "step": 2472
    },
    {
      "epoch": 0.4944,
      "loss_align": 0.2350243330001831,
      "loss_contrastive": 0.1305655837059021,
      "loss_gen": 0.29577919840812683,
      "neg_sim_mean": 0.7955412268638611,
      "pos_sim_mean": 0.7649756669998169,
      "step": 2472
    },
    {
      "epoch": 0.4946,
      "grad_norm": 2.004284620285034,
      "learning_rate": 5.964322452740445e-05,
      "loss": 0.5465,
      "step": 2473
    },
    {
      "epoch": 0.4946,
      "loss_align": 0.12136560678482056,
      "loss_contrastive": 0.07198387384414673,
      "loss_gen": 0.28111347556114197,
      "neg_sim_mean": 0.8506182432174683,
      "pos_sim_mean": 0.8786343932151794,
      "step": 2473
    },
    {
      "epoch": 0.4948,
      "grad_norm": 2.0184428691864014,
      "learning_rate": 5.960897095227541e-05,
      "loss": 0.4111,
      "step": 2474
    },
    {
      "epoch": 0.4948,
      "loss_align": 0.14635133743286133,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19969724118709564,
      "neg_sim_mean": 0.7038881778717041,
      "pos_sim_mean": 0.8536486625671387,
      "step": 2474
    },
    {
      "epoch": 0.495,
      "grad_norm": 1.5345505475997925,
      "learning_rate": 5.9574712693850654e-05,
      "loss": 0.346,
      "step": 2475
    },
    {
      "epoch": 0.495,
      "loss_align": 0.1626027226448059,
      "loss_contrastive": 0.12286752462387085,
      "loss_gen": 0.15092243254184723,
      "neg_sim_mean": 0.860264778137207,
      "pos_sim_mean": 0.8373972773551941,
      "step": 2475
    },
    {
      "epoch": 0.4952,
      "grad_norm": 1.5349650382995605,
      "learning_rate": 5.9540449768827246e-05,
      "loss": 0.3283,
      "step": 2476
    },
    {
      "epoch": 0.4952,
      "loss_align": 0.14652979373931885,
      "loss_contrastive": 0.09320700168609619,
      "loss_gen": 0.30074289441108704,
      "neg_sim_mean": 0.8466771841049194,
      "pos_sim_mean": 0.8534702062606812,
      "step": 2476
    },
    {
      "epoch": 0.4954,
      "grad_norm": 1.8851397037506104,
      "learning_rate": 5.950618219390451e-05,
      "loss": 0.4585,
      "step": 2477
    },
    {
      "epoch": 0.4954,
      "loss_align": 0.1407185196876526,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007949752733111382,
      "neg_sim_mean": 0.33198827505111694,
      "pos_sim_mean": 0.8592814803123474,
      "step": 2477
    },
    {
      "epoch": 0.4956,
      "grad_norm": 0.5033396482467651,
      "learning_rate": 5.9471909985784066e-05,
      "loss": 0.1487,
      "step": 2478
    },
    {
      "epoch": 0.4956,
      "loss_align": 0.22290098667144775,
      "loss_contrastive": 0.0652737021446228,
      "loss_gen": 0.3861401677131653,
      "neg_sim_mean": 0.7423726916313171,
      "pos_sim_mean": 0.7770990133285522,
      "step": 2478
    },
    {
      "epoch": 0.4958,
      "grad_norm": 2.518205404281616,
      "learning_rate": 5.943763316116977e-05,
      "loss": 0.6169,
      "step": 2479
    },
    {
      "epoch": 0.4958,
      "loss_align": 0.15733832120895386,
      "loss_contrastive": 0.03724640607833862,
      "loss_gen": 0.12503953278064728,
      "neg_sim_mean": 0.7799080610275269,
      "pos_sim_mean": 0.8426616787910461,
      "step": 2479
    },
    {
      "epoch": 0.496,
      "grad_norm": 1.4400547742843628,
      "learning_rate": 5.9403351736767696e-05,
      "loss": 0.2868,
      "step": 2480
    },
    {
      "epoch": 0.496,
      "loss_align": 0.09512686729431152,
      "loss_contrastive": 0.0,
      "loss_gen": 0.030632568523287773,
      "neg_sim_mean": 0.321998655796051,
      "pos_sim_mean": 0.9048731327056885,
      "step": 2480
    },
    {
      "epoch": 0.4962,
      "grad_norm": 1.0223382711410522,
      "learning_rate": 5.9369065729286245e-05,
      "loss": 0.1258,
      "step": 2481
    },
    {
      "epoch": 0.4962,
      "loss_align": 0.2333717942237854,
      "loss_contrastive": 0.1387043595314026,
      "loss_gen": 0.20727509260177612,
      "neg_sim_mean": 0.8053325414657593,
      "pos_sim_mean": 0.7666282057762146,
      "step": 2481
    },
    {
      "epoch": 0.4964,
      "grad_norm": 1.874448537826538,
      "learning_rate": 5.933477515543595e-05,
      "loss": 0.4573,
      "step": 2482
    },
    {
      "epoch": 0.4964,
      "loss_align": 0.22280919551849365,
      "loss_contrastive": 0.09152209758758545,
      "loss_gen": 0.14931759238243103,
      "neg_sim_mean": 0.7687128782272339,
      "pos_sim_mean": 0.7771908044815063,
      "step": 2482
    },
    {
      "epoch": 0.4966,
      "grad_norm": 1.6777504682540894,
      "learning_rate": 5.930048003192965e-05,
      "loss": 0.3831,
      "step": 2483
    },
    {
      "epoch": 0.4966,
      "loss_align": 0.13213413953781128,
      "loss_contrastive": 0.05707746744155884,
      "loss_gen": 0.20511287450790405,
      "neg_sim_mean": 0.8249433040618896,
      "pos_sim_mean": 0.8678658604621887,
      "step": 2483
    },
    {
      "epoch": 0.4968,
      "grad_norm": 1.8134716749191284,
      "learning_rate": 5.926618037548237e-05,
      "loss": 0.3441,
      "step": 2484
    },
    {
      "epoch": 0.4968,
      "loss_align": 0.06028693914413452,
      "loss_contrastive": 0.056061625480651855,
      "loss_gen": 0.18531954288482666,
      "neg_sim_mean": 0.8957746624946594,
      "pos_sim_mean": 0.9397130608558655,
      "step": 2484
    },
    {
      "epoch": 0.497,
      "grad_norm": 1.6526422500610352,
      "learning_rate": 5.923187620281135e-05,
      "loss": 0.2523,
      "step": 2485
    },
    {
      "epoch": 0.497,
      "loss_align": 0.17140090465545654,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23789753019809723,
      "neg_sim_mean": 0.6382396817207336,
      "pos_sim_mean": 0.8285990953445435,
      "step": 2485
    },
    {
      "epoch": 0.4972,
      "grad_norm": 1.89521062374115,
      "learning_rate": 5.9197567530636014e-05,
      "loss": 0.4093,
      "step": 2486
    },
    {
      "epoch": 0.4972,
      "loss_align": 0.11615091562271118,
      "loss_contrastive": 0.08825230598449707,
      "loss_gen": 0.22097395360469818,
      "neg_sim_mean": 0.872101366519928,
      "pos_sim_mean": 0.8838490843772888,
      "step": 2486
    },
    {
      "epoch": 0.4974,
      "grad_norm": 1.9066541194915771,
      "learning_rate": 5.916325437567799e-05,
      "loss": 0.3477,
      "step": 2487
    },
    {
      "epoch": 0.4974,
      "loss_align": 0.2875770926475525,
      "loss_contrastive": 0.127760112285614,
      "loss_gen": 0.27205994725227356,
      "neg_sim_mean": 0.7401829957962036,
      "pos_sim_mean": 0.7124229073524475,
      "step": 2487
    },
    {
      "epoch": 0.4976,
      "grad_norm": 1.9887901544570923,
      "learning_rate": 5.912893675466111e-05,
      "loss": 0.575,
      "step": 2488
    },
    {
      "epoch": 0.4976,
      "loss_align": 0.07981234788894653,
      "loss_contrastive": 0.0407291054725647,
      "loss_gen": 0.2521282136440277,
      "neg_sim_mean": 0.8609167337417603,
      "pos_sim_mean": 0.9201876521110535,
      "step": 2488
    },
    {
      "epoch": 0.4978,
      "grad_norm": 1.8110992908477783,
      "learning_rate": 5.9094614684311345e-05,
      "loss": 0.3368,
      "step": 2489
    },
    {
      "epoch": 0.4978,
      "loss_align": 0.1443905234336853,
      "loss_contrastive": 0.07563269138336182,
      "loss_gen": 0.2226051241159439,
      "neg_sim_mean": 0.8312421441078186,
      "pos_sim_mean": 0.8556094765663147,
      "step": 2489
    },
    {
      "epoch": 0.498,
      "grad_norm": 1.911974310874939,
      "learning_rate": 5.906028818135687e-05,
      "loss": 0.3761,
      "step": 2490
    },
    {
      "epoch": 0.498,
      "loss_align": 0.10007154941558838,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03581148386001587,
      "neg_sim_mean": 0.35501569509506226,
      "pos_sim_mean": 0.8999284505844116,
      "step": 2490
    },
    {
      "epoch": 0.4982,
      "grad_norm": 1.3039166927337646,
      "learning_rate": 5.902595726252801e-05,
      "loss": 0.1359,
      "step": 2491
    },
    {
      "epoch": 0.4982,
      "loss_align": 0.13480228185653687,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2136726975440979,
      "neg_sim_mean": 0.5653153657913208,
      "pos_sim_mean": 0.8651977181434631,
      "step": 2491
    },
    {
      "epoch": 0.4984,
      "grad_norm": 1.6343711614608765,
      "learning_rate": 5.899162194455722e-05,
      "loss": 0.3485,
      "step": 2492
    },
    {
      "epoch": 0.4984,
      "loss_align": 0.2218794822692871,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2924191653728485,
      "neg_sim_mean": 0.6659560203552246,
      "pos_sim_mean": 0.7781205177307129,
      "step": 2492
    },
    {
      "epoch": 0.4986,
      "grad_norm": 2.0479319095611572,
      "learning_rate": 5.8957282244179124e-05,
      "loss": 0.5143,
      "step": 2493
    },
    {
      "epoch": 0.4986,
      "loss_align": 0.47383975982666016,
      "loss_contrastive": 0.3161351978778839,
      "loss_gen": 0.22318831086158752,
      "neg_sim_mean": 0.7422954440116882,
      "pos_sim_mean": 0.5261602401733398,
      "step": 2493
    },
    {
      "epoch": 0.4988,
      "grad_norm": 2.2938926219940186,
      "learning_rate": 5.892293817813047e-05,
      "loss": 0.735,
      "step": 2494
    },
    {
      "epoch": 0.4988,
      "loss_align": 0.12017720937728882,
      "loss_contrastive": 0.0,
      "loss_gen": 0.017246145755052567,
      "neg_sim_mean": 0.3946777582168579,
      "pos_sim_mean": 0.8798227906227112,
      "step": 2494
    },
    {
      "epoch": 0.499,
      "grad_norm": 0.6981150507926941,
      "learning_rate": 5.888858976315016e-05,
      "loss": 0.1374,
      "step": 2495
    },
    {
      "epoch": 0.499,
      "loss_align": 0.14486682415008545,
      "loss_contrastive": 0.036858558654785156,
      "loss_gen": 0.22295844554901123,
      "neg_sim_mean": 0.7919917106628418,
      "pos_sim_mean": 0.8551331758499146,
      "step": 2495
    },
    {
      "epoch": 0.4992,
      "grad_norm": 1.5809404850006104,
      "learning_rate": 5.885423701597917e-05,
      "loss": 0.3722,
      "step": 2496
    },
    {
      "epoch": 0.4992,
      "loss_align": 0.2060648798942566,
      "loss_contrastive": 0.09614378213882446,
      "loss_gen": 0.16263331472873688,
      "neg_sim_mean": 0.79007887840271,
      "pos_sim_mean": 0.7939351201057434,
      "step": 2496
    },
    {
      "epoch": 0.4994,
      "grad_norm": 1.9012106657028198,
      "learning_rate": 5.8819879953360625e-05,
      "loss": 0.3802,
      "step": 2497
    },
    {
      "epoch": 0.4994,
      "loss_align": 0.11074662208557129,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10034841299057007,
      "neg_sim_mean": 0.6508688926696777,
      "pos_sim_mean": 0.8892533779144287,
      "step": 2497
    },
    {
      "epoch": 0.4996,
      "grad_norm": 1.1821539402008057,
      "learning_rate": 5.8785518592039734e-05,
      "loss": 0.2111,
      "step": 2498
    },
    {
      "epoch": 0.4996,
      "loss_align": 0.1756860613822937,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2637503743171692,
      "neg_sim_mean": 0.6869386434555054,
      "pos_sim_mean": 0.8243139386177063,
      "step": 2498
    },
    {
      "epoch": 0.4998,
      "grad_norm": 1.9582117795944214,
      "learning_rate": 5.875115294876381e-05,
      "loss": 0.4394,
      "step": 2499
    },
    {
      "epoch": 0.4998,
      "loss_align": 0.310186505317688,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1003657728433609,
      "neg_sim_mean": 0.573229193687439,
      "pos_sim_mean": 0.689813494682312,
      "step": 2499
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.3539055585861206,
      "learning_rate": 5.8716783040282244e-05,
      "loss": 0.4106,
      "step": 2500
    },
    {
      "epoch": 0.5,
      "loss_align": 0.161321759223938,
      "loss_contrastive": 0.02378106117248535,
      "loss_gen": 0.9095543026924133,
      "neg_sim_mean": 0.7624592781066895,
      "pos_sim_mean": 0.838678240776062,
      "step": 2500
    },
    {
      "epoch": 0.5002,
      "grad_norm": 3.087839126586914,
      "learning_rate": 5.868240888334653e-05,
      "loss": 1.0737,
      "step": 2501
    },
    {
      "epoch": 0.5002,
      "loss_align": 0.1845446228981018,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15362046658992767,
      "neg_sim_mean": 0.5865867137908936,
      "pos_sim_mean": 0.8154553771018982,
      "step": 2501
    },
    {
      "epoch": 0.5004,
      "grad_norm": 1.6145741939544678,
      "learning_rate": 5.8648030494710196e-05,
      "loss": 0.3382,
      "step": 2502
    },
    {
      "epoch": 0.5004,
      "loss_align": 0.3308592438697815,
      "loss_contrastive": 0.12862330675125122,
      "loss_gen": 0.23026348650455475,
      "neg_sim_mean": 0.6977640390396118,
      "pos_sim_mean": 0.6691407561302185,
      "step": 2502
    },
    {
      "epoch": 0.5006,
      "grad_norm": 2.0708701610565186,
      "learning_rate": 5.8613647891128845e-05,
      "loss": 0.5766,
      "step": 2503
    },
    {
      "epoch": 0.5006,
      "loss_align": 0.1918807029724121,
      "loss_contrastive": 0.14336156845092773,
      "loss_gen": 0.18210391700267792,
      "neg_sim_mean": 0.8514808416366577,
      "pos_sim_mean": 0.8081192970275879,
      "step": 2503
    },
    {
      "epoch": 0.5008,
      "grad_norm": 1.7846875190734863,
      "learning_rate": 5.857926108936015e-05,
      "loss": 0.3912,
      "step": 2504
    },
    {
      "epoch": 0.5008,
      "loss_align": 0.3586409091949463,
      "loss_contrastive": 0.1034708023071289,
      "loss_gen": 0.3780929744243622,
      "neg_sim_mean": 0.6448298692703247,
      "pos_sim_mean": 0.6413590908050537,
      "step": 2504
    },
    {
      "epoch": 0.501,
      "grad_norm": 2.670975923538208,
      "learning_rate": 5.8544870106163844e-05,
      "loss": 0.7492,
      "step": 2505
    },
    {
      "epoch": 0.501,
      "loss_align": 0.1337013840675354,
      "loss_contrastive": 0.07204991579055786,
      "loss_gen": 0.3948737680912018,
      "neg_sim_mean": 0.8383485078811646,
      "pos_sim_mean": 0.8662986159324646,
      "step": 2505
    },
    {
      "epoch": 0.5012,
      "grad_norm": 2.0908310413360596,
      "learning_rate": 5.851047495830163e-05,
      "loss": 0.5372,
      "step": 2506
    },
    {
      "epoch": 0.5012,
      "loss_align": 0.0927247405052185,
      "loss_contrastive": 0.027511537075042725,
      "loss_gen": 0.39363399147987366,
      "neg_sim_mean": 0.8347867727279663,
      "pos_sim_mean": 0.9072752594947815,
      "step": 2506
    },
    {
      "epoch": 0.5014,
      "grad_norm": 2.2385952472686768,
      "learning_rate": 5.847607566253732e-05,
      "loss": 0.4897,
      "step": 2507
    },
    {
      "epoch": 0.5014,
      "loss_align": 0.125038743019104,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00332306744530797,
      "neg_sim_mean": 0.5617380142211914,
      "pos_sim_mean": 0.874961256980896,
      "step": 2507
    },
    {
      "epoch": 0.5016,
      "grad_norm": 0.5547266602516174,
      "learning_rate": 5.844167223563669e-05,
      "loss": 0.1284,
      "step": 2508
    },
    {
      "epoch": 0.5016,
      "loss_align": 0.09608173370361328,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19362236559391022,
      "neg_sim_mean": 0.7031738758087158,
      "pos_sim_mean": 0.9039182662963867,
      "step": 2508
    },
    {
      "epoch": 0.5018,
      "grad_norm": 1.706099510192871,
      "learning_rate": 5.840726469436758e-05,
      "loss": 0.2897,
      "step": 2509
    },
    {
      "epoch": 0.5018,
      "loss_align": 0.10321354866027832,
      "loss_contrastive": 0.0882577896118164,
      "loss_gen": 0.25015854835510254,
      "neg_sim_mean": 0.8850442171096802,
      "pos_sim_mean": 0.8967864513397217,
      "step": 2509
    },
    {
      "epoch": 0.502,
      "grad_norm": 1.7818994522094727,
      "learning_rate": 5.837285305549978e-05,
      "loss": 0.364,
      "step": 2510
    },
    {
      "epoch": 0.502,
      "loss_align": 0.12085521221160889,
      "loss_contrastive": 0.014141619205474854,
      "loss_gen": 0.26670244336128235,
      "neg_sim_mean": 0.7932863831520081,
      "pos_sim_mean": 0.8791447877883911,
      "step": 2510
    },
    {
      "epoch": 0.5022,
      "grad_norm": 2.332022190093994,
      "learning_rate": 5.833843733580512e-05,
      "loss": 0.3893,
      "step": 2511
    },
    {
      "epoch": 0.5022,
      "loss_align": 0.09530293941497803,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004693892784416676,
      "neg_sim_mean": 0.37501561641693115,
      "pos_sim_mean": 0.904697060585022,
      "step": 2511
    },
    {
      "epoch": 0.5024,
      "grad_norm": 0.4173548221588135,
      "learning_rate": 5.83040175520574e-05,
      "loss": 0.1,
      "step": 2512
    },
    {
      "epoch": 0.5024,
      "loss_align": 0.08556520938873291,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16177429258823395,
      "neg_sim_mean": 0.7908028364181519,
      "pos_sim_mean": 0.9144347906112671,
      "step": 2512
    },
    {
      "epoch": 0.5026,
      "grad_norm": 1.2594444751739502,
      "learning_rate": 5.826959372103239e-05,
      "loss": 0.2473,
      "step": 2513
    },
    {
      "epoch": 0.5026,
      "loss_align": 0.1750887632369995,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007110774051398039,
      "neg_sim_mean": 0.4017478823661804,
      "pos_sim_mean": 0.8249112367630005,
      "step": 2513
    },
    {
      "epoch": 0.5028,
      "grad_norm": 0.6115837693214417,
      "learning_rate": 5.8235165859507864e-05,
      "loss": 0.1822,
      "step": 2514
    },
    {
      "epoch": 0.5028,
      "loss_align": 0.20134586095809937,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25343793630599976,
      "neg_sim_mean": 0.6890106201171875,
      "pos_sim_mean": 0.7986541390419006,
      "step": 2514
    },
    {
      "epoch": 0.503,
      "grad_norm": 2.3127236366271973,
      "learning_rate": 5.820073398426356e-05,
      "loss": 0.4548,
      "step": 2515
    },
    {
      "epoch": 0.503,
      "loss_align": 0.33111119270324707,
      "loss_contrastive": 0.12327432632446289,
      "loss_gen": 0.41146615147590637,
      "neg_sim_mean": 0.6921631097793579,
      "pos_sim_mean": 0.6688888072967529,
      "step": 2515
    },
    {
      "epoch": 0.5032,
      "grad_norm": 2.6150894165039062,
      "learning_rate": 5.816629811208112e-05,
      "loss": 0.7574,
      "step": 2516
    },
    {
      "epoch": 0.5032,
      "loss_align": 0.22123920917510986,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09455814212560654,
      "neg_sim_mean": 0.43196237087249756,
      "pos_sim_mean": 0.7787607908248901,
      "step": 2516
    },
    {
      "epoch": 0.5034,
      "grad_norm": 1.2782034873962402,
      "learning_rate": 5.813185825974419e-05,
      "loss": 0.3158,
      "step": 2517
    },
    {
      "epoch": 0.5034,
      "loss_align": 0.20931458473205566,
      "loss_contrastive": 0.05470395088195801,
      "loss_gen": 0.09548963606357574,
      "neg_sim_mean": 0.7453893423080444,
      "pos_sim_mean": 0.7906854152679443,
      "step": 2517
    },
    {
      "epoch": 0.5036,
      "grad_norm": 1.3575667142868042,
      "learning_rate": 5.809741444403831e-05,
      "loss": 0.3114,
      "step": 2518
    },
    {
      "epoch": 0.5036,
      "loss_align": 0.14567869901657104,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21043726801872253,
      "neg_sim_mean": 0.5223484039306641,
      "pos_sim_mean": 0.854321300983429,
      "step": 2518
    },
    {
      "epoch": 0.5038,
      "grad_norm": 1.8847979307174683,
      "learning_rate": 5.806296668175104e-05,
      "loss": 0.3561,
      "step": 2519
    },
    {
      "epoch": 0.5038,
      "loss_align": 0.14501750469207764,
      "loss_contrastive": 0.0,
      "loss_gen": 0.41892319917678833,
      "neg_sim_mean": 0.7456141710281372,
      "pos_sim_mean": 0.8549824953079224,
      "step": 2519
    },
    {
      "epoch": 0.504,
      "grad_norm": 2.1853137016296387,
      "learning_rate": 5.8028514989671724e-05,
      "loss": 0.5639,
      "step": 2520
    },
    {
      "epoch": 0.504,
      "loss_align": 0.1829315423965454,
      "loss_contrastive": 0.14547717571258545,
      "loss_gen": 0.24860014021396637,
      "neg_sim_mean": 0.8625456094741821,
      "pos_sim_mean": 0.8170684576034546,
      "step": 2520
    },
    {
      "epoch": 0.5042,
      "grad_norm": 1.6595934629440308,
      "learning_rate": 5.799405938459175e-05,
      "loss": 0.449,
      "step": 2521
    },
    {
      "epoch": 0.5042,
      "loss_align": 0.1251300573348999,
      "loss_contrastive": 0.04266852140426636,
      "loss_gen": 0.19506052136421204,
      "neg_sim_mean": 0.8175384402275085,
      "pos_sim_mean": 0.8748699426651001,
      "step": 2521
    },
    {
      "epoch": 0.5044,
      "grad_norm": 1.5062576532363892,
      "learning_rate": 5.795959988330434e-05,
      "loss": 0.3253,
      "step": 2522
    },
    {
      "epoch": 0.5044,
      "loss_align": 0.3298414349555969,
      "loss_contrastive": 0.04811280965805054,
      "loss_gen": 0.14476685225963593,
      "neg_sim_mean": 0.6182713508605957,
      "pos_sim_mean": 0.6701585650444031,
      "step": 2522
    },
    {
      "epoch": 0.5046,
      "grad_norm": 2.317875862121582,
      "learning_rate": 5.792513650260465e-05,
      "loss": 0.4804,
      "step": 2523
    },
    {
      "epoch": 0.5046,
      "loss_align": 0.15092098712921143,
      "loss_contrastive": 0.05272054672241211,
      "loss_gen": 0.20072491466999054,
      "neg_sim_mean": 0.8017995357513428,
      "pos_sim_mean": 0.8490790128707886,
      "step": 2523
    },
    {
      "epoch": 0.5048,
      "grad_norm": 1.5672389268875122,
      "learning_rate": 5.78906692592897e-05,
      "loss": 0.358,
      "step": 2524
    },
    {
      "epoch": 0.5048,
      "loss_align": 0.20814096927642822,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21714110672473907,
      "neg_sim_mean": 0.6563109159469604,
      "pos_sim_mean": 0.7918590307235718,
      "step": 2524
    },
    {
      "epoch": 0.505,
      "grad_norm": 1.5941152572631836,
      "learning_rate": 5.7856198170158394e-05,
      "loss": 0.4253,
      "step": 2525
    },
    {
      "epoch": 0.505,
      "loss_align": 0.1324414610862732,
      "loss_contrastive": 0.0823015570640564,
      "loss_gen": 0.16520951688289642,
      "neg_sim_mean": 0.8498600721359253,
      "pos_sim_mean": 0.8675585389137268,
      "step": 2525
    },
    {
      "epoch": 0.5052,
      "grad_norm": 1.3437724113464355,
      "learning_rate": 5.782172325201155e-05,
      "loss": 0.3075,
      "step": 2526
    },
    {
      "epoch": 0.5052,
      "loss_align": 0.08862566947937012,
      "loss_contrastive": 0.07549047470092773,
      "loss_gen": 0.28084367513656616,
      "neg_sim_mean": 0.8868647813796997,
      "pos_sim_mean": 0.9113743305206299,
      "step": 2526
    },
    {
      "epoch": 0.5054,
      "grad_norm": 1.6913551092147827,
      "learning_rate": 5.778724452165181e-05,
      "loss": 0.3785,
      "step": 2527
    },
    {
      "epoch": 0.5054,
      "loss_align": 0.1493627429008484,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1944943219423294,
      "neg_sim_mean": 0.7239381670951843,
      "pos_sim_mean": 0.8506372570991516,
      "step": 2527
    },
    {
      "epoch": 0.5056,
      "grad_norm": 1.6009973287582397,
      "learning_rate": 5.7752761995883694e-05,
      "loss": 0.3439,
      "step": 2528
    },
    {
      "epoch": 0.5056,
      "loss_align": 0.18366920948028564,
      "loss_contrastive": 0.09323441982269287,
      "loss_gen": 0.26835840940475464,
      "neg_sim_mean": 0.8095651865005493,
      "pos_sim_mean": 0.8163307905197144,
      "step": 2528
    },
    {
      "epoch": 0.5058,
      "grad_norm": 1.9791443347930908,
      "learning_rate": 5.771827569151357e-05,
      "loss": 0.4632,
      "step": 2529
    },
    {
      "epoch": 0.5058,
      "loss_align": 0.14102190732955933,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17109806835651398,
      "neg_sim_mean": 0.7275177240371704,
      "pos_sim_mean": 0.8589780926704407,
      "step": 2529
    },
    {
      "epoch": 0.506,
      "grad_norm": 1.699363112449646,
      "learning_rate": 5.7683785625349616e-05,
      "loss": 0.3121,
      "step": 2530
    },
    {
      "epoch": 0.506,
      "loss_align": 0.1880812644958496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26506856083869934,
      "neg_sim_mean": 0.6969848275184631,
      "pos_sim_mean": 0.8119187355041504,
      "step": 2530
    },
    {
      "epoch": 0.5062,
      "grad_norm": 2.090172290802002,
      "learning_rate": 5.764929181420191e-05,
      "loss": 0.4531,
      "step": 2531
    },
    {
      "epoch": 0.5062,
      "loss_align": 0.19348007440567017,
      "loss_contrastive": 0.08430153131484985,
      "loss_gen": 0.11907421797513962,
      "neg_sim_mean": 0.7908214330673218,
      "pos_sim_mean": 0.8065199255943298,
      "step": 2531
    },
    {
      "epoch": 0.5064,
      "grad_norm": 1.6081441640853882,
      "learning_rate": 5.761479427488229e-05,
      "loss": 0.3227,
      "step": 2532
    },
    {
      "epoch": 0.5064,
      "loss_align": 0.19650661945343018,
      "loss_contrastive": 0.06966161727905273,
      "loss_gen": 0.19185148179531097,
      "neg_sim_mean": 0.7731549739837646,
      "pos_sim_mean": 0.8034933805465698,
      "step": 2532
    },
    {
      "epoch": 0.5066,
      "grad_norm": 1.6290565729141235,
      "learning_rate": 5.758029302420446e-05,
      "loss": 0.3967,
      "step": 2533
    },
    {
      "epoch": 0.5066,
      "loss_align": 0.14784657955169678,
      "loss_contrastive": 0.037973761558532715,
      "loss_gen": 0.321518212556839,
      "neg_sim_mean": 0.790127158164978,
      "pos_sim_mean": 0.8521534204483032,
      "step": 2533
    },
    {
      "epoch": 0.5068,
      "grad_norm": 2.0314667224884033,
      "learning_rate": 5.754578807898388e-05,
      "loss": 0.4739,
      "step": 2534
    },
    {
      "epoch": 0.5068,
      "loss_align": 0.19030332565307617,
      "loss_contrastive": 0.052276015281677246,
      "loss_gen": 0.09445372223854065,
      "neg_sim_mean": 0.7619726657867432,
      "pos_sim_mean": 0.8096966743469238,
      "step": 2534
    },
    {
      "epoch": 0.507,
      "grad_norm": 1.0957820415496826,
      "learning_rate": 5.751127945603786e-05,
      "loss": 0.291,
      "step": 2535
    },
    {
      "epoch": 0.507,
      "loss_align": 0.15610671043395996,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2090223729610443,
      "neg_sim_mean": 0.7322764992713928,
      "pos_sim_mean": 0.84389328956604,
      "step": 2535
    },
    {
      "epoch": 0.5072,
      "grad_norm": 1.8830047845840454,
      "learning_rate": 5.747676717218549e-05,
      "loss": 0.3651,
      "step": 2536
    },
    {
      "epoch": 0.5072,
      "loss_align": 0.3178459405899048,
      "loss_contrastive": 0.14618372917175293,
      "loss_gen": 0.36630779504776,
      "neg_sim_mean": 0.7283377647399902,
      "pos_sim_mean": 0.6821540594100952,
      "step": 2536
    },
    {
      "epoch": 0.5074,
      "grad_norm": 2.4406630992889404,
      "learning_rate": 5.744225124424761e-05,
      "loss": 0.7017,
      "step": 2537
    },
    {
      "epoch": 0.5074,
      "loss_align": 0.1484823226928711,
      "loss_contrastive": 0.003119945526123047,
      "loss_gen": 0.2788923978805542,
      "neg_sim_mean": 0.754637598991394,
      "pos_sim_mean": 0.8515176773071289,
      "step": 2537
    },
    {
      "epoch": 0.5076,
      "grad_norm": 1.7930127382278442,
      "learning_rate": 5.7407731689046904e-05,
      "loss": 0.4277,
      "step": 2538
    },
    {
      "epoch": 0.5076,
      "loss_align": 0.18108820915222168,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29555752873420715,
      "neg_sim_mean": 0.6485950946807861,
      "pos_sim_mean": 0.8189117908477783,
      "step": 2538
    },
    {
      "epoch": 0.5078,
      "grad_norm": 1.9653894901275635,
      "learning_rate": 5.737320852340775e-05,
      "loss": 0.4766,
      "step": 2539
    },
    {
      "epoch": 0.5078,
      "loss_align": 0.11080825328826904,
      "loss_contrastive": 0.008694291114807129,
      "loss_gen": 0.25930824875831604,
      "neg_sim_mean": 0.7978860139846802,
      "pos_sim_mean": 0.889191746711731,
      "step": 2539
    },
    {
      "epoch": 0.508,
      "grad_norm": 1.7533302307128906,
      "learning_rate": 5.733868176415633e-05,
      "loss": 0.3712,
      "step": 2540
    },
    {
      "epoch": 0.508,
      "loss_align": 0.1385713815689087,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17508916556835175,
      "neg_sim_mean": 0.6178305149078369,
      "pos_sim_mean": 0.8614286184310913,
      "step": 2540
    },
    {
      "epoch": 0.5082,
      "grad_norm": 1.2991654872894287,
      "learning_rate": 5.730415142812059e-05,
      "loss": 0.3137,
      "step": 2541
    },
    {
      "epoch": 0.5082,
      "loss_align": 0.1388789415359497,
      "loss_contrastive": 0.006661653518676758,
      "loss_gen": 0.20762769877910614,
      "neg_sim_mean": 0.7677826881408691,
      "pos_sim_mean": 0.8611210584640503,
      "step": 2541
    },
    {
      "epoch": 0.5084,
      "grad_norm": 2.182053565979004,
      "learning_rate": 5.726961753213016e-05,
      "loss": 0.3473,
      "step": 2542
    },
    {
      "epoch": 0.5084,
      "loss_align": 0.21647489070892334,
      "loss_contrastive": 0.08789068460464478,
      "loss_gen": 0.12658949196338654,
      "neg_sim_mean": 0.7714157700538635,
      "pos_sim_mean": 0.7835251092910767,
      "step": 2542
    },
    {
      "epoch": 0.5086,
      "grad_norm": 1.3396700620651245,
      "learning_rate": 5.723508009301646e-05,
      "loss": 0.3536,
      "step": 2543
    },
    {
      "epoch": 0.5086,
      "loss_align": 0.13345956802368164,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03014105185866356,
      "neg_sim_mean": 0.4091402292251587,
      "pos_sim_mean": 0.8665404319763184,
      "step": 2543
    },
    {
      "epoch": 0.5088,
      "grad_norm": 1.0420262813568115,
      "learning_rate": 5.7200539127612604e-05,
      "loss": 0.1636,
      "step": 2544
    },
    {
      "epoch": 0.5088,
      "loss_align": 0.30094897747039795,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04380595311522484,
      "neg_sim_mean": 0.3055760860443115,
      "pos_sim_mean": 0.699051022529602,
      "step": 2544
    },
    {
      "epoch": 0.509,
      "grad_norm": 1.3174805641174316,
      "learning_rate": 5.716599465275348e-05,
      "loss": 0.3448,
      "step": 2545
    },
    {
      "epoch": 0.509,
      "loss_align": 0.12468612194061279,
      "loss_contrastive": 0.07145959138870239,
      "loss_gen": 0.11913775652647018,
      "neg_sim_mean": 0.8467734456062317,
      "pos_sim_mean": 0.8753138780593872,
      "step": 2545
    },
    {
      "epoch": 0.5092,
      "grad_norm": 1.3116956949234009,
      "learning_rate": 5.713144668527559e-05,
      "loss": 0.2524,
      "step": 2546
    },
    {
      "epoch": 0.5092,
      "loss_align": 0.06800425052642822,
      "loss_contrastive": 0.023617148399353027,
      "loss_gen": 0.24619348347187042,
      "neg_sim_mean": 0.8556128740310669,
      "pos_sim_mean": 0.9319957494735718,
      "step": 2546
    },
    {
      "epoch": 0.5094,
      "grad_norm": 1.8178584575653076,
      "learning_rate": 5.709689524201722e-05,
      "loss": 0.317,
      "step": 2547
    },
    {
      "epoch": 0.5094,
      "loss_align": 0.11919665336608887,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2188919335603714,
      "neg_sim_mean": 0.6875827312469482,
      "pos_sim_mean": 0.8808033466339111,
      "step": 2547
    },
    {
      "epoch": 0.5096,
      "grad_norm": 1.708235263824463,
      "learning_rate": 5.706234033981834e-05,
      "loss": 0.3381,
      "step": 2548
    },
    {
      "epoch": 0.5096,
      "loss_align": 0.2408522367477417,
      "loss_contrastive": 0.07344919443130493,
      "loss_gen": 0.2304888814687729,
      "neg_sim_mean": 0.7325969338417053,
      "pos_sim_mean": 0.7591477632522583,
      "step": 2548
    },
    {
      "epoch": 0.5098,
      "grad_norm": 1.9554437398910522,
      "learning_rate": 5.7027781995520545e-05,
      "loss": 0.4802,
      "step": 2549
    },
    {
      "epoch": 0.5098,
      "loss_align": 0.10677039623260498,
      "loss_contrastive": 0.09814131259918213,
      "loss_gen": 0.13110622763633728,
      "neg_sim_mean": 0.8913708925247192,
      "pos_sim_mean": 0.893229603767395,
      "step": 2549
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.3752131462097168,
      "learning_rate": 5.699322022596722e-05,
      "loss": 0.2497,
      "step": 2550
    },
    {
      "epoch": 0.51,
      "loss_align": 0.21763432025909424,
      "loss_contrastive": 0.09873014688491821,
      "loss_gen": 0.13373103737831116,
      "neg_sim_mean": 0.7810958027839661,
      "pos_sim_mean": 0.7823656797409058,
      "step": 2550
    },
    {
      "epoch": 0.5102,
      "grad_norm": 1.480638027191162,
      "learning_rate": 5.695865504800327e-05,
      "loss": 0.3632,
      "step": 2551
    },
    {
      "epoch": 0.5102,
      "loss_align": 0.1574852466583252,
      "loss_contrastive": 0.12381184101104736,
      "loss_gen": 0.21003656089305878,
      "neg_sim_mean": 0.8663265705108643,
      "pos_sim_mean": 0.8425147533416748,
      "step": 2551
    },
    {
      "epoch": 0.5104,
      "grad_norm": 1.582563877105713,
      "learning_rate": 5.6924086478475415e-05,
      "loss": 0.3824,
      "step": 2552
    },
    {
      "epoch": 0.5104,
      "loss_align": 0.17042434215545654,
      "loss_contrastive": 0.04975998401641846,
      "loss_gen": 0.28104114532470703,
      "neg_sim_mean": 0.779335618019104,
      "pos_sim_mean": 0.8295756578445435,
      "step": 2552
    },
    {
      "epoch": 0.5106,
      "grad_norm": 2.095510959625244,
      "learning_rate": 5.68895145342319e-05,
      "loss": 0.4574,
      "step": 2553
    },
    {
      "epoch": 0.5106,
      "loss_align": 0.24041813611984253,
      "loss_contrastive": 0.10193896293640137,
      "loss_gen": 0.3129440248012543,
      "neg_sim_mean": 0.7615208029747009,
      "pos_sim_mean": 0.7595818638801575,
      "step": 2553
    },
    {
      "epoch": 0.5108,
      "grad_norm": 2.3175978660583496,
      "learning_rate": 5.685493923212273e-05,
      "loss": 0.5656,
      "step": 2554
    },
    {
      "epoch": 0.5108,
      "loss_align": 0.1365756392478943,
      "loss_contrastive": 0.0,
      "loss_gen": 0.41753140091896057,
      "neg_sim_mean": 0.6634721755981445,
      "pos_sim_mean": 0.8634243607521057,
      "step": 2554
    },
    {
      "epoch": 0.511,
      "grad_norm": 1.7929447889328003,
      "learning_rate": 5.682036058899942e-05,
      "loss": 0.5541,
      "step": 2555
    },
    {
      "epoch": 0.511,
      "loss_align": 0.1511911153793335,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2563576400279999,
      "neg_sim_mean": 0.658307671546936,
      "pos_sim_mean": 0.8488088846206665,
      "step": 2555
    },
    {
      "epoch": 0.5112,
      "grad_norm": 2.3158626556396484,
      "learning_rate": 5.6785778621715225e-05,
      "loss": 0.4075,
      "step": 2556
    },
    {
      "epoch": 0.5112,
      "loss_align": 0.07805287837982178,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012934322468936443,
      "neg_sim_mean": 0.5121415853500366,
      "pos_sim_mean": 0.9219471216201782,
      "step": 2556
    },
    {
      "epoch": 0.5114,
      "grad_norm": 0.7470036149024963,
      "learning_rate": 5.6751193347124965e-05,
      "loss": 0.091,
      "step": 2557
    },
    {
      "epoch": 0.5114,
      "loss_align": 0.15322881937026978,
      "loss_contrastive": 0.08918088674545288,
      "loss_gen": 0.1941179782152176,
      "neg_sim_mean": 0.8359520435333252,
      "pos_sim_mean": 0.8467711806297302,
      "step": 2557
    },
    {
      "epoch": 0.5116,
      "grad_norm": 1.8929846286773682,
      "learning_rate": 5.671660478208507e-05,
      "loss": 0.358,
      "step": 2558
    },
    {
      "epoch": 0.5116,
      "loss_align": 0.1525578498840332,
      "loss_contrastive": 0.07609844207763672,
      "loss_gen": 0.3258845806121826,
      "neg_sim_mean": 0.8235405683517456,
      "pos_sim_mean": 0.8474421501159668,
      "step": 2558
    },
    {
      "epoch": 0.5118,
      "grad_norm": 1.989766001701355,
      "learning_rate": 5.668201294345363e-05,
      "loss": 0.4876,
      "step": 2559
    },
    {
      "epoch": 0.5118,
      "loss_align": 0.23592698574066162,
      "loss_contrastive": 0.03190577030181885,
      "loss_gen": 0.15085166692733765,
      "neg_sim_mean": 0.6959787607192993,
      "pos_sim_mean": 0.7640730142593384,
      "step": 2559
    },
    {
      "epoch": 0.512,
      "grad_norm": 1.727888584136963,
      "learning_rate": 5.664741784809022e-05,
      "loss": 0.3906,
      "step": 2560
    },
    {
      "epoch": 0.512,
      "loss_align": 0.20474910736083984,
      "loss_contrastive": 0.12426948547363281,
      "loss_gen": 0.4289080500602722,
      "neg_sim_mean": 0.8195203542709351,
      "pos_sim_mean": 0.7952508926391602,
      "step": 2560
    },
    {
      "epoch": 0.5122,
      "grad_norm": 2.174589157104492,
      "learning_rate": 5.661281951285613e-05,
      "loss": 0.6486,
      "step": 2561
    },
    {
      "epoch": 0.5122,
      "loss_align": 0.08208280801773071,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2452501356601715,
      "neg_sim_mean": 0.8013927340507507,
      "pos_sim_mean": 0.9179171919822693,
      "step": 2561
    },
    {
      "epoch": 0.5124,
      "grad_norm": 1.6913095712661743,
      "learning_rate": 5.6578217954614134e-05,
      "loss": 0.3273,
      "step": 2562
    },
    {
      "epoch": 0.5124,
      "loss_align": 0.16762274503707886,
      "loss_contrastive": 0.11414486169815063,
      "loss_gen": 0.2836439609527588,
      "neg_sim_mean": 0.8465220928192139,
      "pos_sim_mean": 0.8323772549629211,
      "step": 2562
    },
    {
      "epoch": 0.5126,
      "grad_norm": 1.7373884916305542,
      "learning_rate": 5.654361319022862e-05,
      "loss": 0.465,
      "step": 2563
    },
    {
      "epoch": 0.5126,
      "loss_align": 0.10644030570983887,
      "loss_contrastive": 0.0,
      "loss_gen": 0.275846928358078,
      "neg_sim_mean": 0.7219070792198181,
      "pos_sim_mean": 0.8935596942901611,
      "step": 2563
    },
    {
      "epoch": 0.5128,
      "grad_norm": 1.7705918550491333,
      "learning_rate": 5.6509005236565525e-05,
      "loss": 0.3823,
      "step": 2564
    },
    {
      "epoch": 0.5128,
      "loss_align": 0.137853741645813,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009573007933795452,
      "neg_sim_mean": 0.3859879672527313,
      "pos_sim_mean": 0.862146258354187,
      "step": 2564
    },
    {
      "epoch": 0.513,
      "grad_norm": 0.6496090888977051,
      "learning_rate": 5.6474394110492344e-05,
      "loss": 0.1474,
      "step": 2565
    },
    {
      "epoch": 0.513,
      "loss_align": 0.07610279321670532,
      "loss_contrastive": 0.0,
      "loss_gen": 0.020092535763978958,
      "neg_sim_mean": 0.2923898696899414,
      "pos_sim_mean": 0.9238972067832947,
      "step": 2565
    },
    {
      "epoch": 0.5132,
      "grad_norm": 0.9110246896743774,
      "learning_rate": 5.643977982887815e-05,
      "loss": 0.0962,
      "step": 2566
    },
    {
      "epoch": 0.5132,
      "loss_align": 0.107200026512146,
      "loss_contrastive": 0.06070506572723389,
      "loss_gen": 0.2557602524757385,
      "neg_sim_mean": 0.85350501537323,
      "pos_sim_mean": 0.892799973487854,
      "step": 2566
    },
    {
      "epoch": 0.5134,
      "grad_norm": 1.624068260192871,
      "learning_rate": 5.6405162408593484e-05,
      "loss": 0.3702,
      "step": 2567
    },
    {
      "epoch": 0.5134,
      "loss_align": 0.18187707662582397,
      "loss_contrastive": 0.11274939775466919,
      "loss_gen": 0.3241450786590576,
      "neg_sim_mean": 0.8308722972869873,
      "pos_sim_mean": 0.818122923374176,
      "step": 2567
    },
    {
      "epoch": 0.5136,
      "grad_norm": 2.0087599754333496,
      "learning_rate": 5.6370541866510474e-05,
      "loss": 0.5196,
      "step": 2568
    },
    {
      "epoch": 0.5136,
      "loss_align": 0.22084146738052368,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02896200306713581,
      "neg_sim_mean": 0.5362613201141357,
      "pos_sim_mean": 0.7791585326194763,
      "step": 2568
    },
    {
      "epoch": 0.5138,
      "grad_norm": 1.601431965827942,
      "learning_rate": 5.633591821950274e-05,
      "loss": 0.2498,
      "step": 2569
    },
    {
      "epoch": 0.5138,
      "loss_align": 0.33311641216278076,
      "loss_contrastive": 0.03498953580856323,
      "loss_gen": 0.5488748550415039,
      "neg_sim_mean": 0.6018730998039246,
      "pos_sim_mean": 0.6668835878372192,
      "step": 2569
    },
    {
      "epoch": 0.514,
      "grad_norm": 3.024773597717285,
      "learning_rate": 5.630129148444543e-05,
      "loss": 0.8862,
      "step": 2570
    },
    {
      "epoch": 0.514,
      "loss_align": 0.12238490581512451,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005821104161441326,
      "neg_sim_mean": 0.3412155508995056,
      "pos_sim_mean": 0.8776150941848755,
      "step": 2570
    },
    {
      "epoch": 0.5142,
      "grad_norm": 0.5116593837738037,
      "learning_rate": 5.6266661678215216e-05,
      "loss": 0.1282,
      "step": 2571
    },
    {
      "epoch": 0.5142,
      "loss_align": 0.19822454452514648,
      "loss_contrastive": 0.053620100021362305,
      "loss_gen": 0.1716219037771225,
      "neg_sim_mean": 0.7553955316543579,
      "pos_sim_mean": 0.8017754554748535,
      "step": 2571
    },
    {
      "epoch": 0.5144,
      "grad_norm": 1.8048075437545776,
      "learning_rate": 5.623202881769023e-05,
      "loss": 0.3763,
      "step": 2572
    },
    {
      "epoch": 0.5144,
      "loss_align": 0.14411181211471558,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008230863139033318,
      "neg_sim_mean": 0.3631799817085266,
      "pos_sim_mean": 0.8558881878852844,
      "step": 2572
    },
    {
      "epoch": 0.5146,
      "grad_norm": 0.6978446841239929,
      "learning_rate": 5.619739291975009e-05,
      "loss": 0.1523,
      "step": 2573
    },
    {
      "epoch": 0.5146,
      "loss_align": 0.17497318983078003,
      "loss_contrastive": 0.11528009176254272,
      "loss_gen": 0.23268011212348938,
      "neg_sim_mean": 0.8403068780899048,
      "pos_sim_mean": 0.82502681016922,
      "step": 2573
    },
    {
      "epoch": 0.5148,
      "grad_norm": 2.096641778945923,
      "learning_rate": 5.616275400127594e-05,
      "loss": 0.4215,
      "step": 2574
    },
    {
      "epoch": 0.5148,
      "loss_align": 0.20106840133666992,
      "loss_contrastive": 0.0,
      "loss_gen": 0.31742584705352783,
      "neg_sim_mean": 0.5442924499511719,
      "pos_sim_mean": 0.7989315986633301,
      "step": 2574
    },
    {
      "epoch": 0.515,
      "grad_norm": 2.177746534347534,
      "learning_rate": 5.612811207915034e-05,
      "loss": 0.5185,
      "step": 2575
    },
    {
      "epoch": 0.515,
      "loss_align": 0.38097864389419556,
      "loss_contrastive": 0.146725594997406,
      "loss_gen": 0.2171180099248886,
      "neg_sim_mean": 0.6657469272613525,
      "pos_sim_mean": 0.6190213561058044,
      "step": 2575
    },
    {
      "epoch": 0.5152,
      "grad_norm": 2.0092263221740723,
      "learning_rate": 5.6093467170257374e-05,
      "loss": 0.6157,
      "step": 2576
    },
    {
      "epoch": 0.5152,
      "loss_align": 0.1291877031326294,
      "loss_contrastive": 0.00632089376449585,
      "loss_gen": 0.3644924759864807,
      "neg_sim_mean": 0.7771331667900085,
      "pos_sim_mean": 0.8708122968673706,
      "step": 2576
    },
    {
      "epoch": 0.5154,
      "grad_norm": 2.3969409465789795,
      "learning_rate": 5.6058819291482534e-05,
      "loss": 0.4944,
      "step": 2577
    },
    {
      "epoch": 0.5154,
      "loss_align": 0.04863846302032471,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005091287195682526,
      "neg_sim_mean": 0.29540929198265076,
      "pos_sim_mean": 0.9513615369796753,
      "step": 2577
    },
    {
      "epoch": 0.5156,
      "grad_norm": 0.36565908789634705,
      "learning_rate": 5.6024168459712764e-05,
      "loss": 0.0537,
      "step": 2578
    },
    {
      "epoch": 0.5156,
      "loss_align": 0.2012476921081543,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4757422208786011,
      "neg_sim_mean": 0.4009852409362793,
      "pos_sim_mean": 0.7987523078918457,
      "step": 2578
    },
    {
      "epoch": 0.5158,
      "grad_norm": 2.664494276046753,
      "learning_rate": 5.598951469183649e-05,
      "loss": 0.677,
      "step": 2579
    },
    {
      "epoch": 0.5158,
      "loss_align": 0.12998366355895996,
      "loss_contrastive": 0.023714661598205566,
      "loss_gen": 0.24152685701847076,
      "neg_sim_mean": 0.7937309741973877,
      "pos_sim_mean": 0.87001633644104,
      "step": 2579
    },
    {
      "epoch": 0.516,
      "grad_norm": 1.8427115678787231,
      "learning_rate": 5.595485800474349e-05,
      "loss": 0.3744,
      "step": 2580
    },
    {
      "epoch": 0.516,
      "loss_align": 0.25192368030548096,
      "loss_contrastive": 0.1330258846282959,
      "loss_gen": 0.2929757833480835,
      "neg_sim_mean": 0.781102180480957,
      "pos_sim_mean": 0.748076319694519,
      "step": 2580
    },
    {
      "epoch": 0.5162,
      "grad_norm": 1.9402403831481934,
      "learning_rate": 5.5920198415325064e-05,
      "loss": 0.5609,
      "step": 2581
    },
    {
      "epoch": 0.5162,
      "loss_align": 0.15499961376190186,
      "loss_contrastive": 0.10414361953735352,
      "loss_gen": 0.10966435819864273,
      "neg_sim_mean": 0.8491439819335938,
      "pos_sim_mean": 0.8450003862380981,
      "step": 2581
    },
    {
      "epoch": 0.5164,
      "grad_norm": 1.5968416929244995,
      "learning_rate": 5.588553594047382e-05,
      "loss": 0.2772,
      "step": 2582
    },
    {
      "epoch": 0.5164,
      "loss_align": 0.21354937553405762,
      "loss_contrastive": 0.06832373142242432,
      "loss_gen": 0.3737206757068634,
      "neg_sim_mean": 0.7547743320465088,
      "pos_sim_mean": 0.7864506244659424,
      "step": 2582
    },
    {
      "epoch": 0.5166,
      "grad_norm": 2.5718746185302734,
      "learning_rate": 5.585087059708388e-05,
      "loss": 0.5955,
      "step": 2583
    },
    {
      "epoch": 0.5166,
      "loss_align": 0.3528110980987549,
      "loss_contrastive": 0.2242712378501892,
      "loss_gen": 0.2207673192024231,
      "neg_sim_mean": 0.7714601159095764,
      "pos_sim_mean": 0.6471889019012451,
      "step": 2583
    },
    {
      "epoch": 0.5168,
      "grad_norm": 2.315049409866333,
      "learning_rate": 5.5816202402050675e-05,
      "loss": 0.6005,
      "step": 2584
    },
    {
      "epoch": 0.5168,
      "loss_align": 0.055565834045410156,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2614533007144928,
      "neg_sim_mean": 0.5389714241027832,
      "pos_sim_mean": 0.9444341659545898,
      "step": 2584
    },
    {
      "epoch": 0.517,
      "grad_norm": 1.9238858222961426,
      "learning_rate": 5.578153137227109e-05,
      "loss": 0.317,
      "step": 2585
    },
    {
      "epoch": 0.517,
      "loss_align": 0.08588540554046631,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2840796113014221,
      "neg_sim_mean": 0.7203062772750854,
      "pos_sim_mean": 0.9141145944595337,
      "step": 2585
    },
    {
      "epoch": 0.5172,
      "grad_norm": 1.9290003776550293,
      "learning_rate": 5.574685752464334e-05,
      "loss": 0.37,
      "step": 2586
    },
    {
      "epoch": 0.5172,
      "loss_align": 0.179620623588562,
      "loss_contrastive": 0.006030559539794922,
      "loss_gen": 0.1827893853187561,
      "neg_sim_mean": 0.726409912109375,
      "pos_sim_mean": 0.820379376411438,
      "step": 2586
    },
    {
      "epoch": 0.5174,
      "grad_norm": 1.6221923828125,
      "learning_rate": 5.5712180876067045e-05,
      "loss": 0.3631,
      "step": 2587
    },
    {
      "epoch": 0.5174,
      "loss_align": 0.14389896392822266,
      "loss_contrastive": 0.04679572582244873,
      "loss_gen": 0.24172766506671906,
      "neg_sim_mean": 0.8028967380523682,
      "pos_sim_mean": 0.8561010360717773,
      "step": 2587
    },
    {
      "epoch": 0.5176,
      "grad_norm": 1.992453694343567,
      "learning_rate": 5.567750144344318e-05,
      "loss": 0.3912,
      "step": 2588
    },
    {
      "epoch": 0.5176,
      "loss_align": 0.11646974086761475,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16848663985729218,
      "neg_sim_mean": 0.6358020305633545,
      "pos_sim_mean": 0.8835302591323853,
      "step": 2588
    },
    {
      "epoch": 0.5178,
      "grad_norm": 1.4588302373886108,
      "learning_rate": 5.564281924367408e-05,
      "loss": 0.285,
      "step": 2589
    },
    {
      "epoch": 0.5178,
      "loss_align": 0.17286145687103271,
      "loss_contrastive": 0.017841577529907227,
      "loss_gen": 0.21300382912158966,
      "neg_sim_mean": 0.7449800968170166,
      "pos_sim_mean": 0.8271385431289673,
      "step": 2589
    },
    {
      "epoch": 0.518,
      "grad_norm": 1.7439326047897339,
      "learning_rate": 5.560813429366345e-05,
      "loss": 0.388,
      "step": 2590
    },
    {
      "epoch": 0.518,
      "loss_align": 0.19572186470031738,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002655001822859049,
      "neg_sim_mean": 0.576209306716919,
      "pos_sim_mean": 0.8042781352996826,
      "step": 2590
    },
    {
      "epoch": 0.5182,
      "grad_norm": 0.6538193225860596,
      "learning_rate": 5.557344661031627e-05,
      "loss": 0.1984,
      "step": 2591
    },
    {
      "epoch": 0.5182,
      "loss_align": 0.20566511154174805,
      "loss_contrastive": 0.15011781454086304,
      "loss_gen": 0.38163504004478455,
      "neg_sim_mean": 0.8444526791572571,
      "pos_sim_mean": 0.794334888458252,
      "step": 2591
    },
    {
      "epoch": 0.5184,
      "grad_norm": 2.205519914627075,
      "learning_rate": 5.5538756210538933e-05,
      "loss": 0.6053,
      "step": 2592
    },
    {
      "epoch": 0.5184,
      "loss_align": 0.2815040349960327,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02051473781466484,
      "neg_sim_mean": 0.3482019305229187,
      "pos_sim_mean": 0.7184959650039673,
      "step": 2592
    },
    {
      "epoch": 0.5186,
      "grad_norm": 1.086554765701294,
      "learning_rate": 5.550406311123911e-05,
      "loss": 0.302,
      "step": 2593
    },
    {
      "epoch": 0.5186,
      "loss_align": 0.19826960563659668,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21687579154968262,
      "neg_sim_mean": 0.6030311584472656,
      "pos_sim_mean": 0.8017303943634033,
      "step": 2593
    },
    {
      "epoch": 0.5188,
      "grad_norm": 1.5450972318649292,
      "learning_rate": 5.5469367329325784e-05,
      "loss": 0.4151,
      "step": 2594
    },
    {
      "epoch": 0.5188,
      "loss_align": 0.22637534141540527,
      "loss_contrastive": 0.16253674030303955,
      "loss_gen": 0.14215409755706787,
      "neg_sim_mean": 0.8361613750457764,
      "pos_sim_mean": 0.7736246585845947,
      "step": 2594
    },
    {
      "epoch": 0.519,
      "grad_norm": 1.7040282487869263,
      "learning_rate": 5.543466888170926e-05,
      "loss": 0.388,
      "step": 2595
    },
    {
      "epoch": 0.519,
      "loss_align": 0.17107701301574707,
      "loss_contrastive": 0.10493707656860352,
      "loss_gen": 0.3861408531665802,
      "neg_sim_mean": 0.8338600397109985,
      "pos_sim_mean": 0.8289229869842529,
      "step": 2595
    },
    {
      "epoch": 0.5192,
      "grad_norm": 2.0624749660491943,
      "learning_rate": 5.539996778530115e-05,
      "loss": 0.5698,
      "step": 2596
    },
    {
      "epoch": 0.5192,
      "loss_align": 0.18767642974853516,
      "loss_contrastive": 0.10750561952590942,
      "loss_gen": 0.19495806097984314,
      "neg_sim_mean": 0.8198291659355164,
      "pos_sim_mean": 0.8123235702514648,
      "step": 2596
    },
    {
      "epoch": 0.5194,
      "grad_norm": 1.645702838897705,
      "learning_rate": 5.5365264057014335e-05,
      "loss": 0.3955,
      "step": 2597
    },
    {
      "epoch": 0.5194,
      "loss_align": 0.1140713095664978,
      "loss_contrastive": 0.08341723680496216,
      "loss_gen": 0.15947376191616058,
      "neg_sim_mean": 0.8693459033966064,
      "pos_sim_mean": 0.8859286904335022,
      "step": 2597
    },
    {
      "epoch": 0.5196,
      "grad_norm": 1.6366361379623413,
      "learning_rate": 5.5330557713763e-05,
      "loss": 0.2836,
      "step": 2598
    },
    {
      "epoch": 0.5196,
      "loss_align": 0.19250184297561646,
      "loss_contrastive": 0.11255031824111938,
      "loss_gen": 0.20795747637748718,
      "neg_sim_mean": 0.820048451423645,
      "pos_sim_mean": 0.8074981570243835,
      "step": 2598
    },
    {
      "epoch": 0.5198,
      "grad_norm": 1.5485550165176392,
      "learning_rate": 5.52958487724626e-05,
      "loss": 0.414,
      "step": 2599
    },
    {
      "epoch": 0.5198,
      "loss_align": 0.1461753249168396,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2724801003932953,
      "neg_sim_mean": 0.7417832612991333,
      "pos_sim_mean": 0.8538246750831604,
      "step": 2599
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.126878499984741,
      "learning_rate": 5.5261137250029835e-05,
      "loss": 0.4187,
      "step": 2600
    },
    {
      "epoch": 0.52,
      "loss_align": 0.25003910064697266,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15192784368991852,
      "neg_sim_mean": 0.6289017200469971,
      "pos_sim_mean": 0.7499608993530273,
      "step": 2600
    },
    {
      "epoch": 0.5202,
      "grad_norm": 1.5316908359527588,
      "learning_rate": 5.522642316338268e-05,
      "loss": 0.402,
      "step": 2601
    },
    {
      "epoch": 0.5202,
      "loss_align": 0.11595559120178223,
      "loss_contrastive": 0.026565909385681152,
      "loss_gen": 0.18679891526699066,
      "neg_sim_mean": 0.810610294342041,
      "pos_sim_mean": 0.8840444087982178,
      "step": 2601
    },
    {
      "epoch": 0.5204,
      "grad_norm": 1.8434031009674072,
      "learning_rate": 5.519170652944037e-05,
      "loss": 0.3059,
      "step": 2602
    },
    {
      "epoch": 0.5204,
      "loss_align": 0.12084317207336426,
      "loss_contrastive": 0.07475012540817261,
      "loss_gen": 0.3117727041244507,
      "neg_sim_mean": 0.8539069294929504,
      "pos_sim_mean": 0.8791568279266357,
      "step": 2602
    },
    {
      "epoch": 0.5206,
      "grad_norm": 1.9302546977996826,
      "learning_rate": 5.515698736512337e-05,
      "loss": 0.4416,
      "step": 2603
    },
    {
      "epoch": 0.5206,
      "loss_align": 0.254585862159729,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003304120386019349,
      "neg_sim_mean": 0.4379636347293854,
      "pos_sim_mean": 0.745414137840271,
      "step": 2603
    },
    {
      "epoch": 0.5208,
      "grad_norm": 0.6927515268325806,
      "learning_rate": 5.512226568735338e-05,
      "loss": 0.2579,
      "step": 2604
    },
    {
      "epoch": 0.5208,
      "loss_align": 0.14116275310516357,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005181395914405584,
      "neg_sim_mean": 0.28830790519714355,
      "pos_sim_mean": 0.8588372468948364,
      "step": 2604
    },
    {
      "epoch": 0.521,
      "grad_norm": 0.586381733417511,
      "learning_rate": 5.508754151305332e-05,
      "loss": 0.1463,
      "step": 2605
    },
    {
      "epoch": 0.521,
      "loss_align": 0.1938772201538086,
      "loss_contrastive": 0.1387159824371338,
      "loss_gen": 0.16121266782283783,
      "neg_sim_mean": 0.8448387384414673,
      "pos_sim_mean": 0.8061227798461914,
      "step": 2605
    },
    {
      "epoch": 0.5212,
      "grad_norm": 2.0344042778015137,
      "learning_rate": 5.5052814859147315e-05,
      "loss": 0.3717,
      "step": 2606
    },
    {
      "epoch": 0.5212,
      "loss_align": 0.07721936702728271,
      "loss_contrastive": 0.08781349658966064,
      "loss_gen": 0.2823294997215271,
      "neg_sim_mean": 0.91059410572052,
      "pos_sim_mean": 0.9227806329727173,
      "step": 2606
    },
    {
      "epoch": 0.5214,
      "grad_norm": 1.7001605033874512,
      "learning_rate": 5.5018085742560744e-05,
      "loss": 0.3701,
      "step": 2607
    },
    {
      "epoch": 0.5214,
      "loss_align": 0.15867769718170166,
      "loss_contrastive": 0.10751783847808838,
      "loss_gen": 0.355040580034256,
      "neg_sim_mean": 0.8488401174545288,
      "pos_sim_mean": 0.8413223028182983,
      "step": 2607
    },
    {
      "epoch": 0.5216,
      "grad_norm": 2.683729648590088,
      "learning_rate": 5.498335418022015e-05,
      "loss": 0.5266,
      "step": 2608
    },
    {
      "epoch": 0.5216,
      "loss_align": 0.022553563117980957,
      "loss_contrastive": 0.025359511375427246,
      "loss_gen": 0.18991470336914062,
      "neg_sim_mean": 0.9028059244155884,
      "pos_sim_mean": 0.977446436882019,
      "step": 2608
    },
    {
      "epoch": 0.5218,
      "grad_norm": 1.4980230331420898,
      "learning_rate": 5.494862018905326e-05,
      "loss": 0.2155,
      "step": 2609
    },
    {
      "epoch": 0.5218,
      "loss_align": 0.19176805019378662,
      "loss_contrastive": 0.09598922729492188,
      "loss_gen": 0.479491651058197,
      "neg_sim_mean": 0.8042211532592773,
      "pos_sim_mean": 0.8082319498062134,
      "step": 2609
    },
    {
      "epoch": 0.522,
      "grad_norm": 2.290719509124756,
      "learning_rate": 5.4913883785988993e-05,
      "loss": 0.6828,
      "step": 2610
    },
    {
      "epoch": 0.522,
      "loss_align": 0.196305513381958,
      "loss_contrastive": 0.03406262397766113,
      "loss_gen": 0.24839594960212708,
      "neg_sim_mean": 0.7377570867538452,
      "pos_sim_mean": 0.803694486618042,
      "step": 2610
    },
    {
      "epoch": 0.5222,
      "grad_norm": 2.1277811527252197,
      "learning_rate": 5.487914498795747e-05,
      "loss": 0.4488,
      "step": 2611
    },
    {
      "epoch": 0.5222,
      "loss_align": 0.08515053987503052,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0023827017284929752,
      "neg_sim_mean": 0.33620262145996094,
      "pos_sim_mean": 0.9148494601249695,
      "step": 2611
    },
    {
      "epoch": 0.5224,
      "grad_norm": 0.33108994364738464,
      "learning_rate": 5.4844403811889965e-05,
      "loss": 0.0875,
      "step": 2612
    },
    {
      "epoch": 0.5224,
      "loss_align": 0.09983515739440918,
      "loss_contrastive": 0.052675843238830566,
      "loss_gen": 0.2787242829799652,
      "neg_sim_mean": 0.8528406620025635,
      "pos_sim_mean": 0.9001648426055908,
      "step": 2612
    },
    {
      "epoch": 0.5226,
      "grad_norm": 1.732050895690918,
      "learning_rate": 5.480966027471889e-05,
      "loss": 0.3849,
      "step": 2613
    },
    {
      "epoch": 0.5226,
      "loss_align": 0.13421636819839478,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005709195043891668,
      "neg_sim_mean": 0.2862699031829834,
      "pos_sim_mean": 0.8657836318016052,
      "step": 2613
    },
    {
      "epoch": 0.5228,
      "grad_norm": 0.5311042070388794,
      "learning_rate": 5.4774914393377816e-05,
      "loss": 0.1399,
      "step": 2614
    },
    {
      "epoch": 0.5228,
      "loss_align": 0.07803845405578613,
      "loss_contrastive": 0.07182741165161133,
      "loss_gen": 0.37835294008255005,
      "neg_sim_mean": 0.8937889337539673,
      "pos_sim_mean": 0.9219615459442139,
      "step": 2614
    },
    {
      "epoch": 0.523,
      "grad_norm": 1.7624014616012573,
      "learning_rate": 5.474016618480147e-05,
      "loss": 0.465,
      "step": 2615
    },
    {
      "epoch": 0.523,
      "loss_align": 0.10227340459823608,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16980773210525513,
      "neg_sim_mean": 0.6661771535873413,
      "pos_sim_mean": 0.8977265954017639,
      "step": 2615
    },
    {
      "epoch": 0.5232,
      "grad_norm": 1.551196813583374,
      "learning_rate": 5.470541566592573e-05,
      "loss": 0.2721,
      "step": 2616
    },
    {
      "epoch": 0.5232,
      "loss_align": 0.18350011110305786,
      "loss_contrastive": 0.03521990776062012,
      "loss_gen": 0.11473468691110611,
      "neg_sim_mean": 0.7517197728157043,
      "pos_sim_mean": 0.8164998888969421,
      "step": 2616
    },
    {
      "epoch": 0.5234,
      "grad_norm": 1.60942542552948,
      "learning_rate": 5.467066285368754e-05,
      "loss": 0.3025,
      "step": 2617
    },
    {
      "epoch": 0.5234,
      "loss_align": 0.2290213704109192,
      "loss_contrastive": 0.06494903564453125,
      "loss_gen": 0.38407328724861145,
      "neg_sim_mean": 0.7359276413917542,
      "pos_sim_mean": 0.7709786295890808,
      "step": 2617
    },
    {
      "epoch": 0.5236,
      "grad_norm": 2.497753858566284,
      "learning_rate": 5.463590776502501e-05,
      "loss": 0.6209,
      "step": 2618
    },
    {
      "epoch": 0.5236,
      "loss_align": 0.18514394760131836,
      "loss_contrastive": 0.12005650997161865,
      "loss_gen": 0.4152386486530304,
      "neg_sim_mean": 0.8349125385284424,
      "pos_sim_mean": 0.8148560523986816,
      "step": 2618
    },
    {
      "epoch": 0.5238,
      "grad_norm": 2.4248545169830322,
      "learning_rate": 5.4601150416877367e-05,
      "loss": 0.6148,
      "step": 2619
    },
    {
      "epoch": 0.5238,
      "loss_align": 0.2341700792312622,
      "loss_contrastive": 0.07333111763000488,
      "loss_gen": 0.5117067694664001,
      "neg_sim_mean": 0.7391610145568848,
      "pos_sim_mean": 0.7658299207687378,
      "step": 2619
    },
    {
      "epoch": 0.524,
      "grad_norm": 2.633035182952881,
      "learning_rate": 5.456639082618489e-05,
      "loss": 0.7547,
      "step": 2620
    },
    {
      "epoch": 0.524,
      "loss_align": 0.22035717964172363,
      "loss_contrastive": 0.0,
      "loss_gen": 0.026450812816619873,
      "neg_sim_mean": 0.328032910823822,
      "pos_sim_mean": 0.7796428203582764,
      "step": 2620
    },
    {
      "epoch": 0.5242,
      "grad_norm": 1.0063656568527222,
      "learning_rate": 5.453162900988902e-05,
      "loss": 0.2468,
      "step": 2621
    },
    {
      "epoch": 0.5242,
      "loss_align": 0.1048506498336792,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22281062602996826,
      "neg_sim_mean": 0.6371034383773804,
      "pos_sim_mean": 0.8951493501663208,
      "step": 2621
    },
    {
      "epoch": 0.5244,
      "grad_norm": 1.7295171022415161,
      "learning_rate": 5.449686498493219e-05,
      "loss": 0.3277,
      "step": 2622
    },
    {
      "epoch": 0.5244,
      "loss_align": 0.10789203643798828,
      "loss_contrastive": 0.06612163782119751,
      "loss_gen": 0.3770469129085541,
      "neg_sim_mean": 0.8582295775413513,
      "pos_sim_mean": 0.8921079635620117,
      "step": 2622
    },
    {
      "epoch": 0.5246,
      "grad_norm": 2.3147835731506348,
      "learning_rate": 5.446209876825803e-05,
      "loss": 0.4929,
      "step": 2623
    },
    {
      "epoch": 0.5246,
      "loss_align": 0.16791534423828125,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008026485331356525,
      "neg_sim_mean": 0.542312502861023,
      "pos_sim_mean": 0.8320846557617188,
      "step": 2623
    },
    {
      "epoch": 0.5248,
      "grad_norm": 0.6940951943397522,
      "learning_rate": 5.442733037681111e-05,
      "loss": 0.1759,
      "step": 2624
    },
    {
      "epoch": 0.5248,
      "loss_align": 0.14831042289733887,
      "loss_contrastive": 0.037755370140075684,
      "loss_gen": 0.26769715547561646,
      "neg_sim_mean": 0.7894449234008789,
      "pos_sim_mean": 0.8516895771026611,
      "step": 2624
    },
    {
      "epoch": 0.525,
      "grad_norm": 1.9960126876831055,
      "learning_rate": 5.439255982753717e-05,
      "loss": 0.4205,
      "step": 2625
    },
    {
      "epoch": 0.525,
      "loss_align": 0.20638507604599,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1702314019203186,
      "neg_sim_mean": 0.652682363986969,
      "pos_sim_mean": 0.79361492395401,
      "step": 2625
    },
    {
      "epoch": 0.5252,
      "grad_norm": 1.5667479038238525,
      "learning_rate": 5.435778713738292e-05,
      "loss": 0.3766,
      "step": 2626
    },
    {
      "epoch": 0.5252,
      "loss_align": 0.23991858959197998,
      "loss_contrastive": 0.115456223487854,
      "loss_gen": 0.20519904792308807,
      "neg_sim_mean": 0.7755376100540161,
      "pos_sim_mean": 0.76008141040802,
      "step": 2626
    },
    {
      "epoch": 0.5254,
      "grad_norm": 1.7408708333969116,
      "learning_rate": 5.432301232329615e-05,
      "loss": 0.459,
      "step": 2627
    },
    {
      "epoch": 0.5254,
      "loss_align": 0.18302088975906372,
      "loss_contrastive": 0.05839616060256958,
      "loss_gen": 0.19197165966033936,
      "neg_sim_mean": 0.775375247001648,
      "pos_sim_mean": 0.8169791102409363,
      "step": 2627
    },
    {
      "epoch": 0.5256,
      "grad_norm": 2.3838376998901367,
      "learning_rate": 5.428823540222569e-05,
      "loss": 0.382,
      "step": 2628
    },
    {
      "epoch": 0.5256,
      "loss_align": 0.16984856128692627,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0021098002325743437,
      "neg_sim_mean": 0.47404515743255615,
      "pos_sim_mean": 0.8301514387130737,
      "step": 2628
    },
    {
      "epoch": 0.5258,
      "grad_norm": 0.4496786892414093,
      "learning_rate": 5.4253456391121405e-05,
      "loss": 0.172,
      "step": 2629
    },
    {
      "epoch": 0.5258,
      "loss_align": 0.17754316329956055,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14805185794830322,
      "neg_sim_mean": 0.604122519493103,
      "pos_sim_mean": 0.8224568367004395,
      "step": 2629
    },
    {
      "epoch": 0.526,
      "grad_norm": 1.6338274478912354,
      "learning_rate": 5.421867530693414e-05,
      "loss": 0.3256,
      "step": 2630
    },
    {
      "epoch": 0.526,
      "loss_align": 0.12658393383026123,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028983699157834053,
      "neg_sim_mean": 0.3995784521102905,
      "pos_sim_mean": 0.8734160661697388,
      "step": 2630
    },
    {
      "epoch": 0.5262,
      "grad_norm": 1.2131447792053223,
      "learning_rate": 5.418389216661579e-05,
      "loss": 0.1556,
      "step": 2631
    },
    {
      "epoch": 0.5262,
      "loss_align": 0.1593698263168335,
      "loss_contrastive": 0.029041826725006104,
      "loss_gen": 0.14507809281349182,
      "neg_sim_mean": 0.7696719765663147,
      "pos_sim_mean": 0.8406301736831665,
      "step": 2631
    },
    {
      "epoch": 0.5264,
      "grad_norm": 1.6461181640625,
      "learning_rate": 5.41491069871192e-05,
      "loss": 0.3079,
      "step": 2632
    },
    {
      "epoch": 0.5264,
      "loss_align": 0.1811160445213318,
      "loss_contrastive": 0.0,
      "loss_gen": 0.32631590962409973,
      "neg_sim_mean": 0.6398190259933472,
      "pos_sim_mean": 0.8188839554786682,
      "step": 2632
    },
    {
      "epoch": 0.5266,
      "grad_norm": 2.048076868057251,
      "learning_rate": 5.411431978539828e-05,
      "loss": 0.5074,
      "step": 2633
    },
    {
      "epoch": 0.5266,
      "loss_align": 0.14768749475479126,
      "loss_contrastive": 0.029620349407196045,
      "loss_gen": 0.1691431999206543,
      "neg_sim_mean": 0.7819328308105469,
      "pos_sim_mean": 0.8523125052452087,
      "step": 2633
    },
    {
      "epoch": 0.5268,
      "grad_norm": 1.6934726238250732,
      "learning_rate": 5.407953057840789e-05,
      "loss": 0.3204,
      "step": 2634
    },
    {
      "epoch": 0.5268,
      "loss_align": 0.18060076236724854,
      "loss_contrastive": 0.05112910270690918,
      "loss_gen": 0.19140629470348358,
      "neg_sim_mean": 0.7705283164978027,
      "pos_sim_mean": 0.8193992376327515,
      "step": 2634
    },
    {
      "epoch": 0.527,
      "grad_norm": 1.8066704273223877,
      "learning_rate": 5.4044739383103835e-05,
      "loss": 0.3781,
      "step": 2635
    },
    {
      "epoch": 0.527,
      "loss_align": 0.2028491497039795,
      "loss_contrastive": 0.20811796188354492,
      "loss_gen": 0.21291007101535797,
      "neg_sim_mean": 0.9052687883377075,
      "pos_sim_mean": 0.7971508502960205,
      "step": 2635
    },
    {
      "epoch": 0.5272,
      "grad_norm": 1.578943133354187,
      "learning_rate": 5.4009946216442944e-05,
      "loss": 0.4407,
      "step": 2636
    },
    {
      "epoch": 0.5272,
      "loss_align": 0.2822631597518921,
      "loss_contrastive": 0.11946368217468262,
      "loss_gen": 0.15189962089061737,
      "neg_sim_mean": 0.7372004985809326,
      "pos_sim_mean": 0.7177368402481079,
      "step": 2636
    },
    {
      "epoch": 0.5274,
      "grad_norm": 1.8195050954818726,
      "learning_rate": 5.3975151095382995e-05,
      "loss": 0.4485,
      "step": 2637
    },
    {
      "epoch": 0.5274,
      "loss_align": 0.17154526710510254,
      "loss_contrastive": 0.10974293947219849,
      "loss_gen": 0.19490791857242584,
      "neg_sim_mean": 0.838197648525238,
      "pos_sim_mean": 0.8284547328948975,
      "step": 2637
    },
    {
      "epoch": 0.5276,
      "grad_norm": 1.735748052597046,
      "learning_rate": 5.394035403688268e-05,
      "loss": 0.3796,
      "step": 2638
    },
    {
      "epoch": 0.5276,
      "loss_align": 0.215684175491333,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22889651358127594,
      "neg_sim_mean": 0.604625940322876,
      "pos_sim_mean": 0.784315824508667,
      "step": 2638
    },
    {
      "epoch": 0.5278,
      "grad_norm": 1.7713059186935425,
      "learning_rate": 5.390555505790168e-05,
      "loss": 0.4446,
      "step": 2639
    },
    {
      "epoch": 0.5278,
      "loss_align": 0.34823179244995117,
      "loss_contrastive": 0.13881951570510864,
      "loss_gen": 0.3375096917152405,
      "neg_sim_mean": 0.6905876994132996,
      "pos_sim_mean": 0.6517682075500488,
      "step": 2639
    },
    {
      "epoch": 0.528,
      "grad_norm": 2.2271366119384766,
      "learning_rate": 5.3870754175400595e-05,
      "loss": 0.7024,
      "step": 2640
    },
    {
      "epoch": 0.528,
      "loss_align": 0.1913793683052063,
      "loss_contrastive": 0.019972026348114014,
      "loss_gen": 0.23071663081645966,
      "neg_sim_mean": 0.7285926342010498,
      "pos_sim_mean": 0.8086206316947937,
      "step": 2640
    },
    {
      "epoch": 0.5282,
      "grad_norm": 1.7695622444152832,
      "learning_rate": 5.383595140634093e-05,
      "loss": 0.4245,
      "step": 2641
    },
    {
      "epoch": 0.5282,
      "loss_align": 0.1185755729675293,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007659465540200472,
      "neg_sim_mean": 0.29306644201278687,
      "pos_sim_mean": 0.8814244270324707,
      "step": 2641
    },
    {
      "epoch": 0.5284,
      "grad_norm": 0.5858862400054932,
      "learning_rate": 5.3801146767685165e-05,
      "loss": 0.1262,
      "step": 2642
    },
    {
      "epoch": 0.5284,
      "loss_align": 0.09776842594146729,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3356524705886841,
      "neg_sim_mean": 0.7125699520111084,
      "pos_sim_mean": 0.9022315740585327,
      "step": 2642
    },
    {
      "epoch": 0.5286,
      "grad_norm": 2.7166502475738525,
      "learning_rate": 5.3766340276396646e-05,
      "loss": 0.4334,
      "step": 2643
    },
    {
      "epoch": 0.5286,
      "loss_align": 0.12203562259674072,
      "loss_contrastive": 0.040309131145477295,
      "loss_gen": 0.42598608136177063,
      "neg_sim_mean": 0.8182734847068787,
      "pos_sim_mean": 0.8779643774032593,
      "step": 2643
    },
    {
      "epoch": 0.5288,
      "grad_norm": 2.520648956298828,
      "learning_rate": 5.373153194943962e-05,
      "loss": 0.5529,
      "step": 2644
    },
    {
      "epoch": 0.5288,
      "loss_align": 0.14282584190368652,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12400476634502411,
      "neg_sim_mean": 0.6663167476654053,
      "pos_sim_mean": 0.8571741580963135,
      "step": 2644
    },
    {
      "epoch": 0.529,
      "grad_norm": 1.3276244401931763,
      "learning_rate": 5.369672180377926e-05,
      "loss": 0.2668,
      "step": 2645
    },
    {
      "epoch": 0.529,
      "loss_align": 0.44924283027648926,
      "loss_contrastive": 0.3839329183101654,
      "loss_gen": 0.17774657905101776,
      "neg_sim_mean": 0.8346900939941406,
      "pos_sim_mean": 0.5507571697235107,
      "step": 2645
    },
    {
      "epoch": 0.5292,
      "grad_norm": 1.6823649406433105,
      "learning_rate": 5.366190985638159e-05,
      "loss": 0.6731,
      "step": 2646
    },
    {
      "epoch": 0.5292,
      "loss_align": 0.2815595865249634,
      "loss_contrastive": 0.07955801486968994,
      "loss_gen": 0.2759420871734619,
      "neg_sim_mean": 0.6979984045028687,
      "pos_sim_mean": 0.7184404134750366,
      "step": 2646
    },
    {
      "epoch": 0.5294,
      "grad_norm": 1.9645214080810547,
      "learning_rate": 5.362709612421355e-05,
      "loss": 0.567,
      "step": 2647
    },
    {
      "epoch": 0.5294,
      "loss_align": 0.2160816192626953,
      "loss_contrastive": 0.01989215612411499,
      "loss_gen": 0.3475140929222107,
      "neg_sim_mean": 0.7038105130195618,
      "pos_sim_mean": 0.7839183807373047,
      "step": 2647
    },
    {
      "epoch": 0.5296,
      "grad_norm": 2.4750821590423584,
      "learning_rate": 5.3592280624242917e-05,
      "loss": 0.566,
      "step": 2648
    },
    {
      "epoch": 0.5296,
      "loss_align": 0.1359093189239502,
      "loss_contrastive": 0.0034844279289245605,
      "loss_gen": 0.23590919375419617,
      "neg_sim_mean": 0.7675750851631165,
      "pos_sim_mean": 0.8640906810760498,
      "step": 2648
    },
    {
      "epoch": 0.5298,
      "grad_norm": 1.997243046760559,
      "learning_rate": 5.3557463373438357e-05,
      "loss": 0.3722,
      "step": 2649
    },
    {
      "epoch": 0.5298,
      "loss_align": 0.1440805196762085,
      "loss_contrastive": 0.09020525217056274,
      "loss_gen": 0.25173139572143555,
      "neg_sim_mean": 0.8461247086524963,
      "pos_sim_mean": 0.8559194803237915,
      "step": 2649
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.8629127740859985,
      "learning_rate": 5.352264438876935e-05,
      "loss": 0.4066,
      "step": 2650
    },
    {
      "epoch": 0.53,
      "loss_align": 0.1427779197692871,
      "loss_contrastive": 0.009578108787536621,
      "loss_gen": 0.35174134373664856,
      "neg_sim_mean": 0.7668001651763916,
      "pos_sim_mean": 0.8572220802307129,
      "step": 2650
    },
    {
      "epoch": 0.5302,
      "grad_norm": 2.4526426792144775,
      "learning_rate": 5.348782368720626e-05,
      "loss": 0.4957,
      "step": 2651
    },
    {
      "epoch": 0.5302,
      "loss_align": 0.21836340427398682,
      "loss_contrastive": 0.18490099906921387,
      "loss_gen": 0.301203191280365,
      "neg_sim_mean": 0.8665375709533691,
      "pos_sim_mean": 0.7816365957260132,
      "step": 2651
    },
    {
      "epoch": 0.5304,
      "grad_norm": 1.9434136152267456,
      "learning_rate": 5.345300128572031e-05,
      "loss": 0.5418,
      "step": 2652
    },
    {
      "epoch": 0.5304,
      "loss_align": 0.1357704997062683,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09957689791917801,
      "neg_sim_mean": 0.7300986647605896,
      "pos_sim_mean": 0.8642295002937317,
      "step": 2652
    },
    {
      "epoch": 0.5306,
      "grad_norm": 1.211912989616394,
      "learning_rate": 5.3418177201283434e-05,
      "loss": 0.2353,
      "step": 2653
    },
    {
      "epoch": 0.5306,
      "loss_align": 0.19513309001922607,
      "loss_contrastive": 0.12691336870193481,
      "loss_gen": 0.2060525119304657,
      "neg_sim_mean": 0.8317802548408508,
      "pos_sim_mean": 0.8048669099807739,
      "step": 2653
    },
    {
      "epoch": 0.5308,
      "grad_norm": 1.8288706541061401,
      "learning_rate": 5.3383351450868544e-05,
      "loss": 0.4164,
      "step": 2654
    },
    {
      "epoch": 0.5308,
      "loss_align": 0.186761736869812,
      "loss_contrastive": 0.05624467134475708,
      "loss_gen": 0.34072479605674744,
      "neg_sim_mean": 0.7694829106330872,
      "pos_sim_mean": 0.813238263130188,
      "step": 2654
    },
    {
      "epoch": 0.531,
      "grad_norm": 2.2373640537261963,
      "learning_rate": 5.3348524051449254e-05,
      "loss": 0.5342,
      "step": 2655
    },
    {
      "epoch": 0.531,
      "loss_align": 0.2142544984817505,
      "loss_contrastive": 0.0260847806930542,
      "loss_gen": 0.35358232259750366,
      "neg_sim_mean": 0.7118302583694458,
      "pos_sim_mean": 0.7857455015182495,
      "step": 2655
    },
    {
      "epoch": 0.5312,
      "grad_norm": 2.0655879974365234,
      "learning_rate": 5.3313695020000024e-05,
      "loss": 0.571,
      "step": 2656
    },
    {
      "epoch": 0.5312,
      "loss_align": 0.2112255096435547,
      "loss_contrastive": 0.053382277488708496,
      "loss_gen": 0.19400006532669067,
      "neg_sim_mean": 0.7421567440032959,
      "pos_sim_mean": 0.7887744903564453,
      "step": 2656
    },
    {
      "epoch": 0.5314,
      "grad_norm": 2.2887794971466064,
      "learning_rate": 5.3278864373496085e-05,
      "loss": 0.4116,
      "step": 2657
    },
    {
      "epoch": 0.5314,
      "loss_align": 0.2030743956565857,
      "loss_contrastive": 0.13381332159042358,
      "loss_gen": 0.37281402945518494,
      "neg_sim_mean": 0.83073890209198,
      "pos_sim_mean": 0.7969256043434143,
      "step": 2657
    },
    {
      "epoch": 0.5316,
      "grad_norm": 1.9894620180130005,
      "learning_rate": 5.3244032128913476e-05,
      "loss": 0.5919,
      "step": 2658
    },
    {
      "epoch": 0.5316,
      "loss_align": 0.09779053926467896,
      "loss_contrastive": 0.09859824180603027,
      "loss_gen": 0.18567226827144623,
      "neg_sim_mean": 0.9008076786994934,
      "pos_sim_mean": 0.902209460735321,
      "step": 2658
    },
    {
      "epoch": 0.5318,
      "grad_norm": 1.5298458337783813,
      "learning_rate": 5.3209198303229027e-05,
      "loss": 0.2953,
      "step": 2659
    },
    {
      "epoch": 0.5318,
      "loss_align": 0.15739262104034424,
      "loss_contrastive": 0.05761992931365967,
      "loss_gen": 0.2542271912097931,
      "neg_sim_mean": 0.8002272844314575,
      "pos_sim_mean": 0.8426073789596558,
      "step": 2659
    },
    {
      "epoch": 0.532,
      "grad_norm": 1.922645926475525,
      "learning_rate": 5.3174362913420306e-05,
      "loss": 0.4185,
      "step": 2660
    },
    {
      "epoch": 0.532,
      "loss_align": 0.08811342716217041,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19996467232704163,
      "neg_sim_mean": 0.7914434671401978,
      "pos_sim_mean": 0.9118865728378296,
      "step": 2660
    },
    {
      "epoch": 0.5322,
      "grad_norm": 1.5137875080108643,
      "learning_rate": 5.313952597646568e-05,
      "loss": 0.2881,
      "step": 2661
    },
    {
      "epoch": 0.5322,
      "loss_align": 0.16852515935897827,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19918976724147797,
      "neg_sim_mean": 0.6511814594268799,
      "pos_sim_mean": 0.8314748406410217,
      "step": 2661
    },
    {
      "epoch": 0.5324,
      "grad_norm": 1.6332989931106567,
      "learning_rate": 5.3104687509344206e-05,
      "loss": 0.3677,
      "step": 2662
    },
    {
      "epoch": 0.5324,
      "loss_align": 0.12404662370681763,
      "loss_contrastive": 0.07767635583877563,
      "loss_gen": 0.1758720576763153,
      "neg_sim_mean": 0.8536297082901001,
      "pos_sim_mean": 0.8759533762931824,
      "step": 2662
    },
    {
      "epoch": 0.5326,
      "grad_norm": 1.6763046979904175,
      "learning_rate": 5.306984752903578e-05,
      "loss": 0.3092,
      "step": 2663
    },
    {
      "epoch": 0.5326,
      "loss_align": 0.16936421394348145,
      "loss_contrastive": 0.08250594139099121,
      "loss_gen": 0.1083005741238594,
      "neg_sim_mean": 0.8131417036056519,
      "pos_sim_mean": 0.8306357860565186,
      "step": 2663
    },
    {
      "epoch": 0.5328,
      "grad_norm": 1.3551450967788696,
      "learning_rate": 5.3035006052520955e-05,
      "loss": 0.2876,
      "step": 2664
    },
    {
      "epoch": 0.5328,
      "loss_align": 0.05902582406997681,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20471855998039246,
      "neg_sim_mean": 0.7579680681228638,
      "pos_sim_mean": 0.9409741759300232,
      "step": 2664
    },
    {
      "epoch": 0.533,
      "grad_norm": 1.7892299890518188,
      "learning_rate": 5.300016309678104e-05,
      "loss": 0.2637,
      "step": 2665
    },
    {
      "epoch": 0.533,
      "loss_align": 0.09952998161315918,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25153595209121704,
      "neg_sim_mean": 0.7933966517448425,
      "pos_sim_mean": 0.9004700183868408,
      "step": 2665
    },
    {
      "epoch": 0.5332,
      "grad_norm": 1.5172334909439087,
      "learning_rate": 5.296531867879809e-05,
      "loss": 0.3511,
      "step": 2666
    },
    {
      "epoch": 0.5332,
      "loss_align": 0.10285341739654541,
      "loss_contrastive": 0.05907547473907471,
      "loss_gen": 0.33892127871513367,
      "neg_sim_mean": 0.8562220335006714,
      "pos_sim_mean": 0.8971465826034546,
      "step": 2666
    },
    {
      "epoch": 0.5334,
      "grad_norm": 1.9305726289749146,
      "learning_rate": 5.293047281555482e-05,
      "loss": 0.4489,
      "step": 2667
    },
    {
      "epoch": 0.5334,
      "loss_align": 0.08842611312866211,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2555544972419739,
      "neg_sim_mean": 0.38922080397605896,
      "pos_sim_mean": 0.9115738868713379,
      "step": 2667
    },
    {
      "epoch": 0.5336,
      "grad_norm": 1.2806472778320312,
      "learning_rate": 5.289562552403472e-05,
      "loss": 0.344,
      "step": 2668
    },
    {
      "epoch": 0.5336,
      "loss_align": 0.26186156272888184,
      "loss_contrastive": 0.2111855149269104,
      "loss_gen": 0.26594051718711853,
      "neg_sim_mean": 0.8493239283561707,
      "pos_sim_mean": 0.7381384372711182,
      "step": 2668
    },
    {
      "epoch": 0.5338,
      "grad_norm": 2.0426223278045654,
      "learning_rate": 5.286077682122191e-05,
      "loss": 0.5531,
      "step": 2669
    },
    {
      "epoch": 0.5338,
      "loss_align": 0.1621670126914978,
      "loss_contrastive": 0.09388405084609985,
      "loss_gen": 0.2715786099433899,
      "neg_sim_mean": 0.8317170143127441,
      "pos_sim_mean": 0.8378329873085022,
      "step": 2669
    },
    {
      "epoch": 0.534,
      "grad_norm": 1.8998429775238037,
      "learning_rate": 5.2825926724101236e-05,
      "loss": 0.445,
      "step": 2670
    },
    {
      "epoch": 0.534,
      "loss_align": 0.16192841529846191,
      "loss_contrastive": 0.0,
      "loss_gen": 0.023724481463432312,
      "neg_sim_mean": 0.5004870295524597,
      "pos_sim_mean": 0.8380715847015381,
      "step": 2670
    },
    {
      "epoch": 0.5342,
      "grad_norm": 1.0891447067260742,
      "learning_rate": 5.279107524965819e-05,
      "loss": 0.1857,
      "step": 2671
    },
    {
      "epoch": 0.5342,
      "loss_align": 0.14922738075256348,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2366442233324051,
      "neg_sim_mean": 0.5187557935714722,
      "pos_sim_mean": 0.8507726192474365,
      "step": 2671
    },
    {
      "epoch": 0.5344,
      "grad_norm": 2.0231831073760986,
      "learning_rate": 5.275622241487899e-05,
      "loss": 0.3859,
      "step": 2672
    },
    {
      "epoch": 0.5344,
      "loss_align": 0.09478020668029785,
      "loss_contrastive": 0.07543528079986572,
      "loss_gen": 0.2563786208629608,
      "neg_sim_mean": 0.88065505027771,
      "pos_sim_mean": 0.9052197933197021,
      "step": 2672
    },
    {
      "epoch": 0.5346,
      "grad_norm": 1.7924307584762573,
      "learning_rate": 5.272136823675046e-05,
      "loss": 0.3602,
      "step": 2673
    },
    {
      "epoch": 0.5346,
      "loss_align": 0.13703930377960205,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0127710634842515,
      "neg_sim_mean": 0.46342378854751587,
      "pos_sim_mean": 0.862960696220398,
      "step": 2673
    },
    {
      "epoch": 0.5348,
      "grad_norm": 0.7557847499847412,
      "learning_rate": 5.268651273226011e-05,
      "loss": 0.1498,
      "step": 2674
    },
    {
      "epoch": 0.5348,
      "loss_align": 0.09615534543991089,
      "loss_contrastive": 0.02985966205596924,
      "loss_gen": 0.252657026052475,
      "neg_sim_mean": 0.8337042927742004,
      "pos_sim_mean": 0.9038446545600891,
      "step": 2674
    },
    {
      "epoch": 0.535,
      "grad_norm": 1.5066834688186646,
      "learning_rate": 5.265165591839609e-05,
      "loss": 0.3524,
      "step": 2675
    },
    {
      "epoch": 0.535,
      "loss_align": 0.0808790922164917,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007989429868757725,
      "neg_sim_mean": 0.4041464626789093,
      "pos_sim_mean": 0.9191209077835083,
      "step": 2675
    },
    {
      "epoch": 0.5352,
      "grad_norm": 0.524796724319458,
      "learning_rate": 5.26167978121472e-05,
      "loss": 0.0889,
      "step": 2676
    },
    {
      "epoch": 0.5352,
      "loss_align": 0.1401262879371643,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22341632843017578,
      "neg_sim_mean": 0.7468715906143188,
      "pos_sim_mean": 0.8598737120628357,
      "step": 2676
    },
    {
      "epoch": 0.5354,
      "grad_norm": 1.6489640474319458,
      "learning_rate": 5.258193843050283e-05,
      "loss": 0.3635,
      "step": 2677
    },
    {
      "epoch": 0.5354,
      "loss_align": 0.15719115734100342,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20131660997867584,
      "neg_sim_mean": 0.5381140112876892,
      "pos_sim_mean": 0.8428088426589966,
      "step": 2677
    },
    {
      "epoch": 0.5356,
      "grad_norm": 1.5169373750686646,
      "learning_rate": 5.2547077790453045e-05,
      "loss": 0.3585,
      "step": 2678
    },
    {
      "epoch": 0.5356,
      "loss_align": 0.11016345024108887,
      "loss_contrastive": 0.034722208976745605,
      "loss_gen": 0.1317707598209381,
      "neg_sim_mean": 0.8245587348937988,
      "pos_sim_mean": 0.8898365497589111,
      "step": 2678
    },
    {
      "epoch": 0.5358,
      "grad_norm": 1.4140186309814453,
      "learning_rate": 5.2512215908988484e-05,
      "loss": 0.2461,
      "step": 2679
    },
    {
      "epoch": 0.5358,
      "loss_align": 0.27923303842544556,
      "loss_contrastive": 0.11287647485733032,
      "loss_gen": 0.23849229514598846,
      "neg_sim_mean": 0.7336434125900269,
      "pos_sim_mean": 0.7207669615745544,
      "step": 2679
    },
    {
      "epoch": 0.536,
      "grad_norm": 2.116007089614868,
      "learning_rate": 5.247735280310041e-05,
      "loss": 0.5313,
      "step": 2680
    },
    {
      "epoch": 0.536,
      "loss_align": 0.2003907561302185,
      "loss_contrastive": 0.00671619176864624,
      "loss_gen": 0.23081330955028534,
      "neg_sim_mean": 0.7063254117965698,
      "pos_sim_mean": 0.7996092438697815,
      "step": 2680
    },
    {
      "epoch": 0.5362,
      "grad_norm": 1.8022489547729492,
      "learning_rate": 5.244248848978067e-05,
      "loss": 0.432,
      "step": 2681
    },
    {
      "epoch": 0.5362,
      "loss_align": 0.18796443939208984,
      "loss_contrastive": 0.1513562798500061,
      "loss_gen": 0.2165888547897339,
      "neg_sim_mean": 0.8633918166160583,
      "pos_sim_mean": 0.8120355606079102,
      "step": 2681
    },
    {
      "epoch": 0.5364,
      "grad_norm": 2.1635379791259766,
      "learning_rate": 5.240762298602171e-05,
      "loss": 0.4227,
      "step": 2682
    },
    {
      "epoch": 0.5364,
      "loss_align": 0.2721215486526489,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23011985421180725,
      "neg_sim_mean": 0.6117434501647949,
      "pos_sim_mean": 0.7278784513473511,
      "step": 2682
    },
    {
      "epoch": 0.5366,
      "grad_norm": 1.355137825012207,
      "learning_rate": 5.237275630881657e-05,
      "loss": 0.5022,
      "step": 2683
    },
    {
      "epoch": 0.5366,
      "loss_align": 0.15967875719070435,
      "loss_contrastive": 0.03270381689071655,
      "loss_gen": 0.17782817780971527,
      "neg_sim_mean": 0.7730250358581543,
      "pos_sim_mean": 0.8403212428092957,
      "step": 2683
    },
    {
      "epoch": 0.5368,
      "grad_norm": 1.8736392259597778,
      "learning_rate": 5.233788847515881e-05,
      "loss": 0.3414,
      "step": 2684
    },
    {
      "epoch": 0.5368,
      "loss_align": 0.06940680742263794,
      "loss_contrastive": 0.0,
      "loss_gen": 0.237477108836174,
      "neg_sim_mean": 0.8181537389755249,
      "pos_sim_mean": 0.9305931925773621,
      "step": 2684
    },
    {
      "epoch": 0.537,
      "grad_norm": 1.9503732919692993,
      "learning_rate": 5.230301950204262e-05,
      "loss": 0.3069,
      "step": 2685
    },
    {
      "epoch": 0.537,
      "loss_align": 0.1332213282585144,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02634056657552719,
      "neg_sim_mean": 0.26683446764945984,
      "pos_sim_mean": 0.8667786717414856,
      "step": 2685
    },
    {
      "epoch": 0.5372,
      "grad_norm": 0.9523817896842957,
      "learning_rate": 5.226814940646269e-05,
      "loss": 0.1596,
      "step": 2686
    },
    {
      "epoch": 0.5372,
      "loss_align": 0.10698974132537842,
      "loss_contrastive": 0.034919559955596924,
      "loss_gen": 0.1283920705318451,
      "neg_sim_mean": 0.8279297947883606,
      "pos_sim_mean": 0.8930102586746216,
      "step": 2686
    },
    {
      "epoch": 0.5374,
      "grad_norm": 1.140051007270813,
      "learning_rate": 5.223327820541431e-05,
      "loss": 0.2396,
      "step": 2687
    },
    {
      "epoch": 0.5374,
      "loss_align": 0.20221161842346191,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009436964057385921,
      "neg_sim_mean": 0.41982898116111755,
      "pos_sim_mean": 0.7977883815765381,
      "step": 2687
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.6667237281799316,
      "learning_rate": 5.219840591589325e-05,
      "loss": 0.2116,
      "step": 2688
    },
    {
      "epoch": 0.5376,
      "loss_align": 0.14525353908538818,
      "loss_contrastive": 0.054062485694885254,
      "loss_gen": 0.24191880226135254,
      "neg_sim_mean": 0.8088089227676392,
      "pos_sim_mean": 0.8547464609146118,
      "step": 2688
    },
    {
      "epoch": 0.5378,
      "grad_norm": 2.258793592453003,
      "learning_rate": 5.2163532554895855e-05,
      "loss": 0.3937,
      "step": 2689
    },
    {
      "epoch": 0.5378,
      "loss_align": 0.22098839282989502,
      "loss_contrastive": 0.21925044059753418,
      "loss_gen": 0.34651243686676025,
      "neg_sim_mean": 0.8982620239257812,
      "pos_sim_mean": 0.779011607170105,
      "step": 2689
    },
    {
      "epoch": 0.538,
      "grad_norm": 2.3821234703063965,
      "learning_rate": 5.212865813941899e-05,
      "loss": 0.5938,
      "step": 2690
    },
    {
      "epoch": 0.538,
      "loss_align": 0.21067607402801514,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11567190289497375,
      "neg_sim_mean": 0.6532903909683228,
      "pos_sim_mean": 0.7893239259719849,
      "step": 2690
    },
    {
      "epoch": 0.5382,
      "grad_norm": 1.4450242519378662,
      "learning_rate": 5.209378268645998e-05,
      "loss": 0.3263,
      "step": 2691
    },
    {
      "epoch": 0.5382,
      "loss_align": 0.11602741479873657,
      "loss_contrastive": 0.1002076268196106,
      "loss_gen": 0.20223543047904968,
      "neg_sim_mean": 0.8841801881790161,
      "pos_sim_mean": 0.8839725852012634,
      "step": 2691
    },
    {
      "epoch": 0.5384,
      "grad_norm": 1.6876150369644165,
      "learning_rate": 5.205890621301676e-05,
      "loss": 0.3303,
      "step": 2692
    },
    {
      "epoch": 0.5384,
      "loss_align": 0.07323336601257324,
      "loss_contrastive": 0.06082618236541748,
      "loss_gen": 0.062313493341207504,
      "neg_sim_mean": 0.8875927925109863,
      "pos_sim_mean": 0.9267666339874268,
      "step": 2692
    },
    {
      "epoch": 0.5386,
      "grad_norm": 0.8318004012107849,
      "learning_rate": 5.202402873608763e-05,
      "loss": 0.1428,
      "step": 2693
    },
    {
      "epoch": 0.5386,
      "loss_align": 0.2287987470626831,
      "loss_contrastive": 0.15488195419311523,
      "loss_gen": 0.2353103905916214,
      "neg_sim_mean": 0.8260831832885742,
      "pos_sim_mean": 0.7712012529373169,
      "step": 2693
    },
    {
      "epoch": 0.5388,
      "grad_norm": 1.9118404388427734,
      "learning_rate": 5.19891502726715e-05,
      "loss": 0.4827,
      "step": 2694
    },
    {
      "epoch": 0.5388,
      "loss_align": 0.3456169366836548,
      "loss_contrastive": 0.3363533616065979,
      "loss_gen": 0.1405947357416153,
      "neg_sim_mean": 0.8907364010810852,
      "pos_sim_mean": 0.6543830633163452,
      "step": 2694
    },
    {
      "epoch": 0.539,
      "grad_norm": 1.680350422859192,
      "learning_rate": 5.1954270839767684e-05,
      "loss": 0.5266,
      "step": 2695
    },
    {
      "epoch": 0.539,
      "loss_align": 0.13658976554870605,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12559811770915985,
      "neg_sim_mean": 0.7271515727043152,
      "pos_sim_mean": 0.863410234451294,
      "step": 2695
    },
    {
      "epoch": 0.5392,
      "grad_norm": 1.3736509084701538,
      "learning_rate": 5.191939045437601e-05,
      "loss": 0.2622,
      "step": 2696
    },
    {
      "epoch": 0.5392,
      "loss_align": 0.2031957507133484,
      "loss_contrastive": 0.13365375995635986,
      "loss_gen": 0.17235781252384186,
      "neg_sim_mean": 0.8304579854011536,
      "pos_sim_mean": 0.7968042492866516,
      "step": 2696
    },
    {
      "epoch": 0.5394,
      "grad_norm": 1.9954578876495361,
      "learning_rate": 5.188450913349674e-05,
      "loss": 0.3916,
      "step": 2697
    },
    {
      "epoch": 0.5394,
      "loss_align": 0.1592589020729065,
      "loss_contrastive": 0.11272275447845459,
      "loss_gen": 0.354112833738327,
      "neg_sim_mean": 0.8534638285636902,
      "pos_sim_mean": 0.8407410979270935,
      "step": 2697
    },
    {
      "epoch": 0.5396,
      "grad_norm": 2.0968058109283447,
      "learning_rate": 5.18496268941306e-05,
      "loss": 0.5269,
      "step": 2698
    },
    {
      "epoch": 0.5396,
      "loss_align": 0.0968506932258606,
      "loss_contrastive": 0.06579846143722534,
      "loss_gen": 0.23382867872714996,
      "neg_sim_mean": 0.8689477443695068,
      "pos_sim_mean": 0.9031493067741394,
      "step": 2698
    },
    {
      "epoch": 0.5398,
      "grad_norm": 1.7340362071990967,
      "learning_rate": 5.1814743753278795e-05,
      "loss": 0.3386,
      "step": 2699
    },
    {
      "epoch": 0.5398,
      "loss_align": 0.14685803651809692,
      "loss_contrastive": 0.11260199546813965,
      "loss_gen": 0.31609803438186646,
      "neg_sim_mean": 0.8657439351081848,
      "pos_sim_mean": 0.8531419634819031,
      "step": 2699
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.694791316986084,
      "learning_rate": 5.1779859727942924e-05,
      "loss": 0.4765,
      "step": 2700
    },
    {
      "epoch": 0.54,
      "loss_align": 0.15347027778625488,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2586764097213745,
      "neg_sim_mean": 0.738739550113678,
      "pos_sim_mean": 0.8465297222137451,
      "step": 2700
    },
    {
      "epoch": 0.5402,
      "grad_norm": 2.038964033126831,
      "learning_rate": 5.174497483512506e-05,
      "loss": 0.4121,
      "step": 2701
    },
    {
      "epoch": 0.5402,
      "loss_align": 0.1531212329864502,
      "loss_contrastive": 0.08187055587768555,
      "loss_gen": 0.18394319713115692,
      "neg_sim_mean": 0.8287492990493774,
      "pos_sim_mean": 0.8468787670135498,
      "step": 2701
    },
    {
      "epoch": 0.5404,
      "grad_norm": 1.728468418121338,
      "learning_rate": 5.171008909182765e-05,
      "loss": 0.3469,
      "step": 2702
    },
    {
      "epoch": 0.5404,
      "loss_align": 0.11590003967285156,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2523233890533447,
      "neg_sim_mean": 0.740006685256958,
      "pos_sim_mean": 0.8840999603271484,
      "step": 2702
    },
    {
      "epoch": 0.5406,
      "grad_norm": 1.6711935997009277,
      "learning_rate": 5.167520251505358e-05,
      "loss": 0.3682,
      "step": 2703
    },
    {
      "epoch": 0.5406,
      "loss_align": 0.15729796886444092,
      "loss_contrastive": 0.09757864475250244,
      "loss_gen": 0.12015479058027267,
      "neg_sim_mean": 0.8402806520462036,
      "pos_sim_mean": 0.8427020311355591,
      "step": 2703
    },
    {
      "epoch": 0.5408,
      "grad_norm": 1.2023036479949951,
      "learning_rate": 5.164031512180616e-05,
      "loss": 0.2892,
      "step": 2704
    },
    {
      "epoch": 0.5408,
      "loss_align": 0.16265004873275757,
      "loss_contrastive": 0.13430720567703247,
      "loss_gen": 0.35908064246177673,
      "neg_sim_mean": 0.871657133102417,
      "pos_sim_mean": 0.8373499512672424,
      "step": 2704
    },
    {
      "epoch": 0.541,
      "grad_norm": 2.362122058868408,
      "learning_rate": 5.1605426929089085e-05,
      "loss": 0.5378,
      "step": 2705
    },
    {
      "epoch": 0.541,
      "loss_align": 0.12944912910461426,
      "loss_contrastive": 0.012820243835449219,
      "loss_gen": 0.22789093852043152,
      "neg_sim_mean": 0.783371090888977,
      "pos_sim_mean": 0.8705508708953857,
      "step": 2705
    },
    {
      "epoch": 0.5412,
      "grad_norm": 1.8181288242340088,
      "learning_rate": 5.157053795390642e-05,
      "loss": 0.3589,
      "step": 2706
    },
    {
      "epoch": 0.5412,
      "loss_align": 0.13125205039978027,
      "loss_contrastive": 0.04459017515182495,
      "loss_gen": 0.4066831171512604,
      "neg_sim_mean": 0.8133381009101868,
      "pos_sim_mean": 0.8687479496002197,
      "step": 2706
    },
    {
      "epoch": 0.5414,
      "grad_norm": 2.236640214920044,
      "learning_rate": 5.153564821326264e-05,
      "loss": 0.5433,
      "step": 2707
    },
    {
      "epoch": 0.5414,
      "loss_align": 0.2608835697174072,
      "loss_contrastive": 0.03530949354171753,
      "loss_gen": 0.11765657365322113,
      "neg_sim_mean": 0.6744258999824524,
      "pos_sim_mean": 0.7391164302825928,
      "step": 2707
    },
    {
      "epoch": 0.5416,
      "grad_norm": 1.4157122373580933,
      "learning_rate": 5.150075772416256e-05,
      "loss": 0.3828,
      "step": 2708
    },
    {
      "epoch": 0.5416,
      "loss_align": 0.19582802057266235,
      "loss_contrastive": 0.15399795770645142,
      "loss_gen": 0.20762275159358978,
      "neg_sim_mean": 0.8581699132919312,
      "pos_sim_mean": 0.8041719794273376,
      "step": 2708
    },
    {
      "epoch": 0.5418,
      "grad_norm": 1.7827305793762207,
      "learning_rate": 5.1465866503611426e-05,
      "loss": 0.4219,
      "step": 2709
    },
    {
      "epoch": 0.5418,
      "loss_align": 0.13215512037277222,
      "loss_contrastive": 0.09450870752334595,
      "loss_gen": 0.48267921805381775,
      "neg_sim_mean": 0.8623535633087158,
      "pos_sim_mean": 0.8678448796272278,
      "step": 2709
    },
    {
      "epoch": 0.542,
      "grad_norm": 2.5794081687927246,
      "learning_rate": 5.143097456861474e-05,
      "loss": 0.6262,
      "step": 2710
    },
    {
      "epoch": 0.542,
      "loss_align": 0.10428428649902344,
      "loss_contrastive": 0.025569915771484375,
      "loss_gen": 0.17658714950084686,
      "neg_sim_mean": 0.821285605430603,
      "pos_sim_mean": 0.8957157135009766,
      "step": 2710
    },
    {
      "epoch": 0.5422,
      "grad_norm": 1.5249221324920654,
      "learning_rate": 5.139608193617845e-05,
      "loss": 0.2839,
      "step": 2711
    },
    {
      "epoch": 0.5422,
      "loss_align": 0.25263476371765137,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11460396647453308,
      "neg_sim_mean": 0.6330602765083313,
      "pos_sim_mean": 0.7473652362823486,
      "step": 2711
    },
    {
      "epoch": 0.5424,
      "grad_norm": 1.6246899366378784,
      "learning_rate": 5.136118862330876e-05,
      "loss": 0.3672,
      "step": 2712
    },
    {
      "epoch": 0.5424,
      "loss_align": 0.1882082223892212,
      "loss_contrastive": 0.06439697742462158,
      "loss_gen": 0.20742729306221008,
      "neg_sim_mean": 0.7761887311935425,
      "pos_sim_mean": 0.8117917776107788,
      "step": 2712
    },
    {
      "epoch": 0.5426,
      "grad_norm": 1.9141755104064941,
      "learning_rate": 5.13262946470123e-05,
      "loss": 0.4034,
      "step": 2713
    },
    {
      "epoch": 0.5426,
      "loss_align": 0.23622512817382812,
      "loss_contrastive": 0.09064072370529175,
      "loss_gen": 0.2450471818447113,
      "neg_sim_mean": 0.7544155716896057,
      "pos_sim_mean": 0.7637748718261719,
      "step": 2713
    },
    {
      "epoch": 0.5428,
      "grad_norm": 2.1055521965026855,
      "learning_rate": 5.129140002429594e-05,
      "loss": 0.4921,
      "step": 2714
    },
    {
      "epoch": 0.5428,
      "loss_align": 0.15881198644638062,
      "loss_contrastive": 0.012331187725067139,
      "loss_gen": 0.23017920553684235,
      "neg_sim_mean": 0.7535191774368286,
      "pos_sim_mean": 0.8411880135536194,
      "step": 2714
    },
    {
      "epoch": 0.543,
      "grad_norm": 1.807870864868164,
      "learning_rate": 5.125650477216688e-05,
      "loss": 0.3905,
      "step": 2715
    },
    {
      "epoch": 0.543,
      "loss_align": 0.26548415422439575,
      "loss_contrastive": 0.08456790447235107,
      "loss_gen": 0.23456498980522156,
      "neg_sim_mean": 0.7190837264060974,
      "pos_sim_mean": 0.7345158457756042,
      "step": 2715
    },
    {
      "epoch": 0.5432,
      "grad_norm": 1.833478331565857,
      "learning_rate": 5.1221608907632665e-05,
      "loss": 0.5102,
      "step": 2716
    },
    {
      "epoch": 0.5432,
      "loss_align": 0.22800028324127197,
      "loss_contrastive": 0.15359580516815186,
      "loss_gen": 0.10776704549789429,
      "neg_sim_mean": 0.825595498085022,
      "pos_sim_mean": 0.771999716758728,
      "step": 2716
    },
    {
      "epoch": 0.5434,
      "grad_norm": 1.1342191696166992,
      "learning_rate": 5.11867124477011e-05,
      "loss": 0.3542,
      "step": 2717
    },
    {
      "epoch": 0.5434,
      "loss_align": 0.09308457374572754,
      "loss_contrastive": 0.10166585445404053,
      "loss_gen": 0.24325579404830933,
      "neg_sim_mean": 0.9085812568664551,
      "pos_sim_mean": 0.9069154262542725,
      "step": 2717
    },
    {
      "epoch": 0.5436,
      "grad_norm": 2.2755343914031982,
      "learning_rate": 5.115181540938032e-05,
      "loss": 0.3485,
      "step": 2718
    },
    {
      "epoch": 0.5436,
      "loss_align": 0.09961903095245361,
      "loss_contrastive": 0.0,
      "loss_gen": 0.240853413939476,
      "neg_sim_mean": 0.4604021906852722,
      "pos_sim_mean": 0.9003809690475464,
      "step": 2718
    },
    {
      "epoch": 0.5438,
      "grad_norm": 2.013190746307373,
      "learning_rate": 5.111691780967869e-05,
      "loss": 0.3405,
      "step": 2719
    },
    {
      "epoch": 0.5438,
      "loss_align": 0.04764223098754883,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2574118673801422,
      "neg_sim_mean": 0.767946720123291,
      "pos_sim_mean": 0.9523577690124512,
      "step": 2719
    },
    {
      "epoch": 0.544,
      "grad_norm": 2.321579933166504,
      "learning_rate": 5.1082019665604895e-05,
      "loss": 0.3051,
      "step": 2720
    },
    {
      "epoch": 0.544,
      "loss_align": 0.1318226456642151,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1892867386341095,
      "neg_sim_mean": 0.7497615814208984,
      "pos_sim_mean": 0.8681773543357849,
      "step": 2720
    },
    {
      "epoch": 0.5442,
      "grad_norm": 1.6339242458343506,
      "learning_rate": 5.104712099416785e-05,
      "loss": 0.3211,
      "step": 2721
    },
    {
      "epoch": 0.5442,
      "loss_align": 0.11756503582000732,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14608418941497803,
      "neg_sim_mean": 0.7661615610122681,
      "pos_sim_mean": 0.8824349641799927,
      "step": 2721
    },
    {
      "epoch": 0.5444,
      "grad_norm": 1.5537152290344238,
      "learning_rate": 5.101222181237676e-05,
      "loss": 0.2636,
      "step": 2722
    },
    {
      "epoch": 0.5444,
      "loss_align": 0.07752346992492676,
      "loss_contrastive": 0.03359532356262207,
      "loss_gen": 0.1183570995926857,
      "neg_sim_mean": 0.8560718297958374,
      "pos_sim_mean": 0.9224765300750732,
      "step": 2722
    },
    {
      "epoch": 0.5446,
      "grad_norm": 1.1894450187683105,
      "learning_rate": 5.0977322137241066e-05,
      "loss": 0.1999,
      "step": 2723
    },
    {
      "epoch": 0.5446,
      "loss_align": 0.1307191252708435,
      "loss_contrastive": 0.10038679838180542,
      "loss_gen": 0.5035619139671326,
      "neg_sim_mean": 0.869667649269104,
      "pos_sim_mean": 0.8692808747291565,
      "step": 2723
    },
    {
      "epoch": 0.5448,
      "grad_norm": 2.6773009300231934,
      "learning_rate": 5.094242198577042e-05,
      "loss": 0.6463,
      "step": 2724
    },
    {
      "epoch": 0.5448,
      "loss_align": 0.10856235027313232,
      "loss_contrastive": 0.11358046531677246,
      "loss_gen": 0.3836471438407898,
      "neg_sim_mean": 0.9050180912017822,
      "pos_sim_mean": 0.8914376497268677,
      "step": 2724
    },
    {
      "epoch": 0.545,
      "grad_norm": 1.9763636589050293,
      "learning_rate": 5.090752137497474e-05,
      "loss": 0.5058,
      "step": 2725
    },
    {
      "epoch": 0.545,
      "loss_align": 0.13681018352508545,
      "loss_contrastive": 0.0,
      "loss_gen": 0.016820671036839485,
      "neg_sim_mean": 0.34710168838500977,
      "pos_sim_mean": 0.8631898164749146,
      "step": 2725
    },
    {
      "epoch": 0.5452,
      "grad_norm": 0.8952056765556335,
      "learning_rate": 5.0872620321864185e-05,
      "loss": 0.1536,
      "step": 2726
    },
    {
      "epoch": 0.5452,
      "loss_align": 0.19241809844970703,
      "loss_contrastive": 0.1250261664390564,
      "loss_gen": 0.2658000588417053,
      "neg_sim_mean": 0.8326080441474915,
      "pos_sim_mean": 0.807581901550293,
      "step": 2726
    },
    {
      "epoch": 0.5454,
      "grad_norm": 1.8380266427993774,
      "learning_rate": 5.0837718843449075e-05,
      "loss": 0.4732,
      "step": 2727
    },
    {
      "epoch": 0.5454,
      "loss_align": 0.16895604133605957,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011563987471163273,
      "neg_sim_mean": 0.31533879041671753,
      "pos_sim_mean": 0.8310439586639404,
      "step": 2727
    },
    {
      "epoch": 0.5456,
      "grad_norm": 0.8670591115951538,
      "learning_rate": 5.080281695673998e-05,
      "loss": 0.1805,
      "step": 2728
    },
    {
      "epoch": 0.5456,
      "loss_align": 0.12212049961090088,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2030223160982132,
      "neg_sim_mean": 0.7565688490867615,
      "pos_sim_mean": 0.8778795003890991,
      "step": 2728
    },
    {
      "epoch": 0.5458,
      "grad_norm": 2.2615675926208496,
      "learning_rate": 5.076791467874765e-05,
      "loss": 0.3251,
      "step": 2729
    },
    {
      "epoch": 0.5458,
      "loss_align": 0.14552807807922363,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1368543654680252,
      "neg_sim_mean": 0.5863628387451172,
      "pos_sim_mean": 0.8544719219207764,
      "step": 2729
    },
    {
      "epoch": 0.546,
      "grad_norm": 1.5326567888259888,
      "learning_rate": 5.073301202648304e-05,
      "loss": 0.2824,
      "step": 2730
    },
    {
      "epoch": 0.546,
      "loss_align": 0.09797072410583496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3156820833683014,
      "neg_sim_mean": 0.7147814035415649,
      "pos_sim_mean": 0.902029275894165,
      "step": 2730
    },
    {
      "epoch": 0.5462,
      "grad_norm": 2.077021598815918,
      "learning_rate": 5.0698109016957274e-05,
      "loss": 0.4137,
      "step": 2731
    },
    {
      "epoch": 0.5462,
      "loss_align": 0.1756305694580078,
      "loss_contrastive": 0.0469130277633667,
      "loss_gen": 0.4124896228313446,
      "neg_sim_mean": 0.771282434463501,
      "pos_sim_mean": 0.8243694305419922,
      "step": 2731
    },
    {
      "epoch": 0.5464,
      "grad_norm": 2.1899619102478027,
      "learning_rate": 5.066320566718165e-05,
      "loss": 0.5937,
      "step": 2732
    },
    {
      "epoch": 0.5464,
      "loss_align": 0.1145625114440918,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17055454850196838,
      "neg_sim_mean": 0.7769697308540344,
      "pos_sim_mean": 0.8854374885559082,
      "step": 2732
    },
    {
      "epoch": 0.5466,
      "grad_norm": 1.4522563219070435,
      "learning_rate": 5.062830199416764e-05,
      "loss": 0.2851,
      "step": 2733
    },
    {
      "epoch": 0.5466,
      "loss_align": 0.31136858463287354,
      "loss_contrastive": 0.05308049917221069,
      "loss_gen": 0.14415009319782257,
      "neg_sim_mean": 0.6417118906974792,
      "pos_sim_mean": 0.6886314153671265,
      "step": 2733
    },
    {
      "epoch": 0.5468,
      "grad_norm": 1.537230372428894,
      "learning_rate": 5.0593398014926866e-05,
      "loss": 0.4619,
      "step": 2734
    },
    {
      "epoch": 0.5468,
      "loss_align": 0.2287747859954834,
      "loss_contrastive": 0.09346163272857666,
      "loss_gen": 0.3489055633544922,
      "neg_sim_mean": 0.7646868228912354,
      "pos_sim_mean": 0.7712252140045166,
      "step": 2734
    },
    {
      "epoch": 0.547,
      "grad_norm": 2.1638967990875244,
      "learning_rate": 5.055849374647112e-05,
      "loss": 0.5889,
      "step": 2735
    },
    {
      "epoch": 0.547,
      "loss_align": 0.11448132991790771,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16176268458366394,
      "neg_sim_mean": 0.7484562397003174,
      "pos_sim_mean": 0.8855186700820923,
      "step": 2735
    },
    {
      "epoch": 0.5472,
      "grad_norm": 1.6314408779144287,
      "learning_rate": 5.052358920581229e-05,
      "loss": 0.2762,
      "step": 2736
    },
    {
      "epoch": 0.5472,
      "loss_align": 0.17951369285583496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.038937732577323914,
      "neg_sim_mean": 0.26938456296920776,
      "pos_sim_mean": 0.820486307144165,
      "step": 2736
    },
    {
      "epoch": 0.5474,
      "grad_norm": 1.2484090328216553,
      "learning_rate": 5.048868440996246e-05,
      "loss": 0.2185,
      "step": 2737
    },
    {
      "epoch": 0.5474,
      "loss_align": 0.1144341230392456,
      "loss_contrastive": 0.06165808439254761,
      "loss_gen": 0.18398092687129974,
      "neg_sim_mean": 0.8472239375114441,
      "pos_sim_mean": 0.8855658769607544,
      "step": 2737
    },
    {
      "epoch": 0.5476,
      "grad_norm": 1.4506992101669312,
      "learning_rate": 5.0453779375933765e-05,
      "loss": 0.3058,
      "step": 2738
    },
    {
      "epoch": 0.5476,
      "loss_align": 0.16359645128250122,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002188138896599412,
      "neg_sim_mean": 0.2968983054161072,
      "pos_sim_mean": 0.8364035487174988,
      "step": 2738
    },
    {
      "epoch": 0.5478,
      "grad_norm": 0.5681937336921692,
      "learning_rate": 5.041887412073854e-05,
      "loss": 0.1658,
      "step": 2739
    },
    {
      "epoch": 0.5478,
      "loss_align": 0.17603325843811035,
      "loss_contrastive": 0.13036751747131348,
      "loss_gen": 0.25832265615463257,
      "neg_sim_mean": 0.8543342351913452,
      "pos_sim_mean": 0.8239667415618896,
      "step": 2739
    },
    {
      "epoch": 0.548,
      "grad_norm": 2.141634225845337,
      "learning_rate": 5.0383968661389146e-05,
      "loss": 0.45,
      "step": 2740
    },
    {
      "epoch": 0.548,
      "loss_align": 0.14358770847320557,
      "loss_contrastive": 0.0553135871887207,
      "loss_gen": 0.24960015714168549,
      "neg_sim_mean": 0.8117258548736572,
      "pos_sim_mean": 0.8564122915267944,
      "step": 2740
    },
    {
      "epoch": 0.5482,
      "grad_norm": 1.8015207052230835,
      "learning_rate": 5.034906301489808e-05,
      "loss": 0.3998,
      "step": 2741
    },
    {
      "epoch": 0.5482,
      "loss_align": 0.09922462701797485,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11826194822788239,
      "neg_sim_mean": 0.8000879287719727,
      "pos_sim_mean": 0.9007753729820251,
      "step": 2741
    },
    {
      "epoch": 0.5484,
      "grad_norm": 1.4785470962524414,
      "learning_rate": 5.0314157198277954e-05,
      "loss": 0.2175,
      "step": 2742
    },
    {
      "epoch": 0.5484,
      "loss_align": 0.23490959405899048,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23890337347984314,
      "neg_sim_mean": 0.6209625005722046,
      "pos_sim_mean": 0.7650904059410095,
      "step": 2742
    },
    {
      "epoch": 0.5486,
      "grad_norm": 2.1820337772369385,
      "learning_rate": 5.027925122854141e-05,
      "loss": 0.4738,
      "step": 2743
    },
    {
      "epoch": 0.5486,
      "loss_align": 0.22386735677719116,
      "loss_contrastive": 0.0,
      "loss_gen": 0.026663396507501602,
      "neg_sim_mean": 0.4720209836959839,
      "pos_sim_mean": 0.7761326432228088,
      "step": 2743
    },
    {
      "epoch": 0.5488,
      "grad_norm": 0.9764576554298401,
      "learning_rate": 5.024434512270123e-05,
      "loss": 0.2505,
      "step": 2744
    },
    {
      "epoch": 0.5488,
      "loss_align": 0.11080867052078247,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0016681697452440858,
      "neg_sim_mean": 0.37075552344322205,
      "pos_sim_mean": 0.8891913294792175,
      "step": 2744
    },
    {
      "epoch": 0.549,
      "grad_norm": 0.4545491933822632,
      "learning_rate": 5.02094388977702e-05,
      "loss": 0.1125,
      "step": 2745
    },
    {
      "epoch": 0.549,
      "loss_align": 0.1649072766304016,
      "loss_contrastive": 0.08962547779083252,
      "loss_gen": 0.29302889108657837,
      "neg_sim_mean": 0.824718177318573,
      "pos_sim_mean": 0.8350927233695984,
      "step": 2745
    },
    {
      "epoch": 0.5492,
      "grad_norm": 1.7475104331970215,
      "learning_rate": 5.017453257076119e-05,
      "loss": 0.4687,
      "step": 2746
    },
    {
      "epoch": 0.5492,
      "loss_align": 0.15115797519683838,
      "loss_contrastive": 0.003695845603942871,
      "loss_gen": 0.18599410355091095,
      "neg_sim_mean": 0.7525378465652466,
      "pos_sim_mean": 0.8488420248031616,
      "step": 2746
    },
    {
      "epoch": 0.5494,
      "grad_norm": 1.74275541305542,
      "learning_rate": 5.013962615868714e-05,
      "loss": 0.3376,
      "step": 2747
    },
    {
      "epoch": 0.5494,
      "loss_align": 0.1466052532196045,
      "loss_contrastive": 0.03509652614593506,
      "loss_gen": 0.13113170862197876,
      "neg_sim_mean": 0.7884912490844727,
      "pos_sim_mean": 0.8533947467803955,
      "step": 2747
    },
    {
      "epoch": 0.5496,
      "grad_norm": 1.1099361181259155,
      "learning_rate": 5.010471967856096e-05,
      "loss": 0.2819,
      "step": 2748
    },
    {
      "epoch": 0.5496,
      "loss_align": 0.07954525947570801,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1296410858631134,
      "neg_sim_mean": 0.5822263956069946,
      "pos_sim_mean": 0.920454740524292,
      "step": 2748
    },
    {
      "epoch": 0.5498,
      "grad_norm": 1.374294400215149,
      "learning_rate": 5.006981314739573e-05,
      "loss": 0.2092,
      "step": 2749
    },
    {
      "epoch": 0.5498,
      "loss_align": 0.1465383768081665,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18357880413532257,
      "neg_sim_mean": 0.6765458583831787,
      "pos_sim_mean": 0.8534616231918335,
      "step": 2749
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.6597861051559448,
      "learning_rate": 5.003490658220438e-05,
      "loss": 0.3301,
      "step": 2750
    },
    {
      "epoch": 0.55,
      "loss_align": 0.10498964786529541,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25379493832588196,
      "neg_sim_mean": 0.7365992069244385,
      "pos_sim_mean": 0.8950103521347046,
      "step": 2750
    },
    {
      "epoch": 0.5502,
      "grad_norm": 1.9106520414352417,
      "learning_rate": 5e-05,
      "loss": 0.3588,
      "step": 2751
    },
    {
      "epoch": 0.5502,
      "loss_align": 0.17092227935791016,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0009964235359802842,
      "neg_sim_mean": 0.3695247173309326,
      "pos_sim_mean": 0.8290777206420898,
      "step": 2751
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.5581628084182739,
      "learning_rate": 4.996509341779563e-05,
      "loss": 0.1719,
      "step": 2752
    },
    {
      "epoch": 0.5504,
      "loss_align": 0.19461572170257568,
      "loss_contrastive": 0.12213492393493652,
      "loss_gen": 0.26527297496795654,
      "neg_sim_mean": 0.8275191783905029,
      "pos_sim_mean": 0.8053842782974243,
      "step": 2752
    },
    {
      "epoch": 0.5506,
      "grad_norm": 2.2760093212127686,
      "learning_rate": 4.993018685260428e-05,
      "loss": 0.4745,
      "step": 2753
    },
    {
      "epoch": 0.5506,
      "loss_align": 0.1437474489212036,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004964910447597504,
      "neg_sim_mean": 0.458869993686676,
      "pos_sim_mean": 0.8562525510787964,
      "step": 2753
    },
    {
      "epoch": 0.5508,
      "grad_norm": 0.5486909747123718,
      "learning_rate": 4.989528032143903e-05,
      "loss": 0.1487,
      "step": 2754
    },
    {
      "epoch": 0.5508,
      "loss_align": 0.1134108304977417,
      "loss_contrastive": 0.09637236595153809,
      "loss_gen": 0.21803000569343567,
      "neg_sim_mean": 0.8829615116119385,
      "pos_sim_mean": 0.8865891695022583,
      "step": 2754
    },
    {
      "epoch": 0.551,
      "grad_norm": 1.653968334197998,
      "learning_rate": 4.986037384131288e-05,
      "loss": 0.343,
      "step": 2755
    },
    {
      "epoch": 0.551,
      "loss_align": 0.2055683135986328,
      "loss_contrastive": 0.08315622806549072,
      "loss_gen": 0.24145303666591644,
      "neg_sim_mean": 0.777587890625,
      "pos_sim_mean": 0.7944316864013672,
      "step": 2755
    },
    {
      "epoch": 0.5512,
      "grad_norm": 1.898584008216858,
      "learning_rate": 4.9825467429238834e-05,
      "loss": 0.457,
      "step": 2756
    },
    {
      "epoch": 0.5512,
      "loss_align": 0.20343607664108276,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09354093670845032,
      "neg_sim_mean": 0.6603685617446899,
      "pos_sim_mean": 0.7965639233589172,
      "step": 2756
    },
    {
      "epoch": 0.5514,
      "grad_norm": 1.3867281675338745,
      "learning_rate": 4.979056110222981e-05,
      "loss": 0.297,
      "step": 2757
    },
    {
      "epoch": 0.5514,
      "loss_align": 0.13846343755722046,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04349179193377495,
      "neg_sim_mean": 0.43371814489364624,
      "pos_sim_mean": 0.8615365624427795,
      "step": 2757
    },
    {
      "epoch": 0.5516,
      "grad_norm": 1.4808822870254517,
      "learning_rate": 4.9755654877298786e-05,
      "loss": 0.182,
      "step": 2758
    },
    {
      "epoch": 0.5516,
      "loss_align": 0.16518127918243408,
      "loss_contrastive": 0.03988838195800781,
      "loss_gen": 0.198199063539505,
      "neg_sim_mean": 0.7747070789337158,
      "pos_sim_mean": 0.8348187208175659,
      "step": 2758
    },
    {
      "epoch": 0.5518,
      "grad_norm": 1.9298717975616455,
      "learning_rate": 4.97207487714586e-05,
      "loss": 0.3682,
      "step": 2759
    },
    {
      "epoch": 0.5518,
      "loss_align": 0.1506863832473755,
      "loss_contrastive": 0.07884854078292847,
      "loss_gen": 0.1795637458562851,
      "neg_sim_mean": 0.8281621336936951,
      "pos_sim_mean": 0.8493136167526245,
      "step": 2759
    },
    {
      "epoch": 0.552,
      "grad_norm": 1.7273385524749756,
      "learning_rate": 4.968584280172206e-05,
      "loss": 0.3397,
      "step": 2760
    },
    {
      "epoch": 0.552,
      "loss_align": 0.17226475477218628,
      "loss_contrastive": 0.025569021701812744,
      "loss_gen": 0.1508112996816635,
      "neg_sim_mean": 0.7533042430877686,
      "pos_sim_mean": 0.8277352452278137,
      "step": 2760
    },
    {
      "epoch": 0.5522,
      "grad_norm": 1.289482593536377,
      "learning_rate": 4.965093698510193e-05,
      "loss": 0.3261,
      "step": 2761
    },
    {
      "epoch": 0.5522,
      "loss_align": 0.12149453163146973,
      "loss_contrastive": 0.0,
      "loss_gen": 0.020796649158000946,
      "neg_sim_mean": 0.4570668935775757,
      "pos_sim_mean": 0.8785054683685303,
      "step": 2761
    },
    {
      "epoch": 0.5524,
      "grad_norm": 0.9904000759124756,
      "learning_rate": 4.9616031338610866e-05,
      "loss": 0.1423,
      "step": 2762
    },
    {
      "epoch": 0.5524,
      "loss_align": 0.2256321907043457,
      "loss_contrastive": 0.0,
      "loss_gen": 0.029490087181329727,
      "neg_sim_mean": 0.33404216170310974,
      "pos_sim_mean": 0.7743678092956543,
      "step": 2762
    },
    {
      "epoch": 0.5526,
      "grad_norm": 1.2598700523376465,
      "learning_rate": 4.958112587926147e-05,
      "loss": 0.2551,
      "step": 2763
    },
    {
      "epoch": 0.5526,
      "loss_align": 0.1671549677848816,
      "loss_contrastive": 0.11684244871139526,
      "loss_gen": 0.3192521631717682,
      "neg_sim_mean": 0.8496874570846558,
      "pos_sim_mean": 0.8328450322151184,
      "step": 2763
    },
    {
      "epoch": 0.5528,
      "grad_norm": 1.7598298788070679,
      "learning_rate": 4.954622062406623e-05,
      "loss": 0.5004,
      "step": 2764
    },
    {
      "epoch": 0.5528,
      "loss_align": 0.07117366790771484,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001331679173745215,
      "neg_sim_mean": 0.3566698431968689,
      "pos_sim_mean": 0.9288263320922852,
      "step": 2764
    },
    {
      "epoch": 0.553,
      "grad_norm": 0.3205520510673523,
      "learning_rate": 4.951131559003756e-05,
      "loss": 0.0725,
      "step": 2765
    },
    {
      "epoch": 0.553,
      "loss_align": 0.10614889860153198,
      "loss_contrastive": 0.0882992148399353,
      "loss_gen": 0.3431870639324188,
      "neg_sim_mean": 0.8821502923965454,
      "pos_sim_mean": 0.893851101398468,
      "step": 2765
    },
    {
      "epoch": 0.5532,
      "grad_norm": 2.1995294094085693,
      "learning_rate": 4.947641079418773e-05,
      "loss": 0.4599,
      "step": 2766
    },
    {
      "epoch": 0.5532,
      "loss_align": 0.1296771764755249,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15585742890834808,
      "neg_sim_mean": 0.6693854331970215,
      "pos_sim_mean": 0.8703228235244751,
      "step": 2766
    },
    {
      "epoch": 0.5534,
      "grad_norm": 1.5768505334854126,
      "learning_rate": 4.9441506253528895e-05,
      "loss": 0.2855,
      "step": 2767
    },
    {
      "epoch": 0.5534,
      "loss_align": 0.15567052364349365,
      "loss_contrastive": 0.07308065891265869,
      "loss_gen": 0.15007270872592926,
      "neg_sim_mean": 0.8174101114273071,
      "pos_sim_mean": 0.8443294763565063,
      "step": 2767
    },
    {
      "epoch": 0.5536,
      "grad_norm": 1.5417487621307373,
      "learning_rate": 4.9406601985073146e-05,
      "loss": 0.3145,
      "step": 2768
    },
    {
      "epoch": 0.5536,
      "loss_align": 0.14453601837158203,
      "loss_contrastive": 0.030903220176696777,
      "loss_gen": 0.18680432438850403,
      "neg_sim_mean": 0.7863671779632568,
      "pos_sim_mean": 0.855463981628418,
      "step": 2768
    },
    {
      "epoch": 0.5538,
      "grad_norm": 2.11120343208313,
      "learning_rate": 4.9371698005832365e-05,
      "loss": 0.335,
      "step": 2769
    },
    {
      "epoch": 0.5538,
      "loss_align": 0.09912121295928955,
      "loss_contrastive": 0.024320602416992188,
      "loss_gen": 0.15689219534397125,
      "neg_sim_mean": 0.8251993656158447,
      "pos_sim_mean": 0.9008787870407104,
      "step": 2769
    },
    {
      "epoch": 0.554,
      "grad_norm": 1.4502110481262207,
      "learning_rate": 4.933679433281836e-05,
      "loss": 0.2589,
      "step": 2770
    },
    {
      "epoch": 0.554,
      "loss_align": 0.18348956108093262,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0007684117881581187,
      "neg_sim_mean": 0.4750640392303467,
      "pos_sim_mean": 0.8165104389190674,
      "step": 2770
    },
    {
      "epoch": 0.5542,
      "grad_norm": 0.6348202228546143,
      "learning_rate": 4.9301890983042744e-05,
      "loss": 0.1843,
      "step": 2771
    },
    {
      "epoch": 0.5542,
      "loss_align": 0.08191907405853271,
      "loss_contrastive": 0.05451691150665283,
      "loss_gen": 0.16321878135204315,
      "neg_sim_mean": 0.8725978136062622,
      "pos_sim_mean": 0.9180809259414673,
      "step": 2771
    },
    {
      "epoch": 0.5544,
      "grad_norm": 1.4198191165924072,
      "learning_rate": 4.926698797351697e-05,
      "loss": 0.2517,
      "step": 2772
    },
    {
      "epoch": 0.5544,
      "loss_align": 0.2241363525390625,
      "loss_contrastive": 0.09682631492614746,
      "loss_gen": 0.29472285509109497,
      "neg_sim_mean": 0.772689938545227,
      "pos_sim_mean": 0.7758636474609375,
      "step": 2772
    },
    {
      "epoch": 0.5546,
      "grad_norm": 2.1018152236938477,
      "learning_rate": 4.9232085321252354e-05,
      "loss": 0.5305,
      "step": 2773
    },
    {
      "epoch": 0.5546,
      "loss_align": 0.17400115728378296,
      "loss_contrastive": 0.10455542802810669,
      "loss_gen": 0.20917761325836182,
      "neg_sim_mean": 0.8305542469024658,
      "pos_sim_mean": 0.825998842716217,
      "step": 2773
    },
    {
      "epoch": 0.5548,
      "grad_norm": 1.8341883420944214,
      "learning_rate": 4.919718304326004e-05,
      "loss": 0.3957,
      "step": 2774
    },
    {
      "epoch": 0.5548,
      "loss_align": 0.18488001823425293,
      "loss_contrastive": 0.033532142639160156,
      "loss_gen": 0.7608080506324768,
      "neg_sim_mean": 0.7486521005630493,
      "pos_sim_mean": 0.8151199817657471,
      "step": 2774
    },
    {
      "epoch": 0.555,
      "grad_norm": 3.102931022644043,
      "learning_rate": 4.916228115655094e-05,
      "loss": 0.9497,
      "step": 2775
    },
    {
      "epoch": 0.555,
      "loss_align": 0.1311638355255127,
      "loss_contrastive": 0.04533511400222778,
      "loss_gen": 0.34517166018486023,
      "neg_sim_mean": 0.8141712546348572,
      "pos_sim_mean": 0.8688361644744873,
      "step": 2775
    },
    {
      "epoch": 0.5552,
      "grad_norm": 2.1121695041656494,
      "learning_rate": 4.912737967813583e-05,
      "loss": 0.4818,
      "step": 2776
    },
    {
      "epoch": 0.5552,
      "loss_align": 0.13537567853927612,
      "loss_contrastive": 0.10299068689346313,
      "loss_gen": 0.2780408263206482,
      "neg_sim_mean": 0.8676149845123291,
      "pos_sim_mean": 0.8646243214607239,
      "step": 2776
    },
    {
      "epoch": 0.5554,
      "grad_norm": 2.137882709503174,
      "learning_rate": 4.909247862502527e-05,
      "loss": 0.4258,
      "step": 2777
    },
    {
      "epoch": 0.5554,
      "loss_align": 0.5361131429672241,
      "loss_contrastive": 0.393909752368927,
      "loss_gen": 0.14876806735992432,
      "neg_sim_mean": 0.7577966451644897,
      "pos_sim_mean": 0.46388688683509827,
      "step": 2777
    },
    {
      "epoch": 0.5556,
      "grad_norm": 2.054673194885254,
      "learning_rate": 4.90575780142296e-05,
      "loss": 0.7322,
      "step": 2778
    },
    {
      "epoch": 0.5556,
      "loss_align": 0.13797348737716675,
      "loss_contrastive": 0.11388802528381348,
      "loss_gen": 0.28066205978393555,
      "neg_sim_mean": 0.8759145140647888,
      "pos_sim_mean": 0.8620265126228333,
      "step": 2778
    },
    {
      "epoch": 0.5558,
      "grad_norm": 1.902287483215332,
      "learning_rate": 4.9022677862758945e-05,
      "loss": 0.4323,
      "step": 2779
    },
    {
      "epoch": 0.5558,
      "loss_align": 0.23926931619644165,
      "loss_contrastive": 0.1764240860939026,
      "loss_gen": 0.26965823769569397,
      "neg_sim_mean": 0.837154746055603,
      "pos_sim_mean": 0.7607306838035583,
      "step": 2779
    },
    {
      "epoch": 0.556,
      "grad_norm": 2.1936020851135254,
      "learning_rate": 4.8987778187623245e-05,
      "loss": 0.5301,
      "step": 2780
    },
    {
      "epoch": 0.556,
      "loss_align": 0.18733203411102295,
      "loss_contrastive": 0.11286407709121704,
      "loss_gen": 0.27939969301223755,
      "neg_sim_mean": 0.8255320191383362,
      "pos_sim_mean": 0.812667965888977,
      "step": 2780
    },
    {
      "epoch": 0.5562,
      "grad_norm": 1.9670488834381104,
      "learning_rate": 4.895287900583216e-05,
      "loss": 0.4803,
      "step": 2781
    },
    {
      "epoch": 0.5562,
      "loss_align": 0.1955662965774536,
      "loss_contrastive": 0.07461833953857422,
      "loss_gen": 0.2082948535680771,
      "neg_sim_mean": 0.7790520191192627,
      "pos_sim_mean": 0.8044337034225464,
      "step": 2781
    },
    {
      "epoch": 0.5564,
      "grad_norm": 1.9109222888946533,
      "learning_rate": 4.891798033439511e-05,
      "loss": 0.4128,
      "step": 2782
    },
    {
      "epoch": 0.5564,
      "loss_align": 0.158355712890625,
      "loss_contrastive": 0.08798837661743164,
      "loss_gen": 0.25215232372283936,
      "neg_sim_mean": 0.8296326398849487,
      "pos_sim_mean": 0.841644287109375,
      "step": 2782
    },
    {
      "epoch": 0.5566,
      "grad_norm": 1.841680645942688,
      "learning_rate": 4.888308219032132e-05,
      "loss": 0.4211,
      "step": 2783
    },
    {
      "epoch": 0.5566,
      "loss_align": 0.19893646240234375,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3298119008541107,
      "neg_sim_mean": 0.6949924230575562,
      "pos_sim_mean": 0.8010635375976562,
      "step": 2783
    },
    {
      "epoch": 0.5568,
      "grad_norm": 1.8652411699295044,
      "learning_rate": 4.88481845906197e-05,
      "loss": 0.5287,
      "step": 2784
    },
    {
      "epoch": 0.5568,
      "loss_align": 0.11619365215301514,
      "loss_contrastive": 0.07020342350006104,
      "loss_gen": 0.18989181518554688,
      "neg_sim_mean": 0.854009747505188,
      "pos_sim_mean": 0.8838063478469849,
      "step": 2784
    },
    {
      "epoch": 0.557,
      "grad_norm": 1.9435017108917236,
      "learning_rate": 4.8813287552298916e-05,
      "loss": 0.3145,
      "step": 2785
    },
    {
      "epoch": 0.557,
      "loss_align": 0.1908506155014038,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19627472758293152,
      "neg_sim_mean": 0.642524003982544,
      "pos_sim_mean": 0.8091493844985962,
      "step": 2785
    },
    {
      "epoch": 0.5572,
      "grad_norm": 1.9638113975524902,
      "learning_rate": 4.8778391092367346e-05,
      "loss": 0.3871,
      "step": 2786
    },
    {
      "epoch": 0.5572,
      "loss_align": 0.09317493438720703,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3222600519657135,
      "neg_sim_mean": 0.7352430820465088,
      "pos_sim_mean": 0.906825065612793,
      "step": 2786
    },
    {
      "epoch": 0.5574,
      "grad_norm": 2.2503864765167236,
      "learning_rate": 4.874349522783313e-05,
      "loss": 0.4154,
      "step": 2787
    },
    {
      "epoch": 0.5574,
      "loss_align": 0.14733999967575073,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0058901929296553135,
      "neg_sim_mean": 0.49705371260643005,
      "pos_sim_mean": 0.8526600003242493,
      "step": 2787
    },
    {
      "epoch": 0.5576,
      "grad_norm": 0.6536923050880432,
      "learning_rate": 4.870859997570407e-05,
      "loss": 0.1532,
      "step": 2788
    },
    {
      "epoch": 0.5576,
      "loss_align": 0.11208856105804443,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3315238356590271,
      "neg_sim_mean": 0.6616241335868835,
      "pos_sim_mean": 0.8879114389419556,
      "step": 2788
    },
    {
      "epoch": 0.5578,
      "grad_norm": 2.564601182937622,
      "learning_rate": 4.8673705352987694e-05,
      "loss": 0.4436,
      "step": 2789
    },
    {
      "epoch": 0.5578,
      "loss_align": 0.22122114896774292,
      "loss_contrastive": 0.05723446607589722,
      "loss_gen": 0.34866631031036377,
      "neg_sim_mean": 0.7360132932662964,
      "pos_sim_mean": 0.7787788510322571,
      "step": 2789
    },
    {
      "epoch": 0.558,
      "grad_norm": 2.2467987537384033,
      "learning_rate": 4.863881137669123e-05,
      "loss": 0.5768,
      "step": 2790
    },
    {
      "epoch": 0.558,
      "loss_align": 0.15105515718460083,
      "loss_contrastive": 0.0325809121131897,
      "loss_gen": 0.15157319605350494,
      "neg_sim_mean": 0.781525731086731,
      "pos_sim_mean": 0.8489448428153992,
      "step": 2790
    },
    {
      "epoch": 0.5582,
      "grad_norm": 1.665103793144226,
      "learning_rate": 4.860391806382157e-05,
      "loss": 0.3065,
      "step": 2791
    },
    {
      "epoch": 0.5582,
      "loss_align": 0.2590233087539673,
      "loss_contrastive": 0.13431406021118164,
      "loss_gen": 0.23986314237117767,
      "neg_sim_mean": 0.7752907276153564,
      "pos_sim_mean": 0.7409766912460327,
      "step": 2791
    },
    {
      "epoch": 0.5584,
      "grad_norm": 2.6505126953125,
      "learning_rate": 4.856902543138527e-05,
      "loss": 0.515,
      "step": 2792
    },
    {
      "epoch": 0.5584,
      "loss_align": 0.22176021337509155,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0043590012937784195,
      "neg_sim_mean": 0.2815009355545044,
      "pos_sim_mean": 0.7782397866249084,
      "step": 2792
    },
    {
      "epoch": 0.5586,
      "grad_norm": 0.5436860918998718,
      "learning_rate": 4.853413349638859e-05,
      "loss": 0.2261,
      "step": 2793
    },
    {
      "epoch": 0.5586,
      "loss_align": 0.1441659927368164,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010883846320211887,
      "neg_sim_mean": 0.2765774428844452,
      "pos_sim_mean": 0.8558340072631836,
      "step": 2793
    },
    {
      "epoch": 0.5588,
      "grad_norm": 0.7640896439552307,
      "learning_rate": 4.8499242275837444e-05,
      "loss": 0.155,
      "step": 2794
    },
    {
      "epoch": 0.5588,
      "loss_align": 0.09377777576446533,
      "loss_contrastive": 0.047579407691955566,
      "loss_gen": 0.16953535377979279,
      "neg_sim_mean": 0.8538016080856323,
      "pos_sim_mean": 0.9062222242355347,
      "step": 2794
    },
    {
      "epoch": 0.559,
      "grad_norm": 1.834154725074768,
      "learning_rate": 4.846435178673736e-05,
      "loss": 0.269,
      "step": 2795
    },
    {
      "epoch": 0.559,
      "loss_align": 0.16188502311706543,
      "loss_contrastive": 0.022538423538208008,
      "loss_gen": 0.26670315861701965,
      "neg_sim_mean": 0.7606533765792847,
      "pos_sim_mean": 0.8381149768829346,
      "step": 2795
    },
    {
      "epoch": 0.5592,
      "grad_norm": 1.723173975944519,
      "learning_rate": 4.8429462046093585e-05,
      "loss": 0.4313,
      "step": 2796
    },
    {
      "epoch": 0.5592,
      "loss_align": 0.17081785202026367,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01859365962445736,
      "neg_sim_mean": 0.500198245048523,
      "pos_sim_mean": 0.8291821479797363,
      "step": 2796
    },
    {
      "epoch": 0.5594,
      "grad_norm": 0.9559557437896729,
      "learning_rate": 4.839457307091093e-05,
      "loss": 0.1894,
      "step": 2797
    },
    {
      "epoch": 0.5594,
      "loss_align": 0.18455004692077637,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17849251627922058,
      "neg_sim_mean": 0.6907068490982056,
      "pos_sim_mean": 0.8154499530792236,
      "step": 2797
    },
    {
      "epoch": 0.5596,
      "grad_norm": 1.5112853050231934,
      "learning_rate": 4.835968487819384e-05,
      "loss": 0.363,
      "step": 2798
    },
    {
      "epoch": 0.5596,
      "loss_align": 0.17603802680969238,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1844293177127838,
      "neg_sim_mean": 0.7188267707824707,
      "pos_sim_mean": 0.8239619731903076,
      "step": 2798
    },
    {
      "epoch": 0.5598,
      "grad_norm": 2.0446062088012695,
      "learning_rate": 4.832479748494643e-05,
      "loss": 0.3605,
      "step": 2799
    },
    {
      "epoch": 0.5598,
      "loss_align": 0.09774327278137207,
      "loss_contrastive": 0.0,
      "loss_gen": 0.030493449419736862,
      "neg_sim_mean": 0.417580246925354,
      "pos_sim_mean": 0.9022567272186279,
      "step": 2799
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.2374595403671265,
      "learning_rate": 4.8289910908172376e-05,
      "loss": 0.1282,
      "step": 2800
    },
    {
      "epoch": 0.56,
      "loss_align": 0.18158364295959473,
      "loss_contrastive": 0.1060633659362793,
      "loss_gen": 0.24276745319366455,
      "neg_sim_mean": 0.8244796991348267,
      "pos_sim_mean": 0.8184163570404053,
      "step": 2800
    },
    {
      "epoch": 0.5602,
      "grad_norm": 2.080686092376709,
      "learning_rate": 4.825502516487497e-05,
      "loss": 0.4371,
      "step": 2801
    },
    {
      "epoch": 0.5602,
      "loss_align": 0.2390344738960266,
      "loss_contrastive": 0.10824257135391235,
      "loss_gen": 0.3962920904159546,
      "neg_sim_mean": 0.7692080736160278,
      "pos_sim_mean": 0.7609655261039734,
      "step": 2801
    },
    {
      "epoch": 0.5604,
      "grad_norm": 2.370454788208008,
      "learning_rate": 4.822014027205708e-05,
      "loss": 0.6483,
      "step": 2802
    },
    {
      "epoch": 0.5604,
      "loss_align": 0.12093937397003174,
      "loss_contrastive": 0.029818713665008545,
      "loss_gen": 0.19463174045085907,
      "neg_sim_mean": 0.8088793158531189,
      "pos_sim_mean": 0.8790606260299683,
      "step": 2802
    },
    {
      "epoch": 0.5606,
      "grad_norm": 1.7131050825119019,
      "learning_rate": 4.8185256246721216e-05,
      "loss": 0.3191,
      "step": 2803
    },
    {
      "epoch": 0.5606,
      "loss_align": 0.15128785371780396,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09633444249629974,
      "neg_sim_mean": 0.7102939486503601,
      "pos_sim_mean": 0.848712146282196,
      "step": 2803
    },
    {
      "epoch": 0.5608,
      "grad_norm": 1.1646009683609009,
      "learning_rate": 4.815037310586941e-05,
      "loss": 0.2476,
      "step": 2804
    },
    {
      "epoch": 0.5608,
      "loss_align": 0.0964055061340332,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006574952509254217,
      "neg_sim_mean": 0.32994604110717773,
      "pos_sim_mean": 0.9035944938659668,
      "step": 2804
    },
    {
      "epoch": 0.561,
      "grad_norm": 0.5425634980201721,
      "learning_rate": 4.811549086650327e-05,
      "loss": 0.103,
      "step": 2805
    },
    {
      "epoch": 0.561,
      "loss_align": 0.13581275939941406,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013984821736812592,
      "neg_sim_mean": 0.2593018412590027,
      "pos_sim_mean": 0.8641872406005859,
      "step": 2805
    },
    {
      "epoch": 0.5612,
      "grad_norm": 0.7373700141906738,
      "learning_rate": 4.8080609545624004e-05,
      "loss": 0.1498,
      "step": 2806
    },
    {
      "epoch": 0.5612,
      "loss_align": 0.10406696796417236,
      "loss_contrastive": 0.07075119018554688,
      "loss_gen": 0.23482070863246918,
      "neg_sim_mean": 0.8666841983795166,
      "pos_sim_mean": 0.8959330320358276,
      "step": 2806
    },
    {
      "epoch": 0.5614,
      "grad_norm": 1.778591275215149,
      "learning_rate": 4.804572916023233e-05,
      "loss": 0.3474,
      "step": 2807
    },
    {
      "epoch": 0.5614,
      "loss_align": 0.2996305823326111,
      "loss_contrastive": 0.11976087093353271,
      "loss_gen": 0.34328845143318176,
      "neg_sim_mean": 0.7201302647590637,
      "pos_sim_mean": 0.7003694176673889,
      "step": 2807
    },
    {
      "epoch": 0.5616,
      "grad_norm": 2.192110300064087,
      "learning_rate": 4.801084972732851e-05,
      "loss": 0.6573,
      "step": 2808
    },
    {
      "epoch": 0.5616,
      "loss_align": 0.14760833978652954,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13716264069080353,
      "neg_sim_mean": 0.7323167324066162,
      "pos_sim_mean": 0.8523916602134705,
      "step": 2808
    },
    {
      "epoch": 0.5618,
      "grad_norm": 1.3749674558639526,
      "learning_rate": 4.797597126391238e-05,
      "loss": 0.2848,
      "step": 2809
    },
    {
      "epoch": 0.5618,
      "loss_align": 0.1393783688545227,
      "loss_contrastive": 0.13875246047973633,
      "loss_gen": 0.47577202320098877,
      "neg_sim_mean": 0.8993740677833557,
      "pos_sim_mean": 0.8606216311454773,
      "step": 2809
    },
    {
      "epoch": 0.562,
      "grad_norm": 2.458113431930542,
      "learning_rate": 4.794109378698327e-05,
      "loss": 0.6318,
      "step": 2810
    },
    {
      "epoch": 0.562,
      "loss_align": 0.13149148225784302,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3933923542499542,
      "neg_sim_mean": 0.6523954272270203,
      "pos_sim_mean": 0.868508517742157,
      "step": 2810
    },
    {
      "epoch": 0.5622,
      "grad_norm": 1.8665478229522705,
      "learning_rate": 4.790621731354003e-05,
      "loss": 0.5249,
      "step": 2811
    },
    {
      "epoch": 0.5622,
      "loss_align": 0.18866056203842163,
      "loss_contrastive": 0.13667577505111694,
      "loss_gen": 0.5263465642929077,
      "neg_sim_mean": 0.8480151891708374,
      "pos_sim_mean": 0.8113394379615784,
      "step": 2811
    },
    {
      "epoch": 0.5624,
      "grad_norm": 2.337714195251465,
      "learning_rate": 4.7871341860581025e-05,
      "loss": 0.7314,
      "step": 2812
    },
    {
      "epoch": 0.5624,
      "loss_align": 0.08065474033355713,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2666231393814087,
      "neg_sim_mean": 0.5713340044021606,
      "pos_sim_mean": 0.9193452596664429,
      "step": 2812
    },
    {
      "epoch": 0.5626,
      "grad_norm": 1.7539259195327759,
      "learning_rate": 4.783646744510416e-05,
      "loss": 0.3473,
      "step": 2813
    },
    {
      "epoch": 0.5626,
      "loss_align": 0.0872427225112915,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16053028404712677,
      "neg_sim_mean": 0.5480160117149353,
      "pos_sim_mean": 0.9127572774887085,
      "step": 2813
    },
    {
      "epoch": 0.5628,
      "grad_norm": 1.2482739686965942,
      "learning_rate": 4.7801594084106763e-05,
      "loss": 0.2478,
      "step": 2814
    },
    {
      "epoch": 0.5628,
      "loss_align": 0.20031172037124634,
      "loss_contrastive": 0.10716843605041504,
      "loss_gen": 0.20454907417297363,
      "neg_sim_mean": 0.8068566918373108,
      "pos_sim_mean": 0.7996882796287537,
      "step": 2814
    },
    {
      "epoch": 0.563,
      "grad_norm": 1.3747549057006836,
      "learning_rate": 4.77667217945857e-05,
      "loss": 0.4177,
      "step": 2815
    },
    {
      "epoch": 0.563,
      "loss_align": 0.05514967441558838,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2536981999874115,
      "neg_sim_mean": 0.6909192800521851,
      "pos_sim_mean": 0.9448503255844116,
      "step": 2815
    },
    {
      "epoch": 0.5632,
      "grad_norm": 1.795030117034912,
      "learning_rate": 4.773185059353732e-05,
      "loss": 0.3088,
      "step": 2816
    },
    {
      "epoch": 0.5632,
      "loss_align": 0.13460123538970947,
      "loss_contrastive": 0.08930337429046631,
      "loss_gen": 0.2224356234073639,
      "neg_sim_mean": 0.8547021150588989,
      "pos_sim_mean": 0.8653987646102905,
      "step": 2816
    },
    {
      "epoch": 0.5634,
      "grad_norm": 1.5218051671981812,
      "learning_rate": 4.769698049795738e-05,
      "loss": 0.3678,
      "step": 2817
    },
    {
      "epoch": 0.5634,
      "loss_align": 0.1475815773010254,
      "loss_contrastive": 0.0,
      "loss_gen": 0.40547776222229004,
      "neg_sim_mean": 0.6231275796890259,
      "pos_sim_mean": 0.8524184226989746,
      "step": 2817
    },
    {
      "epoch": 0.5636,
      "grad_norm": 2.484821319580078,
      "learning_rate": 4.7662111524841213e-05,
      "loss": 0.5531,
      "step": 2818
    },
    {
      "epoch": 0.5636,
      "loss_align": 0.10962474346160889,
      "loss_contrastive": 0.002787172794342041,
      "loss_gen": 0.3835071623325348,
      "neg_sim_mean": 0.7931624054908752,
      "pos_sim_mean": 0.8903752565383911,
      "step": 2818
    },
    {
      "epoch": 0.5638,
      "grad_norm": 2.2180016040802,
      "learning_rate": 4.7627243691183453e-05,
      "loss": 0.4935,
      "step": 2819
    },
    {
      "epoch": 0.5638,
      "loss_align": 0.1197238564491272,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01643706113100052,
      "neg_sim_mean": 0.37865930795669556,
      "pos_sim_mean": 0.8802761435508728,
      "step": 2819
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.7951399087905884,
      "learning_rate": 4.7592377013978306e-05,
      "loss": 0.1362,
      "step": 2820
    },
    {
      "epoch": 0.564,
      "loss_align": 0.08166950941085815,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009526987560093403,
      "neg_sim_mean": 0.428605318069458,
      "pos_sim_mean": 0.9183304905891418,
      "step": 2820
    },
    {
      "epoch": 0.5642,
      "grad_norm": 0.7285760641098022,
      "learning_rate": 4.755751151021934e-05,
      "loss": 0.0912,
      "step": 2821
    },
    {
      "epoch": 0.5642,
      "loss_align": 0.2321944236755371,
      "loss_contrastive": 0.08939254283905029,
      "loss_gen": 0.16806340217590332,
      "neg_sim_mean": 0.7571980953216553,
      "pos_sim_mean": 0.7678055763244629,
      "step": 2821
    },
    {
      "epoch": 0.5644,
      "grad_norm": 1.6376383304595947,
      "learning_rate": 4.7522647196899605e-05,
      "loss": 0.411,
      "step": 2822
    },
    {
      "epoch": 0.5644,
      "loss_align": 0.11403286457061768,
      "loss_contrastive": 0.07693248987197876,
      "loss_gen": 0.2871274948120117,
      "neg_sim_mean": 0.8628996014595032,
      "pos_sim_mean": 0.8859671354293823,
      "step": 2822
    },
    {
      "epoch": 0.5646,
      "grad_norm": 2.121915817260742,
      "learning_rate": 4.748778409101153e-05,
      "loss": 0.4104,
      "step": 2823
    },
    {
      "epoch": 0.5646,
      "loss_align": 0.2677854299545288,
      "loss_contrastive": 0.02525341510772705,
      "loss_gen": 0.25150296092033386,
      "neg_sim_mean": 0.6574679613113403,
      "pos_sim_mean": 0.7322145700454712,
      "step": 2823
    },
    {
      "epoch": 0.5648,
      "grad_norm": 2.1336958408355713,
      "learning_rate": 4.745292220954696e-05,
      "loss": 0.5223,
      "step": 2824
    },
    {
      "epoch": 0.5648,
      "loss_align": 0.2624976634979248,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3633226752281189,
      "neg_sim_mean": 0.5481489300727844,
      "pos_sim_mean": 0.7375023365020752,
      "step": 2824
    },
    {
      "epoch": 0.565,
      "grad_norm": 2.6380019187927246,
      "learning_rate": 4.7418061569497173e-05,
      "loss": 0.6258,
      "step": 2825
    },
    {
      "epoch": 0.565,
      "loss_align": 0.1600027084350586,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0023949644528329372,
      "neg_sim_mean": 0.33898645639419556,
      "pos_sim_mean": 0.8399972915649414,
      "step": 2825
    },
    {
      "epoch": 0.5652,
      "grad_norm": 0.46397659182548523,
      "learning_rate": 4.738320218785281e-05,
      "loss": 0.1624,
      "step": 2826
    },
    {
      "epoch": 0.5652,
      "loss_align": 0.11131846904754639,
      "loss_contrastive": 0.02309560775756836,
      "loss_gen": 0.3132546842098236,
      "neg_sim_mean": 0.8117771148681641,
      "pos_sim_mean": 0.8886815309524536,
      "step": 2826
    },
    {
      "epoch": 0.5654,
      "grad_norm": 2.071423053741455,
      "learning_rate": 4.734834408160392e-05,
      "loss": 0.4273,
      "step": 2827
    },
    {
      "epoch": 0.5654,
      "loss_align": 0.18349242210388184,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2937716543674469,
      "neg_sim_mean": 0.6687144637107849,
      "pos_sim_mean": 0.8165075778961182,
      "step": 2827
    },
    {
      "epoch": 0.5656,
      "grad_norm": 2.083550453186035,
      "learning_rate": 4.7313487267739896e-05,
      "loss": 0.4773,
      "step": 2828
    },
    {
      "epoch": 0.5656,
      "loss_align": 0.12773144245147705,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0012521239696070552,
      "neg_sim_mean": 0.36028069257736206,
      "pos_sim_mean": 0.872268557548523,
      "step": 2828
    },
    {
      "epoch": 0.5658,
      "grad_norm": 0.38779333233833313,
      "learning_rate": 4.7278631763249554e-05,
      "loss": 0.129,
      "step": 2829
    },
    {
      "epoch": 0.5658,
      "loss_align": 0.0965430736541748,
      "loss_contrastive": 0.06994795799255371,
      "loss_gen": 0.19721731543540955,
      "neg_sim_mean": 0.873404860496521,
      "pos_sim_mean": 0.9034569263458252,
      "step": 2829
    },
    {
      "epoch": 0.566,
      "grad_norm": 1.6996997594833374,
      "learning_rate": 4.724377758512103e-05,
      "loss": 0.3022,
      "step": 2830
    },
    {
      "epoch": 0.566,
      "loss_align": 0.15753257274627686,
      "loss_contrastive": 0.09120619297027588,
      "loss_gen": 0.18249092996120453,
      "neg_sim_mean": 0.8336735963821411,
      "pos_sim_mean": 0.8424674272537231,
      "step": 2830
    },
    {
      "epoch": 0.5662,
      "grad_norm": 1.7636891603469849,
      "learning_rate": 4.720892475034181e-05,
      "loss": 0.351,
      "step": 2831
    },
    {
      "epoch": 0.5662,
      "loss_align": 0.19486957788467407,
      "loss_contrastive": 0.013717114925384521,
      "loss_gen": 0.13646076619625092,
      "neg_sim_mean": 0.7188475131988525,
      "pos_sim_mean": 0.8051304221153259,
      "step": 2831
    },
    {
      "epoch": 0.5664,
      "grad_norm": 1.4727202653884888,
      "learning_rate": 4.7174073275898776e-05,
      "loss": 0.333,
      "step": 2832
    },
    {
      "epoch": 0.5664,
      "loss_align": 0.0959089994430542,
      "loss_contrastive": 0.030666589736938477,
      "loss_gen": 0.241437628865242,
      "neg_sim_mean": 0.8347575664520264,
      "pos_sim_mean": 0.9040910005569458,
      "step": 2832
    },
    {
      "epoch": 0.5666,
      "grad_norm": 1.7474671602249146,
      "learning_rate": 4.7139223178778095e-05,
      "loss": 0.341,
      "step": 2833
    },
    {
      "epoch": 0.5666,
      "loss_align": 0.25260746479034424,
      "loss_contrastive": 0.0014470815658569336,
      "loss_gen": 0.2778034508228302,
      "neg_sim_mean": 0.6488395929336548,
      "pos_sim_mean": 0.7473925352096558,
      "step": 2833
    },
    {
      "epoch": 0.5668,
      "grad_norm": 2.210836410522461,
      "learning_rate": 4.710437447596528e-05,
      "loss": 0.5306,
      "step": 2834
    },
    {
      "epoch": 0.5668,
      "loss_align": 0.14737987518310547,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006991033907979727,
      "neg_sim_mean": 0.2836593985557556,
      "pos_sim_mean": 0.8526201248168945,
      "step": 2834
    },
    {
      "epoch": 0.567,
      "grad_norm": 0.5537954568862915,
      "learning_rate": 4.706952718444517e-05,
      "loss": 0.1544,
      "step": 2835
    },
    {
      "epoch": 0.567,
      "loss_align": 0.18511748313903809,
      "loss_contrastive": 0.10943233966827393,
      "loss_gen": 0.26333731412887573,
      "neg_sim_mean": 0.8243148326873779,
      "pos_sim_mean": 0.8148825168609619,
      "step": 2835
    },
    {
      "epoch": 0.5672,
      "grad_norm": 1.8292464017868042,
      "learning_rate": 4.703468132120193e-05,
      "loss": 0.4616,
      "step": 2836
    },
    {
      "epoch": 0.5672,
      "loss_align": 0.17994242906570435,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24307270348072052,
      "neg_sim_mean": 0.6848500967025757,
      "pos_sim_mean": 0.8200575709342957,
      "step": 2836
    },
    {
      "epoch": 0.5674,
      "grad_norm": 2.038841962814331,
      "learning_rate": 4.699983690321898e-05,
      "loss": 0.423,
      "step": 2837
    },
    {
      "epoch": 0.5674,
      "loss_align": 0.16505956649780273,
      "loss_contrastive": 0.09586089849472046,
      "loss_gen": 0.11968282610177994,
      "neg_sim_mean": 0.8308013081550598,
      "pos_sim_mean": 0.8349404335021973,
      "step": 2837
    },
    {
      "epoch": 0.5676,
      "grad_norm": 1.5673167705535889,
      "learning_rate": 4.696499394747906e-05,
      "loss": 0.2962,
      "step": 2838
    },
    {
      "epoch": 0.5676,
      "loss_align": 0.10604894161224365,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005657948553562164,
      "neg_sim_mean": 0.3443988263607025,
      "pos_sim_mean": 0.8939510583877563,
      "step": 2838
    },
    {
      "epoch": 0.5678,
      "grad_norm": 0.4731493592262268,
      "learning_rate": 4.693015247096423e-05,
      "loss": 0.1117,
      "step": 2839
    },
    {
      "epoch": 0.5678,
      "loss_align": 0.12727242708206177,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1370868682861328,
      "neg_sim_mean": 0.7701835632324219,
      "pos_sim_mean": 0.8727275729179382,
      "step": 2839
    },
    {
      "epoch": 0.568,
      "grad_norm": 1.2786086797714233,
      "learning_rate": 4.68953124906558e-05,
      "loss": 0.2644,
      "step": 2840
    },
    {
      "epoch": 0.568,
      "loss_align": 0.09621560573577881,
      "loss_contrastive": 0.06564998626708984,
      "loss_gen": 0.2928105890750885,
      "neg_sim_mean": 0.8694343566894531,
      "pos_sim_mean": 0.9037843942642212,
      "step": 2840
    },
    {
      "epoch": 0.5682,
      "grad_norm": 1.8607087135314941,
      "learning_rate": 4.6860474023534335e-05,
      "loss": 0.3969,
      "step": 2841
    },
    {
      "epoch": 0.5682,
      "loss_align": 0.1470186710357666,
      "loss_contrastive": 0.0,
      "loss_gen": 0.28270193934440613,
      "neg_sim_mean": 0.6726130843162537,
      "pos_sim_mean": 0.8529813289642334,
      "step": 2841
    },
    {
      "epoch": 0.5684,
      "grad_norm": 1.8901784420013428,
      "learning_rate": 4.68256370865797e-05,
      "loss": 0.4297,
      "step": 2842
    },
    {
      "epoch": 0.5684,
      "loss_align": 0.1432788372039795,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008075576275587082,
      "neg_sim_mean": 0.2924758493900299,
      "pos_sim_mean": 0.8567211627960205,
      "step": 2842
    },
    {
      "epoch": 0.5686,
      "grad_norm": 0.6655915379524231,
      "learning_rate": 4.679080169677097e-05,
      "loss": 0.1514,
      "step": 2843
    },
    {
      "epoch": 0.5686,
      "loss_align": 0.1318870186805725,
      "loss_contrastive": 0.0,
      "loss_gen": 0.08712554723024368,
      "neg_sim_mean": 0.741144597530365,
      "pos_sim_mean": 0.8681129813194275,
      "step": 2843
    },
    {
      "epoch": 0.5688,
      "grad_norm": 1.0125231742858887,
      "learning_rate": 4.675596787108653e-05,
      "loss": 0.219,
      "step": 2844
    },
    {
      "epoch": 0.5688,
      "loss_align": 0.10518074035644531,
      "loss_contrastive": 0.0829230546951294,
      "loss_gen": 0.4461624026298523,
      "neg_sim_mean": 0.8777422904968262,
      "pos_sim_mean": 0.8948192596435547,
      "step": 2844
    },
    {
      "epoch": 0.569,
      "grad_norm": 2.2110588550567627,
      "learning_rate": 4.672113562650394e-05,
      "loss": 0.5613,
      "step": 2845
    },
    {
      "epoch": 0.569,
      "loss_align": 0.09846925735473633,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3532547950744629,
      "neg_sim_mean": 0.7743083834648132,
      "pos_sim_mean": 0.9015307426452637,
      "step": 2845
    },
    {
      "epoch": 0.5692,
      "grad_norm": 1.788325309753418,
      "learning_rate": 4.668630498000001e-05,
      "loss": 0.4517,
      "step": 2846
    },
    {
      "epoch": 0.5692,
      "loss_align": 0.15048962831497192,
      "loss_contrastive": 0.12369471788406372,
      "loss_gen": 0.1834566742181778,
      "neg_sim_mean": 0.8732050657272339,
      "pos_sim_mean": 0.8495103716850281,
      "step": 2846
    },
    {
      "epoch": 0.5694,
      "grad_norm": 1.6254240274429321,
      "learning_rate": 4.665147594855076e-05,
      "loss": 0.3488,
      "step": 2847
    },
    {
      "epoch": 0.5694,
      "loss_align": 0.10445177555084229,
      "loss_contrastive": 0.05401456356048584,
      "loss_gen": 0.2715966999530792,
      "neg_sim_mean": 0.8495627641677856,
      "pos_sim_mean": 0.8955482244491577,
      "step": 2847
    },
    {
      "epoch": 0.5696,
      "grad_norm": 1.6617690324783325,
      "learning_rate": 4.661664854913147e-05,
      "loss": 0.3825,
      "step": 2848
    },
    {
      "epoch": 0.5696,
      "loss_align": 0.3360743522644043,
      "loss_contrastive": 0.09492123126983643,
      "loss_gen": 0.16244974732398987,
      "neg_sim_mean": 0.6588468551635742,
      "pos_sim_mean": 0.6639256477355957,
      "step": 2848
    },
    {
      "epoch": 0.5698,
      "grad_norm": 2.278902053833008,
      "learning_rate": 4.658182279871657e-05,
      "loss": 0.5099,
      "step": 2849
    },
    {
      "epoch": 0.5698,
      "loss_align": 0.2379286289215088,
      "loss_contrastive": 0.06257492303848267,
      "loss_gen": 0.38536134362220764,
      "neg_sim_mean": 0.724646270275116,
      "pos_sim_mean": 0.7620713710784912,
      "step": 2849
    },
    {
      "epoch": 0.57,
      "grad_norm": 2.284756898880005,
      "learning_rate": 4.654699871427971e-05,
      "loss": 0.6308,
      "step": 2850
    },
    {
      "epoch": 0.57,
      "loss_align": 0.11086058616638184,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3973788917064667,
      "neg_sim_mean": 0.7721835374832153,
      "pos_sim_mean": 0.8891394138336182,
      "step": 2850
    },
    {
      "epoch": 0.5702,
      "grad_norm": 2.2918078899383545,
      "learning_rate": 4.6512176312793736e-05,
      "loss": 0.5082,
      "step": 2851
    },
    {
      "epoch": 0.5702,
      "loss_align": 0.13864117860794067,
      "loss_contrastive": 0.14347195625305176,
      "loss_gen": 0.16322387754917145,
      "neg_sim_mean": 0.9048307538032532,
      "pos_sim_mean": 0.8613588213920593,
      "step": 2851
    },
    {
      "epoch": 0.5704,
      "grad_norm": 1.3610113859176636,
      "learning_rate": 4.647735561123066e-05,
      "loss": 0.3191,
      "step": 2852
    },
    {
      "epoch": 0.5704,
      "loss_align": 0.09150004386901855,
      "loss_contrastive": 4.357099533081055e-05,
      "loss_gen": 0.33561551570892334,
      "neg_sim_mean": 0.8085435032844543,
      "pos_sim_mean": 0.9084999561309814,
      "step": 2852
    },
    {
      "epoch": 0.5706,
      "grad_norm": 2.212611675262451,
      "learning_rate": 4.6442536626561675e-05,
      "loss": 0.4271,
      "step": 2853
    },
    {
      "epoch": 0.5706,
      "loss_align": 0.08748733997344971,
      "loss_contrastive": 0.06552779674530029,
      "loss_gen": 0.2600296139717102,
      "neg_sim_mean": 0.8780404329299927,
      "pos_sim_mean": 0.9125126600265503,
      "step": 2853
    },
    {
      "epoch": 0.5708,
      "grad_norm": 1.9751783609390259,
      "learning_rate": 4.6407719375757095e-05,
      "loss": 0.3554,
      "step": 2854
    },
    {
      "epoch": 0.5708,
      "loss_align": 0.30468499660491943,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001798390643671155,
      "neg_sim_mean": 0.3497823476791382,
      "pos_sim_mean": 0.6953150033950806,
      "step": 2854
    },
    {
      "epoch": 0.571,
      "grad_norm": 0.7344871759414673,
      "learning_rate": 4.637290387578647e-05,
      "loss": 0.3065,
      "step": 2855
    },
    {
      "epoch": 0.571,
      "loss_align": 0.10919886827468872,
      "loss_contrastive": 0.1027265191078186,
      "loss_gen": 0.21247361600399017,
      "neg_sim_mean": 0.893527626991272,
      "pos_sim_mean": 0.8908011317253113,
      "step": 2855
    },
    {
      "epoch": 0.5712,
      "grad_norm": 1.638787865638733,
      "learning_rate": 4.633809014361843e-05,
      "loss": 0.334,
      "step": 2856
    },
    {
      "epoch": 0.5712,
      "loss_align": 0.14010465145111084,
      "loss_contrastive": 0.03581047058105469,
      "loss_gen": 0.2754242718219757,
      "neg_sim_mean": 0.7957057952880859,
      "pos_sim_mean": 0.8598953485488892,
      "step": 2856
    },
    {
      "epoch": 0.5714,
      "grad_norm": 1.8633743524551392,
      "learning_rate": 4.630327819622075e-05,
      "loss": 0.4198,
      "step": 2857
    },
    {
      "epoch": 0.5714,
      "loss_align": 0.10588610172271729,
      "loss_contrastive": 0.06172525882720947,
      "loss_gen": 0.23075559735298157,
      "neg_sim_mean": 0.8558391332626343,
      "pos_sim_mean": 0.8941138982772827,
      "step": 2857
    },
    {
      "epoch": 0.5716,
      "grad_norm": 1.4107695817947388,
      "learning_rate": 4.626846805056039e-05,
      "loss": 0.344,
      "step": 2858
    },
    {
      "epoch": 0.5716,
      "loss_align": 0.28257715702056885,
      "loss_contrastive": 0.09974759817123413,
      "loss_gen": 0.12903277575969696,
      "neg_sim_mean": 0.7171704173088074,
      "pos_sim_mean": 0.7174228429794312,
      "step": 2858
    },
    {
      "epoch": 0.5718,
      "grad_norm": 1.5820392370224,
      "learning_rate": 4.623365972360337e-05,
      "loss": 0.4236,
      "step": 2859
    },
    {
      "epoch": 0.5718,
      "loss_align": 0.15807819366455078,
      "loss_contrastive": 0.037835896015167236,
      "loss_gen": 0.2921866476535797,
      "neg_sim_mean": 0.7797576785087585,
      "pos_sim_mean": 0.8419218063354492,
      "step": 2859
    },
    {
      "epoch": 0.572,
      "grad_norm": 1.8736218214035034,
      "learning_rate": 4.619885323231484e-05,
      "loss": 0.4548,
      "step": 2860
    },
    {
      "epoch": 0.572,
      "loss_align": 0.07278239727020264,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16325679421424866,
      "neg_sim_mean": 0.7708105444908142,
      "pos_sim_mean": 0.9272176027297974,
      "step": 2860
    },
    {
      "epoch": 0.5722,
      "grad_norm": 1.5690103769302368,
      "learning_rate": 4.616404859365907e-05,
      "loss": 0.236,
      "step": 2861
    },
    {
      "epoch": 0.5722,
      "loss_align": 0.14611518383026123,
      "loss_contrastive": 0.10158193111419678,
      "loss_gen": 0.40115097165107727,
      "neg_sim_mean": 0.8554667234420776,
      "pos_sim_mean": 0.8538848161697388,
      "step": 2861
    },
    {
      "epoch": 0.5724,
      "grad_norm": 2.273566484451294,
      "learning_rate": 4.6129245824599424e-05,
      "loss": 0.5595,
      "step": 2862
    },
    {
      "epoch": 0.5724,
      "loss_align": 0.08513152599334717,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2972533106803894,
      "neg_sim_mean": 0.7342830896377563,
      "pos_sim_mean": 0.9148684740066528,
      "step": 2862
    },
    {
      "epoch": 0.5726,
      "grad_norm": 1.7103945016860962,
      "learning_rate": 4.609444494209834e-05,
      "loss": 0.3824,
      "step": 2863
    },
    {
      "epoch": 0.5726,
      "loss_align": 0.2275415062904358,
      "loss_contrastive": 0.06650066375732422,
      "loss_gen": 0.08436927199363708,
      "neg_sim_mean": 0.7389591336250305,
      "pos_sim_mean": 0.7724584937095642,
      "step": 2863
    },
    {
      "epoch": 0.5728,
      "grad_norm": 1.292712688446045,
      "learning_rate": 4.6059645963117334e-05,
      "loss": 0.3199,
      "step": 2864
    },
    {
      "epoch": 0.5728,
      "loss_align": 0.11724382638931274,
      "loss_contrastive": 0.036941468715667725,
      "loss_gen": 0.3736802339553833,
      "neg_sim_mean": 0.8196976184844971,
      "pos_sim_mean": 0.8827561736106873,
      "step": 2864
    },
    {
      "epoch": 0.573,
      "grad_norm": 1.6773626804351807,
      "learning_rate": 4.602484890461702e-05,
      "loss": 0.4954,
      "step": 2865
    },
    {
      "epoch": 0.573,
      "loss_align": 0.1308383345603943,
      "loss_contrastive": 0.0762222409248352,
      "loss_gen": 0.10535385459661484,
      "neg_sim_mean": 0.845383882522583,
      "pos_sim_mean": 0.8691616654396057,
      "step": 2865
    },
    {
      "epoch": 0.5732,
      "grad_norm": 1.1908490657806396,
      "learning_rate": 4.599005378355706e-05,
      "loss": 0.2453,
      "step": 2866
    },
    {
      "epoch": 0.5732,
      "loss_align": 0.11289417743682861,
      "loss_contrastive": 0.10741269588470459,
      "loss_gen": 0.1788819581270218,
      "neg_sim_mean": 0.8945184946060181,
      "pos_sim_mean": 0.8871058225631714,
      "step": 2866
    },
    {
      "epoch": 0.5734,
      "grad_norm": 1.7675570249557495,
      "learning_rate": 4.595526061689617e-05,
      "loss": 0.3047,
      "step": 2867
    },
    {
      "epoch": 0.5734,
      "loss_align": 0.26302504539489746,
      "loss_contrastive": 0.005255997180938721,
      "loss_gen": 0.11601275205612183,
      "neg_sim_mean": 0.6422309279441833,
      "pos_sim_mean": 0.7369749546051025,
      "step": 2867
    },
    {
      "epoch": 0.5736,
      "grad_norm": 1.5232993364334106,
      "learning_rate": 4.592046942159213e-05,
      "loss": 0.3797,
      "step": 2868
    },
    {
      "epoch": 0.5736,
      "loss_align": 0.11558687686920166,
      "loss_contrastive": 0.08182287216186523,
      "loss_gen": 0.1484704613685608,
      "neg_sim_mean": 0.8662359714508057,
      "pos_sim_mean": 0.8844131231307983,
      "step": 2868
    },
    {
      "epoch": 0.5738,
      "grad_norm": 1.6554230451583862,
      "learning_rate": 4.588568021460172e-05,
      "loss": 0.2739,
      "step": 2869
    },
    {
      "epoch": 0.5738,
      "loss_align": 0.10943984985351562,
      "loss_contrastive": 0.10464906692504883,
      "loss_gen": 0.21285390853881836,
      "neg_sim_mean": 0.8952091932296753,
      "pos_sim_mean": 0.8905601501464844,
      "step": 2869
    },
    {
      "epoch": 0.574,
      "grad_norm": 1.6072733402252197,
      "learning_rate": 4.58508930128808e-05,
      "loss": 0.3349,
      "step": 2870
    },
    {
      "epoch": 0.574,
      "loss_align": 0.16000628471374512,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11002515256404877,
      "neg_sim_mean": 0.7040717005729675,
      "pos_sim_mean": 0.8399937152862549,
      "step": 2870
    },
    {
      "epoch": 0.5742,
      "grad_norm": 1.3405941724777222,
      "learning_rate": 4.5816107833384234e-05,
      "loss": 0.27,
      "step": 2871
    },
    {
      "epoch": 0.5742,
      "loss_align": 0.12450408935546875,
      "loss_contrastive": 0.08580517768859863,
      "loss_gen": 0.2107107937335968,
      "neg_sim_mean": 0.861301064491272,
      "pos_sim_mean": 0.8754959106445312,
      "step": 2871
    },
    {
      "epoch": 0.5744,
      "grad_norm": 1.9923025369644165,
      "learning_rate": 4.578132469306587e-05,
      "loss": 0.3455,
      "step": 2872
    },
    {
      "epoch": 0.5744,
      "loss_align": 0.2551853656768799,
      "loss_contrastive": 0.16907083988189697,
      "loss_gen": 0.1269671469926834,
      "neg_sim_mean": 0.8138854503631592,
      "pos_sim_mean": 0.7448146343231201,
      "step": 2872
    },
    {
      "epoch": 0.5746,
      "grad_norm": 1.7293751239776611,
      "learning_rate": 4.57465436088786e-05,
      "loss": 0.4024,
      "step": 2873
    },
    {
      "epoch": 0.5746,
      "loss_align": 0.1156163215637207,
      "loss_contrastive": 0.005155444145202637,
      "loss_gen": 0.12989823520183563,
      "neg_sim_mean": 0.789539098739624,
      "pos_sim_mean": 0.8843836784362793,
      "step": 2873
    },
    {
      "epoch": 0.5748,
      "grad_norm": 1.3117581605911255,
      "learning_rate": 4.571176459777431e-05,
      "loss": 0.2461,
      "step": 2874
    },
    {
      "epoch": 0.5748,
      "loss_align": 0.07287561893463135,
      "loss_contrastive": 0.005363881587982178,
      "loss_gen": 0.20383058488368988,
      "neg_sim_mean": 0.8324882388114929,
      "pos_sim_mean": 0.9271243810653687,
      "step": 2874
    },
    {
      "epoch": 0.575,
      "grad_norm": 1.7809184789657593,
      "learning_rate": 4.5676987676703865e-05,
      "loss": 0.2773,
      "step": 2875
    },
    {
      "epoch": 0.575,
      "loss_align": 0.057833075523376465,
      "loss_contrastive": 0.0919538140296936,
      "loss_gen": 0.08949416875839233,
      "neg_sim_mean": 0.9341207146644592,
      "pos_sim_mean": 0.9421669244766235,
      "step": 2875
    },
    {
      "epoch": 0.5752,
      "grad_norm": 0.916742742061615,
      "learning_rate": 4.564221286261709e-05,
      "loss": 0.1584,
      "step": 2876
    },
    {
      "epoch": 0.5752,
      "loss_align": 0.13584959506988525,
      "loss_contrastive": 0.053034067153930664,
      "loss_gen": 0.4671606719493866,
      "neg_sim_mean": 0.8171844482421875,
      "pos_sim_mean": 0.8641504049301147,
      "step": 2876
    },
    {
      "epoch": 0.5754,
      "grad_norm": 2.6478238105773926,
      "learning_rate": 4.560744017246284e-05,
      "loss": 0.6094,
      "step": 2877
    },
    {
      "epoch": 0.5754,
      "loss_align": 0.2562227249145508,
      "loss_contrastive": 0.0172501802444458,
      "loss_gen": 0.36797088384628296,
      "neg_sim_mean": 0.6610274314880371,
      "pos_sim_mean": 0.7437772750854492,
      "step": 2877
    },
    {
      "epoch": 0.5756,
      "grad_norm": 2.277270793914795,
      "learning_rate": 4.557266962318889e-05,
      "loss": 0.6263,
      "step": 2878
    },
    {
      "epoch": 0.5756,
      "loss_align": 0.20071172714233398,
      "loss_contrastive": 0.1658562421798706,
      "loss_gen": 0.28926852345466614,
      "neg_sim_mean": 0.8651444911956787,
      "pos_sim_mean": 0.799288272857666,
      "step": 2878
    },
    {
      "epoch": 0.5758,
      "grad_norm": 2.0399529933929443,
      "learning_rate": 4.553790123174197e-05,
      "loss": 0.5099,
      "step": 2879
    },
    {
      "epoch": 0.5758,
      "loss_align": 0.11212879419326782,
      "loss_contrastive": 0.05140429735183716,
      "loss_gen": 0.35052594542503357,
      "neg_sim_mean": 0.8392754793167114,
      "pos_sim_mean": 0.8878712058067322,
      "step": 2879
    },
    {
      "epoch": 0.576,
      "grad_norm": 2.024427652359009,
      "learning_rate": 4.550313501506781e-05,
      "loss": 0.4688,
      "step": 2880
    },
    {
      "epoch": 0.576,
      "loss_align": 0.11411356925964355,
      "loss_contrastive": 0.0009034872055053711,
      "loss_gen": 0.08647321164608002,
      "neg_sim_mean": 0.7867898941040039,
      "pos_sim_mean": 0.8858864307403564,
      "step": 2880
    },
    {
      "epoch": 0.5762,
      "grad_norm": 1.409043550491333,
      "learning_rate": 4.5468370990111006e-05,
      "loss": 0.2007,
      "step": 2881
    },
    {
      "epoch": 0.5762,
      "loss_align": 0.3117483854293823,
      "loss_contrastive": 0.06228339672088623,
      "loss_gen": 0.26550617814064026,
      "neg_sim_mean": 0.650534987449646,
      "pos_sim_mean": 0.6882516145706177,
      "step": 2881
    },
    {
      "epoch": 0.5764,
      "grad_norm": 2.1486990451812744,
      "learning_rate": 4.543360917381512e-05,
      "loss": 0.5847,
      "step": 2882
    },
    {
      "epoch": 0.5764,
      "loss_align": 0.16158342361450195,
      "loss_contrastive": 0.020028650760650635,
      "loss_gen": 0.19190742075443268,
      "neg_sim_mean": 0.7584452033042908,
      "pos_sim_mean": 0.838416576385498,
      "step": 2882
    },
    {
      "epoch": 0.5766,
      "grad_norm": 2.1712536811828613,
      "learning_rate": 4.539884958312265e-05,
      "loss": 0.3559,
      "step": 2883
    },
    {
      "epoch": 0.5766,
      "loss_align": 0.13713926076889038,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3081701993942261,
      "neg_sim_mean": 0.6936994791030884,
      "pos_sim_mean": 0.8628607392311096,
      "step": 2883
    },
    {
      "epoch": 0.5768,
      "grad_norm": 1.6790090799331665,
      "learning_rate": 4.5364092234974996e-05,
      "loss": 0.4453,
      "step": 2884
    },
    {
      "epoch": 0.5768,
      "loss_align": 0.09813714027404785,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17060400545597076,
      "neg_sim_mean": 0.722464919090271,
      "pos_sim_mean": 0.9018628597259521,
      "step": 2884
    },
    {
      "epoch": 0.577,
      "grad_norm": 1.635040044784546,
      "learning_rate": 4.532933714631248e-05,
      "loss": 0.2687,
      "step": 2885
    },
    {
      "epoch": 0.577,
      "loss_align": 0.08975261449813843,
      "loss_contrastive": 0.014969229698181152,
      "loss_gen": 0.24720259010791779,
      "neg_sim_mean": 0.8252165913581848,
      "pos_sim_mean": 0.9102473855018616,
      "step": 2885
    },
    {
      "epoch": 0.5772,
      "grad_norm": 2.19191312789917,
      "learning_rate": 4.529458433407429e-05,
      "loss": 0.3388,
      "step": 2886
    },
    {
      "epoch": 0.5772,
      "loss_align": 0.15161633491516113,
      "loss_contrastive": 0.09373986721038818,
      "loss_gen": 0.3614479601383209,
      "neg_sim_mean": 0.8421235084533691,
      "pos_sim_mean": 0.8483836650848389,
      "step": 2886
    },
    {
      "epoch": 0.5774,
      "grad_norm": 2.159393310546875,
      "learning_rate": 4.5259833815198535e-05,
      "loss": 0.5243,
      "step": 2887
    },
    {
      "epoch": 0.5774,
      "loss_align": 0.1423119306564331,
      "loss_contrastive": 0.09150373935699463,
      "loss_gen": 0.3469310998916626,
      "neg_sim_mean": 0.8491917848587036,
      "pos_sim_mean": 0.8576880693435669,
      "step": 2887
    },
    {
      "epoch": 0.5776,
      "grad_norm": 2.2618680000305176,
      "learning_rate": 4.522508560662219e-05,
      "loss": 0.5002,
      "step": 2888
    },
    {
      "epoch": 0.5776,
      "loss_align": 0.24057310819625854,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013881624676287174,
      "neg_sim_mean": 0.4644436836242676,
      "pos_sim_mean": 0.7594268918037415,
      "step": 2888
    },
    {
      "epoch": 0.5778,
      "grad_norm": 0.9415478706359863,
      "learning_rate": 4.5190339725281136e-05,
      "loss": 0.2545,
      "step": 2889
    },
    {
      "epoch": 0.5778,
      "loss_align": 0.08173644542694092,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17206032574176788,
      "neg_sim_mean": 0.8058069348335266,
      "pos_sim_mean": 0.9182635545730591,
      "step": 2889
    },
    {
      "epoch": 0.578,
      "grad_norm": 1.4694122076034546,
      "learning_rate": 4.515559618811005e-05,
      "loss": 0.2538,
      "step": 2890
    },
    {
      "epoch": 0.578,
      "loss_align": 0.1741679310798645,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24669185280799866,
      "neg_sim_mean": 0.7119326591491699,
      "pos_sim_mean": 0.8258320689201355,
      "step": 2890
    },
    {
      "epoch": 0.5782,
      "grad_norm": 2.235982656478882,
      "learning_rate": 4.512085501204253e-05,
      "loss": 0.4209,
      "step": 2891
    },
    {
      "epoch": 0.5782,
      "loss_align": 0.24303865432739258,
      "loss_contrastive": 0.1441284418106079,
      "loss_gen": 0.1612107902765274,
      "neg_sim_mean": 0.8010897636413574,
      "pos_sim_mean": 0.7569613456726074,
      "step": 2891
    },
    {
      "epoch": 0.5784,
      "grad_norm": 1.5397132635116577,
      "learning_rate": 4.508611621401102e-05,
      "loss": 0.4215,
      "step": 2892
    },
    {
      "epoch": 0.5784,
      "loss_align": 0.16706502437591553,
      "loss_contrastive": 0.18484127521514893,
      "loss_gen": 0.09205549955368042,
      "neg_sim_mean": 0.9177762269973755,
      "pos_sim_mean": 0.8329349756240845,
      "step": 2892
    },
    {
      "epoch": 0.5786,
      "grad_norm": 1.3083707094192505,
      "learning_rate": 4.505137981094675e-05,
      "loss": 0.2813,
      "step": 2893
    },
    {
      "epoch": 0.5786,
      "loss_align": 0.09352648258209229,
      "loss_contrastive": 0.08390593528747559,
      "loss_gen": 0.23125232756137848,
      "neg_sim_mean": 0.8903794288635254,
      "pos_sim_mean": 0.9064735174179077,
      "step": 2893
    },
    {
      "epoch": 0.5788,
      "grad_norm": 1.6565405130386353,
      "learning_rate": 4.501664581977986e-05,
      "loss": 0.3348,
      "step": 2894
    },
    {
      "epoch": 0.5788,
      "loss_align": 0.2665647268295288,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19590745866298676,
      "neg_sim_mean": 0.616485595703125,
      "pos_sim_mean": 0.7334352731704712,
      "step": 2894
    },
    {
      "epoch": 0.579,
      "grad_norm": 2.126251697540283,
      "learning_rate": 4.498191425743925e-05,
      "loss": 0.4625,
      "step": 2895
    },
    {
      "epoch": 0.579,
      "loss_align": 0.07570546865463257,
      "loss_contrastive": 0.0439336895942688,
      "loss_gen": 0.3002474308013916,
      "neg_sim_mean": 0.8682281970977783,
      "pos_sim_mean": 0.9242945313453674,
      "step": 2895
    },
    {
      "epoch": 0.5792,
      "grad_norm": 1.7489856481552124,
      "learning_rate": 4.494718514085268e-05,
      "loss": 0.3812,
      "step": 2896
    },
    {
      "epoch": 0.5792,
      "loss_align": 0.13011282682418823,
      "loss_contrastive": 0.11274832487106323,
      "loss_gen": 0.182122603058815,
      "neg_sim_mean": 0.8826354742050171,
      "pos_sim_mean": 0.8698871731758118,
      "step": 2896
    },
    {
      "epoch": 0.5794,
      "grad_norm": 1.309718132019043,
      "learning_rate": 4.491245848694669e-05,
      "loss": 0.3258,
      "step": 2897
    },
    {
      "epoch": 0.5794,
      "loss_align": 0.17369699478149414,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1425008624792099,
      "neg_sim_mean": 0.5662230849266052,
      "pos_sim_mean": 0.8263030052185059,
      "step": 2897
    },
    {
      "epoch": 0.5796,
      "grad_norm": 1.5406478643417358,
      "learning_rate": 4.487773431264664e-05,
      "loss": 0.3162,
      "step": 2898
    },
    {
      "epoch": 0.5796,
      "loss_align": 0.08724379539489746,
      "loss_contrastive": 0.029340386390686035,
      "loss_gen": 0.2813635468482971,
      "neg_sim_mean": 0.8420965671539307,
      "pos_sim_mean": 0.9127562046051025,
      "step": 2898
    },
    {
      "epoch": 0.5798,
      "grad_norm": 2.1337692737579346,
      "learning_rate": 4.484301263487665e-05,
      "loss": 0.3721,
      "step": 2899
    },
    {
      "epoch": 0.5798,
      "loss_align": 0.13180607557296753,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007665008772164583,
      "neg_sim_mean": 0.36519551277160645,
      "pos_sim_mean": 0.8681939244270325,
      "step": 2899
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.5959495902061462,
      "learning_rate": 4.4808293470559643e-05,
      "loss": 0.1395,
      "step": 2900
    },
    {
      "epoch": 0.58,
      "loss_align": 0.13987040519714355,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19447195529937744,
      "neg_sim_mean": 0.6249330043792725,
      "pos_sim_mean": 0.8601295948028564,
      "step": 2900
    },
    {
      "epoch": 0.5802,
      "grad_norm": 1.5628844499588013,
      "learning_rate": 4.477357683661734e-05,
      "loss": 0.3343,
      "step": 2901
    },
    {
      "epoch": 0.5802,
      "loss_align": 0.06804823875427246,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19816815853118896,
      "neg_sim_mean": 0.5194757580757141,
      "pos_sim_mean": 0.9319517612457275,
      "step": 2901
    },
    {
      "epoch": 0.5804,
      "grad_norm": 1.660947322845459,
      "learning_rate": 4.473886274997018e-05,
      "loss": 0.2662,
      "step": 2902
    },
    {
      "epoch": 0.5804,
      "loss_align": 0.08299219608306885,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11356157064437866,
      "neg_sim_mean": 0.7144152522087097,
      "pos_sim_mean": 0.9170078039169312,
      "step": 2902
    },
    {
      "epoch": 0.5806,
      "grad_norm": 1.2998594045639038,
      "learning_rate": 4.4704151227537414e-05,
      "loss": 0.1966,
      "step": 2903
    },
    {
      "epoch": 0.5806,
      "loss_align": 0.2367376685142517,
      "loss_contrastive": 0.10928881168365479,
      "loss_gen": 0.14674367010593414,
      "neg_sim_mean": 0.7725511193275452,
      "pos_sim_mean": 0.7632623314857483,
      "step": 2903
    },
    {
      "epoch": 0.5808,
      "grad_norm": 1.5688707828521729,
      "learning_rate": 4.466944228623701e-05,
      "loss": 0.3966,
      "step": 2904
    },
    {
      "epoch": 0.5808,
      "loss_align": 0.13712090253829956,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23608969151973724,
      "neg_sim_mean": 0.7414940595626831,
      "pos_sim_mean": 0.8628790974617004,
      "step": 2904
    },
    {
      "epoch": 0.581,
      "grad_norm": 1.7564184665679932,
      "learning_rate": 4.463473594298567e-05,
      "loss": 0.3732,
      "step": 2905
    },
    {
      "epoch": 0.581,
      "loss_align": 0.10786408185958862,
      "loss_contrastive": 0.10055917501449585,
      "loss_gen": 0.15285848081111908,
      "neg_sim_mean": 0.8926950693130493,
      "pos_sim_mean": 0.8921359181404114,
      "step": 2905
    },
    {
      "epoch": 0.5812,
      "grad_norm": 1.434814214706421,
      "learning_rate": 4.460003221469886e-05,
      "loss": 0.2728,
      "step": 2906
    },
    {
      "epoch": 0.5812,
      "loss_align": 0.1374826431274414,
      "loss_contrastive": 0.0,
      "loss_gen": 0.06503885984420776,
      "neg_sim_mean": 0.423769474029541,
      "pos_sim_mean": 0.8625173568725586,
      "step": 2906
    },
    {
      "epoch": 0.5814,
      "grad_norm": 1.4306057691574097,
      "learning_rate": 4.4565331118290756e-05,
      "loss": 0.2025,
      "step": 2907
    },
    {
      "epoch": 0.5814,
      "loss_align": 0.10608971118927002,
      "loss_contrastive": 0.1291825771331787,
      "loss_gen": 0.16731053590774536,
      "neg_sim_mean": 0.9230928421020508,
      "pos_sim_mean": 0.89391028881073,
      "step": 2907
    },
    {
      "epoch": 0.5816,
      "grad_norm": 1.5270575284957886,
      "learning_rate": 4.453063267067424e-05,
      "loss": 0.2889,
      "step": 2908
    },
    {
      "epoch": 0.5816,
      "loss_align": 0.08312177658081055,
      "loss_contrastive": 0.0017646551132202148,
      "loss_gen": 0.12159968167543411,
      "neg_sim_mean": 0.8186428546905518,
      "pos_sim_mean": 0.9168782234191895,
      "step": 2908
    },
    {
      "epoch": 0.5818,
      "grad_norm": 1.5710135698318481,
      "learning_rate": 4.44959368887609e-05,
      "loss": 0.2049,
      "step": 2909
    },
    {
      "epoch": 0.5818,
      "loss_align": 0.19109660387039185,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14723964035511017,
      "neg_sim_mean": 0.5189675688743591,
      "pos_sim_mean": 0.8089033961296082,
      "step": 2909
    },
    {
      "epoch": 0.582,
      "grad_norm": 1.58189857006073,
      "learning_rate": 4.446124378946107e-05,
      "loss": 0.3383,
      "step": 2910
    },
    {
      "epoch": 0.582,
      "loss_align": 0.15336978435516357,
      "loss_contrastive": 0.03111565113067627,
      "loss_gen": 0.22687427699565887,
      "neg_sim_mean": 0.7777458429336548,
      "pos_sim_mean": 0.8466302156448364,
      "step": 2910
    },
    {
      "epoch": 0.5822,
      "grad_norm": 2.004889726638794,
      "learning_rate": 4.442655338968373e-05,
      "loss": 0.384,
      "step": 2911
    },
    {
      "epoch": 0.5822,
      "loss_align": 0.09817361831665039,
      "loss_contrastive": 0.03776133060455322,
      "loss_gen": 0.23417295515537262,
      "neg_sim_mean": 0.8395876884460449,
      "pos_sim_mean": 0.9018263816833496,
      "step": 2911
    },
    {
      "epoch": 0.5824,
      "grad_norm": 1.3722987174987793,
      "learning_rate": 4.4391865706336555e-05,
      "loss": 0.3369,
      "step": 2912
    },
    {
      "epoch": 0.5824,
      "loss_align": 0.11522984504699707,
      "loss_contrastive": 0.06113767623901367,
      "loss_gen": 0.13008376955986023,
      "neg_sim_mean": 0.8459078073501587,
      "pos_sim_mean": 0.8847701549530029,
      "step": 2912
    },
    {
      "epoch": 0.5826,
      "grad_norm": 1.7386590242385864,
      "learning_rate": 4.435718075632592e-05,
      "loss": 0.2527,
      "step": 2913
    },
    {
      "epoch": 0.5826,
      "loss_align": 0.0314670205116272,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22799749672412872,
      "neg_sim_mean": 0.7123435735702515,
      "pos_sim_mean": 0.9685329794883728,
      "step": 2913
    },
    {
      "epoch": 0.5828,
      "grad_norm": 1.514278531074524,
      "learning_rate": 4.432249855655681e-05,
      "loss": 0.2595,
      "step": 2914
    },
    {
      "epoch": 0.5828,
      "loss_align": 0.21372568607330322,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27816903591156006,
      "neg_sim_mean": 0.6365270614624023,
      "pos_sim_mean": 0.7862743139266968,
      "step": 2914
    },
    {
      "epoch": 0.583,
      "grad_norm": 2.0365614891052246,
      "learning_rate": 4.428781912393298e-05,
      "loss": 0.4919,
      "step": 2915
    },
    {
      "epoch": 0.583,
      "loss_align": 0.18302655220031738,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23962551355361938,
      "neg_sim_mean": 0.6752578020095825,
      "pos_sim_mean": 0.8169734477996826,
      "step": 2915
    },
    {
      "epoch": 0.5832,
      "grad_norm": 2.481698513031006,
      "learning_rate": 4.425314247535668e-05,
      "loss": 0.4227,
      "step": 2916
    },
    {
      "epoch": 0.5832,
      "loss_align": 0.24938809871673584,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27880311012268066,
      "neg_sim_mean": 0.646392285823822,
      "pos_sim_mean": 0.7506119012832642,
      "step": 2916
    },
    {
      "epoch": 0.5834,
      "grad_norm": 2.478733539581299,
      "learning_rate": 4.421846862772893e-05,
      "loss": 0.5282,
      "step": 2917
    },
    {
      "epoch": 0.5834,
      "loss_align": 0.20692741870880127,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02030266262590885,
      "neg_sim_mean": 0.42682772874832153,
      "pos_sim_mean": 0.7930725812911987,
      "step": 2917
    },
    {
      "epoch": 0.5836,
      "grad_norm": 1.0945205688476562,
      "learning_rate": 4.418379759794934e-05,
      "loss": 0.2272,
      "step": 2918
    },
    {
      "epoch": 0.5836,
      "loss_align": 0.11016654968261719,
      "loss_contrastive": 0.10277307033538818,
      "loss_gen": 0.31874844431877136,
      "neg_sim_mean": 0.8926064968109131,
      "pos_sim_mean": 0.8898334503173828,
      "step": 2918
    },
    {
      "epoch": 0.5838,
      "grad_norm": 1.9018610715866089,
      "learning_rate": 4.414912940291613e-05,
      "loss": 0.4412,
      "step": 2919
    },
    {
      "epoch": 0.5838,
      "loss_align": 0.13286137580871582,
      "loss_contrastive": 0.11803269386291504,
      "loss_gen": 0.3334321081638336,
      "neg_sim_mean": 0.8851712942123413,
      "pos_sim_mean": 0.8671386241912842,
      "step": 2919
    },
    {
      "epoch": 0.584,
      "grad_norm": 2.5348517894744873,
      "learning_rate": 4.411446405952618e-05,
      "loss": 0.4805,
      "step": 2920
    },
    {
      "epoch": 0.584,
      "loss_align": 0.09839892387390137,
      "loss_contrastive": 0.052240192890167236,
      "loss_gen": 0.15648645162582397,
      "neg_sim_mean": 0.853841245174408,
      "pos_sim_mean": 0.9016010761260986,
      "step": 2920
    },
    {
      "epoch": 0.5842,
      "grad_norm": 1.5353803634643555,
      "learning_rate": 4.407980158467495e-05,
      "loss": 0.2612,
      "step": 2921
    },
    {
      "epoch": 0.5842,
      "loss_align": 0.15961390733718872,
      "loss_contrastive": 0.11896532773971558,
      "loss_gen": 0.21700608730316162,
      "neg_sim_mean": 0.859351396560669,
      "pos_sim_mean": 0.8403860926628113,
      "step": 2921
    },
    {
      "epoch": 0.5844,
      "grad_norm": 2.1813087463378906,
      "learning_rate": 4.404514199525651e-05,
      "loss": 0.3909,
      "step": 2922
    },
    {
      "epoch": 0.5844,
      "loss_align": 0.08273506164550781,
      "loss_contrastive": 0.015566706657409668,
      "loss_gen": 0.3276313841342926,
      "neg_sim_mean": 0.832831621170044,
      "pos_sim_mean": 0.9172649383544922,
      "step": 2922
    },
    {
      "epoch": 0.5846,
      "grad_norm": 2.2797017097473145,
      "learning_rate": 4.401048530816352e-05,
      "loss": 0.4122,
      "step": 2923
    },
    {
      "epoch": 0.5846,
      "loss_align": 0.2092280387878418,
      "loss_contrastive": 0.09511727094650269,
      "loss_gen": 0.2998206317424774,
      "neg_sim_mean": 0.785889208316803,
      "pos_sim_mean": 0.7907719612121582,
      "step": 2923
    },
    {
      "epoch": 0.5848,
      "grad_norm": 2.053436517715454,
      "learning_rate": 4.397583154028725e-05,
      "loss": 0.5205,
      "step": 2924
    },
    {
      "epoch": 0.5848,
      "loss_align": 0.19924521446228027,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0361415334045887,
      "neg_sim_mean": 0.4311976432800293,
      "pos_sim_mean": 0.8007547855377197,
      "step": 2924
    },
    {
      "epoch": 0.585,
      "grad_norm": 1.50087308883667,
      "learning_rate": 4.394118070851749e-05,
      "loss": 0.2354,
      "step": 2925
    },
    {
      "epoch": 0.585,
      "loss_align": 0.12414509057998657,
      "loss_contrastive": 0.057019591331481934,
      "loss_gen": 0.27887603640556335,
      "neg_sim_mean": 0.8328744769096375,
      "pos_sim_mean": 0.8758549094200134,
      "step": 2925
    },
    {
      "epoch": 0.5852,
      "grad_norm": 1.8932300806045532,
      "learning_rate": 4.390653282974264e-05,
      "loss": 0.4099,
      "step": 2926
    },
    {
      "epoch": 0.5852,
      "loss_align": 0.13165390491485596,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0033380796667188406,
      "neg_sim_mean": 0.4149022698402405,
      "pos_sim_mean": 0.868346095085144,
      "step": 2926
    },
    {
      "epoch": 0.5854,
      "grad_norm": 0.5347163081169128,
      "learning_rate": 4.387188792084967e-05,
      "loss": 0.135,
      "step": 2927
    },
    {
      "epoch": 0.5854,
      "loss_align": 0.1607292890548706,
      "loss_contrastive": 0.005716919898986816,
      "loss_gen": 0.16832087934017181,
      "neg_sim_mean": 0.7449876070022583,
      "pos_sim_mean": 0.8392707109451294,
      "step": 2927
    },
    {
      "epoch": 0.5856,
      "grad_norm": 1.6726785898208618,
      "learning_rate": 4.383724599872407e-05,
      "loss": 0.3297,
      "step": 2928
    },
    {
      "epoch": 0.5856,
      "loss_align": 0.09227395057678223,
      "loss_contrastive": 0.04056048393249512,
      "loss_gen": 0.27703821659088135,
      "neg_sim_mean": 0.848286509513855,
      "pos_sim_mean": 0.9077260494232178,
      "step": 2928
    },
    {
      "epoch": 0.5858,
      "grad_norm": 2.4267311096191406,
      "learning_rate": 4.380260708024991e-05,
      "loss": 0.3742,
      "step": 2929
    },
    {
      "epoch": 0.5858,
      "loss_align": 0.09751105308532715,
      "loss_contrastive": 0.029307663440704346,
      "loss_gen": 0.19962000846862793,
      "neg_sim_mean": 0.8317965865135193,
      "pos_sim_mean": 0.9024889469146729,
      "step": 2929
    },
    {
      "epoch": 0.586,
      "grad_norm": 1.8317509889602661,
      "learning_rate": 4.376797118230979e-05,
      "loss": 0.3006,
      "step": 2930
    },
    {
      "epoch": 0.586,
      "loss_align": 0.07960128784179688,
      "loss_contrastive": 0.0,
      "loss_gen": 0.015803014859557152,
      "neg_sim_mean": 0.21872833371162415,
      "pos_sim_mean": 0.9203987121582031,
      "step": 2930
    },
    {
      "epoch": 0.5862,
      "grad_norm": 0.7929801940917969,
      "learning_rate": 4.373333832178478e-05,
      "loss": 0.0954,
      "step": 2931
    },
    {
      "epoch": 0.5862,
      "loss_align": 0.13897645473480225,
      "loss_contrastive": 0.08078992366790771,
      "loss_gen": 0.21495383977890015,
      "neg_sim_mean": 0.8418134450912476,
      "pos_sim_mean": 0.8610235452651978,
      "step": 2931
    },
    {
      "epoch": 0.5864,
      "grad_norm": 2.0886335372924805,
      "learning_rate": 4.369870851555456e-05,
      "loss": 0.3636,
      "step": 2932
    },
    {
      "epoch": 0.5864,
      "loss_align": 0.10287576913833618,
      "loss_contrastive": 0.08972489833831787,
      "loss_gen": 0.42347076535224915,
      "neg_sim_mean": 0.8868491053581238,
      "pos_sim_mean": 0.8971242308616638,
      "step": 2932
    },
    {
      "epoch": 0.5866,
      "grad_norm": 2.270416498184204,
      "learning_rate": 4.366408178049728e-05,
      "loss": 0.5371,
      "step": 2933
    },
    {
      "epoch": 0.5866,
      "loss_align": 0.12864381074905396,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003247741376981139,
      "neg_sim_mean": 0.17818450927734375,
      "pos_sim_mean": 0.871356189250946,
      "step": 2933
    },
    {
      "epoch": 0.5868,
      "grad_norm": 0.5028045773506165,
      "learning_rate": 4.362945813348955e-05,
      "loss": 0.1319,
      "step": 2934
    },
    {
      "epoch": 0.5868,
      "loss_align": 0.1637522578239441,
      "loss_contrastive": 0.1338837742805481,
      "loss_gen": 0.31324905157089233,
      "neg_sim_mean": 0.8701314926147461,
      "pos_sim_mean": 0.8362477421760559,
      "step": 2934
    },
    {
      "epoch": 0.587,
      "grad_norm": 1.5838298797607422,
      "learning_rate": 4.3594837591406535e-05,
      "loss": 0.4931,
      "step": 2935
    },
    {
      "epoch": 0.587,
      "loss_align": 0.11743748188018799,
      "loss_contrastive": 0.036178767681121826,
      "loss_gen": 0.3032273054122925,
      "neg_sim_mean": 0.8187412619590759,
      "pos_sim_mean": 0.882562518119812,
      "step": 2935
    },
    {
      "epoch": 0.5872,
      "grad_norm": 2.1981265544891357,
      "learning_rate": 4.356022017112187e-05,
      "loss": 0.425,
      "step": 2936
    },
    {
      "epoch": 0.5872,
      "loss_align": 0.18704605102539062,
      "loss_contrastive": 0.027679264545440674,
      "loss_gen": 0.4718828797340393,
      "neg_sim_mean": 0.7406331896781921,
      "pos_sim_mean": 0.8129539489746094,
      "step": 2936
    },
    {
      "epoch": 0.5874,
      "grad_norm": 2.188091993331909,
      "learning_rate": 4.352560588950766e-05,
      "loss": 0.6623,
      "step": 2937
    },
    {
      "epoch": 0.5874,
      "loss_align": 0.08909046649932861,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2236669808626175,
      "neg_sim_mean": 0.7970911264419556,
      "pos_sim_mean": 0.9109095335006714,
      "step": 2937
    },
    {
      "epoch": 0.5876,
      "grad_norm": 1.5605337619781494,
      "learning_rate": 4.349099476343448e-05,
      "loss": 0.3128,
      "step": 2938
    },
    {
      "epoch": 0.5876,
      "loss_align": 0.057132601737976074,
      "loss_contrastive": 0.05614423751831055,
      "loss_gen": 0.2934165298938751,
      "neg_sim_mean": 0.8990116119384766,
      "pos_sim_mean": 0.9428673982620239,
      "step": 2938
    },
    {
      "epoch": 0.5878,
      "grad_norm": 1.7088907957077026,
      "learning_rate": 4.345638680977139e-05,
      "loss": 0.3573,
      "step": 2939
    },
    {
      "epoch": 0.5878,
      "loss_align": 0.1375035047531128,
      "loss_contrastive": 0.1195942759513855,
      "loss_gen": 0.21757087111473083,
      "neg_sim_mean": 0.8820907473564148,
      "pos_sim_mean": 0.8624964952468872,
      "step": 2939
    },
    {
      "epoch": 0.588,
      "grad_norm": 1.9476487636566162,
      "learning_rate": 4.342178204538588e-05,
      "loss": 0.3694,
      "step": 2940
    },
    {
      "epoch": 0.588,
      "loss_align": 0.18230408430099487,
      "loss_contrastive": 0.016624391078948975,
      "loss_gen": 0.18699511885643005,
      "neg_sim_mean": 0.7343202829360962,
      "pos_sim_mean": 0.8176959156990051,
      "step": 2940
    },
    {
      "epoch": 0.5882,
      "grad_norm": 1.6504037380218506,
      "learning_rate": 4.3387180487143876e-05,
      "loss": 0.3713,
      "step": 2941
    },
    {
      "epoch": 0.5882,
      "loss_align": 0.14305245876312256,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002770355436950922,
      "neg_sim_mean": 0.37793081998825073,
      "pos_sim_mean": 0.8569475412368774,
      "step": 2941
    },
    {
      "epoch": 0.5884,
      "grad_norm": 0.5052962303161621,
      "learning_rate": 4.335258215190979e-05,
      "loss": 0.1458,
      "step": 2942
    },
    {
      "epoch": 0.5884,
      "loss_align": 0.09059262275695801,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0029181286226958036,
      "neg_sim_mean": 0.24482367932796478,
      "pos_sim_mean": 0.909407377243042,
      "step": 2942
    },
    {
      "epoch": 0.5886,
      "grad_norm": 0.3471231758594513,
      "learning_rate": 4.3317987056546394e-05,
      "loss": 0.0935,
      "step": 2943
    },
    {
      "epoch": 0.5886,
      "loss_align": 0.10311740636825562,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0018316601635888219,
      "neg_sim_mean": 0.43337926268577576,
      "pos_sim_mean": 0.8968825936317444,
      "step": 2943
    },
    {
      "epoch": 0.5888,
      "grad_norm": 0.37862688302993774,
      "learning_rate": 4.328339521791493e-05,
      "loss": 0.1049,
      "step": 2944
    },
    {
      "epoch": 0.5888,
      "loss_align": 0.14727407693862915,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2051185518503189,
      "neg_sim_mean": 0.6707121729850769,
      "pos_sim_mean": 0.8527259230613708,
      "step": 2944
    },
    {
      "epoch": 0.589,
      "grad_norm": 1.633732557296753,
      "learning_rate": 4.324880665287505e-05,
      "loss": 0.3524,
      "step": 2945
    },
    {
      "epoch": 0.589,
      "loss_align": 0.14172828197479248,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15209037065505981,
      "neg_sim_mean": 0.7233953475952148,
      "pos_sim_mean": 0.8582717180252075,
      "step": 2945
    },
    {
      "epoch": 0.5892,
      "grad_norm": 1.6784700155258179,
      "learning_rate": 4.321422137828479e-05,
      "loss": 0.2938,
      "step": 2946
    },
    {
      "epoch": 0.5892,
      "loss_align": 0.16045725345611572,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11939414590597153,
      "neg_sim_mean": 0.6540365219116211,
      "pos_sim_mean": 0.8395427465438843,
      "step": 2946
    },
    {
      "epoch": 0.5894,
      "grad_norm": 1.2931550741195679,
      "learning_rate": 4.317963941100058e-05,
      "loss": 0.2799,
      "step": 2947
    },
    {
      "epoch": 0.5894,
      "loss_align": 0.06191962957382202,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11898072808980942,
      "neg_sim_mean": 0.8039844036102295,
      "pos_sim_mean": 0.938080370426178,
      "step": 2947
    },
    {
      "epoch": 0.5896,
      "grad_norm": 1.1946988105773926,
      "learning_rate": 4.314506076787729e-05,
      "loss": 0.1809,
      "step": 2948
    },
    {
      "epoch": 0.5896,
      "loss_align": 0.05883902311325073,
      "loss_contrastive": 0.040629446506500244,
      "loss_gen": 0.046088721603155136,
      "neg_sim_mean": 0.8817903995513916,
      "pos_sim_mean": 0.9411609768867493,
      "step": 2948
    },
    {
      "epoch": 0.5898,
      "grad_norm": 0.9190599918365479,
      "learning_rate": 4.31104854657681e-05,
      "loss": 0.1098,
      "step": 2949
    },
    {
      "epoch": 0.5898,
      "loss_align": 0.2542690634727478,
      "loss_contrastive": 0.20008403062820435,
      "loss_gen": 0.5389471054077148,
      "neg_sim_mean": 0.8458149433135986,
      "pos_sim_mean": 0.7457309365272522,
      "step": 2949
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.706779718399048,
      "learning_rate": 4.307591352152459e-05,
      "loss": 0.8172,
      "step": 2950
    },
    {
      "epoch": 0.59,
      "loss_align": 0.08019095659255981,
      "loss_contrastive": 0.009863018989562988,
      "loss_gen": 0.2639683783054352,
      "neg_sim_mean": 0.8296720385551453,
      "pos_sim_mean": 0.9198090434074402,
      "step": 2950
    },
    {
      "epoch": 0.5902,
      "grad_norm": 1.933971881866455,
      "learning_rate": 4.3041344951996746e-05,
      "loss": 0.3453,
      "step": 2951
    },
    {
      "epoch": 0.5902,
      "loss_align": 0.13467860221862793,
      "loss_contrastive": 0.07512062788009644,
      "loss_gen": 0.14532002806663513,
      "neg_sim_mean": 0.8404420018196106,
      "pos_sim_mean": 0.8653213977813721,
      "step": 2951
    },
    {
      "epoch": 0.5904,
      "grad_norm": 1.5440597534179688,
      "learning_rate": 4.300677977403281e-05,
      "loss": 0.289,
      "step": 2952
    },
    {
      "epoch": 0.5904,
      "loss_align": 0.0893399715423584,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27211132645606995,
      "neg_sim_mean": 0.795899510383606,
      "pos_sim_mean": 0.9106600284576416,
      "step": 2952
    },
    {
      "epoch": 0.5906,
      "grad_norm": 2.0941224098205566,
      "learning_rate": 4.297221800447946e-05,
      "loss": 0.3615,
      "step": 2953
    },
    {
      "epoch": 0.5906,
      "loss_align": 0.08701825141906738,
      "loss_contrastive": 0.07900220155715942,
      "loss_gen": 0.1600562036037445,
      "neg_sim_mean": 0.8919839262962341,
      "pos_sim_mean": 0.9129817485809326,
      "step": 2953
    },
    {
      "epoch": 0.5908,
      "grad_norm": 1.4553126096725464,
      "learning_rate": 4.2937659660181675e-05,
      "loss": 0.2566,
      "step": 2954
    },
    {
      "epoch": 0.5908,
      "loss_align": 0.22678756713867188,
      "loss_contrastive": 0.09804189205169678,
      "loss_gen": 0.1882973164319992,
      "neg_sim_mean": 0.771254301071167,
      "pos_sim_mean": 0.7732124328613281,
      "step": 2954
    },
    {
      "epoch": 0.591,
      "grad_norm": 1.838149905204773,
      "learning_rate": 4.2903104757982785e-05,
      "loss": 0.4268,
      "step": 2955
    },
    {
      "epoch": 0.591,
      "loss_align": 0.08348977565765381,
      "loss_contrastive": 0.0845177173614502,
      "loss_gen": 0.31321442127227783,
      "neg_sim_mean": 0.9010279178619385,
      "pos_sim_mean": 0.9165102243423462,
      "step": 2955
    },
    {
      "epoch": 0.5912,
      "grad_norm": 2.370147705078125,
      "learning_rate": 4.2868553314724425e-05,
      "loss": 0.4068,
      "step": 2956
    },
    {
      "epoch": 0.5912,
      "loss_align": 0.12107336521148682,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007850401103496552,
      "neg_sim_mean": 0.47729361057281494,
      "pos_sim_mean": 0.8789266347885132,
      "step": 2956
    },
    {
      "epoch": 0.5914,
      "grad_norm": 0.6469433903694153,
      "learning_rate": 4.283400534724653e-05,
      "loss": 0.1289,
      "step": 2957
    },
    {
      "epoch": 0.5914,
      "loss_align": 0.13392329216003418,
      "loss_contrastive": 0.03122556209564209,
      "loss_gen": 0.2842964231967926,
      "neg_sim_mean": 0.79730224609375,
      "pos_sim_mean": 0.8660767078399658,
      "step": 2957
    },
    {
      "epoch": 0.5916,
      "grad_norm": 1.8954206705093384,
      "learning_rate": 4.2799460872387394e-05,
      "loss": 0.422,
      "step": 2958
    },
    {
      "epoch": 0.5916,
      "loss_align": 0.20397520065307617,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26805102825164795,
      "neg_sim_mean": 0.6414048671722412,
      "pos_sim_mean": 0.7960247993469238,
      "step": 2958
    },
    {
      "epoch": 0.5918,
      "grad_norm": 2.207566976547241,
      "learning_rate": 4.276491990698355e-05,
      "loss": 0.472,
      "step": 2959
    },
    {
      "epoch": 0.5918,
      "loss_align": 0.12816143035888672,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02168317511677742,
      "neg_sim_mean": 0.4863801598548889,
      "pos_sim_mean": 0.8718385696411133,
      "step": 2959
    },
    {
      "epoch": 0.592,
      "grad_norm": 1.1340724229812622,
      "learning_rate": 4.273038246786986e-05,
      "loss": 0.1498,
      "step": 2960
    },
    {
      "epoch": 0.592,
      "loss_align": 0.1350705623626709,
      "loss_contrastive": 0.011488795280456543,
      "loss_gen": 0.04042327404022217,
      "neg_sim_mean": 0.7764182090759277,
      "pos_sim_mean": 0.8649294376373291,
      "step": 2960
    },
    {
      "epoch": 0.5922,
      "grad_norm": 0.9135616421699524,
      "learning_rate": 4.269584857187943e-05,
      "loss": 0.1769,
      "step": 2961
    },
    {
      "epoch": 0.5922,
      "loss_align": 0.12382996082305908,
      "loss_contrastive": 0.0971406102180481,
      "loss_gen": 0.1818072497844696,
      "neg_sim_mean": 0.8733106255531311,
      "pos_sim_mean": 0.8761700391769409,
      "step": 2961
    },
    {
      "epoch": 0.5924,
      "grad_norm": 1.75827157497406,
      "learning_rate": 4.266131823584367e-05,
      "loss": 0.3173,
      "step": 2962
    },
    {
      "epoch": 0.5924,
      "loss_align": 0.09881091117858887,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26336318254470825,
      "neg_sim_mean": 0.6109010577201843,
      "pos_sim_mean": 0.9011890888214111,
      "step": 2962
    },
    {
      "epoch": 0.5926,
      "grad_norm": 1.5901747941970825,
      "learning_rate": 4.262679147659227e-05,
      "loss": 0.3622,
      "step": 2963
    },
    {
      "epoch": 0.5926,
      "loss_align": 0.11459732055664062,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011883285827934742,
      "neg_sim_mean": 0.3657357096672058,
      "pos_sim_mean": 0.8854026794433594,
      "step": 2963
    },
    {
      "epoch": 0.5928,
      "grad_norm": 0.9134882092475891,
      "learning_rate": 4.25922683109531e-05,
      "loss": 0.1265,
      "step": 2964
    },
    {
      "epoch": 0.5928,
      "loss_align": 0.13391512632369995,
      "loss_contrastive": 0.07908767461776733,
      "loss_gen": 0.1551862508058548,
      "neg_sim_mean": 0.8451725244522095,
      "pos_sim_mean": 0.8660848736763,
      "step": 2964
    },
    {
      "epoch": 0.593,
      "grad_norm": 1.4901478290557861,
      "learning_rate": 4.255774875575239e-05,
      "loss": 0.2986,
      "step": 2965
    },
    {
      "epoch": 0.593,
      "loss_align": 0.13792914152145386,
      "loss_contrastive": 0.09952425956726074,
      "loss_gen": 0.19724616408348083,
      "neg_sim_mean": 0.861595094203949,
      "pos_sim_mean": 0.8620708584785461,
      "step": 2965
    },
    {
      "epoch": 0.5932,
      "grad_norm": 1.8269866704940796,
      "learning_rate": 4.252323282781453e-05,
      "loss": 0.3471,
      "step": 2966
    },
    {
      "epoch": 0.5932,
      "loss_align": 0.08672487735748291,
      "loss_contrastive": 0.00038254261016845703,
      "loss_gen": 0.20721250772476196,
      "neg_sim_mean": 0.8136576414108276,
      "pos_sim_mean": 0.9132751226425171,
      "step": 2966
    },
    {
      "epoch": 0.5934,
      "grad_norm": 1.6956512928009033,
      "learning_rate": 4.2488720543962146e-05,
      "loss": 0.294,
      "step": 2967
    },
    {
      "epoch": 0.5934,
      "loss_align": 0.13714134693145752,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15706264972686768,
      "neg_sim_mean": 0.5628901124000549,
      "pos_sim_mean": 0.8628586530685425,
      "step": 2967
    },
    {
      "epoch": 0.5936,
      "grad_norm": 1.5904791355133057,
      "learning_rate": 4.2454211921016126e-05,
      "loss": 0.2942,
      "step": 2968
    },
    {
      "epoch": 0.5936,
      "loss_align": 0.1713653802871704,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10467816889286041,
      "neg_sim_mean": 0.5487302541732788,
      "pos_sim_mean": 0.8286346197128296,
      "step": 2968
    },
    {
      "epoch": 0.5938,
      "grad_norm": 1.333074688911438,
      "learning_rate": 4.241970697579557e-05,
      "loss": 0.276,
      "step": 2969
    },
    {
      "epoch": 0.5938,
      "loss_align": 0.18575704097747803,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009905710816383362,
      "neg_sim_mean": 0.2960929870605469,
      "pos_sim_mean": 0.814242959022522,
      "step": 2969
    },
    {
      "epoch": 0.594,
      "grad_norm": 0.740537166595459,
      "learning_rate": 4.238520572511773e-05,
      "loss": 0.1957,
      "step": 2970
    },
    {
      "epoch": 0.594,
      "loss_align": 0.12523555755615234,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23627501726150513,
      "neg_sim_mean": 0.7416194677352905,
      "pos_sim_mean": 0.8747644424438477,
      "step": 2970
    },
    {
      "epoch": 0.5942,
      "grad_norm": 2.025832176208496,
      "learning_rate": 4.23507081857981e-05,
      "loss": 0.3615,
      "step": 2971
    },
    {
      "epoch": 0.5942,
      "loss_align": 0.12814390659332275,
      "loss_contrastive": 0.10920792818069458,
      "loss_gen": 0.13581551611423492,
      "neg_sim_mean": 0.8810639977455139,
      "pos_sim_mean": 0.8718560934066772,
      "step": 2971
    },
    {
      "epoch": 0.5944,
      "grad_norm": 1.259069561958313,
      "learning_rate": 4.2316214374650396e-05,
      "loss": 0.2771,
      "step": 2972
    },
    {
      "epoch": 0.5944,
      "loss_align": 0.1841716766357422,
      "loss_contrastive": 0.0,
      "loss_gen": 0.267536461353302,
      "neg_sim_mean": 0.6211611032485962,
      "pos_sim_mean": 0.8158283233642578,
      "step": 2972
    },
    {
      "epoch": 0.5946,
      "grad_norm": 1.7772067785263062,
      "learning_rate": 4.228172430848644e-05,
      "loss": 0.4517,
      "step": 2973
    },
    {
      "epoch": 0.5946,
      "loss_align": 0.11519670486450195,
      "loss_contrastive": 0.0032494068145751953,
      "loss_gen": 0.28428807854652405,
      "neg_sim_mean": 0.7880526781082153,
      "pos_sim_mean": 0.884803295135498,
      "step": 2973
    },
    {
      "epoch": 0.5948,
      "grad_norm": 2.0383100509643555,
      "learning_rate": 4.224723800411631e-05,
      "loss": 0.3999,
      "step": 2974
    },
    {
      "epoch": 0.5948,
      "loss_align": 0.12462878227233887,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0017566761234775186,
      "neg_sim_mean": 0.32880890369415283,
      "pos_sim_mean": 0.8753712177276611,
      "step": 2974
    },
    {
      "epoch": 0.595,
      "grad_norm": 0.440975546836853,
      "learning_rate": 4.2212755478348196e-05,
      "loss": 0.1264,
      "step": 2975
    },
    {
      "epoch": 0.595,
      "loss_align": 0.11318254470825195,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011031334288418293,
      "neg_sim_mean": 0.2405986785888672,
      "pos_sim_mean": 0.886817455291748,
      "step": 2975
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.4552696943283081,
      "learning_rate": 4.2178276747988446e-05,
      "loss": 0.1143,
      "step": 2976
    },
    {
      "epoch": 0.5952,
      "loss_align": 0.3594817519187927,
      "loss_contrastive": 0.09737282991409302,
      "loss_gen": 0.2944304347038269,
      "neg_sim_mean": 0.6378910541534424,
      "pos_sim_mean": 0.6405182480812073,
      "step": 2976
    },
    {
      "epoch": 0.5954,
      "grad_norm": 2.064176321029663,
      "learning_rate": 4.214380182984163e-05,
      "loss": 0.6656,
      "step": 2977
    },
    {
      "epoch": 0.5954,
      "loss_align": 0.10534441471099854,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2201785445213318,
      "neg_sim_mean": 0.753217875957489,
      "pos_sim_mean": 0.8946555852890015,
      "step": 2977
    },
    {
      "epoch": 0.5956,
      "grad_norm": 1.9085553884506226,
      "learning_rate": 4.210933074071033e-05,
      "loss": 0.3255,
      "step": 2978
    },
    {
      "epoch": 0.5956,
      "loss_align": 0.26915842294692993,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010674962773919106,
      "neg_sim_mean": 0.3869238793849945,
      "pos_sim_mean": 0.7308415770530701,
      "step": 2978
    },
    {
      "epoch": 0.5958,
      "grad_norm": 1.158752202987671,
      "learning_rate": 4.2074863497395377e-05,
      "loss": 0.2798,
      "step": 2979
    },
    {
      "epoch": 0.5958,
      "loss_align": 0.11782807111740112,
      "loss_contrastive": 0.09719246625900269,
      "loss_gen": 0.1673688292503357,
      "neg_sim_mean": 0.8793643712997437,
      "pos_sim_mean": 0.8821719288825989,
      "step": 2979
    },
    {
      "epoch": 0.596,
      "grad_norm": 1.6825599670410156,
      "learning_rate": 4.2040400116695666e-05,
      "loss": 0.2969,
      "step": 2980
    },
    {
      "epoch": 0.596,
      "loss_align": 0.11817419528961182,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0023777151945978403,
      "neg_sim_mean": 0.4099771976470947,
      "pos_sim_mean": 0.8818258047103882,
      "step": 2980
    },
    {
      "epoch": 0.5962,
      "grad_norm": 0.3808690905570984,
      "learning_rate": 4.2005940615408264e-05,
      "loss": 0.1206,
      "step": 2981
    },
    {
      "epoch": 0.5962,
      "loss_align": 0.09334838390350342,
      "loss_contrastive": 0.07385993003845215,
      "loss_gen": 0.12733280658721924,
      "neg_sim_mean": 0.8805115222930908,
      "pos_sim_mean": 0.9066516160964966,
      "step": 2981
    },
    {
      "epoch": 0.5964,
      "grad_norm": 1.3842772245407104,
      "learning_rate": 4.197148501032829e-05,
      "loss": 0.2295,
      "step": 2982
    },
    {
      "epoch": 0.5964,
      "loss_align": 0.14193332195281982,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15327778458595276,
      "neg_sim_mean": 0.6755477786064148,
      "pos_sim_mean": 0.8580666780471802,
      "step": 2982
    },
    {
      "epoch": 0.5966,
      "grad_norm": 1.6213037967681885,
      "learning_rate": 4.193703331824897e-05,
      "loss": 0.2952,
      "step": 2983
    },
    {
      "epoch": 0.5966,
      "loss_align": 0.08667492866516113,
      "loss_contrastive": 0.0414767861366272,
      "loss_gen": 0.23189066350460052,
      "neg_sim_mean": 0.8548018336296082,
      "pos_sim_mean": 0.9133250713348389,
      "step": 2983
    },
    {
      "epoch": 0.5968,
      "grad_norm": 1.8193823099136353,
      "learning_rate": 4.190258555596168e-05,
      "loss": 0.3235,
      "step": 2984
    },
    {
      "epoch": 0.5968,
      "loss_align": 0.15809392929077148,
      "loss_contrastive": 0.002264261245727539,
      "loss_gen": 0.1127854511141777,
      "neg_sim_mean": 0.7441703081130981,
      "pos_sim_mean": 0.8419060707092285,
      "step": 2984
    },
    {
      "epoch": 0.597,
      "grad_norm": 1.4606457948684692,
      "learning_rate": 4.1868141740255823e-05,
      "loss": 0.2712,
      "step": 2985
    },
    {
      "epoch": 0.597,
      "loss_align": 0.08282029628753662,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3638535141944885,
      "neg_sim_mean": 0.6945931911468506,
      "pos_sim_mean": 0.9171797037124634,
      "step": 2985
    },
    {
      "epoch": 0.5972,
      "grad_norm": 2.2685186862945557,
      "learning_rate": 4.1833701887918904e-05,
      "loss": 0.4467,
      "step": 2986
    },
    {
      "epoch": 0.5972,
      "loss_align": 0.16253447532653809,
      "loss_contrastive": 0.0856286883354187,
      "loss_gen": 0.2928083539009094,
      "neg_sim_mean": 0.8230941891670227,
      "pos_sim_mean": 0.8374655246734619,
      "step": 2986
    },
    {
      "epoch": 0.5974,
      "grad_norm": 1.9274358749389648,
      "learning_rate": 4.1799266015736454e-05,
      "loss": 0.4656,
      "step": 2987
    },
    {
      "epoch": 0.5974,
      "loss_align": 0.19890880584716797,
      "loss_contrastive": 0.057607054710388184,
      "loss_gen": 0.1928209364414215,
      "neg_sim_mean": 0.7586982250213623,
      "pos_sim_mean": 0.801091194152832,
      "step": 2987
    },
    {
      "epoch": 0.5976,
      "grad_norm": 1.9314734935760498,
      "learning_rate": 4.176483414049214e-05,
      "loss": 0.3986,
      "step": 2988
    },
    {
      "epoch": 0.5976,
      "loss_align": 0.25234270095825195,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14638444781303406,
      "neg_sim_mean": 0.5903274416923523,
      "pos_sim_mean": 0.747657299041748,
      "step": 2988
    },
    {
      "epoch": 0.5978,
      "grad_norm": 1.1733112335205078,
      "learning_rate": 4.1730406278967624e-05,
      "loss": 0.3987,
      "step": 2989
    },
    {
      "epoch": 0.5978,
      "loss_align": 0.2955016493797302,
      "loss_contrastive": 0.11485475301742554,
      "loss_gen": 0.21839112043380737,
      "neg_sim_mean": 0.7193530797958374,
      "pos_sim_mean": 0.7044983506202698,
      "step": 2989
    },
    {
      "epoch": 0.598,
      "grad_norm": 2.4407050609588623,
      "learning_rate": 4.169598244794261e-05,
      "loss": 0.5277,
      "step": 2990
    },
    {
      "epoch": 0.598,
      "loss_align": 0.13435125350952148,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2609579563140869,
      "neg_sim_mean": 0.7439982891082764,
      "pos_sim_mean": 0.8656487464904785,
      "step": 2990
    },
    {
      "epoch": 0.5982,
      "grad_norm": 1.5174373388290405,
      "learning_rate": 4.166156266419489e-05,
      "loss": 0.3953,
      "step": 2991
    },
    {
      "epoch": 0.5982,
      "loss_align": 0.1382620930671692,
      "loss_contrastive": 0.012677192687988281,
      "loss_gen": 0.24731303751468658,
      "neg_sim_mean": 0.7744150757789612,
      "pos_sim_mean": 0.8617379069328308,
      "step": 2991
    },
    {
      "epoch": 0.5984,
      "grad_norm": 2.071125030517578,
      "learning_rate": 4.162714694450023e-05,
      "loss": 0.3871,
      "step": 2992
    },
    {
      "epoch": 0.5984,
      "loss_align": 0.10732930898666382,
      "loss_contrastive": 0.0,
      "loss_gen": 0.029527029022574425,
      "neg_sim_mean": 0.3243855834007263,
      "pos_sim_mean": 0.8926706910133362,
      "step": 2992
    },
    {
      "epoch": 0.5986,
      "grad_norm": 1.2221992015838623,
      "learning_rate": 4.1592735305632427e-05,
      "loss": 0.1369,
      "step": 2993
    },
    {
      "epoch": 0.5986,
      "loss_align": 0.20401418209075928,
      "loss_contrastive": 0.06714820861816406,
      "loss_gen": 0.16186746954917908,
      "neg_sim_mean": 0.7631340026855469,
      "pos_sim_mean": 0.7959858179092407,
      "step": 2993
    },
    {
      "epoch": 0.5988,
      "grad_norm": 1.5465365648269653,
      "learning_rate": 4.155832776436331e-05,
      "loss": 0.3739,
      "step": 2994
    },
    {
      "epoch": 0.5988,
      "loss_align": 0.029070138931274414,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011703980853781104,
      "neg_sim_mean": 0.2599021792411804,
      "pos_sim_mean": 0.9709298610687256,
      "step": 2994
    },
    {
      "epoch": 0.599,
      "grad_norm": 0.250725656747818,
      "learning_rate": 4.15239243374627e-05,
      "loss": 0.0302,
      "step": 2995
    },
    {
      "epoch": 0.599,
      "loss_align": 0.15630662441253662,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1301371306180954,
      "neg_sim_mean": 0.6844784617424011,
      "pos_sim_mean": 0.8436933755874634,
      "step": 2995
    },
    {
      "epoch": 0.5992,
      "grad_norm": 1.4898802042007446,
      "learning_rate": 4.1489525041698387e-05,
      "loss": 0.2864,
      "step": 2996
    },
    {
      "epoch": 0.5992,
      "loss_align": 0.25881946086883545,
      "loss_contrastive": 0.09316360950469971,
      "loss_gen": 0.22850151360034943,
      "neg_sim_mean": 0.7343441247940063,
      "pos_sim_mean": 0.7411805391311646,
      "step": 2996
    },
    {
      "epoch": 0.5994,
      "grad_norm": 1.562119960784912,
      "learning_rate": 4.1455129893836174e-05,
      "loss": 0.4985,
      "step": 2997
    },
    {
      "epoch": 0.5994,
      "loss_align": 0.09890604019165039,
      "loss_contrastive": 0.09961962699890137,
      "loss_gen": 0.2225441038608551,
      "neg_sim_mean": 0.9007135629653931,
      "pos_sim_mean": 0.9010939598083496,
      "step": 2997
    },
    {
      "epoch": 0.5996,
      "grad_norm": 1.6785165071487427,
      "learning_rate": 4.1420738910639855e-05,
      "loss": 0.3334,
      "step": 2998
    },
    {
      "epoch": 0.5996,
      "loss_align": 0.09263825416564941,
      "loss_contrastive": 0.0,
      "loss_gen": 0.026360005140304565,
      "neg_sim_mean": 0.3359234035015106,
      "pos_sim_mean": 0.9073617458343506,
      "step": 2998
    },
    {
      "epoch": 0.5998,
      "grad_norm": 1.0408438444137573,
      "learning_rate": 4.1386352108871174e-05,
      "loss": 0.119,
      "step": 2999
    },
    {
      "epoch": 0.5998,
      "loss_align": 0.18706512451171875,
      "loss_contrastive": 0.02227407693862915,
      "loss_gen": 0.442810595035553,
      "neg_sim_mean": 0.7352089285850525,
      "pos_sim_mean": 0.8129348754882812,
      "step": 2999
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.73258376121521,
      "learning_rate": 4.135196950528982e-05,
      "loss": 0.6325,
      "step": 3000
    },
    {
      "epoch": 0.6,
      "loss_align": 0.17396843433380127,
      "loss_contrastive": 0.07039636373519897,
      "loss_gen": 0.26646432280540466,
      "neg_sim_mean": 0.7964279055595398,
      "pos_sim_mean": 0.8260315656661987,
      "step": 3000
    },
    {
      "epoch": 0.6002,
      "grad_norm": 2.0253007411956787,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.4489,
      "step": 3001
    },
    {
      "epoch": 0.6002,
      "loss_align": 0.20654785633087158,
      "loss_contrastive": 0.16253817081451416,
      "loss_gen": 0.24928703904151917,
      "neg_sim_mean": 0.8559902906417847,
      "pos_sim_mean": 0.7934521436691284,
      "step": 3001
    },
    {
      "epoch": 0.6004,
      "grad_norm": 1.92239248752594,
      "learning_rate": 4.1283216959717754e-05,
      "loss": 0.4753,
      "step": 3002
    },
    {
      "epoch": 0.6004,
      "loss_align": 0.10726988315582275,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2549617886543274,
      "neg_sim_mean": 0.5819926261901855,
      "pos_sim_mean": 0.8927301168441772,
      "step": 3002
    },
    {
      "epoch": 0.6006,
      "grad_norm": 2.02827787399292,
      "learning_rate": 4.1248847051236195e-05,
      "loss": 0.3622,
      "step": 3003
    },
    {
      "epoch": 0.6006,
      "loss_align": 0.2240457534790039,
      "loss_contrastive": 0.033921778202056885,
      "loss_gen": 0.49956995248794556,
      "neg_sim_mean": 0.7098760008811951,
      "pos_sim_mean": 0.7759542465209961,
      "step": 3003
    },
    {
      "epoch": 0.6008,
      "grad_norm": 2.589972496032715,
      "learning_rate": 4.1214481407960284e-05,
      "loss": 0.7277,
      "step": 3004
    },
    {
      "epoch": 0.6008,
      "loss_align": 0.22614961862564087,
      "loss_contrastive": 0.0268782377243042,
      "loss_gen": 0.17129245400428772,
      "neg_sim_mean": 0.7007285952568054,
      "pos_sim_mean": 0.7738503813743591,
      "step": 3004
    },
    {
      "epoch": 0.601,
      "grad_norm": 1.9585838317871094,
      "learning_rate": 4.118012004663939e-05,
      "loss": 0.4007,
      "step": 3005
    },
    {
      "epoch": 0.601,
      "loss_align": 0.10182547569274902,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1137542799115181,
      "neg_sim_mean": 0.743815541267395,
      "pos_sim_mean": 0.898174524307251,
      "step": 3005
    },
    {
      "epoch": 0.6012,
      "grad_norm": 1.4338785409927368,
      "learning_rate": 4.114576298402084e-05,
      "loss": 0.2156,
      "step": 3006
    },
    {
      "epoch": 0.6012,
      "loss_align": 0.08748698234558105,
      "loss_contrastive": 0.03400319814682007,
      "loss_gen": 0.28660130500793457,
      "neg_sim_mean": 0.8465161919593811,
      "pos_sim_mean": 0.912513017654419,
      "step": 3006
    },
    {
      "epoch": 0.6014,
      "grad_norm": 2.4220240116119385,
      "learning_rate": 4.1111410236849855e-05,
      "loss": 0.3782,
      "step": 3007
    },
    {
      "epoch": 0.6014,
      "loss_align": 0.07599532604217529,
      "loss_contrastive": 0.07905739545822144,
      "loss_gen": 0.3518604636192322,
      "neg_sim_mean": 0.9030620455741882,
      "pos_sim_mean": 0.9240046739578247,
      "step": 3007
    },
    {
      "epoch": 0.6016,
      "grad_norm": 1.7463289499282837,
      "learning_rate": 4.1077061821869537e-05,
      "loss": 0.4373,
      "step": 3008
    },
    {
      "epoch": 0.6016,
      "loss_align": 0.12037128210067749,
      "loss_contrastive": 0.09899109601974487,
      "loss_gen": 0.2665926218032837,
      "neg_sim_mean": 0.8786197900772095,
      "pos_sim_mean": 0.8796287178993225,
      "step": 3008
    },
    {
      "epoch": 0.6018,
      "grad_norm": 2.4456324577331543,
      "learning_rate": 4.104271775582089e-05,
      "loss": 0.3988,
      "step": 3009
    },
    {
      "epoch": 0.6018,
      "loss_align": 0.1645498275756836,
      "loss_contrastive": 0.08175837993621826,
      "loss_gen": 0.3560808598995209,
      "neg_sim_mean": 0.8172085285186768,
      "pos_sim_mean": 0.8354501724243164,
      "step": 3009
    },
    {
      "epoch": 0.602,
      "grad_norm": 1.9438190460205078,
      "learning_rate": 4.100837805544279e-05,
      "loss": 0.5304,
      "step": 3010
    },
    {
      "epoch": 0.602,
      "loss_align": 0.12766343355178833,
      "loss_contrastive": 0.0,
      "loss_gen": 0.05734148994088173,
      "neg_sim_mean": 0.5600018501281738,
      "pos_sim_mean": 0.8723365664482117,
      "step": 3010
    },
    {
      "epoch": 0.6022,
      "grad_norm": 1.0154683589935303,
      "learning_rate": 4.0974042737472006e-05,
      "loss": 0.185,
      "step": 3011
    },
    {
      "epoch": 0.6022,
      "loss_align": 0.18688690662384033,
      "loss_contrastive": 0.04597806930541992,
      "loss_gen": 0.22146733105182648,
      "neg_sim_mean": 0.7590911388397217,
      "pos_sim_mean": 0.8131130933761597,
      "step": 3011
    },
    {
      "epoch": 0.6024,
      "grad_norm": 2.3587558269500732,
      "learning_rate": 4.093971181864313e-05,
      "loss": 0.4139,
      "step": 3012
    },
    {
      "epoch": 0.6024,
      "loss_align": 0.2529793381690979,
      "loss_contrastive": 0.027403295040130615,
      "loss_gen": 0.2403901368379593,
      "neg_sim_mean": 0.6744239330291748,
      "pos_sim_mean": 0.7470206618309021,
      "step": 3012
    },
    {
      "epoch": 0.6026,
      "grad_norm": 1.9005399942398071,
      "learning_rate": 4.0905385315688667e-05,
      "loss": 0.4967,
      "step": 3013
    },
    {
      "epoch": 0.6026,
      "loss_align": 0.19728386402130127,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007528005167841911,
      "neg_sim_mean": 0.34530192613601685,
      "pos_sim_mean": 0.8027161359786987,
      "step": 3013
    },
    {
      "epoch": 0.6028,
      "grad_norm": 0.712558925151825,
      "learning_rate": 4.087106324533891e-05,
      "loss": 0.2048,
      "step": 3014
    },
    {
      "epoch": 0.6028,
      "loss_align": 0.16443294286727905,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2711881995201111,
      "neg_sim_mean": 0.7121098637580872,
      "pos_sim_mean": 0.835567057132721,
      "step": 3014
    },
    {
      "epoch": 0.603,
      "grad_norm": 2.1333839893341064,
      "learning_rate": 4.0836745624322023e-05,
      "loss": 0.4356,
      "step": 3015
    },
    {
      "epoch": 0.603,
      "loss_align": 0.14178025722503662,
      "loss_contrastive": 0.022719264030456543,
      "loss_gen": 0.12244050949811935,
      "neg_sim_mean": 0.780938982963562,
      "pos_sim_mean": 0.8582197427749634,
      "step": 3015
    },
    {
      "epoch": 0.6032,
      "grad_norm": 1.275109052658081,
      "learning_rate": 4.080243246936399e-05,
      "loss": 0.2669,
      "step": 3016
    },
    {
      "epoch": 0.6032,
      "loss_align": 0.121135413646698,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17487667500972748,
      "neg_sim_mean": 0.7332245111465454,
      "pos_sim_mean": 0.878864586353302,
      "step": 3016
    },
    {
      "epoch": 0.6034,
      "grad_norm": 1.3209809064865112,
      "learning_rate": 4.076812379718866e-05,
      "loss": 0.296,
      "step": 3017
    },
    {
      "epoch": 0.6034,
      "loss_align": 0.11552739143371582,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01059484388679266,
      "neg_sim_mean": 0.2973489463329315,
      "pos_sim_mean": 0.8844726085662842,
      "step": 3017
    },
    {
      "epoch": 0.6036,
      "grad_norm": 0.6626250147819519,
      "learning_rate": 4.0733819624517634e-05,
      "loss": 0.1261,
      "step": 3018
    },
    {
      "epoch": 0.6036,
      "loss_align": 0.15996146202087402,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17195379734039307,
      "neg_sim_mean": 0.6391645073890686,
      "pos_sim_mean": 0.840038537979126,
      "step": 3018
    },
    {
      "epoch": 0.6038,
      "grad_norm": 1.6288907527923584,
      "learning_rate": 4.069951996807034e-05,
      "loss": 0.3319,
      "step": 3019
    },
    {
      "epoch": 0.6038,
      "loss_align": 0.08319461345672607,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14343805611133575,
      "neg_sim_mean": 0.5626333951950073,
      "pos_sim_mean": 0.9168053865432739,
      "step": 3019
    },
    {
      "epoch": 0.604,
      "grad_norm": 1.2052232027053833,
      "learning_rate": 4.0665224844564053e-05,
      "loss": 0.2266,
      "step": 3020
    },
    {
      "epoch": 0.604,
      "loss_align": 0.18862706422805786,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2537584602832794,
      "neg_sim_mean": 0.7051807641983032,
      "pos_sim_mean": 0.8113729357719421,
      "step": 3020
    },
    {
      "epoch": 0.6042,
      "grad_norm": 1.8771438598632812,
      "learning_rate": 4.063093427071376e-05,
      "loss": 0.4424,
      "step": 3021
    },
    {
      "epoch": 0.6042,
      "loss_align": 0.13183975219726562,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14903664588928223,
      "neg_sim_mean": 0.6048867702484131,
      "pos_sim_mean": 0.8681602478027344,
      "step": 3021
    },
    {
      "epoch": 0.6044,
      "grad_norm": 1.5315591096878052,
      "learning_rate": 4.0596648263232315e-05,
      "loss": 0.2809,
      "step": 3022
    },
    {
      "epoch": 0.6044,
      "loss_align": 0.23580670356750488,
      "loss_contrastive": 0.0076531171798706055,
      "loss_gen": 0.2650575637817383,
      "neg_sim_mean": 0.6718463897705078,
      "pos_sim_mean": 0.7641932964324951,
      "step": 3022
    },
    {
      "epoch": 0.6046,
      "grad_norm": 1.9444162845611572,
      "learning_rate": 4.056236683883025e-05,
      "loss": 0.5018,
      "step": 3023
    },
    {
      "epoch": 0.6046,
      "loss_align": 0.18041670322418213,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03040844015777111,
      "neg_sim_mean": 0.36936283111572266,
      "pos_sim_mean": 0.8195832967758179,
      "step": 3023
    },
    {
      "epoch": 0.6048,
      "grad_norm": 1.0589630603790283,
      "learning_rate": 4.0528090014215945e-05,
      "loss": 0.2108,
      "step": 3024
    },
    {
      "epoch": 0.6048,
      "loss_align": 0.17004787921905518,
      "loss_contrastive": 0.015178322792053223,
      "loss_gen": 0.29723915457725525,
      "neg_sim_mean": 0.7451304197311401,
      "pos_sim_mean": 0.8299521207809448,
      "step": 3024
    },
    {
      "epoch": 0.605,
      "grad_norm": 1.8557639122009277,
      "learning_rate": 4.04938178060955e-05,
      "loss": 0.4691,
      "step": 3025
    },
    {
      "epoch": 0.605,
      "loss_align": 0.1668565273284912,
      "loss_contrastive": 0.08185279369354248,
      "loss_gen": 0.16794565320014954,
      "neg_sim_mean": 0.8149962425231934,
      "pos_sim_mean": 0.8331434726715088,
      "step": 3025
    },
    {
      "epoch": 0.6052,
      "grad_norm": 1.4989944696426392,
      "learning_rate": 4.045955023117276e-05,
      "loss": 0.3446,
      "step": 3026
    },
    {
      "epoch": 0.6052,
      "loss_align": 0.11192518472671509,
      "loss_contrastive": 0.1087905764579773,
      "loss_gen": 0.3015197217464447,
      "neg_sim_mean": 0.8968653678894043,
      "pos_sim_mean": 0.8880748152732849,
      "step": 3026
    },
    {
      "epoch": 0.6054,
      "grad_norm": 1.810145616531372,
      "learning_rate": 4.042528730614936e-05,
      "loss": 0.4265,
      "step": 3027
    },
    {
      "epoch": 0.6054,
      "loss_align": 0.08651560544967651,
      "loss_contrastive": 0.08778047561645508,
      "loss_gen": 0.571017861366272,
      "neg_sim_mean": 0.9012648463249207,
      "pos_sim_mean": 0.9134843945503235,
      "step": 3027
    },
    {
      "epoch": 0.6056,
      "grad_norm": 2.9417574405670166,
      "learning_rate": 4.039102904772459e-05,
      "loss": 0.6681,
      "step": 3028
    },
    {
      "epoch": 0.6056,
      "loss_align": 0.16626393795013428,
      "loss_contrastive": 0.16304731369018555,
      "loss_gen": 0.36165526509284973,
      "neg_sim_mean": 0.8967833518981934,
      "pos_sim_mean": 0.8337360620498657,
      "step": 3028
    },
    {
      "epoch": 0.6058,
      "grad_norm": 2.1343932151794434,
      "learning_rate": 4.0356775472595546e-05,
      "loss": 0.5475,
      "step": 3029
    },
    {
      "epoch": 0.6058,
      "loss_align": 0.08213520050048828,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09863612800836563,
      "neg_sim_mean": 0.7915875911712646,
      "pos_sim_mean": 0.9178647994995117,
      "step": 3029
    },
    {
      "epoch": 0.606,
      "grad_norm": 1.2853128910064697,
      "learning_rate": 4.032252659745699e-05,
      "loss": 0.1808,
      "step": 3030
    },
    {
      "epoch": 0.606,
      "loss_align": 0.09125363826751709,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0018379359971731901,
      "neg_sim_mean": 0.3974336087703705,
      "pos_sim_mean": 0.9087463617324829,
      "step": 3030
    },
    {
      "epoch": 0.6062,
      "grad_norm": 0.32561349868774414,
      "learning_rate": 4.028828243900141e-05,
      "loss": 0.0931,
      "step": 3031
    },
    {
      "epoch": 0.6062,
      "loss_align": 0.1545426845550537,
      "loss_contrastive": 0.0,
      "loss_gen": 0.07996582984924316,
      "neg_sim_mean": 0.7432929873466492,
      "pos_sim_mean": 0.8454573154449463,
      "step": 3031
    },
    {
      "epoch": 0.6064,
      "grad_norm": 1.314265251159668,
      "learning_rate": 4.0254043013918976e-05,
      "loss": 0.2345,
      "step": 3032
    },
    {
      "epoch": 0.6064,
      "loss_align": 0.37924718856811523,
      "loss_contrastive": 0.32317596673965454,
      "loss_gen": 0.26722919940948486,
      "neg_sim_mean": 0.8439287543296814,
      "pos_sim_mean": 0.6207528114318848,
      "step": 3032
    },
    {
      "epoch": 0.6066,
      "grad_norm": 2.012989044189453,
      "learning_rate": 4.02198083388976e-05,
      "loss": 0.6853,
      "step": 3033
    },
    {
      "epoch": 0.6066,
      "loss_align": 0.20353448390960693,
      "loss_contrastive": 0.09117132425308228,
      "loss_gen": 0.16615864634513855,
      "neg_sim_mean": 0.7876368165016174,
      "pos_sim_mean": 0.7964655160903931,
      "step": 3033
    },
    {
      "epoch": 0.6068,
      "grad_norm": 1.5581954717636108,
      "learning_rate": 4.018557843062282e-05,
      "loss": 0.3806,
      "step": 3034
    },
    {
      "epoch": 0.6068,
      "loss_align": 0.10470902919769287,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24891938269138336,
      "neg_sim_mean": 0.6485523581504822,
      "pos_sim_mean": 0.8952909708023071,
      "step": 3034
    },
    {
      "epoch": 0.607,
      "grad_norm": 1.572929859161377,
      "learning_rate": 4.015135330577787e-05,
      "loss": 0.3536,
      "step": 3035
    },
    {
      "epoch": 0.607,
      "loss_align": 0.09480893611907959,
      "loss_contrastive": 0.05086171627044678,
      "loss_gen": 0.1972845494747162,
      "neg_sim_mean": 0.8560527563095093,
      "pos_sim_mean": 0.9051910638809204,
      "step": 3035
    },
    {
      "epoch": 0.6072,
      "grad_norm": 2.0134735107421875,
      "learning_rate": 4.0117132981043693e-05,
      "loss": 0.2982,
      "step": 3036
    },
    {
      "epoch": 0.6072,
      "loss_align": 0.054825544357299805,
      "loss_contrastive": 0.051081180572509766,
      "loss_gen": 0.03288602828979492,
      "neg_sim_mean": 0.896255612373352,
      "pos_sim_mean": 0.9451744556427002,
      "step": 3036
    },
    {
      "epoch": 0.6074,
      "grad_norm": 0.7603894472122192,
      "learning_rate": 4.008291747309884e-05,
      "loss": 0.0938,
      "step": 3037
    },
    {
      "epoch": 0.6074,
      "loss_align": 0.12322467565536499,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013253008015453815,
      "neg_sim_mean": 0.2978566884994507,
      "pos_sim_mean": 0.876775324344635,
      "step": 3037
    },
    {
      "epoch": 0.6076,
      "grad_norm": 0.8166120648384094,
      "learning_rate": 4.004870679861953e-05,
      "loss": 0.1365,
      "step": 3038
    },
    {
      "epoch": 0.6076,
      "loss_align": 0.16429072618484497,
      "loss_contrastive": 0.06167024374008179,
      "loss_gen": 0.21127983927726746,
      "neg_sim_mean": 0.7973794937133789,
      "pos_sim_mean": 0.835709273815155,
      "step": 3038
    },
    {
      "epoch": 0.6078,
      "grad_norm": 1.9380465745925903,
      "learning_rate": 4.001450097427966e-05,
      "loss": 0.383,
      "step": 3039
    },
    {
      "epoch": 0.6078,
      "loss_align": 0.113558828830719,
      "loss_contrastive": 0.03309148550033569,
      "loss_gen": 0.25814980268478394,
      "neg_sim_mean": 0.8195326328277588,
      "pos_sim_mean": 0.886441171169281,
      "step": 3039
    },
    {
      "epoch": 0.608,
      "grad_norm": 1.9019110202789307,
      "learning_rate": 3.9980300016750694e-05,
      "loss": 0.3757,
      "step": 3040
    },
    {
      "epoch": 0.608,
      "loss_align": 0.12646842002868652,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26211991906166077,
      "neg_sim_mean": 0.6717888712882996,
      "pos_sim_mean": 0.8735315799713135,
      "step": 3040
    },
    {
      "epoch": 0.6082,
      "grad_norm": 2.1522719860076904,
      "learning_rate": 3.9946103942701777e-05,
      "loss": 0.3886,
      "step": 3041
    },
    {
      "epoch": 0.6082,
      "loss_align": 0.2023448348045349,
      "loss_contrastive": 0.11986511945724487,
      "loss_gen": 0.17174063622951508,
      "neg_sim_mean": 0.817520260810852,
      "pos_sim_mean": 0.7976551651954651,
      "step": 3041
    },
    {
      "epoch": 0.6084,
      "grad_norm": 1.5413644313812256,
      "learning_rate": 3.9911912768799655e-05,
      "loss": 0.3885,
      "step": 3042
    },
    {
      "epoch": 0.6084,
      "loss_align": 0.08999514579772949,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004756464157253504,
      "neg_sim_mean": 0.29507601261138916,
      "pos_sim_mean": 0.9100048542022705,
      "step": 3042
    },
    {
      "epoch": 0.6086,
      "grad_norm": 0.49131718277931213,
      "learning_rate": 3.9877726511708715e-05,
      "loss": 0.0948,
      "step": 3043
    },
    {
      "epoch": 0.6086,
      "loss_align": 0.16565853357315063,
      "loss_contrastive": 0.059295594692230225,
      "loss_gen": 0.2634049355983734,
      "neg_sim_mean": 0.7936370372772217,
      "pos_sim_mean": 0.8343414664268494,
      "step": 3043
    },
    {
      "epoch": 0.6088,
      "grad_norm": 1.7894669771194458,
      "learning_rate": 3.98435451880909e-05,
      "loss": 0.4362,
      "step": 3044
    },
    {
      "epoch": 0.6088,
      "loss_align": 0.0973777174949646,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15575610101222992,
      "neg_sim_mean": 0.6953991651535034,
      "pos_sim_mean": 0.9026222825050354,
      "step": 3044
    },
    {
      "epoch": 0.609,
      "grad_norm": 1.3497101068496704,
      "learning_rate": 3.9809368814605766e-05,
      "loss": 0.2531,
      "step": 3045
    },
    {
      "epoch": 0.609,
      "loss_align": 0.1273963451385498,
      "loss_contrastive": 0.030572056770324707,
      "loss_gen": 0.2756554186344147,
      "neg_sim_mean": 0.803175687789917,
      "pos_sim_mean": 0.8726036548614502,
      "step": 3045
    },
    {
      "epoch": 0.6092,
      "grad_norm": 2.1070947647094727,
      "learning_rate": 3.9775197407910485e-05,
      "loss": 0.4067,
      "step": 3046
    },
    {
      "epoch": 0.6092,
      "loss_align": 0.2355351448059082,
      "loss_contrastive": 0.1446441411972046,
      "loss_gen": 0.14675745368003845,
      "neg_sim_mean": 0.8091089725494385,
      "pos_sim_mean": 0.7644648551940918,
      "step": 3046
    },
    {
      "epoch": 0.6094,
      "grad_norm": 1.6896826028823853,
      "learning_rate": 3.974103098465977e-05,
      "loss": 0.3996,
      "step": 3047
    },
    {
      "epoch": 0.6094,
      "loss_align": 0.3684653639793396,
      "loss_contrastive": 0.13842928409576416,
      "loss_gen": 0.2393369972705841,
      "neg_sim_mean": 0.6699638962745667,
      "pos_sim_mean": 0.6315346360206604,
      "step": 3047
    },
    {
      "epoch": 0.6096,
      "grad_norm": 2.6458356380462646,
      "learning_rate": 3.9706869561505946e-05,
      "loss": 0.6244,
      "step": 3048
    },
    {
      "epoch": 0.6096,
      "loss_align": 0.08953934907913208,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12055765837430954,
      "neg_sim_mean": 0.7511990666389465,
      "pos_sim_mean": 0.9104606509208679,
      "step": 3048
    },
    {
      "epoch": 0.6098,
      "grad_norm": 1.435805320739746,
      "learning_rate": 3.967271315509884e-05,
      "loss": 0.2101,
      "step": 3049
    },
    {
      "epoch": 0.6098,
      "loss_align": 0.11734116077423096,
      "loss_contrastive": 0.054434895515441895,
      "loss_gen": 0.17502430081367493,
      "neg_sim_mean": 0.837093710899353,
      "pos_sim_mean": 0.882658839225769,
      "step": 3049
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5484836101531982,
      "learning_rate": 3.963856178208588e-05,
      "loss": 0.2989,
      "step": 3050
    },
    {
      "epoch": 0.61,
      "loss_align": 0.13397979736328125,
      "loss_contrastive": 0.040352463722229004,
      "loss_gen": 0.19216300547122955,
      "neg_sim_mean": 0.8063726425170898,
      "pos_sim_mean": 0.8660202026367188,
      "step": 3050
    },
    {
      "epoch": 0.6102,
      "grad_norm": 1.835624098777771,
      "learning_rate": 3.960441545911204e-05,
      "loss": 0.331,
      "step": 3051
    },
    {
      "epoch": 0.6102,
      "loss_align": 0.19795989990234375,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3214419484138489,
      "neg_sim_mean": 0.5433002710342407,
      "pos_sim_mean": 0.8020401000976562,
      "step": 3051
    },
    {
      "epoch": 0.6104,
      "grad_norm": 2.050417423248291,
      "learning_rate": 3.9570274202819815e-05,
      "loss": 0.5194,
      "step": 3052
    },
    {
      "epoch": 0.6104,
      "loss_align": 0.0864221453666687,
      "loss_contrastive": 0.037464141845703125,
      "loss_gen": 0.22080294787883759,
      "neg_sim_mean": 0.8510419726371765,
      "pos_sim_mean": 0.9135778546333313,
      "step": 3052
    },
    {
      "epoch": 0.6106,
      "grad_norm": 1.6483492851257324,
      "learning_rate": 3.9536138029849244e-05,
      "loss": 0.3117,
      "step": 3053
    },
    {
      "epoch": 0.6106,
      "loss_align": 0.29914265871047974,
      "loss_contrastive": 0.17092525959014893,
      "loss_gen": 0.3110210597515106,
      "neg_sim_mean": 0.7717825770378113,
      "pos_sim_mean": 0.7008573412895203,
      "step": 3053
    },
    {
      "epoch": 0.6108,
      "grad_norm": 2.144818067550659,
      "learning_rate": 3.950200695683788e-05,
      "loss": 0.6307,
      "step": 3054
    },
    {
      "epoch": 0.6108,
      "loss_align": 0.1385454535484314,
      "loss_contrastive": 0.0,
      "loss_gen": 0.021008484065532684,
      "neg_sim_mean": 0.3264753818511963,
      "pos_sim_mean": 0.8614545464515686,
      "step": 3054
    },
    {
      "epoch": 0.611,
      "grad_norm": 0.7665054202079773,
      "learning_rate": 3.9467881000420796e-05,
      "loss": 0.1596,
      "step": 3055
    },
    {
      "epoch": 0.611,
      "loss_align": 0.106192946434021,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14101967215538025,
      "neg_sim_mean": 0.479633092880249,
      "pos_sim_mean": 0.893807053565979,
      "step": 3055
    },
    {
      "epoch": 0.6112,
      "grad_norm": 1.2753592729568481,
      "learning_rate": 3.943376017723057e-05,
      "loss": 0.2472,
      "step": 3056
    },
    {
      "epoch": 0.6112,
      "loss_align": 0.17641258239746094,
      "loss_contrastive": 0.0378495454788208,
      "loss_gen": 0.2698139548301697,
      "neg_sim_mean": 0.761436939239502,
      "pos_sim_mean": 0.8235874176025391,
      "step": 3056
    },
    {
      "epoch": 0.6114,
      "grad_norm": 1.993514060974121,
      "learning_rate": 3.939964450389728e-05,
      "loss": 0.4508,
      "step": 3057
    },
    {
      "epoch": 0.6114,
      "loss_align": 0.15212863683700562,
      "loss_contrastive": 0.11034637689590454,
      "loss_gen": 0.20508623123168945,
      "neg_sim_mean": 0.858217716217041,
      "pos_sim_mean": 0.8478713631629944,
      "step": 3057
    },
    {
      "epoch": 0.6116,
      "grad_norm": 2.148800849914551,
      "learning_rate": 3.936553399704848e-05,
      "loss": 0.3705,
      "step": 3058
    },
    {
      "epoch": 0.6116,
      "loss_align": 0.21017098426818848,
      "loss_contrastive": 0.02149033546447754,
      "loss_gen": 0.3676621913909912,
      "neg_sim_mean": 0.7113193273544312,
      "pos_sim_mean": 0.7898290157318115,
      "step": 3058
    },
    {
      "epoch": 0.6118,
      "grad_norm": 2.031782627105713,
      "learning_rate": 3.9331428673309204e-05,
      "loss": 0.5804,
      "step": 3059
    },
    {
      "epoch": 0.6118,
      "loss_align": 0.09779852628707886,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13271935284137726,
      "neg_sim_mean": 0.5492206811904907,
      "pos_sim_mean": 0.9022014737129211,
      "step": 3059
    },
    {
      "epoch": 0.612,
      "grad_norm": 1.5761618614196777,
      "learning_rate": 3.9297328549302e-05,
      "loss": 0.2305,
      "step": 3060
    },
    {
      "epoch": 0.612,
      "loss_align": 0.14305925369262695,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10723090916872025,
      "neg_sim_mean": 0.6534595489501953,
      "pos_sim_mean": 0.856940746307373,
      "step": 3060
    },
    {
      "epoch": 0.6122,
      "grad_norm": 0.9416699409484863,
      "learning_rate": 3.926323364164684e-05,
      "loss": 0.2503,
      "step": 3061
    },
    {
      "epoch": 0.6122,
      "loss_align": 0.12425220012664795,
      "loss_contrastive": 0.04732072353363037,
      "loss_gen": 0.1842324286699295,
      "neg_sim_mean": 0.8230684995651245,
      "pos_sim_mean": 0.875747799873352,
      "step": 3061
    },
    {
      "epoch": 0.6124,
      "grad_norm": 1.496216058731079,
      "learning_rate": 3.922914396696118e-05,
      "loss": 0.3142,
      "step": 3062
    },
    {
      "epoch": 0.6124,
      "loss_align": 0.08725762367248535,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1810186356306076,
      "neg_sim_mean": 0.7478008270263672,
      "pos_sim_mean": 0.9127423763275146,
      "step": 3062
    },
    {
      "epoch": 0.6126,
      "grad_norm": 1.7114378213882446,
      "learning_rate": 3.91950595418599e-05,
      "loss": 0.2683,
      "step": 3063
    },
    {
      "epoch": 0.6126,
      "loss_align": 0.13422703742980957,
      "loss_contrastive": 0.04024708271026611,
      "loss_gen": 0.19409486651420593,
      "neg_sim_mean": 0.8060200214385986,
      "pos_sim_mean": 0.8657729625701904,
      "step": 3063
    },
    {
      "epoch": 0.6128,
      "grad_norm": 1.8915350437164307,
      "learning_rate": 3.916098038295534e-05,
      "loss": 0.3332,
      "step": 3064
    },
    {
      "epoch": 0.6128,
      "loss_align": 0.17302954196929932,
      "loss_contrastive": 0.09232848882675171,
      "loss_gen": 0.3667012155056,
      "neg_sim_mean": 0.8192989230155945,
      "pos_sim_mean": 0.8269704580307007,
      "step": 3064
    },
    {
      "epoch": 0.613,
      "grad_norm": 2.372328519821167,
      "learning_rate": 3.912690650685726e-05,
      "loss": 0.5508,
      "step": 3065
    },
    {
      "epoch": 0.613,
      "loss_align": 0.1256769895553589,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028869636356830597,
      "neg_sim_mean": 0.21344570815563202,
      "pos_sim_mean": 0.8743230104446411,
      "step": 3065
    },
    {
      "epoch": 0.6132,
      "grad_norm": 1.0877385139465332,
      "learning_rate": 3.9092837930172884e-05,
      "loss": 0.1545,
      "step": 3066
    },
    {
      "epoch": 0.6132,
      "loss_align": 0.1389119029045105,
      "loss_contrastive": 0.06733965873718262,
      "loss_gen": 0.1884162575006485,
      "neg_sim_mean": 0.8284277319908142,
      "pos_sim_mean": 0.8610880970954895,
      "step": 3066
    },
    {
      "epoch": 0.6134,
      "grad_norm": 1.7581862211227417,
      "learning_rate": 3.905877466950679e-05,
      "loss": 0.3354,
      "step": 3067
    },
    {
      "epoch": 0.6134,
      "loss_align": 0.32898133993148804,
      "loss_contrastive": 0.029638826847076416,
      "loss_gen": 0.24239473044872284,
      "neg_sim_mean": 0.6006574630737305,
      "pos_sim_mean": 0.671018660068512,
      "step": 3067
    },
    {
      "epoch": 0.6136,
      "grad_norm": 2.1658337116241455,
      "learning_rate": 3.9024716741460986e-05,
      "loss": 0.5749,
      "step": 3068
    },
    {
      "epoch": 0.6136,
      "loss_align": 0.07843214273452759,
      "loss_contrastive": 0.060131609439849854,
      "loss_gen": 0.18012982606887817,
      "neg_sim_mean": 0.8816994428634644,
      "pos_sim_mean": 0.9215678572654724,
      "step": 3068
    },
    {
      "epoch": 0.6138,
      "grad_norm": 1.6116828918457031,
      "learning_rate": 3.899066416263493e-05,
      "loss": 0.2658,
      "step": 3069
    },
    {
      "epoch": 0.6138,
      "loss_align": 0.09876829385757446,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2631104290485382,
      "neg_sim_mean": 0.7096273899078369,
      "pos_sim_mean": 0.9012317061424255,
      "step": 3069
    },
    {
      "epoch": 0.614,
      "grad_norm": 2.1021265983581543,
      "learning_rate": 3.895661694962542e-05,
      "loss": 0.3619,
      "step": 3070
    },
    {
      "epoch": 0.614,
      "loss_align": 0.12632715702056885,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006107219494879246,
      "neg_sim_mean": 0.30382239818573,
      "pos_sim_mean": 0.8736728429794312,
      "step": 3070
    },
    {
      "epoch": 0.6142,
      "grad_norm": 0.46969854831695557,
      "learning_rate": 3.892257511902664e-05,
      "loss": 0.1324,
      "step": 3071
    },
    {
      "epoch": 0.6142,
      "loss_align": 0.041567087173461914,
      "loss_contrastive": 0.08118784427642822,
      "loss_gen": 0.12522874772548676,
      "neg_sim_mean": 0.9396207332611084,
      "pos_sim_mean": 0.9584329128265381,
      "step": 3071
    },
    {
      "epoch": 0.6144,
      "grad_norm": 1.3686861991882324,
      "learning_rate": 3.8888538687430184e-05,
      "loss": 0.1765,
      "step": 3072
    },
    {
      "epoch": 0.6144,
      "loss_align": 0.13353204727172852,
      "loss_contrastive": 0.11757057905197144,
      "loss_gen": 0.25860288739204407,
      "neg_sim_mean": 0.884038507938385,
      "pos_sim_mean": 0.8664679527282715,
      "step": 3072
    },
    {
      "epoch": 0.6146,
      "grad_norm": 2.1074836254119873,
      "learning_rate": 3.885450767142498e-05,
      "loss": 0.4062,
      "step": 3073
    },
    {
      "epoch": 0.6146,
      "loss_align": 0.2744797468185425,
      "loss_contrastive": 0.03039371967315674,
      "loss_gen": 0.3001633584499359,
      "neg_sim_mean": 0.6559139490127563,
      "pos_sim_mean": 0.7255202531814575,
      "step": 3073
    },
    {
      "epoch": 0.6148,
      "grad_norm": 1.9690666198730469,
      "learning_rate": 3.882048208759735e-05,
      "loss": 0.5783,
      "step": 3074
    },
    {
      "epoch": 0.6148,
      "loss_align": 0.10028231143951416,
      "loss_contrastive": 0.08073759078979492,
      "loss_gen": 0.3204226493835449,
      "neg_sim_mean": 0.8804552555084229,
      "pos_sim_mean": 0.8997176885604858,
      "step": 3074
    },
    {
      "epoch": 0.615,
      "grad_norm": 2.192202568054199,
      "learning_rate": 3.878646195253095e-05,
      "loss": 0.4304,
      "step": 3075
    },
    {
      "epoch": 0.615,
      "loss_align": 0.11559879779815674,
      "loss_contrastive": 0.0,
      "loss_gen": 0.025031324476003647,
      "neg_sim_mean": 0.28084084391593933,
      "pos_sim_mean": 0.8844012022018433,
      "step": 3075
    },
    {
      "epoch": 0.6152,
      "grad_norm": 1.082011342048645,
      "learning_rate": 3.875244728280676e-05,
      "loss": 0.1406,
      "step": 3076
    },
    {
      "epoch": 0.6152,
      "loss_align": 0.1032480001449585,
      "loss_contrastive": 0.022231221199035645,
      "loss_gen": 0.1728050261735916,
      "neg_sim_mean": 0.8189831972122192,
      "pos_sim_mean": 0.8967519998550415,
      "step": 3076
    },
    {
      "epoch": 0.6154,
      "grad_norm": 1.3188923597335815,
      "learning_rate": 3.871843809500313e-05,
      "loss": 0.2787,
      "step": 3077
    },
    {
      "epoch": 0.6154,
      "loss_align": 0.14500021934509277,
      "loss_contrastive": 0.0975039005279541,
      "loss_gen": 0.24963709712028503,
      "neg_sim_mean": 0.8525036573410034,
      "pos_sim_mean": 0.8549997806549072,
      "step": 3077
    },
    {
      "epoch": 0.6156,
      "grad_norm": 1.9986978769302368,
      "learning_rate": 3.868443440569571e-05,
      "loss": 0.4063,
      "step": 3078
    },
    {
      "epoch": 0.6156,
      "loss_align": 0.04985535144805908,
      "loss_contrastive": 0.006865143775939941,
      "loss_gen": 0.16731911897659302,
      "neg_sim_mean": 0.857009768486023,
      "pos_sim_mean": 0.9501446485519409,
      "step": 3078
    },
    {
      "epoch": 0.6158,
      "grad_norm": 1.8024036884307861,
      "learning_rate": 3.865043623145751e-05,
      "loss": 0.218,
      "step": 3079
    },
    {
      "epoch": 0.6158,
      "loss_align": 0.1563248634338379,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14418180286884308,
      "neg_sim_mean": 0.571858823299408,
      "pos_sim_mean": 0.8436751365661621,
      "step": 3079
    },
    {
      "epoch": 0.616,
      "grad_norm": 1.475623607635498,
      "learning_rate": 3.8616443588858805e-05,
      "loss": 0.3005,
      "step": 3080
    },
    {
      "epoch": 0.616,
      "loss_align": 0.20170670747756958,
      "loss_contrastive": 0.0546761155128479,
      "loss_gen": 0.1950068175792694,
      "neg_sim_mean": 0.7529693841934204,
      "pos_sim_mean": 0.7982932925224304,
      "step": 3080
    },
    {
      "epoch": 0.6162,
      "grad_norm": 1.7116400003433228,
      "learning_rate": 3.858245649446721e-05,
      "loss": 0.4033,
      "step": 3081
    },
    {
      "epoch": 0.6162,
      "loss_align": 0.09144806861877441,
      "loss_contrastive": 0.048609256744384766,
      "loss_gen": 0.3164806663990021,
      "neg_sim_mean": 0.8571611642837524,
      "pos_sim_mean": 0.9085519313812256,
      "step": 3081
    },
    {
      "epoch": 0.6164,
      "grad_norm": 2.1426734924316406,
      "learning_rate": 3.854847496484763e-05,
      "loss": 0.4138,
      "step": 3082
    },
    {
      "epoch": 0.6164,
      "loss_align": 0.1173701286315918,
      "loss_contrastive": 0.0,
      "loss_gen": 0.015820808708667755,
      "neg_sim_mean": 0.27857497334480286,
      "pos_sim_mean": 0.8826298713684082,
      "step": 3082
    },
    {
      "epoch": 0.6166,
      "grad_norm": 0.8055036067962646,
      "learning_rate": 3.8514499016562214e-05,
      "loss": 0.1332,
      "step": 3083
    },
    {
      "epoch": 0.6166,
      "loss_align": 0.12230640649795532,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0041192201897501945,
      "neg_sim_mean": 0.3353162407875061,
      "pos_sim_mean": 0.8776935935020447,
      "step": 3083
    },
    {
      "epoch": 0.6168,
      "grad_norm": 0.5156547427177429,
      "learning_rate": 3.848052866617049e-05,
      "loss": 0.1264,
      "step": 3084
    },
    {
      "epoch": 0.6168,
      "loss_align": 0.09467136859893799,
      "loss_contrastive": 0.051356494426727295,
      "loss_gen": 0.2100343555212021,
      "neg_sim_mean": 0.8566851019859314,
      "pos_sim_mean": 0.905328631401062,
      "step": 3084
    },
    {
      "epoch": 0.617,
      "grad_norm": 1.5795292854309082,
      "learning_rate": 3.844656393022912e-05,
      "loss": 0.3109,
      "step": 3085
    },
    {
      "epoch": 0.617,
      "loss_align": 0.14327514171600342,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2590189576148987,
      "neg_sim_mean": 0.7162895202636719,
      "pos_sim_mean": 0.8567248582839966,
      "step": 3085
    },
    {
      "epoch": 0.6172,
      "grad_norm": 1.6832425594329834,
      "learning_rate": 3.841260482529214e-05,
      "loss": 0.4023,
      "step": 3086
    },
    {
      "epoch": 0.6172,
      "loss_align": 0.1180945634841919,
      "loss_contrastive": 0.1298975944519043,
      "loss_gen": 0.6453143954277039,
      "neg_sim_mean": 0.9118030071258545,
      "pos_sim_mean": 0.8819054365158081,
      "step": 3086
    },
    {
      "epoch": 0.6174,
      "grad_norm": 3.303945302963257,
      "learning_rate": 3.83786513679108e-05,
      "loss": 0.779,
      "step": 3087
    },
    {
      "epoch": 0.6174,
      "loss_align": 0.41920435428619385,
      "loss_contrastive": 0.16417357325553894,
      "loss_gen": 0.11646058410406113,
      "neg_sim_mean": 0.6449692249298096,
      "pos_sim_mean": 0.5807956457138062,
      "step": 3087
    },
    {
      "epoch": 0.6176,
      "grad_norm": 1.864845871925354,
      "learning_rate": 3.834470357463362e-05,
      "loss": 0.5554,
      "step": 3088
    },
    {
      "epoch": 0.6176,
      "loss_align": 0.02447950839996338,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12779481709003448,
      "neg_sim_mean": 0.603477418422699,
      "pos_sim_mean": 0.9755204916000366,
      "step": 3088
    },
    {
      "epoch": 0.6178,
      "grad_norm": 0.9751842021942139,
      "learning_rate": 3.831076146200632e-05,
      "loss": 0.1523,
      "step": 3089
    },
    {
      "epoch": 0.6178,
      "loss_align": 0.0687834620475769,
      "loss_contrastive": 0.0,
      "loss_gen": 0.018439415842294693,
      "neg_sim_mean": 0.37449097633361816,
      "pos_sim_mean": 0.9312165379524231,
      "step": 3089
    },
    {
      "epoch": 0.618,
      "grad_norm": 0.9350916743278503,
      "learning_rate": 3.827682504657187e-05,
      "loss": 0.0872,
      "step": 3090
    },
    {
      "epoch": 0.618,
      "loss_align": 0.18641722202301025,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00902440957725048,
      "neg_sim_mean": 0.31996485590934753,
      "pos_sim_mean": 0.8135827779769897,
      "step": 3090
    },
    {
      "epoch": 0.6182,
      "grad_norm": 0.7751882076263428,
      "learning_rate": 3.82428943448705e-05,
      "loss": 0.1954,
      "step": 3091
    },
    {
      "epoch": 0.6182,
      "loss_align": 0.24422776699066162,
      "loss_contrastive": 0.0,
      "loss_gen": 0.026641566306352615,
      "neg_sim_mean": 0.4903624355792999,
      "pos_sim_mean": 0.7557722330093384,
      "step": 3091
    },
    {
      "epoch": 0.6184,
      "grad_norm": 0.8073838353157043,
      "learning_rate": 3.820896937343959e-05,
      "loss": 0.2709,
      "step": 3092
    },
    {
      "epoch": 0.6184,
      "loss_align": 0.13214147090911865,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010589917190372944,
      "neg_sim_mean": 0.4789116084575653,
      "pos_sim_mean": 0.8678585290908813,
      "step": 3092
    },
    {
      "epoch": 0.6186,
      "grad_norm": 0.6943581104278564,
      "learning_rate": 3.817505014881378e-05,
      "loss": 0.1427,
      "step": 3093
    },
    {
      "epoch": 0.6186,
      "loss_align": 0.07766139507293701,
      "loss_contrastive": 0.03095453977584839,
      "loss_gen": 0.18555855751037598,
      "neg_sim_mean": 0.8532931208610535,
      "pos_sim_mean": 0.922338604927063,
      "step": 3093
    },
    {
      "epoch": 0.6188,
      "grad_norm": 1.7215183973312378,
      "learning_rate": 3.814113668752486e-05,
      "loss": 0.2669,
      "step": 3094
    },
    {
      "epoch": 0.6188,
      "loss_align": 0.16524124145507812,
      "loss_contrastive": 0.0,
      "loss_gen": 0.290546715259552,
      "neg_sim_mean": 0.7127130031585693,
      "pos_sim_mean": 0.8347587585449219,
      "step": 3094
    },
    {
      "epoch": 0.619,
      "grad_norm": 2.089254856109619,
      "learning_rate": 3.8107229006101855e-05,
      "loss": 0.4558,
      "step": 3095
    },
    {
      "epoch": 0.619,
      "loss_align": 0.1793363094329834,
      "loss_contrastive": 0.10940051078796387,
      "loss_gen": 0.19670924544334412,
      "neg_sim_mean": 0.8300641775131226,
      "pos_sim_mean": 0.8206636905670166,
      "step": 3095
    },
    {
      "epoch": 0.6192,
      "grad_norm": 1.8884479999542236,
      "learning_rate": 3.807332712107097e-05,
      "loss": 0.3892,
      "step": 3096
    },
    {
      "epoch": 0.6192,
      "loss_align": 0.16138583421707153,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17704761028289795,
      "neg_sim_mean": 0.7169772386550903,
      "pos_sim_mean": 0.8386141657829285,
      "step": 3096
    },
    {
      "epoch": 0.6194,
      "grad_norm": 1.8306772708892822,
      "learning_rate": 3.8039431048955535e-05,
      "loss": 0.3384,
      "step": 3097
    },
    {
      "epoch": 0.6194,
      "loss_align": 0.15431344509124756,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012010505422949791,
      "neg_sim_mean": 0.31559067964553833,
      "pos_sim_mean": 0.8456865549087524,
      "step": 3097
    },
    {
      "epoch": 0.6196,
      "grad_norm": 0.7773668766021729,
      "learning_rate": 3.800554080627613e-05,
      "loss": 0.1663,
      "step": 3098
    },
    {
      "epoch": 0.6196,
      "loss_align": 0.07946580648422241,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011517333332449198,
      "neg_sim_mean": 0.26903098821640015,
      "pos_sim_mean": 0.9205341935157776,
      "step": 3098
    },
    {
      "epoch": 0.6198,
      "grad_norm": 0.4525519907474518,
      "learning_rate": 3.797165640955041e-05,
      "loss": 0.0806,
      "step": 3099
    },
    {
      "epoch": 0.6198,
      "loss_align": 0.07108259201049805,
      "loss_contrastive": 0.07616353034973145,
      "loss_gen": 0.1965189129114151,
      "neg_sim_mean": 0.9050809144973755,
      "pos_sim_mean": 0.928917407989502,
      "step": 3099
    },
    {
      "epoch": 0.62,
      "grad_norm": 2.064516067504883,
      "learning_rate": 3.7937777875293244e-05,
      "loss": 0.2767,
      "step": 3100
    },
    {
      "epoch": 0.62,
      "loss_align": 0.14424097537994385,
      "loss_contrastive": 0.051053643226623535,
      "loss_gen": 0.28303831815719604,
      "neg_sim_mean": 0.8068126440048218,
      "pos_sim_mean": 0.8557590246200562,
      "step": 3100
    },
    {
      "epoch": 0.6202,
      "grad_norm": 1.7971789836883545,
      "learning_rate": 3.790390522001662e-05,
      "loss": 0.4334,
      "step": 3101
    },
    {
      "epoch": 0.6202,
      "loss_align": 0.29979562759399414,
      "loss_contrastive": 0.04426997900009155,
      "loss_gen": 0.3424761891365051,
      "neg_sim_mean": 0.6444743275642395,
      "pos_sim_mean": 0.7002043724060059,
      "step": 3101
    },
    {
      "epoch": 0.6204,
      "grad_norm": 3.0250113010406494,
      "learning_rate": 3.787003846022964e-05,
      "loss": 0.6476,
      "step": 3102
    },
    {
      "epoch": 0.6204,
      "loss_align": 0.10646986961364746,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2833710312843323,
      "neg_sim_mean": 0.7813458442687988,
      "pos_sim_mean": 0.8935301303863525,
      "step": 3102
    },
    {
      "epoch": 0.6206,
      "grad_norm": 2.002847671508789,
      "learning_rate": 3.783617761243856e-05,
      "loss": 0.3898,
      "step": 3103
    },
    {
      "epoch": 0.6206,
      "loss_align": 0.33846187591552734,
      "loss_contrastive": 0.2072538137435913,
      "loss_gen": 0.3039616048336029,
      "neg_sim_mean": 0.768791913986206,
      "pos_sim_mean": 0.6615381240844727,
      "step": 3103
    },
    {
      "epoch": 0.6208,
      "grad_norm": 2.798973560333252,
      "learning_rate": 3.780232269314672e-05,
      "loss": 0.6673,
      "step": 3104
    },
    {
      "epoch": 0.6208,
      "loss_align": 0.14970779418945312,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001071006408892572,
      "neg_sim_mean": 0.3808927834033966,
      "pos_sim_mean": 0.8502922058105469,
      "step": 3104
    },
    {
      "epoch": 0.621,
      "grad_norm": 0.4541179835796356,
      "learning_rate": 3.776847371885464e-05,
      "loss": 0.1508,
      "step": 3105
    },
    {
      "epoch": 0.621,
      "loss_align": 0.09900104999542236,
      "loss_contrastive": 0.052875518798828125,
      "loss_gen": 0.2882353663444519,
      "neg_sim_mean": 0.8538744449615479,
      "pos_sim_mean": 0.9009989500045776,
      "step": 3105
    },
    {
      "epoch": 0.6212,
      "grad_norm": 1.8475661277770996,
      "learning_rate": 3.773463070605987e-05,
      "loss": 0.3936,
      "step": 3106
    },
    {
      "epoch": 0.6212,
      "loss_align": 0.06611645221710205,
      "loss_contrastive": 0.0,
      "loss_gen": 0.014572421088814735,
      "neg_sim_mean": 0.38161593675613403,
      "pos_sim_mean": 0.933883547782898,
      "step": 3106
    },
    {
      "epoch": 0.6214,
      "grad_norm": 0.7607415914535522,
      "learning_rate": 3.77007936712571e-05,
      "loss": 0.0807,
      "step": 3107
    },
    {
      "epoch": 0.6214,
      "loss_align": 0.47359633445739746,
      "loss_contrastive": 0.3669578731060028,
      "loss_gen": 0.4851757287979126,
      "neg_sim_mean": 0.7933615446090698,
      "pos_sim_mean": 0.5264036655426025,
      "step": 3107
    },
    {
      "epoch": 0.6216,
      "grad_norm": 2.9112234115600586,
      "learning_rate": 3.7666962630938084e-05,
      "loss": 1.0028,
      "step": 3108
    },
    {
      "epoch": 0.6216,
      "loss_align": 0.15229713916778564,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0029156673699617386,
      "neg_sim_mean": 0.4106307625770569,
      "pos_sim_mean": 0.8477028608322144,
      "step": 3108
    },
    {
      "epoch": 0.6218,
      "grad_norm": 0.5254287123680115,
      "learning_rate": 3.7633137601591644e-05,
      "loss": 0.1552,
      "step": 3109
    },
    {
      "epoch": 0.6218,
      "loss_align": 0.12047916650772095,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0031295327935367823,
      "neg_sim_mean": 0.31261423230171204,
      "pos_sim_mean": 0.879520833492279,
      "step": 3109
    },
    {
      "epoch": 0.622,
      "grad_norm": 0.4358479380607605,
      "learning_rate": 3.759931859970374e-05,
      "loss": 0.1236,
      "step": 3110
    },
    {
      "epoch": 0.622,
      "loss_align": 0.07573556900024414,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0007821789477020502,
      "neg_sim_mean": 0.33029380440711975,
      "pos_sim_mean": 0.9242644309997559,
      "step": 3110
    },
    {
      "epoch": 0.6222,
      "grad_norm": 0.3368149995803833,
      "learning_rate": 3.756550564175727e-05,
      "loss": 0.0765,
      "step": 3111
    },
    {
      "epoch": 0.6222,
      "loss_align": 0.17317354679107666,
      "loss_contrastive": 0.002651393413543701,
      "loss_gen": 0.10677437484264374,
      "neg_sim_mean": 0.7294778227806091,
      "pos_sim_mean": 0.8268264532089233,
      "step": 3111
    },
    {
      "epoch": 0.6224,
      "grad_norm": 1.6691150665283203,
      "learning_rate": 3.7531698744232305e-05,
      "loss": 0.2803,
      "step": 3112
    },
    {
      "epoch": 0.6224,
      "loss_align": 0.19818705320358276,
      "loss_contrastive": 0.12946945428848267,
      "loss_gen": 0.2673877775669098,
      "neg_sim_mean": 0.831282377243042,
      "pos_sim_mean": 0.8018129467964172,
      "step": 3112
    },
    {
      "epoch": 0.6226,
      "grad_norm": 2.0426025390625,
      "learning_rate": 3.7497897923605895e-05,
      "loss": 0.4811,
      "step": 3113
    },
    {
      "epoch": 0.6226,
      "loss_align": 0.13259762525558472,
      "loss_contrastive": 0.091488778591156,
      "loss_gen": 0.1636800318956375,
      "neg_sim_mean": 0.8588911294937134,
      "pos_sim_mean": 0.8674023747444153,
      "step": 3113
    },
    {
      "epoch": 0.6228,
      "grad_norm": 1.7258305549621582,
      "learning_rate": 3.746410319635217e-05,
      "loss": 0.3073,
      "step": 3114
    },
    {
      "epoch": 0.6228,
      "loss_align": 0.1285465955734253,
      "loss_contrastive": 0.03844892978668213,
      "loss_gen": 0.2758771777153015,
      "neg_sim_mean": 0.8099023103713989,
      "pos_sim_mean": 0.8714534044265747,
      "step": 3114
    },
    {
      "epoch": 0.623,
      "grad_norm": 2.1317052841186523,
      "learning_rate": 3.743031457894226e-05,
      "loss": 0.409,
      "step": 3115
    },
    {
      "epoch": 0.623,
      "loss_align": 0.17800837755203247,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012157720513641834,
      "neg_sim_mean": 0.393699586391449,
      "pos_sim_mean": 0.8219916224479675,
      "step": 3115
    },
    {
      "epoch": 0.6232,
      "grad_norm": 0.8822212815284729,
      "learning_rate": 3.739653208784432e-05,
      "loss": 0.1902,
      "step": 3116
    },
    {
      "epoch": 0.6232,
      "loss_align": 0.08328795433044434,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2509285509586334,
      "neg_sim_mean": 0.7453141212463379,
      "pos_sim_mean": 0.9167120456695557,
      "step": 3116
    },
    {
      "epoch": 0.6234,
      "grad_norm": 1.7618650197982788,
      "learning_rate": 3.736275573952354e-05,
      "loss": 0.3342,
      "step": 3117
    },
    {
      "epoch": 0.6234,
      "loss_align": 0.1667606234550476,
      "loss_contrastive": 0.03872197866439819,
      "loss_gen": 0.2758937180042267,
      "neg_sim_mean": 0.7719613313674927,
      "pos_sim_mean": 0.8332393765449524,
      "step": 3117
    },
    {
      "epoch": 0.6236,
      "grad_norm": 1.7943005561828613,
      "learning_rate": 3.7328985550442084e-05,
      "loss": 0.4473,
      "step": 3118
    },
    {
      "epoch": 0.6236,
      "loss_align": 0.18710017204284668,
      "loss_contrastive": 0.0828215479850769,
      "loss_gen": 0.2535189688205719,
      "neg_sim_mean": 0.7957213521003723,
      "pos_sim_mean": 0.8128998279571533,
      "step": 3118
    },
    {
      "epoch": 0.6238,
      "grad_norm": 1.8845000267028809,
      "learning_rate": 3.729522153705916e-05,
      "loss": 0.4506,
      "step": 3119
    },
    {
      "epoch": 0.6238,
      "loss_align": 0.11190509796142578,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0012207025429233909,
      "neg_sim_mean": 0.3395634889602661,
      "pos_sim_mean": 0.8880949020385742,
      "step": 3119
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.49782663583755493,
      "learning_rate": 3.72614637158309e-05,
      "loss": 0.1131,
      "step": 3120
    },
    {
      "epoch": 0.624,
      "loss_align": 0.08139157295227051,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24797329306602478,
      "neg_sim_mean": 0.5910342335700989,
      "pos_sim_mean": 0.9186084270477295,
      "step": 3120
    },
    {
      "epoch": 0.6242,
      "grad_norm": 1.6882411241531372,
      "learning_rate": 3.7227712103210486e-05,
      "loss": 0.3294,
      "step": 3121
    },
    {
      "epoch": 0.6242,
      "loss_align": 0.33825498819351196,
      "loss_contrastive": 0.09010040760040283,
      "loss_gen": 0.2980720102787018,
      "neg_sim_mean": 0.651845395565033,
      "pos_sim_mean": 0.661745011806488,
      "step": 3121
    },
    {
      "epoch": 0.6244,
      "grad_norm": 2.143843650817871,
      "learning_rate": 3.7193966715648024e-05,
      "loss": 0.6471,
      "step": 3122
    },
    {
      "epoch": 0.6244,
      "loss_align": 0.1508668065071106,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1961510330438614,
      "neg_sim_mean": 0.6367039084434509,
      "pos_sim_mean": 0.8491331934928894,
      "step": 3122
    },
    {
      "epoch": 0.6246,
      "grad_norm": 1.7618889808654785,
      "learning_rate": 3.716022756959061e-05,
      "loss": 0.347,
      "step": 3123
    },
    {
      "epoch": 0.6246,
      "loss_align": 0.1350330114364624,
      "loss_contrastive": 0.09678494930267334,
      "loss_gen": 0.34487470984458923,
      "neg_sim_mean": 0.861751914024353,
      "pos_sim_mean": 0.8649669885635376,
      "step": 3123
    },
    {
      "epoch": 0.6248,
      "grad_norm": 2.012699842453003,
      "learning_rate": 3.712649468148232e-05,
      "loss": 0.4915,
      "step": 3124
    },
    {
      "epoch": 0.6248,
      "loss_align": 0.14908182621002197,
      "loss_contrastive": 0.05540817975997925,
      "loss_gen": 0.2625446915626526,
      "neg_sim_mean": 0.8063263297080994,
      "pos_sim_mean": 0.850918173789978,
      "step": 3124
    },
    {
      "epoch": 0.625,
      "grad_norm": 1.7944715023040771,
      "learning_rate": 3.7092768067764124e-05,
      "loss": 0.4183,
      "step": 3125
    },
    {
      "epoch": 0.625,
      "loss_align": 0.090809166431427,
      "loss_contrastive": 0.04583442211151123,
      "loss_gen": 0.44922301173210144,
      "neg_sim_mean": 0.8550252318382263,
      "pos_sim_mean": 0.909190833568573,
      "step": 3125
    },
    {
      "epoch": 0.6252,
      "grad_norm": 2.12589168548584,
      "learning_rate": 3.705904774487396e-05,
      "loss": 0.5455,
      "step": 3126
    },
    {
      "epoch": 0.6252,
      "loss_align": 0.10877764225006104,
      "loss_contrastive": 0.0,
      "loss_gen": 0.375219464302063,
      "neg_sim_mean": 0.5165186524391174,
      "pos_sim_mean": 0.891222357749939,
      "step": 3126
    },
    {
      "epoch": 0.6254,
      "grad_norm": 2.17720103263855,
      "learning_rate": 3.702533372924673e-05,
      "loss": 0.484,
      "step": 3127
    },
    {
      "epoch": 0.6254,
      "loss_align": 0.36168503761291504,
      "loss_contrastive": 0.22378253936767578,
      "loss_gen": 0.146162748336792,
      "neg_sim_mean": 0.7620974779129028,
      "pos_sim_mean": 0.638314962387085,
      "step": 3127
    },
    {
      "epoch": 0.6256,
      "grad_norm": 1.7068551778793335,
      "learning_rate": 3.699162603731423e-05,
      "loss": 0.5347,
      "step": 3128
    },
    {
      "epoch": 0.6256,
      "loss_align": 0.19235515594482422,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26823970675468445,
      "neg_sim_mean": 0.6806575059890747,
      "pos_sim_mean": 0.8076448440551758,
      "step": 3128
    },
    {
      "epoch": 0.6258,
      "grad_norm": 1.9215552806854248,
      "learning_rate": 3.6957924685505167e-05,
      "loss": 0.4606,
      "step": 3129
    },
    {
      "epoch": 0.6258,
      "loss_align": 0.11784791946411133,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20297719538211823,
      "neg_sim_mean": 0.4276284873485565,
      "pos_sim_mean": 0.8821520805358887,
      "step": 3129
    },
    {
      "epoch": 0.626,
      "grad_norm": 1.4978902339935303,
      "learning_rate": 3.692422969024516e-05,
      "loss": 0.3208,
      "step": 3130
    },
    {
      "epoch": 0.626,
      "loss_align": 0.12042146921157837,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18795493245124817,
      "neg_sim_mean": 0.5167943239212036,
      "pos_sim_mean": 0.8795785307884216,
      "step": 3130
    },
    {
      "epoch": 0.6262,
      "grad_norm": 1.714025616645813,
      "learning_rate": 3.6890541067956776e-05,
      "loss": 0.3084,
      "step": 3131
    },
    {
      "epoch": 0.6262,
      "loss_align": 0.20513230562210083,
      "loss_contrastive": 0.16287845373153687,
      "loss_gen": 0.20897839963436127,
      "neg_sim_mean": 0.8577461242675781,
      "pos_sim_mean": 0.7948676943778992,
      "step": 3131
    },
    {
      "epoch": 0.6264,
      "grad_norm": 1.7169603109359741,
      "learning_rate": 3.68568588350594e-05,
      "loss": 0.4337,
      "step": 3132
    },
    {
      "epoch": 0.6264,
      "loss_align": 0.10717308521270752,
      "loss_contrastive": 0.043637633323669434,
      "loss_gen": 0.21250587701797485,
      "neg_sim_mean": 0.836464524269104,
      "pos_sim_mean": 0.8928269147872925,
      "step": 3132
    },
    {
      "epoch": 0.6266,
      "grad_norm": 1.6650694608688354,
      "learning_rate": 3.682318300796938e-05,
      "loss": 0.3249,
      "step": 3133
    },
    {
      "epoch": 0.6266,
      "loss_align": 0.14327317476272583,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29169490933418274,
      "neg_sim_mean": 0.6324244737625122,
      "pos_sim_mean": 0.8567268252372742,
      "step": 3133
    },
    {
      "epoch": 0.6268,
      "grad_norm": 2.067474603652954,
      "learning_rate": 3.678951360309988e-05,
      "loss": 0.435,
      "step": 3134
    },
    {
      "epoch": 0.6268,
      "loss_align": 0.14611828327178955,
      "loss_contrastive": 0.02679741382598877,
      "loss_gen": 0.24329395592212677,
      "neg_sim_mean": 0.7806791067123413,
      "pos_sim_mean": 0.8538817167282104,
      "step": 3134
    },
    {
      "epoch": 0.627,
      "grad_norm": 2.102051258087158,
      "learning_rate": 3.6755850636860954e-05,
      "loss": 0.3926,
      "step": 3135
    },
    {
      "epoch": 0.627,
      "loss_align": 0.15981626510620117,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004156283102929592,
      "neg_sim_mean": 0.3152209520339966,
      "pos_sim_mean": 0.8401837348937988,
      "step": 3135
    },
    {
      "epoch": 0.6272,
      "grad_norm": 0.502034068107605,
      "learning_rate": 3.6722194125659556e-05,
      "loss": 0.164,
      "step": 3136
    },
    {
      "epoch": 0.6272,
      "loss_align": 0.1420682668685913,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22862614691257477,
      "neg_sim_mean": 0.6041799783706665,
      "pos_sim_mean": 0.8579317331314087,
      "step": 3136
    },
    {
      "epoch": 0.6274,
      "grad_norm": 1.593436360359192,
      "learning_rate": 3.668854408589945e-05,
      "loss": 0.3707,
      "step": 3137
    },
    {
      "epoch": 0.6274,
      "loss_align": 0.1512744426727295,
      "loss_contrastive": 0.06292688846588135,
      "loss_gen": 0.41168344020843506,
      "neg_sim_mean": 0.811652421951294,
      "pos_sim_mean": 0.8487255573272705,
      "step": 3137
    },
    {
      "epoch": 0.6276,
      "grad_norm": 2.100093126296997,
      "learning_rate": 3.665490053398123e-05,
      "loss": 0.5705,
      "step": 3138
    },
    {
      "epoch": 0.6276,
      "loss_align": 0.12960433959960938,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005580197554081678,
      "neg_sim_mean": 0.3468894362449646,
      "pos_sim_mean": 0.8703956604003906,
      "step": 3138
    },
    {
      "epoch": 0.6278,
      "grad_norm": 0.5423784255981445,
      "learning_rate": 3.662126348630237e-05,
      "loss": 0.1352,
      "step": 3139
    },
    {
      "epoch": 0.6278,
      "loss_align": 0.08853781223297119,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16607649624347687,
      "neg_sim_mean": 0.7572181224822998,
      "pos_sim_mean": 0.9114621877670288,
      "step": 3139
    },
    {
      "epoch": 0.628,
      "grad_norm": 1.3666906356811523,
      "learning_rate": 3.6587632959257165e-05,
      "loss": 0.2546,
      "step": 3140
    },
    {
      "epoch": 0.628,
      "loss_align": 0.03834962844848633,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13193120062351227,
      "neg_sim_mean": 0.7771239280700684,
      "pos_sim_mean": 0.9616503715515137,
      "step": 3140
    },
    {
      "epoch": 0.6282,
      "grad_norm": 1.2955968379974365,
      "learning_rate": 3.655400896923672e-05,
      "loss": 0.1703,
      "step": 3141
    },
    {
      "epoch": 0.6282,
      "loss_align": 0.13568967580795288,
      "loss_contrastive": 0.09128731489181519,
      "loss_gen": 0.3704778552055359,
      "neg_sim_mean": 0.8555976152420044,
      "pos_sim_mean": 0.8643103241920471,
      "step": 3141
    },
    {
      "epoch": 0.6284,
      "grad_norm": 2.1444931030273438,
      "learning_rate": 3.652039153262895e-05,
      "loss": 0.5171,
      "step": 3142
    },
    {
      "epoch": 0.6284,
      "loss_align": 0.07049727439880371,
      "loss_contrastive": 0.047257304191589355,
      "loss_gen": 0.1213877871632576,
      "neg_sim_mean": 0.8767600059509277,
      "pos_sim_mean": 0.9295027256011963,
      "step": 3142
    },
    {
      "epoch": 0.6286,
      "grad_norm": 1.1801677942276,
      "learning_rate": 3.6486780665818606e-05,
      "loss": 0.1976,
      "step": 3143
    },
    {
      "epoch": 0.6286,
      "loss_align": 0.05364394187927246,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13791483640670776,
      "neg_sim_mean": 0.694165825843811,
      "pos_sim_mean": 0.9463560581207275,
      "step": 3143
    },
    {
      "epoch": 0.6288,
      "grad_norm": 1.1532254219055176,
      "learning_rate": 3.645317638518721e-05,
      "loss": 0.1916,
      "step": 3144
    },
    {
      "epoch": 0.6288,
      "loss_align": 0.14197778701782227,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4470529854297638,
      "neg_sim_mean": 0.6445128917694092,
      "pos_sim_mean": 0.8580222129821777,
      "step": 3144
    },
    {
      "epoch": 0.629,
      "grad_norm": 2.328803062438965,
      "learning_rate": 3.641957870711306e-05,
      "loss": 0.589,
      "step": 3145
    },
    {
      "epoch": 0.629,
      "loss_align": 0.1168927550315857,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2351636290550232,
      "neg_sim_mean": 0.7762773036956787,
      "pos_sim_mean": 0.8831072449684143,
      "step": 3145
    },
    {
      "epoch": 0.6292,
      "grad_norm": 1.7194774150848389,
      "learning_rate": 3.6385987647971285e-05,
      "loss": 0.3521,
      "step": 3146
    },
    {
      "epoch": 0.6292,
      "loss_align": 0.1582164168357849,
      "loss_contrastive": 0.06607824563980103,
      "loss_gen": 0.33406955003738403,
      "neg_sim_mean": 0.8078618049621582,
      "pos_sim_mean": 0.8417835831642151,
      "step": 3146
    },
    {
      "epoch": 0.6294,
      "grad_norm": 2.3612663745880127,
      "learning_rate": 3.635240322413374e-05,
      "loss": 0.5002,
      "step": 3147
    },
    {
      "epoch": 0.6294,
      "loss_align": 0.11489063501358032,
      "loss_contrastive": 0.07641500234603882,
      "loss_gen": 0.21060268580913544,
      "neg_sim_mean": 0.8615243434906006,
      "pos_sim_mean": 0.8851093649864197,
      "step": 3147
    },
    {
      "epoch": 0.6296,
      "grad_norm": 1.674999475479126,
      "learning_rate": 3.6318825451969087e-05,
      "loss": 0.3347,
      "step": 3148
    },
    {
      "epoch": 0.6296,
      "loss_align": 0.08425843715667725,
      "loss_contrastive": 0.09161800146102905,
      "loss_gen": 0.3181529939174652,
      "neg_sim_mean": 0.9073595404624939,
      "pos_sim_mean": 0.9157415628433228,
      "step": 3148
    },
    {
      "epoch": 0.6298,
      "grad_norm": 1.9366905689239502,
      "learning_rate": 3.6285254347842676e-05,
      "loss": 0.4134,
      "step": 3149
    },
    {
      "epoch": 0.6298,
      "loss_align": 0.13464581966400146,
      "loss_contrastive": 0.04840672016143799,
      "loss_gen": 0.2169199138879776,
      "neg_sim_mean": 0.8137608766555786,
      "pos_sim_mean": 0.8653541803359985,
      "step": 3149
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.79615318775177,
      "learning_rate": 3.625168992811671e-05,
      "loss": 0.3574,
      "step": 3150
    },
    {
      "epoch": 0.63,
      "loss_align": 0.11606335639953613,
      "loss_contrastive": 0.07164692878723145,
      "loss_gen": 0.30474674701690674,
      "neg_sim_mean": 0.8555835485458374,
      "pos_sim_mean": 0.8839366436004639,
      "step": 3150
    },
    {
      "epoch": 0.6302,
      "grad_norm": 2.014575719833374,
      "learning_rate": 3.6218132209150045e-05,
      "loss": 0.4294,
      "step": 3151
    },
    {
      "epoch": 0.6302,
      "loss_align": 0.18457138538360596,
      "loss_contrastive": 0.0,
      "loss_gen": 0.36662182211875916,
      "neg_sim_mean": 0.6382672786712646,
      "pos_sim_mean": 0.815428614616394,
      "step": 3151
    },
    {
      "epoch": 0.6304,
      "grad_norm": 1.9800843000411987,
      "learning_rate": 3.618458120729832e-05,
      "loss": 0.5512,
      "step": 3152
    },
    {
      "epoch": 0.6304,
      "loss_align": 0.07791638374328613,
      "loss_contrastive": 0.0371975302696228,
      "loss_gen": 0.15642815828323364,
      "neg_sim_mean": 0.8592811226844788,
      "pos_sim_mean": 0.9220836162567139,
      "step": 3152
    },
    {
      "epoch": 0.6306,
      "grad_norm": 1.533412218093872,
      "learning_rate": 3.615103693891388e-05,
      "loss": 0.2388,
      "step": 3153
    },
    {
      "epoch": 0.6306,
      "loss_align": 0.10798949003219604,
      "loss_contrastive": 0.0,
      "loss_gen": 0.36716583371162415,
      "neg_sim_mean": 0.7265351414680481,
      "pos_sim_mean": 0.892010509967804,
      "step": 3153
    },
    {
      "epoch": 0.6308,
      "grad_norm": 2.1565678119659424,
      "learning_rate": 3.61174994203458e-05,
      "loss": 0.4752,
      "step": 3154
    },
    {
      "epoch": 0.6308,
      "loss_align": 0.1455613374710083,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10917074233293533,
      "neg_sim_mean": 0.710628867149353,
      "pos_sim_mean": 0.8544386625289917,
      "step": 3154
    },
    {
      "epoch": 0.631,
      "grad_norm": 1.4644761085510254,
      "learning_rate": 3.608396866793988e-05,
      "loss": 0.2547,
      "step": 3155
    },
    {
      "epoch": 0.631,
      "loss_align": 0.14152151346206665,
      "loss_contrastive": 0.09775751829147339,
      "loss_gen": 0.2847691774368286,
      "neg_sim_mean": 0.8562359809875488,
      "pos_sim_mean": 0.8584784865379333,
      "step": 3155
    },
    {
      "epoch": 0.6312,
      "grad_norm": 1.7794359922409058,
      "learning_rate": 3.605044469803854e-05,
      "loss": 0.438,
      "step": 3156
    },
    {
      "epoch": 0.6312,
      "loss_align": 0.08569735288619995,
      "loss_contrastive": 0.047246694564819336,
      "loss_gen": 0.12818512320518494,
      "neg_sim_mean": 0.8615493178367615,
      "pos_sim_mean": 0.9143026471138,
      "step": 3156
    },
    {
      "epoch": 0.6314,
      "grad_norm": 1.4977970123291016,
      "learning_rate": 3.601692752698102e-05,
      "loss": 0.2196,
      "step": 3157
    },
    {
      "epoch": 0.6314,
      "loss_align": 0.21402889490127563,
      "loss_contrastive": 0.0004844069480895996,
      "loss_gen": 0.23138363659381866,
      "neg_sim_mean": 0.686455488204956,
      "pos_sim_mean": 0.7859711050987244,
      "step": 3157
    },
    {
      "epoch": 0.6316,
      "grad_norm": 1.9438554048538208,
      "learning_rate": 3.598341717110313e-05,
      "loss": 0.4455,
      "step": 3158
    },
    {
      "epoch": 0.6316,
      "loss_align": 0.15995466709136963,
      "loss_contrastive": 0.0,
      "loss_gen": 0.607002317905426,
      "neg_sim_mean": 0.6338565349578857,
      "pos_sim_mean": 0.8400453329086304,
      "step": 3158
    },
    {
      "epoch": 0.6318,
      "grad_norm": 3.1098506450653076,
      "learning_rate": 3.594991364673745e-05,
      "loss": 0.767,
      "step": 3159
    },
    {
      "epoch": 0.6318,
      "loss_align": 0.23462867736816406,
      "loss_contrastive": 0.08955341577529907,
      "loss_gen": 0.258626252412796,
      "neg_sim_mean": 0.7549247145652771,
      "pos_sim_mean": 0.7653713226318359,
      "step": 3159
    },
    {
      "epoch": 0.632,
      "grad_norm": 1.8382103443145752,
      "learning_rate": 3.591641697021317e-05,
      "loss": 0.504,
      "step": 3160
    },
    {
      "epoch": 0.632,
      "loss_align": 0.1464185118675232,
      "loss_contrastive": 0.012741506099700928,
      "loss_gen": 0.16131296753883362,
      "neg_sim_mean": 0.7663229703903198,
      "pos_sim_mean": 0.8535814881324768,
      "step": 3160
    },
    {
      "epoch": 0.6322,
      "grad_norm": 1.9306560754776,
      "learning_rate": 3.588292715785617e-05,
      "loss": 0.3093,
      "step": 3161
    },
    {
      "epoch": 0.6322,
      "loss_align": 0.19584637880325317,
      "loss_contrastive": 0.0677141547203064,
      "loss_gen": 0.3340640664100647,
      "neg_sim_mean": 0.7718677520751953,
      "pos_sim_mean": 0.8041536211967468,
      "step": 3161
    },
    {
      "epoch": 0.6324,
      "grad_norm": 2.167069673538208,
      "learning_rate": 3.584944422598899e-05,
      "loss": 0.538,
      "step": 3162
    },
    {
      "epoch": 0.6324,
      "loss_align": 0.07750558853149414,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11834254115819931,
      "neg_sim_mean": 0.4939820468425751,
      "pos_sim_mean": 0.9224944114685059,
      "step": 3162
    },
    {
      "epoch": 0.6326,
      "grad_norm": 1.1292396783828735,
      "learning_rate": 3.581596819093079e-05,
      "loss": 0.1958,
      "step": 3163
    },
    {
      "epoch": 0.6326,
      "loss_align": 0.07014477252960205,
      "loss_contrastive": 0.055831968784332275,
      "loss_gen": 0.26117968559265137,
      "neg_sim_mean": 0.8856871724128723,
      "pos_sim_mean": 0.929855227470398,
      "step": 3163
    },
    {
      "epoch": 0.6328,
      "grad_norm": 1.7857962846755981,
      "learning_rate": 3.578249906899738e-05,
      "loss": 0.338,
      "step": 3164
    },
    {
      "epoch": 0.6328,
      "loss_align": 0.1410425901412964,
      "loss_contrastive": 0.09091031551361084,
      "loss_gen": 0.23410704731941223,
      "neg_sim_mean": 0.8498677015304565,
      "pos_sim_mean": 0.8589574098587036,
      "step": 3164
    },
    {
      "epoch": 0.633,
      "grad_norm": 1.8943599462509155,
      "learning_rate": 3.5749036876501194e-05,
      "loss": 0.3861,
      "step": 3165
    },
    {
      "epoch": 0.633,
      "loss_align": 0.09732949733734131,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03279693424701691,
      "neg_sim_mean": 0.32837700843811035,
      "pos_sim_mean": 0.9026705026626587,
      "step": 3165
    },
    {
      "epoch": 0.6332,
      "grad_norm": 1.1159855127334595,
      "learning_rate": 3.5715581629751326e-05,
      "loss": 0.1301,
      "step": 3166
    },
    {
      "epoch": 0.6332,
      "loss_align": 0.1302068829536438,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3230174481868744,
      "neg_sim_mean": 0.5678645372390747,
      "pos_sim_mean": 0.8697931170463562,
      "step": 3166
    },
    {
      "epoch": 0.6334,
      "grad_norm": 2.1815133094787598,
      "learning_rate": 3.5682133345053445e-05,
      "loss": 0.4532,
      "step": 3167
    },
    {
      "epoch": 0.6334,
      "loss_align": 0.04925870895385742,
      "loss_contrastive": 0.05602717399597168,
      "loss_gen": 0.2848782241344452,
      "neg_sim_mean": 0.9067684412002563,
      "pos_sim_mean": 0.9507412910461426,
      "step": 3167
    },
    {
      "epoch": 0.6336,
      "grad_norm": 2.133113384246826,
      "learning_rate": 3.564869203870982e-05,
      "loss": 0.3409,
      "step": 3168
    },
    {
      "epoch": 0.6336,
      "loss_align": 0.11240679025650024,
      "loss_contrastive": 0.08620595932006836,
      "loss_gen": 0.2168276011943817,
      "neg_sim_mean": 0.8737991452217102,
      "pos_sim_mean": 0.8875932097434998,
      "step": 3168
    },
    {
      "epoch": 0.6338,
      "grad_norm": 1.9172691106796265,
      "learning_rate": 3.561525772701937e-05,
      "loss": 0.3396,
      "step": 3169
    },
    {
      "epoch": 0.6338,
      "loss_align": 0.24953103065490723,
      "loss_contrastive": 0.10713452100753784,
      "loss_gen": 0.28926318883895874,
      "neg_sim_mean": 0.7576034665107727,
      "pos_sim_mean": 0.7504689693450928,
      "step": 3169
    },
    {
      "epoch": 0.634,
      "grad_norm": 2.7983663082122803,
      "learning_rate": 3.558183042627755e-05,
      "loss": 0.5517,
      "step": 3170
    },
    {
      "epoch": 0.634,
      "loss_align": 0.08463692665100098,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22338509559631348,
      "neg_sim_mean": 0.7838334441184998,
      "pos_sim_mean": 0.915363073348999,
      "step": 3170
    },
    {
      "epoch": 0.6342,
      "grad_norm": 1.768447995185852,
      "learning_rate": 3.554841015277641e-05,
      "loss": 0.308,
      "step": 3171
    },
    {
      "epoch": 0.6342,
      "loss_align": 0.1471540927886963,
      "loss_contrastive": 0.0985112190246582,
      "loss_gen": 0.2040013074874878,
      "neg_sim_mean": 0.851357102394104,
      "pos_sim_mean": 0.8528459072113037,
      "step": 3171
    },
    {
      "epoch": 0.6344,
      "grad_norm": 1.7448477745056152,
      "learning_rate": 3.5514996922804636e-05,
      "loss": 0.363,
      "step": 3172
    },
    {
      "epoch": 0.6344,
      "loss_align": 0.16292667388916016,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004234509076923132,
      "neg_sim_mean": 0.44692757725715637,
      "pos_sim_mean": 0.8370733261108398,
      "step": 3172
    },
    {
      "epoch": 0.6346,
      "grad_norm": 0.8204357028007507,
      "learning_rate": 3.5481590752647383e-05,
      "loss": 0.1672,
      "step": 3173
    },
    {
      "epoch": 0.6346,
      "loss_align": 0.13822686672210693,
      "loss_contrastive": 0.003998398780822754,
      "loss_gen": 0.20955316722393036,
      "neg_sim_mean": 0.7657715082168579,
      "pos_sim_mean": 0.8617731332778931,
      "step": 3173
    },
    {
      "epoch": 0.6348,
      "grad_norm": 1.485589861869812,
      "learning_rate": 3.544819165858642e-05,
      "loss": 0.3483,
      "step": 3174
    },
    {
      "epoch": 0.6348,
      "loss_align": 0.1614396572113037,
      "loss_contrastive": 0.08838784694671631,
      "loss_gen": 0.25163742899894714,
      "neg_sim_mean": 0.8269481658935547,
      "pos_sim_mean": 0.8385603427886963,
      "step": 3174
    },
    {
      "epoch": 0.635,
      "grad_norm": 2.1451146602630615,
      "learning_rate": 3.5414799656900054e-05,
      "loss": 0.4237,
      "step": 3175
    },
    {
      "epoch": 0.635,
      "loss_align": 0.13206171989440918,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013093357905745506,
      "neg_sim_mean": 0.40100082755088806,
      "pos_sim_mean": 0.8679382801055908,
      "step": 3175
    },
    {
      "epoch": 0.6352,
      "grad_norm": 0.9514434337615967,
      "learning_rate": 3.5381414763863166e-05,
      "loss": 0.1452,
      "step": 3176
    },
    {
      "epoch": 0.6352,
      "loss_align": 0.11975950002670288,
      "loss_contrastive": 0.08505946397781372,
      "loss_gen": 0.5201488733291626,
      "neg_sim_mean": 0.8652999401092529,
      "pos_sim_mean": 0.8802404999732971,
      "step": 3176
    },
    {
      "epoch": 0.6354,
      "grad_norm": 2.742748260498047,
      "learning_rate": 3.534803699574714e-05,
      "loss": 0.6501,
      "step": 3177
    },
    {
      "epoch": 0.6354,
      "loss_align": 0.10106396675109863,
      "loss_contrastive": 0.10283935070037842,
      "loss_gen": 0.15022291243076324,
      "neg_sim_mean": 0.9017753601074219,
      "pos_sim_mean": 0.8989360332489014,
      "step": 3177
    },
    {
      "epoch": 0.6356,
      "grad_norm": 1.4485772848129272,
      "learning_rate": 3.531466636881987e-05,
      "loss": 0.2636,
      "step": 3178
    },
    {
      "epoch": 0.6356,
      "loss_align": 0.05254983901977539,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004747269209474325,
      "neg_sim_mean": 0.44076794385910034,
      "pos_sim_mean": 0.9474501609802246,
      "step": 3178
    },
    {
      "epoch": 0.6358,
      "grad_norm": 0.3414071798324585,
      "learning_rate": 3.528130289934583e-05,
      "loss": 0.0573,
      "step": 3179
    },
    {
      "epoch": 0.6358,
      "loss_align": 0.08617883920669556,
      "loss_contrastive": 0.0743061900138855,
      "loss_gen": 0.32383811473846436,
      "neg_sim_mean": 0.888127326965332,
      "pos_sim_mean": 0.9138211607933044,
      "step": 3179
    },
    {
      "epoch": 0.636,
      "grad_norm": 2.045417308807373,
      "learning_rate": 3.524794660358593e-05,
      "loss": 0.4189,
      "step": 3180
    },
    {
      "epoch": 0.636,
      "loss_align": 0.3312290906906128,
      "loss_contrastive": 0.16012811660766602,
      "loss_gen": 0.19585959613323212,
      "neg_sim_mean": 0.7288990020751953,
      "pos_sim_mean": 0.6687709093093872,
      "step": 3180
    },
    {
      "epoch": 0.6362,
      "grad_norm": 2.2934279441833496,
      "learning_rate": 3.5214597497797684e-05,
      "loss": 0.5463,
      "step": 3181
    },
    {
      "epoch": 0.6362,
      "loss_align": 0.051571011543273926,
      "loss_contrastive": 0.030840694904327393,
      "loss_gen": 0.23736584186553955,
      "neg_sim_mean": 0.8792696595191956,
      "pos_sim_mean": 0.9484289884567261,
      "step": 3181
    },
    {
      "epoch": 0.6364,
      "grad_norm": 2.0105507373809814,
      "learning_rate": 3.518125559823496e-05,
      "loss": 0.2926,
      "step": 3182
    },
    {
      "epoch": 0.6364,
      "loss_align": 0.10392963886260986,
      "loss_contrastive": 0.034096598625183105,
      "loss_gen": 0.2518807351589203,
      "neg_sim_mean": 0.8301669359207153,
      "pos_sim_mean": 0.8960703611373901,
      "step": 3182
    },
    {
      "epoch": 0.6366,
      "grad_norm": 1.8700765371322632,
      "learning_rate": 3.5147920921148267e-05,
      "loss": 0.3599,
      "step": 3183
    },
    {
      "epoch": 0.6366,
      "loss_align": 0.14769649505615234,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17740334570407867,
      "neg_sim_mean": 0.7453010678291321,
      "pos_sim_mean": 0.8523035049438477,
      "step": 3183
    },
    {
      "epoch": 0.6368,
      "grad_norm": 1.2448164224624634,
      "learning_rate": 3.511459348278448e-05,
      "loss": 0.3251,
      "step": 3184
    },
    {
      "epoch": 0.6368,
      "loss_align": 0.0676000714302063,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4817265570163727,
      "neg_sim_mean": 0.54836505651474,
      "pos_sim_mean": 0.9323999285697937,
      "step": 3184
    },
    {
      "epoch": 0.637,
      "grad_norm": 2.6423442363739014,
      "learning_rate": 3.508127329938699e-05,
      "loss": 0.5493,
      "step": 3185
    },
    {
      "epoch": 0.637,
      "loss_align": 0.12626618146896362,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002873108023777604,
      "neg_sim_mean": 0.2330191731452942,
      "pos_sim_mean": 0.8737338185310364,
      "step": 3185
    },
    {
      "epoch": 0.6372,
      "grad_norm": 0.5323014855384827,
      "learning_rate": 3.504796038719567e-05,
      "loss": 0.1291,
      "step": 3186
    },
    {
      "epoch": 0.6372,
      "loss_align": 0.06866323947906494,
      "loss_contrastive": 0.0,
      "loss_gen": 0.119706891477108,
      "neg_sim_mean": 0.7977412939071655,
      "pos_sim_mean": 0.9313367605209351,
      "step": 3186
    },
    {
      "epoch": 0.6374,
      "grad_norm": 1.3662868738174438,
      "learning_rate": 3.501465476244681e-05,
      "loss": 0.1884,
      "step": 3187
    },
    {
      "epoch": 0.6374,
      "loss_align": 0.0700029730796814,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24732370674610138,
      "neg_sim_mean": 0.7455918788909912,
      "pos_sim_mean": 0.9299970269203186,
      "step": 3187
    },
    {
      "epoch": 0.6376,
      "grad_norm": 2.41754412651062,
      "learning_rate": 3.498135644137318e-05,
      "loss": 0.3173,
      "step": 3188
    },
    {
      "epoch": 0.6376,
      "loss_align": 0.17756348848342896,
      "loss_contrastive": 0.0692186951637268,
      "loss_gen": 0.4203680157661438,
      "neg_sim_mean": 0.7916551828384399,
      "pos_sim_mean": 0.822436511516571,
      "step": 3188
    },
    {
      "epoch": 0.6378,
      "grad_norm": 2.3859405517578125,
      "learning_rate": 3.494806544020398e-05,
      "loss": 0.6062,
      "step": 3189
    },
    {
      "epoch": 0.6378,
      "loss_align": 0.07883083820343018,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21605730056762695,
      "neg_sim_mean": 0.7809939384460449,
      "pos_sim_mean": 0.9211691617965698,
      "step": 3189
    },
    {
      "epoch": 0.638,
      "grad_norm": 1.4442856311798096,
      "learning_rate": 3.491478177516484e-05,
      "loss": 0.2949,
      "step": 3190
    },
    {
      "epoch": 0.638,
      "loss_align": 0.10218566656112671,
      "loss_contrastive": 0.06098204851150513,
      "loss_gen": 0.2307751476764679,
      "neg_sim_mean": 0.8587963581085205,
      "pos_sim_mean": 0.8978143334388733,
      "step": 3190
    },
    {
      "epoch": 0.6382,
      "grad_norm": 1.8859376907348633,
      "learning_rate": 3.488150546247778e-05,
      "loss": 0.3403,
      "step": 3191
    },
    {
      "epoch": 0.6382,
      "loss_align": 0.07424241304397583,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2420809417963028,
      "neg_sim_mean": 0.8244249820709229,
      "pos_sim_mean": 0.9257575869560242,
      "step": 3191
    },
    {
      "epoch": 0.6384,
      "grad_norm": 1.7268602848052979,
      "learning_rate": 3.484823651836131e-05,
      "loss": 0.3163,
      "step": 3192
    },
    {
      "epoch": 0.6384,
      "loss_align": 0.09931910037994385,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1795218288898468,
      "neg_sim_mean": 0.7015854716300964,
      "pos_sim_mean": 0.9006808996200562,
      "step": 3192
    },
    {
      "epoch": 0.6386,
      "grad_norm": 1.6086385250091553,
      "learning_rate": 3.481497495903029e-05,
      "loss": 0.2788,
      "step": 3193
    },
    {
      "epoch": 0.6386,
      "loss_align": 0.10695362091064453,
      "loss_contrastive": 0.09595030546188354,
      "loss_gen": 0.26043885946273804,
      "neg_sim_mean": 0.8889966607093811,
      "pos_sim_mean": 0.8930463790893555,
      "step": 3193
    },
    {
      "epoch": 0.6388,
      "grad_norm": 1.748966097831726,
      "learning_rate": 3.4781720800696004e-05,
      "loss": 0.3789,
      "step": 3194
    },
    {
      "epoch": 0.6388,
      "loss_align": 0.11498337984085083,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009248773567378521,
      "neg_sim_mean": 0.2694721817970276,
      "pos_sim_mean": 0.8850166201591492,
      "step": 3194
    },
    {
      "epoch": 0.639,
      "grad_norm": 0.5751810669898987,
      "learning_rate": 3.4748474059566125e-05,
      "loss": 0.1242,
      "step": 3195
    },
    {
      "epoch": 0.639,
      "loss_align": 0.05651140213012695,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21126922965049744,
      "neg_sim_mean": 0.8285399675369263,
      "pos_sim_mean": 0.943488597869873,
      "step": 3195
    },
    {
      "epoch": 0.6392,
      "grad_norm": 1.6459541320800781,
      "learning_rate": 3.471523475184472e-05,
      "loss": 0.2678,
      "step": 3196
    },
    {
      "epoch": 0.6392,
      "loss_align": 0.12308764457702637,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0038548882585018873,
      "neg_sim_mean": 0.3047386109828949,
      "pos_sim_mean": 0.8769123554229736,
      "step": 3196
    },
    {
      "epoch": 0.6394,
      "grad_norm": 0.46403032541275024,
      "learning_rate": 3.46820028937322e-05,
      "loss": 0.1269,
      "step": 3197
    },
    {
      "epoch": 0.6394,
      "loss_align": 0.16691923141479492,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012683475390076637,
      "neg_sim_mean": 0.24851132929325104,
      "pos_sim_mean": 0.8330807685852051,
      "step": 3197
    },
    {
      "epoch": 0.6396,
      "grad_norm": 0.8531700968742371,
      "learning_rate": 3.4648778501425405e-05,
      "loss": 0.1796,
      "step": 3198
    },
    {
      "epoch": 0.6396,
      "loss_align": 0.30098170042037964,
      "loss_contrastive": 0.0,
      "loss_gen": 0.014548513107001781,
      "neg_sim_mean": 0.4752732515335083,
      "pos_sim_mean": 0.6990182995796204,
      "step": 3198
    },
    {
      "epoch": 0.6398,
      "grad_norm": 2.463113307952881,
      "learning_rate": 3.461556159111748e-05,
      "loss": 0.3155,
      "step": 3199
    },
    {
      "epoch": 0.6398,
      "loss_align": 0.06078922748565674,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23004856705665588,
      "neg_sim_mean": 0.7518966794013977,
      "pos_sim_mean": 0.9392107725143433,
      "step": 3199
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.7439571619033813,
      "learning_rate": 3.4582352178997935e-05,
      "loss": 0.2908,
      "step": 3200
    },
    {
      "epoch": 0.64,
      "loss_align": 0.1527121663093567,
      "loss_contrastive": 0.10206788778305054,
      "loss_gen": 0.3156280517578125,
      "neg_sim_mean": 0.8493556976318359,
      "pos_sim_mean": 0.8472878336906433,
      "step": 3200
    },
    {
      "epoch": 0.6402,
      "grad_norm": 1.9347995519638062,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.4806,
      "step": 3201
    },
    {
      "epoch": 0.6402,
      "loss_align": 0.07497954368591309,
      "loss_contrastive": 0.05693715810775757,
      "loss_gen": 0.3837115466594696,
      "neg_sim_mean": 0.8819575905799866,
      "pos_sim_mean": 0.9250204563140869,
      "step": 3201
    },
    {
      "epoch": 0.6404,
      "grad_norm": 2.3105289936065674,
      "learning_rate": 3.45159559140638e-05,
      "loss": 0.4655,
      "step": 3202
    },
    {
      "epoch": 0.6404,
      "loss_align": 0.05461287498474121,
      "loss_contrastive": 0.001316070556640625,
      "loss_gen": 0.23063386976718903,
      "neg_sim_mean": 0.8467031717300415,
      "pos_sim_mean": 0.9453871250152588,
      "step": 3202
    },
    {
      "epoch": 0.6406,
      "grad_norm": 1.618553638458252,
      "learning_rate": 3.4482769093609944e-05,
      "loss": 0.2854,
      "step": 3203
    },
    {
      "epoch": 0.6406,
      "loss_align": 0.2982026934623718,
      "loss_contrastive": 0.1602267622947693,
      "loss_gen": 0.16284926235675812,
      "neg_sim_mean": 0.7620240449905396,
      "pos_sim_mean": 0.7017973065376282,
      "step": 3203
    },
    {
      "epoch": 0.6408,
      "grad_norm": 1.6709794998168945,
      "learning_rate": 3.444958983606592e-05,
      "loss": 0.4803,
      "step": 3204
    },
    {
      "epoch": 0.6408,
      "loss_align": 0.12402534484863281,
      "loss_contrastive": 0.04323655366897583,
      "loss_gen": 0.13645589351654053,
      "neg_sim_mean": 0.8192111849784851,
      "pos_sim_mean": 0.8759746551513672,
      "step": 3204
    },
    {
      "epoch": 0.641,
      "grad_norm": 1.3783526420593262,
      "learning_rate": 3.441641815760291e-05,
      "loss": 0.2657,
      "step": 3205
    },
    {
      "epoch": 0.641,
      "loss_align": 0.1279388666152954,
      "loss_contrastive": 0.08805131912231445,
      "loss_gen": 0.2795214056968689,
      "neg_sim_mean": 0.8601124286651611,
      "pos_sim_mean": 0.8720611333847046,
      "step": 3205
    },
    {
      "epoch": 0.6412,
      "grad_norm": 2.30670428276062,
      "learning_rate": 3.438325407438837e-05,
      "loss": 0.418,
      "step": 3206
    },
    {
      "epoch": 0.6412,
      "loss_align": 0.07047086954116821,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1743624061346054,
      "neg_sim_mean": 0.741062581539154,
      "pos_sim_mean": 0.9295291304588318,
      "step": 3206
    },
    {
      "epoch": 0.6414,
      "grad_norm": 1.4249716997146606,
      "learning_rate": 3.435009760258608e-05,
      "loss": 0.2448,
      "step": 3207
    },
    {
      "epoch": 0.6414,
      "loss_align": 0.11697471141815186,
      "loss_contrastive": 0.08551424741744995,
      "loss_gen": 0.29806047677993774,
      "neg_sim_mean": 0.8685395121574402,
      "pos_sim_mean": 0.8830252885818481,
      "step": 3207
    },
    {
      "epoch": 0.6416,
      "grad_norm": 1.741495132446289,
      "learning_rate": 3.431694875835613e-05,
      "loss": 0.4253,
      "step": 3208
    },
    {
      "epoch": 0.6416,
      "loss_align": 0.12613177299499512,
      "loss_contrastive": 0.03357195854187012,
      "loss_gen": 0.06546229124069214,
      "neg_sim_mean": 0.8074401617050171,
      "pos_sim_mean": 0.8738682270050049,
      "step": 3208
    },
    {
      "epoch": 0.6418,
      "grad_norm": 1.1196974515914917,
      "learning_rate": 3.428380755785481e-05,
      "loss": 0.1956,
      "step": 3209
    },
    {
      "epoch": 0.6418,
      "loss_align": 0.045876383781433105,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26817387342453003,
      "neg_sim_mean": 0.4308667480945587,
      "pos_sim_mean": 0.9541236162185669,
      "step": 3209
    },
    {
      "epoch": 0.642,
      "grad_norm": 2.0744621753692627,
      "learning_rate": 3.425067401723477e-05,
      "loss": 0.3141,
      "step": 3210
    },
    {
      "epoch": 0.642,
      "loss_align": 0.10255551338195801,
      "loss_contrastive": 0.06595218181610107,
      "loss_gen": 0.15501974523067474,
      "neg_sim_mean": 0.8633966445922852,
      "pos_sim_mean": 0.897444486618042,
      "step": 3210
    },
    {
      "epoch": 0.6422,
      "grad_norm": 1.4132118225097656,
      "learning_rate": 3.4217548152644885e-05,
      "loss": 0.2655,
      "step": 3211
    },
    {
      "epoch": 0.6422,
      "loss_align": 0.09911060333251953,
      "loss_contrastive": 0.05869698524475098,
      "loss_gen": 0.4260668158531189,
      "neg_sim_mean": 0.8595863580703735,
      "pos_sim_mean": 0.9008893966674805,
      "step": 3211
    },
    {
      "epoch": 0.6424,
      "grad_norm": 2.0427772998809814,
      "learning_rate": 3.4184429980230305e-05,
      "loss": 0.5322,
      "step": 3212
    },
    {
      "epoch": 0.6424,
      "loss_align": 0.07248044013977051,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1761683225631714,
      "neg_sim_mean": 0.555509626865387,
      "pos_sim_mean": 0.9275195598602295,
      "step": 3212
    },
    {
      "epoch": 0.6426,
      "grad_norm": 1.5225210189819336,
      "learning_rate": 3.4151319516132416e-05,
      "loss": 0.2486,
      "step": 3213
    },
    {
      "epoch": 0.6426,
      "loss_align": 0.07719051837921143,
      "loss_contrastive": 0.095306396484375,
      "loss_gen": 0.3527385890483856,
      "neg_sim_mean": 0.9181158542633057,
      "pos_sim_mean": 0.9228094816207886,
      "step": 3213
    },
    {
      "epoch": 0.6428,
      "grad_norm": 1.9716840982437134,
      "learning_rate": 3.4118216776488864e-05,
      "loss": 0.4414,
      "step": 3214
    },
    {
      "epoch": 0.6428,
      "loss_align": 0.24510151147842407,
      "loss_contrastive": 0.05074197053909302,
      "loss_gen": 0.5182799100875854,
      "neg_sim_mean": 0.705640435218811,
      "pos_sim_mean": 0.7548984885215759,
      "step": 3214
    },
    {
      "epoch": 0.643,
      "grad_norm": 2.4290425777435303,
      "learning_rate": 3.408512177743353e-05,
      "loss": 0.7695,
      "step": 3215
    },
    {
      "epoch": 0.643,
      "loss_align": 0.07698690891265869,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04286837577819824,
      "neg_sim_mean": 0.31258147954940796,
      "pos_sim_mean": 0.9230130910873413,
      "step": 3215
    },
    {
      "epoch": 0.6432,
      "grad_norm": 1.1120485067367554,
      "learning_rate": 3.40520345350965e-05,
      "loss": 0.1199,
      "step": 3216
    },
    {
      "epoch": 0.6432,
      "loss_align": 0.18568044900894165,
      "loss_contrastive": 0.07036060094833374,
      "loss_gen": 0.20551440119743347,
      "neg_sim_mean": 0.7846801280975342,
      "pos_sim_mean": 0.8143195509910583,
      "step": 3216
    },
    {
      "epoch": 0.6434,
      "grad_norm": 1.8980042934417725,
      "learning_rate": 3.401895506560411e-05,
      "loss": 0.3996,
      "step": 3217
    },
    {
      "epoch": 0.6434,
      "loss_align": 0.13924896717071533,
      "loss_contrastive": 0.048903703689575195,
      "loss_gen": 0.27684202790260315,
      "neg_sim_mean": 0.809654712677002,
      "pos_sim_mean": 0.8607510328292847,
      "step": 3217
    },
    {
      "epoch": 0.6436,
      "grad_norm": 1.6288813352584839,
      "learning_rate": 3.3985883385078874e-05,
      "loss": 0.422,
      "step": 3218
    },
    {
      "epoch": 0.6436,
      "loss_align": 0.17090380191802979,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001889128121547401,
      "neg_sim_mean": 0.356766939163208,
      "pos_sim_mean": 0.8290961980819702,
      "step": 3218
    },
    {
      "epoch": 0.6438,
      "grad_norm": 0.8571726083755493,
      "learning_rate": 3.3952819509639534e-05,
      "loss": 0.1728,
      "step": 3219
    },
    {
      "epoch": 0.6438,
      "loss_align": 0.08371984958648682,
      "loss_contrastive": 0.006265699863433838,
      "loss_gen": 0.2682413160800934,
      "neg_sim_mean": 0.8225458264350891,
      "pos_sim_mean": 0.9162801504135132,
      "step": 3219
    },
    {
      "epoch": 0.644,
      "grad_norm": 2.0981457233428955,
      "learning_rate": 3.3919763455401015e-05,
      "loss": 0.3527,
      "step": 3220
    },
    {
      "epoch": 0.644,
      "loss_align": 0.15130174160003662,
      "loss_contrastive": 0.022101223468780518,
      "loss_gen": 0.12662248313426971,
      "neg_sim_mean": 0.770799458026886,
      "pos_sim_mean": 0.8486982583999634,
      "step": 3220
    },
    {
      "epoch": 0.6442,
      "grad_norm": 1.438070297241211,
      "learning_rate": 3.388671523847445e-05,
      "loss": 0.2806,
      "step": 3221
    },
    {
      "epoch": 0.6442,
      "loss_align": 0.07265347242355347,
      "loss_contrastive": 0.07329052686691284,
      "loss_gen": 0.5784471035003662,
      "neg_sim_mean": 0.9006370306015015,
      "pos_sim_mean": 0.9273465275764465,
      "step": 3221
    },
    {
      "epoch": 0.6444,
      "grad_norm": 2.5398216247558594,
      "learning_rate": 3.385367487496713e-05,
      "loss": 0.6599,
      "step": 3222
    },
    {
      "epoch": 0.6444,
      "loss_align": 0.10850870609283447,
      "loss_contrastive": 0.102572500705719,
      "loss_gen": 0.2061036378145218,
      "neg_sim_mean": 0.8940637707710266,
      "pos_sim_mean": 0.8914912939071655,
      "step": 3222
    },
    {
      "epoch": 0.6446,
      "grad_norm": 1.5999952554702759,
      "learning_rate": 3.3820642380982525e-05,
      "loss": 0.3269,
      "step": 3223
    },
    {
      "epoch": 0.6446,
      "loss_align": 0.08357280492782593,
      "loss_contrastive": 0.08419382572174072,
      "loss_gen": 0.21838724613189697,
      "neg_sim_mean": 0.9006209969520569,
      "pos_sim_mean": 0.9164271950721741,
      "step": 3223
    },
    {
      "epoch": 0.6448,
      "grad_norm": 1.6523733139038086,
      "learning_rate": 3.378761777262028e-05,
      "loss": 0.3121,
      "step": 3224
    },
    {
      "epoch": 0.6448,
      "loss_align": 0.08544623851776123,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011404263786971569,
      "neg_sim_mean": 0.39050281047821045,
      "pos_sim_mean": 0.9145537614822388,
      "step": 3224
    },
    {
      "epoch": 0.645,
      "grad_norm": 0.7061457633972168,
      "learning_rate": 3.375460106597619e-05,
      "loss": 0.0969,
      "step": 3225
    },
    {
      "epoch": 0.645,
      "loss_align": 0.057611703872680664,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2027961015701294,
      "neg_sim_mean": 0.7217674851417542,
      "pos_sim_mean": 0.9423882961273193,
      "step": 3225
    },
    {
      "epoch": 0.6452,
      "grad_norm": 1.5327718257904053,
      "learning_rate": 3.372159227714218e-05,
      "loss": 0.2604,
      "step": 3226
    },
    {
      "epoch": 0.6452,
      "loss_align": 0.05024552345275879,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0016690510092303157,
      "neg_sim_mean": 0.3760572671890259,
      "pos_sim_mean": 0.9497544765472412,
      "step": 3226
    },
    {
      "epoch": 0.6454,
      "grad_norm": 0.3037053644657135,
      "learning_rate": 3.3688591422206334e-05,
      "loss": 0.0519,
      "step": 3227
    },
    {
      "epoch": 0.6454,
      "loss_align": 0.1760295033454895,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004117908887565136,
      "neg_sim_mean": 0.4873432517051697,
      "pos_sim_mean": 0.8239704966545105,
      "step": 3227
    },
    {
      "epoch": 0.6456,
      "grad_norm": 0.6391785740852356,
      "learning_rate": 3.3655598517252885e-05,
      "loss": 0.1801,
      "step": 3228
    },
    {
      "epoch": 0.6456,
      "loss_align": 0.10518288612365723,
      "loss_contrastive": 0.08937114477157593,
      "loss_gen": 0.2242691069841385,
      "neg_sim_mean": 0.8841882348060608,
      "pos_sim_mean": 0.8948171138763428,
      "step": 3228
    },
    {
      "epoch": 0.6458,
      "grad_norm": 1.8995345830917358,
      "learning_rate": 3.362261357836216e-05,
      "loss": 0.3402,
      "step": 3229
    },
    {
      "epoch": 0.6458,
      "loss_align": 0.07789218425750732,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1248963475227356,
      "neg_sim_mean": 0.8215090036392212,
      "pos_sim_mean": 0.9221078157424927,
      "step": 3229
    },
    {
      "epoch": 0.646,
      "grad_norm": 1.4327768087387085,
      "learning_rate": 3.358963662161062e-05,
      "loss": 0.2028,
      "step": 3230
    },
    {
      "epoch": 0.646,
      "loss_align": 0.07620900869369507,
      "loss_contrastive": 0.0,
      "loss_gen": 0.019060250371694565,
      "neg_sim_mean": 0.22551557421684265,
      "pos_sim_mean": 0.9237909913063049,
      "step": 3230
    },
    {
      "epoch": 0.6462,
      "grad_norm": 0.8247203230857849,
      "learning_rate": 3.355666766307084e-05,
      "loss": 0.0953,
      "step": 3231
    },
    {
      "epoch": 0.6462,
      "loss_align": 0.10977298021316528,
      "loss_contrastive": 0.11993497610092163,
      "loss_gen": 0.3898980915546417,
      "neg_sim_mean": 0.9101619720458984,
      "pos_sim_mean": 0.8902270197868347,
      "step": 3231
    },
    {
      "epoch": 0.6464,
      "grad_norm": 2.3979156017303467,
      "learning_rate": 3.3523706718811473e-05,
      "loss": 0.5141,
      "step": 3232
    },
    {
      "epoch": 0.6464,
      "loss_align": 0.07239639759063721,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16323022544384003,
      "neg_sim_mean": 0.7899861931800842,
      "pos_sim_mean": 0.9276036024093628,
      "step": 3232
    },
    {
      "epoch": 0.6466,
      "grad_norm": 1.552547574043274,
      "learning_rate": 3.349075380489731e-05,
      "loss": 0.2356,
      "step": 3233
    },
    {
      "epoch": 0.6466,
      "loss_align": 0.14829391241073608,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01732291281223297,
      "neg_sim_mean": 0.5307000875473022,
      "pos_sim_mean": 0.8517060875892639,
      "step": 3233
    },
    {
      "epoch": 0.6468,
      "grad_norm": 0.979547917842865,
      "learning_rate": 3.34578089373892e-05,
      "loss": 0.1656,
      "step": 3234
    },
    {
      "epoch": 0.6468,
      "loss_align": 0.06898707151412964,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2007090300321579,
      "neg_sim_mean": 0.607190728187561,
      "pos_sim_mean": 0.9310129284858704,
      "step": 3234
    },
    {
      "epoch": 0.647,
      "grad_norm": 1.6006418466567993,
      "learning_rate": 3.342487213234404e-05,
      "loss": 0.2697,
      "step": 3235
    },
    {
      "epoch": 0.647,
      "loss_align": 0.11117511987686157,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008531935513019562,
      "neg_sim_mean": 0.2881784737110138,
      "pos_sim_mean": 0.8888248801231384,
      "step": 3235
    },
    {
      "epoch": 0.6472,
      "grad_norm": 0.6076663732528687,
      "learning_rate": 3.339194340581485e-05,
      "loss": 0.1197,
      "step": 3236
    },
    {
      "epoch": 0.6472,
      "loss_align": 0.10524576902389526,
      "loss_contrastive": 0.11205929517745972,
      "loss_gen": 0.15774691104888916,
      "neg_sim_mean": 0.9068135023117065,
      "pos_sim_mean": 0.8947542309761047,
      "step": 3236
    },
    {
      "epoch": 0.6474,
      "grad_norm": 1.8257699012756348,
      "learning_rate": 3.335902277385067e-05,
      "loss": 0.2764,
      "step": 3237
    },
    {
      "epoch": 0.6474,
      "loss_align": 0.08454382419586182,
      "loss_contrastive": 0.02390986680984497,
      "loss_gen": 0.20808351039886475,
      "neg_sim_mean": 0.8393660187721252,
      "pos_sim_mean": 0.9154561758041382,
      "step": 3237
    },
    {
      "epoch": 0.6476,
      "grad_norm": 1.7438173294067383,
      "learning_rate": 3.332611025249665e-05,
      "loss": 0.2955,
      "step": 3238
    },
    {
      "epoch": 0.6476,
      "loss_align": 0.09664666652679443,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23671574890613556,
      "neg_sim_mean": 0.6819654703140259,
      "pos_sim_mean": 0.9033533334732056,
      "step": 3238
    },
    {
      "epoch": 0.6478,
      "grad_norm": 1.6880030632019043,
      "learning_rate": 3.329320585779393e-05,
      "loss": 0.3334,
      "step": 3239
    },
    {
      "epoch": 0.6478,
      "loss_align": 0.06817758083343506,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009646356105804443,
      "neg_sim_mean": 0.2763281464576721,
      "pos_sim_mean": 0.9318224191665649,
      "step": 3239
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.6475438475608826,
      "learning_rate": 3.326030960577972e-05,
      "loss": 0.0778,
      "step": 3240
    },
    {
      "epoch": 0.648,
      "loss_align": 0.1489155888557434,
      "loss_contrastive": 0.07539719343185425,
      "loss_gen": 0.11678431928157806,
      "neg_sim_mean": 0.8264815807342529,
      "pos_sim_mean": 0.8510844111442566,
      "step": 3240
    },
    {
      "epoch": 0.6482,
      "grad_norm": 1.4482659101486206,
      "learning_rate": 3.322742151248725e-05,
      "loss": 0.2747,
      "step": 3241
    },
    {
      "epoch": 0.6482,
      "loss_align": 0.23366320133209229,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23546147346496582,
      "neg_sim_mean": 0.6637581586837769,
      "pos_sim_mean": 0.7663367986679077,
      "step": 3241
    },
    {
      "epoch": 0.6484,
      "grad_norm": 2.0621328353881836,
      "learning_rate": 3.319454159394578e-05,
      "loss": 0.4691,
      "step": 3242
    },
    {
      "epoch": 0.6484,
      "loss_align": 0.10255420207977295,
      "loss_contrastive": 0.08509594202041626,
      "loss_gen": 0.38501664996147156,
      "neg_sim_mean": 0.8825417160987854,
      "pos_sim_mean": 0.897445797920227,
      "step": 3242
    },
    {
      "epoch": 0.6486,
      "grad_norm": 2.2880136966705322,
      "learning_rate": 3.31616698661806e-05,
      "loss": 0.4978,
      "step": 3243
    },
    {
      "epoch": 0.6486,
      "loss_align": 0.15061217546463013,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27383658289909363,
      "neg_sim_mean": 0.6884242296218872,
      "pos_sim_mean": 0.8493878245353699,
      "step": 3243
    },
    {
      "epoch": 0.6488,
      "grad_norm": 2.15458083152771,
      "learning_rate": 3.312880634521295e-05,
      "loss": 0.4244,
      "step": 3244
    },
    {
      "epoch": 0.6488,
      "loss_align": 0.09705650806427002,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2888648509979248,
      "neg_sim_mean": 0.7647615671157837,
      "pos_sim_mean": 0.90294349193573,
      "step": 3244
    },
    {
      "epoch": 0.649,
      "grad_norm": 1.9276087284088135,
      "learning_rate": 3.309595104706015e-05,
      "loss": 0.3859,
      "step": 3245
    },
    {
      "epoch": 0.649,
      "loss_align": 0.13918828964233398,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17550459504127502,
      "neg_sim_mean": 0.5777829885482788,
      "pos_sim_mean": 0.860811710357666,
      "step": 3245
    },
    {
      "epoch": 0.6492,
      "grad_norm": 2.0544233322143555,
      "learning_rate": 3.3063103987735433e-05,
      "loss": 0.3147,
      "step": 3246
    },
    {
      "epoch": 0.6492,
      "loss_align": 0.11514729261398315,
      "loss_contrastive": 0.07936131954193115,
      "loss_gen": 0.11789686977863312,
      "neg_sim_mean": 0.8642140030860901,
      "pos_sim_mean": 0.8848527073860168,
      "step": 3246
    },
    {
      "epoch": 0.6494,
      "grad_norm": 1.1855536699295044,
      "learning_rate": 3.3030265183248096e-05,
      "loss": 0.2426,
      "step": 3247
    },
    {
      "epoch": 0.6494,
      "loss_align": 0.29532772302627563,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17445166409015656,
      "neg_sim_mean": 0.5503469705581665,
      "pos_sim_mean": 0.7046722769737244,
      "step": 3247
    },
    {
      "epoch": 0.6496,
      "grad_norm": 1.589262843132019,
      "learning_rate": 3.299743464960337e-05,
      "loss": 0.4698,
      "step": 3248
    },
    {
      "epoch": 0.6496,
      "loss_align": 0.2933756113052368,
      "loss_contrastive": 0.06917071342468262,
      "loss_gen": 0.21485044062137604,
      "neg_sim_mean": 0.6757950782775879,
      "pos_sim_mean": 0.7066243886947632,
      "step": 3248
    },
    {
      "epoch": 0.6498,
      "grad_norm": 1.901427984237671,
      "learning_rate": 3.296461240280242e-05,
      "loss": 0.5165,
      "step": 3249
    },
    {
      "epoch": 0.6498,
      "loss_align": 0.143751859664917,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0028460428584367037,
      "neg_sim_mean": 0.5834373235702515,
      "pos_sim_mean": 0.856248140335083,
      "step": 3249
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.6390999555587769,
      "learning_rate": 3.293179845884245e-05,
      "loss": 0.1466,
      "step": 3250
    },
    {
      "epoch": 0.65,
      "loss_align": 0.2591162323951721,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3362663984298706,
      "neg_sim_mean": 0.5801352262496948,
      "pos_sim_mean": 0.7408837676048279,
      "step": 3250
    },
    {
      "epoch": 0.6502,
      "grad_norm": 2.130051612854004,
      "learning_rate": 3.289899283371657e-05,
      "loss": 0.5954,
      "step": 3251
    },
    {
      "epoch": 0.6502,
      "loss_align": 0.1839134693145752,
      "loss_contrastive": 0.11088943481445312,
      "loss_gen": 0.3716084659099579,
      "neg_sim_mean": 0.82697594165802,
      "pos_sim_mean": 0.8160865306854248,
      "step": 3251
    },
    {
      "epoch": 0.6504,
      "grad_norm": 2.1843578815460205,
      "learning_rate": 3.286619554341384e-05,
      "loss": 0.5688,
      "step": 3252
    },
    {
      "epoch": 0.6504,
      "loss_align": 0.12706422805786133,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007479288149625063,
      "neg_sim_mean": 0.22616034746170044,
      "pos_sim_mean": 0.8729357719421387,
      "step": 3252
    },
    {
      "epoch": 0.6506,
      "grad_norm": 0.6145608425140381,
      "learning_rate": 3.2833406603919244e-05,
      "loss": 0.1345,
      "step": 3253
    },
    {
      "epoch": 0.6506,
      "loss_align": 0.1255580186843872,
      "loss_contrastive": 0.08870810270309448,
      "loss_gen": 0.13157354295253754,
      "neg_sim_mean": 0.8631500601768494,
      "pos_sim_mean": 0.8744419813156128,
      "step": 3253
    },
    {
      "epoch": 0.6508,
      "grad_norm": 1.5372060537338257,
      "learning_rate": 3.280062603121373e-05,
      "loss": 0.2678,
      "step": 3254
    },
    {
      "epoch": 0.6508,
      "loss_align": 0.3440220355987549,
      "loss_contrastive": 0.13694322109222412,
      "loss_gen": 0.25228390097618103,
      "neg_sim_mean": 0.6929211616516113,
      "pos_sim_mean": 0.6559779644012451,
      "step": 3254
    },
    {
      "epoch": 0.651,
      "grad_norm": 2.2997162342071533,
      "learning_rate": 3.276785384127415e-05,
      "loss": 0.6127,
      "step": 3255
    },
    {
      "epoch": 0.651,
      "loss_align": 0.12100350856781006,
      "loss_contrastive": 0.04073220491409302,
      "loss_gen": 0.21657031774520874,
      "neg_sim_mean": 0.819728672504425,
      "pos_sim_mean": 0.8789964914321899,
      "step": 3255
    },
    {
      "epoch": 0.6512,
      "grad_norm": 1.6201740503311157,
      "learning_rate": 3.273509005007327e-05,
      "loss": 0.3425,
      "step": 3256
    },
    {
      "epoch": 0.6512,
      "loss_align": 0.06278324127197266,
      "loss_contrastive": 0.0212862491607666,
      "loss_gen": 0.2031845897436142,
      "neg_sim_mean": 0.858502984046936,
      "pos_sim_mean": 0.9372167587280273,
      "step": 3256
    },
    {
      "epoch": 0.6514,
      "grad_norm": 1.600265383720398,
      "learning_rate": 3.270233467357976e-05,
      "loss": 0.2685,
      "step": 3257
    },
    {
      "epoch": 0.6514,
      "loss_align": 0.1548311710357666,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1608024537563324,
      "neg_sim_mean": 0.7336614727973938,
      "pos_sim_mean": 0.8451688289642334,
      "step": 3257
    },
    {
      "epoch": 0.6516,
      "grad_norm": 1.5503920316696167,
      "learning_rate": 3.26695877277582e-05,
      "loss": 0.3156,
      "step": 3258
    },
    {
      "epoch": 0.6516,
      "loss_align": 0.04059022665023804,
      "loss_contrastive": 0.06800389289855957,
      "loss_gen": 0.2866053581237793,
      "neg_sim_mean": 0.9274136424064636,
      "pos_sim_mean": 0.959409773349762,
      "step": 3258
    },
    {
      "epoch": 0.6518,
      "grad_norm": 1.8542993068695068,
      "learning_rate": 3.263684922856905e-05,
      "loss": 0.3354,
      "step": 3259
    },
    {
      "epoch": 0.6518,
      "loss_align": 0.15256774425506592,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2757236063480377,
      "neg_sim_mean": 0.6939165592193604,
      "pos_sim_mean": 0.8474322557449341,
      "step": 3259
    },
    {
      "epoch": 0.652,
      "grad_norm": 1.7481558322906494,
      "learning_rate": 3.2604119191968654e-05,
      "loss": 0.4283,
      "step": 3260
    },
    {
      "epoch": 0.652,
      "loss_align": 0.11061286926269531,
      "loss_contrastive": 0.0,
      "loss_gen": 0.276991069316864,
      "neg_sim_mean": 0.7482316493988037,
      "pos_sim_mean": 0.8893871307373047,
      "step": 3260
    },
    {
      "epoch": 0.6522,
      "grad_norm": 2.293924331665039,
      "learning_rate": 3.257139763390925e-05,
      "loss": 0.3876,
      "step": 3261
    },
    {
      "epoch": 0.6522,
      "loss_align": 0.1383669376373291,
      "loss_contrastive": 0.05524849891662598,
      "loss_gen": 0.1371360570192337,
      "neg_sim_mean": 0.816881537437439,
      "pos_sim_mean": 0.8616330623626709,
      "step": 3261
    },
    {
      "epoch": 0.6524,
      "grad_norm": 1.267046570777893,
      "learning_rate": 3.2538684570338905e-05,
      "loss": 0.2821,
      "step": 3262
    },
    {
      "epoch": 0.6524,
      "loss_align": 0.19428324699401855,
      "loss_contrastive": 0.014873087406158447,
      "loss_gen": 0.26774561405181885,
      "neg_sim_mean": 0.720589816570282,
      "pos_sim_mean": 0.8057167530059814,
      "step": 3262
    },
    {
      "epoch": 0.6526,
      "grad_norm": 2.0494251251220703,
      "learning_rate": 3.250598001720157e-05,
      "loss": 0.4638,
      "step": 3263
    },
    {
      "epoch": 0.6526,
      "loss_align": 0.11662125587463379,
      "loss_contrastive": 0.021758079528808594,
      "loss_gen": 0.18149082362651825,
      "neg_sim_mean": 0.8051367998123169,
      "pos_sim_mean": 0.8833787441253662,
      "step": 3263
    },
    {
      "epoch": 0.6528,
      "grad_norm": 1.5475248098373413,
      "learning_rate": 3.247328399043706e-05,
      "loss": 0.3007,
      "step": 3264
    },
    {
      "epoch": 0.6528,
      "loss_align": 0.21350455284118652,
      "loss_contrastive": 0.15622198581695557,
      "loss_gen": 0.3542604148387909,
      "neg_sim_mean": 0.8427174091339111,
      "pos_sim_mean": 0.7864954471588135,
      "step": 3264
    },
    {
      "epoch": 0.653,
      "grad_norm": 1.9322768449783325,
      "learning_rate": 3.2440596505981004e-05,
      "loss": 0.5865,
      "step": 3265
    },
    {
      "epoch": 0.653,
      "loss_align": 0.1530745029449463,
      "loss_contrastive": 0.07587319612503052,
      "loss_gen": 0.24600815773010254,
      "neg_sim_mean": 0.8227986693382263,
      "pos_sim_mean": 0.8469254970550537,
      "step": 3265
    },
    {
      "epoch": 0.6532,
      "grad_norm": 1.787232518196106,
      "learning_rate": 3.240791757976491e-05,
      "loss": 0.4082,
      "step": 3266
    },
    {
      "epoch": 0.6532,
      "loss_align": 0.0952674150466919,
      "loss_contrastive": 0.09657144546508789,
      "loss_gen": 0.15407010912895203,
      "neg_sim_mean": 0.9013040065765381,
      "pos_sim_mean": 0.9047325849533081,
      "step": 3266
    },
    {
      "epoch": 0.6534,
      "grad_norm": 1.3495395183563232,
      "learning_rate": 3.2375247227716077e-05,
      "loss": 0.2609,
      "step": 3267
    },
    {
      "epoch": 0.6534,
      "loss_align": 0.15600574016571045,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24530081450939178,
      "neg_sim_mean": 0.39354467391967773,
      "pos_sim_mean": 0.8439942598342896,
      "step": 3267
    },
    {
      "epoch": 0.6536,
      "grad_norm": 1.8903383016586304,
      "learning_rate": 3.2342585465757624e-05,
      "loss": 0.4013,
      "step": 3268
    },
    {
      "epoch": 0.6536,
      "loss_align": 0.08013385534286499,
      "loss_contrastive": 0.10654139518737793,
      "loss_gen": 0.15057919919490814,
      "neg_sim_mean": 0.926407516002655,
      "pos_sim_mean": 0.919866144657135,
      "step": 3268
    },
    {
      "epoch": 0.6538,
      "grad_norm": 1.3895223140716553,
      "learning_rate": 3.230993230980853e-05,
      "loss": 0.2435,
      "step": 3269
    },
    {
      "epoch": 0.6538,
      "loss_align": 0.12835276126861572,
      "loss_contrastive": 0.07251119613647461,
      "loss_gen": 0.13082484900951385,
      "neg_sim_mean": 0.844158411026001,
      "pos_sim_mean": 0.8716472387313843,
      "step": 3269
    },
    {
      "epoch": 0.654,
      "grad_norm": 1.1314256191253662,
      "learning_rate": 3.2277287775783525e-05,
      "loss": 0.2679,
      "step": 3270
    },
    {
      "epoch": 0.654,
      "loss_align": 0.2254301905632019,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004521870985627174,
      "neg_sim_mean": 0.41962677240371704,
      "pos_sim_mean": 0.7745698094367981,
      "step": 3270
    },
    {
      "epoch": 0.6542,
      "grad_norm": 0.6425173878669739,
      "learning_rate": 3.224465187959316e-05,
      "loss": 0.23,
      "step": 3271
    },
    {
      "epoch": 0.6542,
      "loss_align": 0.1890740990638733,
      "loss_contrastive": 0.052663207054138184,
      "loss_gen": 0.22791218757629395,
      "neg_sim_mean": 0.763589084148407,
      "pos_sim_mean": 0.8109259009361267,
      "step": 3271
    },
    {
      "epoch": 0.6544,
      "grad_norm": 1.804114818572998,
      "learning_rate": 3.2212024637143755e-05,
      "loss": 0.4233,
      "step": 3272
    },
    {
      "epoch": 0.6544,
      "loss_align": 0.13727480173110962,
      "loss_contrastive": 0.08481663465499878,
      "loss_gen": 0.5445607304573059,
      "neg_sim_mean": 0.8475418090820312,
      "pos_sim_mean": 0.8627251982688904,
      "step": 3272
    },
    {
      "epoch": 0.6546,
      "grad_norm": 3.5757226943969727,
      "learning_rate": 3.217940606433747e-05,
      "loss": 0.692,
      "step": 3273
    },
    {
      "epoch": 0.6546,
      "loss_align": 0.07241559028625488,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02998283877968788,
      "neg_sim_mean": 0.3230556845664978,
      "pos_sim_mean": 0.9275844097137451,
      "step": 3273
    },
    {
      "epoch": 0.6548,
      "grad_norm": 0.9432430863380432,
      "learning_rate": 3.214679617707218e-05,
      "loss": 0.1024,
      "step": 3274
    },
    {
      "epoch": 0.6548,
      "loss_align": 0.08659499883651733,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19606980681419373,
      "neg_sim_mean": 0.7330545783042908,
      "pos_sim_mean": 0.9134050011634827,
      "step": 3274
    },
    {
      "epoch": 0.655,
      "grad_norm": 1.871311068534851,
      "learning_rate": 3.2114194991241535e-05,
      "loss": 0.2827,
      "step": 3275
    },
    {
      "epoch": 0.655,
      "loss_align": 0.2597140669822693,
      "loss_contrastive": 0.13306468725204468,
      "loss_gen": 0.20736289024353027,
      "neg_sim_mean": 0.7733505964279175,
      "pos_sim_mean": 0.7402859330177307,
      "step": 3275
    },
    {
      "epoch": 0.6552,
      "grad_norm": 2.2675883769989014,
      "learning_rate": 3.2081602522734986e-05,
      "loss": 0.483,
      "step": 3276
    },
    {
      "epoch": 0.6552,
      "loss_align": 0.07668715715408325,
      "loss_contrastive": 0.07048195600509644,
      "loss_gen": 0.19807104766368866,
      "neg_sim_mean": 0.8937947750091553,
      "pos_sim_mean": 0.9233128428459167,
      "step": 3276
    },
    {
      "epoch": 0.6554,
      "grad_norm": 1.8213797807693481,
      "learning_rate": 3.204901878743769e-05,
      "loss": 0.2832,
      "step": 3277
    },
    {
      "epoch": 0.6554,
      "loss_align": 0.07600104808807373,
      "loss_contrastive": 0.012925684452056885,
      "loss_gen": 0.1720004379749298,
      "neg_sim_mean": 0.8369246125221252,
      "pos_sim_mean": 0.9239989519119263,
      "step": 3277
    },
    {
      "epoch": 0.6556,
      "grad_norm": 1.5378419160842896,
      "learning_rate": 3.201644380123056e-05,
      "loss": 0.2496,
      "step": 3278
    },
    {
      "epoch": 0.6556,
      "loss_align": 0.11420196294784546,
      "loss_contrastive": 0.039511263370513916,
      "loss_gen": 0.21568705141544342,
      "neg_sim_mean": 0.8253092765808105,
      "pos_sim_mean": 0.8857980370521545,
      "step": 3278
    },
    {
      "epoch": 0.6558,
      "grad_norm": 1.8603053092956543,
      "learning_rate": 3.1983877579990274e-05,
      "loss": 0.3346,
      "step": 3279
    },
    {
      "epoch": 0.6558,
      "loss_align": 0.18329322338104248,
      "loss_contrastive": 0.09170329570770264,
      "loss_gen": 0.42523452639579773,
      "neg_sim_mean": 0.8084100484848022,
      "pos_sim_mean": 0.8167067766189575,
      "step": 3279
    },
    {
      "epoch": 0.656,
      "grad_norm": 2.484457492828369,
      "learning_rate": 3.195132013958918e-05,
      "loss": 0.6195,
      "step": 3280
    },
    {
      "epoch": 0.656,
      "loss_align": 0.09096205234527588,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21397538483142853,
      "neg_sim_mean": 0.7154486775398254,
      "pos_sim_mean": 0.9090379476547241,
      "step": 3280
    },
    {
      "epoch": 0.6562,
      "grad_norm": 1.6346030235290527,
      "learning_rate": 3.1918771495895396e-05,
      "loss": 0.3049,
      "step": 3281
    },
    {
      "epoch": 0.6562,
      "loss_align": 0.1675052046775818,
      "loss_contrastive": 0.011655986309051514,
      "loss_gen": 0.24711862206459045,
      "neg_sim_mean": 0.7441507577896118,
      "pos_sim_mean": 0.8324947953224182,
      "step": 3281
    },
    {
      "epoch": 0.6564,
      "grad_norm": 1.597135305404663,
      "learning_rate": 3.188623166477272e-05,
      "loss": 0.416,
      "step": 3282
    },
    {
      "epoch": 0.6564,
      "loss_align": 0.09644728899002075,
      "loss_contrastive": 0.07572561502456665,
      "loss_gen": 0.13864196836948395,
      "neg_sim_mean": 0.879278302192688,
      "pos_sim_mean": 0.9035527110099792,
      "step": 3282
    },
    {
      "epoch": 0.6566,
      "grad_norm": 1.6034152507781982,
      "learning_rate": 3.185370066208069e-05,
      "loss": 0.2442,
      "step": 3283
    },
    {
      "epoch": 0.6566,
      "loss_align": 0.114013671875,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15563145279884338,
      "neg_sim_mean": 0.6741316318511963,
      "pos_sim_mean": 0.885986328125,
      "step": 3283
    },
    {
      "epoch": 0.6568,
      "grad_norm": 1.6581565141677856,
      "learning_rate": 3.1821178503674514e-05,
      "loss": 0.2696,
      "step": 3284
    },
    {
      "epoch": 0.6568,
      "loss_align": 0.15585362911224365,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1949733942747116,
      "neg_sim_mean": 0.6729371547698975,
      "pos_sim_mean": 0.8441463708877563,
      "step": 3284
    },
    {
      "epoch": 0.657,
      "grad_norm": 1.6613402366638184,
      "learning_rate": 3.178866520540509e-05,
      "loss": 0.3508,
      "step": 3285
    },
    {
      "epoch": 0.657,
      "loss_align": 0.13673871755599976,
      "loss_contrastive": 0.09178465604782104,
      "loss_gen": 0.288311630487442,
      "neg_sim_mean": 0.8550459146499634,
      "pos_sim_mean": 0.8632612824440002,
      "step": 3285
    },
    {
      "epoch": 0.6572,
      "grad_norm": 1.5457323789596558,
      "learning_rate": 3.1756160783119016e-05,
      "loss": 0.4361,
      "step": 3286
    },
    {
      "epoch": 0.6572,
      "loss_align": 0.07888507843017578,
      "loss_contrastive": 0.07522708177566528,
      "loss_gen": 0.0990777239203453,
      "neg_sim_mean": 0.8963419795036316,
      "pos_sim_mean": 0.9211149215698242,
      "step": 3286
    },
    {
      "epoch": 0.6574,
      "grad_norm": 1.237524151802063,
      "learning_rate": 3.1723665252658564e-05,
      "loss": 0.187,
      "step": 3287
    },
    {
      "epoch": 0.6574,
      "loss_align": 0.1002967357635498,
      "loss_contrastive": 0.027317404747009277,
      "loss_gen": 0.2137220799922943,
      "neg_sim_mean": 0.8270206451416016,
      "pos_sim_mean": 0.8997032642364502,
      "step": 3287
    },
    {
      "epoch": 0.6576,
      "grad_norm": 1.733736276626587,
      "learning_rate": 3.169117862986163e-05,
      "loss": 0.3173,
      "step": 3288
    },
    {
      "epoch": 0.6576,
      "loss_align": 0.12976187467575073,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19027744233608246,
      "neg_sim_mean": 0.7671117782592773,
      "pos_sim_mean": 0.8702381253242493,
      "step": 3288
    },
    {
      "epoch": 0.6578,
      "grad_norm": 1.4867655038833618,
      "learning_rate": 3.16587009305618e-05,
      "loss": 0.32,
      "step": 3289
    },
    {
      "epoch": 0.6578,
      "loss_align": 0.1422213912010193,
      "loss_contrastive": 0.07639420032501221,
      "loss_gen": 0.13206356763839722,
      "neg_sim_mean": 0.834172785282135,
      "pos_sim_mean": 0.8577786087989807,
      "step": 3289
    },
    {
      "epoch": 0.658,
      "grad_norm": 1.57552969455719,
      "learning_rate": 3.162623217058834e-05,
      "loss": 0.2835,
      "step": 3290
    },
    {
      "epoch": 0.658,
      "loss_align": 0.21812212467193604,
      "loss_contrastive": 0.14979833364486694,
      "loss_gen": 0.1654207706451416,
      "neg_sim_mean": 0.831676185131073,
      "pos_sim_mean": 0.781877875328064,
      "step": 3290
    },
    {
      "epoch": 0.6582,
      "grad_norm": 1.5950359106063843,
      "learning_rate": 3.1593772365766105e-05,
      "loss": 0.4015,
      "step": 3291
    },
    {
      "epoch": 0.6582,
      "loss_align": 0.05875670909881592,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17684458196163177,
      "neg_sim_mean": 0.7729843258857727,
      "pos_sim_mean": 0.9412432909011841,
      "step": 3291
    },
    {
      "epoch": 0.6584,
      "grad_norm": 1.545267939567566,
      "learning_rate": 3.156132153191562e-05,
      "loss": 0.2356,
      "step": 3292
    },
    {
      "epoch": 0.6584,
      "loss_align": 0.2135148048400879,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3700248897075653,
      "neg_sim_mean": 0.5291265845298767,
      "pos_sim_mean": 0.7864851951599121,
      "step": 3292
    },
    {
      "epoch": 0.6586,
      "grad_norm": 2.169595718383789,
      "learning_rate": 3.152887968485303e-05,
      "loss": 0.5835,
      "step": 3293
    },
    {
      "epoch": 0.6586,
      "loss_align": 0.11267971992492676,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14625583589076996,
      "neg_sim_mean": 0.7662654519081116,
      "pos_sim_mean": 0.8873202800750732,
      "step": 3293
    },
    {
      "epoch": 0.6588,
      "grad_norm": 1.3619835376739502,
      "learning_rate": 3.149644684039008e-05,
      "loss": 0.2589,
      "step": 3294
    },
    {
      "epoch": 0.6588,
      "loss_align": 0.10736358165740967,
      "loss_contrastive": 0.041783928871154785,
      "loss_gen": 0.2228836715221405,
      "neg_sim_mean": 0.8344203233718872,
      "pos_sim_mean": 0.8926364183425903,
      "step": 3294
    },
    {
      "epoch": 0.659,
      "grad_norm": 1.8043731451034546,
      "learning_rate": 3.146402301433417e-05,
      "loss": 0.3353,
      "step": 3295
    },
    {
      "epoch": 0.659,
      "loss_align": 0.13912194967269897,
      "loss_contrastive": 0.07410234212875366,
      "loss_gen": 0.1652025282382965,
      "neg_sim_mean": 0.8349803686141968,
      "pos_sim_mean": 0.860878050327301,
      "step": 3295
    },
    {
      "epoch": 0.6592,
      "grad_norm": 1.7903780937194824,
      "learning_rate": 3.1431608222488275e-05,
      "loss": 0.3132,
      "step": 3296
    },
    {
      "epoch": 0.6592,
      "loss_align": 0.12332862615585327,
      "loss_contrastive": 0.045767366886138916,
      "loss_gen": 0.25559544563293457,
      "neg_sim_mean": 0.8224387168884277,
      "pos_sim_mean": 0.8766713738441467,
      "step": 3296
    },
    {
      "epoch": 0.6594,
      "grad_norm": 2.0547900199890137,
      "learning_rate": 3.139920248065095e-05,
      "loss": 0.3844,
      "step": 3297
    },
    {
      "epoch": 0.6594,
      "loss_align": 0.19433283805847168,
      "loss_contrastive": 0.03150099515914917,
      "loss_gen": 0.2869061827659607,
      "neg_sim_mean": 0.7371681332588196,
      "pos_sim_mean": 0.8056671619415283,
      "step": 3297
    },
    {
      "epoch": 0.6596,
      "grad_norm": 2.4499499797821045,
      "learning_rate": 3.136680580461635e-05,
      "loss": 0.485,
      "step": 3298
    },
    {
      "epoch": 0.6596,
      "loss_align": 0.0819736123085022,
      "loss_contrastive": 0.03963106870651245,
      "loss_gen": 0.27394258975982666,
      "neg_sim_mean": 0.8576574325561523,
      "pos_sim_mean": 0.9180263876914978,
      "step": 3298
    },
    {
      "epoch": 0.6598,
      "grad_norm": 1.695374608039856,
      "learning_rate": 3.1334418210174263e-05,
      "loss": 0.3607,
      "step": 3299
    },
    {
      "epoch": 0.6598,
      "loss_align": 0.04391598701477051,
      "loss_contrastive": 0.058315932750701904,
      "loss_gen": 0.25706061720848083,
      "neg_sim_mean": 0.9143999218940735,
      "pos_sim_mean": 0.9560840129852295,
      "step": 3299
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.9095343351364136,
      "learning_rate": 3.130203971310999e-05,
      "loss": 0.308,
      "step": 3300
    },
    {
      "epoch": 0.66,
      "loss_align": 0.1601300835609436,
      "loss_contrastive": 0.011209726333618164,
      "loss_gen": 0.33535048365592957,
      "neg_sim_mean": 0.7510796189308167,
      "pos_sim_mean": 0.8398699164390564,
      "step": 3300
    },
    {
      "epoch": 0.6602,
      "grad_norm": 2.292924642562866,
      "learning_rate": 3.12696703292044e-05,
      "loss": 0.4968,
      "step": 3301
    },
    {
      "epoch": 0.6602,
      "loss_align": 0.0522117018699646,
      "loss_contrastive": 0.08324909210205078,
      "loss_gen": 0.2830534279346466,
      "neg_sim_mean": 0.9310373663902283,
      "pos_sim_mean": 0.9477882981300354,
      "step": 3301
    },
    {
      "epoch": 0.6604,
      "grad_norm": 2.0402636528015137,
      "learning_rate": 3.123731007423396e-05,
      "loss": 0.3453,
      "step": 3302
    },
    {
      "epoch": 0.6604,
      "loss_align": 0.1932593584060669,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17915625870227814,
      "neg_sim_mean": 0.5551106929779053,
      "pos_sim_mean": 0.8067406415939331,
      "step": 3302
    },
    {
      "epoch": 0.6606,
      "grad_norm": 1.3273048400878906,
      "learning_rate": 3.1204958963970666e-05,
      "loss": 0.3724,
      "step": 3303
    },
    {
      "epoch": 0.6606,
      "loss_align": 0.13458633422851562,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028766561299562454,
      "neg_sim_mean": 0.21012884378433228,
      "pos_sim_mean": 0.8654136657714844,
      "step": 3303
    },
    {
      "epoch": 0.6608,
      "grad_norm": 1.2314832210540771,
      "learning_rate": 3.1172617014182035e-05,
      "loss": 0.1634,
      "step": 3304
    },
    {
      "epoch": 0.6608,
      "loss_align": 0.13516128063201904,
      "loss_contrastive": 0.06038236618041992,
      "loss_gen": 0.17594242095947266,
      "neg_sim_mean": 0.825221061706543,
      "pos_sim_mean": 0.864838719367981,
      "step": 3304
    },
    {
      "epoch": 0.661,
      "grad_norm": 1.5262078046798706,
      "learning_rate": 3.114028424063118e-05,
      "loss": 0.3183,
      "step": 3305
    },
    {
      "epoch": 0.661,
      "loss_align": 0.06728661060333252,
      "loss_contrastive": 0.08841466903686523,
      "loss_gen": 0.20323719084262848,
      "neg_sim_mean": 0.9211280345916748,
      "pos_sim_mean": 0.9327133893966675,
      "step": 3305
    },
    {
      "epoch": 0.6612,
      "grad_norm": 1.9101454019546509,
      "learning_rate": 3.110796065907665e-05,
      "loss": 0.2811,
      "step": 3306
    },
    {
      "epoch": 0.6612,
      "loss_align": 0.20181089639663696,
      "loss_contrastive": 0.0781438946723938,
      "loss_gen": 0.12412495911121368,
      "neg_sim_mean": 0.7763329744338989,
      "pos_sim_mean": 0.798189103603363,
      "step": 3306
    },
    {
      "epoch": 0.6614,
      "grad_norm": 1.342677116394043,
      "learning_rate": 3.107564628527261e-05,
      "loss": 0.3353,
      "step": 3307
    },
    {
      "epoch": 0.6614,
      "loss_align": 0.11662429571151733,
      "loss_contrastive": 0.031263887882232666,
      "loss_gen": 0.3668666183948517,
      "neg_sim_mean": 0.8146395683288574,
      "pos_sim_mean": 0.8833757042884827,
      "step": 3307
    },
    {
      "epoch": 0.6616,
      "grad_norm": 2.424659490585327,
      "learning_rate": 3.104334113496865e-05,
      "loss": 0.4872,
      "step": 3308
    },
    {
      "epoch": 0.6616,
      "loss_align": 0.22007018327713013,
      "loss_contrastive": 0.0,
      "loss_gen": 0.019145043566823006,
      "neg_sim_mean": 0.3143535256385803,
      "pos_sim_mean": 0.7799298167228699,
      "step": 3308
    },
    {
      "epoch": 0.6618,
      "grad_norm": 1.0890190601348877,
      "learning_rate": 3.101104522390995e-05,
      "loss": 0.2392,
      "step": 3309
    },
    {
      "epoch": 0.6618,
      "loss_align": 0.14975279569625854,
      "loss_contrastive": 0.04004853963851929,
      "loss_gen": 0.4739305377006531,
      "neg_sim_mean": 0.7902957201004028,
      "pos_sim_mean": 0.8502472043037415,
      "step": 3309
    },
    {
      "epoch": 0.662,
      "grad_norm": 2.682056188583374,
      "learning_rate": 3.097875856783713e-05,
      "loss": 0.6285,
      "step": 3310
    },
    {
      "epoch": 0.662,
      "loss_align": 0.15117162466049194,
      "loss_contrastive": 0.08446329832077026,
      "loss_gen": 0.22066406905651093,
      "neg_sim_mean": 0.8332916498184204,
      "pos_sim_mean": 0.8488283753395081,
      "step": 3310
    },
    {
      "epoch": 0.6622,
      "grad_norm": 1.9764872789382935,
      "learning_rate": 3.09464811824863e-05,
      "loss": 0.382,
      "step": 3311
    },
    {
      "epoch": 0.6622,
      "loss_align": 0.13991159200668335,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27576929330825806,
      "neg_sim_mean": 0.5492456555366516,
      "pos_sim_mean": 0.8600884079933167,
      "step": 3311
    },
    {
      "epoch": 0.6624,
      "grad_norm": 1.7473903894424438,
      "learning_rate": 3.0914213083589086e-05,
      "loss": 0.4157,
      "step": 3312
    },
    {
      "epoch": 0.6624,
      "loss_align": 0.12863695621490479,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16524173319339752,
      "neg_sim_mean": 0.7316100597381592,
      "pos_sim_mean": 0.8713630437850952,
      "step": 3312
    },
    {
      "epoch": 0.6626,
      "grad_norm": 1.63594651222229,
      "learning_rate": 3.088195428687254e-05,
      "loss": 0.2939,
      "step": 3313
    },
    {
      "epoch": 0.6626,
      "loss_align": 0.14202934503555298,
      "loss_contrastive": 0.10122990608215332,
      "loss_gen": 0.19994866847991943,
      "neg_sim_mean": 0.8592005372047424,
      "pos_sim_mean": 0.857970654964447,
      "step": 3313
    },
    {
      "epoch": 0.6628,
      "grad_norm": 1.3291651010513306,
      "learning_rate": 3.084970480805927e-05,
      "loss": 0.3541,
      "step": 3314
    },
    {
      "epoch": 0.6628,
      "loss_align": 0.17810434103012085,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02736176736652851,
      "neg_sim_mean": 0.5885660648345947,
      "pos_sim_mean": 0.8218956589698792,
      "step": 3314
    },
    {
      "epoch": 0.663,
      "grad_norm": 1.122103214263916,
      "learning_rate": 3.081746466286719e-05,
      "loss": 0.2055,
      "step": 3315
    },
    {
      "epoch": 0.663,
      "loss_align": 0.045443058013916016,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14056475460529327,
      "neg_sim_mean": 0.8149725198745728,
      "pos_sim_mean": 0.954556941986084,
      "step": 3315
    },
    {
      "epoch": 0.6632,
      "grad_norm": 1.4154729843139648,
      "learning_rate": 3.078523386700982e-05,
      "loss": 0.186,
      "step": 3316
    },
    {
      "epoch": 0.6632,
      "loss_align": 0.05541050434112549,
      "loss_contrastive": 0.014433860778808594,
      "loss_gen": 0.09866299480199814,
      "neg_sim_mean": 0.8590233325958252,
      "pos_sim_mean": 0.9445894956588745,
      "step": 3316
    },
    {
      "epoch": 0.6634,
      "grad_norm": 1.153332233428955,
      "learning_rate": 3.075301243619603e-05,
      "loss": 0.1558,
      "step": 3317
    },
    {
      "epoch": 0.6634,
      "loss_align": 0.0767432451248169,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14538763463497162,
      "neg_sim_mean": 0.8215803503990173,
      "pos_sim_mean": 0.9232567548751831,
      "step": 3317
    },
    {
      "epoch": 0.6636,
      "grad_norm": 1.5453646183013916,
      "learning_rate": 3.072080038613018e-05,
      "loss": 0.2221,
      "step": 3318
    },
    {
      "epoch": 0.6636,
      "loss_align": 0.09622645378112793,
      "loss_contrastive": 0.0,
      "loss_gen": 0.35355058312416077,
      "neg_sim_mean": 0.5979874134063721,
      "pos_sim_mean": 0.9037735462188721,
      "step": 3318
    },
    {
      "epoch": 0.6638,
      "grad_norm": 2.3947067260742188,
      "learning_rate": 3.0688597732512e-05,
      "loss": 0.4498,
      "step": 3319
    },
    {
      "epoch": 0.6638,
      "loss_align": 0.25242120027542114,
      "loss_contrastive": 0.07569867372512817,
      "loss_gen": 0.2212110161781311,
      "neg_sim_mean": 0.7232774496078491,
      "pos_sim_mean": 0.7475787997245789,
      "step": 3319
    },
    {
      "epoch": 0.664,
      "grad_norm": 1.7600535154342651,
      "learning_rate": 3.065640449103669e-05,
      "loss": 0.4827,
      "step": 3320
    },
    {
      "epoch": 0.664,
      "loss_align": 0.09658467769622803,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13759134709835052,
      "neg_sim_mean": 0.7692450284957886,
      "pos_sim_mean": 0.903415322303772,
      "step": 3320
    },
    {
      "epoch": 0.6642,
      "grad_norm": 1.3528861999511719,
      "learning_rate": 3.062422067739485e-05,
      "loss": 0.2342,
      "step": 3321
    },
    {
      "epoch": 0.6642,
      "loss_align": 0.20357871055603027,
      "loss_contrastive": 0.09342646598815918,
      "loss_gen": 0.3696438670158386,
      "neg_sim_mean": 0.789847731590271,
      "pos_sim_mean": 0.7964212894439697,
      "step": 3321
    },
    {
      "epoch": 0.6644,
      "grad_norm": 2.212167501449585,
      "learning_rate": 3.059204630727247e-05,
      "loss": 0.5844,
      "step": 3322
    },
    {
      "epoch": 0.6644,
      "loss_align": 0.09851217269897461,
      "loss_contrastive": 0.052971601486206055,
      "loss_gen": 0.40924572944641113,
      "neg_sim_mean": 0.8544594049453735,
      "pos_sim_mean": 0.9014878273010254,
      "step": 3322
    },
    {
      "epoch": 0.6646,
      "grad_norm": 2.228761672973633,
      "learning_rate": 3.055988139635096e-05,
      "loss": 0.5141,
      "step": 3323
    },
    {
      "epoch": 0.6646,
      "loss_align": 0.11019283533096313,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13945291936397552,
      "neg_sim_mean": 0.7252452969551086,
      "pos_sim_mean": 0.8898071646690369,
      "step": 3323
    },
    {
      "epoch": 0.6648,
      "grad_norm": 1.7412431240081787,
      "learning_rate": 3.052772596030708e-05,
      "loss": 0.2496,
      "step": 3324
    },
    {
      "epoch": 0.6648,
      "loss_align": 0.14278340339660645,
      "loss_contrastive": 0.025202453136444092,
      "loss_gen": 0.17237234115600586,
      "neg_sim_mean": 0.7824190258979797,
      "pos_sim_mean": 0.8572165966033936,
      "step": 3324
    },
    {
      "epoch": 0.665,
      "grad_norm": 1.4627891778945923,
      "learning_rate": 3.049558001481302e-05,
      "loss": 0.3182,
      "step": 3325
    },
    {
      "epoch": 0.665,
      "loss_align": 0.116241455078125,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10390177369117737,
      "neg_sim_mean": 0.7634247541427612,
      "pos_sim_mean": 0.883758544921875,
      "step": 3325
    },
    {
      "epoch": 0.6652,
      "grad_norm": 1.0219646692276,
      "learning_rate": 3.046344357553632e-05,
      "loss": 0.2201,
      "step": 3326
    },
    {
      "epoch": 0.6652,
      "loss_align": 0.2827550172805786,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10618800669908524,
      "neg_sim_mean": 0.5991517305374146,
      "pos_sim_mean": 0.7172449827194214,
      "step": 3326
    },
    {
      "epoch": 0.6654,
      "grad_norm": 1.7740614414215088,
      "learning_rate": 3.043131665813988e-05,
      "loss": 0.3889,
      "step": 3327
    },
    {
      "epoch": 0.6654,
      "loss_align": 0.1355680227279663,
      "loss_contrastive": 0.06552648544311523,
      "loss_gen": 0.44422727823257446,
      "neg_sim_mean": 0.829958438873291,
      "pos_sim_mean": 0.8644319772720337,
      "step": 3327
    },
    {
      "epoch": 0.6656,
      "grad_norm": 2.620426893234253,
      "learning_rate": 3.0399199278281988e-05,
      "loss": 0.5877,
      "step": 3328
    },
    {
      "epoch": 0.6656,
      "loss_align": 0.10084891319274902,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1867973804473877,
      "neg_sim_mean": 0.3994087278842926,
      "pos_sim_mean": 0.899151086807251,
      "step": 3328
    },
    {
      "epoch": 0.6658,
      "grad_norm": 1.441920280456543,
      "learning_rate": 3.0367091451616254e-05,
      "loss": 0.2876,
      "step": 3329
    },
    {
      "epoch": 0.6658,
      "loss_align": 0.19979876279830933,
      "loss_contrastive": 0.00942540168762207,
      "loss_gen": 0.2513168752193451,
      "neg_sim_mean": 0.7096266150474548,
      "pos_sim_mean": 0.8002012372016907,
      "step": 3329
    },
    {
      "epoch": 0.666,
      "grad_norm": 1.9705407619476318,
      "learning_rate": 3.033499319379163e-05,
      "loss": 0.4522,
      "step": 3330
    },
    {
      "epoch": 0.666,
      "loss_align": 0.07859116792678833,
      "loss_contrastive": 0.05493366718292236,
      "loss_gen": 0.1931895911693573,
      "neg_sim_mean": 0.8763424754142761,
      "pos_sim_mean": 0.9214088320732117,
      "step": 3330
    },
    {
      "epoch": 0.6662,
      "grad_norm": 1.509851336479187,
      "learning_rate": 3.0302904520452447e-05,
      "loss": 0.2784,
      "step": 3331
    },
    {
      "epoch": 0.6662,
      "loss_align": 0.0720331072807312,
      "loss_contrastive": 0.03807491064071655,
      "loss_gen": 0.18598195910453796,
      "neg_sim_mean": 0.8660417795181274,
      "pos_sim_mean": 0.9279668927192688,
      "step": 3331
    },
    {
      "epoch": 0.6664,
      "grad_norm": 1.5278910398483276,
      "learning_rate": 3.0270825447238316e-05,
      "loss": 0.2626,
      "step": 3332
    },
    {
      "epoch": 0.6664,
      "loss_align": 0.08861649036407471,
      "loss_contrastive": 0.09118157625198364,
      "loss_gen": 0.1477668285369873,
      "neg_sim_mean": 0.902565062046051,
      "pos_sim_mean": 0.9113835096359253,
      "step": 3332
    },
    {
      "epoch": 0.6666,
      "grad_norm": 1.5797818899154663,
      "learning_rate": 3.023875598978419e-05,
      "loss": 0.2473,
      "step": 3333
    },
    {
      "epoch": 0.6666,
      "loss_align": 0.050609290599823,
      "loss_contrastive": 0.05345410108566284,
      "loss_gen": 0.1528952270746231,
      "neg_sim_mean": 0.9028447866439819,
      "pos_sim_mean": 0.949390709400177,
      "step": 3333
    },
    {
      "epoch": 0.6668,
      "grad_norm": 1.3978930711746216,
      "learning_rate": 3.0206696163720317e-05,
      "loss": 0.2099,
      "step": 3334
    },
    {
      "epoch": 0.6668,
      "loss_align": 0.18253493309020996,
      "loss_contrastive": 0.07044088840484619,
      "loss_gen": 0.32151466608047485,
      "neg_sim_mean": 0.7879059314727783,
      "pos_sim_mean": 0.81746506690979,
      "step": 3334
    },
    {
      "epoch": 0.667,
      "grad_norm": 2.1634445190429688,
      "learning_rate": 3.0174645984672295e-05,
      "loss": 0.5125,
      "step": 3335
    },
    {
      "epoch": 0.667,
      "loss_align": 0.053889691829681396,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0062624067068099976,
      "neg_sim_mean": 0.43795686960220337,
      "pos_sim_mean": 0.9461103081703186,
      "step": 3335
    },
    {
      "epoch": 0.6672,
      "grad_norm": 0.43352439999580383,
      "learning_rate": 3.0142605468260978e-05,
      "loss": 0.0602,
      "step": 3336
    },
    {
      "epoch": 0.6672,
      "loss_align": 0.19039487838745117,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19459964334964752,
      "neg_sim_mean": 0.5151098966598511,
      "pos_sim_mean": 0.8096051216125488,
      "step": 3336
    },
    {
      "epoch": 0.6674,
      "grad_norm": 2.354283332824707,
      "learning_rate": 3.011057463010252e-05,
      "loss": 0.385,
      "step": 3337
    },
    {
      "epoch": 0.6674,
      "loss_align": 0.10601156949996948,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2576685845851898,
      "neg_sim_mean": 0.6736723780632019,
      "pos_sim_mean": 0.8939884305000305,
      "step": 3337
    },
    {
      "epoch": 0.6676,
      "grad_norm": 1.939115047454834,
      "learning_rate": 3.007855348580837e-05,
      "loss": 0.3637,
      "step": 3338
    },
    {
      "epoch": 0.6676,
      "loss_align": 0.12779772281646729,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25277799367904663,
      "neg_sim_mean": 0.7114071846008301,
      "pos_sim_mean": 0.8722022771835327,
      "step": 3338
    },
    {
      "epoch": 0.6678,
      "grad_norm": 1.9862308502197266,
      "learning_rate": 3.0046542050985237e-05,
      "loss": 0.3806,
      "step": 3339
    },
    {
      "epoch": 0.6678,
      "loss_align": 0.2076106071472168,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24232159554958344,
      "neg_sim_mean": 0.5975801944732666,
      "pos_sim_mean": 0.7923893928527832,
      "step": 3339
    },
    {
      "epoch": 0.668,
      "grad_norm": 1.8845880031585693,
      "learning_rate": 3.001454034123512e-05,
      "loss": 0.4499,
      "step": 3340
    },
    {
      "epoch": 0.668,
      "loss_align": 0.09459781646728516,
      "loss_contrastive": 0.08714133501052856,
      "loss_gen": 0.2646706998348236,
      "neg_sim_mean": 0.8925434947013855,
      "pos_sim_mean": 0.9054021835327148,
      "step": 3340
    },
    {
      "epoch": 0.6682,
      "grad_norm": 1.9762787818908691,
      "learning_rate": 2.9982548372155263e-05,
      "loss": 0.3697,
      "step": 3341
    },
    {
      "epoch": 0.6682,
      "loss_align": 0.21638530492782593,
      "loss_contrastive": 0.0,
      "loss_gen": 0.36136260628700256,
      "neg_sim_mean": 0.5978848934173584,
      "pos_sim_mean": 0.7836146950721741,
      "step": 3341
    },
    {
      "epoch": 0.6684,
      "grad_norm": 2.404122829437256,
      "learning_rate": 2.9950566159338144e-05,
      "loss": 0.5777,
      "step": 3342
    },
    {
      "epoch": 0.6684,
      "loss_align": 0.2014368772506714,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02772507071495056,
      "neg_sim_mean": 0.37037158012390137,
      "pos_sim_mean": 0.7985631227493286,
      "step": 3342
    },
    {
      "epoch": 0.6686,
      "grad_norm": 1.2074922323226929,
      "learning_rate": 2.9918593718371508e-05,
      "loss": 0.2292,
      "step": 3343
    },
    {
      "epoch": 0.6686,
      "loss_align": 0.1743319034576416,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004134925547987223,
      "neg_sim_mean": 0.4143323302268982,
      "pos_sim_mean": 0.8256680965423584,
      "step": 3343
    },
    {
      "epoch": 0.6688,
      "grad_norm": 0.5631797313690186,
      "learning_rate": 2.9886631064838354e-05,
      "loss": 0.1785,
      "step": 3344
    },
    {
      "epoch": 0.6688,
      "loss_align": 0.11582916975021362,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2233780026435852,
      "neg_sim_mean": 0.7095006704330444,
      "pos_sim_mean": 0.8841708302497864,
      "step": 3344
    },
    {
      "epoch": 0.669,
      "grad_norm": 1.598426342010498,
      "learning_rate": 2.9854678214316873e-05,
      "loss": 0.3392,
      "step": 3345
    },
    {
      "epoch": 0.669,
      "loss_align": 0.09028983116149902,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13258680701255798,
      "neg_sim_mean": 0.7696762084960938,
      "pos_sim_mean": 0.909710168838501,
      "step": 3345
    },
    {
      "epoch": 0.6692,
      "grad_norm": 1.3730361461639404,
      "learning_rate": 2.9822735182380496e-05,
      "loss": 0.2229,
      "step": 3346
    },
    {
      "epoch": 0.6692,
      "loss_align": 0.09549033641815186,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3115769028663635,
      "neg_sim_mean": 0.8033316135406494,
      "pos_sim_mean": 0.9045096635818481,
      "step": 3346
    },
    {
      "epoch": 0.6694,
      "grad_norm": 2.0005884170532227,
      "learning_rate": 2.9790801984597883e-05,
      "loss": 0.4071,
      "step": 3347
    },
    {
      "epoch": 0.6694,
      "loss_align": 0.14886397123336792,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0033550704829394817,
      "neg_sim_mean": 0.39656805992126465,
      "pos_sim_mean": 0.8511360287666321,
      "step": 3347
    },
    {
      "epoch": 0.6696,
      "grad_norm": 0.8413963913917542,
      "learning_rate": 2.9758878636532883e-05,
      "loss": 0.1522,
      "step": 3348
    },
    {
      "epoch": 0.6696,
      "loss_align": 0.15122032165527344,
      "loss_contrastive": 0.0,
      "loss_gen": 0.198456808924675,
      "neg_sim_mean": 0.7308030128479004,
      "pos_sim_mean": 0.8487796783447266,
      "step": 3348
    },
    {
      "epoch": 0.6698,
      "grad_norm": 2.0274131298065186,
      "learning_rate": 2.972696515374455e-05,
      "loss": 0.3497,
      "step": 3349
    },
    {
      "epoch": 0.6698,
      "loss_align": 0.2302694320678711,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24571926891803741,
      "neg_sim_mean": 0.6454553008079529,
      "pos_sim_mean": 0.7697305679321289,
      "step": 3349
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.679084300994873,
      "learning_rate": 2.9695061551787108e-05,
      "loss": 0.476,
      "step": 3350
    },
    {
      "epoch": 0.67,
      "loss_align": 0.2113017439842224,
      "loss_contrastive": 0.0,
      "loss_gen": 0.253819078207016,
      "neg_sim_mean": 0.6517310738563538,
      "pos_sim_mean": 0.7886982560157776,
      "step": 3350
    },
    {
      "epoch": 0.6702,
      "grad_norm": 1.8742748498916626,
      "learning_rate": 2.9663167846209998e-05,
      "loss": 0.4651,
      "step": 3351
    },
    {
      "epoch": 0.6702,
      "loss_align": 0.19453680515289307,
      "loss_contrastive": 0.0,
      "loss_gen": 0.32911384105682373,
      "neg_sim_mean": 0.6574846506118774,
      "pos_sim_mean": 0.8054631948471069,
      "step": 3351
    },
    {
      "epoch": 0.6704,
      "grad_norm": 2.030522346496582,
      "learning_rate": 2.9631284052557827e-05,
      "loss": 0.5237,
      "step": 3352
    },
    {
      "epoch": 0.6704,
      "loss_align": 0.09247618913650513,
      "loss_contrastive": 0.08233171701431274,
      "loss_gen": 0.2857789099216461,
      "neg_sim_mean": 0.8898555040359497,
      "pos_sim_mean": 0.9075238108634949,
      "step": 3352
    },
    {
      "epoch": 0.6706,
      "grad_norm": 1.8069044351577759,
      "learning_rate": 2.9599410186370363e-05,
      "loss": 0.3881,
      "step": 3353
    },
    {
      "epoch": 0.6706,
      "loss_align": 0.1982334852218628,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24554574489593506,
      "neg_sim_mean": 0.7001843452453613,
      "pos_sim_mean": 0.8017665147781372,
      "step": 3353
    },
    {
      "epoch": 0.6708,
      "grad_norm": 1.879756212234497,
      "learning_rate": 2.9567546263182556e-05,
      "loss": 0.4438,
      "step": 3354
    },
    {
      "epoch": 0.6708,
      "loss_align": 0.08096158504486084,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3010927736759186,
      "neg_sim_mean": 0.7121298313140869,
      "pos_sim_mean": 0.9190384149551392,
      "step": 3354
    },
    {
      "epoch": 0.671,
      "grad_norm": 2.0812909603118896,
      "learning_rate": 2.9535692298524474e-05,
      "loss": 0.3821,
      "step": 3355
    },
    {
      "epoch": 0.671,
      "loss_align": 0.08490234613418579,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0017675911076366901,
      "neg_sim_mean": 0.2374604046344757,
      "pos_sim_mean": 0.9150976538658142,
      "step": 3355
    },
    {
      "epoch": 0.6712,
      "grad_norm": 0.42865562438964844,
      "learning_rate": 2.950384830792136e-05,
      "loss": 0.0867,
      "step": 3356
    },
    {
      "epoch": 0.6712,
      "loss_align": 0.14173567295074463,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1438327133655548,
      "neg_sim_mean": 0.6824131011962891,
      "pos_sim_mean": 0.8582643270492554,
      "step": 3356
    },
    {
      "epoch": 0.6714,
      "grad_norm": 1.4862416982650757,
      "learning_rate": 2.9472014306893603e-05,
      "loss": 0.2856,
      "step": 3357
    },
    {
      "epoch": 0.6714,
      "loss_align": 0.13637840747833252,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02584596537053585,
      "neg_sim_mean": 0.5341928005218506,
      "pos_sim_mean": 0.8636215925216675,
      "step": 3357
    },
    {
      "epoch": 0.6716,
      "grad_norm": 1.1730728149414062,
      "learning_rate": 2.94401903109567e-05,
      "loss": 0.1622,
      "step": 3358
    },
    {
      "epoch": 0.6716,
      "loss_align": 0.18192517757415771,
      "loss_contrastive": 0.08444380760192871,
      "loss_gen": 0.3617797791957855,
      "neg_sim_mean": 0.8025186061859131,
      "pos_sim_mean": 0.8180748224258423,
      "step": 3358
    },
    {
      "epoch": 0.6718,
      "grad_norm": 2.2874557971954346,
      "learning_rate": 2.940837633562127e-05,
      "loss": 0.5538,
      "step": 3359
    },
    {
      "epoch": 0.6718,
      "loss_align": 0.09462606906890869,
      "loss_contrastive": 0.0593792200088501,
      "loss_gen": 0.22639213502407074,
      "neg_sim_mean": 0.8647531270980835,
      "pos_sim_mean": 0.9053739309310913,
      "step": 3359
    },
    {
      "epoch": 0.672,
      "grad_norm": 1.5965474843978882,
      "learning_rate": 2.9376572396393048e-05,
      "loss": 0.3281,
      "step": 3360
    },
    {
      "epoch": 0.672,
      "loss_align": 0.11025679111480713,
      "loss_contrastive": 0.13360804319381714,
      "loss_gen": 0.23286081850528717,
      "neg_sim_mean": 0.9233512282371521,
      "pos_sim_mean": 0.8897432088851929,
      "step": 3360
    },
    {
      "epoch": 0.6722,
      "grad_norm": 1.665869951248169,
      "learning_rate": 2.934477850877292e-05,
      "loss": 0.3592,
      "step": 3361
    },
    {
      "epoch": 0.6722,
      "loss_align": 0.1416383981704712,
      "loss_contrastive": 0.006726503372192383,
      "loss_gen": 0.1525278389453888,
      "neg_sim_mean": 0.7650880813598633,
      "pos_sim_mean": 0.8583616018295288,
      "step": 3361
    },
    {
      "epoch": 0.6724,
      "grad_norm": 1.5419753789901733,
      "learning_rate": 2.9312994688256824e-05,
      "loss": 0.295,
      "step": 3362
    },
    {
      "epoch": 0.6724,
      "loss_align": 0.1270129680633545,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16677901148796082,
      "neg_sim_mean": 0.5882503986358643,
      "pos_sim_mean": 0.8729870319366455,
      "step": 3362
    },
    {
      "epoch": 0.6726,
      "grad_norm": 1.6739513874053955,
      "learning_rate": 2.9281220950335796e-05,
      "loss": 0.2938,
      "step": 3363
    },
    {
      "epoch": 0.6726,
      "loss_align": 0.1095658540725708,
      "loss_contrastive": 0.0,
      "loss_gen": 0.41843363642692566,
      "neg_sim_mean": 0.7751765251159668,
      "pos_sim_mean": 0.8904341459274292,
      "step": 3363
    },
    {
      "epoch": 0.6728,
      "grad_norm": 1.9902127981185913,
      "learning_rate": 2.924945731049599e-05,
      "loss": 0.528,
      "step": 3364
    },
    {
      "epoch": 0.6728,
      "loss_align": 0.1489274501800537,
      "loss_contrastive": 0.033765196800231934,
      "loss_gen": 0.28477340936660767,
      "neg_sim_mean": 0.7848377227783203,
      "pos_sim_mean": 0.8510725498199463,
      "step": 3364
    },
    {
      "epoch": 0.673,
      "grad_norm": 1.8711296319961548,
      "learning_rate": 2.9217703784218608e-05,
      "loss": 0.4378,
      "step": 3365
    },
    {
      "epoch": 0.673,
      "loss_align": 0.13469278812408447,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003096936969086528,
      "neg_sim_mean": 0.2818814516067505,
      "pos_sim_mean": 0.8653072118759155,
      "step": 3365
    },
    {
      "epoch": 0.6732,
      "grad_norm": 0.43428224325180054,
      "learning_rate": 2.918596038697995e-05,
      "loss": 0.1378,
      "step": 3366
    },
    {
      "epoch": 0.6732,
      "loss_align": 0.08588409423828125,
      "loss_contrastive": 0.0,
      "loss_gen": 0.014249572530388832,
      "neg_sim_mean": 0.37829142808914185,
      "pos_sim_mean": 0.9141159057617188,
      "step": 3366
    },
    {
      "epoch": 0.6734,
      "grad_norm": 0.8186710476875305,
      "learning_rate": 2.9154227134251337e-05,
      "loss": 0.1001,
      "step": 3367
    },
    {
      "epoch": 0.6734,
      "loss_align": 0.2661762237548828,
      "loss_contrastive": 0.0,
      "loss_gen": 0.025807272642850876,
      "neg_sim_mean": 0.24988260865211487,
      "pos_sim_mean": 0.7338237762451172,
      "step": 3367
    },
    {
      "epoch": 0.6736,
      "grad_norm": 1.1788266897201538,
      "learning_rate": 2.9122504041499182e-05,
      "loss": 0.292,
      "step": 3368
    },
    {
      "epoch": 0.6736,
      "loss_align": 0.07523548603057861,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010713068768382072,
      "neg_sim_mean": 0.34482046961784363,
      "pos_sim_mean": 0.9247645139694214,
      "step": 3368
    },
    {
      "epoch": 0.6738,
      "grad_norm": 0.6284713745117188,
      "learning_rate": 2.9090791124184935e-05,
      "loss": 0.0859,
      "step": 3369
    },
    {
      "epoch": 0.6738,
      "loss_align": 0.16234588623046875,
      "loss_contrastive": 0.10298192501068115,
      "loss_gen": 0.34417834877967834,
      "neg_sim_mean": 0.8406360149383545,
      "pos_sim_mean": 0.8376541137695312,
      "step": 3369
    },
    {
      "epoch": 0.674,
      "grad_norm": 2.2478675842285156,
      "learning_rate": 2.9059088397765088e-05,
      "loss": 0.5189,
      "step": 3370
    },
    {
      "epoch": 0.674,
      "loss_align": 0.09414803981781006,
      "loss_contrastive": 0.0,
      "loss_gen": 0.35427284240722656,
      "neg_sim_mean": 0.7796583771705627,
      "pos_sim_mean": 0.9058519601821899,
      "step": 3370
    },
    {
      "epoch": 0.6742,
      "grad_norm": 2.1758947372436523,
      "learning_rate": 2.9027395877691144e-05,
      "loss": 0.4484,
      "step": 3371
    },
    {
      "epoch": 0.6742,
      "loss_align": 0.14346957206726074,
      "loss_contrastive": 0.0,
      "loss_gen": 0.37267953157424927,
      "neg_sim_mean": 0.70869380235672,
      "pos_sim_mean": 0.8565304279327393,
      "step": 3371
    },
    {
      "epoch": 0.6744,
      "grad_norm": 2.050511598587036,
      "learning_rate": 2.899571357940969e-05,
      "loss": 0.5161,
      "step": 3372
    },
    {
      "epoch": 0.6744,
      "loss_align": 0.11765676736831665,
      "loss_contrastive": 0.09799367189407349,
      "loss_gen": 0.2551738917827606,
      "neg_sim_mean": 0.8803368806838989,
      "pos_sim_mean": 0.8823432326316833,
      "step": 3372
    },
    {
      "epoch": 0.6746,
      "grad_norm": 1.7167785167694092,
      "learning_rate": 2.896404151836227e-05,
      "loss": 0.3846,
      "step": 3373
    },
    {
      "epoch": 0.6746,
      "loss_align": 0.19103968143463135,
      "loss_contrastive": 0.04731649160385132,
      "loss_gen": 0.24282915890216827,
      "neg_sim_mean": 0.7562767863273621,
      "pos_sim_mean": 0.8089603185653687,
      "step": 3373
    },
    {
      "epoch": 0.6748,
      "grad_norm": 2.260607957839966,
      "learning_rate": 2.893237970998547e-05,
      "loss": 0.4395,
      "step": 3374
    },
    {
      "epoch": 0.6748,
      "loss_align": 0.15414273738861084,
      "loss_contrastive": 0.02448904514312744,
      "loss_gen": 0.3653522729873657,
      "neg_sim_mean": 0.7703462839126587,
      "pos_sim_mean": 0.8458572626113892,
      "step": 3374
    },
    {
      "epoch": 0.675,
      "grad_norm": 2.975543737411499,
      "learning_rate": 2.8900728169710867e-05,
      "loss": 0.5224,
      "step": 3375
    },
    {
      "epoch": 0.675,
      "loss_align": 0.06775212287902832,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3185766339302063,
      "neg_sim_mean": 0.831936240196228,
      "pos_sim_mean": 0.9322478771209717,
      "step": 3375
    },
    {
      "epoch": 0.6752,
      "grad_norm": 2.5309231281280518,
      "learning_rate": 2.886908691296504e-05,
      "loss": 0.3863,
      "step": 3376
    },
    {
      "epoch": 0.6752,
      "loss_align": 0.11945962905883789,
      "loss_contrastive": 0.04498088359832764,
      "loss_gen": 0.30662593245506287,
      "neg_sim_mean": 0.8255212306976318,
      "pos_sim_mean": 0.8805403709411621,
      "step": 3376
    },
    {
      "epoch": 0.6754,
      "grad_norm": 2.2801244258880615,
      "learning_rate": 2.8837455955169547e-05,
      "loss": 0.4315,
      "step": 3377
    },
    {
      "epoch": 0.6754,
      "loss_align": 0.08296942710876465,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0009160003392025828,
      "neg_sim_mean": 0.5686359405517578,
      "pos_sim_mean": 0.9170305728912354,
      "step": 3377
    },
    {
      "epoch": 0.6756,
      "grad_norm": 0.385868102312088,
      "learning_rate": 2.8805835311740932e-05,
      "loss": 0.0839,
      "step": 3378
    },
    {
      "epoch": 0.6756,
      "loss_align": 0.08476173877716064,
      "loss_contrastive": 0.08532547950744629,
      "loss_gen": 0.20513863861560822,
      "neg_sim_mean": 0.9005637168884277,
      "pos_sim_mean": 0.9152382612228394,
      "step": 3378
    },
    {
      "epoch": 0.6758,
      "grad_norm": 2.162356376647949,
      "learning_rate": 2.877422499809072e-05,
      "loss": 0.3001,
      "step": 3379
    },
    {
      "epoch": 0.6758,
      "loss_align": 0.17545998096466064,
      "loss_contrastive": 0.09327220916748047,
      "loss_gen": 0.2711605429649353,
      "neg_sim_mean": 0.8178122043609619,
      "pos_sim_mean": 0.8245400190353394,
      "step": 3379
    },
    {
      "epoch": 0.676,
      "grad_norm": 2.058840751647949,
      "learning_rate": 2.874262502962537e-05,
      "loss": 0.4578,
      "step": 3380
    },
    {
      "epoch": 0.676,
      "loss_align": 0.06639939546585083,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002749982988461852,
      "neg_sim_mean": 0.26083338260650635,
      "pos_sim_mean": 0.9336006045341492,
      "step": 3380
    },
    {
      "epoch": 0.6762,
      "grad_norm": 0.334207147359848,
      "learning_rate": 2.8711035421746367e-05,
      "loss": 0.0691,
      "step": 3381
    },
    {
      "epoch": 0.6762,
      "loss_align": 0.15892934799194336,
      "loss_contrastive": 0.09412980079650879,
      "loss_gen": 0.20760822296142578,
      "neg_sim_mean": 0.8352004289627075,
      "pos_sim_mean": 0.8410706520080566,
      "step": 3381
    },
    {
      "epoch": 0.6764,
      "grad_norm": 1.6968648433685303,
      "learning_rate": 2.8679456189850074e-05,
      "loss": 0.3778,
      "step": 3382
    },
    {
      "epoch": 0.6764,
      "loss_align": 0.22920560836791992,
      "loss_contrastive": 0.10975468158721924,
      "loss_gen": 0.15695379674434662,
      "neg_sim_mean": 0.7805490493774414,
      "pos_sim_mean": 0.7707943916320801,
      "step": 3382
    },
    {
      "epoch": 0.6766,
      "grad_norm": 2.3743906021118164,
      "learning_rate": 2.8647887349327828e-05,
      "loss": 0.3993,
      "step": 3383
    },
    {
      "epoch": 0.6766,
      "loss_align": 0.101112961769104,
      "loss_contrastive": 0.06796067953109741,
      "loss_gen": 0.23864951729774475,
      "neg_sim_mean": 0.8668476939201355,
      "pos_sim_mean": 0.898887038230896,
      "step": 3383
    },
    {
      "epoch": 0.6768,
      "grad_norm": 1.8636775016784668,
      "learning_rate": 2.8616328915565904e-05,
      "loss": 0.3479,
      "step": 3384
    },
    {
      "epoch": 0.6768,
      "loss_align": 0.07994741201400757,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16980411112308502,
      "neg_sim_mean": 0.7297854423522949,
      "pos_sim_mean": 0.9200525879859924,
      "step": 3384
    },
    {
      "epoch": 0.677,
      "grad_norm": 1.6691135168075562,
      "learning_rate": 2.858478090394549e-05,
      "loss": 0.2498,
      "step": 3385
    },
    {
      "epoch": 0.677,
      "loss_align": 0.08114230632781982,
      "loss_contrastive": 0.08455115556716919,
      "loss_gen": 0.2618134319782257,
      "neg_sim_mean": 0.9034088253974915,
      "pos_sim_mean": 0.9188576936721802,
      "step": 3385
    },
    {
      "epoch": 0.6772,
      "grad_norm": 1.940236210823059,
      "learning_rate": 2.8553243329842714e-05,
      "loss": 0.3531,
      "step": 3386
    },
    {
      "epoch": 0.6772,
      "loss_align": 0.10465764999389648,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0009433639934286475,
      "neg_sim_mean": 0.572964072227478,
      "pos_sim_mean": 0.8953423500061035,
      "step": 3386
    },
    {
      "epoch": 0.6774,
      "grad_norm": 0.29925447702407837,
      "learning_rate": 2.8521716208628595e-05,
      "loss": 0.1056,
      "step": 3387
    },
    {
      "epoch": 0.6774,
      "loss_align": 0.09221571683883667,
      "loss_contrastive": 0.11642646789550781,
      "loss_gen": 0.14155548810958862,
      "neg_sim_mean": 0.9242107272148132,
      "pos_sim_mean": 0.9077842831611633,
      "step": 3387
    },
    {
      "epoch": 0.6776,
      "grad_norm": 1.4987291097640991,
      "learning_rate": 2.8490199555669078e-05,
      "loss": 0.2477,
      "step": 3388
    },
    {
      "epoch": 0.6776,
      "loss_align": 0.18206441402435303,
      "loss_contrastive": 0.10634303092956543,
      "loss_gen": 0.2879297435283661,
      "neg_sim_mean": 0.8242785930633545,
      "pos_sim_mean": 0.817935585975647,
      "step": 3388
    },
    {
      "epoch": 0.6778,
      "grad_norm": 2.2295734882354736,
      "learning_rate": 2.8458693386325e-05,
      "loss": 0.4828,
      "step": 3389
    },
    {
      "epoch": 0.6778,
      "loss_align": 0.06751757860183716,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0033497849944978952,
      "neg_sim_mean": 0.3852616250514984,
      "pos_sim_mean": 0.9324824213981628,
      "step": 3389
    },
    {
      "epoch": 0.678,
      "grad_norm": 0.34917518496513367,
      "learning_rate": 2.8427197715952047e-05,
      "loss": 0.0709,
      "step": 3390
    },
    {
      "epoch": 0.678,
      "loss_align": 0.10842025279998779,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23547448217868805,
      "neg_sim_mean": 0.7750945687294006,
      "pos_sim_mean": 0.8915797472000122,
      "step": 3390
    },
    {
      "epoch": 0.6782,
      "grad_norm": 1.8447654247283936,
      "learning_rate": 2.8395712559900877e-05,
      "loss": 0.3439,
      "step": 3391
    },
    {
      "epoch": 0.6782,
      "loss_align": 0.09543341398239136,
      "loss_contrastive": 0.050939857959747314,
      "loss_gen": 0.22464290261268616,
      "neg_sim_mean": 0.855506420135498,
      "pos_sim_mean": 0.9045665860176086,
      "step": 3391
    },
    {
      "epoch": 0.6784,
      "grad_norm": 1.6904544830322266,
      "learning_rate": 2.8364237933516963e-05,
      "loss": 0.3262,
      "step": 3392
    },
    {
      "epoch": 0.6784,
      "loss_align": 0.08615511655807495,
      "loss_contrastive": 0.08294808864593506,
      "loss_gen": 0.2447279989719391,
      "neg_sim_mean": 0.8967929482460022,
      "pos_sim_mean": 0.913844883441925,
      "step": 3392
    },
    {
      "epoch": 0.6786,
      "grad_norm": 1.964637279510498,
      "learning_rate": 2.8332773852140644e-05,
      "loss": 0.3408,
      "step": 3393
    },
    {
      "epoch": 0.6786,
      "loss_align": 0.061171531677246094,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011065475642681122,
      "neg_sim_mean": 0.29024171829223633,
      "pos_sim_mean": 0.9388284683227539,
      "step": 3393
    },
    {
      "epoch": 0.6788,
      "grad_norm": 0.668973982334137,
      "learning_rate": 2.8301320331107128e-05,
      "loss": 0.0722,
      "step": 3394
    },
    {
      "epoch": 0.6788,
      "loss_align": 0.13427042961120605,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004841361660510302,
      "neg_sim_mean": 0.48643895983695984,
      "pos_sim_mean": 0.865729570388794,
      "step": 3394
    },
    {
      "epoch": 0.679,
      "grad_norm": 0.5570862293243408,
      "learning_rate": 2.826987738574649e-05,
      "loss": 0.1391,
      "step": 3395
    },
    {
      "epoch": 0.679,
      "loss_align": 0.20993268489837646,
      "loss_contrastive": 0.0,
      "loss_gen": 0.021761514246463776,
      "neg_sim_mean": 0.5137375593185425,
      "pos_sim_mean": 0.7900673151016235,
      "step": 3395
    },
    {
      "epoch": 0.6792,
      "grad_norm": 1.1246380805969238,
      "learning_rate": 2.823844503138363e-05,
      "loss": 0.2317,
      "step": 3396
    },
    {
      "epoch": 0.6792,
      "loss_align": 0.09715437889099121,
      "loss_contrastive": 0.0,
      "loss_gen": 0.30663055181503296,
      "neg_sim_mean": 0.7617292404174805,
      "pos_sim_mean": 0.9028456211090088,
      "step": 3396
    },
    {
      "epoch": 0.6794,
      "grad_norm": 1.896705150604248,
      "learning_rate": 2.8207023283338303e-05,
      "loss": 0.4038,
      "step": 3397
    },
    {
      "epoch": 0.6794,
      "loss_align": 0.13402950763702393,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2171606868505478,
      "neg_sim_mean": 0.5246769189834595,
      "pos_sim_mean": 0.8659704923629761,
      "step": 3397
    },
    {
      "epoch": 0.6796,
      "grad_norm": 1.7641558647155762,
      "learning_rate": 2.817561215692508e-05,
      "loss": 0.3512,
      "step": 3398
    },
    {
      "epoch": 0.6796,
      "loss_align": 0.11483979225158691,
      "loss_contrastive": 0.028401613235473633,
      "loss_gen": 0.14036786556243896,
      "neg_sim_mean": 0.8135617971420288,
      "pos_sim_mean": 0.8851602077484131,
      "step": 3398
    },
    {
      "epoch": 0.6798,
      "grad_norm": 1.2214068174362183,
      "learning_rate": 2.8144211667453368e-05,
      "loss": 0.2586,
      "step": 3399
    },
    {
      "epoch": 0.6798,
      "loss_align": 0.10810267925262451,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18180127441883087,
      "neg_sim_mean": 0.7889828681945801,
      "pos_sim_mean": 0.8918973207473755,
      "step": 3399
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.578680396080017,
      "learning_rate": 2.811282183022736e-05,
      "loss": 0.2899,
      "step": 3400
    },
    {
      "epoch": 0.68,
      "loss_align": 0.21783095598220825,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19884319603443146,
      "neg_sim_mean": 0.6468854546546936,
      "pos_sim_mean": 0.7821690440177917,
      "step": 3400
    },
    {
      "epoch": 0.6802,
      "grad_norm": 1.6698029041290283,
      "learning_rate": 2.8081442660546125e-05,
      "loss": 0.4167,
      "step": 3401
    },
    {
      "epoch": 0.6802,
      "loss_align": 0.24285966157913208,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012330824509263039,
      "neg_sim_mean": 0.3589578866958618,
      "pos_sim_mean": 0.7571403384208679,
      "step": 3401
    },
    {
      "epoch": 0.6804,
      "grad_norm": 1.137399673461914,
      "learning_rate": 2.8050074173703465e-05,
      "loss": 0.2552,
      "step": 3402
    },
    {
      "epoch": 0.6804,
      "loss_align": 0.10073065757751465,
      "loss_contrastive": 0.031801819801330566,
      "loss_gen": 0.14653828740119934,
      "neg_sim_mean": 0.831071138381958,
      "pos_sim_mean": 0.8992693424224854,
      "step": 3402
    },
    {
      "epoch": 0.6806,
      "grad_norm": 1.3936188220977783,
      "learning_rate": 2.8018716384988032e-05,
      "loss": 0.2511,
      "step": 3403
    },
    {
      "epoch": 0.6806,
      "loss_align": 0.08301728963851929,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23639211058616638,
      "neg_sim_mean": 0.6835774183273315,
      "pos_sim_mean": 0.9169827103614807,
      "step": 3403
    },
    {
      "epoch": 0.6808,
      "grad_norm": 1.7281311750411987,
      "learning_rate": 2.7987369309683148e-05,
      "loss": 0.3194,
      "step": 3404
    },
    {
      "epoch": 0.6808,
      "loss_align": 0.12812262773513794,
      "loss_contrastive": 0.0070528388023376465,
      "loss_gen": 0.14200210571289062,
      "neg_sim_mean": 0.7789301872253418,
      "pos_sim_mean": 0.8718773722648621,
      "step": 3404
    },
    {
      "epoch": 0.681,
      "grad_norm": 1.4672490358352661,
      "learning_rate": 2.795603296306708e-05,
      "loss": 0.271,
      "step": 3405
    },
    {
      "epoch": 0.681,
      "loss_align": 0.11615002155303955,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1482049971818924,
      "neg_sim_mean": 0.5979833006858826,
      "pos_sim_mean": 0.8838499784469604,
      "step": 3405
    },
    {
      "epoch": 0.6812,
      "grad_norm": 1.6685625314712524,
      "learning_rate": 2.7924707360412746e-05,
      "loss": 0.2644,
      "step": 3406
    },
    {
      "epoch": 0.6812,
      "loss_align": 0.13596266508102417,
      "loss_contrastive": 0.12393712997436523,
      "loss_gen": 0.37317416071891785,
      "neg_sim_mean": 0.8879744410514832,
      "pos_sim_mean": 0.8640373349189758,
      "step": 3406
    },
    {
      "epoch": 0.6814,
      "grad_norm": 2.160196542739868,
      "learning_rate": 2.789339251698787e-05,
      "loss": 0.524,
      "step": 3407
    },
    {
      "epoch": 0.6814,
      "loss_align": 0.12205815315246582,
      "loss_contrastive": 0.058544933795928955,
      "loss_gen": 0.14556804299354553,
      "neg_sim_mean": 0.8364867568016052,
      "pos_sim_mean": 0.8779418468475342,
      "step": 3407
    },
    {
      "epoch": 0.6816,
      "grad_norm": 1.4563308954238892,
      "learning_rate": 2.7862088448054936e-05,
      "loss": 0.2747,
      "step": 3408
    },
    {
      "epoch": 0.6816,
      "loss_align": 0.39670705795288086,
      "loss_contrastive": 0.10160112380981445,
      "loss_gen": 0.1594584435224533,
      "neg_sim_mean": 0.6048940420150757,
      "pos_sim_mean": 0.6032929420471191,
      "step": 3408
    },
    {
      "epoch": 0.6818,
      "grad_norm": 2.214811325073242,
      "learning_rate": 2.7830795168871127e-05,
      "loss": 0.5684,
      "step": 3409
    },
    {
      "epoch": 0.6818,
      "loss_align": 0.18257194757461548,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3299241065979004,
      "neg_sim_mean": 0.6623356342315674,
      "pos_sim_mean": 0.8174280524253845,
      "step": 3409
    },
    {
      "epoch": 0.682,
      "grad_norm": 1.9766429662704468,
      "learning_rate": 2.7799512694688468e-05,
      "loss": 0.5125,
      "step": 3410
    },
    {
      "epoch": 0.682,
      "loss_align": 0.06758183240890503,
      "loss_contrastive": 0.0752759575843811,
      "loss_gen": 0.3424835205078125,
      "neg_sim_mean": 0.9076941013336182,
      "pos_sim_mean": 0.932418167591095,
      "step": 3410
    },
    {
      "epoch": 0.6822,
      "grad_norm": 2.1849074363708496,
      "learning_rate": 2.776824104075364e-05,
      "loss": 0.4191,
      "step": 3411
    },
    {
      "epoch": 0.6822,
      "loss_align": 0.20868003368377686,
      "loss_contrastive": 0.06766915321350098,
      "loss_gen": 0.39963045716285706,
      "neg_sim_mean": 0.7589890956878662,
      "pos_sim_mean": 0.7913199663162231,
      "step": 3411
    },
    {
      "epoch": 0.6824,
      "grad_norm": 2.1396729946136475,
      "learning_rate": 2.773698022230804e-05,
      "loss": 0.6164,
      "step": 3412
    },
    {
      "epoch": 0.6824,
      "loss_align": 0.07486104965209961,
      "loss_contrastive": 0.00660097599029541,
      "loss_gen": 0.10991939157247543,
      "neg_sim_mean": 0.8317399024963379,
      "pos_sim_mean": 0.9251389503479004,
      "step": 3412
    },
    {
      "epoch": 0.6826,
      "grad_norm": 1.1991255283355713,
      "learning_rate": 2.7705730254587804e-05,
      "loss": 0.1856,
      "step": 3413
    },
    {
      "epoch": 0.6826,
      "loss_align": 0.13305306434631348,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16908276081085205,
      "neg_sim_mean": 0.5468578934669495,
      "pos_sim_mean": 0.8669469356536865,
      "step": 3413
    },
    {
      "epoch": 0.6828,
      "grad_norm": 1.465922236442566,
      "learning_rate": 2.7674491152823822e-05,
      "loss": 0.3021,
      "step": 3414
    },
    {
      "epoch": 0.6828,
      "loss_align": 0.11713427305221558,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2727007269859314,
      "neg_sim_mean": 0.7724540829658508,
      "pos_sim_mean": 0.8828657269477844,
      "step": 3414
    },
    {
      "epoch": 0.683,
      "grad_norm": 1.9505128860473633,
      "learning_rate": 2.7643262932241642e-05,
      "loss": 0.3898,
      "step": 3415
    },
    {
      "epoch": 0.683,
      "loss_align": 0.21688520908355713,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18427053093910217,
      "neg_sim_mean": 0.6497105360031128,
      "pos_sim_mean": 0.7831147909164429,
      "step": 3415
    },
    {
      "epoch": 0.6832,
      "grad_norm": 1.7864269018173218,
      "learning_rate": 2.761204560806152e-05,
      "loss": 0.4012,
      "step": 3416
    },
    {
      "epoch": 0.6832,
      "loss_align": 0.11698335409164429,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0058569032698869705,
      "neg_sim_mean": 0.318698912858963,
      "pos_sim_mean": 0.8830166459083557,
      "step": 3416
    },
    {
      "epoch": 0.6834,
      "grad_norm": 0.5028095841407776,
      "learning_rate": 2.7580839195498398e-05,
      "loss": 0.1228,
      "step": 3417
    },
    {
      "epoch": 0.6834,
      "loss_align": 0.1257982850074768,
      "loss_contrastive": 0.0,
      "loss_gen": 0.160805806517601,
      "neg_sim_mean": 0.7712968587875366,
      "pos_sim_mean": 0.8742017149925232,
      "step": 3417
    },
    {
      "epoch": 0.6836,
      "grad_norm": 1.527279019355774,
      "learning_rate": 2.7549643709761896e-05,
      "loss": 0.2866,
      "step": 3418
    },
    {
      "epoch": 0.6836,
      "loss_align": 0.16960978507995605,
      "loss_contrastive": 0.043817877769470215,
      "loss_gen": 0.36288881301879883,
      "neg_sim_mean": 0.7742080688476562,
      "pos_sim_mean": 0.830390214920044,
      "step": 3418
    },
    {
      "epoch": 0.6838,
      "grad_norm": 2.012173891067505,
      "learning_rate": 2.75184591660563e-05,
      "loss": 0.5378,
      "step": 3419
    },
    {
      "epoch": 0.6838,
      "loss_align": 0.07018506526947021,
      "loss_contrastive": 0.08749890327453613,
      "loss_gen": 0.30994662642478943,
      "neg_sim_mean": 0.917313814163208,
      "pos_sim_mean": 0.9298149347305298,
      "step": 3419
    },
    {
      "epoch": 0.684,
      "grad_norm": 2.132032632827759,
      "learning_rate": 2.7487285579580637e-05,
      "loss": 0.3906,
      "step": 3420
    },
    {
      "epoch": 0.684,
      "loss_align": 0.12465643882751465,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18841077387332916,
      "neg_sim_mean": 0.7518355846405029,
      "pos_sim_mean": 0.8753435611724854,
      "step": 3420
    },
    {
      "epoch": 0.6842,
      "grad_norm": 1.7168525457382202,
      "learning_rate": 2.7456122965528475e-05,
      "loss": 0.3131,
      "step": 3421
    },
    {
      "epoch": 0.6842,
      "loss_align": 0.07352840900421143,
      "loss_contrastive": 0.06404870748519897,
      "loss_gen": 0.2847365736961365,
      "neg_sim_mean": 0.8905202746391296,
      "pos_sim_mean": 0.9264715909957886,
      "step": 3421
    },
    {
      "epoch": 0.6844,
      "grad_norm": 1.5698449611663818,
      "learning_rate": 2.742497133908812e-05,
      "loss": 0.366,
      "step": 3422
    },
    {
      "epoch": 0.6844,
      "loss_align": 0.1829977035522461,
      "loss_contrastive": 0.11728107929229736,
      "loss_gen": 0.18139207363128662,
      "neg_sim_mean": 0.8342833518981934,
      "pos_sim_mean": 0.8170022964477539,
      "step": 3422
    },
    {
      "epoch": 0.6846,
      "grad_norm": 1.9235801696777344,
      "learning_rate": 2.739383071544246e-05,
      "loss": 0.3785,
      "step": 3423
    },
    {
      "epoch": 0.6846,
      "loss_align": 0.10307061672210693,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1626463532447815,
      "neg_sim_mean": 0.6906442046165466,
      "pos_sim_mean": 0.8969293832778931,
      "step": 3423
    },
    {
      "epoch": 0.6848,
      "grad_norm": 1.503133773803711,
      "learning_rate": 2.7362701109769118e-05,
      "loss": 0.2657,
      "step": 3424
    },
    {
      "epoch": 0.6848,
      "loss_align": 0.06741392612457275,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18542669713497162,
      "neg_sim_mean": 0.8108056783676147,
      "pos_sim_mean": 0.9325860738754272,
      "step": 3424
    },
    {
      "epoch": 0.685,
      "grad_norm": 2.938655138015747,
      "learning_rate": 2.733158253724024e-05,
      "loss": 0.2528,
      "step": 3425
    },
    {
      "epoch": 0.685,
      "loss_align": 0.10147011280059814,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22974839806556702,
      "neg_sim_mean": 0.7876161336898804,
      "pos_sim_mean": 0.8985298871994019,
      "step": 3425
    },
    {
      "epoch": 0.6852,
      "grad_norm": 1.7260609865188599,
      "learning_rate": 2.7300475013022663e-05,
      "loss": 0.3312,
      "step": 3426
    },
    {
      "epoch": 0.6852,
      "loss_align": 0.36352258920669556,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24295289814472198,
      "neg_sim_mean": 0.3866434395313263,
      "pos_sim_mean": 0.6364774107933044,
      "step": 3426
    },
    {
      "epoch": 0.6854,
      "grad_norm": 2.0602493286132812,
      "learning_rate": 2.7269378552277808e-05,
      "loss": 0.6065,
      "step": 3427
    },
    {
      "epoch": 0.6854,
      "loss_align": 0.07280004024505615,
      "loss_contrastive": 0.0008307695388793945,
      "loss_gen": 0.27307823300361633,
      "neg_sim_mean": 0.8280307054519653,
      "pos_sim_mean": 0.9271999597549438,
      "step": 3427
    },
    {
      "epoch": 0.6856,
      "grad_norm": 1.7816624641418457,
      "learning_rate": 2.723829317016169e-05,
      "loss": 0.346,
      "step": 3428
    },
    {
      "epoch": 0.6856,
      "loss_align": 0.21657752990722656,
      "loss_contrastive": 0.1346365213394165,
      "loss_gen": 0.281126469373703,
      "neg_sim_mean": 0.818058967590332,
      "pos_sim_mean": 0.7834224700927734,
      "step": 3428
    },
    {
      "epoch": 0.6858,
      "grad_norm": 1.700732707977295,
      "learning_rate": 2.7207218881825014e-05,
      "loss": 0.5139,
      "step": 3429
    },
    {
      "epoch": 0.6858,
      "loss_align": 0.11777585744857788,
      "loss_contrastive": 0.06219393014907837,
      "loss_gen": 0.1639142632484436,
      "neg_sim_mean": 0.8444180488586426,
      "pos_sim_mean": 0.8822241425514221,
      "step": 3429
    },
    {
      "epoch": 0.686,
      "grad_norm": 1.4799379110336304,
      "learning_rate": 2.7176155702412937e-05,
      "loss": 0.2892,
      "step": 3430
    },
    {
      "epoch": 0.686,
      "loss_align": 0.14508628845214844,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0023910347372293472,
      "neg_sim_mean": 0.2534506618976593,
      "pos_sim_mean": 0.8549137115478516,
      "step": 3430
    },
    {
      "epoch": 0.6862,
      "grad_norm": 0.5890758633613586,
      "learning_rate": 2.7145103647065308e-05,
      "loss": 0.1475,
      "step": 3431
    },
    {
      "epoch": 0.6862,
      "loss_align": 0.11703610420227051,
      "loss_contrastive": 0.10419094562530518,
      "loss_gen": 0.3173746168613434,
      "neg_sim_mean": 0.8871548175811768,
      "pos_sim_mean": 0.8829638957977295,
      "step": 3431
    },
    {
      "epoch": 0.6864,
      "grad_norm": 1.7117260694503784,
      "learning_rate": 2.7114062730916512e-05,
      "loss": 0.4469,
      "step": 3432
    },
    {
      "epoch": 0.6864,
      "loss_align": 0.06870710849761963,
      "loss_contrastive": 0.04378300905227661,
      "loss_gen": 0.18897612392902374,
      "neg_sim_mean": 0.8750758767127991,
      "pos_sim_mean": 0.9312928915023804,
      "step": 3432
    },
    {
      "epoch": 0.6866,
      "grad_norm": 1.6283080577850342,
      "learning_rate": 2.7083032969095507e-05,
      "loss": 0.2629,
      "step": 3433
    },
    {
      "epoch": 0.6866,
      "loss_align": 0.25912559032440186,
      "loss_contrastive": 0.12117505073547363,
      "loss_gen": 0.2117423117160797,
      "neg_sim_mean": 0.7620494365692139,
      "pos_sim_mean": 0.7408744096755981,
      "step": 3433
    },
    {
      "epoch": 0.6868,
      "grad_norm": 2.062220335006714,
      "learning_rate": 2.7052014376725848e-05,
      "loss": 0.4854,
      "step": 3434
    },
    {
      "epoch": 0.6868,
      "loss_align": 0.09243381023406982,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20536214113235474,
      "neg_sim_mean": 0.5748485922813416,
      "pos_sim_mean": 0.9075661897659302,
      "step": 3434
    },
    {
      "epoch": 0.687,
      "grad_norm": 1.6360974311828613,
      "learning_rate": 2.702100696892561e-05,
      "loss": 0.2978,
      "step": 3435
    },
    {
      "epoch": 0.687,
      "loss_align": 0.16698020696640015,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17759288847446442,
      "neg_sim_mean": 0.7028251886367798,
      "pos_sim_mean": 0.8330197930335999,
      "step": 3435
    },
    {
      "epoch": 0.6872,
      "grad_norm": 1.5356106758117676,
      "learning_rate": 2.699001076080742e-05,
      "loss": 0.3446,
      "step": 3436
    },
    {
      "epoch": 0.6872,
      "loss_align": 0.04980725049972534,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19603486359119415,
      "neg_sim_mean": 0.6954912543296814,
      "pos_sim_mean": 0.9501927495002747,
      "step": 3436
    },
    {
      "epoch": 0.6874,
      "grad_norm": 1.619284987449646,
      "learning_rate": 2.6959025767478463e-05,
      "loss": 0.2458,
      "step": 3437
    },
    {
      "epoch": 0.6874,
      "loss_align": 0.09327912330627441,
      "loss_contrastive": 0.0,
      "loss_gen": 0.372534841299057,
      "neg_sim_mean": 0.7804826498031616,
      "pos_sim_mean": 0.9067208766937256,
      "step": 3437
    },
    {
      "epoch": 0.6876,
      "grad_norm": 2.1323533058166504,
      "learning_rate": 2.6928052004040438e-05,
      "loss": 0.4658,
      "step": 3438
    },
    {
      "epoch": 0.6876,
      "loss_align": 0.18933796882629395,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01566108502447605,
      "neg_sim_mean": 0.3784201741218567,
      "pos_sim_mean": 0.810662031173706,
      "step": 3438
    },
    {
      "epoch": 0.6878,
      "grad_norm": 0.9713300466537476,
      "learning_rate": 2.6897089485589583e-05,
      "loss": 0.205,
      "step": 3439
    },
    {
      "epoch": 0.6878,
      "loss_align": 0.09298348426818848,
      "loss_contrastive": 0.01736980676651001,
      "loss_gen": 0.1314307451248169,
      "neg_sim_mean": 0.8243862986564636,
      "pos_sim_mean": 0.9070165157318115,
      "step": 3439
    },
    {
      "epoch": 0.688,
      "grad_norm": 1.2490739822387695,
      "learning_rate": 2.686613822721666e-05,
      "loss": 0.2265,
      "step": 3440
    },
    {
      "epoch": 0.688,
      "loss_align": 0.15628957748413086,
      "loss_contrastive": 0.1740436553955078,
      "loss_gen": 0.3334304988384247,
      "neg_sim_mean": 0.917754054069519,
      "pos_sim_mean": 0.8437104225158691,
      "step": 3440
    },
    {
      "epoch": 0.6882,
      "grad_norm": 2.175788640975952,
      "learning_rate": 2.6835198244006927e-05,
      "loss": 0.5106,
      "step": 3441
    },
    {
      "epoch": 0.6882,
      "loss_align": 0.09198015928268433,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19972467422485352,
      "neg_sim_mean": 0.7399550080299377,
      "pos_sim_mean": 0.9080198407173157,
      "step": 3441
    },
    {
      "epoch": 0.6884,
      "grad_norm": 1.8353500366210938,
      "learning_rate": 2.680426955104014e-05,
      "loss": 0.2917,
      "step": 3442
    },
    {
      "epoch": 0.6884,
      "loss_align": 0.08474743366241455,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0049431961961090565,
      "neg_sim_mean": 0.5055130124092102,
      "pos_sim_mean": 0.9152525663375854,
      "step": 3442
    },
    {
      "epoch": 0.6886,
      "grad_norm": 0.4375078082084656,
      "learning_rate": 2.6773352163390618e-05,
      "loss": 0.0897,
      "step": 3443
    },
    {
      "epoch": 0.6886,
      "loss_align": 0.0948636531829834,
      "loss_contrastive": 0.015242815017700195,
      "loss_gen": 0.21707724034786224,
      "neg_sim_mean": 0.8203791379928589,
      "pos_sim_mean": 0.9051363468170166,
      "step": 3443
    },
    {
      "epoch": 0.6888,
      "grad_norm": 1.8467937707901,
      "learning_rate": 2.674244609612708e-05,
      "loss": 0.3138,
      "step": 3444
    },
    {
      "epoch": 0.6888,
      "loss_align": 0.08916735649108887,
      "loss_contrastive": 0.025954484939575195,
      "loss_gen": 0.22963082790374756,
      "neg_sim_mean": 0.8367871046066284,
      "pos_sim_mean": 0.9108326435089111,
      "step": 3444
    },
    {
      "epoch": 0.689,
      "grad_norm": 1.7243611812591553,
      "learning_rate": 2.6711551364312787e-05,
      "loss": 0.3219,
      "step": 3445
    },
    {
      "epoch": 0.689,
      "loss_align": 0.11489701271057129,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26909857988357544,
      "neg_sim_mean": 0.6176342964172363,
      "pos_sim_mean": 0.8851029872894287,
      "step": 3445
    },
    {
      "epoch": 0.6892,
      "grad_norm": 1.841691017150879,
      "learning_rate": 2.668066798300545e-05,
      "loss": 0.384,
      "step": 3446
    },
    {
      "epoch": 0.6892,
      "loss_align": 0.16481339931488037,
      "loss_contrastive": 0.0886925458908081,
      "loss_gen": 0.20938193798065186,
      "neg_sim_mean": 0.8238791227340698,
      "pos_sim_mean": 0.8351866006851196,
      "step": 3446
    },
    {
      "epoch": 0.6894,
      "grad_norm": 1.9215047359466553,
      "learning_rate": 2.664979596725724e-05,
      "loss": 0.3848,
      "step": 3447
    },
    {
      "epoch": 0.6894,
      "loss_align": 0.07664042711257935,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27100035548210144,
      "neg_sim_mean": 0.5598363280296326,
      "pos_sim_mean": 0.9233595728874207,
      "step": 3447
    },
    {
      "epoch": 0.6896,
      "grad_norm": 1.7607449293136597,
      "learning_rate": 2.6618935332114818e-05,
      "loss": 0.3476,
      "step": 3448
    },
    {
      "epoch": 0.6896,
      "loss_align": 0.09672021865844727,
      "loss_contrastive": 0.0,
      "loss_gen": 0.039769548922777176,
      "neg_sim_mean": 0.24641326069831848,
      "pos_sim_mean": 0.9032797813415527,
      "step": 3448
    },
    {
      "epoch": 0.6898,
      "grad_norm": 1.181328296661377,
      "learning_rate": 2.658808609261928e-05,
      "loss": 0.1365,
      "step": 3449
    },
    {
      "epoch": 0.6898,
      "loss_align": 0.0725783109664917,
      "loss_contrastive": 0.054811060428619385,
      "loss_gen": 0.24100883305072784,
      "neg_sim_mean": 0.8822327256202698,
      "pos_sim_mean": 0.9274216890335083,
      "step": 3449
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.2437868118286133,
      "learning_rate": 2.6557248263806174e-05,
      "loss": 0.3202,
      "step": 3450
    },
    {
      "epoch": 0.69,
      "loss_align": 0.11312931776046753,
      "loss_contrastive": 0.046498656272888184,
      "loss_gen": 0.1606484353542328,
      "neg_sim_mean": 0.8333693146705627,
      "pos_sim_mean": 0.8868706822395325,
      "step": 3450
    },
    {
      "epoch": 0.6902,
      "grad_norm": 1.7608932256698608,
      "learning_rate": 2.6526421860705473e-05,
      "loss": 0.2794,
      "step": 3451
    },
    {
      "epoch": 0.6902,
      "loss_align": 0.16839230060577393,
      "loss_contrastive": 0.06305408477783203,
      "loss_gen": 0.19077542424201965,
      "neg_sim_mean": 0.7946617603302002,
      "pos_sim_mean": 0.8316076993942261,
      "step": 3451
    },
    {
      "epoch": 0.6904,
      "grad_norm": 1.6081231832504272,
      "learning_rate": 2.649560689834158e-05,
      "loss": 0.3667,
      "step": 3452
    },
    {
      "epoch": 0.6904,
      "loss_align": 0.1420304775238037,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13126690685749054,
      "neg_sim_mean": 0.6652359962463379,
      "pos_sim_mean": 0.8579695224761963,
      "step": 3452
    },
    {
      "epoch": 0.6906,
      "grad_norm": 1.7422021627426147,
      "learning_rate": 2.6464803391733374e-05,
      "loss": 0.2733,
      "step": 3453
    },
    {
      "epoch": 0.6906,
      "loss_align": 0.0903211236000061,
      "loss_contrastive": 0.0,
      "loss_gen": 0.019953535869717598,
      "neg_sim_mean": 0.3662835359573364,
      "pos_sim_mean": 0.9096788763999939,
      "step": 3453
    },
    {
      "epoch": 0.6908,
      "grad_norm": 1.0075154304504395,
      "learning_rate": 2.6434011355894074e-05,
      "loss": 0.1103,
      "step": 3454
    },
    {
      "epoch": 0.6908,
      "loss_align": 0.09165704250335693,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0020515169017016888,
      "neg_sim_mean": 0.24655264616012573,
      "pos_sim_mean": 0.9083429574966431,
      "step": 3454
    },
    {
      "epoch": 0.691,
      "grad_norm": 0.38744205236434937,
      "learning_rate": 2.6403230805831368e-05,
      "loss": 0.0937,
      "step": 3455
    },
    {
      "epoch": 0.691,
      "loss_align": 0.15745973587036133,
      "loss_contrastive": 0.10068738460540771,
      "loss_gen": 0.181993305683136,
      "neg_sim_mean": 0.8432276248931885,
      "pos_sim_mean": 0.8425402641296387,
      "step": 3455
    },
    {
      "epoch": 0.6912,
      "grad_norm": 1.5172570943832397,
      "learning_rate": 2.6372461756547306e-05,
      "loss": 0.3515,
      "step": 3456
    },
    {
      "epoch": 0.6912,
      "loss_align": 0.2003650665283203,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013536060228943825,
      "neg_sim_mean": 0.3427734375,
      "pos_sim_mean": 0.7996349334716797,
      "step": 3456
    },
    {
      "epoch": 0.6914,
      "grad_norm": 0.8934257626533508,
      "learning_rate": 2.634170422303835e-05,
      "loss": 0.2139,
      "step": 3457
    },
    {
      "epoch": 0.6914,
      "loss_align": 0.13663917779922485,
      "loss_contrastive": 0.08295834064483643,
      "loss_gen": 0.3149959444999695,
      "neg_sim_mean": 0.8463191390037537,
      "pos_sim_mean": 0.8633608222007751,
      "step": 3457
    },
    {
      "epoch": 0.6916,
      "grad_norm": 1.7545316219329834,
      "learning_rate": 2.6310958220295358e-05,
      "loss": 0.4616,
      "step": 3458
    },
    {
      "epoch": 0.6916,
      "loss_align": 0.1368611454963684,
      "loss_contrastive": 0.006790041923522949,
      "loss_gen": 0.23251979053020477,
      "neg_sim_mean": 0.7699288725852966,
      "pos_sim_mean": 0.8631388545036316,
      "step": 3458
    },
    {
      "epoch": 0.6918,
      "grad_norm": 1.6362534761428833,
      "learning_rate": 2.6280223763303546e-05,
      "loss": 0.3702,
      "step": 3459
    },
    {
      "epoch": 0.6918,
      "loss_align": 0.11132794618606567,
      "loss_contrastive": 0.11704748868942261,
      "loss_gen": 0.13344815373420715,
      "neg_sim_mean": 0.905719518661499,
      "pos_sim_mean": 0.8886720538139343,
      "step": 3459
    },
    {
      "epoch": 0.692,
      "grad_norm": 1.381658911705017,
      "learning_rate": 2.6249500867042522e-05,
      "loss": 0.2588,
      "step": 3460
    },
    {
      "epoch": 0.692,
      "loss_align": 0.08666831254959106,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19723887741565704,
      "neg_sim_mean": 0.606766939163208,
      "pos_sim_mean": 0.9133316874504089,
      "step": 3460
    },
    {
      "epoch": 0.6922,
      "grad_norm": 2.056624174118042,
      "learning_rate": 2.6218789546486234e-05,
      "loss": 0.2839,
      "step": 3461
    },
    {
      "epoch": 0.6922,
      "loss_align": 0.05445981025695801,
      "loss_contrastive": 0.0,
      "loss_gen": 0.277124285697937,
      "neg_sim_mean": 0.7821149230003357,
      "pos_sim_mean": 0.945540189743042,
      "step": 3461
    },
    {
      "epoch": 0.6924,
      "grad_norm": 2.0355935096740723,
      "learning_rate": 2.618808981660304e-05,
      "loss": 0.3316,
      "step": 3462
    },
    {
      "epoch": 0.6924,
      "loss_align": 0.18606603145599365,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2775539457798004,
      "neg_sim_mean": 0.6718041896820068,
      "pos_sim_mean": 0.8139339685440063,
      "step": 3462
    },
    {
      "epoch": 0.6926,
      "grad_norm": 1.8111754655838013,
      "learning_rate": 2.6157401692355597e-05,
      "loss": 0.4636,
      "step": 3463
    },
    {
      "epoch": 0.6926,
      "loss_align": 0.14874815940856934,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15165257453918457,
      "neg_sim_mean": 0.6914589405059814,
      "pos_sim_mean": 0.8512518405914307,
      "step": 3463
    },
    {
      "epoch": 0.6928,
      "grad_norm": 1.6905653476715088,
      "learning_rate": 2.612672518870093e-05,
      "loss": 0.3004,
      "step": 3464
    },
    {
      "epoch": 0.6928,
      "loss_align": 0.14176225662231445,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003447312628850341,
      "neg_sim_mean": 0.3426169157028198,
      "pos_sim_mean": 0.8582377433776855,
      "step": 3464
    },
    {
      "epoch": 0.693,
      "grad_norm": 0.5287066698074341,
      "learning_rate": 2.6096060320590393e-05,
      "loss": 0.1452,
      "step": 3465
    },
    {
      "epoch": 0.693,
      "loss_align": 0.14003020524978638,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2506279945373535,
      "neg_sim_mean": 0.29492565989494324,
      "pos_sim_mean": 0.8599697947502136,
      "step": 3465
    },
    {
      "epoch": 0.6932,
      "grad_norm": 2.0435726642608643,
      "learning_rate": 2.6065407102969664e-05,
      "loss": 0.3907,
      "step": 3466
    },
    {
      "epoch": 0.6932,
      "loss_align": 0.11845290660858154,
      "loss_contrastive": 0.040843963623046875,
      "loss_gen": 0.2725331485271454,
      "neg_sim_mean": 0.8223910331726074,
      "pos_sim_mean": 0.8815470933914185,
      "step": 3466
    },
    {
      "epoch": 0.6934,
      "grad_norm": 2.0903913974761963,
      "learning_rate": 2.603476555077875e-05,
      "loss": 0.3959,
      "step": 3467
    },
    {
      "epoch": 0.6934,
      "loss_align": 0.14113378524780273,
      "loss_contrastive": 0.0,
      "loss_gen": 0.027836207300424576,
      "neg_sim_mean": 0.2931380271911621,
      "pos_sim_mean": 0.8588662147521973,
      "step": 3467
    },
    {
      "epoch": 0.6936,
      "grad_norm": 1.2177820205688477,
      "learning_rate": 2.600413567895198e-05,
      "loss": 0.169,
      "step": 3468
    },
    {
      "epoch": 0.6936,
      "loss_align": 0.08097362518310547,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2022063285112381,
      "neg_sim_mean": 0.7807432413101196,
      "pos_sim_mean": 0.9190263748168945,
      "step": 3468
    },
    {
      "epoch": 0.6938,
      "grad_norm": 1.5387729406356812,
      "learning_rate": 2.5973517502417966e-05,
      "loss": 0.2832,
      "step": 3469
    },
    {
      "epoch": 0.6938,
      "loss_align": 0.11893290281295776,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4275219738483429,
      "neg_sim_mean": 0.6471670269966125,
      "pos_sim_mean": 0.8810670971870422,
      "step": 3469
    },
    {
      "epoch": 0.694,
      "grad_norm": 2.2147202491760254,
      "learning_rate": 2.5942911036099655e-05,
      "loss": 0.5465,
      "step": 3470
    },
    {
      "epoch": 0.694,
      "loss_align": 0.13928091526031494,
      "loss_contrastive": 0.004301965236663818,
      "loss_gen": 0.15649528801441193,
      "neg_sim_mean": 0.765021026134491,
      "pos_sim_mean": 0.8607190847396851,
      "step": 3470
    },
    {
      "epoch": 0.6942,
      "grad_norm": 1.548514485359192,
      "learning_rate": 2.591231629491423e-05,
      "loss": 0.2963,
      "step": 3471
    },
    {
      "epoch": 0.6942,
      "loss_align": 0.10161960124969482,
      "loss_contrastive": 0.08610665798187256,
      "loss_gen": 0.18037720024585724,
      "neg_sim_mean": 0.8844870328903198,
      "pos_sim_mean": 0.8983803987503052,
      "step": 3471
    },
    {
      "epoch": 0.6944,
      "grad_norm": 1.4846594333648682,
      "learning_rate": 2.5881733293773242e-05,
      "loss": 0.2923,
      "step": 3472
    },
    {
      "epoch": 0.6944,
      "loss_align": 0.2861713171005249,
      "loss_contrastive": 0.08800172805786133,
      "loss_gen": 0.31849586963653564,
      "neg_sim_mean": 0.7018303871154785,
      "pos_sim_mean": 0.7138286828994751,
      "step": 3472
    },
    {
      "epoch": 0.6946,
      "grad_norm": 2.1490683555603027,
      "learning_rate": 2.585116204758248e-05,
      "loss": 0.6152,
      "step": 3473
    },
    {
      "epoch": 0.6946,
      "loss_align": 0.07892799377441406,
      "loss_contrastive": 0.0022951364517211914,
      "loss_gen": 0.12152645736932755,
      "neg_sim_mean": 0.8233671188354492,
      "pos_sim_mean": 0.9210720062255859,
      "step": 3473
    },
    {
      "epoch": 0.6948,
      "grad_norm": 1.2497509717941284,
      "learning_rate": 2.5820602571241947e-05,
      "loss": 0.2007,
      "step": 3474
    },
    {
      "epoch": 0.6948,
      "loss_align": 0.0664106011390686,
      "loss_contrastive": 0.03794616460800171,
      "loss_gen": 0.38697147369384766,
      "neg_sim_mean": 0.8715355396270752,
      "pos_sim_mean": 0.9335893988609314,
      "step": 3474
    },
    {
      "epoch": 0.695,
      "grad_norm": 2.4494805335998535,
      "learning_rate": 2.5790054879645965e-05,
      "loss": 0.4579,
      "step": 3475
    },
    {
      "epoch": 0.695,
      "loss_align": 0.10520148277282715,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26462656259536743,
      "neg_sim_mean": 0.782158374786377,
      "pos_sim_mean": 0.8947985172271729,
      "step": 3475
    },
    {
      "epoch": 0.6952,
      "grad_norm": 1.9782196283340454,
      "learning_rate": 2.575951898768315e-05,
      "loss": 0.3698,
      "step": 3476
    },
    {
      "epoch": 0.6952,
      "loss_align": 0.08801233768463135,
      "loss_contrastive": 0.03370559215545654,
      "loss_gen": 0.3081095218658447,
      "neg_sim_mean": 0.8456932306289673,
      "pos_sim_mean": 0.9119876623153687,
      "step": 3476
    },
    {
      "epoch": 0.6954,
      "grad_norm": 2.4105169773101807,
      "learning_rate": 2.5728994910236304e-05,
      "loss": 0.4002,
      "step": 3477
    },
    {
      "epoch": 0.6954,
      "loss_align": 0.22093665599822998,
      "loss_contrastive": 0.10591632127761841,
      "loss_gen": 0.1269683688879013,
      "neg_sim_mean": 0.7849796414375305,
      "pos_sim_mean": 0.77906334400177,
      "step": 3477
    },
    {
      "epoch": 0.6956,
      "grad_norm": 1.7739970684051514,
      "learning_rate": 2.5698482662182493e-05,
      "loss": 0.3606,
      "step": 3478
    },
    {
      "epoch": 0.6956,
      "loss_align": 0.16837584972381592,
      "loss_contrastive": 0.042408645153045654,
      "loss_gen": 0.2218329757452011,
      "neg_sim_mean": 0.7740327715873718,
      "pos_sim_mean": 0.8316241502761841,
      "step": 3478
    },
    {
      "epoch": 0.6958,
      "grad_norm": 1.5699598789215088,
      "learning_rate": 2.5667982258393014e-05,
      "loss": 0.3953,
      "step": 3479
    },
    {
      "epoch": 0.6958,
      "loss_align": 0.14431476593017578,
      "loss_contrastive": 0.04634559154510498,
      "loss_gen": 0.24892380833625793,
      "neg_sim_mean": 0.8020308017730713,
      "pos_sim_mean": 0.8556852340698242,
      "step": 3479
    },
    {
      "epoch": 0.696,
      "grad_norm": 1.8693948984146118,
      "learning_rate": 2.5637493713733374e-05,
      "loss": 0.3988,
      "step": 3480
    },
    {
      "epoch": 0.696,
      "loss_align": 0.18824875354766846,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01116205658763647,
      "neg_sim_mean": 0.41783571243286133,
      "pos_sim_mean": 0.8117512464523315,
      "step": 3480
    },
    {
      "epoch": 0.6962,
      "grad_norm": 0.7925358414649963,
      "learning_rate": 2.560701704306336e-05,
      "loss": 0.1994,
      "step": 3481
    },
    {
      "epoch": 0.6962,
      "loss_align": 0.0705728530883789,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24244563281536102,
      "neg_sim_mean": 0.30844587087631226,
      "pos_sim_mean": 0.9294271469116211,
      "step": 3481
    },
    {
      "epoch": 0.6964,
      "grad_norm": 1.887631893157959,
      "learning_rate": 2.557655226123693e-05,
      "loss": 0.313,
      "step": 3482
    },
    {
      "epoch": 0.6964,
      "loss_align": 0.10841923952102661,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21787501871585846,
      "neg_sim_mean": 0.5982205867767334,
      "pos_sim_mean": 0.8915807604789734,
      "step": 3482
    },
    {
      "epoch": 0.6966,
      "grad_norm": 1.9046001434326172,
      "learning_rate": 2.5546099383102207e-05,
      "loss": 0.3263,
      "step": 3483
    },
    {
      "epoch": 0.6966,
      "loss_align": 0.1028822660446167,
      "loss_contrastive": 0.03329336643218994,
      "loss_gen": 0.25004109740257263,
      "neg_sim_mean": 0.8304110765457153,
      "pos_sim_mean": 0.8971177339553833,
      "step": 3483
    },
    {
      "epoch": 0.6968,
      "grad_norm": 1.8610599040985107,
      "learning_rate": 2.551565842350157e-05,
      "loss": 0.3569,
      "step": 3484
    },
    {
      "epoch": 0.6968,
      "loss_align": 0.13716256618499756,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15302492678165436,
      "neg_sim_mean": 0.7126594185829163,
      "pos_sim_mean": 0.8628374338150024,
      "step": 3484
    },
    {
      "epoch": 0.697,
      "grad_norm": 1.655576467514038,
      "learning_rate": 2.5485229397271565e-05,
      "loss": 0.2902,
      "step": 3485
    },
    {
      "epoch": 0.697,
      "loss_align": 0.05912017822265625,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011873469920828938,
      "neg_sim_mean": 0.2727503180503845,
      "pos_sim_mean": 0.9408798217773438,
      "step": 3485
    },
    {
      "epoch": 0.6972,
      "grad_norm": 0.3271963894367218,
      "learning_rate": 2.545481231924296e-05,
      "loss": 0.0603,
      "step": 3486
    },
    {
      "epoch": 0.6972,
      "loss_align": 0.06890261173248291,
      "loss_contrastive": 0.03147232532501221,
      "loss_gen": 0.3891139328479767,
      "neg_sim_mean": 0.8625696897506714,
      "pos_sim_mean": 0.9310973882675171,
      "step": 3486
    },
    {
      "epoch": 0.6974,
      "grad_norm": 2.3542168140411377,
      "learning_rate": 2.5424407204240653e-05,
      "loss": 0.4618,
      "step": 3487
    },
    {
      "epoch": 0.6974,
      "loss_align": 0.14181774854660034,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001327495207078755,
      "neg_sim_mean": 0.5205811858177185,
      "pos_sim_mean": 0.8581822514533997,
      "step": 3487
    },
    {
      "epoch": 0.6976,
      "grad_norm": 0.45801860094070435,
      "learning_rate": 2.5394014067083728e-05,
      "loss": 0.1431,
      "step": 3488
    },
    {
      "epoch": 0.6976,
      "loss_align": 0.07138323783874512,
      "loss_contrastive": 0.05663096904754639,
      "loss_gen": 0.32230278849601746,
      "neg_sim_mean": 0.8852477073669434,
      "pos_sim_mean": 0.9286167621612549,
      "step": 3488
    },
    {
      "epoch": 0.6978,
      "grad_norm": 2.143214225769043,
      "learning_rate": 2.536363292258543e-05,
      "loss": 0.4005,
      "step": 3489
    },
    {
      "epoch": 0.6978,
      "loss_align": 0.15292805433273315,
      "loss_contrastive": 0.10694658756256104,
      "loss_gen": 0.47538429498672485,
      "neg_sim_mean": 0.85401850938797,
      "pos_sim_mean": 0.8470719456672668,
      "step": 3489
    },
    {
      "epoch": 0.698,
      "grad_norm": 2.7215511798858643,
      "learning_rate": 2.533326378555314e-05,
      "loss": 0.6411,
      "step": 3490
    },
    {
      "epoch": 0.698,
      "loss_align": 0.1448444128036499,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18869000673294067,
      "neg_sim_mean": 0.709854245185852,
      "pos_sim_mean": 0.8551555871963501,
      "step": 3490
    },
    {
      "epoch": 0.6982,
      "grad_norm": 1.5739878416061401,
      "learning_rate": 2.5302906670788462e-05,
      "loss": 0.3335,
      "step": 3491
    },
    {
      "epoch": 0.6982,
      "loss_align": 0.11310023069381714,
      "loss_contrastive": 0.020151197910308838,
      "loss_gen": 0.31118953227996826,
      "neg_sim_mean": 0.8070509433746338,
      "pos_sim_mean": 0.8868997693061829,
      "step": 3491
    },
    {
      "epoch": 0.6984,
      "grad_norm": 2.2103872299194336,
      "learning_rate": 2.527256159308703e-05,
      "loss": 0.4267,
      "step": 3492
    },
    {
      "epoch": 0.6984,
      "loss_align": 0.11459630727767944,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13014966249465942,
      "neg_sim_mean": 0.7197328805923462,
      "pos_sim_mean": 0.8854036927223206,
      "step": 3492
    },
    {
      "epoch": 0.6986,
      "grad_norm": 1.368425726890564,
      "learning_rate": 2.524222856723869e-05,
      "loss": 0.2447,
      "step": 3493
    },
    {
      "epoch": 0.6986,
      "loss_align": 0.08379685878753662,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2961612641811371,
      "neg_sim_mean": 0.7761412262916565,
      "pos_sim_mean": 0.9162031412124634,
      "step": 3493
    },
    {
      "epoch": 0.6988,
      "grad_norm": 2.126962661743164,
      "learning_rate": 2.5211907608027362e-05,
      "loss": 0.38,
      "step": 3494
    },
    {
      "epoch": 0.6988,
      "loss_align": 0.17266017198562622,
      "loss_contrastive": 0.04299074411392212,
      "loss_gen": 0.3910781741142273,
      "neg_sim_mean": 0.770330548286438,
      "pos_sim_mean": 0.8273398280143738,
      "step": 3494
    },
    {
      "epoch": 0.699,
      "grad_norm": 2.2577199935913086,
      "learning_rate": 2.518159873023116e-05,
      "loss": 0.5689,
      "step": 3495
    },
    {
      "epoch": 0.699,
      "loss_align": 0.1508086919784546,
      "loss_contrastive": 0.11670708656311035,
      "loss_gen": 0.21107803285121918,
      "neg_sim_mean": 0.8658983707427979,
      "pos_sim_mean": 0.8491913080215454,
      "step": 3495
    },
    {
      "epoch": 0.6992,
      "grad_norm": 1.8043010234832764,
      "learning_rate": 2.5151301948622237e-05,
      "loss": 0.3759,
      "step": 3496
    },
    {
      "epoch": 0.6992,
      "loss_align": 0.20133590698242188,
      "loss_contrastive": 0.0333784818649292,
      "loss_gen": 0.17898906767368317,
      "neg_sim_mean": 0.7320425510406494,
      "pos_sim_mean": 0.7986640930175781,
      "step": 3496
    },
    {
      "epoch": 0.6994,
      "grad_norm": 1.928348183631897,
      "learning_rate": 2.5121017277966874e-05,
      "loss": 0.3843,
      "step": 3497
    },
    {
      "epoch": 0.6994,
      "loss_align": 0.09035617113113403,
      "loss_contrastive": 0.09029406309127808,
      "loss_gen": 0.13486376404762268,
      "neg_sim_mean": 0.8999378681182861,
      "pos_sim_mean": 0.909643828868866,
      "step": 3497
    },
    {
      "epoch": 0.6996,
      "grad_norm": 1.3132961988449097,
      "learning_rate": 2.509074473302546e-05,
      "loss": 0.2361,
      "step": 3498
    },
    {
      "epoch": 0.6996,
      "loss_align": 0.17546093463897705,
      "loss_contrastive": 0.17912226915359497,
      "loss_gen": 0.07847264409065247,
      "neg_sim_mean": 0.90366131067276,
      "pos_sim_mean": 0.824539065361023,
      "step": 3498
    },
    {
      "epoch": 0.6998,
      "grad_norm": 1.1859514713287354,
      "learning_rate": 2.506048432855247e-05,
      "loss": 0.2754,
      "step": 3499
    },
    {
      "epoch": 0.6998,
      "loss_align": 0.13693344593048096,
      "loss_contrastive": 0.04591524600982666,
      "loss_gen": 0.19479170441627502,
      "neg_sim_mean": 0.8089817762374878,
      "pos_sim_mean": 0.863066554069519,
      "step": 3499
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.6089060306549072,
      "learning_rate": 2.5030236079296444e-05,
      "loss": 0.3372,
      "step": 3500
    },
    {
      "epoch": 0.7,
      "loss_align": 0.20611923933029175,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1579541563987732,
      "neg_sim_mean": 0.556200385093689,
      "pos_sim_mean": 0.7938807606697083,
      "step": 3500
    },
    {
      "epoch": 0.7002,
      "grad_norm": 1.9654850959777832,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.3641,
      "step": 3501
    },
    {
      "epoch": 0.7002,
      "loss_align": 0.1633281707763672,
      "loss_contrastive": 0.0,
      "loss_gen": 0.28886038064956665,
      "neg_sim_mean": 0.590107798576355,
      "pos_sim_mean": 0.8366718292236328,
      "step": 3501
    },
    {
      "epoch": 0.7004,
      "grad_norm": 2.1036672592163086,
      "learning_rate": 2.4969776105399878e-05,
      "loss": 0.4522,
      "step": 3502
    },
    {
      "epoch": 0.7004,
      "loss_align": 0.10909509658813477,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3629816770553589,
      "neg_sim_mean": 0.7019964456558228,
      "pos_sim_mean": 0.8909049034118652,
      "step": 3502
    },
    {
      "epoch": 0.7006,
      "grad_norm": 2.435274124145508,
      "learning_rate": 2.49395644102268e-05,
      "loss": 0.4721,
      "step": 3503
    },
    {
      "epoch": 0.7006,
      "loss_align": 0.07138442993164062,
      "loss_contrastive": 0.02351248264312744,
      "loss_gen": 0.1919330507516861,
      "neg_sim_mean": 0.8521280288696289,
      "pos_sim_mean": 0.9286155700683594,
      "step": 3503
    },
    {
      "epoch": 0.7008,
      "grad_norm": 1.526808261871338,
      "learning_rate": 2.4909364929205576e-05,
      "loss": 0.2661,
      "step": 3504
    },
    {
      "epoch": 0.7008,
      "loss_align": 0.1395295262336731,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26272663474082947,
      "neg_sim_mean": 0.46515482664108276,
      "pos_sim_mean": 0.8604704737663269,
      "step": 3504
    },
    {
      "epoch": 0.701,
      "grad_norm": 1.8614174127578735,
      "learning_rate": 2.48791776770551e-05,
      "loss": 0.4023,
      "step": 3505
    },
    {
      "epoch": 0.701,
      "loss_align": 0.21910059452056885,
      "loss_contrastive": 0.0556110143661499,
      "loss_gen": 0.23246657848358154,
      "neg_sim_mean": 0.7365103960037231,
      "pos_sim_mean": 0.7808994054794312,
      "step": 3505
    },
    {
      "epoch": 0.7012,
      "grad_norm": 2.0965983867645264,
      "learning_rate": 2.4849002668488245e-05,
      "loss": 0.4582,
      "step": 3506
    },
    {
      "epoch": 0.7012,
      "loss_align": 0.13029831647872925,
      "loss_contrastive": 0.06104558706283569,
      "loss_gen": 0.33591708540916443,
      "neg_sim_mean": 0.8307472467422485,
      "pos_sim_mean": 0.8697016835212708,
      "step": 3506
    },
    {
      "epoch": 0.7014,
      "grad_norm": 2.148163080215454,
      "learning_rate": 2.4818839918211962e-05,
      "loss": 0.4735,
      "step": 3507
    },
    {
      "epoch": 0.7014,
      "loss_align": 0.15508681535720825,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2661462724208832,
      "neg_sim_mean": 0.6324313879013062,
      "pos_sim_mean": 0.8449131846427917,
      "step": 3507
    },
    {
      "epoch": 0.7016,
      "grad_norm": 1.825021743774414,
      "learning_rate": 2.4788689440927194e-05,
      "loss": 0.4212,
      "step": 3508
    },
    {
      "epoch": 0.7016,
      "loss_align": 0.11793345212936401,
      "loss_contrastive": 0.12372291088104248,
      "loss_gen": 0.1267920285463333,
      "neg_sim_mean": 0.9057894349098206,
      "pos_sim_mean": 0.882066547870636,
      "step": 3508
    },
    {
      "epoch": 0.7018,
      "grad_norm": 1.4504921436309814,
      "learning_rate": 2.4758551251328922e-05,
      "loss": 0.2596,
      "step": 3509
    },
    {
      "epoch": 0.7018,
      "loss_align": 0.10180771350860596,
      "loss_contrastive": 0.0,
      "loss_gen": 0.213450625538826,
      "neg_sim_mean": 0.47505998611450195,
      "pos_sim_mean": 0.898192286491394,
      "step": 3509
    },
    {
      "epoch": 0.702,
      "grad_norm": 2.279571056365967,
      "learning_rate": 2.4728425364106135e-05,
      "loss": 0.3153,
      "step": 3510
    },
    {
      "epoch": 0.702,
      "loss_align": 0.10731852054595947,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2622006833553314,
      "neg_sim_mean": 0.7874417304992676,
      "pos_sim_mean": 0.8926814794540405,
      "step": 3510
    },
    {
      "epoch": 0.7022,
      "grad_norm": 1.9003863334655762,
      "learning_rate": 2.469831179394182e-05,
      "loss": 0.3695,
      "step": 3511
    },
    {
      "epoch": 0.7022,
      "loss_align": 0.11545002460479736,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26420170068740845,
      "neg_sim_mean": 0.6783421635627747,
      "pos_sim_mean": 0.8845499753952026,
      "step": 3511
    },
    {
      "epoch": 0.7024,
      "grad_norm": 2.075380802154541,
      "learning_rate": 2.466821055551297e-05,
      "loss": 0.3797,
      "step": 3512
    },
    {
      "epoch": 0.7024,
      "loss_align": 0.09791874885559082,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11029555648565292,
      "neg_sim_mean": 0.7339851260185242,
      "pos_sim_mean": 0.9020812511444092,
      "step": 3512
    },
    {
      "epoch": 0.7026,
      "grad_norm": 1.146673560142517,
      "learning_rate": 2.4638121663490545e-05,
      "loss": 0.2082,
      "step": 3513
    },
    {
      "epoch": 0.7026,
      "loss_align": 0.09703147411346436,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11728544533252716,
      "neg_sim_mean": 0.7902402877807617,
      "pos_sim_mean": 0.9029685258865356,
      "step": 3513
    },
    {
      "epoch": 0.7028,
      "grad_norm": 1.4336096048355103,
      "learning_rate": 2.4608045132539537e-05,
      "loss": 0.2143,
      "step": 3514
    },
    {
      "epoch": 0.7028,
      "loss_align": 0.07672351598739624,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12590263783931732,
      "neg_sim_mean": 0.7513541579246521,
      "pos_sim_mean": 0.9232764840126038,
      "step": 3514
    },
    {
      "epoch": 0.703,
      "grad_norm": 1.676230788230896,
      "learning_rate": 2.4577980977318865e-05,
      "loss": 0.2026,
      "step": 3515
    },
    {
      "epoch": 0.703,
      "loss_align": 0.09320151805877686,
      "loss_contrastive": 0.05016970634460449,
      "loss_gen": 0.4771798253059387,
      "neg_sim_mean": 0.8569681644439697,
      "pos_sim_mean": 0.9067984819412231,
      "step": 3515
    },
    {
      "epoch": 0.7032,
      "grad_norm": 2.279665470123291,
      "learning_rate": 2.4547929212481435e-05,
      "loss": 0.5764,
      "step": 3516
    },
    {
      "epoch": 0.7032,
      "loss_align": 0.10737228393554688,
      "loss_contrastive": 0.05507391691207886,
      "loss_gen": 0.3103698790073395,
      "neg_sim_mean": 0.8477016091346741,
      "pos_sim_mean": 0.8926277160644531,
      "step": 3516
    },
    {
      "epoch": 0.7034,
      "grad_norm": 2.2426066398620605,
      "learning_rate": 2.4517889852674113e-05,
      "loss": 0.4244,
      "step": 3517
    },
    {
      "epoch": 0.7034,
      "loss_align": 0.06115293502807617,
      "loss_contrastive": 0.03078216314315796,
      "loss_gen": 0.2771633565425873,
      "neg_sim_mean": 0.8696292042732239,
      "pos_sim_mean": 0.9388470649719238,
      "step": 3517
    },
    {
      "epoch": 0.7036,
      "grad_norm": 1.9623875617980957,
      "learning_rate": 2.4487862912537718e-05,
      "loss": 0.342,
      "step": 3518
    },
    {
      "epoch": 0.7036,
      "loss_align": 0.09012317657470703,
      "loss_contrastive": 0.03437602519989014,
      "loss_gen": 0.13254551589488983,
      "neg_sim_mean": 0.8442528247833252,
      "pos_sim_mean": 0.909876823425293,
      "step": 3518
    },
    {
      "epoch": 0.7038,
      "grad_norm": 1.4510118961334229,
      "learning_rate": 2.4457848406707013e-05,
      "loss": 0.2268,
      "step": 3519
    },
    {
      "epoch": 0.7038,
      "loss_align": 0.10045915842056274,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15227968990802765,
      "neg_sim_mean": 0.5665785074234009,
      "pos_sim_mean": 0.8995408415794373,
      "step": 3519
    },
    {
      "epoch": 0.704,
      "grad_norm": 1.475974678993225,
      "learning_rate": 2.442784634981071e-05,
      "loss": 0.2527,
      "step": 3520
    },
    {
      "epoch": 0.704,
      "loss_align": 0.11620038747787476,
      "loss_contrastive": 0.0,
      "loss_gen": 0.35597068071365356,
      "neg_sim_mean": 0.7601172924041748,
      "pos_sim_mean": 0.8837996125221252,
      "step": 3520
    },
    {
      "epoch": 0.7042,
      "grad_norm": 2.2620415687561035,
      "learning_rate": 2.4397856756471432e-05,
      "loss": 0.4722,
      "step": 3521
    },
    {
      "epoch": 0.7042,
      "loss_align": 0.14567041397094727,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2677718698978424,
      "neg_sim_mean": 0.7187625169754028,
      "pos_sim_mean": 0.8543295860290527,
      "step": 3521
    },
    {
      "epoch": 0.7044,
      "grad_norm": 1.904251217842102,
      "learning_rate": 2.4367879641305758e-05,
      "loss": 0.4134,
      "step": 3522
    },
    {
      "epoch": 0.7044,
      "loss_align": 0.12725555896759033,
      "loss_contrastive": 0.06757402420043945,
      "loss_gen": 0.19190546870231628,
      "neg_sim_mean": 0.8403184413909912,
      "pos_sim_mean": 0.8727444410324097,
      "step": 3522
    },
    {
      "epoch": 0.7046,
      "grad_norm": 1.505159854888916,
      "learning_rate": 2.4337915018924146e-05,
      "loss": 0.3273,
      "step": 3523
    },
    {
      "epoch": 0.7046,
      "loss_align": 0.08954095840454102,
      "loss_contrastive": 0.011023163795471191,
      "loss_gen": 0.28545594215393066,
      "neg_sim_mean": 0.8214821815490723,
      "pos_sim_mean": 0.910459041595459,
      "step": 3523
    },
    {
      "epoch": 0.7048,
      "grad_norm": 1.9422783851623535,
      "learning_rate": 2.4307962903931025e-05,
      "loss": 0.3763,
      "step": 3524
    },
    {
      "epoch": 0.7048,
      "loss_align": 0.10818654298782349,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028615368530154228,
      "neg_sim_mean": 0.3455737829208374,
      "pos_sim_mean": 0.8918134570121765,
      "step": 3524
    },
    {
      "epoch": 0.705,
      "grad_norm": 1.0435189008712769,
      "learning_rate": 2.4278023310924673e-05,
      "loss": 0.1368,
      "step": 3525
    },
    {
      "epoch": 0.705,
      "loss_align": 0.1097944974899292,
      "loss_contrastive": 0.11368995904922485,
      "loss_gen": 0.2153102308511734,
      "neg_sim_mean": 0.9038954377174377,
      "pos_sim_mean": 0.8902055025100708,
      "step": 3525
    },
    {
      "epoch": 0.7052,
      "grad_norm": 1.7062500715255737,
      "learning_rate": 2.4248096254497288e-05,
      "loss": 0.3387,
      "step": 3526
    },
    {
      "epoch": 0.7052,
      "loss_align": 0.09321677684783936,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2288418710231781,
      "neg_sim_mean": 0.6598719954490662,
      "pos_sim_mean": 0.9067832231521606,
      "step": 3526
    },
    {
      "epoch": 0.7054,
      "grad_norm": 1.8196755647659302,
      "learning_rate": 2.4218181749234953e-05,
      "loss": 0.3221,
      "step": 3527
    },
    {
      "epoch": 0.7054,
      "loss_align": 0.12134993076324463,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007005624007433653,
      "neg_sim_mean": 0.4522576928138733,
      "pos_sim_mean": 0.8786500692367554,
      "step": 3527
    },
    {
      "epoch": 0.7056,
      "grad_norm": 0.6761770844459534,
      "learning_rate": 2.418827980971763e-05,
      "loss": 0.1284,
      "step": 3528
    },
    {
      "epoch": 0.7056,
      "loss_align": 0.09285527467727661,
      "loss_contrastive": 0.0,
      "loss_gen": 0.021181395277380943,
      "neg_sim_mean": 0.47486168146133423,
      "pos_sim_mean": 0.9071447253227234,
      "step": 3528
    },
    {
      "epoch": 0.7058,
      "grad_norm": 0.910896360874176,
      "learning_rate": 2.4158390450519157e-05,
      "loss": 0.114,
      "step": 3529
    },
    {
      "epoch": 0.7058,
      "loss_align": 0.1153188943862915,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028673343360424042,
      "neg_sim_mean": 0.3505420684814453,
      "pos_sim_mean": 0.8846811056137085,
      "step": 3529
    },
    {
      "epoch": 0.706,
      "grad_norm": 1.0430293083190918,
      "learning_rate": 2.412851368620726e-05,
      "loss": 0.144,
      "step": 3530
    },
    {
      "epoch": 0.706,
      "loss_align": 0.06047248840332031,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21229735016822815,
      "neg_sim_mean": 0.8152914047241211,
      "pos_sim_mean": 0.9395275115966797,
      "step": 3530
    },
    {
      "epoch": 0.7062,
      "grad_norm": 1.8815473318099976,
      "learning_rate": 2.4098649531343497e-05,
      "loss": 0.2728,
      "step": 3531
    },
    {
      "epoch": 0.7062,
      "loss_align": 0.09828341007232666,
      "loss_contrastive": 0.09374284744262695,
      "loss_gen": 0.09898234158754349,
      "neg_sim_mean": 0.8954594135284424,
      "pos_sim_mean": 0.9017165899276733,
      "step": 3531
    },
    {
      "epoch": 0.7064,
      "grad_norm": 1.381798267364502,
      "learning_rate": 2.4068798000483307e-05,
      "loss": 0.2085,
      "step": 3532
    },
    {
      "epoch": 0.7064,
      "loss_align": 0.054912269115448,
      "loss_contrastive": 0.0,
      "loss_gen": 0.39709973335266113,
      "neg_sim_mean": 0.6527049541473389,
      "pos_sim_mean": 0.945087730884552,
      "step": 3532
    },
    {
      "epoch": 0.7066,
      "grad_norm": 2.009767532348633,
      "learning_rate": 2.403895910817593e-05,
      "loss": 0.452,
      "step": 3533
    },
    {
      "epoch": 0.7066,
      "loss_align": 0.17698752880096436,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04095235466957092,
      "neg_sim_mean": 0.3340006172657013,
      "pos_sim_mean": 0.8230124711990356,
      "step": 3533
    },
    {
      "epoch": 0.7068,
      "grad_norm": 1.383782982826233,
      "learning_rate": 2.4009132868964522e-05,
      "loss": 0.2179,
      "step": 3534
    },
    {
      "epoch": 0.7068,
      "loss_align": 0.09675592184066772,
      "loss_contrastive": 0.013151943683624268,
      "loss_gen": 0.13363829255104065,
      "neg_sim_mean": 0.8163959980010986,
      "pos_sim_mean": 0.9032440781593323,
      "step": 3534
    },
    {
      "epoch": 0.707,
      "grad_norm": 1.412352442741394,
      "learning_rate": 2.3979319297386032e-05,
      "loss": 0.232,
      "step": 3535
    },
    {
      "epoch": 0.707,
      "loss_align": 0.11685991287231445,
      "loss_contrastive": 0.06336146593093872,
      "loss_gen": 0.14630644023418427,
      "neg_sim_mean": 0.8465015292167664,
      "pos_sim_mean": 0.8831400871276855,
      "step": 3535
    },
    {
      "epoch": 0.7072,
      "grad_norm": 1.3263437747955322,
      "learning_rate": 2.39495184079712e-05,
      "loss": 0.2708,
      "step": 3536
    },
    {
      "epoch": 0.7072,
      "loss_align": 0.08682143688201904,
      "loss_contrastive": 0.07608187198638916,
      "loss_gen": 0.15005634725093842,
      "neg_sim_mean": 0.8892604112625122,
      "pos_sim_mean": 0.913178563117981,
      "step": 3536
    },
    {
      "epoch": 0.7074,
      "grad_norm": 1.6812636852264404,
      "learning_rate": 2.3919730215244614e-05,
      "loss": 0.246,
      "step": 3537
    },
    {
      "epoch": 0.7074,
      "loss_align": 0.1903247833251953,
      "loss_contrastive": 0.038634300231933594,
      "loss_gen": 0.22747235000133514,
      "neg_sim_mean": 0.7483094930648804,
      "pos_sim_mean": 0.8096752166748047,
      "step": 3537
    },
    {
      "epoch": 0.7076,
      "grad_norm": 1.8176517486572266,
      "learning_rate": 2.3889954733724707e-05,
      "loss": 0.4224,
      "step": 3538
    },
    {
      "epoch": 0.7076,
      "loss_align": 0.16297411918640137,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18693608045578003,
      "neg_sim_mean": 0.677804172039032,
      "pos_sim_mean": 0.8370258808135986,
      "step": 3538
    },
    {
      "epoch": 0.7078,
      "grad_norm": 1.769405484199524,
      "learning_rate": 2.3860191977923672e-05,
      "loss": 0.3499,
      "step": 3539
    },
    {
      "epoch": 0.7078,
      "loss_align": 0.10900306701660156,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03169795498251915,
      "neg_sim_mean": 0.31524553894996643,
      "pos_sim_mean": 0.8909969329833984,
      "step": 3539
    },
    {
      "epoch": 0.708,
      "grad_norm": 1.1322251558303833,
      "learning_rate": 2.383044196234753e-05,
      "loss": 0.1407,
      "step": 3540
    },
    {
      "epoch": 0.708,
      "loss_align": 0.2093803882598877,
      "loss_contrastive": 0.07647484540939331,
      "loss_gen": 0.17471817135810852,
      "neg_sim_mean": 0.7670944333076477,
      "pos_sim_mean": 0.7906196117401123,
      "step": 3540
    },
    {
      "epoch": 0.7082,
      "grad_norm": 2.1463778018951416,
      "learning_rate": 2.3800704701496053e-05,
      "loss": 0.3933,
      "step": 3541
    },
    {
      "epoch": 0.7082,
      "loss_align": 0.10592693090438843,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0019292758079245687,
      "neg_sim_mean": 0.27943432331085205,
      "pos_sim_mean": 0.8940730690956116,
      "step": 3541
    },
    {
      "epoch": 0.7084,
      "grad_norm": 0.42556506395339966,
      "learning_rate": 2.3770980209862813e-05,
      "loss": 0.1079,
      "step": 3542
    },
    {
      "epoch": 0.7084,
      "loss_align": 0.12028670310974121,
      "loss_contrastive": 0.021781206130981445,
      "loss_gen": 0.24845047295093536,
      "neg_sim_mean": 0.8014944791793823,
      "pos_sim_mean": 0.8797132968902588,
      "step": 3542
    },
    {
      "epoch": 0.7086,
      "grad_norm": 1.6513351202011108,
      "learning_rate": 2.374126850193521e-05,
      "loss": 0.3714,
      "step": 3543
    },
    {
      "epoch": 0.7086,
      "loss_align": 0.10278761386871338,
      "loss_contrastive": 0.042168378829956055,
      "loss_gen": 0.22963298857212067,
      "neg_sim_mean": 0.8393807411193848,
      "pos_sim_mean": 0.8972123861312866,
      "step": 3543
    },
    {
      "epoch": 0.7088,
      "grad_norm": 1.9452935457229614,
      "learning_rate": 2.371156959219436e-05,
      "loss": 0.3375,
      "step": 3544
    },
    {
      "epoch": 0.7088,
      "loss_align": 0.21213680505752563,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011231115087866783,
      "neg_sim_mean": 0.26201993227005005,
      "pos_sim_mean": 0.7878631949424744,
      "step": 3544
    },
    {
      "epoch": 0.709,
      "grad_norm": 0.8505608439445496,
      "learning_rate": 2.3681883495115116e-05,
      "loss": 0.2234,
      "step": 3545
    },
    {
      "epoch": 0.709,
      "loss_align": 0.08040803670883179,
      "loss_contrastive": 0.0809057354927063,
      "loss_gen": 0.28423479199409485,
      "neg_sim_mean": 0.9004976749420166,
      "pos_sim_mean": 0.9195919632911682,
      "step": 3545
    },
    {
      "epoch": 0.7092,
      "grad_norm": 1.6855465173721313,
      "learning_rate": 2.3652210225166122e-05,
      "loss": 0.3744,
      "step": 3546
    },
    {
      "epoch": 0.7092,
      "loss_align": 0.0915604829788208,
      "loss_contrastive": 0.09195661544799805,
      "loss_gen": 0.37749889492988586,
      "neg_sim_mean": 0.9003961086273193,
      "pos_sim_mean": 0.9084395170211792,
      "step": 3546
    },
    {
      "epoch": 0.7094,
      "grad_norm": 2.010530948638916,
      "learning_rate": 2.3622549796809806e-05,
      "loss": 0.4801,
      "step": 3547
    },
    {
      "epoch": 0.7094,
      "loss_align": 0.08014720678329468,
      "loss_contrastive": 0.0017529726028442383,
      "loss_gen": 0.15162444114685059,
      "neg_sim_mean": 0.8216057419776917,
      "pos_sim_mean": 0.9198527932167053,
      "step": 3547
    },
    {
      "epoch": 0.7096,
      "grad_norm": 1.4732388257980347,
      "learning_rate": 2.3592902224502282e-05,
      "loss": 0.232,
      "step": 3548
    },
    {
      "epoch": 0.7096,
      "loss_align": 0.228690505027771,
      "loss_contrastive": 0.05595254898071289,
      "loss_gen": 0.16723069548606873,
      "neg_sim_mean": 0.727262020111084,
      "pos_sim_mean": 0.771309494972229,
      "step": 3548
    },
    {
      "epoch": 0.7098,
      "grad_norm": 1.6742557287216187,
      "learning_rate": 2.3563267522693415e-05,
      "loss": 0.4026,
      "step": 3549
    },
    {
      "epoch": 0.7098,
      "loss_align": 0.09535646438598633,
      "loss_contrastive": 0.0,
      "loss_gen": 0.016953611746430397,
      "neg_sim_mean": 0.3102656602859497,
      "pos_sim_mean": 0.9046435356140137,
      "step": 3549
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.7809540629386902,
      "learning_rate": 2.3533645705826806e-05,
      "loss": 0.1123,
      "step": 3550
    },
    {
      "epoch": 0.71,
      "loss_align": 0.04757273197174072,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007017249241471291,
      "neg_sim_mean": 0.35978004336357117,
      "pos_sim_mean": 0.9524272680282593,
      "step": 3550
    },
    {
      "epoch": 0.7102,
      "grad_norm": 0.4699353277683258,
      "learning_rate": 2.350403678833976e-05,
      "loss": 0.0546,
      "step": 3551
    },
    {
      "epoch": 0.7102,
      "loss_align": 0.10629314184188843,
      "loss_contrastive": 0.03563416004180908,
      "loss_gen": 0.05427803844213486,
      "neg_sim_mean": 0.8293409943580627,
      "pos_sim_mean": 0.8937068581581116,
      "step": 3551
    },
    {
      "epoch": 0.7104,
      "grad_norm": 1.0477933883666992,
      "learning_rate": 2.347444078466329e-05,
      "loss": 0.1648,
      "step": 3552
    },
    {
      "epoch": 0.7104,
      "loss_align": 0.12715530395507812,
      "loss_contrastive": 0.06037569046020508,
      "loss_gen": 0.2196081131696701,
      "neg_sim_mean": 0.833220362663269,
      "pos_sim_mean": 0.8728446960449219,
      "step": 3552
    },
    {
      "epoch": 0.7106,
      "grad_norm": 1.557986855506897,
      "learning_rate": 2.344485770922218e-05,
      "loss": 0.354,
      "step": 3553
    },
    {
      "epoch": 0.7106,
      "loss_align": 0.07652515172958374,
      "loss_contrastive": 0.039816975593566895,
      "loss_gen": 0.10013417154550552,
      "neg_sim_mean": 0.8632918000221252,
      "pos_sim_mean": 0.9234748482704163,
      "step": 3553
    },
    {
      "epoch": 0.7108,
      "grad_norm": 1.214355230331421,
      "learning_rate": 2.3415287576434808e-05,
      "loss": 0.1814,
      "step": 3554
    },
    {
      "epoch": 0.7108,
      "loss_align": 0.17846602201461792,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18100914359092712,
      "neg_sim_mean": 0.5578665137290955,
      "pos_sim_mean": 0.8215339779853821,
      "step": 3554
    },
    {
      "epoch": 0.711,
      "grad_norm": 1.854306936264038,
      "learning_rate": 2.3385730400713318e-05,
      "loss": 0.3595,
      "step": 3555
    },
    {
      "epoch": 0.711,
      "loss_align": 0.09131991863250732,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002521851332858205,
      "neg_sim_mean": 0.3391804099082947,
      "pos_sim_mean": 0.9086800813674927,
      "step": 3555
    },
    {
      "epoch": 0.7112,
      "grad_norm": 0.3626093566417694,
      "learning_rate": 2.33561861964635e-05,
      "loss": 0.0938,
      "step": 3556
    },
    {
      "epoch": 0.7112,
      "loss_align": 0.08100420236587524,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20575791597366333,
      "neg_sim_mean": 0.48389846086502075,
      "pos_sim_mean": 0.9189957976341248,
      "step": 3556
    },
    {
      "epoch": 0.7114,
      "grad_norm": 1.4470564126968384,
      "learning_rate": 2.332665497808487e-05,
      "loss": 0.2868,
      "step": 3557
    },
    {
      "epoch": 0.7114,
      "loss_align": 0.1262950897216797,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0017188789788633585,
      "neg_sim_mean": 0.3221183717250824,
      "pos_sim_mean": 0.8737049102783203,
      "step": 3557
    },
    {
      "epoch": 0.7116,
      "grad_norm": 0.46165382862091064,
      "learning_rate": 2.3297136759970577e-05,
      "loss": 0.128,
      "step": 3558
    },
    {
      "epoch": 0.7116,
      "loss_align": 0.09597325325012207,
      "loss_contrastive": 0.11611151695251465,
      "loss_gen": 0.3558970093727112,
      "neg_sim_mean": 0.9201382398605347,
      "pos_sim_mean": 0.9040267467498779,
      "step": 3558
    },
    {
      "epoch": 0.7118,
      "grad_norm": 2.640052080154419,
      "learning_rate": 2.326763155650744e-05,
      "loss": 0.4658,
      "step": 3559
    },
    {
      "epoch": 0.7118,
      "loss_align": 0.12987804412841797,
      "loss_contrastive": 0.11121511459350586,
      "loss_gen": 0.26987430453300476,
      "neg_sim_mean": 0.88133704662323,
      "pos_sim_mean": 0.870121955871582,
      "step": 3559
    },
    {
      "epoch": 0.712,
      "grad_norm": 2.3266842365264893,
      "learning_rate": 2.323813938207593e-05,
      "loss": 0.4131,
      "step": 3560
    },
    {
      "epoch": 0.712,
      "loss_align": 0.09305310249328613,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10963944345712662,
      "neg_sim_mean": 0.7876068353652954,
      "pos_sim_mean": 0.9069468975067139,
      "step": 3560
    },
    {
      "epoch": 0.7122,
      "grad_norm": 1.4364311695098877,
      "learning_rate": 2.3208660251050158e-05,
      "loss": 0.2027,
      "step": 3561
    },
    {
      "epoch": 0.7122,
      "loss_align": 0.14767348766326904,
      "loss_contrastive": 0.08052980899810791,
      "loss_gen": 0.18985334038734436,
      "neg_sim_mean": 0.832856297492981,
      "pos_sim_mean": 0.852326512336731,
      "step": 3561
    },
    {
      "epoch": 0.7124,
      "grad_norm": 1.9970667362213135,
      "learning_rate": 2.3179194177797952e-05,
      "loss": 0.3472,
      "step": 3562
    },
    {
      "epoch": 0.7124,
      "loss_align": 0.0792609453201294,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22914163768291473,
      "neg_sim_mean": 0.7222909927368164,
      "pos_sim_mean": 0.9207390546798706,
      "step": 3562
    },
    {
      "epoch": 0.7126,
      "grad_norm": 1.9158588647842407,
      "learning_rate": 2.3149741176680666e-05,
      "loss": 0.3084,
      "step": 3563
    },
    {
      "epoch": 0.7126,
      "loss_align": 0.13298219442367554,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29943788051605225,
      "neg_sim_mean": 0.5563496351242065,
      "pos_sim_mean": 0.8670178055763245,
      "step": 3563
    },
    {
      "epoch": 0.7128,
      "grad_norm": 2.007781505584717,
      "learning_rate": 2.312030126205335e-05,
      "loss": 0.4324,
      "step": 3564
    },
    {
      "epoch": 0.7128,
      "loss_align": 0.13587355613708496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14681386947631836,
      "neg_sim_mean": 0.7047923803329468,
      "pos_sim_mean": 0.864126443862915,
      "step": 3564
    },
    {
      "epoch": 0.713,
      "grad_norm": 1.722487211227417,
      "learning_rate": 2.3090874448264637e-05,
      "loss": 0.2827,
      "step": 3565
    },
    {
      "epoch": 0.713,
      "loss_align": 0.1871623992919922,
      "loss_contrastive": 0.04319411516189575,
      "loss_gen": 0.18579448759555817,
      "neg_sim_mean": 0.7560316920280457,
      "pos_sim_mean": 0.8128376007080078,
      "step": 3565
    },
    {
      "epoch": 0.7132,
      "grad_norm": 1.7754693031311035,
      "learning_rate": 2.3061460749656844e-05,
      "loss": 0.3781,
      "step": 3566
    },
    {
      "epoch": 0.7132,
      "loss_align": 0.17146623134613037,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0006932411342859268,
      "neg_sim_mean": 0.3858552873134613,
      "pos_sim_mean": 0.8285337686538696,
      "step": 3566
    },
    {
      "epoch": 0.7134,
      "grad_norm": 0.7323982119560242,
      "learning_rate": 2.3032060180565828e-05,
      "loss": 0.1722,
      "step": 3567
    },
    {
      "epoch": 0.7134,
      "loss_align": 0.11365127563476562,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2704242169857025,
      "neg_sim_mean": 0.6722453236579895,
      "pos_sim_mean": 0.8863487243652344,
      "step": 3567
    },
    {
      "epoch": 0.7136,
      "grad_norm": 1.7653813362121582,
      "learning_rate": 2.3002672755321074e-05,
      "loss": 0.3841,
      "step": 3568
    },
    {
      "epoch": 0.7136,
      "loss_align": 0.1170114278793335,
      "loss_contrastive": 0.09161925315856934,
      "loss_gen": 0.24927771091461182,
      "neg_sim_mean": 0.8746078014373779,
      "pos_sim_mean": 0.8829885721206665,
      "step": 3568
    },
    {
      "epoch": 0.7138,
      "grad_norm": 1.7709964513778687,
      "learning_rate": 2.2973298488245648e-05,
      "loss": 0.3773,
      "step": 3569
    },
    {
      "epoch": 0.7138,
      "loss_align": 0.10028982162475586,
      "loss_contrastive": 0.07355242967605591,
      "loss_gen": 0.24549910426139832,
      "neg_sim_mean": 0.8732625842094421,
      "pos_sim_mean": 0.8997101783752441,
      "step": 3569
    },
    {
      "epoch": 0.714,
      "grad_norm": 1.86119544506073,
      "learning_rate": 2.294393739365621e-05,
      "loss": 0.3546,
      "step": 3570
    },
    {
      "epoch": 0.714,
      "loss_align": 0.14242023229599,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0043594855815172195,
      "neg_sim_mean": 0.2824220359325409,
      "pos_sim_mean": 0.85757976770401,
      "step": 3570
    },
    {
      "epoch": 0.7142,
      "grad_norm": 0.578569233417511,
      "learning_rate": 2.2914589485863014e-05,
      "loss": 0.1468,
      "step": 3571
    },
    {
      "epoch": 0.7142,
      "loss_align": 0.08179068565368652,
      "loss_contrastive": 0.03944307565689087,
      "loss_gen": 0.4444408416748047,
      "neg_sim_mean": 0.8576523661613464,
      "pos_sim_mean": 0.9182093143463135,
      "step": 3571
    },
    {
      "epoch": 0.7144,
      "grad_norm": 2.4200947284698486,
      "learning_rate": 2.2885254779169864e-05,
      "loss": 0.531,
      "step": 3572
    },
    {
      "epoch": 0.7144,
      "loss_align": 0.15737712383270264,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10027981549501419,
      "neg_sim_mean": 0.7199329137802124,
      "pos_sim_mean": 0.8426228761672974,
      "step": 3572
    },
    {
      "epoch": 0.7146,
      "grad_norm": 1.1704994440078735,
      "learning_rate": 2.2855933287874138e-05,
      "loss": 0.2577,
      "step": 3573
    },
    {
      "epoch": 0.7146,
      "loss_align": 0.154862642288208,
      "loss_contrastive": 0.04050612449645996,
      "loss_gen": 0.14295217394828796,
      "neg_sim_mean": 0.785643458366394,
      "pos_sim_mean": 0.845137357711792,
      "step": 3573
    },
    {
      "epoch": 0.7148,
      "grad_norm": 1.73844313621521,
      "learning_rate": 2.282662502626678e-05,
      "loss": 0.3027,
      "step": 3574
    },
    {
      "epoch": 0.7148,
      "loss_align": 0.11155188083648682,
      "loss_contrastive": 0.05828016996383667,
      "loss_gen": 0.1903184950351715,
      "neg_sim_mean": 0.8467282652854919,
      "pos_sim_mean": 0.8884481191635132,
      "step": 3574
    },
    {
      "epoch": 0.715,
      "grad_norm": 1.5411726236343384,
      "learning_rate": 2.2797330008632257e-05,
      "loss": 0.3089,
      "step": 3575
    },
    {
      "epoch": 0.715,
      "loss_align": 0.03532606363296509,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2969697415828705,
      "neg_sim_mean": 0.8631401658058167,
      "pos_sim_mean": 0.9646739363670349,
      "step": 3575
    },
    {
      "epoch": 0.7152,
      "grad_norm": 2.244131326675415,
      "learning_rate": 2.2768048249248648e-05,
      "loss": 0.3323,
      "step": 3576
    },
    {
      "epoch": 0.7152,
      "loss_align": 0.1253756880760193,
      "loss_contrastive": 0.04518812894821167,
      "loss_gen": 0.20757970213890076,
      "neg_sim_mean": 0.8198124170303345,
      "pos_sim_mean": 0.8746243119239807,
      "step": 3576
    },
    {
      "epoch": 0.7154,
      "grad_norm": 2.4485225677490234,
      "learning_rate": 2.2738779762387497e-05,
      "loss": 0.3384,
      "step": 3577
    },
    {
      "epoch": 0.7154,
      "loss_align": 0.1063385009765625,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01665741577744484,
      "neg_sim_mean": 0.4856472909450531,
      "pos_sim_mean": 0.8936614990234375,
      "step": 3577
    },
    {
      "epoch": 0.7156,
      "grad_norm": 0.816967785358429,
      "learning_rate": 2.270952456231392e-05,
      "loss": 0.123,
      "step": 3578
    },
    {
      "epoch": 0.7156,
      "loss_align": 0.2884284257888794,
      "loss_contrastive": 0.05709344148635864,
      "loss_gen": 0.19122719764709473,
      "neg_sim_mean": 0.6686649918556213,
      "pos_sim_mean": 0.7115715742111206,
      "step": 3578
    },
    {
      "epoch": 0.7158,
      "grad_norm": 1.9891623258590698,
      "learning_rate": 2.2680282663286552e-05,
      "loss": 0.4865,
      "step": 3579
    },
    {
      "epoch": 0.7158,
      "loss_align": 0.06642413139343262,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27000662684440613,
      "neg_sim_mean": 0.7623249292373657,
      "pos_sim_mean": 0.9335758686065674,
      "step": 3579
    },
    {
      "epoch": 0.716,
      "grad_norm": 1.7936971187591553,
      "learning_rate": 2.265105407955752e-05,
      "loss": 0.3364,
      "step": 3580
    },
    {
      "epoch": 0.716,
      "loss_align": 0.07975077629089355,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015069348737597466,
      "neg_sim_mean": 0.2158132940530777,
      "pos_sim_mean": 0.9202492237091064,
      "step": 3580
    },
    {
      "epoch": 0.7162,
      "grad_norm": 0.26047971844673157,
      "learning_rate": 2.2621838825372493e-05,
      "loss": 0.0813,
      "step": 3581
    },
    {
      "epoch": 0.7162,
      "loss_align": 0.1046518087387085,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20839807391166687,
      "neg_sim_mean": 0.477374792098999,
      "pos_sim_mean": 0.8953481912612915,
      "step": 3581
    },
    {
      "epoch": 0.7164,
      "grad_norm": 1.9099061489105225,
      "learning_rate": 2.2592636914970634e-05,
      "loss": 0.313,
      "step": 3582
    },
    {
      "epoch": 0.7164,
      "loss_align": 0.063118577003479,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13279248774051666,
      "neg_sim_mean": 0.8264578580856323,
      "pos_sim_mean": 0.936881422996521,
      "step": 3582
    },
    {
      "epoch": 0.7166,
      "grad_norm": 1.62734055519104,
      "learning_rate": 2.256344836258459e-05,
      "loss": 0.1959,
      "step": 3583
    },
    {
      "epoch": 0.7166,
      "loss_align": 0.13216376304626465,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0010534825269132853,
      "neg_sim_mean": 0.3432930111885071,
      "pos_sim_mean": 0.8678362369537354,
      "step": 3583
    },
    {
      "epoch": 0.7168,
      "grad_norm": 0.5202685594558716,
      "learning_rate": 2.2534273182440514e-05,
      "loss": 0.1332,
      "step": 3584
    },
    {
      "epoch": 0.7168,
      "loss_align": 0.15372222661972046,
      "loss_contrastive": 0.09726256132125854,
      "loss_gen": 0.2587911784648895,
      "neg_sim_mean": 0.8435403108596802,
      "pos_sim_mean": 0.8462777733802795,
      "step": 3584
    },
    {
      "epoch": 0.717,
      "grad_norm": 2.063119411468506,
      "learning_rate": 2.250511138875801e-05,
      "loss": 0.4242,
      "step": 3585
    },
    {
      "epoch": 0.717,
      "loss_align": 0.21755588054656982,
      "loss_contrastive": 0.03930705785751343,
      "loss_gen": 0.23562443256378174,
      "neg_sim_mean": 0.7217511534690857,
      "pos_sim_mean": 0.7824441194534302,
      "step": 3585
    },
    {
      "epoch": 0.7172,
      "grad_norm": 1.6711843013763428,
      "learning_rate": 2.247596299575022e-05,
      "loss": 0.4579,
      "step": 3586
    },
    {
      "epoch": 0.7172,
      "loss_align": 0.17892104387283325,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14639998972415924,
      "neg_sim_mean": 0.5465642213821411,
      "pos_sim_mean": 0.8210789561271667,
      "step": 3586
    },
    {
      "epoch": 0.7174,
      "grad_norm": 2.0173633098602295,
      "learning_rate": 2.24468280176237e-05,
      "loss": 0.3253,
      "step": 3587
    },
    {
      "epoch": 0.7174,
      "loss_align": 0.07989829778671265,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0024860030971467495,
      "neg_sim_mean": 0.30646881461143494,
      "pos_sim_mean": 0.9201017022132874,
      "step": 3587
    },
    {
      "epoch": 0.7176,
      "grad_norm": 0.38409069180488586,
      "learning_rate": 2.241770646857849e-05,
      "loss": 0.0824,
      "step": 3588
    },
    {
      "epoch": 0.7176,
      "loss_align": 0.12137037515640259,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2594341039657593,
      "neg_sim_mean": 0.6720607280731201,
      "pos_sim_mean": 0.8786296248435974,
      "step": 3588
    },
    {
      "epoch": 0.7178,
      "grad_norm": 1.8015128374099731,
      "learning_rate": 2.2388598362808073e-05,
      "loss": 0.3808,
      "step": 3589
    },
    {
      "epoch": 0.7178,
      "loss_align": 0.1615215539932251,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0010751368245109916,
      "neg_sim_mean": 0.3548072278499603,
      "pos_sim_mean": 0.8384784460067749,
      "step": 3589
    },
    {
      "epoch": 0.718,
      "grad_norm": 0.5877722501754761,
      "learning_rate": 2.235950371449938e-05,
      "loss": 0.1626,
      "step": 3590
    },
    {
      "epoch": 0.718,
      "loss_align": 0.16502773761749268,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002006073947995901,
      "neg_sim_mean": 0.20915603637695312,
      "pos_sim_mean": 0.8349722623825073,
      "step": 3590
    },
    {
      "epoch": 0.7182,
      "grad_norm": 0.5465683341026306,
      "learning_rate": 2.23304225378328e-05,
      "loss": 0.167,
      "step": 3591
    },
    {
      "epoch": 0.7182,
      "loss_align": 0.13949620723724365,
      "loss_contrastive": 0.028984129428863525,
      "loss_gen": 0.23659130930900574,
      "neg_sim_mean": 0.789487898349762,
      "pos_sim_mean": 0.8605037927627563,
      "step": 3591
    },
    {
      "epoch": 0.7184,
      "grad_norm": 1.8377918004989624,
      "learning_rate": 2.2301354846982146e-05,
      "loss": 0.3796,
      "step": 3592
    },
    {
      "epoch": 0.7184,
      "loss_align": 0.1543712615966797,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18122512102127075,
      "neg_sim_mean": 0.5841653347015381,
      "pos_sim_mean": 0.8456287384033203,
      "step": 3592
    },
    {
      "epoch": 0.7186,
      "grad_norm": 1.5677146911621094,
      "learning_rate": 2.227230065611465e-05,
      "loss": 0.3356,
      "step": 3593
    },
    {
      "epoch": 0.7186,
      "loss_align": 0.11418318748474121,
      "loss_contrastive": 0.054933786392211914,
      "loss_gen": 0.17540360987186432,
      "neg_sim_mean": 0.8407505750656128,
      "pos_sim_mean": 0.8858168125152588,
      "step": 3593
    },
    {
      "epoch": 0.7188,
      "grad_norm": 1.466637372970581,
      "learning_rate": 2.224325997939095e-05,
      "loss": 0.2962,
      "step": 3594
    },
    {
      "epoch": 0.7188,
      "loss_align": 0.09764707088470459,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12765179574489594,
      "neg_sim_mean": 0.4648110866546631,
      "pos_sim_mean": 0.9023529291152954,
      "step": 3594
    },
    {
      "epoch": 0.719,
      "grad_norm": 1.649381160736084,
      "learning_rate": 2.2214232830965172e-05,
      "loss": 0.2253,
      "step": 3595
    },
    {
      "epoch": 0.719,
      "loss_align": 0.06168818473815918,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0018568815430626273,
      "neg_sim_mean": 0.3410798907279968,
      "pos_sim_mean": 0.9383118152618408,
      "step": 3595
    },
    {
      "epoch": 0.7192,
      "grad_norm": 0.3262237012386322,
      "learning_rate": 2.218521922498476e-05,
      "loss": 0.0635,
      "step": 3596
    },
    {
      "epoch": 0.7192,
      "loss_align": 0.15657442808151245,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005136819556355476,
      "neg_sim_mean": 0.29209011793136597,
      "pos_sim_mean": 0.8434255719184875,
      "step": 3596
    },
    {
      "epoch": 0.7194,
      "grad_norm": 0.5676231384277344,
      "learning_rate": 2.215621917559062e-05,
      "loss": 0.1617,
      "step": 3597
    },
    {
      "epoch": 0.7194,
      "loss_align": 0.08594262599945068,
      "loss_contrastive": 0.09129023551940918,
      "loss_gen": 0.20647570490837097,
      "neg_sim_mean": 0.9053475856781006,
      "pos_sim_mean": 0.9140573740005493,
      "step": 3597
    },
    {
      "epoch": 0.7196,
      "grad_norm": 1.6652977466583252,
      "learning_rate": 2.2127232696916965e-05,
      "loss": 0.3034,
      "step": 3598
    },
    {
      "epoch": 0.7196,
      "loss_align": 0.0716736912727356,
      "loss_contrastive": 0.07857352495193481,
      "loss_gen": 0.23707862198352814,
      "neg_sim_mean": 0.9068998098373413,
      "pos_sim_mean": 0.9283263087272644,
      "step": 3598
    },
    {
      "epoch": 0.7198,
      "grad_norm": 1.9084863662719727,
      "learning_rate": 2.209825980309151e-05,
      "loss": 0.3182,
      "step": 3599
    },
    {
      "epoch": 0.7198,
      "loss_align": 0.14030134677886963,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00890621542930603,
      "neg_sim_mean": 0.37128227949142456,
      "pos_sim_mean": 0.8596986532211304,
      "step": 3599
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.7125905752182007,
      "learning_rate": 2.2069300508235275e-05,
      "loss": 0.1492,
      "step": 3600
    },
    {
      "epoch": 0.72,
      "loss_align": 0.15434187650680542,
      "loss_contrastive": 0.0,
      "loss_gen": 0.020596126094460487,
      "neg_sim_mean": 0.33842334151268005,
      "pos_sim_mean": 0.8456581234931946,
      "step": 3600
    },
    {
      "epoch": 0.7202,
      "grad_norm": 1.1228687763214111,
      "learning_rate": 2.2040354826462668e-05,
      "loss": 0.1749,
      "step": 3601
    },
    {
      "epoch": 0.7202,
      "loss_align": 0.16380596160888672,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19779038429260254,
      "neg_sim_mean": 0.5788715481758118,
      "pos_sim_mean": 0.8361940383911133,
      "step": 3601
    },
    {
      "epoch": 0.7204,
      "grad_norm": 2.1120710372924805,
      "learning_rate": 2.2011422771881458e-05,
      "loss": 0.3616,
      "step": 3602
    },
    {
      "epoch": 0.7204,
      "loss_align": 0.14243686199188232,
      "loss_contrastive": 0.08564209938049316,
      "loss_gen": 0.32999661564826965,
      "neg_sim_mean": 0.8432052135467529,
      "pos_sim_mean": 0.8575631380081177,
      "step": 3602
    },
    {
      "epoch": 0.7206,
      "grad_norm": 1.9635306596755981,
      "learning_rate": 2.1982504358592776e-05,
      "loss": 0.4827,
      "step": 3603
    },
    {
      "epoch": 0.7206,
      "loss_align": 0.08550846576690674,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1915510594844818,
      "neg_sim_mean": 0.601833701133728,
      "pos_sim_mean": 0.9144915342330933,
      "step": 3603
    },
    {
      "epoch": 0.7208,
      "grad_norm": 1.5429534912109375,
      "learning_rate": 2.19535996006911e-05,
      "loss": 0.2771,
      "step": 3604
    },
    {
      "epoch": 0.7208,
      "loss_align": 0.1394432783126831,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23364847898483276,
      "neg_sim_mean": 0.5846390724182129,
      "pos_sim_mean": 0.8605567216873169,
      "step": 3604
    },
    {
      "epoch": 0.721,
      "grad_norm": 1.9477213621139526,
      "learning_rate": 2.1924708512264282e-05,
      "loss": 0.3731,
      "step": 3605
    },
    {
      "epoch": 0.721,
      "loss_align": 0.19578540325164795,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17453059554100037,
      "neg_sim_mean": 0.6589581966400146,
      "pos_sim_mean": 0.804214596748352,
      "step": 3605
    },
    {
      "epoch": 0.7212,
      "grad_norm": 1.61221444606781,
      "learning_rate": 2.1895831107393484e-05,
      "loss": 0.3703,
      "step": 3606
    },
    {
      "epoch": 0.7212,
      "loss_align": 0.17922884225845337,
      "loss_contrastive": 0.013848841190338135,
      "loss_gen": 0.15243422985076904,
      "neg_sim_mean": 0.7346199750900269,
      "pos_sim_mean": 0.8207711577415466,
      "step": 3606
    },
    {
      "epoch": 0.7214,
      "grad_norm": 1.8153256177902222,
      "learning_rate": 2.1866967400153183e-05,
      "loss": 0.3333,
      "step": 3607
    },
    {
      "epoch": 0.7214,
      "loss_align": 0.2203471064567566,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19611604511737823,
      "neg_sim_mean": 0.6688362956047058,
      "pos_sim_mean": 0.7796528935432434,
      "step": 3607
    },
    {
      "epoch": 0.7216,
      "grad_norm": 1.7718852758407593,
      "learning_rate": 2.183811740461118e-05,
      "loss": 0.4165,
      "step": 3608
    },
    {
      "epoch": 0.7216,
      "loss_align": 0.07912600040435791,
      "loss_contrastive": 0.045577406883239746,
      "loss_gen": 0.2372353971004486,
      "neg_sim_mean": 0.8664513826370239,
      "pos_sim_mean": 0.9208739995956421,
      "step": 3608
    },
    {
      "epoch": 0.7218,
      "grad_norm": 1.9675153493881226,
      "learning_rate": 2.1809281134828664e-05,
      "loss": 0.3218,
      "step": 3609
    },
    {
      "epoch": 0.7218,
      "loss_align": 0.12642526626586914,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013372446410357952,
      "neg_sim_mean": 0.319888174533844,
      "pos_sim_mean": 0.8735747337341309,
      "step": 3609
    },
    {
      "epoch": 0.722,
      "grad_norm": 0.7980465292930603,
      "learning_rate": 2.1780458604860056e-05,
      "loss": 0.1398,
      "step": 3610
    },
    {
      "epoch": 0.722,
      "loss_align": 0.09605085849761963,
      "loss_contrastive": 0.0,
      "loss_gen": 0.30668753385543823,
      "neg_sim_mean": 0.6168733835220337,
      "pos_sim_mean": 0.9039491415023804,
      "step": 3610
    },
    {
      "epoch": 0.7222,
      "grad_norm": 1.7694547176361084,
      "learning_rate": 2.1751649828753106e-05,
      "loss": 0.4027,
      "step": 3611
    },
    {
      "epoch": 0.7222,
      "loss_align": 0.14097094535827637,
      "loss_contrastive": 0.11357879638671875,
      "loss_gen": 0.10347964614629745,
      "neg_sim_mean": 0.8726078271865845,
      "pos_sim_mean": 0.8590290546417236,
      "step": 3611
    },
    {
      "epoch": 0.7224,
      "grad_norm": 1.177154302597046,
      "learning_rate": 2.1722854820548872e-05,
      "loss": 0.2581,
      "step": 3612
    },
    {
      "epoch": 0.7224,
      "loss_align": 0.06720101833343506,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16359484195709229,
      "neg_sim_mean": 0.7328272461891174,
      "pos_sim_mean": 0.9327989816665649,
      "step": 3612
    },
    {
      "epoch": 0.7226,
      "grad_norm": 1.5585671663284302,
      "learning_rate": 2.1694073594281662e-05,
      "loss": 0.2308,
      "step": 3613
    },
    {
      "epoch": 0.7226,
      "loss_align": 0.23726820945739746,
      "loss_contrastive": 0.08273321390151978,
      "loss_gen": 0.09816528856754303,
      "neg_sim_mean": 0.7454649806022644,
      "pos_sim_mean": 0.7627317905426025,
      "step": 3613
    },
    {
      "epoch": 0.7228,
      "grad_norm": 1.502880573272705,
      "learning_rate": 2.166530616397913e-05,
      "loss": 0.3454,
      "step": 3614
    },
    {
      "epoch": 0.7228,
      "loss_align": 0.09875667095184326,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008555738255381584,
      "neg_sim_mean": 0.3505643308162689,
      "pos_sim_mean": 0.9012433290481567,
      "step": 3614
    },
    {
      "epoch": 0.723,
      "grad_norm": 0.635529100894928,
      "learning_rate": 2.1636552543662186e-05,
      "loss": 0.1073,
      "step": 3615
    },
    {
      "epoch": 0.723,
      "loss_align": 0.13356554508209229,
      "loss_contrastive": 0.06613337993621826,
      "loss_gen": 0.13941025733947754,
      "neg_sim_mean": 0.8325678110122681,
      "pos_sim_mean": 0.8664344549179077,
      "step": 3615
    },
    {
      "epoch": 0.7232,
      "grad_norm": 1.2629450559616089,
      "learning_rate": 2.160781274734495e-05,
      "loss": 0.2809,
      "step": 3616
    },
    {
      "epoch": 0.7232,
      "loss_align": 0.09259068965911865,
      "loss_contrastive": 0.008993864059448242,
      "loss_gen": 0.5221163630485535,
      "neg_sim_mean": 0.8164031505584717,
      "pos_sim_mean": 0.9074093103408813,
      "step": 3616
    },
    {
      "epoch": 0.7234,
      "grad_norm": 2.350389242172241,
      "learning_rate": 2.157908678903487e-05,
      "loss": 0.6158,
      "step": 3617
    },
    {
      "epoch": 0.7234,
      "loss_align": 0.08100152015686035,
      "loss_contrastive": 0.029970765113830566,
      "loss_gen": 0.1611129492521286,
      "neg_sim_mean": 0.8489692211151123,
      "pos_sim_mean": 0.9189984798431396,
      "step": 3617
    },
    {
      "epoch": 0.7236,
      "grad_norm": 1.488892674446106,
      "learning_rate": 2.1550374682732606e-05,
      "loss": 0.2457,
      "step": 3618
    },
    {
      "epoch": 0.7236,
      "loss_align": 0.16013747453689575,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09966415166854858,
      "neg_sim_mean": 0.44601351022720337,
      "pos_sim_mean": 0.8398625254631042,
      "step": 3618
    },
    {
      "epoch": 0.7238,
      "grad_norm": 1.3733181953430176,
      "learning_rate": 2.152167644243213e-05,
      "loss": 0.2598,
      "step": 3619
    },
    {
      "epoch": 0.7238,
      "loss_align": 0.07587748765945435,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19536854326725006,
      "neg_sim_mean": 0.751369833946228,
      "pos_sim_mean": 0.9241225123405457,
      "step": 3619
    },
    {
      "epoch": 0.724,
      "grad_norm": 1.7699761390686035,
      "learning_rate": 2.14929920821206e-05,
      "loss": 0.2712,
      "step": 3620
    },
    {
      "epoch": 0.724,
      "loss_align": 0.11589998006820679,
      "loss_contrastive": 0.02397829294204712,
      "loss_gen": 0.2578277289867401,
      "neg_sim_mean": 0.8080782890319824,
      "pos_sim_mean": 0.8841000199317932,
      "step": 3620
    },
    {
      "epoch": 0.7242,
      "grad_norm": 2.1209161281585693,
      "learning_rate": 2.1464321615778422e-05,
      "loss": 0.3766,
      "step": 3621
    },
    {
      "epoch": 0.7242,
      "loss_align": 0.22943371534347534,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15419313311576843,
      "neg_sim_mean": 0.6171053647994995,
      "pos_sim_mean": 0.7705662846565247,
      "step": 3621
    },
    {
      "epoch": 0.7244,
      "grad_norm": 1.6457935571670532,
      "learning_rate": 2.1435665057379235e-05,
      "loss": 0.3836,
      "step": 3622
    },
    {
      "epoch": 0.7244,
      "loss_align": 0.08249807357788086,
      "loss_contrastive": 0.06244945526123047,
      "loss_gen": 0.33794718980789185,
      "neg_sim_mean": 0.8799513578414917,
      "pos_sim_mean": 0.9175019264221191,
      "step": 3622
    },
    {
      "epoch": 0.7246,
      "grad_norm": 1.9831849336624146,
      "learning_rate": 2.1407022420889872e-05,
      "loss": 0.4279,
      "step": 3623
    },
    {
      "epoch": 0.7246,
      "loss_align": 0.20229995250701904,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20589697360992432,
      "neg_sim_mean": 0.4631405770778656,
      "pos_sim_mean": 0.797700047492981,
      "step": 3623
    },
    {
      "epoch": 0.7248,
      "grad_norm": 2.1014561653137207,
      "learning_rate": 2.1378393720270468e-05,
      "loss": 0.4082,
      "step": 3624
    },
    {
      "epoch": 0.7248,
      "loss_align": 0.09212899208068848,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16356487572193146,
      "neg_sim_mean": 0.7623740434646606,
      "pos_sim_mean": 0.9078710079193115,
      "step": 3624
    },
    {
      "epoch": 0.725,
      "grad_norm": 1.4804809093475342,
      "learning_rate": 2.134977896947425e-05,
      "loss": 0.2557,
      "step": 3625
    },
    {
      "epoch": 0.725,
      "loss_align": 0.07834231853485107,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19541656970977783,
      "neg_sim_mean": 0.5211453437805176,
      "pos_sim_mean": 0.9216576814651489,
      "step": 3625
    },
    {
      "epoch": 0.7252,
      "grad_norm": 1.525750756263733,
      "learning_rate": 2.132117818244771e-05,
      "loss": 0.2738,
      "step": 3626
    },
    {
      "epoch": 0.7252,
      "loss_align": 0.1255049705505371,
      "loss_contrastive": 0.0,
      "loss_gen": 0.131883442401886,
      "neg_sim_mean": 0.7275994420051575,
      "pos_sim_mean": 0.8744950294494629,
      "step": 3626
    },
    {
      "epoch": 0.7254,
      "grad_norm": 1.3064055442810059,
      "learning_rate": 2.1292591373130518e-05,
      "loss": 0.2574,
      "step": 3627
    },
    {
      "epoch": 0.7254,
      "loss_align": 0.0617527961730957,
      "loss_contrastive": 0.05406278371810913,
      "loss_gen": 0.21894024312496185,
      "neg_sim_mean": 0.8923099637031555,
      "pos_sim_mean": 0.9382472038269043,
      "step": 3627
    },
    {
      "epoch": 0.7256,
      "grad_norm": 1.7576874494552612,
      "learning_rate": 2.126401855545556e-05,
      "loss": 0.2872,
      "step": 3628
    },
    {
      "epoch": 0.7256,
      "loss_align": 0.12048351764678955,
      "loss_contrastive": 0.05663275718688965,
      "loss_gen": 0.11377844959497452,
      "neg_sim_mean": 0.8361492156982422,
      "pos_sim_mean": 0.8795164823532104,
      "step": 3628
    },
    {
      "epoch": 0.7258,
      "grad_norm": 1.5182571411132812,
      "learning_rate": 2.1235459743348872e-05,
      "loss": 0.2411,
      "step": 3629
    },
    {
      "epoch": 0.7258,
      "loss_align": 0.100527822971344,
      "loss_contrastive": 0.06850212812423706,
      "loss_gen": 0.2704058289527893,
      "neg_sim_mean": 0.8679742813110352,
      "pos_sim_mean": 0.899472177028656,
      "step": 3629
    },
    {
      "epoch": 0.726,
      "grad_norm": 1.9177602529525757,
      "learning_rate": 2.1206914950729672e-05,
      "loss": 0.3792,
      "step": 3630
    },
    {
      "epoch": 0.726,
      "loss_align": 0.18295705318450928,
      "loss_contrastive": 0.05288594961166382,
      "loss_gen": 0.26428207755088806,
      "neg_sim_mean": 0.7699288725852966,
      "pos_sim_mean": 0.8170429468154907,
      "step": 3630
    },
    {
      "epoch": 0.7262,
      "grad_norm": 2.0937368869781494,
      "learning_rate": 2.117838419151034e-05,
      "loss": 0.4536,
      "step": 3631
    },
    {
      "epoch": 0.7262,
      "loss_align": 0.07658243179321289,
      "loss_contrastive": 0.022840380668640137,
      "loss_gen": 0.23916859924793243,
      "neg_sim_mean": 0.8462579250335693,
      "pos_sim_mean": 0.9234175682067871,
      "step": 3631
    },
    {
      "epoch": 0.7264,
      "grad_norm": 2.0593814849853516,
      "learning_rate": 2.114986747959643e-05,
      "loss": 0.3185,
      "step": 3632
    },
    {
      "epoch": 0.7264,
      "loss_align": 0.09284698963165283,
      "loss_contrastive": 0.051635682582855225,
      "loss_gen": 0.18873436748981476,
      "neg_sim_mean": 0.8587886691093445,
      "pos_sim_mean": 0.9071530103683472,
      "step": 3632
    },
    {
      "epoch": 0.7266,
      "grad_norm": 1.7519958019256592,
      "learning_rate": 2.112136482888663e-05,
      "loss": 0.2878,
      "step": 3633
    },
    {
      "epoch": 0.7266,
      "loss_align": 0.12798964977264404,
      "loss_contrastive": 0.04257047176361084,
      "loss_gen": 0.16826868057250977,
      "neg_sim_mean": 0.8145807981491089,
      "pos_sim_mean": 0.872010350227356,
      "step": 3633
    },
    {
      "epoch": 0.7268,
      "grad_norm": 1.5679831504821777,
      "learning_rate": 2.1092876253272792e-05,
      "loss": 0.3014,
      "step": 3634
    },
    {
      "epoch": 0.7268,
      "loss_align": 0.18154650926589966,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3762083649635315,
      "neg_sim_mean": 0.7072988748550415,
      "pos_sim_mean": 0.8184534907341003,
      "step": 3634
    },
    {
      "epoch": 0.727,
      "grad_norm": 2.16359543800354,
      "learning_rate": 2.1064401766639895e-05,
      "loss": 0.5578,
      "step": 3635
    },
    {
      "epoch": 0.727,
      "loss_align": 0.1030280590057373,
      "loss_contrastive": 0.06807750463485718,
      "loss_gen": 0.20992596447467804,
      "neg_sim_mean": 0.865049421787262,
      "pos_sim_mean": 0.8969719409942627,
      "step": 3635
    },
    {
      "epoch": 0.7272,
      "grad_norm": 1.7174476385116577,
      "learning_rate": 2.103594138286607e-05,
      "loss": 0.3211,
      "step": 3636
    },
    {
      "epoch": 0.7272,
      "loss_align": 0.08177626132965088,
      "loss_contrastive": 0.05743002891540527,
      "loss_gen": 0.21149201691150665,
      "neg_sim_mean": 0.8756537437438965,
      "pos_sim_mean": 0.9182237386703491,
      "step": 3636
    },
    {
      "epoch": 0.7274,
      "grad_norm": 1.7316651344299316,
      "learning_rate": 2.100749511582254e-05,
      "loss": 0.3002,
      "step": 3637
    },
    {
      "epoch": 0.7274,
      "loss_align": 0.17448484897613525,
      "loss_contrastive": 0.13369661569595337,
      "loss_gen": 0.3647521138191223,
      "neg_sim_mean": 0.8592117428779602,
      "pos_sim_mean": 0.8255151510238647,
      "step": 3637
    },
    {
      "epoch": 0.7276,
      "grad_norm": 2.0969314575195312,
      "learning_rate": 2.0979062979373697e-05,
      "loss": 0.5553,
      "step": 3638
    },
    {
      "epoch": 0.7276,
      "loss_align": 0.11222672462463379,
      "loss_contrastive": 0.08095544576644897,
      "loss_gen": 0.2525885999202728,
      "neg_sim_mean": 0.8687286972999573,
      "pos_sim_mean": 0.8877732753753662,
      "step": 3638
    },
    {
      "epoch": 0.7278,
      "grad_norm": 2.171757698059082,
      "learning_rate": 2.095064498737701e-05,
      "loss": 0.3745,
      "step": 3639
    },
    {
      "epoch": 0.7278,
      "loss_align": 0.11896264553070068,
      "loss_contrastive": 0.09059315919876099,
      "loss_gen": 0.4568716585636139,
      "neg_sim_mean": 0.8716304898262024,
      "pos_sim_mean": 0.8810373544692993,
      "step": 3639
    },
    {
      "epoch": 0.728,
      "grad_norm": 3.065706968307495,
      "learning_rate": 2.0922241153683066e-05,
      "loss": 0.5867,
      "step": 3640
    },
    {
      "epoch": 0.728,
      "loss_align": 0.5369352102279663,
      "loss_contrastive": 0.2066739797592163,
      "loss_gen": 0.35973289608955383,
      "neg_sim_mean": 0.5697388052940369,
      "pos_sim_mean": 0.4630648195743561,
      "step": 3640
    },
    {
      "epoch": 0.7282,
      "grad_norm": 3.285412073135376,
      "learning_rate": 2.0893851492135537e-05,
      "loss": 0.9215,
      "step": 3641
    },
    {
      "epoch": 0.7282,
      "loss_align": 0.06327897310256958,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1535925418138504,
      "neg_sim_mean": 0.8257704973220825,
      "pos_sim_mean": 0.9367210268974304,
      "step": 3641
    },
    {
      "epoch": 0.7284,
      "grad_norm": 1.5883816480636597,
      "learning_rate": 2.0865476016571207e-05,
      "loss": 0.2169,
      "step": 3642
    },
    {
      "epoch": 0.7284,
      "loss_align": 0.06200826168060303,
      "loss_contrastive": 0.0,
      "loss_gen": 0.207776740193367,
      "neg_sim_mean": 0.79222571849823,
      "pos_sim_mean": 0.937991738319397,
      "step": 3642
    },
    {
      "epoch": 0.7286,
      "grad_norm": 1.6208471059799194,
      "learning_rate": 2.083711474081993e-05,
      "loss": 0.2698,
      "step": 3643
    },
    {
      "epoch": 0.7286,
      "loss_align": 0.1156458854675293,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004297063685953617,
      "neg_sim_mean": 0.43758898973464966,
      "pos_sim_mean": 0.8843541145324707,
      "step": 3643
    },
    {
      "epoch": 0.7288,
      "grad_norm": 0.4807950258255005,
      "learning_rate": 2.080876767870466e-05,
      "loss": 0.1199,
      "step": 3644
    },
    {
      "epoch": 0.7288,
      "loss_align": 0.1993720531463623,
      "loss_contrastive": 0.03689831495285034,
      "loss_gen": 0.13194790482521057,
      "neg_sim_mean": 0.7375262379646301,
      "pos_sim_mean": 0.8006279468536377,
      "step": 3644
    },
    {
      "epoch": 0.729,
      "grad_norm": 1.5752477645874023,
      "learning_rate": 2.07804348440414e-05,
      "loss": 0.3357,
      "step": 3645
    },
    {
      "epoch": 0.729,
      "loss_align": 0.08577620983123779,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22616270184516907,
      "neg_sim_mean": 0.7278270721435547,
      "pos_sim_mean": 0.9142237901687622,
      "step": 3645
    },
    {
      "epoch": 0.7292,
      "grad_norm": 1.6113454103469849,
      "learning_rate": 2.0752116250639225e-05,
      "loss": 0.3119,
      "step": 3646
    },
    {
      "epoch": 0.7292,
      "loss_align": 0.10973000526428223,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23434685170650482,
      "neg_sim_mean": 0.6384326219558716,
      "pos_sim_mean": 0.8902699947357178,
      "step": 3646
    },
    {
      "epoch": 0.7294,
      "grad_norm": 1.9880465269088745,
      "learning_rate": 2.0723811912300295e-05,
      "loss": 0.3441,
      "step": 3647
    },
    {
      "epoch": 0.7294,
      "loss_align": 0.10147988796234131,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004108544439077377,
      "neg_sim_mean": 0.27352359890937805,
      "pos_sim_mean": 0.8985201120376587,
      "step": 3647
    },
    {
      "epoch": 0.7296,
      "grad_norm": 0.436681866645813,
      "learning_rate": 2.0695521842819788e-05,
      "loss": 0.1056,
      "step": 3648
    },
    {
      "epoch": 0.7296,
      "loss_align": 0.11019623279571533,
      "loss_contrastive": 0.05649232864379883,
      "loss_gen": 0.3181915879249573,
      "neg_sim_mean": 0.8462960720062256,
      "pos_sim_mean": 0.8898037672042847,
      "step": 3648
    },
    {
      "epoch": 0.7298,
      "grad_norm": 2.125711679458618,
      "learning_rate": 2.066724605598594e-05,
      "loss": 0.4352,
      "step": 3649
    },
    {
      "epoch": 0.7298,
      "loss_align": 0.08016979694366455,
      "loss_contrastive": 0.05654299259185791,
      "loss_gen": 0.5554536581039429,
      "neg_sim_mean": 0.8763731718063354,
      "pos_sim_mean": 0.9198302030563354,
      "step": 3649
    },
    {
      "epoch": 0.73,
      "grad_norm": 2.3735246658325195,
      "learning_rate": 2.0638984565580023e-05,
      "loss": 0.6424,
      "step": 3650
    },
    {
      "epoch": 0.73,
      "loss_align": 0.1417860984802246,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011405035853385925,
      "neg_sim_mean": 0.4034775495529175,
      "pos_sim_mean": 0.8582139015197754,
      "step": 3650
    },
    {
      "epoch": 0.7302,
      "grad_norm": 0.8037242889404297,
      "learning_rate": 2.061073738537635e-05,
      "loss": 0.1532,
      "step": 3651
    },
    {
      "epoch": 0.7302,
      "loss_align": 0.14683514833450317,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0207161083817482,
      "neg_sim_mean": 0.3660847544670105,
      "pos_sim_mean": 0.8531648516654968,
      "step": 3651
    },
    {
      "epoch": 0.7304,
      "grad_norm": 1.3669227361679077,
      "learning_rate": 2.058250452914225e-05,
      "loss": 0.1676,
      "step": 3652
    },
    {
      "epoch": 0.7304,
      "loss_align": 0.08068680763244629,
      "loss_contrastive": 0.02425968647003174,
      "loss_gen": 0.26171043515205383,
      "neg_sim_mean": 0.8435728549957275,
      "pos_sim_mean": 0.9193131923675537,
      "step": 3652
    },
    {
      "epoch": 0.7306,
      "grad_norm": 1.9008262157440186,
      "learning_rate": 2.0554286010638076e-05,
      "loss": 0.3453,
      "step": 3653
    },
    {
      "epoch": 0.7306,
      "loss_align": 0.12747687101364136,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01800237037241459,
      "neg_sim_mean": 0.26470762491226196,
      "pos_sim_mean": 0.8725231289863586,
      "step": 3653
    },
    {
      "epoch": 0.7308,
      "grad_norm": 0.9607683420181274,
      "learning_rate": 2.0526081843617183e-05,
      "loss": 0.1455,
      "step": 3654
    },
    {
      "epoch": 0.7308,
      "loss_align": 0.06911325454711914,
      "loss_contrastive": 0.0938032865524292,
      "loss_gen": 0.23882828652858734,
      "neg_sim_mean": 0.9246900081634521,
      "pos_sim_mean": 0.9308867454528809,
      "step": 3654
    },
    {
      "epoch": 0.731,
      "grad_norm": 1.6723864078521729,
      "learning_rate": 2.0497892041825955e-05,
      "loss": 0.3192,
      "step": 3655
    },
    {
      "epoch": 0.731,
      "loss_align": 0.1657322645187378,
      "loss_contrastive": 0.0,
      "loss_gen": 0.029676632955670357,
      "neg_sim_mean": 0.2621373236179352,
      "pos_sim_mean": 0.8342677354812622,
      "step": 3655
    },
    {
      "epoch": 0.7312,
      "grad_norm": 1.186947226524353,
      "learning_rate": 2.0469716619003725e-05,
      "loss": 0.1954,
      "step": 3656
    },
    {
      "epoch": 0.7312,
      "loss_align": 0.0545506477355957,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1941317319869995,
      "neg_sim_mean": 0.8397750854492188,
      "pos_sim_mean": 0.9454493522644043,
      "step": 3656
    },
    {
      "epoch": 0.7314,
      "grad_norm": 1.5840970277786255,
      "learning_rate": 2.04415555888829e-05,
      "loss": 0.2487,
      "step": 3657
    },
    {
      "epoch": 0.7314,
      "loss_align": 0.07180476188659668,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17934364080429077,
      "neg_sim_mean": 0.8039759397506714,
      "pos_sim_mean": 0.9281952381134033,
      "step": 3657
    },
    {
      "epoch": 0.7316,
      "grad_norm": 1.8834421634674072,
      "learning_rate": 2.04134089651888e-05,
      "loss": 0.2511,
      "step": 3658
    },
    {
      "epoch": 0.7316,
      "loss_align": 0.1628018617630005,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0046516661532223225,
      "neg_sim_mean": 0.3100869655609131,
      "pos_sim_mean": 0.8371981382369995,
      "step": 3658
    },
    {
      "epoch": 0.7318,
      "grad_norm": 0.6094536781311035,
      "learning_rate": 2.0385276761639765e-05,
      "loss": 0.1675,
      "step": 3659
    },
    {
      "epoch": 0.7318,
      "loss_align": 0.1463719606399536,
      "loss_contrastive": 0.02346813678741455,
      "loss_gen": 0.203136146068573,
      "neg_sim_mean": 0.777096152305603,
      "pos_sim_mean": 0.8536280393600464,
      "step": 3659
    },
    {
      "epoch": 0.732,
      "grad_norm": 1.7169976234436035,
      "learning_rate": 2.035715899194704e-05,
      "loss": 0.3523,
      "step": 3660
    },
    {
      "epoch": 0.732,
      "loss_align": 0.09150171279907227,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1879991739988327,
      "neg_sim_mean": 0.7784081101417542,
      "pos_sim_mean": 0.9084982872009277,
      "step": 3660
    },
    {
      "epoch": 0.7322,
      "grad_norm": 1.6647628545761108,
      "learning_rate": 2.0329055669814934e-05,
      "loss": 0.2795,
      "step": 3661
    },
    {
      "epoch": 0.7322,
      "loss_align": 0.17894220352172852,
      "loss_contrastive": 0.0,
      "loss_gen": 0.32497820258140564,
      "neg_sim_mean": 0.6044092178344727,
      "pos_sim_mean": 0.8210577964782715,
      "step": 3661
    },
    {
      "epoch": 0.7324,
      "grad_norm": 1.8319870233535767,
      "learning_rate": 2.0300966808940647e-05,
      "loss": 0.5039,
      "step": 3662
    },
    {
      "epoch": 0.7324,
      "loss_align": 0.08562779426574707,
      "loss_contrastive": 0.0,
      "loss_gen": 0.015435673296451569,
      "neg_sim_mean": 0.4143538475036621,
      "pos_sim_mean": 0.9143722057342529,
      "step": 3662
    },
    {
      "epoch": 0.7326,
      "grad_norm": 0.8245528936386108,
      "learning_rate": 2.027289242301435e-05,
      "loss": 0.1011,
      "step": 3663
    },
    {
      "epoch": 0.7326,
      "loss_align": 0.10183048248291016,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00427537877112627,
      "neg_sim_mean": 0.28180956840515137,
      "pos_sim_mean": 0.8981695175170898,
      "step": 3663
    },
    {
      "epoch": 0.7328,
      "grad_norm": 0.44198495149612427,
      "learning_rate": 2.0244832525719155e-05,
      "loss": 0.1061,
      "step": 3664
    },
    {
      "epoch": 0.7328,
      "loss_align": 0.13704192638397217,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12816591560840607,
      "neg_sim_mean": 0.7123830318450928,
      "pos_sim_mean": 0.8629580736160278,
      "step": 3664
    },
    {
      "epoch": 0.733,
      "grad_norm": 1.5760995149612427,
      "learning_rate": 2.0216787130731103e-05,
      "loss": 0.2652,
      "step": 3665
    },
    {
      "epoch": 0.733,
      "loss_align": 0.07801258563995361,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18078278005123138,
      "neg_sim_mean": 0.766403079032898,
      "pos_sim_mean": 0.9219874143600464,
      "step": 3665
    },
    {
      "epoch": 0.7332,
      "grad_norm": 1.721720576286316,
      "learning_rate": 2.0188756251719203e-05,
      "loss": 0.2588,
      "step": 3666
    },
    {
      "epoch": 0.7332,
      "loss_align": 0.11591917276382446,
      "loss_contrastive": 0.0,
      "loss_gen": 0.30156880617141724,
      "neg_sim_mean": 0.7557924389839172,
      "pos_sim_mean": 0.8840808272361755,
      "step": 3666
    },
    {
      "epoch": 0.7334,
      "grad_norm": 1.811987280845642,
      "learning_rate": 2.016073990234536e-05,
      "loss": 0.4175,
      "step": 3667
    },
    {
      "epoch": 0.7334,
      "loss_align": 0.14443594217300415,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20696333050727844,
      "neg_sim_mean": 0.6230404376983643,
      "pos_sim_mean": 0.8555640578269958,
      "step": 3667
    },
    {
      "epoch": 0.7336,
      "grad_norm": 1.730450987815857,
      "learning_rate": 2.0132738096264415e-05,
      "loss": 0.3514,
      "step": 3668
    },
    {
      "epoch": 0.7336,
      "loss_align": 0.09183913469314575,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15146616101264954,
      "neg_sim_mean": 0.8066072463989258,
      "pos_sim_mean": 0.9081608653068542,
      "step": 3668
    },
    {
      "epoch": 0.7338,
      "grad_norm": 1.2986892461776733,
      "learning_rate": 2.0104750847124075e-05,
      "loss": 0.2433,
      "step": 3669
    },
    {
      "epoch": 0.7338,
      "loss_align": 0.08510661125183105,
      "loss_contrastive": 0.0072844624519348145,
      "loss_gen": 0.17870968580245972,
      "neg_sim_mean": 0.8221778273582458,
      "pos_sim_mean": 0.914893388748169,
      "step": 3669
    },
    {
      "epoch": 0.734,
      "grad_norm": 1.5433690547943115,
      "learning_rate": 2.007677816856498e-05,
      "loss": 0.2647,
      "step": 3670
    },
    {
      "epoch": 0.734,
      "loss_align": 0.09421384334564209,
      "loss_contrastive": 0.09580457210540771,
      "loss_gen": 0.2590268850326538,
      "neg_sim_mean": 0.9015907049179077,
      "pos_sim_mean": 0.9057861566543579,
      "step": 3670
    },
    {
      "epoch": 0.7342,
      "grad_norm": 1.737075924873352,
      "learning_rate": 2.0048820074220715e-05,
      "loss": 0.3647,
      "step": 3671
    },
    {
      "epoch": 0.7342,
      "loss_align": 0.15662819147109985,
      "loss_contrastive": 0.008056104183197021,
      "loss_gen": 0.1857445240020752,
      "neg_sim_mean": 0.7514278888702393,
      "pos_sim_mean": 0.8433718085289001,
      "step": 3671
    },
    {
      "epoch": 0.7344,
      "grad_norm": 1.7419512271881104,
      "learning_rate": 2.002087657771769e-05,
      "loss": 0.3433,
      "step": 3672
    },
    {
      "epoch": 0.7344,
      "loss_align": 0.02744615077972412,
      "loss_contrastive": 0.009297966957092285,
      "loss_gen": 0.1566399186849594,
      "neg_sim_mean": 0.8818517923355103,
      "pos_sim_mean": 0.9725538492202759,
      "step": 3672
    },
    {
      "epoch": 0.7346,
      "grad_norm": 1.8881107568740845,
      "learning_rate": 1.9992947692675228e-05,
      "loss": 0.1852,
      "step": 3673
    },
    {
      "epoch": 0.7346,
      "loss_align": 0.09013485908508301,
      "loss_contrastive": 0.08652061223983765,
      "loss_gen": 0.3165191411972046,
      "neg_sim_mean": 0.8963857293128967,
      "pos_sim_mean": 0.909865140914917,
      "step": 3673
    },
    {
      "epoch": 0.7348,
      "grad_norm": 1.988041639328003,
      "learning_rate": 1.9965033432705537e-05,
      "loss": 0.417,
      "step": 3674
    },
    {
      "epoch": 0.7348,
      "loss_align": 0.07073938846588135,
      "loss_contrastive": 0.025051772594451904,
      "loss_gen": 0.17810450494289398,
      "neg_sim_mean": 0.8543123602867126,
      "pos_sim_mean": 0.9292606115341187,
      "step": 3674
    },
    {
      "epoch": 0.735,
      "grad_norm": 1.7054636478424072,
      "learning_rate": 1.9937133811413666e-05,
      "loss": 0.2519,
      "step": 3675
    },
    {
      "epoch": 0.735,
      "loss_align": 0.1069953441619873,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011864305706694722,
      "neg_sim_mean": 0.4198337495326996,
      "pos_sim_mean": 0.8930046558380127,
      "step": 3675
    },
    {
      "epoch": 0.7352,
      "grad_norm": 0.38752928376197815,
      "learning_rate": 1.9909248842397584e-05,
      "loss": 0.1082,
      "step": 3676
    },
    {
      "epoch": 0.7352,
      "loss_align": 0.07834690809249878,
      "loss_contrastive": 0.0864332914352417,
      "loss_gen": 0.1304367184638977,
      "neg_sim_mean": 0.908086359500885,
      "pos_sim_mean": 0.9216530919075012,
      "step": 3676
    },
    {
      "epoch": 0.7354,
      "grad_norm": 1.6190582513809204,
      "learning_rate": 1.988137853924808e-05,
      "loss": 0.2192,
      "step": 3677
    },
    {
      "epoch": 0.7354,
      "loss_align": 0.10976117849349976,
      "loss_contrastive": 0.05691426992416382,
      "loss_gen": 0.11375325173139572,
      "neg_sim_mean": 0.8471530675888062,
      "pos_sim_mean": 0.8902388215065002,
      "step": 3677
    },
    {
      "epoch": 0.7356,
      "grad_norm": 1.2122336626052856,
      "learning_rate": 1.9853522915548777e-05,
      "loss": 0.2303,
      "step": 3678
    },
    {
      "epoch": 0.7356,
      "loss_align": 0.20774924755096436,
      "loss_contrastive": 0.0,
      "loss_gen": 0.35498401522636414,
      "neg_sim_mean": 0.5448593497276306,
      "pos_sim_mean": 0.7922507524490356,
      "step": 3678
    },
    {
      "epoch": 0.7358,
      "grad_norm": 2.4675915241241455,
      "learning_rate": 1.9825681984876172e-05,
      "loss": 0.5627,
      "step": 3679
    },
    {
      "epoch": 0.7358,
      "loss_align": 0.1030157208442688,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006865859962999821,
      "neg_sim_mean": 0.4606668949127197,
      "pos_sim_mean": 0.8969842791557312,
      "step": 3679
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.5730878710746765,
      "learning_rate": 1.9797855760799612e-05,
      "loss": 0.1099,
      "step": 3680
    },
    {
      "epoch": 0.736,
      "loss_align": 0.0673443078994751,
      "loss_contrastive": 0.05946773290634155,
      "loss_gen": 0.14085297286510468,
      "neg_sim_mean": 0.8921234011650085,
      "pos_sim_mean": 0.9326556921005249,
      "step": 3680
    },
    {
      "epoch": 0.7362,
      "grad_norm": 1.562560796737671,
      "learning_rate": 1.977004425688126e-05,
      "loss": 0.2153,
      "step": 3681
    },
    {
      "epoch": 0.7362,
      "loss_align": 0.11943769454956055,
      "loss_contrastive": 0.0,
      "loss_gen": 0.34606337547302246,
      "neg_sim_mean": 0.745166540145874,
      "pos_sim_mean": 0.8805623054504395,
      "step": 3681
    },
    {
      "epoch": 0.7364,
      "grad_norm": 2.4936275482177734,
      "learning_rate": 1.9742247486676096e-05,
      "loss": 0.4655,
      "step": 3682
    },
    {
      "epoch": 0.7364,
      "loss_align": 0.17840635776519775,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0017497787484899163,
      "neg_sim_mean": 0.3343670070171356,
      "pos_sim_mean": 0.8215936422348022,
      "step": 3682
    },
    {
      "epoch": 0.7366,
      "grad_norm": 0.7608959078788757,
      "learning_rate": 1.9714465463731934e-05,
      "loss": 0.1802,
      "step": 3683
    },
    {
      "epoch": 0.7366,
      "loss_align": 0.12575548887252808,
      "loss_contrastive": 0.0693628191947937,
      "loss_gen": 0.24514475464820862,
      "neg_sim_mean": 0.8436073064804077,
      "pos_sim_mean": 0.8742445111274719,
      "step": 3683
    },
    {
      "epoch": 0.7368,
      "grad_norm": 1.6409287452697754,
      "learning_rate": 1.9686698201589393e-05,
      "loss": 0.3792,
      "step": 3684
    },
    {
      "epoch": 0.7368,
      "loss_align": 0.1096041202545166,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1270279586315155,
      "neg_sim_mean": 0.6939575672149658,
      "pos_sim_mean": 0.8903958797454834,
      "step": 3684
    },
    {
      "epoch": 0.737,
      "grad_norm": 1.375792145729065,
      "learning_rate": 1.9658945713781883e-05,
      "loss": 0.2366,
      "step": 3685
    },
    {
      "epoch": 0.737,
      "loss_align": 0.1485193967819214,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006390579976141453,
      "neg_sim_mean": 0.26091468334198,
      "pos_sim_mean": 0.8514806032180786,
      "step": 3685
    },
    {
      "epoch": 0.7372,
      "grad_norm": 0.619106650352478,
      "learning_rate": 1.9631208013835678e-05,
      "loss": 0.1549,
      "step": 3686
    },
    {
      "epoch": 0.7372,
      "loss_align": 0.19801115989685059,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002529817633330822,
      "neg_sim_mean": 0.3746342360973358,
      "pos_sim_mean": 0.8019888401031494,
      "step": 3686
    },
    {
      "epoch": 0.7374,
      "grad_norm": 0.6900413036346436,
      "learning_rate": 1.9603485115269744e-05,
      "loss": 0.2005,
      "step": 3687
    },
    {
      "epoch": 0.7374,
      "loss_align": 0.09211838245391846,
      "loss_contrastive": 0.027719736099243164,
      "loss_gen": 0.11912372708320618,
      "neg_sim_mean": 0.8356013298034668,
      "pos_sim_mean": 0.9078816175460815,
      "step": 3687
    },
    {
      "epoch": 0.7376,
      "grad_norm": 1.2264318466186523,
      "learning_rate": 1.9575777031595905e-05,
      "loss": 0.2146,
      "step": 3688
    },
    {
      "epoch": 0.7376,
      "loss_align": 0.09181791543960571,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02190118096768856,
      "neg_sim_mean": 0.3394765853881836,
      "pos_sim_mean": 0.9081820845603943,
      "step": 3688
    },
    {
      "epoch": 0.7378,
      "grad_norm": 1.0350497961044312,
      "learning_rate": 1.9548083776318727e-05,
      "loss": 0.1137,
      "step": 3689
    },
    {
      "epoch": 0.7378,
      "loss_align": 0.04344356060028076,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15461046993732452,
      "neg_sim_mean": 0.5111315250396729,
      "pos_sim_mean": 0.9565564393997192,
      "step": 3689
    },
    {
      "epoch": 0.738,
      "grad_norm": 1.7019109725952148,
      "learning_rate": 1.9520405362935594e-05,
      "loss": 0.1981,
      "step": 3690
    },
    {
      "epoch": 0.738,
      "loss_align": 0.16251325607299805,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21549662947654724,
      "neg_sim_mean": 0.7052812576293945,
      "pos_sim_mean": 0.837486743927002,
      "step": 3690
    },
    {
      "epoch": 0.7382,
      "grad_norm": 1.7422499656677246,
      "learning_rate": 1.9492741804936622e-05,
      "loss": 0.378,
      "step": 3691
    },
    {
      "epoch": 0.7382,
      "loss_align": 0.07279562950134277,
      "loss_contrastive": 0.07336580753326416,
      "loss_gen": 0.28858840465545654,
      "neg_sim_mean": 0.9005701541900635,
      "pos_sim_mean": 0.9272043704986572,
      "step": 3691
    },
    {
      "epoch": 0.7384,
      "grad_norm": 2.211151361465454,
      "learning_rate": 1.946509311580469e-05,
      "loss": 0.3702,
      "step": 3692
    },
    {
      "epoch": 0.7384,
      "loss_align": 0.1386629343032837,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16190311312675476,
      "neg_sim_mean": 0.6447662115097046,
      "pos_sim_mean": 0.8613370656967163,
      "step": 3692
    },
    {
      "epoch": 0.7386,
      "grad_norm": 1.4016116857528687,
      "learning_rate": 1.9437459309015427e-05,
      "loss": 0.3006,
      "step": 3693
    },
    {
      "epoch": 0.7386,
      "loss_align": 0.0843038558959961,
      "loss_contrastive": 0.03802168369293213,
      "loss_gen": 0.3687899708747864,
      "neg_sim_mean": 0.8537178039550781,
      "pos_sim_mean": 0.9156961441040039,
      "step": 3693
    },
    {
      "epoch": 0.7388,
      "grad_norm": 1.9166287183761597,
      "learning_rate": 1.94098403980372e-05,
      "loss": 0.4577,
      "step": 3694
    },
    {
      "epoch": 0.7388,
      "loss_align": 0.10286694765090942,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1529339849948883,
      "neg_sim_mean": 0.7247970104217529,
      "pos_sim_mean": 0.8971330523490906,
      "step": 3694
    },
    {
      "epoch": 0.739,
      "grad_norm": 1.2924810647964478,
      "learning_rate": 1.938223639633119e-05,
      "loss": 0.2558,
      "step": 3695
    },
    {
      "epoch": 0.739,
      "loss_align": 0.14850342273712158,
      "loss_contrastive": 0.029791831970214844,
      "loss_gen": 0.3453047275543213,
      "neg_sim_mean": 0.7812883853912354,
      "pos_sim_mean": 0.8514965772628784,
      "step": 3695
    },
    {
      "epoch": 0.7392,
      "grad_norm": 2.010547637939453,
      "learning_rate": 1.9354647317351188e-05,
      "loss": 0.4974,
      "step": 3696
    },
    {
      "epoch": 0.7392,
      "loss_align": 0.13272684812545776,
      "loss_contrastive": 0.12666040658950806,
      "loss_gen": 0.19403207302093506,
      "neg_sim_mean": 0.8939335346221924,
      "pos_sim_mean": 0.8672731518745422,
      "step": 3696
    },
    {
      "epoch": 0.7394,
      "grad_norm": 1.4992231130599976,
      "learning_rate": 1.9327073174543803e-05,
      "loss": 0.342,
      "step": 3697
    },
    {
      "epoch": 0.7394,
      "loss_align": 0.10416567325592041,
      "loss_contrastive": 0.10372322797775269,
      "loss_gen": 0.21988990902900696,
      "neg_sim_mean": 0.8995575308799744,
      "pos_sim_mean": 0.8958343267440796,
      "step": 3697
    },
    {
      "epoch": 0.7396,
      "grad_norm": 1.5868585109710693,
      "learning_rate": 1.9299513981348316e-05,
      "loss": 0.3365,
      "step": 3698
    },
    {
      "epoch": 0.7396,
      "loss_align": 0.09906810522079468,
      "loss_contrastive": 0.028699278831481934,
      "loss_gen": 0.18075603246688843,
      "neg_sim_mean": 0.8296311497688293,
      "pos_sim_mean": 0.9009318947792053,
      "step": 3698
    },
    {
      "epoch": 0.7398,
      "grad_norm": 1.947264552116394,
      "learning_rate": 1.9271969751196776e-05,
      "loss": 0.2833,
      "step": 3699
    },
    {
      "epoch": 0.7398,
      "loss_align": 0.10484945774078369,
      "loss_contrastive": 0.09008324146270752,
      "loss_gen": 0.5189245939254761,
      "neg_sim_mean": 0.8852337598800659,
      "pos_sim_mean": 0.8951505422592163,
      "step": 3699
    },
    {
      "epoch": 0.74,
      "grad_norm": 3.1489081382751465,
      "learning_rate": 1.9244440497513893e-05,
      "loss": 0.6346,
      "step": 3700
    },
    {
      "epoch": 0.74,
      "loss_align": 0.08365237712860107,
      "loss_contrastive": 0.05867987871170044,
      "loss_gen": 0.37368351221084595,
      "neg_sim_mean": 0.8750274777412415,
      "pos_sim_mean": 0.9163476228713989,
      "step": 3700
    },
    {
      "epoch": 0.7402,
      "grad_norm": 2.489893674850464,
      "learning_rate": 1.9216926233717085e-05,
      "loss": 0.4644,
      "step": 3701
    },
    {
      "epoch": 0.7402,
      "loss_align": 0.09653061628341675,
      "loss_contrastive": 0.10732567310333252,
      "loss_gen": 0.13272441923618317,
      "neg_sim_mean": 0.9107950329780579,
      "pos_sim_mean": 0.9034693837165833,
      "step": 3701
    },
    {
      "epoch": 0.7404,
      "grad_norm": 1.4531912803649902,
      "learning_rate": 1.918942697321648e-05,
      "loss": 0.2421,
      "step": 3702
    },
    {
      "epoch": 0.7404,
      "loss_align": 0.10368925333023071,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002580236876383424,
      "neg_sim_mean": 0.3299426734447479,
      "pos_sim_mean": 0.8963107466697693,
      "step": 3702
    },
    {
      "epoch": 0.7406,
      "grad_norm": 0.4267100691795349,
      "learning_rate": 1.916194272941488e-05,
      "loss": 0.1063,
      "step": 3703
    },
    {
      "epoch": 0.7406,
      "loss_align": 0.1203814148902893,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0029639217536896467,
      "neg_sim_mean": 0.33919259905815125,
      "pos_sim_mean": 0.8796185851097107,
      "step": 3703
    },
    {
      "epoch": 0.7408,
      "grad_norm": 0.393047571182251,
      "learning_rate": 1.9134473515707757e-05,
      "loss": 0.1233,
      "step": 3704
    },
    {
      "epoch": 0.7408,
      "loss_align": 0.042461514472961426,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13742992281913757,
      "neg_sim_mean": 0.5797561407089233,
      "pos_sim_mean": 0.9575384855270386,
      "step": 3704
    },
    {
      "epoch": 0.741,
      "grad_norm": 1.2582556009292603,
      "learning_rate": 1.910701934548329e-05,
      "loss": 0.1799,
      "step": 3705
    },
    {
      "epoch": 0.741,
      "loss_align": 0.08041739463806152,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3008759915828705,
      "neg_sim_mean": 0.7530463933944702,
      "pos_sim_mean": 0.9195826053619385,
      "step": 3705
    },
    {
      "epoch": 0.7412,
      "grad_norm": 1.8146424293518066,
      "learning_rate": 1.9079580232122303e-05,
      "loss": 0.3813,
      "step": 3706
    },
    {
      "epoch": 0.7412,
      "loss_align": 0.12786978483200073,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13541200757026672,
      "neg_sim_mean": 0.6663631200790405,
      "pos_sim_mean": 0.8721302151679993,
      "step": 3706
    },
    {
      "epoch": 0.7414,
      "grad_norm": 1.566952109336853,
      "learning_rate": 1.9052156188998282e-05,
      "loss": 0.2633,
      "step": 3707
    },
    {
      "epoch": 0.7414,
      "loss_align": 0.05185139179229736,
      "loss_contrastive": 0.027277112007141113,
      "loss_gen": 0.08199995756149292,
      "neg_sim_mean": 0.8754256963729858,
      "pos_sim_mean": 0.9481486082077026,
      "step": 3707
    },
    {
      "epoch": 0.7416,
      "grad_norm": 0.9739813208580017,
      "learning_rate": 1.9024747229477363e-05,
      "loss": 0.1371,
      "step": 3708
    },
    {
      "epoch": 0.7416,
      "loss_align": 0.08431220054626465,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028162676841020584,
      "neg_sim_mean": 0.2821047008037567,
      "pos_sim_mean": 0.9156877994537354,
      "step": 3708
    },
    {
      "epoch": 0.7418,
      "grad_norm": 1.0226553678512573,
      "learning_rate": 1.899735336691837e-05,
      "loss": 0.1125,
      "step": 3709
    },
    {
      "epoch": 0.7418,
      "loss_align": 0.16258221864700317,
      "loss_contrastive": 0.12237763404846191,
      "loss_gen": 0.2682333290576935,
      "neg_sim_mean": 0.8597953915596008,
      "pos_sim_mean": 0.8374177813529968,
      "step": 3709
    },
    {
      "epoch": 0.742,
      "grad_norm": 2.382131576538086,
      "learning_rate": 1.896997461467272e-05,
      "loss": 0.4455,
      "step": 3710
    },
    {
      "epoch": 0.742,
      "loss_align": 0.07380616664886475,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0022609189618378878,
      "neg_sim_mean": 0.3536030054092407,
      "pos_sim_mean": 0.9261938333511353,
      "step": 3710
    },
    {
      "epoch": 0.7422,
      "grad_norm": 0.3567456305027008,
      "learning_rate": 1.8942610986084486e-05,
      "loss": 0.0761,
      "step": 3711
    },
    {
      "epoch": 0.7422,
      "loss_align": 0.086589515209198,
      "loss_contrastive": 0.0,
      "loss_gen": 0.014580477960407734,
      "neg_sim_mean": 0.24322713911533356,
      "pos_sim_mean": 0.913410484790802,
      "step": 3711
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.8294432163238525,
      "learning_rate": 1.8915262494490365e-05,
      "loss": 0.1012,
      "step": 3712
    },
    {
      "epoch": 0.7424,
      "loss_align": 0.13741672039031982,
      "loss_contrastive": 0.06208544969558716,
      "loss_gen": 0.36549943685531616,
      "neg_sim_mean": 0.8246687054634094,
      "pos_sim_mean": 0.8625832796096802,
      "step": 3712
    },
    {
      "epoch": 0.7426,
      "grad_norm": 2.077716112136841,
      "learning_rate": 1.8887929153219687e-05,
      "loss": 0.5104,
      "step": 3713
    },
    {
      "epoch": 0.7426,
      "loss_align": 0.18205654621124268,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2529843747615814,
      "neg_sim_mean": 0.6640803813934326,
      "pos_sim_mean": 0.8179434537887573,
      "step": 3713
    },
    {
      "epoch": 0.7428,
      "grad_norm": 2.1880695819854736,
      "learning_rate": 1.8860610975594382e-05,
      "loss": 0.435,
      "step": 3714
    },
    {
      "epoch": 0.7428,
      "loss_align": 0.12040525674819946,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0019520999630913138,
      "neg_sim_mean": 0.3345851004123688,
      "pos_sim_mean": 0.8795947432518005,
      "step": 3714
    },
    {
      "epoch": 0.743,
      "grad_norm": 0.3811946213245392,
      "learning_rate": 1.8833307974929004e-05,
      "loss": 0.1224,
      "step": 3715
    },
    {
      "epoch": 0.743,
      "loss_align": 0.10405546426773071,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01096628699451685,
      "neg_sim_mean": 0.25518226623535156,
      "pos_sim_mean": 0.8959445357322693,
      "step": 3715
    },
    {
      "epoch": 0.7432,
      "grad_norm": 0.7314945459365845,
      "learning_rate": 1.8806020164530702e-05,
      "loss": 0.115,
      "step": 3716
    },
    {
      "epoch": 0.7432,
      "loss_align": 0.09555399417877197,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1905670166015625,
      "neg_sim_mean": 0.5280685424804688,
      "pos_sim_mean": 0.904446005821228,
      "step": 3716
    },
    {
      "epoch": 0.7434,
      "grad_norm": 1.5571751594543457,
      "learning_rate": 1.8778747557699224e-05,
      "loss": 0.2861,
      "step": 3717
    },
    {
      "epoch": 0.7434,
      "loss_align": 0.07434248924255371,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1391507387161255,
      "neg_sim_mean": 0.6168971657752991,
      "pos_sim_mean": 0.9256575107574463,
      "step": 3717
    },
    {
      "epoch": 0.7436,
      "grad_norm": 1.5352723598480225,
      "learning_rate": 1.8751490167726887e-05,
      "loss": 0.2135,
      "step": 3718
    },
    {
      "epoch": 0.7436,
      "loss_align": 0.1452723741531372,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12584394216537476,
      "neg_sim_mean": 0.7157503962516785,
      "pos_sim_mean": 0.8547276258468628,
      "step": 3718
    },
    {
      "epoch": 0.7438,
      "grad_norm": 1.4607188701629639,
      "learning_rate": 1.8724248007898647e-05,
      "loss": 0.2711,
      "step": 3719
    },
    {
      "epoch": 0.7438,
      "loss_align": 0.09508764743804932,
      "loss_contrastive": 0.05615377426147461,
      "loss_gen": 0.25282663106918335,
      "neg_sim_mean": 0.8610661029815674,
      "pos_sim_mean": 0.9049123525619507,
      "step": 3719
    },
    {
      "epoch": 0.744,
      "grad_norm": 1.7664076089859009,
      "learning_rate": 1.869702109149199e-05,
      "loss": 0.3547,
      "step": 3720
    },
    {
      "epoch": 0.744,
      "loss_align": 0.1006171703338623,
      "loss_contrastive": 0.12078911066055298,
      "loss_gen": 0.2997649908065796,
      "neg_sim_mean": 0.9201719164848328,
      "pos_sim_mean": 0.8993828296661377,
      "step": 3720
    },
    {
      "epoch": 0.7442,
      "grad_norm": 1.6743305921554565,
      "learning_rate": 1.866980943177699e-05,
      "loss": 0.4149,
      "step": 3721
    },
    {
      "epoch": 0.7442,
      "loss_align": 0.13481223583221436,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17435550689697266,
      "neg_sim_mean": 0.6362755298614502,
      "pos_sim_mean": 0.8651877641677856,
      "step": 3721
    },
    {
      "epoch": 0.7444,
      "grad_norm": 1.7456332445144653,
      "learning_rate": 1.8642613042016245e-05,
      "loss": 0.3092,
      "step": 3722
    },
    {
      "epoch": 0.7444,
      "loss_align": 0.0804823637008667,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09611241519451141,
      "neg_sim_mean": 0.7924894690513611,
      "pos_sim_mean": 0.9195176362991333,
      "step": 3722
    },
    {
      "epoch": 0.7446,
      "grad_norm": 1.378245234489441,
      "learning_rate": 1.8615431935464982e-05,
      "loss": 0.1766,
      "step": 3723
    },
    {
      "epoch": 0.7446,
      "loss_align": 0.1655104160308838,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00938358809798956,
      "neg_sim_mean": 0.5081899762153625,
      "pos_sim_mean": 0.8344895839691162,
      "step": 3723
    },
    {
      "epoch": 0.7448,
      "grad_norm": 0.8489950299263,
      "learning_rate": 1.858826612537093e-05,
      "loss": 0.1749,
      "step": 3724
    },
    {
      "epoch": 0.7448,
      "loss_align": 0.09241002798080444,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0014459526864811778,
      "neg_sim_mean": 0.4302911162376404,
      "pos_sim_mean": 0.9075899720191956,
      "step": 3724
    },
    {
      "epoch": 0.745,
      "grad_norm": 0.6572996973991394,
      "learning_rate": 1.8561115624974375e-05,
      "loss": 0.0939,
      "step": 3725
    },
    {
      "epoch": 0.745,
      "loss_align": 0.14186084270477295,
      "loss_contrastive": 0.08630591630935669,
      "loss_gen": 0.1392282396554947,
      "neg_sim_mean": 0.8444450497627258,
      "pos_sim_mean": 0.858139157295227,
      "step": 3725
    },
    {
      "epoch": 0.7452,
      "grad_norm": 1.5367416143417358,
      "learning_rate": 1.8533980447508137e-05,
      "loss": 0.2914,
      "step": 3726
    },
    {
      "epoch": 0.7452,
      "loss_align": 0.14501595497131348,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24393326044082642,
      "neg_sim_mean": 0.6541365385055542,
      "pos_sim_mean": 0.8549840450286865,
      "step": 3726
    },
    {
      "epoch": 0.7454,
      "grad_norm": 2.094822883605957,
      "learning_rate": 1.850686060619756e-05,
      "loss": 0.3889,
      "step": 3727
    },
    {
      "epoch": 0.7454,
      "loss_align": 0.2321922779083252,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22077573835849762,
      "neg_sim_mean": 0.6464233994483948,
      "pos_sim_mean": 0.7678077220916748,
      "step": 3727
    },
    {
      "epoch": 0.7456,
      "grad_norm": 2.121894121170044,
      "learning_rate": 1.847975611426056e-05,
      "loss": 0.453,
      "step": 3728
    },
    {
      "epoch": 0.7456,
      "loss_align": 0.14776122570037842,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2603372633457184,
      "neg_sim_mean": 0.6659681797027588,
      "pos_sim_mean": 0.8522387742996216,
      "step": 3728
    },
    {
      "epoch": 0.7458,
      "grad_norm": 1.920996904373169,
      "learning_rate": 1.845266698490752e-05,
      "loss": 0.4081,
      "step": 3729
    },
    {
      "epoch": 0.7458,
      "loss_align": 0.0691676139831543,
      "loss_contrastive": 0.03555935621261597,
      "loss_gen": 0.16185559332370758,
      "neg_sim_mean": 0.8663917183876038,
      "pos_sim_mean": 0.9308323860168457,
      "step": 3729
    },
    {
      "epoch": 0.746,
      "grad_norm": 1.3715431690216064,
      "learning_rate": 1.842559323134136e-05,
      "loss": 0.2353,
      "step": 3730
    },
    {
      "epoch": 0.746,
      "loss_align": 0.112437903881073,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0013010342372581363,
      "neg_sim_mean": 0.4550311267375946,
      "pos_sim_mean": 0.887562096118927,
      "step": 3730
    },
    {
      "epoch": 0.7462,
      "grad_norm": 0.4020695090293884,
      "learning_rate": 1.8398534866757454e-05,
      "loss": 0.1137,
      "step": 3731
    },
    {
      "epoch": 0.7462,
      "loss_align": 0.11823630332946777,
      "loss_contrastive": 0.06417262554168701,
      "loss_gen": 0.2540798485279083,
      "neg_sim_mean": 0.8459362983703613,
      "pos_sim_mean": 0.8817636966705322,
      "step": 3731
    },
    {
      "epoch": 0.7464,
      "grad_norm": 1.7695238590240479,
      "learning_rate": 1.837149190434378e-05,
      "loss": 0.38,
      "step": 3732
    },
    {
      "epoch": 0.7464,
      "loss_align": 0.0649867057800293,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22096166014671326,
      "neg_sim_mean": 0.8278571367263794,
      "pos_sim_mean": 0.9350132942199707,
      "step": 3732
    },
    {
      "epoch": 0.7466,
      "grad_norm": 2.4237918853759766,
      "learning_rate": 1.834446435728072e-05,
      "loss": 0.2859,
      "step": 3733
    },
    {
      "epoch": 0.7466,
      "loss_align": 0.12248754501342773,
      "loss_contrastive": 0.027144432067871094,
      "loss_gen": 0.1694832444190979,
      "neg_sim_mean": 0.8046568632125854,
      "pos_sim_mean": 0.8775124549865723,
      "step": 3733
    },
    {
      "epoch": 0.7468,
      "grad_norm": 1.6014909744262695,
      "learning_rate": 1.8317452238741177e-05,
      "loss": 0.2952,
      "step": 3734
    },
    {
      "epoch": 0.7468,
      "loss_align": 0.20536184310913086,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011991928331553936,
      "neg_sim_mean": 0.26223793625831604,
      "pos_sim_mean": 0.7946381568908691,
      "step": 3734
    },
    {
      "epoch": 0.747,
      "grad_norm": 1.2654087543487549,
      "learning_rate": 1.8290455561890528e-05,
      "loss": 0.2066,
      "step": 3735
    },
    {
      "epoch": 0.747,
      "loss_align": 0.1092602014541626,
      "loss_contrastive": 0.12452322244644165,
      "loss_gen": 0.3676563799381256,
      "neg_sim_mean": 0.9152629971504211,
      "pos_sim_mean": 0.8907397985458374,
      "step": 3735
    },
    {
      "epoch": 0.7472,
      "grad_norm": 2.450153350830078,
      "learning_rate": 1.8263474339886628e-05,
      "loss": 0.4919,
      "step": 3736
    },
    {
      "epoch": 0.7472,
      "loss_align": 0.11954700946807861,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0037930572871118784,
      "neg_sim_mean": 0.20430058240890503,
      "pos_sim_mean": 0.8804529905319214,
      "step": 3736
    },
    {
      "epoch": 0.7474,
      "grad_norm": 0.6372732520103455,
      "learning_rate": 1.823650858587978e-05,
      "loss": 0.1233,
      "step": 3737
    },
    {
      "epoch": 0.7474,
      "loss_align": 0.15176397562026978,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0010381921892985702,
      "neg_sim_mean": 0.25234052538871765,
      "pos_sim_mean": 0.8482360243797302,
      "step": 3737
    },
    {
      "epoch": 0.7476,
      "grad_norm": 0.5471977591514587,
      "learning_rate": 1.8209558313012794e-05,
      "loss": 0.1528,
      "step": 3738
    },
    {
      "epoch": 0.7476,
      "loss_align": 0.12472796440124512,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20574097335338593,
      "neg_sim_mean": 0.6852880716323853,
      "pos_sim_mean": 0.8752720355987549,
      "step": 3738
    },
    {
      "epoch": 0.7478,
      "grad_norm": 1.7130461931228638,
      "learning_rate": 1.8182623534420907e-05,
      "loss": 0.3305,
      "step": 3739
    },
    {
      "epoch": 0.7478,
      "loss_align": 0.1450190544128418,
      "loss_contrastive": 0.056974172592163086,
      "loss_gen": 0.2057488113641739,
      "neg_sim_mean": 0.8119550943374634,
      "pos_sim_mean": 0.8549809455871582,
      "step": 3739
    },
    {
      "epoch": 0.748,
      "grad_norm": 1.7730271816253662,
      "learning_rate": 1.8155704263231776e-05,
      "loss": 0.3576,
      "step": 3740
    },
    {
      "epoch": 0.748,
      "loss_align": 0.10706448554992676,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011492943158373237,
      "neg_sim_mean": 0.4005323052406311,
      "pos_sim_mean": 0.8929355144500732,
      "step": 3740
    },
    {
      "epoch": 0.7482,
      "grad_norm": 0.4017960727214813,
      "learning_rate": 1.8128800512565513e-05,
      "loss": 0.1082,
      "step": 3741
    },
    {
      "epoch": 0.7482,
      "loss_align": 0.07040965557098389,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12890838086605072,
      "neg_sim_mean": 0.6882588863372803,
      "pos_sim_mean": 0.9295903444290161,
      "step": 3741
    },
    {
      "epoch": 0.7484,
      "grad_norm": 1.441953182220459,
      "learning_rate": 1.8101912295534728e-05,
      "loss": 0.1993,
      "step": 3742
    },
    {
      "epoch": 0.7484,
      "loss_align": 0.1345205307006836,
      "loss_contrastive": 0.06734812259674072,
      "loss_gen": 0.1293361783027649,
      "neg_sim_mean": 0.8328275680541992,
      "pos_sim_mean": 0.8654794692993164,
      "step": 3742
    },
    {
      "epoch": 0.7486,
      "grad_norm": 1.7375887632369995,
      "learning_rate": 1.807503962524439e-05,
      "loss": 0.2719,
      "step": 3743
    },
    {
      "epoch": 0.7486,
      "loss_align": 0.11507207155227661,
      "loss_contrastive": 0.043508946895599365,
      "loss_gen": 0.2318553328514099,
      "neg_sim_mean": 0.8284368515014648,
      "pos_sim_mean": 0.8849279284477234,
      "step": 3743
    },
    {
      "epoch": 0.7488,
      "grad_norm": 1.6327533721923828,
      "learning_rate": 1.80481825147919e-05,
      "loss": 0.3521,
      "step": 3744
    },
    {
      "epoch": 0.7488,
      "loss_align": 0.12761199474334717,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0009434142266400158,
      "neg_sim_mean": 0.4301048815250397,
      "pos_sim_mean": 0.8723880052566528,
      "step": 3744
    },
    {
      "epoch": 0.749,
      "grad_norm": 0.6517345309257507,
      "learning_rate": 1.8021340977267105e-05,
      "loss": 0.1286,
      "step": 3745
    },
    {
      "epoch": 0.749,
      "loss_align": 0.08163964748382568,
      "loss_contrastive": 0.017911076545715332,
      "loss_gen": 0.13302068412303925,
      "neg_sim_mean": 0.8362714052200317,
      "pos_sim_mean": 0.9183603525161743,
      "step": 3745
    },
    {
      "epoch": 0.7492,
      "grad_norm": 1.4986001253128052,
      "learning_rate": 1.7994515025752217e-05,
      "loss": 0.2168,
      "step": 3746
    },
    {
      "epoch": 0.7492,
      "loss_align": 0.12831640243530273,
      "loss_contrastive": 0.10459375381469727,
      "loss_gen": 0.21458423137664795,
      "neg_sim_mean": 0.8762773275375366,
      "pos_sim_mean": 0.8716835975646973,
      "step": 3746
    },
    {
      "epoch": 0.7494,
      "grad_norm": 1.6253575086593628,
      "learning_rate": 1.7967704673321918e-05,
      "loss": 0.3555,
      "step": 3747
    },
    {
      "epoch": 0.7494,
      "loss_align": 0.34910881519317627,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23619802296161652,
      "neg_sim_mean": 0.5390329360961914,
      "pos_sim_mean": 0.6508911848068237,
      "step": 3747
    },
    {
      "epoch": 0.7496,
      "grad_norm": 2.4602200984954834,
      "learning_rate": 1.7940909933043242e-05,
      "loss": 0.5853,
      "step": 3748
    },
    {
      "epoch": 0.7496,
      "loss_align": 0.11814498901367188,
      "loss_contrastive": 0.012873411178588867,
      "loss_gen": 0.46162307262420654,
      "neg_sim_mean": 0.7947283983230591,
      "pos_sim_mean": 0.8818550109863281,
      "step": 3748
    },
    {
      "epoch": 0.7498,
      "grad_norm": 2.494636297225952,
      "learning_rate": 1.7914130817975594e-05,
      "loss": 0.5813,
      "step": 3749
    },
    {
      "epoch": 0.7498,
      "loss_align": 0.11051082611083984,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1902904063463211,
      "neg_sim_mean": 0.44971132278442383,
      "pos_sim_mean": 0.8894891738891602,
      "step": 3749
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.9183820486068726,
      "learning_rate": 1.788736734117078e-05,
      "loss": 0.3008,
      "step": 3750
    },
    {
      "epoch": 0.75,
      "loss_align": 0.15962493419647217,
      "loss_contrastive": 0.12959414720535278,
      "loss_gen": 0.19990912079811096,
      "neg_sim_mean": 0.8699691891670227,
      "pos_sim_mean": 0.8403750658035278,
      "step": 3750
    },
    {
      "epoch": 0.7502,
      "grad_norm": 1.7710683345794678,
      "learning_rate": 1.7860619515673033e-05,
      "loss": 0.3751,
      "step": 3751
    },
    {
      "epoch": 0.7502,
      "loss_align": 0.10458743572235107,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003748113289475441,
      "neg_sim_mean": 0.34446489810943604,
      "pos_sim_mean": 0.8954125642776489,
      "step": 3751
    },
    {
      "epoch": 0.7504,
      "grad_norm": 0.4263855814933777,
      "learning_rate": 1.78338873545189e-05,
      "loss": 0.1083,
      "step": 3752
    },
    {
      "epoch": 0.7504,
      "loss_align": 0.09414398670196533,
      "loss_contrastive": 0.08183848857879639,
      "loss_gen": 0.1649811863899231,
      "neg_sim_mean": 0.8876944780349731,
      "pos_sim_mean": 0.9058560132980347,
      "step": 3752
    },
    {
      "epoch": 0.7506,
      "grad_norm": 1.8275789022445679,
      "learning_rate": 1.7807170870737316e-05,
      "loss": 0.2689,
      "step": 3753
    },
    {
      "epoch": 0.7506,
      "loss_align": 0.06010448932647705,
      "loss_contrastive": 0.08225250244140625,
      "loss_gen": 0.17996695637702942,
      "neg_sim_mean": 0.9221479892730713,
      "pos_sim_mean": 0.939895510673523,
      "step": 3753
    },
    {
      "epoch": 0.7508,
      "grad_norm": 1.5699275732040405,
      "learning_rate": 1.7780470077349566e-05,
      "loss": 0.2499,
      "step": 3754
    },
    {
      "epoch": 0.7508,
      "loss_align": 0.10129296779632568,
      "loss_contrastive": 0.022729992866516113,
      "loss_gen": 0.1453993171453476,
      "neg_sim_mean": 0.8214370012283325,
      "pos_sim_mean": 0.8987070322036743,
      "step": 3754
    },
    {
      "epoch": 0.751,
      "grad_norm": 1.7533901929855347,
      "learning_rate": 1.7753784987369287e-05,
      "loss": 0.2494,
      "step": 3755
    },
    {
      "epoch": 0.751,
      "loss_align": 0.18965399265289307,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009838675148785114,
      "neg_sim_mean": 0.34700679779052734,
      "pos_sim_mean": 0.8103460073471069,
      "step": 3755
    },
    {
      "epoch": 0.7512,
      "grad_norm": 0.7397071719169617,
      "learning_rate": 1.7727115613802465e-05,
      "loss": 0.1995,
      "step": 3756
    },
    {
      "epoch": 0.7512,
      "loss_align": 0.10781663656234741,
      "loss_contrastive": 0.04497736692428589,
      "loss_gen": 0.20451560616493225,
      "neg_sim_mean": 0.8371607065200806,
      "pos_sim_mean": 0.8921833634376526,
      "step": 3756
    },
    {
      "epoch": 0.7514,
      "grad_norm": 1.9606088399887085,
      "learning_rate": 1.770046196964747e-05,
      "loss": 0.3177,
      "step": 3757
    },
    {
      "epoch": 0.7514,
      "loss_align": 0.08939123153686523,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3231511116027832,
      "neg_sim_mean": 0.6981031894683838,
      "pos_sim_mean": 0.9106087684631348,
      "step": 3757
    },
    {
      "epoch": 0.7516,
      "grad_norm": 2.0263302326202393,
      "learning_rate": 1.767382406789491e-05,
      "loss": 0.4125,
      "step": 3758
    },
    {
      "epoch": 0.7516,
      "loss_align": 0.07547259330749512,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2304784059524536,
      "neg_sim_mean": 0.7302095890045166,
      "pos_sim_mean": 0.9245274066925049,
      "step": 3758
    },
    {
      "epoch": 0.7518,
      "grad_norm": 2.037224531173706,
      "learning_rate": 1.76472019215278e-05,
      "loss": 0.306,
      "step": 3759
    },
    {
      "epoch": 0.7518,
      "loss_align": 0.10154217481613159,
      "loss_contrastive": 0.029242396354675293,
      "loss_gen": 0.20590321719646454,
      "neg_sim_mean": 0.8277001976966858,
      "pos_sim_mean": 0.8984578251838684,
      "step": 3759
    },
    {
      "epoch": 0.752,
      "grad_norm": 1.4654245376586914,
      "learning_rate": 1.762059554352143e-05,
      "loss": 0.311,
      "step": 3760
    },
    {
      "epoch": 0.752,
      "loss_align": 0.07353830337524414,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09098716080188751,
      "neg_sim_mean": 0.43889373540878296,
      "pos_sim_mean": 0.9264616966247559,
      "step": 3760
    },
    {
      "epoch": 0.7522,
      "grad_norm": 1.3667058944702148,
      "learning_rate": 1.7594004946843456e-05,
      "loss": 0.1645,
      "step": 3761
    },
    {
      "epoch": 0.7522,
      "loss_align": 0.07776886224746704,
      "loss_contrastive": 0.006423652172088623,
      "loss_gen": 0.3847631812095642,
      "neg_sim_mean": 0.8286547660827637,
      "pos_sim_mean": 0.922231137752533,
      "step": 3761
    },
    {
      "epoch": 0.7524,
      "grad_norm": 2.2059013843536377,
      "learning_rate": 1.75674301444538e-05,
      "loss": 0.4633,
      "step": 3762
    },
    {
      "epoch": 0.7524,
      "loss_align": 0.17258769273757935,
      "loss_contrastive": 0.026553571224212646,
      "loss_gen": 0.19891361892223358,
      "neg_sim_mean": 0.7539658546447754,
      "pos_sim_mean": 0.8274123072624207,
      "step": 3762
    },
    {
      "epoch": 0.7526,
      "grad_norm": 2.0469095706939697,
      "learning_rate": 1.75408711493047e-05,
      "loss": 0.3747,
      "step": 3763
    },
    {
      "epoch": 0.7526,
      "loss_align": 0.1345958113670349,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0005796464392915368,
      "neg_sim_mean": 0.37899476289749146,
      "pos_sim_mean": 0.8654041886329651,
      "step": 3763
    },
    {
      "epoch": 0.7528,
      "grad_norm": 0.6122547388076782,
      "learning_rate": 1.7514327974340678e-05,
      "loss": 0.1352,
      "step": 3764
    },
    {
      "epoch": 0.7528,
      "loss_align": 0.15005052089691162,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26158782839775085,
      "neg_sim_mean": 0.6859198212623596,
      "pos_sim_mean": 0.8499494791030884,
      "step": 3764
    },
    {
      "epoch": 0.753,
      "grad_norm": 1.811453104019165,
      "learning_rate": 1.7487800632498545e-05,
      "loss": 0.4116,
      "step": 3765
    },
    {
      "epoch": 0.753,
      "loss_align": 0.09916329383850098,
      "loss_contrastive": 0.0762183666229248,
      "loss_gen": 0.21988634765148163,
      "neg_sim_mean": 0.8770550489425659,
      "pos_sim_mean": 0.900836706161499,
      "step": 3765
    },
    {
      "epoch": 0.7532,
      "grad_norm": 1.9031685590744019,
      "learning_rate": 1.746128913670746e-05,
      "loss": 0.3282,
      "step": 3766
    },
    {
      "epoch": 0.7532,
      "loss_align": 0.05491781234741211,
      "loss_contrastive": 0.03704017400741577,
      "loss_gen": 0.2880081832408905,
      "neg_sim_mean": 0.8821223378181458,
      "pos_sim_mean": 0.9450821876525879,
      "step": 3766
    },
    {
      "epoch": 0.7534,
      "grad_norm": 1.8986282348632812,
      "learning_rate": 1.7434793499888747e-05,
      "loss": 0.3474,
      "step": 3767
    },
    {
      "epoch": 0.7534,
      "loss_align": 0.06300544738769531,
      "loss_contrastive": 0.08912181854248047,
      "loss_gen": 0.15519601106643677,
      "neg_sim_mean": 0.9261163473129272,
      "pos_sim_mean": 0.9369945526123047,
      "step": 3767
    },
    {
      "epoch": 0.7536,
      "grad_norm": 1.5261746644973755,
      "learning_rate": 1.740831373495607e-05,
      "loss": 0.2289,
      "step": 3768
    },
    {
      "epoch": 0.7536,
      "loss_align": 0.08589589595794678,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3207349181175232,
      "neg_sim_mean": 0.7718465328216553,
      "pos_sim_mean": 0.9141041040420532,
      "step": 3768
    },
    {
      "epoch": 0.7538,
      "grad_norm": 2.0558388233184814,
      "learning_rate": 1.7381849854815358e-05,
      "loss": 0.4066,
      "step": 3769
    },
    {
      "epoch": 0.7538,
      "loss_align": 0.05090361833572388,
      "loss_contrastive": 0.06587165594100952,
      "loss_gen": 0.3005145192146301,
      "neg_sim_mean": 0.9149680137634277,
      "pos_sim_mean": 0.9490963816642761,
      "step": 3769
    },
    {
      "epoch": 0.754,
      "grad_norm": 1.840667486190796,
      "learning_rate": 1.7355401872364758e-05,
      "loss": 0.3593,
      "step": 3770
    },
    {
      "epoch": 0.754,
      "loss_align": 0.08350622653961182,
      "loss_contrastive": 0.0,
      "loss_gen": 0.28644421696662903,
      "neg_sim_mean": 0.6375567317008972,
      "pos_sim_mean": 0.9164937734603882,
      "step": 3770
    },
    {
      "epoch": 0.7542,
      "grad_norm": 2.1093742847442627,
      "learning_rate": 1.7328969800494726e-05,
      "loss": 0.37,
      "step": 3771
    },
    {
      "epoch": 0.7542,
      "loss_align": 0.0729404091835022,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18482725322246552,
      "neg_sim_mean": 0.7314654588699341,
      "pos_sim_mean": 0.9270595908164978,
      "step": 3771
    },
    {
      "epoch": 0.7544,
      "grad_norm": 1.812638282775879,
      "learning_rate": 1.7302553652087926e-05,
      "loss": 0.2578,
      "step": 3772
    },
    {
      "epoch": 0.7544,
      "loss_align": 0.10673654079437256,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15997442603111267,
      "neg_sim_mean": 0.6709950566291809,
      "pos_sim_mean": 0.8932634592056274,
      "step": 3772
    },
    {
      "epoch": 0.7546,
      "grad_norm": 1.5406625270843506,
      "learning_rate": 1.727615344001926e-05,
      "loss": 0.2667,
      "step": 3773
    },
    {
      "epoch": 0.7546,
      "loss_align": 0.08964109420776367,
      "loss_contrastive": 0.06972062587738037,
      "loss_gen": 0.20996546745300293,
      "neg_sim_mean": 0.8800795078277588,
      "pos_sim_mean": 0.9103589057922363,
      "step": 3773
    },
    {
      "epoch": 0.7548,
      "grad_norm": 1.9547574520111084,
      "learning_rate": 1.7249769177155878e-05,
      "loss": 0.308,
      "step": 3774
    },
    {
      "epoch": 0.7548,
      "loss_align": 0.10167515277862549,
      "loss_contrastive": 0.09503936767578125,
      "loss_gen": 0.14406874775886536,
      "neg_sim_mean": 0.8933641910552979,
      "pos_sim_mean": 0.8983248472213745,
      "step": 3774
    },
    {
      "epoch": 0.755,
      "grad_norm": 1.5625661611557007,
      "learning_rate": 1.722340087635714e-05,
      "loss": 0.2571,
      "step": 3775
    },
    {
      "epoch": 0.755,
      "loss_align": 0.08214688301086426,
      "loss_contrastive": 0.053882479667663574,
      "loss_gen": 0.23182250559329987,
      "neg_sim_mean": 0.8717355728149414,
      "pos_sim_mean": 0.9178531169891357,
      "step": 3775
    },
    {
      "epoch": 0.7552,
      "grad_norm": 1.796238899230957,
      "learning_rate": 1.7197048550474643e-05,
      "loss": 0.3204,
      "step": 3776
    },
    {
      "epoch": 0.7552,
      "loss_align": 0.06231188774108887,
      "loss_contrastive": 0.04570317268371582,
      "loss_gen": 0.10252003371715546,
      "neg_sim_mean": 0.883391261100769,
      "pos_sim_mean": 0.9376881122589111,
      "step": 3776
    },
    {
      "epoch": 0.7554,
      "grad_norm": 1.1798741817474365,
      "learning_rate": 1.7170712212352187e-05,
      "loss": 0.1703,
      "step": 3777
    },
    {
      "epoch": 0.7554,
      "loss_align": 0.04728424549102783,
      "loss_contrastive": 0.010821938514709473,
      "loss_gen": 0.23559539020061493,
      "neg_sim_mean": 0.8635376691818237,
      "pos_sim_mean": 0.9527157545089722,
      "step": 3777
    },
    {
      "epoch": 0.7556,
      "grad_norm": 2.1287336349487305,
      "learning_rate": 1.7144391874825782e-05,
      "loss": 0.2842,
      "step": 3778
    },
    {
      "epoch": 0.7556,
      "loss_align": 0.12107813358306885,
      "loss_contrastive": 0.11210864782333374,
      "loss_gen": 0.4062701463699341,
      "neg_sim_mean": 0.891030490398407,
      "pos_sim_mean": 0.8789218664169312,
      "step": 3778
    },
    {
      "epoch": 0.7558,
      "grad_norm": 2.137545585632324,
      "learning_rate": 1.7118087550723633e-05,
      "loss": 0.5408,
      "step": 3779
    },
    {
      "epoch": 0.7558,
      "loss_align": 0.06280064582824707,
      "loss_contrastive": 0.05301809310913086,
      "loss_gen": 0.24260498583316803,
      "neg_sim_mean": 0.8902174234390259,
      "pos_sim_mean": 0.9371993541717529,
      "step": 3779
    },
    {
      "epoch": 0.756,
      "grad_norm": 1.7507734298706055,
      "learning_rate": 1.709179925286617e-05,
      "loss": 0.3118,
      "step": 3780
    },
    {
      "epoch": 0.756,
      "loss_align": 0.1479862928390503,
      "loss_contrastive": 0.05605965852737427,
      "loss_gen": 0.46753910183906555,
      "neg_sim_mean": 0.8080733418464661,
      "pos_sim_mean": 0.8520137071609497,
      "step": 3780
    },
    {
      "epoch": 0.7562,
      "grad_norm": 2.5818798542022705,
      "learning_rate": 1.7065526994065973e-05,
      "loss": 0.6223,
      "step": 3781
    },
    {
      "epoch": 0.7562,
      "loss_align": 0.07181864976882935,
      "loss_contrastive": 0.09636670351028442,
      "loss_gen": 0.2544485330581665,
      "neg_sim_mean": 0.9245480298995972,
      "pos_sim_mean": 0.9281813502311707,
      "step": 3781
    },
    {
      "epoch": 0.7564,
      "grad_norm": 1.7684732675552368,
      "learning_rate": 1.7039270787127832e-05,
      "loss": 0.3378,
      "step": 3782
    },
    {
      "epoch": 0.7564,
      "loss_align": 0.1310511827468872,
      "loss_contrastive": 0.0,
      "loss_gen": 0.028397776186466217,
      "neg_sim_mean": 0.1936642825603485,
      "pos_sim_mean": 0.8689488172531128,
      "step": 3782
    },
    {
      "epoch": 0.7566,
      "grad_norm": 1.0068912506103516,
      "learning_rate": 1.70130306448487e-05,
      "loss": 0.1594,
      "step": 3783
    },
    {
      "epoch": 0.7566,
      "loss_align": 0.1446303129196167,
      "loss_contrastive": 0.0,
      "loss_gen": 0.31277433037757874,
      "neg_sim_mean": 0.7505919337272644,
      "pos_sim_mean": 0.8553696870803833,
      "step": 3783
    },
    {
      "epoch": 0.7568,
      "grad_norm": 2.1735212802886963,
      "learning_rate": 1.6986806580017694e-05,
      "loss": 0.4574,
      "step": 3784
    },
    {
      "epoch": 0.7568,
      "loss_align": 0.10359549522399902,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21721012890338898,
      "neg_sim_mean": 0.6708730459213257,
      "pos_sim_mean": 0.896404504776001,
      "step": 3784
    },
    {
      "epoch": 0.757,
      "grad_norm": 1.9630566835403442,
      "learning_rate": 1.6960598605416117e-05,
      "loss": 0.3208,
      "step": 3785
    },
    {
      "epoch": 0.757,
      "loss_align": 0.12147116661071777,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2161942422389984,
      "neg_sim_mean": 0.5942456722259521,
      "pos_sim_mean": 0.8785288333892822,
      "step": 3785
    },
    {
      "epoch": 0.7572,
      "grad_norm": 1.9740842580795288,
      "learning_rate": 1.6934406733817414e-05,
      "loss": 0.3377,
      "step": 3786
    },
    {
      "epoch": 0.7572,
      "loss_align": 0.15096962451934814,
      "loss_contrastive": 0.07669484615325928,
      "loss_gen": 0.12865686416625977,
      "neg_sim_mean": 0.8257251977920532,
      "pos_sim_mean": 0.8490303754806519,
      "step": 3786
    },
    {
      "epoch": 0.7574,
      "grad_norm": 1.6382248401641846,
      "learning_rate": 1.6908230977987182e-05,
      "loss": 0.2888,
      "step": 3787
    },
    {
      "epoch": 0.7574,
      "loss_align": 0.10770481824874878,
      "loss_contrastive": 0.06764411926269531,
      "loss_gen": 0.2474360167980194,
      "neg_sim_mean": 0.8599392771720886,
      "pos_sim_mean": 0.8922951817512512,
      "step": 3787
    },
    {
      "epoch": 0.7576,
      "grad_norm": 1.8819833993911743,
      "learning_rate": 1.6882071350683166e-05,
      "loss": 0.3633,
      "step": 3788
    },
    {
      "epoch": 0.7576,
      "loss_align": 0.0687519907951355,
      "loss_contrastive": 0.030123353004455566,
      "loss_gen": 0.17893670499324799,
      "neg_sim_mean": 0.8613713383674622,
      "pos_sim_mean": 0.9312480092048645,
      "step": 3788
    },
    {
      "epoch": 0.7578,
      "grad_norm": 1.4653303623199463,
      "learning_rate": 1.685592786465524e-05,
      "loss": 0.2513,
      "step": 3789
    },
    {
      "epoch": 0.7578,
      "loss_align": 0.10531401634216309,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13546910881996155,
      "neg_sim_mean": 0.6123824119567871,
      "pos_sim_mean": 0.8946859836578369,
      "step": 3789
    },
    {
      "epoch": 0.758,
      "grad_norm": 1.5271894931793213,
      "learning_rate": 1.6829800532645447e-05,
      "loss": 0.2408,
      "step": 3790
    },
    {
      "epoch": 0.758,
      "loss_align": 0.0921066403388977,
      "loss_contrastive": 0.07145345211029053,
      "loss_gen": 0.30265912413597107,
      "neg_sim_mean": 0.8793467879295349,
      "pos_sim_mean": 0.9078933596611023,
      "step": 3790
    },
    {
      "epoch": 0.7582,
      "grad_norm": 1.4426155090332031,
      "learning_rate": 1.680368936738792e-05,
      "loss": 0.4033,
      "step": 3791
    },
    {
      "epoch": 0.7582,
      "loss_align": 0.14242219924926758,
      "loss_contrastive": 0.029545068740844727,
      "loss_gen": 0.1825668215751648,
      "neg_sim_mean": 0.7871228456497192,
      "pos_sim_mean": 0.8575778007507324,
      "step": 3791
    },
    {
      "epoch": 0.7584,
      "grad_norm": 1.7127442359924316,
      "learning_rate": 1.6777594381608936e-05,
      "loss": 0.3285,
      "step": 3792
    },
    {
      "epoch": 0.7584,
      "loss_align": 0.1295703649520874,
      "loss_contrastive": 0.09399306774139404,
      "loss_gen": 0.26371529698371887,
      "neg_sim_mean": 0.8644226789474487,
      "pos_sim_mean": 0.8704296350479126,
      "step": 3792
    },
    {
      "epoch": 0.7586,
      "grad_norm": 2.0226008892059326,
      "learning_rate": 1.6751515588026828e-05,
      "loss": 0.4046,
      "step": 3793
    },
    {
      "epoch": 0.7586,
      "loss_align": 0.13576948642730713,
      "loss_contrastive": 0.10764509439468384,
      "loss_gen": 0.12455331534147263,
      "neg_sim_mean": 0.8718755841255188,
      "pos_sim_mean": 0.8642305135726929,
      "step": 3793
    },
    {
      "epoch": 0.7588,
      "grad_norm": 1.6710233688354492,
      "learning_rate": 1.6725452999352138e-05,
      "loss": 0.2732,
      "step": 3794
    },
    {
      "epoch": 0.7588,
      "loss_align": 0.08764922618865967,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007481282576918602,
      "neg_sim_mean": 0.25402119755744934,
      "pos_sim_mean": 0.9123507738113403,
      "step": 3794
    },
    {
      "epoch": 0.759,
      "grad_norm": 0.5269196033477783,
      "learning_rate": 1.669940662828742e-05,
      "loss": 0.0951,
      "step": 3795
    },
    {
      "epoch": 0.759,
      "loss_align": 0.09455347061157227,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18113651871681213,
      "neg_sim_mean": 0.8034203052520752,
      "pos_sim_mean": 0.9054465293884277,
      "step": 3795
    },
    {
      "epoch": 0.7592,
      "grad_norm": 1.418755292892456,
      "learning_rate": 1.667337648752738e-05,
      "loss": 0.2757,
      "step": 3796
    },
    {
      "epoch": 0.7592,
      "loss_align": 0.25138044357299805,
      "loss_contrastive": 0.13283979892730713,
      "loss_gen": 0.17941197752952576,
      "neg_sim_mean": 0.7814593315124512,
      "pos_sim_mean": 0.748619556427002,
      "step": 3796
    },
    {
      "epoch": 0.7594,
      "grad_norm": 1.7906734943389893,
      "learning_rate": 1.6647362589758787e-05,
      "loss": 0.4467,
      "step": 3797
    },
    {
      "epoch": 0.7594,
      "loss_align": 0.1352940797805786,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15803015232086182,
      "neg_sim_mean": 0.6929124593734741,
      "pos_sim_mean": 0.8647059202194214,
      "step": 3797
    },
    {
      "epoch": 0.7596,
      "grad_norm": 1.5949482917785645,
      "learning_rate": 1.6621364947660473e-05,
      "loss": 0.2933,
      "step": 3798
    },
    {
      "epoch": 0.7596,
      "loss_align": 0.03859710693359375,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03475503996014595,
      "neg_sim_mean": 0.32841598987579346,
      "pos_sim_mean": 0.9614028930664062,
      "step": 3798
    },
    {
      "epoch": 0.7598,
      "grad_norm": 1.189405083656311,
      "learning_rate": 1.659538357390341e-05,
      "loss": 0.0734,
      "step": 3799
    },
    {
      "epoch": 0.7598,
      "loss_align": 0.06730204820632935,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25733911991119385,
      "neg_sim_mean": 0.8121468424797058,
      "pos_sim_mean": 0.9326979517936707,
      "step": 3799
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.8687552213668823,
      "learning_rate": 1.6569418481150595e-05,
      "loss": 0.3246,
      "step": 3800
    },
    {
      "epoch": 0.76,
      "loss_align": 0.08054190874099731,
      "loss_contrastive": 0.045364439487457275,
      "loss_gen": 0.21669508516788483,
      "neg_sim_mean": 0.864822506904602,
      "pos_sim_mean": 0.9194580912590027,
      "step": 3800
    },
    {
      "epoch": 0.7602,
      "grad_norm": 1.9540493488311768,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 0.3027,
      "step": 3801
    },
    {
      "epoch": 0.7602,
      "loss_align": 0.2418966293334961,
      "loss_contrastive": 0.04119539260864258,
      "loss_gen": 0.1668219417333603,
      "neg_sim_mean": 0.6992987394332886,
      "pos_sim_mean": 0.7581033706665039,
      "step": 3801
    },
    {
      "epoch": 0.7604,
      "grad_norm": 1.9962776899337769,
      "learning_rate": 1.6517537189270044e-05,
      "loss": 0.4137,
      "step": 3802
    },
    {
      "epoch": 0.7604,
      "loss_align": 0.120516836643219,
      "loss_contrastive": 0.0,
      "loss_gen": 0.452848345041275,
      "neg_sim_mean": 0.7276585102081299,
      "pos_sim_mean": 0.879483163356781,
      "step": 3802
    },
    {
      "epoch": 0.7606,
      "grad_norm": 2.363016366958618,
      "learning_rate": 1.6491621015428587e-05,
      "loss": 0.5734,
      "step": 3803
    },
    {
      "epoch": 0.7606,
      "loss_align": 0.08026432991027832,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23906688392162323,
      "neg_sim_mean": 0.582964301109314,
      "pos_sim_mean": 0.9197356700897217,
      "step": 3803
    },
    {
      "epoch": 0.7608,
      "grad_norm": 1.710074782371521,
      "learning_rate": 1.6465721173164002e-05,
      "loss": 0.3193,
      "step": 3804
    },
    {
      "epoch": 0.7608,
      "loss_align": 0.08266603946685791,
      "loss_contrastive": 0.03322911262512207,
      "loss_gen": 0.2724078297615051,
      "neg_sim_mean": 0.8505630493164062,
      "pos_sim_mean": 0.9173339605331421,
      "step": 3804
    },
    {
      "epoch": 0.761,
      "grad_norm": 1.8379673957824707,
      "learning_rate": 1.643983767509954e-05,
      "loss": 0.3591,
      "step": 3805
    },
    {
      "epoch": 0.761,
      "loss_align": 0.1670771837234497,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12136530876159668,
      "neg_sim_mean": 0.39412742853164673,
      "pos_sim_mean": 0.8329228162765503,
      "step": 3805
    },
    {
      "epoch": 0.7612,
      "grad_norm": 1.6829644441604614,
      "learning_rate": 1.6413970533850498e-05,
      "loss": 0.2884,
      "step": 3806
    },
    {
      "epoch": 0.7612,
      "loss_align": 0.12569820880889893,
      "loss_contrastive": 0.0026711225509643555,
      "loss_gen": 0.2812901437282562,
      "neg_sim_mean": 0.7769728899002075,
      "pos_sim_mean": 0.8743017911911011,
      "step": 3806
    },
    {
      "epoch": 0.7614,
      "grad_norm": 2.3726682662963867,
      "learning_rate": 1.638811976202421e-05,
      "loss": 0.4073,
      "step": 3807
    },
    {
      "epoch": 0.7614,
      "loss_align": 0.0846560001373291,
      "loss_contrastive": 0.0,
      "loss_gen": 0.39596688747406006,
      "neg_sim_mean": 0.6751925349235535,
      "pos_sim_mean": 0.9153439998626709,
      "step": 3807
    },
    {
      "epoch": 0.7616,
      "grad_norm": 2.4937684535980225,
      "learning_rate": 1.6362285372220016e-05,
      "loss": 0.4806,
      "step": 3808
    },
    {
      "epoch": 0.7616,
      "loss_align": 0.18658298254013062,
      "loss_contrastive": 0.15406864881515503,
      "loss_gen": 0.1896950751543045,
      "neg_sim_mean": 0.8674856424331665,
      "pos_sim_mean": 0.8134170174598694,
      "step": 3808
    },
    {
      "epoch": 0.7618,
      "grad_norm": 1.583885908126831,
      "learning_rate": 1.6336467377029308e-05,
      "loss": 0.3948,
      "step": 3809
    },
    {
      "epoch": 0.7618,
      "loss_align": 0.09911590814590454,
      "loss_contrastive": 0.03192669153213501,
      "loss_gen": 0.18324817717075348,
      "neg_sim_mean": 0.8328107595443726,
      "pos_sim_mean": 0.9008840918540955,
      "step": 3809
    },
    {
      "epoch": 0.762,
      "grad_norm": 1.4935839176177979,
      "learning_rate": 1.6310665789035467e-05,
      "loss": 0.2862,
      "step": 3810
    },
    {
      "epoch": 0.762,
      "loss_align": 0.08196282386779785,
      "loss_contrastive": 0.04834723472595215,
      "loss_gen": 0.1927415281534195,
      "neg_sim_mean": 0.8663843870162964,
      "pos_sim_mean": 0.9180371761322021,
      "step": 3810
    },
    {
      "epoch": 0.7622,
      "grad_norm": 1.441598892211914,
      "learning_rate": 1.6284880620813848e-05,
      "loss": 0.2805,
      "step": 3811
    },
    {
      "epoch": 0.7622,
      "loss_align": 0.08367037773132324,
      "loss_contrastive": 0.02395176887512207,
      "loss_gen": 0.3411369025707245,
      "neg_sim_mean": 0.8402813673019409,
      "pos_sim_mean": 0.9163296222686768,
      "step": 3811
    },
    {
      "epoch": 0.7624,
      "grad_norm": 1.9624137878417969,
      "learning_rate": 1.6259111884931815e-05,
      "loss": 0.4277,
      "step": 3812
    },
    {
      "epoch": 0.7624,
      "loss_align": 0.11864298582077026,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20955723524093628,
      "neg_sim_mean": 0.7267211675643921,
      "pos_sim_mean": 0.8813570141792297,
      "step": 3812
    },
    {
      "epoch": 0.7626,
      "grad_norm": 1.4923810958862305,
      "learning_rate": 1.6233359593948777e-05,
      "loss": 0.3282,
      "step": 3813
    },
    {
      "epoch": 0.7626,
      "loss_align": 0.1876436471939087,
      "loss_contrastive": 0.13124704360961914,
      "loss_gen": 0.3359888792037964,
      "neg_sim_mean": 0.8436033725738525,
      "pos_sim_mean": 0.8123563528060913,
      "step": 3813
    },
    {
      "epoch": 0.7628,
      "grad_norm": 2.0995309352874756,
      "learning_rate": 1.6207623760416073e-05,
      "loss": 0.5394,
      "step": 3814
    },
    {
      "epoch": 0.7628,
      "loss_align": 0.16595619916915894,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011044858023524284,
      "neg_sim_mean": 0.42988377809524536,
      "pos_sim_mean": 0.8340438008308411,
      "step": 3814
    },
    {
      "epoch": 0.763,
      "grad_norm": 0.798317015171051,
      "learning_rate": 1.618190439687704e-05,
      "loss": 0.177,
      "step": 3815
    },
    {
      "epoch": 0.763,
      "loss_align": 0.13607239723205566,
      "loss_contrastive": 0.12747204303741455,
      "loss_gen": 0.23744913935661316,
      "neg_sim_mean": 0.891399621963501,
      "pos_sim_mean": 0.8639276027679443,
      "step": 3815
    },
    {
      "epoch": 0.7632,
      "grad_norm": 1.811675786972046,
      "learning_rate": 1.615620151586697e-05,
      "loss": 0.3888,
      "step": 3816
    },
    {
      "epoch": 0.7632,
      "loss_align": 0.05031108856201172,
      "loss_contrastive": 0.004008829593658447,
      "loss_gen": 0.16226381063461304,
      "neg_sim_mean": 0.8536977171897888,
      "pos_sim_mean": 0.9496889114379883,
      "step": 3816
    },
    {
      "epoch": 0.7634,
      "grad_norm": 1.4068396091461182,
      "learning_rate": 1.613051512991314e-05,
      "loss": 0.2131,
      "step": 3817
    },
    {
      "epoch": 0.7634,
      "loss_align": 0.15845227241516113,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25806739926338196,
      "neg_sim_mean": 0.5744668245315552,
      "pos_sim_mean": 0.8415477275848389,
      "step": 3817
    },
    {
      "epoch": 0.7636,
      "grad_norm": 1.6825920343399048,
      "learning_rate": 1.6104845251534774e-05,
      "loss": 0.4165,
      "step": 3818
    },
    {
      "epoch": 0.7636,
      "loss_align": 0.060640692710876465,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001326584373600781,
      "neg_sim_mean": 0.3071044683456421,
      "pos_sim_mean": 0.9393593072891235,
      "step": 3818
    },
    {
      "epoch": 0.7638,
      "grad_norm": 0.3229312598705292,
      "learning_rate": 1.60791918932431e-05,
      "loss": 0.062,
      "step": 3819
    },
    {
      "epoch": 0.7638,
      "loss_align": 0.0737341046333313,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1856003999710083,
      "neg_sim_mean": 0.5393267869949341,
      "pos_sim_mean": 0.9262658953666687,
      "step": 3819
    },
    {
      "epoch": 0.764,
      "grad_norm": 1.5190504789352417,
      "learning_rate": 1.605355506754121e-05,
      "loss": 0.2593,
      "step": 3820
    },
    {
      "epoch": 0.764,
      "loss_align": 0.16660332679748535,
      "loss_contrastive": 0.0,
      "loss_gen": 0.015825364738702774,
      "neg_sim_mean": 0.3101691007614136,
      "pos_sim_mean": 0.8333966732025146,
      "step": 3820
    },
    {
      "epoch": 0.7642,
      "grad_norm": 0.9352576732635498,
      "learning_rate": 1.602793478692419e-05,
      "loss": 0.1824,
      "step": 3821
    },
    {
      "epoch": 0.7642,
      "loss_align": 0.0930752158164978,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17039982974529266,
      "neg_sim_mean": 0.7713475823402405,
      "pos_sim_mean": 0.9069247841835022,
      "step": 3821
    },
    {
      "epoch": 0.7644,
      "grad_norm": 1.457008957862854,
      "learning_rate": 1.600233106387904e-05,
      "loss": 0.2635,
      "step": 3822
    },
    {
      "epoch": 0.7644,
      "loss_align": 0.07484853267669678,
      "loss_contrastive": 0.09280157089233398,
      "loss_gen": 0.37686681747436523,
      "neg_sim_mean": 0.9179530143737793,
      "pos_sim_mean": 0.9251514673233032,
      "step": 3822
    },
    {
      "epoch": 0.7646,
      "grad_norm": 2.1717443466186523,
      "learning_rate": 1.597674391088474e-05,
      "loss": 0.4629,
      "step": 3823
    },
    {
      "epoch": 0.7646,
      "loss_align": 0.10075891017913818,
      "loss_contrastive": 0.08440864086151123,
      "loss_gen": 0.3555505573749542,
      "neg_sim_mean": 0.8836497068405151,
      "pos_sim_mean": 0.8992410898208618,
      "step": 3823
    },
    {
      "epoch": 0.7648,
      "grad_norm": 1.8632938861846924,
      "learning_rate": 1.5951173340412133e-05,
      "loss": 0.4664,
      "step": 3824
    },
    {
      "epoch": 0.7648,
      "loss_align": 0.1305485963821411,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0029541479889303446,
      "neg_sim_mean": 0.43060797452926636,
      "pos_sim_mean": 0.8694514036178589,
      "step": 3824
    },
    {
      "epoch": 0.765,
      "grad_norm": 0.4766358435153961,
      "learning_rate": 1.5925619364924017e-05,
      "loss": 0.1335,
      "step": 3825
    },
    {
      "epoch": 0.765,
      "loss_align": 0.08901786804199219,
      "loss_contrastive": 0.11203241348266602,
      "loss_gen": 0.20807139575481415,
      "neg_sim_mean": 0.9230145215988159,
      "pos_sim_mean": 0.9109821319580078,
      "step": 3825
    },
    {
      "epoch": 0.7652,
      "grad_norm": 1.9601328372955322,
      "learning_rate": 1.5900081996875083e-05,
      "loss": 0.3105,
      "step": 3826
    },
    {
      "epoch": 0.7652,
      "loss_align": 0.0780336856842041,
      "loss_contrastive": 0.0,
      "loss_gen": 0.07973562180995941,
      "neg_sim_mean": 0.8032032251358032,
      "pos_sim_mean": 0.9219663143157959,
      "step": 3826
    },
    {
      "epoch": 0.7654,
      "grad_norm": 1.0677770376205444,
      "learning_rate": 1.587456124871191e-05,
      "loss": 0.1578,
      "step": 3827
    },
    {
      "epoch": 0.7654,
      "loss_align": 0.11678570508956909,
      "loss_contrastive": 0.07122987508773804,
      "loss_gen": 0.09413684904575348,
      "neg_sim_mean": 0.854444146156311,
      "pos_sim_mean": 0.8832142949104309,
      "step": 3827
    },
    {
      "epoch": 0.7656,
      "grad_norm": 1.605096697807312,
      "learning_rate": 1.5849057132873062e-05,
      "loss": 0.2195,
      "step": 3828
    },
    {
      "epoch": 0.7656,
      "loss_align": 0.07728976011276245,
      "loss_contrastive": 0.04492312669754028,
      "loss_gen": 0.15183185040950775,
      "neg_sim_mean": 0.8676333427429199,
      "pos_sim_mean": 0.9227102398872375,
      "step": 3828
    },
    {
      "epoch": 0.7658,
      "grad_norm": 1.8255008459091187,
      "learning_rate": 1.582356966178888e-05,
      "loss": 0.2345,
      "step": 3829
    },
    {
      "epoch": 0.7658,
      "loss_align": 0.099906325340271,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001457012491300702,
      "neg_sim_mean": 0.3712715804576874,
      "pos_sim_mean": 0.900093674659729,
      "step": 3829
    },
    {
      "epoch": 0.766,
      "grad_norm": 0.4533233940601349,
      "learning_rate": 1.5798098847881666e-05,
      "loss": 0.1014,
      "step": 3830
    },
    {
      "epoch": 0.766,
      "loss_align": 0.0838814377784729,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24494212865829468,
      "neg_sim_mean": 0.762688159942627,
      "pos_sim_mean": 0.9161185622215271,
      "step": 3830
    },
    {
      "epoch": 0.7662,
      "grad_norm": 1.625876784324646,
      "learning_rate": 1.5772644703565565e-05,
      "loss": 0.3288,
      "step": 3831
    },
    {
      "epoch": 0.7662,
      "loss_align": 0.10130304098129272,
      "loss_contrastive": 0.03784739971160889,
      "loss_gen": 0.23516248166561127,
      "neg_sim_mean": 0.8365443348884583,
      "pos_sim_mean": 0.8986969590187073,
      "step": 3831
    },
    {
      "epoch": 0.7664,
      "grad_norm": 1.8718817234039307,
      "learning_rate": 1.5747207241246652e-05,
      "loss": 0.341,
      "step": 3832
    },
    {
      "epoch": 0.7664,
      "loss_align": 0.11543828248977661,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012860167771577835,
      "neg_sim_mean": 0.633520781993866,
      "pos_sim_mean": 0.8845617175102234,
      "step": 3832
    },
    {
      "epoch": 0.7666,
      "grad_norm": 0.7312657833099365,
      "learning_rate": 1.5721786473322825e-05,
      "loss": 0.1283,
      "step": 3833
    },
    {
      "epoch": 0.7666,
      "loss_align": 0.2248934507369995,
      "loss_contrastive": 0.03236353397369385,
      "loss_gen": 0.2956223487854004,
      "neg_sim_mean": 0.7074700593948364,
      "pos_sim_mean": 0.7751065492630005,
      "step": 3833
    },
    {
      "epoch": 0.7668,
      "grad_norm": 2.042665719985962,
      "learning_rate": 1.5696382412183852e-05,
      "loss": 0.5244,
      "step": 3834
    },
    {
      "epoch": 0.7668,
      "loss_align": 0.09362286329269409,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20688214898109436,
      "neg_sim_mean": 0.7433910369873047,
      "pos_sim_mean": 0.9063771367073059,
      "step": 3834
    },
    {
      "epoch": 0.767,
      "grad_norm": 1.639872431755066,
      "learning_rate": 1.567099507021137e-05,
      "loss": 0.3005,
      "step": 3835
    },
    {
      "epoch": 0.767,
      "loss_align": 0.2597876787185669,
      "loss_contrastive": 0.18968641757965088,
      "loss_gen": 0.11418227106332779,
      "neg_sim_mean": 0.8298987150192261,
      "pos_sim_mean": 0.7402123212814331,
      "step": 3835
    },
    {
      "epoch": 0.7672,
      "grad_norm": 1.4226638078689575,
      "learning_rate": 1.5645624459778856e-05,
      "loss": 0.3967,
      "step": 3836
    },
    {
      "epoch": 0.7672,
      "loss_align": 0.1722351312637329,
      "loss_contrastive": 0.0,
      "loss_gen": 0.32722729444503784,
      "neg_sim_mean": 0.25638526678085327,
      "pos_sim_mean": 0.8277648687362671,
      "step": 3836
    },
    {
      "epoch": 0.7674,
      "grad_norm": 1.9984885454177856,
      "learning_rate": 1.5620270593251635e-05,
      "loss": 0.4995,
      "step": 3837
    },
    {
      "epoch": 0.7674,
      "loss_align": 0.12006211280822754,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24289272725582123,
      "neg_sim_mean": 0.7660133242607117,
      "pos_sim_mean": 0.8799378871917725,
      "step": 3837
    },
    {
      "epoch": 0.7676,
      "grad_norm": 1.7377887964248657,
      "learning_rate": 1.5594933482986883e-05,
      "loss": 0.363,
      "step": 3838
    },
    {
      "epoch": 0.7676,
      "loss_align": 0.12749207019805908,
      "loss_contrastive": 0.0621107816696167,
      "loss_gen": 0.37798818945884705,
      "neg_sim_mean": 0.8346186876296997,
      "pos_sim_mean": 0.8725079298019409,
      "step": 3838
    },
    {
      "epoch": 0.7678,
      "grad_norm": 2.065993070602417,
      "learning_rate": 1.556961314133359e-05,
      "loss": 0.5129,
      "step": 3839
    },
    {
      "epoch": 0.7678,
      "loss_align": 0.06149870157241821,
      "loss_contrastive": 0.0,
      "loss_gen": 0.020775720477104187,
      "neg_sim_mean": 0.341189980506897,
      "pos_sim_mean": 0.9385012984275818,
      "step": 3839
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.9342060685157776,
      "learning_rate": 1.554430958063259e-05,
      "loss": 0.0823,
      "step": 3840
    },
    {
      "epoch": 0.768,
      "loss_align": 0.1762709617614746,
      "loss_contrastive": 0.11503982543945312,
      "loss_gen": 0.24658203125,
      "neg_sim_mean": 0.8387688398361206,
      "pos_sim_mean": 0.8237290382385254,
      "step": 3840
    },
    {
      "epoch": 0.7682,
      "grad_norm": 2.055704116821289,
      "learning_rate": 1.551902281321651e-05,
      "loss": 0.4367,
      "step": 3841
    },
    {
      "epoch": 0.7682,
      "loss_align": 0.12281084060668945,
      "loss_contrastive": 0.07942038774490356,
      "loss_gen": 0.1770865023136139,
      "neg_sim_mean": 0.8566095232963562,
      "pos_sim_mean": 0.8771891593933105,
      "step": 3841
    },
    {
      "epoch": 0.7684,
      "grad_norm": 1.7309560775756836,
      "learning_rate": 1.5493752851409844e-05,
      "loss": 0.3094,
      "step": 3842
    },
    {
      "epoch": 0.7684,
      "loss_align": 0.1777052879333496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2534567713737488,
      "neg_sim_mean": 0.719499945640564,
      "pos_sim_mean": 0.8222947120666504,
      "step": 3842
    },
    {
      "epoch": 0.7686,
      "grad_norm": 1.8288016319274902,
      "learning_rate": 1.5468499707528854e-05,
      "loss": 0.4312,
      "step": 3843
    },
    {
      "epoch": 0.7686,
      "loss_align": 0.08659142255783081,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2950283885002136,
      "neg_sim_mean": 0.6744851469993591,
      "pos_sim_mean": 0.9134085774421692,
      "step": 3843
    },
    {
      "epoch": 0.7688,
      "grad_norm": 2.1589643955230713,
      "learning_rate": 1.5443263393881617e-05,
      "loss": 0.3816,
      "step": 3844
    },
    {
      "epoch": 0.7688,
      "loss_align": 0.10349857807159424,
      "loss_contrastive": 0.05094635486602783,
      "loss_gen": 0.1158999353647232,
      "neg_sim_mean": 0.8474477529525757,
      "pos_sim_mean": 0.8965014219284058,
      "step": 3844
    },
    {
      "epoch": 0.769,
      "grad_norm": 1.327307939529419,
      "learning_rate": 1.5418043922768e-05,
      "loss": 0.2255,
      "step": 3845
    },
    {
      "epoch": 0.769,
      "loss_align": 0.11687368154525757,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005936918780207634,
      "neg_sim_mean": 0.31550508737564087,
      "pos_sim_mean": 0.8831263184547424,
      "step": 3845
    },
    {
      "epoch": 0.7692,
      "grad_norm": 0.612533450126648,
      "learning_rate": 1.5392841306479666e-05,
      "loss": 0.1228,
      "step": 3846
    },
    {
      "epoch": 0.7692,
      "loss_align": 0.1217467188835144,
      "loss_contrastive": 0.059742271900177,
      "loss_gen": 0.17292732000350952,
      "neg_sim_mean": 0.8379955291748047,
      "pos_sim_mean": 0.8782532811164856,
      "step": 3846
    },
    {
      "epoch": 0.7694,
      "grad_norm": 1.6306345462799072,
      "learning_rate": 1.5367655557300065e-05,
      "loss": 0.3018,
      "step": 3847
    },
    {
      "epoch": 0.7694,
      "loss_align": 0.19240665435791016,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010724401101469994,
      "neg_sim_mean": 0.34941908717155457,
      "pos_sim_mean": 0.8075933456420898,
      "step": 3847
    },
    {
      "epoch": 0.7696,
      "grad_norm": 0.8540608882904053,
      "learning_rate": 1.534248668750443e-05,
      "loss": 0.2031,
      "step": 3848
    },
    {
      "epoch": 0.7696,
      "loss_align": 0.047219038009643555,
      "loss_contrastive": 0.007435202598571777,
      "loss_gen": 0.2127908319234848,
      "neg_sim_mean": 0.8602161407470703,
      "pos_sim_mean": 0.9527809619903564,
      "step": 3848
    },
    {
      "epoch": 0.7698,
      "grad_norm": 1.6919323205947876,
      "learning_rate": 1.531733470935976e-05,
      "loss": 0.2609,
      "step": 3849
    },
    {
      "epoch": 0.7698,
      "loss_align": 0.21289336681365967,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22470727562904358,
      "neg_sim_mean": 0.6095129251480103,
      "pos_sim_mean": 0.7871066331863403,
      "step": 3849
    },
    {
      "epoch": 0.77,
      "grad_norm": 2.193540573120117,
      "learning_rate": 1.529219963512481e-05,
      "loss": 0.4376,
      "step": 3850
    },
    {
      "epoch": 0.77,
      "loss_align": 0.08425873517990112,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1661829948425293,
      "neg_sim_mean": 0.7597168684005737,
      "pos_sim_mean": 0.9157412648200989,
      "step": 3850
    },
    {
      "epoch": 0.7702,
      "grad_norm": 1.4921300411224365,
      "learning_rate": 1.526708147705013e-05,
      "loss": 0.2504,
      "step": 3851
    },
    {
      "epoch": 0.7702,
      "loss_align": 0.09372258186340332,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09327264130115509,
      "neg_sim_mean": 0.7448102235794067,
      "pos_sim_mean": 0.9062774181365967,
      "step": 3851
    },
    {
      "epoch": 0.7704,
      "grad_norm": 1.2534494400024414,
      "learning_rate": 1.5241980247378008e-05,
      "loss": 0.187,
      "step": 3852
    },
    {
      "epoch": 0.7704,
      "loss_align": 0.21355831623077393,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3216644823551178,
      "neg_sim_mean": 0.5089995861053467,
      "pos_sim_mean": 0.7864416837692261,
      "step": 3852
    },
    {
      "epoch": 0.7706,
      "grad_norm": 2.5324900150299072,
      "learning_rate": 1.5216895958342458e-05,
      "loss": 0.5352,
      "step": 3853
    },
    {
      "epoch": 0.7706,
      "loss_align": 0.09663254022598267,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0770561695098877,
      "neg_sim_mean": 0.7543349266052246,
      "pos_sim_mean": 0.9033674597740173,
      "step": 3853
    },
    {
      "epoch": 0.7708,
      "grad_norm": 1.074099063873291,
      "learning_rate": 1.519182862216929e-05,
      "loss": 0.1737,
      "step": 3854
    },
    {
      "epoch": 0.7708,
      "loss_align": 0.11422085762023926,
      "loss_contrastive": 0.05733007192611694,
      "loss_gen": 0.5328923463821411,
      "neg_sim_mean": 0.8431091904640198,
      "pos_sim_mean": 0.8857791423797607,
      "step": 3854
    },
    {
      "epoch": 0.771,
      "grad_norm": 2.4537789821624756,
      "learning_rate": 1.5166778251075963e-05,
      "loss": 0.654,
      "step": 3855
    },
    {
      "epoch": 0.771,
      "loss_align": 0.10319268703460693,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26668182015419006,
      "neg_sim_mean": 0.6044868230819702,
      "pos_sim_mean": 0.8968073129653931,
      "step": 3855
    },
    {
      "epoch": 0.7712,
      "grad_norm": 2.23736572265625,
      "learning_rate": 1.5141744857271778e-05,
      "loss": 0.3699,
      "step": 3856
    },
    {
      "epoch": 0.7712,
      "loss_align": 0.09572470188140869,
      "loss_contrastive": 0.03361475467681885,
      "loss_gen": 0.12911419570446014,
      "neg_sim_mean": 0.8378900289535522,
      "pos_sim_mean": 0.9042752981185913,
      "step": 3856
    },
    {
      "epoch": 0.7714,
      "grad_norm": 1.6033287048339844,
      "learning_rate": 1.5116728452957684e-05,
      "loss": 0.2289,
      "step": 3857
    },
    {
      "epoch": 0.7714,
      "loss_align": 0.08775907754898071,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13290613889694214,
      "neg_sim_mean": 0.5309900045394897,
      "pos_sim_mean": 0.9122409224510193,
      "step": 3857
    },
    {
      "epoch": 0.7716,
      "grad_norm": 1.2887018918991089,
      "learning_rate": 1.5091729050326375e-05,
      "loss": 0.2207,
      "step": 3858
    },
    {
      "epoch": 0.7716,
      "loss_align": 0.10369855165481567,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1940479427576065,
      "neg_sim_mean": 0.5355511903762817,
      "pos_sim_mean": 0.8963014483451843,
      "step": 3858
    },
    {
      "epoch": 0.7718,
      "grad_norm": 1.8434664011001587,
      "learning_rate": 1.5066746661562253e-05,
      "loss": 0.2977,
      "step": 3859
    },
    {
      "epoch": 0.7718,
      "loss_align": 0.15225857496261597,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004202648065984249,
      "neg_sim_mean": 0.40316277742385864,
      "pos_sim_mean": 0.847741425037384,
      "step": 3859
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.6739619374275208,
      "learning_rate": 1.5041781298841423e-05,
      "loss": 0.1565,
      "step": 3860
    },
    {
      "epoch": 0.772,
      "loss_align": 0.10213404893875122,
      "loss_contrastive": 0.03235989809036255,
      "loss_gen": 0.2723368704319,
      "neg_sim_mean": 0.8302258253097534,
      "pos_sim_mean": 0.8978659510612488,
      "step": 3860
    },
    {
      "epoch": 0.7722,
      "grad_norm": 2.1301519870758057,
      "learning_rate": 1.5016832974331724e-05,
      "loss": 0.3784,
      "step": 3861
    },
    {
      "epoch": 0.7722,
      "loss_align": 0.10747289657592773,
      "loss_contrastive": 0.05319476127624512,
      "loss_gen": 0.36657747626304626,
      "neg_sim_mean": 0.8457218408584595,
      "pos_sim_mean": 0.8925271034240723,
      "step": 3861
    },
    {
      "epoch": 0.7724,
      "grad_norm": 2.5658535957336426,
      "learning_rate": 1.4991901700192656e-05,
      "loss": 0.4804,
      "step": 3862
    },
    {
      "epoch": 0.7724,
      "loss_align": 0.09261703491210938,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21074211597442627,
      "neg_sim_mean": 0.8051217198371887,
      "pos_sim_mean": 0.9073829650878906,
      "step": 3862
    },
    {
      "epoch": 0.7726,
      "grad_norm": 1.7949668169021606,
      "learning_rate": 1.496698748857543e-05,
      "loss": 0.3034,
      "step": 3863
    },
    {
      "epoch": 0.7726,
      "loss_align": 0.11803865432739258,
      "loss_contrastive": 0.06916689872741699,
      "loss_gen": 0.16518718004226685,
      "neg_sim_mean": 0.8511282205581665,
      "pos_sim_mean": 0.8819613456726074,
      "step": 3863
    },
    {
      "epoch": 0.7728,
      "grad_norm": 1.4757481813430786,
      "learning_rate": 1.4942090351622883e-05,
      "loss": 0.2915,
      "step": 3864
    },
    {
      "epoch": 0.7728,
      "loss_align": 0.05670076608657837,
      "loss_contrastive": 0.06069678068161011,
      "loss_gen": 0.2412409484386444,
      "neg_sim_mean": 0.9039959907531738,
      "pos_sim_mean": 0.9432992339134216,
      "step": 3864
    },
    {
      "epoch": 0.773,
      "grad_norm": 1.6738702058792114,
      "learning_rate": 1.4917210301469632e-05,
      "loss": 0.3052,
      "step": 3865
    },
    {
      "epoch": 0.773,
      "loss_align": 0.10297739505767822,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18346399068832397,
      "neg_sim_mean": 0.5121457576751709,
      "pos_sim_mean": 0.8970226049423218,
      "step": 3865
    },
    {
      "epoch": 0.7732,
      "grad_norm": 2.130568265914917,
      "learning_rate": 1.4892347350241881e-05,
      "loss": 0.2864,
      "step": 3866
    },
    {
      "epoch": 0.7732,
      "loss_align": 0.13261055946350098,
      "loss_contrastive": 0.11025816202163696,
      "loss_gen": 0.1289595365524292,
      "neg_sim_mean": 0.8776475787162781,
      "pos_sim_mean": 0.867389440536499,
      "step": 3866
    },
    {
      "epoch": 0.7734,
      "grad_norm": 1.4628032445907593,
      "learning_rate": 1.4867501510057546e-05,
      "loss": 0.2748,
      "step": 3867
    },
    {
      "epoch": 0.7734,
      "loss_align": 0.08035844564437866,
      "loss_contrastive": 0.08422684669494629,
      "loss_gen": 0.259131520986557,
      "neg_sim_mean": 0.9038683772087097,
      "pos_sim_mean": 0.9196415543556213,
      "step": 3867
    },
    {
      "epoch": 0.7736,
      "grad_norm": 1.9915058612823486,
      "learning_rate": 1.484267279302618e-05,
      "loss": 0.3496,
      "step": 3868
    },
    {
      "epoch": 0.7736,
      "loss_align": 0.1276097297668457,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2091253399848938,
      "neg_sim_mean": 0.7448043823242188,
      "pos_sim_mean": 0.8723902702331543,
      "step": 3868
    },
    {
      "epoch": 0.7738,
      "grad_norm": 1.908713698387146,
      "learning_rate": 1.4817861211248996e-05,
      "loss": 0.3367,
      "step": 3869
    },
    {
      "epoch": 0.7738,
      "loss_align": 0.09062838554382324,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24898244440555573,
      "neg_sim_mean": 0.6704093217849731,
      "pos_sim_mean": 0.9093716144561768,
      "step": 3869
    },
    {
      "epoch": 0.774,
      "grad_norm": 1.4500079154968262,
      "learning_rate": 1.4793066776818842e-05,
      "loss": 0.3396,
      "step": 3870
    },
    {
      "epoch": 0.774,
      "loss_align": 0.0917176604270935,
      "loss_contrastive": 0.09272617101669312,
      "loss_gen": 0.35228803753852844,
      "neg_sim_mean": 0.9010084867477417,
      "pos_sim_mean": 0.9082823395729065,
      "step": 3870
    },
    {
      "epoch": 0.7742,
      "grad_norm": 2.241588830947876,
      "learning_rate": 1.4768289501820265e-05,
      "loss": 0.4551,
      "step": 3871
    },
    {
      "epoch": 0.7742,
      "loss_align": 0.07798761129379272,
      "loss_contrastive": 0.034538447856903076,
      "loss_gen": 0.1701984703540802,
      "neg_sim_mean": 0.8565508127212524,
      "pos_sim_mean": 0.9220123887062073,
      "step": 3871
    },
    {
      "epoch": 0.7744,
      "grad_norm": 1.705881118774414,
      "learning_rate": 1.4743529398329391e-05,
      "loss": 0.2523,
      "step": 3872
    },
    {
      "epoch": 0.7744,
      "loss_align": 0.15400409698486328,
      "loss_contrastive": 0.006754636764526367,
      "loss_gen": 0.12785805761814117,
      "neg_sim_mean": 0.7527505159378052,
      "pos_sim_mean": 0.8459959030151367,
      "step": 3872
    },
    {
      "epoch": 0.7746,
      "grad_norm": 1.3718723058700562,
      "learning_rate": 1.4718786478413982e-05,
      "loss": 0.2827,
      "step": 3873
    },
    {
      "epoch": 0.7746,
      "loss_align": 0.15351390838623047,
      "loss_contrastive": 0.0,
      "loss_gen": 0.021007947623729706,
      "neg_sim_mean": 0.30707502365112305,
      "pos_sim_mean": 0.8464860916137695,
      "step": 3873
    },
    {
      "epoch": 0.7748,
      "grad_norm": 1.03420090675354,
      "learning_rate": 1.469406075413342e-05,
      "loss": 0.1745,
      "step": 3874
    },
    {
      "epoch": 0.7748,
      "loss_align": 0.16172897815704346,
      "loss_contrastive": 0.05935782194137573,
      "loss_gen": 0.21440380811691284,
      "neg_sim_mean": 0.7976288199424744,
      "pos_sim_mean": 0.8382710218429565,
      "step": 3874
    },
    {
      "epoch": 0.775,
      "grad_norm": 1.6880325078964233,
      "learning_rate": 1.4669352237538765e-05,
      "loss": 0.3833,
      "step": 3875
    },
    {
      "epoch": 0.775,
      "loss_align": 0.09879720211029053,
      "loss_contrastive": 0.0009868144989013672,
      "loss_gen": 0.2794865071773529,
      "neg_sim_mean": 0.8021895885467529,
      "pos_sim_mean": 0.9012027978897095,
      "step": 3875
    },
    {
      "epoch": 0.7752,
      "grad_norm": 1.6827278137207031,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.3784,
      "step": 3876
    },
    {
      "epoch": 0.7752,
      "loss_align": 0.1096181869506836,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22269141674041748,
      "neg_sim_mean": 0.786140501499176,
      "pos_sim_mean": 0.8903818130493164,
      "step": 3876
    },
    {
      "epoch": 0.7754,
      "grad_norm": 1.8895580768585205,
      "learning_rate": 1.4619986875569247e-05,
      "loss": 0.3323,
      "step": 3877
    },
    {
      "epoch": 0.7754,
      "loss_align": 0.12000095844268799,
      "loss_contrastive": 0.12181901931762695,
      "loss_gen": 0.2760797441005707,
      "neg_sim_mean": 0.901818037033081,
      "pos_sim_mean": 0.879999041557312,
      "step": 3877
    },
    {
      "epoch": 0.7756,
      "grad_norm": 2.0567615032196045,
      "learning_rate": 1.4595330054254459e-05,
      "loss": 0.4107,
      "step": 3878
    },
    {
      "epoch": 0.7756,
      "loss_align": 0.2185288667678833,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02234664186835289,
      "neg_sim_mean": 0.5809170603752136,
      "pos_sim_mean": 0.7814711332321167,
      "step": 3878
    },
    {
      "epoch": 0.7758,
      "grad_norm": 1.1738033294677734,
      "learning_rate": 1.4570690488745687e-05,
      "loss": 0.2409,
      "step": 3879
    },
    {
      "epoch": 0.7758,
      "loss_align": 0.05005991458892822,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17654167115688324,
      "neg_sim_mean": 0.7749471664428711,
      "pos_sim_mean": 0.9499400854110718,
      "step": 3879
    },
    {
      "epoch": 0.776,
      "grad_norm": 1.6009796857833862,
      "learning_rate": 1.4546068191051987e-05,
      "loss": 0.2266,
      "step": 3880
    },
    {
      "epoch": 0.776,
      "loss_align": 0.10688960552215576,
      "loss_contrastive": 0.04685616493225098,
      "loss_gen": 0.17327852547168732,
      "neg_sim_mean": 0.8399665355682373,
      "pos_sim_mean": 0.8931103944778442,
      "step": 3880
    },
    {
      "epoch": 0.7762,
      "grad_norm": 1.7127752304077148,
      "learning_rate": 1.4521463173173965e-05,
      "loss": 0.2858,
      "step": 3881
    },
    {
      "epoch": 0.7762,
      "loss_align": 0.22542035579681396,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16821186244487762,
      "neg_sim_mean": 0.4996880888938904,
      "pos_sim_mean": 0.774579644203186,
      "step": 3881
    },
    {
      "epoch": 0.7764,
      "grad_norm": 2.007974863052368,
      "learning_rate": 1.449687544710378e-05,
      "loss": 0.3936,
      "step": 3882
    },
    {
      "epoch": 0.7764,
      "loss_align": 0.18153971433639526,
      "loss_contrastive": 0.0,
      "loss_gen": 0.252715528011322,
      "neg_sim_mean": 0.6848863959312439,
      "pos_sim_mean": 0.8184602856636047,
      "step": 3882
    },
    {
      "epoch": 0.7766,
      "grad_norm": 1.9195526838302612,
      "learning_rate": 1.447230502482519e-05,
      "loss": 0.4343,
      "step": 3883
    },
    {
      "epoch": 0.7766,
      "loss_align": 0.020703256130218506,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12006033957004547,
      "neg_sim_mean": 0.6980698108673096,
      "pos_sim_mean": 0.9792967438697815,
      "step": 3883
    },
    {
      "epoch": 0.7768,
      "grad_norm": 0.9832040667533875,
      "learning_rate": 1.4447751918313552e-05,
      "loss": 0.1408,
      "step": 3884
    },
    {
      "epoch": 0.7768,
      "loss_align": 0.10544037818908691,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0033007545862346888,
      "neg_sim_mean": 0.4745694398880005,
      "pos_sim_mean": 0.8945596218109131,
      "step": 3884
    },
    {
      "epoch": 0.777,
      "grad_norm": 0.5337718725204468,
      "learning_rate": 1.4423216139535734e-05,
      "loss": 0.1087,
      "step": 3885
    },
    {
      "epoch": 0.777,
      "loss_align": 0.1190413236618042,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14556777477264404,
      "neg_sim_mean": 0.7039259076118469,
      "pos_sim_mean": 0.8809586763381958,
      "step": 3885
    },
    {
      "epoch": 0.7772,
      "grad_norm": 1.7731198072433472,
      "learning_rate": 1.439869770045018e-05,
      "loss": 0.2646,
      "step": 3886
    },
    {
      "epoch": 0.7772,
      "loss_align": 0.09802252054214478,
      "loss_contrastive": 0.0,
      "loss_gen": 0.307957261800766,
      "neg_sim_mean": 0.6424083709716797,
      "pos_sim_mean": 0.9019774794578552,
      "step": 3886
    },
    {
      "epoch": 0.7774,
      "grad_norm": 2.1793789863586426,
      "learning_rate": 1.4374196613006874e-05,
      "loss": 0.406,
      "step": 3887
    },
    {
      "epoch": 0.7774,
      "loss_align": 0.17748135328292847,
      "loss_contrastive": 0.0,
      "loss_gen": 0.017775189131498337,
      "neg_sim_mean": 0.375770628452301,
      "pos_sim_mean": 0.8225186467170715,
      "step": 3887
    },
    {
      "epoch": 0.7776,
      "grad_norm": 1.1030343770980835,
      "learning_rate": 1.4349712889147355e-05,
      "loss": 0.1953,
      "step": 3888
    },
    {
      "epoch": 0.7776,
      "loss_align": 0.1053779125213623,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19935745000839233,
      "neg_sim_mean": 0.6614482998847961,
      "pos_sim_mean": 0.8946220874786377,
      "step": 3888
    },
    {
      "epoch": 0.7778,
      "grad_norm": 1.6284353733062744,
      "learning_rate": 1.432524654080467e-05,
      "loss": 0.3047,
      "step": 3889
    },
    {
      "epoch": 0.7778,
      "loss_align": 0.09897661209106445,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18071384727954865,
      "neg_sim_mean": 0.5381255149841309,
      "pos_sim_mean": 0.9010233879089355,
      "step": 3889
    },
    {
      "epoch": 0.778,
      "grad_norm": 1.9150148630142212,
      "learning_rate": 1.4300797579903474e-05,
      "loss": 0.2797,
      "step": 3890
    },
    {
      "epoch": 0.778,
      "loss_align": 0.1139599084854126,
      "loss_contrastive": 0.0701521635055542,
      "loss_gen": 0.09098448604345322,
      "neg_sim_mean": 0.8561922311782837,
      "pos_sim_mean": 0.8860400915145874,
      "step": 3890
    },
    {
      "epoch": 0.7782,
      "grad_norm": 1.1117682456970215,
      "learning_rate": 1.4276366018359844e-05,
      "loss": 0.2134,
      "step": 3891
    },
    {
      "epoch": 0.7782,
      "loss_align": 0.31535547971725464,
      "loss_contrastive": 0.0952838659286499,
      "loss_gen": 0.24661152064800262,
      "neg_sim_mean": 0.6799283623695374,
      "pos_sim_mean": 0.6846445202827454,
      "step": 3891
    },
    {
      "epoch": 0.7784,
      "grad_norm": 2.143993377685547,
      "learning_rate": 1.4251951868081438e-05,
      "loss": 0.5734,
      "step": 3892
    },
    {
      "epoch": 0.7784,
      "loss_align": 0.17435109615325928,
      "loss_contrastive": 0.10596036911010742,
      "loss_gen": 0.37940508127212524,
      "neg_sim_mean": 0.8316092491149902,
      "pos_sim_mean": 0.8256489038467407,
      "step": 3892
    },
    {
      "epoch": 0.7786,
      "grad_norm": 2.5164334774017334,
      "learning_rate": 1.4227555140967402e-05,
      "loss": 0.5665,
      "step": 3893
    },
    {
      "epoch": 0.7786,
      "loss_align": 0.08452284336090088,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2011639028787613,
      "neg_sim_mean": 0.7418380379676819,
      "pos_sim_mean": 0.9154771566390991,
      "step": 3893
    },
    {
      "epoch": 0.7788,
      "grad_norm": 1.6773139238357544,
      "learning_rate": 1.420317584890844e-05,
      "loss": 0.2857,
      "step": 3894
    },
    {
      "epoch": 0.7788,
      "loss_align": 0.12369394302368164,
      "loss_contrastive": 0.0,
      "loss_gen": 0.32914578914642334,
      "neg_sim_mean": 0.6486858129501343,
      "pos_sim_mean": 0.8763060569763184,
      "step": 3894
    },
    {
      "epoch": 0.779,
      "grad_norm": 2.4669673442840576,
      "learning_rate": 1.4178814003786705e-05,
      "loss": 0.4528,
      "step": 3895
    },
    {
      "epoch": 0.779,
      "loss_align": 0.06796622276306152,
      "loss_contrastive": 0.0,
      "loss_gen": 0.255964994430542,
      "neg_sim_mean": 0.7719535231590271,
      "pos_sim_mean": 0.9320337772369385,
      "step": 3895
    },
    {
      "epoch": 0.7792,
      "grad_norm": 2.078911781311035,
      "learning_rate": 1.4154469617475863e-05,
      "loss": 0.3239,
      "step": 3896
    },
    {
      "epoch": 0.7792,
      "loss_align": 0.3926540017127991,
      "loss_contrastive": 0.056152164936065674,
      "loss_gen": 0.338746041059494,
      "neg_sim_mean": 0.5634981393814087,
      "pos_sim_mean": 0.6073459982872009,
      "step": 3896
    },
    {
      "epoch": 0.7794,
      "grad_norm": 2.366886854171753,
      "learning_rate": 1.4130142701841075e-05,
      "loss": 0.7381,
      "step": 3897
    },
    {
      "epoch": 0.7794,
      "loss_align": 0.051650166511535645,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4138968288898468,
      "neg_sim_mean": 0.8322457671165466,
      "pos_sim_mean": 0.9483498334884644,
      "step": 3897
    },
    {
      "epoch": 0.7796,
      "grad_norm": 2.088341236114502,
      "learning_rate": 1.4105833268738966e-05,
      "loss": 0.4655,
      "step": 3898
    },
    {
      "epoch": 0.7796,
      "loss_align": 0.37430429458618164,
      "loss_contrastive": 0.037293851375579834,
      "loss_gen": 0.401073694229126,
      "neg_sim_mean": 0.5629895329475403,
      "pos_sim_mean": 0.6256957054138184,
      "step": 3898
    },
    {
      "epoch": 0.7798,
      "grad_norm": 2.563608407974243,
      "learning_rate": 1.4081541330017705e-05,
      "loss": 0.7799,
      "step": 3899
    },
    {
      "epoch": 0.7798,
      "loss_align": 0.07698869705200195,
      "loss_contrastive": 0.04191172122955322,
      "loss_gen": 0.1221519410610199,
      "neg_sim_mean": 0.8649230003356934,
      "pos_sim_mean": 0.923011302947998,
      "step": 3899
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.1179317235946655,
      "learning_rate": 1.4057266897516841e-05,
      "loss": 0.2042,
      "step": 3900
    },
    {
      "epoch": 0.78,
      "loss_align": 0.11627352237701416,
      "loss_contrastive": 0.09818106889724731,
      "loss_gen": 0.2523432672023773,
      "neg_sim_mean": 0.8819075226783752,
      "pos_sim_mean": 0.8837264776229858,
      "step": 3900
    },
    {
      "epoch": 0.7802,
      "grad_norm": 1.3936196565628052,
      "learning_rate": 1.4033009983067452e-05,
      "loss": 0.3804,
      "step": 3901
    },
    {
      "epoch": 0.7802,
      "loss_align": 0.09854072332382202,
      "loss_contrastive": 0.07376015186309814,
      "loss_gen": 0.35181012749671936,
      "neg_sim_mean": 0.8752194046974182,
      "pos_sim_mean": 0.901459276676178,
      "step": 3901
    },
    {
      "epoch": 0.7804,
      "grad_norm": 2.663877487182617,
      "learning_rate": 1.4008770598492071e-05,
      "loss": 0.4592,
      "step": 3902
    },
    {
      "epoch": 0.7804,
      "loss_align": 0.10663759708404541,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21557654440402985,
      "neg_sim_mean": 0.7410355806350708,
      "pos_sim_mean": 0.8933624029159546,
      "step": 3902
    },
    {
      "epoch": 0.7806,
      "grad_norm": 2.0519626140594482,
      "learning_rate": 1.3984548755604654e-05,
      "loss": 0.3222,
      "step": 3903
    },
    {
      "epoch": 0.7806,
      "loss_align": 0.2089337706565857,
      "loss_contrastive": 0.11540532112121582,
      "loss_gen": 0.1402512788772583,
      "neg_sim_mean": 0.8064715266227722,
      "pos_sim_mean": 0.7910662293434143,
      "step": 3903
    },
    {
      "epoch": 0.7808,
      "grad_norm": 1.8796348571777344,
      "learning_rate": 1.3960344466210667e-05,
      "loss": 0.363,
      "step": 3904
    },
    {
      "epoch": 0.7808,
      "loss_align": 0.08484798669815063,
      "loss_contrastive": 0.03911381959915161,
      "loss_gen": 0.32215750217437744,
      "neg_sim_mean": 0.8542658090591431,
      "pos_sim_mean": 0.9151520133018494,
      "step": 3904
    },
    {
      "epoch": 0.781,
      "grad_norm": 1.7152364253997803,
      "learning_rate": 1.3936157742106976e-05,
      "loss": 0.4117,
      "step": 3905
    },
    {
      "epoch": 0.781,
      "loss_align": 0.14927220344543457,
      "loss_contrastive": 0.03029000759124756,
      "loss_gen": 0.3237389922142029,
      "neg_sim_mean": 0.7810177803039551,
      "pos_sim_mean": 0.8507277965545654,
      "step": 3905
    },
    {
      "epoch": 0.7812,
      "grad_norm": 1.9999040365219116,
      "learning_rate": 1.3911988595081893e-05,
      "loss": 0.4766,
      "step": 3906
    },
    {
      "epoch": 0.7812,
      "loss_align": 0.05962419509887695,
      "loss_contrastive": 0.0,
      "loss_gen": 0.031058337539434433,
      "neg_sim_mean": 0.3051861822605133,
      "pos_sim_mean": 0.940375804901123,
      "step": 3906
    },
    {
      "epoch": 0.7814,
      "grad_norm": 1.0926017761230469,
      "learning_rate": 1.3887837036915169e-05,
      "loss": 0.0907,
      "step": 3907
    },
    {
      "epoch": 0.7814,
      "loss_align": 0.11985152959823608,
      "loss_contrastive": 0.03899866342544556,
      "loss_gen": 0.2614745795726776,
      "neg_sim_mean": 0.8191471099853516,
      "pos_sim_mean": 0.8801484704017639,
      "step": 3907
    },
    {
      "epoch": 0.7816,
      "grad_norm": 1.886445164680481,
      "learning_rate": 1.386370307937797e-05,
      "loss": 0.386,
      "step": 3908
    },
    {
      "epoch": 0.7816,
      "loss_align": 0.15727829933166504,
      "loss_contrastive": 0.03596186637878418,
      "loss_gen": 0.18252171576023102,
      "neg_sim_mean": 0.7786835432052612,
      "pos_sim_mean": 0.842721700668335,
      "step": 3908
    },
    {
      "epoch": 0.7818,
      "grad_norm": 1.7591434717178345,
      "learning_rate": 1.3839586734232906e-05,
      "loss": 0.3441,
      "step": 3909
    },
    {
      "epoch": 0.7818,
      "loss_align": 0.10831433534622192,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16626085340976715,
      "neg_sim_mean": 0.6316497325897217,
      "pos_sim_mean": 0.8916856646537781,
      "step": 3909
    },
    {
      "epoch": 0.782,
      "grad_norm": 1.4261994361877441,
      "learning_rate": 1.3815488013233984e-05,
      "loss": 0.2746,
      "step": 3910
    },
    {
      "epoch": 0.782,
      "loss_align": 0.09451401233673096,
      "loss_contrastive": 0.07961273193359375,
      "loss_gen": 0.3079468905925751,
      "neg_sim_mean": 0.8850986957550049,
      "pos_sim_mean": 0.905485987663269,
      "step": 3910
    },
    {
      "epoch": 0.7822,
      "grad_norm": 1.642250895500183,
      "learning_rate": 1.3791406928126638e-05,
      "loss": 0.412,
      "step": 3911
    },
    {
      "epoch": 0.7822,
      "loss_align": 0.1008073091506958,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16873708367347717,
      "neg_sim_mean": 0.5803655385971069,
      "pos_sim_mean": 0.8991926908493042,
      "step": 3911
    },
    {
      "epoch": 0.7824,
      "grad_norm": 1.6725184917449951,
      "learning_rate": 1.3767343490647665e-05,
      "loss": 0.2695,
      "step": 3912
    },
    {
      "epoch": 0.7824,
      "loss_align": 0.13690400123596191,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015495545230805874,
      "neg_sim_mean": 0.3215295970439911,
      "pos_sim_mean": 0.8630959987640381,
      "step": 3912
    },
    {
      "epoch": 0.7826,
      "grad_norm": 0.7396819591522217,
      "learning_rate": 1.3743297712525333e-05,
      "loss": 0.1385,
      "step": 3913
    },
    {
      "epoch": 0.7826,
      "loss_align": 0.1452745795249939,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0023068757727742195,
      "neg_sim_mean": 0.46879029273986816,
      "pos_sim_mean": 0.8547254204750061,
      "step": 3913
    },
    {
      "epoch": 0.7828,
      "grad_norm": 0.5943177938461304,
      "learning_rate": 1.371926960547924e-05,
      "loss": 0.1476,
      "step": 3914
    },
    {
      "epoch": 0.7828,
      "loss_align": 0.17634332180023193,
      "loss_contrastive": 0.07695060968399048,
      "loss_gen": 0.1887204945087433,
      "neg_sim_mean": 0.8006072640419006,
      "pos_sim_mean": 0.8236566781997681,
      "step": 3914
    },
    {
      "epoch": 0.783,
      "grad_norm": 2.0615322589874268,
      "learning_rate": 1.3695259181220404e-05,
      "loss": 0.3743,
      "step": 3915
    },
    {
      "epoch": 0.783,
      "loss_align": 0.20252108573913574,
      "loss_contrastive": 0.12308394908905029,
      "loss_gen": 0.1808348000049591,
      "neg_sim_mean": 0.8205628395080566,
      "pos_sim_mean": 0.7974789142608643,
      "step": 3915
    },
    {
      "epoch": 0.7832,
      "grad_norm": 1.9439373016357422,
      "learning_rate": 1.367126645145121e-05,
      "loss": 0.3981,
      "step": 3916
    },
    {
      "epoch": 0.7832,
      "loss_align": 0.12214481830596924,
      "loss_contrastive": 0.0925910472869873,
      "loss_gen": 0.43363258242607117,
      "neg_sim_mean": 0.8704462051391602,
      "pos_sim_mean": 0.8778551816940308,
      "step": 3916
    },
    {
      "epoch": 0.7834,
      "grad_norm": 2.6322824954986572,
      "learning_rate": 1.3647291427865416e-05,
      "loss": 0.5669,
      "step": 3917
    },
    {
      "epoch": 0.7834,
      "loss_align": 0.10501372814178467,
      "loss_contrastive": 0.016417622566223145,
      "loss_gen": 0.14061880111694336,
      "neg_sim_mean": 0.8114038705825806,
      "pos_sim_mean": 0.8949862718582153,
      "step": 3917
    },
    {
      "epoch": 0.7836,
      "grad_norm": 1.4520878791809082,
      "learning_rate": 1.3623334122148162e-05,
      "loss": 0.2476,
      "step": 3918
    },
    {
      "epoch": 0.7836,
      "loss_align": 0.10301041603088379,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18506132066249847,
      "neg_sim_mean": 0.7671118378639221,
      "pos_sim_mean": 0.8969895839691162,
      "step": 3918
    },
    {
      "epoch": 0.7838,
      "grad_norm": 1.8540323972702026,
      "learning_rate": 1.3599394545975951e-05,
      "loss": 0.2881,
      "step": 3919
    },
    {
      "epoch": 0.7838,
      "loss_align": 0.1250896453857422,
      "loss_contrastive": 0.0046332478523254395,
      "loss_gen": 0.18223035335540771,
      "neg_sim_mean": 0.7795435786247253,
      "pos_sim_mean": 0.8749103546142578,
      "step": 3919
    },
    {
      "epoch": 0.784,
      "grad_norm": 1.6991627216339111,
      "learning_rate": 1.3575472711016635e-05,
      "loss": 0.3079,
      "step": 3920
    },
    {
      "epoch": 0.784,
      "loss_align": 0.059720635414123535,
      "loss_contrastive": 0.09822219610214233,
      "loss_gen": 0.18752919137477875,
      "neg_sim_mean": 0.9385015368461609,
      "pos_sim_mean": 0.9402793645858765,
      "step": 3920
    },
    {
      "epoch": 0.7842,
      "grad_norm": 1.607566237449646,
      "learning_rate": 1.3551568628929434e-05,
      "loss": 0.259,
      "step": 3921
    },
    {
      "epoch": 0.7842,
      "loss_align": 0.13583683967590332,
      "loss_contrastive": 0.07685726881027222,
      "loss_gen": 0.1934565007686615,
      "neg_sim_mean": 0.841020405292511,
      "pos_sim_mean": 0.8641631603240967,
      "step": 3921
    },
    {
      "epoch": 0.7844,
      "grad_norm": 1.6540660858154297,
      "learning_rate": 1.3527682311364887e-05,
      "loss": 0.3385,
      "step": 3922
    },
    {
      "epoch": 0.7844,
      "loss_align": 0.08075308799743652,
      "loss_contrastive": 0.05494874715805054,
      "loss_gen": 0.2480246126651764,
      "neg_sim_mean": 0.8741956353187561,
      "pos_sim_mean": 0.9192469120025635,
      "step": 3922
    },
    {
      "epoch": 0.7846,
      "grad_norm": 1.8546963930130005,
      "learning_rate": 1.3503813769964924e-05,
      "loss": 0.3354,
      "step": 3923
    },
    {
      "epoch": 0.7846,
      "loss_align": 0.07442641258239746,
      "loss_contrastive": 0.056449830532073975,
      "loss_gen": 0.17082156240940094,
      "neg_sim_mean": 0.8820233941078186,
      "pos_sim_mean": 0.9255735874176025,
      "step": 3923
    },
    {
      "epoch": 0.7848,
      "grad_norm": 1.6662888526916504,
      "learning_rate": 1.3479963016362767e-05,
      "loss": 0.252,
      "step": 3924
    },
    {
      "epoch": 0.7848,
      "loss_align": 0.10275399684906006,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1986331343650818,
      "neg_sim_mean": 0.5654385089874268,
      "pos_sim_mean": 0.8972460031509399,
      "step": 3924
    },
    {
      "epoch": 0.785,
      "grad_norm": 1.8951090574264526,
      "learning_rate": 1.3456130062183003e-05,
      "loss": 0.3014,
      "step": 3925
    },
    {
      "epoch": 0.785,
      "loss_align": 0.1052556037902832,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003740811487659812,
      "neg_sim_mean": 0.43148139119148254,
      "pos_sim_mean": 0.8947443962097168,
      "step": 3925
    },
    {
      "epoch": 0.7852,
      "grad_norm": 0.4325660467147827,
      "learning_rate": 1.3432314919041478e-05,
      "loss": 0.109,
      "step": 3926
    },
    {
      "epoch": 0.7852,
      "loss_align": 0.09098529815673828,
      "loss_contrastive": 0.0734596848487854,
      "loss_gen": 0.22710241377353668,
      "neg_sim_mean": 0.8824743628501892,
      "pos_sim_mean": 0.9090147018432617,
      "step": 3926
    },
    {
      "epoch": 0.7854,
      "grad_norm": 1.9379373788833618,
      "learning_rate": 1.3408517598545444e-05,
      "loss": 0.3269,
      "step": 3927
    },
    {
      "epoch": 0.7854,
      "loss_align": 0.12551504373550415,
      "loss_contrastive": 0.11536318063735962,
      "loss_gen": 0.24855172634124756,
      "neg_sim_mean": 0.8898481130599976,
      "pos_sim_mean": 0.8744849562644958,
      "step": 3927
    },
    {
      "epoch": 0.7856,
      "grad_norm": 2.0686798095703125,
      "learning_rate": 1.3384738112293416e-05,
      "loss": 0.3879,
      "step": 3928
    },
    {
      "epoch": 0.7856,
      "loss_align": 0.10343772172927856,
      "loss_contrastive": 0.04799109697341919,
      "loss_gen": 0.15916700661182404,
      "neg_sim_mean": 0.8445533514022827,
      "pos_sim_mean": 0.8965622782707214,
      "step": 3928
    },
    {
      "epoch": 0.7858,
      "grad_norm": 1.5155490636825562,
      "learning_rate": 1.3360976471875225e-05,
      "loss": 0.2684,
      "step": 3929
    },
    {
      "epoch": 0.7858,
      "loss_align": 0.09589576721191406,
      "loss_contrastive": 0.08187395334243774,
      "loss_gen": 0.192582905292511,
      "neg_sim_mean": 0.8859781622886658,
      "pos_sim_mean": 0.9041042327880859,
      "step": 3929
    },
    {
      "epoch": 0.786,
      "grad_norm": 1.5743149518966675,
      "learning_rate": 1.3337232688872009e-05,
      "loss": 0.2983,
      "step": 3930
    },
    {
      "epoch": 0.786,
      "loss_align": 0.0801382064819336,
      "loss_contrastive": 0.0310899019241333,
      "loss_gen": 0.14554741978645325,
      "neg_sim_mean": 0.8509516716003418,
      "pos_sim_mean": 0.9198617935180664,
      "step": 3930
    },
    {
      "epoch": 0.7862,
      "grad_norm": 1.685009241104126,
      "learning_rate": 1.3313506774856177e-05,
      "loss": 0.2294,
      "step": 3931
    },
    {
      "epoch": 0.7862,
      "loss_align": 0.15004974603652954,
      "loss_contrastive": 0.0,
      "loss_gen": 0.035001497715711594,
      "neg_sim_mean": 0.39051634073257446,
      "pos_sim_mean": 0.8499502539634705,
      "step": 3931
    },
    {
      "epoch": 0.7864,
      "grad_norm": 1.4396721124649048,
      "learning_rate": 1.3289798741391485e-05,
      "loss": 0.1851,
      "step": 3932
    },
    {
      "epoch": 0.7864,
      "loss_align": 0.13389557600021362,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1843530684709549,
      "neg_sim_mean": 0.7551105618476868,
      "pos_sim_mean": 0.8661044239997864,
      "step": 3932
    },
    {
      "epoch": 0.7866,
      "grad_norm": 1.567537546157837,
      "learning_rate": 1.3266108600032929e-05,
      "loss": 0.3182,
      "step": 3933
    },
    {
      "epoch": 0.7866,
      "loss_align": 0.06557750701904297,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18443875014781952,
      "neg_sim_mean": 0.614448070526123,
      "pos_sim_mean": 0.934422492980957,
      "step": 3933
    },
    {
      "epoch": 0.7868,
      "grad_norm": 1.673240303993225,
      "learning_rate": 1.3242436362326804e-05,
      "loss": 0.25,
      "step": 3934
    },
    {
      "epoch": 0.7868,
      "loss_align": 0.10903829336166382,
      "loss_contrastive": 0.11909466981887817,
      "loss_gen": 0.4612697660923004,
      "neg_sim_mean": 0.9100563526153564,
      "pos_sim_mean": 0.8909617066383362,
      "step": 3934
    },
    {
      "epoch": 0.787,
      "grad_norm": 2.481872081756592,
      "learning_rate": 1.3218782039810634e-05,
      "loss": 0.5846,
      "step": 3935
    },
    {
      "epoch": 0.787,
      "loss_align": 0.12850189208984375,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20230227708816528,
      "neg_sim_mean": 0.6249971985816956,
      "pos_sim_mean": 0.8714981079101562,
      "step": 3935
    },
    {
      "epoch": 0.7872,
      "grad_norm": 1.581877589225769,
      "learning_rate": 1.3195145644013285e-05,
      "loss": 0.3308,
      "step": 3936
    },
    {
      "epoch": 0.7872,
      "loss_align": 0.0692596435546875,
      "loss_contrastive": 0.018163442611694336,
      "loss_gen": 0.17055028676986694,
      "neg_sim_mean": 0.8489037752151489,
      "pos_sim_mean": 0.9307403564453125,
      "step": 3936
    },
    {
      "epoch": 0.7874,
      "grad_norm": 1.8931024074554443,
      "learning_rate": 1.317152718645484e-05,
      "loss": 0.242,
      "step": 3937
    },
    {
      "epoch": 0.7874,
      "loss_align": 0.1202704906463623,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10659009963274002,
      "neg_sim_mean": 0.7597299814224243,
      "pos_sim_mean": 0.8797295093536377,
      "step": 3937
    },
    {
      "epoch": 0.7876,
      "grad_norm": 1.3148431777954102,
      "learning_rate": 1.3147926678646649e-05,
      "loss": 0.2269,
      "step": 3938
    },
    {
      "epoch": 0.7876,
      "loss_align": 0.08000558614730835,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3507840931415558,
      "neg_sim_mean": 0.7600745558738708,
      "pos_sim_mean": 0.9199944138526917,
      "step": 3938
    },
    {
      "epoch": 0.7878,
      "grad_norm": 2.293165922164917,
      "learning_rate": 1.312434413209131e-05,
      "loss": 0.4308,
      "step": 3939
    },
    {
      "epoch": 0.7878,
      "loss_align": 0.057299911975860596,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18436288833618164,
      "neg_sim_mean": 0.6500759720802307,
      "pos_sim_mean": 0.9427000880241394,
      "step": 3939
    },
    {
      "epoch": 0.788,
      "grad_norm": 1.5070040225982666,
      "learning_rate": 1.3100779558282672e-05,
      "loss": 0.2417,
      "step": 3940
    },
    {
      "epoch": 0.788,
      "loss_align": 0.17699432373046875,
      "loss_contrastive": 0.07066655158996582,
      "loss_gen": 0.1591460257768631,
      "neg_sim_mean": 0.7936722040176392,
      "pos_sim_mean": 0.8230056762695312,
      "step": 3940
    },
    {
      "epoch": 0.7882,
      "grad_norm": 1.7894926071166992,
      "learning_rate": 1.3077232968705805e-05,
      "loss": 0.3446,
      "step": 3941
    },
    {
      "epoch": 0.7882,
      "loss_align": 0.10720884799957275,
      "loss_contrastive": 0.09737968444824219,
      "loss_gen": 0.32509419322013855,
      "neg_sim_mean": 0.8901708126068115,
      "pos_sim_mean": 0.8927911520004272,
      "step": 3941
    },
    {
      "epoch": 0.7884,
      "grad_norm": 1.9246819019317627,
      "learning_rate": 1.3053704374837061e-05,
      "loss": 0.444,
      "step": 3942
    },
    {
      "epoch": 0.7884,
      "loss_align": 0.1072571873664856,
      "loss_contrastive": 0.05224651098251343,
      "loss_gen": 0.165167897939682,
      "neg_sim_mean": 0.8449892997741699,
      "pos_sim_mean": 0.8927428126335144,
      "step": 3942
    },
    {
      "epoch": 0.7886,
      "grad_norm": 1.845580816268921,
      "learning_rate": 1.303019378814399e-05,
      "loss": 0.2787,
      "step": 3943
    },
    {
      "epoch": 0.7886,
      "loss_align": 0.14635241031646729,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013864297419786453,
      "neg_sim_mean": 0.43447041511535645,
      "pos_sim_mean": 0.8536475896835327,
      "step": 3943
    },
    {
      "epoch": 0.7888,
      "grad_norm": 0.8854367136955261,
      "learning_rate": 1.3006701220085338e-05,
      "loss": 0.1602,
      "step": 3944
    },
    {
      "epoch": 0.7888,
      "loss_align": 0.04924076795578003,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12796099483966827,
      "neg_sim_mean": 0.837632954120636,
      "pos_sim_mean": 0.95075923204422,
      "step": 3944
    },
    {
      "epoch": 0.789,
      "grad_norm": 1.4827525615692139,
      "learning_rate": 1.2983226682111093e-05,
      "loss": 0.1772,
      "step": 3945
    },
    {
      "epoch": 0.789,
      "loss_align": 0.05093371868133545,
      "loss_contrastive": 0.0,
      "loss_gen": 0.026214776560664177,
      "neg_sim_mean": 0.33308732509613037,
      "pos_sim_mean": 0.9490662813186646,
      "step": 3945
    },
    {
      "epoch": 0.7892,
      "grad_norm": 0.9854158759117126,
      "learning_rate": 1.29597701856625e-05,
      "loss": 0.0771,
      "step": 3946
    },
    {
      "epoch": 0.7892,
      "loss_align": 0.09063327312469482,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19795209169387817,
      "neg_sim_mean": 0.7229331731796265,
      "pos_sim_mean": 0.9093667268753052,
      "step": 3946
    },
    {
      "epoch": 0.7894,
      "grad_norm": 1.851166844367981,
      "learning_rate": 1.2936331742171942e-05,
      "loss": 0.2886,
      "step": 3947
    },
    {
      "epoch": 0.7894,
      "loss_align": 0.3373991847038269,
      "loss_contrastive": 0.05658203363418579,
      "loss_gen": 0.009079059585928917,
      "neg_sim_mean": 0.619182825088501,
      "pos_sim_mean": 0.6626008152961731,
      "step": 3947
    },
    {
      "epoch": 0.7896,
      "grad_norm": 1.2851336002349854,
      "learning_rate": 1.2912911363063046e-05,
      "loss": 0.3533,
      "step": 3948
    },
    {
      "epoch": 0.7896,
      "loss_align": 0.0760124921798706,
      "loss_contrastive": 0.06946957111358643,
      "loss_gen": 0.2873969078063965,
      "neg_sim_mean": 0.8934570550918579,
      "pos_sim_mean": 0.9239875078201294,
      "step": 3948
    },
    {
      "epoch": 0.7898,
      "grad_norm": 1.9499698877334595,
      "learning_rate": 1.2889509059750604e-05,
      "loss": 0.3717,
      "step": 3949
    },
    {
      "epoch": 0.7898,
      "loss_align": 0.15574276447296143,
      "loss_contrastive": 0.09935343265533447,
      "loss_gen": 0.22461821138858795,
      "neg_sim_mean": 0.8436106443405151,
      "pos_sim_mean": 0.8442572355270386,
      "step": 3949
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.959773302078247,
      "learning_rate": 1.2866124843640615e-05,
      "loss": 0.3923,
      "step": 3950
    },
    {
      "epoch": 0.79,
      "loss_align": 0.07438445091247559,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3498360812664032,
      "neg_sim_mean": 0.772831916809082,
      "pos_sim_mean": 0.9256155490875244,
      "step": 3950
    },
    {
      "epoch": 0.7902,
      "grad_norm": 1.7960489988327026,
      "learning_rate": 1.2842758726130283e-05,
      "loss": 0.4242,
      "step": 3951
    },
    {
      "epoch": 0.7902,
      "loss_align": 0.08812326192855835,
      "loss_contrastive": 0.0657166838645935,
      "loss_gen": 0.48836904764175415,
      "neg_sim_mean": 0.8775933980941772,
      "pos_sim_mean": 0.9118767380714417,
      "step": 3951
    },
    {
      "epoch": 0.7904,
      "grad_norm": 2.5730819702148438,
      "learning_rate": 1.2819410718607972e-05,
      "loss": 0.5844,
      "step": 3952
    },
    {
      "epoch": 0.7904,
      "loss_align": 0.053531527519226074,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0049067335203289986,
      "neg_sim_mean": 0.2592167854309082,
      "pos_sim_mean": 0.9464684724807739,
      "step": 3952
    },
    {
      "epoch": 0.7906,
      "grad_norm": 0.4252273142337799,
      "learning_rate": 1.2796080832453184e-05,
      "loss": 0.0584,
      "step": 3953
    },
    {
      "epoch": 0.7906,
      "loss_align": 0.07662862539291382,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015255421167239547,
      "neg_sim_mean": 0.29064488410949707,
      "pos_sim_mean": 0.9233713746070862,
      "step": 3953
    },
    {
      "epoch": 0.7908,
      "grad_norm": 0.37569645047187805,
      "learning_rate": 1.277276907903664e-05,
      "loss": 0.0782,
      "step": 3954
    },
    {
      "epoch": 0.7908,
      "loss_align": 0.11271095275878906,
      "loss_contrastive": 0.06251382827758789,
      "loss_gen": 0.14472614228725433,
      "neg_sim_mean": 0.8498028516769409,
      "pos_sim_mean": 0.8872890472412109,
      "step": 3954
    },
    {
      "epoch": 0.791,
      "grad_norm": 1.4396793842315674,
      "learning_rate": 1.2749475469720195e-05,
      "loss": 0.2649,
      "step": 3955
    },
    {
      "epoch": 0.791,
      "loss_align": 0.1348186731338501,
      "loss_contrastive": 0.10370039939880371,
      "loss_gen": 0.10406678169965744,
      "neg_sim_mean": 0.8688817024230957,
      "pos_sim_mean": 0.8651813268661499,
      "step": 3955
    },
    {
      "epoch": 0.7912,
      "grad_norm": 1.3629618883132935,
      "learning_rate": 1.2726200015856892e-05,
      "loss": 0.2513,
      "step": 3956
    },
    {
      "epoch": 0.7912,
      "loss_align": 0.0693160891532898,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010154354386031628,
      "neg_sim_mean": 0.27196648716926575,
      "pos_sim_mean": 0.9306839108467102,
      "step": 3956
    },
    {
      "epoch": 0.7914,
      "grad_norm": 0.6471776962280273,
      "learning_rate": 1.2702942728790895e-05,
      "loss": 0.0795,
      "step": 3957
    },
    {
      "epoch": 0.7914,
      "loss_align": 0.13675087690353394,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09509073197841644,
      "neg_sim_mean": 0.5972065925598145,
      "pos_sim_mean": 0.8632491230964661,
      "step": 3957
    },
    {
      "epoch": 0.7916,
      "grad_norm": 1.4319860935211182,
      "learning_rate": 1.2679703619857525e-05,
      "loss": 0.2318,
      "step": 3958
    },
    {
      "epoch": 0.7916,
      "loss_align": 0.1615663766860962,
      "loss_contrastive": 0.0,
      "loss_gen": 0.016339046880602837,
      "neg_sim_mean": 0.45490312576293945,
      "pos_sim_mean": 0.8384336233139038,
      "step": 3958
    },
    {
      "epoch": 0.7918,
      "grad_norm": 0.8977392911911011,
      "learning_rate": 1.2656482700383237e-05,
      "loss": 0.1779,
      "step": 3959
    },
    {
      "epoch": 0.7918,
      "loss_align": 0.12304699420928955,
      "loss_contrastive": 0.11528414487838745,
      "loss_gen": 0.16671472787857056,
      "neg_sim_mean": 0.89223712682724,
      "pos_sim_mean": 0.8769530057907104,
      "step": 3959
    },
    {
      "epoch": 0.792,
      "grad_norm": 1.3216814994812012,
      "learning_rate": 1.2633279981685608e-05,
      "loss": 0.3036,
      "step": 3960
    },
    {
      "epoch": 0.792,
      "loss_align": 0.14237761497497559,
      "loss_contrastive": 0.08362972736358643,
      "loss_gen": 0.4907315969467163,
      "neg_sim_mean": 0.8412520885467529,
      "pos_sim_mean": 0.8576223850250244,
      "step": 3960
    },
    {
      "epoch": 0.7922,
      "grad_norm": 2.611711025238037,
      "learning_rate": 1.2610095475073414e-05,
      "loss": 0.6431,
      "step": 3961
    },
    {
      "epoch": 0.7922,
      "loss_align": 0.1041102409362793,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25694364309310913,
      "neg_sim_mean": 0.5645180344581604,
      "pos_sim_mean": 0.8958897590637207,
      "step": 3961
    },
    {
      "epoch": 0.7924,
      "grad_norm": 1.6891875267028809,
      "learning_rate": 1.2586929191846453e-05,
      "loss": 0.3611,
      "step": 3962
    },
    {
      "epoch": 0.7924,
      "loss_align": 0.23860961198806763,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001143189030699432,
      "neg_sim_mean": 0.318531334400177,
      "pos_sim_mean": 0.7613903880119324,
      "step": 3962
    },
    {
      "epoch": 0.7926,
      "grad_norm": 0.6696960926055908,
      "learning_rate": 1.2563781143295706e-05,
      "loss": 0.2398,
      "step": 3963
    },
    {
      "epoch": 0.7926,
      "loss_align": 0.09635066986083984,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20404338836669922,
      "neg_sim_mean": 0.7855589985847473,
      "pos_sim_mean": 0.9036493301391602,
      "step": 3963
    },
    {
      "epoch": 0.7928,
      "grad_norm": 1.6316320896148682,
      "learning_rate": 1.254065134070323e-05,
      "loss": 0.3004,
      "step": 3964
    },
    {
      "epoch": 0.7928,
      "loss_align": 0.1826760172843933,
      "loss_contrastive": 0.09104615449905396,
      "loss_gen": 0.19206035137176514,
      "neg_sim_mean": 0.8083701133728027,
      "pos_sim_mean": 0.8173239827156067,
      "step": 3964
    },
    {
      "epoch": 0.793,
      "grad_norm": 1.7421495914459229,
      "learning_rate": 1.2517539795342248e-05,
      "loss": 0.3857,
      "step": 3965
    },
    {
      "epoch": 0.793,
      "loss_align": 0.11013787984848022,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17769883573055267,
      "neg_sim_mean": 0.6993017196655273,
      "pos_sim_mean": 0.8898621201515198,
      "step": 3965
    },
    {
      "epoch": 0.7932,
      "grad_norm": 1.5148671865463257,
      "learning_rate": 1.2494446518477022e-05,
      "loss": 0.2878,
      "step": 3966
    },
    {
      "epoch": 0.7932,
      "loss_align": 0.0761491060256958,
      "loss_contrastive": 0.0,
      "loss_gen": 0.023626552894711494,
      "neg_sim_mean": 0.26714393496513367,
      "pos_sim_mean": 0.9238508939743042,
      "step": 3966
    },
    {
      "epoch": 0.7934,
      "grad_norm": 1.0491032600402832,
      "learning_rate": 1.2471371521362946e-05,
      "loss": 0.0998,
      "step": 3967
    },
    {
      "epoch": 0.7934,
      "loss_align": 0.11988067626953125,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01117792446166277,
      "neg_sim_mean": 0.3757924437522888,
      "pos_sim_mean": 0.8801193237304688,
      "step": 3967
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.7450970411300659,
      "learning_rate": 1.2448314815246487e-05,
      "loss": 0.1311,
      "step": 3968
    },
    {
      "epoch": 0.7936,
      "loss_align": 0.11643046140670776,
      "loss_contrastive": 0.09472501277923584,
      "loss_gen": 0.11455994844436646,
      "neg_sim_mean": 0.8782945275306702,
      "pos_sim_mean": 0.8835695385932922,
      "step": 3968
    },
    {
      "epoch": 0.7938,
      "grad_norm": 1.2048187255859375,
      "learning_rate": 1.24252764113652e-05,
      "loss": 0.2424,
      "step": 3969
    },
    {
      "epoch": 0.7938,
      "loss_align": 0.16638755798339844,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15125204622745514,
      "neg_sim_mean": 0.6030240058898926,
      "pos_sim_mean": 0.8336124420166016,
      "step": 3969
    },
    {
      "epoch": 0.794,
      "grad_norm": 1.4141513109207153,
      "learning_rate": 1.240225632094773e-05,
      "loss": 0.3176,
      "step": 3970
    },
    {
      "epoch": 0.794,
      "loss_align": 0.07835352420806885,
      "loss_contrastive": 0.03252971172332764,
      "loss_gen": 0.10622211545705795,
      "neg_sim_mean": 0.8541761636734009,
      "pos_sim_mean": 0.9216464757919312,
      "step": 3970
    },
    {
      "epoch": 0.7942,
      "grad_norm": 1.140032410621643,
      "learning_rate": 1.2379254555213788e-05,
      "loss": 0.1885,
      "step": 3971
    },
    {
      "epoch": 0.7942,
      "loss_align": 0.15147465467453003,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3021697998046875,
      "neg_sim_mean": 0.6402087211608887,
      "pos_sim_mean": 0.84852534532547,
      "step": 3971
    },
    {
      "epoch": 0.7944,
      "grad_norm": 2.1977617740631104,
      "learning_rate": 1.2356271125374152e-05,
      "loss": 0.4536,
      "step": 3972
    },
    {
      "epoch": 0.7944,
      "loss_align": 0.31548798084259033,
      "loss_contrastive": 0.0661388635635376,
      "loss_gen": 0.15504813194274902,
      "neg_sim_mean": 0.6506508588790894,
      "pos_sim_mean": 0.6845120191574097,
      "step": 3972
    },
    {
      "epoch": 0.7946,
      "grad_norm": 1.7889230251312256,
      "learning_rate": 1.2333306042630671e-05,
      "loss": 0.4785,
      "step": 3973
    },
    {
      "epoch": 0.7946,
      "loss_align": 0.060893893241882324,
      "loss_contrastive": 0.022060632705688477,
      "loss_gen": 0.18163836002349854,
      "neg_sim_mean": 0.8611667156219482,
      "pos_sim_mean": 0.9391061067581177,
      "step": 3973
    },
    {
      "epoch": 0.7948,
      "grad_norm": 1.4310288429260254,
      "learning_rate": 1.2310359318176229e-05,
      "loss": 0.2452,
      "step": 3974
    },
    {
      "epoch": 0.7948,
      "loss_align": 0.16596275568008423,
      "loss_contrastive": 0.12445861101150513,
      "loss_gen": 0.11264987289905548,
      "neg_sim_mean": 0.858495831489563,
      "pos_sim_mean": 0.8340372443199158,
      "step": 3974
    },
    {
      "epoch": 0.795,
      "grad_norm": 1.2117960453033447,
      "learning_rate": 1.2287430963194807e-05,
      "loss": 0.2935,
      "step": 3975
    },
    {
      "epoch": 0.795,
      "loss_align": 0.08172672986984253,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2281871885061264,
      "neg_sim_mean": 0.7040283679962158,
      "pos_sim_mean": 0.9182732701301575,
      "step": 3975
    },
    {
      "epoch": 0.7952,
      "grad_norm": 1.7288548946380615,
      "learning_rate": 1.22645209888614e-05,
      "loss": 0.3099,
      "step": 3976
    },
    {
      "epoch": 0.7952,
      "loss_align": 0.15231633186340332,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012235606089234352,
      "neg_sim_mean": 0.27722883224487305,
      "pos_sim_mean": 0.8476836681365967,
      "step": 3976
    },
    {
      "epoch": 0.7954,
      "grad_norm": 0.8359599113464355,
      "learning_rate": 1.2241629406342048e-05,
      "loss": 0.1646,
      "step": 3977
    },
    {
      "epoch": 0.7954,
      "loss_align": 0.130407452583313,
      "loss_contrastive": 0.10000532865524292,
      "loss_gen": 0.261929452419281,
      "neg_sim_mean": 0.869597852230072,
      "pos_sim_mean": 0.869592547416687,
      "step": 3977
    },
    {
      "epoch": 0.7956,
      "grad_norm": 2.389406681060791,
      "learning_rate": 1.2218756226793826e-05,
      "loss": 0.4043,
      "step": 3978
    },
    {
      "epoch": 0.7956,
      "loss_align": 0.09644293785095215,
      "loss_contrastive": 0.018778085708618164,
      "loss_gen": 0.30474355816841125,
      "neg_sim_mean": 0.8223351240158081,
      "pos_sim_mean": 0.9035570621490479,
      "step": 3978
    },
    {
      "epoch": 0.7958,
      "grad_norm": 1.9798206090927124,
      "learning_rate": 1.219590146136485e-05,
      "loss": 0.4034,
      "step": 3979
    },
    {
      "epoch": 0.7958,
      "loss_align": 0.22582179307937622,
      "loss_contrastive": 0.0,
      "loss_gen": 0.021669618785381317,
      "neg_sim_mean": 0.6109121441841125,
      "pos_sim_mean": 0.7741782069206238,
      "step": 3979
    },
    {
      "epoch": 0.796,
      "grad_norm": 0.979708194732666,
      "learning_rate": 1.217306512119425e-05,
      "loss": 0.2475,
      "step": 3980
    },
    {
      "epoch": 0.796,
      "loss_align": 0.14131736755371094,
      "loss_contrastive": 0.010733604431152344,
      "loss_gen": 0.2723905146121979,
      "neg_sim_mean": 0.7694162130355835,
      "pos_sim_mean": 0.8586826324462891,
      "step": 3980
    },
    {
      "epoch": 0.7962,
      "grad_norm": 1.9127840995788574,
      "learning_rate": 1.2150247217412186e-05,
      "loss": 0.415,
      "step": 3981
    },
    {
      "epoch": 0.7962,
      "loss_align": 0.11311006546020508,
      "loss_contrastive": 0.02760028839111328,
      "loss_gen": 0.1210271492600441,
      "neg_sim_mean": 0.8144901990890503,
      "pos_sim_mean": 0.8868899345397949,
      "step": 3981
    },
    {
      "epoch": 0.7964,
      "grad_norm": 1.5055900812149048,
      "learning_rate": 1.212744776113982e-05,
      "loss": 0.2374,
      "step": 3982
    },
    {
      "epoch": 0.7964,
      "loss_align": 0.10483062267303467,
      "loss_contrastive": 0.061015546321868896,
      "loss_gen": 0.2561185359954834,
      "neg_sim_mean": 0.8561848998069763,
      "pos_sim_mean": 0.8951693773269653,
      "step": 3982
    },
    {
      "epoch": 0.7966,
      "grad_norm": 2.075014114379883,
      "learning_rate": 1.2104666763489325e-05,
      "loss": 0.3683,
      "step": 3983
    },
    {
      "epoch": 0.7966,
      "loss_align": 0.08433365821838379,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0019872579723596573,
      "neg_sim_mean": 0.3070604205131531,
      "pos_sim_mean": 0.9156663417816162,
      "step": 3983
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.33291229605674744,
      "learning_rate": 1.2081904235563906e-05,
      "loss": 0.0863,
      "step": 3984
    },
    {
      "epoch": 0.7968,
      "loss_align": 0.06538057327270508,
      "loss_contrastive": 0.02840900421142578,
      "loss_gen": 0.3586442172527313,
      "neg_sim_mean": 0.8630284070968628,
      "pos_sim_mean": 0.9346194267272949,
      "step": 3984
    },
    {
      "epoch": 0.797,
      "grad_norm": 2.1745529174804688,
      "learning_rate": 1.2059160188457724e-05,
      "loss": 0.4274,
      "step": 3985
    },
    {
      "epoch": 0.797,
      "loss_align": 0.07074743509292603,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24153053760528564,
      "neg_sim_mean": 0.695128321647644,
      "pos_sim_mean": 0.929252564907074,
      "step": 3985
    },
    {
      "epoch": 0.7972,
      "grad_norm": 2.061142683029175,
      "learning_rate": 1.203643463325596e-05,
      "loss": 0.3123,
      "step": 3986
    },
    {
      "epoch": 0.7972,
      "loss_align": 0.09470033645629883,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0076258210465312,
      "neg_sim_mean": 0.21728919446468353,
      "pos_sim_mean": 0.9052996635437012,
      "step": 3986
    },
    {
      "epoch": 0.7974,
      "grad_norm": 0.5940829515457153,
      "learning_rate": 1.2013727581034783e-05,
      "loss": 0.1023,
      "step": 3987
    },
    {
      "epoch": 0.7974,
      "loss_align": 0.11882495880126953,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0017480992246419191,
      "neg_sim_mean": 0.4101206660270691,
      "pos_sim_mean": 0.8811750411987305,
      "step": 3987
    },
    {
      "epoch": 0.7976,
      "grad_norm": 0.47597381472587585,
      "learning_rate": 1.1991039042861291e-05,
      "loss": 0.1206,
      "step": 3988
    },
    {
      "epoch": 0.7976,
      "loss_align": 0.051224589347839355,
      "loss_contrastive": 0.06837832927703857,
      "loss_gen": 0.1386961191892624,
      "neg_sim_mean": 0.9171537160873413,
      "pos_sim_mean": 0.9487754106521606,
      "step": 3988
    },
    {
      "epoch": 0.7978,
      "grad_norm": 1.57319176197052,
      "learning_rate": 1.1968369029793641e-05,
      "loss": 0.1981,
      "step": 3989
    },
    {
      "epoch": 0.7978,
      "loss_align": 0.2119375467300415,
      "loss_contrastive": 0.05110865831375122,
      "loss_gen": 0.20582418143749237,
      "neg_sim_mean": 0.7391710877418518,
      "pos_sim_mean": 0.7880624532699585,
      "step": 3989
    },
    {
      "epoch": 0.798,
      "grad_norm": 2.063976287841797,
      "learning_rate": 1.194571755288092e-05,
      "loss": 0.4239,
      "step": 3990
    },
    {
      "epoch": 0.798,
      "loss_align": 0.1315007209777832,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17834050953388214,
      "neg_sim_mean": 0.761915385723114,
      "pos_sim_mean": 0.8684992790222168,
      "step": 3990
    },
    {
      "epoch": 0.7982,
      "grad_norm": 1.8696460723876953,
      "learning_rate": 1.1923084623163172e-05,
      "loss": 0.3098,
      "step": 3991
    },
    {
      "epoch": 0.7982,
      "loss_align": 0.1481562852859497,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2715532183647156,
      "neg_sim_mean": 0.7360208034515381,
      "pos_sim_mean": 0.8518437147140503,
      "step": 3991
    },
    {
      "epoch": 0.7984,
      "grad_norm": 1.6624279022216797,
      "learning_rate": 1.1900470251671414e-05,
      "loss": 0.4197,
      "step": 3992
    },
    {
      "epoch": 0.7984,
      "loss_align": 0.10863083600997925,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11390058696269989,
      "neg_sim_mean": 0.5972061157226562,
      "pos_sim_mean": 0.8913691639900208,
      "step": 3992
    },
    {
      "epoch": 0.7986,
      "grad_norm": 1.3509609699249268,
      "learning_rate": 1.18778744494276e-05,
      "loss": 0.2225,
      "step": 3993
    },
    {
      "epoch": 0.7986,
      "loss_align": 0.12337136268615723,
      "loss_contrastive": 0.0,
      "loss_gen": 0.06950563937425613,
      "neg_sim_mean": 0.6332629323005676,
      "pos_sim_mean": 0.8766286373138428,
      "step": 3993
    },
    {
      "epoch": 0.7988,
      "grad_norm": 1.0665814876556396,
      "learning_rate": 1.1855297227444689e-05,
      "loss": 0.1929,
      "step": 3994
    },
    {
      "epoch": 0.7988,
      "loss_align": 0.12568926811218262,
      "loss_contrastive": 0.07446855306625366,
      "loss_gen": 0.4729175865650177,
      "neg_sim_mean": 0.8487792611122131,
      "pos_sim_mean": 0.8743107318878174,
      "step": 3994
    },
    {
      "epoch": 0.799,
      "grad_norm": 2.8486948013305664,
      "learning_rate": 1.1832738596726517e-05,
      "loss": 0.6075,
      "step": 3995
    },
    {
      "epoch": 0.799,
      "loss_align": 0.07185298204421997,
      "loss_contrastive": 0.07373315095901489,
      "loss_gen": 0.09293245524168015,
      "neg_sim_mean": 0.901880145072937,
      "pos_sim_mean": 0.92814701795578,
      "step": 3995
    },
    {
      "epoch": 0.7992,
      "grad_norm": 0.9648770093917847,
      "learning_rate": 1.1810198568267905e-05,
      "loss": 0.1736,
      "step": 3996
    },
    {
      "epoch": 0.7992,
      "loss_align": 0.06092554330825806,
      "loss_contrastive": 0.1054871678352356,
      "loss_gen": 0.25106489658355713,
      "neg_sim_mean": 0.9445616006851196,
      "pos_sim_mean": 0.9390744566917419,
      "step": 3996
    },
    {
      "epoch": 0.7994,
      "grad_norm": 1.7575170993804932,
      "learning_rate": 1.178767715305455e-05,
      "loss": 0.3246,
      "step": 3997
    },
    {
      "epoch": 0.7994,
      "loss_align": 0.10697460174560547,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007735197898000479,
      "neg_sim_mean": 0.3000640869140625,
      "pos_sim_mean": 0.8930253982543945,
      "step": 3997
    },
    {
      "epoch": 0.7996,
      "grad_norm": 0.71042799949646,
      "learning_rate": 1.1765174362063153e-05,
      "loss": 0.1147,
      "step": 3998
    },
    {
      "epoch": 0.7996,
      "loss_align": 0.17628705501556396,
      "loss_contrastive": 0.1417030692100525,
      "loss_gen": 0.3257290720939636,
      "neg_sim_mean": 0.8654159903526306,
      "pos_sim_mean": 0.823712944984436,
      "step": 3998
    },
    {
      "epoch": 0.7998,
      "grad_norm": 2.2916951179504395,
      "learning_rate": 1.1742690206261292e-05,
      "loss": 0.519,
      "step": 3999
    },
    {
      "epoch": 0.7998,
      "loss_align": 0.08682429790496826,
      "loss_contrastive": 0.01916658878326416,
      "loss_gen": 0.21141456067562103,
      "neg_sim_mean": 0.832342267036438,
      "pos_sim_mean": 0.9131757020950317,
      "step": 3999
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5323574542999268,
      "learning_rate": 1.1720224696607474e-05,
      "loss": 0.3005,
      "step": 4000
    },
    {
      "epoch": 0.8,
      "loss_align": 0.26458489894866943,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20100289583206177,
      "neg_sim_mean": 0.6284118890762329,
      "pos_sim_mean": 0.7354151010513306,
      "step": 4000
    },
    {
      "epoch": 0.8002,
      "grad_norm": 2.2196366786956787,
      "learning_rate": 1.1697777844051105e-05,
      "loss": 0.4656,
      "step": 4001
    },
    {
      "epoch": 0.8002,
      "loss_align": 0.12147074937820435,
      "loss_contrastive": 0.0015624165534973145,
      "loss_gen": 0.20992673933506012,
      "neg_sim_mean": 0.7800916433334351,
      "pos_sim_mean": 0.8785292506217957,
      "step": 4001
    },
    {
      "epoch": 0.8004,
      "grad_norm": 1.6743565797805786,
      "learning_rate": 1.1675349659532513e-05,
      "loss": 0.3316,
      "step": 4002
    },
    {
      "epoch": 0.8004,
      "loss_align": 0.14052510261535645,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0036552157253026962,
      "neg_sim_mean": 0.39991551637649536,
      "pos_sim_mean": 0.8594748973846436,
      "step": 4002
    },
    {
      "epoch": 0.8006,
      "grad_norm": 0.695057213306427,
      "learning_rate": 1.1652940153982917e-05,
      "loss": 0.1442,
      "step": 4003
    },
    {
      "epoch": 0.8006,
      "loss_align": 0.0339130163192749,
      "loss_contrastive": 0.08038902282714844,
      "loss_gen": 0.0750112533569336,
      "neg_sim_mean": 0.9464759826660156,
      "pos_sim_mean": 0.9660869836807251,
      "step": 4003
    },
    {
      "epoch": 0.8008,
      "grad_norm": 0.8044894337654114,
      "learning_rate": 1.1630549338324453e-05,
      "loss": 0.1186,
      "step": 4004
    },
    {
      "epoch": 0.8008,
      "loss_align": 0.09566259384155273,
      "loss_contrastive": 0.09388399124145508,
      "loss_gen": 0.15132878720760345,
      "neg_sim_mean": 0.8982213735580444,
      "pos_sim_mean": 0.9043374061584473,
      "step": 4004
    },
    {
      "epoch": 0.801,
      "grad_norm": 1.4674935340881348,
      "learning_rate": 1.160817722347014e-05,
      "loss": 0.2583,
      "step": 4005
    },
    {
      "epoch": 0.801,
      "loss_align": 0.03654611110687256,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19484005868434906,
      "neg_sim_mean": 0.7311919927597046,
      "pos_sim_mean": 0.9634538888931274,
      "step": 4005
    },
    {
      "epoch": 0.8012,
      "grad_norm": 1.873274326324463,
      "learning_rate": 1.1585823820323843e-05,
      "loss": 0.2314,
      "step": 4006
    },
    {
      "epoch": 0.8012,
      "loss_align": 0.09361898899078369,
      "loss_contrastive": 0.07749044895172119,
      "loss_gen": 0.18286903202533722,
      "neg_sim_mean": 0.8838714361190796,
      "pos_sim_mean": 0.9063810110092163,
      "step": 4006
    },
    {
      "epoch": 0.8014,
      "grad_norm": 1.3684316873550415,
      "learning_rate": 1.1563489139780342e-05,
      "loss": 0.2858,
      "step": 4007
    },
    {
      "epoch": 0.8014,
      "loss_align": 0.08116883039474487,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2202291488647461,
      "neg_sim_mean": 0.5709580183029175,
      "pos_sim_mean": 0.9188311696052551,
      "step": 4007
    },
    {
      "epoch": 0.8016,
      "grad_norm": 2.036287307739258,
      "learning_rate": 1.1541173192725318e-05,
      "loss": 0.3014,
      "step": 4008
    },
    {
      "epoch": 0.8016,
      "loss_align": 0.051506757736206055,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25983676314353943,
      "neg_sim_mean": 0.6487870216369629,
      "pos_sim_mean": 0.948493242263794,
      "step": 4008
    },
    {
      "epoch": 0.8018,
      "grad_norm": 2.118288993835449,
      "learning_rate": 1.1518875990035278e-05,
      "loss": 0.3113,
      "step": 4009
    },
    {
      "epoch": 0.8018,
      "loss_align": 0.12869799137115479,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20246131718158722,
      "neg_sim_mean": 0.7198513746261597,
      "pos_sim_mean": 0.8713020086288452,
      "step": 4009
    },
    {
      "epoch": 0.802,
      "grad_norm": 1.6079626083374023,
      "learning_rate": 1.1496597542577603e-05,
      "loss": 0.3312,
      "step": 4010
    },
    {
      "epoch": 0.802,
      "loss_align": 0.0737454891204834,
      "loss_contrastive": 0.03691685199737549,
      "loss_gen": 0.12726587057113647,
      "neg_sim_mean": 0.8631713390350342,
      "pos_sim_mean": 0.9262545108795166,
      "step": 4010
    },
    {
      "epoch": 0.8022,
      "grad_norm": 1.320427417755127,
      "learning_rate": 1.1474337861210543e-05,
      "loss": 0.2054,
      "step": 4011
    },
    {
      "epoch": 0.8022,
      "loss_align": 0.11155223846435547,
      "loss_contrastive": 0.034588634967803955,
      "loss_gen": 0.17961090803146362,
      "neg_sim_mean": 0.8230363726615906,
      "pos_sim_mean": 0.8884477615356445,
      "step": 4011
    },
    {
      "epoch": 0.8024,
      "grad_norm": 1.6593667268753052,
      "learning_rate": 1.1452096956783181e-05,
      "loss": 0.2953,
      "step": 4012
    },
    {
      "epoch": 0.8024,
      "loss_align": 0.07624804973602295,
      "loss_contrastive": 0.07814228534698486,
      "loss_gen": 0.06817261129617691,
      "neg_sim_mean": 0.901894211769104,
      "pos_sim_mean": 0.923751950263977,
      "step": 4012
    },
    {
      "epoch": 0.8026,
      "grad_norm": 1.0943478345870972,
      "learning_rate": 1.1429874840135491e-05,
      "loss": 0.1538,
      "step": 4013
    },
    {
      "epoch": 0.8026,
      "loss_align": 0.07302594184875488,
      "loss_contrastive": 0.018593251705169678,
      "loss_gen": 0.3624363839626312,
      "neg_sim_mean": 0.8455672860145569,
      "pos_sim_mean": 0.9269740581512451,
      "step": 4013
    },
    {
      "epoch": 0.8028,
      "grad_norm": 1.9728561639785767,
      "learning_rate": 1.1407671522098262e-05,
      "loss": 0.4377,
      "step": 4014
    },
    {
      "epoch": 0.8028,
      "loss_align": 0.12398171424865723,
      "loss_contrastive": 0.07754009962081909,
      "loss_gen": 0.3034832179546356,
      "neg_sim_mean": 0.853558361530304,
      "pos_sim_mean": 0.8760182857513428,
      "step": 4014
    },
    {
      "epoch": 0.803,
      "grad_norm": 2.2005255222320557,
      "learning_rate": 1.1385487013493095e-05,
      "loss": 0.4368,
      "step": 4015
    },
    {
      "epoch": 0.803,
      "loss_align": 0.09462004899978638,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10361921042203903,
      "neg_sim_mean": 0.7487933039665222,
      "pos_sim_mean": 0.9053799510002136,
      "step": 4015
    },
    {
      "epoch": 0.8032,
      "grad_norm": 1.2528952360153198,
      "learning_rate": 1.1363321325132447e-05,
      "loss": 0.1982,
      "step": 4016
    },
    {
      "epoch": 0.8032,
      "loss_align": 0.10440170764923096,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13292276859283447,
      "neg_sim_mean": 0.7776066064834595,
      "pos_sim_mean": 0.895598292350769,
      "step": 4016
    },
    {
      "epoch": 0.8034,
      "grad_norm": 1.4201325178146362,
      "learning_rate": 1.1341174467819637e-05,
      "loss": 0.2373,
      "step": 4017
    },
    {
      "epoch": 0.8034,
      "loss_align": 0.1589418649673462,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20953866839408875,
      "neg_sim_mean": 0.5991101861000061,
      "pos_sim_mean": 0.8410581350326538,
      "step": 4017
    },
    {
      "epoch": 0.8036,
      "grad_norm": 1.820033311843872,
      "learning_rate": 1.1319046452348759e-05,
      "loss": 0.3685,
      "step": 4018
    },
    {
      "epoch": 0.8036,
      "loss_align": 0.11193382740020752,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11793867498636246,
      "neg_sim_mean": 0.726864218711853,
      "pos_sim_mean": 0.8880661725997925,
      "step": 4018
    },
    {
      "epoch": 0.8038,
      "grad_norm": 1.5415849685668945,
      "learning_rate": 1.129693728950474e-05,
      "loss": 0.2299,
      "step": 4019
    },
    {
      "epoch": 0.8038,
      "loss_align": 0.07120245695114136,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2420448511838913,
      "neg_sim_mean": 0.820833146572113,
      "pos_sim_mean": 0.9287975430488586,
      "step": 4019
    },
    {
      "epoch": 0.804,
      "grad_norm": 1.9214191436767578,
      "learning_rate": 1.1274846990063315e-05,
      "loss": 0.3132,
      "step": 4020
    },
    {
      "epoch": 0.804,
      "loss_align": 0.0933496356010437,
      "loss_contrastive": 0.022907912731170654,
      "loss_gen": 0.13568277657032013,
      "neg_sim_mean": 0.829558253288269,
      "pos_sim_mean": 0.9066503643989563,
      "step": 4020
    },
    {
      "epoch": 0.8042,
      "grad_norm": 1.7505295276641846,
      "learning_rate": 1.1252775564791024e-05,
      "loss": 0.2318,
      "step": 4021
    },
    {
      "epoch": 0.8042,
      "loss_align": 0.12163728475570679,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1531406044960022,
      "neg_sim_mean": 0.49163156747817993,
      "pos_sim_mean": 0.8783627152442932,
      "step": 4021
    },
    {
      "epoch": 0.8044,
      "grad_norm": 1.9716160297393799,
      "learning_rate": 1.123072302444521e-05,
      "loss": 0.2748,
      "step": 4022
    },
    {
      "epoch": 0.8044,
      "loss_align": 0.08512091636657715,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18638001382350922,
      "neg_sim_mean": 0.6523838043212891,
      "pos_sim_mean": 0.9148790836334229,
      "step": 4022
    },
    {
      "epoch": 0.8046,
      "grad_norm": 1.9489394426345825,
      "learning_rate": 1.120868937977404e-05,
      "loss": 0.2715,
      "step": 4023
    },
    {
      "epoch": 0.8046,
      "loss_align": 0.09473001956939697,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1925632804632187,
      "neg_sim_mean": 0.8032236099243164,
      "pos_sim_mean": 0.905269980430603,
      "step": 4023
    },
    {
      "epoch": 0.8048,
      "grad_norm": 1.893740177154541,
      "learning_rate": 1.1186674641516415e-05,
      "loss": 0.2873,
      "step": 4024
    },
    {
      "epoch": 0.8048,
      "loss_align": 0.12249970436096191,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1595190167427063,
      "neg_sim_mean": 0.7536405324935913,
      "pos_sim_mean": 0.8775002956390381,
      "step": 4024
    },
    {
      "epoch": 0.805,
      "grad_norm": 1.9208948612213135,
      "learning_rate": 1.1164678820402058e-05,
      "loss": 0.282,
      "step": 4025
    },
    {
      "epoch": 0.805,
      "loss_align": 0.06029343605041504,
      "loss_contrastive": 0.0,
      "loss_gen": 0.137465238571167,
      "neg_sim_mean": 0.5346488356590271,
      "pos_sim_mean": 0.939706563949585,
      "step": 4025
    },
    {
      "epoch": 0.8052,
      "grad_norm": 1.5514925718307495,
      "learning_rate": 1.1142701927151456e-05,
      "loss": 0.1978,
      "step": 4026
    },
    {
      "epoch": 0.8052,
      "loss_align": 0.08722031116485596,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3504941165447235,
      "neg_sim_mean": 0.7419463396072388,
      "pos_sim_mean": 0.912779688835144,
      "step": 4026
    },
    {
      "epoch": 0.8054,
      "grad_norm": 2.1529860496520996,
      "learning_rate": 1.11207439724759e-05,
      "loss": 0.4377,
      "step": 4027
    },
    {
      "epoch": 0.8054,
      "loss_align": 0.06963992118835449,
      "loss_contrastive": 0.07484036684036255,
      "loss_gen": 0.2199896275997162,
      "neg_sim_mean": 0.9052004218101501,
      "pos_sim_mean": 0.9303600788116455,
      "step": 4027
    },
    {
      "epoch": 0.8056,
      "grad_norm": 1.714900016784668,
      "learning_rate": 1.1098804967077425e-05,
      "loss": 0.2986,
      "step": 4028
    },
    {
      "epoch": 0.8056,
      "loss_align": 0.16692036390304565,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14738143980503082,
      "neg_sim_mean": 0.4918473958969116,
      "pos_sim_mean": 0.8330796360969543,
      "step": 4028
    },
    {
      "epoch": 0.8058,
      "grad_norm": 1.7681641578674316,
      "learning_rate": 1.1076884921648834e-05,
      "loss": 0.3143,
      "step": 4029
    },
    {
      "epoch": 0.8058,
      "loss_align": 0.09960603713989258,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2778199315071106,
      "neg_sim_mean": 0.6209235191345215,
      "pos_sim_mean": 0.9003939628601074,
      "step": 4029
    },
    {
      "epoch": 0.806,
      "grad_norm": 2.1098215579986572,
      "learning_rate": 1.1054983846873685e-05,
      "loss": 0.3774,
      "step": 4030
    },
    {
      "epoch": 0.806,
      "loss_align": 0.09719789028167725,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12343025207519531,
      "neg_sim_mean": 0.5826106071472168,
      "pos_sim_mean": 0.9028021097183228,
      "step": 4030
    },
    {
      "epoch": 0.8062,
      "grad_norm": 1.6248202323913574,
      "learning_rate": 1.1033101753426283e-05,
      "loss": 0.2206,
      "step": 4031
    },
    {
      "epoch": 0.8062,
      "loss_align": 0.19181281328201294,
      "loss_contrastive": 0.08206534385681152,
      "loss_gen": 0.590217113494873,
      "neg_sim_mean": 0.7902525067329407,
      "pos_sim_mean": 0.8081871867179871,
      "step": 4031
    },
    {
      "epoch": 0.8064,
      "grad_norm": 2.9786055088043213,
      "learning_rate": 1.1011238651971745e-05,
      "loss": 0.7919,
      "step": 4032
    },
    {
      "epoch": 0.8064,
      "loss_align": 0.14930498600006104,
      "loss_contrastive": 0.06782639026641846,
      "loss_gen": 0.14916683733463287,
      "neg_sim_mean": 0.8185213804244995,
      "pos_sim_mean": 0.850695013999939,
      "step": 4032
    },
    {
      "epoch": 0.8066,
      "grad_norm": 1.6967432498931885,
      "learning_rate": 1.0989394553165832e-05,
      "loss": 0.3066,
      "step": 4033
    },
    {
      "epoch": 0.8066,
      "loss_align": 0.16920709609985352,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015919868601486087,
      "neg_sim_mean": 0.33636021614074707,
      "pos_sim_mean": 0.8307929039001465,
      "step": 4033
    },
    {
      "epoch": 0.8068,
      "grad_norm": 0.46352505683898926,
      "learning_rate": 1.0967569467655103e-05,
      "loss": 0.1708,
      "step": 4034
    },
    {
      "epoch": 0.8068,
      "loss_align": 0.12089061737060547,
      "loss_contrastive": 0.006608068943023682,
      "loss_gen": 0.2076326310634613,
      "neg_sim_mean": 0.7857174277305603,
      "pos_sim_mean": 0.8791093826293945,
      "step": 4034
    },
    {
      "epoch": 0.807,
      "grad_norm": 1.6902563571929932,
      "learning_rate": 1.0945763406076836e-05,
      "loss": 0.3293,
      "step": 4035
    },
    {
      "epoch": 0.807,
      "loss_align": 0.07379424571990967,
      "loss_contrastive": 0.09892475605010986,
      "loss_gen": 0.24217632412910461,
      "neg_sim_mean": 0.9251304864883423,
      "pos_sim_mean": 0.9262057542800903,
      "step": 4035
    },
    {
      "epoch": 0.8072,
      "grad_norm": 1.8445016145706177,
      "learning_rate": 1.0923976379059058e-05,
      "loss": 0.3278,
      "step": 4036
    },
    {
      "epoch": 0.8072,
      "loss_align": 0.17374634742736816,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14227452874183655,
      "neg_sim_mean": 0.5104979872703552,
      "pos_sim_mean": 0.8262536525726318,
      "step": 4036
    },
    {
      "epoch": 0.8074,
      "grad_norm": 1.7944483757019043,
      "learning_rate": 1.0902208397220497e-05,
      "loss": 0.316,
      "step": 4037
    },
    {
      "epoch": 0.8074,
      "loss_align": 0.15317916870117188,
      "loss_contrastive": 0.0948367714881897,
      "loss_gen": 0.18875756859779358,
      "neg_sim_mean": 0.8416575789451599,
      "pos_sim_mean": 0.8468208312988281,
      "step": 4037
    },
    {
      "epoch": 0.8076,
      "grad_norm": 1.6508936882019043,
      "learning_rate": 1.0880459471170596e-05,
      "loss": 0.3533,
      "step": 4038
    },
    {
      "epoch": 0.8076,
      "loss_align": 0.07824462652206421,
      "loss_contrastive": 0.06374245882034302,
      "loss_gen": 0.19458332657814026,
      "neg_sim_mean": 0.8854978084564209,
      "pos_sim_mean": 0.9217553734779358,
      "step": 4038
    },
    {
      "epoch": 0.8078,
      "grad_norm": 1.743802785873413,
      "learning_rate": 1.0858729611509516e-05,
      "loss": 0.2805,
      "step": 4039
    },
    {
      "epoch": 0.8078,
      "loss_align": 0.07334733009338379,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14142534136772156,
      "neg_sim_mean": 0.7815710306167603,
      "pos_sim_mean": 0.9266526699066162,
      "step": 4039
    },
    {
      "epoch": 0.808,
      "grad_norm": 1.36106538772583,
      "learning_rate": 1.0837018828828133e-05,
      "loss": 0.2148,
      "step": 4040
    },
    {
      "epoch": 0.808,
      "loss_align": 0.1092749834060669,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1760202795267105,
      "neg_sim_mean": 0.6686131954193115,
      "pos_sim_mean": 0.8907250165939331,
      "step": 4040
    },
    {
      "epoch": 0.8082,
      "grad_norm": 1.441774845123291,
      "learning_rate": 1.0815327133708015e-05,
      "loss": 0.2853,
      "step": 4041
    },
    {
      "epoch": 0.8082,
      "loss_align": 0.08360695838928223,
      "loss_contrastive": 0.05111539363861084,
      "loss_gen": 0.2020265907049179,
      "neg_sim_mean": 0.8675084114074707,
      "pos_sim_mean": 0.9163930416107178,
      "step": 4041
    },
    {
      "epoch": 0.8084,
      "grad_norm": 2.121033191680908,
      "learning_rate": 1.0793654536721431e-05,
      "loss": 0.2918,
      "step": 4042
    },
    {
      "epoch": 0.8084,
      "loss_align": 0.075431227684021,
      "loss_contrastive": 0.08615028858184814,
      "loss_gen": 0.28237971663475037,
      "neg_sim_mean": 0.9107190370559692,
      "pos_sim_mean": 0.924568772315979,
      "step": 4042
    },
    {
      "epoch": 0.8086,
      "grad_norm": 1.9526925086975098,
      "learning_rate": 1.077200104843134e-05,
      "loss": 0.3681,
      "step": 4043
    },
    {
      "epoch": 0.8086,
      "loss_align": 0.143110454082489,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3346349000930786,
      "neg_sim_mean": 0.7376213669776917,
      "pos_sim_mean": 0.856889545917511,
      "step": 4043
    },
    {
      "epoch": 0.8088,
      "grad_norm": 2.1690785884857178,
      "learning_rate": 1.0750366679391393e-05,
      "loss": 0.4777,
      "step": 4044
    },
    {
      "epoch": 0.8088,
      "loss_align": 0.11163544654846191,
      "loss_contrastive": 0.0,
      "loss_gen": 0.45992007851600647,
      "neg_sim_mean": 0.6607591509819031,
      "pos_sim_mean": 0.8883645534515381,
      "step": 4044
    },
    {
      "epoch": 0.809,
      "grad_norm": 2.6281213760375977,
      "learning_rate": 1.0728751440145906e-05,
      "loss": 0.5716,
      "step": 4045
    },
    {
      "epoch": 0.809,
      "loss_align": 0.07693135738372803,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27353399991989136,
      "neg_sim_mean": 0.7419191598892212,
      "pos_sim_mean": 0.923068642616272,
      "step": 4045
    },
    {
      "epoch": 0.8092,
      "grad_norm": 1.7976535558700562,
      "learning_rate": 1.0707155341229901e-05,
      "loss": 0.3505,
      "step": 4046
    },
    {
      "epoch": 0.8092,
      "loss_align": 0.16043007373809814,
      "loss_contrastive": 0.0,
      "loss_gen": 0.06278149783611298,
      "neg_sim_mean": 0.726209819316864,
      "pos_sim_mean": 0.8395699262619019,
      "step": 4046
    },
    {
      "epoch": 0.8094,
      "grad_norm": 0.9888391494750977,
      "learning_rate": 1.0685578393169055e-05,
      "loss": 0.2232,
      "step": 4047
    },
    {
      "epoch": 0.8094,
      "loss_align": 0.09286320209503174,
      "loss_contrastive": 0.07847368717193604,
      "loss_gen": 0.14996249973773956,
      "neg_sim_mean": 0.8856104612350464,
      "pos_sim_mean": 0.9071367979049683,
      "step": 4047
    },
    {
      "epoch": 0.8096,
      "grad_norm": 1.3249013423919678,
      "learning_rate": 1.0664020606479701e-05,
      "loss": 0.2522,
      "step": 4048
    },
    {
      "epoch": 0.8096,
      "loss_align": 0.0917205810546875,
      "loss_contrastive": 0.07481896877288818,
      "loss_gen": 0.44178083539009094,
      "neg_sim_mean": 0.8830983638763428,
      "pos_sim_mean": 0.9082794189453125,
      "step": 4048
    },
    {
      "epoch": 0.8098,
      "grad_norm": 2.457047939300537,
      "learning_rate": 1.064248199166884e-05,
      "loss": 0.5425,
      "step": 4049
    },
    {
      "epoch": 0.8098,
      "loss_align": 0.09134936332702637,
      "loss_contrastive": 0.09582805633544922,
      "loss_gen": 0.17551776766777039,
      "neg_sim_mean": 0.9044786691665649,
      "pos_sim_mean": 0.9086506366729736,
      "step": 4049
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6394455432891846,
      "learning_rate": 1.0620962559234143e-05,
      "loss": 0.2784,
      "step": 4050
    },
    {
      "epoch": 0.81,
      "loss_align": 0.1486455202102661,
      "loss_contrastive": 0.0,
      "loss_gen": 0.08934246748685837,
      "neg_sim_mean": 0.6823980808258057,
      "pos_sim_mean": 0.8513544797897339,
      "step": 4050
    },
    {
      "epoch": 0.8102,
      "grad_norm": 1.066150426864624,
      "learning_rate": 1.0599462319663905e-05,
      "loss": 0.238,
      "step": 4051
    },
    {
      "epoch": 0.8102,
      "loss_align": 0.14078980684280396,
      "loss_contrastive": 0.024854063987731934,
      "loss_gen": 0.2349085509777069,
      "neg_sim_mean": 0.7840642333030701,
      "pos_sim_mean": 0.859210193157196,
      "step": 4051
    },
    {
      "epoch": 0.8104,
      "grad_norm": 1.7274612188339233,
      "learning_rate": 1.0577981283437095e-05,
      "loss": 0.3787,
      "step": 4052
    },
    {
      "epoch": 0.8104,
      "loss_align": 0.18337881565093994,
      "loss_contrastive": 0.09223878383636475,
      "loss_gen": 0.24287152290344238,
      "neg_sim_mean": 0.8088599443435669,
      "pos_sim_mean": 0.8166211843490601,
      "step": 4052
    },
    {
      "epoch": 0.8106,
      "grad_norm": 1.53484308719635,
      "learning_rate": 1.0556519461023301e-05,
      "loss": 0.4373,
      "step": 4053
    },
    {
      "epoch": 0.8106,
      "loss_align": 0.07885384559631348,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16286297142505646,
      "neg_sim_mean": 0.28602927923202515,
      "pos_sim_mean": 0.9211461544036865,
      "step": 4053
    },
    {
      "epoch": 0.8108,
      "grad_norm": 1.8127306699752808,
      "learning_rate": 1.053507686288276e-05,
      "loss": 0.2417,
      "step": 4054
    },
    {
      "epoch": 0.8108,
      "loss_align": 0.0975947380065918,
      "loss_contrastive": 0.04672032594680786,
      "loss_gen": 0.2013404220342636,
      "neg_sim_mean": 0.8491255640983582,
      "pos_sim_mean": 0.9024052619934082,
      "step": 4054
    },
    {
      "epoch": 0.811,
      "grad_norm": 1.8782695531845093,
      "learning_rate": 1.0513653499466313e-05,
      "loss": 0.3045,
      "step": 4055
    },
    {
      "epoch": 0.811,
      "loss_align": 0.07064270973205566,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00643885275349021,
      "neg_sim_mean": 0.24157996475696564,
      "pos_sim_mean": 0.9293572902679443,
      "step": 4055
    },
    {
      "epoch": 0.8112,
      "grad_norm": 0.48104825615882874,
      "learning_rate": 1.049224938121548e-05,
      "loss": 0.0771,
      "step": 4056
    },
    {
      "epoch": 0.8112,
      "loss_align": 0.15598857402801514,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010080612264573574,
      "neg_sim_mean": 0.3449152112007141,
      "pos_sim_mean": 0.8440114259719849,
      "step": 4056
    },
    {
      "epoch": 0.8114,
      "grad_norm": 0.7299876809120178,
      "learning_rate": 1.047086451856235e-05,
      "loss": 0.1661,
      "step": 4057
    },
    {
      "epoch": 0.8114,
      "loss_align": 0.10186338424682617,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14502090215682983,
      "neg_sim_mean": 0.7774651050567627,
      "pos_sim_mean": 0.8981366157531738,
      "step": 4057
    },
    {
      "epoch": 0.8116,
      "grad_norm": 1.3325601816177368,
      "learning_rate": 1.0449498921929667e-05,
      "loss": 0.2469,
      "step": 4058
    },
    {
      "epoch": 0.8116,
      "loss_align": 0.06711846590042114,
      "loss_contrastive": 0.04189687967300415,
      "loss_gen": 0.21171268820762634,
      "neg_sim_mean": 0.8747783899307251,
      "pos_sim_mean": 0.9328815340995789,
      "step": 4058
    },
    {
      "epoch": 0.8118,
      "grad_norm": 1.794642686843872,
      "learning_rate": 1.0428152601730718e-05,
      "loss": 0.2839,
      "step": 4059
    },
    {
      "epoch": 0.8118,
      "loss_align": 0.09770357608795166,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19903680682182312,
      "neg_sim_mean": 0.7575734257698059,
      "pos_sim_mean": 0.9022964239120483,
      "step": 4059
    },
    {
      "epoch": 0.812,
      "grad_norm": 1.8079596757888794,
      "learning_rate": 1.0406825568369477e-05,
      "loss": 0.2967,
      "step": 4060
    },
    {
      "epoch": 0.812,
      "loss_align": 0.11152440309524536,
      "loss_contrastive": 0.08434170484542847,
      "loss_gen": 0.1128150075674057,
      "neg_sim_mean": 0.8728172779083252,
      "pos_sim_mean": 0.8884755969047546,
      "step": 4060
    },
    {
      "epoch": 0.8122,
      "grad_norm": 1.4698643684387207,
      "learning_rate": 1.0385517832240471e-05,
      "loss": 0.2345,
      "step": 4061
    },
    {
      "epoch": 0.8122,
      "loss_align": 0.10013139247894287,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009764671325683594,
      "neg_sim_mean": 0.2877747118473053,
      "pos_sim_mean": 0.8998686075210571,
      "step": 4061
    },
    {
      "epoch": 0.8124,
      "grad_norm": 0.6477330923080444,
      "learning_rate": 1.0364229403728831e-05,
      "loss": 0.1099,
      "step": 4062
    },
    {
      "epoch": 0.8124,
      "loss_align": 0.03922182321548462,
      "loss_contrastive": 0.04109007120132446,
      "loss_gen": 0.19709350168704987,
      "neg_sim_mean": 0.9018682241439819,
      "pos_sim_mean": 0.9607781767845154,
      "step": 4062
    },
    {
      "epoch": 0.8126,
      "grad_norm": 1.341720461845398,
      "learning_rate": 1.034296029321028e-05,
      "loss": 0.2412,
      "step": 4063
    },
    {
      "epoch": 0.8126,
      "loss_align": 0.12631756067276,
      "loss_contrastive": 0.0,
      "loss_gen": 0.329797625541687,
      "neg_sim_mean": 0.7513464093208313,
      "pos_sim_mean": 0.87368243932724,
      "step": 4063
    },
    {
      "epoch": 0.8128,
      "grad_norm": 2.054877996444702,
      "learning_rate": 1.0321710511051109e-05,
      "loss": 0.4561,
      "step": 4064
    },
    {
      "epoch": 0.8128,
      "loss_align": 0.09875476360321045,
      "loss_contrastive": 0.08810794353485107,
      "loss_gen": 0.14161086082458496,
      "neg_sim_mean": 0.8893531560897827,
      "pos_sim_mean": 0.9012452363967896,
      "step": 4064
    },
    {
      "epoch": 0.813,
      "grad_norm": 1.1437104940414429,
      "learning_rate": 1.0300480067608231e-05,
      "loss": 0.2509,
      "step": 4065
    },
    {
      "epoch": 0.813,
      "loss_align": 0.09944349527359009,
      "loss_contrastive": 2.4437904357910156e-05,
      "loss_gen": 0.15958335995674133,
      "neg_sim_mean": 0.8005809187889099,
      "pos_sim_mean": 0.9005565047264099,
      "step": 4065
    },
    {
      "epoch": 0.8132,
      "grad_norm": 1.6722638607025146,
      "learning_rate": 1.0279268973229089e-05,
      "loss": 0.259,
      "step": 4066
    },
    {
      "epoch": 0.8132,
      "loss_align": 0.10996860265731812,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006269092205911875,
      "neg_sim_mean": 0.4117206335067749,
      "pos_sim_mean": 0.8900313973426819,
      "step": 4066
    },
    {
      "epoch": 0.8134,
      "grad_norm": 0.5354406833648682,
      "learning_rate": 1.0258077238251735e-05,
      "loss": 0.1162,
      "step": 4067
    },
    {
      "epoch": 0.8134,
      "loss_align": 0.08424854278564453,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010947700589895248,
      "neg_sim_mean": 0.29009339213371277,
      "pos_sim_mean": 0.9157514572143555,
      "step": 4067
    },
    {
      "epoch": 0.8136,
      "grad_norm": 0.7228217124938965,
      "learning_rate": 1.0236904873004722e-05,
      "loss": 0.0952,
      "step": 4068
    },
    {
      "epoch": 0.8136,
      "loss_align": 0.13531279563903809,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12144995480775833,
      "neg_sim_mean": 0.5875031352043152,
      "pos_sim_mean": 0.8646872043609619,
      "step": 4068
    },
    {
      "epoch": 0.8138,
      "grad_norm": 1.2766730785369873,
      "learning_rate": 1.0215751887807228e-05,
      "loss": 0.2568,
      "step": 4069
    },
    {
      "epoch": 0.8138,
      "loss_align": 0.07761764526367188,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01636183261871338,
      "neg_sim_mean": 0.3103899359703064,
      "pos_sim_mean": 0.9223823547363281,
      "step": 4069
    },
    {
      "epoch": 0.814,
      "grad_norm": 0.8594719767570496,
      "learning_rate": 1.0194618292968972e-05,
      "loss": 0.094,
      "step": 4070
    },
    {
      "epoch": 0.814,
      "loss_align": 0.08367729187011719,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009046225808560848,
      "neg_sim_mean": 0.32291167974472046,
      "pos_sim_mean": 0.9163227081298828,
      "step": 4070
    },
    {
      "epoch": 0.8142,
      "grad_norm": 0.585586667060852,
      "learning_rate": 1.0173504098790187e-05,
      "loss": 0.0927,
      "step": 4071
    },
    {
      "epoch": 0.8142,
      "loss_align": 0.13436543941497803,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14637087285518646,
      "neg_sim_mean": 0.7454783916473389,
      "pos_sim_mean": 0.865634560585022,
      "step": 4071
    },
    {
      "epoch": 0.8144,
      "grad_norm": 1.4085942506790161,
      "learning_rate": 1.0152409315561695e-05,
      "loss": 0.2807,
      "step": 4072
    },
    {
      "epoch": 0.8144,
      "loss_align": 0.06642317771911621,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00965672917664051,
      "neg_sim_mean": 0.35977333784103394,
      "pos_sim_mean": 0.9335768222808838,
      "step": 4072
    },
    {
      "epoch": 0.8146,
      "grad_norm": 0.7048994898796082,
      "learning_rate": 1.0131333953564826e-05,
      "loss": 0.0761,
      "step": 4073
    },
    {
      "epoch": 0.8146,
      "loss_align": 0.1540919542312622,
      "loss_contrastive": 0.09246951341629028,
      "loss_gen": 0.39611852169036865,
      "neg_sim_mean": 0.8383775353431702,
      "pos_sim_mean": 0.8459080457687378,
      "step": 4073
    },
    {
      "epoch": 0.8148,
      "grad_norm": 2.574791193008423,
      "learning_rate": 1.0110278023071446e-05,
      "loss": 0.5613,
      "step": 4074
    },
    {
      "epoch": 0.8148,
      "loss_align": 0.07887685298919678,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21955513954162598,
      "neg_sim_mean": 0.7996984124183655,
      "pos_sim_mean": 0.9211231470108032,
      "step": 4074
    },
    {
      "epoch": 0.815,
      "grad_norm": 1.4401867389678955,
      "learning_rate": 1.0089241534343984e-05,
      "loss": 0.2984,
      "step": 4075
    },
    {
      "epoch": 0.815,
      "loss_align": 0.08717596530914307,
      "loss_contrastive": 0.04050111770629883,
      "loss_gen": 0.15318073332309723,
      "neg_sim_mean": 0.8533251285552979,
      "pos_sim_mean": 0.9128240346908569,
      "step": 4075
    },
    {
      "epoch": 0.8152,
      "grad_norm": 1.5031284093856812,
      "learning_rate": 1.006822449763537e-05,
      "loss": 0.2452,
      "step": 4076
    },
    {
      "epoch": 0.8152,
      "loss_align": 0.12179070711135864,
      "loss_contrastive": 0.06463736295700073,
      "loss_gen": 0.20274285972118378,
      "neg_sim_mean": 0.8428466320037842,
      "pos_sim_mean": 0.8782092928886414,
      "step": 4076
    },
    {
      "epoch": 0.8154,
      "grad_norm": 1.4774138927459717,
      "learning_rate": 1.0047226923189024e-05,
      "loss": 0.3323,
      "step": 4077
    },
    {
      "epoch": 0.8154,
      "loss_align": 0.0723523497581482,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010608043521642685,
      "neg_sim_mean": 0.4000113010406494,
      "pos_sim_mean": 0.9276476502418518,
      "step": 4077
    },
    {
      "epoch": 0.8156,
      "grad_norm": 0.6557332277297974,
      "learning_rate": 1.0026248821238916e-05,
      "loss": 0.083,
      "step": 4078
    },
    {
      "epoch": 0.8156,
      "loss_align": 0.09206652641296387,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1635044515132904,
      "neg_sim_mean": 0.775046169757843,
      "pos_sim_mean": 0.9079334735870361,
      "step": 4078
    },
    {
      "epoch": 0.8158,
      "grad_norm": 1.682582139968872,
      "learning_rate": 1.0005290202009531e-05,
      "loss": 0.2556,
      "step": 4079
    },
    {
      "epoch": 0.8158,
      "loss_align": 0.07239794731140137,
      "loss_contrastive": 0.02033323049545288,
      "loss_gen": 0.34460073709487915,
      "neg_sim_mean": 0.8479352593421936,
      "pos_sim_mean": 0.9276020526885986,
      "step": 4079
    },
    {
      "epoch": 0.816,
      "grad_norm": 1.928642749786377,
      "learning_rate": 9.984351075715848e-06,
      "loss": 0.4194,
      "step": 4080
    },
    {
      "epoch": 0.816,
      "loss_align": 0.12299579381942749,
      "loss_contrastive": 0.09079533815383911,
      "loss_gen": 0.19620738923549652,
      "neg_sim_mean": 0.8677995204925537,
      "pos_sim_mean": 0.8770042061805725,
      "step": 4080
    },
    {
      "epoch": 0.8162,
      "grad_norm": 1.5469454526901245,
      "learning_rate": 9.963431452563332e-06,
      "loss": 0.3301,
      "step": 4081
    },
    {
      "epoch": 0.8162,
      "loss_align": 0.19977569580078125,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013559011742472649,
      "neg_sim_mean": 0.4776769280433655,
      "pos_sim_mean": 0.8002243041992188,
      "step": 4081
    },
    {
      "epoch": 0.8164,
      "grad_norm": 0.997526228427887,
      "learning_rate": 9.942531342747952e-06,
      "loss": 0.2133,
      "step": 4082
    },
    {
      "epoch": 0.8164,
      "loss_align": 0.06062793731689453,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22421100735664368,
      "neg_sim_mean": 0.6520125269889832,
      "pos_sim_mean": 0.9393720626831055,
      "step": 4082
    },
    {
      "epoch": 0.8166,
      "grad_norm": 2.026371955871582,
      "learning_rate": 9.921650756456164e-06,
      "loss": 0.2848,
      "step": 4083
    },
    {
      "epoch": 0.8166,
      "loss_align": 0.10022377967834473,
      "loss_contrastive": 0.02207803726196289,
      "loss_gen": 0.20212826132774353,
      "neg_sim_mean": 0.8218542337417603,
      "pos_sim_mean": 0.8997762203216553,
      "step": 4083
    },
    {
      "epoch": 0.8168,
      "grad_norm": 1.8081213235855103,
      "learning_rate": 9.900789703864933e-06,
      "loss": 0.305,
      "step": 4084
    },
    {
      "epoch": 0.8168,
      "loss_align": 0.1760404109954834,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2777880132198334,
      "neg_sim_mean": 0.35219046473503113,
      "pos_sim_mean": 0.8239595890045166,
      "step": 4084
    },
    {
      "epoch": 0.817,
      "grad_norm": 1.8413453102111816,
      "learning_rate": 9.879948195141681e-06,
      "loss": 0.4538,
      "step": 4085
    },
    {
      "epoch": 0.817,
      "loss_align": 0.12594181299209595,
      "loss_contrastive": 0.03693181276321411,
      "loss_gen": 0.17093542218208313,
      "neg_sim_mean": 0.8109899759292603,
      "pos_sim_mean": 0.874058187007904,
      "step": 4085
    },
    {
      "epoch": 0.8172,
      "grad_norm": 1.8845821619033813,
      "learning_rate": 9.859126240444283e-06,
      "loss": 0.3013,
      "step": 4086
    },
    {
      "epoch": 0.8172,
      "loss_align": 0.07969403266906738,
      "loss_contrastive": 0.04930472373962402,
      "loss_gen": 0.3269960582256317,
      "neg_sim_mean": 0.8696106672286987,
      "pos_sim_mean": 0.9203059673309326,
      "step": 4086
    },
    {
      "epoch": 0.8174,
      "grad_norm": 2.220247507095337,
      "learning_rate": 9.838323849921121e-06,
      "loss": 0.4126,
      "step": 4087
    },
    {
      "epoch": 0.8174,
      "loss_align": 0.24366897344589233,
      "loss_contrastive": 0.07168400287628174,
      "loss_gen": 0.4104423224925995,
      "neg_sim_mean": 0.7280150055885315,
      "pos_sim_mean": 0.7563310265541077,
      "step": 4087
    },
    {
      "epoch": 0.8176,
      "grad_norm": 2.274678945541382,
      "learning_rate": 9.81754103371101e-06,
      "loss": 0.6627,
      "step": 4088
    },
    {
      "epoch": 0.8176,
      "loss_align": 0.1308203935623169,
      "loss_contrastive": 0.0805310606956482,
      "loss_gen": 0.3479991853237152,
      "neg_sim_mean": 0.8497106432914734,
      "pos_sim_mean": 0.8691796064376831,
      "step": 4088
    },
    {
      "epoch": 0.8178,
      "grad_norm": 2.8428168296813965,
      "learning_rate": 9.796777801943269e-06,
      "loss": 0.4885,
      "step": 4089
    },
    {
      "epoch": 0.8178,
      "loss_align": 0.16279858350753784,
      "loss_contrastive": 0.0,
      "loss_gen": 0.36604663729667664,
      "neg_sim_mean": 0.7028496861457825,
      "pos_sim_mean": 0.8372014164924622,
      "step": 4089
    },
    {
      "epoch": 0.818,
      "grad_norm": 2.644223690032959,
      "learning_rate": 9.77603416473763e-06,
      "loss": 0.5288,
      "step": 4090
    },
    {
      "epoch": 0.818,
      "loss_align": 0.11182117462158203,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18697167932987213,
      "neg_sim_mean": 0.6439580917358398,
      "pos_sim_mean": 0.888178825378418,
      "step": 4090
    },
    {
      "epoch": 0.8182,
      "grad_norm": 1.9180783033370972,
      "learning_rate": 9.755310132204298e-06,
      "loss": 0.2988,
      "step": 4091
    },
    {
      "epoch": 0.8182,
      "loss_align": 0.07807743549346924,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011177553795278072,
      "neg_sim_mean": 0.45261555910110474,
      "pos_sim_mean": 0.9219225645065308,
      "step": 4091
    },
    {
      "epoch": 0.8184,
      "grad_norm": 0.7167166471481323,
      "learning_rate": 9.734605714443906e-06,
      "loss": 0.0893,
      "step": 4092
    },
    {
      "epoch": 0.8184,
      "loss_align": 0.08847308158874512,
      "loss_contrastive": 0.020484447479248047,
      "loss_gen": 0.4711916744709015,
      "neg_sim_mean": 0.832011342048645,
      "pos_sim_mean": 0.9115269184112549,
      "step": 4092
    },
    {
      "epoch": 0.8186,
      "grad_norm": 2.3335769176483154,
      "learning_rate": 9.713920921547532e-06,
      "loss": 0.5621,
      "step": 4093
    },
    {
      "epoch": 0.8186,
      "loss_align": 0.2308167815208435,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001151581178419292,
      "neg_sim_mean": 0.37822240591049194,
      "pos_sim_mean": 0.7691832184791565,
      "step": 4093
    },
    {
      "epoch": 0.8188,
      "grad_norm": 0.697180449962616,
      "learning_rate": 9.69325576359672e-06,
      "loss": 0.232,
      "step": 4094
    },
    {
      "epoch": 0.8188,
      "loss_align": 0.08564072847366333,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010408874601125717,
      "neg_sim_mean": 0.25495272874832153,
      "pos_sim_mean": 0.9143592715263367,
      "step": 4094
    },
    {
      "epoch": 0.819,
      "grad_norm": 0.6721802353858948,
      "learning_rate": 9.672610250663389e-06,
      "loss": 0.096,
      "step": 4095
    },
    {
      "epoch": 0.819,
      "loss_align": 0.19551682472229004,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001202707295306027,
      "neg_sim_mean": 0.2874680161476135,
      "pos_sim_mean": 0.80448317527771,
      "step": 4095
    },
    {
      "epoch": 0.8192,
      "grad_norm": 0.582557737827301,
      "learning_rate": 9.651984392809914e-06,
      "loss": 0.1967,
      "step": 4096
    },
    {
      "epoch": 0.8192,
      "loss_align": 0.09583824872970581,
      "loss_contrastive": 0.003133714199066162,
      "loss_gen": 0.26978161931037903,
      "neg_sim_mean": 0.8072954416275024,
      "pos_sim_mean": 0.9041617512702942,
      "step": 4096
    },
    {
      "epoch": 0.8194,
      "grad_norm": 1.7091329097747803,
      "learning_rate": 9.631378200089081e-06,
      "loss": 0.366,
      "step": 4097
    },
    {
      "epoch": 0.8194,
      "loss_align": 0.10126352310180664,
      "loss_contrastive": 0.03593850135803223,
      "loss_gen": 0.15257884562015533,
      "neg_sim_mean": 0.8346749544143677,
      "pos_sim_mean": 0.8987364768981934,
      "step": 4097
    },
    {
      "epoch": 0.8196,
      "grad_norm": 1.7895357608795166,
      "learning_rate": 9.610791682544124e-06,
      "loss": 0.2582,
      "step": 4098
    },
    {
      "epoch": 0.8196,
      "loss_align": 0.12692582607269287,
      "loss_contrastive": 0.02275264263153076,
      "loss_gen": 0.19860631227493286,
      "neg_sim_mean": 0.79582679271698,
      "pos_sim_mean": 0.8730741739273071,
      "step": 4098
    },
    {
      "epoch": 0.8198,
      "grad_norm": 1.7214206457138062,
      "learning_rate": 9.590224850208646e-06,
      "loss": 0.3283,
      "step": 4099
    },
    {
      "epoch": 0.8198,
      "loss_align": 0.048911094665527344,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17824430763721466,
      "neg_sim_mean": 0.733078122138977,
      "pos_sim_mean": 0.9510889053344727,
      "step": 4099
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.5226982831954956,
      "learning_rate": 9.569677713106674e-06,
      "loss": 0.2272,
      "step": 4100
    },
    {
      "epoch": 0.82,
      "loss_align": 0.2695683240890503,
      "loss_contrastive": 0.0,
      "loss_gen": 0.018412450328469276,
      "neg_sim_mean": 0.4589442014694214,
      "pos_sim_mean": 0.7304316759109497,
      "step": 4100
    },
    {
      "epoch": 0.8202,
      "grad_norm": 1.1634840965270996,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.288,
      "step": 4101
    },
    {
      "epoch": 0.8202,
      "loss_align": 0.18628907203674316,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0036295352037996054,
      "neg_sim_mean": 0.41097456216812134,
      "pos_sim_mean": 0.8137109279632568,
      "step": 4101
    },
    {
      "epoch": 0.8204,
      "grad_norm": 0.5931064486503601,
      "learning_rate": 9.528642564651342e-06,
      "loss": 0.1899,
      "step": 4102
    },
    {
      "epoch": 0.8204,
      "loss_align": 0.1051337718963623,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3807080090045929,
      "neg_sim_mean": 0.6980811357498169,
      "pos_sim_mean": 0.8948662281036377,
      "step": 4102
    },
    {
      "epoch": 0.8206,
      "grad_norm": 2.398359537124634,
      "learning_rate": 9.508154573298012e-06,
      "loss": 0.4858,
      "step": 4103
    },
    {
      "epoch": 0.8206,
      "loss_align": 0.12052774429321289,
      "loss_contrastive": 0.054251015186309814,
      "loss_gen": 0.13568809628486633,
      "neg_sim_mean": 0.833723247051239,
      "pos_sim_mean": 0.8794722557067871,
      "step": 4103
    },
    {
      "epoch": 0.8208,
      "grad_norm": 1.7669004201889038,
      "learning_rate": 9.48768631717824e-06,
      "loss": 0.2627,
      "step": 4104
    },
    {
      "epoch": 0.8208,
      "loss_align": 0.19292914867401123,
      "loss_contrastive": 0.12377142906188965,
      "loss_gen": 0.16632233560085297,
      "neg_sim_mean": 0.8308422565460205,
      "pos_sim_mean": 0.8070708513259888,
      "step": 4104
    },
    {
      "epoch": 0.821,
      "grad_norm": 1.9257276058197021,
      "learning_rate": 9.467237806268009e-06,
      "loss": 0.3741,
      "step": 4105
    },
    {
      "epoch": 0.821,
      "loss_align": 0.09137672185897827,
      "loss_contrastive": 0.08947896957397461,
      "loss_gen": 0.26921701431274414,
      "neg_sim_mean": 0.8981022238731384,
      "pos_sim_mean": 0.9086232781410217,
      "step": 4105
    },
    {
      "epoch": 0.8212,
      "grad_norm": 1.9787076711654663,
      "learning_rate": 9.446809050533678e-06,
      "loss": 0.3713,
      "step": 4106
    },
    {
      "epoch": 0.8212,
      "loss_align": 0.08777493238449097,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006626185029745102,
      "neg_sim_mean": 0.32087138295173645,
      "pos_sim_mean": 0.912225067615509,
      "step": 4106
    },
    {
      "epoch": 0.8214,
      "grad_norm": 0.5318350195884705,
      "learning_rate": 9.426400059931955e-06,
      "loss": 0.0944,
      "step": 4107
    },
    {
      "epoch": 0.8214,
      "loss_align": 0.08301335573196411,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14958904683589935,
      "neg_sim_mean": 0.8148134350776672,
      "pos_sim_mean": 0.9169866442680359,
      "step": 4107
    },
    {
      "epoch": 0.8216,
      "grad_norm": 1.3704978227615356,
      "learning_rate": 9.406010844409957e-06,
      "loss": 0.2326,
      "step": 4108
    },
    {
      "epoch": 0.8216,
      "loss_align": 0.17003250122070312,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2067379355430603,
      "neg_sim_mean": 0.7198553681373596,
      "pos_sim_mean": 0.8299674987792969,
      "step": 4108
    },
    {
      "epoch": 0.8218,
      "grad_norm": 1.8861751556396484,
      "learning_rate": 9.38564141390514e-06,
      "loss": 0.3768,
      "step": 4109
    },
    {
      "epoch": 0.8218,
      "loss_align": 0.1094256043434143,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24971002340316772,
      "neg_sim_mean": 0.756759762763977,
      "pos_sim_mean": 0.8905743956565857,
      "step": 4109
    },
    {
      "epoch": 0.822,
      "grad_norm": 1.995943546295166,
      "learning_rate": 9.365291778345303e-06,
      "loss": 0.3591,
      "step": 4110
    },
    {
      "epoch": 0.822,
      "loss_align": 0.17839622497558594,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002599820727482438,
      "neg_sim_mean": 0.34527722001075745,
      "pos_sim_mean": 0.8216037750244141,
      "step": 4110
    },
    {
      "epoch": 0.8222,
      "grad_norm": 0.65254807472229,
      "learning_rate": 9.344961947648623e-06,
      "loss": 0.181,
      "step": 4111
    },
    {
      "epoch": 0.8222,
      "loss_align": 0.09239798784255981,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0012051266385242343,
      "neg_sim_mean": 0.3175053596496582,
      "pos_sim_mean": 0.9076020121574402,
      "step": 4111
    },
    {
      "epoch": 0.8224,
      "grad_norm": 0.39395326375961304,
      "learning_rate": 9.324651931723599e-06,
      "loss": 0.0936,
      "step": 4112
    },
    {
      "epoch": 0.8224,
      "loss_align": 0.15761971473693848,
      "loss_contrastive": 0.07465028762817383,
      "loss_gen": 0.17709797620773315,
      "neg_sim_mean": 0.8170305490493774,
      "pos_sim_mean": 0.8423802852630615,
      "step": 4112
    },
    {
      "epoch": 0.8226,
      "grad_norm": 1.656498908996582,
      "learning_rate": 9.304361740469103e-06,
      "loss": 0.3437,
      "step": 4113
    },
    {
      "epoch": 0.8226,
      "loss_align": 0.1377672553062439,
      "loss_contrastive": 0.0,
      "loss_gen": 0.6153014302253723,
      "neg_sim_mean": 0.6892063617706299,
      "pos_sim_mean": 0.8622327446937561,
      "step": 4113
    },
    {
      "epoch": 0.8228,
      "grad_norm": 2.6111061573028564,
      "learning_rate": 9.284091383774313e-06,
      "loss": 0.7531,
      "step": 4114
    },
    {
      "epoch": 0.8228,
      "loss_align": 0.09128010272979736,
      "loss_contrastive": 0.009088754653930664,
      "loss_gen": 0.3415979743003845,
      "neg_sim_mean": 0.8178086280822754,
      "pos_sim_mean": 0.9087198972702026,
      "step": 4114
    },
    {
      "epoch": 0.823,
      "grad_norm": 1.9089137315750122,
      "learning_rate": 9.26384087151876e-06,
      "loss": 0.434,
      "step": 4115
    },
    {
      "epoch": 0.823,
      "loss_align": 0.15923357009887695,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10289072245359421,
      "neg_sim_mean": 0.629615843296051,
      "pos_sim_mean": 0.840766429901123,
      "step": 4115
    },
    {
      "epoch": 0.8232,
      "grad_norm": 1.3957833051681519,
      "learning_rate": 9.243610213572285e-06,
      "loss": 0.2621,
      "step": 4116
    },
    {
      "epoch": 0.8232,
      "loss_align": 0.11699646711349487,
      "loss_contrastive": 0.07265836000442505,
      "loss_gen": 0.22706760466098785,
      "neg_sim_mean": 0.8556618690490723,
      "pos_sim_mean": 0.8830035328865051,
      "step": 4116
    },
    {
      "epoch": 0.8234,
      "grad_norm": 2.0561866760253906,
      "learning_rate": 9.223399419795093e-06,
      "loss": 0.3528,
      "step": 4117
    },
    {
      "epoch": 0.8234,
      "loss_align": 0.06611025333404541,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3139171302318573,
      "neg_sim_mean": 0.5161212086677551,
      "pos_sim_mean": 0.9338897466659546,
      "step": 4117
    },
    {
      "epoch": 0.8236,
      "grad_norm": 2.0455360412597656,
      "learning_rate": 9.203208500037663e-06,
      "loss": 0.38,
      "step": 4118
    },
    {
      "epoch": 0.8236,
      "loss_align": 0.13167822360992432,
      "loss_contrastive": 0.07392805814743042,
      "loss_gen": 0.2918337881565094,
      "neg_sim_mean": 0.8422498106956482,
      "pos_sim_mean": 0.8683217763900757,
      "step": 4118
    },
    {
      "epoch": 0.8238,
      "grad_norm": 2.057657480239868,
      "learning_rate": 9.183037464140804e-06,
      "loss": 0.4324,
      "step": 4119
    },
    {
      "epoch": 0.8238,
      "loss_align": 0.0770573616027832,
      "loss_contrastive": 0.0,
      "loss_gen": 0.016035109758377075,
      "neg_sim_mean": 0.406991571187973,
      "pos_sim_mean": 0.9229426383972168,
      "step": 4119
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.7803817987442017,
      "learning_rate": 9.162886321935632e-06,
      "loss": 0.0931,
      "step": 4120
    },
    {
      "epoch": 0.824,
      "loss_align": 0.10942530632019043,
      "loss_contrastive": 0.038648903369903564,
      "loss_gen": 0.12254168093204498,
      "neg_sim_mean": 0.8292235732078552,
      "pos_sim_mean": 0.8905746936798096,
      "step": 4120
    },
    {
      "epoch": 0.8242,
      "grad_norm": 1.5588237047195435,
      "learning_rate": 9.142755083243576e-06,
      "loss": 0.2366,
      "step": 4121
    },
    {
      "epoch": 0.8242,
      "loss_align": 0.14977937936782837,
      "loss_contrastive": 0.010336339473724365,
      "loss_gen": 0.2964237332344055,
      "neg_sim_mean": 0.7605569362640381,
      "pos_sim_mean": 0.8502206206321716,
      "step": 4121
    },
    {
      "epoch": 0.8244,
      "grad_norm": 2.0321624279022217,
      "learning_rate": 9.122643757876354e-06,
      "loss": 0.4474,
      "step": 4122
    },
    {
      "epoch": 0.8244,
      "loss_align": 0.0859825611114502,
      "loss_contrastive": 0.005442023277282715,
      "loss_gen": 0.21111464500427246,
      "neg_sim_mean": 0.8194594383239746,
      "pos_sim_mean": 0.9140174388885498,
      "step": 4122
    },
    {
      "epoch": 0.8246,
      "grad_norm": 1.8500807285308838,
      "learning_rate": 9.102552355635979e-06,
      "loss": 0.2978,
      "step": 4123
    },
    {
      "epoch": 0.8246,
      "loss_align": 0.06161701679229736,
      "loss_contrastive": 0.020012378692626953,
      "loss_gen": 0.32725924253463745,
      "neg_sim_mean": 0.8583953380584717,
      "pos_sim_mean": 0.9383829832077026,
      "step": 4123
    },
    {
      "epoch": 0.8248,
      "grad_norm": 1.8656343221664429,
      "learning_rate": 9.08248088631476e-06,
      "loss": 0.3913,
      "step": 4124
    },
    {
      "epoch": 0.8248,
      "loss_align": 0.12855875492095947,
      "loss_contrastive": 0.10780799388885498,
      "loss_gen": 0.2023852914571762,
      "neg_sim_mean": 0.8792492151260376,
      "pos_sim_mean": 0.8714412450790405,
      "step": 4124
    },
    {
      "epoch": 0.825,
      "grad_norm": 1.7136679887771606,
      "learning_rate": 9.06242935969528e-06,
      "loss": 0.3439,
      "step": 4125
    },
    {
      "epoch": 0.825,
      "loss_align": 0.16188490390777588,
      "loss_contrastive": 0.0,
      "loss_gen": 0.021006852388381958,
      "neg_sim_mean": 0.320332407951355,
      "pos_sim_mean": 0.8381150960922241,
      "step": 4125
    },
    {
      "epoch": 0.8252,
      "grad_norm": 0.9826540946960449,
      "learning_rate": 9.042397785550405e-06,
      "loss": 0.1829,
      "step": 4126
    },
    {
      "epoch": 0.8252,
      "loss_align": 0.15273809432983398,
      "loss_contrastive": 0.1510937213897705,
      "loss_gen": 0.3408151865005493,
      "neg_sim_mean": 0.8983556032180786,
      "pos_sim_mean": 0.847261905670166,
      "step": 4126
    },
    {
      "epoch": 0.8254,
      "grad_norm": 2.5531716346740723,
      "learning_rate": 9.022386173643305e-06,
      "loss": 0.5117,
      "step": 4127
    },
    {
      "epoch": 0.8254,
      "loss_align": 0.11675220727920532,
      "loss_contrastive": 0.0197107195854187,
      "loss_gen": 0.6446084380149841,
      "neg_sim_mean": 0.8029584884643555,
      "pos_sim_mean": 0.8832477927207947,
      "step": 4127
    },
    {
      "epoch": 0.8256,
      "grad_norm": 2.8727385997772217,
      "learning_rate": 9.002394533727382e-06,
      "loss": 0.7637,
      "step": 4128
    },
    {
      "epoch": 0.8256,
      "loss_align": 0.14194488525390625,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24375152587890625,
      "neg_sim_mean": 0.6005905270576477,
      "pos_sim_mean": 0.8580551147460938,
      "step": 4128
    },
    {
      "epoch": 0.8258,
      "grad_norm": 2.2269489765167236,
      "learning_rate": 8.982422875546332e-06,
      "loss": 0.3857,
      "step": 4129
    },
    {
      "epoch": 0.8258,
      "loss_align": 0.11954665184020996,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2833157479763031,
      "neg_sim_mean": 0.7386261820793152,
      "pos_sim_mean": 0.88045334815979,
      "step": 4129
    },
    {
      "epoch": 0.826,
      "grad_norm": 2.1412789821624756,
      "learning_rate": 8.962471208834055e-06,
      "loss": 0.4029,
      "step": 4130
    },
    {
      "epoch": 0.826,
      "loss_align": 0.03040909767150879,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22347897291183472,
      "neg_sim_mean": 0.8066902160644531,
      "pos_sim_mean": 0.9695909023284912,
      "step": 4130
    },
    {
      "epoch": 0.8262,
      "grad_norm": 1.9280521869659424,
      "learning_rate": 8.9425395433148e-06,
      "loss": 0.2539,
      "step": 4131
    },
    {
      "epoch": 0.8262,
      "loss_align": 0.23430776596069336,
      "loss_contrastive": 0.00219881534576416,
      "loss_gen": 0.2771291136741638,
      "neg_sim_mean": 0.6678910255432129,
      "pos_sim_mean": 0.7656922340393066,
      "step": 4131
    },
    {
      "epoch": 0.8264,
      "grad_norm": 1.9645483493804932,
      "learning_rate": 8.922627888703e-06,
      "loss": 0.5117,
      "step": 4132
    },
    {
      "epoch": 0.8264,
      "loss_align": 0.10045367479324341,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008217915892601013,
      "neg_sim_mean": 0.5430605411529541,
      "pos_sim_mean": 0.8995463252067566,
      "step": 4132
    },
    {
      "epoch": 0.8266,
      "grad_norm": 0.7984235286712646,
      "learning_rate": 8.902736254703348e-06,
      "loss": 0.1087,
      "step": 4133
    },
    {
      "epoch": 0.8266,
      "loss_align": 0.09127664566040039,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19769619405269623,
      "neg_sim_mean": 0.777694821357727,
      "pos_sim_mean": 0.9087233543395996,
      "step": 4133
    },
    {
      "epoch": 0.8268,
      "grad_norm": 1.6311500072479248,
      "learning_rate": 8.882864651010798e-06,
      "loss": 0.289,
      "step": 4134
    },
    {
      "epoch": 0.8268,
      "loss_align": 0.12724721431732178,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008074905723333359,
      "neg_sim_mean": 0.29116782546043396,
      "pos_sim_mean": 0.8727527856826782,
      "step": 4134
    },
    {
      "epoch": 0.827,
      "grad_norm": 0.5873173475265503,
      "learning_rate": 8.863013087310501e-06,
      "loss": 0.1353,
      "step": 4135
    },
    {
      "epoch": 0.827,
      "loss_align": 0.17755985260009766,
      "loss_contrastive": 0.1695052981376648,
      "loss_gen": 0.2022176831960678,
      "neg_sim_mean": 0.8919454216957092,
      "pos_sim_mean": 0.8224401473999023,
      "step": 4135
    },
    {
      "epoch": 0.8272,
      "grad_norm": 1.5130113363265991,
      "learning_rate": 8.843181573277902e-06,
      "loss": 0.4001,
      "step": 4136
    },
    {
      "epoch": 0.8272,
      "loss_align": 0.09116816520690918,
      "loss_contrastive": 0.0754970908164978,
      "loss_gen": 0.18715247511863708,
      "neg_sim_mean": 0.8843289017677307,
      "pos_sim_mean": 0.9088318347930908,
      "step": 4136
    },
    {
      "epoch": 0.8274,
      "grad_norm": 1.4900487661361694,
      "learning_rate": 8.823370118578628e-06,
      "loss": 0.2874,
      "step": 4137
    },
    {
      "epoch": 0.8274,
      "loss_align": 0.05764573812484741,
      "loss_contrastive": 0.0025684237480163574,
      "loss_gen": 0.29602423310279846,
      "neg_sim_mean": 0.844922661781311,
      "pos_sim_mean": 0.9423542618751526,
      "step": 4137
    },
    {
      "epoch": 0.8276,
      "grad_norm": 1.9276989698410034,
      "learning_rate": 8.803578732868545e-06,
      "loss": 0.354,
      "step": 4138
    },
    {
      "epoch": 0.8276,
      "loss_align": 0.044349849224090576,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2686036229133606,
      "neg_sim_mean": 0.7738207578659058,
      "pos_sim_mean": 0.9556501507759094,
      "step": 4138
    },
    {
      "epoch": 0.8278,
      "grad_norm": 1.6497812271118164,
      "learning_rate": 8.783807425793721e-06,
      "loss": 0.313,
      "step": 4139
    },
    {
      "epoch": 0.8278,
      "loss_align": 0.21676850318908691,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001193295931443572,
      "neg_sim_mean": 0.42108041048049927,
      "pos_sim_mean": 0.7832314968109131,
      "step": 4139
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.9680687189102173,
      "learning_rate": 8.764056206990445e-06,
      "loss": 0.218,
      "step": 4140
    },
    {
      "epoch": 0.828,
      "loss_align": 0.06791132688522339,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011549675837159157,
      "neg_sim_mean": 0.2968299090862274,
      "pos_sim_mean": 0.9320886731147766,
      "step": 4140
    },
    {
      "epoch": 0.8282,
      "grad_norm": 0.3254607319831848,
      "learning_rate": 8.744325086085248e-06,
      "loss": 0.0691,
      "step": 4141
    },
    {
      "epoch": 0.8282,
      "loss_align": 0.2002071738243103,
      "loss_contrastive": 0.08994930982589722,
      "loss_gen": 0.249847412109375,
      "neg_sim_mean": 0.789742112159729,
      "pos_sim_mean": 0.7997928261756897,
      "step": 4141
    },
    {
      "epoch": 0.8284,
      "grad_norm": 2.4023845195770264,
      "learning_rate": 8.72461407269482e-06,
      "loss": 0.4608,
      "step": 4142
    },
    {
      "epoch": 0.8284,
      "loss_align": 0.11620444059371948,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1463739573955536,
      "neg_sim_mean": 0.7664194107055664,
      "pos_sim_mean": 0.8837955594062805,
      "step": 4142
    },
    {
      "epoch": 0.8286,
      "grad_norm": 1.3474122285842896,
      "learning_rate": 8.704923176426072e-06,
      "loss": 0.2626,
      "step": 4143
    },
    {
      "epoch": 0.8286,
      "loss_align": 0.03936499357223511,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3445695638656616,
      "neg_sim_mean": 0.8112413883209229,
      "pos_sim_mean": 0.9606350064277649,
      "step": 4143
    },
    {
      "epoch": 0.8288,
      "grad_norm": 1.6099939346313477,
      "learning_rate": 8.685252406876115e-06,
      "loss": 0.3839,
      "step": 4144
    },
    {
      "epoch": 0.8288,
      "loss_align": 0.22873848676681519,
      "loss_contrastive": 0.053953349590301514,
      "loss_gen": 0.14031390845775604,
      "neg_sim_mean": 0.7252148389816284,
      "pos_sim_mean": 0.7712615132331848,
      "step": 4144
    },
    {
      "epoch": 0.829,
      "grad_norm": 1.8336621522903442,
      "learning_rate": 8.665601773632226e-06,
      "loss": 0.3755,
      "step": 4145
    },
    {
      "epoch": 0.829,
      "loss_align": 0.13563621044158936,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22931157052516937,
      "neg_sim_mean": 0.6842512488365173,
      "pos_sim_mean": 0.8643637895584106,
      "step": 4145
    },
    {
      "epoch": 0.8292,
      "grad_norm": 2.256150245666504,
      "learning_rate": 8.645971286271904e-06,
      "loss": 0.3649,
      "step": 4146
    },
    {
      "epoch": 0.8292,
      "loss_align": 0.10402745008468628,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22170105576515198,
      "neg_sim_mean": 0.7716913223266602,
      "pos_sim_mean": 0.8959725499153137,
      "step": 4146
    },
    {
      "epoch": 0.8294,
      "grad_norm": 1.88276207447052,
      "learning_rate": 8.626360954362816e-06,
      "loss": 0.3257,
      "step": 4147
    },
    {
      "epoch": 0.8294,
      "loss_align": 0.06636285781860352,
      "loss_contrastive": 0.04710513353347778,
      "loss_gen": 0.18051113188266754,
      "neg_sim_mean": 0.8807422518730164,
      "pos_sim_mean": 0.9336371421813965,
      "step": 4147
    },
    {
      "epoch": 0.8296,
      "grad_norm": 1.609529972076416,
      "learning_rate": 8.606770787462775e-06,
      "loss": 0.2525,
      "step": 4148
    },
    {
      "epoch": 0.8296,
      "loss_align": 0.1581425666809082,
      "loss_contrastive": 0.10007995367050171,
      "loss_gen": 0.21247901022434235,
      "neg_sim_mean": 0.8419373631477356,
      "pos_sim_mean": 0.8418574333190918,
      "step": 4148
    },
    {
      "epoch": 0.8298,
      "grad_norm": 1.9132025241851807,
      "learning_rate": 8.587200795119793e-06,
      "loss": 0.3826,
      "step": 4149
    },
    {
      "epoch": 0.8298,
      "loss_align": 0.254230260848999,
      "loss_contrastive": 0.01183938980102539,
      "loss_gen": 0.35495874285697937,
      "neg_sim_mean": 0.6576091051101685,
      "pos_sim_mean": 0.745769739151001,
      "step": 4149
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.2974753379821777,
      "learning_rate": 8.567650986872062e-06,
      "loss": 0.6106,
      "step": 4150
    },
    {
      "epoch": 0.83,
      "loss_align": 0.17229318618774414,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17291221022605896,
      "neg_sim_mean": 0.585472047328949,
      "pos_sim_mean": 0.8277068138122559,
      "step": 4150
    },
    {
      "epoch": 0.8302,
      "grad_norm": 2.1943769454956055,
      "learning_rate": 8.548121372247918e-06,
      "loss": 0.3452,
      "step": 4151
    },
    {
      "epoch": 0.8302,
      "loss_align": 0.09667927026748657,
      "loss_contrastive": 0.03819417953491211,
      "loss_gen": 0.21147353947162628,
      "neg_sim_mean": 0.8415148854255676,
      "pos_sim_mean": 0.9033207297325134,
      "step": 4151
    },
    {
      "epoch": 0.8304,
      "grad_norm": 1.5854065418243408,
      "learning_rate": 8.528611960765853e-06,
      "loss": 0.3127,
      "step": 4152
    },
    {
      "epoch": 0.8304,
      "loss_align": 0.13854968547821045,
      "loss_contrastive": 0.06079190969467163,
      "loss_gen": 0.24217605590820312,
      "neg_sim_mean": 0.8222422003746033,
      "pos_sim_mean": 0.8614503145217896,
      "step": 4152
    },
    {
      "epoch": 0.8306,
      "grad_norm": 1.9665452241897583,
      "learning_rate": 8.509122761934518e-06,
      "loss": 0.388,
      "step": 4153
    },
    {
      "epoch": 0.8306,
      "loss_align": 0.0944218635559082,
      "loss_contrastive": 0.04770463705062866,
      "loss_gen": 0.14689192175865173,
      "neg_sim_mean": 0.8532827496528625,
      "pos_sim_mean": 0.9055781364440918,
      "step": 4153
    },
    {
      "epoch": 0.8308,
      "grad_norm": 1.3867814540863037,
      "learning_rate": 8.48965378525271e-06,
      "loss": 0.247,
      "step": 4154
    },
    {
      "epoch": 0.8308,
      "loss_align": 0.09222710132598877,
      "loss_contrastive": 0.05073058605194092,
      "loss_gen": 0.2450275421142578,
      "neg_sim_mean": 0.8585034608840942,
      "pos_sim_mean": 0.9077728986740112,
      "step": 4154
    },
    {
      "epoch": 0.831,
      "grad_norm": 1.9855620861053467,
      "learning_rate": 8.470205040209361e-06,
      "loss": 0.3433,
      "step": 4155
    },
    {
      "epoch": 0.831,
      "loss_align": 0.256691038608551,
      "loss_contrastive": 0.0,
      "loss_gen": 0.019826285541057587,
      "neg_sim_mean": 0.42670947313308716,
      "pos_sim_mean": 0.743308961391449,
      "step": 4155
    },
    {
      "epoch": 0.8312,
      "grad_norm": 1.0989885330200195,
      "learning_rate": 8.450776536283594e-06,
      "loss": 0.2765,
      "step": 4156
    },
    {
      "epoch": 0.8312,
      "loss_align": 0.03834211826324463,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27886030077934265,
      "neg_sim_mean": 0.7278075814247131,
      "pos_sim_mean": 0.9616578817367554,
      "step": 4156
    },
    {
      "epoch": 0.8314,
      "grad_norm": 1.924149990081787,
      "learning_rate": 8.431368282944584e-06,
      "loss": 0.3172,
      "step": 4157
    },
    {
      "epoch": 0.8314,
      "loss_align": 0.1458677053451538,
      "loss_contrastive": 0.08595061302185059,
      "loss_gen": 0.2637149393558502,
      "neg_sim_mean": 0.8400828838348389,
      "pos_sim_mean": 0.8541322946548462,
      "step": 4157
    },
    {
      "epoch": 0.8316,
      "grad_norm": 1.8443238735198975,
      "learning_rate": 8.411980289651689e-06,
      "loss": 0.4199,
      "step": 4158
    },
    {
      "epoch": 0.8316,
      "loss_align": 0.12443256378173828,
      "loss_contrastive": 0.01344287395477295,
      "loss_gen": 0.2668125629425049,
      "neg_sim_mean": 0.7890102863311768,
      "pos_sim_mean": 0.8755674362182617,
      "step": 4158
    },
    {
      "epoch": 0.8318,
      "grad_norm": 2.165665864944458,
      "learning_rate": 8.392612565854375e-06,
      "loss": 0.3929,
      "step": 4159
    },
    {
      "epoch": 0.8318,
      "loss_align": 0.13931536674499512,
      "loss_contrastive": 0.06779676675796509,
      "loss_gen": 0.14437177777290344,
      "neg_sim_mean": 0.8284813761711121,
      "pos_sim_mean": 0.8606846332550049,
      "step": 4159
    },
    {
      "epoch": 0.832,
      "grad_norm": 1.6326186656951904,
      "learning_rate": 8.373265120992252e-06,
      "loss": 0.2918,
      "step": 4160
    },
    {
      "epoch": 0.832,
      "loss_align": 0.06023216247558594,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1448802947998047,
      "neg_sim_mean": 0.6815075874328613,
      "pos_sim_mean": 0.9397678375244141,
      "step": 4160
    },
    {
      "epoch": 0.8322,
      "grad_norm": 1.5938493013381958,
      "learning_rate": 8.353937964495029e-06,
      "loss": 0.2051,
      "step": 4161
    },
    {
      "epoch": 0.8322,
      "loss_align": 0.08202791213989258,
      "loss_contrastive": 0.0071721673011779785,
      "loss_gen": 0.21112561225891113,
      "neg_sim_mean": 0.8251442313194275,
      "pos_sim_mean": 0.9179720878601074,
      "step": 4161
    },
    {
      "epoch": 0.8324,
      "grad_norm": 1.608572006225586,
      "learning_rate": 8.334631105782514e-06,
      "loss": 0.294,
      "step": 4162
    },
    {
      "epoch": 0.8324,
      "loss_align": 0.05323386192321777,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1373065859079361,
      "neg_sim_mean": 0.7916868329048157,
      "pos_sim_mean": 0.9467661380767822,
      "step": 4162
    },
    {
      "epoch": 0.8326,
      "grad_norm": 1.8847124576568604,
      "learning_rate": 8.315344554264642e-06,
      "loss": 0.1905,
      "step": 4163
    },
    {
      "epoch": 0.8326,
      "loss_align": 0.08753037452697754,
      "loss_contrastive": 0.07616806030273438,
      "loss_gen": 0.2556472718715668,
      "neg_sim_mean": 0.8886376619338989,
      "pos_sim_mean": 0.9124696254730225,
      "step": 4163
    },
    {
      "epoch": 0.8328,
      "grad_norm": 1.6840941905975342,
      "learning_rate": 8.296078319341443e-06,
      "loss": 0.3523,
      "step": 4164
    },
    {
      "epoch": 0.8328,
      "loss_align": 0.1642659306526184,
      "loss_contrastive": 0.10739409923553467,
      "loss_gen": 0.068879634141922,
      "neg_sim_mean": 0.8431281447410583,
      "pos_sim_mean": 0.8357340693473816,
      "step": 4164
    },
    {
      "epoch": 0.833,
      "grad_norm": 1.092696189880371,
      "learning_rate": 8.276832410403051e-06,
      "loss": 0.246,
      "step": 4165
    },
    {
      "epoch": 0.833,
      "loss_align": 0.08774220943450928,
      "loss_contrastive": 0.029793202877044678,
      "loss_gen": 0.19944240152835846,
      "neg_sim_mean": 0.8420509696006775,
      "pos_sim_mean": 0.9122577905654907,
      "step": 4165
    },
    {
      "epoch": 0.8332,
      "grad_norm": 1.6886060237884521,
      "learning_rate": 8.257606836829678e-06,
      "loss": 0.2908,
      "step": 4166
    },
    {
      "epoch": 0.8332,
      "loss_align": 0.11306703090667725,
      "loss_contrastive": 0.026763856410980225,
      "loss_gen": 0.2552373707294464,
      "neg_sim_mean": 0.8136968016624451,
      "pos_sim_mean": 0.8869329690933228,
      "step": 4166
    },
    {
      "epoch": 0.8334,
      "grad_norm": 2.3088037967681885,
      "learning_rate": 8.238401607991647e-06,
      "loss": 0.3715,
      "step": 4167
    },
    {
      "epoch": 0.8334,
      "loss_align": 0.11581242084503174,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003687923075631261,
      "neg_sim_mean": 0.4170967936515808,
      "pos_sim_mean": 0.8841875791549683,
      "step": 4167
    },
    {
      "epoch": 0.8336,
      "grad_norm": 0.4659267067909241,
      "learning_rate": 8.21921673324933e-06,
      "loss": 0.1195,
      "step": 4168
    },
    {
      "epoch": 0.8336,
      "loss_align": 0.09057986736297607,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24460282921791077,
      "neg_sim_mean": 0.7821141481399536,
      "pos_sim_mean": 0.9094201326370239,
      "step": 4168
    },
    {
      "epoch": 0.8338,
      "grad_norm": 1.6010290384292603,
      "learning_rate": 8.20005222195323e-06,
      "loss": 0.3352,
      "step": 4169
    },
    {
      "epoch": 0.8338,
      "loss_align": 0.11421585083007812,
      "loss_contrastive": 0.0785788893699646,
      "loss_gen": 0.08692692220211029,
      "neg_sim_mean": 0.8643630146980286,
      "pos_sim_mean": 0.8857841491699219,
      "step": 4169
    },
    {
      "epoch": 0.834,
      "grad_norm": 1.410081386566162,
      "learning_rate": 8.180908083443884e-06,
      "loss": 0.2106,
      "step": 4170
    },
    {
      "epoch": 0.834,
      "loss_align": 0.10088849067687988,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11431806534528732,
      "neg_sim_mean": 0.721757173538208,
      "pos_sim_mean": 0.8991115093231201,
      "step": 4170
    },
    {
      "epoch": 0.8342,
      "grad_norm": 1.229414701461792,
      "learning_rate": 8.16178432705192e-06,
      "loss": 0.2152,
      "step": 4171
    },
    {
      "epoch": 0.8342,
      "loss_align": 0.07260847091674805,
      "loss_contrastive": 0.08316195011138916,
      "loss_gen": 0.30190375447273254,
      "neg_sim_mean": 0.9105534553527832,
      "pos_sim_mean": 0.927391529083252,
      "step": 4171
    },
    {
      "epoch": 0.8344,
      "grad_norm": 1.9399516582489014,
      "learning_rate": 8.142680962098015e-06,
      "loss": 0.3845,
      "step": 4172
    },
    {
      "epoch": 0.8344,
      "loss_align": 0.08200633525848389,
      "loss_contrastive": 0.0801154375076294,
      "loss_gen": 0.11426256597042084,
      "neg_sim_mean": 0.8981090784072876,
      "pos_sim_mean": 0.9179936647415161,
      "step": 4172
    },
    {
      "epoch": 0.8346,
      "grad_norm": 1.2082937955856323,
      "learning_rate": 8.123597997892918e-06,
      "loss": 0.2059,
      "step": 4173
    },
    {
      "epoch": 0.8346,
      "loss_align": 0.1067209243774414,
      "loss_contrastive": 0.10630321502685547,
      "loss_gen": 0.3770611882209778,
      "neg_sim_mean": 0.8995822668075562,
      "pos_sim_mean": 0.8932790756225586,
      "step": 4173
    },
    {
      "epoch": 0.8348,
      "grad_norm": 2.4039552211761475,
      "learning_rate": 8.104535443737438e-06,
      "loss": 0.4965,
      "step": 4174
    },
    {
      "epoch": 0.8348,
      "loss_align": 0.10938990116119385,
      "loss_contrastive": 0.033405184745788574,
      "loss_gen": 0.366607129573822,
      "neg_sim_mean": 0.8240152597427368,
      "pos_sim_mean": 0.8906100988388062,
      "step": 4174
    },
    {
      "epoch": 0.835,
      "grad_norm": 2.1759586334228516,
      "learning_rate": 8.085493308922431e-06,
      "loss": 0.48,
      "step": 4175
    },
    {
      "epoch": 0.835,
      "loss_align": 0.08651447296142578,
      "loss_contrastive": 0.00975567102432251,
      "loss_gen": 0.3134973645210266,
      "neg_sim_mean": 0.8232411742210388,
      "pos_sim_mean": 0.9134855270385742,
      "step": 4175
    },
    {
      "epoch": 0.8352,
      "grad_norm": 1.6945393085479736,
      "learning_rate": 8.066471602728803e-06,
      "loss": 0.4012,
      "step": 4176
    },
    {
      "epoch": 0.8352,
      "loss_align": 0.09095853567123413,
      "loss_contrastive": 0.006681561470031738,
      "loss_gen": 0.30294641852378845,
      "neg_sim_mean": 0.8157230019569397,
      "pos_sim_mean": 0.9090414643287659,
      "step": 4176
    },
    {
      "epoch": 0.8354,
      "grad_norm": 2.19482159614563,
      "learning_rate": 8.047470334427503e-06,
      "loss": 0.3947,
      "step": 4177
    },
    {
      "epoch": 0.8354,
      "loss_align": 0.11114907264709473,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16985848546028137,
      "neg_sim_mean": 0.5903232097625732,
      "pos_sim_mean": 0.8888509273529053,
      "step": 4177
    },
    {
      "epoch": 0.8356,
      "grad_norm": 1.663502812385559,
      "learning_rate": 8.028489513279502e-06,
      "loss": 0.281,
      "step": 4178
    },
    {
      "epoch": 0.8356,
      "loss_align": 0.07952272891998291,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18884222209453583,
      "neg_sim_mean": 0.5843319892883301,
      "pos_sim_mean": 0.9204772710800171,
      "step": 4178
    },
    {
      "epoch": 0.8358,
      "grad_norm": 1.5571643114089966,
      "learning_rate": 8.009529148535855e-06,
      "loss": 0.2684,
      "step": 4179
    },
    {
      "epoch": 0.8358,
      "loss_align": 0.15138697624206543,
      "loss_contrastive": 0.002575695514678955,
      "loss_gen": 0.39574629068374634,
      "neg_sim_mean": 0.7511886954307556,
      "pos_sim_mean": 0.8486130237579346,
      "step": 4179
    },
    {
      "epoch": 0.836,
      "grad_norm": 2.372316837310791,
      "learning_rate": 7.99058924943759e-06,
      "loss": 0.5474,
      "step": 4180
    },
    {
      "epoch": 0.836,
      "loss_align": 0.1850430965423584,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008922256529331207,
      "neg_sim_mean": 0.3795701265335083,
      "pos_sim_mean": 0.8149569034576416,
      "step": 4180
    },
    {
      "epoch": 0.8362,
      "grad_norm": 0.7972760796546936,
      "learning_rate": 7.971669825215788e-06,
      "loss": 0.194,
      "step": 4181
    },
    {
      "epoch": 0.8362,
      "loss_align": 0.10910099744796753,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4229642450809479,
      "neg_sim_mean": 0.7413798570632935,
      "pos_sim_mean": 0.8908990025520325,
      "step": 4181
    },
    {
      "epoch": 0.8364,
      "grad_norm": 2.3317394256591797,
      "learning_rate": 7.952770885091548e-06,
      "loss": 0.5321,
      "step": 4182
    },
    {
      "epoch": 0.8364,
      "loss_align": 0.09096449613571167,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1905808299779892,
      "neg_sim_mean": 0.5107645392417908,
      "pos_sim_mean": 0.9090355038642883,
      "step": 4182
    },
    {
      "epoch": 0.8366,
      "grad_norm": 1.7855709791183472,
      "learning_rate": 7.933892438275986e-06,
      "loss": 0.2815,
      "step": 4183
    },
    {
      "epoch": 0.8366,
      "loss_align": 0.07685530185699463,
      "loss_contrastive": 0.0,
      "loss_gen": 0.015540560707449913,
      "neg_sim_mean": 0.29143643379211426,
      "pos_sim_mean": 0.9231446981430054,
      "step": 4183
    },
    {
      "epoch": 0.8368,
      "grad_norm": 0.8365806937217712,
      "learning_rate": 7.915034493970219e-06,
      "loss": 0.0924,
      "step": 4184
    },
    {
      "epoch": 0.8368,
      "loss_align": 0.06232106685638428,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25349026918411255,
      "neg_sim_mean": 0.4888104498386383,
      "pos_sim_mean": 0.9376789331436157,
      "step": 4184
    },
    {
      "epoch": 0.837,
      "grad_norm": 1.9264063835144043,
      "learning_rate": 7.89619706136539e-06,
      "loss": 0.3158,
      "step": 4185
    },
    {
      "epoch": 0.837,
      "loss_align": 0.09385412931442261,
      "loss_contrastive": 0.08647292852401733,
      "loss_gen": 0.20966514945030212,
      "neg_sim_mean": 0.8926187753677368,
      "pos_sim_mean": 0.9061458706855774,
      "step": 4185
    },
    {
      "epoch": 0.8372,
      "grad_norm": 1.9039249420166016,
      "learning_rate": 7.877380149642626e-06,
      "loss": 0.3139,
      "step": 4186
    },
    {
      "epoch": 0.8372,
      "loss_align": 0.08680558204650879,
      "loss_contrastive": 0.0,
      "loss_gen": 0.190378338098526,
      "neg_sim_mean": 0.5591743588447571,
      "pos_sim_mean": 0.9131944179534912,
      "step": 4186
    },
    {
      "epoch": 0.8374,
      "grad_norm": 1.6936308145523071,
      "learning_rate": 7.858583767973072e-06,
      "loss": 0.2772,
      "step": 4187
    },
    {
      "epoch": 0.8374,
      "loss_align": 0.12401032447814941,
      "loss_contrastive": 0.0,
      "loss_gen": 0.5346542596817017,
      "neg_sim_mean": 0.5011280179023743,
      "pos_sim_mean": 0.8759896755218506,
      "step": 4187
    },
    {
      "epoch": 0.8376,
      "grad_norm": 2.3081634044647217,
      "learning_rate": 7.839807925517834e-06,
      "loss": 0.6587,
      "step": 4188
    },
    {
      "epoch": 0.8376,
      "loss_align": 0.07815736532211304,
      "loss_contrastive": 0.009686112403869629,
      "loss_gen": 0.5041135549545288,
      "neg_sim_mean": 0.8315287232398987,
      "pos_sim_mean": 0.921842634677887,
      "step": 4188
    },
    {
      "epoch": 0.8378,
      "grad_norm": 2.4269745349884033,
      "learning_rate": 7.82105263142806e-06,
      "loss": 0.5834,
      "step": 4189
    },
    {
      "epoch": 0.8378,
      "loss_align": 0.09558892250061035,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0069291559047997,
      "neg_sim_mean": 0.33210116624832153,
      "pos_sim_mean": 0.9044110774993896,
      "step": 4189
    },
    {
      "epoch": 0.838,
      "grad_norm": 0.47485432028770447,
      "learning_rate": 7.802317894844835e-06,
      "loss": 0.1025,
      "step": 4190
    },
    {
      "epoch": 0.838,
      "loss_align": 0.058283090591430664,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18785910308361053,
      "neg_sim_mean": 0.8388553857803345,
      "pos_sim_mean": 0.9417169094085693,
      "step": 4190
    },
    {
      "epoch": 0.8382,
      "grad_norm": 1.4973679780960083,
      "learning_rate": 7.783603724899257e-06,
      "loss": 0.2461,
      "step": 4191
    },
    {
      "epoch": 0.8382,
      "loss_align": 0.04630237817764282,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1177104264497757,
      "neg_sim_mean": 0.636529803276062,
      "pos_sim_mean": 0.9536976218223572,
      "step": 4191
    },
    {
      "epoch": 0.8384,
      "grad_norm": 1.561155915260315,
      "learning_rate": 7.764910130712349e-06,
      "loss": 0.164,
      "step": 4192
    },
    {
      "epoch": 0.8384,
      "loss_align": 0.09959161281585693,
      "loss_contrastive": 0.042798638343811035,
      "loss_gen": 0.24417740106582642,
      "neg_sim_mean": 0.8432070016860962,
      "pos_sim_mean": 0.9004083871841431,
      "step": 4192
    },
    {
      "epoch": 0.8386,
      "grad_norm": 1.8731030225753784,
      "learning_rate": 7.746237121395183e-06,
      "loss": 0.3489,
      "step": 4193
    },
    {
      "epoch": 0.8386,
      "loss_align": 0.1688859462738037,
      "loss_contrastive": 0.0,
      "loss_gen": 0.39816731214523315,
      "neg_sim_mean": 0.6707593202590942,
      "pos_sim_mean": 0.8311140537261963,
      "step": 4193
    },
    {
      "epoch": 0.8388,
      "grad_norm": 2.025738000869751,
      "learning_rate": 7.727584706048735e-06,
      "loss": 0.5671,
      "step": 4194
    },
    {
      "epoch": 0.8388,
      "loss_align": 0.18587905168533325,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15695174038410187,
      "neg_sim_mean": 0.5930105447769165,
      "pos_sim_mean": 0.8141209483146667,
      "step": 4194
    },
    {
      "epoch": 0.839,
      "grad_norm": 1.5288608074188232,
      "learning_rate": 7.70895289376397e-06,
      "loss": 0.3428,
      "step": 4195
    },
    {
      "epoch": 0.839,
      "loss_align": 0.09833675622940063,
      "loss_contrastive": 0.08296817541122437,
      "loss_gen": 0.23267889022827148,
      "neg_sim_mean": 0.8846313953399658,
      "pos_sim_mean": 0.9016632437705994,
      "step": 4195
    },
    {
      "epoch": 0.8392,
      "grad_norm": 1.7640751600265503,
      "learning_rate": 7.690341693621805e-06,
      "loss": 0.341,
      "step": 4196
    },
    {
      "epoch": 0.8392,
      "loss_align": 0.21951693296432495,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02578231319785118,
      "neg_sim_mean": 0.22787857055664062,
      "pos_sim_mean": 0.780483067035675,
      "step": 4196
    },
    {
      "epoch": 0.8394,
      "grad_norm": 1.367358922958374,
      "learning_rate": 7.671751114693104e-06,
      "loss": 0.2453,
      "step": 4197
    },
    {
      "epoch": 0.8394,
      "loss_align": 0.18079841136932373,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26073431968688965,
      "neg_sim_mean": 0.6364648342132568,
      "pos_sim_mean": 0.8192015886306763,
      "step": 4197
    },
    {
      "epoch": 0.8396,
      "grad_norm": 2.005559206008911,
      "learning_rate": 7.653181166038715e-06,
      "loss": 0.4415,
      "step": 4198
    },
    {
      "epoch": 0.8396,
      "loss_align": 0.1021222472190857,
      "loss_contrastive": 0.03190487623214722,
      "loss_gen": 0.23892074823379517,
      "neg_sim_mean": 0.8297826051712036,
      "pos_sim_mean": 0.8978777527809143,
      "step": 4198
    },
    {
      "epoch": 0.8398,
      "grad_norm": 1.99225652217865,
      "learning_rate": 7.63463185670939e-06,
      "loss": 0.3449,
      "step": 4199
    },
    {
      "epoch": 0.8398,
      "loss_align": 0.3274611830711365,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0008970377966761589,
      "neg_sim_mean": 0.40306660532951355,
      "pos_sim_mean": 0.6725388169288635,
      "step": 4199
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.1151399612426758,
      "learning_rate": 7.6161031957458494e-06,
      "loss": 0.3284,
      "step": 4200
    },
    {
      "epoch": 0.84,
      "loss_align": 0.08134597539901733,
      "loss_contrastive": 0.08584398031234741,
      "loss_gen": 0.24097022414207458,
      "neg_sim_mean": 0.9044979810714722,
      "pos_sim_mean": 0.9186540246009827,
      "step": 4200
    },
    {
      "epoch": 0.8402,
      "grad_norm": 1.7494707107543945,
      "learning_rate": 7.597595192178702e-06,
      "loss": 0.3326,
      "step": 4201
    },
    {
      "epoch": 0.8402,
      "loss_align": 0.07877284288406372,
      "loss_contrastive": 0.05611485242843628,
      "loss_gen": 0.10608722269535065,
      "neg_sim_mean": 0.8773419857025146,
      "pos_sim_mean": 0.9212271571159363,
      "step": 4201
    },
    {
      "epoch": 0.8404,
      "grad_norm": 1.5346529483795166,
      "learning_rate": 7.579107855028561e-06,
      "loss": 0.1916,
      "step": 4202
    },
    {
      "epoch": 0.8404,
      "loss_align": 0.1229095458984375,
      "loss_contrastive": 0.08520358800888062,
      "loss_gen": 0.2558598816394806,
      "neg_sim_mean": 0.8622940182685852,
      "pos_sim_mean": 0.8770904541015625,
      "step": 4202
    },
    {
      "epoch": 0.8406,
      "grad_norm": 1.4929134845733643,
      "learning_rate": 7.560641193305912e-06,
      "loss": 0.389,
      "step": 4203
    },
    {
      "epoch": 0.8406,
      "loss_align": 0.16551220417022705,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0019626652356237173,
      "neg_sim_mean": 0.3784446120262146,
      "pos_sim_mean": 0.834487795829773,
      "step": 4203
    },
    {
      "epoch": 0.8408,
      "grad_norm": 0.742917537689209,
      "learning_rate": 7.5421952160111884e-06,
      "loss": 0.1675,
      "step": 4204
    },
    {
      "epoch": 0.8408,
      "loss_align": 0.10557723045349121,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10515065491199493,
      "neg_sim_mean": 0.6987009048461914,
      "pos_sim_mean": 0.8944227695465088,
      "step": 4204
    },
    {
      "epoch": 0.841,
      "grad_norm": 1.2553017139434814,
      "learning_rate": 7.523769932134739e-06,
      "loss": 0.2107,
      "step": 4205
    },
    {
      "epoch": 0.841,
      "loss_align": 0.0789804458618164,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09847459942102432,
      "neg_sim_mean": 0.7901406288146973,
      "pos_sim_mean": 0.9210195541381836,
      "step": 4205
    },
    {
      "epoch": 0.8412,
      "grad_norm": 1.1990106105804443,
      "learning_rate": 7.505365350656812e-06,
      "loss": 0.1775,
      "step": 4206
    },
    {
      "epoch": 0.8412,
      "loss_align": 0.1697101593017578,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22255799174308777,
      "neg_sim_mean": 0.6975568532943726,
      "pos_sim_mean": 0.8302898406982422,
      "step": 4206
    },
    {
      "epoch": 0.8414,
      "grad_norm": 1.73560631275177,
      "learning_rate": 7.486981480547567e-06,
      "loss": 0.3923,
      "step": 4207
    },
    {
      "epoch": 0.8414,
      "loss_align": 0.14468348026275635,
      "loss_contrastive": 0.10971736907958984,
      "loss_gen": 0.4987058937549591,
      "neg_sim_mean": 0.8650338649749756,
      "pos_sim_mean": 0.8553165197372437,
      "step": 4207
    },
    {
      "epoch": 0.8416,
      "grad_norm": 2.500633478164673,
      "learning_rate": 7.468618330767113e-06,
      "loss": 0.6566,
      "step": 4208
    },
    {
      "epoch": 0.8416,
      "loss_align": 0.14619368314743042,
      "loss_contrastive": 0.0071032047271728516,
      "loss_gen": 0.24773328006267548,
      "neg_sim_mean": 0.7609094977378845,
      "pos_sim_mean": 0.8538063168525696,
      "step": 4208
    },
    {
      "epoch": 0.8418,
      "grad_norm": 1.9906922578811646,
      "learning_rate": 7.450275910265414e-06,
      "loss": 0.3948,
      "step": 4209
    },
    {
      "epoch": 0.8418,
      "loss_align": 0.08158349990844727,
      "loss_contrastive": 0.08861994743347168,
      "loss_gen": 0.3427216410636902,
      "neg_sim_mean": 0.9070364236831665,
      "pos_sim_mean": 0.9184165000915527,
      "step": 4209
    },
    {
      "epoch": 0.842,
      "grad_norm": 2.0863380432128906,
      "learning_rate": 7.43195422798233e-06,
      "loss": 0.4349,
      "step": 4210
    },
    {
      "epoch": 0.842,
      "loss_align": 0.13321542739868164,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16739428043365479,
      "neg_sim_mean": 0.5634947419166565,
      "pos_sim_mean": 0.8667845726013184,
      "step": 4210
    },
    {
      "epoch": 0.8422,
      "grad_norm": 1.5628101825714111,
      "learning_rate": 7.413653292847617e-06,
      "loss": 0.3006,
      "step": 4211
    },
    {
      "epoch": 0.8422,
      "loss_align": 0.11247450113296509,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009270727634429932,
      "neg_sim_mean": 0.28201958537101746,
      "pos_sim_mean": 0.8875254988670349,
      "step": 4211
    },
    {
      "epoch": 0.8424,
      "grad_norm": 0.6753353476524353,
      "learning_rate": 7.395373113780962e-06,
      "loss": 0.1217,
      "step": 4212
    },
    {
      "epoch": 0.8424,
      "loss_align": 0.08883118629455566,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011591195361688733,
      "neg_sim_mean": 0.32119929790496826,
      "pos_sim_mean": 0.9111688137054443,
      "step": 4212
    },
    {
      "epoch": 0.8426,
      "grad_norm": 0.4879647493362427,
      "learning_rate": 7.377113699691878e-06,
      "loss": 0.09,
      "step": 4213
    },
    {
      "epoch": 0.8426,
      "loss_align": 0.16338706016540527,
      "loss_contrastive": 0.09422624111175537,
      "loss_gen": 0.2654076814651489,
      "neg_sim_mean": 0.8308391571044922,
      "pos_sim_mean": 0.8366129398345947,
      "step": 4213
    },
    {
      "epoch": 0.8428,
      "grad_norm": 1.6060514450073242,
      "learning_rate": 7.358875059479792e-06,
      "loss": 0.4401,
      "step": 4214
    },
    {
      "epoch": 0.8428,
      "loss_align": 0.08162802457809448,
      "loss_contrastive": 0.0,
      "loss_gen": 0.064305379986763,
      "neg_sim_mean": 0.6548978686332703,
      "pos_sim_mean": 0.9183719754219055,
      "step": 4214
    },
    {
      "epoch": 0.843,
      "grad_norm": 0.97359299659729,
      "learning_rate": 7.34065720203399e-06,
      "loss": 0.1459,
      "step": 4215
    },
    {
      "epoch": 0.843,
      "loss_align": 0.06514537334442139,
      "loss_contrastive": 0.05090653896331787,
      "loss_gen": 0.12039238214492798,
      "neg_sim_mean": 0.8857611417770386,
      "pos_sim_mean": 0.9348546266555786,
      "step": 4215
    },
    {
      "epoch": 0.8432,
      "grad_norm": 1.5026311874389648,
      "learning_rate": 7.322460136233622e-06,
      "loss": 0.1916,
      "step": 4216
    },
    {
      "epoch": 0.8432,
      "loss_align": 0.12453454732894897,
      "loss_contrastive": 0.09940510988235474,
      "loss_gen": 0.19022557139396667,
      "neg_sim_mean": 0.8748705387115479,
      "pos_sim_mean": 0.875465452671051,
      "step": 4216
    },
    {
      "epoch": 0.8434,
      "grad_norm": 1.4743545055389404,
      "learning_rate": 7.3042838709477475e-06,
      "loss": 0.3267,
      "step": 4217
    },
    {
      "epoch": 0.8434,
      "loss_align": 0.10135364532470703,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2567868232727051,
      "neg_sim_mean": 0.7073067426681519,
      "pos_sim_mean": 0.898646354675293,
      "step": 4217
    },
    {
      "epoch": 0.8436,
      "grad_norm": 1.7073698043823242,
      "learning_rate": 7.286128415035248e-06,
      "loss": 0.3581,
      "step": 4218
    },
    {
      "epoch": 0.8436,
      "loss_align": 0.08268964290618896,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11531038582324982,
      "neg_sim_mean": 0.8100451827049255,
      "pos_sim_mean": 0.917310357093811,
      "step": 4218
    },
    {
      "epoch": 0.8438,
      "grad_norm": 1.1040241718292236,
      "learning_rate": 7.267993777344856e-06,
      "loss": 0.198,
      "step": 4219
    },
    {
      "epoch": 0.8438,
      "loss_align": 0.09763652086257935,
      "loss_contrastive": 0.024457931518554688,
      "loss_gen": 0.1756369173526764,
      "neg_sim_mean": 0.8268213868141174,
      "pos_sim_mean": 0.9023634791374207,
      "step": 4219
    },
    {
      "epoch": 0.844,
      "grad_norm": 1.45977783203125,
      "learning_rate": 7.249879966715173e-06,
      "loss": 0.2762,
      "step": 4220
    },
    {
      "epoch": 0.844,
      "loss_align": 0.04163360595703125,
      "loss_contrastive": 0.03585648536682129,
      "loss_gen": 0.23714910447597504,
      "neg_sim_mean": 0.8942228555679321,
      "pos_sim_mean": 0.9583663940429688,
      "step": 4220
    },
    {
      "epoch": 0.8442,
      "grad_norm": 1.7542998790740967,
      "learning_rate": 7.2317869919746705e-06,
      "loss": 0.2831,
      "step": 4221
    },
    {
      "epoch": 0.8442,
      "loss_align": 0.08748972415924072,
      "loss_contrastive": 0.0589679479598999,
      "loss_gen": 0.14165015518665314,
      "neg_sim_mean": 0.8714781999588013,
      "pos_sim_mean": 0.9125102758407593,
      "step": 4221
    },
    {
      "epoch": 0.8444,
      "grad_norm": 1.5302106142044067,
      "learning_rate": 7.213714861941628e-06,
      "loss": 0.2362,
      "step": 4222
    },
    {
      "epoch": 0.8444,
      "loss_align": 0.10131686925888062,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00357560021802783,
      "neg_sim_mean": 0.3602033257484436,
      "pos_sim_mean": 0.8986831307411194,
      "step": 4222
    },
    {
      "epoch": 0.8446,
      "grad_norm": 0.48068907856941223,
      "learning_rate": 7.195663585424195e-06,
      "loss": 0.1049,
      "step": 4223
    },
    {
      "epoch": 0.8446,
      "loss_align": 0.09198582172393799,
      "loss_contrastive": 0.04302924871444702,
      "loss_gen": 0.34748712182044983,
      "neg_sim_mean": 0.8510434031486511,
      "pos_sim_mean": 0.908014178276062,
      "step": 4223
    },
    {
      "epoch": 0.8448,
      "grad_norm": 2.449479818344116,
      "learning_rate": 7.1776331712203385e-06,
      "loss": 0.4446,
      "step": 4224
    },
    {
      "epoch": 0.8448,
      "loss_align": 0.02973484992980957,
      "loss_contrastive": 0.0670403242111206,
      "loss_gen": 0.04855524003505707,
      "neg_sim_mean": 0.9373054504394531,
      "pos_sim_mean": 0.9702651500701904,
      "step": 4224
    },
    {
      "epoch": 0.845,
      "grad_norm": 0.660975456237793,
      "learning_rate": 7.159623628117857e-06,
      "loss": 0.0863,
      "step": 4225
    },
    {
      "epoch": 0.845,
      "loss_align": 0.10700905323028564,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2085990607738495,
      "neg_sim_mean": 0.7558614015579224,
      "pos_sim_mean": 0.8929909467697144,
      "step": 4225
    },
    {
      "epoch": 0.8452,
      "grad_norm": 1.6844666004180908,
      "learning_rate": 7.1416349648943894e-06,
      "loss": 0.3156,
      "step": 4226
    },
    {
      "epoch": 0.8452,
      "loss_align": 0.08645033836364746,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3205868601799011,
      "neg_sim_mean": 0.7048966884613037,
      "pos_sim_mean": 0.9135496616363525,
      "step": 4226
    },
    {
      "epoch": 0.8454,
      "grad_norm": 1.9200035333633423,
      "learning_rate": 7.123667190317396e-06,
      "loss": 0.407,
      "step": 4227
    },
    {
      "epoch": 0.8454,
      "loss_align": 0.08708280324935913,
      "loss_contrastive": 0.06493312120437622,
      "loss_gen": 0.1902620494365692,
      "neg_sim_mean": 0.8778502941131592,
      "pos_sim_mean": 0.9129171967506409,
      "step": 4227
    },
    {
      "epoch": 0.8456,
      "grad_norm": 1.8601619005203247,
      "learning_rate": 7.105720313144143e-06,
      "loss": 0.2851,
      "step": 4228
    },
    {
      "epoch": 0.8456,
      "loss_align": 0.09710961580276489,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008371547795832157,
      "neg_sim_mean": 0.474377304315567,
      "pos_sim_mean": 0.9028903841972351,
      "step": 4228
    },
    {
      "epoch": 0.8458,
      "grad_norm": 0.6261098384857178,
      "learning_rate": 7.087794342121723e-06,
      "loss": 0.1055,
      "step": 4229
    },
    {
      "epoch": 0.8458,
      "loss_align": 0.15235698223114014,
      "loss_contrastive": 0.10392117500305176,
      "loss_gen": 0.3166288435459137,
      "neg_sim_mean": 0.8515641689300537,
      "pos_sim_mean": 0.8476430177688599,
      "step": 4229
    },
    {
      "epoch": 0.846,
      "grad_norm": 2.2917144298553467,
      "learning_rate": 7.0698892859870245e-06,
      "loss": 0.4815,
      "step": 4230
    },
    {
      "epoch": 0.846,
      "loss_align": 0.14247024059295654,
      "loss_contrastive": 0.09981739521026611,
      "loss_gen": 0.17470785975456238,
      "neg_sim_mean": 0.8573471307754517,
      "pos_sim_mean": 0.8575297594070435,
      "step": 4230
    },
    {
      "epoch": 0.8462,
      "grad_norm": 1.7741541862487793,
      "learning_rate": 7.052005153466779e-06,
      "loss": 0.3292,
      "step": 4231
    },
    {
      "epoch": 0.8462,
      "loss_align": 0.0914989709854126,
      "loss_contrastive": 0.0,
      "loss_gen": 0.28844183683395386,
      "neg_sim_mean": 0.8018720746040344,
      "pos_sim_mean": 0.9085010290145874,
      "step": 4231
    },
    {
      "epoch": 0.8464,
      "grad_norm": 2.29274845123291,
      "learning_rate": 7.034141953277485e-06,
      "loss": 0.3799,
      "step": 4232
    },
    {
      "epoch": 0.8464,
      "loss_align": 0.05674588680267334,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01977636106312275,
      "neg_sim_mean": 0.36368218064308167,
      "pos_sim_mean": 0.9432541131973267,
      "step": 4232
    },
    {
      "epoch": 0.8466,
      "grad_norm": 0.8633778691291809,
      "learning_rate": 7.01629969412545e-06,
      "loss": 0.0765,
      "step": 4233
    },
    {
      "epoch": 0.8466,
      "loss_align": 0.1067575216293335,
      "loss_contrastive": 0.0,
      "loss_gen": 0.016512691974639893,
      "neg_sim_mean": 0.307528018951416,
      "pos_sim_mean": 0.8932424783706665,
      "step": 4233
    },
    {
      "epoch": 0.8468,
      "grad_norm": 0.8163774609565735,
      "learning_rate": 6.998478384706769e-06,
      "loss": 0.1233,
      "step": 4234
    },
    {
      "epoch": 0.8468,
      "loss_align": 0.0797266960144043,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004809355828911066,
      "neg_sim_mean": 0.2659045159816742,
      "pos_sim_mean": 0.9202733039855957,
      "step": 4234
    },
    {
      "epoch": 0.847,
      "grad_norm": 0.5500025153160095,
      "learning_rate": 6.980678033707333e-06,
      "loss": 0.0845,
      "step": 4235
    },
    {
      "epoch": 0.847,
      "loss_align": 0.13562166690826416,
      "loss_contrastive": 0.09274446964263916,
      "loss_gen": 0.20940403640270233,
      "neg_sim_mean": 0.8571227788925171,
      "pos_sim_mean": 0.8643783330917358,
      "step": 4235
    },
    {
      "epoch": 0.8472,
      "grad_norm": 2.0111982822418213,
      "learning_rate": 6.962898649802823e-06,
      "loss": 0.3562,
      "step": 4236
    },
    {
      "epoch": 0.8472,
      "loss_align": 0.07767307758331299,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25740036368370056,
      "neg_sim_mean": 0.8195412755012512,
      "pos_sim_mean": 0.922326922416687,
      "step": 4236
    },
    {
      "epoch": 0.8474,
      "grad_norm": 1.712483286857605,
      "learning_rate": 6.945140241658687e-06,
      "loss": 0.3351,
      "step": 4237
    },
    {
      "epoch": 0.8474,
      "loss_align": 0.13088464736938477,
      "loss_contrastive": 0.08333998918533325,
      "loss_gen": 0.20338314771652222,
      "neg_sim_mean": 0.8524553179740906,
      "pos_sim_mean": 0.8691153526306152,
      "step": 4237
    },
    {
      "epoch": 0.8476,
      "grad_norm": 1.5832992792129517,
      "learning_rate": 6.927402817930168e-06,
      "loss": 0.3443,
      "step": 4238
    },
    {
      "epoch": 0.8476,
      "loss_align": 0.0623776912689209,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10242577642202377,
      "neg_sim_mean": 0.7070280313491821,
      "pos_sim_mean": 0.9376223087310791,
      "step": 4238
    },
    {
      "epoch": 0.8478,
      "grad_norm": 1.2155938148498535,
      "learning_rate": 6.909686387262254e-06,
      "loss": 0.1648,
      "step": 4239
    },
    {
      "epoch": 0.8478,
      "loss_align": 0.08754849433898926,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2874910831451416,
      "neg_sim_mean": 0.6504915952682495,
      "pos_sim_mean": 0.9124515056610107,
      "step": 4239
    },
    {
      "epoch": 0.848,
      "grad_norm": 1.8609446287155151,
      "learning_rate": 6.891990958289723e-06,
      "loss": 0.375,
      "step": 4240
    },
    {
      "epoch": 0.848,
      "loss_align": 0.13266193866729736,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0068005044013261795,
      "neg_sim_mean": 0.3643733859062195,
      "pos_sim_mean": 0.8673380613327026,
      "step": 4240
    },
    {
      "epoch": 0.8482,
      "grad_norm": 0.6148459911346436,
      "learning_rate": 6.874316539637127e-06,
      "loss": 0.1395,
      "step": 4241
    },
    {
      "epoch": 0.8482,
      "loss_align": 0.11236560344696045,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02627580240368843,
      "neg_sim_mean": 0.32309532165527344,
      "pos_sim_mean": 0.8876343965530396,
      "step": 4241
    },
    {
      "epoch": 0.8484,
      "grad_norm": 0.9039224982261658,
      "learning_rate": 6.8566631399187505e-06,
      "loss": 0.1386,
      "step": 4242
    },
    {
      "epoch": 0.8484,
      "loss_align": 0.0617411732673645,
      "loss_contrastive": 0.0,
      "loss_gen": 0.34586894512176514,
      "neg_sim_mean": 0.7633145451545715,
      "pos_sim_mean": 0.9382588267326355,
      "step": 4242
    },
    {
      "epoch": 0.8486,
      "grad_norm": 2.193739891052246,
      "learning_rate": 6.839030767738652e-06,
      "loss": 0.4076,
      "step": 4243
    },
    {
      "epoch": 0.8486,
      "loss_align": 0.09537816047668457,
      "loss_contrastive": 0.06418663263320923,
      "loss_gen": 0.32827678322792053,
      "neg_sim_mean": 0.8688084483146667,
      "pos_sim_mean": 0.9046218395233154,
      "step": 4243
    },
    {
      "epoch": 0.8488,
      "grad_norm": 1.6016861200332642,
      "learning_rate": 6.821419431690629e-06,
      "loss": 0.4314,
      "step": 4244
    },
    {
      "epoch": 0.8488,
      "loss_align": 0.07132542133331299,
      "loss_contrastive": 0.0101814866065979,
      "loss_gen": 0.15487797558307648,
      "neg_sim_mean": 0.838856041431427,
      "pos_sim_mean": 0.928674578666687,
      "step": 4244
    },
    {
      "epoch": 0.849,
      "grad_norm": 1.7111109495162964,
      "learning_rate": 6.803829140358237e-06,
      "loss": 0.2274,
      "step": 4245
    },
    {
      "epoch": 0.849,
      "loss_align": 0.09903007745742798,
      "loss_contrastive": 0.0807848572731018,
      "loss_gen": 0.5000956654548645,
      "neg_sim_mean": 0.8817547559738159,
      "pos_sim_mean": 0.900969922542572,
      "step": 4245
    },
    {
      "epoch": 0.8492,
      "grad_norm": 2.3991503715515137,
      "learning_rate": 6.786259902314768e-06,
      "loss": 0.6088,
      "step": 4246
    },
    {
      "epoch": 0.8492,
      "loss_align": 0.10433149337768555,
      "loss_contrastive": 0.06823688745498657,
      "loss_gen": 0.20683296024799347,
      "neg_sim_mean": 0.8639053702354431,
      "pos_sim_mean": 0.8956685066223145,
      "step": 4246
    },
    {
      "epoch": 0.8494,
      "grad_norm": 1.691886305809021,
      "learning_rate": 6.7687117261232616e-06,
      "loss": 0.3194,
      "step": 4247
    },
    {
      "epoch": 0.8494,
      "loss_align": 0.05274796485900879,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2625252306461334,
      "neg_sim_mean": 0.8397880792617798,
      "pos_sim_mean": 0.9472520351409912,
      "step": 4247
    },
    {
      "epoch": 0.8496,
      "grad_norm": 1.771600604057312,
      "learning_rate": 6.751184620336471e-06,
      "loss": 0.3153,
      "step": 4248
    },
    {
      "epoch": 0.8496,
      "loss_align": 0.126481831073761,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0898851826786995,
      "neg_sim_mean": 0.5635575652122498,
      "pos_sim_mean": 0.873518168926239,
      "step": 4248
    },
    {
      "epoch": 0.8498,
      "grad_norm": 1.2575072050094604,
      "learning_rate": 6.7336785934969e-06,
      "loss": 0.2164,
      "step": 4249
    },
    {
      "epoch": 0.8498,
      "loss_align": 0.07246840000152588,
      "loss_contrastive": 0.05796569585800171,
      "loss_gen": 0.26038622856140137,
      "neg_sim_mean": 0.8854972720146179,
      "pos_sim_mean": 0.9275315999984741,
      "step": 4249
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.5661585330963135,
      "learning_rate": 6.716193654136788e-06,
      "loss": 0.3398,
      "step": 4250
    },
    {
      "epoch": 0.85,
      "loss_align": 0.13138794898986816,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2549609839916229,
      "neg_sim_mean": 0.6897655725479126,
      "pos_sim_mean": 0.8686120510101318,
      "step": 4250
    },
    {
      "epoch": 0.8502,
      "grad_norm": 1.8795539140701294,
      "learning_rate": 6.698729810778065e-06,
      "loss": 0.3863,
      "step": 4251
    },
    {
      "epoch": 0.8502,
      "loss_align": 0.33165186643600464,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007364292163401842,
      "neg_sim_mean": 0.3900056481361389,
      "pos_sim_mean": 0.6683481335639954,
      "step": 4251
    },
    {
      "epoch": 0.8504,
      "grad_norm": 1.4114207029342651,
      "learning_rate": 6.681287071932407e-06,
      "loss": 0.339,
      "step": 4252
    },
    {
      "epoch": 0.8504,
      "loss_align": 0.12491500377655029,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013105952180922031,
      "neg_sim_mean": 0.2779256999492645,
      "pos_sim_mean": 0.8750849962234497,
      "step": 4252
    },
    {
      "epoch": 0.8506,
      "grad_norm": 0.7758616805076599,
      "learning_rate": 6.663865446101192e-06,
      "loss": 0.138,
      "step": 4253
    },
    {
      "epoch": 0.8506,
      "loss_align": 0.0746985673904419,
      "loss_contrastive": 0.003476738929748535,
      "loss_gen": 0.21408917009830475,
      "neg_sim_mean": 0.8287781476974487,
      "pos_sim_mean": 0.9253014326095581,
      "step": 4253
    },
    {
      "epoch": 0.8508,
      "grad_norm": 1.613343358039856,
      "learning_rate": 6.646464941775499e-06,
      "loss": 0.2892,
      "step": 4254
    },
    {
      "epoch": 0.8508,
      "loss_align": 0.052738189697265625,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15811757743358612,
      "neg_sim_mean": 0.8258525729179382,
      "pos_sim_mean": 0.9472618103027344,
      "step": 4254
    },
    {
      "epoch": 0.851,
      "grad_norm": 1.6023756265640259,
      "learning_rate": 6.629085567436133e-06,
      "loss": 0.2109,
      "step": 4255
    },
    {
      "epoch": 0.851,
      "loss_align": 0.07312148809432983,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009022501297295094,
      "neg_sim_mean": 0.30133533477783203,
      "pos_sim_mean": 0.9268785119056702,
      "step": 4255
    },
    {
      "epoch": 0.8512,
      "grad_norm": 0.7767323851585388,
      "learning_rate": 6.611727331553586e-06,
      "loss": 0.0821,
      "step": 4256
    },
    {
      "epoch": 0.8512,
      "loss_align": 0.10447812080383301,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2743392288684845,
      "neg_sim_mean": 0.7486468553543091,
      "pos_sim_mean": 0.895521879196167,
      "step": 4256
    },
    {
      "epoch": 0.8514,
      "grad_norm": 1.9467496871948242,
      "learning_rate": 6.594390242588044e-06,
      "loss": 0.3788,
      "step": 4257
    },
    {
      "epoch": 0.8514,
      "loss_align": 0.08411675691604614,
      "loss_contrastive": 0.08807098865509033,
      "loss_gen": 0.42161184549331665,
      "neg_sim_mean": 0.9039542078971863,
      "pos_sim_mean": 0.9158832430839539,
      "step": 4257
    },
    {
      "epoch": 0.8516,
      "grad_norm": 2.380204677581787,
      "learning_rate": 6.577074308989406e-06,
      "loss": 0.5163,
      "step": 4258
    },
    {
      "epoch": 0.8516,
      "loss_align": 0.11685848236083984,
      "loss_contrastive": 0.044531822204589844,
      "loss_gen": 0.27036526799201965,
      "neg_sim_mean": 0.8276733160018921,
      "pos_sim_mean": 0.8831415176391602,
      "step": 4258
    },
    {
      "epoch": 0.8518,
      "grad_norm": 2.157306671142578,
      "learning_rate": 6.559779539197231e-06,
      "loss": 0.3926,
      "step": 4259
    },
    {
      "epoch": 0.8518,
      "loss_align": 0.10821312665939331,
      "loss_contrastive": 0.0,
      "loss_gen": 0.187205970287323,
      "neg_sim_mean": 0.7676867842674255,
      "pos_sim_mean": 0.8917868733406067,
      "step": 4259
    },
    {
      "epoch": 0.852,
      "grad_norm": 1.7603545188903809,
      "learning_rate": 6.542505941640803e-06,
      "loss": 0.2954,
      "step": 4260
    },
    {
      "epoch": 0.852,
      "loss_align": 0.13057368993759155,
      "loss_contrastive": 0.04132211208343506,
      "loss_gen": 0.2675340473651886,
      "neg_sim_mean": 0.8107483983039856,
      "pos_sim_mean": 0.8694263100624084,
      "step": 4260
    },
    {
      "epoch": 0.8522,
      "grad_norm": 2.060285806655884,
      "learning_rate": 6.52525352473905e-06,
      "loss": 0.4031,
      "step": 4261
    },
    {
      "epoch": 0.8522,
      "loss_align": 0.11782020330429077,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10690971463918686,
      "neg_sim_mean": 0.7120300531387329,
      "pos_sim_mean": 0.8821797966957092,
      "step": 4261
    },
    {
      "epoch": 0.8524,
      "grad_norm": 1.320199966430664,
      "learning_rate": 6.508022296900601e-06,
      "loss": 0.2247,
      "step": 4262
    },
    {
      "epoch": 0.8524,
      "loss_align": 0.1605851650238037,
      "loss_contrastive": 0.0,
      "loss_gen": 0.05968248471617699,
      "neg_sim_mean": 0.628993809223175,
      "pos_sim_mean": 0.8394148349761963,
      "step": 4262
    },
    {
      "epoch": 0.8526,
      "grad_norm": 1.4133036136627197,
      "learning_rate": 6.490812266523716e-06,
      "loss": 0.2203,
      "step": 4263
    },
    {
      "epoch": 0.8526,
      "loss_align": 0.14780080318450928,
      "loss_contrastive": 0.0037978291511535645,
      "loss_gen": 0.30705443024635315,
      "neg_sim_mean": 0.7559970021247864,
      "pos_sim_mean": 0.8521991968154907,
      "step": 4263
    },
    {
      "epoch": 0.8528,
      "grad_norm": 2.3492236137390137,
      "learning_rate": 6.47362344199639e-06,
      "loss": 0.4553,
      "step": 4264
    },
    {
      "epoch": 0.8528,
      "loss_align": 0.07500636577606201,
      "loss_contrastive": 0.0,
      "loss_gen": 0.025028115138411522,
      "neg_sim_mean": 0.3238895535469055,
      "pos_sim_mean": 0.924993634223938,
      "step": 4264
    },
    {
      "epoch": 0.853,
      "grad_norm": 1.2049349546432495,
      "learning_rate": 6.456455831696234e-06,
      "loss": 0.1,
      "step": 4265
    },
    {
      "epoch": 0.853,
      "loss_align": 0.06330353021621704,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4309447109699249,
      "neg_sim_mean": 0.8291031122207642,
      "pos_sim_mean": 0.936696469783783,
      "step": 4265
    },
    {
      "epoch": 0.8532,
      "grad_norm": 2.2390129566192627,
      "learning_rate": 6.439309443990532e-06,
      "loss": 0.4942,
      "step": 4266
    },
    {
      "epoch": 0.8532,
      "loss_align": 0.07795572280883789,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11513745784759521,
      "neg_sim_mean": 0.8208645582199097,
      "pos_sim_mean": 0.9220442771911621,
      "step": 4266
    },
    {
      "epoch": 0.8534,
      "grad_norm": 1.3399659395217896,
      "learning_rate": 6.422184287236227e-06,
      "loss": 0.1931,
      "step": 4267
    },
    {
      "epoch": 0.8534,
      "loss_align": 0.19629734754562378,
      "loss_contrastive": 0.006806910037994385,
      "loss_gen": 0.16246110200881958,
      "neg_sim_mean": 0.7105095386505127,
      "pos_sim_mean": 0.8037026524543762,
      "step": 4267
    },
    {
      "epoch": 0.8536,
      "grad_norm": 1.6573987007141113,
      "learning_rate": 6.405080369779898e-06,
      "loss": 0.3596,
      "step": 4268
    },
    {
      "epoch": 0.8536,
      "loss_align": 0.11024576425552368,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26290374994277954,
      "neg_sim_mean": 0.6671456098556519,
      "pos_sim_mean": 0.8897542357444763,
      "step": 4268
    },
    {
      "epoch": 0.8538,
      "grad_norm": 2.1981592178344727,
      "learning_rate": 6.3879976999578154e-06,
      "loss": 0.3731,
      "step": 4269
    },
    {
      "epoch": 0.8538,
      "loss_align": 0.07431787252426147,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02426139824092388,
      "neg_sim_mean": 0.3932308852672577,
      "pos_sim_mean": 0.9256821274757385,
      "step": 4269
    },
    {
      "epoch": 0.854,
      "grad_norm": 1.122033715248108,
      "learning_rate": 6.370936286095841e-06,
      "loss": 0.0986,
      "step": 4270
    },
    {
      "epoch": 0.854,
      "loss_align": 0.07928532361984253,
      "loss_contrastive": 0.0,
      "loss_gen": 0.022958777844905853,
      "neg_sim_mean": 0.22478832304477692,
      "pos_sim_mean": 0.9207146763801575,
      "step": 4270
    },
    {
      "epoch": 0.8542,
      "grad_norm": 1.030746579170227,
      "learning_rate": 6.353896136509524e-06,
      "loss": 0.1022,
      "step": 4271
    },
    {
      "epoch": 0.8542,
      "loss_align": 0.11085778474807739,
      "loss_contrastive": 0.044142723083496094,
      "loss_gen": 0.19519905745983124,
      "neg_sim_mean": 0.8332849144935608,
      "pos_sim_mean": 0.8891422152519226,
      "step": 4271
    },
    {
      "epoch": 0.8544,
      "grad_norm": 1.6672255992889404,
      "learning_rate": 6.336877259504004e-06,
      "loss": 0.3114,
      "step": 4272
    },
    {
      "epoch": 0.8544,
      "loss_align": 0.08340072631835938,
      "loss_contrastive": 0.017890214920043945,
      "loss_gen": 0.18168304860591888,
      "neg_sim_mean": 0.8344894647598267,
      "pos_sim_mean": 0.9165992736816406,
      "step": 4272
    },
    {
      "epoch": 0.8546,
      "grad_norm": 1.5876753330230713,
      "learning_rate": 6.319879663374067e-06,
      "loss": 0.2672,
      "step": 4273
    },
    {
      "epoch": 0.8546,
      "loss_align": 0.0736885666847229,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2510376572608948,
      "neg_sim_mean": 0.7545585632324219,
      "pos_sim_mean": 0.9263114333152771,
      "step": 4273
    },
    {
      "epoch": 0.8548,
      "grad_norm": 1.5943821668624878,
      "learning_rate": 6.30290335640416e-06,
      "loss": 0.3247,
      "step": 4274
    },
    {
      "epoch": 0.8548,
      "loss_align": 0.17838120460510254,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015977013390511274,
      "neg_sim_mean": 0.2922121286392212,
      "pos_sim_mean": 0.8216187953948975,
      "step": 4274
    },
    {
      "epoch": 0.855,
      "grad_norm": 0.6086339950561523,
      "learning_rate": 6.28594834686832e-06,
      "loss": 0.18,
      "step": 4275
    },
    {
      "epoch": 0.855,
      "loss_align": 0.10955154895782471,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007636526599526405,
      "neg_sim_mean": 0.4601312577724457,
      "pos_sim_mean": 0.8904484510421753,
      "step": 4275
    },
    {
      "epoch": 0.8552,
      "grad_norm": 0.7162770628929138,
      "learning_rate": 6.269014643030213e-06,
      "loss": 0.1172,
      "step": 4276
    },
    {
      "epoch": 0.8552,
      "loss_align": 0.09699952602386475,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2393471896648407,
      "neg_sim_mean": 0.7993828058242798,
      "pos_sim_mean": 0.9030004739761353,
      "step": 4276
    },
    {
      "epoch": 0.8554,
      "grad_norm": 2.0352861881256104,
      "learning_rate": 6.252102253143122e-06,
      "loss": 0.3363,
      "step": 4277
    },
    {
      "epoch": 0.8554,
      "loss_align": 0.11080491542816162,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16966699063777924,
      "neg_sim_mean": 0.7014739513397217,
      "pos_sim_mean": 0.8891950845718384,
      "step": 4277
    },
    {
      "epoch": 0.8556,
      "grad_norm": 1.7376445531845093,
      "learning_rate": 6.235211185449919e-06,
      "loss": 0.2805,
      "step": 4278
    },
    {
      "epoch": 0.8556,
      "loss_align": 0.06035125255584717,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001995772123336792,
      "neg_sim_mean": 0.30585765838623047,
      "pos_sim_mean": 0.9396487474441528,
      "step": 4278
    },
    {
      "epoch": 0.8558,
      "grad_norm": 0.3239366412162781,
      "learning_rate": 6.21834144818314e-06,
      "loss": 0.0623,
      "step": 4279
    },
    {
      "epoch": 0.8558,
      "loss_align": 0.1303797960281372,
      "loss_contrastive": 0.03813743591308594,
      "loss_gen": 0.2942292094230652,
      "neg_sim_mean": 0.8077576160430908,
      "pos_sim_mean": 0.8696202039718628,
      "step": 4279
    },
    {
      "epoch": 0.856,
      "grad_norm": 2.2527215480804443,
      "learning_rate": 6.201493049564882e-06,
      "loss": 0.4292,
      "step": 4280
    },
    {
      "epoch": 0.856,
      "loss_align": 0.0811498761177063,
      "loss_contrastive": 0.05960017442703247,
      "loss_gen": 0.23704786598682404,
      "neg_sim_mean": 0.8784502744674683,
      "pos_sim_mean": 0.9188501238822937,
      "step": 4280
    },
    {
      "epoch": 0.8562,
      "grad_norm": 2.0235166549682617,
      "learning_rate": 6.184665997806832e-06,
      "loss": 0.3253,
      "step": 4281
    },
    {
      "epoch": 0.8562,
      "loss_align": 0.0789405107498169,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18136318027973175,
      "neg_sim_mean": 0.7687867283821106,
      "pos_sim_mean": 0.9210594892501831,
      "step": 4281
    },
    {
      "epoch": 0.8564,
      "grad_norm": 1.4742830991744995,
      "learning_rate": 6.167860301110284e-06,
      "loss": 0.2603,
      "step": 4282
    },
    {
      "epoch": 0.8564,
      "loss_align": 0.09140408039093018,
      "loss_contrastive": 0.05083000659942627,
      "loss_gen": 0.2787744104862213,
      "neg_sim_mean": 0.8594259023666382,
      "pos_sim_mean": 0.9085959196090698,
      "step": 4282
    },
    {
      "epoch": 0.8566,
      "grad_norm": 1.9888619184494019,
      "learning_rate": 6.151075967666164e-06,
      "loss": 0.3763,
      "step": 4283
    },
    {
      "epoch": 0.8566,
      "loss_align": 0.2009803056716919,
      "loss_contrastive": 0.04283648729324341,
      "loss_gen": 0.2745225727558136,
      "neg_sim_mean": 0.7418561577796936,
      "pos_sim_mean": 0.7990196943283081,
      "step": 4283
    },
    {
      "epoch": 0.8568,
      "grad_norm": 2.000629425048828,
      "learning_rate": 6.134313005654929e-06,
      "loss": 0.4806,
      "step": 4284
    },
    {
      "epoch": 0.8568,
      "loss_align": 0.07787841558456421,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24734818935394287,
      "neg_sim_mean": 0.5414342284202576,
      "pos_sim_mean": 0.9221215844154358,
      "step": 4284
    },
    {
      "epoch": 0.857,
      "grad_norm": 1.5307400226593018,
      "learning_rate": 6.117571423246654e-06,
      "loss": 0.3252,
      "step": 4285
    },
    {
      "epoch": 0.857,
      "loss_align": 0.08175945281982422,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004645425360649824,
      "neg_sim_mean": 0.4295259118080139,
      "pos_sim_mean": 0.9182405471801758,
      "step": 4285
    },
    {
      "epoch": 0.8572,
      "grad_norm": 0.426140159368515,
      "learning_rate": 6.100851228600973e-06,
      "loss": 0.0864,
      "step": 4286
    },
    {
      "epoch": 0.8572,
      "loss_align": 0.0811428427696228,
      "loss_contrastive": 0.0,
      "loss_gen": 0.08746059983968735,
      "neg_sim_mean": 0.49842870235443115,
      "pos_sim_mean": 0.9188571572303772,
      "step": 4286
    },
    {
      "epoch": 0.8574,
      "grad_norm": 1.3298568725585938,
      "learning_rate": 6.084152429867113e-06,
      "loss": 0.1686,
      "step": 4287
    },
    {
      "epoch": 0.8574,
      "loss_align": 0.11344295740127563,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12105493992567062,
      "neg_sim_mean": 0.7629090547561646,
      "pos_sim_mean": 0.8865570425987244,
      "step": 4287
    },
    {
      "epoch": 0.8576,
      "grad_norm": 1.21040678024292,
      "learning_rate": 6.067475035183861e-06,
      "loss": 0.2345,
      "step": 4288
    },
    {
      "epoch": 0.8576,
      "loss_align": 0.2575507164001465,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11749569326639175,
      "neg_sim_mean": 0.6319098472595215,
      "pos_sim_mean": 0.7424492835998535,
      "step": 4288
    },
    {
      "epoch": 0.8578,
      "grad_norm": 1.7480217218399048,
      "learning_rate": 6.0508190526795845e-06,
      "loss": 0.375,
      "step": 4289
    },
    {
      "epoch": 0.8578,
      "loss_align": 0.08756792545318604,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17664365470409393,
      "neg_sim_mean": 0.747085452079773,
      "pos_sim_mean": 0.912432074546814,
      "step": 4289
    },
    {
      "epoch": 0.858,
      "grad_norm": 1.5882046222686768,
      "learning_rate": 6.034184490472195e-06,
      "loss": 0.2642,
      "step": 4290
    },
    {
      "epoch": 0.858,
      "loss_align": 0.09341037273406982,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2213371992111206,
      "neg_sim_mean": 0.6884739995002747,
      "pos_sim_mean": 0.9065896272659302,
      "step": 4290
    },
    {
      "epoch": 0.8582,
      "grad_norm": 1.988671898841858,
      "learning_rate": 6.017571356669183e-06,
      "loss": 0.3147,
      "step": 4291
    },
    {
      "epoch": 0.8582,
      "loss_align": 0.09939843416213989,
      "loss_contrastive": 0.06428796052932739,
      "loss_gen": 0.19399455189704895,
      "neg_sim_mean": 0.8648895025253296,
      "pos_sim_mean": 0.9006015658378601,
      "step": 4291
    },
    {
      "epoch": 0.8584,
      "grad_norm": 1.6261351108551025,
      "learning_rate": 6.000979659367578e-06,
      "loss": 0.3011,
      "step": 4292
    },
    {
      "epoch": 0.8584,
      "loss_align": 0.087749183177948,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21674668788909912,
      "neg_sim_mean": 0.6472885608673096,
      "pos_sim_mean": 0.912250816822052,
      "step": 4292
    },
    {
      "epoch": 0.8586,
      "grad_norm": 1.5629920959472656,
      "learning_rate": 5.98440940665399e-06,
      "loss": 0.3045,
      "step": 4293
    },
    {
      "epoch": 0.8586,
      "loss_align": 0.09066665172576904,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0023927849251776934,
      "neg_sim_mean": 0.34557944536209106,
      "pos_sim_mean": 0.909333348274231,
      "step": 4293
    },
    {
      "epoch": 0.8588,
      "grad_norm": 0.38242006301879883,
      "learning_rate": 5.9678606066045525e-06,
      "loss": 0.0931,
      "step": 4294
    },
    {
      "epoch": 0.8588,
      "loss_align": 0.19353878498077393,
      "loss_contrastive": 0.10626912117004395,
      "loss_gen": 0.24142684042453766,
      "neg_sim_mean": 0.8127303123474121,
      "pos_sim_mean": 0.8064612150192261,
      "step": 4294
    },
    {
      "epoch": 0.859,
      "grad_norm": 2.0415215492248535,
      "learning_rate": 5.951333267284942e-06,
      "loss": 0.4477,
      "step": 4295
    },
    {
      "epoch": 0.859,
      "loss_align": 0.07389622926712036,
      "loss_contrastive": 0.029511570930480957,
      "loss_gen": 0.21838895976543427,
      "neg_sim_mean": 0.8556153178215027,
      "pos_sim_mean": 0.9261037707328796,
      "step": 4295
    },
    {
      "epoch": 0.8592,
      "grad_norm": 1.6481468677520752,
      "learning_rate": 5.934827396750392e-06,
      "loss": 0.2958,
      "step": 4296
    },
    {
      "epoch": 0.8592,
      "loss_align": 0.18341457843780518,
      "loss_contrastive": 0.0,
      "loss_gen": 0.30666568875312805,
      "neg_sim_mean": 0.7025852203369141,
      "pos_sim_mean": 0.8165854215621948,
      "step": 4296
    },
    {
      "epoch": 0.8594,
      "grad_norm": 2.438375949859619,
      "learning_rate": 5.918343003045657e-06,
      "loss": 0.4901,
      "step": 4297
    },
    {
      "epoch": 0.8594,
      "loss_align": 0.26951098442077637,
      "loss_contrastive": 0.15279698371887207,
      "loss_gen": 0.18800845742225647,
      "neg_sim_mean": 0.7832859754562378,
      "pos_sim_mean": 0.7304890155792236,
      "step": 4297
    },
    {
      "epoch": 0.8596,
      "grad_norm": 2.1539955139160156,
      "learning_rate": 5.901880094205036e-06,
      "loss": 0.4759,
      "step": 4298
    },
    {
      "epoch": 0.8596,
      "loss_align": 0.0670163631439209,
      "loss_contrastive": 0.0,
      "loss_gen": 0.022291364148259163,
      "neg_sim_mean": 0.38501647114753723,
      "pos_sim_mean": 0.9329836368560791,
      "step": 4298
    },
    {
      "epoch": 0.8598,
      "grad_norm": 1.0053800344467163,
      "learning_rate": 5.885438678252342e-06,
      "loss": 0.0893,
      "step": 4299
    },
    {
      "epoch": 0.8598,
      "loss_align": 0.0697629451751709,
      "loss_contrastive": 0.011726081371307373,
      "loss_gen": 0.18171121180057526,
      "neg_sim_mean": 0.8419631123542786,
      "pos_sim_mean": 0.9302370548248291,
      "step": 4299
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7436076402664185,
      "learning_rate": 5.8690187632009285e-06,
      "loss": 0.2529,
      "step": 4300
    },
    {
      "epoch": 0.86,
      "loss_align": 0.11003154516220093,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09401088207960129,
      "neg_sim_mean": 0.7746832370758057,
      "pos_sim_mean": 0.8899684548377991,
      "step": 4300
    },
    {
      "epoch": 0.8602,
      "grad_norm": 1.2138749361038208,
      "learning_rate": 5.852620357053651e-06,
      "loss": 0.204,
      "step": 4301
    },
    {
      "epoch": 0.8602,
      "loss_align": 0.08554834127426147,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015207859687507153,
      "neg_sim_mean": 0.39917123317718506,
      "pos_sim_mean": 0.9144516587257385,
      "step": 4301
    },
    {
      "epoch": 0.8604,
      "grad_norm": 0.3979667127132416,
      "learning_rate": 5.836243467802916e-06,
      "loss": 0.0871,
      "step": 4302
    },
    {
      "epoch": 0.8604,
      "loss_align": 0.11224806308746338,
      "loss_contrastive": 0.05083280801773071,
      "loss_gen": 0.21524101495742798,
      "neg_sim_mean": 0.8385847210884094,
      "pos_sim_mean": 0.8877519369125366,
      "step": 4302
    },
    {
      "epoch": 0.8606,
      "grad_norm": 1.6347708702087402,
      "learning_rate": 5.819888103430598e-06,
      "loss": 0.3336,
      "step": 4303
    },
    {
      "epoch": 0.8606,
      "loss_align": 0.0848846435546875,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1453808844089508,
      "neg_sim_mean": 0.7289007902145386,
      "pos_sim_mean": 0.9151153564453125,
      "step": 4303
    },
    {
      "epoch": 0.8608,
      "grad_norm": 1.3343209028244019,
      "learning_rate": 5.803554271908124e-06,
      "loss": 0.2303,
      "step": 4304
    },
    {
      "epoch": 0.8608,
      "loss_align": 0.13875746726989746,
      "loss_contrastive": 0.0,
      "loss_gen": 0.218353733420372,
      "neg_sim_mean": 0.6739197969436646,
      "pos_sim_mean": 0.8612425327301025,
      "step": 4304
    },
    {
      "epoch": 0.861,
      "grad_norm": 2.013474464416504,
      "learning_rate": 5.787241981196384e-06,
      "loss": 0.3571,
      "step": 4305
    },
    {
      "epoch": 0.861,
      "loss_align": 0.08112579584121704,
      "loss_contrastive": 0.0,
      "loss_gen": 0.34566375613212585,
      "neg_sim_mean": 0.6248827576637268,
      "pos_sim_mean": 0.918874204158783,
      "step": 4305
    },
    {
      "epoch": 0.8612,
      "grad_norm": 2.3592395782470703,
      "learning_rate": 5.770951239245803e-06,
      "loss": 0.4268,
      "step": 4306
    },
    {
      "epoch": 0.8612,
      "loss_align": 0.13355231285095215,
      "loss_contrastive": 0.0012699365615844727,
      "loss_gen": 0.2906586825847626,
      "neg_sim_mean": 0.7677175998687744,
      "pos_sim_mean": 0.8664476871490479,
      "step": 4306
    },
    {
      "epoch": 0.8614,
      "grad_norm": 2.229245901107788,
      "learning_rate": 5.754682053996291e-06,
      "loss": 0.4244,
      "step": 4307
    },
    {
      "epoch": 0.8614,
      "loss_align": 0.04930448532104492,
      "loss_contrastive": 0.0,
      "loss_gen": 0.31406861543655396,
      "neg_sim_mean": 0.6299437284469604,
      "pos_sim_mean": 0.9506955146789551,
      "step": 4307
    },
    {
      "epoch": 0.8616,
      "grad_norm": 2.5140159130096436,
      "learning_rate": 5.738434433377243e-06,
      "loss": 0.3634,
      "step": 4308
    },
    {
      "epoch": 0.8616,
      "loss_align": 0.10217863321304321,
      "loss_contrastive": 0.04515713453292847,
      "loss_gen": 0.05108832195401192,
      "neg_sim_mean": 0.8429784774780273,
      "pos_sim_mean": 0.8978213667869568,
      "step": 4308
    },
    {
      "epoch": 0.8618,
      "grad_norm": 0.8673205971717834,
      "learning_rate": 5.722208385307559e-06,
      "loss": 0.1587,
      "step": 4309
    },
    {
      "epoch": 0.8618,
      "loss_align": 0.08053088188171387,
      "loss_contrastive": 0.026018381118774414,
      "loss_gen": 0.22300252318382263,
      "neg_sim_mean": 0.8454874753952026,
      "pos_sim_mean": 0.9194691181182861,
      "step": 4309
    },
    {
      "epoch": 0.862,
      "grad_norm": 2.214452028274536,
      "learning_rate": 5.706003917695618e-06,
      "loss": 0.3067,
      "step": 4310
    },
    {
      "epoch": 0.862,
      "loss_align": 0.08205819129943848,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02446146309375763,
      "neg_sim_mean": 0.34544551372528076,
      "pos_sim_mean": 0.9179418087005615,
      "step": 4310
    },
    {
      "epoch": 0.8622,
      "grad_norm": 0.8897231817245483,
      "learning_rate": 5.689821038439263e-06,
      "loss": 0.1065,
      "step": 4311
    },
    {
      "epoch": 0.8622,
      "loss_align": 0.07571792602539062,
      "loss_contrastive": 0.06448984146118164,
      "loss_gen": 0.2156161665916443,
      "neg_sim_mean": 0.8887718915939331,
      "pos_sim_mean": 0.9242820739746094,
      "step": 4311
    },
    {
      "epoch": 0.8624,
      "grad_norm": 1.9049437046051025,
      "learning_rate": 5.673659755425859e-06,
      "loss": 0.2991,
      "step": 4312
    },
    {
      "epoch": 0.8624,
      "loss_align": 0.06799161434173584,
      "loss_contrastive": 0.0005896687507629395,
      "loss_gen": 0.23037882149219513,
      "neg_sim_mean": 0.8325980305671692,
      "pos_sim_mean": 0.9320083856582642,
      "step": 4312
    },
    {
      "epoch": 0.8626,
      "grad_norm": 1.7083433866500854,
      "learning_rate": 5.657520076532208e-06,
      "loss": 0.2984,
      "step": 4313
    },
    {
      "epoch": 0.8626,
      "loss_align": 0.12205874919891357,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18306100368499756,
      "neg_sim_mean": 0.6054599285125732,
      "pos_sim_mean": 0.8779412508010864,
      "step": 4313
    },
    {
      "epoch": 0.8628,
      "grad_norm": 2.0259363651275635,
      "learning_rate": 5.641402009624591e-06,
      "loss": 0.3051,
      "step": 4314
    },
    {
      "epoch": 0.8628,
      "loss_align": 0.1367243528366089,
      "loss_contrastive": 0.0853571891784668,
      "loss_gen": 0.17676572501659393,
      "neg_sim_mean": 0.8486328125,
      "pos_sim_mean": 0.8632756471633911,
      "step": 4314
    },
    {
      "epoch": 0.863,
      "grad_norm": 1.6668305397033691,
      "learning_rate": 5.625305562558764e-06,
      "loss": 0.3237,
      "step": 4315
    },
    {
      "epoch": 0.863,
      "loss_align": 0.14370125532150269,
      "loss_contrastive": 0.1093793511390686,
      "loss_gen": 0.18432502448558807,
      "neg_sim_mean": 0.865678071975708,
      "pos_sim_mean": 0.8562987446784973,
      "step": 4315
    },
    {
      "epoch": 0.8632,
      "grad_norm": 1.4608113765716553,
      "learning_rate": 5.6092307431799384e-06,
      "loss": 0.3412,
      "step": 4316
    },
    {
      "epoch": 0.8632,
      "loss_align": 0.09799718856811523,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24250957369804382,
      "neg_sim_mean": 0.7459198236465454,
      "pos_sim_mean": 0.9020028114318848,
      "step": 4316
    },
    {
      "epoch": 0.8634,
      "grad_norm": 1.5515156984329224,
      "learning_rate": 5.593177559322777e-06,
      "loss": 0.3405,
      "step": 4317
    },
    {
      "epoch": 0.8634,
      "loss_align": 0.11115056276321411,
      "loss_contrastive": 0.04572087526321411,
      "loss_gen": 0.2788323760032654,
      "neg_sim_mean": 0.8345702886581421,
      "pos_sim_mean": 0.8888494372367859,
      "step": 4317
    },
    {
      "epoch": 0.8636,
      "grad_norm": 2.096158504486084,
      "learning_rate": 5.577146018811419e-06,
      "loss": 0.3955,
      "step": 4318
    },
    {
      "epoch": 0.8636,
      "loss_align": 0.11131209135055542,
      "loss_contrastive": 0.13767075538635254,
      "loss_gen": 0.2158484011888504,
      "neg_sim_mean": 0.9263586401939392,
      "pos_sim_mean": 0.8886879086494446,
      "step": 4318
    },
    {
      "epoch": 0.8638,
      "grad_norm": 1.9862295389175415,
      "learning_rate": 5.5611361294594325e-06,
      "loss": 0.3437,
      "step": 4319
    },
    {
      "epoch": 0.8638,
      "loss_align": 0.14252150058746338,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0031961300410330296,
      "neg_sim_mean": 0.2933764159679413,
      "pos_sim_mean": 0.8574784994125366,
      "step": 4319
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.5231720209121704,
      "learning_rate": 5.545147899069836e-06,
      "loss": 0.1457,
      "step": 4320
    },
    {
      "epoch": 0.864,
      "loss_align": 0.22318267822265625,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1452483981847763,
      "neg_sim_mean": 0.5006409883499146,
      "pos_sim_mean": 0.7768173217773438,
      "step": 4320
    },
    {
      "epoch": 0.8642,
      "grad_norm": 1.5373567342758179,
      "learning_rate": 5.529181335435124e-06,
      "loss": 0.3684,
      "step": 4321
    },
    {
      "epoch": 0.8642,
      "loss_align": 0.07035869359970093,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13362494111061096,
      "neg_sim_mean": 0.7937973737716675,
      "pos_sim_mean": 0.9296413064002991,
      "step": 4321
    },
    {
      "epoch": 0.8644,
      "grad_norm": 1.3190006017684937,
      "learning_rate": 5.513236446337189e-06,
      "loss": 0.204,
      "step": 4322
    },
    {
      "epoch": 0.8644,
      "loss_align": 0.13593751192092896,
      "loss_contrastive": 0.0,
      "loss_gen": 0.39846932888031006,
      "neg_sim_mean": 0.6072659492492676,
      "pos_sim_mean": 0.864062488079071,
      "step": 4322
    },
    {
      "epoch": 0.8646,
      "grad_norm": 2.1574227809906006,
      "learning_rate": 5.497313239547375e-06,
      "loss": 0.5344,
      "step": 4323
    },
    {
      "epoch": 0.8646,
      "loss_align": 0.05596107244491577,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11237429082393646,
      "neg_sim_mean": 0.7803330421447754,
      "pos_sim_mean": 0.9440389275550842,
      "step": 4323
    },
    {
      "epoch": 0.8648,
      "grad_norm": 1.164365291595459,
      "learning_rate": 5.481411722826479e-06,
      "loss": 0.1683,
      "step": 4324
    },
    {
      "epoch": 0.8648,
      "loss_align": 0.0458756685256958,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0313769206404686,
      "neg_sim_mean": 0.29983192682266235,
      "pos_sim_mean": 0.9541243314743042,
      "step": 4324
    },
    {
      "epoch": 0.865,
      "grad_norm": 1.1501868963241577,
      "learning_rate": 5.46553190392467e-06,
      "loss": 0.0773,
      "step": 4325
    },
    {
      "epoch": 0.865,
      "loss_align": 0.07482093572616577,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007650475483387709,
      "neg_sim_mean": 0.24983423948287964,
      "pos_sim_mean": 0.9251790642738342,
      "step": 4325
    },
    {
      "epoch": 0.8652,
      "grad_norm": 0.6119229793548584,
      "learning_rate": 5.449673790581611e-06,
      "loss": 0.0825,
      "step": 4326
    },
    {
      "epoch": 0.8652,
      "loss_align": 0.08100438117980957,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001356937806122005,
      "neg_sim_mean": 0.2506915032863617,
      "pos_sim_mean": 0.9189956188201904,
      "step": 4326
    },
    {
      "epoch": 0.8654,
      "grad_norm": 0.3461291193962097,
      "learning_rate": 5.433837390526342e-06,
      "loss": 0.0824,
      "step": 4327
    },
    {
      "epoch": 0.8654,
      "loss_align": 0.06787627935409546,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3411531150341034,
      "neg_sim_mean": 0.8264170289039612,
      "pos_sim_mean": 0.9321237206459045,
      "step": 4327
    },
    {
      "epoch": 0.8656,
      "grad_norm": 2.237673759460449,
      "learning_rate": 5.418022711477333e-06,
      "loss": 0.409,
      "step": 4328
    },
    {
      "epoch": 0.8656,
      "loss_align": 0.1531912088394165,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1339206099510193,
      "neg_sim_mean": 0.6610912084579468,
      "pos_sim_mean": 0.8468087911605835,
      "step": 4328
    },
    {
      "epoch": 0.8658,
      "grad_norm": 1.4152823686599731,
      "learning_rate": 5.402229761142463e-06,
      "loss": 0.2871,
      "step": 4329
    },
    {
      "epoch": 0.8658,
      "loss_align": 0.074310302734375,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21000446379184723,
      "neg_sim_mean": 0.7918828725814819,
      "pos_sim_mean": 0.925689697265625,
      "step": 4329
    },
    {
      "epoch": 0.866,
      "grad_norm": 1.4639179706573486,
      "learning_rate": 5.3864585472190255e-06,
      "loss": 0.2843,
      "step": 4330
    },
    {
      "epoch": 0.866,
      "loss_align": 0.11299854516983032,
      "loss_contrastive": 0.06449019908905029,
      "loss_gen": 0.23686878383159637,
      "neg_sim_mean": 0.8514916300773621,
      "pos_sim_mean": 0.8870014548301697,
      "step": 4330
    },
    {
      "epoch": 0.8662,
      "grad_norm": 1.6396000385284424,
      "learning_rate": 5.370709077393721e-06,
      "loss": 0.3576,
      "step": 4331
    },
    {
      "epoch": 0.8662,
      "loss_align": 0.0968393087387085,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14898572862148285,
      "neg_sim_mean": 0.7505970597267151,
      "pos_sim_mean": 0.9031606912612915,
      "step": 4331
    },
    {
      "epoch": 0.8664,
      "grad_norm": 1.2030977010726929,
      "learning_rate": 5.354981359342659e-06,
      "loss": 0.2458,
      "step": 4332
    },
    {
      "epoch": 0.8664,
      "loss_align": 0.08651202917098999,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012223382480442524,
      "neg_sim_mean": 0.35214492678642273,
      "pos_sim_mean": 0.91348797082901,
      "step": 4332
    },
    {
      "epoch": 0.8666,
      "grad_norm": 0.7252390384674072,
      "learning_rate": 5.33927540073133e-06,
      "loss": 0.0987,
      "step": 4333
    },
    {
      "epoch": 0.8666,
      "loss_align": 0.1843419075012207,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15976203978061676,
      "neg_sim_mean": 0.7130373120307922,
      "pos_sim_mean": 0.8156580924987793,
      "step": 4333
    },
    {
      "epoch": 0.8668,
      "grad_norm": 1.5425512790679932,
      "learning_rate": 5.323591209214613e-06,
      "loss": 0.3441,
      "step": 4334
    },
    {
      "epoch": 0.8668,
      "loss_align": 0.13671714067459106,
      "loss_contrastive": 0.08251422643661499,
      "loss_gen": 0.3784676790237427,
      "neg_sim_mean": 0.845797061920166,
      "pos_sim_mean": 0.8632828593254089,
      "step": 4334
    },
    {
      "epoch": 0.867,
      "grad_norm": 2.536461591720581,
      "learning_rate": 5.307928792436811e-06,
      "loss": 0.5251,
      "step": 4335
    },
    {
      "epoch": 0.867,
      "loss_align": 0.12767308950424194,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4605104625225067,
      "neg_sim_mean": 0.6843960881233215,
      "pos_sim_mean": 0.8723269104957581,
      "step": 4335
    },
    {
      "epoch": 0.8672,
      "grad_norm": 2.8576724529266357,
      "learning_rate": 5.292288158031594e-06,
      "loss": 0.5882,
      "step": 4336
    },
    {
      "epoch": 0.8672,
      "loss_align": 0.10658204555511475,
      "loss_contrastive": 0.04950261116027832,
      "loss_gen": 0.2954008877277374,
      "neg_sim_mean": 0.8429205417633057,
      "pos_sim_mean": 0.8934179544448853,
      "step": 4336
    },
    {
      "epoch": 0.8674,
      "grad_norm": 1.8916016817092896,
      "learning_rate": 5.2766693136220125e-06,
      "loss": 0.4079,
      "step": 4337
    },
    {
      "epoch": 0.8674,
      "loss_align": 0.07806944847106934,
      "loss_contrastive": 0.051084816455841064,
      "loss_gen": 0.25286924839019775,
      "neg_sim_mean": 0.8730153441429138,
      "pos_sim_mean": 0.9219305515289307,
      "step": 4337
    },
    {
      "epoch": 0.8676,
      "grad_norm": 1.739734411239624,
      "learning_rate": 5.261072266820499e-06,
      "loss": 0.3371,
      "step": 4338
    },
    {
      "epoch": 0.8676,
      "loss_align": 0.04977315664291382,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20933715999126434,
      "neg_sim_mean": 0.8357183337211609,
      "pos_sim_mean": 0.9502268433570862,
      "step": 4338
    },
    {
      "epoch": 0.8678,
      "grad_norm": 1.5806856155395508,
      "learning_rate": 5.245497025228874e-06,
      "loss": 0.2591,
      "step": 4339
    },
    {
      "epoch": 0.8678,
      "loss_align": 0.1653062105178833,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13551557064056396,
      "neg_sim_mean": 0.33777233958244324,
      "pos_sim_mean": 0.8346937894821167,
      "step": 4339
    },
    {
      "epoch": 0.868,
      "grad_norm": 1.299930453300476,
      "learning_rate": 5.229943596438297e-06,
      "loss": 0.3008,
      "step": 4340
    },
    {
      "epoch": 0.868,
      "loss_align": 0.04197216033935547,
      "loss_contrastive": 0.021476387977600098,
      "loss_gen": 0.15318167209625244,
      "neg_sim_mean": 0.8795042037963867,
      "pos_sim_mean": 0.9580278396606445,
      "step": 4340
    },
    {
      "epoch": 0.8682,
      "grad_norm": 1.29178786277771,
      "learning_rate": 5.214411988029355e-06,
      "loss": 0.1977,
      "step": 4341
    },
    {
      "epoch": 0.8682,
      "loss_align": 0.0541454553604126,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013536844402551651,
      "neg_sim_mean": 0.4823378622531891,
      "pos_sim_mean": 0.9458545446395874,
      "step": 4341
    },
    {
      "epoch": 0.8684,
      "grad_norm": 0.6895930171012878,
      "learning_rate": 5.198902207571954e-06,
      "loss": 0.0677,
      "step": 4342
    },
    {
      "epoch": 0.8684,
      "loss_align": 0.07385587692260742,
      "loss_contrastive": 0.03815305233001709,
      "loss_gen": 0.12451212853193283,
      "neg_sim_mean": 0.8642971515655518,
      "pos_sim_mean": 0.9261441230773926,
      "step": 4342
    },
    {
      "epoch": 0.8686,
      "grad_norm": 1.4319103956222534,
      "learning_rate": 5.183414262625363e-06,
      "loss": 0.2029,
      "step": 4343
    },
    {
      "epoch": 0.8686,
      "loss_align": 0.16050374507904053,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29300224781036377,
      "neg_sim_mean": 0.6143698692321777,
      "pos_sim_mean": 0.8394962549209595,
      "step": 4343
    },
    {
      "epoch": 0.8688,
      "grad_norm": 1.9094833135604858,
      "learning_rate": 5.1679481607382065e-06,
      "loss": 0.4535,
      "step": 4344
    },
    {
      "epoch": 0.8688,
      "loss_align": 0.06936466693878174,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011933818459510803,
      "neg_sim_mean": 0.30139803886413574,
      "pos_sim_mean": 0.9306353330612183,
      "step": 4344
    },
    {
      "epoch": 0.869,
      "grad_norm": 0.2974351942539215,
      "learning_rate": 5.152503909448503e-06,
      "loss": 0.0706,
      "step": 4345
    },
    {
      "epoch": 0.869,
      "loss_align": 0.14040303230285645,
      "loss_contrastive": 0.0004928112030029297,
      "loss_gen": 0.18048423528671265,
      "neg_sim_mean": 0.7600897550582886,
      "pos_sim_mean": 0.8595969676971436,
      "step": 4345
    },
    {
      "epoch": 0.8692,
      "grad_norm": 1.7995797395706177,
      "learning_rate": 5.137081516283581e-06,
      "loss": 0.3209,
      "step": 4346
    },
    {
      "epoch": 0.8692,
      "loss_align": 0.20945191383361816,
      "loss_contrastive": 0.09244680404663086,
      "loss_gen": 0.471161425113678,
      "neg_sim_mean": 0.7829948663711548,
      "pos_sim_mean": 0.7905480861663818,
      "step": 4346
    },
    {
      "epoch": 0.8694,
      "grad_norm": 2.511803150177002,
      "learning_rate": 5.1216809887601245e-06,
      "loss": 0.6917,
      "step": 4347
    },
    {
      "epoch": 0.8694,
      "loss_align": 0.06627130508422852,
      "loss_contrastive": 0.0,
      "loss_gen": 0.30309557914733887,
      "neg_sim_mean": 0.8287932276725769,
      "pos_sim_mean": 0.9337286949157715,
      "step": 4347
    },
    {
      "epoch": 0.8696,
      "grad_norm": 2.029371976852417,
      "learning_rate": 5.106302334384172e-06,
      "loss": 0.3694,
      "step": 4348
    },
    {
      "epoch": 0.8696,
      "loss_align": 0.08885574340820312,
      "loss_contrastive": 0.09436941146850586,
      "loss_gen": 0.2509065270423889,
      "neg_sim_mean": 0.9055136442184448,
      "pos_sim_mean": 0.9111442565917969,
      "step": 4348
    },
    {
      "epoch": 0.8698,
      "grad_norm": 2.104679822921753,
      "learning_rate": 5.090945560651073e-06,
      "loss": 0.3511,
      "step": 4349
    },
    {
      "epoch": 0.8698,
      "loss_align": 0.13367891311645508,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1026897132396698,
      "neg_sim_mean": 0.5265883803367615,
      "pos_sim_mean": 0.8663210868835449,
      "step": 4349
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.3615872859954834,
      "learning_rate": 5.075610675045567e-06,
      "loss": 0.2364,
      "step": 4350
    },
    {
      "epoch": 0.87,
      "loss_align": 0.08770966529846191,
      "loss_contrastive": 0.0,
      "loss_gen": 0.492288202047348,
      "neg_sim_mean": 0.7382473945617676,
      "pos_sim_mean": 0.9122903347015381,
      "step": 4350
    },
    {
      "epoch": 0.8702,
      "grad_norm": 2.444180965423584,
      "learning_rate": 5.060297685041659e-06,
      "loss": 0.58,
      "step": 4351
    },
    {
      "epoch": 0.8702,
      "loss_align": 0.10639643669128418,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009008248336613178,
      "neg_sim_mean": 0.3026157021522522,
      "pos_sim_mean": 0.8936035633087158,
      "step": 4351
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.7253499031066895,
      "learning_rate": 5.045006598102725e-06,
      "loss": 0.1154,
      "step": 4352
    },
    {
      "epoch": 0.8704,
      "loss_align": 0.06704545021057129,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11422190070152283,
      "neg_sim_mean": 0.7316628694534302,
      "pos_sim_mean": 0.9329545497894287,
      "step": 4352
    },
    {
      "epoch": 0.8706,
      "grad_norm": 1.3571839332580566,
      "learning_rate": 5.029737421681446e-06,
      "loss": 0.1813,
      "step": 4353
    },
    {
      "epoch": 0.8706,
      "loss_align": 0.19903242588043213,
      "loss_contrastive": 0.07841551303863525,
      "loss_gen": 0.08426649123430252,
      "neg_sim_mean": 0.7793830633163452,
      "pos_sim_mean": 0.8009675741195679,
      "step": 4353
    },
    {
      "epoch": 0.8708,
      "grad_norm": 1.2560368776321411,
      "learning_rate": 5.014490163219853e-06,
      "loss": 0.2927,
      "step": 4354
    },
    {
      "epoch": 0.8708,
      "loss_align": 0.08573085069656372,
      "loss_contrastive": 0.005160033702850342,
      "loss_gen": 0.263640820980072,
      "neg_sim_mean": 0.8194291591644287,
      "pos_sim_mean": 0.9142691493034363,
      "step": 4354
    },
    {
      "epoch": 0.871,
      "grad_norm": 1.7837896347045898,
      "learning_rate": 4.9992648301492695e-06,
      "loss": 0.35,
      "step": 4355
    },
    {
      "epoch": 0.871,
      "loss_align": 0.08216989040374756,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1998719424009323,
      "neg_sim_mean": 0.7930819988250732,
      "pos_sim_mean": 0.9178301095962524,
      "step": 4355
    },
    {
      "epoch": 0.8712,
      "grad_norm": 1.6323916912078857,
      "learning_rate": 4.984061429890324e-06,
      "loss": 0.282,
      "step": 4356
    },
    {
      "epoch": 0.8712,
      "loss_align": 0.08434385061264038,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0013371760724112391,
      "neg_sim_mean": 0.43179404735565186,
      "pos_sim_mean": 0.9156561493873596,
      "step": 4356
    },
    {
      "epoch": 0.8714,
      "grad_norm": 0.4424164593219757,
      "learning_rate": 4.9688799698529855e-06,
      "loss": 0.0857,
      "step": 4357
    },
    {
      "epoch": 0.8714,
      "loss_align": 0.09712982177734375,
      "loss_contrastive": 0.0,
      "loss_gen": 0.31396135687828064,
      "neg_sim_mean": 0.6502366065979004,
      "pos_sim_mean": 0.9028701782226562,
      "step": 4357
    },
    {
      "epoch": 0.8716,
      "grad_norm": 2.1701064109802246,
      "learning_rate": 4.9537204574365e-06,
      "loss": 0.4111,
      "step": 4358
    },
    {
      "epoch": 0.8716,
      "loss_align": 0.10142922401428223,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09143944829702377,
      "neg_sim_mean": 0.7047620415687561,
      "pos_sim_mean": 0.8985707759857178,
      "step": 4358
    },
    {
      "epoch": 0.8718,
      "grad_norm": 1.5575369596481323,
      "learning_rate": 4.938582900029437e-06,
      "loss": 0.1929,
      "step": 4359
    },
    {
      "epoch": 0.8718,
      "loss_align": 0.1756352186203003,
      "loss_contrastive": 0.11048030853271484,
      "loss_gen": 0.5134177207946777,
      "neg_sim_mean": 0.8348450660705566,
      "pos_sim_mean": 0.8243647813796997,
      "step": 4359
    },
    {
      "epoch": 0.872,
      "grad_norm": 2.5518999099731445,
      "learning_rate": 4.92346730500966e-06,
      "loss": 0.7023,
      "step": 4360
    },
    {
      "epoch": 0.872,
      "loss_align": 0.06767559051513672,
      "loss_contrastive": 0.009600996971130371,
      "loss_gen": 0.22120428085327148,
      "neg_sim_mean": 0.8419253826141357,
      "pos_sim_mean": 0.9323244094848633,
      "step": 4360
    },
    {
      "epoch": 0.8722,
      "grad_norm": 1.802000880241394,
      "learning_rate": 4.908373679744316e-06,
      "loss": 0.29,
      "step": 4361
    },
    {
      "epoch": 0.8722,
      "loss_align": 0.059961915016174316,
      "loss_contrastive": 0.07404512166976929,
      "loss_gen": 0.10967127233743668,
      "neg_sim_mean": 0.9140831828117371,
      "pos_sim_mean": 0.9400380849838257,
      "step": 4361
    },
    {
      "epoch": 0.8724,
      "grad_norm": 1.145763635635376,
      "learning_rate": 4.893302031589864e-06,
      "loss": 0.1785,
      "step": 4362
    },
    {
      "epoch": 0.8724,
      "loss_align": 0.07095324993133545,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14969483017921448,
      "neg_sim_mean": 0.7853134870529175,
      "pos_sim_mean": 0.9290467500686646,
      "step": 4362
    },
    {
      "epoch": 0.8726,
      "grad_norm": 1.418824315071106,
      "learning_rate": 4.878252367892033e-06,
      "loss": 0.2206,
      "step": 4363
    },
    {
      "epoch": 0.8726,
      "loss_align": 0.0588221549987793,
      "loss_contrastive": 0.027333378791809082,
      "loss_gen": 0.2647409737110138,
      "neg_sim_mean": 0.8685111999511719,
      "pos_sim_mean": 0.9411778450012207,
      "step": 4363
    },
    {
      "epoch": 0.8728,
      "grad_norm": 1.931024193763733,
      "learning_rate": 4.863224695985857e-06,
      "loss": 0.3268,
      "step": 4364
    },
    {
      "epoch": 0.8728,
      "loss_align": 0.09383261203765869,
      "loss_contrastive": 0.09848928451538086,
      "loss_gen": 0.20410796999931335,
      "neg_sim_mean": 0.9046566486358643,
      "pos_sim_mean": 0.9061673879623413,
      "step": 4364
    },
    {
      "epoch": 0.873,
      "grad_norm": 1.6993939876556396,
      "learning_rate": 4.848219023195644e-06,
      "loss": 0.3098,
      "step": 4365
    },
    {
      "epoch": 0.873,
      "loss_align": 0.1143389344215393,
      "loss_contrastive": 0.0,
      "loss_gen": 0.28875732421875,
      "neg_sim_mean": 0.7464065551757812,
      "pos_sim_mean": 0.8856610655784607,
      "step": 4365
    },
    {
      "epoch": 0.8732,
      "grad_norm": 1.9507368803024292,
      "learning_rate": 4.833235356834959e-06,
      "loss": 0.4031,
      "step": 4366
    },
    {
      "epoch": 0.8732,
      "loss_align": 0.16759097576141357,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15068325400352478,
      "neg_sim_mean": 0.48172760009765625,
      "pos_sim_mean": 0.8324090242385864,
      "step": 4366
    },
    {
      "epoch": 0.8734,
      "grad_norm": 1.5479280948638916,
      "learning_rate": 4.818273704206677e-06,
      "loss": 0.3183,
      "step": 4367
    },
    {
      "epoch": 0.8734,
      "loss_align": 0.09633314609527588,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0018060019938275218,
      "neg_sim_mean": 0.372902512550354,
      "pos_sim_mean": 0.9036668539047241,
      "step": 4367
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.48645636439323425,
      "learning_rate": 4.8033340726029165e-06,
      "loss": 0.0981,
      "step": 4368
    },
    {
      "epoch": 0.8736,
      "loss_align": 0.17630153894424438,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2806035280227661,
      "neg_sim_mean": 0.5440201163291931,
      "pos_sim_mean": 0.8236984610557556,
      "step": 4368
    },
    {
      "epoch": 0.8738,
      "grad_norm": 2.0946667194366455,
      "learning_rate": 4.788416469305068e-06,
      "loss": 0.4569,
      "step": 4369
    },
    {
      "epoch": 0.8738,
      "loss_align": 0.05628645420074463,
      "loss_contrastive": 0.05687373876571655,
      "loss_gen": 0.5313292145729065,
      "neg_sim_mean": 0.900587260723114,
      "pos_sim_mean": 0.9437135457992554,
      "step": 4369
    },
    {
      "epoch": 0.874,
      "grad_norm": 2.3242006301879883,
      "learning_rate": 4.7735209015838e-06,
      "loss": 0.5944,
      "step": 4370
    },
    {
      "epoch": 0.874,
      "loss_align": 0.13786178827285767,
      "loss_contrastive": 0.043748676776885986,
      "loss_gen": 0.2522573471069336,
      "neg_sim_mean": 0.8058868646621704,
      "pos_sim_mean": 0.8621382117271423,
      "step": 4370
    },
    {
      "epoch": 0.8742,
      "grad_norm": 1.6334909200668335,
      "learning_rate": 4.758647376699032e-06,
      "loss": 0.3954,
      "step": 4371
    },
    {
      "epoch": 0.8742,
      "loss_align": 0.1471981406211853,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17197012901306152,
      "neg_sim_mean": 0.7121539115905762,
      "pos_sim_mean": 0.8528018593788147,
      "step": 4371
    },
    {
      "epoch": 0.8744,
      "grad_norm": 1.6940228939056396,
      "learning_rate": 4.743795901899928e-06,
      "loss": 0.3192,
      "step": 4372
    },
    {
      "epoch": 0.8744,
      "loss_align": 0.0976184606552124,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2834177613258362,
      "neg_sim_mean": 0.7309873104095459,
      "pos_sim_mean": 0.9023815393447876,
      "step": 4372
    },
    {
      "epoch": 0.8746,
      "grad_norm": 1.8689320087432861,
      "learning_rate": 4.728966484424912e-06,
      "loss": 0.381,
      "step": 4373
    },
    {
      "epoch": 0.8746,
      "loss_align": 0.06574362516403198,
      "loss_contrastive": 0.07748514413833618,
      "loss_gen": 0.2704789936542511,
      "neg_sim_mean": 0.9117414951324463,
      "pos_sim_mean": 0.934256374835968,
      "step": 4373
    },
    {
      "epoch": 0.8748,
      "grad_norm": 1.962459683418274,
      "learning_rate": 4.714159131501689e-06,
      "loss": 0.3455,
      "step": 4374
    },
    {
      "epoch": 0.8748,
      "loss_align": 0.08675968647003174,
      "loss_contrastive": 0.0,
      "loss_gen": 0.350124716758728,
      "neg_sim_mean": 0.8039854764938354,
      "pos_sim_mean": 0.9132403135299683,
      "step": 4374
    },
    {
      "epoch": 0.875,
      "grad_norm": 2.078461170196533,
      "learning_rate": 4.699373850347161e-06,
      "loss": 0.4369,
      "step": 4375
    },
    {
      "epoch": 0.875,
      "loss_align": 0.06490635871887207,
      "loss_contrastive": 0.0,
      "loss_gen": 0.08880606293678284,
      "neg_sim_mean": 0.7949931621551514,
      "pos_sim_mean": 0.9350936412811279,
      "step": 4375
    },
    {
      "epoch": 0.8752,
      "grad_norm": 1.261290431022644,
      "learning_rate": 4.684610648167503e-06,
      "loss": 0.1537,
      "step": 4376
    },
    {
      "epoch": 0.8752,
      "loss_align": 0.09251987934112549,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13672752678394318,
      "neg_sim_mean": 0.49408313632011414,
      "pos_sim_mean": 0.9074801206588745,
      "step": 4376
    },
    {
      "epoch": 0.8754,
      "grad_norm": 1.355333685874939,
      "learning_rate": 4.6698695321581165e-06,
      "loss": 0.2292,
      "step": 4377
    },
    {
      "epoch": 0.8754,
      "loss_align": 0.06452786922454834,
      "loss_contrastive": 0.025105774402618408,
      "loss_gen": 0.3038351833820343,
      "neg_sim_mean": 0.8605778813362122,
      "pos_sim_mean": 0.9354721307754517,
      "step": 4377
    },
    {
      "epoch": 0.8756,
      "grad_norm": 1.9572889804840088,
      "learning_rate": 4.655150509503642e-06,
      "loss": 0.3714,
      "step": 4378
    },
    {
      "epoch": 0.8756,
      "loss_align": 0.11486709117889404,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1698390543460846,
      "neg_sim_mean": 0.704414963722229,
      "pos_sim_mean": 0.885132908821106,
      "step": 4378
    },
    {
      "epoch": 0.8758,
      "grad_norm": 1.751051902770996,
      "learning_rate": 4.640453587377957e-06,
      "loss": 0.2847,
      "step": 4379
    },
    {
      "epoch": 0.8758,
      "loss_align": 0.10280376672744751,
      "loss_contrastive": 0.0,
      "loss_gen": 0.018037449568510056,
      "neg_sim_mean": 0.35339850187301636,
      "pos_sim_mean": 0.8971962332725525,
      "step": 4379
    },
    {
      "epoch": 0.876,
      "grad_norm": 0.9089983105659485,
      "learning_rate": 4.625778772944156e-06,
      "loss": 0.1208,
      "step": 4380
    },
    {
      "epoch": 0.876,
      "loss_align": 0.15951108932495117,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2204579859972,
      "neg_sim_mean": 0.5328608155250549,
      "pos_sim_mean": 0.8404889106750488,
      "step": 4380
    },
    {
      "epoch": 0.8762,
      "grad_norm": 1.7847223281860352,
      "learning_rate": 4.611126073354571e-06,
      "loss": 0.38,
      "step": 4381
    },
    {
      "epoch": 0.8762,
      "loss_align": 0.11658310890197754,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0308074913918972,
      "neg_sim_mean": 0.40508151054382324,
      "pos_sim_mean": 0.8834168910980225,
      "step": 4381
    },
    {
      "epoch": 0.8764,
      "grad_norm": 1.2180978059768677,
      "learning_rate": 4.596495495750741e-06,
      "loss": 0.1474,
      "step": 4382
    },
    {
      "epoch": 0.8764,
      "loss_align": 0.11213433742523193,
      "loss_contrastive": 0.039240241050720215,
      "loss_gen": 0.25632137060165405,
      "neg_sim_mean": 0.8271058797836304,
      "pos_sim_mean": 0.8878656625747681,
      "step": 4382
    },
    {
      "epoch": 0.8766,
      "grad_norm": 1.813786268234253,
      "learning_rate": 4.5818870472634444e-06,
      "loss": 0.3732,
      "step": 4383
    },
    {
      "epoch": 0.8766,
      "loss_align": 0.07211184501647949,
      "loss_contrastive": 0.03821265697479248,
      "loss_gen": 0.13152353465557098,
      "neg_sim_mean": 0.8661007881164551,
      "pos_sim_mean": 0.9278881549835205,
      "step": 4383
    },
    {
      "epoch": 0.8768,
      "grad_norm": 1.5320948362350464,
      "learning_rate": 4.567300735012653e-06,
      "loss": 0.2082,
      "step": 4384
    },
    {
      "epoch": 0.8768,
      "loss_align": 0.16298913955688477,
      "loss_contrastive": 0.020889997482299805,
      "loss_gen": 0.3398286700248718,
      "neg_sim_mean": 0.7579008340835571,
      "pos_sim_mean": 0.8370108604431152,
      "step": 4384
    },
    {
      "epoch": 0.877,
      "grad_norm": 2.0759198665618896,
      "learning_rate": 4.5527365661075625e-06,
      "loss": 0.5053,
      "step": 4385
    },
    {
      "epoch": 0.877,
      "loss_align": 0.10164022445678711,
      "loss_contrastive": 0.0388033390045166,
      "loss_gen": 0.1770685315132141,
      "neg_sim_mean": 0.8371630907058716,
      "pos_sim_mean": 0.8983597755432129,
      "step": 4385
    },
    {
      "epoch": 0.8772,
      "grad_norm": 1.3798267841339111,
      "learning_rate": 4.538194547646574e-06,
      "loss": 0.2834,
      "step": 4386
    },
    {
      "epoch": 0.8772,
      "loss_align": 0.058226704597473145,
      "loss_contrastive": 0.08411228656768799,
      "loss_gen": 0.20381249487400055,
      "neg_sim_mean": 0.9258855581283569,
      "pos_sim_mean": 0.9417732954025269,
      "step": 4386
    },
    {
      "epoch": 0.8774,
      "grad_norm": 1.9163353443145752,
      "learning_rate": 4.523674686717283e-06,
      "loss": 0.2721,
      "step": 4387
    },
    {
      "epoch": 0.8774,
      "loss_align": 0.13792532682418823,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1923350840806961,
      "neg_sim_mean": 0.7439114451408386,
      "pos_sim_mean": 0.8620746731758118,
      "step": 4387
    },
    {
      "epoch": 0.8776,
      "grad_norm": 1.8625844717025757,
      "learning_rate": 4.509176990396497e-06,
      "loss": 0.3303,
      "step": 4388
    },
    {
      "epoch": 0.8776,
      "loss_align": 0.09166944026947021,
      "loss_contrastive": 0.028063178062438965,
      "loss_gen": 0.46187159419059753,
      "neg_sim_mean": 0.8363937139511108,
      "pos_sim_mean": 0.9083305597305298,
      "step": 4388
    },
    {
      "epoch": 0.8778,
      "grad_norm": 2.1687393188476562,
      "learning_rate": 4.494701465750217e-06,
      "loss": 0.5569,
      "step": 4389
    },
    {
      "epoch": 0.8778,
      "loss_align": 0.10497987270355225,
      "loss_contrastive": 0.0508725643157959,
      "loss_gen": 0.16582943499088287,
      "neg_sim_mean": 0.8458926677703857,
      "pos_sim_mean": 0.8950201272964478,
      "step": 4389
    },
    {
      "epoch": 0.878,
      "grad_norm": 1.5550271272659302,
      "learning_rate": 4.480248119833641e-06,
      "loss": 0.2769,
      "step": 4390
    },
    {
      "epoch": 0.878,
      "loss_align": 0.10510754585266113,
      "loss_contrastive": 0.0,
      "loss_gen": 0.38030412793159485,
      "neg_sim_mean": 0.6694542169570923,
      "pos_sim_mean": 0.8948924541473389,
      "step": 4390
    },
    {
      "epoch": 0.8782,
      "grad_norm": 2.0368998050689697,
      "learning_rate": 4.465816959691149e-06,
      "loss": 0.4854,
      "step": 4391
    },
    {
      "epoch": 0.8782,
      "loss_align": 0.19443893432617188,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2942100167274475,
      "neg_sim_mean": 0.6248445510864258,
      "pos_sim_mean": 0.8055610656738281,
      "step": 4391
    },
    {
      "epoch": 0.8784,
      "grad_norm": 2.0386574268341064,
      "learning_rate": 4.45140799235631e-06,
      "loss": 0.4886,
      "step": 4392
    },
    {
      "epoch": 0.8784,
      "loss_align": 0.06174808740615845,
      "loss_contrastive": 0.06253141164779663,
      "loss_gen": 0.19412270188331604,
      "neg_sim_mean": 0.9007833003997803,
      "pos_sim_mean": 0.9382519125938416,
      "step": 4392
    },
    {
      "epoch": 0.8786,
      "grad_norm": 1.5514272451400757,
      "learning_rate": 4.43702122485189e-06,
      "loss": 0.2634,
      "step": 4393
    },
    {
      "epoch": 0.8786,
      "loss_align": 0.06524580717086792,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2606043815612793,
      "neg_sim_mean": 0.820085883140564,
      "pos_sim_mean": 0.9347541928291321,
      "step": 4393
    },
    {
      "epoch": 0.8788,
      "grad_norm": 1.8662890195846558,
      "learning_rate": 4.422656664189817e-06,
      "loss": 0.3259,
      "step": 4394
    },
    {
      "epoch": 0.8788,
      "loss_align": 0.14520913362503052,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11277806013822556,
      "neg_sim_mean": 0.6122508645057678,
      "pos_sim_mean": 0.8547908663749695,
      "step": 4394
    },
    {
      "epoch": 0.879,
      "grad_norm": 1.7269643545150757,
      "learning_rate": 4.408314317371221e-06,
      "loss": 0.258,
      "step": 4395
    },
    {
      "epoch": 0.879,
      "loss_align": 0.1251683235168457,
      "loss_contrastive": 0.07636874914169312,
      "loss_gen": 0.2581214904785156,
      "neg_sim_mean": 0.8512004017829895,
      "pos_sim_mean": 0.8748316764831543,
      "step": 4395
    },
    {
      "epoch": 0.8792,
      "grad_norm": 1.8377540111541748,
      "learning_rate": 4.3939941913863525e-06,
      "loss": 0.3925,
      "step": 4396
    },
    {
      "epoch": 0.8792,
      "loss_align": 0.25124120712280273,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12302281707525253,
      "neg_sim_mean": 0.4672239124774933,
      "pos_sim_mean": 0.7487587928771973,
      "step": 4396
    },
    {
      "epoch": 0.8794,
      "grad_norm": 1.7317445278167725,
      "learning_rate": 4.379696293214696e-06,
      "loss": 0.3743,
      "step": 4397
    },
    {
      "epoch": 0.8794,
      "loss_align": 0.12329745292663574,
      "loss_contrastive": 0.023378968238830566,
      "loss_gen": 0.2748594284057617,
      "neg_sim_mean": 0.8000814914703369,
      "pos_sim_mean": 0.8767025470733643,
      "step": 4397
    },
    {
      "epoch": 0.8796,
      "grad_norm": 1.7653334140777588,
      "learning_rate": 4.365420629824863e-06,
      "loss": 0.401,
      "step": 4398
    },
    {
      "epoch": 0.8796,
      "loss_align": 0.05543529987335205,
      "loss_contrastive": 0.05461174249649048,
      "loss_gen": 0.27771416306495667,
      "neg_sim_mean": 0.8991764187812805,
      "pos_sim_mean": 0.944564700126648,
      "step": 4398
    },
    {
      "epoch": 0.8798,
      "grad_norm": 1.784814476966858,
      "learning_rate": 4.351167208174639e-06,
      "loss": 0.3397,
      "step": 4399
    },
    {
      "epoch": 0.8798,
      "loss_align": 0.0789991021156311,
      "loss_contrastive": 0.0452040433883667,
      "loss_gen": 0.15901407599449158,
      "neg_sim_mean": 0.8662049174308777,
      "pos_sim_mean": 0.9210008978843689,
      "step": 4399
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.3671022653579712,
      "learning_rate": 4.33693603521097e-06,
      "loss": 0.2434,
      "step": 4400
    },
    {
      "epoch": 0.88,
      "loss_align": 0.05790126323699951,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23276036977767944,
      "neg_sim_mean": 0.7654645442962646,
      "pos_sim_mean": 0.9420987367630005,
      "step": 4400
    },
    {
      "epoch": 0.8802,
      "grad_norm": 2.014188766479492,
      "learning_rate": 4.322727117869951e-06,
      "loss": 0.2907,
      "step": 4401
    },
    {
      "epoch": 0.8802,
      "loss_align": 0.08629065752029419,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22513532638549805,
      "neg_sim_mean": 0.7811252474784851,
      "pos_sim_mean": 0.9137093424797058,
      "step": 4401
    },
    {
      "epoch": 0.8804,
      "grad_norm": 1.8389508724212646,
      "learning_rate": 4.308540463076849e-06,
      "loss": 0.3114,
      "step": 4402
    },
    {
      "epoch": 0.8804,
      "loss_align": 0.11788254976272583,
      "loss_contrastive": 0.05047339200973511,
      "loss_gen": 0.3816159665584564,
      "neg_sim_mean": 0.8325908184051514,
      "pos_sim_mean": 0.8821174502372742,
      "step": 4402
    },
    {
      "epoch": 0.8806,
      "grad_norm": 2.35546875,
      "learning_rate": 4.294376077746059e-06,
      "loss": 0.5056,
      "step": 4403
    },
    {
      "epoch": 0.8806,
      "loss_align": 0.13842123746871948,
      "loss_contrastive": 0.05156618356704712,
      "loss_gen": 0.6710172891616821,
      "neg_sim_mean": 0.8131449222564697,
      "pos_sim_mean": 0.8615787625312805,
      "step": 4403
    },
    {
      "epoch": 0.8808,
      "grad_norm": 3.436518669128418,
      "learning_rate": 4.2802339687811385e-06,
      "loss": 0.8156,
      "step": 4404
    },
    {
      "epoch": 0.8808,
      "loss_align": 0.07405710220336914,
      "loss_contrastive": 0.06669449806213379,
      "loss_gen": 0.22761604189872742,
      "neg_sim_mean": 0.8926373720169067,
      "pos_sim_mean": 0.9259428977966309,
      "step": 4404
    },
    {
      "epoch": 0.881,
      "grad_norm": 1.5113950967788696,
      "learning_rate": 4.2661141430747505e-06,
      "loss": 0.3097,
      "step": 4405
    },
    {
      "epoch": 0.881,
      "loss_align": 0.0776146650314331,
      "loss_contrastive": 0.07502758502960205,
      "loss_gen": 0.21323063969612122,
      "neg_sim_mean": 0.897412896156311,
      "pos_sim_mean": 0.9223853349685669,
      "step": 4405
    },
    {
      "epoch": 0.8812,
      "grad_norm": 1.8226516246795654,
      "learning_rate": 4.2520166075087635e-06,
      "loss": 0.2998,
      "step": 4406
    },
    {
      "epoch": 0.8812,
      "loss_align": 0.09675407409667969,
      "loss_contrastive": 0.0,
      "loss_gen": 0.32107946276664734,
      "neg_sim_mean": 0.656279981136322,
      "pos_sim_mean": 0.9032459259033203,
      "step": 4406
    },
    {
      "epoch": 0.8814,
      "grad_norm": 2.057076930999756,
      "learning_rate": 4.237941368954124e-06,
      "loss": 0.4178,
      "step": 4407
    },
    {
      "epoch": 0.8814,
      "loss_align": 0.07239854335784912,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17252640426158905,
      "neg_sim_mean": 0.7116745114326477,
      "pos_sim_mean": 0.9276014566421509,
      "step": 4407
    },
    {
      "epoch": 0.8816,
      "grad_norm": 1.7668451070785522,
      "learning_rate": 4.223888434270939e-06,
      "loss": 0.2449,
      "step": 4408
    },
    {
      "epoch": 0.8816,
      "loss_align": 0.10148084163665771,
      "loss_contrastive": 0.09174072742462158,
      "loss_gen": 0.38524550199508667,
      "neg_sim_mean": 0.890259861946106,
      "pos_sim_mean": 0.8985191583633423,
      "step": 4408
    },
    {
      "epoch": 0.8818,
      "grad_norm": 1.9626133441925049,
      "learning_rate": 4.209857810308437e-06,
      "loss": 0.4977,
      "step": 4409
    },
    {
      "epoch": 0.8818,
      "loss_align": 0.07410889863967896,
      "loss_contrastive": 0.0,
      "loss_gen": 0.382414311170578,
      "neg_sim_mean": 0.7292137145996094,
      "pos_sim_mean": 0.925891101360321,
      "step": 4409
    },
    {
      "epoch": 0.882,
      "grad_norm": 1.8774532079696655,
      "learning_rate": 4.195849503904975e-06,
      "loss": 0.4565,
      "step": 4410
    },
    {
      "epoch": 0.882,
      "loss_align": 0.09947609901428223,
      "loss_contrastive": 0.0063245296478271484,
      "loss_gen": 0.19799144566059113,
      "neg_sim_mean": 0.806848406791687,
      "pos_sim_mean": 0.9005239009857178,
      "step": 4410
    },
    {
      "epoch": 0.8822,
      "grad_norm": 1.631811261177063,
      "learning_rate": 4.181863521888019e-06,
      "loss": 0.2982,
      "step": 4411
    },
    {
      "epoch": 0.8822,
      "loss_align": 0.08041554689407349,
      "loss_contrastive": 0.03399139642715454,
      "loss_gen": 0.27435487508773804,
      "neg_sim_mean": 0.8535758256912231,
      "pos_sim_mean": 0.9195844531059265,
      "step": 4411
    },
    {
      "epoch": 0.8824,
      "grad_norm": 1.451574683189392,
      "learning_rate": 4.167899871074194e-06,
      "loss": 0.3588,
      "step": 4412
    },
    {
      "epoch": 0.8824,
      "loss_align": 0.09649050235748291,
      "loss_contrastive": 0.0,
      "loss_gen": 0.06031284108757973,
      "neg_sim_mean": 0.7860340476036072,
      "pos_sim_mean": 0.9035094976425171,
      "step": 4412
    },
    {
      "epoch": 0.8826,
      "grad_norm": 1.1170912981033325,
      "learning_rate": 4.153958558269189e-06,
      "loss": 0.1568,
      "step": 4413
    },
    {
      "epoch": 0.8826,
      "loss_align": 0.08119726181030273,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10451744496822357,
      "neg_sim_mean": 0.6536039710044861,
      "pos_sim_mean": 0.9188027381896973,
      "step": 4413
    },
    {
      "epoch": 0.8828,
      "grad_norm": 1.1455634832382202,
      "learning_rate": 4.140039590267836e-06,
      "loss": 0.1857,
      "step": 4414
    },
    {
      "epoch": 0.8828,
      "loss_align": 0.10182374715805054,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15215659141540527,
      "neg_sim_mean": 0.664687991142273,
      "pos_sim_mean": 0.8981762528419495,
      "step": 4414
    },
    {
      "epoch": 0.883,
      "grad_norm": 1.4154465198516846,
      "learning_rate": 4.126142973854069e-06,
      "loss": 0.254,
      "step": 4415
    },
    {
      "epoch": 0.883,
      "loss_align": 0.08391839265823364,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17202533781528473,
      "neg_sim_mean": 0.7795212268829346,
      "pos_sim_mean": 0.9160816073417664,
      "step": 4415
    },
    {
      "epoch": 0.8832,
      "grad_norm": 1.5833343267440796,
      "learning_rate": 4.112268715800943e-06,
      "loss": 0.2559,
      "step": 4416
    },
    {
      "epoch": 0.8832,
      "loss_align": 0.05586087703704834,
      "loss_contrastive": 0.05202603340148926,
      "loss_gen": 0.09325321763753891,
      "neg_sim_mean": 0.896165132522583,
      "pos_sim_mean": 0.9441391229629517,
      "step": 4416
    },
    {
      "epoch": 0.8834,
      "grad_norm": 1.1083983182907104,
      "learning_rate": 4.098416822870593e-06,
      "loss": 0.1554,
      "step": 4417
    },
    {
      "epoch": 0.8834,
      "loss_align": 0.05933678150177002,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19500713050365448,
      "neg_sim_mean": 0.557432234287262,
      "pos_sim_mean": 0.94066321849823,
      "step": 4417
    },
    {
      "epoch": 0.8836,
      "grad_norm": 1.192264437675476,
      "learning_rate": 4.084587301814269e-06,
      "loss": 0.2543,
      "step": 4418
    },
    {
      "epoch": 0.8836,
      "loss_align": 0.07436716556549072,
      "loss_contrastive": 0.10201072692871094,
      "loss_gen": 0.5856239795684814,
      "neg_sim_mean": 0.9276435375213623,
      "pos_sim_mean": 0.9256328344345093,
      "step": 4418
    },
    {
      "epoch": 0.8838,
      "grad_norm": 2.528045654296875,
      "learning_rate": 4.0707801593723e-06,
      "loss": 0.6722,
      "step": 4419
    },
    {
      "epoch": 0.8838,
      "loss_align": 0.13627266883850098,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3120323717594147,
      "neg_sim_mean": 0.5311498641967773,
      "pos_sim_mean": 0.863727331161499,
      "step": 4419
    },
    {
      "epoch": 0.884,
      "grad_norm": 2.401634454727173,
      "learning_rate": 4.056995402274122e-06,
      "loss": 0.4483,
      "step": 4420
    },
    {
      "epoch": 0.884,
      "loss_align": 0.10706144571304321,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12218696624040604,
      "neg_sim_mean": 0.7642634510993958,
      "pos_sim_mean": 0.8929385542869568,
      "step": 4420
    },
    {
      "epoch": 0.8842,
      "grad_norm": 1.1722594499588013,
      "learning_rate": 4.043233037238281e-06,
      "loss": 0.2292,
      "step": 4421
    },
    {
      "epoch": 0.8842,
      "loss_align": 0.07221585512161255,
      "loss_contrastive": 0.065060555934906,
      "loss_gen": 0.17619861662387848,
      "neg_sim_mean": 0.8928446769714355,
      "pos_sim_mean": 0.9277841448783875,
      "step": 4421
    },
    {
      "epoch": 0.8844,
      "grad_norm": 1.4859693050384521,
      "learning_rate": 4.029493070972362e-06,
      "loss": 0.2562,
      "step": 4422
    },
    {
      "epoch": 0.8844,
      "loss_align": 0.07007741928100586,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12976954877376556,
      "neg_sim_mean": 0.35346758365631104,
      "pos_sim_mean": 0.9299225807189941,
      "step": 4422
    },
    {
      "epoch": 0.8846,
      "grad_norm": 1.7696834802627563,
      "learning_rate": 4.015775510173064e-06,
      "loss": 0.1998,
      "step": 4423
    },
    {
      "epoch": 0.8846,
      "loss_align": 0.15108942985534668,
      "loss_contrastive": 0.013415098190307617,
      "loss_gen": 0.3351760506629944,
      "neg_sim_mean": 0.762325644493103,
      "pos_sim_mean": 0.8489105701446533,
      "step": 4423
    },
    {
      "epoch": 0.8848,
      "grad_norm": 2.1814606189727783,
      "learning_rate": 4.002080361526156e-06,
      "loss": 0.4879,
      "step": 4424
    },
    {
      "epoch": 0.8848,
      "loss_align": 0.08558011054992676,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2195431888103485,
      "neg_sim_mean": 0.6143567562103271,
      "pos_sim_mean": 0.9144198894500732,
      "step": 4424
    },
    {
      "epoch": 0.885,
      "grad_norm": 1.5365849733352661,
      "learning_rate": 3.9884076317064814e-06,
      "loss": 0.3051,
      "step": 4425
    },
    {
      "epoch": 0.885,
      "loss_align": 0.08440136909484863,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0201172586530447,
      "neg_sim_mean": 0.3388056755065918,
      "pos_sim_mean": 0.9155986309051514,
      "step": 4425
    },
    {
      "epoch": 0.8852,
      "grad_norm": 0.9047520756721497,
      "learning_rate": 3.974757327377981e-06,
      "loss": 0.1045,
      "step": 4426
    },
    {
      "epoch": 0.8852,
      "loss_align": 0.1578836441040039,
      "loss_contrastive": 0.05463993549346924,
      "loss_gen": 0.23839743435382843,
      "neg_sim_mean": 0.7967562675476074,
      "pos_sim_mean": 0.8421163558959961,
      "step": 4426
    },
    {
      "epoch": 0.8854,
      "grad_norm": 1.9157496690750122,
      "learning_rate": 3.961129455193641e-06,
      "loss": 0.4028,
      "step": 4427
    },
    {
      "epoch": 0.8854,
      "loss_align": 0.11135852336883545,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21566203236579895,
      "neg_sim_mean": 0.7136531472206116,
      "pos_sim_mean": 0.8886414766311646,
      "step": 4427
    },
    {
      "epoch": 0.8856,
      "grad_norm": 1.6944767236709595,
      "learning_rate": 3.947524021795518e-06,
      "loss": 0.327,
      "step": 4428
    },
    {
      "epoch": 0.8856,
      "loss_align": 0.062309980392456055,
      "loss_contrastive": 0.05489093065261841,
      "loss_gen": 0.1807691603899002,
      "neg_sim_mean": 0.8925809264183044,
      "pos_sim_mean": 0.937690019607544,
      "step": 4428
    },
    {
      "epoch": 0.8858,
      "grad_norm": 1.6283326148986816,
      "learning_rate": 3.933941033814736e-06,
      "loss": 0.2497,
      "step": 4429
    },
    {
      "epoch": 0.8858,
      "loss_align": 0.14591842889785767,
      "loss_contrastive": 0.047632455825805664,
      "loss_gen": 0.9314051270484924,
      "neg_sim_mean": 0.8017140030860901,
      "pos_sim_mean": 0.8540815711021423,
      "step": 4429
    },
    {
      "epoch": 0.886,
      "grad_norm": 3.237333059310913,
      "learning_rate": 3.920380497871473e-06,
      "loss": 1.083,
      "step": 4430
    },
    {
      "epoch": 0.886,
      "loss_align": 0.10732114315032959,
      "loss_contrastive": 0.0,
      "loss_gen": 0.022432250902056694,
      "neg_sim_mean": 0.32740092277526855,
      "pos_sim_mean": 0.8926788568496704,
      "step": 4430
    },
    {
      "epoch": 0.8862,
      "grad_norm": 0.9894196391105652,
      "learning_rate": 3.90684242057498e-06,
      "loss": 0.1298,
      "step": 4431
    },
    {
      "epoch": 0.8862,
      "loss_align": 0.15630745887756348,
      "loss_contrastive": 0.07533526420593262,
      "loss_gen": 0.38073867559432983,
      "neg_sim_mean": 0.8190277814865112,
      "pos_sim_mean": 0.8436925411224365,
      "step": 4431
    },
    {
      "epoch": 0.8864,
      "grad_norm": 2.294543743133545,
      "learning_rate": 3.8933268085235395e-06,
      "loss": 0.5461,
      "step": 4432
    },
    {
      "epoch": 0.8864,
      "loss_align": 0.087638258934021,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14397713541984558,
      "neg_sim_mean": 0.79697585105896,
      "pos_sim_mean": 0.912361741065979,
      "step": 4432
    },
    {
      "epoch": 0.8866,
      "grad_norm": 1.5030103921890259,
      "learning_rate": 3.879833668304505e-06,
      "loss": 0.2316,
      "step": 4433
    },
    {
      "epoch": 0.8866,
      "loss_align": 0.11070764064788818,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19573020935058594,
      "neg_sim_mean": 0.7338976263999939,
      "pos_sim_mean": 0.8892923593521118,
      "step": 4433
    },
    {
      "epoch": 0.8868,
      "grad_norm": 1.5132453441619873,
      "learning_rate": 3.866363006494255e-06,
      "loss": 0.3064,
      "step": 4434
    },
    {
      "epoch": 0.8868,
      "loss_align": 0.06971263885498047,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24151870608329773,
      "neg_sim_mean": 0.8120753765106201,
      "pos_sim_mean": 0.9302873611450195,
      "step": 4434
    },
    {
      "epoch": 0.887,
      "grad_norm": 1.7702749967575073,
      "learning_rate": 3.85291482965825e-06,
      "loss": 0.3112,
      "step": 4435
    },
    {
      "epoch": 0.887,
      "loss_align": 0.1210627555847168,
      "loss_contrastive": 0.08216369152069092,
      "loss_gen": 0.5157897472381592,
      "neg_sim_mean": 0.8611009120941162,
      "pos_sim_mean": 0.8789372444152832,
      "step": 4435
    },
    {
      "epoch": 0.8872,
      "grad_norm": 2.671109437942505,
      "learning_rate": 3.839489144350955e-06,
      "loss": 0.6467,
      "step": 4436
    },
    {
      "epoch": 0.8872,
      "loss_align": 0.11831682920455933,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0017439507646486163,
      "neg_sim_mean": 0.48540765047073364,
      "pos_sim_mean": 0.8816831707954407,
      "step": 4436
    },
    {
      "epoch": 0.8874,
      "grad_norm": 0.49656200408935547,
      "learning_rate": 3.826085957115888e-06,
      "loss": 0.1201,
      "step": 4437
    },
    {
      "epoch": 0.8874,
      "loss_align": 0.12580031156539917,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22713470458984375,
      "neg_sim_mean": 0.7391842603683472,
      "pos_sim_mean": 0.8741996884346008,
      "step": 4437
    },
    {
      "epoch": 0.8876,
      "grad_norm": 1.91112220287323,
      "learning_rate": 3.8127052744855953e-06,
      "loss": 0.3529,
      "step": 4438
    },
    {
      "epoch": 0.8876,
      "loss_align": 0.055290937423706055,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13255168497562408,
      "neg_sim_mean": 0.782850980758667,
      "pos_sim_mean": 0.944709062576294,
      "step": 4438
    },
    {
      "epoch": 0.8878,
      "grad_norm": 1.3070870637893677,
      "learning_rate": 3.799347102981665e-06,
      "loss": 0.1878,
      "step": 4439
    },
    {
      "epoch": 0.8878,
      "loss_align": 0.10362333059310913,
      "loss_contrastive": 0.0,
      "loss_gen": 0.014368250034749508,
      "neg_sim_mean": 0.2833442687988281,
      "pos_sim_mean": 0.8963766694068909,
      "step": 4439
    },
    {
      "epoch": 0.888,
      "grad_norm": 0.7729837894439697,
      "learning_rate": 3.7860114491147016e-06,
      "loss": 0.118,
      "step": 4440
    },
    {
      "epoch": 0.888,
      "loss_align": 0.13789713382720947,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003093292471021414,
      "neg_sim_mean": 0.49932894110679626,
      "pos_sim_mean": 0.8621028661727905,
      "step": 4440
    },
    {
      "epoch": 0.8882,
      "grad_norm": 0.5621035695075989,
      "learning_rate": 3.772698319384349e-06,
      "loss": 0.141,
      "step": 4441
    },
    {
      "epoch": 0.8882,
      "loss_align": 0.05042219161987305,
      "loss_contrastive": 0.0,
      "loss_gen": 0.030414938926696777,
      "neg_sim_mean": 0.29234862327575684,
      "pos_sim_mean": 0.949577808380127,
      "step": 4441
    },
    {
      "epoch": 0.8884,
      "grad_norm": 1.141162633895874,
      "learning_rate": 3.7594077202792566e-06,
      "loss": 0.0808,
      "step": 4442
    },
    {
      "epoch": 0.8884,
      "loss_align": 0.22150647640228271,
      "loss_contrastive": 0.08161133527755737,
      "loss_gen": 0.16535955667495728,
      "neg_sim_mean": 0.7601048350334167,
      "pos_sim_mean": 0.7784935235977173,
      "step": 4442
    },
    {
      "epoch": 0.8886,
      "grad_norm": 1.8271863460540771,
      "learning_rate": 3.7461396582771035e-06,
      "loss": 0.3967,
      "step": 4443
    },
    {
      "epoch": 0.8886,
      "loss_align": 0.0685957670211792,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18898804485797882,
      "neg_sim_mean": 0.7990878820419312,
      "pos_sim_mean": 0.9314042329788208,
      "step": 4443
    },
    {
      "epoch": 0.8888,
      "grad_norm": 1.5318745374679565,
      "learning_rate": 3.732894139844578e-06,
      "loss": 0.2576,
      "step": 4444
    },
    {
      "epoch": 0.8888,
      "loss_align": 0.15725815296173096,
      "loss_contrastive": 0.03746843338012695,
      "loss_gen": 0.27914372086524963,
      "neg_sim_mean": 0.7802102565765381,
      "pos_sim_mean": 0.842741847038269,
      "step": 4444
    },
    {
      "epoch": 0.889,
      "grad_norm": 2.2079179286956787,
      "learning_rate": 3.7196711714373945e-06,
      "loss": 0.4409,
      "step": 4445
    },
    {
      "epoch": 0.889,
      "loss_align": 0.06848084926605225,
      "loss_contrastive": 0.06391918659210205,
      "loss_gen": 0.2986138164997101,
      "neg_sim_mean": 0.8954383134841919,
      "pos_sim_mean": 0.9315191507339478,
      "step": 4445
    },
    {
      "epoch": 0.8892,
      "grad_norm": 1.9547114372253418,
      "learning_rate": 3.7064707595002635e-06,
      "loss": 0.3748,
      "step": 4446
    },
    {
      "epoch": 0.8892,
      "loss_align": 0.08234262466430664,
      "loss_contrastive": 0.015141129493713379,
      "loss_gen": 0.2141040563583374,
      "neg_sim_mean": 0.8327984809875488,
      "pos_sim_mean": 0.9176573753356934,
      "step": 4446
    },
    {
      "epoch": 0.8894,
      "grad_norm": 1.3332613706588745,
      "learning_rate": 3.6932929104669055e-06,
      "loss": 0.2983,
      "step": 4447
    },
    {
      "epoch": 0.8894,
      "loss_align": 0.08920681476593018,
      "loss_contrastive": 0.0,
      "loss_gen": 0.32416781783103943,
      "neg_sim_mean": 0.8023853302001953,
      "pos_sim_mean": 0.9107931852340698,
      "step": 4447
    },
    {
      "epoch": 0.8896,
      "grad_norm": 1.995553731918335,
      "learning_rate": 3.680137630760039e-06,
      "loss": 0.4134,
      "step": 4448
    },
    {
      "epoch": 0.8896,
      "loss_align": 0.09720224142074585,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005843708757311106,
      "neg_sim_mean": 0.24599823355674744,
      "pos_sim_mean": 0.9027977585792542,
      "step": 4448
    },
    {
      "epoch": 0.8898,
      "grad_norm": 0.6361315846443176,
      "learning_rate": 3.6670049267913952e-06,
      "loss": 0.103,
      "step": 4449
    },
    {
      "epoch": 0.8898,
      "loss_align": 0.12035942077636719,
      "loss_contrastive": 0.0,
      "loss_gen": 0.37188979983329773,
      "neg_sim_mean": 0.7134018540382385,
      "pos_sim_mean": 0.8796405792236328,
      "step": 4449
    },
    {
      "epoch": 0.89,
      "grad_norm": 3.4024336338043213,
      "learning_rate": 3.6538948049616882e-06,
      "loss": 0.4922,
      "step": 4450
    },
    {
      "epoch": 0.89,
      "loss_align": 0.05141925811767578,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2370343804359436,
      "neg_sim_mean": 0.8179652690887451,
      "pos_sim_mean": 0.9485807418823242,
      "step": 4450
    },
    {
      "epoch": 0.8902,
      "grad_norm": 1.79256272315979,
      "learning_rate": 3.6408072716606346e-06,
      "loss": 0.2885,
      "step": 4451
    },
    {
      "epoch": 0.8902,
      "loss_align": 0.10717082023620605,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0067542376928031445,
      "neg_sim_mean": 0.5285384654998779,
      "pos_sim_mean": 0.892829179763794,
      "step": 4451
    },
    {
      "epoch": 0.8904,
      "grad_norm": 0.6452047824859619,
      "learning_rate": 3.627742333266937e-06,
      "loss": 0.1139,
      "step": 4452
    },
    {
      "epoch": 0.8904,
      "loss_align": 0.19809967279434204,
      "loss_contrastive": 0.0,
      "loss_gen": 0.020050734281539917,
      "neg_sim_mean": 0.4254743456840515,
      "pos_sim_mean": 0.801900327205658,
      "step": 4452
    },
    {
      "epoch": 0.8906,
      "grad_norm": 1.0898096561431885,
      "learning_rate": 3.6146999961482853e-06,
      "loss": 0.2182,
      "step": 4453
    },
    {
      "epoch": 0.8906,
      "loss_align": 0.14956414699554443,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0020764057990163565,
      "neg_sim_mean": 0.415005624294281,
      "pos_sim_mean": 0.8504358530044556,
      "step": 4453
    },
    {
      "epoch": 0.8908,
      "grad_norm": 0.5962371826171875,
      "learning_rate": 3.601680266661367e-06,
      "loss": 0.1516,
      "step": 4454
    },
    {
      "epoch": 0.8908,
      "loss_align": 0.09880876541137695,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18280841410160065,
      "neg_sim_mean": 0.8009874820709229,
      "pos_sim_mean": 0.901191234588623,
      "step": 4454
    },
    {
      "epoch": 0.891,
      "grad_norm": 1.6715131998062134,
      "learning_rate": 3.5886831511518336e-06,
      "loss": 0.2816,
      "step": 4455
    },
    {
      "epoch": 0.891,
      "loss_align": 0.0679396390914917,
      "loss_contrastive": 0.01955866813659668,
      "loss_gen": 0.14460255205631256,
      "neg_sim_mean": 0.8516190052032471,
      "pos_sim_mean": 0.9320603609085083,
      "step": 4455
    },
    {
      "epoch": 0.8912,
      "grad_norm": 1.7278449535369873,
      "learning_rate": 3.575708655954324e-06,
      "loss": 0.2149,
      "step": 4456
    },
    {
      "epoch": 0.8912,
      "loss_align": 0.1121903657913208,
      "loss_contrastive": 0.009623825550079346,
      "loss_gen": 0.30267444252967834,
      "neg_sim_mean": 0.7974334359169006,
      "pos_sim_mean": 0.8878096342086792,
      "step": 4456
    },
    {
      "epoch": 0.8914,
      "grad_norm": 1.9033117294311523,
      "learning_rate": 3.5627567873924516e-06,
      "loss": 0.416,
      "step": 4457
    },
    {
      "epoch": 0.8914,
      "loss_align": 0.09098589420318604,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19254787266254425,
      "neg_sim_mean": 0.3473922610282898,
      "pos_sim_mean": 0.909014105796814,
      "step": 4457
    },
    {
      "epoch": 0.8916,
      "grad_norm": 1.9179612398147583,
      "learning_rate": 3.549827551778778e-06,
      "loss": 0.2835,
      "step": 4458
    },
    {
      "epoch": 0.8916,
      "loss_align": 0.07684135437011719,
      "loss_contrastive": 0.0,
      "loss_gen": 0.141330748796463,
      "neg_sim_mean": 0.3152596950531006,
      "pos_sim_mean": 0.9231586456298828,
      "step": 4458
    },
    {
      "epoch": 0.8918,
      "grad_norm": 1.2984371185302734,
      "learning_rate": 3.536920955414885e-06,
      "loss": 0.2182,
      "step": 4459
    },
    {
      "epoch": 0.8918,
      "loss_align": 0.11438620090484619,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21725277602672577,
      "neg_sim_mean": 0.6038063764572144,
      "pos_sim_mean": 0.8856137990951538,
      "step": 4459
    },
    {
      "epoch": 0.892,
      "grad_norm": 1.7428585290908813,
      "learning_rate": 3.524037004591274e-06,
      "loss": 0.3316,
      "step": 4460
    },
    {
      "epoch": 0.892,
      "loss_align": 0.10587143898010254,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16937488317489624,
      "neg_sim_mean": 0.6028141975402832,
      "pos_sim_mean": 0.8941285610198975,
      "step": 4460
    },
    {
      "epoch": 0.8922,
      "grad_norm": 1.779439091682434,
      "learning_rate": 3.511175705587433e-06,
      "loss": 0.2752,
      "step": 4461
    },
    {
      "epoch": 0.8922,
      "loss_align": 0.11581110954284668,
      "loss_contrastive": 0.011733412742614746,
      "loss_gen": 0.2478797733783722,
      "neg_sim_mean": 0.7959222793579102,
      "pos_sim_mean": 0.8841888904571533,
      "step": 4461
    },
    {
      "epoch": 0.8924,
      "grad_norm": 1.582787275314331,
      "learning_rate": 3.498337064671803e-06,
      "loss": 0.3651,
      "step": 4462
    },
    {
      "epoch": 0.8924,
      "loss_align": 0.11451631784439087,
      "loss_contrastive": 0.021906793117523193,
      "loss_gen": 0.24572636187076569,
      "neg_sim_mean": 0.8073904514312744,
      "pos_sim_mean": 0.8854836821556091,
      "step": 4462
    },
    {
      "epoch": 0.8926,
      "grad_norm": 1.9001715183258057,
      "learning_rate": 3.4855210881017674e-06,
      "loss": 0.3629,
      "step": 4463
    },
    {
      "epoch": 0.8926,
      "loss_align": 0.09939181804656982,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009180665016174316,
      "neg_sim_mean": 0.2839260995388031,
      "pos_sim_mean": 0.9006081819534302,
      "step": 4463
    },
    {
      "epoch": 0.8928,
      "grad_norm": 0.6594322919845581,
      "learning_rate": 3.472727782123697e-06,
      "loss": 0.1086,
      "step": 4464
    },
    {
      "epoch": 0.8928,
      "loss_align": 0.1226511001586914,
      "loss_contrastive": 0.07126927375793457,
      "loss_gen": 0.15062153339385986,
      "neg_sim_mean": 0.8486181497573853,
      "pos_sim_mean": 0.8773488998413086,
      "step": 4464
    },
    {
      "epoch": 0.893,
      "grad_norm": 1.4609754085540771,
      "learning_rate": 3.459957152972887e-06,
      "loss": 0.2818,
      "step": 4465
    },
    {
      "epoch": 0.893,
      "loss_align": 0.14390206336975098,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0666852742433548,
      "neg_sim_mean": 0.35736289620399475,
      "pos_sim_mean": 0.856097936630249,
      "step": 4465
    },
    {
      "epoch": 0.8932,
      "grad_norm": 1.0269556045532227,
      "learning_rate": 3.4472092068735916e-06,
      "loss": 0.2106,
      "step": 4466
    },
    {
      "epoch": 0.8932,
      "loss_align": 0.11512815952301025,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0167789738625288,
      "neg_sim_mean": 0.3334037661552429,
      "pos_sim_mean": 0.8848718404769897,
      "step": 4466
    },
    {
      "epoch": 0.8934,
      "grad_norm": 0.8789572715759277,
      "learning_rate": 3.434483950038986e-06,
      "loss": 0.1319,
      "step": 4467
    },
    {
      "epoch": 0.8934,
      "loss_align": 0.09878242015838623,
      "loss_contrastive": 0.025573372840881348,
      "loss_gen": 0.25162455439567566,
      "neg_sim_mean": 0.8267909288406372,
      "pos_sim_mean": 0.9012175798416138,
      "step": 4467
    },
    {
      "epoch": 0.8936,
      "grad_norm": 1.798046588897705,
      "learning_rate": 3.4217813886712247e-06,
      "loss": 0.3535,
      "step": 4468
    },
    {
      "epoch": 0.8936,
      "loss_align": 0.15130245685577393,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1733977049589157,
      "neg_sim_mean": 0.7149182558059692,
      "pos_sim_mean": 0.8486975431442261,
      "step": 4468
    },
    {
      "epoch": 0.8938,
      "grad_norm": 1.7606199979782104,
      "learning_rate": 3.4091015289613783e-06,
      "loss": 0.3247,
      "step": 4469
    },
    {
      "epoch": 0.8938,
      "loss_align": 0.056169331073760986,
      "loss_contrastive": 0.06071704626083374,
      "loss_gen": 0.15515093505382538,
      "neg_sim_mean": 0.9045476913452148,
      "pos_sim_mean": 0.943830668926239,
      "step": 4469
    },
    {
      "epoch": 0.894,
      "grad_norm": 1.491104006767273,
      "learning_rate": 3.3964443770894528e-06,
      "loss": 0.2186,
      "step": 4470
    },
    {
      "epoch": 0.894,
      "loss_align": 0.0658944845199585,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20237010717391968,
      "neg_sim_mean": 0.8248746395111084,
      "pos_sim_mean": 0.9341055154800415,
      "step": 4470
    },
    {
      "epoch": 0.8942,
      "grad_norm": 1.74616539478302,
      "learning_rate": 3.3838099392243916e-06,
      "loss": 0.2683,
      "step": 4471
    },
    {
      "epoch": 0.8942,
      "loss_align": 0.10575079917907715,
      "loss_contrastive": 0.06791460514068604,
      "loss_gen": 0.19057977199554443,
      "neg_sim_mean": 0.862163782119751,
      "pos_sim_mean": 0.8942492008209229,
      "step": 4471
    },
    {
      "epoch": 0.8944,
      "grad_norm": 1.6286996603012085,
      "learning_rate": 3.371198221524069e-06,
      "loss": 0.3045,
      "step": 4472
    },
    {
      "epoch": 0.8944,
      "loss_align": 0.15774530172348022,
      "loss_contrastive": 0.17535513639450073,
      "loss_gen": 0.23497705161571503,
      "neg_sim_mean": 0.9176098108291626,
      "pos_sim_mean": 0.8422546982765198,
      "step": 4472
    },
    {
      "epoch": 0.8946,
      "grad_norm": 1.851880431175232,
      "learning_rate": 3.358609230135268e-06,
      "loss": 0.4138,
      "step": 4473
    },
    {
      "epoch": 0.8946,
      "loss_align": 0.11673927307128906,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0037191780284047127,
      "neg_sim_mean": 0.2570275068283081,
      "pos_sim_mean": 0.8832607269287109,
      "step": 4473
    },
    {
      "epoch": 0.8948,
      "grad_norm": 0.5843889713287354,
      "learning_rate": 3.3460429711937414e-06,
      "loss": 0.1205,
      "step": 4474
    },
    {
      "epoch": 0.8948,
      "loss_align": 0.09807413816452026,
      "loss_contrastive": 0.10368388891220093,
      "loss_gen": 0.2772959768772125,
      "neg_sim_mean": 0.9056097269058228,
      "pos_sim_mean": 0.9019258618354797,
      "step": 4474
    },
    {
      "epoch": 0.895,
      "grad_norm": 1.9961071014404297,
      "learning_rate": 3.3334994508241015e-06,
      "loss": 0.3878,
      "step": 4475
    },
    {
      "epoch": 0.895,
      "loss_align": 0.07110309600830078,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21464121341705322,
      "neg_sim_mean": 0.795527994632721,
      "pos_sim_mean": 0.9288969039916992,
      "step": 4475
    },
    {
      "epoch": 0.8952,
      "grad_norm": 1.840489149093628,
      "learning_rate": 3.3209786751399187e-06,
      "loss": 0.2857,
      "step": 4476
    },
    {
      "epoch": 0.8952,
      "loss_align": 0.16961216926574707,
      "loss_contrastive": 0.10015392303466797,
      "loss_gen": 0.1834603101015091,
      "neg_sim_mean": 0.830541729927063,
      "pos_sim_mean": 0.8303878307342529,
      "step": 4476
    },
    {
      "epoch": 0.8954,
      "grad_norm": 1.8006199598312378,
      "learning_rate": 3.308480650243662e-06,
      "loss": 0.3651,
      "step": 4477
    },
    {
      "epoch": 0.8954,
      "loss_align": 0.11042451858520508,
      "loss_contrastive": 0.0014138221740722656,
      "loss_gen": 0.16481326520442963,
      "neg_sim_mean": 0.7909892797470093,
      "pos_sim_mean": 0.8895754814147949,
      "step": 4477
    },
    {
      "epoch": 0.8956,
      "grad_norm": 1.5653010606765747,
      "learning_rate": 3.2960053822267246e-06,
      "loss": 0.2754,
      "step": 4478
    },
    {
      "epoch": 0.8956,
      "loss_align": 0.08694636821746826,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25229257345199585,
      "neg_sim_mean": 0.8036528825759888,
      "pos_sim_mean": 0.9130536317825317,
      "step": 4478
    },
    {
      "epoch": 0.8958,
      "grad_norm": 1.6763474941253662,
      "learning_rate": 3.2835528771693992e-06,
      "loss": 0.3392,
      "step": 4479
    },
    {
      "epoch": 0.8958,
      "loss_align": 0.08055758476257324,
      "loss_contrastive": 0.013528823852539062,
      "loss_gen": 0.2474992424249649,
      "neg_sim_mean": 0.8329712152481079,
      "pos_sim_mean": 0.9194424152374268,
      "step": 4479
    },
    {
      "epoch": 0.896,
      "grad_norm": 1.8245973587036133,
      "learning_rate": 3.271123141140886e-06,
      "loss": 0.3297,
      "step": 4480
    },
    {
      "epoch": 0.896,
      "loss_align": 0.1497199535369873,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007117625791579485,
      "neg_sim_mean": 0.371659517288208,
      "pos_sim_mean": 0.8502800464630127,
      "step": 4480
    },
    {
      "epoch": 0.8962,
      "grad_norm": 0.6872684359550476,
      "learning_rate": 3.258716180199278e-06,
      "loss": 0.1568,
      "step": 4481
    },
    {
      "epoch": 0.8962,
      "loss_align": 0.09856438636779785,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013828017748892307,
      "neg_sim_mean": 0.3219057321548462,
      "pos_sim_mean": 0.9014356136322021,
      "step": 4481
    },
    {
      "epoch": 0.8964,
      "grad_norm": 0.8018675446510315,
      "learning_rate": 3.246332000391583e-06,
      "loss": 0.1124,
      "step": 4482
    },
    {
      "epoch": 0.8964,
      "loss_align": 0.09055554866790771,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23194435238838196,
      "neg_sim_mean": 0.7800674438476562,
      "pos_sim_mean": 0.9094444513320923,
      "step": 4482
    },
    {
      "epoch": 0.8966,
      "grad_norm": 2.0878171920776367,
      "learning_rate": 3.2339706077537167e-06,
      "loss": 0.3225,
      "step": 4483
    },
    {
      "epoch": 0.8966,
      "loss_align": 0.13443952798843384,
      "loss_contrastive": 0.0,
      "loss_gen": 0.023159364238381386,
      "neg_sim_mean": 0.32913607358932495,
      "pos_sim_mean": 0.8655604720115662,
      "step": 4483
    },
    {
      "epoch": 0.8968,
      "grad_norm": 0.9139935970306396,
      "learning_rate": 3.2216320083104434e-06,
      "loss": 0.1576,
      "step": 4484
    },
    {
      "epoch": 0.8968,
      "loss_align": 0.14513146877288818,
      "loss_contrastive": 0.05836296081542969,
      "loss_gen": 0.3271481692790985,
      "neg_sim_mean": 0.8132314682006836,
      "pos_sim_mean": 0.8548685312271118,
      "step": 4484
    },
    {
      "epoch": 0.897,
      "grad_norm": 2.335576057434082,
      "learning_rate": 3.2093162080754637e-06,
      "loss": 0.4793,
      "step": 4485
    },
    {
      "epoch": 0.897,
      "loss_align": 0.0711284875869751,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27127522230148315,
      "neg_sim_mean": 0.6519634127616882,
      "pos_sim_mean": 0.9288715124130249,
      "step": 4485
    },
    {
      "epoch": 0.8972,
      "grad_norm": 1.4116084575653076,
      "learning_rate": 3.197023213051337e-06,
      "loss": 0.3424,
      "step": 4486
    },
    {
      "epoch": 0.8972,
      "loss_align": 0.12048840522766113,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004724144469946623,
      "neg_sim_mean": 0.3146391212940216,
      "pos_sim_mean": 0.8795115947723389,
      "step": 4486
    },
    {
      "epoch": 0.8974,
      "grad_norm": 0.5525318384170532,
      "learning_rate": 3.1847530292295313e-06,
      "loss": 0.1252,
      "step": 4487
    },
    {
      "epoch": 0.8974,
      "loss_align": 0.08042681217193604,
      "loss_contrastive": 0.02725815773010254,
      "loss_gen": 0.2324068695306778,
      "neg_sim_mean": 0.8468313217163086,
      "pos_sim_mean": 0.919573187828064,
      "step": 4487
    },
    {
      "epoch": 0.8976,
      "grad_norm": 1.9060248136520386,
      "learning_rate": 3.172505662590386e-06,
      "loss": 0.3161,
      "step": 4488
    },
    {
      "epoch": 0.8976,
      "loss_align": 0.06766045093536377,
      "loss_contrastive": 0.07715356349945068,
      "loss_gen": 0.22351253032684326,
      "neg_sim_mean": 0.909493088722229,
      "pos_sim_mean": 0.9323395490646362,
      "step": 4488
    },
    {
      "epoch": 0.8978,
      "grad_norm": 1.5335264205932617,
      "learning_rate": 3.160281119103109e-06,
      "loss": 0.3004,
      "step": 4489
    },
    {
      "epoch": 0.8978,
      "loss_align": 0.06651973724365234,
      "loss_contrastive": 0.025217711925506592,
      "loss_gen": 0.13056929409503937,
      "neg_sim_mean": 0.8586979508399963,
      "pos_sim_mean": 0.9334802627563477,
      "step": 4489
    },
    {
      "epoch": 0.898,
      "grad_norm": 1.5489236116409302,
      "learning_rate": 3.148079404725801e-06,
      "loss": 0.2001,
      "step": 4490
    },
    {
      "epoch": 0.898,
      "loss_align": 0.09391570091247559,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20907989144325256,
      "neg_sim_mean": 0.699838399887085,
      "pos_sim_mean": 0.9060842990875244,
      "step": 4490
    },
    {
      "epoch": 0.8982,
      "grad_norm": 1.7409782409667969,
      "learning_rate": 3.1359005254054273e-06,
      "loss": 0.303,
      "step": 4491
    },
    {
      "epoch": 0.8982,
      "loss_align": 0.14781486988067627,
      "loss_contrastive": 0.09905701875686646,
      "loss_gen": 0.2883399426937103,
      "neg_sim_mean": 0.8512421250343323,
      "pos_sim_mean": 0.8521851301193237,
      "step": 4491
    },
    {
      "epoch": 0.8984,
      "grad_norm": 2.2774858474731445,
      "learning_rate": 3.1237444870778287e-06,
      "loss": 0.448,
      "step": 4492
    },
    {
      "epoch": 0.8984,
      "loss_align": 0.0669703483581543,
      "loss_contrastive": 0.017481744289398193,
      "loss_gen": 0.11905995011329651,
      "neg_sim_mean": 0.850511372089386,
      "pos_sim_mean": 0.9330296516418457,
      "step": 4492
    },
    {
      "epoch": 0.8986,
      "grad_norm": 1.6431373357772827,
      "learning_rate": 3.1116112956677046e-06,
      "loss": 0.1881,
      "step": 4493
    },
    {
      "epoch": 0.8986,
      "loss_align": 0.13318336009979248,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14541715383529663,
      "neg_sim_mean": 0.5636596083641052,
      "pos_sim_mean": 0.8668166399002075,
      "step": 4493
    },
    {
      "epoch": 0.8988,
      "grad_norm": 1.6791678667068481,
      "learning_rate": 3.0995009570886303e-06,
      "loss": 0.2786,
      "step": 4494
    },
    {
      "epoch": 0.8988,
      "loss_align": 0.0568622350692749,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10796202719211578,
      "neg_sim_mean": 0.46417346596717834,
      "pos_sim_mean": 0.9431377649307251,
      "step": 4494
    },
    {
      "epoch": 0.899,
      "grad_norm": 1.0531363487243652,
      "learning_rate": 3.087413477243034e-06,
      "loss": 0.1648,
      "step": 4495
    },
    {
      "epoch": 0.899,
      "loss_align": 0.08933401107788086,
      "loss_contrastive": 0.09081101417541504,
      "loss_gen": 0.3972245156764984,
      "neg_sim_mean": 0.9014769792556763,
      "pos_sim_mean": 0.9106659889221191,
      "step": 4495
    },
    {
      "epoch": 0.8992,
      "grad_norm": 2.300489902496338,
      "learning_rate": 3.0753488620222037e-06,
      "loss": 0.4975,
      "step": 4496
    },
    {
      "epoch": 0.8992,
      "loss_align": 0.10153383016586304,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00196368177421391,
      "neg_sim_mean": 0.30337029695510864,
      "pos_sim_mean": 0.898466169834137,
      "step": 4496
    },
    {
      "epoch": 0.8994,
      "grad_norm": 0.3722669780254364,
      "learning_rate": 3.0633071173062967e-06,
      "loss": 0.1035,
      "step": 4497
    },
    {
      "epoch": 0.8994,
      "loss_align": 0.11444950103759766,
      "loss_contrastive": 0.017824769020080566,
      "loss_gen": 0.29342231154441833,
      "neg_sim_mean": 0.803375244140625,
      "pos_sim_mean": 0.8855504989624023,
      "step": 4497
    },
    {
      "epoch": 0.8996,
      "grad_norm": 2.2428202629089355,
      "learning_rate": 3.051288248964307e-06,
      "loss": 0.41,
      "step": 4498
    },
    {
      "epoch": 0.8996,
      "loss_align": 0.08165156841278076,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15937119722366333,
      "neg_sim_mean": 0.8120843172073364,
      "pos_sim_mean": 0.9183484315872192,
      "step": 4498
    },
    {
      "epoch": 0.8998,
      "grad_norm": 1.8546404838562012,
      "learning_rate": 3.039292262854088e-06,
      "loss": 0.241,
      "step": 4499
    },
    {
      "epoch": 0.8998,
      "loss_align": 0.11257076263427734,
      "loss_contrastive": 0.10916554927825928,
      "loss_gen": 0.27149078249931335,
      "neg_sim_mean": 0.896594762802124,
      "pos_sim_mean": 0.8874292373657227,
      "step": 4499
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.799774408340454,
      "learning_rate": 3.0273191648223287e-06,
      "loss": 0.3972,
      "step": 4500
    },
    {
      "epoch": 0.9,
      "loss_align": 0.18693864345550537,
      "loss_contrastive": 0.08771014213562012,
      "loss_gen": 0.21051347255706787,
      "neg_sim_mean": 0.8007714748382568,
      "pos_sim_mean": 0.8130613565444946,
      "step": 4500
    },
    {
      "epoch": 0.9002,
      "grad_norm": 1.673595905303955,
      "learning_rate": 3.0153689607045845e-06,
      "loss": 0.408,
      "step": 4501
    },
    {
      "epoch": 0.9002,
      "loss_align": 0.08431023359298706,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0026810872368514538,
      "neg_sim_mean": 0.41714584827423096,
      "pos_sim_mean": 0.9156897664070129,
      "step": 4501
    },
    {
      "epoch": 0.9004,
      "grad_norm": 0.4379250109195709,
      "learning_rate": 3.0034416563252287e-06,
      "loss": 0.087,
      "step": 4502
    },
    {
      "epoch": 0.9004,
      "loss_align": 0.11111372709274292,
      "loss_contrastive": 0.097087562084198,
      "loss_gen": 0.10330591350793839,
      "neg_sim_mean": 0.8859738111495972,
      "pos_sim_mean": 0.8888862729072571,
      "step": 4502
    },
    {
      "epoch": 0.9006,
      "grad_norm": 1.339682936668396,
      "learning_rate": 2.99153725749749e-06,
      "loss": 0.2261,
      "step": 4503
    },
    {
      "epoch": 0.9006,
      "loss_align": 0.08340096473693848,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01824457198381424,
      "neg_sim_mean": 0.3779846727848053,
      "pos_sim_mean": 0.9165990352630615,
      "step": 4503
    },
    {
      "epoch": 0.9008,
      "grad_norm": 0.9308257102966309,
      "learning_rate": 2.9796557700234318e-06,
      "loss": 0.1016,
      "step": 4504
    },
    {
      "epoch": 0.9008,
      "loss_align": 0.08998429775238037,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10382342338562012,
      "neg_sim_mean": 0.8010267019271851,
      "pos_sim_mean": 0.9100157022476196,
      "step": 4504
    },
    {
      "epoch": 0.901,
      "grad_norm": 1.3082019090652466,
      "learning_rate": 2.967797199693928e-06,
      "loss": 0.1938,
      "step": 4505
    },
    {
      "epoch": 0.901,
      "loss_align": 0.05826389789581299,
      "loss_contrastive": 0.028893470764160156,
      "loss_gen": 0.11119816452264786,
      "neg_sim_mean": 0.8706295490264893,
      "pos_sim_mean": 0.941736102104187,
      "step": 4505
    },
    {
      "epoch": 0.9012,
      "grad_norm": 1.092234492301941,
      "learning_rate": 2.9559615522887273e-06,
      "loss": 0.1729,
      "step": 4506
    },
    {
      "epoch": 0.9012,
      "loss_align": 0.1395452618598938,
      "loss_contrastive": 0.11194008588790894,
      "loss_gen": 0.13996735215187073,
      "neg_sim_mean": 0.8723948001861572,
      "pos_sim_mean": 0.8604547381401062,
      "step": 4506
    },
    {
      "epoch": 0.9014,
      "grad_norm": 1.2177759408950806,
      "learning_rate": 2.9441488335763656e-06,
      "loss": 0.2929,
      "step": 4507
    },
    {
      "epoch": 0.9014,
      "loss_align": 0.09591996669769287,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22790057957172394,
      "neg_sim_mean": 0.7835734486579895,
      "pos_sim_mean": 0.9040800333023071,
      "step": 4507
    },
    {
      "epoch": 0.9016,
      "grad_norm": 1.685754418373108,
      "learning_rate": 2.9323590493142205e-06,
      "loss": 0.3238,
      "step": 4508
    },
    {
      "epoch": 0.9016,
      "loss_align": 0.1011890172958374,
      "loss_contrastive": 0.0640261173248291,
      "loss_gen": 0.37426498532295227,
      "neg_sim_mean": 0.8628370761871338,
      "pos_sim_mean": 0.8988109827041626,
      "step": 4508
    },
    {
      "epoch": 0.9018,
      "grad_norm": 2.1006298065185547,
      "learning_rate": 2.9205922052484958e-06,
      "loss": 0.4831,
      "step": 4509
    },
    {
      "epoch": 0.9018,
      "loss_align": 0.13191258907318115,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005448134616017342,
      "neg_sim_mean": 0.5282483100891113,
      "pos_sim_mean": 0.8680874109268188,
      "step": 4509
    },
    {
      "epoch": 0.902,
      "grad_norm": 0.5309595465660095,
      "learning_rate": 2.9088483071141983e-06,
      "loss": 0.1374,
      "step": 4510
    },
    {
      "epoch": 0.902,
      "loss_align": 0.055438876152038574,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23114445805549622,
      "neg_sim_mean": 0.8377888202667236,
      "pos_sim_mean": 0.9445611238479614,
      "step": 4510
    },
    {
      "epoch": 0.9022,
      "grad_norm": 1.758642315864563,
      "learning_rate": 2.8971273606351658e-06,
      "loss": 0.2866,
      "step": 4511
    },
    {
      "epoch": 0.9022,
      "loss_align": 0.08267742395401001,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15042130649089813,
      "neg_sim_mean": 0.7779898047447205,
      "pos_sim_mean": 0.91732257604599,
      "step": 4511
    },
    {
      "epoch": 0.9024,
      "grad_norm": 1.369461178779602,
      "learning_rate": 2.8854293715240456e-06,
      "loss": 0.2331,
      "step": 4512
    },
    {
      "epoch": 0.9024,
      "loss_align": 0.08424937725067139,
      "loss_contrastive": 0.0,
      "loss_gen": 0.30474725365638733,
      "neg_sim_mean": 0.6499695777893066,
      "pos_sim_mean": 0.9157506227493286,
      "step": 4512
    },
    {
      "epoch": 0.9026,
      "grad_norm": 1.9776124954223633,
      "learning_rate": 2.873754345482299e-06,
      "loss": 0.389,
      "step": 4513
    },
    {
      "epoch": 0.9026,
      "loss_align": 0.17961692810058594,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22613869607448578,
      "neg_sim_mean": 0.7142484188079834,
      "pos_sim_mean": 0.8203830718994141,
      "step": 4513
    },
    {
      "epoch": 0.9028,
      "grad_norm": 1.7093868255615234,
      "learning_rate": 2.8621022882001856e-06,
      "loss": 0.4058,
      "step": 4514
    },
    {
      "epoch": 0.9028,
      "loss_align": 0.07094287872314453,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10875924676656723,
      "neg_sim_mean": 0.7966642379760742,
      "pos_sim_mean": 0.9290571212768555,
      "step": 4514
    },
    {
      "epoch": 0.903,
      "grad_norm": 1.3726823329925537,
      "learning_rate": 2.850473205356774e-06,
      "loss": 0.1797,
      "step": 4515
    },
    {
      "epoch": 0.903,
      "loss_align": 0.12112832069396973,
      "loss_contrastive": 0.057196199893951416,
      "loss_gen": 0.23557740449905396,
      "neg_sim_mean": 0.8360678553581238,
      "pos_sim_mean": 0.8788716793060303,
      "step": 4515
    },
    {
      "epoch": 0.9032,
      "grad_norm": 1.84222412109375,
      "learning_rate": 2.8388671026199522e-06,
      "loss": 0.3636,
      "step": 4516
    },
    {
      "epoch": 0.9032,
      "loss_align": 0.18342947959899902,
      "loss_contrastive": 0.11022579669952393,
      "loss_gen": 0.11573934555053711,
      "neg_sim_mean": 0.826796293258667,
      "pos_sim_mean": 0.816570520401001,
      "step": 4516
    },
    {
      "epoch": 0.9034,
      "grad_norm": 1.4657586812973022,
      "learning_rate": 2.8272839856463783e-06,
      "loss": 0.3124,
      "step": 4517
    },
    {
      "epoch": 0.9034,
      "loss_align": 0.08852905035018921,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008921893313527107,
      "neg_sim_mean": 0.2927681803703308,
      "pos_sim_mean": 0.9114709496498108,
      "step": 4517
    },
    {
      "epoch": 0.9036,
      "grad_norm": 0.5981618165969849,
      "learning_rate": 2.8157238600815372e-06,
      "loss": 0.0975,
      "step": 4518
    },
    {
      "epoch": 0.9036,
      "loss_align": 0.10090857744216919,
      "loss_contrastive": 0.053527891635894775,
      "loss_gen": 0.16935068368911743,
      "neg_sim_mean": 0.8526192903518677,
      "pos_sim_mean": 0.8990914225578308,
      "step": 4518
    },
    {
      "epoch": 0.9038,
      "grad_norm": 1.5642412900924683,
      "learning_rate": 2.804186731559677e-06,
      "loss": 0.2767,
      "step": 4519
    },
    {
      "epoch": 0.9038,
      "loss_align": 0.10272473096847534,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011257770471274853,
      "neg_sim_mean": 0.37869930267333984,
      "pos_sim_mean": 0.8972752690315247,
      "step": 4519
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.7180716395378113,
      "learning_rate": 2.7926726057038666e-06,
      "loss": 0.114,
      "step": 4520
    },
    {
      "epoch": 0.904,
      "loss_align": 0.1148521900177002,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002991270739585161,
      "neg_sim_mean": 0.16183117032051086,
      "pos_sim_mean": 0.8851478099822998,
      "step": 4520
    },
    {
      "epoch": 0.9042,
      "grad_norm": 0.513883113861084,
      "learning_rate": 2.7811814881259503e-06,
      "loss": 0.1178,
      "step": 4521
    },
    {
      "epoch": 0.9042,
      "loss_align": 0.11007606983184814,
      "loss_contrastive": 0.024799466133117676,
      "loss_gen": 0.2472524791955948,
      "neg_sim_mean": 0.8147233724594116,
      "pos_sim_mean": 0.8899239301681519,
      "step": 4521
    },
    {
      "epoch": 0.9044,
      "grad_norm": 2.058635950088501,
      "learning_rate": 2.7697133844265533e-06,
      "loss": 0.3603,
      "step": 4522
    },
    {
      "epoch": 0.9044,
      "loss_align": 0.10939228534698486,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19191549718379974,
      "neg_sim_mean": 0.6948847770690918,
      "pos_sim_mean": 0.8906077146530151,
      "step": 4522
    },
    {
      "epoch": 0.9046,
      "grad_norm": 1.5491856336593628,
      "learning_rate": 2.758268300195094e-06,
      "loss": 0.3013,
      "step": 4523
    },
    {
      "epoch": 0.9046,
      "loss_align": 0.11644423007965088,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13917089998722076,
      "neg_sim_mean": 0.7777715921401978,
      "pos_sim_mean": 0.8835557699203491,
      "step": 4523
    },
    {
      "epoch": 0.9048,
      "grad_norm": 1.2533214092254639,
      "learning_rate": 2.746846241009765e-06,
      "loss": 0.2556,
      "step": 4524
    },
    {
      "epoch": 0.9048,
      "loss_align": 0.06824976205825806,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003744601272046566,
      "neg_sim_mean": 0.36125001311302185,
      "pos_sim_mean": 0.9317502379417419,
      "step": 4524
    },
    {
      "epoch": 0.905,
      "grad_norm": 0.4064747989177704,
      "learning_rate": 2.735447212437531e-06,
      "loss": 0.072,
      "step": 4525
    },
    {
      "epoch": 0.905,
      "loss_align": 0.06651192903518677,
      "loss_contrastive": 0.04262596368789673,
      "loss_gen": 0.2673202455043793,
      "neg_sim_mean": 0.876114010810852,
      "pos_sim_mean": 0.9334880709648132,
      "step": 4525
    },
    {
      "epoch": 0.9052,
      "grad_norm": 1.675253987312317,
      "learning_rate": 2.724071220034158e-06,
      "loss": 0.3389,
      "step": 4526
    },
    {
      "epoch": 0.9052,
      "loss_align": 0.10886788368225098,
      "loss_contrastive": 0.09286379814147949,
      "loss_gen": 0.37930265069007874,
      "neg_sim_mean": 0.8839958906173706,
      "pos_sim_mean": 0.891132116317749,
      "step": 4526
    },
    {
      "epoch": 0.9054,
      "grad_norm": 2.0578932762145996,
      "learning_rate": 2.712718269344161e-06,
      "loss": 0.4993,
      "step": 4527
    },
    {
      "epoch": 0.9054,
      "loss_align": 0.11428076028823853,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15855908393859863,
      "neg_sim_mean": 0.7198590636253357,
      "pos_sim_mean": 0.8857192397117615,
      "step": 4527
    },
    {
      "epoch": 0.9056,
      "grad_norm": 1.832442045211792,
      "learning_rate": 2.701388365900831e-06,
      "loss": 0.2728,
      "step": 4528
    },
    {
      "epoch": 0.9056,
      "loss_align": 0.11755573749542236,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25459024310112,
      "neg_sim_mean": 0.7598881721496582,
      "pos_sim_mean": 0.8824442625045776,
      "step": 4528
    },
    {
      "epoch": 0.9058,
      "grad_norm": 2.144118309020996,
      "learning_rate": 2.690081515226206e-06,
      "loss": 0.3721,
      "step": 4529
    },
    {
      "epoch": 0.9058,
      "loss_align": 0.12108683586120605,
      "loss_contrastive": 0.011783123016357422,
      "loss_gen": 0.2718813717365265,
      "neg_sim_mean": 0.7906962633132935,
      "pos_sim_mean": 0.878913164138794,
      "step": 4529
    },
    {
      "epoch": 0.906,
      "grad_norm": 1.912163257598877,
      "learning_rate": 2.6787977228311334e-06,
      "loss": 0.3944,
      "step": 4530
    },
    {
      "epoch": 0.906,
      "loss_align": 0.09296858310699463,
      "loss_contrastive": 0.10506939888000488,
      "loss_gen": 0.1369996815919876,
      "neg_sim_mean": 0.9121007919311523,
      "pos_sim_mean": 0.9070314168930054,
      "step": 4530
    },
    {
      "epoch": 0.9062,
      "grad_norm": 1.3333022594451904,
      "learning_rate": 2.667536994215186e-06,
      "loss": 0.2426,
      "step": 4531
    },
    {
      "epoch": 0.9062,
      "loss_align": 0.11128711700439453,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24730680882930756,
      "neg_sim_mean": 0.7886766791343689,
      "pos_sim_mean": 0.8887128829956055,
      "step": 4531
    },
    {
      "epoch": 0.9064,
      "grad_norm": 1.7806342840194702,
      "learning_rate": 2.656299334866702e-06,
      "loss": 0.3586,
      "step": 4532
    },
    {
      "epoch": 0.9064,
      "loss_align": 0.09277737140655518,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002470368519425392,
      "neg_sim_mean": 0.4137924909591675,
      "pos_sim_mean": 0.9072226285934448,
      "step": 4532
    },
    {
      "epoch": 0.9066,
      "grad_norm": 0.439256876707077,
      "learning_rate": 2.6450847502627884e-06,
      "loss": 0.0952,
      "step": 4533
    },
    {
      "epoch": 0.9066,
      "loss_align": 0.09303325414657593,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3195514380931854,
      "neg_sim_mean": 0.6547958254814148,
      "pos_sim_mean": 0.9069667458534241,
      "step": 4533
    },
    {
      "epoch": 0.9068,
      "grad_norm": 2.1613171100616455,
      "learning_rate": 2.6338932458692843e-06,
      "loss": 0.4126,
      "step": 4534
    },
    {
      "epoch": 0.9068,
      "loss_align": 0.0967913269996643,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012515044771134853,
      "neg_sim_mean": 0.26485493779182434,
      "pos_sim_mean": 0.9032086730003357,
      "step": 4534
    },
    {
      "epoch": 0.907,
      "grad_norm": 0.7005767226219177,
      "learning_rate": 2.6227248271408157e-06,
      "loss": 0.1093,
      "step": 4535
    },
    {
      "epoch": 0.907,
      "loss_align": 0.09689927101135254,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22663304209709167,
      "neg_sim_mean": 0.7005035877227783,
      "pos_sim_mean": 0.9031007289886475,
      "step": 4535
    },
    {
      "epoch": 0.9072,
      "grad_norm": 1.5715136528015137,
      "learning_rate": 2.611579499520722e-06,
      "loss": 0.3235,
      "step": 4536
    },
    {
      "epoch": 0.9072,
      "loss_align": 0.10056620836257935,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0014838934876024723,
      "neg_sim_mean": 0.31459784507751465,
      "pos_sim_mean": 0.8994337916374207,
      "step": 4536
    },
    {
      "epoch": 0.9074,
      "grad_norm": 0.438541442155838,
      "learning_rate": 2.6004572684410923e-06,
      "loss": 0.1021,
      "step": 4537
    },
    {
      "epoch": 0.9074,
      "loss_align": 0.04128432273864746,
      "loss_contrastive": 0.0,
      "loss_gen": 0.08800207078456879,
      "neg_sim_mean": 0.8455657958984375,
      "pos_sim_mean": 0.9587156772613525,
      "step": 4537
    },
    {
      "epoch": 0.9076,
      "grad_norm": 1.5854434967041016,
      "learning_rate": 2.5893581393227674e-06,
      "loss": 0.1293,
      "step": 4538
    },
    {
      "epoch": 0.9076,
      "loss_align": 0.08251422643661499,
      "loss_contrastive": 0.0,
      "loss_gen": 0.026252465322613716,
      "neg_sim_mean": 0.4044012427330017,
      "pos_sim_mean": 0.917485773563385,
      "step": 4538
    },
    {
      "epoch": 0.9078,
      "grad_norm": 1.4649182558059692,
      "learning_rate": 2.5782821175753422e-06,
      "loss": 0.1088,
      "step": 4539
    },
    {
      "epoch": 0.9078,
      "loss_align": 0.10931658744812012,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10975723713636398,
      "neg_sim_mean": 0.7654924392700195,
      "pos_sim_mean": 0.8906834125518799,
      "step": 4539
    },
    {
      "epoch": 0.908,
      "grad_norm": 1.3178941011428833,
      "learning_rate": 2.567229208597127e-06,
      "loss": 0.2191,
      "step": 4540
    },
    {
      "epoch": 0.908,
      "loss_align": 0.08830946683883667,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0031709789764136076,
      "neg_sim_mean": 0.3725827932357788,
      "pos_sim_mean": 0.9116905331611633,
      "step": 4540
    },
    {
      "epoch": 0.9082,
      "grad_norm": 0.4925350546836853,
      "learning_rate": 2.5561994177751737e-06,
      "loss": 0.0915,
      "step": 4541
    },
    {
      "epoch": 0.9082,
      "loss_align": 0.06184518337249756,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1355201154947281,
      "neg_sim_mean": 0.5327656269073486,
      "pos_sim_mean": 0.9381548166275024,
      "step": 4541
    },
    {
      "epoch": 0.9084,
      "grad_norm": 1.3287855386734009,
      "learning_rate": 2.5451927504852757e-06,
      "loss": 0.1974,
      "step": 4542
    },
    {
      "epoch": 0.9084,
      "loss_align": 0.10724294185638428,
      "loss_contrastive": 0.029347360134124756,
      "loss_gen": 0.16858729720115662,
      "neg_sim_mean": 0.8221043944358826,
      "pos_sim_mean": 0.8927570581436157,
      "step": 4542
    },
    {
      "epoch": 0.9086,
      "grad_norm": 1.746212124824524,
      "learning_rate": 2.534209212091937e-06,
      "loss": 0.2794,
      "step": 4543
    },
    {
      "epoch": 0.9086,
      "loss_align": 0.08449280261993408,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13306593894958496,
      "neg_sim_mean": 0.7056853771209717,
      "pos_sim_mean": 0.9155071973800659,
      "step": 4543
    },
    {
      "epoch": 0.9088,
      "grad_norm": 1.5717159509658813,
      "learning_rate": 2.523248807948403e-06,
      "loss": 0.2176,
      "step": 4544
    },
    {
      "epoch": 0.9088,
      "loss_align": 0.11172956228256226,
      "loss_contrastive": 0.0,
      "loss_gen": 0.37026476860046387,
      "neg_sim_mean": 0.7742573022842407,
      "pos_sim_mean": 0.8882704377174377,
      "step": 4544
    },
    {
      "epoch": 0.909,
      "grad_norm": 2.5186002254486084,
      "learning_rate": 2.5123115433966614e-06,
      "loss": 0.482,
      "step": 4545
    },
    {
      "epoch": 0.909,
      "loss_align": 0.14442622661590576,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2736373841762543,
      "neg_sim_mean": 0.3933939039707184,
      "pos_sim_mean": 0.8555737733840942,
      "step": 4545
    },
    {
      "epoch": 0.9092,
      "grad_norm": 2.099456787109375,
      "learning_rate": 2.501397423767382e-06,
      "loss": 0.4181,
      "step": 4546
    },
    {
      "epoch": 0.9092,
      "loss_align": 0.0699813961982727,
      "loss_contrastive": 0.07740062475204468,
      "loss_gen": 0.11632473021745682,
      "neg_sim_mean": 0.9074192047119141,
      "pos_sim_mean": 0.9300186038017273,
      "step": 4546
    },
    {
      "epoch": 0.9094,
      "grad_norm": 1.3606468439102173,
      "learning_rate": 2.4905064543799705e-06,
      "loss": 0.1956,
      "step": 4547
    },
    {
      "epoch": 0.9094,
      "loss_align": 0.1278972029685974,
      "loss_contrastive": 0.0,
      "loss_gen": 0.015523020178079605,
      "neg_sim_mean": 0.25343018770217896,
      "pos_sim_mean": 0.8721027970314026,
      "step": 4547
    },
    {
      "epoch": 0.9096,
      "grad_norm": 0.9415231943130493,
      "learning_rate": 2.4796386405425643e-06,
      "loss": 0.1434,
      "step": 4548
    },
    {
      "epoch": 0.9096,
      "loss_align": 0.18673038482666016,
      "loss_contrastive": 0.0,
      "loss_gen": 0.308743417263031,
      "neg_sim_mean": 0.651120662689209,
      "pos_sim_mean": 0.8132696151733398,
      "step": 4548
    },
    {
      "epoch": 0.9098,
      "grad_norm": 2.1920905113220215,
      "learning_rate": 2.468793987551998e-06,
      "loss": 0.4955,
      "step": 4549
    },
    {
      "epoch": 0.9098,
      "loss_align": 0.09989845752716064,
      "loss_contrastive": 0.06996381282806396,
      "loss_gen": 0.39637288451194763,
      "neg_sim_mean": 0.8700653314590454,
      "pos_sim_mean": 0.9001015424728394,
      "step": 4549
    },
    {
      "epoch": 0.91,
      "grad_norm": 2.0162336826324463,
      "learning_rate": 2.4579725006938337e-06,
      "loss": 0.5047,
      "step": 4550
    },
    {
      "epoch": 0.91,
      "loss_align": 0.08149659633636475,
      "loss_contrastive": 0.048238933086395264,
      "loss_gen": 0.17326219379901886,
      "neg_sim_mean": 0.8667423129081726,
      "pos_sim_mean": 0.9185034036636353,
      "step": 4550
    },
    {
      "epoch": 0.9102,
      "grad_norm": 1.6896950006484985,
      "learning_rate": 2.4471741852423237e-06,
      "loss": 0.2605,
      "step": 4551
    },
    {
      "epoch": 0.9102,
      "loss_align": 0.20744049549102783,
      "loss_contrastive": 0.0,
      "loss_gen": 0.04538159817457199,
      "neg_sim_mean": 0.537384033203125,
      "pos_sim_mean": 0.7925595045089722,
      "step": 4551
    },
    {
      "epoch": 0.9104,
      "grad_norm": 2.7693400382995605,
      "learning_rate": 2.436399046460436e-06,
      "loss": 0.2528,
      "step": 4552
    },
    {
      "epoch": 0.9104,
      "loss_align": 0.08262574672698975,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16152770817279816,
      "neg_sim_mean": 0.7978000044822693,
      "pos_sim_mean": 0.9173742532730103,
      "step": 4552
    },
    {
      "epoch": 0.9106,
      "grad_norm": 1.9809476137161255,
      "learning_rate": 2.4256470895998363e-06,
      "loss": 0.2442,
      "step": 4553
    },
    {
      "epoch": 0.9106,
      "loss_align": 0.0720747709274292,
      "loss_contrastive": 0.059251368045806885,
      "loss_gen": 0.2872375547885895,
      "neg_sim_mean": 0.8871765732765198,
      "pos_sim_mean": 0.9279252290725708,
      "step": 4553
    },
    {
      "epoch": 0.9108,
      "grad_norm": 2.0303735733032227,
      "learning_rate": 2.4149183199009216e-06,
      "loss": 0.3664,
      "step": 4554
    },
    {
      "epoch": 0.9108,
      "loss_align": 0.0457226037979126,
      "loss_contrastive": 0.015123724937438965,
      "loss_gen": 0.27172818779945374,
      "neg_sim_mean": 0.8694010972976685,
      "pos_sim_mean": 0.9542773962020874,
      "step": 4554
    },
    {
      "epoch": 0.911,
      "grad_norm": 1.9645227193832397,
      "learning_rate": 2.404212742592743e-06,
      "loss": 0.3193,
      "step": 4555
    },
    {
      "epoch": 0.911,
      "loss_align": 0.1585170030593872,
      "loss_contrastive": 0.0,
      "loss_gen": 0.321357399225235,
      "neg_sim_mean": 0.5194098949432373,
      "pos_sim_mean": 0.8414829969406128,
      "step": 4555
    },
    {
      "epoch": 0.9112,
      "grad_norm": 2.139200448989868,
      "learning_rate": 2.3935303628930707e-06,
      "loss": 0.4799,
      "step": 4556
    },
    {
      "epoch": 0.9112,
      "loss_align": 0.20495635271072388,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0011020614765584469,
      "neg_sim_mean": 0.36863040924072266,
      "pos_sim_mean": 0.7950436472892761,
      "step": 4556
    },
    {
      "epoch": 0.9114,
      "grad_norm": 0.6160252094268799,
      "learning_rate": 2.3828711860083674e-06,
      "loss": 0.2061,
      "step": 4557
    },
    {
      "epoch": 0.9114,
      "loss_align": 0.0961376428604126,
      "loss_contrastive": 0.0,
      "loss_gen": 0.272884339094162,
      "neg_sim_mean": 0.6984564065933228,
      "pos_sim_mean": 0.9038623571395874,
      "step": 4557
    },
    {
      "epoch": 0.9116,
      "grad_norm": 2.006680727005005,
      "learning_rate": 2.3722352171337834e-06,
      "loss": 0.369,
      "step": 4558
    },
    {
      "epoch": 0.9116,
      "loss_align": 0.09466445446014404,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2443021833896637,
      "neg_sim_mean": 0.7817896604537964,
      "pos_sim_mean": 0.905335545539856,
      "step": 4558
    },
    {
      "epoch": 0.9118,
      "grad_norm": 1.7018154859542847,
      "learning_rate": 2.361622461453178e-06,
      "loss": 0.339,
      "step": 4559
    },
    {
      "epoch": 0.9118,
      "loss_align": 0.05968892574310303,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23160691559314728,
      "neg_sim_mean": 0.8116892576217651,
      "pos_sim_mean": 0.940311074256897,
      "step": 4559
    },
    {
      "epoch": 0.912,
      "grad_norm": 1.6570228338241577,
      "learning_rate": 2.351032924139063e-06,
      "loss": 0.2913,
      "step": 4560
    },
    {
      "epoch": 0.912,
      "loss_align": 0.15691113471984863,
      "loss_contrastive": 0.1077876091003418,
      "loss_gen": 0.39256906509399414,
      "neg_sim_mean": 0.8508764505386353,
      "pos_sim_mean": 0.8430888652801514,
      "step": 4560
    },
    {
      "epoch": 0.9122,
      "grad_norm": 2.357665538787842,
      "learning_rate": 2.340466610352654e-06,
      "loss": 0.5624,
      "step": 4561
    },
    {
      "epoch": 0.9122,
      "loss_align": 0.19630634784698486,
      "loss_contrastive": 0.0,
      "loss_gen": 0.40012800693511963,
      "neg_sim_mean": 0.5430376529693604,
      "pos_sim_mean": 0.8036936521530151,
      "step": 4561
    },
    {
      "epoch": 0.9124,
      "grad_norm": 2.5834381580352783,
      "learning_rate": 2.3299235252438434e-06,
      "loss": 0.5964,
      "step": 4562
    },
    {
      "epoch": 0.9124,
      "loss_align": 0.05497395992279053,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20664608478546143,
      "neg_sim_mean": 0.2776280641555786,
      "pos_sim_mean": 0.9450260400772095,
      "step": 4562
    },
    {
      "epoch": 0.9126,
      "grad_norm": 1.7564704418182373,
      "learning_rate": 2.319403673951204e-06,
      "loss": 0.2616,
      "step": 4563
    },
    {
      "epoch": 0.9126,
      "loss_align": 0.15136855840682983,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007583873812109232,
      "neg_sim_mean": 0.3194439113140106,
      "pos_sim_mean": 0.8486314415931702,
      "step": 4563
    },
    {
      "epoch": 0.9128,
      "grad_norm": 0.624131441116333,
      "learning_rate": 2.3089070616019835e-06,
      "loss": 0.159,
      "step": 4564
    },
    {
      "epoch": 0.9128,
      "loss_align": 0.07986277341842651,
      "loss_contrastive": 0.05605798959732056,
      "loss_gen": 0.2701430916786194,
      "neg_sim_mean": 0.8761951923370361,
      "pos_sim_mean": 0.9201372265815735,
      "step": 4564
    },
    {
      "epoch": 0.913,
      "grad_norm": 2.0039126873016357,
      "learning_rate": 2.2984336933121075e-06,
      "loss": 0.3567,
      "step": 4565
    },
    {
      "epoch": 0.913,
      "loss_align": 0.08514666557312012,
      "loss_contrastive": 0.031162142753601074,
      "loss_gen": 0.32454565167427063,
      "neg_sim_mean": 0.846015453338623,
      "pos_sim_mean": 0.9148533344268799,
      "step": 4565
    },
    {
      "epoch": 0.9132,
      "grad_norm": 1.6850265264511108,
      "learning_rate": 2.2879835741861586e-06,
      "loss": 0.4134,
      "step": 4566
    },
    {
      "epoch": 0.9132,
      "loss_align": 0.11796492338180542,
      "loss_contrastive": 0.0,
      "loss_gen": 0.28730544447898865,
      "neg_sim_mean": 0.7537323236465454,
      "pos_sim_mean": 0.8820350766181946,
      "step": 4566
    },
    {
      "epoch": 0.9134,
      "grad_norm": 2.5393590927124023,
      "learning_rate": 2.2775567093174023e-06,
      "loss": 0.4053,
      "step": 4567
    },
    {
      "epoch": 0.9134,
      "loss_align": 0.15626227855682373,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27191412448883057,
      "neg_sim_mean": 0.599732518196106,
      "pos_sim_mean": 0.8437377214431763,
      "step": 4567
    },
    {
      "epoch": 0.9136,
      "grad_norm": 2.339341640472412,
      "learning_rate": 2.2671531037877723e-06,
      "loss": 0.4282,
      "step": 4568
    },
    {
      "epoch": 0.9136,
      "loss_align": 0.04545736312866211,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2587733268737793,
      "neg_sim_mean": 0.840448260307312,
      "pos_sim_mean": 0.9545426368713379,
      "step": 4568
    },
    {
      "epoch": 0.9138,
      "grad_norm": 1.7136521339416504,
      "learning_rate": 2.2567727626678527e-06,
      "loss": 0.3042,
      "step": 4569
    },
    {
      "epoch": 0.9138,
      "loss_align": 0.15148264169692993,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2798152267932892,
      "neg_sim_mean": 0.5313713550567627,
      "pos_sim_mean": 0.8485173583030701,
      "step": 4569
    },
    {
      "epoch": 0.914,
      "grad_norm": 2.1189827919006348,
      "learning_rate": 2.2464156910168954e-06,
      "loss": 0.4313,
      "step": 4570
    },
    {
      "epoch": 0.914,
      "loss_align": 0.03517591953277588,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011548653244972229,
      "neg_sim_mean": 0.31727665662765503,
      "pos_sim_mean": 0.9648240804672241,
      "step": 4570
    },
    {
      "epoch": 0.9142,
      "grad_norm": 0.6187030673027039,
      "learning_rate": 2.2360818938828187e-06,
      "loss": 0.0467,
      "step": 4571
    },
    {
      "epoch": 0.9142,
      "loss_align": 0.07000160217285156,
      "loss_contrastive": 0.09887200593948364,
      "loss_gen": 0.10561726987361908,
      "neg_sim_mean": 0.9288703799247742,
      "pos_sim_mean": 0.9299983978271484,
      "step": 4571
    },
    {
      "epoch": 0.9144,
      "grad_norm": 1.1806612014770508,
      "learning_rate": 2.2257713763021827e-06,
      "loss": 0.1875,
      "step": 4572
    },
    {
      "epoch": 0.9144,
      "loss_align": 0.17455601692199707,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16467498242855072,
      "neg_sim_mean": 0.6249656677246094,
      "pos_sim_mean": 0.8254439830780029,
      "step": 4572
    },
    {
      "epoch": 0.9146,
      "grad_norm": 1.520877718925476,
      "learning_rate": 2.2154841433002062e-06,
      "loss": 0.3392,
      "step": 4573
    },
    {
      "epoch": 0.9146,
      "loss_align": 0.3582957983016968,
      "loss_contrastive": 0.1691138744354248,
      "loss_gen": 0.16360916197299957,
      "neg_sim_mean": 0.7108180522918701,
      "pos_sim_mean": 0.6417042016983032,
      "step": 4573
    },
    {
      "epoch": 0.9148,
      "grad_norm": 2.1150596141815186,
      "learning_rate": 2.205220199890767e-06,
      "loss": 0.5422,
      "step": 4574
    },
    {
      "epoch": 0.9148,
      "loss_align": 0.07315719127655029,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1621292680501938,
      "neg_sim_mean": 0.8002058863639832,
      "pos_sim_mean": 0.9268428087234497,
      "step": 4574
    },
    {
      "epoch": 0.915,
      "grad_norm": 1.6603354215621948,
      "learning_rate": 2.194979551076387e-06,
      "loss": 0.2353,
      "step": 4575
    },
    {
      "epoch": 0.915,
      "loss_align": 0.10603535175323486,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2457987517118454,
      "neg_sim_mean": 0.7879025340080261,
      "pos_sim_mean": 0.8939646482467651,
      "step": 4575
    },
    {
      "epoch": 0.9152,
      "grad_norm": 2.228606939315796,
      "learning_rate": 2.1847622018482283e-06,
      "loss": 0.3518,
      "step": 4576
    },
    {
      "epoch": 0.9152,
      "loss_align": 0.1068679690361023,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16175365447998047,
      "neg_sim_mean": 0.7603048086166382,
      "pos_sim_mean": 0.8931320309638977,
      "step": 4576
    },
    {
      "epoch": 0.9154,
      "grad_norm": 1.5835806131362915,
      "learning_rate": 2.174568157186102e-06,
      "loss": 0.2686,
      "step": 4577
    },
    {
      "epoch": 0.9154,
      "loss_align": 0.10870784521102905,
      "loss_contrastive": 0.0,
      "loss_gen": 0.08837481588125229,
      "neg_sim_mean": 0.7555477023124695,
      "pos_sim_mean": 0.891292154788971,
      "step": 4577
    },
    {
      "epoch": 0.9156,
      "grad_norm": 1.257973551750183,
      "learning_rate": 2.164397422058473e-06,
      "loss": 0.1971,
      "step": 4578
    },
    {
      "epoch": 0.9156,
      "loss_align": 0.07870465517044067,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005101201590150595,
      "neg_sim_mean": 0.31420958042144775,
      "pos_sim_mean": 0.9212953448295593,
      "step": 4578
    },
    {
      "epoch": 0.9158,
      "grad_norm": 0.47641611099243164,
      "learning_rate": 2.154250001422431e-06,
      "loss": 0.0838,
      "step": 4579
    },
    {
      "epoch": 0.9158,
      "loss_align": 0.07970893383026123,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19478653371334076,
      "neg_sim_mean": 0.8131222724914551,
      "pos_sim_mean": 0.9202910661697388,
      "step": 4579
    },
    {
      "epoch": 0.916,
      "grad_norm": 1.8184856176376343,
      "learning_rate": 2.1441259002236924e-06,
      "loss": 0.2745,
      "step": 4580
    },
    {
      "epoch": 0.916,
      "loss_align": 0.10592883825302124,
      "loss_contrastive": 0.0,
      "loss_gen": 0.010565079748630524,
      "neg_sim_mean": 0.3406815826892853,
      "pos_sim_mean": 0.8940711617469788,
      "step": 4580
    },
    {
      "epoch": 0.9162,
      "grad_norm": 0.7503231167793274,
      "learning_rate": 2.134025123396638e-06,
      "loss": 0.1165,
      "step": 4581
    },
    {
      "epoch": 0.9162,
      "loss_align": 0.10625618696212769,
      "loss_contrastive": 0.09958404302597046,
      "loss_gen": 0.43865349888801575,
      "neg_sim_mean": 0.8933278322219849,
      "pos_sim_mean": 0.8937438130378723,
      "step": 4581
    },
    {
      "epoch": 0.9164,
      "grad_norm": 2.60475754737854,
      "learning_rate": 2.123947675864252e-06,
      "loss": 0.5569,
      "step": 4582
    },
    {
      "epoch": 0.9164,
      "loss_align": 0.1350550651550293,
      "loss_contrastive": 0.04832804203033447,
      "loss_gen": 0.24134284257888794,
      "neg_sim_mean": 0.8132729530334473,
      "pos_sim_mean": 0.8649449348449707,
      "step": 4582
    },
    {
      "epoch": 0.9166,
      "grad_norm": 2.1011500358581543,
      "learning_rate": 2.113893562538166e-06,
      "loss": 0.3822,
      "step": 4583
    },
    {
      "epoch": 0.9166,
      "loss_align": 0.11041134595870972,
      "loss_contrastive": 0.057983219623565674,
      "loss_gen": 0.19162696599960327,
      "neg_sim_mean": 0.847571849822998,
      "pos_sim_mean": 0.8895886540412903,
      "step": 4583
    },
    {
      "epoch": 0.9168,
      "grad_norm": 1.4637619256973267,
      "learning_rate": 2.103862788318628e-06,
      "loss": 0.309,
      "step": 4584
    },
    {
      "epoch": 0.9168,
      "loss_align": 0.1513615846633911,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2675165832042694,
      "neg_sim_mean": 0.6784840822219849,
      "pos_sim_mean": 0.8486384153366089,
      "step": 4584
    },
    {
      "epoch": 0.917,
      "grad_norm": 2.252406358718872,
      "learning_rate": 2.093855358094521e-06,
      "loss": 0.4189,
      "step": 4585
    },
    {
      "epoch": 0.917,
      "loss_align": 0.10194802284240723,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002347705652937293,
      "neg_sim_mean": 0.4594554901123047,
      "pos_sim_mean": 0.8980519771575928,
      "step": 4585
    },
    {
      "epoch": 0.9172,
      "grad_norm": 0.43128564953804016,
      "learning_rate": 2.0838712767433375e-06,
      "loss": 0.1043,
      "step": 4586
    },
    {
      "epoch": 0.9172,
      "loss_align": 0.06552666425704956,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11986688524484634,
      "neg_sim_mean": 0.8326080441474915,
      "pos_sim_mean": 0.9344733357429504,
      "step": 4586
    },
    {
      "epoch": 0.9174,
      "grad_norm": 1.261297583580017,
      "learning_rate": 2.0739105491312027e-06,
      "loss": 0.1854,
      "step": 4587
    },
    {
      "epoch": 0.9174,
      "loss_align": 0.09045130014419556,
      "loss_contrastive": 0.028014004230499268,
      "loss_gen": 0.19452637434005737,
      "neg_sim_mean": 0.8375626802444458,
      "pos_sim_mean": 0.9095486998558044,
      "step": 4587
    },
    {
      "epoch": 0.9176,
      "grad_norm": 1.6227179765701294,
      "learning_rate": 2.0639731801128605e-06,
      "loss": 0.2883,
      "step": 4588
    },
    {
      "epoch": 0.9176,
      "loss_align": 0.050435543060302734,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09619534015655518,
      "neg_sim_mean": 0.6855653524398804,
      "pos_sim_mean": 0.9495644569396973,
      "step": 4588
    },
    {
      "epoch": 0.9178,
      "grad_norm": 1.223272442817688,
      "learning_rate": 2.054059174531653e-06,
      "loss": 0.1466,
      "step": 4589
    },
    {
      "epoch": 0.9178,
      "loss_align": 0.09943532943725586,
      "loss_contrastive": 0.013204693794250488,
      "loss_gen": 0.24479764699935913,
      "neg_sim_mean": 0.8137693405151367,
      "pos_sim_mean": 0.9005646705627441,
      "step": 4589
    },
    {
      "epoch": 0.918,
      "grad_norm": 2.1365532875061035,
      "learning_rate": 2.0441685372195484e-06,
      "loss": 0.3458,
      "step": 4590
    },
    {
      "epoch": 0.918,
      "loss_align": 0.10677981376647949,
      "loss_contrastive": 0.11483430862426758,
      "loss_gen": 0.18372510373592377,
      "neg_sim_mean": 0.9080544710159302,
      "pos_sim_mean": 0.8932201862335205,
      "step": 4590
    },
    {
      "epoch": 0.9182,
      "grad_norm": 1.4569731950759888,
      "learning_rate": 2.0343012729971243e-06,
      "loss": 0.3043,
      "step": 4591
    },
    {
      "epoch": 0.9182,
      "loss_align": 0.2325853705406189,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004319286439567804,
      "neg_sim_mean": 0.29619860649108887,
      "pos_sim_mean": 0.7674146294593811,
      "step": 4591
    },
    {
      "epoch": 0.9184,
      "grad_norm": 0.7708874344825745,
      "learning_rate": 2.024457386673567e-06,
      "loss": 0.2369,
      "step": 4592
    },
    {
      "epoch": 0.9184,
      "loss_align": 0.07702887058258057,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006926549132913351,
      "neg_sim_mean": 0.31956201791763306,
      "pos_sim_mean": 0.9229711294174194,
      "step": 4592
    },
    {
      "epoch": 0.9186,
      "grad_norm": 0.5174062252044678,
      "learning_rate": 2.0146368830466667e-06,
      "loss": 0.084,
      "step": 4593
    },
    {
      "epoch": 0.9186,
      "loss_align": 0.09588193893432617,
      "loss_contrastive": 0.05656653642654419,
      "loss_gen": 0.14808787405490875,
      "neg_sim_mean": 0.8606845736503601,
      "pos_sim_mean": 0.9041180610656738,
      "step": 4593
    },
    {
      "epoch": 0.9188,
      "grad_norm": 1.6092641353607178,
      "learning_rate": 2.0048397669028163e-06,
      "loss": 0.2508,
      "step": 4594
    },
    {
      "epoch": 0.9188,
      "loss_align": 0.11036086082458496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17662213742733002,
      "neg_sim_mean": 0.43490976095199585,
      "pos_sim_mean": 0.889639139175415,
      "step": 4594
    },
    {
      "epoch": 0.919,
      "grad_norm": 1.7526354789733887,
      "learning_rate": 1.995066043017013e-06,
      "loss": 0.287,
      "step": 4595
    },
    {
      "epoch": 0.919,
      "loss_align": 0.10055792331695557,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09941133856773376,
      "neg_sim_mean": 0.7556731700897217,
      "pos_sim_mean": 0.8994420766830444,
      "step": 4595
    },
    {
      "epoch": 0.9192,
      "grad_norm": 1.2524477243423462,
      "learning_rate": 1.985315716152847e-06,
      "loss": 0.2,
      "step": 4596
    },
    {
      "epoch": 0.9192,
      "loss_align": 0.09351217746734619,
      "loss_contrastive": 0.06887143850326538,
      "loss_gen": 0.2208334505558014,
      "neg_sim_mean": 0.8753592371940613,
      "pos_sim_mean": 0.9064878225326538,
      "step": 4596
    },
    {
      "epoch": 0.9194,
      "grad_norm": 1.7035200595855713,
      "learning_rate": 1.9755887910625105e-06,
      "loss": 0.3226,
      "step": 4597
    },
    {
      "epoch": 0.9194,
      "loss_align": 0.12152612209320068,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0055665019899606705,
      "neg_sim_mean": 0.34807467460632324,
      "pos_sim_mean": 0.8784738779067993,
      "step": 4597
    },
    {
      "epoch": 0.9196,
      "grad_norm": 0.5250226259231567,
      "learning_rate": 1.9658852724868005e-06,
      "loss": 0.1271,
      "step": 4598
    },
    {
      "epoch": 0.9196,
      "loss_align": 0.126889169216156,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2872171401977539,
      "neg_sim_mean": 0.6034026741981506,
      "pos_sim_mean": 0.873110830783844,
      "step": 4598
    },
    {
      "epoch": 0.9198,
      "grad_norm": 1.9979175329208374,
      "learning_rate": 1.956205165155078e-06,
      "loss": 0.4141,
      "step": 4599
    },
    {
      "epoch": 0.9198,
      "loss_align": 0.07617634534835815,
      "loss_contrastive": 0.0,
      "loss_gen": 0.021176772192120552,
      "neg_sim_mean": 0.3639066815376282,
      "pos_sim_mean": 0.9238236546516418,
      "step": 4599
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.8708427548408508,
      "learning_rate": 1.946548473785309e-06,
      "loss": 0.0974,
      "step": 4600
    },
    {
      "epoch": 0.92,
      "loss_align": 0.12592828273773193,
      "loss_contrastive": 0.016688823699951172,
      "loss_gen": 0.2037210911512375,
      "neg_sim_mean": 0.7907605171203613,
      "pos_sim_mean": 0.8740717172622681,
      "step": 4600
    },
    {
      "epoch": 0.9202,
      "grad_norm": 1.8668887615203857,
      "learning_rate": 1.9369152030840556e-06,
      "loss": 0.3317,
      "step": 4601
    },
    {
      "epoch": 0.9202,
      "loss_align": 0.05766063928604126,
      "loss_contrastive": 0.06742620468139648,
      "loss_gen": 0.045004814863204956,
      "neg_sim_mean": 0.9097655415534973,
      "pos_sim_mean": 0.9423393607139587,
      "step": 4601
    },
    {
      "epoch": 0.9204,
      "grad_norm": 0.7796604037284851,
      "learning_rate": 1.9273053577464618e-06,
      "loss": 0.1108,
      "step": 4602
    },
    {
      "epoch": 0.9204,
      "loss_align": 0.11087846755981445,
      "loss_contrastive": 0.07023382186889648,
      "loss_gen": 0.2711620628833771,
      "neg_sim_mean": 0.8593553304672241,
      "pos_sim_mean": 0.8891215324401855,
      "step": 4602
    },
    {
      "epoch": 0.9206,
      "grad_norm": 2.018441915512085,
      "learning_rate": 1.917718942456237e-06,
      "loss": 0.3905,
      "step": 4603
    },
    {
      "epoch": 0.9206,
      "loss_align": 0.14506971836090088,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09609062969684601,
      "neg_sim_mean": 0.7444525957107544,
      "pos_sim_mean": 0.8549302816390991,
      "step": 4603
    },
    {
      "epoch": 0.9208,
      "grad_norm": 1.356497049331665,
      "learning_rate": 1.9081559618856937e-06,
      "loss": 0.2412,
      "step": 4604
    },
    {
      "epoch": 0.9208,
      "loss_align": 0.061316609382629395,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11357472091913223,
      "neg_sim_mean": 0.7747585773468018,
      "pos_sim_mean": 0.9386833906173706,
      "step": 4604
    },
    {
      "epoch": 0.921,
      "grad_norm": 1.2781405448913574,
      "learning_rate": 1.8986164206957035e-06,
      "loss": 0.1749,
      "step": 4605
    },
    {
      "epoch": 0.921,
      "loss_align": 0.06146878004074097,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16913285851478577,
      "neg_sim_mean": 0.5257772207260132,
      "pos_sim_mean": 0.938531219959259,
      "step": 4605
    },
    {
      "epoch": 0.9212,
      "grad_norm": 1.6189849376678467,
      "learning_rate": 1.8891003235357308e-06,
      "loss": 0.2306,
      "step": 4606
    },
    {
      "epoch": 0.9212,
      "loss_align": 0.04492461681365967,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17084437608718872,
      "neg_sim_mean": 0.791081964969635,
      "pos_sim_mean": 0.9550753831863403,
      "step": 4606
    },
    {
      "epoch": 0.9214,
      "grad_norm": 1.5828436613082886,
      "learning_rate": 1.8796076750438096e-06,
      "loss": 0.2158,
      "step": 4607
    },
    {
      "epoch": 0.9214,
      "loss_align": 0.10700523853302002,
      "loss_contrastive": 0.04510849714279175,
      "loss_gen": 0.4161803722381592,
      "neg_sim_mean": 0.8381032347679138,
      "pos_sim_mean": 0.89299476146698,
      "step": 4607
    },
    {
      "epoch": 0.9216,
      "grad_norm": 2.417524814605713,
      "learning_rate": 1.8701384798465281e-06,
      "loss": 0.5286,
      "step": 4608
    },
    {
      "epoch": 0.9216,
      "loss_align": 0.055527329444885254,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13598164916038513,
      "neg_sim_mean": 0.5305452346801758,
      "pos_sim_mean": 0.9444726705551147,
      "step": 4608
    },
    {
      "epoch": 0.9218,
      "grad_norm": 1.3106704950332642,
      "learning_rate": 1.8606927425590614e-06,
      "loss": 0.1915,
      "step": 4609
    },
    {
      "epoch": 0.9218,
      "loss_align": 0.06662344932556152,
      "loss_contrastive": 0.07274651527404785,
      "loss_gen": 0.24758489429950714,
      "neg_sim_mean": 0.9061230421066284,
      "pos_sim_mean": 0.9333765506744385,
      "step": 4609
    },
    {
      "epoch": 0.922,
      "grad_norm": 2.235044002532959,
      "learning_rate": 1.8512704677851488e-06,
      "loss": 0.3229,
      "step": 4610
    },
    {
      "epoch": 0.922,
      "loss_align": 0.09048062562942505,
      "loss_contrastive": 0.07022404670715332,
      "loss_gen": 0.13962748646736145,
      "neg_sim_mean": 0.8797433972358704,
      "pos_sim_mean": 0.909519374370575,
      "step": 4610
    },
    {
      "epoch": 0.9222,
      "grad_norm": 1.7711917161941528,
      "learning_rate": 1.841871660117095e-06,
      "loss": 0.2385,
      "step": 4611
    },
    {
      "epoch": 0.9222,
      "loss_align": 0.08265280723571777,
      "loss_contrastive": 0.061635375022888184,
      "loss_gen": 0.3274856209754944,
      "neg_sim_mean": 0.8789825439453125,
      "pos_sim_mean": 0.9173471927642822,
      "step": 4611
    },
    {
      "epoch": 0.9224,
      "grad_norm": 2.2649893760681152,
      "learning_rate": 1.832496324135763e-06,
      "loss": 0.4175,
      "step": 4612
    },
    {
      "epoch": 0.9224,
      "loss_align": 0.12423509359359741,
      "loss_contrastive": 0.08338373899459839,
      "loss_gen": 0.252375990152359,
      "neg_sim_mean": 0.8591486215591431,
      "pos_sim_mean": 0.8757649064064026,
      "step": 4612
    },
    {
      "epoch": 0.9226,
      "grad_norm": 2.0677998065948486,
      "learning_rate": 1.8231444644105756e-06,
      "loss": 0.3866,
      "step": 4613
    },
    {
      "epoch": 0.9226,
      "loss_align": 0.13575756549835205,
      "loss_contrastive": 0.024812400341033936,
      "loss_gen": 0.15152646601200104,
      "neg_sim_mean": 0.789054811000824,
      "pos_sim_mean": 0.864242434501648,
      "step": 4613
    },
    {
      "epoch": 0.9228,
      "grad_norm": 1.4894070625305176,
      "learning_rate": 1.8138160854995145e-06,
      "loss": 0.2903,
      "step": 4614
    },
    {
      "epoch": 0.9228,
      "loss_align": 0.07223290205001831,
      "loss_contrastive": 0.059763550758361816,
      "loss_gen": 0.2715745270252228,
      "neg_sim_mean": 0.8875306248664856,
      "pos_sim_mean": 0.9277670979499817,
      "step": 4614
    },
    {
      "epoch": 0.923,
      "grad_norm": 1.4636518955230713,
      "learning_rate": 1.8045111919491208e-06,
      "loss": 0.351,
      "step": 4615
    },
    {
      "epoch": 0.923,
      "loss_align": 0.05251502990722656,
      "loss_contrastive": 0.04626345634460449,
      "loss_gen": 0.24637682735919952,
      "neg_sim_mean": 0.89374840259552,
      "pos_sim_mean": 0.9474849700927734,
      "step": 4615
    },
    {
      "epoch": 0.9232,
      "grad_norm": 1.7530630826950073,
      "learning_rate": 1.7952297882945003e-06,
      "loss": 0.3044,
      "step": 4616
    },
    {
      "epoch": 0.9232,
      "loss_align": 0.0780329704284668,
      "loss_contrastive": 0.06385713815689087,
      "loss_gen": 0.21130803227424622,
      "neg_sim_mean": 0.8858241438865662,
      "pos_sim_mean": 0.9219670295715332,
      "step": 4616
    },
    {
      "epoch": 0.9234,
      "grad_norm": 1.4219142198562622,
      "learning_rate": 1.7859718790592727e-06,
      "loss": 0.297,
      "step": 4617
    },
    {
      "epoch": 0.9234,
      "loss_align": 0.14697718620300293,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10542219132184982,
      "neg_sim_mean": 0.532501757144928,
      "pos_sim_mean": 0.8530228137969971,
      "step": 4617
    },
    {
      "epoch": 0.9236,
      "grad_norm": 1.4291106462478638,
      "learning_rate": 1.7767374687556405e-06,
      "loss": 0.2524,
      "step": 4618
    },
    {
      "epoch": 0.9236,
      "loss_align": 0.136580228805542,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21103954315185547,
      "neg_sim_mean": 0.7492728233337402,
      "pos_sim_mean": 0.863419771194458,
      "step": 4618
    },
    {
      "epoch": 0.9238,
      "grad_norm": 2.225430488586426,
      "learning_rate": 1.7675265618843362e-06,
      "loss": 0.3476,
      "step": 4619
    },
    {
      "epoch": 0.9238,
      "loss_align": 0.09422892332077026,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29803702235221863,
      "neg_sim_mean": 0.7747371792793274,
      "pos_sim_mean": 0.9057710766792297,
      "step": 4619
    },
    {
      "epoch": 0.924,
      "grad_norm": 1.9010077714920044,
      "learning_rate": 1.758339162934658e-06,
      "loss": 0.3923,
      "step": 4620
    },
    {
      "epoch": 0.924,
      "loss_align": 0.07039976119995117,
      "loss_contrastive": 0.07483184337615967,
      "loss_gen": 0.1629018634557724,
      "neg_sim_mean": 0.9044320583343506,
      "pos_sim_mean": 0.9296002388000488,
      "step": 4620
    },
    {
      "epoch": 0.9242,
      "grad_norm": 1.638246774673462,
      "learning_rate": 1.7491752763844293e-06,
      "loss": 0.2423,
      "step": 4621
    },
    {
      "epoch": 0.9242,
      "loss_align": 0.15704095363616943,
      "loss_contrastive": 0.035859882831573486,
      "loss_gen": 0.1788739711046219,
      "neg_sim_mean": 0.7788189053535461,
      "pos_sim_mean": 0.8429590463638306,
      "step": 4621
    },
    {
      "epoch": 0.9244,
      "grad_norm": 1.4690606594085693,
      "learning_rate": 1.740034906700011e-06,
      "loss": 0.3402,
      "step": 4622
    },
    {
      "epoch": 0.9244,
      "loss_align": 0.07948464155197144,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23715077340602875,
      "neg_sim_mean": 0.6798976063728333,
      "pos_sim_mean": 0.9205153584480286,
      "step": 4622
    },
    {
      "epoch": 0.9246,
      "grad_norm": 1.514579176902771,
      "learning_rate": 1.730918058336306e-06,
      "loss": 0.3166,
      "step": 4623
    },
    {
      "epoch": 0.9246,
      "loss_align": 0.08106529712677002,
      "loss_contrastive": 0.05137866735458374,
      "loss_gen": 0.3838323652744293,
      "neg_sim_mean": 0.8703133463859558,
      "pos_sim_mean": 0.91893470287323,
      "step": 4623
    },
    {
      "epoch": 0.9248,
      "grad_norm": 1.962743878364563,
      "learning_rate": 1.7218247357367656e-06,
      "loss": 0.4711,
      "step": 4624
    },
    {
      "epoch": 0.9248,
      "loss_align": 0.11707788705825806,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24817225337028503,
      "neg_sim_mean": 0.6780103445053101,
      "pos_sim_mean": 0.8829221129417419,
      "step": 4624
    },
    {
      "epoch": 0.925,
      "grad_norm": 1.6958268880844116,
      "learning_rate": 1.7127549433333556e-06,
      "loss": 0.3653,
      "step": 4625
    },
    {
      "epoch": 0.925,
      "loss_align": 0.12087726593017578,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2621965706348419,
      "neg_sim_mean": 0.738796591758728,
      "pos_sim_mean": 0.8791227340698242,
      "step": 4625
    },
    {
      "epoch": 0.9252,
      "grad_norm": 1.9990136623382568,
      "learning_rate": 1.70370868554659e-06,
      "loss": 0.3831,
      "step": 4626
    },
    {
      "epoch": 0.9252,
      "loss_align": 0.23964518308639526,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0013758308487012982,
      "neg_sim_mean": 0.4968959391117096,
      "pos_sim_mean": 0.7603548169136047,
      "step": 4626
    },
    {
      "epoch": 0.9254,
      "grad_norm": 0.9139307141304016,
      "learning_rate": 1.6946859667854975e-06,
      "loss": 0.241,
      "step": 4627
    },
    {
      "epoch": 0.9254,
      "loss_align": 0.16214728355407715,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00885141920298338,
      "neg_sim_mean": 0.23643290996551514,
      "pos_sim_mean": 0.8378527164459229,
      "step": 4627
    },
    {
      "epoch": 0.9256,
      "grad_norm": 0.6565831303596497,
      "learning_rate": 1.6856867914476494e-06,
      "loss": 0.171,
      "step": 4628
    },
    {
      "epoch": 0.9256,
      "loss_align": 0.08527302742004395,
      "loss_contrastive": 0.05002540349960327,
      "loss_gen": 0.0712241530418396,
      "neg_sim_mean": 0.8647523522377014,
      "pos_sim_mean": 0.914726972579956,
      "step": 4628
    },
    {
      "epoch": 0.9258,
      "grad_norm": 1.147141933441162,
      "learning_rate": 1.6767111639191202e-06,
      "loss": 0.1625,
      "step": 4629
    },
    {
      "epoch": 0.9258,
      "loss_align": 0.2306954264640808,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2972947657108307,
      "neg_sim_mean": 0.666964054107666,
      "pos_sim_mean": 0.7693045735359192,
      "step": 4629
    },
    {
      "epoch": 0.926,
      "grad_norm": 2.3389155864715576,
      "learning_rate": 1.6677590885745386e-06,
      "loss": 0.528,
      "step": 4630
    },
    {
      "epoch": 0.926,
      "loss_align": 0.12135350704193115,
      "loss_contrastive": 0.045606374740600586,
      "loss_gen": 0.2914930284023285,
      "neg_sim_mean": 0.8242528438568115,
      "pos_sim_mean": 0.8786464929580688,
      "step": 4630
    },
    {
      "epoch": 0.9262,
      "grad_norm": 1.952634334564209,
      "learning_rate": 1.6588305697770313e-06,
      "loss": 0.4183,
      "step": 4631
    },
    {
      "epoch": 0.9262,
      "loss_align": 0.08442014455795288,
      "loss_contrastive": 0.044695913791656494,
      "loss_gen": 0.36593717336654663,
      "neg_sim_mean": 0.8602757453918457,
      "pos_sim_mean": 0.9155798554420471,
      "step": 4631
    },
    {
      "epoch": 0.9264,
      "grad_norm": 2.258843421936035,
      "learning_rate": 1.6499256118782503e-06,
      "loss": 0.4557,
      "step": 4632
    },
    {
      "epoch": 0.9264,
      "loss_align": 0.12631946802139282,
      "loss_contrastive": 0.033016204833984375,
      "loss_gen": 0.13987323641777039,
      "neg_sim_mean": 0.8066967129707336,
      "pos_sim_mean": 0.8736805319786072,
      "step": 4632
    },
    {
      "epoch": 0.9266,
      "grad_norm": 1.7623430490493774,
      "learning_rate": 1.6410442192183573e-06,
      "loss": 0.2702,
      "step": 4633
    },
    {
      "epoch": 0.9266,
      "loss_align": 0.101437509059906,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006927459500730038,
      "neg_sim_mean": 0.3094792366027832,
      "pos_sim_mean": 0.898562490940094,
      "step": 4633
    },
    {
      "epoch": 0.9268,
      "grad_norm": 0.5922091603279114,
      "learning_rate": 1.6321863961260453e-06,
      "loss": 0.1084,
      "step": 4634
    },
    {
      "epoch": 0.9268,
      "loss_align": 0.08488786220550537,
      "loss_contrastive": 0.036202430725097656,
      "loss_gen": 0.2281721830368042,
      "neg_sim_mean": 0.8513145446777344,
      "pos_sim_mean": 0.9151121377944946,
      "step": 4634
    },
    {
      "epoch": 0.927,
      "grad_norm": 1.8729748725891113,
      "learning_rate": 1.6233521469185054e-06,
      "loss": 0.3174,
      "step": 4635
    },
    {
      "epoch": 0.927,
      "loss_align": 0.20294028520584106,
      "loss_contrastive": 0.05363941192626953,
      "loss_gen": 0.31538644433021545,
      "neg_sim_mean": 0.7506991028785706,
      "pos_sim_mean": 0.7970597147941589,
      "step": 4635
    },
    {
      "epoch": 0.9272,
      "grad_norm": 2.3498594760894775,
      "learning_rate": 1.6145414759014431e-06,
      "loss": 0.5248,
      "step": 4636
    },
    {
      "epoch": 0.9272,
      "loss_align": 0.09052121639251709,
      "loss_contrastive": 0.07447266578674316,
      "loss_gen": 0.47997406125068665,
      "neg_sim_mean": 0.8839514255523682,
      "pos_sim_mean": 0.9094787836074829,
      "step": 4636
    },
    {
      "epoch": 0.9274,
      "grad_norm": 2.4967970848083496,
      "learning_rate": 1.6057543873690683e-06,
      "loss": 0.5794,
      "step": 4637
    },
    {
      "epoch": 0.9274,
      "loss_align": 0.07116210460662842,
      "loss_contrastive": 0.10115337371826172,
      "loss_gen": 0.3373955190181732,
      "neg_sim_mean": 0.9299912452697754,
      "pos_sim_mean": 0.9288378953933716,
      "step": 4637
    },
    {
      "epoch": 0.9276,
      "grad_norm": 2.4622762203216553,
      "learning_rate": 1.596990885604105e-06,
      "loss": 0.4207,
      "step": 4638
    },
    {
      "epoch": 0.9276,
      "loss_align": 0.10181248188018799,
      "loss_contrastive": 0.0,
      "loss_gen": 0.004340982995927334,
      "neg_sim_mean": 0.345611572265625,
      "pos_sim_mean": 0.898187518119812,
      "step": 4638
    },
    {
      "epoch": 0.9278,
      "grad_norm": 0.5412060618400574,
      "learning_rate": 1.5882509748777808e-06,
      "loss": 0.1062,
      "step": 4639
    },
    {
      "epoch": 0.9278,
      "loss_align": 0.07822787761688232,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14134372770786285,
      "neg_sim_mean": 0.78399658203125,
      "pos_sim_mean": 0.9217721223831177,
      "step": 4639
    },
    {
      "epoch": 0.928,
      "grad_norm": 1.5652693510055542,
      "learning_rate": 1.5795346594498161e-06,
      "loss": 0.2196,
      "step": 4640
    },
    {
      "epoch": 0.928,
      "loss_align": 0.13356143236160278,
      "loss_contrastive": 0.06621009111404419,
      "loss_gen": 0.07483147829771042,
      "neg_sim_mean": 0.8326486349105835,
      "pos_sim_mean": 0.8664385676383972,
      "step": 4640
    },
    {
      "epoch": 0.9282,
      "grad_norm": 1.302243709564209,
      "learning_rate": 1.5708419435684462e-06,
      "loss": 0.2163,
      "step": 4641
    },
    {
      "epoch": 0.9282,
      "loss_align": 0.06297683715820312,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3472985029220581,
      "neg_sim_mean": 0.7850130200386047,
      "pos_sim_mean": 0.9370231628417969,
      "step": 4641
    },
    {
      "epoch": 0.9284,
      "grad_norm": 2.147474765777588,
      "learning_rate": 1.562172831470382e-06,
      "loss": 0.4103,
      "step": 4642
    },
    {
      "epoch": 0.9284,
      "loss_align": 0.05017268657684326,
      "loss_contrastive": 0.06560587882995605,
      "loss_gen": 0.24112510681152344,
      "neg_sim_mean": 0.9154331684112549,
      "pos_sim_mean": 0.9498273134231567,
      "step": 4642
    },
    {
      "epoch": 0.9286,
      "grad_norm": 1.8915728330612183,
      "learning_rate": 1.5535273273808548e-06,
      "loss": 0.2992,
      "step": 4643
    },
    {
      "epoch": 0.9286,
      "loss_align": 0.09119582176208496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.023610806092619896,
      "neg_sim_mean": 0.29651573300361633,
      "pos_sim_mean": 0.908804178237915,
      "step": 4643
    },
    {
      "epoch": 0.9288,
      "grad_norm": 0.8926008343696594,
      "learning_rate": 1.5449054355135717e-06,
      "loss": 0.1148,
      "step": 4644
    },
    {
      "epoch": 0.9288,
      "loss_align": 0.10771715641021729,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23385313153266907,
      "neg_sim_mean": 0.7525129318237305,
      "pos_sim_mean": 0.8922828435897827,
      "step": 4644
    },
    {
      "epoch": 0.929,
      "grad_norm": 1.786754846572876,
      "learning_rate": 1.5363071600707435e-06,
      "loss": 0.3416,
      "step": 4645
    },
    {
      "epoch": 0.929,
      "loss_align": 0.049207210540771484,
      "loss_contrastive": 0.021230697631835938,
      "loss_gen": 0.19858390092849731,
      "neg_sim_mean": 0.8720234632492065,
      "pos_sim_mean": 0.9507927894592285,
      "step": 4645
    },
    {
      "epoch": 0.9292,
      "grad_norm": 1.7388204336166382,
      "learning_rate": 1.5277325052430568e-06,
      "loss": 0.2503,
      "step": 4646
    },
    {
      "epoch": 0.9292,
      "loss_align": 0.16539275646209717,
      "loss_contrastive": 0.1688079833984375,
      "loss_gen": 0.15771113336086273,
      "neg_sim_mean": 0.9034152030944824,
      "pos_sim_mean": 0.8346072435379028,
      "step": 4646
    },
    {
      "epoch": 0.9294,
      "grad_norm": 1.857401728630066,
      "learning_rate": 1.5191814752097023e-06,
      "loss": 0.3434,
      "step": 4647
    },
    {
      "epoch": 0.9294,
      "loss_align": 0.0821843147277832,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0018241681391373277,
      "neg_sim_mean": 0.48936665058135986,
      "pos_sim_mean": 0.9178156852722168,
      "step": 4647
    },
    {
      "epoch": 0.9296,
      "grad_norm": 0.3921898603439331,
      "learning_rate": 1.5106540741383402e-06,
      "loss": 0.084,
      "step": 4648
    },
    {
      "epoch": 0.9296,
      "loss_align": 0.1248471736907959,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012660416774451733,
      "neg_sim_mean": 0.42296767234802246,
      "pos_sim_mean": 0.8751528263092041,
      "step": 4648
    },
    {
      "epoch": 0.9298,
      "grad_norm": 0.8281778693199158,
      "learning_rate": 1.5021503061851349e-06,
      "loss": 0.1375,
      "step": 4649
    },
    {
      "epoch": 0.9298,
      "loss_align": 0.11005115509033203,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008382271975278854,
      "neg_sim_mean": 0.3274696469306946,
      "pos_sim_mean": 0.889948844909668,
      "step": 4649
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.6258558630943298,
      "learning_rate": 1.4936701754947101e-06,
      "loss": 0.1184,
      "step": 4650
    },
    {
      "epoch": 0.93,
      "loss_align": 0.13875877857208252,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007846701890230179,
      "neg_sim_mean": 0.3122328519821167,
      "pos_sim_mean": 0.8612412214279175,
      "step": 4650
    },
    {
      "epoch": 0.9302,
      "grad_norm": 0.6875578165054321,
      "learning_rate": 1.4852136862001764e-06,
      "loss": 0.1466,
      "step": 4651
    },
    {
      "epoch": 0.9302,
      "loss_align": 0.1549246907234192,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01808711513876915,
      "neg_sim_mean": 0.3968774080276489,
      "pos_sim_mean": 0.8450753092765808,
      "step": 4651
    },
    {
      "epoch": 0.9304,
      "grad_norm": 1.0065803527832031,
      "learning_rate": 1.476780842423131e-06,
      "loss": 0.173,
      "step": 4652
    },
    {
      "epoch": 0.9304,
      "loss_align": 0.08465498685836792,
      "loss_contrastive": 0.056124746799468994,
      "loss_gen": 0.23936690390110016,
      "neg_sim_mean": 0.8714697360992432,
      "pos_sim_mean": 0.9153450131416321,
      "step": 4652
    },
    {
      "epoch": 0.9306,
      "grad_norm": 1.923607587814331,
      "learning_rate": 1.4683716482736366e-06,
      "loss": 0.3308,
      "step": 4653
    },
    {
      "epoch": 0.9306,
      "loss_align": 0.10072582960128784,
      "loss_contrastive": 0.06559246778488159,
      "loss_gen": 0.2091739922761917,
      "neg_sim_mean": 0.8648666143417358,
      "pos_sim_mean": 0.8992741703987122,
      "step": 4653
    },
    {
      "epoch": 0.9308,
      "grad_norm": 1.8322315216064453,
      "learning_rate": 1.4599861078502308e-06,
      "loss": 0.3178,
      "step": 4654
    },
    {
      "epoch": 0.9308,
      "loss_align": 0.06740128993988037,
      "loss_contrastive": 0.012897253036499023,
      "loss_gen": 0.10966002941131592,
      "neg_sim_mean": 0.8454959392547607,
      "pos_sim_mean": 0.9325987100601196,
      "step": 4654
    },
    {
      "epoch": 0.931,
      "grad_norm": 1.219822883605957,
      "learning_rate": 1.4516242252399226e-06,
      "loss": 0.1786,
      "step": 4655
    },
    {
      "epoch": 0.931,
      "loss_align": 0.09590369462966919,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2050393968820572,
      "neg_sim_mean": 0.5911041498184204,
      "pos_sim_mean": 0.9040963053703308,
      "step": 4655
    },
    {
      "epoch": 0.9312,
      "grad_norm": 1.7025551795959473,
      "learning_rate": 1.4432860045182017e-06,
      "loss": 0.3009,
      "step": 4656
    },
    {
      "epoch": 0.9312,
      "loss_align": 0.20153456926345825,
      "loss_contrastive": 0.03471660614013672,
      "loss_gen": 0.26460033655166626,
      "neg_sim_mean": 0.7331820130348206,
      "pos_sim_mean": 0.7984654307365417,
      "step": 4656
    },
    {
      "epoch": 0.9314,
      "grad_norm": 1.9112497568130493,
      "learning_rate": 1.4349714497490009e-06,
      "loss": 0.4703,
      "step": 4657
    },
    {
      "epoch": 0.9314,
      "loss_align": 0.17375832796096802,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15984004735946655,
      "neg_sim_mean": 0.5038558840751648,
      "pos_sim_mean": 0.826241672039032,
      "step": 4657
    },
    {
      "epoch": 0.9316,
      "grad_norm": 1.5249583721160889,
      "learning_rate": 1.4266805649847392e-06,
      "loss": 0.3336,
      "step": 4658
    },
    {
      "epoch": 0.9316,
      "loss_align": 0.13562601804733276,
      "loss_contrastive": 0.1216476559638977,
      "loss_gen": 0.07822413742542267,
      "neg_sim_mean": 0.886021614074707,
      "pos_sim_mean": 0.8643739819526672,
      "step": 4658
    },
    {
      "epoch": 0.9318,
      "grad_norm": 1.1070791482925415,
      "learning_rate": 1.4184133542663014e-06,
      "loss": 0.2284,
      "step": 4659
    },
    {
      "epoch": 0.9318,
      "loss_align": 0.07247436046600342,
      "loss_contrastive": 0.06920802593231201,
      "loss_gen": 0.2931874692440033,
      "neg_sim_mean": 0.8967336416244507,
      "pos_sim_mean": 0.9275256395339966,
      "step": 4659
    },
    {
      "epoch": 0.932,
      "grad_norm": 2.030592441558838,
      "learning_rate": 1.4101698216230253e-06,
      "loss": 0.374,
      "step": 4660
    },
    {
      "epoch": 0.932,
      "loss_align": 0.13325506448745728,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19121776521205902,
      "neg_sim_mean": 0.3873211145401001,
      "pos_sim_mean": 0.8667449355125427,
      "step": 4660
    },
    {
      "epoch": 0.9322,
      "grad_norm": 1.6315997838974,
      "learning_rate": 1.4019499710726913e-06,
      "loss": 0.3245,
      "step": 4661
    },
    {
      "epoch": 0.9322,
      "loss_align": 0.05696231126785278,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19288252294063568,
      "neg_sim_mean": 0.37828677892684937,
      "pos_sim_mean": 0.9430376887321472,
      "step": 4661
    },
    {
      "epoch": 0.9324,
      "grad_norm": 1.467821717262268,
      "learning_rate": 1.3937538066215672e-06,
      "loss": 0.2498,
      "step": 4662
    },
    {
      "epoch": 0.9324,
      "loss_align": 0.35516107082366943,
      "loss_contrastive": 0.07821482419967651,
      "loss_gen": 0.04156442731618881,
      "neg_sim_mean": 0.6230537295341492,
      "pos_sim_mean": 0.6448389291763306,
      "step": 4662
    },
    {
      "epoch": 0.9326,
      "grad_norm": 2.2847559452056885,
      "learning_rate": 1.385581332264363e-06,
      "loss": 0.4061,
      "step": 4663
    },
    {
      "epoch": 0.9326,
      "loss_align": 0.1428769826889038,
      "loss_contrastive": 0.05394637584686279,
      "loss_gen": 0.22076739370822906,
      "neg_sim_mean": 0.8110693693161011,
      "pos_sim_mean": 0.8571230173110962,
      "step": 4663
    },
    {
      "epoch": 0.9328,
      "grad_norm": 1.7668414115905762,
      "learning_rate": 1.3774325519842423e-06,
      "loss": 0.3701,
      "step": 4664
    },
    {
      "epoch": 0.9328,
      "loss_align": 0.12943178415298462,
      "loss_contrastive": 0.0,
      "loss_gen": 0.015275358222424984,
      "neg_sim_mean": 0.32775986194610596,
      "pos_sim_mean": 0.8705682158470154,
      "step": 4664
    },
    {
      "epoch": 0.933,
      "grad_norm": 0.9123729467391968,
      "learning_rate": 1.369307469752823e-06,
      "loss": 0.1447,
      "step": 4665
    },
    {
      "epoch": 0.933,
      "loss_align": 0.18542325496673584,
      "loss_contrastive": 0.04154020547866821,
      "loss_gen": 0.3990918695926666,
      "neg_sim_mean": 0.7561169266700745,
      "pos_sim_mean": 0.8145767450332642,
      "step": 4665
    },
    {
      "epoch": 0.9332,
      "grad_norm": 2.086158275604248,
      "learning_rate": 1.3612060895301759e-06,
      "loss": 0.5895,
      "step": 4666
    },
    {
      "epoch": 0.9332,
      "loss_align": 0.10832750797271729,
      "loss_contrastive": 0.0703279972076416,
      "loss_gen": 0.20671291649341583,
      "neg_sim_mean": 0.8620004653930664,
      "pos_sim_mean": 0.8916724920272827,
      "step": 4666
    },
    {
      "epoch": 0.9334,
      "grad_norm": 1.989689588546753,
      "learning_rate": 1.3531284152647982e-06,
      "loss": 0.3235,
      "step": 4667
    },
    {
      "epoch": 0.9334,
      "loss_align": 0.06105220317840576,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2588486075401306,
      "neg_sim_mean": 0.4649081528186798,
      "pos_sim_mean": 0.9389477968215942,
      "step": 4667
    },
    {
      "epoch": 0.9336,
      "grad_norm": 2.056884288787842,
      "learning_rate": 1.3450744508936686e-06,
      "loss": 0.3199,
      "step": 4668
    },
    {
      "epoch": 0.9336,
      "loss_align": 0.1333785057067871,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3162604868412018,
      "neg_sim_mean": 0.6307858824729919,
      "pos_sim_mean": 0.8666214942932129,
      "step": 4668
    },
    {
      "epoch": 0.9338,
      "grad_norm": 1.8852834701538086,
      "learning_rate": 1.3370442003421913e-06,
      "loss": 0.4496,
      "step": 4669
    },
    {
      "epoch": 0.9338,
      "loss_align": 0.07403779029846191,
      "loss_contrastive": 0.0,
      "loss_gen": 0.012759136967360973,
      "neg_sim_mean": 0.3203873038291931,
      "pos_sim_mean": 0.9259622097015381,
      "step": 4669
    },
    {
      "epoch": 0.934,
      "grad_norm": 0.7169362902641296,
      "learning_rate": 1.3290376675242023e-06,
      "loss": 0.0868,
      "step": 4670
    },
    {
      "epoch": 0.934,
      "loss_align": 0.10228490829467773,
      "loss_contrastive": 0.07514834403991699,
      "loss_gen": 0.16582399606704712,
      "neg_sim_mean": 0.8728634119033813,
      "pos_sim_mean": 0.8977150917053223,
      "step": 4670
    },
    {
      "epoch": 0.9342,
      "grad_norm": 1.848710536956787,
      "learning_rate": 1.3210548563419856e-06,
      "loss": 0.2771,
      "step": 4671
    },
    {
      "epoch": 0.9342,
      "loss_align": 0.0502399206161499,
      "loss_contrastive": 0.10002857446670532,
      "loss_gen": 0.17523370683193207,
      "neg_sim_mean": 0.9497886300086975,
      "pos_sim_mean": 0.9497600793838501,
      "step": 4671
    },
    {
      "epoch": 0.9344,
      "grad_norm": 1.6874322891235352,
      "learning_rate": 1.3130957706862789e-06,
      "loss": 0.2375,
      "step": 4672
    },
    {
      "epoch": 0.9344,
      "loss_align": 0.09025245904922485,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2650184631347656,
      "neg_sim_mean": 0.7602255344390869,
      "pos_sim_mean": 0.9097475409507751,
      "step": 4672
    },
    {
      "epoch": 0.9346,
      "grad_norm": 1.672279715538025,
      "learning_rate": 1.3051604144362406e-06,
      "loss": 0.3553,
      "step": 4673
    },
    {
      "epoch": 0.9346,
      "loss_align": 0.05189478397369385,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21459443867206573,
      "neg_sim_mean": 0.3985642194747925,
      "pos_sim_mean": 0.9481052160263062,
      "step": 4673
    },
    {
      "epoch": 0.9348,
      "grad_norm": 1.8778527975082397,
      "learning_rate": 1.29724879145946e-06,
      "loss": 0.2665,
      "step": 4674
    },
    {
      "epoch": 0.9348,
      "loss_align": 0.08870571851730347,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25088492035865784,
      "neg_sim_mean": 0.560901403427124,
      "pos_sim_mean": 0.9112942814826965,
      "step": 4674
    },
    {
      "epoch": 0.935,
      "grad_norm": 1.862368106842041,
      "learning_rate": 1.2893609056119748e-06,
      "loss": 0.3396,
      "step": 4675
    },
    {
      "epoch": 0.935,
      "loss_align": 0.12110710144042969,
      "loss_contrastive": 0.033214271068573,
      "loss_gen": 0.4897650480270386,
      "neg_sim_mean": 0.8121071457862854,
      "pos_sim_mean": 0.8788928985595703,
      "step": 4675
    },
    {
      "epoch": 0.9352,
      "grad_norm": 2.1031832695007324,
      "learning_rate": 1.2814967607382432e-06,
      "loss": 0.6149,
      "step": 4676
    },
    {
      "epoch": 0.9352,
      "loss_align": 0.08698546886444092,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15232975780963898,
      "neg_sim_mean": 0.7929672598838806,
      "pos_sim_mean": 0.9130145311355591,
      "step": 4676
    },
    {
      "epoch": 0.9354,
      "grad_norm": 1.8313926458358765,
      "learning_rate": 1.2736563606711382e-06,
      "loss": 0.2393,
      "step": 4677
    },
    {
      "epoch": 0.9354,
      "loss_align": 0.09104388952255249,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15571878850460052,
      "neg_sim_mean": 0.6458713412284851,
      "pos_sim_mean": 0.9089561104774475,
      "step": 4677
    },
    {
      "epoch": 0.9356,
      "grad_norm": 1.5331575870513916,
      "learning_rate": 1.265839709232003e-06,
      "loss": 0.2468,
      "step": 4678
    },
    {
      "epoch": 0.9356,
      "loss_align": 0.1070629358291626,
      "loss_contrastive": 0.08438509702682495,
      "loss_gen": 0.11394135653972626,
      "neg_sim_mean": 0.8773221373558044,
      "pos_sim_mean": 0.8929370641708374,
      "step": 4678
    },
    {
      "epoch": 0.9358,
      "grad_norm": 1.1574677228927612,
      "learning_rate": 1.258046810230562e-06,
      "loss": 0.2311,
      "step": 4679
    },
    {
      "epoch": 0.9358,
      "loss_align": 0.07235479354858398,
      "loss_contrastive": 0.007110893726348877,
      "loss_gen": 0.1616760641336441,
      "neg_sim_mean": 0.834756076335907,
      "pos_sim_mean": 0.927645206451416,
      "step": 4679
    },
    {
      "epoch": 0.936,
      "grad_norm": 1.5577583312988281,
      "learning_rate": 1.2502776674649774e-06,
      "loss": 0.2349,
      "step": 4680
    },
    {
      "epoch": 0.936,
      "loss_align": 0.1187671422958374,
      "loss_contrastive": 0.0,
      "loss_gen": 0.023862674832344055,
      "neg_sim_mean": 0.4045112729072571,
      "pos_sim_mean": 0.8812328577041626,
      "step": 4680
    },
    {
      "epoch": 0.9362,
      "grad_norm": 1.0347988605499268,
      "learning_rate": 1.2425322847218368e-06,
      "loss": 0.1426,
      "step": 4681
    },
    {
      "epoch": 0.9362,
      "loss_align": 0.037944793701171875,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2207234650850296,
      "neg_sim_mean": 0.772607684135437,
      "pos_sim_mean": 0.9620552062988281,
      "step": 4681
    },
    {
      "epoch": 0.9364,
      "grad_norm": 1.5886579751968384,
      "learning_rate": 1.2348106657761537e-06,
      "loss": 0.2587,
      "step": 4682
    },
    {
      "epoch": 0.9364,
      "loss_align": 0.192823588848114,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0344403050839901,
      "neg_sim_mean": 0.36438339948654175,
      "pos_sim_mean": 0.807176411151886,
      "step": 4682
    },
    {
      "epoch": 0.9366,
      "grad_norm": 1.253564476966858,
      "learning_rate": 1.2271128143913457e-06,
      "loss": 0.2273,
      "step": 4683
    },
    {
      "epoch": 0.9366,
      "loss_align": 0.0965346097946167,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3867705464363098,
      "neg_sim_mean": 0.6546158194541931,
      "pos_sim_mean": 0.9034653902053833,
      "step": 4683
    },
    {
      "epoch": 0.9368,
      "grad_norm": 2.2771925926208496,
      "learning_rate": 1.2194387343192504e-06,
      "loss": 0.4833,
      "step": 4684
    },
    {
      "epoch": 0.9368,
      "loss_align": 0.08296585083007812,
      "loss_contrastive": 0.06023263931274414,
      "loss_gen": 0.2786761522293091,
      "neg_sim_mean": 0.8772667646408081,
      "pos_sim_mean": 0.9170341491699219,
      "step": 4684
    },
    {
      "epoch": 0.937,
      "grad_norm": 1.9088157415390015,
      "learning_rate": 1.2117884293001258e-06,
      "loss": 0.3689,
      "step": 4685
    },
    {
      "epoch": 0.937,
      "loss_align": 0.10782361030578613,
      "loss_contrastive": 0.027219772338867188,
      "loss_gen": 0.20238444209098816,
      "neg_sim_mean": 0.8193961381912231,
      "pos_sim_mean": 0.8921763896942139,
      "step": 4685
    },
    {
      "epoch": 0.9372,
      "grad_norm": 1.844347357749939,
      "learning_rate": 1.2041619030626284e-06,
      "loss": 0.3135,
      "step": 4686
    },
    {
      "epoch": 0.9372,
      "loss_align": 0.08073878288269043,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21280191838741302,
      "neg_sim_mean": 0.7233489751815796,
      "pos_sim_mean": 0.9192612171173096,
      "step": 4686
    },
    {
      "epoch": 0.9374,
      "grad_norm": 1.7287373542785645,
      "learning_rate": 1.1965591593238513e-06,
      "loss": 0.2935,
      "step": 4687
    },
    {
      "epoch": 0.9374,
      "loss_align": 0.0704851746559143,
      "loss_contrastive": 0.08444225788116455,
      "loss_gen": 0.17759722471237183,
      "neg_sim_mean": 0.9139570593833923,
      "pos_sim_mean": 0.9295148253440857,
      "step": 4687
    },
    {
      "epoch": 0.9376,
      "grad_norm": 1.5751162767410278,
      "learning_rate": 1.1889802017892636e-06,
      "loss": 0.2582,
      "step": 4688
    },
    {
      "epoch": 0.9376,
      "loss_align": 0.10552096366882324,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09976296871900558,
      "neg_sim_mean": 0.4265249967575073,
      "pos_sim_mean": 0.8944790363311768,
      "step": 4688
    },
    {
      "epoch": 0.9378,
      "grad_norm": 1.339275598526001,
      "learning_rate": 1.1814250341527611e-06,
      "loss": 0.2053,
      "step": 4689
    },
    {
      "epoch": 0.9378,
      "loss_align": 0.14157450199127197,
      "loss_contrastive": 0.0,
      "loss_gen": 0.025417465716600418,
      "neg_sim_mean": 0.2791503071784973,
      "pos_sim_mean": 0.858425498008728,
      "step": 4689
    },
    {
      "epoch": 0.938,
      "grad_norm": 0.992810070514679,
      "learning_rate": 1.1738936600966367e-06,
      "loss": 0.167,
      "step": 4690
    },
    {
      "epoch": 0.938,
      "loss_align": 0.07701313495635986,
      "loss_contrastive": 0.029761195182800293,
      "loss_gen": 0.1208307296037674,
      "neg_sim_mean": 0.8527480363845825,
      "pos_sim_mean": 0.9229868650436401,
      "step": 4690
    },
    {
      "epoch": 0.9382,
      "grad_norm": 1.4316291809082031,
      "learning_rate": 1.166386083291604e-06,
      "loss": 0.2014,
      "step": 4691
    },
    {
      "epoch": 0.9382,
      "loss_align": 0.2281213402748108,
      "loss_contrastive": 0.09336203336715698,
      "loss_gen": 0.19368137419223785,
      "neg_sim_mean": 0.7652406692504883,
      "pos_sim_mean": 0.7718786597251892,
      "step": 4691
    },
    {
      "epoch": 0.9384,
      "grad_norm": 1.8361479043960571,
      "learning_rate": 1.1589023073967587e-06,
      "loss": 0.433,
      "step": 4692
    },
    {
      "epoch": 0.9384,
      "loss_align": 0.1178199052810669,
      "loss_contrastive": 0.08157902956008911,
      "loss_gen": 0.49025124311447144,
      "neg_sim_mean": 0.8637591004371643,
      "pos_sim_mean": 0.8821800947189331,
      "step": 4692
    },
    {
      "epoch": 0.9386,
      "grad_norm": 2.260732650756836,
      "learning_rate": 1.1514423360595938e-06,
      "loss": 0.6179,
      "step": 4693
    },
    {
      "epoch": 0.9386,
      "loss_align": 0.04102146625518799,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0033837012015283108,
      "neg_sim_mean": 0.39343053102493286,
      "pos_sim_mean": 0.958978533744812,
      "step": 4693
    },
    {
      "epoch": 0.9388,
      "grad_norm": 0.3564338684082031,
      "learning_rate": 1.1440061729160235e-06,
      "loss": 0.0444,
      "step": 4694
    },
    {
      "epoch": 0.9388,
      "loss_align": 0.09968078136444092,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16744112968444824,
      "neg_sim_mean": 0.7559612989425659,
      "pos_sim_mean": 0.9003192186355591,
      "step": 4694
    },
    {
      "epoch": 0.939,
      "grad_norm": 1.5293453931808472,
      "learning_rate": 1.136593821590326e-06,
      "loss": 0.2671,
      "step": 4695
    },
    {
      "epoch": 0.939,
      "loss_align": 0.14357268810272217,
      "loss_contrastive": 0.01708853244781494,
      "loss_gen": 0.2226296216249466,
      "neg_sim_mean": 0.7735158205032349,
      "pos_sim_mean": 0.8564273118972778,
      "step": 4695
    },
    {
      "epoch": 0.9392,
      "grad_norm": 1.7059111595153809,
      "learning_rate": 1.1292052856952062e-06,
      "loss": 0.3683,
      "step": 4696
    },
    {
      "epoch": 0.9392,
      "loss_align": 0.07115697860717773,
      "loss_contrastive": 0.07017463445663452,
      "loss_gen": 0.08214063942432404,
      "neg_sim_mean": 0.8990176320075989,
      "pos_sim_mean": 0.9288430213928223,
      "step": 4696
    },
    {
      "epoch": 0.9394,
      "grad_norm": 1.027955174446106,
      "learning_rate": 1.1218405688317446e-06,
      "loss": 0.1617,
      "step": 4697
    },
    {
      "epoch": 0.9394,
      "loss_align": 0.14847052097320557,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16980834305286407,
      "neg_sim_mean": 0.50102698802948,
      "pos_sim_mean": 0.8515294790267944,
      "step": 4697
    },
    {
      "epoch": 0.9396,
      "grad_norm": 1.6279667615890503,
      "learning_rate": 1.1144996745894032e-06,
      "loss": 0.3183,
      "step": 4698
    },
    {
      "epoch": 0.9396,
      "loss_align": 0.07765650749206543,
      "loss_contrastive": 0.02573835849761963,
      "loss_gen": 0.26645174622535706,
      "neg_sim_mean": 0.8480818271636963,
      "pos_sim_mean": 0.9223434925079346,
      "step": 4698
    },
    {
      "epoch": 0.9398,
      "grad_norm": 1.6649794578552246,
      "learning_rate": 1.1071826065460588e-06,
      "loss": 0.3472,
      "step": 4699
    },
    {
      "epoch": 0.9398,
      "loss_align": 0.04068613052368164,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2333637923002243,
      "neg_sim_mean": 0.8145536780357361,
      "pos_sim_mean": 0.9593138694763184,
      "step": 4699
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.6595569849014282,
      "learning_rate": 1.0998893682679479e-06,
      "loss": 0.274,
      "step": 4700
    },
    {
      "epoch": 0.94,
      "loss_align": 0.054396629333496094,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09435374289751053,
      "neg_sim_mean": 0.7156066298484802,
      "pos_sim_mean": 0.9456033706665039,
      "step": 4700
    },
    {
      "epoch": 0.9402,
      "grad_norm": 1.2583194971084595,
      "learning_rate": 1.0926199633097157e-06,
      "loss": 0.1488,
      "step": 4701
    },
    {
      "epoch": 0.9402,
      "loss_align": 0.10108602046966553,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0034916927106678486,
      "neg_sim_mean": 0.4766885042190552,
      "pos_sim_mean": 0.8989139795303345,
      "step": 4701
    },
    {
      "epoch": 0.9404,
      "grad_norm": 0.46652087569236755,
      "learning_rate": 1.0853743952143836e-06,
      "loss": 0.1046,
      "step": 4702
    },
    {
      "epoch": 0.9404,
      "loss_align": 0.06913620233535767,
      "loss_contrastive": 0.06804871559143066,
      "loss_gen": 0.27177193760871887,
      "neg_sim_mean": 0.8989124894142151,
      "pos_sim_mean": 0.9308637976646423,
      "step": 4702
    },
    {
      "epoch": 0.9406,
      "grad_norm": 2.21512770652771,
      "learning_rate": 1.078152667513349e-06,
      "loss": 0.3491,
      "step": 4703
    },
    {
      "epoch": 0.9406,
      "loss_align": 0.11842882633209229,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17923986911773682,
      "neg_sim_mean": 0.6198574304580688,
      "pos_sim_mean": 0.8815711736679077,
      "step": 4703
    },
    {
      "epoch": 0.9408,
      "grad_norm": 1.7498761415481567,
      "learning_rate": 1.0709547837263966e-06,
      "loss": 0.2977,
      "step": 4704
    },
    {
      "epoch": 0.9408,
      "loss_align": 0.1616213321685791,
      "loss_contrastive": 0.09462648630142212,
      "loss_gen": 0.23439329862594604,
      "neg_sim_mean": 0.8330051302909851,
      "pos_sim_mean": 0.8383786678314209,
      "step": 4704
    },
    {
      "epoch": 0.941,
      "grad_norm": 2.1182327270507812,
      "learning_rate": 1.0637807473616812e-06,
      "loss": 0.4074,
      "step": 4705
    },
    {
      "epoch": 0.941,
      "loss_align": 0.05283033847808838,
      "loss_contrastive": 0.03264045715332031,
      "loss_gen": 0.18968644738197327,
      "neg_sim_mean": 0.879810094833374,
      "pos_sim_mean": 0.9471696615219116,
      "step": 4705
    },
    {
      "epoch": 0.9412,
      "grad_norm": 1.7785838842391968,
      "learning_rate": 1.0566305619157502e-06,
      "loss": 0.2464,
      "step": 4706
    },
    {
      "epoch": 0.9412,
      "loss_align": 0.13491147756576538,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1628495752811432,
      "neg_sim_mean": 0.5245827436447144,
      "pos_sim_mean": 0.8650885224342346,
      "step": 4706
    },
    {
      "epoch": 0.9414,
      "grad_norm": 1.5640567541122437,
      "learning_rate": 1.0495042308735103e-06,
      "loss": 0.2978,
      "step": 4707
    },
    {
      "epoch": 0.9414,
      "loss_align": 0.09389239549636841,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01574573665857315,
      "neg_sim_mean": 0.30780139565467834,
      "pos_sim_mean": 0.9061076045036316,
      "step": 4707
    },
    {
      "epoch": 0.9416,
      "grad_norm": 0.8295141458511353,
      "learning_rate": 1.0424017577082556e-06,
      "loss": 0.1096,
      "step": 4708
    },
    {
      "epoch": 0.9416,
      "loss_align": 0.1391310691833496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17170344293117523,
      "neg_sim_mean": 0.324365496635437,
      "pos_sim_mean": 0.8608689308166504,
      "step": 4708
    },
    {
      "epoch": 0.9418,
      "grad_norm": 1.4583927392959595,
      "learning_rate": 1.0353231458816337e-06,
      "loss": 0.3108,
      "step": 4709
    },
    {
      "epoch": 0.9418,
      "loss_align": 0.07083690166473389,
      "loss_contrastive": 0.00863426923751831,
      "loss_gen": 0.16260313987731934,
      "neg_sim_mean": 0.8377973437309265,
      "pos_sim_mean": 0.9291630983352661,
      "step": 4709
    },
    {
      "epoch": 0.942,
      "grad_norm": 1.7441990375518799,
      "learning_rate": 1.0282683988436792e-06,
      "loss": 0.2345,
      "step": 4710
    },
    {
      "epoch": 0.942,
      "loss_align": 0.2040412425994873,
      "loss_contrastive": 0.07378578186035156,
      "loss_gen": 0.25022566318511963,
      "neg_sim_mean": 0.7697445154190063,
      "pos_sim_mean": 0.7959587574005127,
      "step": 4710
    },
    {
      "epoch": 0.9422,
      "grad_norm": 1.6041245460510254,
      "learning_rate": 1.0212375200327973e-06,
      "loss": 0.4631,
      "step": 4711
    },
    {
      "epoch": 0.9422,
      "loss_align": 0.17282891273498535,
      "loss_contrastive": 0.04605746269226074,
      "loss_gen": 0.16207876801490784,
      "neg_sim_mean": 0.7732285261154175,
      "pos_sim_mean": 0.8271710872650146,
      "step": 4711
    },
    {
      "epoch": 0.9424,
      "grad_norm": 1.6975144147872925,
      "learning_rate": 1.0142305128757468e-06,
      "loss": 0.3404,
      "step": 4712
    },
    {
      "epoch": 0.9424,
      "loss_align": 0.3133392930030823,
      "loss_contrastive": 0.14202600717544556,
      "loss_gen": 0.222150981426239,
      "neg_sim_mean": 0.7286866903305054,
      "pos_sim_mean": 0.6866607069969177,
      "step": 4712
    },
    {
      "epoch": 0.9426,
      "grad_norm": 1.7855305671691895,
      "learning_rate": 1.007247380787657e-06,
      "loss": 0.5525,
      "step": 4713
    },
    {
      "epoch": 0.9426,
      "loss_align": 0.04658854007720947,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005018853582441807,
      "neg_sim_mean": 0.33022409677505493,
      "pos_sim_mean": 0.9534114599227905,
      "step": 4713
    },
    {
      "epoch": 0.9428,
      "grad_norm": 0.4613606035709381,
      "learning_rate": 1.0002881271720222e-06,
      "loss": 0.0516,
      "step": 4714
    },
    {
      "epoch": 0.9428,
      "loss_align": 0.19936823844909668,
      "loss_contrastive": 0.08765256404876709,
      "loss_gen": 0.15814705193042755,
      "neg_sim_mean": 0.7882843017578125,
      "pos_sim_mean": 0.8006317615509033,
      "step": 4714
    },
    {
      "epoch": 0.943,
      "grad_norm": 1.5203661918640137,
      "learning_rate": 9.93352755420701e-07,
      "loss": 0.368,
      "step": 4715
    },
    {
      "epoch": 0.943,
      "loss_align": 0.14168238639831543,
      "loss_contrastive": 0.1148691177368164,
      "loss_gen": 0.2712627947330475,
      "neg_sim_mean": 0.8731867074966431,
      "pos_sim_mean": 0.8583176136016846,
      "step": 4715
    },
    {
      "epoch": 0.9432,
      "grad_norm": 1.9442613124847412,
      "learning_rate": 9.864412689139123e-07,
      "loss": 0.4267,
      "step": 4716
    },
    {
      "epoch": 0.9432,
      "loss_align": 0.08829867839813232,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26854264736175537,
      "neg_sim_mean": 0.6011219620704651,
      "pos_sim_mean": 0.9117013216018677,
      "step": 4716
    },
    {
      "epoch": 0.9434,
      "grad_norm": 2.002214193344116,
      "learning_rate": 9.795536710202169e-07,
      "loss": 0.3568,
      "step": 4717
    },
    {
      "epoch": 0.9434,
      "loss_align": 0.07882952690124512,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17725059390068054,
      "neg_sim_mean": 0.7968220114707947,
      "pos_sim_mean": 0.9211704730987549,
      "step": 4717
    },
    {
      "epoch": 0.9436,
      "grad_norm": 1.3728398084640503,
      "learning_rate": 9.726899650965625e-07,
      "loss": 0.2561,
      "step": 4718
    },
    {
      "epoch": 0.9436,
      "loss_align": 0.09267973899841309,
      "loss_contrastive": 0.03951364755630493,
      "loss_gen": 0.09968994557857513,
      "neg_sim_mean": 0.8468338847160339,
      "pos_sim_mean": 0.9073202610015869,
      "step": 4718
    },
    {
      "epoch": 0.9438,
      "grad_norm": 1.1568009853363037,
      "learning_rate": 9.65850154488218e-07,
      "loss": 0.1971,
      "step": 4719
    },
    {
      "epoch": 0.9438,
      "loss_align": 0.11996018886566162,
      "loss_contrastive": 0.13102424144744873,
      "loss_gen": 0.2513478398323059,
      "neg_sim_mean": 0.9110640287399292,
      "pos_sim_mean": 0.8800398111343384,
      "step": 4719
    },
    {
      "epoch": 0.944,
      "grad_norm": 1.8771356344223022,
      "learning_rate": 9.590342425288446e-07,
      "loss": 0.387,
      "step": 4720
    },
    {
      "epoch": 0.944,
      "loss_align": 0.053104519844055176,
      "loss_contrastive": 0.057513654232025146,
      "loss_gen": 0.22410525381565094,
      "neg_sim_mean": 0.9044091105461121,
      "pos_sim_mean": 0.9468954801559448,
      "step": 4720
    },
    {
      "epoch": 0.9442,
      "grad_norm": 1.879837155342102,
      "learning_rate": 9.522422325404235e-07,
      "loss": 0.2841,
      "step": 4721
    },
    {
      "epoch": 0.9442,
      "loss_align": 0.08372366428375244,
      "loss_contrastive": 0.04934406280517578,
      "loss_gen": 0.22706159949302673,
      "neg_sim_mean": 0.8656203746795654,
      "pos_sim_mean": 0.9162763357162476,
      "step": 4721
    },
    {
      "epoch": 0.9444,
      "grad_norm": 1.9766193628311157,
      "learning_rate": 9.454741278333012e-07,
      "loss": 0.3167,
      "step": 4722
    },
    {
      "epoch": 0.9444,
      "loss_align": 0.06281352043151855,
      "loss_contrastive": 0.021306157112121582,
      "loss_gen": 0.20586545765399933,
      "neg_sim_mean": 0.8584926128387451,
      "pos_sim_mean": 0.9371864795684814,
      "step": 4722
    },
    {
      "epoch": 0.9446,
      "grad_norm": 1.4653925895690918,
      "learning_rate": 9.387299317061616e-07,
      "loss": 0.2712,
      "step": 4723
    },
    {
      "epoch": 0.9446,
      "loss_align": 0.14046621322631836,
      "loss_contrastive": 0.025653362274169922,
      "loss_gen": 0.3060367703437805,
      "neg_sim_mean": 0.7851871252059937,
      "pos_sim_mean": 0.8595337867736816,
      "step": 4723
    },
    {
      "epoch": 0.9448,
      "grad_norm": 2.1018598079681396,
      "learning_rate": 9.320096474460527e-07,
      "loss": 0.4496,
      "step": 4724
    },
    {
      "epoch": 0.9448,
      "loss_align": 0.08474433422088623,
      "loss_contrastive": 0.005733132362365723,
      "loss_gen": 0.10311448574066162,
      "neg_sim_mean": 0.8209887742996216,
      "pos_sim_mean": 0.9152556657791138,
      "step": 4724
    },
    {
      "epoch": 0.945,
      "grad_norm": 1.345597267150879,
      "learning_rate": 9.253132783283547e-07,
      "loss": 0.1885,
      "step": 4725
    },
    {
      "epoch": 0.945,
      "loss_align": 0.09736233949661255,
      "loss_contrastive": 0.07338482141494751,
      "loss_gen": 0.17317995429039001,
      "neg_sim_mean": 0.876022458076477,
      "pos_sim_mean": 0.9026376605033875,
      "step": 4725
    },
    {
      "epoch": 0.9452,
      "grad_norm": 1.3984134197235107,
      "learning_rate": 9.186408276168013e-07,
      "loss": 0.2793,
      "step": 4726
    },
    {
      "epoch": 0.9452,
      "loss_align": 0.0610121488571167,
      "loss_contrastive": 0.0027045607566833496,
      "loss_gen": 0.14701618254184723,
      "neg_sim_mean": 0.8416923880577087,
      "pos_sim_mean": 0.9389878511428833,
      "step": 4726
    },
    {
      "epoch": 0.9454,
      "grad_norm": 1.5162824392318726,
      "learning_rate": 9.119922985634633e-07,
      "loss": 0.2084,
      "step": 4727
    },
    {
      "epoch": 0.9454,
      "loss_align": 0.1736128330230713,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005315159913152456,
      "neg_sim_mean": 0.30392730236053467,
      "pos_sim_mean": 0.8263871669769287,
      "step": 4727
    },
    {
      "epoch": 0.9456,
      "grad_norm": 0.663174569606781,
      "learning_rate": 9.053676944087541e-07,
      "loss": 0.1789,
      "step": 4728
    },
    {
      "epoch": 0.9456,
      "loss_align": 0.15859413146972656,
      "loss_contrastive": 0.0,
      "loss_gen": 0.024029146879911423,
      "neg_sim_mean": 0.4897115230560303,
      "pos_sim_mean": 0.8414058685302734,
      "step": 4728
    },
    {
      "epoch": 0.9458,
      "grad_norm": 1.1249940395355225,
      "learning_rate": 8.987670183814134e-07,
      "loss": 0.1826,
      "step": 4729
    },
    {
      "epoch": 0.9458,
      "loss_align": 0.08187389373779297,
      "loss_contrastive": 0.02437281608581543,
      "loss_gen": 0.25871843099594116,
      "neg_sim_mean": 0.8424988985061646,
      "pos_sim_mean": 0.918126106262207,
      "step": 4729
    },
    {
      "epoch": 0.946,
      "grad_norm": 1.8017590045928955,
      "learning_rate": 8.921902736985399e-07,
      "loss": 0.3435,
      "step": 4730
    },
    {
      "epoch": 0.946,
      "loss_align": 0.22162896394729614,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29531756043434143,
      "neg_sim_mean": 0.6593581438064575,
      "pos_sim_mean": 0.7783710360527039,
      "step": 4730
    },
    {
      "epoch": 0.9462,
      "grad_norm": 2.192699909210205,
      "learning_rate": 8.856374635655695e-07,
      "loss": 0.5169,
      "step": 4731
    },
    {
      "epoch": 0.9462,
      "loss_align": 0.11672359704971313,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1804942786693573,
      "neg_sim_mean": 0.6985342502593994,
      "pos_sim_mean": 0.8832764029502869,
      "step": 4731
    },
    {
      "epoch": 0.9464,
      "grad_norm": 1.6576321125030518,
      "learning_rate": 8.791085911762476e-07,
      "loss": 0.2972,
      "step": 4732
    },
    {
      "epoch": 0.9464,
      "loss_align": 0.11683845520019531,
      "loss_contrastive": 0.0,
      "loss_gen": 0.07881030440330505,
      "neg_sim_mean": 0.726140558719635,
      "pos_sim_mean": 0.8831615447998047,
      "step": 4732
    },
    {
      "epoch": 0.9466,
      "grad_norm": 1.368126392364502,
      "learning_rate": 8.726036597126619e-07,
      "loss": 0.1956,
      "step": 4733
    },
    {
      "epoch": 0.9466,
      "loss_align": 0.10756182670593262,
      "loss_contrastive": 0.05751115083694458,
      "loss_gen": 0.27057507634162903,
      "neg_sim_mean": 0.849949300289154,
      "pos_sim_mean": 0.8924381732940674,
      "step": 4733
    },
    {
      "epoch": 0.9468,
      "grad_norm": 1.9374676942825317,
      "learning_rate": 8.661226723452543e-07,
      "loss": 0.385,
      "step": 4734
    },
    {
      "epoch": 0.9468,
      "loss_align": 0.16851961612701416,
      "loss_contrastive": 7.814168930053711e-05,
      "loss_gen": 0.16646523773670197,
      "neg_sim_mean": 0.7315585017204285,
      "pos_sim_mean": 0.8314803838729858,
      "step": 4734
    },
    {
      "epoch": 0.947,
      "grad_norm": 1.7851715087890625,
      "learning_rate": 8.596656322327645e-07,
      "loss": 0.335,
      "step": 4735
    },
    {
      "epoch": 0.947,
      "loss_align": 0.1171499490737915,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2698991000652313,
      "neg_sim_mean": 0.7086224555969238,
      "pos_sim_mean": 0.8828500509262085,
      "step": 4735
    },
    {
      "epoch": 0.9472,
      "grad_norm": 1.8186215162277222,
      "learning_rate": 8.53232542522292e-07,
      "loss": 0.387,
      "step": 4736
    },
    {
      "epoch": 0.9472,
      "loss_align": 0.1418609619140625,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003637815359979868,
      "neg_sim_mean": 0.4532568156719208,
      "pos_sim_mean": 0.8581390380859375,
      "step": 4736
    },
    {
      "epoch": 0.9474,
      "grad_norm": 0.6668870449066162,
      "learning_rate": 8.468234063492286e-07,
      "loss": 0.1455,
      "step": 4737
    },
    {
      "epoch": 0.9474,
      "loss_align": 0.33750009536743164,
      "loss_contrastive": 0.07124549150466919,
      "loss_gen": 0.1885063350200653,
      "neg_sim_mean": 0.6337453722953796,
      "pos_sim_mean": 0.6624999046325684,
      "step": 4737
    },
    {
      "epoch": 0.9476,
      "grad_norm": 2.0613608360290527,
      "learning_rate": 8.404382268373145e-07,
      "loss": 0.5346,
      "step": 4738
    },
    {
      "epoch": 0.9476,
      "loss_align": 0.19342005252838135,
      "loss_contrastive": 0.037909507751464844,
      "loss_gen": 0.21853753924369812,
      "neg_sim_mean": 0.7444894313812256,
      "pos_sim_mean": 0.8065799474716187,
      "step": 4738
    },
    {
      "epoch": 0.9478,
      "grad_norm": 1.6959407329559326,
      "learning_rate": 8.340770070986214e-07,
      "loss": 0.4165,
      "step": 4739
    },
    {
      "epoch": 0.9478,
      "loss_align": 0.05360245704650879,
      "loss_contrastive": 0.01702094078063965,
      "loss_gen": 0.42111754417419434,
      "neg_sim_mean": 0.863418459892273,
      "pos_sim_mean": 0.9463975429534912,
      "step": 4739
    },
    {
      "epoch": 0.948,
      "grad_norm": 2.638650417327881,
      "learning_rate": 8.277397502335194e-07,
      "loss": 0.4768,
      "step": 4740
    },
    {
      "epoch": 0.948,
      "loss_align": 0.10105019807815552,
      "loss_contrastive": 0.08791917562484741,
      "loss_gen": 0.23268242180347443,
      "neg_sim_mean": 0.886868953704834,
      "pos_sim_mean": 0.8989498019218445,
      "step": 4740
    },
    {
      "epoch": 0.9482,
      "grad_norm": 2.1225318908691406,
      "learning_rate": 8.214264593307098e-07,
      "loss": 0.3443,
      "step": 4741
    },
    {
      "epoch": 0.9482,
      "loss_align": 0.10076916217803955,
      "loss_contrastive": 0.08787596225738525,
      "loss_gen": 0.168049156665802,
      "neg_sim_mean": 0.8871067762374878,
      "pos_sim_mean": 0.8992308378219604,
      "step": 4741
    },
    {
      "epoch": 0.9484,
      "grad_norm": 1.5397207736968994,
      "learning_rate": 8.151371374672146e-07,
      "loss": 0.2794,
      "step": 4742
    },
    {
      "epoch": 0.9484,
      "loss_align": 0.12585431337356567,
      "loss_contrastive": 0.034619033336639404,
      "loss_gen": 0.15034781396389008,
      "neg_sim_mean": 0.8087646961212158,
      "pos_sim_mean": 0.8741456866264343,
      "step": 4742
    },
    {
      "epoch": 0.9486,
      "grad_norm": 1.5856853723526,
      "learning_rate": 8.088717877083707e-07,
      "loss": 0.2804,
      "step": 4743
    },
    {
      "epoch": 0.9486,
      "loss_align": 0.06740933656692505,
      "loss_contrastive": 0.0,
      "loss_gen": 0.007583667989820242,
      "neg_sim_mean": 0.2967846095561981,
      "pos_sim_mean": 0.932590663433075,
      "step": 4743
    },
    {
      "epoch": 0.9488,
      "grad_norm": 0.6162890791893005,
      "learning_rate": 8.02630413107841e-07,
      "loss": 0.075,
      "step": 4744
    },
    {
      "epoch": 0.9488,
      "loss_align": 0.10948801040649414,
      "loss_contrastive": 0.03948104381561279,
      "loss_gen": 0.08825856447219849,
      "neg_sim_mean": 0.8299930095672607,
      "pos_sim_mean": 0.8905119895935059,
      "step": 4744
    },
    {
      "epoch": 0.949,
      "grad_norm": 1.321969747543335,
      "learning_rate": 7.964130167075923e-07,
      "loss": 0.2025,
      "step": 4745
    },
    {
      "epoch": 0.949,
      "loss_align": 0.12813204526901245,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006417182739824057,
      "neg_sim_mean": 0.3401185870170593,
      "pos_sim_mean": 0.8718679547309875,
      "step": 4745
    },
    {
      "epoch": 0.9492,
      "grad_norm": 0.6980041265487671,
      "learning_rate": 7.90219601537906e-07,
      "loss": 0.1345,
      "step": 4746
    },
    {
      "epoch": 0.9492,
      "loss_align": 0.15190088748931885,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1341352015733719,
      "neg_sim_mean": 0.5928057432174683,
      "pos_sim_mean": 0.8480991125106812,
      "step": 4746
    },
    {
      "epoch": 0.9494,
      "grad_norm": 1.4974596500396729,
      "learning_rate": 7.840501706173786e-07,
      "loss": 0.286,
      "step": 4747
    },
    {
      "epoch": 0.9494,
      "loss_align": 0.07029163837432861,
      "loss_contrastive": 0.0,
      "loss_gen": 0.4910530149936676,
      "neg_sim_mean": 0.7537086009979248,
      "pos_sim_mean": 0.9297083616256714,
      "step": 4747
    },
    {
      "epoch": 0.9496,
      "grad_norm": 2.1278061866760254,
      "learning_rate": 7.779047269529105e-07,
      "loss": 0.5613,
      "step": 4748
    },
    {
      "epoch": 0.9496,
      "loss_align": 0.08697295188903809,
      "loss_contrastive": 0.05424445867538452,
      "loss_gen": 0.2919604182243347,
      "neg_sim_mean": 0.8672714829444885,
      "pos_sim_mean": 0.9130270481109619,
      "step": 4748
    },
    {
      "epoch": 0.9498,
      "grad_norm": 2.091562271118164,
      "learning_rate": 7.717832735397335e-07,
      "loss": 0.3854,
      "step": 4749
    },
    {
      "epoch": 0.9498,
      "loss_align": 0.09498107433319092,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0743895024061203,
      "neg_sim_mean": 0.776852011680603,
      "pos_sim_mean": 0.9050189256668091,
      "step": 4749
    },
    {
      "epoch": 0.95,
      "grad_norm": 1.0865590572357178,
      "learning_rate": 7.656858133613498e-07,
      "loss": 0.1694,
      "step": 4750
    },
    {
      "epoch": 0.95,
      "loss_align": 0.10809886455535889,
      "loss_contrastive": 0.057242631912231445,
      "loss_gen": 0.17208392918109894,
      "neg_sim_mean": 0.8491437435150146,
      "pos_sim_mean": 0.8919011354446411,
      "step": 4750
    },
    {
      "epoch": 0.9502,
      "grad_norm": 1.6159322261810303,
      "learning_rate": 7.596123493895991e-07,
      "loss": 0.2871,
      "step": 4751
    },
    {
      "epoch": 0.9502,
      "loss_align": 0.07483983039855957,
      "loss_contrastive": 0.06308996677398682,
      "loss_gen": 0.21266727149486542,
      "neg_sim_mean": 0.8882501125335693,
      "pos_sim_mean": 0.9251601696014404,
      "step": 4751
    },
    {
      "epoch": 0.9504,
      "grad_norm": 1.9711625576019287,
      "learning_rate": 7.535628845846076e-07,
      "loss": 0.2951,
      "step": 4752
    },
    {
      "epoch": 0.9504,
      "loss_align": 0.10182195901870728,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008744249120354652,
      "neg_sim_mean": 0.37598490715026855,
      "pos_sim_mean": 0.8981780409812927,
      "step": 4752
    },
    {
      "epoch": 0.9506,
      "grad_norm": 0.6746817827224731,
      "learning_rate": 7.475374218948117e-07,
      "loss": 0.1106,
      "step": 4753
    },
    {
      "epoch": 0.9506,
      "loss_align": 0.07711946964263916,
      "loss_contrastive": 0.0,
      "loss_gen": 0.08904470503330231,
      "neg_sim_mean": 0.7439234852790833,
      "pos_sim_mean": 0.9228805303573608,
      "step": 4753
    },
    {
      "epoch": 0.9508,
      "grad_norm": 1.1262480020523071,
      "learning_rate": 7.415359642569564e-07,
      "loss": 0.1662,
      "step": 4754
    },
    {
      "epoch": 0.9508,
      "loss_align": 0.07307910919189453,
      "loss_contrastive": 0.0416332483291626,
      "loss_gen": 0.41370686888694763,
      "neg_sim_mean": 0.8685541152954102,
      "pos_sim_mean": 0.9269208908081055,
      "step": 4754
    },
    {
      "epoch": 0.951,
      "grad_norm": 2.33921217918396,
      "learning_rate": 7.355585145960742e-07,
      "loss": 0.4918,
      "step": 4755
    },
    {
      "epoch": 0.951,
      "loss_align": 0.1060335636138916,
      "loss_contrastive": 0.00944983959197998,
      "loss_gen": 0.12463566660881042,
      "neg_sim_mean": 0.8034162521362305,
      "pos_sim_mean": 0.8939664363861084,
      "step": 4755
    },
    {
      "epoch": 0.9512,
      "grad_norm": 1.3503912687301636,
      "learning_rate": 7.296050758254957e-07,
      "loss": 0.2318,
      "step": 4756
    },
    {
      "epoch": 0.9512,
      "loss_align": 0.05630922317504883,
      "loss_contrastive": 0.04420042037963867,
      "loss_gen": 0.33219796419143677,
      "neg_sim_mean": 0.8878911733627319,
      "pos_sim_mean": 0.9436907768249512,
      "step": 4756
    },
    {
      "epoch": 0.9514,
      "grad_norm": 2.01639461517334,
      "learning_rate": 7.236756508468611e-07,
      "loss": 0.3938,
      "step": 4757
    },
    {
      "epoch": 0.9514,
      "loss_align": 0.09844464063644409,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10543368756771088,
      "neg_sim_mean": 0.8012765645980835,
      "pos_sim_mean": 0.9015553593635559,
      "step": 4757
    },
    {
      "epoch": 0.9516,
      "grad_norm": 1.195204734802246,
      "learning_rate": 7.177702425500976e-07,
      "loss": 0.2039,
      "step": 4758
    },
    {
      "epoch": 0.9516,
      "loss_align": 0.08211582899093628,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006187832914292812,
      "neg_sim_mean": 0.2700235843658447,
      "pos_sim_mean": 0.9178841710090637,
      "step": 4758
    },
    {
      "epoch": 0.9518,
      "grad_norm": 0.5246988534927368,
      "learning_rate": 7.11888853813436e-07,
      "loss": 0.0883,
      "step": 4759
    },
    {
      "epoch": 0.9518,
      "loss_align": 0.14173918962478638,
      "loss_contrastive": 0.08262717723846436,
      "loss_gen": 0.142142191529274,
      "neg_sim_mean": 0.8408879637718201,
      "pos_sim_mean": 0.8582608103752136,
      "step": 4759
    },
    {
      "epoch": 0.952,
      "grad_norm": 1.8998841047286987,
      "learning_rate": 7.060314875033835e-07,
      "loss": 0.2938,
      "step": 4760
    },
    {
      "epoch": 0.952,
      "loss_align": 0.08481240272521973,
      "loss_contrastive": 0.10441410541534424,
      "loss_gen": 0.34105515480041504,
      "neg_sim_mean": 0.9196016788482666,
      "pos_sim_mean": 0.9151875972747803,
      "step": 4760
    },
    {
      "epoch": 0.9522,
      "grad_norm": 2.3066320419311523,
      "learning_rate": 7.001981464747565e-07,
      "loss": 0.4384,
      "step": 4761
    },
    {
      "epoch": 0.9522,
      "loss_align": 0.1055566668510437,
      "loss_contrastive": 0.04792267084121704,
      "loss_gen": 0.3067420721054077,
      "neg_sim_mean": 0.8423659801483154,
      "pos_sim_mean": 0.8944433331489563,
      "step": 4761
    },
    {
      "epoch": 0.9524,
      "grad_norm": 2.1576995849609375,
      "learning_rate": 6.943888335706472e-07,
      "loss": 0.418,
      "step": 4762
    },
    {
      "epoch": 0.9524,
      "loss_align": 0.05720454454421997,
      "loss_contrastive": 0.09222346544265747,
      "loss_gen": 0.21487925946712494,
      "neg_sim_mean": 0.9350188970565796,
      "pos_sim_mean": 0.94279545545578,
      "step": 4762
    },
    {
      "epoch": 0.9526,
      "grad_norm": 1.3003993034362793,
      "learning_rate": 6.88603551622452e-07,
      "loss": 0.2832,
      "step": 4763
    },
    {
      "epoch": 0.9526,
      "loss_align": 0.10167992115020752,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1461724042892456,
      "neg_sim_mean": 0.4284546375274658,
      "pos_sim_mean": 0.8983200788497925,
      "step": 4763
    },
    {
      "epoch": 0.9528,
      "grad_norm": 1.5363165140151978,
      "learning_rate": 6.828423034498488e-07,
      "loss": 0.2479,
      "step": 4764
    },
    {
      "epoch": 0.9528,
      "loss_align": 0.146742045879364,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14069753885269165,
      "neg_sim_mean": 0.7359368801116943,
      "pos_sim_mean": 0.853257954120636,
      "step": 4764
    },
    {
      "epoch": 0.953,
      "grad_norm": 1.8005810976028442,
      "learning_rate": 6.771050918607913e-07,
      "loss": 0.2874,
      "step": 4765
    },
    {
      "epoch": 0.953,
      "loss_align": 0.1624075174331665,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1823900043964386,
      "neg_sim_mean": 0.5740561485290527,
      "pos_sim_mean": 0.8375924825668335,
      "step": 4765
    },
    {
      "epoch": 0.9532,
      "grad_norm": 1.925869107246399,
      "learning_rate": 6.713919196515317e-07,
      "loss": 0.3448,
      "step": 4766
    },
    {
      "epoch": 0.9532,
      "loss_align": 0.07595998048782349,
      "loss_contrastive": 0.04572427272796631,
      "loss_gen": 0.05016598850488663,
      "neg_sim_mean": 0.8697642683982849,
      "pos_sim_mean": 0.9240400195121765,
      "step": 4766
    },
    {
      "epoch": 0.9534,
      "grad_norm": 0.8020073771476746,
      "learning_rate": 6.657027896065982e-07,
      "loss": 0.1316,
      "step": 4767
    },
    {
      "epoch": 0.9534,
      "loss_align": 0.08532166481018066,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006611245684325695,
      "neg_sim_mean": 0.3395134210586548,
      "pos_sim_mean": 0.9146783351898193,
      "step": 4767
    },
    {
      "epoch": 0.9536,
      "grad_norm": 0.5974728465080261,
      "learning_rate": 6.60037704498806e-07,
      "loss": 0.0919,
      "step": 4768
    },
    {
      "epoch": 0.9536,
      "loss_align": 0.07454919815063477,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1664736568927765,
      "neg_sim_mean": 0.6483102440834045,
      "pos_sim_mean": 0.9254508018493652,
      "step": 4768
    },
    {
      "epoch": 0.9538,
      "grad_norm": 1.4886656999588013,
      "learning_rate": 6.543966670892465e-07,
      "loss": 0.241,
      "step": 4769
    },
    {
      "epoch": 0.9538,
      "loss_align": 0.09544563293457031,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3350779414176941,
      "neg_sim_mean": 0.6033941507339478,
      "pos_sim_mean": 0.9045543670654297,
      "step": 4769
    },
    {
      "epoch": 0.954,
      "grad_norm": 2.287959575653076,
      "learning_rate": 6.487796801272983e-07,
      "loss": 0.4305,
      "step": 4770
    },
    {
      "epoch": 0.954,
      "loss_align": 0.12744617462158203,
      "loss_contrastive": 0.0,
      "loss_gen": 0.021544067189097404,
      "neg_sim_mean": 0.33970773220062256,
      "pos_sim_mean": 0.872553825378418,
      "step": 4770
    },
    {
      "epoch": 0.9542,
      "grad_norm": 1.0430822372436523,
      "learning_rate": 6.431867463506048e-07,
      "loss": 0.149,
      "step": 4771
    },
    {
      "epoch": 0.9542,
      "loss_align": 0.07686072587966919,
      "loss_contrastive": 0.026939094066619873,
      "loss_gen": 0.303196519613266,
      "neg_sim_mean": 0.8500783443450928,
      "pos_sim_mean": 0.9231392741203308,
      "step": 4771
    },
    {
      "epoch": 0.9544,
      "grad_norm": 2.142878770828247,
      "learning_rate": 6.376178684850965e-07,
      "loss": 0.3833,
      "step": 4772
    },
    {
      "epoch": 0.9544,
      "loss_align": 0.24858903884887695,
      "loss_contrastive": 0.12736427783966064,
      "loss_gen": 0.24070248007774353,
      "neg_sim_mean": 0.7787752151489258,
      "pos_sim_mean": 0.751410961151123,
      "step": 4772
    },
    {
      "epoch": 0.9546,
      "grad_norm": 1.7186492681503296,
      "learning_rate": 6.3207304924498e-07,
      "loss": 0.5046,
      "step": 4773
    },
    {
      "epoch": 0.9546,
      "loss_align": 0.06442952156066895,
      "loss_contrastive": 0.041576087474823,
      "loss_gen": 0.27957072854042053,
      "neg_sim_mean": 0.8771465420722961,
      "pos_sim_mean": 0.935570478439331,
      "step": 4773
    },
    {
      "epoch": 0.9548,
      "grad_norm": 1.9666624069213867,
      "learning_rate": 6.265522913327327e-07,
      "loss": 0.349,
      "step": 4774
    },
    {
      "epoch": 0.9548,
      "loss_align": 0.03850758075714111,
      "loss_contrastive": 0.04097557067871094,
      "loss_gen": 0.17009098827838898,
      "neg_sim_mean": 0.9024679660797119,
      "pos_sim_mean": 0.9614924192428589,
      "step": 4774
    },
    {
      "epoch": 0.955,
      "grad_norm": 1.4668769836425781,
      "learning_rate": 6.210555974391074e-07,
      "loss": 0.2135,
      "step": 4775
    },
    {
      "epoch": 0.955,
      "loss_align": 0.07525819540023804,
      "loss_contrastive": 0.012620270252227783,
      "loss_gen": 0.25156939029693604,
      "neg_sim_mean": 0.8373620510101318,
      "pos_sim_mean": 0.924741804599762,
      "step": 4775
    },
    {
      "epoch": 0.9552,
      "grad_norm": 2.1022627353668213,
      "learning_rate": 6.15582970243117e-07,
      "loss": 0.3283,
      "step": 4776
    },
    {
      "epoch": 0.9552,
      "loss_align": 0.1591780185699463,
      "loss_contrastive": 0.0,
      "loss_gen": 0.006164156831800938,
      "neg_sim_mean": 0.5646845102310181,
      "pos_sim_mean": 0.8408219814300537,
      "step": 4776
    },
    {
      "epoch": 0.9554,
      "grad_norm": 0.8307960033416748,
      "learning_rate": 6.101344124120556e-07,
      "loss": 0.1653,
      "step": 4777
    },
    {
      "epoch": 0.9554,
      "loss_align": 0.15368413925170898,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2952684462070465,
      "neg_sim_mean": 0.7069392800331116,
      "pos_sim_mean": 0.846315860748291,
      "step": 4777
    },
    {
      "epoch": 0.9556,
      "grad_norm": 1.8571243286132812,
      "learning_rate": 6.047099266014877e-07,
      "loss": 0.449,
      "step": 4778
    },
    {
      "epoch": 0.9556,
      "loss_align": 0.1336788535118103,
      "loss_contrastive": 0.018667995929718018,
      "loss_gen": 0.20744718611240387,
      "neg_sim_mean": 0.7849891185760498,
      "pos_sim_mean": 0.8663211464881897,
      "step": 4778
    },
    {
      "epoch": 0.9558,
      "grad_norm": 1.7314752340316772,
      "learning_rate": 5.993095154552431e-07,
      "loss": 0.3434,
      "step": 4779
    },
    {
      "epoch": 0.9558,
      "loss_align": 0.07521456480026245,
      "loss_contrastive": 0.0,
      "loss_gen": 0.05135137215256691,
      "neg_sim_mean": 0.8104127645492554,
      "pos_sim_mean": 0.9247854351997375,
      "step": 4779
    },
    {
      "epoch": 0.956,
      "grad_norm": 1.0385563373565674,
      "learning_rate": 5.939331816054161e-07,
      "loss": 0.1266,
      "step": 4780
    },
    {
      "epoch": 0.956,
      "loss_align": 0.09360325336456299,
      "loss_contrastive": 0.0,
      "loss_gen": 0.014652957208454609,
      "neg_sim_mean": 0.3554801344871521,
      "pos_sim_mean": 0.906396746635437,
      "step": 4780
    },
    {
      "epoch": 0.9562,
      "grad_norm": 0.820074737071991,
      "learning_rate": 5.885809276723608e-07,
      "loss": 0.1083,
      "step": 4781
    },
    {
      "epoch": 0.9562,
      "loss_align": 0.08333432674407959,
      "loss_contrastive": 0.036114633083343506,
      "loss_gen": 0.14901557564735413,
      "neg_sim_mean": 0.852780282497406,
      "pos_sim_mean": 0.9166656732559204,
      "step": 4781
    },
    {
      "epoch": 0.9564,
      "grad_norm": 1.511174201965332,
      "learning_rate": 5.832527562647016e-07,
      "loss": 0.2367,
      "step": 4782
    },
    {
      "epoch": 0.9564,
      "loss_align": 0.04838883876800537,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003186781657859683,
      "neg_sim_mean": 0.3482881486415863,
      "pos_sim_mean": 0.9516111612319946,
      "step": 4782
    },
    {
      "epoch": 0.9566,
      "grad_norm": 0.3189263939857483,
      "learning_rate": 5.779486699793335e-07,
      "loss": 0.0516,
      "step": 4783
    },
    {
      "epoch": 0.9566,
      "loss_align": 0.18639570474624634,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02948440983891487,
      "neg_sim_mean": 0.358745276927948,
      "pos_sim_mean": 0.8136042952537537,
      "step": 4783
    },
    {
      "epoch": 0.9568,
      "grad_norm": 1.313215970993042,
      "learning_rate": 5.726686714013996e-07,
      "loss": 0.2159,
      "step": 4784
    },
    {
      "epoch": 0.9568,
      "loss_align": 0.08286166191101074,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1983845978975296,
      "neg_sim_mean": 0.7684016823768616,
      "pos_sim_mean": 0.9171383380889893,
      "step": 4784
    },
    {
      "epoch": 0.957,
      "grad_norm": 1.633759617805481,
      "learning_rate": 5.674127631043025e-07,
      "loss": 0.2812,
      "step": 4785
    },
    {
      "epoch": 0.957,
      "loss_align": 0.14965999126434326,
      "loss_contrastive": 0.010412096977233887,
      "loss_gen": 0.14495863020420074,
      "neg_sim_mean": 0.7607520818710327,
      "pos_sim_mean": 0.8503400087356567,
      "step": 4785
    },
    {
      "epoch": 0.9572,
      "grad_norm": 1.4893966913223267,
      "learning_rate": 5.621809476497098e-07,
      "loss": 0.2959,
      "step": 4786
    },
    {
      "epoch": 0.9572,
      "loss_align": 0.11691510677337646,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3436855673789978,
      "neg_sim_mean": 0.6162798404693604,
      "pos_sim_mean": 0.8830848932266235,
      "step": 4786
    },
    {
      "epoch": 0.9574,
      "grad_norm": 1.8026615381240845,
      "learning_rate": 5.569732275875428e-07,
      "loss": 0.4606,
      "step": 4787
    },
    {
      "epoch": 0.9574,
      "loss_align": 0.08027952909469604,
      "loss_contrastive": 0.07225698232650757,
      "loss_gen": 0.1412438601255417,
      "neg_sim_mean": 0.8919774293899536,
      "pos_sim_mean": 0.919720470905304,
      "step": 4787
    },
    {
      "epoch": 0.9576,
      "grad_norm": 1.5840578079223633,
      "learning_rate": 5.51789605455988e-07,
      "loss": 0.2302,
      "step": 4788
    },
    {
      "epoch": 0.9576,
      "loss_align": 0.08261728286743164,
      "loss_contrastive": 0.022785067558288574,
      "loss_gen": 0.2715022563934326,
      "neg_sim_mean": 0.840167760848999,
      "pos_sim_mean": 0.9173827171325684,
      "step": 4788
    },
    {
      "epoch": 0.9578,
      "grad_norm": 2.1100358963012695,
      "learning_rate": 5.466300837814797e-07,
      "loss": 0.3569,
      "step": 4789
    },
    {
      "epoch": 0.9578,
      "loss_align": 0.14818787574768066,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15426281094551086,
      "neg_sim_mean": 0.6091797351837158,
      "pos_sim_mean": 0.8518121242523193,
      "step": 4789
    },
    {
      "epoch": 0.958,
      "grad_norm": 1.5769096612930298,
      "learning_rate": 5.414946650786957e-07,
      "loss": 0.3025,
      "step": 4790
    },
    {
      "epoch": 0.958,
      "loss_align": 0.08499801158905029,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13207992911338806,
      "neg_sim_mean": 0.7700088024139404,
      "pos_sim_mean": 0.9150019884109497,
      "step": 4790
    },
    {
      "epoch": 0.9582,
      "grad_norm": 1.3965225219726562,
      "learning_rate": 5.363833518505834e-07,
      "loss": 0.2171,
      "step": 4791
    },
    {
      "epoch": 0.9582,
      "loss_align": 0.10656684637069702,
      "loss_contrastive": 0.03208649158477783,
      "loss_gen": 0.16270262002944946,
      "neg_sim_mean": 0.8255196213722229,
      "pos_sim_mean": 0.893433153629303,
      "step": 4791
    },
    {
      "epoch": 0.9584,
      "grad_norm": 1.7659157514572144,
      "learning_rate": 5.312961465883392e-07,
      "loss": 0.2731,
      "step": 4792
    },
    {
      "epoch": 0.9584,
      "loss_align": 0.15956169366836548,
      "loss_contrastive": 0.06586259603500366,
      "loss_gen": 0.2340548187494278,
      "neg_sim_mean": 0.8063008785247803,
      "pos_sim_mean": 0.8404383063316345,
      "step": 4792
    },
    {
      "epoch": 0.9586,
      "grad_norm": 1.7459737062454224,
      "learning_rate": 5.262330517713965e-07,
      "loss": 0.4015,
      "step": 4793
    },
    {
      "epoch": 0.9586,
      "loss_align": 0.12973207235336304,
      "loss_contrastive": 0.009536802768707275,
      "loss_gen": 0.3313312232494354,
      "neg_sim_mean": 0.7798047065734863,
      "pos_sim_mean": 0.870267927646637,
      "step": 4793
    },
    {
      "epoch": 0.9588,
      "grad_norm": 2.6312544345855713,
      "learning_rate": 5.211940698674534e-07,
      "loss": 0.4622,
      "step": 4794
    },
    {
      "epoch": 0.9588,
      "loss_align": 0.12342584133148193,
      "loss_contrastive": 0.0,
      "loss_gen": 0.016800375655293465,
      "neg_sim_mean": 0.4065224528312683,
      "pos_sim_mean": 0.8765741586685181,
      "step": 4794
    },
    {
      "epoch": 0.959,
      "grad_norm": 0.9430164098739624,
      "learning_rate": 5.161792033324398e-07,
      "loss": 0.1402,
      "step": 4795
    },
    {
      "epoch": 0.959,
      "loss_align": 0.06242263317108154,
      "loss_contrastive": 0.09768038988113403,
      "loss_gen": 0.2278856635093689,
      "neg_sim_mean": 0.9352577328681946,
      "pos_sim_mean": 0.9375773668289185,
      "step": 4795
    },
    {
      "epoch": 0.9592,
      "grad_norm": 2.167382001876831,
      "learning_rate": 5.111884546105506e-07,
      "loss": 0.302,
      "step": 4796
    },
    {
      "epoch": 0.9592,
      "loss_align": 0.03293049335479736,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11391760408878326,
      "neg_sim_mean": 0.7717845439910889,
      "pos_sim_mean": 0.9670695066452026,
      "step": 4796
    },
    {
      "epoch": 0.9594,
      "grad_norm": 1.0688400268554688,
      "learning_rate": 5.062218261342122e-07,
      "loss": 0.1468,
      "step": 4797
    },
    {
      "epoch": 0.9594,
      "loss_align": 0.05864906311035156,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00421225605532527,
      "neg_sim_mean": 0.28801947832107544,
      "pos_sim_mean": 0.9413509368896484,
      "step": 4797
    },
    {
      "epoch": 0.9596,
      "grad_norm": 0.3825148642063141,
      "learning_rate": 5.012793203240995e-07,
      "loss": 0.0629,
      "step": 4798
    },
    {
      "epoch": 0.9596,
      "loss_align": 0.09252184629440308,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11396748572587967,
      "neg_sim_mean": 0.5966615676879883,
      "pos_sim_mean": 0.9074781537055969,
      "step": 4798
    },
    {
      "epoch": 0.9598,
      "grad_norm": 1.047629475593567,
      "learning_rate": 4.963609395891299e-07,
      "loss": 0.2065,
      "step": 4799
    },
    {
      "epoch": 0.9598,
      "loss_align": 0.0555342435836792,
      "loss_contrastive": 0.07578569650650024,
      "loss_gen": 0.22929804027080536,
      "neg_sim_mean": 0.9202514290809631,
      "pos_sim_mean": 0.9444657564163208,
      "step": 4799
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.5849441289901733,
      "learning_rate": 4.914666863264528e-07,
      "loss": 0.2939,
      "step": 4800
    },
    {
      "epoch": 0.96,
      "loss_align": 0.07707548141479492,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0023845136165618896,
      "neg_sim_mean": 0.35313040018081665,
      "pos_sim_mean": 0.9229245185852051,
      "step": 4800
    },
    {
      "epoch": 0.9602,
      "grad_norm": 0.4005782902240753,
      "learning_rate": 4.865965629214819e-07,
      "loss": 0.0795,
      "step": 4801
    },
    {
      "epoch": 0.9602,
      "loss_align": 0.08500826358795166,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0036841409746557474,
      "neg_sim_mean": 0.31757795810699463,
      "pos_sim_mean": 0.9149917364120483,
      "step": 4801
    },
    {
      "epoch": 0.9604,
      "grad_norm": 0.5109148621559143,
      "learning_rate": 4.817505717478576e-07,
      "loss": 0.0887,
      "step": 4802
    },
    {
      "epoch": 0.9604,
      "loss_align": 0.05152541399002075,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21146366000175476,
      "neg_sim_mean": 0.8344360589981079,
      "pos_sim_mean": 0.9484745860099792,
      "step": 4802
    },
    {
      "epoch": 0.9606,
      "grad_norm": 1.6652964353561401,
      "learning_rate": 4.769287151674406e-07,
      "loss": 0.263,
      "step": 4803
    },
    {
      "epoch": 0.9606,
      "loss_align": 0.16444730758666992,
      "loss_contrastive": 0.11272108554840088,
      "loss_gen": 0.18475307524204254,
      "neg_sim_mean": 0.848273754119873,
      "pos_sim_mean": 0.8355526924133301,
      "step": 4803
    },
    {
      "epoch": 0.9608,
      "grad_norm": 1.6738747358322144,
      "learning_rate": 4.7213099553035654e-07,
      "loss": 0.3627,
      "step": 4804
    },
    {
      "epoch": 0.9608,
      "loss_align": 0.09261941909790039,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0018405638402327895,
      "neg_sim_mean": 0.5235685110092163,
      "pos_sim_mean": 0.9073805809020996,
      "step": 4804
    },
    {
      "epoch": 0.961,
      "grad_norm": 0.4483906924724579,
      "learning_rate": 4.673574151749571e-07,
      "loss": 0.0945,
      "step": 4805
    },
    {
      "epoch": 0.961,
      "loss_align": 0.11359065771102905,
      "loss_contrastive": 0.0,
      "loss_gen": 0.10477929562330246,
      "neg_sim_mean": 0.7820922136306763,
      "pos_sim_mean": 0.886409342288971,
      "step": 4805
    },
    {
      "epoch": 0.9612,
      "grad_norm": 1.574380874633789,
      "learning_rate": 4.6260797642782014e-07,
      "loss": 0.2184,
      "step": 4806
    },
    {
      "epoch": 0.9612,
      "loss_align": 0.065510094165802,
      "loss_contrastive": 0.03137969970703125,
      "loss_gen": 0.11599435657262802,
      "neg_sim_mean": 0.8658695816993713,
      "pos_sim_mean": 0.934489905834198,
      "step": 4806
    },
    {
      "epoch": 0.9614,
      "grad_norm": 1.457277536392212,
      "learning_rate": 4.578826816037718e-07,
      "loss": 0.1853,
      "step": 4807
    },
    {
      "epoch": 0.9614,
      "loss_align": 0.08561533689498901,
      "loss_contrastive": 0.06314665079116821,
      "loss_gen": 0.25401341915130615,
      "neg_sim_mean": 0.8775312900543213,
      "pos_sim_mean": 0.914384663105011,
      "step": 4807
    },
    {
      "epoch": 0.9616,
      "grad_norm": 1.5821589231491089,
      "learning_rate": 4.5318153300585865e-07,
      "loss": 0.3472,
      "step": 4808
    },
    {
      "epoch": 0.9616,
      "loss_align": 0.03730142116546631,
      "loss_contrastive": 0.04503822326660156,
      "loss_gen": 0.12153004109859467,
      "neg_sim_mean": 0.9077367782592773,
      "pos_sim_mean": 0.9626985788345337,
      "step": 4808
    },
    {
      "epoch": 0.9618,
      "grad_norm": 1.731233835220337,
      "learning_rate": 4.4850453292536454e-07,
      "loss": 0.1642,
      "step": 4809
    },
    {
      "epoch": 0.9618,
      "loss_align": 0.07840275764465332,
      "loss_contrastive": 0.09887969493865967,
      "loss_gen": 0.29940691590309143,
      "neg_sim_mean": 0.9204769134521484,
      "pos_sim_mean": 0.9215972423553467,
      "step": 4809
    },
    {
      "epoch": 0.962,
      "grad_norm": 1.894585371017456,
      "learning_rate": 4.438516836417994e-07,
      "loss": 0.3897,
      "step": 4810
    },
    {
      "epoch": 0.962,
      "loss_align": 0.2492661476135254,
      "loss_contrastive": 0.14363765716552734,
      "loss_gen": 0.2641984522342682,
      "neg_sim_mean": 0.794371485710144,
      "pos_sim_mean": 0.7507338523864746,
      "step": 4810
    },
    {
      "epoch": 0.9622,
      "grad_norm": 2.2196097373962402,
      "learning_rate": 4.392229874229159e-07,
      "loss": 0.5307,
      "step": 4811
    },
    {
      "epoch": 0.9622,
      "loss_align": 0.12972640991210938,
      "loss_contrastive": 0.09075981378555298,
      "loss_gen": 0.39452946186065674,
      "neg_sim_mean": 0.8610333800315857,
      "pos_sim_mean": 0.8702735900878906,
      "step": 4811
    },
    {
      "epoch": 0.9624,
      "grad_norm": 2.4471540451049805,
      "learning_rate": 4.3461844652467607e-07,
      "loss": 0.5351,
      "step": 4812
    },
    {
      "epoch": 0.9624,
      "loss_align": 0.16696274280548096,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12440820038318634,
      "neg_sim_mean": 0.5741417407989502,
      "pos_sim_mean": 0.833037257194519,
      "step": 4812
    },
    {
      "epoch": 0.9626,
      "grad_norm": 1.2078638076782227,
      "learning_rate": 4.300380631912737e-07,
      "loss": 0.2914,
      "step": 4813
    },
    {
      "epoch": 0.9626,
      "loss_align": 0.08281838893890381,
      "loss_contrastive": 0.07093417644500732,
      "loss_gen": 0.4197724759578705,
      "neg_sim_mean": 0.8881157636642456,
      "pos_sim_mean": 0.9171816110610962,
      "step": 4813
    },
    {
      "epoch": 0.9628,
      "grad_norm": 2.869763135910034,
      "learning_rate": 4.254818396551341e-07,
      "loss": 0.5111,
      "step": 4814
    },
    {
      "epoch": 0.9628,
      "loss_align": 0.05877053737640381,
      "loss_contrastive": 0.0,
      "loss_gen": 0.34045642614364624,
      "neg_sim_mean": 0.8100311160087585,
      "pos_sim_mean": 0.9412294626235962,
      "step": 4814
    },
    {
      "epoch": 0.963,
      "grad_norm": 1.711240530014038,
      "learning_rate": 4.2094977813691426e-07,
      "loss": 0.3992,
      "step": 4815
    },
    {
      "epoch": 0.963,
      "loss_align": 0.0765993595123291,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09213325381278992,
      "neg_sim_mean": 0.7901325225830078,
      "pos_sim_mean": 0.9234006404876709,
      "step": 4815
    },
    {
      "epoch": 0.9632,
      "grad_norm": 1.373015284538269,
      "learning_rate": 4.1644188084548063e-07,
      "loss": 0.1687,
      "step": 4816
    },
    {
      "epoch": 0.9632,
      "loss_align": 0.07229679822921753,
      "loss_contrastive": 0.04222041368484497,
      "loss_gen": 0.16591359674930573,
      "neg_sim_mean": 0.8699235916137695,
      "pos_sim_mean": 0.9277032017707825,
      "step": 4816
    },
    {
      "epoch": 0.9634,
      "grad_norm": 1.6418066024780273,
      "learning_rate": 4.1195814997792013e-07,
      "loss": 0.2433,
      "step": 4817
    },
    {
      "epoch": 0.9634,
      "loss_align": 0.10076320171356201,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1935405731201172,
      "neg_sim_mean": 0.7830877900123596,
      "pos_sim_mean": 0.899236798286438,
      "step": 4817
    },
    {
      "epoch": 0.9636,
      "grad_norm": 1.6678847074508667,
      "learning_rate": 4.074985877195625e-07,
      "loss": 0.2943,
      "step": 4818
    },
    {
      "epoch": 0.9636,
      "loss_align": 0.09401130676269531,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0016506197862327099,
      "neg_sim_mean": 0.4548937678337097,
      "pos_sim_mean": 0.9059886932373047,
      "step": 4818
    },
    {
      "epoch": 0.9638,
      "grad_norm": 0.40892118215560913,
      "learning_rate": 4.030631962439302e-07,
      "loss": 0.0957,
      "step": 4819
    },
    {
      "epoch": 0.9638,
      "loss_align": 0.1417943239212036,
      "loss_contrastive": 0.001538991928100586,
      "loss_gen": 0.19734512269496918,
      "neg_sim_mean": 0.7597446441650391,
      "pos_sim_mean": 0.8582056760787964,
      "step": 4819
    },
    {
      "epoch": 0.964,
      "grad_norm": 2.03340220451355,
      "learning_rate": 3.986519777127884e-07,
      "loss": 0.3393,
      "step": 4820
    },
    {
      "epoch": 0.964,
      "loss_align": 0.0623326301574707,
      "loss_contrastive": 0.08233803510665894,
      "loss_gen": 0.1669296771287918,
      "neg_sim_mean": 0.9200053811073303,
      "pos_sim_mean": 0.9376673698425293,
      "step": 4820
    },
    {
      "epoch": 0.9642,
      "grad_norm": 1.6018650531768799,
      "learning_rate": 3.9426493427611177e-07,
      "loss": 0.2391,
      "step": 4821
    },
    {
      "epoch": 0.9642,
      "loss_align": 0.09154355525970459,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20044943690299988,
      "neg_sim_mean": 0.6417019367218018,
      "pos_sim_mean": 0.9084564447402954,
      "step": 4821
    },
    {
      "epoch": 0.9644,
      "grad_norm": 1.9472706317901611,
      "learning_rate": 3.899020680720844e-07,
      "loss": 0.292,
      "step": 4822
    },
    {
      "epoch": 0.9644,
      "loss_align": 0.11530017852783203,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25849127769470215,
      "neg_sim_mean": 0.7236069440841675,
      "pos_sim_mean": 0.884699821472168,
      "step": 4822
    },
    {
      "epoch": 0.9646,
      "grad_norm": 2.471416711807251,
      "learning_rate": 3.8556338122711643e-07,
      "loss": 0.3738,
      "step": 4823
    },
    {
      "epoch": 0.9646,
      "loss_align": 0.09512776136398315,
      "loss_contrastive": 0.122059166431427,
      "loss_gen": 0.30994299054145813,
      "neg_sim_mean": 0.9269313812255859,
      "pos_sim_mean": 0.9048722386360168,
      "step": 4823
    },
    {
      "epoch": 0.9648,
      "grad_norm": 2.094409227371216,
      "learning_rate": 3.812488758558386e-07,
      "loss": 0.4197,
      "step": 4824
    },
    {
      "epoch": 0.9648,
      "loss_align": 0.0498051643371582,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11802302300930023,
      "neg_sim_mean": 0.7711867094039917,
      "pos_sim_mean": 0.9501948356628418,
      "step": 4824
    },
    {
      "epoch": 0.965,
      "grad_norm": 1.0482807159423828,
      "learning_rate": 3.769585540610798e-07,
      "loss": 0.1678,
      "step": 4825
    },
    {
      "epoch": 0.965,
      "loss_align": 0.07076603174209595,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0007240260019898415,
      "neg_sim_mean": 0.4964302182197571,
      "pos_sim_mean": 0.929233968257904,
      "step": 4825
    },
    {
      "epoch": 0.9652,
      "grad_norm": 0.3939134180545807,
      "learning_rate": 3.7269241793390085e-07,
      "loss": 0.0715,
      "step": 4826
    },
    {
      "epoch": 0.9652,
      "loss_align": 0.10137152671813965,
      "loss_contrastive": 0.09921014308929443,
      "loss_gen": 0.15421514213085175,
      "neg_sim_mean": 0.8978385925292969,
      "pos_sim_mean": 0.8986284732818604,
      "step": 4826
    },
    {
      "epoch": 0.9654,
      "grad_norm": 1.3495913743972778,
      "learning_rate": 3.684504695535496e-07,
      "loss": 0.2675,
      "step": 4827
    },
    {
      "epoch": 0.9654,
      "loss_align": 0.39764928817749023,
      "loss_contrastive": 0.0,
      "loss_gen": 0.02599240094423294,
      "neg_sim_mean": 0.4421212077140808,
      "pos_sim_mean": 0.6023507118225098,
      "step": 4827
    },
    {
      "epoch": 0.9656,
      "grad_norm": 1.961642861366272,
      "learning_rate": 3.6423271098751655e-07,
      "loss": 0.4236,
      "step": 4828
    },
    {
      "epoch": 0.9656,
      "loss_align": 0.1473715901374817,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12804894149303436,
      "neg_sim_mean": 0.6345233917236328,
      "pos_sim_mean": 0.8526284098625183,
      "step": 4828
    },
    {
      "epoch": 0.9658,
      "grad_norm": 1.466895580291748,
      "learning_rate": 3.600391442914741e-07,
      "loss": 0.2754,
      "step": 4829
    },
    {
      "epoch": 0.9658,
      "loss_align": 0.05701792240142822,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00629033287987113,
      "neg_sim_mean": 0.3658510446548462,
      "pos_sim_mean": 0.9429820775985718,
      "step": 4829
    },
    {
      "epoch": 0.966,
      "grad_norm": 0.46513494849205017,
      "learning_rate": 3.558697715093207e-07,
      "loss": 0.0633,
      "step": 4830
    },
    {
      "epoch": 0.966,
      "loss_align": 0.11378216743469238,
      "loss_contrastive": 0.06813406944274902,
      "loss_gen": 0.18782059848308563,
      "neg_sim_mean": 0.8543518781661987,
      "pos_sim_mean": 0.8862178325653076,
      "step": 4830
    },
    {
      "epoch": 0.9662,
      "grad_norm": 1.5270096063613892,
      "learning_rate": 3.517245946731529e-07,
      "loss": 0.3098,
      "step": 4831
    },
    {
      "epoch": 0.9662,
      "loss_align": 0.06407129764556885,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15722593665122986,
      "neg_sim_mean": 0.6752480268478394,
      "pos_sim_mean": 0.9359287023544312,
      "step": 4831
    },
    {
      "epoch": 0.9664,
      "grad_norm": 1.4653992652893066,
      "learning_rate": 3.4760361580328804e-07,
      "loss": 0.2213,
      "step": 4832
    },
    {
      "epoch": 0.9664,
      "loss_align": 0.1366150975227356,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002043024869635701,
      "neg_sim_mean": 0.3470672369003296,
      "pos_sim_mean": 0.8633849024772644,
      "step": 4832
    },
    {
      "epoch": 0.9666,
      "grad_norm": 0.5409271717071533,
      "learning_rate": 3.435068369082306e-07,
      "loss": 0.1387,
      "step": 4833
    },
    {
      "epoch": 0.9666,
      "loss_align": 0.06590795516967773,
      "loss_contrastive": 0.07320719957351685,
      "loss_gen": 0.21884757280349731,
      "neg_sim_mean": 0.9072992205619812,
      "pos_sim_mean": 0.9340920448303223,
      "step": 4833
    },
    {
      "epoch": 0.9668,
      "grad_norm": 1.6008275747299194,
      "learning_rate": 3.394342599847111e-07,
      "loss": 0.2935,
      "step": 4834
    },
    {
      "epoch": 0.9668,
      "loss_align": 0.06776690483093262,
      "loss_contrastive": 0.0,
      "loss_gen": 0.28844156861305237,
      "neg_sim_mean": 0.7238051295280457,
      "pos_sim_mean": 0.9322330951690674,
      "step": 4834
    },
    {
      "epoch": 0.967,
      "grad_norm": 2.026709794998169,
      "learning_rate": 3.3538588701765296e-07,
      "loss": 0.3562,
      "step": 4835
    },
    {
      "epoch": 0.967,
      "loss_align": 0.07966220378875732,
      "loss_contrastive": 0.0,
      "loss_gen": 0.34795230627059937,
      "neg_sim_mean": 0.7799411416053772,
      "pos_sim_mean": 0.9203377962112427,
      "step": 4835
    },
    {
      "epoch": 0.9672,
      "grad_norm": 2.20297908782959,
      "learning_rate": 3.3136171998017775e-07,
      "loss": 0.4276,
      "step": 4836
    },
    {
      "epoch": 0.9672,
      "loss_align": 0.0698169469833374,
      "loss_contrastive": 0.05979132652282715,
      "loss_gen": 0.18380188941955566,
      "neg_sim_mean": 0.8899743556976318,
      "pos_sim_mean": 0.9301830530166626,
      "step": 4836
    },
    {
      "epoch": 0.9674,
      "grad_norm": 1.6751632690429688,
      "learning_rate": 3.2736176083362216e-07,
      "loss": 0.2608,
      "step": 4837
    },
    {
      "epoch": 0.9674,
      "loss_align": 0.12532520294189453,
      "loss_contrastive": 0.10936963558197021,
      "loss_gen": 0.25725141167640686,
      "neg_sim_mean": 0.8840444087982178,
      "pos_sim_mean": 0.8746747970581055,
      "step": 4837
    },
    {
      "epoch": 0.9676,
      "grad_norm": 2.251587152481079,
      "learning_rate": 3.2338601152751e-07,
      "loss": 0.3957,
      "step": 4838
    },
    {
      "epoch": 0.9676,
      "loss_align": 0.10631829500198364,
      "loss_contrastive": 0.0,
      "loss_gen": 0.008140391670167446,
      "neg_sim_mean": 0.2585943937301636,
      "pos_sim_mean": 0.8936817049980164,
      "step": 4838
    },
    {
      "epoch": 0.9678,
      "grad_norm": 0.6219130158424377,
      "learning_rate": 3.1943447399958027e-07,
      "loss": 0.1145,
      "step": 4839
    },
    {
      "epoch": 0.9678,
      "loss_align": 0.17771482467651367,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18360397219657898,
      "neg_sim_mean": 0.5476735830307007,
      "pos_sim_mean": 0.8222851753234863,
      "step": 4839
    },
    {
      "epoch": 0.968,
      "grad_norm": 1.6555277109146118,
      "learning_rate": 3.1550715017575895e-07,
      "loss": 0.3613,
      "step": 4840
    },
    {
      "epoch": 0.968,
      "loss_align": 0.17960309982299805,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1457088589668274,
      "neg_sim_mean": 0.6599730849266052,
      "pos_sim_mean": 0.820396900177002,
      "step": 4840
    },
    {
      "epoch": 0.9682,
      "grad_norm": 1.6400114297866821,
      "learning_rate": 3.1160404197018154e-07,
      "loss": 0.3253,
      "step": 4841
    },
    {
      "epoch": 0.9682,
      "loss_align": 0.1648087501525879,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20845746994018555,
      "neg_sim_mean": 0.6722837686538696,
      "pos_sim_mean": 0.8351912498474121,
      "step": 4841
    },
    {
      "epoch": 0.9684,
      "grad_norm": 1.6110570430755615,
      "learning_rate": 3.0772515128517085e-07,
      "loss": 0.3733,
      "step": 4842
    },
    {
      "epoch": 0.9684,
      "loss_align": 0.10426032543182373,
      "loss_contrastive": 0.07590627670288086,
      "loss_gen": 0.24568144977092743,
      "neg_sim_mean": 0.8716459274291992,
      "pos_sim_mean": 0.8957396745681763,
      "step": 4842
    },
    {
      "epoch": 0.9686,
      "grad_norm": 2.1142611503601074,
      "learning_rate": 3.038704800112535e-07,
      "loss": 0.3591,
      "step": 4843
    },
    {
      "epoch": 0.9686,
      "loss_align": 0.08928573131561279,
      "loss_contrastive": 0.0,
      "loss_gen": 0.255738765001297,
      "neg_sim_mean": 0.646081805229187,
      "pos_sim_mean": 0.9107142686843872,
      "step": 4843
    },
    {
      "epoch": 0.9688,
      "grad_norm": 1.5761834383010864,
      "learning_rate": 3.0004003002714887e-07,
      "loss": 0.345,
      "step": 4844
    },
    {
      "epoch": 0.9688,
      "loss_align": 0.20351195335388184,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13361971080303192,
      "neg_sim_mean": 0.5667271614074707,
      "pos_sim_mean": 0.7964880466461182,
      "step": 4844
    },
    {
      "epoch": 0.969,
      "grad_norm": 1.512333869934082,
      "learning_rate": 2.962338031997691e-07,
      "loss": 0.3371,
      "step": 4845
    },
    {
      "epoch": 0.969,
      "loss_align": 0.14752542972564697,
      "loss_contrastive": 0.14816343784332275,
      "loss_gen": 0.25970518589019775,
      "neg_sim_mean": 0.9006379842758179,
      "pos_sim_mean": 0.852474570274353,
      "step": 4845
    },
    {
      "epoch": 0.9692,
      "grad_norm": 1.7292462587356567,
      "learning_rate": 2.924518013842303e-07,
      "loss": 0.425,
      "step": 4846
    },
    {
      "epoch": 0.9692,
      "loss_align": 0.1470344066619873,
      "loss_contrastive": 0.03939443826675415,
      "loss_gen": 0.14718298614025116,
      "neg_sim_mean": 0.7923600077629089,
      "pos_sim_mean": 0.8529655933380127,
      "step": 4846
    },
    {
      "epoch": 0.9694,
      "grad_norm": 1.479263186454773,
      "learning_rate": 2.886940264238247e-07,
      "loss": 0.2989,
      "step": 4847
    },
    {
      "epoch": 0.9694,
      "loss_align": 0.06925439834594727,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12592625617980957,
      "neg_sim_mean": 0.8302435278892517,
      "pos_sim_mean": 0.9307456016540527,
      "step": 4847
    },
    {
      "epoch": 0.9696,
      "grad_norm": 1.4516924619674683,
      "learning_rate": 2.849604801500538e-07,
      "loss": 0.1952,
      "step": 4848
    },
    {
      "epoch": 0.9696,
      "loss_align": 0.0778929591178894,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15036092698574066,
      "neg_sim_mean": 0.7655991911888123,
      "pos_sim_mean": 0.9221070408821106,
      "step": 4848
    },
    {
      "epoch": 0.9698,
      "grad_norm": 1.3751074075698853,
      "learning_rate": 2.81251164382601e-07,
      "loss": 0.2283,
      "step": 4849
    },
    {
      "epoch": 0.9698,
      "loss_align": 0.18147170543670654,
      "loss_contrastive": 0.0,
      "loss_gen": 0.001146801863797009,
      "neg_sim_mean": 0.2946605682373047,
      "pos_sim_mean": 0.8185282945632935,
      "step": 4849
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.6184422373771667,
      "learning_rate": 2.775660809293368e-07,
      "loss": 0.1826,
      "step": 4850
    },
    {
      "epoch": 0.97,
      "loss_align": 0.07184553146362305,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20281895995140076,
      "neg_sim_mean": 0.8279282450675964,
      "pos_sim_mean": 0.928154468536377,
      "step": 4850
    },
    {
      "epoch": 0.9702,
      "grad_norm": 1.8110419511795044,
      "learning_rate": 2.7390523158633554e-07,
      "loss": 0.2747,
      "step": 4851
    },
    {
      "epoch": 0.9702,
      "loss_align": 0.12105631828308105,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2574714422225952,
      "neg_sim_mean": 0.7461727857589722,
      "pos_sim_mean": 0.878943681716919,
      "step": 4851
    },
    {
      "epoch": 0.9704,
      "grad_norm": 1.9692679643630981,
      "learning_rate": 2.702686181378367e-07,
      "loss": 0.3785,
      "step": 4852
    },
    {
      "epoch": 0.9704,
      "loss_align": 0.06858760118484497,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12219733744859695,
      "neg_sim_mean": 0.7175818085670471,
      "pos_sim_mean": 0.931412398815155,
      "step": 4852
    },
    {
      "epoch": 0.9706,
      "grad_norm": 1.1018823385238647,
      "learning_rate": 2.666562423562946e-07,
      "loss": 0.1908,
      "step": 4853
    },
    {
      "epoch": 0.9706,
      "loss_align": 0.06791788339614868,
      "loss_contrastive": 0.07288819551467896,
      "loss_gen": 0.2266560047864914,
      "neg_sim_mean": 0.9049702882766724,
      "pos_sim_mean": 0.9320821166038513,
      "step": 4853
    },
    {
      "epoch": 0.9708,
      "grad_norm": 1.563900351524353,
      "learning_rate": 2.630681060023343e-07,
      "loss": 0.3033,
      "step": 4854
    },
    {
      "epoch": 0.9708,
      "loss_align": 0.16019487380981445,
      "loss_contrastive": 0.006314635276794434,
      "loss_gen": 0.22011882066726685,
      "neg_sim_mean": 0.7461197376251221,
      "pos_sim_mean": 0.8398051261901855,
      "step": 4854
    },
    {
      "epoch": 0.971,
      "grad_norm": 1.6488991975784302,
      "learning_rate": 2.5950421082476804e-07,
      "loss": 0.3811,
      "step": 4855
    },
    {
      "epoch": 0.971,
      "loss_align": 0.1379162073135376,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22516514360904694,
      "neg_sim_mean": 0.5038949251174927,
      "pos_sim_mean": 0.8620837926864624,
      "step": 4855
    },
    {
      "epoch": 0.9712,
      "grad_norm": 1.775884747505188,
      "learning_rate": 2.5596455856058963e-07,
      "loss": 0.3631,
      "step": 4856
    },
    {
      "epoch": 0.9712,
      "loss_align": 0.11957716941833496,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0018450088100507855,
      "neg_sim_mean": 0.2965332269668579,
      "pos_sim_mean": 0.880422830581665,
      "step": 4856
    },
    {
      "epoch": 0.9714,
      "grad_norm": 0.7282166481018066,
      "learning_rate": 2.5244915093499134e-07,
      "loss": 0.1214,
      "step": 4857
    },
    {
      "epoch": 0.9714,
      "loss_align": 0.153622567653656,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20409202575683594,
      "neg_sim_mean": 0.5206726789474487,
      "pos_sim_mean": 0.846377432346344,
      "step": 4857
    },
    {
      "epoch": 0.9716,
      "grad_norm": 1.7464408874511719,
      "learning_rate": 2.489579896613359e-07,
      "loss": 0.3577,
      "step": 4858
    },
    {
      "epoch": 0.9716,
      "loss_align": 0.10611635446548462,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24986515939235687,
      "neg_sim_mean": 0.7666043639183044,
      "pos_sim_mean": 0.8938836455345154,
      "step": 4858
    },
    {
      "epoch": 0.9718,
      "grad_norm": 1.9561320543289185,
      "learning_rate": 2.4549107644117885e-07,
      "loss": 0.356,
      "step": 4859
    },
    {
      "epoch": 0.9718,
      "loss_align": 0.11593812704086304,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2497752159833908,
      "neg_sim_mean": 0.7371823787689209,
      "pos_sim_mean": 0.884061872959137,
      "step": 4859
    },
    {
      "epoch": 0.972,
      "grad_norm": 2.031991481781006,
      "learning_rate": 2.4204841296424085e-07,
      "loss": 0.3657,
      "step": 4860
    },
    {
      "epoch": 0.972,
      "loss_align": 0.12501561641693115,
      "loss_contrastive": 0.0,
      "loss_gen": 0.22950716316699982,
      "neg_sim_mean": 0.6857802271842957,
      "pos_sim_mean": 0.8749843835830688,
      "step": 4860
    },
    {
      "epoch": 0.9722,
      "grad_norm": 1.515211820602417,
      "learning_rate": 2.386300009084408e-07,
      "loss": 0.3545,
      "step": 4861
    },
    {
      "epoch": 0.9722,
      "loss_align": 0.10453236103057861,
      "loss_contrastive": 0.07387721538543701,
      "loss_gen": 0.21221643686294556,
      "neg_sim_mean": 0.8693448305130005,
      "pos_sim_mean": 0.8954676389694214,
      "step": 4861
    },
    {
      "epoch": 0.9724,
      "grad_norm": 1.8636064529418945,
      "learning_rate": 2.3523584193986814e-07,
      "loss": 0.3256,
      "step": 4862
    },
    {
      "epoch": 0.9724,
      "loss_align": 0.05369436740875244,
      "loss_contrastive": 0.0,
      "loss_gen": 0.29732686281204224,
      "neg_sim_mean": 0.8309389352798462,
      "pos_sim_mean": 0.9463056325912476,
      "step": 4862
    },
    {
      "epoch": 0.9726,
      "grad_norm": 2.271289110183716,
      "learning_rate": 2.3186593771280517e-07,
      "loss": 0.351,
      "step": 4863
    },
    {
      "epoch": 0.9726,
      "loss_align": 0.10139989852905273,
      "loss_contrastive": 0.023377060890197754,
      "loss_gen": 0.1646520048379898,
      "neg_sim_mean": 0.8219771385192871,
      "pos_sim_mean": 0.8986001014709473,
      "step": 4863
    },
    {
      "epoch": 0.9728,
      "grad_norm": 1.5172336101531982,
      "learning_rate": 2.285202898696881e-07,
      "loss": 0.2689,
      "step": 4864
    },
    {
      "epoch": 0.9728,
      "loss_align": 0.09021824598312378,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2722724974155426,
      "neg_sim_mean": 0.7889785766601562,
      "pos_sim_mean": 0.9097817540168762,
      "step": 4864
    },
    {
      "epoch": 0.973,
      "grad_norm": 1.7528966665267944,
      "learning_rate": 2.251989000411514e-07,
      "loss": 0.3625,
      "step": 4865
    },
    {
      "epoch": 0.973,
      "loss_align": 0.09221798181533813,
      "loss_contrastive": 0.0,
      "loss_gen": 0.12604805827140808,
      "neg_sim_mean": 0.5368655323982239,
      "pos_sim_mean": 0.9077820181846619,
      "step": 4865
    },
    {
      "epoch": 0.9732,
      "grad_norm": 1.1668140888214111,
      "learning_rate": 2.219017698460002e-07,
      "loss": 0.2183,
      "step": 4866
    },
    {
      "epoch": 0.9732,
      "loss_align": 0.08236563205718994,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009557412005960941,
      "neg_sim_mean": 0.46159741282463074,
      "pos_sim_mean": 0.9176343679428101,
      "step": 4866
    },
    {
      "epoch": 0.9734,
      "grad_norm": 0.7099573016166687,
      "learning_rate": 2.186289008912157e-07,
      "loss": 0.0919,
      "step": 4867
    },
    {
      "epoch": 0.9734,
      "loss_align": 0.09644150733947754,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1856020838022232,
      "neg_sim_mean": 0.7549800276756287,
      "pos_sim_mean": 0.9035584926605225,
      "step": 4867
    },
    {
      "epoch": 0.9736,
      "grad_norm": 1.9099318981170654,
      "learning_rate": 2.153802947719552e-07,
      "loss": 0.282,
      "step": 4868
    },
    {
      "epoch": 0.9736,
      "loss_align": 0.1039804220199585,
      "loss_contrastive": 0.0,
      "loss_gen": 0.28594785928726196,
      "neg_sim_mean": 0.7619727849960327,
      "pos_sim_mean": 0.8960195779800415,
      "step": 4868
    },
    {
      "epoch": 0.9738,
      "grad_norm": 2.1475985050201416,
      "learning_rate": 2.1215595307154667e-07,
      "loss": 0.3899,
      "step": 4869
    },
    {
      "epoch": 0.9738,
      "loss_align": 0.04711568355560303,
      "loss_contrastive": 0.0,
      "loss_gen": 0.31623023748397827,
      "neg_sim_mean": 0.8451333045959473,
      "pos_sim_mean": 0.952884316444397,
      "step": 4869
    },
    {
      "epoch": 0.974,
      "grad_norm": 2.0856986045837402,
      "learning_rate": 2.0895587736149414e-07,
      "loss": 0.3633,
      "step": 4870
    },
    {
      "epoch": 0.974,
      "loss_align": 0.09955692291259766,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0018462626030668616,
      "neg_sim_mean": 0.6180239915847778,
      "pos_sim_mean": 0.9004430770874023,
      "step": 4870
    },
    {
      "epoch": 0.9742,
      "grad_norm": 0.5118704438209534,
      "learning_rate": 2.057800692014833e-07,
      "loss": 0.1014,
      "step": 4871
    },
    {
      "epoch": 0.9742,
      "loss_align": 0.13995635509490967,
      "loss_contrastive": 0.08670389652252197,
      "loss_gen": 0.13796961307525635,
      "neg_sim_mean": 0.8467475175857544,
      "pos_sim_mean": 0.8600436449050903,
      "step": 4871
    },
    {
      "epoch": 0.9744,
      "grad_norm": 1.9042494297027588,
      "learning_rate": 2.0262853013935378e-07,
      "loss": 0.2883,
      "step": 4872
    },
    {
      "epoch": 0.9744,
      "loss_align": 0.12974274158477783,
      "loss_contrastive": 0.09411907196044922,
      "loss_gen": 0.106314517557621,
      "neg_sim_mean": 0.8643763065338135,
      "pos_sim_mean": 0.8702572584152222,
      "step": 4872
    },
    {
      "epoch": 0.9746,
      "grad_norm": 1.4851932525634766,
      "learning_rate": 1.9950126171114359e-07,
      "loss": 0.2474,
      "step": 4873
    },
    {
      "epoch": 0.9746,
      "loss_align": 0.09570431709289551,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1809098720550537,
      "neg_sim_mean": 0.780593752861023,
      "pos_sim_mean": 0.9042956829071045,
      "step": 4873
    },
    {
      "epoch": 0.9748,
      "grad_norm": 1.847175121307373,
      "learning_rate": 1.9639826544102792e-07,
      "loss": 0.2766,
      "step": 4874
    },
    {
      "epoch": 0.9748,
      "loss_align": 0.09592300653457642,
      "loss_contrastive": 0.0,
      "loss_gen": 0.26646241545677185,
      "neg_sim_mean": 0.5849562883377075,
      "pos_sim_mean": 0.9040769934654236,
      "step": 4874
    },
    {
      "epoch": 0.975,
      "grad_norm": 2.1843361854553223,
      "learning_rate": 1.9331954284137476e-07,
      "loss": 0.3624,
      "step": 4875
    },
    {
      "epoch": 0.975,
      "loss_align": 0.05684959888458252,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005184540059417486,
      "neg_sim_mean": 0.34258899092674255,
      "pos_sim_mean": 0.9431504011154175,
      "step": 4875
    },
    {
      "epoch": 0.9752,
      "grad_norm": 0.478314071893692,
      "learning_rate": 1.9026509541272275e-07,
      "loss": 0.062,
      "step": 4876
    },
    {
      "epoch": 0.9752,
      "loss_align": 0.08868086338043213,
      "loss_contrastive": 0.12362062931060791,
      "loss_gen": 0.3054610788822174,
      "neg_sim_mean": 0.9349397420883179,
      "pos_sim_mean": 0.9113191366195679,
      "step": 4876
    },
    {
      "epoch": 0.9754,
      "grad_norm": 1.9304147958755493,
      "learning_rate": 1.872349246437699e-07,
      "loss": 0.409,
      "step": 4877
    },
    {
      "epoch": 0.9754,
      "loss_align": 0.09478038549423218,
      "loss_contrastive": 0.04015529155731201,
      "loss_gen": 0.4668140113353729,
      "neg_sim_mean": 0.8453748822212219,
      "pos_sim_mean": 0.9052196145057678,
      "step": 4877
    },
    {
      "epoch": 0.9756,
      "grad_norm": 2.1868605613708496,
      "learning_rate": 1.8422903201137932e-07,
      "loss": 0.5664,
      "step": 4878
    },
    {
      "epoch": 0.9756,
      "loss_align": 0.08682703971862793,
      "loss_contrastive": 0.03945577144622803,
      "loss_gen": 0.25935807824134827,
      "neg_sim_mean": 0.8526287078857422,
      "pos_sim_mean": 0.9131729602813721,
      "step": 4878
    },
    {
      "epoch": 0.9758,
      "grad_norm": 2.0588388442993164,
      "learning_rate": 1.8124741898058462e-07,
      "loss": 0.3509,
      "step": 4879
    },
    {
      "epoch": 0.9758,
      "loss_align": 0.10366278886795044,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2964438796043396,
      "neg_sim_mean": 0.6130379438400269,
      "pos_sim_mean": 0.8963372111320496,
      "step": 4879
    },
    {
      "epoch": 0.976,
      "grad_norm": 2.1570451259613037,
      "learning_rate": 1.7829008700460115e-07,
      "loss": 0.4001,
      "step": 4880
    },
    {
      "epoch": 0.976,
      "loss_align": 0.06025993824005127,
      "loss_contrastive": 0.0,
      "loss_gen": 0.09441261738538742,
      "neg_sim_mean": 0.7663818597793579,
      "pos_sim_mean": 0.9397400617599487,
      "step": 4880
    },
    {
      "epoch": 0.9762,
      "grad_norm": 1.0309782028198242,
      "learning_rate": 1.753570375247815e-07,
      "loss": 0.1547,
      "step": 4881
    },
    {
      "epoch": 0.9762,
      "loss_align": 0.10028207302093506,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003139637177810073,
      "neg_sim_mean": 0.33920907974243164,
      "pos_sim_mean": 0.8997179269790649,
      "step": 4881
    },
    {
      "epoch": 0.9764,
      "grad_norm": 0.43999797105789185,
      "learning_rate": 1.7244827197067103e-07,
      "loss": 0.1034,
      "step": 4882
    },
    {
      "epoch": 0.9764,
      "loss_align": 0.08263075351715088,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015048946952447295,
      "neg_sim_mean": 0.23823189735412598,
      "pos_sim_mean": 0.9173692464828491,
      "step": 4882
    },
    {
      "epoch": 0.9766,
      "grad_norm": 0.36745792627334595,
      "learning_rate": 1.6956379175995797e-07,
      "loss": 0.0841,
      "step": 4883
    },
    {
      "epoch": 0.9766,
      "loss_align": 0.01727813482284546,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15720438957214355,
      "neg_sim_mean": 0.8197181224822998,
      "pos_sim_mean": 0.9827218651771545,
      "step": 4883
    },
    {
      "epoch": 0.9768,
      "grad_norm": 1.3261117935180664,
      "learning_rate": 1.6670359829850657e-07,
      "loss": 0.1745,
      "step": 4884
    },
    {
      "epoch": 0.9768,
      "loss_align": 0.06701314449310303,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16409380733966827,
      "neg_sim_mean": 0.4328814744949341,
      "pos_sim_mean": 0.932986855506897,
      "step": 4884
    },
    {
      "epoch": 0.977,
      "grad_norm": 1.7329844236373901,
      "learning_rate": 1.6386769298034065e-07,
      "loss": 0.2311,
      "step": 4885
    },
    {
      "epoch": 0.977,
      "loss_align": 0.12079018354415894,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11475842446088791,
      "neg_sim_mean": 0.7027705311775208,
      "pos_sim_mean": 0.8792098164558411,
      "step": 4885
    },
    {
      "epoch": 0.9772,
      "grad_norm": 1.4349956512451172,
      "learning_rate": 1.6105607718764347e-07,
      "loss": 0.2355,
      "step": 4886
    },
    {
      "epoch": 0.9772,
      "loss_align": 0.07675039768218994,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013235439546406269,
      "neg_sim_mean": 0.3330158591270447,
      "pos_sim_mean": 0.9232496023178101,
      "step": 4886
    },
    {
      "epoch": 0.9774,
      "grad_norm": 0.7854308485984802,
      "learning_rate": 1.5826875229076333e-07,
      "loss": 0.09,
      "step": 4887
    },
    {
      "epoch": 0.9774,
      "loss_align": 0.09559661149978638,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1183793842792511,
      "neg_sim_mean": 0.79634690284729,
      "pos_sim_mean": 0.9044033885002136,
      "step": 4887
    },
    {
      "epoch": 0.9776,
      "grad_norm": 1.5470486879348755,
      "learning_rate": 1.5550571964820793e-07,
      "loss": 0.214,
      "step": 4888
    },
    {
      "epoch": 0.9776,
      "loss_align": 0.06625854969024658,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16001100838184357,
      "neg_sim_mean": 0.7947027683258057,
      "pos_sim_mean": 0.9337414503097534,
      "step": 4888
    },
    {
      "epoch": 0.9778,
      "grad_norm": 1.3685318231582642,
      "learning_rate": 1.5276698060665007e-07,
      "loss": 0.2263,
      "step": 4889
    },
    {
      "epoch": 0.9778,
      "loss_align": 0.11071407794952393,
      "loss_contrastive": 0.0992090106010437,
      "loss_gen": 0.09365391731262207,
      "neg_sim_mean": 0.8884949088096619,
      "pos_sim_mean": 0.8892859220504761,
      "step": 4889
    },
    {
      "epoch": 0.978,
      "grad_norm": 1.262711763381958,
      "learning_rate": 1.500525365009109e-07,
      "loss": 0.2163,
      "step": 4890
    },
    {
      "epoch": 0.978,
      "loss_align": 0.08515298366546631,
      "loss_contrastive": 0.0,
      "loss_gen": 0.37372028827667236,
      "neg_sim_mean": 0.8009686470031738,
      "pos_sim_mean": 0.9148470163345337,
      "step": 4890
    },
    {
      "epoch": 0.9782,
      "grad_norm": 2.0994622707366943,
      "learning_rate": 1.4736238865398765e-07,
      "loss": 0.4589,
      "step": 4891
    },
    {
      "epoch": 0.9782,
      "loss_align": 0.06226789951324463,
      "loss_contrastive": 0.008873581886291504,
      "loss_gen": 0.2003084123134613,
      "neg_sim_mean": 0.846605658531189,
      "pos_sim_mean": 0.9377321004867554,
      "step": 4891
    },
    {
      "epoch": 0.9784,
      "grad_norm": 1.5928822755813599,
      "learning_rate": 1.446965383770149e-07,
      "loss": 0.2636,
      "step": 4892
    },
    {
      "epoch": 0.9784,
      "loss_align": 0.11937230825424194,
      "loss_contrastive": 0.0,
      "loss_gen": 0.1967778354883194,
      "neg_sim_mean": 0.6768730878829956,
      "pos_sim_mean": 0.8806276917457581,
      "step": 4892
    },
    {
      "epoch": 0.9786,
      "grad_norm": 1.9493463039398193,
      "learning_rate": 1.420549869693033e-07,
      "loss": 0.3162,
      "step": 4893
    },
    {
      "epoch": 0.9786,
      "loss_align": 0.15713322162628174,
      "loss_contrastive": 0.0,
      "loss_gen": 0.03225381299853325,
      "neg_sim_mean": 0.282474160194397,
      "pos_sim_mean": 0.8428667783737183,
      "step": 4893
    },
    {
      "epoch": 0.9788,
      "grad_norm": 1.1183520555496216,
      "learning_rate": 1.3943773571831186e-07,
      "loss": 0.1894,
      "step": 4894
    },
    {
      "epoch": 0.9788,
      "loss_align": 0.08110570907592773,
      "loss_contrastive": 0.05272972583770752,
      "loss_gen": 0.13555335998535156,
      "neg_sim_mean": 0.8716239929199219,
      "pos_sim_mean": 0.9188942909240723,
      "step": 4894
    },
    {
      "epoch": 0.979,
      "grad_norm": 1.3772400617599487,
      "learning_rate": 1.36844785899648e-07,
      "loss": 0.223,
      "step": 4895
    },
    {
      "epoch": 0.979,
      "loss_align": 0.08360660076141357,
      "loss_contrastive": 0.04836767911911011,
      "loss_gen": 0.31779998540878296,
      "neg_sim_mean": 0.8647610545158386,
      "pos_sim_mean": 0.9163933992385864,
      "step": 4895
    },
    {
      "epoch": 0.9792,
      "grad_norm": 1.8761754035949707,
      "learning_rate": 1.342761387770952e-07,
      "loss": 0.4072,
      "step": 4896
    },
    {
      "epoch": 0.9792,
      "loss_align": 0.058406054973602295,
      "loss_contrastive": 0.0,
      "loss_gen": 0.019525468349456787,
      "neg_sim_mean": 0.3294000029563904,
      "pos_sim_mean": 0.9415939450263977,
      "step": 4896
    },
    {
      "epoch": 0.9794,
      "grad_norm": 0.9777734875679016,
      "learning_rate": 1.3173179560257432e-07,
      "loss": 0.0779,
      "step": 4897
    },
    {
      "epoch": 0.9794,
      "loss_align": 0.10663443803787231,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24145542085170746,
      "neg_sim_mean": 0.7157595753669739,
      "pos_sim_mean": 0.8933655619621277,
      "step": 4897
    },
    {
      "epoch": 0.9796,
      "grad_norm": 1.7784312963485718,
      "learning_rate": 1.292117576161711e-07,
      "loss": 0.3481,
      "step": 4898
    },
    {
      "epoch": 0.9796,
      "loss_align": 0.04581904411315918,
      "loss_contrastive": 0.0,
      "loss_gen": 0.14118026196956635,
      "neg_sim_mean": 0.271972119808197,
      "pos_sim_mean": 0.9541809558868408,
      "step": 4898
    },
    {
      "epoch": 0.9798,
      "grad_norm": 1.1932165622711182,
      "learning_rate": 1.267160260461253e-07,
      "loss": 0.187,
      "step": 4899
    },
    {
      "epoch": 0.9798,
      "loss_align": 0.10442495346069336,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002064462285488844,
      "neg_sim_mean": 0.3502151370048523,
      "pos_sim_mean": 0.8955750465393066,
      "step": 4899
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.46265214681625366,
      "learning_rate": 1.2424460210881395e-07,
      "loss": 0.1065,
      "step": 4900
    },
    {
      "epoch": 0.98,
      "loss_align": 0.11079287528991699,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19069598615169525,
      "neg_sim_mean": 0.6050851345062256,
      "pos_sim_mean": 0.889207124710083,
      "step": 4900
    },
    {
      "epoch": 0.9802,
      "grad_norm": 1.6641340255737305,
      "learning_rate": 1.2179748700879012e-07,
      "loss": 0.3015,
      "step": 4901
    },
    {
      "epoch": 0.9802,
      "loss_align": 0.12698334455490112,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15057317912578583,
      "neg_sim_mean": 0.6999756097793579,
      "pos_sim_mean": 0.8730166554450989,
      "step": 4901
    },
    {
      "epoch": 0.9804,
      "grad_norm": 1.5592488050460815,
      "learning_rate": 1.193746819387387e-07,
      "loss": 0.2776,
      "step": 4902
    },
    {
      "epoch": 0.9804,
      "loss_align": 0.09957337379455566,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24084138870239258,
      "neg_sim_mean": 0.6508345603942871,
      "pos_sim_mean": 0.9004266262054443,
      "step": 4902
    },
    {
      "epoch": 0.9806,
      "grad_norm": 1.8466848134994507,
      "learning_rate": 1.1697618807951505e-07,
      "loss": 0.3404,
      "step": 4903
    },
    {
      "epoch": 0.9806,
      "loss_align": 0.1340193748474121,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19245292246341705,
      "neg_sim_mean": 0.5717083215713501,
      "pos_sim_mean": 0.8659806251525879,
      "step": 4903
    },
    {
      "epoch": 0.9808,
      "grad_norm": 2.011171579360962,
      "learning_rate": 1.1460200660011189e-07,
      "loss": 0.3265,
      "step": 4904
    },
    {
      "epoch": 0.9808,
      "loss_align": 0.08430111408233643,
      "loss_contrastive": 0.017946958541870117,
      "loss_gen": 0.3602845370769501,
      "neg_sim_mean": 0.8336458206176758,
      "pos_sim_mean": 0.9156988859176636,
      "step": 4904
    },
    {
      "epoch": 0.981,
      "grad_norm": 2.362614393234253,
      "learning_rate": 1.1225213865767026e-07,
      "loss": 0.4467,
      "step": 4905
    },
    {
      "epoch": 0.981,
      "loss_align": 0.1165703535079956,
      "loss_contrastive": 0.04463601112365723,
      "loss_gen": 0.11885831505060196,
      "neg_sim_mean": 0.8280656337738037,
      "pos_sim_mean": 0.8834296464920044,
      "step": 4905
    },
    {
      "epoch": 0.9812,
      "grad_norm": 1.2612721920013428,
      "learning_rate": 1.0992658539750178e-07,
      "loss": 0.2408,
      "step": 4906
    },
    {
      "epoch": 0.9812,
      "loss_align": 0.24836844205856323,
      "loss_contrastive": 0.0,
      "loss_gen": 0.011111489497125149,
      "neg_sim_mean": 0.30014368891716003,
      "pos_sim_mean": 0.7516315579414368,
      "step": 4906
    },
    {
      "epoch": 0.9814,
      "grad_norm": 0.9813293814659119,
      "learning_rate": 1.0762534795303868e-07,
      "loss": 0.2595,
      "step": 4907
    },
    {
      "epoch": 0.9814,
      "loss_align": 0.12839210033416748,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2666192054748535,
      "neg_sim_mean": 0.4970390200614929,
      "pos_sim_mean": 0.8716078996658325,
      "step": 4907
    },
    {
      "epoch": 0.9816,
      "grad_norm": 1.8125375509262085,
      "learning_rate": 1.053484274458838e-07,
      "loss": 0.395,
      "step": 4908
    },
    {
      "epoch": 0.9816,
      "loss_align": 0.11452877521514893,
      "loss_contrastive": 0.11096960306167603,
      "loss_gen": 0.2637910544872284,
      "neg_sim_mean": 0.8964408040046692,
      "pos_sim_mean": 0.8854712247848511,
      "step": 4908
    },
    {
      "epoch": 0.9818,
      "grad_norm": 1.896032691001892,
      "learning_rate": 1.0309582498577719e-07,
      "loss": 0.3916,
      "step": 4909
    },
    {
      "epoch": 0.9818,
      "loss_align": 0.22491991519927979,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25327399373054504,
      "neg_sim_mean": 0.40832895040512085,
      "pos_sim_mean": 0.7750800848007202,
      "step": 4909
    },
    {
      "epoch": 0.982,
      "grad_norm": 2.244706869125366,
      "learning_rate": 1.008675416706073e-07,
      "loss": 0.4782,
      "step": 4910
    },
    {
      "epoch": 0.982,
      "loss_align": 0.16693115234375,
      "loss_contrastive": 0.0,
      "loss_gen": 0.002608873415738344,
      "neg_sim_mean": 0.3916700482368469,
      "pos_sim_mean": 0.83306884765625,
      "step": 4910
    },
    {
      "epoch": 0.9822,
      "grad_norm": 0.6713588833808899,
      "learning_rate": 9.866357858642205e-08,
      "loss": 0.1695,
      "step": 4911
    },
    {
      "epoch": 0.9822,
      "loss_align": 0.12155574560165405,
      "loss_contrastive": 0.0,
      "loss_gen": 0.219375878572464,
      "neg_sim_mean": 0.7586314082145691,
      "pos_sim_mean": 0.878444254398346,
      "step": 4911
    },
    {
      "epoch": 0.9824,
      "grad_norm": 2.0607621669769287,
      "learning_rate": 9.648393680740108e-08,
      "loss": 0.3409,
      "step": 4912
    },
    {
      "epoch": 0.9824,
      "loss_align": 0.10908371210098267,
      "loss_contrastive": 0.05267685651779175,
      "loss_gen": 0.17188584804534912,
      "neg_sim_mean": 0.8435931205749512,
      "pos_sim_mean": 0.8909162878990173,
      "step": 4912
    },
    {
      "epoch": 0.9826,
      "grad_norm": 1.503830075263977,
      "learning_rate": 9.432861739586684e-08,
      "loss": 0.2873,
      "step": 4913
    },
    {
      "epoch": 0.9826,
      "loss_align": 0.059917330741882324,
      "loss_contrastive": 0.019752204418182373,
      "loss_gen": 0.1133563369512558,
      "neg_sim_mean": 0.8598348498344421,
      "pos_sim_mean": 0.9400826692581177,
      "step": 4913
    },
    {
      "epoch": 0.9828,
      "grad_norm": 1.5060569047927856,
      "learning_rate": 9.219762140231236e-08,
      "loss": 0.1756,
      "step": 4914
    },
    {
      "epoch": 0.9828,
      "loss_align": 0.0770806074142456,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17535914480686188,
      "neg_sim_mean": 0.5505789518356323,
      "pos_sim_mean": 0.9229193925857544,
      "step": 4914
    },
    {
      "epoch": 0.983,
      "grad_norm": 1.782963514328003,
      "learning_rate": 9.009094986534572e-08,
      "loss": 0.2524,
      "step": 4915
    },
    {
      "epoch": 0.983,
      "loss_align": 0.06889224052429199,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2951047420501709,
      "neg_sim_mean": 0.7410809397697449,
      "pos_sim_mean": 0.931107759475708,
      "step": 4915
    },
    {
      "epoch": 0.9832,
      "grad_norm": 2.2047781944274902,
      "learning_rate": 8.800860381173448e-08,
      "loss": 0.364,
      "step": 4916
    },
    {
      "epoch": 0.9832,
      "loss_align": 0.09063506126403809,
      "loss_contrastive": 0.0450480580329895,
      "loss_gen": 0.1096886396408081,
      "neg_sim_mean": 0.8544129729270935,
      "pos_sim_mean": 0.9093649387359619,
      "step": 4916
    },
    {
      "epoch": 0.9834,
      "grad_norm": 1.2599109411239624,
      "learning_rate": 8.595058425640013e-08,
      "loss": 0.2057,
      "step": 4917
    },
    {
      "epoch": 0.9834,
      "loss_align": 0.09247815608978271,
      "loss_contrastive": 0.0,
      "loss_gen": 0.003620708594098687,
      "neg_sim_mean": 0.3983570337295532,
      "pos_sim_mean": 0.9075218439102173,
      "step": 4917
    },
    {
      "epoch": 0.9836,
      "grad_norm": 0.4878045320510864,
      "learning_rate": 8.391689220238475e-08,
      "loss": 0.0961,
      "step": 4918
    },
    {
      "epoch": 0.9836,
      "loss_align": 0.07880836725234985,
      "loss_contrastive": 0.06766539812088013,
      "loss_gen": 0.30700698494911194,
      "neg_sim_mean": 0.8888570070266724,
      "pos_sim_mean": 0.9211916327476501,
      "step": 4918
    },
    {
      "epoch": 0.9838,
      "grad_norm": 1.8452482223510742,
      "learning_rate": 8.190752864088436e-08,
      "loss": 0.3939,
      "step": 4919
    },
    {
      "epoch": 0.9838,
      "loss_align": 0.13540375232696533,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18693022429943085,
      "neg_sim_mean": 0.6794413328170776,
      "pos_sim_mean": 0.8645962476730347,
      "step": 4919
    },
    {
      "epoch": 0.984,
      "grad_norm": 2.131723642349243,
      "learning_rate": 7.99224945512489e-08,
      "loss": 0.3223,
      "step": 4920
    },
    {
      "epoch": 0.984,
      "loss_align": 0.077170729637146,
      "loss_contrastive": 0.0,
      "loss_gen": 0.23922662436962128,
      "neg_sim_mean": 0.7623633146286011,
      "pos_sim_mean": 0.922829270362854,
      "step": 4920
    },
    {
      "epoch": 0.9842,
      "grad_norm": 2.068708896636963,
      "learning_rate": 7.796179090094891e-08,
      "loss": 0.3164,
      "step": 4921
    },
    {
      "epoch": 0.9842,
      "loss_align": 0.1251457929611206,
      "loss_contrastive": 0.0,
      "loss_gen": 0.25019559264183044,
      "neg_sim_mean": 0.7677310109138489,
      "pos_sim_mean": 0.8748542070388794,
      "step": 4921
    },
    {
      "epoch": 0.9844,
      "grad_norm": 1.8887354135513306,
      "learning_rate": 7.60254186456144e-08,
      "loss": 0.3753,
      "step": 4922
    },
    {
      "epoch": 0.9844,
      "loss_align": 0.10421329736709595,
      "loss_contrastive": 0.07419764995574951,
      "loss_gen": 0.256181538105011,
      "neg_sim_mean": 0.8699843287467957,
      "pos_sim_mean": 0.895786702632904,
      "step": 4922
    },
    {
      "epoch": 0.9846,
      "grad_norm": 1.9288020133972168,
      "learning_rate": 7.411337872900714e-08,
      "loss": 0.3693,
      "step": 4923
    },
    {
      "epoch": 0.9846,
      "loss_align": 0.17345881462097168,
      "loss_contrastive": 0.12650102376937866,
      "loss_gen": 0.3069514036178589,
      "neg_sim_mean": 0.8530421853065491,
      "pos_sim_mean": 0.8265411853790283,
      "step": 4923
    },
    {
      "epoch": 0.9848,
      "grad_norm": 2.512319803237915,
      "learning_rate": 7.222567208303721e-08,
      "loss": 0.4956,
      "step": 4924
    },
    {
      "epoch": 0.9848,
      "loss_align": 0.09415853023529053,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3347002863883972,
      "neg_sim_mean": 0.7897564172744751,
      "pos_sim_mean": 0.9058414697647095,
      "step": 4924
    },
    {
      "epoch": 0.985,
      "grad_norm": 2.408144950866699,
      "learning_rate": 7.036229962774089e-08,
      "loss": 0.4289,
      "step": 4925
    },
    {
      "epoch": 0.985,
      "loss_align": 0.08809763193130493,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0045108310878276825,
      "neg_sim_mean": 0.20271036028862,
      "pos_sim_mean": 0.9119023680686951,
      "step": 4925
    },
    {
      "epoch": 0.9852,
      "grad_norm": 0.47364023327827454,
      "learning_rate": 6.852326227130834e-08,
      "loss": 0.0926,
      "step": 4926
    },
    {
      "epoch": 0.9852,
      "loss_align": 0.10779953002929688,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21496053040027618,
      "neg_sim_mean": 0.7520033717155457,
      "pos_sim_mean": 0.8922004699707031,
      "step": 4926
    },
    {
      "epoch": 0.9854,
      "grad_norm": 1.9031610488891602,
      "learning_rate": 6.67085609100615e-08,
      "loss": 0.3228,
      "step": 4927
    },
    {
      "epoch": 0.9854,
      "loss_align": 0.10438203811645508,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015501246089115739,
      "neg_sim_mean": 0.3301540017127991,
      "pos_sim_mean": 0.8956179618835449,
      "step": 4927
    },
    {
      "epoch": 0.9856,
      "grad_norm": 0.3857405185699463,
      "learning_rate": 6.491819642846509e-08,
      "loss": 0.1059,
      "step": 4928
    },
    {
      "epoch": 0.9856,
      "loss_align": 0.08126133680343628,
      "loss_contrastive": 0.05404031276702881,
      "loss_gen": 0.16710945963859558,
      "neg_sim_mean": 0.8727789521217346,
      "pos_sim_mean": 0.9187386631965637,
      "step": 4928
    },
    {
      "epoch": 0.9858,
      "grad_norm": 1.5154438018798828,
      "learning_rate": 6.315216969912663e-08,
      "loss": 0.2549,
      "step": 4929
    },
    {
      "epoch": 0.9858,
      "loss_align": 0.07977592945098877,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15319834649562836,
      "neg_sim_mean": 0.7251538634300232,
      "pos_sim_mean": 0.9202240705490112,
      "step": 4929
    },
    {
      "epoch": 0.986,
      "grad_norm": 1.713473916053772,
      "learning_rate": 6.141048158277429e-08,
      "loss": 0.233,
      "step": 4930
    },
    {
      "epoch": 0.986,
      "loss_align": 0.1649187207221985,
      "loss_contrastive": 0.0,
      "loss_gen": 0.31866517663002014,
      "neg_sim_mean": 0.5579909086227417,
      "pos_sim_mean": 0.8350812792778015,
      "step": 4930
    },
    {
      "epoch": 0.9862,
      "grad_norm": 2.305748462677002,
      "learning_rate": 5.969313292830125e-08,
      "loss": 0.4836,
      "step": 4931
    },
    {
      "epoch": 0.9862,
      "loss_align": 0.07980203628540039,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0019615627825260162,
      "neg_sim_mean": 0.3640437722206116,
      "pos_sim_mean": 0.9201979637145996,
      "step": 4931
    },
    {
      "epoch": 0.9864,
      "grad_norm": 0.32428696751594543,
      "learning_rate": 5.800012457270465e-08,
      "loss": 0.0818,
      "step": 4932
    },
    {
      "epoch": 0.9864,
      "loss_align": 0.13279330730438232,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2355479598045349,
      "neg_sim_mean": 0.7398144006729126,
      "pos_sim_mean": 0.8672066926956177,
      "step": 4932
    },
    {
      "epoch": 0.9866,
      "grad_norm": 1.8061667680740356,
      "learning_rate": 5.633145734114664e-08,
      "loss": 0.3683,
      "step": 4933
    },
    {
      "epoch": 0.9866,
      "loss_align": 0.09800612926483154,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20944739878177643,
      "neg_sim_mean": 0.795532763004303,
      "pos_sim_mean": 0.9019938707351685,
      "step": 4933
    },
    {
      "epoch": 0.9868,
      "grad_norm": 1.6316734552383423,
      "learning_rate": 5.4687132046921107e-08,
      "loss": 0.3075,
      "step": 4934
    },
    {
      "epoch": 0.9868,
      "loss_align": 0.2054029107093811,
      "loss_contrastive": 0.17498493194580078,
      "loss_gen": 0.3562539517879486,
      "neg_sim_mean": 0.8695819973945618,
      "pos_sim_mean": 0.7945970892906189,
      "step": 4934
    },
    {
      "epoch": 0.987,
      "grad_norm": 2.2102413177490234,
      "learning_rate": 5.306714949143699e-08,
      "loss": 0.5827,
      "step": 4935
    },
    {
      "epoch": 0.987,
      "loss_align": 0.05116426944732666,
      "loss_contrastive": 0.03428012132644653,
      "loss_gen": 0.32299140095710754,
      "neg_sim_mean": 0.883115828037262,
      "pos_sim_mean": 0.9488357305526733,
      "step": 4935
    },
    {
      "epoch": 0.9872,
      "grad_norm": 2.223419666290283,
      "learning_rate": 5.1471510464268236e-08,
      "loss": 0.3783,
      "step": 4936
    },
    {
      "epoch": 0.9872,
      "loss_align": 0.10532957315444946,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0039144475013017654,
      "neg_sim_mean": 0.3609117865562439,
      "pos_sim_mean": 0.8946704268455505,
      "step": 4936
    },
    {
      "epoch": 0.9874,
      "grad_norm": 0.5126368403434753,
      "learning_rate": 4.9900215743098334e-08,
      "loss": 0.1092,
      "step": 4937
    },
    {
      "epoch": 0.9874,
      "loss_align": 0.0613093376159668,
      "loss_contrastive": 0.039702415466308594,
      "loss_gen": 0.1731862872838974,
      "neg_sim_mean": 0.8783930540084839,
      "pos_sim_mean": 0.9386906623840332,
      "step": 4937
    },
    {
      "epoch": 0.9876,
      "grad_norm": 1.3939030170440674,
      "learning_rate": 4.835326609376467e-08,
      "loss": 0.2393,
      "step": 4938
    },
    {
      "epoch": 0.9876,
      "loss_align": 0.08587265014648438,
      "loss_contrastive": 0.05512213706970215,
      "loss_gen": 0.372188538312912,
      "neg_sim_mean": 0.8692494630813599,
      "pos_sim_mean": 0.9141273498535156,
      "step": 4938
    },
    {
      "epoch": 0.9878,
      "grad_norm": 2.0577778816223145,
      "learning_rate": 4.683066227023081e-08,
      "loss": 0.4647,
      "step": 4939
    },
    {
      "epoch": 0.9878,
      "loss_align": 0.15100979804992676,
      "loss_contrastive": 0.0,
      "loss_gen": 0.187117800116539,
      "neg_sim_mean": 0.680006742477417,
      "pos_sim_mean": 0.8489902019500732,
      "step": 4939
    },
    {
      "epoch": 0.988,
      "grad_norm": 1.853145956993103,
      "learning_rate": 4.5332405014592014e-08,
      "loss": 0.3381,
      "step": 4940
    },
    {
      "epoch": 0.988,
      "loss_align": 0.03472781181335449,
      "loss_contrastive": 0.0,
      "loss_gen": 0.21129906177520752,
      "neg_sim_mean": 0.8533052206039429,
      "pos_sim_mean": 0.9652721881866455,
      "step": 4940
    },
    {
      "epoch": 0.9882,
      "grad_norm": 1.592930793762207,
      "learning_rate": 4.385849505708084e-08,
      "loss": 0.246,
      "step": 4941
    },
    {
      "epoch": 0.9882,
      "loss_align": 0.07311207056045532,
      "loss_contrastive": 0.0,
      "loss_gen": 0.18098175525665283,
      "neg_sim_mean": 0.755409836769104,
      "pos_sim_mean": 0.9268879294395447,
      "step": 4941
    },
    {
      "epoch": 0.9884,
      "grad_norm": 1.696899175643921,
      "learning_rate": 4.240893311607264e-08,
      "loss": 0.2541,
      "step": 4942
    },
    {
      "epoch": 0.9884,
      "loss_align": 0.08272910118103027,
      "loss_contrastive": 0.0019597411155700684,
      "loss_gen": 0.3827984035015106,
      "neg_sim_mean": 0.8192306160926819,
      "pos_sim_mean": 0.9172708988189697,
      "step": 4942
    },
    {
      "epoch": 0.9886,
      "grad_norm": 2.061441659927368,
      "learning_rate": 4.0983719898052274e-08,
      "loss": 0.4658,
      "step": 4943
    },
    {
      "epoch": 0.9886,
      "loss_align": 0.14048844575881958,
      "loss_contrastive": 0.0,
      "loss_gen": 0.00503808306530118,
      "neg_sim_mean": 0.21935871243476868,
      "pos_sim_mean": 0.8595115542411804,
      "step": 4943
    },
    {
      "epoch": 0.9888,
      "grad_norm": 0.5863053202629089,
      "learning_rate": 3.958285609765855e-08,
      "loss": 0.1455,
      "step": 4944
    },
    {
      "epoch": 0.9888,
      "loss_align": 0.136266827583313,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11053185164928436,
      "neg_sim_mean": 0.580071747303009,
      "pos_sim_mean": 0.863733172416687,
      "step": 4944
    },
    {
      "epoch": 0.989,
      "grad_norm": 1.1942479610443115,
      "learning_rate": 3.8206342397656413e-08,
      "loss": 0.2468,
      "step": 4945
    },
    {
      "epoch": 0.989,
      "loss_align": 0.16579586267471313,
      "loss_contrastive": 0.06511574983596802,
      "loss_gen": 0.2958570420742035,
      "neg_sim_mean": 0.799319863319397,
      "pos_sim_mean": 0.8342041373252869,
      "step": 4945
    },
    {
      "epoch": 0.9892,
      "grad_norm": 2.348963499069214,
      "learning_rate": 3.685417946894254e-08,
      "loss": 0.4695,
      "step": 4946
    },
    {
      "epoch": 0.9892,
      "loss_align": 0.10007888078689575,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009877998381853104,
      "neg_sim_mean": 0.3827856779098511,
      "pos_sim_mean": 0.8999211192131042,
      "step": 4946
    },
    {
      "epoch": 0.9894,
      "grad_norm": 0.6701387166976929,
      "learning_rate": 3.5526367970539765e-08,
      "loss": 0.11,
      "step": 4947
    },
    {
      "epoch": 0.9894,
      "loss_align": 0.09036630392074585,
      "loss_contrastive": 0.0,
      "loss_gen": 0.005182597320526838,
      "neg_sim_mean": 0.3382391929626465,
      "pos_sim_mean": 0.9096336960792542,
      "step": 4947
    },
    {
      "epoch": 0.9896,
      "grad_norm": 0.593469500541687,
      "learning_rate": 3.422290854960819e-08,
      "loss": 0.0955,
      "step": 4948
    },
    {
      "epoch": 0.9896,
      "loss_align": 0.07883590459823608,
      "loss_contrastive": 0.0,
      "loss_gen": 0.013243401423096657,
      "neg_sim_mean": 0.33522331714630127,
      "pos_sim_mean": 0.9211640954017639,
      "step": 4948
    },
    {
      "epoch": 0.9898,
      "grad_norm": 0.7134760618209839,
      "learning_rate": 3.294380184143964e-08,
      "loss": 0.0921,
      "step": 4949
    },
    {
      "epoch": 0.9898,
      "loss_align": 0.07853913307189941,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0040022642351686954,
      "neg_sim_mean": 0.13145087659358978,
      "pos_sim_mean": 0.9214608669281006,
      "step": 4949
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5114030241966248,
      "learning_rate": 3.168904846945764e-08,
      "loss": 0.0825,
      "step": 4950
    },
    {
      "epoch": 0.99,
      "loss_align": 0.0941762924194336,
      "loss_contrastive": 0.027651667594909668,
      "loss_gen": 0.25839918851852417,
      "neg_sim_mean": 0.8334753513336182,
      "pos_sim_mean": 0.9058237075805664,
      "step": 4950
    },
    {
      "epoch": 0.9902,
      "grad_norm": 2.239415168762207,
      "learning_rate": 3.04586490452119e-08,
      "loss": 0.3559,
      "step": 4951
    },
    {
      "epoch": 0.9902,
      "loss_align": 0.10970008373260498,
      "loss_contrastive": 0.0087355375289917,
      "loss_gen": 0.1750326305627823,
      "neg_sim_mean": 0.7990354299545288,
      "pos_sim_mean": 0.890299916267395,
      "step": 4951
    },
    {
      "epoch": 0.9904,
      "grad_norm": 1.7517380714416504,
      "learning_rate": 2.9252604168383824e-08,
      "loss": 0.2858,
      "step": 4952
    },
    {
      "epoch": 0.9904,
      "loss_align": 0.07002055644989014,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15852943062782288,
      "neg_sim_mean": 0.6397075653076172,
      "pos_sim_mean": 0.9299794435501099,
      "step": 4952
    },
    {
      "epoch": 0.9906,
      "grad_norm": 1.5695642232894897,
      "learning_rate": 2.8070914426786553e-08,
      "loss": 0.2285,
      "step": 4953
    },
    {
      "epoch": 0.9906,
      "loss_align": 0.14915740489959717,
      "loss_contrastive": 0.034947872161865234,
      "loss_gen": 0.19441591203212738,
      "neg_sim_mean": 0.7857904434204102,
      "pos_sim_mean": 0.8508425951004028,
      "step": 4953
    },
    {
      "epoch": 0.9908,
      "grad_norm": 1.886791467666626,
      "learning_rate": 2.6913580396359383e-08,
      "loss": 0.3478,
      "step": 4954
    },
    {
      "epoch": 0.9908,
      "loss_align": 0.11453711986541748,
      "loss_contrastive": 0.014202117919921875,
      "loss_gen": 0.163864403963089,
      "neg_sim_mean": 0.7996649742126465,
      "pos_sim_mean": 0.8854628801345825,
      "step": 4954
    },
    {
      "epoch": 0.991,
      "grad_norm": 1.5803216695785522,
      "learning_rate": 2.5780602641167773e-08,
      "loss": 0.2801,
      "step": 4955
    },
    {
      "epoch": 0.991,
      "loss_align": 0.09200823307037354,
      "loss_contrastive": 0.0,
      "loss_gen": 0.30037829279899597,
      "neg_sim_mean": 0.7530156970024109,
      "pos_sim_mean": 0.9079917669296265,
      "step": 4955
    },
    {
      "epoch": 0.9912,
      "grad_norm": 2.478703260421753,
      "learning_rate": 2.467198171342e-08,
      "loss": 0.3924,
      "step": 4956
    },
    {
      "epoch": 0.9912,
      "loss_align": 0.060941100120544434,
      "loss_contrastive": 0.0,
      "loss_gen": 0.11540435999631882,
      "neg_sim_mean": 0.8306533098220825,
      "pos_sim_mean": 0.9390588998794556,
      "step": 4956
    },
    {
      "epoch": 0.9914,
      "grad_norm": 1.4354246854782104,
      "learning_rate": 2.358771815344496e-08,
      "loss": 0.1763,
      "step": 4957
    },
    {
      "epoch": 0.9914,
      "loss_align": 0.10346174240112305,
      "loss_contrastive": 0.08997344970703125,
      "loss_gen": 0.14752715826034546,
      "neg_sim_mean": 0.8865116834640503,
      "pos_sim_mean": 0.896538257598877,
      "step": 4957
    },
    {
      "epoch": 0.9916,
      "grad_norm": 1.4039092063903809,
      "learning_rate": 2.2527812489692158e-08,
      "loss": 0.2618,
      "step": 4958
    },
    {
      "epoch": 0.9916,
      "loss_align": 0.11339056491851807,
      "loss_contrastive": 0.09282970428466797,
      "loss_gen": 0.09838393330574036,
      "neg_sim_mean": 0.879439115524292,
      "pos_sim_mean": 0.8866094350814819,
      "step": 4958
    },
    {
      "epoch": 0.9918,
      "grad_norm": 1.3221487998962402,
      "learning_rate": 2.1492265238748366e-08,
      "loss": 0.2229,
      "step": 4959
    },
    {
      "epoch": 0.9918,
      "loss_align": 0.11204040050506592,
      "loss_contrastive": 0.0,
      "loss_gen": 0.17198219895362854,
      "neg_sim_mean": 0.6126722097396851,
      "pos_sim_mean": 0.8879595994949341,
      "step": 4959
    },
    {
      "epoch": 0.992,
      "grad_norm": 1.4711264371871948,
      "learning_rate": 2.0481076905332075e-08,
      "loss": 0.284,
      "step": 4960
    },
    {
      "epoch": 0.992,
      "loss_align": 0.12848997116088867,
      "loss_contrastive": 0.0,
      "loss_gen": 0.009069976396858692,
      "neg_sim_mean": 0.3543550372123718,
      "pos_sim_mean": 0.8715100288391113,
      "step": 4960
    },
    {
      "epoch": 0.9922,
      "grad_norm": 0.6951436400413513,
      "learning_rate": 1.949424798228239e-08,
      "loss": 0.1376,
      "step": 4961
    },
    {
      "epoch": 0.9922,
      "loss_align": 0.14294421672821045,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2873154282569885,
      "neg_sim_mean": 0.36210280656814575,
      "pos_sim_mean": 0.8570557832717896,
      "step": 4961
    },
    {
      "epoch": 0.9924,
      "grad_norm": 1.8565171957015991,
      "learning_rate": 1.8531778950564572e-08,
      "loss": 0.4303,
      "step": 4962
    },
    {
      "epoch": 0.9924,
      "loss_align": 0.09196138381958008,
      "loss_contrastive": 0.06658554077148438,
      "loss_gen": 0.15410108864307404,
      "neg_sim_mean": 0.8746241331100464,
      "pos_sim_mean": 0.9080386161804199,
      "step": 4962
    },
    {
      "epoch": 0.9926,
      "grad_norm": 1.6255345344543457,
      "learning_rate": 1.759367027927561e-08,
      "loss": 0.2541,
      "step": 4963
    },
    {
      "epoch": 0.9926,
      "loss_align": 0.15535211563110352,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16601122915744781,
      "neg_sim_mean": 0.7028611898422241,
      "pos_sim_mean": 0.8446478843688965,
      "step": 4963
    },
    {
      "epoch": 0.9928,
      "grad_norm": 1.7267045974731445,
      "learning_rate": 1.667992242563865e-08,
      "loss": 0.3214,
      "step": 4964
    },
    {
      "epoch": 0.9928,
      "loss_align": 0.11072540283203125,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2851739823818207,
      "neg_sim_mean": 0.5547927021980286,
      "pos_sim_mean": 0.8892745971679688,
      "step": 4964
    },
    {
      "epoch": 0.993,
      "grad_norm": 1.8721446990966797,
      "learning_rate": 1.5790535835003008e-08,
      "loss": 0.3959,
      "step": 4965
    },
    {
      "epoch": 0.993,
      "loss_align": 0.07871508598327637,
      "loss_contrastive": 0.0,
      "loss_gen": 0.27636009454727173,
      "neg_sim_mean": 0.8092489242553711,
      "pos_sim_mean": 0.9212849140167236,
      "step": 4965
    },
    {
      "epoch": 0.9932,
      "grad_norm": 2.1351404190063477,
      "learning_rate": 1.4925510940844156e-08,
      "loss": 0.3551,
      "step": 4966
    },
    {
      "epoch": 0.9932,
      "loss_align": 0.05638134479522705,
      "loss_contrastive": 0.08947420120239258,
      "loss_gen": 0.2508174479007721,
      "neg_sim_mean": 0.9330928325653076,
      "pos_sim_mean": 0.943618655204773,
      "step": 4966
    },
    {
      "epoch": 0.9934,
      "grad_norm": 1.4929105043411255,
      "learning_rate": 1.4084848164763742e-08,
      "loss": 0.3179,
      "step": 4967
    },
    {
      "epoch": 0.9934,
      "loss_align": 0.027913451194763184,
      "loss_contrastive": 0.025517642498016357,
      "loss_gen": 0.2913263142108917,
      "neg_sim_mean": 0.8976041674613953,
      "pos_sim_mean": 0.9720865488052368,
      "step": 4967
    },
    {
      "epoch": 0.9936,
      "grad_norm": 1.9424196481704712,
      "learning_rate": 1.3268547916495122e-08,
      "loss": 0.3223,
      "step": 4968
    },
    {
      "epoch": 0.9936,
      "loss_align": 0.07092839479446411,
      "loss_contrastive": 0.0,
      "loss_gen": 0.01313831377774477,
      "neg_sim_mean": 0.3146481513977051,
      "pos_sim_mean": 0.9290716052055359,
      "step": 4968
    },
    {
      "epoch": 0.9938,
      "grad_norm": 0.7258053421974182,
      "learning_rate": 1.2476610593892269e-08,
      "loss": 0.0841,
      "step": 4969
    },
    {
      "epoch": 0.9938,
      "loss_align": 0.07857418060302734,
      "loss_contrastive": 0.027655839920043945,
      "loss_gen": 0.4702533483505249,
      "neg_sim_mean": 0.8490816354751587,
      "pos_sim_mean": 0.9214258193969727,
      "step": 4969
    },
    {
      "epoch": 0.994,
      "grad_norm": 2.931459665298462,
      "learning_rate": 1.1709036582935318e-08,
      "loss": 0.5521,
      "step": 4970
    },
    {
      "epoch": 0.994,
      "loss_align": 0.14391648769378662,
      "loss_contrastive": 0.05719888210296631,
      "loss_gen": 0.2966555655002594,
      "neg_sim_mean": 0.8132823705673218,
      "pos_sim_mean": 0.8560835123062134,
      "step": 4970
    },
    {
      "epoch": 0.9942,
      "grad_norm": 1.8246805667877197,
      "learning_rate": 1.096582625772502e-08,
      "loss": 0.4474,
      "step": 4971
    },
    {
      "epoch": 0.9942,
      "loss_align": 0.10463052988052368,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13124720752239227,
      "neg_sim_mean": 0.6711950898170471,
      "pos_sim_mean": 0.8953694701194763,
      "step": 4971
    },
    {
      "epoch": 0.9944,
      "grad_norm": 1.3944905996322632,
      "learning_rate": 1.0246979980499393e-08,
      "loss": 0.2359,
      "step": 4972
    },
    {
      "epoch": 0.9944,
      "loss_align": 0.07027697563171387,
      "loss_contrastive": 0.0,
      "loss_gen": 0.34086066484451294,
      "neg_sim_mean": 0.7443062663078308,
      "pos_sim_mean": 0.9297230243682861,
      "step": 4972
    },
    {
      "epoch": 0.9946,
      "grad_norm": 1.960331678390503,
      "learning_rate": 9.552498101611518e-09,
      "loss": 0.4111,
      "step": 4973
    },
    {
      "epoch": 0.9946,
      "loss_align": 0.04311072826385498,
      "loss_contrastive": 0.0,
      "loss_gen": 0.07727409899234772,
      "neg_sim_mean": 0.8050386905670166,
      "pos_sim_mean": 0.956889271736145,
      "step": 4973
    },
    {
      "epoch": 0.9948,
      "grad_norm": 1.039482593536377,
      "learning_rate": 8.882380959551739e-09,
      "loss": 0.1204,
      "step": 4974
    },
    {
      "epoch": 0.9948,
      "loss_align": 0.10141611099243164,
      "loss_contrastive": 0.10843002796173096,
      "loss_gen": 0.2555685341358185,
      "neg_sim_mean": 0.9070138931274414,
      "pos_sim_mean": 0.8985838890075684,
      "step": 4974
    },
    {
      "epoch": 0.995,
      "grad_norm": 1.9772372245788574,
      "learning_rate": 8.236628880914366e-09,
      "loss": 0.37,
      "step": 4975
    },
    {
      "epoch": 0.995,
      "loss_align": 0.05773437023162842,
      "loss_contrastive": 0.0,
      "loss_gen": 0.24108323454856873,
      "neg_sim_mean": 0.5228793025016785,
      "pos_sim_mean": 0.9422656297683716,
      "step": 4975
    },
    {
      "epoch": 0.9952,
      "grad_norm": 1.7458555698394775,
      "learning_rate": 7.615242180436522e-09,
      "loss": 0.2988,
      "step": 4976
    },
    {
      "epoch": 0.9952,
      "loss_align": 0.056465744972229004,
      "loss_contrastive": 0.027234554290771484,
      "loss_gen": 0.36412665247917175,
      "neg_sim_mean": 0.8707687854766846,
      "pos_sim_mean": 0.943534255027771,
      "step": 4976
    },
    {
      "epoch": 0.9954,
      "grad_norm": 2.3090524673461914,
      "learning_rate": 7.018221160981497e-09,
      "loss": 0.4239,
      "step": 4977
    },
    {
      "epoch": 0.9954,
      "loss_align": 0.07385438680648804,
      "loss_contrastive": 0.021441161632537842,
      "loss_gen": 0.24999164044857025,
      "neg_sim_mean": 0.8475867509841919,
      "pos_sim_mean": 0.926145613193512,
      "step": 4977
    },
    {
      "epoch": 0.9956,
      "grad_norm": 1.7580008506774902,
      "learning_rate": 6.445566113516543e-09,
      "loss": 0.3264,
      "step": 4978
    },
    {
      "epoch": 0.9956,
      "loss_align": 0.13119345903396606,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3552219867706299,
      "neg_sim_mean": 0.6969992518424988,
      "pos_sim_mean": 0.8688065409660339,
      "step": 4978
    },
    {
      "epoch": 0.9958,
      "grad_norm": 1.925750732421875,
      "learning_rate": 5.897277317157279e-09,
      "loss": 0.4864,
      "step": 4979
    },
    {
      "epoch": 0.9958,
      "loss_align": 0.07220423221588135,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2167055308818817,
      "neg_sim_mean": 0.5970543622970581,
      "pos_sim_mean": 0.9277957677841187,
      "step": 4979
    },
    {
      "epoch": 0.996,
      "grad_norm": 1.8828052282333374,
      "learning_rate": 5.373355039128835e-09,
      "loss": 0.2889,
      "step": 4980
    },
    {
      "epoch": 0.996,
      "loss_align": 0.1285630464553833,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16142922639846802,
      "neg_sim_mean": 0.7525016665458679,
      "pos_sim_mean": 0.8714369535446167,
      "step": 4980
    },
    {
      "epoch": 0.9962,
      "grad_norm": 1.3903557062149048,
      "learning_rate": 4.873799534788059e-09,
      "loss": 0.29,
      "step": 4981
    },
    {
      "epoch": 0.9962,
      "loss_align": 0.06574136018753052,
      "loss_contrastive": 0.08682388067245483,
      "loss_gen": 0.22401364147663116,
      "neg_sim_mean": 0.9210824966430664,
      "pos_sim_mean": 0.9342586398124695,
      "step": 4981
    },
    {
      "epoch": 0.9964,
      "grad_norm": 1.884959101676941,
      "learning_rate": 4.398611047612411e-09,
      "loss": 0.3002,
      "step": 4982
    },
    {
      "epoch": 0.9964,
      "loss_align": 0.10434561967849731,
      "loss_contrastive": 0.10428768396377563,
      "loss_gen": 0.3539718687534332,
      "neg_sim_mean": 0.8999420404434204,
      "pos_sim_mean": 0.8956543803215027,
      "step": 4982
    },
    {
      "epoch": 0.9966,
      "grad_norm": 2.015254020690918,
      "learning_rate": 3.947789809194414e-09,
      "loss": 0.4708,
      "step": 4983
    },
    {
      "epoch": 0.9966,
      "loss_align": 0.09293842315673828,
      "loss_contrastive": 0.0,
      "loss_gen": 0.13795587420463562,
      "neg_sim_mean": 0.5571039915084839,
      "pos_sim_mean": 0.9070615768432617,
      "step": 4983
    },
    {
      "epoch": 0.9968,
      "grad_norm": 1.463544249534607,
      "learning_rate": 3.5213360392638563e-09,
      "loss": 0.2309,
      "step": 4984
    },
    {
      "epoch": 0.9968,
      "loss_align": 0.11712515354156494,
      "loss_contrastive": 0.03265976905822754,
      "loss_gen": 0.15829026699066162,
      "neg_sim_mean": 0.8155345916748047,
      "pos_sim_mean": 0.8828748464584351,
      "step": 4984
    },
    {
      "epoch": 0.997,
      "grad_norm": 1.5684808492660522,
      "learning_rate": 3.1192499456766946e-09,
      "loss": 0.2793,
      "step": 4985
    },
    {
      "epoch": 0.997,
      "loss_align": 0.05464434623718262,
      "loss_contrastive": 0.0,
      "loss_gen": 0.025299658998847008,
      "neg_sim_mean": 0.528079628944397,
      "pos_sim_mean": 0.9453556537628174,
      "step": 4985
    },
    {
      "epoch": 0.9972,
      "grad_norm": 1.0919052362442017,
      "learning_rate": 2.741531724392843e-09,
      "loss": 0.0799,
      "step": 4986
    },
    {
      "epoch": 0.9972,
      "loss_align": 0.07732445001602173,
      "loss_contrastive": 0.00857478380203247,
      "loss_gen": 0.0808006003499031,
      "neg_sim_mean": 0.8312503099441528,
      "pos_sim_mean": 0.9226755499839783,
      "step": 4986
    },
    {
      "epoch": 0.9974,
      "grad_norm": 1.127566933631897,
      "learning_rate": 2.3881815595150346e-09,
      "loss": 0.1592,
      "step": 4987
    },
    {
      "epoch": 0.9974,
      "loss_align": 0.11655563116073608,
      "loss_contrastive": 0.02892667055130005,
      "loss_gen": 0.19230428338050842,
      "neg_sim_mean": 0.812371015548706,
      "pos_sim_mean": 0.8834443688392639,
      "step": 4987
    },
    {
      "epoch": 0.9976,
      "grad_norm": 1.539736270904541,
      "learning_rate": 2.0591996232610657e-09,
      "loss": 0.3123,
      "step": 4988
    },
    {
      "epoch": 0.9976,
      "loss_align": 0.09183710813522339,
      "loss_contrastive": 0.0,
      "loss_gen": 0.20743431150913239,
      "neg_sim_mean": 0.7413032650947571,
      "pos_sim_mean": 0.9081628918647766,
      "step": 4988
    },
    {
      "epoch": 0.9978,
      "grad_norm": 1.713462471961975,
      "learning_rate": 1.7545860759693445e-09,
      "loss": 0.2993,
      "step": 4989
    },
    {
      "epoch": 0.9978,
      "loss_align": 0.19595211744308472,
      "loss_contrastive": 0.015168905258178711,
      "loss_gen": 0.44828876852989197,
      "neg_sim_mean": 0.7192167639732361,
      "pos_sim_mean": 0.8040478825569153,
      "step": 4989
    },
    {
      "epoch": 0.998,
      "grad_norm": 2.7692582607269287,
      "learning_rate": 1.4743410661044455e-09,
      "loss": 0.6461,
      "step": 4990
    },
    {
      "epoch": 0.998,
      "loss_align": 0.13080334663391113,
      "loss_contrastive": 0.0,
      "loss_gen": 0.0015452546067535877,
      "neg_sim_mean": 0.5324733853340149,
      "pos_sim_mean": 0.8691966533660889,
      "step": 4990
    },
    {
      "epoch": 0.9982,
      "grad_norm": 0.5620125532150269,
      "learning_rate": 1.2184647302626583e-09,
      "loss": 0.1323,
      "step": 4991
    },
    {
      "epoch": 0.9982,
      "loss_align": 0.07255995273590088,
      "loss_contrastive": 0.0,
      "loss_gen": 0.017502693459391594,
      "neg_sim_mean": 0.2920672297477722,
      "pos_sim_mean": 0.9274400472640991,
      "step": 4991
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.9422104358673096,
      "learning_rate": 9.869571931442334e-10,
      "loss": 0.0901,
      "step": 4992
    },
    {
      "epoch": 0.9984,
      "loss_align": 0.10054183006286621,
      "loss_contrastive": 0.0,
      "loss_gen": 0.16355760395526886,
      "neg_sim_mean": 0.5886119604110718,
      "pos_sim_mean": 0.8994581699371338,
      "step": 4992
    },
    {
      "epoch": 0.9986,
      "grad_norm": 1.6757581233978271,
      "learning_rate": 7.798185675866876e-10,
      "loss": 0.2641,
      "step": 4993
    },
    {
      "epoch": 0.9986,
      "loss_align": 0.09674948453903198,
      "loss_contrastive": 0.04947543144226074,
      "loss_gen": 0.2808825671672821,
      "neg_sim_mean": 0.8527259230613708,
      "pos_sim_mean": 0.903250515460968,
      "step": 4993
    },
    {
      "epoch": 0.9988,
      "grad_norm": 1.7575262784957886,
      "learning_rate": 5.970489545537028e-10,
      "loss": 0.3836,
      "step": 4994
    },
    {
      "epoch": 0.9988,
      "loss_align": 0.13879340887069702,
      "loss_contrastive": 0.09494298696517944,
      "loss_gen": 0.2709422707557678,
      "neg_sim_mean": 0.8561495542526245,
      "pos_sim_mean": 0.861206591129303,
      "step": 4994
    },
    {
      "epoch": 0.999,
      "grad_norm": 2.7239372730255127,
      "learning_rate": 4.386484431184723e-10,
      "loss": 0.4211,
      "step": 4995
    },
    {
      "epoch": 0.999,
      "loss_align": 0.1466512680053711,
      "loss_contrastive": 0.0,
      "loss_gen": 0.2233225554227829,
      "neg_sim_mean": 0.4373018145561218,
      "pos_sim_mean": 0.8533487319946289,
      "step": 4995
    },
    {
      "epoch": 0.9992,
      "grad_norm": 1.9706414937973022,
      "learning_rate": 3.0461711048035415e-10,
      "loss": 0.37,
      "step": 4996
    },
    {
      "epoch": 0.9992,
      "loss_align": 0.20294541120529175,
      "loss_contrastive": 0.0,
      "loss_gen": 0.15427805483341217,
      "neg_sim_mean": 0.6809722185134888,
      "pos_sim_mean": 0.7970545887947083,
      "step": 4996
    },
    {
      "epoch": 0.9994,
      "grad_norm": 1.4482057094573975,
      "learning_rate": 1.9495502197042214e-10,
      "loss": 0.3572,
      "step": 4997
    },
    {
      "epoch": 0.9994,
      "loss_align": 0.04663717746734619,
      "loss_contrastive": 0.047307491302490234,
      "loss_gen": 0.23272249102592468,
      "neg_sim_mean": 0.9006702899932861,
      "pos_sim_mean": 0.9533628225326538,
      "step": 4997
    },
    {
      "epoch": 0.9996,
      "grad_norm": 1.7830379009246826,
      "learning_rate": 1.0966223103481276e-10,
      "loss": 0.285,
      "step": 4998
    },
    {
      "epoch": 0.9996,
      "loss_align": 0.1679222583770752,
      "loss_contrastive": 0.0,
      "loss_gen": 0.19975849986076355,
      "neg_sim_mean": 0.6718106269836426,
      "pos_sim_mean": 0.8320777416229248,
      "step": 4998
    },
    {
      "epoch": 0.9998,
      "grad_norm": 2.119474411010742,
      "learning_rate": 4.873877924582715e-11,
      "loss": 0.3677,
      "step": 4999
    },
    {
      "epoch": 0.9998,
      "loss_align": 0.09345293045043945,
      "loss_contrastive": 0.0,
      "loss_gen": 0.3116375803947449,
      "neg_sim_mean": 0.6539332270622253,
      "pos_sim_mean": 0.9065470695495605,
      "step": 4999
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.0822935104370117,
      "learning_rate": 1.2184696296380082e-11,
      "loss": 0.4051,
      "step": 5000
    },
    {
      "epoch": 1.0,
      "step": 5000,
      "total_flos": 5.282142592578355e+16,
      "train_loss": 0.4946937225859612,
      "train_runtime": 3525.3482,
      "train_samples_per_second": 1.418,
      "train_steps_per_second": 1.418
    }
  ],
  "logging_steps": 1,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.282142592578355e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}